260 Small Places, Large Issues One seemingly paradoxical result of modernisation in many parts of the world is the emergence of 'traditionalist' movements praising the virtue of 'the ancestral culture'. Like cargo cults, such movements may be interpreted as strategies to come to terms with new social and cultural circumstances; adapting to the new without letting go of the old entirely and thereby creating a sense of continuity with the past in a rapidly changing world. In the following two chapters, we look into some such movements and processes in some detail. SUGGESTIONS FOR FURTHER READING David Lan: Guns and Rain: Guerrillas and Spirit Mediums in Zimbabwe. London: James Currey 1985. Akhii Gupta and James Ferguson, eds: Anthropological Locations: Boundaries and Grounds of a Field Science. Berkeley: University of California Press 1997. Eric Wolf: Europe and the People without History. Berkeley: University of California Press 1982. Peter Worsley: The Three Worlds. London: Weidenfeld & Nicolson 1984. 17 ETHNICITY People become aware of their culture when they stand at its boundaries: when they encounter other cultures, or when they become aware of other ways of doing things, or merely of contradictions to their own culture. — Anthony P. Cohen A well-known musician from Finnmark, the Saroi-dominated county in northern Norway, was once asked the following question by a journalist: 'Are you mostly a Sami or mostly a musician?' She tried to be accommodating and gave an answer to the question; if she had been an anthropologist, she would probably have regarded the question as absurd. This chapter, which outlines basic dimensions of ethnicity, explains why. THE CONTEMPORARY UBIQUITY OF ETHNICITY A cursory glance at major anthropological journals and monographs from, say, 1950 to 2000, will quickly reveal a change in the language of the subject. The terminology has generally become more influenced by hermeneutics and literary theory than by natural science during this period. Words such as 'function' and 'social structure' have become less common. Those like 'class', 'infrastructure' and 'contradiction' had a brief spell of popularity in the 1970s, while terms such as 'discourse', 'resistance' and 'symbolic capital' have steadily grown more popular since the early 1980s. Such terminological changes reflect shifts in the dominant perspectives of the subject, but they may also reflect changes in the outside world. The enormous interest in ethnicity which has developed since the late 1960s, the growing interest in nationalism since the early 1980s, and the enormous number of books with 'global' in the title since about 1990, indicate some such changes. For one thing, a term like 'ethnic group', which has largely replaced that of 'tribe', simultaneously expresses that tribal organisation is no longer common and that anthropology no longer works from a rigid boundary between 'us' and 'them'. For ethnic groups (and nations) are omnipresent and exist in the anthropologist's own society as well as elsewhere. Looking at the political situation in the world at the end of the twentieth century, the immediate impression is that most of the serious armed conflicts today have an important ethnic dimension. From Punjab to Northern Ireland, from Tibet to Bosnia, from Sri Lanka to the former Soviet Union, 261 262 Small Places, Large Issues there is conflict and competition between different ethnic groups regarding political sovereignty and control over territories. There are also other kinds of conflicts where ethnic groups emerge as corporate groups. Indigenous peoples and immigrant groups may, for example, demand the right to cultural survival and the right to equality with the majority, usually without demanding their own state. Other ethnic conflicts are not expressed through institutional politics. Clashes between natives and immigrants in Germany, or between blacks and whites in the USA, are usually enacted in day-to-day situations. In other words, ethnicity may be articulated at many levels in complex societies. Ethnicity does not necessarily entail conflict: it may be expressed in quite undramatic ways through everyday definitions of situations, through impression management, in religious cults and other peaceful phenomena. It can be identified at different levels of scale - from dyadic interaction to civil war. The phenomenon of ethnicity is, in other words, a complex one. In everyday language, the concept 'ethnic group' is normally used to describe a minority group which is culturally distinguishable from the majority, and as such the term encompasses groups in very different situations - ranging from New York Jews to the Yanomamö in Brazil. In anthropology, the expression 'ethnic group' may also be used to describe majority groups, and ethnicity concerns the relationship between groups whose members consider each other culturally distinctive. This is still rather vague. Let us therefore be more specific. COMMUNICATING CULTURAL DIFFERENCE It is commonly held, not least among members of ethnic groups with a strong collective identity, that ethnicity has to do with 'objective cultural differences'. This would suggest that ethnicity becomes more important the greater the cultural differences are, and that the phenomenon is caused by the fact that different groups have lived in relative isolation from each other and have developed socioculturally in very different directions. Anthropological research on ethnicity has shown this assumption to be false. In fact, ethnicity is frequently most important in contexts where groups are culturally close and enter into contact with each other regularly. Anthropology may therefore give an answer to a seeming paradox of our time, namely that whereas cultural differences in many regards become less apparent because of increased contact and the general processes of modernisation, ethnic identity and self-consciousness become increasingly important. The more similar people become, it seems, the more they are concerned with remaining distinctive. Bateson (1979) has written that there must always be two of something to create a difference and thereby information. In line with this idea, we Ethnicity 263 might say that the idea of an isolated ethnic group is absurd. It is through contact with others that we discover who we are, and an 'isolated ethnic group' may therefore be compared with the sound of one hand clapping -an absurdity. The fact that two groups are culturally distinctive does not create ethnicity. There must be at least a minimum of contact between their respective members. We therefore have to draw the conclusion that the members of different ethnic groups must have something in common - some basis for interaction - in addition to being different. Ethnicity occurs when cultural differences are made relevant through interaction. It thus concerns what is socially relevant, not which cultural differences are 'actually there'. In an article on ethnic relations in Thailand, Michael Moerman (1965) has shown that many of his informants mention cultural particulars which they presume are characteristic of themselves but which they in fact share with neighbouring peoples. Indeed, a variety of criteria can be used as markers of cultural difference ininterethnic situations -phenotype (appearance or 'race'), language, religion or even clothes. If any such marker is socially recognised as an indicator of an ethnic contrast, it matters little if the 'objective cultural differences' are negligible. In an influential essay on the social nature of ethnicity, Fredrik Barth (1969) criticised a then influential view for being overly concerned with cultural content, or substance, instead of focusing on social processes in the study of ethnicity. Like Moerman (and Leach, Mitchell and others) before him, Barth emphasises that 'cultural traits' do not entail ethnicity, and suggests that the focus of research ought to be the social boundaries between groups rather than the 'cultural stuff they contain. In fact, he says, there may be a continuous flow of people and information across ethnic boundaries even though they are maintained as boundaries. If such divisions are maintained, this must be because they have some social relevance. In the relationship between the sedentary Fur and the nomadic Baggara in the Sudan, for example, there is an economic complementarity. They are mutually dependent on commodity exchange and occupy complementary ecological niches. As Gunnar Haaland (1969) has shown, Fur may become Baggara by changing their way of life, just as Leach (1954) showed that Kachin could become Shan; but the ethnic boundary separating the groups remains untouched in the process. Barth stresses the social process in his model of ethnicity, which provides an image of inter-ethnic relationships as dynamic and negotiable. Ethnicity must therefore be seen as an aspect of a relationship, not as a property of a person or a group. The existence of the ethnic group thus has to be affirmed socially and ideologically through the general recognition, among its members and outsiders, that it is culturally distinctive. In addition, this cultural distinctiveness has to be related to social practices, such as religion, marriage (a rule of endogamy), language or work: for an ethnic identity to survive, it must be embedded in at least some of the social situations the actor goes through. 2 64 Small Places, Large Issues A further dimension of ethnicity, which was not dealt with in the volume edited by Barth but which has become politically important in many contemporary societies, is the appropriation of a shared history (Tonkin et al. 1989), that simultaneously functions as an origin myth, justifies claims to a common culture and serves to depict the ethnic group as an extended kin group. SOCIAL CLASSIFICATION AND STEREOTYPES The studies of urbanisation in the Copperbelt (Chapter 16) are related to the line of reasoning presented systematically by Barth. In The Kálela Dance (1956) Mitchell clearly shows how ethnicity or 'tribalism' is expressed through a practical, everyday classification of persons. In the polyethnic situation of the mining towns of North Rhodesia, ethnicity did not have to be related to rank differences or an ethnically based division of labour. Rather, it expressed an internal cultural differentiation and emic concepts of relevant cultural differences, but it was ultimately based on social organisation in that the urbanites were still tied through kinship to their home village and traditional marriage rules. Ethnic classification, nonetheless, has something to do with the creation of order in the social environment by providing a division into 'kinds' of people. Such a classification may or may not be related to power disparities. For a system of social classification to be effective, the actors have to believe in it. They must be convinced that there are relevant differences which distinguish their group from members of the other groups. A classification of persons into kinds of persons which stresses the commonalities of each kind and neglects their differences, depends on stereotypes to be efficient. Stereotypes are simplistic descriptions of cultural traits in other groups which are conventionally believed to exist. One may have stereotypes about workers, women and royal families; for instance, a common stereotypical idea about male homosexuals amounts to their being 'effeminate'. Phrases like 'Jews are greedy' or 'the French are passionate' express stereotypes: they refuse to take individual variations into account. Ethnic stereotypes are often morally condemning (as in 'Hindus are selfish', 'Never trust an Arab'), and such images of others may strengthen group cohesion, boundaries and one's self-perception. In polyethnic societies, people also commonly hold stereotypes about themselves; an example could be the Creoles of Mauritius (Eriksen 1988), who tend to describe themselves collectively as honest, generous folk in implicit contrast to other ethnic groups. It is impossible to make a general statement about the relationship of stereotypes to 'facts'. They can be exaggerated, overly generalising and ideologically charged descriptions of social facts. When the Creoles of Mauritius describe the Hindus as miserly, this stereotype corresponds to one that Hindus hold of themselves as hard-working and responsible people. Ethnicity 265 Ethnic Anomalies In the study of classification and cosmology, the term anomaly is used to describe animals, plants or other phenomena which do not fit into the social system of classification. The pangolin is an anomaly among the Lele (Douglas 1975), and the duck-billed platypus was seen as an anomaly within the Linnean system of classification when it was discovered in the nineteenth century: it was an egg-laying mammal with a duck's beak. Anomalies occur in social classification as well. An anomaly in an ethnic system of classification may be a poor Syrian in Trinidad, for example, where wealth is seen as a defining characteristic of Syrians. A highly educated Gypsy would appear an anomaly in many European societies, where Gypsy identity is symbolically tied to low education. In many societies, moreover, children of ethnically mixed marriages are considered anomalies. In certain polyethnic societies, entire categories of persons may be seen as anomalous. In Trinidad, the main contrasting pair of the ethnic system of classification is the Afro-Trinidadian/Indian contrast. The children of mixed African/Indian unions are locally known as douglas (which is Bhojpuri for bastard), and tend to be placed in an uncomfortable position oscillating between 'both-anď and 'neither-nor' in the system of classification. In Mauritius, a similar dilemma characterises the situation of the gens de couleur or 'mulattoes'; people of mixed African/European origin. They are not allowed to join the ethnic community of Mauritian whites, nor are they generally perceived as good Creoles by the black Creoles (of African descent). Frequently, they are described by black working-class Creoles as kreol fer blan - as Creoles who try to be white. Seen from a systemic perspectives, ethnic anomalies are helpless 'victims' of a hegemonic system of classification. Seen from an actor perspective, however, they may sometimes be regarded as ethnic entrepreneurs who succeed in being both, say, Indian and African and switching strategically between these identifications (Eriksen 1993a). Stereotypes may also function as self-fulfilling prophecies. A dominating group may, for example, turn a dominated group into school dunces by systematically bombarding them with statements to the effect that they are born stupid. There are also many stereotypes with no clear relationship to social facts. Accusations about cannibalism in many societies are typical examples: they are usually false (Arens 19 78). Stereotypes provide ideological legitimation of ethnic boundaries ('Don't marry one of those!') and strengthen group cohesion ('It's a good thing we're SITUATIONAL ETHNICITY An important insight from the studies of the Copperbelt is the fact that ethnicity is relative and to some extent situational. As Mitchell expresses it (1966), a man may behave as a tribal in some situations and as a wageworker in others. For a North Indian Brahmin in an English city, it would be deeply insulting if a native English person were to classify him in the same category as persons he perceived as 'pitch-black' immigrants from Jamaica (in the broad category of 'non-white immigrants'). He might see himself as a white, 'clean' person, and would perhaps try to convince the native English person that he rightly belonged to a different category from the Jamaican. Such a negotiation over identity may be regarded as a struggle between different views of what the world looks like in ethnic terms, as competition between ideological descriptions of the world, or even as competition between world-views. Some groups, who may be the victims of damaging ethnic stereotypes, may try to argue the irrelevance of ethnic distinctions or to challenge the prevailing stereotypes, much in the manner of the Pan caste of Bisipara (Chapter 10), switching strategically from trying to exploit the system to their own advantage to rejecting the entire caste system. The situational aspect of ethnicity has been explored by many anthropologists. In a study of ethnic relations on the sub-Arctic north Norwegian coast, Harald Eidheim (1971) has shown how ethnicity is made relevant in various ways, in various kinds of situations, between Norwegians and Sami. Here he introduces the concepts of dichotomisation and complementarisa-tion to describe fundamental ethnic processes. Dichotomisation refers to the articulation of ethnic relationships through mutual negations: the Sami define themselves in direct contrast to the non-Sami (usually the Norwegian). Complementarisation, rather, gives an expression of ethnic relationships within a shared language where both groups appear as culturally distinctive and as structurally equivalent. Dichotomisation is expressed through stereotypes where the other group is seen as inferior ('the Sami are dirty', etc.), while complementarisation is expressed through the school system, where both Norwegian and Sami history is taught. In complementarisation, the members of one group will compare themselves directly with the other group, stating, for example, that we have our history, religion, folk music and arts; while they, too, have their history, religion, folk music and arts. In this way, ethnicity contributes to making cultural differences comparable. Eidheim also showed the essential falseness of widespread stereotypes about the Sami. Contrary to what many Norwegians held, they were not 'dirty', 'drunken' or 'pagan'. 266 Small Places, Large Issues not like them'). When they are coupled with a rank system, stereotypes tend to support and strengthen it. Ethnicity 267 In the early 1960s the coastal Sami were a culturally stigmatised group, meaning that they were looked down upon by the dominant Norwegians. For this reason, they undercommunicated their ethnic identity in public contexts, in the shop, on the local steamer and so on. They then presented themselves as Norwegians, and overcommunicated what they saw as Norwegian culture to escape from the stigma. In private 'backstage' situations, however, they always spoke Samí and expressed their common identity. A theoretical point from this kind of ethnicity studies is the fact that ethnicity is relational and processual: it is not a 'thing', but an aspect of a social process. This does not mean that the emotions and cultural heritage attaching individuals to ethnic groups are in some sense not 'real' (A.P. Cohen 1994; Jenkins 1996), but that they become operational only in relationships with others. ETHNIC IDENTITY AND ORGANISATION An important element in ethnic ideologies is the notion of the historical continuity of the group. By appealing to notions of shared tradition and history, such ideologies give the impression that the ethnic group is 'natural' and enjoys cultural continuity over a long period of time. In this way, every ethnic ideology offers a feeling of cultural belongingness and security. Such ideologies also, naturally, have a political dimension. Both ethnic identity (group identification) and ethnic organisation (politics) are important, complementary expressions of ethnicity. How strong the we-feeling is, and what a possible ethnic organisation or corporation has to offer its members, varies greatly, however. The fact that ethnicity occurs in a society gives us no indication of its relative importance in the everyday life or politics of that society. Ethnic identity may vary both situationally and in absolute terms, in the sense that one's ethnic membership may be socially relevant in a negligible number of contexts, and that one's self-perception is made up by many other elements as well as the ethnic one. For example, all North Americans except the indigenous populations are descendants of people who arrived from other continents quite recently, but in most cases this origin has little social relevance. To the Swedish-Americans of Minnesota, it may be the case two or three times a year that their ethnic identity is made visible through public rituals; for the rest of the year, they are ordinary and not hyphenated Americans. By contrast, in everyday life in a highland village in Guatemala the dichotomisation between Ladinos (of mixed origin and Spanish-speaking) and Maya Indians is very powerful, and nearly everything an individual does -from dress to food to language to work to body language - can easily be read by the other villagers as an expression of his or her ethnic membership (Nash 1988). The local community is, as it were, saturated with ethnicity. There is no agreement among specialists as to which is more fundamental, ethnic identity or organisation. Some assume that identity, as 'we-feeling', 268 Small Places, Large Issues is the most basic aspect of ethnicity and that ethnically based politics consists of transferring this collective emotional attachment to a new field. Abner Cohen, who has studied ethnicity in Nigeria, Sierra Leone and London (1969, 1981, 1993), represents the opposite position. In his view, ethnic organisation - the pursuit of group interests - is the very raison d'etre of ethnic identity; he holds that the identity would vanish if it had no organisational focus, and that the ethnic cohesion as well as the we-feeling are actually created through social and political processes, especially in contexts of competition for scarce resources. Research on ethnicity indicates that there are many variations here, although Cohen's view of ethnicity as an essentially political phenomenon is certainly valuable in the study of ethnicity in modern societies. Ethnicity, as it is enacted, is clearly a combination of both dimensions - the symbolic and the social or political. Clearly, ethnic ideologies depend on a cultural 'raw material' as a point of departure, although this 'raw material' may be manipulated. Some aspects of the identity base of ethnicity cannot, however, be manipulated. Obviously, Mauritian Hindu politicians who decide to exploit ethnicity in the quest for office will be unable to persuade a single voter that they are Hindus if they define themselves as Muslims. On the other hand, they may be able to persuade Hindus to vote along ethnic lines and may thus contribute to the formation of a corporate Hindu ideology. It has also been argued that ethnic identities may continue to be important to their carriers even if they do not 'pay off. The Creoles of Mauritius may provide an example: they are poorly organised politically, and define themselves in relation to ostensible cultural traits which make it difficult for them to compete with other groups in the labour market and in politics (Eriksen 1988). An important insight from Cohen and others who have studied ethnicity as a kind of political organisation is nevertheless that ethnicity attains its overall greatest importance when it is expressed as economic and political competition over scarce resources which both or all groups deem valuable. There are thus degrees of ethnicity. Don Handelman (1977) has developed a typology of degrees of ethnic incorporation where he distinguishes between four levels; put in other terms, he classifies inter-ethnic contexts as four distinctive degrees of social and cultural importance. His degrees of incorporation are as follows. In the ethnic category, ethnicity as identity is reproduced over several generations through myths of origin and endogamy. Its social relevance outside the household and kinship levels of organisation is negligible. The ethnic network is an interpersonal system of interaction, accompanied by a flow of value, which follows ethnic lines. For example, ethnic networks beyond the kinship level may be helpful in the search for a job, a house and a spouse. The third level is the ethnic association, whose members are organised cor-poratively in certain respects. It consists of a goal-directed collective Ethnicity 269 organisation which pursues shared goals defined on behalf of the ethnic group as a whole. Finally, Handelman speaks of the ethnic community. Here we have an ethnic group with a clear territorial base. At this, highest level of ethnic incorporation, group members have shared interests in their ethnic identity, their ethnic networks, their ethnic associations and their shared territorial estate. It is obvious that a society where virtually all scarce resources are held by ethnic groups in this way becomes much more 'ethnifieď - ethnicity becomes a more pervasive dimension in politics and everyday life - than societies where neither politics nor the division of labour are strongly correlated with ethnic distinctions. There are thus important differences in degree as to the relative importance of ethnicity. The apparent dualism between ethnic identity and ethnic organisation -personal fulfilment versus 'What's in it for me?' - can be overcome, either by simply conceding that they are two sides of the same coin, or by conceptualising ethnicity as something akin to a total social phenomenon. Drawing on Weiner's concept of inalienable possessions (see Chapter 12), Simon Harrison (2000) does the latter. He discusses the symbols and markers of ethnic identity as core elements of selfhood, which cannot be bartered or traded. When one group tries to appropriate another"s symbols, through 'piracy' or commercialisation (e.g. in the case of Aboriginal art), or in processes of social mobility (e.g. in Sanskritisation, cf. Chapter 10), the original group will try to protect its ethnic estate, which consists of the symbols, knowledge and artifacts that make them who they are. In such situations, identity and politics merge. ETHNICITY AND RANK In polyethnic societies, as in others, a variety of criteria is available for the classification of people, and ethnic status never gives sufficient information to describe a person's position in a system of social classification. Ethnic categories or groups are internally differentiated according to gender, age and (often) class, and there are also other independent criteria for differentiation which have no direct relationship to ethnicity. We can illustrate this with an example from Mauritius, that polyethnic island-state in the southwestern Indian Ocean (Eriksen 1988, 1992,1998). Both Rajiv and Kumar are Hindus around 30 years of age. Their respective positions in Mauritian society are nevertheless very different, and one may almost say that all they have in common, apart from gender and age, is their ethnic identity. Rajiv belongs to the Babojee caste, which makes him a Brahmin; in other words, he is a member of one of the highest castes. He is the son of a wealthy merchant in Vacoas and is eventually expected to take over his father's business, where he now works. Rajiv is also expected to travel to India to find 2 70 Small Places, Large Issues a wife from his own caste within a couple of years. He has a BA degree from the University of Lucknow. He speaks English and French in addition to Creole (his vernacular) and takes evening classes in Hindu. He owns a car. Many of Rajiv's friends and acquaintances belong to other castes; some even belong to other ethnic groups. Kumar belongs to the Ravi Ved caste, which is one of the lowest castes in Mauritius. Like his father, he is an agricultural worker on a sugar estate in Flacq, and the family lives in a small rented house on the estate. Kumar is hardly literate and can make himself understood only in Bhojpuri (a rural Hindi dialect) and Creole, the low-status languages of Mauritius. He is married to a woman from the village and they have two children. Kumar's personal network largely consists of kinsmen, neighbours and colleagues from work, all of them low-caste Hindus like himself. Class Education Ethnicity , . s Gender (property) Figure 17.1 Four perspectives on rank in Mauritian society According to a class analysis, Kumar belongs to the working class, while Rajiv belongs to the bourgeoisie. Their life-worlds are in many ways very different, and in this sense Kumar may have more in common with the Creoles who work in the sugar factory than with Rajiv. Nevertheless, they vote for the same party, the Hindu-dominated Labour Party. There are political organisations and parties in Mauritius which hold that class differences and the rural-urban contrast are more fundamental than ethnic differences. Some of them have enjoyed some success in certain periods; for example, in the 1970s there were two general strikes where Creoles, Hindus and Muslims went on strike together against the state and the local capitalists, regardless of their ethnic membership. This illustrates the fact that the social relevance of ethnicity cannot be taken for granted, but is negotiable. The example from Mauritius shows that ethnicity cuts across class. However, in many societies there is a strong correlation between ethnicity and class, so that some ethnic groups rank higher than others. Many Caribbean societies reveal such a pattern, where skin colour may frequently be read as an indication of class membership. In some places, moreover, it is Ethnicity 271 possible to change one's ethnic membership as one changes one's class membership. In Peru, it may be de facto possible for Indians to become Ladinos if they acquire the way of life and external markers (dress, language) characteristic of the Ladinos (van den Berghe 1975). In Mauritius, Creoles may regard themselves as, and be regarded by others as, gens de couleur ('Coloured') if they acquire certain aspects of what is locally seen as a European way of life. Both class and ethnicity are criteria for social differentiation and rank. As with the caste-class relationship discussed in Chapter 10, neither can be reduced to the other, but they do influence each other. For example, the Chinese of Trinidad, formerly a poor and somewhat despised group, have gradually become one of the richest ethnic groups in the island. As a result, people of mixed origin now tend to overcommunicate the 'Chinee' aspect of their identity. Although it is commonly assumed that ethnicity is ascribed while class is achieved, it may thus be possible to change one's ethnic membership in certain cases. Gender is also a dimension of social differentiation which may influence the significance of ethnicity. Several anthropologists have pointed out how gender may create cross-cutting ties of loyalty, especially among women, who may develop female networks that cut across ethnic boundaries and thereby make them less rigid (Little 19 78). If the relationship between class and ethnicity is difficult to grasp analytically, it is no easier to understand fully the relationship between class and ethnicity on the one hand and gender on the other. It is obvious that a Mauritian woman who belongs to the Franco-Mauritian upper class in many ways ranks higher than a man who is a low-caste Hindu and plantation worker. At the same time, there are conceivable situations where his rank -or power - may be higher than hers, by virtue of his gender. If the public spheres of society - politics, notably - are male dominated, the Hindu worker may be able to participate and compete there, while she will be more or less strictly confined to 'Kinder, Küche und Kirche' (children, kitchen and church), namely the private sphere. Ethnicity and class are to some extent comparable, since both phenomena are related to power discrepancies and competition in public space. Gender, on the contrary, cuts across both ethnicity and class and is not related to either of them in a simple way, yet it is important in a multitude of social situations. This is not the place, therefore, to decide which status is the most 'fundamental' out of class, gender and ethnicity. It is nevertheless a fact that the actors themselves in many societies regard ethnicity and gender as ascribed, and class as achieved. The ethnic status thereby appears as an imperative status, more compelling than other statuses. One can change one's class membership through social mobility; for most people, it is far more difficult (and sometimes impossible) to change one's ethnic membership. 272 Small Places, Large Issues SEGMENTARY IDENTITIES Ethnic organisation can only be including if it is also excluding, and ethnic identity is largely defined by contrasts with others. It has been suggested above that ethnicity cannot by itself define the social identity of a person. This is caused, among other things, by the fact that every person is a member of several groups, and only some of them are formed on the basis of ethnicity. For example, Cris Shore (1993) draws extensively on the theory of ethnic identification when he writes about communist identity construction in central Italy, a non-ethnic kind of identification which was nonetheless very important to the people involved and which resembled ethnicity in its functioning. First, we may imagine the social identities of an actor as a series of Chinese boxes or concentric circles, which includes an increasing number of people as we move from the small to the large scale. As in the segmentary organisation of the Nuer (Chapter 11), the concrete situation decides which group one participates in; which community, rather, that is made relevant. To many of the inhabitants of Mauritius, their social identity as Mauritians became relevant only when the island began to attract large numbers of tourists, as well as becoming industrialised and starting to compete for markets on a global scale. Simultaneously, kinship and ethnic membership continue to play a part in many situations, such as marriage and politics, but in other regards they have been replaced by the more encompassing phenomenon of Mauritian citizenship. In kinship-based societies, a social level like the region or the state may be irrelevant to many of the inhabitants in most social situations, if their needs are by and large satisfied at lower levels of integration, such as those of the household, lineage or village. Any person thus has many complementary social identities, and the context decides which of them is activated at any time. The model of segmentary identities serves as a reminder that identity is not fixed, is not 'innate', but is fashioned in the encounter between an individual and a social situation. The segmentary model may enable us to describe the social identities of a person as, say, citizen of the world, African, Kenyan, Kikuyu, someone from the Murang'a area, member of clan X, member of lineage A. However, as the discussion of segmentary identities in Chapter 11 indicated, individuals also have other identities which cut across a system of concentric circles. For example, apersonmaybe 50yearsold, a man, a spouse, a lawyer, a socialist, a stamp-collector and so on. He or she is a member of many groups with only partially overlapping membership. Which identity is assumed in ;i situation of conflict, when an individual must choose, say, between loyalty to the party and loyalty to the nation or ethnic group, is an empirical question. IDEOLOGICAL USES OF THE PAST As shown previously, myths of origin are powerful devices that have the potential to make sense of the present, legitimise the existing political order Ethnicity 273 and offer group identity. In contemporary ethnic groups and nations, history is used in the same ways. Although written records and archives provide information of a different kind from oral transmissions, they too are ambiguous and open to 'tailoring' and varying interpretations. In a very influential volume entitled The Invention of Tradition (Hobsbawm and Ranger 1983), historians and anthropologists investigate how the past can be manipulated, in order to justify a particular view of the present. As the title indicates, the focus is largely on consciously invented traditions; examples discussed include the Scottish highland tradition and rituals invented by colonial authorities to give the impression that the colonial empire was ancient and 'natural'. However, there is no need to restrict oneself to 'traditions' that are recent creations and that conceal political agendas. The past can be viewed in a multitude of ways, as any comparison of history textbooks from neighbouring countries will reveal. Obviously, the Napoleonic Wars are not described in identical ways to British and French schoolchildren, to mention one example. In many countries, there are heated debates over the 'correct' depiction of past events. As the present changes, so does the past - at least as it is portrayed in authoritative histories. The rise of ethnopolitics in many parts of the world has been an important factor in reassessments of the past, for example in Canadian and American depictions of Native populations. The analytical perspectives on ethnicity outlined here have focused on the social construction and maintenance of ethnic boundaries, the use of history as myth to justify boundaries, the communication of cultural differences (which may be 'real' or 'fictitious') and the situational character of social identification. Several readers may feel that these views on ethnicity, while no doubt analytically useful, are overly dispassionate tools for dealing with a world torn by brutal conflicts often described as ethnic. It is true that anthropological models of ethnic relationships have rarely presupposed power asymmetry, although they do not preclude it. It is also true that the conceptualisation of ethnicity in this chapter has largely stressed the formal properties of ethnic relationships, rather than discussing, say, the civil war in Bosnia, racism in Western Europe or genocide in Guatemala. However, the standard anthropological perspectives on ethnic relationships can help us in understanding such conflicts, provided the analysis also takes the power relationships and the violent dimension into account. It is characteristic of late twentieth-century politics that ethnic identities are used in political legitimation. In Bolivia, rural non-governmental organisations (NGOs) put forward their claims to be the representatives of an oppressed class in the 1970s; the same groups now emphasise that they represent 'Pre-Columbian cultures'. To the poor villagers, this shift makes little difference in practice. The shift from class analysis to the analysis of ethnic relationships mentioned at the outset of this chapter is symptomatic of an ongoing ethnification of politics and identification in many parts of the 2 74 Small Places, Large Issues world, which will be analysed in further detail in the next chapter on nationalism and minority issues. SUGGESTIONS FOR FURTHER READING Marcus Banks: Ethnicity: Anthropological Constructions. London: Routledge 1996. Fredrik Barth: Introduction, in Fredrik Barth, ed., Ethnic Groups and Boundaries. Oslo: Scandinavian University Press 1969. Anthony P. Cohen: The Symbolic Construction of Community. London: Routledge 1985. Hans Vermeulen and Cora Covers, eds: The Anthropology of Ethnicity: Beyond 'Ethnic Groups and Boundaries'. Amsterdam: Het Spinhuis 1994. 18 THE POLITICS OF IDENTITY: NATIONALISM AND MINORITIES — Ernest Gellner tering expressions of national! t ľeoloÍv Th í"? aVoUed 6nC0Un- nation-building is an important ľ f' , ^ľ* °f nationa^m and and de-colonisation National ! , ľCOnnected with colonialism everywhereint^^ aThtľh^a ^ "*** ^ almost not n^an that aD^dt^SilTrT^f4^1^-^^ support, nationalistTdedoT b„nf r T ^ °f for that matto Phenomenon of sucnimtrLc^ ZSha f '""T^ * * ^^ sciences have, since about19RCfl ľ anthropology and other social grips with it analytľcaľy " * """ ^^ l° ^t0 COme to NATIONALISM AND MODERNITY to say, that theľate n 1 t ŕCOrresP°nd *° Political boundaries; that is which woufdTobabrbet f, lrChaÍC SUrVÍVaľ fr0m a rem°te age, -tisation(f^^^^ thatitisactuallyaproducľofmnľ V J ?'* has SraduaIIy become clear was developed in boľtan^S^" ' natÍ°naIÍSm M W kn°W tt Revolution it has a dul 0rľl in ?^ ■"""* the time of the ^ench Romanticism. § " m ^ Fr6nch EnIightenment and German Ä?ÍÍstudy of nationalism and that of «h™*y * definition^^Spec" 1^7 H? ""f ** ^ bUt that is a ****<* special cases of ethnic ideologies. Since most nationalist 275