This website will be closed down during 2014.

Thomas Hylland Eriksen's new site is now open at

You are now being redirected to the new Engaging with the world site.
























Tensions between the ethnic and the post-ethnic

Ethnicity, change and mixed marriages in Mauritius

Thomas Hylland Eriksen

In Hans Vermeulen and Cora Govers, eds., The Politics of Ethnic Consciousness, pp. 250­76. London: Macmillan 1997.

The issue

When I told a Mauritian acquaintance that I planned to carry out research on mixed marriages in Mauritius, he laughed sadly and said that the best place to investigate this phenomenon would probably be near the Pont Colville Deverell - the highest bridge in the island, which has been a favoured spot for double suicides by young couples unhappily in love, unable to marry each other because of rules of ethnic endogamy and, sometimes, caste endogamy. In this article I shall try to show why my acquaintance was wrong (although he was also right in certain respects) through accounting for the current growth in the number of interethnic marriages in Mauritius and indicating some possible long-term effects on ethnic categorisation and organisation in the island.

When dealing with this kind of issue, it can be highly instructive to focus on some of the variations and changes in the experienced structures of relevance (Schütz & Luckmann 1979) or life-worlds Mauritians live within and act upon. A main task here must be to indicate which paths of action different agents are able to take from where they stand in terms of their personal experiences, their surrounding social environment and their perceived opportunities; to put it differently, how their actions are generated by a combination of variables intrinsic to their life-worlds. This kind of approach, which sheds light on intentional elements rather than causal links, does not encourage the development of deterministic or predictive models, but it may enable us to isolate variables which are particularly important for the direction of thought and action.

No serious scholar, it seems, would today defend the classic view that modernisation inevitably leads to the eventual disappearance of ethnic distinctions and ethnic boundaries. On the contrary, the doctrinaire view on the relationship between modernity and ethnicity is presently that modernisation in important ways inspires and contributes to the intensification of ethnic identifications (see e.g. Nash 1988, Roosens 1989). A variety of explanations are offered to justify this position, which is supported by a wealth of empirical cases worldwide (see e.g. Horowitz 1985). It has been shown that cultural homogenisation (usually within the framework of a state) very often leads to traditionalistic counterreactions from groups which feel that their political interests and/or their sense of self (or identity) is threatened. It has also been argued that modern education and capitalist labour markets encourage the formation of collective ethnic identities and indigenous models of ethnic groups as imagined communities, chiefly because such large-scale processes of standardisation enable persons to perceive themselves as members of anonymous communities comprising a multitude of persons whom they will never know, but with whom they share fundamental characteristics (Gellner 1983, Anderson 1991). The dimension of collective nostalgia has also been prominent in some studies; how people in situations of modern alienation develop an intense longing for a mythical Gemeinschaft past and recreate an "authentic" culture in a modern context (A.P. Cohen 1985, Giddens 1990). Further, the functional aspects of political ethnicity have been studied by many scholars, who have indicated that ethnic symbolism and ethnic principles of political organisation are both politically and emotionally functional in situations of rapid social change - they create political legitimacy and simultaneously provide symbols of social identity (A. Cohen 1974, Eriksen 1993).

Notwithstanding its obvious merits, this view to the effect that modernisation more or less inevitably leads to ethnic revitalisation, needs closer scrutiny. In the following examination of some dimensions of Mauritian culture and society, I shall challenge this view, showing rather that revitalisation and homogenisation or creolisation are two complementary, and sometimes opposed, dimensions of the same process, and that it is too simplistic to claim merely that ethnic revitalisation is a necessary outcome of modernisation.

The poly-ethnic character of Mauritian society and politics
The southwestern Indian Ocean island society Mauritius is often described as a quintessential plural society (Benedict 1965, Simmons 1983, Bowman 1991, cf. Eriksen 1988). The island, located just inside the Tropic of Capricorn about 800 kilometres east of Madagascar, has no indigenous population, and its present inhabitants are the descendants of fairly recent immigrants. Frenchmen, Malagasy, East and West Africans, Indians from both northern and southern India, and Chinese arrived in successive waves and for a variety of reasons from the early eighteenth to the early twentieth century to Mauritius, which in colonial times served equally as a port midway between the Cape and India, and as a sugar colony. Mauritius has successively been a Dutch (1670--1710), French (1715--1814) and British (1814--1968) colony. The French influence remains the strongest of the three, not least due to the fact that planters of French origin dominated public life in the island before and throughout British rule.

The ensuing cultural complexity of Mauritius has frequently been commented upon. It is an island where fifteen languages are said to be spoken (cf. Souchon 1982), but where the official language (English) is scarcely used, where four world religions rub shoulders, where the currency is the rupee and the national anthem is usually sung in French. The Mauritian population of slightly over one million is composed of something between four and twelve ethnic groups, the number (or rather, the level of segmentation) depending on the situation. Officially, four "ethnic groups" existed until they were removed from the censuses in 1983, but they still exist in folk representations: the Hindus (52%), the Muslims (17%), the Chinese (3%) and the "General Population" (28%). However, most Mauritians would agree that Tamils (7%), Telugus (2.5%) and possibly Marathis (2%) should not be lumped together with the majority Hindus from Bihar and Uttar Pradesh in northern India, and that the residual category of "General Population" really encompasses at least three distinctive categories; the Creoles of African and Malagasy descent (23--24%), the Franco-Mauritians of French descent (2%), and the gens de couleur of mixed descent and French language (2--3%). When asked about the number of ethnic categories, most Mauritians would immediately list the Hindus, the Tamils, the Muslims, the Creoles, the Chinese and the Franco-Mauritians. In certain situations, for example concerning marriage, the number of endogamous groups is higher still, as caste and clan membership may be relevant.

As can easily be seen, the criteria for distinguishing between ethnic groups are not consistent. One group, which contains both Catholics and Buddhists, is designated on the basis of geographic origin: the Chinese. Two groups are designated on the basis of religion: the Hindus and Muslims; but the Hindu minorities are distinguished on the basis of ancestral language and geographic origin. The fourth official category, the General Population, contains people of various origins and varied physical appearance, but usually Catholic religion.

Several factors have ensured the continued ethnic segregation of the Mauritian population up to or nearly up to the present, although the importance of some of them has been decreasing (see Eriksen 1988, Bowman 1991, Keng 1991 for details). The division of labour has traditionally been ethnically based. The Franco-Mauritians have composed the upper managerial levels in the sugar industry and the highest bureaucratic positions. Most Hindus and Muslims have been field labourers and smallplanters, although there exist a few wealthy Muslim trading families. Most Chinese have been involved in retail trade. The gens de couleur have dominated the liberal professions of lawyers, journalists and the like, while the Creoles have been artisans, factory workers in the sugar industry, fishermen and hawkers.

This ethnic division of labour is still discernable, but it has in important ways been modified since independence in 1968. Notably, the civil service is now dominated by Hindus. The growth of the tourist and textile industries in the 1980s and 1990s has created new job opportunities for many Mauritians. The official unemployment rate in January 1986 was 23%, whereas two years later there was a shortage of labour. In the mid-nineties, Mauritius is importing labour (from Madagascar, China and India) and exporting capital and industries (to Madagascar). The hotels and factories tend to employ their staff according to qualifications rather than ethnic membership, and many of them have foreign managers with no ethnic loyalties in Mauritius.

Religion is also an important ethnic marker in Mauritius. There is a sense of solidarity among Catholics, Hindus and Muslims, and although there is not a one-to-one relationship between religious affiliations and ethnic ones, there are strong correlations. Although some Creoles have converted to Islam and some Tamils and most Chinese have converted to Christianity, the general picture is that religious groups are associated with ethnic groups. The largest ethnic groups in Mauritius - the Hindus, Muslims and Creoles - each represent one major religion, even if there exist anomalies such as Christian Hindus and Muslim Creoles, and even if the main categories are to some extent internally divided by sects and heterodoxies.

Language is a more complex matter than religion. Nearly every ethnic group has its own ancestral language although, as Hookoomsing (1986) has shown, there is no exact one-to-one relationship between ethnic membership and language. Indeed, since the official abolition of ethnic groups in Mauritian population censuses in 1983, "ancestral language" (which was retained in the censuses) has virtually become a synonym for ethnic membership in everyday speech. However, many of the ancestral languages are no longer in active use, and tend to serve as emblems of ethnic membership rather than being vehicles of communication. In practice, a growing majority of the Mauritian population speaks Kreol most of the time or all the time, a French-lexicon creole developed in the second half of the eighteenth century. In rural areas, Bhojpuri (a Hindi dialect) is still spoken, but few young Mauritians speak it at home. French is the dominant language of the media (and American films are dubbed in French), while only the Franco-Mauritians and gens de couleur speak it at home. A fairly large proportion of Sino-Mauritians still speak Hakka (a southern Chinese language) and read Mandarin. However, most (or nearly all) Mauritians are also fluent in Kreol, which has a triple role of ethnic language (for the Creoles), mother-tongue (for most Mauritians) and lingua franca (for all). Presently, it has no official status, although powerful political groups tried to gain official recognition for it during the 1970s and early 1980s and briefly succeeded in 1982.

Party politics has been organised on largely ethnic principles since the electoral reforms and extension of the franchise in 1948, since when Mauritius has in practice been a parliamentary democracy. Important political parties in the brief history of independent Mauritius have been Labour (Parti Travailliste), which is strongly associated with the Hindus, the Creole/Coloured/Franco-Mauritian Parti Mauricien (later Parti Mauricien Social Démocrate), and the Comité d'Action Musulman (Muslim). Already in the late 1960s, however, there were attempts at breaking with the ethnic logic of Mauritian politics, when the MMM (Mouvement Militant Mauricien) was formed by a group of young students and immediately became an important political force. The aim of the MMM was to become a truly national (in the meaning of supra-ethnic) movement, and it did succeed in this for a few years (Oodiah 1989, Bowman 1991). However, since the latter half of the 1970s, the MMM has increasingly in practice become the political vehicle of the non-Hindu populations of Mauritius. Nine months after the 1982 general elections, where the party won a devastating victory, the MMM split into two factions: the MMM "proper" and the new MSM (Mouvement Socialiste Mauricien), the latter being largely a Hindu based party. Later developments in Mauritian politics have been marked by various more or less enduring alliances - one sometimes gets the impression that every leading Mauritian politician has been allied with every other politician at least once - but the fundamental ethnic logic of party politics and voting clearly prevails in the mid-1990s. Alliances tend to be interethnic, not supraethnic.

I have identified four important structuring principles for ethnic identity and boundary maintenance in Mauritius: the labour market, religion, language and party politics. When, as is the case today, one or several of them changes in overall significance as an ethnic mechanism, it is highly likely that there will be repercussions at the level of the overall system and at the micro level. Notably, a move away from ethnic organisation gives people new experiences, which may not support an ethnic world view; and creates new possibilities for action. Below, I will consider some consequences of these changes.

Non-ethnic aspects of Mauritian public life

When dealing with a complex society like Mauritius, it would be simplistic to argue merely either that ethnic distinctions are simply being reproduced or that they are gradually disappearing. Both kinds of processes can be observed, and I shall later indicate how they articulate with each other. At this point, let me briefly mention a few social fields and arenas where ethnic boundaries are either seriously challenged or irrelevant.

Spoken language is generally irrelevant as an ethnic marker in everyday situations. Although it is true that the vernaculars of the Sino- and Franco-Mauritians set them effectively apart, an increasing number of Sino-Mauritians today speak Kreol at home and fail to teach their children to speak Hakka and to read Mandarin. Among the larger groups, Kreol, French and English hold comparable positions (although, it must in all fairness be said, Bhojpuri is still quite widely spoken in rural areas). Kreol is the language spoken most of the time by most Mauritians (Eriksen 1990). French is the preferred written language of most Mauritians, while English is the language of the bureaucracy and the state. Significantly, Mauritian fiction tends to be written in French while political memoirs are usually written in English.

The major sports clubs were traditionally organised along ethnic lines and were called Hindu Cadets, Muslim Scouts etc., but removed the ethnic epithet due to political pressure in the mid-eighties. Although the connection between ethnic membership and sports club membership or support is still evident, there is a clear tendency for teams to become multi-ethnic. Recent international sports events, such as the regional Jeux des Iles de l'Océan Indien, have also contributed to the development of a shared Mauritian identity (Eriksen in press).

With recreational youth clubs, the tendency is the same. Although they still tend to be organised on ethnic lines, the non-religious clubs have been strongly encouraged to accept members (sometimes spoken of as hostages) from another ethnic group than the dominant one.

The educational system, further, definitely serves to homogenise the Mauritian population. Core curricula are largely uniform cross-ethnically, and pupils from different ethnic groups compete for the same scholarships and, increasingly, for the same jobs. In urban Mauritius outside of the capital Port-Louis, neighbourhoods are segregated by class, not by ethnicity; whereas many rural areas are still largely mono-ethnic. As a result, many Mauritians from different ethnic groups share many of the same childhood experiences as they go to school together.

The recent changes in the Mauritian labour market (from the early 1980s onwards) points in the same direction (Eriksen 1994). Recruitment to the labour market no longer clearly follows ethnic lines, and moreover, the state is replacing kinship and ethnic networks as a provider of welfare benefits.

There are also several other fields where the impact of ethnic distinctions may seem to diminish. Trade unions in tourism and the textile industry are not ethnically based; feminism and environmentalism are becoming political movements which naturally create loyalties which cross ethnic boundaries; leisure activities are increasingly disengaged from ethnic or religious organisation; and national TV encourages the development of fields of shared discourse (Eriksen 1992b).

Although the MMM may not in the long term have succeeded in its "battle against communalism", it did succeed in placing problems of ethnic injustice and ethnic particularism permanently on the agenda. In the 1990s, there is a very open discourse about ethnicity in Mauritian mass media and in many kinds of formal and informal social contexts. For example, the popular press now publishes "scandal stories" about ethnic manipulation in politics, people may be inclined to sue employers if they feel bypassed, and a main preoccupation in public debate - from the Legislative Assembly to the pub - is the relationship between the salade de fruit, where the "components" remain discrete, and the compôt de fruit, where they are forcefully mixed together. There are contemporary social and cultural processes supporting both tendencies. In some respects and among some persons, Mauritius is experiencing a powerful ethnic revitalisation; in other respect and among others, ethnicity is becoming irrelevant as a principle of social organisation.

The increase in mixed marriages

The foregoing sketch has provided an overview of some main elements in Mauritian social organisation. From the perspective of the individual agent, ethnic boundedness has appeared "natural" and rational throughout Mauritian history. Many, if not most, important resources have traditionally been channelled through of ethnic and kinship organisation: employment, material and social security, group belongingness, "old age insurance", marriage and political influence. I shall now indicate further that this is changing, by focusing on one important, visible and quantifiable change in the interethnic behaviour of Mauritians; namely, the current changes in marriage strategies and criteria for spouse selection. In 1960, the number of interethnic marriages was nearly negligible; in 1982, the number was 497, while in 1987, the number had risen to 989 cases, being 8.8% of the total number of marriages contracted in the island (Oodiah 1992:59). The number of divorces is also increasing significantly, and doubled from about 300 in 1982 to about 600 in 1992, which also indicates that the social significance of marriage and the family institution is undergoing a transformation in parts of Mauritian society.

In most of Mauritian history from the 18th century on, different kinds of resources have been bundled together in social networks and organisations based on kinship and ethnicity. The family remains very important, and when asked, many young Mauritians will say that they cannot marry outside their ethnic group "even if I wouldn't mind myself", because the family would reject it. In a society where employment opportunities and financial support is channelled through kinship and metaphoric kinship organisation (that is, ethnic organisation), it can be a very serious thing indeed to disobey parental orders. Although the marriage pattern is changing and that individually based "love marriages" (as opposed to arranged marriages) are now widespread, even among Hindus and Muslims, parental authority remains strong.

How do interethnic marriages function in a society where ethnicity is the most important criterion for ordering the social world? There is no simple answer, but through discussing a few selected cases, I hope to indicate the circumstances under which mixed marriages can be viable as well as some variations, and will finally suggest some possible consequences for Mauritian ethnicity in the twenty-first century.

Case 1. Marie-Claude (née Gita) and Jean.
The couple lives in a coastal village dominated by Creoles, but with a sizeable Hindu minority. They were married in 1976 and have three children. She runs a tabagie (sweetshop), and Jean works at the small coffee factory nearby. When she was baptised as Marie-Claude at the local church in order to marry Jean, her widowed mother did not attend the ceremony, and has since remained adamant that "her daughter is no longer her daughter", meaning that Marie-Claude is not allowed into her home and has little contact with her family. Her younger brother Ram explains that he has nothing against Creoles, but that Marie-Claude is responsible for her social alienation herself, since she can no longer be a member of a Hindu family after converting to Christianity. Jean's family, who are Creoles, were only mildly opposed to the marriage, and are on reasonably good terms with their son and daughter-in-law.

This "openness" of the Creole ethnic category requires some comment. Creoles may describe themselves as a "mixed" people since they have no single shared tribal or geographic origin, speak a "mixed" language (sometimes described as a language composed of French words and East African syntax) and have few if any ancient folk traditions exploited in ethnic boundary processes. In addition, the Creoles do not form corporate groups at the lineage, family or ethnic level (Eriksen 1988:121--134). Compared to the other ethnic groups in Mauritius, the Creoles command few corporate resources. This suggests that there are few strong reasons for Creoles to be endogamous, and indeed they have no strong rule of endogamy. On the other hand, Creoles tend to stress their cultural values, including Christianity, and for a non-Creole affine to be fully accepted in the group, he or she must usually convert to Catholicism. Relatively speaking, the Creoles are more open at the social level than at the cultural level.

The case of Jean and Marie-Claude is interesting in at least two respects. First, Jean was a nonconformist already as a teenager, and had few close friends. In other words, he did not have to worry about losing his primary peer network, which is usually a very important source of recognition and personal identity to a Creole, through marrying a Hindu girl. Second, Marie-Claude quickly became economically independent through setting up her tabagie immediately after marrying. Had their personal circumstances been different, the marriage might never have succeeded. It should also be noted that mixed marriages have always occurred in Mauritian villages and that this one had little connection with the ongoing changes in Mauritian society. It was locally perceived as an anomaly, perhaps even as an aberration, and the couple itself did not challenge the ethnic logic of the village organisation as such. They admitted having broken the rules.

Case 2. Françoise and Mahmood
This is a very different case. Françoise was an upper-middle class Franco-Mauritian girl who fell in love with a lower-middle class Muslim boy. When her family found out, she was sent to live with relatives in France for a year so that she might change her mind, but upon her return, she immediately re-established clandestine contact with Mahmood, and with the help of friends, they arranged to spend two weeks together in the neighbouring island of La Réunion, a French département-d'outre-mer. Despite very strong warnings from Françoise's family, they married five years before I met them. They live in a flat in central urban Mauritius.

Mahmood's family were critical of the marriage, but they eventually accepted it and, Mahmood admits, were "both ashamed and proud" that their son should marry a white girl from a posh family. She converted to Islam, but they both describe themselves as "indifferent Muslims". They have one child who has a Muslim name and who will be brought up as a Muslim, although they admit he will not be a "complete Muslim". For Mahmood, the cost of marrying Françoise was minimal, and since she has converted to Islam, she is accepted as a member of his family. He has a clerical job in his uncle's firm in Port-Louis.

To Françoise, the choice was a more consequential one. She lost her birthright to a secure and predictable life surrounded by material wealth and a tight network of Franco-Mauritian friends and relatives. She says she has very few, but very loyal friends left, and that she is often spoken of in Franco-Mauritian circles as a tragic example of a woman gone astray. At one of the last family gatherings she attended, she wore a sari, and her mother commented, "You are dirtying the blood of your family". Later, her mother said, as an argument against the marriage, that Françoise apparently "did not want any friends". She retorted that she did indeed have some friends, whereupon the mother remarked that they were either not from ta societé ("your society", referring to both class and race) or nonconformists (single, gay or professionally idiosyncratic). At the final quarrel, Françoise says, the mother said that she would rather see her daughter as a drug addict than as the wife of a Muslim. Most of the Mauritian Muslims are descendants of indentured labourers, who were servants and labourers working for the Franco-Mauritians, and many Mauritian Christians, like many European Christians, regard Islam as a threat and as an inferior religion.

Since Mahmood and Françoise do not live with his parents, her personal freedom is greater than it would have been otherwise, but she admits that she cannot smoke cigarettes or drink alcohol at home even if she would have liked to. (Being "indifferent Muslims", they serve alcohol to guests.)

The case of Françoise and Mahmood exemplifies a number of general points pertaining to the viability of interethnic marriages.

First, the question of religion can be crucial. She herself remarks that if she had been strongly religious (Christian), the marriage would not have been possible. (This is not a question of gender. A Christian man would also have been obliged to convert.)

Second, if she had regarded herself as socially, psychologically and economically dependent on her family and "sa societé", she would not have been able to marry Mahmood. Her practical and reflexive ability to sever her ties with her family (which she was on basically good terms with until the dramatic events) was a necessary condition for the marriage.

Third, the most difficult aspect of mixed marriages in this kind of setting - the self-defined plural society with no hegemonic group - may be the identity of the children. As Françoise and Mahmood admit, they are worried about their children, who will grow up as anomalies in a society where ethnic distinctions are seen as nearly as fundamental for a person's identity as gender distinctions.

Fourth, this example may remind us that it is nearly tautologically true that to "marry down", classwise, is socially much more problematic than to "marry up", and this pertains to men as well as to women. This variable does not have a strong bearing on interethnic marriages as such, and is just as relevant in monoethnic mixed marriages between bourgeois and proletarians. In Mauritius, where the correlation between class and ethnicity is traditionally strong, the two kinds of variables are often difficult to distinguish. Consider, for example, the elderly Franco-Mauritian who can frequently be seen roaming the streets of Beau-Bassin on his old moped. He carries a revered aristocratic name and belongs to one of the island's most powerful lineages. When, in his youth, it became known that he had fallen in love with a Coloured girl, he was disinherited and literally thrown out of his family, and has since made his living as a junk merchant. Regarded as an anomaly by everyone, he has no primary network.

Finally, it should be noted that there is no convincing sociological explanation for the fact of Françoise and Mahmood falling in love. Neither of them were "misfits" or "radicals" in their respective social environments. The act of falling in love seems to be an independent variable in this regard, but the realisation of their marriage was, as we have seen, dependent on other factors which needed not be present.

Case 3. Vishnu and Shalini.
This third and final example brings out a further dimension of the issue, and can serve as a starting-point for a general discussion on social and cultural dynamics in contemporary Mauritius. Vishnu, who is classified as a Tamil, has petit-bourgeois and proletarian family origins. He grew up in the cosmopolitan town of Rose-Hill, and due to a combination of family efforts and his personal grantwinning abilities, he was able to pursue university studies in France. Upon returning, he was an underemployed intellectual for several years until, in the early 1990s, he became a successful consultant for private enterprises. Shalini, who is a Hindu (in Mauritius, as noted, Tamils are not considered Hindus in an ethnic sense) of high-caste origin, comes from a wealthy merchant family. She and Vishnu had been sweethearts since their teens, studied together in France, and married shortly afterwards. What is striking about their case is, in the Mauritian context, that it is entirely unspectacular. Neither of the two families was opposed to the marriage, although Shalini's parents were for a long time slightly suspicious of Vishnu - more or less in the same way as an upper-middle class European family would have been ambivalent towards the long-haired, but obviously kind and intelligent radical courting their daughter. Vishnu explains, "I have never thought of us as a mixed couple. We have grown up in the same town, been to university together, shared the same experiences and so on." In certain periods, they have depended on Shalini's family financially, and there is no indication that their marriage has weakened kinship bonds.

This example adds several further points to the discussion.

First, and most obviously, the very notion of "mixed marriage" presupposes an ethnically informed epistemology. When I interviewed a married couple of political activists, asking them a naïve question about their mixed marriage, they quickly retorted: "What do you mean, 'mixed marriage'? We have the same class background, the same kind of education and the same political views. What do you see as 'mixed' about our marriage?"

Second, the case of Vishnu and Shalini exemplifies that Mauritian ethnicity is in many regards a matter of degree in the sense that the perceived distance between groups varies. There is no doubt that some groups perceive themselves as closer than others, and that a Hindu--Tamil alliance is less controversial than a Hindu--Creole or Creole--Muslim alliance would have been. Had Vishnu been a Creole, Shalini's family would probably, despite their liberal attitudes, have been emphatically unenthusiastic about the alliance.

Third, the case indirectly brings out some of the complexity of Mauritian society and the ensuing difficulties in generalising about Mauritian ethnicity. Within the life-worlds of Vishnu and Shalini, the Mauritian white-collar world of university academics, writers, journalists and businessmen, "primordial identities" do not necessarily make up an important dimension of social organisation. Such identities can be activated symbolically, which they are in some cases: Hardly anywhere in Mauritius does one see more young women in saris than at the Mahatma Gandhi Institute, a research institution near the university. However, and that is the point here, in this kind of environment, ethnic identity is not perceived as "second nature"; it has to be chosen self-consciously.

Fourth, the very question of ethnic identity as opposed to other forms of identification is made explicit by Vishnu, Shalini and many others in a similar kind of situation - not just urban intellectuals living in mixed marriages, but by others with comparable experiences and outlooks. Vishnu says that when asked what his ancestral language is, he replies that it is Kreol. The next question is, "but aren't you a Tamil?". His answer would be: "My mother-tongue is Kreol. My parents' mother-tongue was Kreol. My grandparents may have known Tamil, but I always heard them speak Kreol. Why do you think I should go further back than that in order to find my ancestral language?" This exchange brings out the main contradiction in current Mauritian identity politics - which can be described as a tension, sometimes a contradiction, between an orientation towards the past and an orientation towards the present and the future. Unlike Françoise and Mahmood, Vishnu and Shalini do not worry about the ethnic identity of their children. Rather, their main concern is that the children should have a good education.

The three examples discussed above reveal great variations between interethnic marriages. In relation to identity politics, which is our main concern here, they could perhaps be graded on a scale.

Jean and Marie-Claude are trapped inside an ethnic system of signification and organisation, and have improvised considerably to carve out an anomalous existence outside it - at a significant cost. The system of ethnic distinctions is able to absorb a great many marriages of this kind, bringing light-skinned Creole children into Mauritian society (who nevertheless remain Creole children), without changing in its structure and modes of legitimation.

Mahmood and Françoise are actively rejecting and opposing practices of ethnic segregation, but are nevertheless faced with subjectively perceived dilemmas of belongingness, personal sacrifices and the children's social identity. They recognise the continued importance of ethnicity and willingly pay a price for deviating from it.

Vishnu and Shalini, for their part, do not see themselves as being "up against" anything. To them, marriage appears as a voluntary contract between two individuals, which does not necessarily involve families or other groups. Their professional networks, informal social life and perspectives on the future do not necessitate collective organisation based on shared ancestry or ideologies of shared culture.

It is important to note here that the differences between the three marriages cannot be reduced merely to "personality differences", but must be seen in relation to differences in life experiences, generating different structures of relevance and different perceived possibilities of choice. In a sense, the outcome is identical in the three examples - an interethnic marriage - but both the social consequences and the very meaning of the term "interethnic marriage" (mariaz miks) varies with the context.

Creolisation and revitalisation

Let us now move a step further, and reflect on the aforementioned tension in Mauritian society; the opposition between what Hannerz (1990) has labelled, in a comparative vein, "cosmopolitans and locals". First of all, it should be emphasised that there is little to be gained from viewing this tension in evolutionary terms. Some individuals define themselves, and act as, "cosmopolitans" because their interpretations of their experiences and life-projects imply that they do so; whereas others define themselves as "locals" for the same kind of reasons. The point to be made in the context of current changes in Mauritian society is that an increasing number of young individuals experience the world and their own lives in ways encouraging a "cosmopolitan" interpretation of their own identity and the surrounding social environment. To rephrase some points made earlier about social change in Mauritius: Many Mauritians nowadays spend their Sundays in front of the TV set, in the shopping mall or at the beach instead of going to a place of worship; they read French photo-novels rather than the Bible, the Gita or the Koran; they go to cafes and discos where they meet other adolescents with a lifestyle similar to their own but a different ethnic identity; they compete on a par with everybody else for jobs and grades; and they end up working next to, and taking lunch breaks with, persons of different ethnic membership.

This "cosmopolitan" tendency is underpinned at the institutional level by new forms of economic organisation, by the increasing application of principles of meritocracy in the educational system and the labour market (particularly in the private sector), by the growing secular public sphere (cafes, newspapers, magazines, professional organisations etc.) and by increased contacts with the outside world through incoming tourism and economic diversification.

Equally importantly, the importance of kinship and family in the social organisation of Mauritius is decreasing because of the individualistic and meritocratic tendencies in the labour market. Just like work, marriage is becoming a relationship between individuals rather than a relationship between groups.

One immediate outcome of this situation, which is no longer a mere scenario but which is visible (and quantifiable) in urban Mauritius, is the growth of the "Creole" ethnic category. As remarked earlier, the Creoles make up an ethnic category which is not based on shared descent, but on "family resemblances" (Wittgenstein 1983) pertaining to their general lifestyle. Ethnic anomalies therefore tend to be classified as Creoles. "Creole" as an ethnic label in Mauritius is actually a "catch-all" label; a truly residual category absorbing everyone who does not fit well into the other categories, which are legitimised through references to notions of purity and descent. The children of a Chinese--Muslim marriages (a few exist) tend to be categorised as "a kind of Creoles", despite the fact that Creoles were initially defined as Mauritians of wholly or partial African or Malagasy descent.

Through this absorbent quality of the Creole social category, it may be remarked, the native term Kreol (when used about people, not about language) is superbly compatible with the analytic term of "creolisation" as used in the work of Hannerz (1992) and others, where it is conceptualised as a continuous process whereby distinctive "packages" of cultural signification melt into new forms. A possible definition of "a Creole" in Mauritius could be "an individual who holds that his or her ancestral language is Kreol" (note that this is my suggestion, not a local one), thereby acknowledging that his or her origins are mixed - if not genetically, then at least culturally. This option is, of course, open to Hindus as well as Muslims, who thereby do not become fully-fledged Creoles, but "Creolised Indo-Mauritians" whose children may be regarded as Creoles. The Creole category is thus open in several respects, but it remains bounded at least partly because most Mauritians define themselves as non-Creoles.

The next logical step, exemplified through Vishnu and Shalini (and many others), transcends the ethnic logic altogether, rejects "Creole identity" for being a residual category created by an obsolete ethnic logic, and claims Mauritian citizenship as the only rational basis for political identity. Within this world view or structure of relevance, shared culture is caused by the ability to communicate rather than by shared origins. It would be possible to argue, in this respect, that the cultural distance between a rural, proletarian Hindu and an urban middle-class Hindu is greater than between an urban middle-class Hindu and an urban middle-class gen de couleur.

Many thousands of Mauritians today live within an experienced reality of this kind, which was impossible only thirty years ago, when the main social institutions of Mauritius were still tightly tied up with ethnic distinctions. In contemporary Mauritius, the boundaries have become fuzzy. Of course, most Mauritians still think and act largely within an ethnic mode of thought. Still, Creoles may bitterly complain that tu pu malbar ("Everything is for the Hindus") when explaining why they can never expect to find employment in the civil service. And still, a Hindu may tell a visitor that "it's funny, but nowadays, a lot of Creoles look almost like Hindus". However, it can also be observed that a lot of Hindus look almost like Creoles, and this, perhaps, pertains especially to the young, who are constantly exposed to the same influences as Creoles in terms of music, dress, food and so on. However, it is clear that Mauritian ethnicity is in the middle of a phase of transformation where ethnicity is changing in significance and relevance. If the tendencies I have sketched were the only ones, the end of ethnicity might have been imminent. But there are other strong tendencies, and I shall briefly describe their relationship to the processes of creolisation taking place in the economy, in the media and in the intimate sphere.

Until a few decades ago, ethnicity was firmly embedded in politics, the economy and informal social interaction in Mauritius. Ethnicity was highly hierarchical. The changes in post-independence Mauritius have been no less than spectacular. The ethnic foundation of politics, although still strong, has repeatedly been challenged. Principles for recruitment to the labour market are no longer unambiguously ethnic. Educational opportunities have spread and have levelled out some profound (including linguistic) cultural differences. New arenas for informal networking, such as discos, have appeared. Most households now have a TV set, and follow the same programmes. And - as an objective marker of the change - today, nearly ten per cent of Mauritian marriages are ethnically mixed. Far from everybody views this development with delight, and the pressure towards conformity and cultural homogenisation is met with powerful counterreactions from different quarters.

Religious leaders from Hinduism, Christianity and Islam preach tolerance and simultaneously stress the importance of having one faith. Some high-profiled political leaders have also campaigned more or less openly for ethnic solidarity in recent years, and are gaining support. One of them, a Hindu leader, spoke at a public meeting in 1992 about the decline of Bhojpuri, linking it to urban decadence, the replacement of the sari and incense with jeans and the pill; and called for a revitalisation of ancient Hindu values. In line with this logic, a Franco-Mauritian whom I met at a party argued that in Mauritius, one had avoided violent ethnic conflict because one had - up to the present - avoided mixed marriages. (Another guest commented, angrily, that this was tantamount to defending apartheid.) "Traditionalism" and the search for roots takes a number of other forms as well.

These kinds of counterreactions against the homogenisation of identities indicate that many Mauritians today reflexively fashion ethnic identities as self-conscious responses to the tendencies towards blurring identity boundaries and cultural creolisation. Why?

There seem to be two distinct kinds of motivation for subscribing to essentialist, ethnic notions of identity in the current situation.

Most obviously, there are large groups of people who have vested political or economic interests in some kind of ethnic segregation. A rich ethnic group such as the Franco-Mauritians is a very clear example - in their case, the very colour of their skin is a ticket to privilege - but among many Hindus, there is also fear that their privileged access to positions in the civil service is threatened by individualism and meritocracy. Through linking these tendencies to a moral decline, they try to gather the support of people who are concerned with leading a decent life in accordance with established values. During a recent electoral campaign, thus, a false rumour to the effect that Prime Minister Jugnauth's son was engaged to a Muslim girl (the Jugnauths are Hindus) was heard in many Hindu dominated villages. It is not adequate to view this kind of rumour purely as an attempt to discredit the Prime Minister as a moral person, a good Hindu and so on. Economic and political interests are also involved, since rural Hindus remain socially and economically organised on the basis of lineage and kinship. To marry a Muslim therefore, in this kind of context, implies selling out the ethnic estate of Hindus (seen as a metaphoric kin group), which amounts to very real economic interests.

This is not to say that purely instrumental motives underlie ethnicist counterreactions against individualism and meritocracy, but findings from parts of Mauritius where the employment structure is different, indicate that the economic dimension is an important one. If no economic and political resources were channeled through ethnic organisation, it is unlikely that calls for ethnic purity would have mass appeal. At the time of writing, it is still uncertain whether they will.

A different context of ethnic revitalisation is found in the urban middle classes. Often accounted for as nostalgia and romanticism in the professional literature, this kind of ideology has a strong appeal in urban areas in Mauritius. Many Mauritians, among them many urban "cosmopolitans", feel an increasing attraction for their ancestral culture as they approach middle age, many even making pilgrimages to their areas of origin in India. The erosion of the past is countered by a reconstruction of the past, whose architects do not necessarily turn this into a political programme aimed at defending their rights at the expense of the rights of others.

This way of reasoning, which is symmetrical or complementary to creolisation, globalisation and cultural homogenisation (cf. Friedman 1994), seems more difficult to undertake in Mauritius than in many other societies. For one thing, few Mauritians are able to trace their origins accurately. About three quarters of the population are the descendants of either slaves or indentured labourers, and their genealogies usually vanish into the mist of myth after a few generations. Others, including many who are opinion leaders by virtue of being writers and journalists, have origins so mixed that any call for purity would seem absolutely meaningless to them. One of them counts as many as eight different sources of origin - from Wales to Canton.

The dual character of contemporary cultural dynamics

Processes of globalisation and creolisation of culture, moreover, are not intrinsically opposed to ethnic fragmentation, and indeed, cultural creolisation or hybridisation can fruitfully be analysed as complementary to ethnic revitalisation. Groups may become more similar at the level of culture, lifestyle, worldview and so on, while simultaneously strengthening their ethnic (social) boundaries and vice versa (Blom 1969). Finally, the contemporary situation of increased social scale in many societies and the introduction of immensely efficient mass communication technology has led to a vast increase in the sheer quantity of ethnic encounters. Where groups could formerly live in a greater or lesser degree of isolation, they are now increasingly brought into contact with each other. They thereby become culturally more similar and, frequently, more aware and more self-conscious of their differences. They start to compete for the same scarce resources, and an ethnification of politics frequently results (see e.g. Young 1993). Cultural homogenisation may in this way inspire ethnic revitalisation.

These and similar general insights into the mechanisms of modern ethnicity, many of them indebted to the perspectives developed in Barth (1969), are very valuable, but they should not disable us from seeing different and sometimes contradictory processes. A real danger when the analytic focus is placed on ethnic revitalisation is neglect of the far less spectacular, but often no less important processes of cross-ethnic integration, which may also be an outcome of cultural homogenisation (see e.g. Roosens 1989). I have therefore chosen to focus on circumstances under which ethnic identities and the reproduction of ethnic boundaries actually do become less socially important, and have also indicated how the dissolution of ethnic boundaries in some fields is intrinsically related to the upsurge and revitalisation of ethnicity in others in the context of the modern state and capitalist economic system. I have also argued that it is by no means granted that the ethnification of politics and identity is bound to win in the end in every society, and that Mauritius may conceivably become a post-ethnic society in a couple of generations - although, of course, it is impossible to predict the outcome of this kind of complex identity proces. Nonetheless, it is intellectually and politically important to make this point at a time in history when ethnic determinism seems to have become an important folk (and analytical!) model in many societies. By focusing on central dimensions of the life-worlds of differently positioned actors in a single society, I have argued that personal experiences may indeed contradict ideologies and practices which reproduce and strengthen ethnic boundary processes - and that the long-term result may be a fundamental transformation of ethnic relations.

Despite the arguments against ethnic revitalisation on a large scale, and despite institutional changes militating against a new strengthening of ethnic organisation in Mauritius, experiences from other parts of the world tell us that objective processes of homogenisation and individualisation are not necessarily sufficient to level out ethnic distinctions. From the analysis presented in this article and elsewhere (Eriksen 1988, 1992a, 1994), a likely scenario for Mauritian politics in the near future may depict it as a tension along two axes: One axis divides the population in ethnic groups with assumed opposed corporate interests. This is the classic plural society model, which still holds good for the civil service, the sugar industry and religion. The other axis, however, divides the population between "cosmopolitans" and "locals"; between post-ethnic and ethnic principles of organisation and signification. The non-ethnic fields include the new urban space of informal interaction (such as shopping malls, fast food restaurants, discos, cafes and clubs), the textile and tourism industries, and the educational system. The media and party politics are at the crossroads between the two logics. Individual life-worlds and social taxonomies may draw on both sets of fields, but in some concrete situations, relating to, for example, marriage and employment, persons may have to choose between the two. If a sufficient number of Mauritians choose to act on a non-ethnic basis (because such a world makes the most sense to them), the accumulated outcome will be a fundamental change in the organisation of Mauritian society. It is indeed difficult to envision a society where, say, over half of the population are "ethnic anomalies" of some kind or other, and where the cohesive and divisive principles of ethnicity (kinship, shared origins, shared ancestral culture) are still functioning for a majority of the population. Such a change, which I have argued the possibility of, would be comparable to the Copernican paradigm shift, where the number of epicycles and anomalies became too large to handle within the existing conceptual framework. A taxonomic system which is continuously contradicted by experience cannot survive indefinitely.

Concluding remarks

The general points made here could be relevant for, and could indeed be adapted to, many societies in the world. To mention but the most obvious example in contemporary European identity politics: One interpretation of the current war in southeastern Europe, which is supported inter alia by intellectuals in cities like Zagreb, Ljubljana, Belgrade and Sarajevo, is that the conflict is not really an ethnic one, but should rather be seen as a conflict between an ethnic political logic and a non-ethnic one. The conflict, seen in this way, divides the people who justify the war through their actions from those who oppose it. Similarly, it has often been remarked in Mauritius that although the two main communalist political leaders - one Hindu, one Creole - are bitter enemies, they are surprisingly similar in their rhetoric and presentation of self.

It may also be a general point with respect to complex modern societies that individualist career structures and the diminishing importance of religion (both as religious content and as ethnic marker) are necessary conditions for mixed marriages to be stable and successful. However, they are not sufficient conditions. The situation in Mauritius is currently characterised by a tension between ethnic and non-ethnic forms of identification, but it is by no means certain that a logic of social classification not based on metaphoric kinship (i.e. an ethnic logic) will win in the long run. One insight from the last thirty years of ethnic studies which remains valid and important, is that ethnic symbolisation and organisation is incredibly malleable and adaptable. However, we should be careful not to conclude that the mere fact that ethnic categorisation and social organisation may reappear at any time, means that it is bound to reappear. In Mauritius, it is a real possibility that ethnicity as we know it today will actually disappear in a few generations. This does not mean that kinship, "race" and religion will be unimportant in the classification of people, but that the categorical fluidity which is already apparent, eventually will make it conceptually and practically impossible to develop enduring corporate groups and unambiguous myths of origin which reproduce the system of mutually exclusive identity categories characteristic of the "plural society". Instead, a variety of criteria will determine a person's social position, and it may well happen that only a small minority of Mauritians will be able to draw on metaphoric kinship (ethnicity) for their group belongingness. Such a change would not, it should be noted, be a direct product of bureaucratic rationalisation, but would have to be accounted for by investigating the life-worlds and experiences of actors. Ethnicity will become less important only if a decreasing proportion of individual experiences lends credibility to an ethnic taxonomy of the world. Thus, we should be careful not to generalise about "Mauritians" from a limited number of detailed cases. Their experiences differ systematically, and although essentially non-ethnic life-worlds are spreading today with urbanisation and globalisation, ethnic bases of identity and social organisation remain strong in other life-worlds.


Thanks are due to Vinesh Y. Hookoomsing, Cora Govers and Hans Vermeulen for their insightful and useful comments on an early version of the article.


Archetti, Eduardo P. (1994) Nationalism, Football and Polo: Tradition and Creolization in the Making of Modern Argentina. Paper presented at the workshop "Locating Cultural Creativity", University of Copenhagen, October 3-5, 1994.
Barth, Fredrik , ed. (1969) Ethnic Groups and Boundaries. The Social Organization of Culture Difference. Oslo: Scandinavian University Press.
Benedict, Burton (1965) Mauritius: Problems of a Plural Society. London: Her Majesty's Stationery Office.
Blom, Jan-Petter (1969) Ethnic and cultural differentiation. In Ethnic Groups and Boundaries. The Social Organization of Culture Difference, Fredrik Barth, ed., pp. 75-85. Oslo: Scandinavian University Press.
Bowman, Larry W. (1991) Mauritius: Democracy and Development in the Indian Ocean. Boulder: Westview.
Cohen, Abner (1974) Two-Dimensional Man. London: Tavistock.
Cohen, Anthony P. (1985) The Symbolic Construction of Community. London: Tavistock.
Eriksen, Thomas Hylland (1988) Communicating Cultural Difference and Identity. Ethnicity and Nationalism in Mauritius. Oslo: Department of Social Anthropology , Occasional Papers, no. 16.
-- (1990) Linguistic diversity and the quest for national identity: The case of Mauritius. Ethnic and Racial Studies, 13(1):1--24.
-- (1992a) Us and Them in Modern Societies: Ethnicity and Nationalism in Trinidad, Mauritius and Beyond. Oslo: Scandinavian University Press.
-- (1992b) Multiple traditions and the question of cultural integration. Ethnos, 57(1-2): 5-30.
-- (1993) Ethnicity and Nationalism: Anthropological Perspectives. London: Pluto.
-- (1994) Nationalism, Mauritian style: Cultural unity and ethnic diversity. Comparative Studies in Society and History 36(3): 549-574.
-- (in press) We and Us: Two modes of group identification. Journal of Peace Research (forthcoming 1995).
Friedman, Jonathan (1994) Cultural Identity and Global Process. London: SAGE.
Gellner, Ernest (1983) Nations and Nationalism. Oxford: Blackwell.
Giddens, Anthony (1990) The Consequences of Modernity. Cambridge: Polity.
Hannerz, Ulf (1990) Cosmopolitans and locals. In Global Culture, Mike Featherstone, ed., pp. 237-252. London: SAGE.
-- (1992) Cultural Complexity. New York: Columbia University Press.
Hookoomsing, Vinesh Y. (1986) Langue et identité ethnique: Les langues ancestrales à l'ile Maurice. Journal of Mauritian Studies, 1 (2): 126--153.
Horowitz, Donald (1985) Ethnic Groups in Conflict. Berkeley: University of California Press.
Keng, Jean-Claude Lau Thi (1991) Inter-Ethnicité et Politique à L'Ile Maurice. Paris: L'Harmattan.
Klass, Morton (1991) Singing with Sai Baba. The Politics of Revitalisation in Trinidad. Boulder: Westview 1991.
Nash, Manning (1988) The Cauldron of Ethnicity in the Modern World. Chicago: University of Chicago Press.
Oodiah, Malenn (1989) MMM 1969--1989: 20 ans d'histoire. Rose-Hill (no publisher stated).
-- (1992) Les années décisives: 1982-1992. Rose-Hill: Océan.
Roosens, Eugeen E. (1989) Creating Ethnicity. London: SAGE.
Schütz, Alfred and Thomas Luckmann (1979) Strukturen der Lebenswelt. Band 1. Frankfurt-am-Main: Suhrkamp.
Simmons, Adele Smith (1983) Modern Mauritius: The Politics of Decolonization. Bloomington: Indiana University Press.
Souchon, Henri (1982) Le mythe de quinze langues pour un population d'un million. In H. Unmole, ed., Proceedings, National Seminar on the Language Issue in Mauritius, pp. 78-83. Le Réduit: University of Mauritius.
Taylor, Charles (1992) Multiculturalism and "the Politics of Recognition", Amy Gutmann, ed. Princeton: Princeton University Press.
Wittgenstein, Ludwig (1983 [1958]) Philosophical Investigations, tr. G.E.M. Anscombe. Oxford: Blackwell.
Young, Crawford, ed. (1993) The Rising Tide of Cultural Pluralism. Madison: University of Wisconsin Press.

©Thomas Hylland Eriksen 1996