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.
























Confessions of a useful idiot,

or Why culture should be brought back in

Thomas Hylland Eriksen

LBC Newsletter (Uppsala University), October 2002


It is about time we face it, whether we like it or not: The relativism—universalism tension in anthropological comparison is simply not going to go away. During the past century it has been re-phrased, temporarily transcended, noisily neglected and even ritually exorcised — by recourse to the primacy of the local, or the mental unity of humanity, or the Universal Declaration of Human Rights, or any other pretext one might think of. But no sooner than it evaporates from academia, it re-emerges elsewhere, often with a vengeance. In recent years for certain, the agenda of the debate has been set not from within academia, but in the mass media and in various political arenas.

Consider the situation twenty years ago. With a few exceptions, anthropologists were then basically cultural relativists. A main academic debate concerned the possibilities of reconciliation between relativism and Marxist evolutionism. The non-Marxists did not share this anxiety, and could happily (and often arrogantly) criticise development aid programmes for not taking cultural specificities into account, spend their lunch hour discussing female circumcision in the Horn of Africa (those who had completed their passage into anthropology were naturally in favour of it) and, in their spare time, defend the right to self-determination of cultural groups everywhere. What we said did not really matter. On the few occasions when the itineraries of anthropologists and others intersected, we were tolerated, but were not taken seriously.

This seems a long time ago now. Gone is the intellectual insularity of the discipline, and gone is the splendid isolation of social anthropology as well, as a small subject professing access to arcane forms of secret knowledge.

As I write, it occurs to me that I began my undergraduate studies in social anthropology exactly twenty years ago. The ensuing two decades have seen the diversification and growth of the discipline — phenomenal in some places — and its increasing engagement with the outside world in at least two significant ways. First, anthropologists now routinely work in complex modern societies, and are thus directly confronted, as professionals, with the value systems and predicaments characterising their own society. Secondly, many anthropologists now have policy related jobs, with state administrations, NGOs and so on. As a result, it is now impossible for the discipline (but perfectly possible, and even often defensible, for many of its practitioners) not to be engaged in questions concerning cultural rights and the relationship between universal values and cultural specificity more generally.

Ironically, we seem to be worse equipped for taking on this task than ever. As Mikael Kurkiala (2002) showed in a previous issue of the LBC Newsletter, the current dominance of social constructivism and the fear of objectifying differences seems to have created an anthropological discipline perfectly able to make sophisticated statements about most issues — except, of course, everything to do with culture. In his cautious and ambivalent article, Kurkiala traces the dominance of constructivism to the internal development of the discipline, to wider intellectual trends and to political realities. He is correct, I think, in identifying several sources for the present situation, although the next question ought to be why constructivism has become the main (in some places the only acceptable) approach now and not, say, in the 1970s. To this I will return in a short while.


Kurkiala’s argument can be outlined as follows. Current sensibilities (both political and academic) militate against any appeal to objectively existing entities. Since everything is socially constructed, moreover, everything might have been different (and probably ought to have been). Anybody claiming that pre-discursive, pre-reflexive aspects of sociocultural reality cannot just be wished away (partly because people are generally unaware of their existence), is therefore likely to promote a thinly disguised racist, sexist or ethnic supremacist argument. This, Kurkiala argues, has made it politically and intellectually contentious to talk about ‘ontological differences’ between categories of people. Everything which is not subject to choice (which at a closer examination turns out to be mere consumer choice) is by default politically suspect in this discourse. Although ‘diversity’ is celebrated (in a Benetton fashion) ‘difference’ is obliterated and — these are my words — more often than not associated with religious fanaticism or exclusivist nationalism. The conflation of difference with inequality, moreover, makes this liberalist argument palatable to the academic left, creating an unholy alliance between the market forces and academic chic.

I have considerable sympathy with Kurkiala’s argument (see Eriksen, 2000). Ethnic nationalism and sexism are political evils. Cultural essentialism is a straitjacket. Puritanist traditionalism does not help the Amazon Indians to defend their rights. Yet there is a fundamental baby-and-bathwater problem here. If all we are allowed to do is to study people’s reflexive constructions of their culture, that means relinquishing the constitutive skills of our craft: the methodical awareness of cultural schemata, internalised values and social arrangements which are taken for granted and therefore unknown to non-specialists, but which it is our task to unveil.

The short answer to the above question ‘why now’ is neo-liberalism as the hegemonic global ideology. This ideology of free trade and free choice is so commonplace and uncontested these days that it is rarely represented as an ideology (but rather as ‘common sense’ — keep in mind Geertz’ ‘Common sense as a cultural system’!). It has promoted an upbeat vision of society as based on the free choices of consenting adults so efficiently that the universities these days are not only run as profit-seeking enterprises, but the dominant mode of thought in the very same universities is perfectly compatible with neo-liberalism itself.

As one of the unwitting accomplices (or useful idiots) of neo-liberalism for about a decade (from around 1985), I should add that there were good reasons, and even better arguments, for a social constructivist approach to cultural identity at the time. Cultural romanticism had been unmasked as a product of the Western middle class with spurious connections to the actual cultures in question. Those cultures were rapidly changing and becoming part of the modern world. Clever entrepreneurs were exploiting Romantic ideas of wholeness and continuity with the past in order to pursue their personal political or economic interests. The horrors of nationalist excesses — from ‘ethnic cleansing’ to the insensitive exclusion of minorities — seemed to offer a choice between a battlefield and a marketplace. Besides, our inherited ethnographic map of the world no longer matched the territory, characterised as it was by creolisation, mobility, diversification and the appearance of modern cultural production (and accompanying reflexivity) in the most unexpected places. We were overripe for a new way of conceptualising variations between life-worlds, many of us eager to throw off the shackles of neo-Marxism, Durkheimianism, cultural romanticism or some other holist orthodoxy. It was only after the revolt against the classic concept of culture that many of us began to sense that something inalienable had been given away, and that the anti-essentialist discourse had begun to merge with neo-liberalism. Kurkiala describes the dilemma wonderfully in his conclusion, when he says that ‘the opposite of difference need not be equality but may equally well be indifference’.

Rebuilding Humpty Dumpty?

Modernity is associated with fragmentation, individuation, Gesellschaft and fast-moving changes. If modernity is everywhere, it thus seems, then there can be no hope for cultural communities based on a sense of sharing and continuity. Yet ‘we have never been modern’. There is by now massive evidence to the effect that in spite of the ubiquity of modernity, systematic cultural differences continue to exist. Collective identities based on assumptions of cultural similarity also exist. Moreover, there is a complicated relationship between the two: sometimes there is a convincing fit between culture and group für sich, but sometimes groups are neatly bounded while the cultures they profess to represent are not.

Anthropologists should react against fraudulent attempts to delineate ‘authentic cultures’ and facile evolutionist rhetoric whereby other cultures merely become poor imitations of one’s own (cf. the current immigrant debates in Western Europe). At the same time, we should point out that cultures do exist and not just persons exerting choice, and that they are morally equal until proven otherwise. If we are engaged in cultural imperialism, we should say so. Human rights work, support for the Salman Rushdies of the world and the protection of oppressed Muslim women in Western societies are obviously forms of cultural imperialism, and should be described as such. However, the only defensible form of cultural imperialism is the enlightened one, which acknowledges the existence of deep differences.

After more than a decade of varying applications of culture concepts — from the questionable (Samuel Huntington) to the horrible (Bosnia) — we cannot relinquish it, but we must be careful in distinguishing between cultural differences and the political exploitation of assumed cultural differences. The late Algerian author Rachid Mimouni phrased the problem accurately when, in a trenchant criticism of political Islam, he argued that the problem of politicised religion was that it took the religiosity out of religion. Many practicing Hindus argue along the same lines against hindutva, the doctrine of political Hinduism. In 20th century European history, the expropriation of German cultural history by the Nazis, and on a smaller scale, the expropriation of Viking symbolism by the Norwegian Quisling regime, made it difficult subsequently to use the same universes of meaning for other purposes. They had been contaminated.

Bringing culture back in

Political responsibility weighs heavily on our shoulders these days; our academic or semi-academic statements about nations, ethnic groups or cultures may immediately be picked up, or assimilated more or less subconsciously, by ideologists and politicians wishing to build their reputation on national chauvinism, ethnic antagonism, enemy images. The liberal academic establishment thus wags a warning finger at those who dare to talk of culture as the cause of conflicts, shaking their heads sadly over those lost souls who have not yet heeded the words of Saint Barth and Saint Gellner, who do not realise that culture is chimerical and fleeting, and that reified culture is a dangerous tool in the hands of non-specialists. It is, as Kurkiala points out, not only intellectually correct, but also politically correct to reject all forms of essentialism.

As a result, we have too easily dismissed the question of the role of culture as a determining factor in ethnicity in our eagerness to make ethnic groups and relationships everywhere comparable, fitting the same analytic matrix. The currently dominant framework for identity studies is limited in so far as it rules out the possibility of a literal reading of the cultural universes in question.

It is widely assumed in the research community that ethnicity can be understood without recourse to cultural differences between groups. Now, few scholars of ethnicity deny that such differences may exist; indeed, Barth himself notes, in his famous ‘Introduction’, that if the patterns of behaviour on either side of the ethnic boundary become identical, the boundary will probably cease to be effective. No, the point is rather that cultural differences are not held to account for the creation and maintenance of ethnic boundaries. Only those differences that are made relevant contribute to defining an ethnic relationship, and in other respects, the variation within each group may be greater than the variation between groups. Just as nationhood may be legitimised by referring to a common religion, language or territory, as the case may be, ethnic markers are seen to be arbitrarily selected from a wide cultural repertoire. Clearly, if one wants to emphasise cultural difference, one is unlikely to pick out, as a symbol of one’s collective identity, a trait that one shares with one’s neighbours. Internal variation is undercommunicated, and conversely, differences vis-à-vis others is overcommunicated. This important point has been used in accounts, for example, of the Bosnian war, where a Serb villager could be said to have more in common, culturally speaking, with a Muslim co-villager than with a Serb from Belgrade. This would hold true of both dialect spoken and way of life in general. However, since religion was singled out as the central identity marker, the effective boundary was drawn not between villagers and city-dwellers, but between religious groups. The boundary was thus arbitrary, it was argued; it served to strengthen ideas of fictitious differences and drew on ethnic stereotypes to do so. The people who responded to this kind of boundary-making in so desperate ways were seen not as normatively directed human robots, or as cultural dupes, but as the passive victims of ideology. (So much, by the way, for the liberation of actors from the strictures of authoritarian theory.)

Nothing comes out of nothing. In one of the most detailed accounts of ethnicity in pre-war Bosnia, Tone Bringa (1995) shows that although cultural differences between the ethnic or religious groups were perhaps negligible, and although relations between Serbs, Muslims and Croats could be cordial, there were nevertheless important social boundary mechanisms between them, not in cosmopolitan Sarajevo perhaps, but in the rural areas. Intermarriage was restricted, the close informal networks of friends tended to be monoethnic, and the discrete groups maintained different myths of origin. The intimate sphere, in other words, seems to have been largely monoethnic and by this token, Bosnia was a plural society in the classic sense; the public arenas were shared, but the private ones were discrete. We may choose not to speak of such features of social reality and everyday life in terms of ‘culture’, but they are neither more nor less invented, or real, than anything else. People do not choose their kin, they cannot choose to do away with their childhood and everything they learnt at a tender age. These are aspects of identity which are not chosen, which are incorporated and implicit. Of course, we relate to them as reflexive agents, but we do so within limitations that are not chosen. Such limitations form the objective foundations of identification, on top of which situational selection and relational identities can be played out.

A one-sided emphasis on the manipulation of symbols, the situational selection of identity and the fleeting and indefinite character of culture seems to suggest that nothing really endures, that the social world is continuously re-created, and that constructivist analytical approaches may tell the whole story about human identification. This kind of view, which is rarely far away in contemporary studies of ethnicity and nationalism, or for that matter in currently fashionable social philosophy, would not just be methodologically individualist, but also, it seems to me, a rather strong expression of voluntarism. Such a view would, to the social scientist, imply that he or she would have to unlearn everything he or she has learnt about socialisation, the transmission of knowledge and skills from one generation to the next, the power of norms, the unconscious importance of religion and language for identity and a sense of community. For how are societies integrated, if not through culture, which must not be seen merely as a socially constructed common heritage but rather as a shared system of communication?

What I am saying is, in effect, that culture has, paradoxically, been bracketed – for methodological and political reasons – in contemporary studies of cultural identification, and I will now suggest how it can be brought back in. In a unipolar post-11 September world, we cannot afford not to.

That which is not chosen

In his introduction to The Invention of Tradition, Eric Hobsbawm (1983) distinguishes between invented traditions and non-invented ones, which have a "real" continuity. Against this dichotomy it has been pointed out that all traditions are in a certain sense invented. That is true, but in another sense it is meaningful to distinguish between those traditions that have been consciously invented for political, usually nation- or empire-building, purposes and those which have arisen under other circumstances. Similarly, it makes sense to distinguish between those aspects of culture which are self-consciously worn as identity labels – the tulips of the Netherlands, the tribal dances of Kenya, the steelbands of Trinidad – and those aspects which are quietly reproduced without forming part of self-identity. By singlemindedly focusing on the loud and conspicuous expressions of culture in interethnic contexts, researchers have not only been able to conclude that ‘culture’ largely exists as a political tool, but by implication, the implicit and incorporated taken-for-granted aspects of culture become neglected.

When we talk of history in late 1990s academic discourse, it is often referred to in the David Lowenthal sense of his celebrated The Past is a Foreign Country (Lowenthal 1985) where he, on the basis of a mass of examples of the commercialisation of cultural history, shows that history is often not the product of the past, but of the present; in other words, history as myth, as legitimation, as a particular, ideologically charged reconstruction of the past. However, as every schoolboy knows, history is not merely historiography, but it is also a sedimented past which works in frequently unacknowledged ways, shaping minds and social circumstances over a longue durée and thereby creating very different conditions for thought and action in different social environments. It is certainly illuminating to study how history is being used; but it is also a great intellectual challenge, which may shed important light on the present, to investigate the effects of history that is not being used for a particular legitimating purpose. What is called for, in other words, is a reorientation back to the study of implicit, non-reflexive, doxic foundations for thought and action; historical depth and cultural sensitivity, that which is beyond strategy and self-consciousness. Let me illustrate this point with an example from my own ethnographic work in Mauritius.

Mauritius, a polyethnic island-state in the south-western Indian Ocean, has been described as a veritable laboratory for social studies, encompassing four world religions and a bewildering number of languages on a smallish, isolated oceanic island between Africa and India. The main ethnic groups are recognised as Hindus, Tamils and Muslims, all of them of Indian origin, Creoles of African, Malagasy and mixed origin, Franco-Mauritians of French descent and Chinese, as well as a number of smaller groups based on finer distinctions, which pop in and out of existence depending on the context. Cultural stereotypes are invoked locally to justify the continued existence of ethnic boundaries. To anybody who has done fieldwork in Mauritius, it is easy to argue — and it can be analytically important to show — that the actual cultural variations in Mauritius do not follow the same lines as ethnic variations do. Linguistic variation is much less than one might expect, and the vast majority of the population speaks Kreol, a French-lexicon Creole. Dialects vary not so much along ethnic lines as along regional lines, a point which has also been made with respect to dialects of Serbo-Croatian in Bosnia. Moreover, religious differences are also less conspicious than one would expect. On the village level, religious notions and practices are similar although villagers may belong to Hinduism and Catholicism, religions which are in theory very different. And one might go on to show that neither diet, household structure, leisure activities or representations of politics vary systematically along ethnic lines. In many respects, differences pertaining to social class and the rural—urban divide are more profound than ethnic differences, so that an urban middle-class Creole would have more in common with an urban middle-class Hindu than with a rural working-class Creole.

Yet ethnic boundaries remain relatively solid in most of the Mauritian population, although there has been a growth in the occurrence of intermarriage in recent years. That in itself is significant, but it is not the point I want to make here. What I would like to call attention to, is rather the striking differences in social mobility between the largest ethnic groups, the Hindus and the Creoles. Since the mid-1980s, Mauritius has gone through a dramatic period of economic change, moving from a monoculture dependent on sugar exports to a diversified economy where manufacturing and tourism have attained growing importance. New job opportunities and a rapid economic growth rate have led to a general increase in the material standard of living. However, in this fast process of change, it has become increasingly clear that the Creoles are being left collectively behind. Their unemployment rate is still fairly high, their level of education is comparatively low, and few enterprises are led and run by Creoles. In my very first academic paper, ‘Creole culture and social change’ (Eriksen, 1986), I tried to explain this by referring to cultural values and features of social organisation, and I shall repeat the argument briefly now. The Hindus have a social organisation based on patrilineal kinship and, to a lesser extent, caste. They form corporate groups, many still practice arranged marriage and are endogamous, and expectations of kinship loyalty are strong. Their everyday morality revolves around notions of frugality, prudence, planning and responsibility for one’s dependents. The Creoles, on the contrary, have a social organisation based on the fragile nuclear family. They have no corporate groups, no collective marriage strategies, and shallow genealogies. Among Creoles, individualism and a certain joie de vivre tend to be strong values. An individual Creole who is professionally successful, is rarely expected to aid his relatives in finding good careers; indeed, in some cases, upwardly mobile Creoles change ethnic membership and start identifying themselves as gens de couleur, Coloureds. In Creole communities, the distrust of formal organisation and hierarchy are strong. This ethic may be traced back to the era of slavery, where the conjugal bond was loose or non-existent, and where individual freedom must have been valued extremely highly. It could also be argued that Creole values are indebted to those of their ancient slavemasters; the values of aristocratic France. Whatever the case may be, the Creoles in general, quite contrary to the Hindus, lack cultural values and organisational resources enabling them to take collective advantage of industrialisation.

This means that in order to understand what is locally spoken of as the malaise Créole, it is not sufficient to look at ethnicity as politics and as the self-conscious communication of cultural difference. Cultural differences exist and may become relevant even when they are not consciously "made relevant". The values and way of life associated with the Creoles are counterproductive in political and economic careers; yet, they are demonstrably so deeply embedded in personal experiences and life-worlds that they cannot be accounted for merely by referring to stereotypes and reflexive identity politics.

As Worsley pointed out years ago, one cannot simply exchange one’s ethnic identity for another; life is not a self-service cafeteria (Worsley 1984). In addition, one cannot easily trade one’s childhood experiences and personal network for others; one doesn’t choose one’s cultural universe. Culture is to some extent chosen and constructed, but it is also to a great extent implicit, it has an element of fate, or destiny. This is an acutely relevant point to make in a world where arranged marriages are seen as an evil patriarchal plot, where all cultural significance is taken away from female circumcision, and where the USA displays a growing blindness to anything smacking of non-US life-worlds. Anthropology can and should offer a recipe for disentangling these issues. If we are going to be cultural imperialists, and we probably have no other option, then at least we should be enlightened and respectful ones. To arrive there, we need to understand culture as something which can neither be exchanged in the marketplace, nor reduced to its political face.


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

Bringa, Tone (1995) Being Muslim the Bosnian Way: Identity and Community in a Central Bosnian Village. Princeton, N.J.: Princeton University Press.Eriksen, Thomas Hylland (1986) Creole Culture and Social Change. Journal of Mauritian Studies, 1(2).

– (2000) Ethnicity and culture: A second look, in Herman Roodenburg and Regina Bendix, eds., Managing Ethnicity. Amsterdam: Het Spinhuis.

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

Kurkiala, Mikael (2002) Business as usual? Critical remarks on the trivialization of difference and diversity. LBC Newsletter, 2.

Lowenthal, David (1985) The Past is a Foreign Country. Cambridge: Cambridge University Press.

Worsley, Peter (1984) The Three Worlds. London: Weidenfeld & Nicolson.