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.

























Walls: Vanishing boundaries of social anthropology

Thomas Hylland Eriksen

Trial lecture for Dr. Polit. degree (chosen topic), 14 September 1991

Later version published in Anthropological Notebooks, 1996


The term 'globalisation' has become a central, if contested, term in contemporary social anthropology. This essay traces the preconditions for its emergence, discusses some widespread misreadings of the concept, and argues that the study of global--local relationships may save anthropology as a comparative discipline from the important, but ultimately irresponsible criticisms of postmodernist deconstruction. The neo-diffusionist stance developed in this essay stresses the need to rephrase some of the central conceptualisations of the discipline while retaining the classic preoccupations of social anthropology; to understand and account for social and cultural variation, and to relate the symbolic realm to social organisation.

* * *


A main controversy in the social anthropology of the 1990s is centred on the concept of globalisation. According to some anthropologists, we are in the middle of a paradigm change where the study of remote connections, dispersed networks, flow and flux is about to replace the holistic study of localised communities as the dominant research strategy; where the subject as a whole is on the verge of becoming a comparative study of modernities, since, it is argued, modernity is the most pervasive and complex socio-cultural phenomenon extant. Others would argue, against this view, that even large-scale social and cultural processes, fast change and highly diversified, unbounded societies may profitably be studied by means of our traditional conceptual and methodological toolboxes. Further, the sceptics would claim, the actual importance of globalisation - seen as uneven modernisation on a global scale - is frequently exaggerated, and many, many millions of earthlings are still relatively unaffected by modern mass media, capitalism, the state and labour migration - just to mention some of the most important features of globalisation.

This essay is concerned with this development, which is related both to changes in the world and to changes in anthropological styles of thought and research. By way of a lengthy introduction on walls, the idea of primitive society and the globalisation of culture, some of these changes will be sketched. Drawing on my own ethnographic material from Mauritius, I shall then raise a few questions concerning the relationship between our field of enquiry - social anthropology - and the contemporary world. In conclusion, I shall defend the view that classical anthropological epistemology and methodology, while far from obsolete, are quite inadequate if we are to fully understand cultural variation in the contemporary world. This argument is no longer an original one, but it concerns us all as social and cultural anthropologists, and no matter our various specialisations, we shall have to deal with it since it crucially impinges on conventional conceptualisations of society and culture.

It should at the outset be remarked that there are currently two dominant ways of handling the current implosion of the classic concepts of culture and society: deconstruction and the comparative study of local--global relationships. The neo-diffusionism implied by the very term globalisation hints at a new universalism, a new attempt at making discrete phenomena comparable. Deconstruction, on the contrary, which in its anthropological variety has taken the shape of a penetrating autocritique, appears as a strong form of relativism whereby all anthropological concepts are ultimately inadequate as tools for making sense of cultural variation. My concern in the following is to show the usefulness of neo-diffusionism, which is seen as a possible escape route from the infinite regress of deconstruction.

On the eve of the last decade of this millennium, every reader of this text (if not exactly every inhabitant of the world) remembers, waves of euphoria were catapulted through invisible globalising networks with their nodal points in television satellites, radio transmitters, computer modems, telephones and fax machines. These waves reached remote areas like Trinidad and the Trobriand islands, and engendered a widespread feeling that a new era in the history of humanity was about to begin. The collective memory of small groups may be short; the collective memory of the cultural world-system is dramatically shorter. The event spurring the euphoric celebrations referred, through its immanent metonymic chains, to a period of little more consequence than a footnote in human history. Nevertheless, the opening and eventual demolition of the Berlin Wall shall defend its place as a frontispiece to this essay, which will revolve around anthropological conceptions of difference and identity in the contemporary world - different in many ways as it is from the late Victorian world of the explorers, missionaries, armchair theorists and ethnographic butterfly-collectors who were the pre-Malinowskian founding fathers of our discipline.
Like the French word mur, the English wall clusters together very different referents. In German and the Scandinavian languages, we distinguish between, on the one hand, Mauer and mur, respectively, and on the other hand, Wand and vegg, the latter referring to the inside or outside walls of a building, the former being a rather more awesome and monumental structure, such as the wall featuring in Sartre's short story 'Le Mur', in Roger Waters' The Wall or the Berlin-Mauer. (One may remark in passing that the segmentary character of the wall as a dividing agent is better taken care of in German and Scandinavian than in French and English.) In the present context, the term will basically refer to the metaphor of eine Mauer, a solid and powerful structure ensuring separation and distinctiveness.

The demolition of primitive society
The demolition of the idea of primitive society is in some interesting ways analogous to the demolition of the Berlin wall: the importance of the event is obvious, although in some excited quarters, its significance has been greatly exaggerated. Besides, it has a strong bearing on the present argument. I shall therefore describe this deconstruction in some detail.
If modern man had not known of actual primitives, he would have had to invent them. In a sense, this is exactly what he (yes: it is usually a 'he') has done, projecting good or bad qualities of his own society, as the case may be, into his fictions of Otherness. Until the rise of modern anthropological research, 'savages' were either depicted as cruel brutes who testified to the superiority of Christianity or the British Empire; or they were envisaged as noble creatures leading humble but immensely rewarding lives in profound, almost instinctive deference of Mother Nature and in love of one another.

With the emergence of the embryonic social sciences in the mid-nineteenth century, so-called scientific theories of primitive society were developed. In the beginning, the myth inherited from Freud, Engels, Morgan and their likes goes, men and women lived in undifferentiated bands. There was no incest taboo, there were no rules of exogamy, formal differentiation of rank or rules of inheritance and succession. Then, due to some evolutionary leap, envisaged as the development of private property, the appearance of the family, the division of labour or the primaeval emergence of the Oedipus complex, humanity fell from innocence, and walls of power were erected; walls pitting man against man, man against woman, cunning against sincerity, greed against humility, the proprietors against the propertyless, the healthy against the lepers, artifice against authenticity. In Europe and elsewhere, literal walls protected the insiders of towns from the outsiders. Eventually, the myth continues, the feudal state evolved, finally transforming itself into the modern capitalist state, which imprisoned the creative labours of man and his fertile natural imagination within the walls of its bureaucratic industrial machine. The organically integrated Gemeinschaft turned into the fragmented and anonymous Gesellschaft (Tönnies); the dead labour seized the living labour at hyperbolically increasing rates (Marx); the iron-cage of bureaucracy (Weber) tightened its grip, the maladies of anomie (Durkheim) were rife, and men and women were enslaved by the clock (Mumford).

Societies which had been spared the mixed blessings of this evolution were destined to form the subject-matter of anthropology or comparative ethnology. Whether or not their inhabitants were fancied as noble savages, they served well as the white mice of anthropology and as live arguments against the anomie, alienation or excessive discipline - pick your choice - of modern society. Thus was the underlying evolutionary model of much classic anthropological research, notwithstanding its sometimes programmatic anti-evolutionary bias. The Boasian relativists lamented the fall from innocence, viewed from their comfortable vantage-point in the hubris and progress of their age; the French and British structural-functionalists refrained from subjecting their own society to crude objectivist explanations, yet applied them rigorously to others; and that master of primitive man, Claude Lévi-Strauss, who has virtually devoted his life to the promotion of 'neolithic intelligence', saw the synchronic study of cultural variation as being essentially equivalent to the diachronic study of history (Lévi-Strauss 1962).

The duality between the West and the Rest, to paraphrase Sahlins (1976), remains a forceful one in our collective mind; it was a constituent element in the formation of our discipline, and it is a wall which has not yet fully been demolished. This duality, on which our discipline has rested comfortably for more than a century, has time and again been challenged; by mild-mannered iconoclasts like Bateson, who proposed new and less dichotomous paths for the mind; by Evans-Pritchard, the founder of the ethnographically based theory of knowledge; by late Malinowskians like Firth, who insisted on the primacy of practice and the particular - and more recently, by American critics of anthropological writings, who draw on interpretations of deconstructivism, contemporary hermeneutics and literary theory to show the limitations of the traditional ethnographic narrative, particularly in its representation of 'The Other'.

In his study of the rise and fall of the idea of primitive society, Adam Kuper seems to close the case by arguing forcefully that the very idea of primitive society has had its time, and that this fact may eventually indicate the end of social anthropology as we have known it (Kuper 1988: 243; but cf. Kuper 1994). The development of an anthropology without the primitive, as Kuper puts it, would imply a collapse of the implicit evolutionism and explicit relativism seeing societies as self-contained systems and presuming radical cultural discontinuity between ourselves and our objects of study. Recent developments within the discipline suggest that this may be about to happen.

First, anthropologists increasingly study modern societies without pretending that they are 'all the same'. Secondly, recent epistemological critiques have indicated the arbitrary nature of typological distinctions between 'kinds' of societies. Thirdly, and that is the main focus here, boundaries between societies are in certain important respects dissolving in a very visible way. One may also add that apartheid and similar ideologies promoting the enforced maintenance of cultural boundaries are clearly inspired by a reading of anthropological notions of insular, integrated and bounded cultures. According to the theory (if not the practice) of apartheid, cultures cannot be ranked, but they should nonetheless be allowed to evolve according to their unique internal logic, and for this reason, modernity should not be allowed to corrupt traditional African societies: they should be kept isolated, in apartheid (which is Afrikaans for apartness). Ask an ANC militant what he thinks of cultural relativism, and the point will be made emically. Let us assume, then, as a premise for this discussion, that the wall separating the monolithic, thoroughly rationalised Us from the intriguing and primaeval Them, is finally about to come tumbling down.

We may flatter ourselves in a Popperian manner and attribute such a turning away from simplistic evolutionary dichotomies to the merciless self-criticism inherent in the anthropological practice. However, it is beyond doubt that changes in the outside world (i.e., aspects of the world which are not constructed by anthropologists) have made important contributions to the increasing uneasiness with which we use dichotomous models dividing the world's societies into two categories. For the Others have caught up with the ethnographic practice, rather than the other way around. They are demonstrably no longer distant 'savages' or acephalous segmentary tribes; they are no longer illiterate half-naked exotics informing the adventurous, but confused traveller that they are 'really' red macaws, and nowadays some of them may even correct faulty details and misinterpretations when they are casually leafing through the ethnographer's draft manuscripts. The walls dividing civilisation from savagery have been demolished, if at all, by insistent savages. Unless we pretend that the tribal world remains intact because a few groups seem to have avoided the onslaught of modernisation, we shall be compelled to realise that many of our dear Azande, Trobriand islanders, Kachin and Swat Pathans are, like ourselves, citizens and wage workers who meet every afternoon in the neighbourhood cafe to discuss the current situation in the Balkans. The vernacular equivalent to the word 'country' forms part of their everyday vocabulary. They tell the anthropologist that 'if we did not have kastom, we would be just like the white man' (Sahlins 1994: 378). The world has in many respects been decolonised, and anthropology must come to terms with this - provided, of course, it is willing to take the challenge.

The globalisation thesis
At the inception of our discipline, there were still white spots on the map. Simultaneously, our subjects for investigation were gradually colonised. They no longer are. Some of them even refuse to offer themselves to the appropriation of anthropologists: they refuse to be our objects of investigation. (This is glossed, among anthropologists, as 'mounting difficulties to research'.) They are citizens in nation-states, and indeed, social and cultural processes channeled through the institutional dimensions of modernity are at this very moment busy ravaging the heartlands of the former anthropological laboratory. Let me briefly mention some general tendencies of social change, although they are already quite familiar from the literature on bureaucratic society. Let me also stress that the processes which I describe are happening now; they have not already happened. That is to say, I speak of an emergent pattern of relationship, not an established one.

First, a monetary economy based on wagework has become the norm, if not a universal practice, in most parts of the world. Such an economy is encouraged by states, which receive important revenue through direct and indirect taxation. States have become the most powerful absentee landlords, and the omnipresence of money integrates an unlimited number of people anonymously into a vast system of exchange. The temporal structure on which this depends is linear and irreversible.

Secondly, formal education is nearly universally recognised as an important means for the achievement of rank, wealth and related benefits. This entails, among other things, literacy, the standardisation of languages and the suppression of minority languages. Two hundred of the original two hundred and fifty Australian languages have been eradicated, which is a testimony not only to literal genocide, but also a strong indication of cultural genocide.

Thirdly, political units of prime importance to the majority of mankind are political parties, organised at a nation-state level with local branches. Position in political parties is ostensibly achieved, not ascribed.

Fourthly, official ideologies in virtually every country in the world are nationalist in character, and everybody is to a varying degree forced to be a citizen. The nation-states require their citizens to adhere to an abstract ideology of metaphoric kinship, and to make personal sacrifices for the betterment of the abstract community of the country. In return, the nation-state presumably offers protection and career opportunities.

This list could have been made much longer, but I shall stop here. The main point is that the fact of the modern nation-state seems to create a uniform and universal framework for social organisation on a very large scale. Of course, everybody is not affected in the same way, but virtually everybody has to cope with aspects of the nation-state and capitalism. Hardly anybody is totally unaffected in the contemporary world. Even our celebrated significant Others, the Yanomamö, are currently negotiating with two nation-states and desperately trying to ensure their survival as a culture-bearing group through transnational networks including the global media, the WCIP (World Council of Indigenous Peoples), and invocations of international law.

At this very general level, there can be no doubt that a massive process of global cultural homogenisation is taking place; that the temperature of hot societies (Lévi-Strauss) is spreading in conformity with the second law of thermodynamics, as it were, rapidly heating up the erstwhile cold societies.

The globalisation of power and the interrelatedness of systems of exchange are no less spectacular. Truly global processes affect the conditions of people living in particular localities. Giddens (1990: 124) has remarked that risks are globally shared in the age of the nuclear bomb and potential ecological disasters. One may also note that the economic conditions in localities often depend on events taking place elsewhere in the global system. If there is an industrial boom in Taiwan, towns in the Midlands will be affected. If oil prices rise, this implies salvation for the Trinidadian economy and disaster for the Mauritian one. Charles Tilly has suggested this delineation of global processes.

A sensible rule of thumb for connectedness might be that the actions of powerholders in one region of a network rapidly (say within a year) and visibly (say in changes actually reported by nearby observers) affect the welfare of at least a significant minority (say a tenth) of the population in another region of the network. Such a criterion indubitably makes our own world a single system; even in the absence of worldwide flows of capital, communications, and manufactured goods, shipments of grain and arms from region to region would suffice to establish the minimum connections. (Tilly 1984: 62)

In a word, the boundaries between societies and cultures, which were never absolute, are becoming increasingly fuzzy in our minds.

A further point, which is immediately relevant to our branch of comparative sociology, is that patterns of consumption seem to merge in certain respects. A precondition for this to happen is the more or less successful implementation of certain institutional dimensions of modernity, notably that of a monetary economy - if not necessarily wagework and literacy. The ever-increasing transnational flow of commodities, be they material or immaterial, seems to create a set of common cultural denominators eradicating local distinctions. The hot-dog (whether halal or not), the pizza and the hamburger are empirically at the apex of world cuisine; identical pop songs are played at apparently identical discotheques in Mauretania and Mauritius; the same Coca-Cola commercials are shown with minimal local variations at cinemas all over the world. And investment capital, military power and world literature are equally being disembedded; they no longer belong to a place (Giddens 1990). Territorial walls have been demolished to this effect. With the development of the jet plane and the satellite dish, distance no longer seems a crucial limiting factor for the flow of cultural signifiers. There are Pakistani migrants settled in Oslo who follow Pakistani news on the Internet and watch new Pakistani films on their video. A future scenario which could be extrapolated from these tendencies might therefore be a world where cultural variation is not related to space (cf. Hannerz 1992) - where the spatial dimension has imploded.

Several social theorists have responded to these emergent changes in the external social world. In anthropology and neighbouring disciplines, Wallerstein's and others' neo-Trotskyite notions of world-systems (Wallerstein 1979) are currently being transformed into theory of culture through a hastily prepared marriage with McLuhan's (1964) and others' theory of modern mass media, injected with insights from recent grand theory on nationalism and the nation-state (see, for example, Robertson 1992; Hannerz 1992; Appadurai 1990; see Worsley 1984, for a slightly different perspective).

According to this cluster of notions, which can loosely be labelled the globalisation thesis, the Entzäuberung (Weber) or disenchantment of the world is about to be completed; the contemporary world is secular, rationalised, continuous and woven together by dense networks of communication and exchange. Jean-François Lyotard (1979), better known as the ambivalent high priest of post-modernity than as a critic of globalisation, sees these tendencies as being inherently levelling and destructive. In Lyotard's view, the social world is being moulded, transformed and reduced to a language intelligible to computers. From a slightly different perspective, the American historian of ideas Francis Fukuyama (1989, 1992), who rose briefly to fame around 1990 for proclaiming the end of history, argued, slightly nostalgically, that profound ideological difference was a thing of the past, and that the liberalist ideology associated with capitalism and bourgeois democracy had virtually become universal. The Hegelian dialectic of history had reached its end, claimed Fukuyama, an employee of the U.S. State Department. A social theorist of Ernest Gellner's kind would fit into the same rough category, as he, too, highlights the apparent disappearance of profound cultural differences. At a conference on nationalism, Gellner remarked that it may well be that people in different areas still utter different noises, but nowadays they say pretty much the same things: the political culture of the world has become one (Gellner 1991; cf. also Gellner 1992). According to such a view, conservative Islamic societies, for example, would appear to be a variant of modern society, or at least as being crucially interdependent with modern societies, rather than something qualitatively different. This assumption, of course, cannot be taken for granted. Nevertheless, the main point is well taken: The world has shrunk, and global links function at the representational level as well as on the economic and macropolitical levels. As Miller (1992) writes, rhetorically: 'What proportion of the world's population today has not been exposed to the concept of class and the reorientation of their cultural perception in the light of this exposure?'

To sum up: The argument of globalisation goes approximately like this. In the post-colonial world, that is to say the world of the micro-chip, general-purpose money, linear time, the satellite dish and the jetplane, culture is no longer restricted to particular places - it has in important respects been disembedded from spatial structures. Transnational and non-localised networks of communication and exchange function alongside localised processes of modernisation and integration into nation-states, to the effect that cultural variation is forcefully being channeled through the universal interfaces of modernity. As Giddens (1990) puts it, rather bluntly: The world is becoming a single place.

However, as every serious anthropologist working within the emerging globalisation paradigm would stress, 'no total homogenization of systems of meaning and expression has occurred, nor does it appear likely that there will be one any time soon' (Hannerz 1990: 237). Global culture

is marked by an organization of diversity rather than by a replication of uniformity. But the world has become one network of social relationships, and between its different regions there is a flow of meaning as well as of people and goods (idem.).

The Rushdie affair from 1989 on exemplifies the 'flow of meaning' aspect mentioned by Hannerz: A book published in Britain provoked a reaction in Iran, and as a consequence, virtually half of the world's population were immediately involved in a public debate on religion versus civil rights.
The globalisation theorists find themselves in a complementary relationship to those romantics who would perhaps, if pressed, prefer to keep their traditional peoples or 'Others' in a state where they would be allowed to remain authentic. Lévi-Strauss himself said, in an interview in L'Espresso in 1986, that he is mainly interested in phenomena which no longer exist. To him, the contemporary world holds no charm because worldwide cultural homogenisation has, in his view, obliterated those radical cultural discontinuities which he has seen it as his task to understand and account for. The German anthropologist Hans-Peter Duerr, who has also devoted himself to understanding radical cultural variation, has increasingly been compelled to use archeological and historical sources when arguing, in a vaguely 'original affluent society' vein, that modernity and the state have shamelessly reduced the depth and scope of the human experience (cf. Duerr 1985).
Extrapolating from the recently coined notion of globalisation, it might be tempting to model the world as a segmentary lineage, its chief constituent segments being nation-states, which occasionally unite at the highest conceivable level of scale, namely the General Assembly of the United Nations. The world according to the globalisation schema could also be conceptualised as a kind of generalised entropic field where all differences between local world-structures are shaved away, where individual identity becomes fuzzy and where individuals become interchangeable at a global level.

The anthropological community was never impressed by this line of reasoning; world-system theory, to mention one important example, never really caught on within our discipline. We knew too much, it seemed, about the peculiarly local dynamics of our field sites to be seduced by general theories of the global. Such anthropological arrogance may sometimes be misplaced, but sometimes it is pertinent, as I shall argue later. Further, the idea that we are not all becoming culturally identical does not require justification from anthropologists who have always concentrated on intrinsic processes taking place in unique localities. In fact, it is an important insight from recent studies of modernities that modernisation and increasing scale in social organisation are marked by a dual process of simultaneous homogenisation and differentiation. Some differences vanish, whereas others emerge. This is plain Parsonian sociology. As Friedman puts it:

Ethnic and cultural fragmentation and modernist homogenization are not two arguments, two opposing views of what is happening in the world today, but two constitutive trends of global reality. (Friedman 1990: 311)

Anthropological gut reactions
The intuitive scepticism displayed by many anthropologists against rather sweeping statements describing the world as being 'one place', has prevented many among us from appreciating that there are in fact profound changes taking place this very moment, which contribute to modifying our subject-matter, in most parts of the world, in substantial ways. One common type of reaction among anthropologists faced with such changes in the social world, can be labelled the plus ça change stance. According to this view, appearances of modernity are frequently superficial, and when all is said and done, the exotic societies really stay the same. Popular ethnographic films such as Trobriand cricket (Leach & Kildea, 1974) would seem to confirm such a view. The film shows how cricket has been introduced into the Trobriand islands apparently without altering the fundamentals of Trobriand culture; rather than adapting the natives, the institution of cricket has itself been adapted to suit the indigenous requirements of the natives. According to such a perspective, we may safely go back to our old toolbox; that is to say, we will not need to modify our vision of the world as consisting of discrete, relatively isolated cultures or societies because of mere processes of diffusion of cultural form or superficial appearances of modernisation.
Another reaction consists in insisting that modernities are peculiar and extraordinary phenomena, and that we should therefore mainly focus on non-modern or non-industrial societies, which display a much wider range of variation than modern ones do. That would be the romantic view as well as the typical positivist view, which both tend to envisage the world as an archipelago of societies or cultures or, less charitably phrased, as a laboratory for cross-cultural measurements. That would also be the view of those who more or less programmatically reject development projects because they interfere with the 'local culture'.

The third typical reaction, related to the previous one, argues on epistemological grounds that social anthropology must be the study of otherness, since our own culture is necessary as a basis for comparison.

The fourth typical reaction, which is the position represented in this essay, consists in discarding the dichotomous classification of societies into the West and the Rest, which also implies that we cease to regard 'modernity', 'nationalism' and so on as uniform social and cultural systems. Instead we should study actual social systems, warts and all.

For like it or not, it is an indubitable fact that we live in a world where the local increasingly bears the mark of the global. Neither anthropology nor the natives benefit from the excessive romanticism, the widespread anti-modern moralism and the view that tropical societies ought to be spared from the influences of the so-called Western world. Societies change, for better or for worse, and instead of pretending that they do not, or proclaiming that they should not, we might get down to looking into what is actually going on. A family network in a Lima shantytown may initially seem less fascinating to an anthropologist than a New Guinea highland clan, but it may on closer investigation turn out to be just as revealing of the variations of human culture and social organisation. A major merit of recent writings on the globalisation of culture indeed consists in reminding those who still rely on the image of the primitive that they are really living in the past. Certain institutional and symbolic dimensions of modernity are nonlocalised (which is not to say that they are omnipresent); most of the world's inhabitants are in different ways citizens; formal schooling and literacy are widespread even in remote areas; wagework has replaced so-called domestic or tribal modes of production on a large scale; consumption based on monetary exchange has replaced non-monetarily based consumption on an even larger scale; and modern mass media have disseminated fragments of shared meaning-structures so widely that it may indeed in certain respects be correct to talk of global culture. There are truly global issues in the world, and they are discussed locally as global issues. The Balkan war has affected the value of the Deutsche Mark, provoked worried editorials in Mauritian newspapers, and is being discussed in cafes and homes all over the world.

On the other hand, it would clearly be nonsense to claim that we are all becoming 'the same'. Such statements reveal, if anything, a lack of ethnographic sensitivity. Nonetheless, the substantial changes which have taken place in the world since the inception of our discipline provide ample reasons for a reassessment of our entire endeavour, which initially drew its raison-d'être - on both sides of the Atlantic - from the assumption that the world was an archipelago of societies or cultures which could be grouped in two rough clusters or categories. The companion idea postulating the uniformity of 'Western culture' as a basis for comparison, absurd from the outset, has become even more curious in the light of globalisation.

The notion that cultural variation remains discontinuous in a systematic way is virtually part of our habitus (cf. Ingold 1993), and anthropologists are also fond of pointing out that predicted cultural 'melting-pots' never came about. Statements of this kind are too general to be of any real value. For example, one may definitely claim that a cultural melting-pot has come about if virtually all that remains of the distinctiveness of a minority is the self-awareness of belonging to a distinctive group.

Taking our cue from the notion of the globalisation of culture, we may thus pose the following questions: - In what respects do the walls come tumbling down, in what respects do they remain unmoved, under which circumstances are new walls erected, and finally: What are the mechanisms of entropy-resistance preventing the dissolution of boundaries between symbolic universes?

Entropy and negentropy in a Mauritian town ward
Writings on the globalisation of culture have been littered with anecdotes revealing disembeddedness of cultural signs or links between the local and the global; there has so far been relatively little contextually sensitive research on the actual implications of such processes (although matters are improving). I shall now provide a brief example, a main purpose of which is to emphasise the continued relevance of anthropological fieldwork in a world increasingly marked by symbolic and social structures too vast in compass to be investigated anthropologically.
Rose-Hill is a Mauritian town of some 40,000 inhabitants, according to statistics. However, boundaries between Mauritian towns are unclear, and it would probably be most accurate to describe Rose-Hill as one of five or six nodes along the nearly continuous urbanised stretch from Port-Louis to Curepipe, where about half of the Mauritian population of a million live.
The quarter of Roches-Brunes, located on the western outskirts of Rose-Hill, is dominated by a municipal housing estate (cité ouvrière), and most of the approximately 1,000 people living in the area are working-class Creoles or blacks. The more imposing dwellings belonging to a few affluent families are located away from the more monotonous cité ouvrière. Apart from the Creoles, some Coloureds (light-skinned Creoles with middle-class aspirations) and Chinese live in the area, as well as a few Hindus and a single Muslim household. Roches-Brunes is not representative of Mauritius with regards to ethnicity, since the largest 'community' islandwide is Hindu.

In describing the relationship between the global and the local, I shall focus on the Rioux household, which I have come to know very well. It is what is commonly described as a matrifocal household, consisting of Mme Rioux, her daughter Aline (20), her two sons, François and Jean, both in their mid-twenties, Aline's baby daughter and a tenant, a young student from the neighbouring island of Rodrigues. Their income is average by local standards. Aline works as a shopgirl in the town centre, her elder brother François is a carpenter's apprentice, and her younger brother Jean is unemployed. The household sometimes receives remittances from other relatives, notably a married daughter who lives in the neighbouring French département La Réunion, and the student from Rodrigues pays a moderate sum in monthly rent.

The living-room in the Rioux' home contains several objects signifying links with distant places. Two posters depicting pop stars (one English, one American) are prominently displayed; so is a cupboard with glass doors, behind which are souvenirs from Paris, Bombay and London. There is a radiocassette and a black-and-white TV set. On the floor next to the TV set, there is a small heap of foreign magazines, some of them in English, which is a language none of the household member masters. Images of Europe are powerful and persuasive in Mauritius; an opinion poll carried out in the mid-seventies indicated that half of the population wished to emigrate if they could (Sofrès 1977).

The mass media consumed in the household confirm the common stereotype of life in 'the Western world' as an easy, glamorous life. Local knowledge of Europe generally suggests it is a continent of affluence and excitement. Many Mauritians have emigrated, the majority to France and Britain. Aline Rioux says she wouldn't emigrate; she has heard too many ugly stories of girls who were forced into prostitution, or who were married to old men living in the countryside. There has, in other words, been a certain feed-back from other parts of the global system. She reads romans-photo, 'photo-novels' of French origin, and occasionally a local magazine. She is very fond of French pop music.

François Rioux plays soccer and follows world politics in the local newspapers; he frequently discusses global issues with his friends. The whole family watches American soap operas on TV; the younger generation go to the cinema to see largely American films dubbed in French about twice a month. They are devoted Catholics and go to Mass every Sunday (actually, Aline goes somewhat more rarely).

The members of the household agree that education is important for a person's opportunities, unless he or she has relatives in high places. François, Jean and Aline are all prepared to compete for jobs and promotion. None of them have completed secondary school.

Seen superficially and in a fragmented way, as I have done now, the world-structures and patterns of consumption of the Rioux household seem comparable to, similar to, that of working classes in many other countries. The globalisation of culture seems predominant in Roches-Brunes, which to an untrained observer like the Lévi-Strauss of Tristes Tropiques (1955), must seem a squalid backyard of civilisation. Scrutinised more closely, however, the lives and world-structures of the Rioux and the others in the cité have a distinctively local character, and cannot at all be understood outside of their local context.

Soccer and pop music may credibly be seen as prime instances of global culture. Eduardo Archetti (1991) tells this anecdote about an incident in Burkina Faso, which may illustrate the global nature of soccer as channelled through mass media. Arriving at the airport in Ouagadougou with his Argentinian passport at a time of political instability, just after Thomas Sankara's assassination, he expected difficulties. At first, the passport official seemed suspicious, glancing through his passport. He then left the room with the passport, re-emerged, opened it on a blank page and stamped it, smiling, while uttering the magical word: 'Maradona'. (Archetti 1991).

Unlike cricket in the Trobriands, soccer in Roches-Brunes follows the same rules as in Britain. Unlike World Cup soccer, however, it is entirely local in character. François Rioux owes much of his reputation in the neighbourhood to his skills as a soccer player. As a result, he is popular with the girls, makes friends with the boys and in fact, he got his present job at least partly because of his personal popularity. Concerning pop music, so arrogantly despised by most anthropologists who encounter it in the field, a similar local context applies. It is played at local parties and in rumshops, and may evoke sentiments and stimulate social relations quite different from its effects in other environments. In the black Jamaican working-class, for example, the world-famous pop singer Michael Jackson was unpopular already in the late 1980s because he was considered 'not sufficiently black'; in the black Mauritian working-class, he is second only to God, not least because he is black.

When watching American TV serials (dubbed in French), the Rioux comment them incessantly. Very often, they compare the characters with people they know; when commenting on rich and miserly men, they might make remarks to the effect that 'Hey, that's just like Lee Foo used to treat my friend,' referring to a neighbourhood Chinese merchant. They always compare the plots and social milieux on the screen with contexts they are familiar with. It should also be noted, significantly, that for a Mauritian Creole, European culture is attractive partly because it is reflexively being contrasted locally with Mauritian Hindu culture. Films and magazines describing middle-class life in Europe or North America, for example, thus make sense and are popular partly because they can be interpreted into into a local dichotomous schema depicting Indian culture as inferior. Overcommunicating what is locally perceived as Europeanness indicates a culturally valued air of superiority compared with the local Indians. In their selective interpretations of aspects of global culture, the inhabitants of Roches-Brunes appropriate them and transform the global into something local.

There are sound reasons that we should treat the thesis of globalisation, in its most general and sweeping forms, with great caution. For one thing, local life-worlds are produced and reproduced locally, and there are social fields where the globalisation of culture has little or no effect - for example in the socialisation of children, where Mauritian Creole custom is strikingly different from the French. For another, there are large parts of the world where the globalising agencies hardly enter. Poverty, it needs to be mentioned, functions as an efficient entropy-resistant mechanism in these matters. The very poor have no access to the shared interfaces of modernity, nor are the agencies of modernity particularly interested in providing them.

Consequences of globalisation for cultural research
Let us now try to draw some general conclusions from the preceding discussion.
First, the institutional dimensions of modernity, to the extent that they are present in a particular locality, facilitate the globalisation of culture and, by extension, create a wide range of potential meaningful communication between people from discrete places. To this effect, walls between societies are being torn down. Further, certain aspects of the globalisation of culture relating to consumption in a wide sense, do not necessarily depend on the implementation of modern social institutions.

Secondly, the disembedding of culture and social relations from the spatial dimension is partial and applies in special contexts only. Most contexts of communication and exchange remain local, although they are influenced by the non-spatial symbolic systems. As Hannerz (1990) wryly observes: Cosmopolitans are dependent on locals for their very identity as cosmopolitans.
Thirdly, facile talk about 'Western society' or 'Western culture' as a monolithic system should no longer be tolerated in anthropological writings. The cultural variations within and between so-called modern societies are spectacular, not least since most societies in the world could be classified as 'more or less modern'. In line with this, processes of 'rationalisation' considered endemic to modern bureaucratic society cannot be taken for granted anywhere. As Sahlins has convincingly argued, it cannot be said of dominant styles of thought in bourgeois societies that they are more or less 'rational' than styles of thought in other societies (Sahlins 1976).
Fourthly, the thesis of globalisation seems relevant for anthropology in two main ways. First, it may inspire research into those limited social and cultural forms which are truly non-spatial, such as international business conglomerates or international advertising, or the 'global switchboards' of the kind envisaged by Hannerz when he speaks of international hotels breaking down barriers between symbolic systems. Secondly, it may increase our sensitivity as anthropological fieldworkers vis-à-vis the impact of processes taking place at a vast scale, yet manifesting themselves locally. Surely, it cannot be without interest to an ethnographer that the leading Mauritian socialist party, the MMM (Mouvement Militant Mauricien), which was initially inspired by French militant groups in the late 1960s, followed the example of the French socialist party in altering its programme in a less anti-capitalist direction in the mid-eighties. In order to grasp this interrelationship, one must know something about French politics, even if one's field of enquiry is in Mauritius.

Fifthly, the lives of humans are in crucial respects created through local practices which take place in locally defined contexts. In this regard, it would be futile to look for particular entropy-resistant mechanisms. In many a regard, societies do remain discrete; the actual impact of globalisation is variable and never all-encompassing. Remarks about international soccer which fall between people from Argentina and Burkina Faso, it must be conceded, form a small part of their respective life-worlds. On the other hand, it becomes increasingly clear that in order to study the local, we must frequently know something about the global for it to make sense. For the local is usually a part of the global in certain respects. To that effect, walls have been demolished. In this regard, we should be careful to avoid a logical error in our reasoning. The fact that the global always manifests itself locally, does not mean that all or most local phenomena have a global element.

Sixthly and finally, the ethical implications of globalisation should not be missed. A deconstruction of the radical distinction between 'us' and 'them' implies that they, rather than being beasts in a natural laboratory, must be taken seriously as human beings during research and analysis. On fieldwork in Trinidad, I was once threatened - only half jokingly - by one of my Trinidadian informants, that if I dared to depict Trinidad as a typical 'plural society' in the tradition from Furnivall and M.G. Smith, he would personally come and beat me up.

New dimensions for anthropological comparison?
Let us finally consider the utility of concepts concerning modern society in anthropological comparison. The danger of our analytical concepts becoming diluted in accuracy by their use in casual political small-talk increases as we try to come to terms with the contemporary world. Such a confusion of everyday terms with analytical concepts is quite common among globalisation theorists. It may, for example, seem plausible to assume that 'nationalism', 'cinema' and 'political party', for example, which form part of native systems of representations and practices in many remote areas, are useful comparative concepts or constants. Individual competition in the labour market, thus, is becoming universal and seems to lend itself to cross-cultural comparison. On the other hand, the local perceptions of labour markets, opportunities and so on, and the actual setup of labour markets, vary - they are context-dependent. As comparative concepts, notions like 'competition in the labour market' may therefore serve as preliminary bridgeheads at best. When we try them out cross-culturally, we will presumably be forced to acknowledge the enormous variations within that category of societies which Lévi-Strauss and others have written off as 'the same kind' or worse, 'Western society'.

In a word, one should not be deluded into believing that the influences of 'global culture' are omnipresent and unambiguous. On the other hand, substantial dimensions for wide-ranging comparisons are becoming viable. Walls have been demolished to this effect. These dimensions for comparison may well reveal greater differences than similarities, and their comparative use may rapidly lead to what Ardener (1989) has called parameter collapse: they may cease to be meaningful as defining concepts and collapse into the defined space, like the concept of society may already have done (see discussion in Ingold 1989). Until such a collapse occurs, we should try to use the concepts in a comparative way; comparing patterns of consumption, nationalist ideologies, labour markets and migration, and so on.

The anthropology of the present should ideally do the same for contemporary society as classical anthropology did for so-called traditional societies; it should insist on, and demonstrate, the enormous variation between the social systems which have hitherto been lumped conveniently together as 'modern society' or even 'Western culture'. Readings of Shakespeare, we have learnt from Laura Bohannan (1966), can take on very different meanings in different societies. This holds true for citizenship, wagework, TV watching, disco dancing and consumption in general as well. However, and that is a fact of cultural homogenisation, concepts and practices relating to citizenship, wagework etc., no matter how context-dependent they are, now intrude into the life-worlds of millions of people whose grandparents were considered savages by anthropologists. Future comparisons cannot, therefore, be conducted as anthropological quasi-experiments. An aspect of Galton's problem, namely the problem of isolating systems in comparison, is now evidently irresolvable (Strathern 1991). Not only are no societies isolated, but they are subjected to such a multitude of influences, and form part of networks of such inconceivable scale that it is impossible to control for such interferences with quasi-experimental setups. Diffusion, in a word, is a constitutive element of the global system (Robertson 1992).

The scope of cultural variation in the world has definitely been reduced, but the diversity remains enormous and must be accounted for. Within the post-colonial, disenchanted parametres of the nation-state, capitalism and globalising systems of symbolisation, we shall find the survival of theoretically incompatible institutions; there will be people 'living in two cultures', as it were; we shall find a wealth of unexpected bricolages, ancient religions comfortable in the nuclear age, we will discover unique local appropriations of prefabricated signifiers - and we shall, certainly, find stable, sleepy local communities oblivious of the external world. Finally, we should not forget, there will always remain desolate backwaters abandoned by the never omnipresent agencies of modernity and globalisation. The primitives, however, are dead. They were killed by their own disenchantment - and by ours.

* * *

Allow me finally to return to my initial metaphor, the Berlin wall. Certainly, like so many other walls dividing people or societies into fixed categories, this wall has been demolished. But as I have argued, walls remain all over the world, even if we succeed in doing away with the wall dividing the civilised from the savages, the intuitively familiar us from the radically different them. There are the walls of class, ethnicity, religion and gender, there are the walls enclosing localities, and there are other, thick walls preventing power and meaning from seeping out. In Germany, walls still separate Ossies from Wessies, Deutsche (Germans) from Ausländer (foreigners). Nowadays, walls like these are invisible and require high ethnographic field competence as well as knowledge about the global system to be fully understood. This is exactly the kind of competence which will be required of us in the future. Never has the Parsonian cliché depicting homogenisation and differentiation as complementary processes (Parsons 1977) been more evidently true than in the contemporary world, a pristine field for comparative anthropological research.


1In André Brink's novel The Wall of the Plague (Brink 1984), the central metaphor for apartheid is a wall which was erected in Carcassonne to protect the healthy from the Black Death.

2'By disembedding I mean the 'lifting out' of social relations from local contexts of interaction and their restructuring across indefinite spans of time-space'. (Giddens 1990:21)

3The idea of an 'urgent anthropology' seems related to this perspective. However, and on the contrary, the proponents of urgent anthropologies, who argue that certain tribal peoples must be studied quickly before they 'disappear', demonstrate that they are aware of important changes taking place in remote corners of the world.

4Commenting on the author of Understanding Media, Umberto Eco remarks: 'Where the apocalyptics saw the end of the world, McLuhan sees the beginning of a new phase of history. This is exactly what happens when a prim vegetarian argues with a user of LSD: The former sees the drug as the end of reason, the latter as the beginning of a new sensitivity. Both agree on the chemical composition of psychedelics.' (Eco 1987:137)

5Fieldwork in Mauritius was undertaken from February through November, 1986, and from December 1991 to March 1992. See Eriksen (1990) for a general ethnography.

6Miller (1992) has described in detail how the North American soap opera 'The Young and the Restless' is appropriated by Trinidadians; indeed, he claims that 'paradoxically an imported soap opera has become a key instrument for forging a highly specific sense of Trinidadian culture.'


Appadurai, Arjun. 1990. 'Disjuncture and Difference in the Global Cultural Economy.' In Global Culture. Nationalism, Globalization and Modernity, ed. Mike Featherstone, pp. 295-310. London: SAGE
Archetti, Eduardo P. 1991. Personal communication
Ardener, Edwin. 1989. The Voice of Prophecy and Other Essays. Oxford: Blackwell
Bohannan, Laura. 1966. 'Shakespeare in the Bush'. Natural History, vol. 75, pp. 28-33
Brink, André. 1984. The Wall of the Plague. London: Faber & Faber.
Duerr, Hans Peter. 1985. Sedna oder die Liebe zum Leben. Frankfurt-am-Main: Suhrkamp
Eco, Umberto. 1987. Travels in Hyperreality, tr. William Weaver. London: Picador
Eriksen, Thomas Hylland. 1990 [1988]. Communicating Cultural Difference and Identity. Ethnicity and Nationalism in Mauritius. Oslo: Occasional Papers in Social Anthropology
Friedman, Jonathan. 1990. 'Being in the World: Globalization and Localization.' In Global Culture. Nationalism, Globalization and Modernity, ed. Mike Featherstone, pp. 311-328. London: SAGE
Fukuyama, Francis. 1989. 'The End of History'. The National Interest, 16 (Summer)
-. 1991. The End of History and the Last Man. New York: The Free Press.
Gellner, Ernest. 1991. 'Stages of Nationalism in Eastern and Central Europe'. Lecture at the conference Nationalism in a Post-Marxist World, Association for the Study of Ethnicity and Nationalism (ASEN), 1 March, 1991
- 1992. Postmodernism, Reason and Religion. London: Routledge
Giddens, Anthony. 1990. The Consequences of Modernity. Cambridge: Polity
Hannerz, Ulf. 1990. 'Cosmopolitans and Locals in World Culture'. In Global Culture. Nationalism, Globalization and Modernity, ed. Mike Featherstone, pp. 237-252. London: SAGE
- 1992. Cultural Complexity. New York: Columbia University Press.
Ingold, Tim (ed.) 1989. The Concept of Society is Theoretically Obsolete. Manchester: Group for Debates in Anthropological Theory.
-. 1993. 'The Art of Translation in a Continuous World'. In Beyond Boundaries, ed. Gisli Pálsson, pp. 210--30. Oxford: Berg.
Kuper, Adam. 1988. The Invention of Primitive Society. Transformations of an Illusion. London: Routledge
-. 1994. Anthropological Futures. In Assessing Cultural Anthropology, ed., Robert Borofsky, pp. 113-18. New York: McGraw-Hill.
Leach, J.W. and G. Kildea. 1974. Trobriand Cricket: An Ingenious Response to Colonialism. Berkeley: University of California Extension Media Center
Lévi-Strauss, Claude. 1955. Tristes Tropiques. Paris: Plon
-. 1962. La pensée sauvage. Paris: Plon
Lyotard, Jean-François. 1979. La condition postmoderne. Paris: Minuit
McLuhan, Marshall. 1964. Understanding Media. New York: McGraw-Hill
Miller, Daniel. 1992. ''The Young and the Restless' in Trinidad: A Case of the Local and the Global in Mass Consumption'. In Consuming Technologies, eds. R. Silverstone and E. Hirsch, pp. 163-82. London: Routledge.
Parsons, Talcott. 1977. The Evolution of Societies. Englewood Cliffs, N.J.: Prentice-Hall.
Robertson, Roland. 1992. Globalization. London: SAGE
Sahlins, Marshall D. 1976. Culture and Practical Reason. Chicago: University of Chicago Press
-. 1994. 'Goodbye to Tristes Tropes: Ethnography in the Context of Modern World History'. In Assessing Cultural Anthropology, ed., Robert Borofsky, pp. 377-94. New York: McGraw-Hill.
Sofrès. 1977. L'Opinion publique à Maurice en avril 1977. Paris: Groupe Métro
Strathern, Marilyn. 1991. Partial Connections. Association for Social Anthropology in Oceania special publication 3. Savage, Md.: Rowman & Littlefield.
Tilly, Charles. 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage
Wallerstein, Immanuel. 1979. The Capitalist World-Economy. Cambridge: Cambridge University Press
Worsley, Peter. 1984. The Three Worlds. London: Weidenfeld & Nicholson

©Thomas Hylland Eriksen 1991