Copyright Sociological Research Online, 2001


John Rex (2001) 'The Basic Elements of a Systematic Theory of Ethnic Relations '
Sociological Research Online, vol. 6, no. 1, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 10/3/2000      Accepted: 21/3/2001      Published:


The theory of ethnic relations has developed ad hoc on an interdisciplinary basis. It has dealt with ethnicity in small communities, larger ethnic groups or "ethnies", ethnic nations, modernising nation states, subordinate nationalisms, the establishment of empires, post- imperial situations, transnational migrant communities and the political problems facing modernising nation states in dealing both with subordinate nationalisms and with migrant ethnic minorities.This paper seeks to deal with these various elements in an interconnected and systematic way setting out the nature of communities, small-scale self contained communities, the enlargement of these communities to form ethnic nations and the relationship between these and the institutions of the modernising nation state. It also seeks however to deal with what I call "The second project of ethnicity" which is migration and finally goes on to look at the complex problems of multi-nationality and multiculturalism in modern nation states

Empire; Ethnic Mobilisation; Ethnic Nations; Ethnies; Migration; Minorityintegration.; Modernising Nations; Post-imperial Nationalism; Primidrialism; Transnational Communities

Introduction: The Aims of the Argument

The study of ethnic relations has in recent years come to play a central role in the social sciences to a large extent replacing class structure and class conflict as a central focus of attention. This has occurred on an interdisciplinary basis involving sociology, political theory, political philosophy, social anthropology and history. It has dealt with the theory of nationalism and the theory of transnational migrant communities and the problem of incorporating them into modernising national societies. Each of these foci of study has developed its own separate theory but what is worse, has sometimes claimed that it incorporates all the others. Thus, for example, the theory of nationalism has claimed to be a general theory of ethnicity, not recognising the difference between national groups and transnational systems of social relations and culture which bind together the members of migrating groups. Again those who work within the discipline of social anthropology may feel that the field is exhaustively dealt with within its own problematic and the limited range of interests.

The object of this paper is not simply to substitute a new general theory for these specialised theories, but while trying to do justice to their insights, to demonstrate the major points of theoretical connection between them by a careful process of conceptual analysis. I shall deal successively with the theoretical concepts involved as follows: the notion of primordiality and small-scale community; the notion of ethnies and ethnic nationalism; the concept of the modern nation state and related forms of nationalism; the analysis of the structure of empires and colonial societies; that of the reconstitution of post-imperial societies; the concepts of economic and political migration and migrant ethnic mobilisation; what is involved in national policy responses to migration; and, finally, with the notion of multicultural societies. I believe that an analysis of the concepts used at these various levels can point the way to an overall systematic theory.

A fully comprehensive review of the literature on ethnicity and nationalism would include many detailed empirical and historical studies of particular cases. The theories mentioned above take one step beyond this attempting to analyse the concepts which are implicit in such studies. A further step, however, is that of discovering the common elements in all of these theories. It draws upon classical sociological theories dealing with forms of social relations, social structures and social systems at a more abstract level to be found in the work of Tonnies, Weber, Marx, Durkheim and Parsons, and seeks to place the theories focussed on particular problems within this framework. It is only at this level that a general theory of ethnic relations can be formulated.

The object of the sections which follow is to look in each case at the general sociological problems involved in the structure of the theories which have been developed to deal with them.

The Theoretical Problems of Community and Primordial Relations

One of the most fundamental problems in theoretical sociology is that of the distinction between "community (gemeinschaft)" and "association (gesellschaft) posed originally as the basis for his sociology by Tonnies (Tonnies 1963). Tonnies based this on what is really a metaphysical distinction between the real and the artificial wills. The same distinction, however, is used by Max Weber when he seeks to distinguish between different types of "solidary social relationships" (Weber 1968)

As he puts it,
"A social relationship will be called communal if and insofar as the orientation of social action….is based upon a subjective feeling of the parties that they belong together. A social relationship will on the other hand be called associative if and insofar as the orientation of social action within it rests on a rationally motivated adjustment of interests or a similarly motivated agreement. (Weber 1968 p 54)

This much clearer definition rests upon Weber's reduction of all social relations, not to varieties of the will as in Tonnies, but rather to actors taking account of each others behaviour in the course of planning their own actions.

The discussion of the nature of ethnicity in sociological and anthropological literature is very reminiscent of this debate about these concepts of community and association. This is particularly true of the distinction made by Geertz between primordial and other social bonds. In a famous paragraph Geertz writes,

"By primordial attachment is meant one that stems from the "givens" of existence or more precisely, as culture is inevitably involved in such matters, the assumed givens of social existence; immediate contiguity and live connection mainly, but beyond them the givenness that stems from being born into a particular religious community, speaking a particular language, or even a dialect of a language, and following particular social practices. These continuities of blood, speech, custom and so on are seen to have an ineffable, and at times overpowering coerciveness in and of themselves. One is bound to one's kinsman, one's neighbour, one's fellow believer ipse facto as the result not merely of personal attraction, tactical necessity, common interest or incurred moral obligation, but in least in great part by virtue of some unaccountable absolute import attributed to the very tie itself. (Geertz 1963. p.109)

There are serious problems involved in Geertz notion that some attachments are "unaccountable". Such notions can have no place in social science. Geertz, however, suggests something else, namely that the attachments are "given" in the very nature of social existence. Such attachments are contrasted with those which are a matter of choice. Apart from the few reported cases of individuals brought up by animals, so called "feral" individuals, it would seem to be the lot of human beings that they are caught up in a network of interlocking social and cultural constraints. It is worth considering in some detail the nature of these constraints not suggesting that they are ever to be found in a pure form in any actual small-scale community, but in order to establish a basic theoretical point of reference.

Primordial Relations and Community

The distinction made by Geertz between primordial and other social relations approaches that between community and association as in Tonnies and Weber. Attachments which are not primordial, at least those based on "common interest" of "tactical necessity" seem to suggest Tonnies gesellschaft or Weber's associative relations. This leads one to ask whether there is not also an equivalence between social attachments which follow from the very nature of social existence and gemeinschaft or communal social relations.

In fact these two ideas do not imply one another. What Geertz's discussion does, however, is to add to the notion of "belonging together" the notion that this feeling arises when individuals are bound together by the necessities of social existence.

Geertz further develops the notion of community by suggesting that it involves a number of different yet mutually supportive and perhaps functionally interrelated types of attachment. The ones he lists are kinship, neighbourhood or territory, shared language, shared religion, and shared customs. He might also have added the social division of labour and a shared history or myth of origin.

Types of Primordial Attachment

Social anthropology has traditionally been based upon the analysis of kinship terminology. It is one of the "givens" of social existence that every individual entering the world finds that there is a set of terms which have the effect of defining his/her rights and duties vis a vis specific other persons or groups of persons. The fact that such terms can refer to groups as well as individuals means that such terms can be quite extensive. This is quite clear in classical anthropology[1], for example in the work of Spencer and Gillen (1968), whose study of classificatory kinship became the basis of Durkheim's Elementary Forms of Religious Life and indeed of his whole theoretical sociology (Durkheim 1915) Such studies were also undertaken in the classical period by the leading social anthropologists like Radcliffe- Brown (1922, 1931, 1950 and 1952).

Anthropologists studying small-scale primitive communities, however, were always aware of the fact that not all those who lived together were kin. Some were merely neighbours. But neighbours were inevitably involved in working together with kin groups in dealing with the problems of a shared immediate environment. This was also made clear in classical anthropology particularly in the work of Raymond Firth (Firth 1929, 1936, 1958).

The next primordial factor is religion. In the small-scale community this is primarily a matter of belief in the supernatural. This includes, firstly, an extension of kinship to include the dead as well as the living so that ancestors play a continuing role in social organisation; secondly a set of beliefs relating to the natural world which are super-added to those of common sense; thirdly an account of origins going beyond those given in ordinary history, the Australian tribes, for example, believing in a "dream time" in which human history was merged with that of the animal kingdom.

Of course religion is not merely a matter of belief and Durkheim, for instance, argued that there was "no religion without a church" (Durkheim 1915). What is the case in the small-scale community, however, is that there is no specialised "church". The collective practices of the general kinship system constitute the only church, and Durkheim went so far as to argue that "worship" which purported to be worship of the divine was in fact worship of society itself.

Finally under the heading of religion we should note that in the theoretical primordial community it does not include many other elements which we take for granted as being part of religion. It does not provide a general philosophy of existence, does not deal with the problem of the disparity between an ideal divinely governed world and the institutions of the secular world, does not address what Weber called the problem of unmerited suffering, and does not offer some kind of moral code (Weber 1968 Vol II) These are all factors which have to be discussed in dealing with the role of religion in larger collectivities.

The fourth primordial form of attachment is that based upon shared language. Any individual born into the community clearly has a special relationship with others who speak the same language. Language is a means first of all for an agreed naming of objects, but it also involves evaluation of those objects, having a built-in set of moral and aesthetic standards.

The fifth primordial form of attachment and one which Geertz's paragraph does not mention is the social division of labour. It is true, of course, that he does see the kinship system as involving a distribution of rights and duties but, additionally to this there is a system of co-operation in work and practical matters, a simple economy, which is one of the "givens" of existence at this level. It is something which is taken for granted as being in the nature of things.

Finally there is the question of a shared history. Although this may be supplemented by a supernatural account of origins as we have seen above, there is also usually some shared account of secular history, albeit an orally transmitted one.

The Relative Closure of the Small Scale Community

The notion of a closed small-scale community involves some difficulties. Even in a small community of a few thousand people the various forms of attachment may attach others from outside the community. This is particularly true in the case of religious and linguistic attachments, because the religion may be believed in or practised by and the language spoken by a wider group. This is even more likely in the case of religion if the religion itself becomes more complex than the elementary form which we have discussed. Moreover it is possible that the boundaries created by one of these forms of attachment may not coincide with that created by others.

Thus even the notion of a small community appears to be fuzzy at the edges and indeed this may sometimes be the case. But one may also expect mechanisms to arise which lead to the religion practised or the language spoken by the local group being made subtly different from that which exists more widely. This may be facilitated if there are those in the community capable of exercising leadership, but, even where no such leadership exists, the force of community itself may serve to counteract the fuzziness at the edges.

The existence of communities with their own individual membership logically must imply the possibility simply of individuals or groups of individuals who have the negative characteristic of being non-members or of belonging to other groups which are out-groups from the point of view of the group from which we started. The existence of such groups does not necessarily imply hostility and conflict. We may find that there are simply a number of co-existing groups or even communities which co-operate with one another. It is, however, commonly the case that such conflict does occur, and that the feeling of belonging within a community may be enhanced by hostility to out-groups.

Community and Ethnicity

So far we have confined ourselves to the discussion of community as such. We should now notice that, when such a community exists based upon mutually supportive forms of attachment, the community is thought of as having an additional characteristic over and above that represented by any of the groups created by the various forms of attachment. Ethnicity is a term which is used to describe this additional characteristic.

The communities which will be of concern to us in this paper are therefore to be thought of as ethnic communities. The tradition of functionalist analysis in social anthropology might lead to the various forms of attachment being seen as functionally interrelated. At a theoretically simple level however it is perhaps best to describe them as mutually supportive.

Attachment to an ethnic group involves not merely a sense of psychological belonging but also a sense of sacredness. This is of course embodied above all in religion but the various specific forms of attachment, kinship itself, sense of territory, language customs and myths of origin are all imbued with it. When we speak of ethnicity we are not dealing with what Durkheim saw as the "profane" world.

The Primordial Community in a Wider Context

Self-chosen and other Attributed Ethnicity

All that we have said so far refers to the way in which a community or ethnic community sees itself. But it may also have a view of the nature of other ethnic communities and may itself be the object of an attributed ethnicity. One of the things of which we have continually to be aware in the study of ethnicity and ethnic relations is that all ethnic situations have this duality. On the one hand there is a feeling of belonging, of primordial attachments of various kinds and an overall sense of ethnicity felt by group members; on the other the nature of their community may be described differently by outsiders. In more complex modern situations what is accepted as a group's ethnicity or ethnic community structure is often simply the ethnicity which government or those who exercise power attribute to it.

The attribution of ethnicity to outsiders is something that refers not only to groups physically located outside the boundaries of a community. It may also occur in relation to groups who live within the same territory but who do not participate in its system of primordial attachments. Such groups are commonly referred to as minorities, a term which does not have a merely numerical connotation but refers to non-participation in social networks. The group concerned may even be a numerical majority.

The Effect of the Existence of an Out-group on a Group's own Self Image

A further and more complex problem is the effect on a group's self-image of the existence of out-groups. A recent paper by Zolberg and Woon (1999) indeed appears to go further than this implicitly arguing that the self-concept of an ethnic community depends above all upon its having an out-group in relation to which it defines itself. Thus instead of dealing with self-chosen community and ethnicity or ethnicity attributed by other groups it refers to the complex matter of what might be called reflexive ethnicity or perhaps reflexive identity.[2] It is within this context that it argues that specific out-groups might play this role in different nations. Islam is seen as playing this role for West European nations, Spanish speaking communities for Americans.

It is perfectly possible that questions of this kind might be asked. It should however be recognized that this reflexive identity has to be set alongside the less complex cases of self-chosen and attributed ethnicity and ethnic identity. It would be quite wrong to assume that the study of such reflexive ethnicity represented a complete study of ethnicity or ethnic relations.

The Instrumentalist or Situational View of Ethnicity

What we have discussed above is a theoretically small enclosed community with definite boundaries, although we did suggest that there was a certain fuzziness at the edges where those bound together by the different forms of primordial attachment were also bound together with others beyond those boundaries. A different issue, however, is raised by Frederick Barth (1959) in his study of the North West Frontier Province in Pakistan and in his later work on Social Groups and Boundaries Barth (1969). He suggests that the question of who is and is not a member of a community (in his study, who is and who is not a Pathan) depends not upon the content of the culture but upon the purpose for which the community acts together. This is the basis of instrumentalist as opposed to the primordialist view of community. Taken to its extreme this would suggest that the ethnic group should be regarded not as a community at all but as rational and purposive association. A more moderate view is that there is indeed a cultural content in an ethnic community but that the boundaries of the group which has that culture depend upon the purpose in hand.

The Expansion of Small-scale Ethnic Communities and the Formation of Ethnies

The problem of instrumentally or situationally determined boundaries is not a one which occurs in the sort of small-scale community we have been discussing. By definition it is a primordial group. The possibility of instrumentally determined boundaries is rather something which occurs when the small-scale community is thought of as expanding to become what we shall call an ethnie. This is one of a number of features of the increasing structural complexity of larger groups. We must now proceed therefore to consider the nature of ethnies and ethnic nations. In doing this we have to embark upon the second part of our discussion suggested above.

The Institutional Structure of Ethnies

It may be argued that very few small-scale local communities of the kind which we have been discussing exist in the contemporary world. To some extent the very notion is an analytic abstraction, but the usefulness of such an abstraction is that, having made it, we may bracket it away, or use it as a basis for comparison with the structure of the larger collectivities which we do encounter. We shall therefore now look at the way in which the various forms of attachment operate in these larger groups, what such groups actually do, and the ways in which they operate.

While classificatory kinship makes it possible for fairly large kinship groups to be constituted the idea of kinship can also be extended to take the form of fictive kinship. Thus the notion of brotherhood, sisterhood or parentage is extended as these terms are used to refer to many others with whom no biological link exists. In territorially based groups, moreover there are likely to be references to the mother or fatherland.

Similarly the notion of territory is extended to refer not simply to groups of neighbours but to the inhabitants of large geographical areas who are thought to share an emotional attachment to such areas.

The content of religious belief and practice becomes more complex in these larger groups. A specialised priesthood may emerge; the religion might come to include a philosophy of existence possibly involving the idea of a divine creator; the notion might arise that there is a disparity between an ideal world and the institutions of "this world" from which people must be saved; pain and suffering have to be explained; and the religion may come to be supportive of a moral code, as is the case with the so-called "religions of the book", Judaism, Christianity and Islam.

Of course all these elements are to be found in the world religions in different combinations and with different emphases, and the larger group or ethny still faces the problem that its religious beliefs and practices are shared with many outsiders. Here again, however, the ethny usually succeeds in adopting a distinctive form of the more widespread religion which belongs especially to its members.

Similar processes occur with regard to language. A language like Arabic is spoken by hundreds of thousands of people across many countries but it is spoken in each ethny with a distinct dialect. A larger unity of language may be preserved in a sacred form used for religious purposes (e.g. Koranic Arabic or Church Latin) and there are dialect continua which connect one dialect community with another in a kind of chain, but this does not mean that any particular dialect community does not contribute to the solidarity of the ethny. Very similar relationships to these exist with regard to customs. Archaic versions of the customs may be ritually preserved and there may be customs continua across the boundaries of the ethny, but the distinctive version of the customs within the ethny contribute to its solidarity.

The division of labour in the larger ethny may well be based far more on economic specialisation than in the small-scale local community. Furthermore it might very well come to involve social stratification of the population. Such differentiation in terms of prestige will also be coupled with the existence of political power and authority.

Finally there is the question of history. This will no longer be based upon a myth of origin and still less upon some notion of dream time. Rather it will be based upon actual political events involving the group's relation to its territory and its neighbours.

Shared Symbols and the Imagined Community

In the primordial community it was suggested that the mutually supporting system of attachments came to be thought of as sacred. In the larger ethny this sense of sacredness is attached to symbols such as flags and anthems, which is a common topic of research amongst Anthony Smith's students.

Another aspect of the unity of this community is that which is suggested by the title of Benedict Anderson's influential book (Anderson 1991). It should be pointed out that Anderson's whole thesis is a very complex one, focussing especially on the interactions made possible by print capitalism, but the implication of his title is important in earlier situations. In the ethny the interrelated set of institutions is understood as a community in the imagination of its members.

Functional and Dysfunctional Relations between Institutions in the Ethnie

In the case of small-scale communities we spoke of the various forms of attachment as mutually supportive or perhaps as functionally related. In the ethny functional relationships may be more important. In speaking of functional relationships one implies that any one set of structures (kin, territorial, religious, linguistic, political and so on) may be thought of as rationally contributing to the effective working of the whole.[3] Such functional relationships, however, do not necessarily come smoothly into being, There is always the possibility of dysfunction or system contradiction between the different structures of attachment or what we should now call institutions as well as conflict between the participating social actors.

Of major importance in the development of an ethny is the conflict between religious and political leaders. In some cases the ethny will be simply led by the religious authorities, but it is in the nature of expanding ethnies that secular political authorities emerge as well. This is a conflict which goes on in all the great civilisations and not least in Europe where the conflict is between the authority of the Pope and the Emperor.

So far as language is concerned there are not usually separate linguistic authorities asserting a lead role for the institutions of language. Rather such authorities are concerned with establishing one single language for the ethny and suppressing local dialects without their playing a lead role in the complex of institutions.

Another theme in the development of the ethny is that of control of a territory. Such control is dependent upon soldiers and their weapons and one type of authority is that of military authority. If this development is closely associated with kinship the system of authority will be what Weber calls "patrimonial" (Weber 1968 Vol I) but there is also the alternative of feudalism in which the ruler offers military protection in return for economic tribute.

The Ethnic Nation

In this complex system of overlapping ties and institutions the boundaries of the ethny will no longer be of the closed kind in a simple small-scale ethnic community. Which institutions play a lead role and which individuals exercise authority may be very much dependent on the purpose in hand. One such purpose will be the control of territory and insofar as this is the case we should speak not simply of ethnies but of ethnic nations. The system of authority in such ethnic nations becomes more sharply organised but it still remains embedded in the various institutional orders of kin, religion, language and so on. This embeddednes is what is challenged by the emergence of modernising nation state which will be discussed in the next section.

Looked at from an instrumentalist point of view the ethny may have as its purpose the creation of a nation with its own state. This is what I would call the first project of ethnicity. But this is not the only purpose which the ethny might have. One other important purpose which I shall discuss below is migration. I call this the second project of ethnicity (Rex 1996 Chapter 5 , Guibernau and Rex 1997) (see Section 5 on Ethnic nations and the Modernising Nation State below) In this case the ethnic group is not tied to a particular territory and does not seek to establish its own state. There may also be national non-migrant communities which do not seek to establish a state. Guibernau (1999) speaks of "nations without states" She refers to situations like that in Catalonia where the Catalan nation has not a state of its own but aspires to achieving autonomy within a larger national state. A more radical distinction would be that which distinguished the pure ethny from the ethnic nation. This is what most sociologists would regard as a nation without a state.

Before we go on to discuss the difference between the ethnically embedded nation and the modernising nation state there is one other point which we should remember. This is the fact that along with the question of the self-chosen definition of the boundaries of ethnicity and that of which of the various institutions plays a lead role, there is always the fact that members of other groups may see the original one differently from the way it sees itself. Whichever institution they themselves see as dominant, other groups may see them as defined by and led by a different institution. Thus while they may see themselves as a national political group outsiders may see them as defined by, say, their religion or their language.

The Social Anthropology of Ethnicity

It is convenient here to note that the range of problems discussed under the heading of ethnicity differs in different disciplines. In the influential work of Thomas Hylland Eriksen, for example there is little discussion of the more political aspect of social relations (Eriksen 1993). Jenkins (1997), however, combines an anthropological perspective with an interest shared with other students of ethnies in national symbols, while Fenton (1999) discusses the relation between ethnicity and class. In all of these cases, while the narrower focus of social anthropology does draw attention to some important aspects of social interaction in ethnic groups, there does seem to be a need to place this within a more comprehensive sociological context. It might also be asked whether Barthian inspired sociology of ethnicity does not require some supplementation by a consideration of primordial elements. While I have concentrated my attention on the work of Eriksen, Jenkins and Fenton there is now a considerable literature dealing with the questions of ethnicity and ethnic identity, including the work of Banks, writing from a specifically anthropological perspective (Banks 1996), while writers like Oommen (1997) deal with these questions from the point of view of political sociology.[4]

Intellectuals and Ethnic Consciousness

A last point to be made here concerns the role of intellectuals in the development of ethnic consciousness. Intellectuals do help to formulate the notion of ethnic solidarity and, as a consequence it is necessary to recognise that a distinction may usefully be made between the solidarity which they perceive and that which is felt and imagined by common people. On the one hand one may have the work of poets, dramatists, musicians, artists and architects at a high level. On the other one may have folk art folk music folklore and folk culture. This is a point of wider relevance however. It deserves mention here however because it does emerge at the level of the ethnic nation.

Ethnic Nations and the Modernising Nation State

In his extensive writings on nationalism, Anthony Smith (1981, 1986) has emphasised the ethnic origins of nationalism.[5] An ethnic community may become an ethnic nation when its prime purpose is to lay claim to a territory but such an ethnic nation remains embedded in network of ethnic institutions. Very different is the notion of the modernising nation state described by Weber (1968 Vol I, Chapter I_) and Gellner (1983). Weber defines the state as "a compulsory political association with continuous operations" whose "administrative staff successfully upholds the claim to the monopoly of the legitimate use of physical force in the enforcement of its order" (p.54)

Defining the Modern Nation State

Gellner, whose writings have been central to the theorisation of the modernising state and its accompanying types of nationalism, emphasises the needs of the economy which brings the modernising state into being and which is corrosive of all pre-existing forms of social bonding

As he puts it,
The economy needs both the new type of central culture and the central state; the culture need the state and the state probably needs the homogeneous cultural branding of its flock, in a situation in which it cannot rely on largely eroded sub-groups either to police its citizens, or to inspire them with that minimum of moral zeal and social identification without which social life becomes very difficult….the mutual relationship of a modern culture and state is something quite new and springs inevitably from the requirements of the modern economy ( Gellner, 1983. p140)

It is not clear who or what group directs the new nation state in Gellner's conception. The emphasis is more on the method which any group might use to govern. It seems to be based upon the idea of the new nation state which emerged from the French Revolution, but has been used by a variety of different groups seeming to modernise their societies.

In this new nation all institutions must ideally be subordinated to central control. The economy must be centrally directed even if it allows freedom within this to individual entrepreneurs; any priesthood must be subject to state control; there must be an official language; and there must be a national education system so that individuals, instead of being trained only for specialised roles, receive the sort of education which enables them to be moved flexibly from one position to another.

The central nation state seeks to erode all previous forms of solidarity particularly those based upon class, status or ethnicity. What an ethnic community might expect, however, is that, if it is territorially concentrated it might be given some degree of autonomy. There will be a debate about devolution, the community concerned being allowed to make decisions for itself so long as they do not run counter the purposes of the centre. Yet even within this the devolved community will itself be partially transformed so that it has more rational forms of government and authority. These questions have all arisen in cases like those of Catalonia within the Spanish state and Scotland within the United Kingdom. Their devolved authority is quite compatible with the existence of a modern nation state.

The national state has also to control not only internal sources of division but also any relationships which its institutions have beyond its borders. It has to retain control of its own economy where this is threatened by the operation of an international system of entrepreneurship. It has to resist the attempts of international religious bodies to control its priests by creating a national church within the larger international church and it has to keep control of the national language.

Gellner and Marxism

Gellner's account of the modern nation state accepts the Marxist notion of economic determinism. It is the transition from an agricultural society to an industrial one which requires the new type of political order. His approach is also functionalist in that changes in any one of its institutions requires appropriate changes in others. His position is, however, clearly non-Marxist in other respects. As he puts it,

"The socio-economic base is decisive. That much is true in Marxism, even if its more specific propositions are false."(Gellner, 1983)

These "more specific propositions" relate to the expected role of the working class. It is the nation rather than the working class which carries through what the change in the economic base demands.

Traditional Loyalties and the Invention of New Traditions

A further implication of Gellner's notion of the new unitary culture is that the state must indoctrinate all of its members with it and it must displace any pre-existing ethnic cultures. The state needs its own culture as a means of winning the loyalty of its citizens. They have to feel identified with it. Hobsbawm, however, sees that the state appeals to a false kind of ethnic nationalism. It purports to be basing itself upon tradition but it invents its own version of that tradition invoking many practices as traditional which are actually of relatively recent origin (Hobsbawm and Ranger, 1984).

Hobsbawm sees this as happening in revolutionary France. Eugene Roosens sees a similar process as occurring amongst the Huron in Canada whose present leaders present a particular version of Huron ethnicity based on those elements which are useful in their present political projects rather than a true ethnic tradition (Roosens 1989)

Other Forms of the modernising state

Gellner sees the modernising nation state as being exemplified above all by the French Revolution. Others, however, reject the notion that the political arrangements which Gellner describes are the product of either the French or the industrial revolution. Thus Liah Greenfeld argues that they were in place in 15th and 16th century England (Greenfeld 1992). Moreover Gellner's theory is stated in abstract terms and does not deal with individual cases in any detail. One particularly interesting case is that of the Kemalite revolution in Turkey which attempted to establish a modern nation state within part of the Ottoman Empire.

None of what has been said above, however, detracts from the importance of the central distinction which Gellner is making between the modernising nation state and its attendant forms of nationalism and any form of ethnic nation.

Nations and Empires

The age of empires

During the past five hundred years most people have been conscious of living, not simply in ethnic nations or modern nation states, but in empires. In the earliest part of this period this took the form of the establishment of the rule by one dominant nation over its neighbours, as was the case in Britain where the English attempted to establish their rule over Wales, Scotland and Ireland. In other cases what came to be thought of as nations were also based upon the rule of one nation over another. Much more far reaching, however, were the longer distance empires. Europe and a large part of the Middle East and Asia were divided up between the Austro- Hungarian or Hapsburg, the Tsarist (later the Soviet) and the Ottoman Empires. Large parts of North and South America, Africa and Asia were ruled by the British, French, Portuguese, Spanish, Dutch and Belgians. Within these longer distance empires the subordinated nations had differing degrees of autonomy. In the French model all of the subordinated peoples were regarded as overseas French. The Portuguese distinguished between some of the colonial native people who were regarded as evoluees who were entitled to be treated as Portuguese, while others were not. The British model, by way of contrast, is usually thought to be one of indirect rule in which traditional rulers continued to govern. Normally however these powers were restricted to certain areas so that the continued existence of the traditional systems did not conflict with the purposes and interests of the imperial power. In all cases of empire it was the imperial government which ruled overall.

The Establishment of Empires

In the ideal and extreme form of empire the imperial power simply extends its rule through establishing its institutions amongst the subordinate people as though they were the subjects of the metropolitan nation state. Such unified control however is not usually possible. The actual actors who carry out the transformation of the subordinate nations, are the soldiers, the bureaucrats, the economic entrepreneurs, the settlers and the missionaries. These various groups become participant actors in the newly created society and are sometimes in conflict with one another. This was clearly true in the overseas empires of the European powers the imperial civil service, the military, the plantocracy and other business interests, the settlers and the missionaries were often in conflict one another and the art of the imperial government often lay in balancing their various interest against each other. It was probably also true in the case of the Central and Eastern European Empires whose rule was over neighbouring territories.

This situation of imperial control is never complete and will clearly encounter resistance from the conquered nation, be it an ethnic or itself a modernised nation. Here, as in dealing with internal ethnic nations, the imperial power may resort to indirect rule or to granting a limited autonomy under which traditional institutions continue to exist insofar as they do not conflict with the purposes of the imperial nation. It is important to recognise that there is always a potentiality for resistance here by the subordinated peoples. The art of the imperial government therefore lies not only in balancing the interests of the different sections of its own people in the conquered territories, as has been said above, but dealing with this potential resistance.

A different approach to the study of imperial/colonial societies in the longer distance empires is suggested by the theorists of the so-called plural colonial society of whom the leading representatives are J.S.Furnivall (1939, 1948) and M.G.Smith (1965).

Furnivall describes colonial society in Indonesia in the following way: on the one hand one sees a plurality of tightly bonded ethnic groups; on the other one has a market place in which individuals from these groups encounter individuals from other groups. Social and cultural bonding within the different groups is intense and emotional. In the market place by contrast one has only what Marx and Engels in The Communist Manifesto called the "callous cash nexus". Furnivall does not refer to Marx but points out that, whereas for Durkheim a market based society in Europe was bound together by a common will, no such common will existed in the colonial market place.

M.G.Smith (1965) sees the nature of the separate groups and what bind them together in a larger society somewhat differently. Each of the ethnic groups, who have been brought together in a colony, including both the native people and various groups of migrants has its own nearly complete set of institutions. This set of institutions is only nearly complete because they do not each have their own political institution. The political institution is controlled by the imperial power.

Smith does not discuss Furnivall's market place but he does have to look at other forms of ties than those simply imposed by the imperial government. He is concerned with such forms of organisation as the plantation and this does seem to be an economic as well as a political institution.[6]

Attempting to apply this notion to more complex societies like the United States which combines some of the institutions and structures which succeeded the plantation system in the South with those of the North in which free migrants from several European societies are united by the adoption of English as a language. Such societies are not homogeneous as functionalist sociology suggests, but neither are they plural. Smith describes them as heterogeneous. There is more than merely a political bond between the groups. They share a common culture ranging over many areas of life. Within this however the various groups, and particularly Black Americans, retain their own culture in relation to limited areas of life. Some of Smith's associates like Leo Kuper and Pierre van den Berghe (Smith and Kuper 1969) have sought to develop these notions particularly as they seek to apply them in South Africa[7]

The Breakdown of Empire and the Post-imperial Situation

Whereas the world of the nineteenth and early twentieth century was a world organised and controlled by empires, (the Ottoman, Austro-Hungarian, Tsarist and Soviet Communist Empires in East Europe and Asia and the overseas empires of Spain, Portugal, Britain, France, Holland and Belgium), the world of the late twentieth century was one in which the old imperial systems had been overthrown, either because of their own economic and political weakness, or through the resistance of subject peoples. At this point the question arises of what types of social bonds or what type of institutions were likely to arise in the formerly subject areas.

Post-imperial Nationalism

Sometimes what has happened is that there has been a simultaneous overthrow of imperial power and of modernising tendencies within these areas. In these circumstances one might well see a return to the ethnic nation. More commonly however some form of the modern nation state survives and this may even be governed by those who had served the imperial power but now seek to represent themselves as the new nationalists. In these cases what we may have is a halfway house which will be further challenged by ethnic nationalism. In these circumstances the various groups who had constituted the imperial/colonial order find their situation fundamentally altered. This is particularly true of settler groups from the metropolitan centre. They may find that they have difficulty in returning to the metropolis, partly because they are nor wanted there, but also because they have developed their own communities with a settler culture different from that of the metropolitan centre.

Intergroup Relations in Post-colonial Plural Societies

M.G.Smith, in discussing what happens in former colonial territories after independence, suggests that there are a number of possible situations. One is that the society will simply break up. A second is that in which one of the ethnic segments takes over the state and governs the others. A third is a differentiation of function in which one will take over government while another pursues an economic role, an alternative which is evident in Malaysia after independence where the Malays control the government and the Chinese control the economy.[8]. This, however, oversimplifies the actual empirical situations which occur. There are class differences within the different groups and there may also be some interpenetration between them so that, for instance, some members of a group which performs an economic function actually enter the government, or members of the governing group engage in business.[9]

Migrant Ethnic Communities

All the discussion in this section has dealt with situations in which the purpose of nationalism has been control of a territory. We must now turn, however, to what I have called "The Second Project of Ethnicity (Rex 1994). Much of the sociological literature, deriving from the work of Anthony Smith and Gellner purports to deal with ethnicity in general terms but actually addresses itself to the problem of nationalism. If we look at this in terms of instrumentalist theory what we have been discussing are situations in which ethnicity serves national purposes in ethnic nations or modernising nation states. What it does not deal with is the ethnic mobilisation of migrant communities, where the purpose of such communities is precisely movement from one territory to another. The theory of nationalism seeks to deal with this by referring to migrant communities as forms of "diasporic" nationalism. This is misleading because the concept of diaspora involves the notion of return to an original homeland whereas in point of fact many migrant communities have no such intention to return. What we have to do therefore is to give an account of ethnic mobilisation in migrant communities to supplement that given for nationalist communities. [10]

The study of ethnic organisation and mobilisation in transnational migrant communities has to be set within the framework of migration theory. Existing theories on an interdisciplinary basis have been usefully reviewed by Douglas Massey and his colleagues. (Massey et al 1993). I have sought to place my own account of ethnic mobilisation within the framework which they suggest looking particularly at the new economic theory which sees migration as a group decision. [11]

I suggest that migration is normally organised by an extended family seeking to improve its estate. When such a family is in migration it has three points of reference. These are the society and culture of the homeland, the society and culture of the land of first migration and then thirdly that of any country of possible onward migration. Each of these has to be analysed in a way parallel to that which we employed in looking at the structure of ethnies and ethnic nations.

It is, however, important in making this analysis to avoid what is called an essentialist approach in which the various societies and cultures are thought of as closed and unchanging systems. Rather each of them is to be seen as changing and developing in complex ways. The society of the homeland changes when it sends migrants abroad, when those migrants exercise their influence on its social relations and institutions and when they sometimes return; the structure of the community in the land of first settlement results from its encounter with the host culture and, usually with a modern type of society; similarly the community in any country of onward migration involves the relationship between "twice migrants" and the institutions of the new land of settlement.

The Changing Homeland Society

There are, of course, traditional authorities and there is a surviving traditional culture in the homeland. This, however, is something which is challenged by the influences of emigrant culture. The emigrants send back remittances which can be used to but land and houses which increase their economic power and their social status. They offer opportunities for spouses from the homeland to enter countries of settlement. They also send their children to the homeland for an education complementary to that in the land of settlement. Finally, political groups which may otherwise be suppressed in the homeland, may have bases for political action in these countries.

Thus, far from the homeland community being closed and stable in a traditional way it is in fact based upon a struggle between the traditional authorities and the way of life they represent and the sort of society which emigrants work to create. Of course this is not to say that this struggle is not sometimes muted because there are those who remain traditional despite emigration both while abroad and on returning home.

The Immigrant Community in the Land of First Settlement

Going to live abroad means that migrants must also orient their behaviour to a society different from that of the homeland. The key unit in the migrant community is, of course, the extended family which will seek to gain maximum advantage from migration by fighting for equality with host society individuals and families. In this fight for equality they share the universal values of a modern society and this would seem to be a central feature of migrant culture. But families compete both individually and collectively together with their ethnic peers. Clearly there is a sense that those who share the same language, religion and customs can be relied upon in a way which distinguishes them from families drawn from the host society. The boundaries of the community which can be relied upon in this way are often strengthened by religious organisation. Thus the life crises of birth death and marriage are attended to through religious services in religious buildings like churches, mosques, synagogues and temples.

Immigrant culture is therefore a complex entity involving from the simultaneous participation of immigrants in the culture of a modern democracy and in a continuing element of ethnic culture. The continuing ethnic culture helps to provide the immigrants with an emotional and moral home which protects then from a situation of anomie while at the same time giving them the organisational means for collective action in support of their rights.

As the children and grandchildren of the original immigrants become more secure they may find other ways of gaining communal protection from anomie and other means of collective action. Inevitably therefore there will be some who will abandon their ethnic culture altogether. The loss of these descendants is a price which most immigrants are prepared to pay. It is part of the overall cost of migration which brings other advantages. The experience is not in fact unlike that of the working class in industrial society in the past. Working class parents fight for equality including the right to social mobility through education. Where this is achieved there is always the chance that the children will develop interests which conflict with those of their parents. What is interesting, however, is the way in which both working class children and the children of migrants do retain some links with the class or ethnicity of their parents. All sorts of hybrid balance are possible in individual cases.

It could be that the need to maintain a separate immigrant culture will disappear and that what we are dealing with is simply a two or three generation problem which will be ended by the assimilation of the third generation. Where migration from a particular point of origin continues, however, a further nuance of the situation will be that of the relationship between earlier settlers, more or less assimilated and new arrivals amongst whom an ethnic culture is still strong. But even amongst the earlier arrivals there are examples which suggest that the experience of discrimination and failure to gain equality could lead to the reassertion of ethnic culture including its religious expressions. This has been particularly true amongst Muslim immigrants in Europe.

In the next section we shall be dealing with different policies towards migrants and towards cultural pluralism and we shall argue that it is actually in the interest of host societies to encourage assimilation as well as protecting the communal cultures of immigrant ethnic minorities. In that way they will not lose valuable recruits to onward migration or to rebellion.

Communities of Onward Migration

Since the dynamic of migration lies in extended families trying to improve their estate it will always be the case that rather than envisaging a return to the homeland or being assimilated in the society of first settlements many migrants will seize any better economic opportunities which occur through onward migration. In this second migration many of the problems experienced in the land of first settlement will recur and will serve to produce a similar migrant culture for several generations.

The Transnational Migrant Community and the Extended Family

If we ask the question "To what society do migrants belong? The answer may not be the homeland, the society of first settlement or that of onward migration. In fact they belong to all of these simultaneously and the most important binding factor is kinship.

The migrant is first of all a member of a kin group and that kin group stretches across all three societies. A migrant from the Punjab living in Britain for example might have relatives in the Punjab itself, in various parts of Britain and Europe and in North America. What he belongs to perhaps primarily is this transnational community of kinship.

Religion and Language as Unifying Factors amongst Migrant Groups

The self-chosen ethnicity of migrant groups may unite several such groups. This will be the case particularly if several groups speak the same language or practice the same religion. Thus several groups may speak Arabic or a South Asian language in European countries or Spanish in the United States and different groups may be able to act together because they share the same religion, (e.g Islam). This tendency toward joint action will be even more likely if the groups concerned are classified and stereotyped in terms of language or religion by the host society. [12]

Cross-border Migrants

In the account of economic migration given above it has been assumed that the migrants are long distance migrants both geographically and culturally. Connections with the homeland are greatly facilitated, however, if the migrants come to settle from adjacent countries. The extreme case is that of seasonal labour in which case the migrants remain primarily as members of the sending society. Less extreme but also important to distinguish from longer distance migration is the case of cross-birder migrants who actually commit themselves to longer term settlement but are able to maintain lively contact through visits and other forms of direct communication.

Economic and Political Migrants

Not all migrants move for economic reasons. In Western Europe for example there was an immigration stop in the early nineteen seventies, and, since then, asylum seekers, refugees and, more generally, those migrating for political reasons have been more important numerically. (see Joly and Cohen, 1989, Joly, 1996)

It is necessary here to include more than those migrants who are able to make a case for asylum. Under the Geneva Convention such a claim rests upon the ability of the claimant to show that he/she would be in personal danger at home. Only a small proportion of political migrants, however, can make this claim. They are not in personal danger but simply come from situations of political conflict or sometimes, ecological disaster. Often their home societies will be in a process of breaking up and in a state of civil war, or a group (like the Roma of Slovakia) may suffer such sustained oppression and discrimination that the only society to which they could return would be one in which the whole political order in their homelands had been radically changed.

In extreme cases the political migrant will have no kin because they have been killed or have disappeared, but usually there is a strong remaining tie to those who are missing and reunion with them may be a primary goal for those who have escaped. There will also be strong political links and for the community of political migrants and the political project of altering the situation back home may be a central theme.

For many of these migrants there is a strong "myth of return". As a result their view of their situation in the land of refuge is that it is temporary. What they need is temporary housing accommodation, the possibility of paid employment, education for their children and medical and other minimal social services. If, however, the prospect of return fades or if there is a chance of complete integration into the society in which they have taken refuge, they may see themselves as in a similar position to economic migrants. With these two alternative possibilities existing, the life of these migrants may well be one of uncertainty and ambiguity.

As we shall see the response of the society which is a reluctant host to these migrants is that they are "bogus asylum seekers" and this response will be made because these hosts are reluctant to fulfil their obligations. There is, however, some truth concealed in this notion, in that the political migrants are not sure themselves what their aims are. Probably they simply want to keep their options open: to return if the situation at home can be changed so as to offer better economic opportunities as well as a sense of belonging than are available in the land of refuge, but to stay and seek assimilation if that gives greater economic security and better prospects for their children.

The Use of the Term "Diaspora" in the Study of Migrant Communities

I have suggested above that there are those who seek to absorb all studies of ethnicity under the heading of nationalism and that they believe that they have accomplished this by the use of the term "diasporic" nationalism. Separately from this there are others who do recognise the distinction between the study of nations and migrant communities, but use the term "diaspora" to cover the latter. My main objection to this is that it is a loosely used term which fails to capture the detailed structural complexity of migrant communities as we have been discussing it above. It also carries the overtones of a more specific use of the concept, namely that, deriving particularly from Jewish history which refers to communities which have suffered a traumatic political experience, have been dispersed and envisage a return to "Zion"[13], One of the better attempts to deal with this question is that of the American anthropologist James Clifford in his article simply entitled "Diasporas". (Clifford 1994) In this article Clifford distinguishes a large number of structurally different situations in relation to which the term might be used. Another contributor to the debate is Paul Gilroy (whom Clifford does discuss) (Gilroy 1993). In Gilroy's case the term is stretched to, and probably beyond, its limits, seeing it as involving not the hankering after an old lost culture and society but the creation of a wholly new one. On this basis he rejects the Africanist notion of a return to Africa by the descendants of those who were forcibly transported across the Atlantic and suggests that there is a new consciousness which is based both upon the culture developed in the Americas and that which arises for migrants who move from colonial to metropolitan societies.

There are real insights which can be derived from the work of writers like Clifford and Gilroy but it would be better if they did not attach themselves to the concept of diaspora. In fact many of them might very well fit in with some of the distinctions made above. If there is to be a general term which covers this whole area it might be better to speak not of diasporas but simply of "transnational communities".

Policy Responses and Host SocietyAttitudes to Immigration

The Range of Responses to the Presence of Immigrants

There are four main policy responses by modern governments to the arrival of immigrants. These are the response which simply demands that they should be excluded, expelled, persecuted or exterminated and the three alternatives represented by the use of immigrants as labour without allowing them rights of citizenship, the assimilationist approach which accepts their settlement provided that they give up their culture and separate social organisation, and that of multiculturalism. No government deliberately adopts the first of these policies in Europe today although there is a substantial minority representation in politics of parties which advocate exclusion and expulsion in France, Austria, Belgium and Switzerland. The second policy is that adopted in the German speaking countries and very widely in Africa and Asia, which, in its German version, is euphemistically called the gastarbeider policy. German policy even denies that the gastarbeider is an immigrant, since the notion of an immigrant suggests someone who will stay in the long term and even be assimilated. The third policy, assimilationism, is commonly regarded as the characteristic French policy. This offers relatively easy access to citizenship to immigrants but has no place for the encouragement of minority cultures or separate political organisation. Finally there is the policy of multiculturalism usually said to be the policy of the Netherlands, Britain and Sweden.

Forms of Multiculturalism

The most radical version of multiculturalism is the Netherlands policy of pillarisation. This evolved as a means of securing religious tolerance by allowing for a separate education system, separate social services, separate media and separate trades unions to the Catholic, Protestant and secular communities. There is now some argument as to whether Muslims constitute a separate pillar, but it is not surprising that the Netherlands has been more willing to allow separate Muslim schools than any other country. The British government accepts the co-existence of diverse ethnic communities without setting up separate institutions, or denying a common and equal citizenship to all individuals.[14] Eventually this allowed for state financial support for Muslim schools along with the Catholic, Protestant and Jewish schools which had enjoyed it for more than a century. In Sweden there was considerable willingness to consult with those whom the government deemed to be the representatives of ethnic minority communities.

In my book Ethnic Minorities in the Modern Nation State, (Rex 1996) [15]. I have shown at length how in practice all countries deviate from the policy towards minorities which they proclaim an ideological level, and I have noted the very considerable literature which claims that even the high sounding policy of multiculturalism can conceal unequal treatment of minorities or their manipulation for its own purposes by governments. Nonetheless it is still possible to set out the difference policy options of exclusion, assimilation, the denial of full citizenship and multiculturalism and, within this, the different forms of multiculturalism have occurred in Europe and America.

The Position of Political Migrants in a Multicultural Society

The settlement of political migrants raises separate policy problems. The Western European countries have dealt with this question by making it difficult for asylum seekers to arrive by imposing visa requirements, by sending them back to their first country of refuge, and by punishing the carriers of claimants. It is also made the status of a Convention refugee difficult to achieve, and life is made difficult for those awaiting decisions and those, whose claims are rejected, but who are given some kind of "exceptional leave to remain". Foreign policy is also directed to creating "safe havens" for potential refugees in their own countries so that they have no need to seek asylum. All in all what prevails is a reluctant acceptance of some international obligations which allow relatively small numbers to succeed with their applications. Countries outside Europe differ in their degree of generosity, but it is noteworthy that the Organisation of African States, which has to deal with a far greater refugee problem than Europe, shows a greater willingness to accept them. A satisfactory analysis of contemporary multicultural societies has to deal with these political migrants as well as with those who immigrate for economic reasons.

Multiculturalism in the United States

Turning to the United States, on the question of multiculturalism it has adopted the distinctively assimilationist policy represented by the notion of the "melting pot". Although this is a country of immigrants, the institutions and, language of the first immigrants, White Anglo-Saxon Protestants has provided the framework within which later immigrants must work. Arthur Schlesinger Junior (1992) in his book The Disuniting of America has forcefully defended this position which sees American society as threatened both by Black separatism and by the spreading of the Spanish language.[16]

On immigration policy as distinct from policy on multiculturalism American policy has been more generous than its European equivalent particularly since the removal of many of the restrictions which had stood in the way of immigrants from Asia, Africa and Latin America. This is still true today so far as legal immigration is concerned, but there is a considerable problem of illegal arrival and settlement by immigrants particularly from across the Mexican border, and refugees arriving from all parts of the world. Since employers need the labour of such immigrants, the flow of illegals cannot be stopped, but there is pressure to prevent their receiving many of the social rights of citizens, such as the attempt in California to deny them education.

Official and Popular Responses to the Arrival of Immigrants

All the above discussion has been about policies of governments which on the whole are benign and progressive. We should also note here, however that such benign and progressive policies are not shared by the population at large. Overwhelmingly the citizens of European countries fail to distinguish different classes of immigrants and would be opposed to them indiscriminately. This is why politicians discussing almost any immigrant or refugee situation will set their public statements in a context which suggests that they are fighting to stave off a potentially overwhelming immigrant flood. In Britain Margaret Thatcher famously declared that the British were afraid of being "swamped"

Not surprisingly this fear of being swamped by a tide of immigrants and asylum seekers has been expressed in a number of European countries in the adoption of anti-immigrant policies by parties of the extreme right, most notably by Haider's Peoples Party in Austria, by the National Front of Le Pen in France and by the Vlaamse Blok in Belgium. Such parties seem able to command about twenty per cent of the vote in elections, but are more influential than that, because the parties of the centre usually move towards adopting their anti-immigrant policies in order to win back the vote, even if they don't actually enter into coalition with them, as has happened in Austria. It is against this background that the attempts by governments to introduce their official policies of assimilation or multiculturalism have to be seen.

Conclusion: The Sociological Elements of a Systematic Theory of Nationalism and Ethnicity

One aim of this paper has simply been that of recording the different types of study which have been concerned with the questions of nationalism and ethnicity, working from the simplest to the more complex (e.g. from discussions of primordial relations through to those concerned with nationalism and ethnicity in an increasingly globalised world) or in the order in which they have been thought to succeed each other historically. These, however, may be looked at at a higher level of abstraction. Looking at the matter in this way, the basic issues which are raised by the specific theories of nationalism and ethnicity are the following: The types of social relations involved and whether these involve instrumental or affective ties; more complex structures of social relations; institutions and functional and dysfunctional relations between them; groups and collectivities such as ethnies and ethnic nations; affective ties, symbolic unity and group consciousness; rationalisation of national structures; relation between rationalising agents and affectively bonded groups; interaction and conflict between collectivities within larger groups such as nations; relations of conflict and subordination between nations; transnational structures of affectively bonded groups; the internal structure of transnational groups; the relation between modern nation states and transnational groups; supranational societies.

My argument is that the problems which have been discussed in specific theories are ones which recur at different levels of scale and complexity. It is for this reason that it is possible to envisage an overall general theory of nationalism and ethnicity transcending the more specific theories which have been used in understanding small-scale communities, nations and transnational communities. Such a theory is not simply another new one competing with those which have been developed by theorists like Geertz and Barth, Anthony Smith and Gellner, or Furnivall and M.G. Smith. Rather it provides the context in which their theories may be better understood.


1 I refer here to the classical period of social anthropology in which the study of kinship played a central role in the study of social structure. Naturally I am not attempting to review all contributions by social anthropologists who studies kinship as one specialised aspect of social structure.

2I shall in general to avoid using the concept of identity because it is often introduced as a kind of joker in the pack to cover all sorts of situations sometimes referring to the subject in a cognitive mapping of the world, sometimes referring to social position, sometimes to social characteristics sometimes to the personality system. It also has something of the same mysteriousness which seems to follow from Geertz's definitions of primordiality. I use the term here only to refer to a sense of belonging. The notion of reflexive identity refers to the particular sense of belonging which arises from relations to out-groups.

3It would be inappropriate here to go into the whole argument about "functionalism" in sociology. My own position is made clear in my book Key Problems of Sociological Theory in which I reject the use of the organic analogy in Radcliffe-Brown's work and offer my own account based upon the Weberian concept of social action

4There are also now a number of edited collections like those of Romanucci and de Vos(1995) and Sollors (1989) which include articles written from the perspective of anthropology and of a political sociology which deal with conflict as well as intergroup accommodation.

5In my discussion of the work of Smith and Gellner I have not sought to follow the exact distinctions which they make but starting from their basic ideas I have sought to develop the implications of these ideas in a logical way.

6I have set out my own account of the relation between forms of economic exploitation on the one hand and the political order on the other. I do think it important to strengthen Smith's pluralist theory with a recognition of the forms of economic exploitation. This, however, can be combined with a recognition of the importance of political relations. The concept of estates rather than classes is a necessary one in the study of colonial societies as it is in the study of mediaeval Europe. (Rex 1981 , and 1983). I also draw upon the work of Max Weber in the analysis of the various forms of economic exploitation (Weber 1961).

7Actually there is more theoretical complexity than this represented in the volume edited by Smith and Kuper (1969), but the main lines of the distinction between homogeneous, heterogeneous and plural societies remain at the centre of the argument.

8I have discussed some the changes which occur after the end of empire for the colonial people which include the development of a purer market based economy than that which existed in colonial times, the marginalisation of some groups on the edges of this economy and the development of new nationalist forms of revolution.(Rex 1981 and 1983).

9My aim here is to provide a theoretical starting point from which it should be possible to go on to look at the complicating details of actual political history.

10I have in mind here the work of the Association for the Study of Ethnicity and nationalism at the London School of Economics which has never addressed the question of transnational communities except under the heading of diasporas.

11Another account of this process is that given by Castles and Miller (1998), while Robin Cohen who had previously discussed the term diaspora in relation to the Jews, Armenians and Africans across the Atlantic, in his more recent work seeks to set the concept of Diaspora within the overall framework of globalisation. (Cohen 1997).

12There are a number of studies of groups united by language or religion in Britain. Floya Anthias has looked at the question of the Greek Cypriot community united by language if divided internally by social class (Anthias 1992) while Modood has discussed the ways in which the Asian and the Muslim population of Great Britain can act together (Modood 1994and Modood et al 1997).

13Originally the term was used apart from the Jewish case to the dispersal of Armenians and the forced migration of Africans to the Americas and their ambitions to return.

14The official policy of the British government was described as one of "integration" defined by the Home Secretary, Roy Jenkins in 1966 as "cultural diversity, coupled with equal opportunity in an atmosphere of mutual tolerance" (Rex and Tomlinson , 1979)

15I would draw attention particularly to the criticisms of multicultural policy by Wieviorka, Radtke, Rath and Schierup and Alund set out in my book Ethnic Minorities in Modern Nation State,(Rex, 1996) and the volume edited by Beatrice Drury and myself (Rex and Drury,1994).

16In saying this I would also note that there are features of Schlesingsr's argument which are less acceptable, notably his claim that American institutions are essentially European rather than American, and his representation of much Black culture as being akin to that of General Idi Amin.


ANDERSON, B, 1991, Imagined Communities: Reflections on the Origin and Spread of Nationalism Revised Edition, London and New York: Verso.

ANTHIAS, F., 1992, Ethnicity, Class, Gender and Migration, Avebury, Aldershot.

BANKS, M., 1996, Ethnicity: Anthropological Perspectives, Routledge, London.

BARTH, F., 1959, Political Leadership amongst the Swot Pathans, London School of Economics Monographs on Social Anthropology, London School of Economics, London.

BARTH, F, 1969. Ethnic Groups and Boundaries, Allen and Unwin, London.

CASTLES, S., and MILLER, M., 1998, The Age of Migration: International Population Movements in the Modern World, MacMillan, Basingstoke and St, Martin's Press, New York.

CLIFFORD, J., 1994, "Diasporas" in Cultural Anthropology Vol 9 No 3, New York.

DURKHEIM, E., (1915), The Elementary Forms of Religious Life, Free Press, Glencoe, Illinois.

COHEN, R., (1997) Global Diasporas, University College London, London.

ERIKSEN, T., 1993, Ethnicity and Nationalism - Anthropological Perspectives, Pluto Press, London.

FENTON, S., 1999, Ethnicity: Racism, Class and Culture, Rowman & Littlefield, Lanham.

FIRTH, R., 1929, Primitive Economics amongst the New Zealand Maori, Routledge, London, 1936, We the Tikopia, Allen and Unwin, London, 1957, Human types, Nelson, London.

FURNIVALL, J., 1939, Netherlands India, Cambridge University Press, Cambridge, 1948, Colonial Policy and Practice, Cambridge University Press, Cambridge.

GEERTZ, C., 1963, Old Societies and New States - The Quest for Modernity in Asia and Africa, Free Press, Glencoe Illinois.

GELLNER, E., 1983, Nations and Nationalism, Blackwells, Oxford.

GILROY, P., 1993, The Black Atlantic, Verso Press, London.

GREENFELD, L., 1993, Nationalism: Five Roads to Modernity, Harvard University Press, Cambridge, Mass, London.

GUIBERNAU, M., 1999, Nations without States, Polity Press, Cambridge.

GUIBERNAU, M., and REX, J., (Editors) 1997, The Ethnicity Reader, Polity Press, Cambridge.

HOBSBAWM, E., and RANGER, T. (Editors) 1983, The invention of tradition, Cambridge University Press, Cambridge.

JENKINS, R., 1997, Rethinking Ethnicity - Arguments and Explorations, Sage Publications, London.

JOLY, D., 1996, Haven or Hell: asylum policies and refugees and refugees in Europe, Macmillan, Basingstoke and St. Martin's Press, New York.

JOLY, D., and COHEN, R., 1989, Reluctant Hosts, Europe and its Refugees, Gower, Aldershot.

MASSEY, D.,ARANGO, J., HUGO, KOUAOUCI, G., PELLEGRINO, A., and TAYLOR, J., 1993, "Theories of international migration: a review and appraisal" in Population and Development Review, Vol 19, No 3. New York.

MODOOD, T., 1994, "The end of a hegemony: the concept of "Black" and British Asians" in Rex, J., and Drury, B., (editors) Op.Cit, 1997, Modood, T., Berthoud, R., Lakey, J., Nazroo, J., Smith, P., Virdee, S., and Beishon, S., Ethnic Minorities in Britain, Policy Studies Institute, London.

OOMMEN, T., 1997, Citzenship, nationality and ethnicity, Polity Press, Cambridge.

RADCLIFFE-BROWN, A., 1922, The Andaman Islanders, Cambridge University Press, Cambridge, 1931, The social organisation of Australian Tribes, Oceania Monographs, Melbourne, 1950 African systems of kinship and marriage, Oxford University Press, London, and 1958, Structure and function in primitive Society, Cohen and West, London.

REX, J., 1961, Key Problems of Sociological Theory, Routledge and Kegan Paul, London.

REX, J, 1981, "A working paradigm for race relations research" in Ethnic and Racial Studies, Vol 4, No 1, Routledge, London.

REX, J., 1983, Race Relations in Sociological Theory, Second Edition with additional material, Routledge and Kegan Paul, London .

REX, J., 1996, Ethnic Minorities in the Modern Nation State, MacMillan, Basingstoke and St.Martin's Press, New York.

REX, J., and DRURY, B., (Editors) 1994, Ethnic Mobilisation in a Multicultural Europe, Avebury, Aldershot.

REX, J., and TOMLINSON, S., 1979, Colonial Immigrants in a British City - A Class Analysis, Routledge and Kegan Paul, London.

ROMANUCCI-Ross, L., and deVOS, G., (editors), 1995, Ethnic Identity: Creation, Conflict and Accommodation, Altamira Press, London.

ROOSENS, E., 1989, Creating Ethnicity - The Process of Ethnogenesis, Sage Publications, London.

SCHLESINGER, A., Jr., 1992, The Disuniting of America, Norton, New York.

SOLLORS, W., (Editor) 1989, The Invention of Ethnicity, Oxford University Press, Oxford.

SPENCER, B., and GILLEN, F., 1968, The Native Tribes of Central Australia, Dover Books, New York.

SMITH, A., 1981,The Ethnic Revival in the Modern World, Cambridge University Press, Cambridge, 1989, The Ethnic Origins of Nationalism, Blackwells, Oxford.

SMITH, M., 1965, The Plural Society in the British West Indies, University of California Press. Berkeley.

SMITH, M., and KUPER, L., 1969, Pluralism in Africa, University of California Press, Berkeley.

TONNIES, F., 1963, Community and Association, Harper and Row, New York.

WEBER, M., 1961, General Economic History, Collier Books, New York.

WEBER, M., 1968, Economy and Society, 3 volumes, Bedmister Press, New York.

ZOLBERG, A., and WOON, L., "Why Islam is like Spanish - Cultural incorporation in Europe and the United States" Politics and Society, Vol 27, No 1, Sage Publications, London.

Copyright Sociological Research Online, 2001