The ancient Greek word ethnos, the root of ‘‘ethnicity,’’ referred to people living and acting together in a manner that we might apply to a ‘‘people’’ or a ‘‘nation’’: a collectivity with a ‘‘way of life’’ – some manners and mores, practices and purposes – in common, whose members share something in terms of ‘‘culture.’’ Thus  the  anthropologist  Frederick  Barth (1969) defined ethnicity as ‘‘the social organization of culture difference.’’ Ethnicity is not only a relatively abstract collective phenomenon, however: it also matters to individuals. To quote another anthropologist, Clifford Geertz, ethnicity is ‘‘personal identity collectively ratified and publicly expressed’’ (1973: 268).

After kinship, ethnicity is perhaps the most ubiquitous way of classifying and organizing humans into collectivities. It requires shared perceptions that certain people are similar to each other and different from others. Ethnic inclusivity and exclusivity build on the cultural differences and similarities that people regard as significant to generate boundaries and dramatize them. How the nuanced complexities of culture are organized into ethnicity is, however, neither obvious nor straightforward. People may appear to differ enormously in terms of culture and yet be able to identify themselves as ethnic fellows: think, for example, about the diversity that is subsumed within Jewishness. Nor does apparent cultural similarity preclude strong ethnic differentiation. Viewed by an anthropologist from Mars, Danes and Norwegians, for example, might look very similar, or even the same; they, however, do not see things this way.

This suggests that our understanding of ethnicity cannot simply depend upon a crude model of discrete and different cultures seen ‘‘in the round.’’ Some cultural themes offer more scope for ethnic identification than others: language, notions of shared descent, historical narratives, locality and co residence, and religion have all proved to be particularly potent ethnic markers. Even so, a common language, for example, or shared religious beliefs and practices do not necessarily do the trick in themselves. Nor do shared space and place: living together may be a potent source of common identification, but space and place can also divide people. They may be a resource for which to compete and the interaction that is necessary for a sense of difference to emerge takes up space: it needs a terrain. Lines are drawn in the sand, and borders and boundaries come to delineate arbitrary group territories.

This further suggests that ethnicity is not a matter of definable degrees or obvious kinds of cultural similarity or difference. There is no checklist with which to determine whether or not members of Group A are really ethnically different to members of Group B, or whether Group C is an ethnic group or some other kind of collectivity. Enumerating cultural traits or characteristics is not a useful way to understand or identify ethnic differences. Human beings are distinguished by their voices, and the base line is always whether a group is seen by its members to be different.

Self definition is not all that matters, however. It is also necessary that a group be categorized as distinctive by others (Jenkins 1997: 51–73). This means that power – whose definition counts in any given situation – is always a lurking presence. There can be no such thing as unilateral ethnicity. Ethnicity always involves ethnic relations: connections and contacts between people who are seen to be different, as well as between those who are seen to be the same. A sense of ethnicity can only arise in the con text of relationships and interaction with others: without difference there is no similarity. Defining us implies – if nothing stronger – an image of them. It is difficult to imagine a meaningful identification, ethnic or whatever, that is not at least recognized by others. It is not enough to assert, ‘‘I am an X,’’ or ‘‘We are Xs,’’ for either of these things to become so.

To say this, however, begs a question: what counts as ‘‘being an X’’ in the contemporary world? Looking at the range of relationships of similarity and difference that might be said to involve ‘‘culture’’ reveals a broad spectrum of possibilities. Neighborhood and locality are among the more immediate. Local senses of belonging that we call ‘‘community’’ – built on an ‘‘us’’ and a ‘‘them,’’ apparently shared understandings, and ways of doing things in common – are well documented (Cohen 1985). Kinship ties may also be invoked as criteria of membership. More abstract regional identities, such as the North–South distinctions that still play so well in England and the United States, are also clearly related. From here it is but a step to the nation (Anderson 1983). While the boundaries of community and region are policed by the informal powers of individuals and groups to accept or reject identity claims, national identity is a formal package that includes citizenship, a passport, political rights and duties within and without the national borders, and so on. This is a domain of formal power and authority. Even here, however, every day practices such as language, taste in food, and perhaps religion may come into the picture: ways of life are still significant.

Descent and kinship may also be important in understandings of the nation (as in the German model of the national identity, defined in terms of ‘‘blood’’ rather than ‘‘soil’’; Bauman 1992). This requires us also to look at ‘‘race,’’ the belief in distinctive populations sharing common ancestors in the remote past, human stocks with their own characteristics. From this point of view, Germans are different from Poles, for example: they are not the same ‘‘kind’’ of people. And although ‘‘racial’’ categories may draw upon the visible features of bodies to assert the ‘‘naturalness’’ of particular similarities and differences, let us remember that ‘‘race’’ is culturally defined, not natural.

The words ‘‘ethnic’’ or ‘‘ethnicity’’ do not appear in the two paragraphs above. Yet in terms of the definition of ethnicity offered earlier, much of the similarity and difference that has been referred to looks something like ethnicity. This suggests some questions. Where does ethnicity end and communal identity, or local identity, or regional identity, or national identity, begin? What is the relationship between community and locality, or locality and region? And what are the differences between all of these things? Where does ‘‘race’’ fit in with them? Are community, locality, region, nation, and ‘‘race’’ even the same kind of thing? The answer is no, and yes. No, in that they appear to be about different things, each evoking its own combinations of criteria of similarity and difference. No, in that some of these criteria are more flexible than others. Locality or citizenship, for example, are easier to change than descent based criteria such as family or ‘‘race.’’ No, in that some of these identities are more likely than others to find expression through ideologies, such as nationalism and racism, which describe the world as it is believed to be and as it should be. But yes, in that the criteria of similarity and difference in each case are cultural. Yes, in that they all contribute to the social organization of a broad and distinctive genre of collective identification, which is not reducible to either kinship or social class, to pick only the most obvious comparisons. And yes, in that they all offer the potential for political organization.

Instead of searching for ever more precise definitions, a better approach might suggest that communal, local, regional, national, and ‘‘racial’’ identities are locally and historically specific variations on a generic principle of collective identification, ethnicity. Each says something about ‘‘the social organization of culture difference’’ and ‘‘the cultural organization of social difference.’’ They are culturally imagined and socially consequential, a way of phrasing the matter which recognizes that distinctions between ‘‘the cultural’’ and ‘‘the social’’ may not be particularly helpful. These communal, local, regional, national, and ‘‘racial’’ identities also offer the possibility of ‘‘collectively ratified personal identity.’’ They may make a considerable personal difference to individuals, both in their sense of self and in their judgment and treatment of others.

This  broad  understanding  of  ethnicity acknowledges that ethnic identification is a contextually variable and relative process. That ethnicity may be negotiable, flexible, and variable in its significance from one situation to another is among the most important lessons of the specialist social science literature (Cornell 1996). Which also means that, depending on cultural context and social situation, ethnicity may not be negotiable. There may not be much of a choice. And when ethnicity matters to people, it has the capacity to really matter, to move them to action and awaken powerful emotions.

That ethnicity can be a source of powerful affect and meaning is at the heart of a longstanding debate between ‘‘primordialists,’’ who believe that ethnic attachments are immutable and irresistible, and ‘‘constructionists,’’ who argue that they are a matter of strategy and negotiation; between what Marcus Banks (1996: 186–7) has evocatively described as models of ‘‘ethnicity in the heart’’ and ‘‘ethnicity in the head.’’ There are several things to bear in mind about this debate. First, the degree to which ethnicity and its variants matter, and to whom, differs demonstrably from epoch to epoch and place to place. There is no consistency with respect to the strength of ethnic attachments, although that humans form ethnic attachments seems to be fairly universal. Nor do we need to resort to notions of essence and nature to explain why, when ethnicity matters to people, it can matter so much: the nature and content of primary socialization, the power of symbols, the implacability of some local histories, and the often considerable consequences of identification are probably sufficient to account for this. What matters is not whether ethnicity is a primordial personal and cultural essence, into which we are born and about which we can do nothing, but that many people fervently believe this to be so and behave accordingly.

Recently, questions have been asked about whether concepts such as ethnicity and identity actually explain behavior (Martin 1995; Brubaker & Cooper 2000). Does ethnicity shape what people do? Does it, in fact, matter? This is partly a response to ambitious postmodern claims about ‘‘identity politics,’’ ‘‘hybridity,’’ and the like. The  argument is that words such as ‘‘identity’’ and ‘‘ethnicity’’ have been bandied about so much that they have become analytically meaningless. While it is easy to sympathize with this view, it is an argument for rehabilitating these concepts, not abandoning them.

The debate is also about the relationship between ethnicity and interests (Goldstein & Rayner 1994): is talking about ethnicity an analytical and political smokescreen to obscure the fact that people are, as they have always done, simply pursuing their material interests? We are back here with Barth’s original account of ethnicity as an emergent property of transactions and negotiations. Now, as then, we have to ask whether it is possible easily to disentangle identification from interests. For example, who I am, whether that is defined individually or collectively, will influence how I define what is in my interests and what is against them. From  another direction, how other people identify me has some bearing on what they perceive my – and, indeed, often their – interests to be. What’s more, my pursuit of particular interests may cause me to be identified in particular ways by myself and by others. Finally, how I identify others may have influences on which interests I pursue.

Interests and ethnicity are entwined in each other, not opposing principles of motivation: where ethnic identification is locally salient, one cannot be understood without the other. Locally – and ethnicity is always a local matter – primary socialization, the affective power of symbols, obstinate history, and the consequences of being identified in a particular way by others conspire to ensure that where it matters, ethnicity really matters. Ethnic attachments do not  determine the  choices that people make, but they cannot be ignored either.


  1. Anderson, (1983) Imagined Communities: Reflections on the Origins and Spread of Nationalism. Verso, London.
  2. Banks, M. (1996) Anthropological Constructions of Ethnicity: An Introductory Guide. Routledge,
  3. Barth, (1969) Introduction. In: Barth, F. (Ed.), Ethnic Groups and Boundaries: The Social Organization of Culture Difference. Universitetsforlaget, Oslo.
  4. Bauman, (1992) Soil, Blood and Identity. Sociological Review 40: 675-701.
  5. Brubaker, & Cooper, F. (2000) Beyond Identity. Theory and Society 29: 1-47.
  6. Cohen, A. (1985) The Symbolic Construction of Communities. Tavistock, London.
  7. Cornell, (1996) The Variable Ties That Bind: Content and Circumstances in Ethnic Processes. Ethnic and Racial Studies 19: 265-89.
  8. Geertz, (1973) The Interpretation of Culture. Basic Books, New York.
  9. Goldstein, J. & Rayner, (1994) The Politics of Identity in Late Modern Society. Theory and Society 23: 367-84.
  10. Jenkins, (1997) Rethinking Ethnicity: Arguments and Explorations. Sage, London.
  11. Martin, C. (1995) The Choices of Identity. Social Identities 1:5-20.

Back to Sociology of Race