Approaches to Ethnicity1 

Ethnic studies received their filip in 1969 with publications from both Frederick Barth and Abner Cohen. The former was a useful corrective to definitions of ethnicity as fixed and defined by cultural traits,2  the latter looked at how a common culture could be used as a base to form an autonomous polity. Both have influenced the course of ethnic studies, largely through their observations of how such identity is maintained and, with Cohen , how ethnicity can be manipulated to advance the interests of a group of people.

The line that ethnic studies in general have taken appears twofold, although the factor of migration seems to be common to the development of both and they both have their roots in racial classificatory schemes which have a history going back at least 400 years.3 In these culture was generally confused with biological features, and most of the racial categories were of a hierarchical nature with the defining race at the top of the pyramid.4 
        The first set of approaches, which I will deal with briefly, largely consist of race relations theories. Specific historical developments led to sociological analysis of the 'meeting' of the races. America, with its background of being host to numerous migrants, proved a fertile ground for theories of assimilation and these in turn reflected "practical and theoretical concerns of American sociology with an immigrant-populated society".5  The term was first introduced in a psychological sense by Simons, in 1901-2, who looked at descriptive accounts of the obstacles to and facilities for social assimilation, the way people grew alike in character. The approach was later expanded upon by Burgess and Park of the Chicago Sociological School who defined the term in 1921 as being "a process of interpenetration and fusion in which persons and groups acquire the memories, sentiments and attitudes of other persons and groups, and, by sharing their experience and history, are incorporated with them in the common cultural life".6  But this definition merely added to the confusion and the word 'assimilation' is variously used as meaning psychological adjustment, cultural adaptation and social process; and discussion still takes place as to whether it refers to a state of affairs or a process.7  As Bash rightly points out,8  it has proved unable to deal with the emergence of black consciousness (what had happened to all those psychologically assimilated negroes?) and assumes a neutral group with the movement towards this dominant culture as a shedding of ethnicity.9  Whatever its drawbacks, this type of approach has been evident in government policy towards 'minority' groups and migrants,10  and informs much sociological discussion of minority problems in such areas as housing.
        The obvious reluctance or inability of some groups to assimilate received attention in modifications to Park's original race relations cycle which started with contact, through competition, adjustment, then assimilation or amalgamation for the migrant/foreign group.11  Others, like Wirth (1941), concentrated on the characteristics of the minority, a concept generally denoting discriminated against migrant or racial groups.12 
        What these studies did do, by concentrating on a presumed common identity based on physical characteristics and/or country of origin, was to make 'ethnicity' (used interchangeably with race, minority, and nationality) a fixed attribute.13 

The second route to the modern approaches to ethnicity may be said to lie through the numerous studies conducted in African townships in the 1940s and 1950s by British anthropologists. Some of these concerned the inter-relationship of African migrants from different backgrounds and discussed this in terms of tribal relations.14 
        If the British and French colonialists had a good idea of what a tribe was in their administrative dealings, the anthropologists were less of one mind. According to Gulliver,15  it was used in the sense of kinship relations (those with a common real or mythical ancestor), as a group which existed only in times of opposition (Evans Pritchard), through socio-economic criteria, or divisions in the population on geographical lines. For Gulliver himself it was both an emotive and regional grouping, characterised by its cultural features, particularistic nature and its attachment to traditional symbols.16 
        What was important here were the efforts, largely stimulated by Gluckman's studies on African townsmen's attitudes to other townsmen, to explain an identity in cultural terms rather than racial. That is not to say that the characteristic of origin did not inform some of these studies. Indeed it tended to confuse them. Thus Epstein could talk of rural tribalism (ie, a political and social unit) and urban tribalism (a meaningful social category, though ultimately based on a group of migrants from a rural tribe).17  It could also lead to such questions as to when did an African trade union leader act as a tribal chief and when as his modern trade union counterpart.18  The separation of rural and tribal from modern is evident in promoting this sort of problem.
        The shift to concentration on ethnic identity and the way people saw each other date from these, it appears. Many of the later treatments of ethnicity seem to owe much to the work of C Cooley and George Mead, both turn of century American social philosophers, whose work gave rise to the symbolic interactionist approach. Here selves had social origins (were socially constructed) and were expressed in social contexts. Identity for Mead was a social phenomenon.19  Equally, those who share meanings or a culture typify or stereotype other groups to enable them to behave towards and prejudge the other's behaviour in a common communicable manner.20 
        Mitchell (1966 and 1974) specifically, built upon this socio-cognitive orientation with his examination of the construct of ethnicity as a perceptual or cognitive phenomenon and the construct of the ethnic group as a behavioural phenomenon. But despite this abstraction ethnicity as a categorical relationship still depended upon some physical characteristic to act as a cue:21  its framework was "a set of shared meanings attached to socially identifiable ethnic cues that provide the actors with sets of expectations of behaviour".22  Mitchell's main problem was the relationship between cognitive and behavioural ethnicity and he examined a survey of ethnic distance constructs with the behaviour of single men in a housing barracks in an African town to see if there was any correlation between the two. His hypothesis that the perception of ethnic identity had a meaning for social action was confirmed, but how this happened was not satisfactorily explained.23 
        Barth's essay in 1969 and his studies of the Fur and Pathans were another step away from the tribal and racial theories and, like Mitchell's treatment of the issue, saw the concept of ethnicity as a category of identification used by the actor and a way of organising interaction between people. His specific concern lay with the maintenance of group boundaries, that was, the sharing of criteria for evaluation and thus identification of others as strangers, the limiting rules of interaction, and the "canalising" of social life. However, by focusing on the organisational forms of social groups Barth had apriori decided that the unity he was looking at was an ethnic group. As Cohen remarked,24  his definition was circular – people acted as a member of an ethnic category because they identified and were identified as one of them, a criticism that could also be levelled at Mitchell. Thus the Fur agriculturists and Baggara pastoralists that Barth studied were already considered ethnic groups as related to a particular mode of production. And should a person go from one to the other he changed ethnic identities. This approach has been criticised as not taking into account the ambiguities and historical shifts in ethnic identity between the two peoples.25  While Barth criticised the traditional view as exemplified by Narroll, where differences between groups were one of traits, biological self perpetuation, and a clearly defined membership,26  his own studies appear to fall into the same trap of giving the identity fixed qualities. His description of groups in (economic) niches through has been developed by others (see below).
        Barth and Gulliver have much in common when they define this kind of group identity as springing from a sharing in certain socio-economic interests with the need to protect them. Tribes, said Gulliver, could also be the result of intergroup competition — they were not just a European creation (through bureaucratic measures or by educating people to identify with certain symbols and a particular culture). Like Barth's ethnic group, Gulliver's tribes appear to be a pot into which different cultures or interests are poured.27 
        But it is not enough to say that an ethnic group is that group whose members share a culture or niche. To explore the subject in a more sociological manner we should ask about the use to which such an identity is put, what kind of organisation if any perpetuates and is perpetuated by 'ethnicity' and so on. Distinction only exists with the existence of one or more other groups and presumes28  political conflict or competition between them at some point.

The study of ethnic groups (or races) has gone through three periods, it has been suggested.29  From the concern with biological problems and the task of classifying mankind with the aim of explaining social phenomena in biological terms, to the cultural frame of reference where significant differences are seen to be languages, customs, beliefs. The most recent period emphasises relationships, and contact and interaction between groups of interest.
        Thus we find the approach of Abner Cohen (1969, 1974, 1981), who located ethnic groups in their historical/political context and, instead of seeing them as Barth did, as givens, he treated them as informal interest groups using their common culture as a basis of organisation and occasionally using their culture's symbolic aspects for political purposes. From this perspective the men in the City of London Rate Tribunal are as much an ethnic group as the Hausa in Ibadan, Nigeria: mutual trust between them rises from their sharing of the same values, norms and language, since they are recruited from the same status group.30  Speed and efficiency result from the same shared 'culture' since none of the actors have to waste time in translation of meanings and assumptions.
        This feature can equally be applied to the Hausa in the Yoruba towns of Nigeria (1969). They dominated the long distance trade and, in Ibadan, clung to their distinct quarter, since, by keeping themselves as an exclusive group, they could co-ordinate collective political action more effectively to secure their economic advantage. Efficiency and domination of their particular economic niche (as Barth would put it) came from shared language and understanding of procedures, as much as from their success in warding off the attacks and strategies to oust them of the Yoruba, who had connected the relationship of Hausa political organisation and spatio-cultural exclusivity with their monopoly of trade.31 
        From a historical analysis of both the Hausa and the Creoles in Sierra Leone (1981) Cohen was able to generalise further about the kinds of organisation necessary to advance or consolidate politico-economic interests. In his scheme of organisational structure, ethnicity is associated with the communal-type organisation where total involvement of personality is demanded, vague ambiguous aims exist, and affinity is necessary for any group's success is distinctiveness, communication, authority structure, ideology, a decision-making procedure and socialisation.32 
        Parkin has further elaborated on the way an ethnic group uses its structural components to communicate messages about the current situation and make public or private political statements. He shows, with an analysis of the Luo, how the kinship network overcomes residential and occupational dispersal of members to disseminate information. The descent group was used for formal announcements and formation of ethnic association. Both helped the Luo to foster and retain political consciousness.33 
        Cohen also saw ethnicity as varying in degree, the importance of group interest, the pressure on members imposed by other groups, and in its cultural forms. He suggested that the ethnic identity of a particular group could be weakened if it found itself in confrontation with other interests, like class. Should such cleavages match tribal group lines this would reinforce their boundaries.
        The difficulty with this principle is in delineating the exact differences between class and ethnic interests. Hannerz points this out in a study of how ethnic groups try to establish economic niches in America. He comments that it is not so obvious that ethnic identity manifestations will be inhibited by "emerging countervailing alignments of power". The elite within a group may ensure continuity of support from followers by distributing jobs to them (a kind of patron-client relationship): shared ethnicity is credit towards the resources for those lower down the group's hierarchy.34  The elite themselves may well ally with the same strata from other niches. For Hannerz,35  ethnicity is both an idiom, with which to promote solidarity as a moral duty, and a fact of origin: he implicitly accepts the Anglo-Saxon/Protestant dominant group ideology that all except themselves are ethnic groups. These latter try to gain access in some way or another to resources, either by finding an unoccupied niche, by raising barriers around their group and making it a special niche for some of its members, or by using the group's strength in other areas. What is interesting in his approach (which owes much to Barth) is the consideration of the 'ideology' of ethnicity as an instrument of subordination.

The ambiguities of ethnicity and its subsequent manipulation by the actors is examined by a number of anthropologists. Since the concept represents an identity category it seems a short step away from appreciating that its dependance upon the actor's consciousness of his or her difference from another group, and the group's imposition of identity upon the actor himself, means that it can be stressed or ignored at convenience, or according to the circumstances.
        Thus Patterson can define ethnicity as "that condition wherein some members of a society in a given social context choose to emphasise as their most meaningful basis of primary, extra-familial identity certain assumed cultural, national or even somatic traits".36  The circumstances in which the members of a group (and here the common links must already exist) choose to do this tend, in this case, to be occasions where it would maximise political and social gains. His cases demonstrating this were two groups of Chinese arriving in the Caribbean. One was imported into Jamaica where eventual success in an economic niche reinforced ethnic consolidation. For those taken into the plantations in Guyana their best advantage lay in playing down their Chinese heritage. Circumstances meant they could not establish themselves in a particular economic sphere due to competition from the Portuguese (who were given undue advantage by the colonialists and European plantation owners). Thus after much intermarriage they were to a large extent absorbed into the majority black population.37  How far any of these groups acted with the choice that Patterson credits them must remain an unanswered question.
        Nagata, in her study of ethnicity in Georgetown, Penang, Malaya, argues that ethnic boundaries change through interpretation of cultural phenomena according to the expressive needs of context (as she puts it).38  The very features of ethnic identity can change. Ethnic groups are, for her, reference groups which vary according to factors in the broader social situation, and some individuals may oscillate between one group to another.39  What one classifies oneself as may depend on whether one wishes to express social distance or solidarity with a group that is stereotyped as having favourable or unfavourable characteristics (and either of these maybe stressed as necessary), out of expediency, or just simply concern for social status. Since it became economically advantageous in Malaya to be a Malay, Nagata notes that Indians of mixed heritage ('klings') appear to participate in both Malay and Indian Chambers of Commerce. What is interesting about this account of personal ethnic identity is that, at the official level, each person is classed as belonging to a particular ethnic group and carries a card to this effect.
        Caplan's account of a hospital strike in India demonstrates not only that choice of ethnic affiliation is situational, constrained by external factors and the aims of the groups concerned,40  but how certain ethnic symbols are used by leaders to unite or to try to gain the support of followers. He also looks at the way these ideologies became manipulated by class interests which lay beyond the immediate situation. Whether to ally with the Christian moral community that the employers at the hospital were trying to persuade the striking employees that they belonged to, or to accept that they were part of the discriminated Tamil population (the Mayalis dominated top positions in the hospital) as the Union rhetoric countered, was not in the final analysis the only set of identities available. The strike soon became part of the class battle taking place outside the hospital by the DMK party against the elites represented by Congress. Demands for ethnic solidarity in times of struggle "are demands for the recognition of moral imperatives"41  — but in this case it was also a question of which ones.

However, in spite of the broadening of the application of the concept 'ethnicity' to include all different kinds of groups, their members gathered together through common interests, values and norms, and to see how groups actually use it, it would be misleading to assume that the characteristic of 'origin' is no longer relevant. As is evident from the numerous studies on migrants and minorities in Britain, this defining attribute still persists as a method of identifying people to study as ethnic groups.42  While theoretically ethnic identity may be considered a flexible phenomenon, practically and methodologically it appears many anthropologists are still very much members of their own dominant group and define themselves as such by distinguishing migrants as outsiders — albeit couched in such jargon as 'ethnic'.43  Where may we ask are the ethnic studies on Australians, French or Germans in London. Are these not classed as cultural associations?

1   I shall take this in the widest sense, rather than just concentrate on specifically political approaches.
2   Eikelmann 1981:158.
3   A number are mentioned in Tischler and Berry (1978). The first systematic attempt was by the French traveler Bernier in the 17th century. He divided the inhabitants of the earth up into four divisions (p.32-33).
4   A typical attitude is no doubt expressed by the following: "We fully admit that the domestic Negroes are improved in intelligence in America, resulting from the imitation of the superior race by which they are surrounded; but much of the improvement in intellect is owing to the mixture of European and Negro blood." (Hunt 1865:33).
      And the white Westeners were not the only ones to betray such ethnocentricity. Chinese 3rd century AD discussions have been found concerning the yellow-haired and green-eyed barbarians who resembled the monkeys from whom they were descended. (In Tischler and Berry op cit:47.)
5   Philpott 1970:9. See also Coser, L (1956, rep 1972:16-20) 'The functions of social conflict' for a discussion of this pragmatic aspect to US sociology.
6   Quoted in Bash 1959:68.
7   Ib:73.
8   Ib:172-173.
9   Nagata 1974:332.
10   Watson 1977:11.
11   Cf.Tischler and Berry op cit:150ff.
12   Mentioned in Krausz 1971:126. Wirth divided minorities up into those who were 'pluralist', ie, wished to preserve their identity in the wider system; 'assimilationist', those wanting to merge fully; 'secessionist', wanting to have full cultural and political independence; or 'militant', those wanting to become dominant.
13   Tischler and Berry (op cit:30-31) point out that the word 'race' has had at least 10 different referents. Eg, citizens of a certain country, those who speak a language (eg, English), religious groups (eg, Jews), a local population (eg, the Basques), a hypothetical pure type (eg, Aryans), a recognisable type (eg, Arab), a biological division, a social race (eg, Mexican) usually distinguished by physical features, and those having a common culture and traditions (eg, Scots).
14  Cf. Epstein 1967:276-277 and Mitchell 1966:48-51.
15  Gulliver 1969:7-11.
16   Ib:5, 12 and 24.
17   Epstein 1967:280.
18   Ib:284.
19   For an account of Mead and Cooley see Abrahamson p.6-22.
20   See Rex (1970:114) for a brief summary of the phenominological (Schutzian version) and symbolic interactionist view of common meanings. Ie, communicate their prejudices intra-group.
21   Mitchell 1966:51-56.
22   Mitchell 1974:14. Physical characteristics included dress, behaviour, etc. Ie, cultural traits.
23   Ib:29. "Fusing of structural and cultural constructions of ethnic behaviour ... must occur in the interpretations of sequences of actions of people involved in social situation" is his final explanation.
      The men appeared to choose people for room-mates whom (in another survey done with different informants) were considered less 'distant'. But Mitchell is unable to say from his data as to whether this choice was on the basis of age, occupation, education or religious affiliation. As he points out, because of the way the Christian missionaries prostelysed there was a close correlation anyway between religious beliefs and area of origin.
      The problem with Mitchell's approach is that it informs us very little other than, perhaps, substantiating his hypothesis that there is a relationship between cognitive and behavioural components — and since these are analyst's devices this seems self-evident anyway.
24   Cohen 1974:xxii-xiv.
25   Eickelmann 1981:158-159.
26   Barth 1969:10-11.
27   Gulliver (1969:35): Tribalism is a flexible concept, a unit of reference for varying interests. The cultural factors (p.31) don't constitute the real factors of differentiation and can change.
28   According to Parkin 1974.
29   E B Reuter (1945), mentioned in Tischler and Berry op cit:16-17.
30   Cohen 1974: xviii-xxiii.
31   This feature of exclusivity was noted by Southall (1973) for the Fulani in Lunsar, who had high status, kept themselves apart from other groups and retained their traditional ways. Southall was concentrating on the retention of traditional culture in the face of industrialisation and urbanisation. However, it would be interesting to note in the light of Cohen's findings, how far their high status was linked with their maintenance of distinction as an ethnic group.
32   Cohen 1974:xvi-xviii.
33   Parkin 1974:120-121. I wonder how far the formation of Clans and establishment of wider kinship links than had occurred in the rural area was a means by which the minority Christian migrants, the urban Toba Batak in Medan, Indonesia, maintained their ethnic identity.
34   Hannerz 1974:38-42. Arrighi, in Foster-Carter (1978:242), suggests that rather than looking at class interests as an independant variable, what about examining the way other phenomena, eg, ethnic groups, affect class interests.
35   Hannerz ib:60-62.
36   Patterson 1975:308.
37   Parkin (op cit:147-151) notes on his elaboration of Cohen's continuum of political ethnicity that the strongest ethnic groups (corporate in organisation) are those with tight control over children and women, ie, encourage ethnic endogamy.
38   Nagata op cit:344.
39   Ib:333.
40   Caplan 1981:79.
41   Ib:64.
42   Cf. Watson, 1977 especially. This may seem tautologous!
43   Certainly the ethnic organisation of the Turkish Cypriots in London appears very little (as described by Ladbury in Watson). There appears little evidence for it acting as an ethnic group in the sense defined by Watson or used by other writers in this particular anthology. To call it then just an ethnic 'category' would be, I think, misleading.

References consulted and cited

Mark ABRAHAMSON (1981) Sociological theory: an introduction to concepts, issues and research. New Jersey: Prentice-Hall.
Frederick BARTH, (ed.) (1969) Introduction to Ethnic groups and boundaries: the social organisation of culture difference, Universitetsforlaget, Oslo.
Harry BASH (1979) Sociology, race and ethnicity, Philadelphia: Gordon and Breach Science Publishers.
Lionel CAPLAN (1981) 'Morality and polyethnic identity in urban S India' in Adrian MAYER, (ed.) Culture and morality, Delhi: OUP.
Abner COHEN (1969) Custom and politics in urban Africa, London: RKP.
— (ed.) (1974) Introduction to Urban Ethnicity, London: Tavistock.
— (1981) 'Variables in ethnicity', in Charles F. KEYES (ed.) Ethnic change, University of Washington Press.
D. EICKELMANN, (1981) The Middle East: an anthropological approach, New Jersey: Prentice Hall.
A. EPSTEIN (1967) 'Urbanisation and social change in Africa', in Current Anthropology, pp.275-284.
A. FOSTER-CARTER, (1978) 'Can we articulate articulation?' in J. CLAMMER (ed.) The new economic anthropology, Macmillan.
P. GULLIVER, (ed.) (1969) Introduction to Tradition and transition in East Africa, RKP.
U. HANNERZ, (1974) 'Ethnicity and opportunity in urban America' in Urban ethnicity, ed. COHEN, op cit.
James HUNT (1865) 'On the Negro's place in nature' in Memoirs read before the anthropological society of London, Vol I, 1863-64, Trubner and Co, pp.1-64.
Ernest KRAUSZ, (1971) Ethnic minorities in Britain, London: MacGibbon and Kee.
S. LADBURY (1977) 'The Turkish Cypriots - ethnic relations in London and Cyprus', in WATSON, op cit.
J. Clyde MITCHELL, (1974) 'Perceptions of ethnicity and ethnic behaviour: an empirical exploration', in COHEN (1974) op cit.
— (1966) 'Theoretical orientations in African urban studies' in M. BANTON, (ed.)The anthropology of complex societies, London: Tavistock, pp.37-60.
Judith NAGATA (1974) 'What is a Malay? situational selection of ethnic identity in a plural society' in American Ethnologist, pp.331-350.
David PARKIN (1974) 'Congregational and interpersonal ideologies in political ethnicity', in COHEN (1974) op cit.
Orlando PATTERSON, (1975) 'Context and choice in ethnic allegiance: a theoretical framework and Caribbean case study' in N. GLAZAR and D. MOYNIHAN (eds.) Ethnicity - theory and experience, Harvard University Press.
S. B. PHILPOTT (1970) 'The implications of migration for sending societies: some theoretical considerations', in R. F. SPENCER and L. KASDAN (eds.) Migration and Anthropology, Seattle: American Ethnological Society, Proceedings of the annual spring meeting 1970.
John REX (1970) 'Race relations and sociological theory' in G. BOWKER and J. CARRIER (eds) Race and ethnic relations, London: Hutchinson.
A. SOUTHALL (ed.) (1973) Introduction to Urban Anthropology, Oxford: OUP.
H. L. TISCHLER and Brewton BERRY (1951, 2nd edn 1978) Race and ethnic relations, Boston: Houghton Mifflin Co.
James WATSON (ed.) (1977) Between two cultures: migrants and minorities in Britain, London: Basil Blackwell.