Материал: Rethinking ethnicity, 2nd Edition by Richard Jenkins

2 From Tribes to Ethnic Groups

Until the 1960s, and in many cases later, social anthropologists studying ethnicity and inter-group relations typically concentrated upon ethnicity as a corporate social phenomenon. They oriented themselves to the identification and understanding of substantial ethnic groups, conceptualized as distinct-from-each-other, culture-bearing collectivities. This approach still tacitly characterizes the working practices of many anthropologists. As an ontology of the human world it emerged from the social anthropological preoccupation – itself a legacy of the long-standing and disproportionate theoretical sway exercised by structural functionalism during the discipline’s formative years – with social structure, social groups and their systematic interrelationships, and social order and integration.

To an audience that has, perhaps, allowed itself to become over-impressed by the postmodern fluidity and evanescence of the human world and its conventions, these concerns may appear to be passé or quaint. However, they continue to inform many of the presumptions and foundational axioms of anthropology as a generalizing study of human social life. Nor are they necessarily ill-founded. The human world is, among other things, a world of institutionalized social groups, that world is to some extent, at least, ordered, and integration is not insignificant. It is the preoccupation with, and the taking for granted of, these dimensions of the human world, to the exclusion or marginalization of other things – non-corporate collectivities, transient and heterodox improvisation, disorder, disintegration, conflict – which is problematic. These preoccupations, and the taking for granted, were what Barth was writing against in the ‘Introduction’ to Ethnic Groups and Boundaries.

During the colonial and immediately post-colonial periods, social anthropology’s theoretical preoccupation with corporate groups and integrated social systems was manifest in an orthodox assumption that the subject matter of the discipline – ‘primitive’ peoples (Firth 1958: 6) – was most commonly organized into tribal groups. Indeed, as in the following quotation from one of the founders of the ethnographic method, the notion of the ‘tribe’, as a real, substantial social entity, was central to the theoretical and the methodological development of social anthropology:

[The modern ethnographer] with his tables of kinship terms, genealogies, maps, plans and diagrams, proves the existence of an extensive and big organisation, shows the constitution of the tribe, of the clan, of the family … The Ethnographer has in the field, according to what has just been said, the duty before him of drawing up all the rules and regularities of tribal life; all that is permanent and fixed; of giving an anatomy of their culture, of depicting the constitution of their society. (Malinowski 1922: 10, 11)

In one move, the concept of the tribe accomplished two important things: first, it distanced tribal society from civilized society, in both common-sense and analytical

18 Rethinking Ethnicity

discourses; second, it provided the anthropologist with a theoretical model of ‘non-civilized’ social organization which could serve both to organize ethnographic data and function as a framework for the cross-cultural comparison of ‘primitive social organization’. Thus, while on the one hand the difference between ‘them’ and ‘us’ was firmly established, on the other, the basic similarity between different sorts of ‘them’ was proclaimed.

One of the earliest anthropologists to question the usefulness of the notion of the tribe was Edmund Leach, in his study of inter-group relations in northern Burma. Leach argued that tribes as discrete bounded entities were essentially analytical models – developed by anthropologists or other outsiders, for their own purposes – rather than locally meaningful principles of everyday social organization in practice:

the mere fact that two groups of people are of different culture does not necessarily imply – as has nearly always been assumed – that they belong to two quite different social systems. (Leach 1954: 17)

Leach argued that although locals might sometimes talk about themselves as if there were clear-cut collective identities and entities, everyday interaction and organization revealed a more complex pattern of overlap and variation, the reality of which was also recognized by locals. While hinting at a more general theoretical problem (ibid.: 281–2), Leach attributed his conceptual difficulty in the first instance to local ethnographic specifics: anthropology ‘in the classic manner which treated culture groups as social isolates’ was impossible in the Kachin Hills, due to the fact that ‘named groups culturally or partly distinct’ were often ‘all jumbled up’ (ibid.: 60).

By the 1960s the notion of the tribe was beginning to be replaced by the, perhaps less embarrassingly colonial, ‘ethnic group’. However, the underlying presumptions had not necessarily changed. The event that most clearly marked an anthropological paradigm shift, from the study of ‘tribal society’ to the social constructionist model of ‘ethnic groups’ which is current today, was the publication of Ethnic Groups and Boundaries (Barth 1969b). For Barth, the issue was not whether local ethnography fitted the model. The problem was inherent in the presumption of ‘distinct named groups’. In saying this his argument was in harmony with a growing body of critique which, contra structural functionalism, had begun to explore the important place of non-corporate collectivities in social life: ‘networks’ (Barnes 1954; Mitchell 1969), ‘non-groups’ (Boissevain 1968), and ‘action sets’ and ‘quasi groups’ (Mayer 1966), for example.

Barth identified four theoretical features of the conventional, taken-for-granted model of the corporate, culturally distinct ethnic group (1969a: 10–11). First, such a group was biologically self-perpetuating; second, members of the group shared basic cultural values, manifest in overt cultural forms; third, the group was a bounded social field of communication and interaction; and, fourth, its members identified themselves, and were identified by others, as belonging to that group:

This ideal type definition is not so far removed in content from the traditional proposition that a race = a culture = a language and that a society = a unit which rejects or discriminates against others … we are led to imagine each group developing its cultural and social form in relative isolation … a world of separate peoples,

From Tribes to Ethnic Groups 19

each with their culture and each organized in a society which can legitimately be isolated for description as an island to itself. (Barth 1969a: 11)

It would be wrong to present the ‘tribe-group’ model that Barth was criticizing as an explicit analytical framework. It was implicitly embedded in most ethnographic studies, without being examined or seriously questioned. But it was there, and the kind of problem that this taken-for-granted understanding of collective identity has bequeathed to latter-day interpreters of the classic ethnographies is well illustrated by the well-known controversy, which has been utterly overtaken by recent events in the region, surrounding the question of whether the Nuer of the southern Sudan – as documented by Evans-Pritchard (1940) – are, in fact, really the Dinka (Newcomer 1972), or whether they are both distinct components of a single plural society (Glickman 1972).1

In Barth’s model, ethnicity and the boundaries of ethnic groups – and hence their ontological status as collectivities – should not be treated as ‘hard’ or accepted uncritically as fixed aspects of the social reality in question. Barth insisted that ethnic identity, and its production and reproduction in routine social interaction, must be treated as essentially problematic features of social reality, emergent properties of everyday life. The ethnographer must examine the practices and processes whereby ethnicity and ethnic boundaries are socially constructed. And the starting point for such an examination must be an acknowledgement that ‘ethnic groups are categories of ascription and identification by the actors themselves’ (Barth 1969a: 10).

To make use of a distinction first developed by Raymond Firth (1961: 28), Barth and his colleagues shifted the theoretical emphasis, from the evocation of tribal identity as a defining feature of social structure, to a recognition of ethnic identity as an aspect of social organization. In addition, as Ronald Cohen pointed out (1978: 384), while the earliest anthropological notions of ethnic groups still typically implied that such groups were isolated, primitive-atavistic, or non-Western, the shift of emphasis signalled by Ethnic Groups and Boundaries involved ‘fundamental changes in anthropological perspectives’, from a Western interest in the uncivilized peoples of the colonies and ex-colonies, to a more equitable concern with the routine heterogeneity of all societies.

Returning to Barth’s critique of the traditional model, several features of the emergent alternative framework are worth highlighting. In the first place, the analysis of ethnicity starts from the definition of the situation held by social actors. Second, the focus of attention then becomes the maintenance of ethnic boundaries in the interaction between ‘us’ and ‘them’ that takes place across the boundary. Third, ethnic identity depends on ascription, both by members of the ethnic group in question and by outsiders. Fourth, ethnicity is not fixed, it is situationally defined. Fifth, ecological issues are particularly influential in determining ethnic identity, inasmuch as competition for economic niches plays an important role in the generation of ethnicity.

The origins of this understanding of ethnicity can be traced as far back as Max Weber, as argued in the previous chapter. In laying it out as explicitly as he did, introducing it to a new audience, and developing it within an anthropological context, Barth sketched out the essentials of the model of ethnicity that has dominated discussion of the topic within European and North American anthropology in subsequent decades:

20 Rethinking Ethnicity

there is little doubt that Barth’s introduction to Ethnic groups and boundaries holds a special place. More than any other text, it has become the symbolic marker of the changes to which it contributed substantially … The shift from a static to an interactional approach was accomplished by differentiating the notion of ethnicity from that of culture. Barth presented ethnicity or ethnic identity as an aspect of social organization, not of culture. (Vermeulen and Govers 1994: 2)

In British social anthropology, the development of the ethnicity model can be exemplified by looking at the work of Michael Banton and Sandra Wallman. In his original formulations Banton (1967) argued that ‘race’2 is used as a role sign. Social groups are unities that appear in response to various stimuli to alignment, and ‘race’ is a visible marker of status and role as functions of group membership. The notion of ethnicity is here combined with a status-attainment model of stratification. The emphasis is upon inter-ethnic behaviour as transactional, albeit within a social context of stratification and/or coercion. Subsequently, Banton developed a ‘rational choice’ model of inter-ethnic relations (1983), which sought to integrate the study of ethnicity and nationalism with the study of ‘race relations’. This analytical framework is founded on four premises: one, human beings act in order to maximize their net advantage; two, social action has a cumulative effect, inasmuch as present actions limit or constrain subsequent actions; three, actors utilize physical or cultural differences to generate social groups and categories; and, four, when relationships between groups which are held to be physically distinctive are determined by an imbalance of power, ‘racial’ categories are created.

Banton’s ideas shared some of the same assumptions as Barth’s, and presumably some of their Weberian roots. However, the relation between the analyses of Barth and Wallman was a good deal closer (Wallman 1978, 1979, 1986). Locating herself explicitly within the Barthian framework, Wallman argued that:

ethnicity is the process by which ‘their’ difference is used to enhance the sense of ‘us’ for purposes of organisation or identification … Because it takes two, ethnicity can only happen at the boundary of ‘us’, in contact or confrontation or by contrast with ‘them’. And as the sense of ‘us’ changes, so the boundary between ‘us’ and ‘them’ shifts. Not only does the boundary shift, but the criteria which mark it change. (Wallman 1979: 3)

By this definition ethnicity is transactional, shifting and essentially impermanent. Ethnic boundaries are always two-sided, and one of the key issues becomes the manipulation of perceived significant differences in their generation. Furthermore, Wallman considered that it is ‘useful to set the ethnic/racial quibble aside and to consider simply how social boundaries are marked’ (1978: 205); having once done so, there is an effective concentration in her work upon the mobilization of social (ethnic) identity as a resource with respect to social support, solidarity, communication and economic cooperation.

Moving away from the narrow orbit of specifically British social anthropology, the ethnicity paradigm was usefully developed by Don Handelman (1977), in an article that spoke to a number of the issues with which this book is concerned. While Handelman acknowledged the Barthian importance of boundaries and the situational negotiability of ethnic membership, he criticized Barth for confusing ‘the idea of “group” with the ascription of social categories which are everpresent

From Tribes to Ethnic Groups 21

features of the process of interaction’ (p. 187), a comment which echoed an earlier sceptical assessment of Barth by Abner Cohen: ‘At most, what we are establishing … is the simple fact that ethnic categories exist’ (1974: xiii). Handelman further suggested that Barth, in assuming or overstressing the necessity of ‘group-ness’ to ethnic identification, failed to appreciate that:

‘Cultural stuff’… and ethnic boundary mutually modify and support one another. The former establishes and legitimizes the contrast of the boundary; while the latter, often in response to external conditions, modifies or alters the relevance to the boundary of aspects of the former. (1977: 200)

Thus, for Handelman, the cultural content of ethnicity was an important aspect of its social organization: a crude dichotomy between the cultural and the social is misleading. Acknowledging the influence of Boissevain’s discussion of ‘non groups’ (1968), Handelman argued further that ethnicity is socially organized or incorporated in differing degrees of intensity and territoriality – differing degrees of group-ness, if you like – on which depend its salience and importance in individual experience. Moving from ‘the casual to the corporate’, Handelman distinguished the ethnic set,3 the ethnic category, the ethnic network, the ethnic association and the ethnic community. Ethnic identities can, for example, organize everyday life without ethnic groups featuring locally as significant social forms.

Handelman was also concerned with the relationship between ethnicity and other identities. He distinguishes ‘lateral’ from ‘hierarchical’ arrangements of categorical membership. Laterally arranged identities are of more or less equal significance and individuals may be categorized in different ways in different situations: their occupation or class may be most important in one situation, their religion in another, and their ethnicity in another. In a hierarchical arrangement, however, one category will tend to dominate or colour all others. ‘Race’ is a potent example of a hierarchically dominant ethnic category set (p. 193).

The understanding of ethnicity that was ushered in by Ethnic Groups and Boundaries has had valuable consequences. Inasmuch as it stresses social processes and the practices of actors, the approach has encouraged a move away from various strands of determinism that were influential during the 1970s and early 1980s. Similarly, its emphasis upon the social construction of ethnic or ‘racial’ categories discourages the intrusion of biologically based conceptions of ‘race’ into social analysis. Finally, the importance that it attaches to the views and self-perceptions of social actors themselves – the ‘folk view’ of ethnic identity – should lead us away from the taken-for-granted ethnocentrism of much social science. These are all important gains, which remain valuable today.

However, the ethnicity paradigm – the ‘basic social anthropological model of ethnicity’ outlined in Chapter 1 – seems often to have suffered at the hands of those who have most wholeheartedly embraced it. Pursuing a theme similar to Handelman’s, Ronald Cohen (1978: 386–7) argues that, by stressing concepts such as ‘group’, ‘category’ and ‘boundary’, and processes of ‘maintenance’, Barth unintentionally contributed to the further reification of the ethnic group as an enduring, substantial corporate entity. In particular, the notion of the ‘boundary’, with its spatial and physical overtones, allows the superficial or the careless to celebrate the situational flexibility of ethnicity without taking on board the more difficult