Материал: Rethinking ethnicity, 2nd Edition by Richard Jenkins

Ethnicity Etcetera 47

usually misrepresented, sometimes to a remarkable degree. For example, as Eriksen reminds us (2002: 53–6) even Barth has been accused, by Abner Cohen, of the sin of primordialism:

Barth … [sees] ethnic categories as classifying persons in terms of their ‘basic most general identity ’… Some writers attribute primordiality to this basic identity.

Ethnicity tends to be conceived by this school of thought as an essentially innate disposition. Barth goes so far as to attribute to it an existence of its own, separate from any social ‘content’. (Cohen 1974: xii)

Geertz, too, has often been widely caricatured, as the purveyor of:

a picture of underived and socially-unconstructed emotions that are unanalysable and overpowering … A more unintelligible and unsociological concept would be hard to imagine. (Eller and Coughlan 1993: 187)

However, turning to what he actually wrote, Geertz explicitly recognizes not only the role of culture in defining the primordial ‘givens’, but also that the:

strength of such primordial bonds, and the types of them that are important, differ from person to person, from society to society, and from time to time. (1973: 259)

Further, Geertz is perfectly clear that what matters analytically is that ties of blood, language and culture are seen by actors to be ineffable and obligatory; that they are seen as natural. This is, in fact, ‘constructed primordiality’ (Cornell and Hartmann 1998: 89–90; see also Gil-White 1999; Hale 2004). Geertz is concerned with the terms in which attachments are understood and mobilized locally, with what people believe. Later in the same piece (1973: 269–70) – making a point which, in all its essentials, has been echoed by Anthony Cohen (1985: 117), with respect to the effect of national integration on local, communal identities – Geertz further argues that, in some respects, these putative ‘primordial attachments’ are actually likely to be stimulated and quickened by the political modernization of nation-building.

Barth, for his part, has never neglected the power and stability of ethnic identifications, the ‘organizing and canalizing effects of ethnic distinctions’ (1969a: 10). As much concerned with the persistence of ethnic boundaries as with their flux, his argument was that under certain, not uncommon, circumstances ethnic change can happen, not that it must. More important perhaps, Barth’s point of view is often presented fossilized in its 1969 Ethnic Groups and Boundaries incarnation. Since then he has explored the importance of ongoing and historically relatively stable ‘streams of tradition’ or ‘universes of discourse’, within the constraints of which ethnic identities are produced and reproduced in practice (Barth 1984, 1989). Most recently, in a discussion which explicitly refers to the ‘privacy of … hearts and minds’ (1994, 29), he acknowledges the need to add micro and macro levels of analysis to the intermediate level upon which his original arguments concentrated. This means looking at what ethnicity means for individuals.

There are other reasons for becoming impatient with the debate between instrumentalism and primordialism. In the evocative words of one recent text, it offers a contrast between ‘ethnicity in the heart’ and ‘ethnicity in the head’ (Banks 1996: 185–7). While this may alert us to the need to acknowledge affect and emotion in

48 Rethinking Ethnicity

our considerations of ethnicity (Epstein 1978), it offers an unreal choice: calculation and emotion are not alternatives. There is no necessary contradiction between instrumental manipulation, on the one hand, and powerful ethnic sentiment, on the other (Cornell and Hartmann 1998: 39–71; Ruane and Todd 2004: 228). They may, in fact, go hand in hand. They may also, of course, conflict. But so may different sentiments, or opposing instrumental goals. And when they do, difficult decisions – of precisely the kind that are the staple of political manoeuvre and manipulation – are the result.

Taking another tack, there is good cause to reject totally any strongly primordialist view. Too much ethnographic evidence exists of the fluidity and flux of ethnic identification, and of the differing degrees to which ethnicity organizes social life in different settings. The theoretical argument in favour of the constructionist view is also well founded. As the discussion in Chapter 1 showed, Barth did not ‘invent’ the constructionist understanding of ethnicity and its utility is not confined to a limited range of cultural contexts. Leach’s classic study of Shan and Kachin in highland Burma (1954) is an ethnographic precursor of the social constructionist position in important respects, and many other studies could be cited to make the point: a wide range of work about pastoralists in East Africa, for example, has offered a similar view (Forster et al. 2000; Schlee 1989; Spear and Waller 1993).

However, we cannot deny the longevity and stubbornness, in certain circumstances, of ethnic attachments (Hale 2004). Pondering the relative failure of the project of social and cultural ‘integration’ with respect to ‘ethnic minorities’ in Britain – and, in the seventeen years since these remarks the situation has, if anything, become more marked – John Rex comments that:

despite the very strong pressure in complex societies for groups to be formed on the basis of congruence of interest, many individuals do in fact stubbornly continue to unite with those with whom they have ties of ethnic sameness, even though such alliances might run contrary to patterns of group formation determined by shared interests. (Rex 1991: 11)

But to acknowledge this point doesn’t necessitate embracing an unreconstructed notion of primordiality or abandoning the social constructionist perspective. It is not stretching the point to regard ethnic differentiation – the social construction of ‘us’ and ‘them’, marked in cultural terms – as a ubiquitous feature of human sociability, and hence of all human societies. The only possible exception to this might be an imaginary and altogether unlikely society existing in total isolation. The debate about whether or not ethnicity is ‘situational’ or ‘primordial’ has confused the ubiquity of a social phenomenon such as ethnicity with ‘naturalness’, implying fixity, determinism and some kind of pre-social power of causation. To suggest that ethnicity is ever-present as one of the ‘givens’ of human social life – or that group identification, for example, is simply part of the human species-specific repertoire, and, thus, ‘human nature’ – is not to endorse any of the arguments of the primordial point of view.

An individual’s sense of ethnic membership may – depending upon context – be internalized during early primary socialization, along with many of the markers of ethnicity such as language, religion, non-verbal behaviour, etc. (Bentley 1987; Fenton 2003: 88). During this period the primary, deep-rooted identities of selfhood,

Ethnicity Etcetera 49

gender and humanness are entered into (Jenkins 2008: Chapter 7). In these senses, identification is an aspect of the emotional, psychological and cognitive constitution of individuality; it is, correspondingly, bound up with the maintenance of personal integrity and security, and may be extremely resistant to change. In a social setting where ethnic differentiation is sufficiently salient and consequential to intrude into the early world of children – over and above the routine acquisition of the ‘cultural stuff’ – ethnicity may be entered into self-consciously, as an integral part of the individually embodied point of view that is selfhood. Thus ethnicity may, depending on local circumstances, be a primary, although not a primordial, dimension of individual identity.

Internal or self-identification – whether by individuals or groups – is, however, not the only ‘mechanism’ of ethnic identity formation. People are not always in a position to ‘choose’ who they are or what their identity means in terms of its social consequences. Power differentials are important here. As will be explored in later chapters, external categorization is an important contributor to ethnicity, not least during primary socialization, but also in a range of other, lifelong settings. Categorization is, furthermore, related to my earlier distinction between the nominal and the virtual: the consequences of ethnic identification across a range of settings – not least whether these are mutually reinforcing or not – are likely to have an important influence on how much ethnicity matters. To put this in another way:

How is intense ethnic solidarity and resistance to change constituted …? The answer does not lie in any essential feature of ethnicity – none exists – but in a particular set of interrelations of elements which form systematic feedback patterns, or a ‘path dependent’ system. (Ruane and Todd 2004: 224)

So the salience of ethnicity within primary socialization, for example, is likely to be a product of local contingencies of place and time. Growing up in Denmark in 2007, for example, is not the same as growing up in Northern Ireland in 2007. And – to return to the distinction between the nominal and the virtual – growing up in Denmark in 2007 is not the same as growing up in Denmark in 1944. Ethnicity, or at least an awareness of it, is likely to figure in different ways, with different social costs and benefits (consequences) attached, in each place and at each time. Other processes of ethnic categorization – which is, after all, what is going on in primary socialization: children know who they are, in large part, because others tell them – also vary with context in their effectiveness and intrusiveness. Thus, although ethnicity may be a primary social identity, its salience, strength and manipulability are situationally contingent (to say which is to come very close to Geertz’s position, outlined above). No matter how apparently strong or inflexible it may be, ethnicity is always socially constructed, in the first instance and in every other.

If the primordial position is so unconvincing, that may be because, as an analytical model, it is largely something of a straw man. The debate about instrumentalism and primordialism has exaggerated the differences between, for example, Geertz and Barth, and does neither justice. It is tempting to agree with Eller and Coughlan (1993) that the notion of ‘the primordial’ should be banished from the social science lexicon. However, the debate remains important and it would be irresponsible to forget it altogether. Crude primordialism is essentially a common-sense

50 Rethinking Ethnicity

view, with enormous power in the world. In sheep’s clothing it occasionally makes an academic showing, indirectly in the guise of sociobiological arguments about kin-based nepotism (van den Berghe 1981, 1995) and, more explicitly, in a recent paper about territoriality (Grosby 1995). Even though these views are not widely supported they cannot simply be ignored. In response, an anthropological perspective, rooted in social constructionist assumptions, has much to offer as a critique of the naturalization of chauvinism and ethnic conflict.

Culture and biology

The controversy about the relationship between culture and biology is in many respects homologous to the debate about instrumentality and primordiality. It is at least as important, and for largely similar reasons: common sense continues to assert the reality and significance of ‘race’ and racism blights human lives. Building upon the example set by Boas (1940), social and cultural anthropologists have long offered an important challenge to dubiously scientific notions of ‘race’ (Benedict 1983 [1942]; Lévi-Strauss 1952). Unfortunately, an ironic consequence of the ascendance of the social constructionist model of ethnicity as – to borrow a notion from Kuhn – a normal science paradigm within the discipline has been that the problem of ‘race’, of the relationship between biology and culture, has tended to recede from the anthropological agenda. Sandra Wallman, as we saw earlier, dismisses the debate about the distinction between ‘race’ and ethnicity as a ‘quibble’ (1978: 205). Elsewhere (1986: 229) she argues that phenotype or physical appearance is just one potential ethnic boundary marker among many. A strikingly similar position is taken by Thomas Hylland Eriksen in his recent textbook: ‘Ideas of ‘race’ may or may not form part of ethnic ideologies, and their presence or absence does not seem to be a decisive factor in interethnic relations’ (2002: 6).

Adopting this position amounts, at least in part, to a laudable refusal to risk inadvertently validating biological models of ethnic difference. Interestingly, however, the naturalization of ethnicity within a sociobiological framework can lead to a similar position. In van den Berghe’s terms, for example, ethnicity is an extension of kinship, a manifestation of an adaptive nepotism between kin that has essentially genetic foundations. Van den Berghe argues that ‘race’ has become nothing more than a ‘special marker of ethnicity’, a visible folk test of likely common ancestry (1981: 240). By this token, it is racism rather than ‘race’ that matters, as an adaptive stratagem of ‘biologically-rooted nepotism’ (although it is instructive to observe the later conflation of ‘race’ and racism in his writing: see van den Berghe 1995).

There are, of course, anthropologists who neither gloss over the relationship between ‘race’ and ethnicity nor submerge the one in the other. As illustrated by the discussion in Chapter 2, one such is Michael Banton, who has written at length about the history of theories of ‘race’ (1998), and consistently explored and theorized the two concepts and the relationship between them. His basic argument is that ‘race’ is a categorical identification denoting ‘them’, based on physical or phenotypical characteristics, while ethnicity is the cultural group identification of ‘us’ (Banton 1983: 1–14; 1988: 1–15). Ethnicity is thus voluntarily embraced, while

Ethnicity Etcetera 51

‘racial’ identifications are imposed. Although Banton is clear that both – albeit perhaps with different force – are socially constructed, Wade (1993, 2002: 18) has criticized him for taking phenotypical variation somewhat for granted, neglecting the social processes of categorization which denote and signify the differences which make a difference. In Wade’s view, physical differences of ‘race’ are always highly – if not completely – socially constructed.

A closely related issue is the anthropological perspective on racism, most particularly the typical anthropological celebration of ethnicity as a ‘good thing’, as a cultural and social resource rather than a liability. Anthropology has also concentrated on minority rather than majority ethnicities. Each of these biases derives in part from the anthropological romance with exotic Otherness, in part from the interpretive cross-cultural advocacy role assumed by many anthropologists, and in part from the continuing legacy of structural functionalism and the theoretical underplaying of conflict. There is an increasing number of exceptions to these generalizations, and clear evidence that world events – in the Balkans, Amazonia, Indonesia … the list would have to be a very long one – in tandem with the developing anthropological concern with nationalism, are finally effecting shifts in these respects within the discipline that are likely to prove durable.1 Nonetheless, racism remains an aspect of the social construction of ethnic differentiation that is relatively little discussed by anthropologists; when it is, it is not always done well (see the exchange between La Fontaine [1986] and Jenkins [1987]).

The arguments about the spurious nature of ‘race’ are too well known to require rehearsal here. The complexities of how one might theorize the relationship between ‘race’ and ethnicity are discussed in subsequent chapters. Here all I want to do is to suggest that the issue of ‘race’ is currently underplayed by anthropologists – even though there are indications of a revival of critical interest, particularly in the United States (Gil-White 2001; Harrison 2002; Wade 2002: 3) – and that adopting a social constructionist model of ethnicity neither requires nor entitles us to indulge in this neglect. Anthropology has, after all, for many years specialized in the study of classificatory systems. To reiterate Wade’s key point (1993, 2002), whatever else it is, ‘race’ is a set of classificatory social constructs of considerable historical and contemporary significance. Exemplary studies of the social construction of ‘racial’ categorization, by non-anthropologists, can be found in Dikötter’s monograph about modern China (1992) and his edited collection about China and Japan (1997). Racism, as an ideology of ethnic identification and a folk cosmology, is also socially constructed, an everyday version of the primordial view. It, too, is an appropriate object for the anthropological attention that, at the moment, it receives all too infrequently.

Redrawing the boundaries

A social constructionist approach to ethnicity and cultural differentiation involves, of necessity, an appreciation that ethnic identification is situationally variable and negotiable. It also involves recognizing the central emphasis that must be placed on the points of view of actors themselves, if we are to understand how processes of social construction and negotiation work. None of the reservations about the