Generations of controversy have done little to alter the centrality of class in the social sciences: it remains an essential theoretical concept, even if its value in operational research has been seriously eroded by changes in the structure of capitalist societies. In anthropology, however, social class understood as a relationship to the means of production has always been less central. Some research traditions have deliberately excluded the study of class differences, preferring to emphasize the unity and homogeneity of bounded cultural units. But even those traditions which have been more open to the study of sociological variation have seldom made much use of class analysis. The concept has been deployed in two principal ways, related through their common debt to the influence of Marx.
The first is that of nineteenth-century evolutionist anthropology, in which the emergence of class society is linked to the rise of private property and the state (Engels 1972 [1884]). Such theories have remained influential in accounts of the development of political institutions, for example in distinguishing between hierarchies based on social class and those based on rank, which do not require the presence of the state (Fried 1967). Distinctions between ruling and subordinate classes are upheld by cultural materialists such as Marvin Harris, despite their lack of sympathy with other aspects of Marxist analysis.
The second major context in which class analysis has entered anthropology is the neo-Marxist school that began in France in the 1960s, and later converged with a variety of approaches influenced by the rejuvenation of political economy. Neo-Marxists argued that polarized classes analogous to those detected by Marx and Engels under early capitalism could also be found across virtually the whole range of precapitalist societies. Thus African societies presented in harmonious coherence by earlier functionalist ethnographers were now shown to be riven by conflict and class struggle. To the extent that male elders appropriated the surplus labour of their juniors and of women, they were to be seen as an exploiting class (or at least they could qualify as a ‘class in itself, for even the most enthusiastic neo-Marxists found it hard to detect class consciousness, ‘class for itself, among such people). This work added little to earlier ethnographies, but was nonetheless valuable in exposing the implicit bias of functionalist accounts.
The irrelevance of the concept of class as far as self-identifications were concerned was always troublesome to anthropologists working in ‘tribal’ and ‘peasant’ societies. Difficulties faced elsewhere in the social sciences, such as the problems posed by the growth of the middle classes in capitalist industrial societies, are posed in accentuated forms in anthropology. The setting and techniques of anthropological research lead the anthropologist away from abstractions that have plausibility at a macro-statistical level. For example, Peter Lloyd (1982) has indicated the difficulties in classifying the occupants of shanty towns around ‘Third-world’ cities as a proletariat. Others have made similar points concerning rural proletariats, where very often it seems that ‘vertical’ links across apparent class boundaries impede the formation of horizontal linkages between those sharing the same ‘objective’ economic situation. Links of kinship, religion,ethnicity and nation have all tended to seem more powerful than links of class. This has not prevented the more dogmatic neo-Marxists in revisiting earlier studies, from seeing all these as surrogates for the one true ontology of class. Yet it is striking that the original ethnographers seldom found this concept useful, preferring to develop more complex notions such as networks, cross-cutting ties, and patron—client relations.
China presents a particularly striking case, given the extreme circumstances of the Maoist period when ‘class label’ had direct and fundamental consequences for the life-chances (or even physical survival) of millions (Watson 1984). Even here anthropologists, when able to carry out fieldwork, have tended to prioritize other forms of grouping, despite the fact that some indigenous pre-socialist concepts of hierarchy can plausibly be glossed in terms of class. Even when the Marxist categories were crudely imposed in the countryside, they were susceptible to local manipulation, and quite different types of class formation in the towns proved even more difficult for the state to control. Some anthropologists, however, argue that the second generation of socialist power is now witnessing the re-emergence of earlier forms of capitalist stratification for which the concept of class remains the appropriate analytic term (Potter and Potter 1990).
On the whole the more successful anthropological accounts of class have been those prepared to shift its definition away from a Marxist rooting in the ownership and control of the means of production. Greater recognition of the importance of status-honour, as emphasized in Weberian approaches, is one important tendency. The Durkheimian inspiration that leads Mary Douglas to adapt the word in its taxo-nomic sense in her identification of ‘consumption classes’ is another. Radical feminists have argued for shifting attention away from relations of production in order to show how the experience of class is structured by race, gender and kinship. For example, Sacks rejects a focus on the individual’s production activities and instead defines membership of the working class as ‘membership in a community that is dependent upon waged labor, but that is unable to subsist or reproduce by such labor alone’ (Sacks 1989: 543). From this perspective, ‘class emerges as a relation to the means of production that is collective rather than individual, a relation of communities to the capitalist state more than of employees to employers’ (Sacks 1989: 547). The drawback, of course, is that in pursuing more culturally sensitive interpretations of stratification and social inequality and in aspiring to a unified theory of class, race and gender, anthropologists will necessarily have to sacrifice the analytic rigour that made class such a popular term in the Marxist tradition. This is precisely what seems to be happening, for example in burgeoning accounts of ideology and resistance.
