Is there a ‘primitive society’?

Social anthropologists traditionally believed that while their subject was comparative and derived its data from fieldwork, it also had a special (though not exclusive) interest in non-Western societies of a type that could be described as ‘primitive’. But in the 1970s and 80s the view developed that this was a totally mistaken view of human society. This was typified by two books by senior and well-established social anthropologists, Professor Adam Kuper’s The Invention of Primitive Society (1988) and the late Professor Edmund Leach’s Social Anthropology (1982). These claim that the very idea of ‘primitive society’ is an illusion, a completely fictitious creation of the discipline of social anthropology itself, and that the notion should be abandoned altogether. Kuper compares the idea of primitive society to that of phlogiston or aether, and says

‘The theory of primitive society is about something which does not and never has existed. One of my reasons for writing this book is to remove the constitution of primitive society from the agenda of anthropology and political theory once and for all. (This is unashamedly a story with a moral)’ (Kuper 1988:8).

According to Kuper, by the beginning of the twentieth century anthropologists had developed a model of primitive society which provided the basis for subsequent theories: ‘Each major anthropologist developed his particular version. The elements were generally stable, though one or other might be given special prominence; but these elements were reordered in every conceivable pattern, the relationships between them systematically transformed to produce a succession of prototypes, ideal types and models’ (ibid., 231).

And what, in Kuper’s view, was this archetypal model which dominated social anthropology in the twentieth century?

‘Primitive society was originally an organic whole. It then split into two or more identical building blocks – exogamous, corporate descent groups. There were no families in the accepted sense. Women and goods were held communally by the men of each group. Marriage took the form of regular exchanges between them. The groups worshipped ancestral spirits. These social forms, no longer extant, were preserved in the languages (especially in kinship terminologies) and in the ceremonies of primitive peoples. After countless generations this system gave way to a form of society based on territorial units, the family, private property and, eventually, the state.’ (ibid., 231).

It is immediately obvious, however, that this model has virtually no relation at all to the notion of ‘primitive society’ to be found in the writings of the major anthropologists of the previous forty years, at least. Evans-Pritchard, for example, defined ‘primitive societies’ as:

‘…those societies which are small in scale with regard to number, territory, and range of social contacts, and which have by comparison with more advanced societies a simple technology and economy and little specialization of social function. Some anthropologists would add further criteria, particularly the absence of literature, and hence of any systematic art, science, or theology.’ (Evans-Pritchard 1951: 8).

In 1954 a group of leading British anthropologists published a volume of talks (previously broadcast in the BBC Third Programme) entitled The Institutions of Primitive Society. The topics included religion, modes of thought, political institutions, family and kinship, law, aesthetics, and economic life. Raymond Firth emphasised the technological aspects of the definition: ‘As I (and I think most of my colleagues) use it, ‘primitive’ is little more than a technological index – a shorthand term for a type of economic life in which the tool system and level of material achievement is fairly simple; little use of metals; no complex mechanical apparatus; no indigenous system of writing’ (Firth 1954:15). But his colleagues refer to many other characteristics besides technology: Evans-Pritchard notes that religious life in highly developed and complex society is not so inextricably involved in the rest of social life as it is in primitive society (Evans-Pritchard 1954:6). On primitive law Peristiany takes as his theme the fact that ‘The more advanced societies like our own have a political superstructure which is organically different from the organization of their constituent elements. Here, it is not possible for a band of kinsmen to act as avengers seeking redress from a group of similar nature, as the state tends to monopolize the legitimate use of force’ (Peristiany 1954:39). Peristiany’s talk is an exposition of the forms of social control in stateless societies. Leach singles out the lack of specialization of labour, the relative self-sufficiency of local communities, non-literacy, and ‘segmentary’ social organization: ‘Many primitive societies…consist of numbers of discrete groups each of which resembles any other except that each group has a different name, a different territory, and a different set of religious rites’ (Leach 1954:78-9).

The complete irrelevance of Kuper’s notion of primitive society is also quite apparent in many other writings of social anthropologists during the last thirty to forty years. Lucy Mair, for example, emphasises the absence of writing, a monetary economy, and political centralization as crucial features of primitive society (Mair 1962:7-13). Mary Douglas stresses lack of differentiation in general, which she associates with technological advances, and works out in some detail the implications of ‘primitive’:

‘Progress means differentiation. Thus primitive means undifferentiated, modern means differentiated. Advance in technology involves differentiation in every sphere, in techniques and materials, in productive and political roles…There are not, in the most small scale types of society, any specialized political institutions. Historical progress is marked by development of diverse judicial, military, police, parliamentary and bureaucratic institutions.’ (Douglas 1966:77-8).

In the rest of the chapter she traces some of the consequences of primitive technology and social organization for modes of thought in particular.

Daryll Forde elaborates at length on the distinctive characteristics of societies with primitive technologies:

‘We may sum up the basic characteristics of primitive economies as follows: preoccupation with the daily and seasonal food supply, limitation of transport, difficulty of storage, overdependence on one or two major resources. These restrictions derive mainly from a low level of technical knowledge, which severely limits productive capacity. Whenever these characteristics are found, certain consequences flow from them. The economic unit is small and, save for occasionally bartered specialities, do not transcend the population of a small village. Such relations are of the personal, face-to-face kind. Everyone has known everyone else from childhood, everyone is related to everyone else…Impersonal commercial relations hardly exist…The small size of the social groups within which production is organized also reduces the opportunity for specialization. Such skills as are practised are known to everyone else of the appropriate age and sex.’ (Forde 1956:333).

Forde also points out that economic processes are embedded in wider social needs, and are inextricably mixed with political, ceremonial, and general festivals:

‘Surpluses cannot be speedily sent to areas of scarcity because of the difficulties of transport. On the other hand, if the surplus is to be used at all, it must somehow be disposed of at once because of the technical difficulty of storage … Often the only way an individual can dispose of a surplus is by holding a lavish feast or simply by giving it to kinsmen and neighbours who will feel bound to repay one day.’ (ibid., 334).

Such economies lack money, and while objects of value are certainly exchanged, ‘When such an object is used as money in a primitive economy, it is usually employed only for a restricted range of transactions. We find that by convention only certain kinds of goods or services can be bought and sold, or that only between certain categories of persons can there be a buying and selling relationship’ (ibid., 334).

Modes of warfare are much less often cited as distinctive of primitive society, but this criterion, too, has been quite clear since 1949, when Turney-High published Primitive War. In this he established that in the conduct of warfare, societies could be grouped into those which had, and those which had not, attained what he called ‘the military horizon’, that is, the ability to conquer other societies and subjugate them politically by means of warfare. Types of warfare and political organization are thus closely associated. (See also Otterbein 1970.)

Professor La Fontaine lists ‘Small scale’, ‘simplicity of social organization’, ‘lack of industrial technology’, and ‘pre-literacy’ as ‘…ways of designating the distinctive characteristics of the kinds of societies forming the bulk of anthropological studies…Their chief distinguishing characteristic is that most people produce their own food and traditionally use no money. They are said to have a subsistence economy’ (La Fontaine 1985:29).

The concept of ‘primitive society’ has therefore been distinguished for many decades by a long list of characteristics which have no connection at all with those of Kuper’s model: smallness of scale in numbers, territory, and range of contacts; simplicity in technology and mode of subsistence; little division of labour and specialization of function; absence of literacy; absence of money and a market system; lack of centralized government and the state; lack of conquest warfare; and organization on the ascriptive principles of kinship, relative age, and gender; and these are only some of the characteristics that could be listed.

Kuper has virtually nothing to say about the effects of technology on population density, permanence of settlement, community size, and all the other consequences of small size and subsistence economy on social organization and modes of thought. While he occasionally refers to the distinction between states and stateless societies, the best he can do to explain the origin of the idea that there are significant differences between societies based on kinship and those based on centralized government is to claim that they are derived from naive stereotypes about ‘blood’ and ‘soil’. This is, to put it charitably, an ineffectual response to the very large literature on state formation which has accumulated in recent years (e.g. Fried 1967; Service 1961, 1975; Claessen and Skalnik 1978; Claessen and Skalnik 1981; Haas 1982; Claessen, van de Velde and Smith 1985; Hallpike 1986).

Kuper’s version of ‘primitive society’ is, in fact, his own illusion, a straw man assembled out of the long discredited speculations of Morgan, Engels, Bachofen, McLennan, Robertson Smith, Maine, Frazer, Tylor, and Lubbock about the earliest forms of human society. Theories about the relative priority in social evolution of totemism, exogamy, matriarchy, patriarchy, ancestor worship, private property, and so on have been effectively dead since the First World War – Lowie, for example, did a comprehensive demolition job in Primitive Society (1921) – and Kuper’s demonstration of this very obvious fact is simply a monument of misdirected effort.

* * * * *

Has Leach been any more successful? He condemns the idea of ‘primitive society’ in the following terms: ‘In my view there is no significant discontinuity in terms either of structure or form between “modern” and “primitive societies”. The social anthropologist can find what he is looking for in either…is it possible to formulate a useful stereotype of what this notional entity “a primitive society” or “a savage (wild) society” is like? The answer is: No!’ (Leach 1982:14). And ‘When modern social anthropologists write about “primitive peoples” it is for want of a better vocabulary: they could just as well be writing about “other” peoples’ (ibid., 55). He goes on to say that ‘There are plenty of useless stereotypes around and since they influence the way anthropologists work up their monographs I ought to say something about them’ (ibid., 141). The three examples of ‘useless stereotypes’ he chooses are primitive society as homogeneous, segmented, and mythopoeic. These are comparatively minor characteristics by comparison with those of technology, economy, size, division of labour, simplicity of political organization, modes of warfare, and lack of literacy, and in any case his attempts to refute even these stereotypes are quite ineffectual.

With regard to homogeneity he objects to ethnographers’ claims that their observations are ‘typical’ of what they did not see because ‘Latterly it has become increasingly apparent that neighbouring small-scale communities, even when they are lumped together under the same “tribal” label, are just as likely to be strongly contrasted as they are to be very much the same’ (ibid.,142), and that ‘homogeneity is introduced as an axiom: it cannot be demonstrated’ (ibid.,142). All this simply evades the very obvious point that in small scale societies of a few hundreds or thousands, with little division of labour, there will be, axiomatically if you like, greater homogeneity in beliefs, values, and customs than in large states with differences of region, class, occupation, education, religion, and ethnic background.

On segmentation Leach says that ‘in many societies virtually everyone is treated as a kinsman of one sort or another’ (ibid.,143) but that there are also numerous exceptions to this. The obvious conclusion is that ‘primitive societies’ do not form a homogeneous class of entities all of which possess n defining characteristics, but rather they are a polythetic class. Since polythetic classes are quite normal in taxonomy, however (Needham 1975), this by itself is no argument against the validity of the concept of ‘primitive society’.

Leach’s comments on the allegedly mythopoeic character of primitive society are very brief and he gives no reason at all for rejecting it. At various other points in his book Leach makes further efforts to dismiss the major differences between primitive and industrial societies:

‘The differences between the primitive societies which anthropologists like to discuss and the modern industrial societies in which most readers of this book are likely to have grown up belong to the level of macro-economics and macro-sociology. At the level of domestic relations with which most people are concerned for most of each working day, the superficial [sic] contrasts are simply different transformations of a single complex of ideas about the proper relationships between men and things and men and other men. It is wholly misleading to suppose that people who seem different from us at this level must on that account be archaic in their mental or social organization.’ (ibid., 118).

This passage seems remarkably evasive: why should ‘macro-economics’ and ‘macro-sociology’ be excluded from consideration, and how does one separate domestic relations from those of the wider society when vengeance, for example, is involved? This is typically required from kinsmen in stateless societies and forbidden by states. Again, how does one separate the norms of everyday reciprocity from the economy as a whole? These norms will obviously be very different in subsistence economies and in those dominated by the market.

‘It is true that literacy seems to generate a sense of rigidly structured social order and a consciousness of historical change which is qualitatively different from the loose structural metaphors and the ‘mythological’ view of the past which anthropologists usually encounter in ‘wild’ (sauvage) societies.’ (ibid., 141).

These are by no means insignificant differences, and he completely ignores the manifold consequences of literacy in government and bureaucracy, communications and trade, the rise of scholarly elites, and the development of specialised literature, philosophy, religion, and science which have been rather more important in history than ‘a sense of rigidly structured social order’. Not only does he evade all the essential issues, but in the course of the book he keeps forgetting that there is not supposed to be any difference between primitive and industrial societies, and distinguishes between them by the traditional criteria:

Law ‘Can societies which lack either law courts or judges be said to have a system of law? It clearly distorts the English conception of law if one says that they have…’ (ibid., 132).

Religion ‘…If anthropologists use the word religion in the sense in which it is ordinarily used by ordinary speakers of English, where it is tied in with such compartmentalized matters as church membership and a professional priesthood, then it would have no application at all to most of the societies which anthropologists usually study⁷(ibid., 133).

Ascription ‘We expect, or say that we expect, that individuals will be appointed to social offices of all kinds in their individual capacity and on the basis of their individual merit…Typically, the anthropologist finds that individuals hold titular offices by virtue of their position in the kinship system. Offices are deemed to be permanently related to one another in a structure of kinship. This is a simple enough idea but when the social structure is made to apply to government, labour relations, the administration of justice, or mystical communication according to context, we find it confusing’ (ibid., 140).

Money ‘One of the characteristics which distinguish ‘modern industrial’ from ‘pre-industrial’ society is that in the former there is a wide circulation of ‘money’, where money means a general medium of exchange and a store of value which serves equally well as a reward for services rendered and as a means of purchasing commodities in the market. By contrast, in ‘pre-industrial’ societies there are often a variety of media of exchange which circulate in different social spheres but which are not interchangeable in any straightforward pattern’ (ibid., 162).

Kinship ‘In the pre-industrial societies, which provide the main field for anthropological inquiry, this kind of distinction [between domestic and public] is either non-existent or quite peripheral to the main field of the ordinary individual’s day to day activities. In such societies it is frequently the case that the individual treats all outsiders as if they were kinsmen of some sort’ (ibid., 164).

The attempts of Kuper and Leach to refute the idea of primitive society are therefore complete failures, which depend on idiosyncratic definitions of ‘primitive’ and on evading or denying all evidence to the contrary.

* * * * *

But we must also ask how this denial of the very existence of ‘primitive society’ was ever possible in the first place. Part of the answer is that the traditional conception of primitive society was vague and fuzzy, and was opposed in a simplistic binary manner to ‘modern industrial society’. While it was generally agreed that simplicity of technology and economy, smallness of scale, the dominance of kinship and other ascriptive principles, and lack of political centralization were significant differences between primitive and industrial societies, there was usually no attempt to work out in any detail how these and other features of primitive society were linked together. This deficiency of theory was reinforced by the diversity of societies considered ‘primitive’: a term which embraced Aboriginal bands, Polynesian chiefdoms, segmentary Nilotic tribes, and even African states was rendered almost completely impotent as an analytical concept.

The obvious method of trying to bring some order and theoretical coherence to the concept of ‘primitive society’ was to use some theory of social evolution, and to classify societies according to a scheme of increasing organizational complexity. Indeed, the very terms which anthropologists used to characterise primitive society – ‘pre-industrial’, ‘pre-literate’, ‘stateless’, ‘simple’ – all have strong evolutionary connotations of which those who use them have been perfectly well aware, while the term ‘primitive’ itself is meaningless outside an evolutionary context. Unfortunately the denial of the idea of social evolution became one of the many obsessions with which anthropology has fettered itself. Evolutionism was closely associated with Marxism, materialism, and laws of historical necessity which most anthropologists found morally and politically repugnant, and schemes of social evolution were also believed to represent our society as superior to all others in an ethnocentric and imperialist scheme of progress. It was often supposed that any theory of social evolution must seek to discover the original forms of social institutions and the first form of human society, and this was rejected as pseudo-historical speculation. No distinction was made between history and evolution, and so it was concluded that since every society existing today has a history which is as long as ours they must be equally ‘evolved’, but in different ways. Again, no distinction was made between types of complexity, and it has been pointed out, for example, that modern Western society has a very simple kin terminology (usually ‘Eskimo’ in type): in our society there is nothing like the complex age-grading systems of East Africa; and our marriage rules cannot compare in complexity with those of the Aborigines. So, it is concluded, ‘primitive society^’ is just as complex as ours, but in different ways.

These conventional assumptions about what any scheme of social evolution must entail are clearly evident in the books of Kuper and Leach.

‘…the orthodox modern view is that there never was such a thing as ‘primitive society’. Certainly, no such thing can be reconstructed now. There is not even a sensible way in which one can specify what a ‘primitive society’ is. The term implies some historical point of reference. It presumably defines a type of society ancestral to more advanced forms, on the analogy of some natural species. But human societies cannot be traced back to a single point of origin, and there is no way of reconstructing prehistoric forms, classifying them and aligning them in a temporal series. There are no fossils of social organization… If it is useful to apply evolutionary theory to social history, then it must direct attention to variation, to adaptation to all sorts of local circumstances, and so to diversification. And it does seem likely that early human societies were indeed rather diverse. Surviving hunter-gatherers certainly do not conform to a single organizational type. Since ecological variations constrain social organization, especially where technology is simple, there must have been considerable differences in social structure between the earliest human societies.’ (Kuper 1988:7).

Leach, like Kuper, is extremely hostile to the possibility of drawing any general conclusions from technology: ‘…the hunter gatherer societies of remote antiquity need not have resembled in the very least [my italics] any of those which are now known to us from direct observation’ (Leach 1982:19). All evidence of archaeology is dismissed in the following terms: ‘Archaeologists are able to demonstrate the historical occurrence of a very small number of isolated, usually uninteresting events. What is more important from my point of view is that they can often demonstrate that particular exercises in conjectural history are invalid; but they cannot do more than that’ (ibid., 50). He is also strongly opposed to the idea that human history has been governed by laws:

‘I am not an historicist. I do not believe in the existence of laws of historical development which will help us either explain or reconstruct the past or to predict the future. I am interested in the details rather than the generalities of the diversity of human culture, but I do not consider that these details are, in any discoverable sense, causally determined.’ (ibid., 49).

These assumptions about evolutionary theory are quite mistaken, however. In the first place a valid theory of social evolution does not require the reconstruction of some initial state of society that was once common to the whole human race, a kind of agnostic version of the Garden of Eden. We certainly have no means of knowing whether the Neanderthals or even the cro-Magnons had the institutions of polyandry or polygyny, exogamy, totemism, or bride-capture, and speculation on such problems is a waste of time because there is no evidence to support it and never will be. But we have considerable evidence from archaeology, history, and ethnography about the consequences of a number of innovations such as agriculture and the domestication of animals, the use of metals, the emergence of the state, conquest warfare, writing, money, and industrialization. Some of these have occurred before others in the course of history, and social evolution is the study of how such innovations occur, why they have the effects that they do, and how they are linked together.

While hunter-gatherers clearly have many differences in cosmologies, rituals, marriage rules and so on they also have many important and distinctive similarities: population density is very low, groups are very small, residence is impermanent, leadership is informal and very restricted in scope, there are no specialized judicial or ritual functionaries and the division of labour in general is very simple, while reciprocity in the form of sharing and gift exchange is the dominant mode of personal interaction. In these respects, at least, it is obvious that modern hunter-gatherer societies must resemble those of pre-agricultural society the world over.

The adoption of agriculture and the domestication of animals leads to a number of relatively well-documented changes in social organization that do not require the support of ‘pseudo-historical speculation’, and the same is true of the emergence of the state and later of industrialization. One does not have to be a cultural materialist to recognize the importance of such innovations in human history, and the emergence of the state, indeed, has relatively little to do with changes in technology. These innovations are not ‘inevitable’, since they depend on the particular historical circumstances of each society, but they are certainly sequential, in the sense that some social forms are the essential preconditions of later ones, and they are for the most part irreversible:

‘Food production, political centralization, the division of labour, literacy, science, intellectual liberalization, appear in a certain historical sequence. They do so because some at least of the later developments in human history seem to presuppose the earlier ones, and could not have preceded them. Human history is a play in which the cast tends to increase over time and within which constraints seem to be imposed on the order in which the characters appear. The theorist of human society cannot introduce them in any old order at will. Some changes are at least relatively irreversible; agriculture, centralization, literacy, science can of course disappear in areas where they were once established, and occasionally such regressions do occur; but, by and large, there does seem to be a kind of overall cumulativeness.’ (Gellner 1988: 13-14).

Nor can these cumulative changes be treated as nothing more than differentiation. In biological evolution there was an historical sequence of habitats from sea to land to air, but fish, reptiles, and birds merely have different capacities which are related to their respective habitats. A bird can do many things which a fish cannot, and vice versa. But the capacities of birds, reptiles and fishes are all far greater than those of the amoeba, for example, and they can do everything that the amoeba can as well. In the same way, modern industrial states have greater capacities than agricultural states, which in turn had greater capacities than stateless societies. This is not ‘ethnocentrism’ but merely the recognition of obvious facts, and it does not oblige us to conclude that our own type of society must therefore be morally superior to all its predecessors.

* * * * *

It might be wondered why, if the whole notion of primitive society is an illusion, it should have been so fundamental to social anthropology. According to Kuper,

‘Its birth may be related to the late Victorian surge of imperialism, and its perhaps terminal decline in the last two decades may be related to the end of the Empire. The rise and fall of nationalism is probably equally relevant. The idea of primitive society fed the common belief that societies were based on blood and soil, and that these principles of descent and territoriality may be equated with race and citizenship, the contrasting components of every imperialism and every nationalism.’ (Kuper 1988: 9).

He also suggest that the idea survived because it generated a specialized set of concepts forming the privileged preserve of a new academic discipline which had a vested interest in maintaining them. Leach is much more forthright:

‘Part of the original appeal of this new arrogant and ethnocentric science was that it fitted perfectly with the ethos of the era of European colonial expansion and- the westward movement of the American frontier, for it rested on the basic premise that all non-Europeans are stupid, childish, barbarous and servile by their very nature. Even today the technical jargon of anthropology is laden with value-loaded implications which stem from their origins in the context of the colonial world…The contemporary primitive peoples… were not regarded as interesting in themselves but only because of what they might, by inference, tell us about the distant past. They were ‘living fossils’; their savage customs were horrid survivals from antiquity which served to illustrate both the stupidity and the depravity of the beast-like behaviour of our primeval ancestors.’ (Leach 1982: 16-17).

Leach attributes the same kind of crude prejudice and contempt for primitive peoples to many of the authors of modern introductory textbooks in anthropology:

‘These samples of human oddity certainly exemplify cultural diversity but the choice of pictures makes it obvious that the people who are to be discussed are ‘primitives’ who lack all the gadgetry of modern technology and modern sanitation. Whatever the authors’ texts may say, the whole layout of such books implies that anthropology is essentially concerned with the lives of people who are in some way inferior and/or deprived.’ (ibid., 123).

This is an extraordinarily prejudiced assessment of the work of generations of anthropologists who have had a sincere intellectual interest in working out how small scale societies without writing, or the state, or money, and with simple technology organize their lives and their view of the world. In this sense there are quite obviously ‘primitive societies’, and when people like Leach and Kuper attribute unworthy motives to those who believe that there is such a thing as primitive society, they invite an inquiry into their own motives for denying its existence. It does not seem unfair or unreasonable to infer that these are basically political, and that Leach and Kuper are dominated by the anti-colonialist, egalitarian, relativistic ethos of Western academic liberalism. One of the more amusing aspects of relativism is that its advocates are under the illusion that everyone is prejudiced except them.

The kind of massive intellectual dishonesty which permeates these two books is not likely to provide the foundations for any scholarly or scientific discipline worthy of the name. According to Leach,

‘…if the findings of anthropology do not have the truth status of either history or science do they have any validity at all? Validity seems to me the wrong word. Social anthropologists should not see themselves as seekers after objective truth; their purpose is to gain insight into other people’s behaviour, or, for that matter, into their own. ‘Insight’ may seem a very vague concept but it is one that we admire in other contexts; it has the quality of deep understanding which, as critics, we attribute to those whom we regard as great artists, dramatists, novelists, composers…’ (ibid., 52).

Kuper’s vision of the New Anthropology is equally negative:

‘It is no longer about the primitive, and no longer particularly or necessarily about ‘the other’. More surprisingly, perhaps, it is no longer centrally concerned with social relations. Mainstream cultural and social anthropology today has abandoned primitive society and, with it, society itself. Instead it is embracing the second tradition of anthropology, the anthropology of Tylor and Frazer, rather than Morgan and Rivers, the anthropology of culture. Meanwhile, on the margin, there is the third tradition of anthropology, which has at its heart the theory of biological evolution. It too has imperial designs and may colonize some of the deserted strongholds of the theory of primitive society.’ (Kuper 1988: 243).

The study of meaning, the sympathetic understanding of other cultures, and the description of human diversity are clearly essential components of social anthropology, but to make them its only aims must inevitably trivialize the subject in the manner so vividly conveyed by these two quotation from Leach and Kuper. Meaning only exists in the context of social relations and institutions, and of the physical environment, and good fieldwork requires not only some of the qualities of the good novelist but those of the good scientist as well. The defining characteristic of ‘science’ is not that it should be modelled on physics or biology (a ‘natural’ science of society is as ridiculous a notion as the idea of a ‘social’ science of chemistry), but that it should try to discover general principles underlying diversity, and support its conclusions by relevant evidence. The mere description of diversity as an end in itself, which Leach advocates, is thoroughly trivial from the scientific point of view. One method of making sense out of the diversity of human society is that of evolutionary theory. There are, of course, other methods of discerning the underlying principles of human society and culture which are not evolutionary – linguistics is one example, and the study of symbolic classification is another – but the recognition that human societies differ qualitatively in an evolutionary manner is one of the basic components of social anthropology, and supplies the essential justification for the concept of primitive society.

References

Claessen, H.J.M. and Skalnik, P. (eds) 1978 The Early State. The Hague:Mouton.

Claessen, H.J.M. and Skalnik, P. (eds) 1981 The Study of the State. The Hague: Mouton.

Claessen, H.J.M., van de Velde, P. and Smith, M.E. 1985 Development and Decline. The evolution of sociopolitical organization. Mass.: Bergin & Garvey.

Douglas, M. 1966. Purity and Danger. An analysis of concepts of purity and taboo.

London: Routledge & Kegan Paul.

Evans-Pritchard, E.E. 1951 Social Anthropology. London: Cohen & West.

Evans-Pritchard, E.E. 1954 ‘Religion’, in The Institutions of Primitive Society,1-11.

Oxford: Blackwell

Firth, R. 1954 ‘Orientations in economic life’, in The Institutions of

PrimitiveSociety, 12-24. Oxford: Blackwell.

Forde, D. 1960 ‘Primitive economies’, in Man, Culture, and Society, ed. H. Shapiro, 330-344. New York: Oxford University Press Galaxy Books.

Fried, M. 1967 The Evolution of Political Society. An essay in political anthropology. New

York: Random House.

Gellner, E. 1988. Plough, Sword, and Book. London: Collins-Harvill.

Haas, J. 1982 The Evolution of the Prehistoric State. Columbia University Press.