This website will be closed down during 2014.

Thomas Hylland Eriksen's new site is now open at

You are now being redirected to the new Engaging with the world site.
























Small places, large issues

An introduction to social and cultural anthropology

Thomas Hylland Eriksen

Pluto Press 1995/2001 (2nd, revised, updated & slightly expanded edition – 3rd, revised and updated version 2010)

Ranging from Pacific islands to the Arctic north and from small villages to modern nation states, this concise and accessible introduction to social anthropology reveals the rich global variation in social life and culture. Also providing a clear overview of anthropology, Dr Eriksen focuses on central topics such as kinship, ethnicity, ritual and political systems, offering a wealth of examples that demonstrate the enormous scope of anthropology and the importance of a comparative perspective. Dr Eriksen emphasises the need to establish interrelations between action and social atructure and between social organisation and cognitive aspects of culture. Unlike previous texts on the subject, Small Places, Large Issues broadens the study to incorporate the anthropology of complex modern societies, thus providing a unique key text for all students of social and cultural anthropology.

The book is distinct from the Norwegian edition in a number of important ways. Most obviously, it is much shorter and more concise (and therefore more affordable in the bookshop). In addition, a number of topical changes have been done. A new secion on the concept of the person has been added, the chapters on politics and the economy have been reorganised and updated, the chapter on economy and ecology has been rewritten extensively, and so on. As I literally rewrote the book, every paragraph has been modified to a greater or lesser extent, although the overall structure will be familiar to readers of the Norwegain edition.

The sample chapter is taken from the first (1995) edition.




Author's preface

1. Social anthropology: Comparison and context
2. Fieldwork and interpretation
3. The social human
4. Local organisation
5. Person and society
6. Kinship
7. Marriage and alliance
8. Gender and age
9. Caste and class
10. Politics and power
11. Exchange
12. Production and technology
13. Religion and ritual
14. Modes of thought
15. The challenge of multiple traditions
16. Ethnicity
17. Nationalism and minorities
18. The local and the global

Epilogue: And so what?



14. Modes of thought

Animals are divided into (a) belonging to the Emperor, (b) embalmed, (c) tame, (d) sucking pigs, (e) sirens, (f) fabulous, (g) stray dogs, (h) included in the present classification, (i) frenzied, (j) innumerable, (k) drawn with a very fine camelhair brush, (l) et cetera, (m) having just broken the water pitcher, (n) that from a long way off look like flies.

-Jorge Luis Borges (quoting from "a certain Chinese encyclopedia")

Whorf's hypothesis and the problem of translation

Benjamin Lee Whorf was an insurance salesman in the USA in the 1920s. A recurrent problem in his job concerned the interpretation of words; the precise meaning of words was often extremely significant concerning indemnity payments. What did it mean, for example, that a fire was self-inflicted? And what did it mean that a drum of petrol was empty? In some cases, it could be empty of petrol, but full of petrol gas and highly explosive. A fire which was caused by an empty petrol drum exploding could, however, not be defined as self-inflicted. Whorf's company lost a bit of money on this kind of cases.

Some years later, Whorf would develop an hypothesis on the relationship between language and the non-linguistic world, which has enjoyed great influence in anthropology. Whorf's teacher in linguistic anthropology, Edward Sapir, had a part in the development of the idea, and the hypothesis is sometimes named the Sapir--Whorf hypothesis, but I shall speak of it as Whorf's hypothesis (Whorf 1956). It postulates that there is an intimate connection between the categories and structure of a language and the ways in which humans are able to experience the world. Whorf paid special attention to the language of the Hopis, which nearly lacked nouns as we know them, and which also lacked the standard verb conjugations common to Indo-European languages. Since the language of the Hopis had these peculiar characteristics, Whorf argued, they would experience the world in a fundamentally different way from the descendants of European settlers in North America, who had brought their languages and grammars to the continent. The language of the Hopis was process-oriented and oriented towards movement whereas English and other European languages were oriented towards things and nouns in general.

Whorf argued that there was an intrinsic connection between the life-world of a people and its language; that every people will develop the linguistic tools it needs to solve tasks perceived as necessary, and that the language of a people will therefore be a significant source of knowledge about their mode of thought, their cosmology and their everyday life.

An immediate implication of Whorf's hypothesis is the problem of cross-cultural translation, one of the perennial problems of anthropology. Is it necessarily possible to translate say, the life-world and culture of the Azande to English? Or could it rather be that their form of life is so closely connected with the Zande language that such a project is doomed to fail - because we will always be forced to interpret them in our own terms, and not in theirs, when we try to describe them in a language other than their own? Whorf did not himself hesitate to describe the differences between the Hopi language and English in comparative, or "etic", terms, and in practice he thus carried out cultural translation. Such translations are necessary for anthropology to be possible, but they are not unproblematic.

The notion of the pre-logical mind

This kind of issues are fundamental to anthropology as a comparative social science. They do not concern research methodology only; they also deal with the question of whether all humans think in roughly the same way, or if there exist culturally specific modes of thought which follow different logics and cannot be faithfully reproduced in a foreign language. When the German explorer von den Steinen reported, in the late 19th century, that the Bororo of Amazonas described themselves as red macaws, many - among them Lévy-Bruhl - drew the conclusion that they were clearly unable to think logically. For how can it be possible to think that one is a parrot and a human being at the same time? The Bororo mode of thought thus had to be pre-logical; they violated Aristotle's principle of contradiction, which states that an object cannot both have and not have one and the same property at the same time and in the same respect. One cannot, in other words, both believe and not believe that one is a parrot. (Later it has become evident that the Bororo by no means contradicted themselves, but rather spoke metaphorically in a way incomprehensible to von den Steinen. He interpreted them in too literary a sense.)

The general problem of translation is still with us, although it has been reformulated many times since the early 1930s. The problem has three main aspects. First: Do "primitive", nonliterate peoples think in a fundamentally different way from ourselves? Secondly: If so, is it possible to understand their life-world and to translate it into a comparative anthropological terminology? Thirdly, is the anthropological terminology inherently culturally embedded, or does it represent a kind of context-free, and therefore comparatively useful, kind of language? There are many ways to approach these issues, and the only answer nearly all anthropologists agree about, is that any differences in modes of thought are not innate - they are not caused by "racial" differences. We must, therefore, study and compare culture and social organisation, even when the topic is the relationship between abstract modes of thought among different peoples.

The mental unity of humanity

One of the central dogmas of anthropology is the principle of the mental unity of humanity. This indicates that the innate characteristics of humanity are roughly the same everywhere - not in the sense that humans are identical, but rather in the sense that inborn differences to not account for cultural variation. If, for example, one had believed that the "races" had varying degrees of intelligence, one might have accepted that there might be inherent genetic causes for the fact that Africans in colonial times were illiterate and engaged in ancestor worship whereas British gentlemen drank port and quoted Shelley. If this had been correct, the entire modern anthropological endeavour would have been superfluous, since it would have been futile to search in culture and social organisation for causes of human variation.

The scientific grounds for claiming that different human groups have systematically varying mental faculties, has never been convincing. The variation within each group has frequently been shown to be greater than the variation between the groups. Within any random sample of individuals, there will be some "smart" and some "stupid" people, some enterprising and some lazy individuals, and so on; but it cannot be shown that say, Sami are intelligent whereas Mbuti are stupid. This is to say that human groups worldwide are endowed with roughly the same innate faculties and potentials, and that cultural variation must be accounted for by referring to events taking place after birth.
Many kinds of cultural variation have been accounted for in this way in previous chapters. Neither the Kula exchange of the Trobrianders, the ancestor cults of the Kaguru or the agricultural technology of the Dogon have been explained through reference to inborn characteristics of the "races". This chapter focuses on variations between different cultural modes of thought, which are some of the most difficult cultural differences both to understand and to account for in comparative terms. I shall begin by discussing whether it may be reasonable to believe in witches, and will then move on to classification, cultural knowledge and the relationship between thought, power and social organisation.

Witchcraft and knowledge among the Azande

The Azande are a patrilineal people of agriculturalists who live largely in southern Sudan, a few hundred kilometres south-west from Nuerland (Evans-Pritchard 1983 [1937]). Their cosmology presumes (in the ethnographic present tense) the existence of a number of spirits of different kinds, including ancestral spirits. In addition, the institution of witchcraft is central to their daily life and worldview. Witchcraft is seen as the individual ability to create misfortune for others in spiritual ways. Only some Azande possess this ability. Unlike magic, which involves medicines and magical formulas, witchcraft is a purely spiritual, generally involuntary activity: the witchcraft power frequently commits its acts while the carrier (the witch) is asleep.

Death and other unfortunate circumstances are usually seen as caused by witchcraft. Traditionally, witches were executed ritually, but at the time of Evans-Pritchard's fieldwork in the late 1920s, this practice had been abandoned, although the belief in witchcraft continued, and even decades later, when many Azande had been proletarised, witchcraft beliefs were common (Reining 1966).
The witchcraft institution provides answers to important questions, and explains why people suffer misfortunes. The notion of witchcraft cannot explain why one catches a fever from snakebite in general, but it does offer an explanation for why a certain person was bitten by a certain snake on a certain day. The scientific doctrine about cause and effect cannot provide explanations of this kind: it cannot tell why the granary had to collapse just when several Azande were resting in its shade. Although the poles supporting the granary were destroyed by termites, the victims held that the accident was ultimately caused by witchcraft.

The notion of witchcraft is not incompatible with a belief in causality. A Zande might agree that certain diseases are caused by bacteria in the drinking water, but he would also want to know why he became ill when his neighbour did not. He would look for the cause in his enemies, whom he would suspect of witchcraft.

Evans-Pritchard suggests that witchcraft is invoked as an explanatory principle "whenever plain reason fails". When somebody is accused of witchcraft, a prince or a witchdoctor consults an oracle to decide the matter. The most important oracle is the poison oracle, which consists of a portion of poison and two fowls. The first fowl is served poison; if it survives, the accused is innocent, but if it dies, he or she is guilty. Then, one double-checks the validity of the verdict by administrating the poison to a second bird.

Evans-Pritchard took witches more seriously than anybody had done earlier, and was concerned to show how the belief in witches made sense, and was perfectly rational, within the Zande world. He was among the earliest to criticise, and discard, the idea that there existed a specifically primitive, "pre-logical" mentality. His aim was to explore the interrelationships between thought and social structure, but not to reduce the former to the latter.

However, at two important points Evans-Pritchard indicates that when all is said and done, the Azande are wrong in assuming that witches exist. First, he introduces a sharp distinction between the witchcraft logic and the scientific logic, and frequently makes statements to the effect that "obviously, witches do not exist". He also distinguishes clearly between mystical notions, notions based on common sense, and scientific notions. Since witchcraft is invisible and (in "our" view) supernatural, a cosmology based on such beliefs falls sqarely into the first category, and must be less valid, on objective grounds, than scientific notions.

Secondly, Evans-Pritchard's monograph ends with a primarily structural functionalist explanation of the witchcraft institution: the belief in witches and similar institutions exist, ultimately, because they contribute to social integration and check deviant behaviour - not because they produce valid insight and understanding.

Winch's criticism

The philosopher Peter Winch, reacting against Evans-Pritchard's distinction between mystical and scientific notions, started a lengthy and heated debate on comparison, rationality and cultural translation when he wrote a paper, in 1964, titled "Understanding a Primitive Society" (Winch 1970).

Winch rejects the idea that there are universal standards available to compare witchcraft beliefs and science. To him, science is just as much as witchcraft based on unverifiable axioms. Winch also claims that Oxford professors are scarcely less superstitious than Azande; they, too, trust blindly in forces they do not fully understand. One of his examples is from meteorology. How many of us really understand the principles of meteorology? Yet, we watch the weather forecasts.
Winch agrees that ideas and notions must be tested in order for their validity to be justified. This, he argues, is done both in scientific experiment and in the Zande consultation of poison oracles, and there is no difference in principle between the two procedures.

Further, Winch claims that scientific experiments are meaningless to someone who is ignorant about the principles of science. For this reason, science - like witchcraft - is not inherently meaningful, but makes sense only within a particular, culturally created frame of reference. Actually, he compares the helplessness of an engineer deprived of his mathematics with the predicament of a Zande without access to his oracles.

To Winch, it is also important to note that the lives of the Azande seem to function well; that their relationship to witchcraft makes their existence meaningful, and that the system by and large is consistent.

The disagreement between Evans-Pritchard and Winch ultimately amounts to divergent views of science. Whereas Evans-Pritchard holds that the Azande are wrong, Winch argues that all knowledge is culturally constructed, and that it can therefore only be deemed right or wrong within its own cultural context. Winch questions the assumption of anthropology to the effect that our comparative concepts are culturally "neutral" - when all is said and done, he suggests, even anthropology is a cultural practice.

Winch draws extensively on Ludwig Wittgenstein's theory of language games (Wittgenstein 1983 [1958]), where the latter argues that knowledge is socially created, and that different systems of knowledge (language games, or in Winch's sense, cultures) are incommensurable and can therefore not be ranked hierarchically nor, strictly speaking, compared. This line of reasoning, which Winch applies not only to anthropological analysis, but also to the anthropologists themselves, can be glossed as a strong version of Whorf's hypothesis, and it seems to render different cultural universes incommensurable for want of a neutral language of comparison.

Let us pose the question differently. Why is it that anthropology as an academic discipline developed in Western Europe and the USA, and not, say, in the Trobriand Islands or Zandeland? As an experiment in thinking, we may imagine a Zande anthropologist who arrives in Britain to look into the local cosmology and cultural perception of death. He would quickly discover that the witchcraft institution is absent in that country, something which clearly must be accounted for. If he is a faithful structural functionalist, he might search for functional causes for the strange denial, on the part of the British, of the existence of witches. Perhaps he would eventually draw the conclusion that the denial of witchcraft, the blind faith in "natural causes of death", strengthened social integration in British society, since it prevented open conflict between families and lineages.

This kind of argument seems to lend support to Winch's relativist position. However, social anthropology did as a matter of fact not develop in Zandeland, but in Britain and other northern countries, and this must also be taken into account. Perhaps the hypothetical example of the Zande anthropologist is best seen as warning against simplistic functionalist explanations, but not as an argument against anthropology as a generalising, comparative discipline. Later in this chapter, I shall suggest some reasons why the Zande did not develop their own comparative science of society and culture.


Durkheim and Mauss were among the earliest to explore the interrelationship between social organisation and patterns of thought. The basic idea in their book Primitive Classification (1963 [1903]) was that thought is a social product, and that different societies thereby produce different kinds of thought. (Unlike Winch, they did not question the privileged position of scientific thought.) A great portion of the book discusses primitive systems of classification, and since its publication, the study of classification has been a central concern in anthropology.

Classification, in the anthropological sense, entails dividing objects, persons, animals and other phenomena according to socially pre-established categories or types. The system of classification is an important part of the knowledge system of any society, and knowledge is always related to social organisation and power. I have just presented arguments against the notion that some kinds of knowledge are "objectively and universally true", and in exploring systems of knowledge, it is necessary to be aware of the interrelationship between knowledge and other parts of the social world, and this includes one's own knowledge.

Just as witchcraft beliefs may seem "irrational" to the ethnocentric observer, alien systems of classification may seem unsystematic to someone who takes the Western system for granted. Ethnographic studies have revealed great variations in the ways other people classify. One famous example is the Karam of highland New Guinea, who do not classify the cassowary as a bird (Bulmer 1967), although Linnaeus (the founder of the scientific system of plant and animal classification) would definitely have classified it as a bird. The cassowary resembles an ostrich: it has feathers and lays eggs, but does not fly. Therefore, the Karam do not consider it a bird. On the other hand, the Karam classify bats together with birds (as flying creatures), even though we "know" that it is "really" a mammal.

For a long time, anthropologists tried to show that the logic of systems of classification was intrinsically connected to the usefulness of plants and animals; that it was simply a functional device for the material reproduction of society. This idea eventually had to be abandoned, and we now turn to showing why.

Classificatory anomalies

The Lele of Kasai (in present-day Zaïre) distinguish meticulously between different classes of animals (Douglas 1975). For example, birds are characterised by feathers, their ability to fly and the laying of eggs, and are thereby distinguished from other animals. However, there are certain animals that do not fit neatly into this logic. The monitor lizard and the tortoise are examples of such exceptions: they lay eggs, but walk on all fours and lack feathers. Douglas describes such "deviant" creatures as anomalies; they fail to fit in. The anomaly, like the liminal phase in Turner's model of the ritual process (Chap. 8), is both outside and inside; it threatens the established order. Anomalous animals are subjected to certain rules; one can only eat them under specific circumstances, women are not allowed to touch them, and so on.

The most important anomalous creature among the Lele is the pangolin (manis tricuspis). It has, the Lele explain, the tail and body of a fish, and it is covered with scales, but it gives birth like a mammal. It has four small legs, and climbs trees (Douglas 1975, p. 33). This animal, it turns out, has an important place in the mythology and ritual life of the Lele. There is a cult of fertility centred on it. The reason, argues Douglas, is that the pangolin is anomalous in a crucial way: in addition to everything else, it gives birth to only one offspring at the time. In this regard, it resembles a human more than an animal. Just as the parents of twins and triplets (who are also anomalies on this score) are seen as mediators between the human and the spiritual worlds, the pangolin is seen as a mediator between humanity and the animal world.

Anomalies are usually associated with danger and pollution. One example, described by Douglas elsewhere (Douglas 1966), is the pig in Middle Eastern religions: Being a cloven-hoofen but not ruminant mammal, it was not classified as edible since edible animals ought to be both cloven-hoofen and ruminant - it was an anomaly. The rather more positive status of the pangolin is caused by the fact, Douglas argues, that the Lele have succeeded in turning a potential curse into a blessing, exploiting the ambiguous status of the animal to their advantage. The pangolin is not economically important, and yet it occupies a central place in Lele cosmology: the pangolin is a mediator.

Totemic classification

When the Bororo spoke of themselves as red macaws, to the bewilderment of von den Steinen, they referred to a system of classification known in the professional literature as totemism. Totemism - the term is of Ojibwa origin - is a knowledge system whereby each social subgroup in a society, usually a clan, has a special, ritual relationship to a class of natural phenomena, usually plants or animals. Totemism has traditionally been particularly widespread in Australia and the Pacific, the Americas and in Africa. For example, the totems of the Algonquin in Quebec include the bear, the fish and thunder in a symbolic system whereby natural phenomena are seen to correspond to aspects of society. The question posed by many anthropologists, from Frazer onwards, was the exact nature of this correspondence.

Malinowski, writing on totemism in the Trobriand Islands, held that totemic plants and animals were chosen because they were inherently useful to the maintenance of society (Malinowski 1974). Radcliffe-Brown, who developed a more complex view on totemism, drew on Durkheim's notion that the attitude towards a totem was caused by a special relationship between the totem and the social order, and that the ultimate function of totemism was to maintain social integration (Radcliffe-Brown 1952 [1929]). The totem is thus a tangible identity marker for a group; Durkheim himself mentions flags as a kind of totem.

Radcliffe-Brown then poses the question of why certain animals and plants are chosen as totems. Like Malinowski and others before him, he assumes that there must be a practical reason, so that, for example, experts in bear hunting take the bear as their totem. In this way, totemism could be seen as a symbolic expression of the division of labour in society.

In a later article, Radcliffe-Brown (1951) raises doubt about his earlier assumption that totemic animals were economically useful to society. At this point, he rather focuses on their symbolic meaning in society. However, he fails to draw a clear conclusion, and Lévi-Strauss is generally credited with resolving the enigma of totemism in anthropology (Lévi-Strauss 1963, 1966). Drawing on an enormous mass of recorded ethnography, largely from North America and Australia, Lévi-Strauss shows that there is no inherent connection between the utilitarian value of a creature and its significance in the totemic system. Instead, he argues, certain animals are chosen because of their mutual relationship - that is, not because of their direct relationship to groups in a segmentary society. The differences between totemic animals (the way they are perceived by the people) correspond to the differences between groups in society (see Fig. 11 --sorry, not available in the electronic version).

Totemic animals contribute to the creation of order; up to this point, Lévi-Strauss agrees with earlier theorists. But, as he puts it, they are not chosen because they are good to eat, but because they are good to think (bons à penser)!

The system of totems and the clans in society are further connected symbolically in two complementary ways, through metaphor and metonymy. A metaphor is a symbol which stands for something else, in the way the milk tree among the Ndembu stands for fertility among women (Chap. 13). A metonym is rather a part which symbolically expresses a whole. Metaphorically, the king may be represented by a lion, metonymically by the crown he wears on his head. The relationship between metaphor and metonymy can be said to correspond to the relationship between melody and harmony (see Leach 1976, Lakoff and Johnson 1980). A metaphor acquires its meaning through its association with the object it represents, while metonymy consists in using a part to represent the whole.

In a totemic system, therefore, each totemic animal stands metonymically for the whole chain of totems, just as each clan stands for the whole society (just as a single word may represent the whole sentence). Simultaneously, of course, the totems are metaphors for each clan. The relationship between the bear and the eagle corresponds to the relationship between the bear clan and the eagle clan. Now, the totems themselves - say, the bear and the eagle - are arbitrary; what counts is the relationship between them.

Undomesticated thinking

A main concern in Lévi-Strauss's work on totemism was to invalidate notions to the effect that there existed a "pre-logical, primitive mode of thought" - although he follows a different path from Evans-Pritchard. The structuralism of Lévi-Strauss seeks to reveal not similarities in actual reasoning, but universal underlying principles for thought and symbolisation.

In La pensée sauvage, "Undomesticated thinking" (misleadingly rendered in English as The Savage Mind, Lévi-Strauss 1966), the fundamental cognitive processes among modern and non-modern peoples are seen as identical. People everywhere think in terms of metaphor and metonymy, and above all, they think through contrasting pairs, so-called binary oppositions. This general model depicting organising principles of thought resembles Bateson's theory of information (Bateson 1972, 1979), where he argues that only differences that make a difference can create knowledge. Both Lévi-Strauss and Bateson are concerned to show that what is essential are relationships rather than the objects themselves.

Lévi-Strauss argues that fundamental thought processes are identical everywhere, but he also indicates that people with different kinds of technology at their disposal will express their thought in very different ways. People who depend on script and numbers clearly think along different lines than nonliterates, he says. Lévi-Strauss compares the literate and nonliterate styles of thinking, and describes the latter as the science of the concrete (la science du concret). When a nonliterate person, living in a society with no script, is to think abstractly, he is forced to align his concepts with concrete, visible objects. Spirits, for example, are abstractions described in terms of their visible manifestations; this explains why many early explorers and missionaries erroneously thought that tribal peoples "worshipped trees and rocks". Originality, in this kind of society, is possible through novel juxtapositions of concepts referring to familiar objects. Lévi-Strauss describes this thought operation as bricolage (a bricoleur can be translated as a handyman, a jack-of-all-trades). This creative, associational and "playful" mode of thought is contrasted with that of the engineer; the abstract science dominant in Western societies, imprisoned and disciplined by writing and numbers.
However, the bricoleur has a limited repertory of symbols at his disposal. The engineer, who creates abstractions from abstractions, may rather try to transcend the familiar. He is tied up - his thought is tamed or domesticated - by writing and numbers, but at the same time he is liberated from the direct communication with natural objects enforced on the "untamed thought" of the bricoleur.

The distinction between bricoleurs and engineers should not be seen as absolute. Today, most societies in the world are "semi-literate", and even Lévi-Strauss himself admits that some modes of thought reminiscent of bricolage, notably in music and poetry, exist even in throughly literate societies. Still, the distinction can be a useful starting-point for an exploration of the interrelationship between knowledge, technology and social organisation.

Writing as technology

In La pensée sauvage Lévi-Strauss distinguishes between what he calls "cold" and "hot" societies. Cold societies see themselves as essentially unchanging, while hot societies are based on an ideology perceiving change as inevitable and potentially beneficial. This distinction corresponds not only to the bricoleur--engineer dichotomy, but also to the distinction between "traditional" and "modern" societies. For the sake of the argument, I shall overstate the contrast between these societal "types" here, but the reader should keep in mind that "modern" and "traditional" are ideal types, and that real societies on the ground are much more complex than this simple dichotomy implies.

The role of script as a form of technology has been discussed by generations of anthropologists (see e.g. Goody 1968, Ong 1982, Finnegan 1988, Street and Besnier 1994). In a number of books, Jack Goody has argued that the introduction of writing may have fundamental effects on thought as well as social organisation, and his idea of the "Great Divide" between nonliterate and literate societies is close kin to Lévi-Strauss' studies of totemic versus historical thinking and the bricoleur--engineer contrast - characteristically, one of Goody's books on literacy is called The Domestication of the Savage Mind (Goody 1977). It could be said that just as Marx turned Hegel on his head (or on his feet!), Goody tries to operationalise and sociologise Lévi-Strauss. Controversial among anthropologists who hold that this kind of distinction is simplistic (e.g. Halverson 1992), Goody's main arguments nevertheless merit to be outlined.

The introduction of writing, Goody argues, enables people to distinguish between concepts and their referents. Writing enables us to turn words into things, to freeze them in time and space. Speech, by contrast, is fleeting and transient, and cannot be fixed for posterity. In this sense, writing entails a reduction of speech: the two are not "the same", and the written version of a statement lacks the extralinguistic context for the statement - facial expression, social situation, tone of voice etc. Writing can indeed be seen as a kind of material culture; like artifacts, it is solid and enduring, and it can be analysed as objectified subjectivity (T. Barth 1991).

Writing arguably liberates thought from the necessity of mnemotechnics; one does not have to remember everything, but can look it up instead. By implication, writing makes the cumulation of vast amounts of knowledge possible in ways orality is unable to. Writing also narrows the meanings of thoughts in the sense that it lends itself, Goody argues, to accurate critical examination in ways which oral statements do not. One may isolate a small bit of human discourse and subject it to thorough examination in ways which cannot be achieved in societies which lack writing. However - and this is a criticism which has repeatedly been levelled against this kind of theory - there are many examples of literate societies where criticism (in the scientific sense) is not encouraged. On the other hand, one may retort that writing is a necessary, but not sufficient condition for science as we know it. This argument, one may agree, goes a long way towards explaining why the Azande did not develop their own comparative science of culture and society - but it does not alone explain why many literate peoples have not done so.

Writing also has great potential importance for social organisation. It has been noted that writing was used at a very early stage (ancient Mesopotamia) for lists, inventories of the amount of grain in the granary, the number of slaves and animals in the city, and so on. As the Christian evangelists witness, censuses were also used very early in the history of writing. Writing thus facilitates not only analytical thought, but also the surveillance of vast numbers of people. It can therefore be regarded as an important kind of technology in the political administration of complex societies.

Finally, a main use of writing in most literate societies has consisted in the building of archives, some of which eventually become history. Lévi-Strauss, commenting on the "totemic void" in Europe and Asia (Lévi-Strauss 1966), concludes that these societies have chosen history instead of totemic myths. He does not see history as inherently "truer" than myth, but rather as a special kind of myth.

The difference between literacy and orality should not be overemphasised: there is by no means a clear-cut distinction. It is nevertheless obvious that the uses of script form an important part of the technology of a society. An abstract ideology such as nationalism, for example (see Chap. 17), is scarcely imaginable without the information technology of writing, which enables members of society to disseminate ideas over a vast area, thus creating bonds of solidarity between millions of individuals who will never know each other personally.

Time and scale

Abstract time, that is the kind of time represented in clocks and calendars, may have effects analogous to those of writing. In the kind of society where most of the readers were raised, it is generally believed that time is something one may have much or little of; something which can be saved, something which "is money", something which can be measured independently of concrete events. Concepts like "one hour" or "one week" are meaningful even if we do not say what they contain by way of events. Time, in this kind of society, is conventionally conceptualised as a line with an arrow at the end, where a moving point called "the present" separates past and future. This kind of abstraction is a cultural invention, neither more nor less. In a certain sense, clocks do not measure time, but create it.

Societies lacking clocks do not "lack time", but rather tend to be organised according to what we may call concrete time (although, as usual, there are very important variations). In this kind of society - historically speaking, the vast majority of human societies - time exists only as embedded in action and process, not as something abstract and autonomous existing outside of the events taking place. Rituals do not take place "at five o'clock", but when all is ready - when the preparations are completed and the guests have arrived. In clockless societies, time is not a scarce resource, since it exists only as events. One cannot "lose" or "kill" time there.

Past and future take on a different meaning in societies with and without an abstract concept of time, respectively. Obviously, peoples without dates and calendars do not date previous events in the same way that we do. Bourdieu, further, has written of the Kabyles that they were shocked to learn of the way the French related to the future (Bourdieu 1963). "The French see themselves as greater than God," they said, "for they believe that they can control the future. But the future belongs to God." Many peoples, further, do not conjugate verbs in the future tense. One philosophically sound way of explaining this may be that events in the world create time, and since no events have yet taken place in the future, the future cannot constitute a time (Tempels 1959).

Linear, quantified, abstract time is not detached from social organisation, but it did not arise mechanically in response to "societal needs". Just as writing, a tool for political control and the advancement of science, was first developed for ritual purposes, the first Europeans to use clocks were monks who needed them to coordinate their prayers. However, abstract time has taken on an important place in the social organisation of contemporary societies. Lewis Mumford has written that the most tyrannical and authoritarian device developed in modern societies was neither the car nor the steam engine, but the clock. The philosopher Henri Bergson, writing in the early decades of this century, was concerned to save the subjective experience of time, la durée, which was threatened by quantified, mechanical time.

Why is it that people living in modern societies have become slaves of the clock, as it were, while others seem to manage perfectly well without it? The answer must be sought in the social organisation of society. If I wish to travel, say, from Oslo to Prague, it would have been extremely inconvenient to have to go to the airport and wait for a day or two so that a sufficient number of passengers to Prague might find their way to the airport. It seems more reasonable that the airline states that the departure will be at 11 a.m., that all of the passengers agree on the meaning of 11 a.m., and thus appear at the airport more or less simultaneously. In other words, the concept of abstract time and the omnipresence of clocks makes it possible to coordinate the actions of a much larger number of people than that which is possible in a society with no shared, quantified notion of time. In other words, both script and abstract time makes social integration at a very high level of scale possible.

Knowledge and power

Evans-Pritchard once wrote that he believed his studies of Azande witchcraft might contribute to the understanding of Communist Russia (Evans-Pritchard 1951). What he meant was that an understanding of the ideological underpinnings of the knowledge system of one society may give clues as to similar structures elsewhere. Definitely, knowledge systems create a particular order in the world, and this does not only concern ideologies of gender, caste, class or ethnicity as dealt with in other chapters, but also the very structuring of experience. In his celebrated novel 1984, George Orwell (1949) describes a society where the language has consciously been changed by the power elite, in order to prevent the citizens from critical thought. In "Newspeak", the word "freedom" has thus lost its meaning of "individual freedom", and can only be used in sentences like "the dog is free from lice". Although such conscious manipulation with language may be rare, there can be no doubt that the kind of insight introduced by Whorf may profitably be used to study ideology and power structures. In our kind of society, the shift from "chairman" to "chairperson" (or simply "chair") and similar changes in language use indicate a growing consciousness about the ideological character of language and concepts.

A different approach to the relationship between knowledge and power is exemplified in the study of so-called secret societies. Initiation into such societies, common in several parts of the world, is accompanied by the acquisition of esoteric, highly valued knowledge. In some societies, such as dynastic China, literacy was seen as esoteric knowledge and kept away from the masses. In Homo Academicus, Bourdieu (1988) actually describes academic knowledge as a political resource of a similar kind. He describes the inaccessible language spoken by academics, the pompous rituals and conventions surrounding academic life in France - allegedly necessary for the "advance of science" - as expressions of symbolic power.

The relationship between knowledge and social organisation can be illuminated in many ways. For example, it is common to assume that culinary differentiation, particularly the development of haute cuisine, is connected with social differentiation and hierarchy. Everything which is taken for granted has a social origin, be it totemic classification, dogmatic belief in the blessings of liberal democracy, belief in God or the idea that one should eat with a knife and a fork. Karl Marx was profoundly aware of this kind of relationship when he wrote, in the mid-nineteenth century, that even the functioning of our five senses is a product of the whole of history up to this day.

This chapter has discussed a number of simple contrasts frequently invoked by anthropologists (especially in the past), between witchcraft accounts and scientific accounts, between the bricoleur and the engineer, between literacy and orality, between abstract linear time and concrete time, and ultimately between large-scale, "modern" and small-scale, "traditional" societies. This kind of dichotomy, which never provided a satisfactory empirical description of the world, has been maintained for generations, at least partly because it facilitates the classification of social and cultural phenomena - if not entire societies. In the remaining chapters, this kind of dichotomous modelling will be subjected to critical scrutiny, and both its strengths and limitations will be made clear.

Suggestions for further reading

E.E. Evans-Pritchard: Witchcraft, Oracles and Magic among the Azande, abridged edition. Oxford: Oxford University Press 1983 (1937).
Alfred Gell: The Anthropology of Time. Oxford: Berg 1992.
Jack Goody: The Domestication of the Savage Mind. Cambridge: Cambridge University Press 1977.
Edmund Leach: Lévi-Strauss. Glasgow: Fontana 1970.
Gisli Pálsson, ed.: Beyond Boundaries. Understanding, Translation and Anthropological Discourse. Oxford: Berg 1993.

©Thomas Hylland Eriksen 1995