Title: Value
Subtitle: The antropological theories of value
Author: David Graeber
Topics: anthropology, value
Date: 2005
Source: Retrieved on 28th November 2021 from davidgraeber.org
Notes: Chapter 27 of James G. Carrier (ed.) A Handbook of Economic Anthropology. Published by Edward Elgar Publishing. ISBN 1-84376-175-0.

Economics might be said to have begun largely as a series of reflections on the origin and nature of value in human society. But those days are long since past. Nowadays, economists tend to limit themselves to producing mathematical models of how economic actors allocate scarce resources in pursuit of profit, or how consumers rank their preferences; they do not ask what those actors are ultimately trying to achieve in life or why consumers want to consume the things they do. The latter sorts of questions, questions of value, have been largely abandoned to anthropologists, sociologists or philosophers.

It is not entirely clear, however, whether an anthropological theory of value actually exists. Anthropologists often talk as if one does, even as if there are quite a few of them. But it is difficult to find anyone willing to describe clearly what such an anthropological theory of value might look like. Instead, one usually finds three different uses of the term, and a feeling that on some ultimate theoretical level they are the same. They are:

  1. ‘values’ in the sociological or philosophical sense. This is the sense in which an anthropologist might say ‘seventeenth-century Hurons placed a high value on individual autonomy’, or a politician might speak of ‘family values’;

  2. ‘value’ in the classic economic sense, in which one might speak of the market value of a house, food processor or ton of pig-iron;

  3. ‘value’ in a more specific linguistic usage, particularly the structural linguistics of Ferdinand de Saussure. He argued that the meaning of a word was essentially a ‘negative value’, a contrast with other words in the same lexicon, as the colour ‘red’ is defined in contrast to ‘yellow’, ‘blue’, ‘brown’, ‘pink’. One might call this ‘value as contrast’ or ‘value as meaningful difference’.

What would it mean to say these are all, ultimately, versions of the same thing? Presumably, several things. It would imply the existence of some kind of symbolic system that defines the world in terms of what is important, meaningful, desirable or worthwhile in it. This system of value would presumably extend to everything from feelings about what one might like to eat for breakfast to what human beings basically owe one another or how one wishes to be remembered after one is dead. And it would imply that there is a way to understand how this system translates into practice.

Most anthropologists use ‘value’ in a way that suggests they believe such symbolic systems do exist. Most nowadays are also pretty certain that old-fashioned ways of talking about bounded ‘societies’ or ‘cultures’ are not the best way to approach the problem, that instead people negotiate their way through a variety of different ‘domains’ or ‘regimes’ of value. But beyond that, there is very little agreement as to how these are actually organised.

This is too bad, because these are important questions, and many critical unresolved issues in social theory turn on the answers. Let me then briefly describe the background in Western economic theory, outline some of the chief ways in which anthropologists have, over the last hundred years or so, used ‘value’ and, finally, examine some recent attempts to create a synthesis.

Value in Western economic theory

The standard history of Western economic theory begins with the mercantilists and physiocrats of the early eighteenth century, then moves on to the political economy school (Adam Smith, David Ricardo, Karl Marx), but sees all of these essentially as precursors. Modern (neoclassical) economics was born from the marginal revolution of the 1870s. It is interesting that what set each of these schools apart from the others was, above all, its views on the origin of value, which over time became increasingly disembodied and subjective. Mercantilists located wealth in precious metals; physiocrats argued that since the ultimate source of value was nature, all social wealth was ultimately derived from agriculture; the political economists claimed that value was a product of human labour (in other words, that it emerged through the body, at exactly the point where our minds become a physical force in nature). For neoclassical economists it transcended the physical altogether, and became simply a subjective measure of desire. From their time on, the value of an object became increasingly indistinguishable from its price: how much potential buyers were willing to give up to acquire some product on the market. It exists only in the eye of the beholder.

It is important to bear in mind that, for earlier economists, value was assumed to be different from price. People were willing to pay money for an item because they saw it as valuable for some other reason. True, prices would also fluctuate owing to the vagaries of supply and demand; but all other things being equal, the market price of a loaf of bread, it was assumed, would tend to gravitate towards what was often called its ‘natural price’, its inherent desirability, measured in relation to the desirability of other items. It was in this sense that value was seen as the regulative principle of prices. Economics could only free itself from such a notion when it purged itself of all moral elements; as a result, the marginal revolution really involved eliminating value from economics entirely.

To understand this, consider Smith’s famous statement of the so-called ‘paradox of value’, which he posed in explaining the distinction, originally posed by Aristotle, between value in use and value in exchange:

The word value, it is to be observed, has two different meanings, and sometimes expresses the utility of some particular object, and sometimes the power of purchasing other goods which the possession of that object conveys. The one may be called ‘value in use’; the other, ‘value in exchange.’ The things which have the greatest value in use have frequently little or no value in exchange; and, on the contrary, those which have the greatest value in exchange have frequently little or no value in use. Nothing is more useful than water: but it will purchase scarce anything; scarce anything can be had in exchange for it. A diamond, on the contrary, has scarce any value in use; but a very great quantity of other goods may frequently be had in exchange for it. (Smith 1776: 28)

Now, the standard line, repeated endlessly in economic textbooks, is that Smith was unable to resolve this paradox because he lacked an adequate theory of supply and demand. His mistake was to, naively, contrast what an economist would now call the total utility, the overall usefulness, of water, with the total utility of diamonds, where he should have been looking at the marginal utility of any unit of water: since the market price of any one unit of a product is the lowest amount a potential buyer is willing to pay for it, and since most people already have access to water, the price they would be willing to pay for an additional unit is likely to be very low. It was because Smith was unable to see this that he was forced to invent the famous labour theory of value, concluding that the high price of diamonds must derive from the fact that it takes all sorts of toil and trouble to produce one, while water falls from the sky. According to the usual account, then, this mistake led economics down a series of hopeless dead ends, ranging from Ricardo’s attempt to calculate the natural price of a commodity through the total number of ‘man hours’ that went into producing it, to pretty much the entire corpus of Marx.

This is all a bit odd because, as others pointed out (for example, Fayazmanesh 1998), there is no reason to believe that Smith even saw this as a paradox, or felt that there was anything to be explained here. He was trying to make a very different sort of point. Smith was, after all, a professor of moral philosophy; economics had, until very recently, been considered a branch of ethics; and the example he used, of diamonds and water, had a long history from theological arguments within the medieval Catholic church.

The paradox of value first appears, in fact, in the works of St Augustine, who in The city of God (IX: 16) noted that there was a great disparity between how one might evaluate things ‘according their own merits’ (by which standard plants are clearly superior to stones, animals to plants, humans to animals) and how humans value them: we would much rather have bread or gold in our house than mice or fleas; people will often spend more money for a horse or jewel than for a slave. For Augustine, this was a result of our fallen nature, because of which we are cursed with endless physical needs and desires. These distort our perceptions. We come to see things through our own needs (use value) rather than their absolute worth, their position on the Great Chain of Being and, hence, proximity to God.

Such reflections were of interest to Scholastic thinkers of the Middle Ages seeking a way to calculate the ‘just price’ of a given commodity. Roman law had defined a just price as anything a buyer and seller were willing to agree on, but this could easily lead to results that flew in the face of any sort of morality. A starving prisoner might be willing to trade his entire fortune for an egg, but that did not make it right to make him do so. A fair or just price, therefore, should have some relation to the ‘intrinsic worth’ or ‘value’ of what you were selling. But how should that be calculated? As theologians from St Thomas Aquinas on were quick to realise, Augustine’s scale of natural perfection was not much help here, since we are dealing with the domain of human needs.[1] But basing intrinsic value on the ability of an object to fulfil human needs did not really work either; first this was not an intrinsic quality, it was relational; second, as the example of water and diamonds revealed, there was little systematic relation between needs and prices anyway. Aquinas concluded that a just price would simply be one the seller would himself have been willing to pay for some commodity; others, notably members of the rival Franciscan school of Duns Scotus, which was far more suspicious of wealth and private property, argued instead that intrinsic value had to be based in its costs of production, which made labour the main source of value.

In The wealth of nations, Smith was clearly drawing on this moral tradition. But as an Enlightenment optimist, he was also committed to the proposition that God had designed the world so that it would essentially run on its own accord for the benefit of humans. Hence his famous argument that the market would, if allowed to work by its own logic, produce an optimal result ‘as by an invisible hand’, an effect which he explicitly said was a result of Divine Providence. Here, too, he was not as much interested in a scientific argument about the reasons for price fluctuations as in a moral argument that in the absence of interference, market prices would indeed always tend to gravitate around the natural price (1776: 51); which in turn meant that people would indeed be justly rewarded for their labours.

The problem was that the latter proposition soon became very hard to defend. Ricardo, who tried to develop Smith’s labour theory of value, was soon to discover what he called an ‘iron law’ that held wages down; this, of course, set the stage for Marx’s critique of the entire wage system as turning human creativity itself into an abstraction that can be bought or sold, necessarily involving alienation, exploitation and the destruction of most of what makes life meaningful or worthwhile. The marginalist revolution simply tossed the problem aside by redefining economics as the study of price formation. After it, any talk of intrinsic value came to be seen as meaningless metaphysics, a primitive idea of some kind of invisible substance that scientific economics could only reject. Value was price, and nothing more. It was purely subjective. By the 1930s, value itself, once considered the main subject of economics, had largely faded away as an object of theoretical reflection. Whatever it was that ultimately motivated people to want the things they did, either it lay outside the purview of economics (a matter for psychologists, perhaps) or, for many, it was inherently unknowable; nothing could be said about it at all.[2] In this way economic theory became the model for a new sort of extreme moral relativism, embodied nowadays in rational-choice approaches to human behaviour that, rather paradoxically, claim a higher moral authority precisely because they are ‘value-free’. In other words, it is no longer necessary to try to prove (like Smith) that the market rewards us justly for our labours, because there is no standard of justice outside of the market itself.

Back to anthropology

The ethical trick largely lies in the word ‘rational’. Market behaviour is by definition rational. In the world as viewed by contemporary, free-market economists (which is, increasingly, identical to the world as viewed by governments and other ruling institutions), anthropologists and their like are therefore relegated to two roles. One is to describe the causes of ‘irrational’ or ‘inefficient’ market behaviour: why people, especially non-Western people, sometimes do not act in the way economic theory says they should. The other is to describe the logic of consumption, which is the one area in which people are not really supposed to be acting rationally, but rather, constructing their unique individual or ethnic identities, or perhaps forging social ties (for example, Douglas and Isherwood 1979; Miller 1987, 1995; Sahlins 1976). Either can lead to employment for anthropologists willing to go that route, either for development agencies or advertising firms. But neither are exactly about value.

The promise of value theory has always been to do much more. It has been to understand the workings of any system of exchange (including free-market capitalism) as part of larger systems of meaning, one containing conceptions of what the cosmos is ultimately about and what is worth pursuing in it. Such systems of meaning meant that the kind of moral and ethical questions that Aquinas or Smith felt were at the heart of the matter could not simply be pushed aside.

I began by observing that those who wish to revive a more comprehensive value theory of this sort generally begin by searching for the common basis between values in the sociological sense, economic value, and the meaningful difference of the linguistic use of the word. There are many reasons to believe such a common basis should exist. For instance, anthropologists, much like sociologists or philosophers, have long been accustomed to speak of values in the plural sense, as one does when, for instance, one says that in Mediterranean societies ‘honour’ is a key value or that, in America, ‘freedom’ is. Within capitalist societies, the word is normally invoked to refer to all those domains of human action that are not governed by the laws of the market: thus we hear about family values, spiritual values, values in the domains of art and political ideals. In other words, ‘values’ begin precisely where (economic) ‘value’ ends. This would certainly imply that we are dealing with two different refractions of the same thing and, therefore, that we would be justified in searching for their common basis.

When economists, or those trying to apply economic approaches to anthropology, actually come to grips with their material, they often come to surprisingly similar conclusions. Neoclassical theory assumes that all human behaviour involves the allocation of resources in pursuit of some kind of scarce good in such a way as to achieve the most product from the least sacrifice. However, anyone who tries to apply this to anthropology will encounter cases where people seem to be vying to sacrifice as much as possible: potlatches, contests of generosity, gratuitous displays of wealth. The usual explanation is that they are trying to ‘maximise’ some other sort of value: prestige, honour, fame, religious merit; precisely the sort of values from which most anthropologists begin.

When anthropologists look for theories of value Every few decades, anthropologists do seem to start thinking along these lines.

It usually seems to occur during moments of crisis. So far, there have been three main moments in which ‘value’ was widely bandied about. The first was the late 1950s, when both British social anthropology and American cultural anthropology had entered a kind of theoretical doldrums, and there was much talk of value theory as a way to break out. The second was the early 1980s, when the great theoretical problem was how to break with structuralism and develop some sort of theory of practice. Arguably, the same thing is happening today.

The reasons for these different instances are, ultimately, similar. Anthropologists had created very powerful models for analysing total social systems, but ran into intractable problems trying to square these models with the reasons why real people do the things they do. One might say this is really the reflection of a much broader dilemma within social theory (see, for example, Caillé 2000), of how to square systemic approaches with individualistic ones. The first begin by imagining some total system or structure – a society, a culture, a world-system – and then try to understand how it is maintained and reproduced over time. The other starts with individual actors pursuing something, and sees society largely as the effect of their actions (here economics and its derivatives, like rational-choice theory, have been the paradigm). Anthropologists have always tended to prefer systemic approaches, but at certain points the theoretical contradictions within a given approach become so overwhelming that there is a sense of crisis. At such points, value tends to become important as a way to bring the advantages of individualistic approaches back in. In the 1950s this happened to British structural–functional anthropology, which for decades had been developing Arthur Radcliffe-Brown’s idea that a society could be seen as a kind of organism, whose parts (institutions) all play some role in maintaining social order. By the 1950s they had taken this about as far as it could go. The problem, they discovered, was to understand (1) how individual members of such a society are motivated to maintain and reproduce it, and (2) how, if they were, societies ever change. Some kind of value theory seemed just the thing to provide the answers. Sir Edward Evans-Pritchard was already playing around with such ideas in 1940 when he described the Nuer concept of cieng (‘home’) as a value in both the linguistic and the political sense (1940: 135–8), but it was Raymond Firth (1964) who really ended up developing the idea, suggesting it might be possible to conceive two analytical levels, the one of ‘social structure’, the fixed world of clans, lineages, age-sets, village moieties and the like that anthropologists had got so good at describing, and another of ‘social organisation’, where individuals pursue value in a more loose, pragmatic fashion.

In the United States the problem was slightly different. Ever since Franz Boas, American anthropology had dedicated itself to the study of something it called ‘culture’, defining ‘cultures’ as symbolic systems, total systems for understanding the world which could not be reduced to individual psychology. When people from different cultures looked at a forest or an ocean or another human being, they literally saw different things. Here the problem was to identify the organising principle of culture. Was it language, or something more like artistic style? Was it based in environment, in certain key symbols, or psychological complexes writ large? Without understanding that, it would be impossible to compare one culture with another. By the 1950s this approach too had something like a crisis.

Into this quandary walked Klyde Kluckhohn, who believed anthropology could be reconceived as a comparative study of values. He fixed on the county of Rimrock, New Mexico, (1951b, 1956; Vogt and Albert 1966) as the first test case for his ‘values project’. Rimrock was divided between five different communities: Navaho, Zuñi, Mormon, Texan and Mexican-American. Its existence, Kluckhohn thought, provided as close as one could get in anthropology to a controlled experiment, a chance to see how five groups of people with profoundly different systems of value adapted to the same environment. He sent off students to study each, remaining behind at Harvard to lead a seminar on values. With help from Florence Kluckhohn, a sociologist, and Edith Albert, a philosopher, he produced a succession of working papers that aimed to hone his theoretical terms.

Kluckhohn defined ‘values’ as ‘conceptions of the desirable’: they were ideas that played some sort of role in influencing the choices people make between different possible courses of action (1951a: 395). By ‘desirable’ he meant that values are not simply what people want (even though desires are largely social, real people want all sorts of different things); they are ideas about what people ought to want. They are the criteria by which people decide whether specific desires are legitimate and worthwhile, or not. So, while values are not necessarily ideas about the meaning of life, they are about what one could justifiably want from it. The problem, though, comes with the second half of the definition: the specific ways that these conceptions relate to behaviour.

In traditional value analysis this is not much of a problem, because traditional value analysis is largely interpretative. One begins by identifying some key term: one observes that members of the Navaho community in Rimrock often talk about the importance of something they call ‘harmony’, or the Texans talk about something they call ‘success’. One then proceeds to interpret precisely what ‘harmony’ or ‘success’ means to them, and then places these definitions in a larger cultural context. The problem is that such terms tend to be highly idiosyncratic: it is difficult to make a systematic comparison between, say, the Texan idea of ‘success’ and the Hindu idea of ‘purity’, and this was precisely what Kluckhohn wanted to do.

To resolve the problem, Kluckhohn proposed a second, less abstract level, ‘value orientations’: ‘assumptions about the ends and purposes of human existence’, the nature of knowledge, ‘what human beings have a right to expect from each other and the gods, about what constitutes fulfillment and frustration’ (Kluckhohn 1949: 358–9). Value orientations thus mixed ideas of the desirable with assumptions about the nature of the world in which one acts. They were also far more uniform, and hence easier to compare. Kluckhohn argued that it should be possible to construct a basic list of existential questions that every culture had to answer: are human beings good or evil? Should their relations with nature be based on harmony, mastery or subjugation? Should one’s ultimate loyalties be to oneself, one’s group or to other individuals? Should one’s time orientation be primarily to the future, the present or the past? All this was quite innovative, and departed radically from the extreme relativism of the American cultural anthropology of the day (basically, Kluckhohn had invented what was later to be called the ‘cosmological approach’), but it never quite worked. He and his students found it very difficult to move from this super-refined level to concrete values like harmony or success, let alone to such mundane questions as what sort of crops people prefer to grow or what sort of marriages they considered to be incestuous.

In the end, the values project came apart. The Rimrock study was published without conclusions; nowadays the project is ignored or seen as yet another dead end. Some (Dumont 1982; Edmonson 1973; compare Nuckolls 1998) suggest that things might have been different if Kluckhohn had lived to see the advent of models from structural linguistics. However, when those models emerged in the 1960s they made all previous debates seem irrelevant. Interest in value did not swell again until the early 1980s, when anthropology was again trying to break out of totalising systems, this time the structuralist models themselves. But by that time ‘values’ was no longer considered an interesting term; instead, the focus was on Saussurean ideas of value.

Saussurean syntheses

It might be surprising that structuralist, Saussurean models of value largely displaced ‘values’ in the old Kluckhohnian sense, since structuralism had little to say about questions of value. Structuralists were concerned with systems of knowledge, the ‘codes’ by which people organise experience, the means by which they interpret messages, rather than what they are trying to accomplish by saying them. The code or system or structure was assumed to exist on some abstract plane outside of time and human action, in the same way that language as an abstract system of rules and meanings exists apart from any particular act of speech. From another perspective, though, the Saussurean approach to value as meaningful difference might seem the perfect thing to mediate between systemic approaches (which analysis of ‘values’ would seem to call for) and individualistic ones, emphasising value in the economic sense. After all, Saussurean approaches insist that the meaning of a term can only be understood in the context of a total system: the meaning of ‘river’, for example, is defined in relation to all the other words in the same lexicon (‘stream’, ‘brook’, ‘inlet’, and so on). Yet at the same time, Saussureans tend to see value as existing in the eye of the beholder, just like neoclassical economists do.

Its approach to both system and individual levels has made this approach attractive to those studying consumption. The distinctions between different sorts of consumer goods provide a map of different sorts of human identity, and what sort of person one is in consumer society is a function of what sort of goods one has. But one can apply Saussurean models to very different sorts of society as well. In her famous analyses of Melanesian ‘gift economies’, Marilyn Strathern (1987, 1988, 1992) has made a number of brilliant and provocative arguments about how objects are valued as ‘ways of making relationships visible’. Such economies are profoundly different from commodity economies, but there too, ‘value’ is a matter of how something is defined in another person’s eyes, as part of some larger system of categories (Graeber 2001).

The early 1980s saw a series of attempts to break out of classical structuralism, usually by trying to develop some sort of theory of action. This eventually paved the way for the various forms of post-structuralism and practice theory that dominate the intellectual field today. For some, this meant returning to economic theories of value (for example, Appadurai 1986; Thomas 1991). For others, it meant exactly the opposite, retooling structuralism so as to create a new and very explicit systemic form. The choice seemed to be largely between the economists and de Saussure.

The most prominent among the latter was the French anthropologist Louis Dumont. He is best known for his analysis of the Indian caste system (Dumont 1980 [1966]), and especially for having been almost single-handedly responsible for popularising the concept of ‘hierarchy’ in the social sciences. His notion of value emerges directly out of his concept of hierarchy. Meaning, he argues, is fundamentally hierarchical, and since all social systems are essentially systems of meaning, they too are organised as total systems around key values. In such societies, whether a Hindu kingdom or Melanesian village, what we would consider material self-interest and the values attached to it are subordinated to larger, cosmological principles; it is only in market societies that this comes to be reversed. Let me explain the argument in greater detail.

Classical structuralism, according to Dumont, was developed as a technique to analyse the formal organisation of ideas. One always proceeds in more or less the same fashion, delimiting a field (whether a series of myths or a social system), identifying its elements, then mapping out the relations between them in terms of certain key conceptual oppositions (for example, raw vs. cooked, pure vs. impure, masculine vs. feminine, consanguinity vs. affinity). One then maps out the ‘relations between the relations’; how these relate to one another. According to Dumont all this is very well and good, but what most structuralists fail to realise is that these ideas are always also ‘values’. This is because, in any such pair of terms, one will be considered superior and will ‘encompass’ the inferior one. All hierarchical relations are based on some such notion of encompassment. One of Dumont’s favourite illustrations is the opposition of right and left. Anthropologists have long noted a tendency, which apparently occurs in the vast majority of the world’s cultures, for the right hand to be treated as somehow morally superior to the left. But, Dumont notes, this is because the right (side) always represents the right (moral). In offering a handshake, Dumont notes, one must normally extend one hand or the other. The right hand put forward thus, in effect, represents one’s person as a whole, including the left hand that is not extended (Dumont 1982; see Tcherkezoff 1983). Hence, at least in that context, the right hand ‘encompasses’ or ‘includes’ the left, which is also its opposite. (This is what he calls ‘encompassing the contrary’.) This principle of hierarchy, he argues, applies to all significant binary oppositions.

Meaning thus arises from making conceptual distinctions that, in turn, are ranked and hence always contain an element of value. Even more important, the social contexts in which these distinctions are put into practice are also ranked. Societies are divided into a series of domains or levels, and higher ones encompass lower ones: they are more universal and thus have more value. In pretty much any society, for instance, domestic affairs, which relate to the interests of a small group of people, will be considered subordinate to political affairs, which represent the concerns of a larger, more inclusive community; and likely as not that political sphere will be considered subordinate to the religious or cosmological one, which represents the concerns of humanity as a whole. Perhaps the most innovative aspect of Dumont’s theory is the way that the relations between different conceptual terms can be inverted on different levels. This can be illustrated with the Indian caste system. On the religious level, where Brahmans represent humanity as a whole before the gods, the operative principle is purity: all castes are ranked according to their purity, and by this standard Brahmans outrank even kings. In the subordinate, political sphere, in which humans relate only to other humans, power is the dominant value, and in that sphere kings are superior to Brahmans, who must do as they say. None the less, Brahmans are ultimately superior, because the sphere in which they are superior is the most encompassing. Running through Dumont’s approach is the Saussurean notion that you have to understand a total system of meaning in order for any particular part of it to make sense. The first step in analysis is to identify some totality. The Dumontians call their project one of ‘comparing wholes’: by which they mean not so much symbolic systems, as societies taken as totalities structured around certain key ‘ideas–values’.

Accordingly, Dumont’s students have contributed a series of concrete analyses of societies in Africa, Melanesia, North Africa and the Indonesian archipelago (Barraud et al. 1994). Each begins by identifying a series of ranked values, which even at their simplest involve a division between the values that regulate exchange between human beings, and those which define human relations with the cosmos as a whole. For the Berbers of the Morroccan Rif, political and economic life centres on exchanges between prominent men locked in endless conflicts over honour; but ultimately honour is encompassed by the importance of baraka, or divine grace. As with the division between purity and power in India, ‘values’ in Kluckhohn’s sense are superior to those that organise the competitive games of political or economic self-interest, whether these be battles over prestige (as in gift economies) or struggles to control material wealth. The distinction corresponds almost exactly to St Augustine’s distinction between ‘objective’ value, the intrinsic merit of a thing, its place in the cosmos, and value as interpreted through the faulty lens of human desire.

According to Dumont (1971), what makes contemporary Western society so unusual is that it inverts all this. Westerners no longer live in a hierarchical society, ranked into a series of ever-more inclusive domains. However, this is not because Westerners value equality (it is an item of faith for Dumontians that equality cannot, itself, be a value) but because for them, the supreme value has become the individual. Each person is assumed to be unique and thus, by definition, incomparable. If all individuals are values unto themselves (an idea which he traces back to Christian ideas about the value of the immortal soul), none can be treated as intrinsically superior to any other. It is this which has allowed the market, as the sphere of individual self-realisation, to become the hierarchically dominant, highest sphere, to which art, religion, science and politics are all increasingly subordinate.

Value as the meaning of actions

For a long time, Marxist anthropology did not have much to say about value (for one exception, see Taussig 1980). However, one of the most promising of the new approaches that began to emerge in the 1980s grew primarily out of Marx’s insight that value ultimately measures the importance not of objects, but of actions. The two most important advocates of this new approach were Terence Turner (1979, 1984, 1987) and Nancy Munn (1986; her phenomenological approach does not derive directly from Marx, but draws on the same dialectical tradition). I have tried to develop their ideas in my own work (Graeber 2001). Let me end, then, with a brief outline of how an action-based value system might work.

Like Smith, Marx did not propose a labour theory of value mainly as a way to explain price fluctuations,[3] but as a way of connecting economic theory with broader moral and philosophical concerns. For Marx, ‘labour’ was more or less identical with human creativity: it is the way human beings exercise their imaginative powers to create their worlds, their social ties as well as their physical environments. The unique thing about capitalism, Marx held, was that it allowed labour to become an abstraction. This was because capitalism alone turns labour into a commodity, something that can be bought or sold, and what an employer who hires a labourer buys is an abstraction, that labourer’s capacity to work. What makes this possible is the use of a specific symbolic medium of value: money. For Marx, money is a symbol. It represents the ‘value’ or importance of labour. It can do so by incorporating it into a total market system, because for Marx the real value of a product is not (as Ricardo claimed) how many hours of work went into making it, but the proportion of the total amount of labour in the entire economy that went into making it. This proportion one can only determine through the market; that is, through the use of money. But even more, money is a symbol that brings into being the very thing that it symbolises: after all, wage labourers only go to work in order to get money. This is why, in Marx’s terms, we can say that money is being ‘fetishised’. Value is the way our actions take on meaning or importance by becoming incorporated into something larger than ourselves. But almost always, this can only happen through some kind of material medium, a token of value like money. Fetishism occurs when we assume that the value comes from the token, rather than ourselves.

In most known human societies there was no such market in labour; but still, one can say something similar happens. Different sorts of labour (perhaps better put: different forms of human creative activity) cluster together. They tend to get reflected back in the form of concrete, material media that are seen as valuable in themselves, and thus end up becoming the actual ends for which action takes place. Tokens of honour inspire honourable behaviour, tokens of piety inspire religious devotion, tokens of wisdom inspire learning and so on. Their value is just that of the actions they represent, but the actors see them as valuable in themselves. Usually, these tokens are unlike money, which is an abstract medium in which all tokens are effectively the same: any dollar bill is to all practical intents and purposes identical to any other. Instead, these tokens are unique, particular things: heirlooms, unique gestures, titles and so on. But this makes sense, too. After all, the same thing happens, to a lesser degree, in our own society. Recall my earlier remark about economic ‘value’ applying within the market, and a more particular sort of ‘values’ applying outside it. Abstract labour, the sort you get in the capitalist workplaces, ends up being materialised in abstract symbols; more concrete forms of labour end up being materialised in more concrete symbols. So housework and childcare, for which one is not paid, becomes a matter of ‘family values’ and is reflected in tokens of love and respect; work for the church becomes a matter of religious values, political activism is inspired by social values and so on. But even here, there is usually some sort of material token through which it all becomes real.

To schematise matters considerably (and readers interested in seeing any of this fleshed out are encouraged to consult Graeber 2001), allow me to suggest the following. First, value is the way actors represent the importance of their own actions to themselves as part of some larger whole (or ‘concrete totality’, as Marx liked to put it). Second, this importance is always seen in comparative terms. Some forms of value are seen as unique and incommensurable; others are ranked (as in categories of kula valuables or the famous Tiv ‘spheres of exchange’; see Strathern and Stewart chap. 14, and Isaac chap.1 supra); for yet others, such as money in market systems, value can be calculated precisely, so that one can know precisely how many of item A are equivalent to one item B. Third, importance is always realised through some kind of material token, and generally is realised somewhere other than the place it is primarily produced. For instance, in non-capitalist societies commonly there is a domestic sphere in which most of the primary work of people-creation takes place, and this is distinguished from a public, political sphere, in which the value generated by that work is realised, but usually in ways that exclude the women and younger people who do the bulk of the work.

Turner has tried to develop such ideas in a series of analyses of the Kayapo, of central Brazil. One of Turner’s key points is that in non-capitalist societies the bulk of social labour is not so much directed at creating material objects as at shaping and reshaping human beings and the relations between them; the Kayapo see material production as a subordinate aspect of the reproduction of people. Hence Kayapo communities are organised as rings of households, surrounding a central, public, political space dominated by a plaza and adolescent men’s houses. The households can be seen as the areas where the bulk of the creative work in raising and socialising children (and, for that matter, adults) takes place, through relations which themselves embody the two key Kayapo values: ‘beauty’ (a kind of total, integrative harmony) and a less articulated value that Turner variously calls ‘power’ or ‘dominance’. These forms of value are ultimately produced in the domestic sphere. However, they are realised primarily in the central, public space, especially through certain forms of public performance, by elders who are themselves ‘elders’ only because they are the peak of a domestic process of creating and socialising people that takes place just offstage, and which is carried out primarily by people other than themselves.

This might seem to resemble the kind of terms a Dumontian might have discovered, but framing it this way emphasises that the process of realisation of value involves some form of public recognition. But this is not to say that people are simply battling over prestige; instead, the range of people who are willing to recognise certain forms of value constitutes the extent of what an actor considers a ‘society’ to consist of. There are any number of directions in which this kind of approach could be developed, though it remains to be seen whether it can resolve the endless paradoxes and moral dilemmas which have dogged the study of material exchange from Aristotle and Augustine onwards.


The study of value, then, invariably takes us beyond what we normally refer to as ‘economics’, for it leads us into moral, aesthetic and symbolic territory that is very hard to reduce to rational calculation and science. In the Western tradition, economics began as a series of questions about the morality of value; it could only claim the status of a science by trying to exorcise value completely. Anthropologists, on the other hand, have tended to see their special expertise as lying in precisely the areas that economics abandoned. However, it appears that anthropologists have only tried to develop explicit theories of value when they find themselves in a crisis brought about by their inability to understand how flesh-and-blood individuals are motivated to maintain and re-create the abstract systems that anthropologists have always been so good at discerning. Since the failure of Kluckhohn’s ‘values project’ in the 1950s, this has usually led anthropologists to work with some variation of economic models, or with linguistic models in the structuralist tradition of de Saussure. I have suggested that there are other possibilities, especially one that treats Marx’s analysis of value as a symbolic analysis and looks at ‘value’ as a way people’s own actions become meaningful to them, how they take on importance by becoming incorporated into some larger system of meaning. We can only wait to see which, if any, of these many strands of value theory are most useful in the future.


Appadurai, A. 1986. Introduction. In The social life of things (ed.) A. Appadurai. Cambridge: Cambridge University Press.

Barraud, C., D. de Coppet, A. Iteanu and R. Jamous 1994. Of relations and the dead: four societies viewed from the angle of their exchanges. (trans. S.J. Suffern). Oxford: Berg.

Caillé, A. 2000. Anthropologie du don. Paris: Éditions de Decouverte/MAUSS.

Douglas, M. and B. Isherwood 1979. The world of goods. London: Routledge.

Dumont, L. 1971. From Mandeville to Marx. Chicago: University of Chicago Press.

Dumont, L. 1980 [1966]. Homo hierarchichus. Chicago: University of Chicago Press.

Dumont, L. 1982. On value. Proceedings of the British Academy 66: 207–41.

Edmonson, M.S. 1973. The anthropology of values. In Culture and life: essays in memory of Klyde Kluckhohn (eds) W. Taylor, J. Eischer, E. Vogt. Carbondale: Southern Illinois University Press.

Evans-Pritchard, E.E. 1940. The Nuer. Oxford: Oxford University Press.

Fayazmanesh, S. 1998. The magical, mystical ‘paradox of value’. Research in the History of Economic Thought and Methodology 16: 123–53.

Firth, R. 1964. The study of values by social anthropologists. In Essays on social organization and values. London: Athlone.

Graeber, D. 2001. Toward an anthropological theory of value. New York: Palgrave.

Kluckhohn, K. 1949. The philosophy of the Navaho Indian. In Ideological differences and world order (ed.) F.S.D. Northrop. New Haven, Conn.: Yale University Press.

Kluckhohn, K. 1951a. Values and value-orientations in the theory of action: an exploration in definition and classification. In Towards a general theory of action (eds) T. Parsons and E. Shils. Cambridge, Mass.: Harvard University Press.

Kluckhohn, K. 1951b. A comparative study of values in five cultures. In Navaho veterans, a study in changing values, E.Z. Vogt ( Papers of the Peabody Museum of American Archaeology and Ethnology 41: vii–ix.) Cambridge, Mass.: Harvard University.

Kluckhohn, K. 1956. Towards a comparison of value-emphases in different cultures. In The state of the social sciences (ed.) L. White. Chicago: University of Chicago Press.

Langholm, O. 1992. Economics in the medieval schools. Leiden: E.J. Brill.

Miller, D. 1987. Material culture and mass consumption. Oxford: Basil Blackwell.

Miller, D. (ed.) 1995. Acknowledging consumption. London: Routledge.

Munn, N. 1986. The fame of Gawa. Cambridge: Cambridge University Press.

Nuckolls, C.W. 1998. Culture: a problem that cannot be solved. Madison: University of Wisconsin Press.

Sahlins, M. 1976. Culture and practical reason. Chicago: University of Chicago Press.

Smith, A. 1776. An enquiry concerning the wealth of nations. www.adamsmith.org.

Strathern, M. 1987. Conclusion. In Dealing with inequality: analysing gender relations in Melanesia and beyond (ed.) M. Strathern. Cambridge: Cambridge University Press.

Strathern, M. 1988. The gender of the gift. Berkeley: University of California Press.

Strathern, M. 1992. Qualified value: the perspective of gift exchange. In Barter, exchange and value (eds) C. Humphrey and S. Hugh-Jones. Cambridge: Cambridge University Press.

Taussig, M. 1980. The devil and commodity fetishism in South America. Chapel Hill: University of North Carolina Press.

Tcherkezoff, S. 1983. Le Roi Nyamwezi, la droite et la gauche. Révision comparative des classifications dualistes. Paris: Atelier d’Anthropologie Sociale, Maison des Sciences de l’Homme. (Published in English as Dual classification reconsidered: Nyamwezi sacred kingship and other examples. Cambridge: Cambridge University Press. 1987.) Thomas, N. 1991. Entangled objects. Cambridge, Mass.: Harvard University Press.

Turner, T. 1979. Anthropology and the politics of indigenous peoples’ struggles. Cambridge Anthropology 5: 1–43.

Turner, T. 1984. Value, production and exploitation in non-capitalist societies. To appear in Critique of pure culture. New York: Berg [forthcoming].

Turner, T. 1987. The Kayapo of southeastern Para. MS.

Vogt, E.Z. and E.M. Albert (eds) 1966. The people of Rimrock: a study of values in five cultures. Cambridge, Mass.: Harvard University Press.

[1] ‘But this one standard which truly measures all things is demand … Articles are not valued according to the dignity of their nature, otherwise a mouse, an animal endowed with sense, should be of greater value than a pearl, a thing without life. But they are priced according as man stands in need of them for his own use’ ( Sententia libri ethicorum V, 9 cited in Langholm 1992: 229).

[2] This is already very much implied by the notion that value was reflected in ‘utility’, which, after all, means not desirability but usefulness. That is, one values something for its ability to get one something else. What that something else is has already been pushed out of consideration. By the 1930s, with the ‘ordinal revolution’, economists discovered they could model consumer behaviour simply as a series of ranked preferences, which eliminated the need for even something as vague as utility.

[3] In fact, one of the great problems for Marxist economists has always been figuring out precisely how the connection between values and prices is supposed to work, the famous ‘transformation problem’.