The Social Life of Things

Few will deny that a commodity is a thoroughly socialized thing. The definitional question is: in what does its sociality consist? The purist answer, routinely attributed to Marx, is that a commodity is a product intended principally for exchange, and that such products emerge, by definition, in the institutional, psychological, and economic conditions of capitalism. Less purist definitions regard commodities as goods intended for exchange, regardless of the form of the exchange. The purist definition forecloses the question prematurely. The looser definitions threaten to equate commodity with gift and many other kinds of thing. In this section, through a critique of the Marxian understanding of the commodity, I shall suggest that commodities are things with a particular type of social potential, that they are distinguishable from "products," "objects," "goods," "artifacts," and other sorts of things – but only in certain respects and from a certain point of view. If my argument holds water, it will follow that it is definitionally useful to regard commodities as existing in a very wide variety of societies (though with a special intensity and salience in

modern, capitalist societies), and that there is an unexpected convergence between Marx and Simmel on the topic of commodities.

The most elaborate and thought-provoking discussion of the idea of the commodity appears in Volume I, Part I, of Marx's Capital, though the idea was widespread in nineteenth-century discussions of political economy. Marx's own reanalysis of the concept of commodity was a central part of his critique of bourgeois political economy and a fulcrum for the transition from his own earlier thought (see especially Marx 1973) on capitalism to the full-fledged analysis of Capital. Today, the conceptual centrality of the idea of commodity has given way to the neoclassical, marginalist conception of "goods," and the word "commodity" is used in neoclassical economics only to refer to a special subclass of primary goods and no longer plays a central analytic role. This is, of course, not the case with Marxian approaches in economics and sociology, or with neo-Ricardian approaches (such as those of Piero Sraffa), where the analysis of the "commodity" still plays a central theoretical role (Sraffa 1961; Seddon 1978).

But in most modern analyses of economy (outside anthropology), the meaning of the term commodity has narrowed to reflect only one part of the heritage of Marx and the early political economists. That is, in most contemporary uses, commodities are special kinds of manufactured goods (or services), which are associated only with capitalist modes of production and are thus to be found only where capitalism has penetrated. Thus even in current debates about proto-industrialization (see, for example, Perlin 1982), the issue is not whether commodities are associated with capitalism, but whether certain organizational and technical forms associated with capitalism are solely of European origin. Commodities are generally seen as typical material representations of the capitalist mode of production, even if they are classified as petty and their capitalist context as incipient.

Yet it is clear that this is to draw on only one strand in Marx's own understanding of the nature of the commodity. The treatment of the commodity in the first hundred or so pages of Capital is arguably one of the most difficult, contradictory, and ambiguous parts of Marx's corpus. It begins with an extremely broad definition of commodity ("A commodity is, in the first place, an object outside us, a thing that by its properties satisfies human wants of some sort or another"). It then moves dialectically through a series of more parsimonious definitions, which permit the gradual elaboration of the basic Marxian approach to use value and exchange value, the problem of equivalence, the circulation and exchange of products, and the significance of money. It is the elaboration of this understanding of the relation-

ship between the commodity form and the money form that allows Marx to make his famous distinction between two forms of circulation of commodities (Commodities-Money-Commodities and Money-Commodities-Money), the latter representing the general formula for capital. In the course of this analytic movement, commodities become intricately tied to money, an impersonal market, and exchange value. Even in the simple form of circulation (tied to use value), commodities are related through the commensuration capabilities of money. Today, in general, the link of commodities to postindustrial social, financial, and exchange forms is taken for granted, even by those who in other regards do not take Marx seriously.

Yet in Marx's own writings, there is the basis for a much broader, more cross-culturally and historically useful approach to commodities, whose spirit is attenuated as soon as he becomes embroiled in the details of his analysis of nineteenth-century industrial capitalism. By this earlier formulation, in order to produce not mere products but commodities, a man must produce use values for others, social use values (Marx 1971:48). This idea was glossed by Engels in a parenthesis he inserted into Marx's text in the following interesting way: "To become a commodity a product must be transferred to another, whom it will serve as a use-value, by means of an exchange" (Marx 1971:48). Though Engels was content with this elucidation, Marx proceeds to make a very complex (and ambiguous) series of distinctions between products and commodities, but for anthropological purposes, the key passage deserves quotation in full:

Every product of labour is, in all states of society, a use-value; but it is only at a definite historical epoch in a society's development that such a product becomes a commodity, viz. at the epoch when the labour spent on the production of a useful article becomes expressed as one of the objective qualities of that article, i.e., as its value. It therefore follows that the elementary value-form is also the primitive form under which a product of labour appears historically as a commodity, and that the gradual transformation of such products into commodities, proceeds pari passu with the development of the value-form. (Marx 1971:67).

The difficulty of distinguishing the logical aspect of this argument from its historical aspect has been noted by Anne Chapman (1980), whose argument I will return to shortly. In the above passage from Capital, the shift from product to commodity is discussed historically. But the resolution is still highly schematic, and it is difficult to specify or test it in any clear way.

The point is that Marx was still imprisoned in two aspects of the mid-nineteenth-century episteme: one could see the economy only in

reference to the problematics of production (Baudrillard 1975); the other regarded the movement to commodity production as evolutionary, unidirectional, and historical. As a result commodities either exist or do not exist, and they are products of a particular sort. Each of these assumptions requires modification.

Despite these epistemic limitations, in his famous discussion of the fetishism of commodities, Marx does note, as he does elsewhere in Capital, that the commodity does not emerge whole-cloth from the product under bourgeois production, but makes its appearance "at an early date in history, though not in the same predominating and characteristic manner as nowadays." (Marx 1971:86). Though it is outside the scope of this essay to explore the difficulties of Marx's own thought on precapitalist, nonstate, nonmonetary economies, we might note that Marx left the door open for the existence of commodities, at least in a primitive form, in many sorts of society.

The definitional strategy I propose is a return to a version of Engels's emendation of Marx's broad definition involving the production of use value for others, which converges with Simmel's emphasis on exchange as the source of economic value. Let us start with the idea that a commodity is any thing intended for exchange. This gets us away from the exclusive preoccupation with the "product," "production," and the original or dominant intention of the "producer" and permits us to focus on the dynamics of exchange. For comparative purposes, then, the question becomes not "What is a commodity?" but rather "What sort of an exchange is commodity exchange?" Here, and as part of the effort to define commodities better, we need to deal with two kinds of exchange that are conventionally contrasted with commodity exchange. The first is barter (sometimes referred to as direct exchange), and the other is the exchange of gifts. Let us start with barter.

Barter as a form of exchange has recently been analyzed by Chapman (1980) in an essay that, among other things, takes issue with Marx's own analysis of the relationship between direct exchange and commodity exchange. Combining aspects of several current definitions of barter (including Chapman's), I would suggest that barter is the exchange of objects for one another without reference to money and with maximum feasible reduction of social, cultural, political, or personal transaction costs. The former criterion distinguishes barter from commodity exchange in the strict Marxist sense, and the latter from gift exchange by virtually any definition.

Chapman is right that, insofar as Marx's theory of value is taken seriously, his treatment of barter poses insoluble theoretical and con-

ceptual problems (Chapman 1980:68–70), for Marx postulated that barter took the form of direct exchange of the product (x use value A = y use value B), as well as direct exchange of the commodity (x commodity A = y commodity B). But this Marxist view of barter, whatever problems it may pose for a Marxist theory of the origin of exchange value, has the virtue of fitting well with Chapman's most persuasive claim – that barter, as either a dominant or a subordinate form of exchange, exists in an extremely wide range of societies. Chapman criticizes Marx for inserting the commodity into barter and wishes to keep them quite separate, on the grounds that commodities assume the use of money objects (and thus congealed labor value), and not just money as a unit of account or measure of equivalence. Commodity exchange, for Chapman, occurs only when a money object intervenes in exchange. Since barter, in her model, excludes such intervention, commodity exchange and barter are formally completely distinct, though they may coexist in some societies (Chapman 1980:67–68).

In her critique of Marx, it seems to me, Chapman takes an unduly constricted view of the role of money in the circulation of commodities. Though Marx ran into difficulties in his own analysis of the relationship between barter and commodity exchange, he was right to see, as did Polanyi, that there was a commonality of spirit between barter and capitalist commodity exchange, a commonality tied (in this view) to the object-centered, relatively impersonal, asocial nature of each. In the various simple forms of barter, we see an effort to exchange things without the constraints of sociality on the one hand, and the complications of money on the other. Barter in the contemporary world is on the increase: one estimate has it that an estimated $12 billion a year in goods and services is bartered in the United States alone. International barter (Pepsico syrup for Russian vodka; Coca-Cola for Korean toothpicks and Bulgarian folklofts are examples) is also developing into a complex alternative economy. In these latter situations, barter is a response to the growing number of barriers to international trade and finance, and has a specific role to play in the larger economy. Barter, as a form of trade, thus links the exchange of commodities in widely different social, technological, and institutional circumstances. Barter may thus be regarded as a special form of commodity exchange, one in which, for any variety of reasons, money plays either no role or a very indirect role (as a mere unit of account). By this definition of barter, it would be difficult to locate any human society in which commodity exchange is completely irrelevant. Barter appears to be the form of commodity exchange in

which the circulation of things is most divorced from social, political, or cultural norms. Yet wherever evidence is available, the determination of what may be bartered, where, when, and by whom, as well as of what drives the demand for the goods of the "other," is a social affair. There is a deep tendency to regard this social regulation as a largely negative matter, so that barter in small-scale societies and in earlier periods is frequently regarded as having been restricted to the relation between communities rather than within communities. Barter is, in this model, held to be in inverse proportion to sociality, and foreign trade, by extension, is seen to have 'preceded' internal trade (Sahlins 1972). But there are good empirical and methodological reasons to question this view.

The notion that trade in nonmonetized, preindustrial economies is generally regarded as antisocial from the point of view of face-to-face communities and thus was frequently restricted to dealings with strangers has as its close counterpart the view that the spirit of the gift and that of the commodity are deeply opposed. In this view, gift exchange and commodity exchange are fundamentally contrastive and mutually exclusive. Though there have been some important recent attempts to mute the exaggerated contrast between Marx and Mauss (Hart 1982; Tambiah 1984), the tendency to see these two modalities of exchange as fundamentally opposed remains a marked feature of anthropological discourse (Dumont 1980; Hyde 1979; Gregory 1982; Sahlins 1972; Taussig 1980).

The exaggeration and reification of the contrast between gift and commodity in anthropological writing has many sources. Among them are the tendency to romanticize small-scale societies; to conflate use value (in Marx's sense) with gemeinschaft (in Toennies's sense); the tendency to forget that capitalist societies, too, operate according to cultural designs; the proclivity to marginalize and underplay the calculative, impersonal and self-aggrandizing features of noncapitalist societies. These tendencies, in turn, are a product of an oversimplified view of the opposition between Mauss and Marx, which, as Keith Hart (1982) has suggested, misses important aspects of the commonalities between them.

Gifts, and the spirit of reciprocity, sociability, and spontaneity in which they are typically exchanged, usually are starkly opposed to the profit-oriented, self-centered, and calculated spirit that fires the circulation of commodities. Further, where gifts link things to persons and embed the flow of things in the flow of social relations, commodities are held to represent the drive – largely free of moral or cultural constraints – of goods for one another, a drive mediated by

money and not by sociality. Many of the essays in this volume, as well as my own argument here, are designed to show that this is a simplified and overdrawn series of contrasts. For the present, though, let me propose one important quality that gift exchange and the circulation of commodities share.

My view of the spirit of gift exchange owes a good deal to Bourdieu (1977), who has extended a hitherto underplayed aspect of Mauss's analysis of the gift (Mauss 1976:70-3), which stresses certain strategic parallels between gift exchange and more ostensibly "economic" practices. Bourdieu's argument, which stresses the temporal dynamics of gifting, makes a shrewd analysis of the common spirit that underlies both gift and commodity circulation:

It is true that the lapse of time interposed is what enables the gift or counter-gift to be seen and experienced as an inaugural act of generosity, without any past or future, i.e., without calculation, then it is clear that in reducing the polythetic to the monotheistic, objectivism destroys the specificity of all practices which, like gift exchange, tend or pretend to put the law of self-interest into abeyance. A rational contract would telescope into an instant a transaction which gift exchange disguises, by stretching it out in time; and because of this, gift exchange is, if not the only mode of commodity circulation practiced, at least the only mode to be fully recognized, in societies which, because they deny "the true soil of their life," as Lukács puts it, have an economy in itself and not for itself. (Bourdieu 1977:171.)

This treatment of gift exchange as a particular form of the circulation of commodities comes out of Bourdieu's critique not only of "objectivist" treatments of social action, but of the sort of ethnocentrism, itself a historical product of capitalism, that assumes a very restricted definition of economic interest.³ Bourdieu suggests that "practice never ceases to conform to economic calculation even when it gives every appearance of disinterestedness by departing from the logic of interested calculation (in the narrow sense) and playing for stakes that are non-material and not easily quantified" (ibid:177).

I take this suggestion to converge, though from a slightly different angle, with the proposals of Tambiah (1984), Baudrillard (1968; 1975; 1981), Sahlins (1976), and Douglas and Isherwood (1981), all of which represent efforts to restore the cultural dimension to societies that are too often represented simply as economies writ large, and to restore the calculative dimension to societies that are too often simply portrayed as solidarity writ small. Part of the difficulty with a cross-cultural analysis of commodities is that, as with other matters in social life, anthropology is excessively dualistic: "us and them"; "materialist and religious"; "objectification of persons" versus "personification of

things"; "market exchange" versus "reciprocity"; and so forth. These oppositions parody both poles and reduce human diversities artificially. One symptom of this problem has been an excessively positivist conception of the commodity, as being a certain kind of thing, thus restricting the debate to the matter of deciding what kind of thing it is. But, in trying to understand what is distinctive about commodity exchange, it does not make sense to distinguish it sharply either from barter on the one hand, or from the exchange of gifts on the other. As Simmel (1978:97-8), suggests, it is important to see the calculative dimension in all these forms of exchange, even if they vary in the form and intensity of sociality associated with them. It remains now to characterize commodity exchange in a comparative and processual manner.

Let us approach commodities as things in a certain situation, a situation that can characterize many different kinds of thing, at different points in their social lives. This means looking at the commodity potential of all things rather than searching fruitlessly for the magic distinction between commodities and other sorts of things. It also means breaking significantly with the production-dominated Marxian view of the commodity and focusing on its total trajectory from production, through exchange/distribution, to consumption.

But how are we to define the commodity situation? I propose that the commodity situation in the social life of any "thing" be defined as the situation in which its exchangeability (past, present, or future) for some other thing is its socially relevant feature. Further, the commodity situation, defined this way, can be disaggregated into: (1) the commodity phase of the social life of any thing; (2) the commodity candidacy of any thing; and (3) the commodity context in which any thing may be placed. Each of these aspects of "commodity-hood" needs some explication.

The idea of the commodity phase in the social life of a thing is a summary way to capture the central insight in Igor Kopytoff's important essay in this volume, where certain things are seen as moving in and out of the commodity state. I shall have more to say on this biographical approach to things in the next section, but let us note for the moment that things can move in and out of the commodity state, that such movements can be slow or fast, reversible or terminal, normative or deviant.⁴ Though the biographical aspect of some things (such as heirlooms, postage stamps, and antiques) may be more noticeable than that of some others (such as steel bars, salt, or sugar), this component is never completely irrelevant.

The commodity candidacy of things is less a temporal than a con-

ceptual feature, and it refers to the standards and criteria (symbolic, classificatory, and moral) that define the exchangeability of things in any particular social and historical context. At first glance, this feature would appear best glossed as the cultural framework within which things are classified, and it is a central preoccupation of Kopystoff's paper in this volume. Yet this gloss conceals a variety of complexities. It is true that in most stable societies, it would be possible to discover a taxonomic structure that defines the world of things, lumping some things together, discriminating between others, attaching meanings and values to these groupings, and providing a basis for rules and practices governing the circulation of these objects. In regard to the economy (that is, to exchange), Paul Bohannan's (1955) account of spheres of exchange among the Tiv is an obvious example of this type, of framework for exchange. But there are two kinds of situations where the standards and criteria that govern exchange are so attenuated as to seem virtually absent. The first is the case of transactions across cultural boundaries, where all that is agreed upon is price (whether monetary or not) and a minimum set of conventions regarding the transaction itself.⁵ The other is the case of those intracultural exchanges where, despite a vast universe of shared understandings, a specific exchange is based on deeply divergent perceptions of the value of the objects being exchanged. The best examples of such intracultural value divergence are to be found in situations of extreme hardship (such as famine or warfare), when exchanges are made whose logic has little to do with the commensuration of sacrifices. Thus a Bengali male who abandons his wife to prostitution in exchange for a meal, or a Turkana woman who sells critical pieces of her personal jewelry for a week's food, are engaging in transactions that may be seen as legitimate in extreme circumstances, but could hardly be regarded as operating under a rich shared framework of valuation between buyer and seller. Another way to characterize such situations is to say that in such contexts, value and price have come almost completely unyoked.

Also, as Simmel has pointed out, from the point of view of the individual and his subjectivity, all exchanges might contain this type of discrepancy between the sacrifices of buyer and seller, discrepancies normally brushed aside because of the host of conventions about exchange that are complied with by both parties (Simmel 1978:80). We may speak, thus, of the cultural framework that defines the commodity candidacy of things, but we must bear in mind that some exchange situations, both inter- and intracultural, are characterized by a shallower set of shared standards of value than others. I therefore

prefer to use the term regimes of value, which does not imply that every act of commodity exchange presupposes a complete cultural sharing of assumptions, but rather that the degree of value coherence may be highly variable from situation to situation, and from commodity to commodity. A regime of value, in this sense, is consistent with both very high and very low sharing of standards by the parties to a particular commodity exchange. Such regimes of value account for the constant transcendence of cultural boundaries by the flow of commodities, where culture is understood as a bounded and localized system of meanings.

Finally, the commodity context refers to the variety of social arenas, within or between cultural units, that help link the commodity candidacy of a thing to the commodity phase of its career. Thus in many societies, marriage transactions might constitute the context in which women are most intensely, and most appropriately, regarded as exchange values. Dealings with strangers might provide contexts for the commoditization of things that are otherwise protected from commoditization. Auctions accentuate the commodity dimension of objects (such as paintings) in a manner that might well be regarded as deeply inappropriate in other contexts. Bazaar settings are likely to encourage commodity flows as domestic settings may not. The variety of such contexts, within and across societies, provides the link between the social environment of the commodity and its temporal and symbolic state. As I have already suggested, the commodity context, as a social matter, may bring together actors from quite different cultural systems who share only the most minimal understandings (from the conceptual point of view) about the objects in question and agree only about the terms of trade. The so-called silent trade phenomenon is the most obvious example of the minimal fit between the cultural and social dimensions of commodity exchange (Price 1980).

Thus, commoditization lies at the complex intersection of temporal, cultural, and social factors. To the degree that some things in a society are frequently to be found in the commodity phase, to fit the requirements of commodity candidacy, and to appear in a commodity context, they are its quintessential commodities. To the degree that many or most things in a society sometimes meet these criteria, the society may be said to be highly commoditized. In modern capitalist societies, it can safely be said that more things are likely to experience a commodity phase in their own careers, more contexts to become legitimate commodity contexts, and the standards of commodity candidacy to embrace a large part of the world of things than in noncapitalist societies. Though Marx was therefore right in seeing modern indus-

trial capitalism as entailing the most intensely commoditized type of society, the comparison of societies in regard to the degree of "commoditization" would be a most complex affair given the definitional approach to commodities taken here. By this definition, the term "commodity" is used in the rest of this essay to refer to things that, at a certain phase in their careers and in a particular context, meet the requirements of commodity candidacy. Keith Hart's recent (1982) analysis of the importance of the growing hegemony of the commodity in the world would fit with the approach suggested here, except that commoditization is here regarded as a differentiated process (affecting matters of phase, context, and categorization, differentially) and the capitalist mode of commoditization is seen as interacting with myriad other indigenous social forms of commoditization.

Three additional sets of distinctions between commodities are worth making here (others appear later in this essay). The first, which is a modified application of a distinction originally made by Jacques Maquet in 1971 in regard to aesthetic productions,⁶ divides commodities into the following four types: (1) commodities by destination, that is, objects intended by their producers principally for exchange; (2) commodities by metamorphosis, things intended for other uses that are placed into the commodity state; (3) a special, sharp case of commodities by metamorphosis are commodities by diversion, objects placed into a commodity state though originally specifically protected from it; (4) ex-commodities, things retrieved, either temporarily or permanently, from the commodity state and placed in some other state. It also seems worthwhile to distinguish "singular" from "homogeneous" commodities in order to discriminate between commodities whose candidacy for the commodity state is precisely a matter of their class characteristics (a perfectly standardized steel bar, indistinguishable in practical terms from any other steel bar) and those whose candidacy are precisely their uniqueness within some class (a Manet rather than a Picasso; one Manet rather than another). Closely related, though not identical, is the distinction between primary and secondary commodities; necessities and luxuries; and what I call mobile versus enclaved commodities. Nevertheless, all efforts at defining commodities are doomed to sterility unless they illuminate commodities in motion. This is the principal aim of the section that follows.

Commodities are frequently represented as mechanical products of production regimes governed by the laws of supply and demand. By

drawing on certain ethnographic examples, I hope to show in this section that the flow of commodities in any given situation is a shifting compromise between socially regulated paths and competitively inspired diversions.

Commodities, as Igor Kopytoff points out, can usefully be regarded as having life histories. In this processual view, the commodity phase of the life history of an object does not exhaust its biography; it is culturally regulated; and its interpretation is open to individual manipulation to some degree. Further, as Kopytoff also points out, the question of what sorts of object may have what sorts of biography is more deeply a matter for social contest and individual taste in modern societies than in smaller-scale, nonmonetized, preindustrial ones. There is, in Kopytoff's model, a perennial and universal tug-of-war between the tendency of all economies to expand the jurisdiction of commoditization and of all cultures to restrict it. Individuals, in this view, can go with either tendency as it suits their interests or matches their sense of moral appropriateness, though in premodern societies the room for maneuver is usually not great. Of the many virtues of Kopytoff's model the most important, in my view, is that it proposes a general processual model of commoditization, in which objects may be moved both into and out of the commodity state. I am less comfortable with the opposition between singularization and commoditization, since some of the most interesting cases (in what Kopytoff agrees are in the middle zone of his ideal-typical contrast) involve the more or less permanent commoditizing of singularities.

Two questions can be raised about this aspect of Kopytoff's argument. One would be that the very definition of what constitutes singularities as opposed to classes is a cultural question, just as there can be unique examples of homogeneous classes (the perfect steel bar) and classes of culturally valued singularities (such as works of art and designer-label clothing). On the other hand, a Marxist critique of this contrast would suggest that it is commoditization as a worldwide historical process that determines in very important ways the shifting relationship between singular and homogeneous things at any given moment in the life of a society. But the important point is that the commodity is not one kind of thing rather than another, but one phase in the life of some things. Here, Kopytoff and I are in full agreement.

This view of commodities and commoditization has several important implications, some of which are touched upon in the course of Kopytoff's argument. Others are discussed later in this essay. But my immediate concern is with one important aspect of this temporal per-

trial capitalism as entailing the most intensely commoditized type of society, the comparison of societies in regard to the degree of "commoditization" would be a most complex affair given the definitional approach to commodities taken here. By this definition, the term "commodity" is used in the rest of this essay to refer to things that, at a certain phase in their careers and in a particular context, meet the requirements of commodity candidacy. Keith Hart's recent (1982) analysis of the importance of the growing hegemony of the commodity in the world would fit with the approach suggested here, except that commoditization is here regarded as a differentiated process (affecting matters of phase, context, and categorization, differentially) and the capitalist mode of commoditization is seen as interacting with myriad other indigenous social forms of commoditization.

Three additional sets of distinctions between commodities are worth making here (others appear later in this essay). The first, which is a modified application of a distinction originally made by Jacques Maquet in 1971 in regard to aesthetic productions,⁶ divides commodities into the following four types: (1) commodities by destination, that is, objects intended by their producers principally for exchange; (2) commodities by metamorphosis, things intended for other uses that are placed into the commodity state; (3) a special, sharp case of commodities by metamorphosis are commodities by diversion, objects placed into a commodity state though originally specifically protected from it; (4) ex-commodities, things retrieved, either temporarily or permanently, from the commodity state and placed in some other state. It also seems worthwhile to distinguish "singular" from "homogeneous" commodities in order to discriminate between commodities whose candidacy for the commodity state is precisely a matter of their class characteristics (a perfectly standardized steel bar, indistinguishable in practical terms from any other steel bar) and those whose candidacy is precisely their uniqueness within some class (a Manet rather than a Picasso; one Manet rather than another). Closely related, though not identical, is the distinction between primary and secondary commodities; necessities and luxuries; and what I call mobile versus enclaved commodities. Nevertheless, all efforts at defining commodities are doomed to sterility unless they illuminate commodities in motion. This is the principal aim of the section that follows.

Commodities are frequently represented as mechanical products of production regimes governed by the laws of supply and demand. By

drawing on certain ethnographic examples, I hope to show in this section that the flow of commodities in any given situation is a shifting compromise between socially regulated paths and competitively inspired diversions.

Commodities, as Igor Kopytoff points out, can usefully be regarded as having life histories. In this processual view, the commodity phase of the life history of an object does not exhaust its biography; it is culturally regulated; and its interpretation is open to individual manipulation to some degree. Further, as Kopytoff also points out, the question of what sorts of object may have what sorts of biography is more deeply a matter for social contest and individual taste in modern societies than in smaller-scale, nonmonetized, preindustrial ones. There is, in Kopytoff's model, a perennial and universal tug-of-war between the tendency of all economies to expand the jurisdiction of commoditization and of all cultures to restrict it. Individuals, in this view, can go with either tendency as it suits their interests or matches their sense of moral appropriateness, though in premodern societies the room for maneuver is usually not great. Of the many virtues of Kopytoff's model the most important, in my view, is that it proposes a general processual model of commoditization, in which objects may be moved both into and out of the commodity state. I am less comfortable with the opposition between singularization and commoditization, since some of the most interesting cases (in what Kopytoff agrees are in the middle zone of his ideal-typical contrast) involve the more or less permanent commoditizing of singularities.

Two questions can be raised about this aspect of Kopytoff's argument. One would be that the very definition of what constitutes singularities as opposed to classes is a cultural question, just as there can be unique examples of homogeneous classes (the perfect steel bar) and classes of culturally valued singularities (such as works of art and designer-label clothing). On the other hand, a Marxist critique of this contrast would suggest that it is commoditization as a worldwide historical process that determines in very important ways the shifting relationship between singular and homogeneous things at any given moment in the life of a society. But the important point is that the commodity is not one kind of thing rather than another, but one phase in the life of some things. Here, Kopytoff and I are in full agreement.

This view of commodities and commoditization has several important implications, some of which are touched upon in the course of Kopytoff's argument. Others are discussed later in this essay. But my immediate concern is with one important aspect of this temporal per-

spective on the commoditization of things, which concerns what I have called paths and diversions. I owe both these terms, and some measure of my understanding of the relationship between them, to Nancy Munn's contribution (Munn 1983) in an important collection of papers on a phenomenon that is of great importance to the topic of this volume, the celebrated kula system of the Western Pacific (Leach and Leach 1983).

The kula is the best-documented example of a non-Western, preindustrial, nonmonetized, translocal exchange system, and with the publication of this recent collection, it becomes, arguably, the most thoughtfully and fruitfully analyzed one. It now appears that Malinowski's classic account of this system (Malinowski 1922) was partial and problematic, though it has laid the foundation for even the most sophisticated recent analyses. The implications of this recent rethinking of the kula phenomenon for the general concerns of this volume are several. Although the essays I shall cite from this volume reflect different vantage points, both ethnographic and theoretical, they do permit some general observations.

The kula is an extremely complex regional system for the circulation of particular kinds of valuables, usually between men of substance, in the Massim group of islands off the eastern tip of New Guinea. The main objects exchanged for one another are of two types: decorated necklaces (which circulate in one direction) and armshells (which circulate in the other). These valuables acquire very specific biographies as they move from place to place and hand to hand, just as the men who exchange them gain and lose reputation as they acquire, hold, and part with these valuables. The term keda (road, route, path, or track) is used in some Massim communities to describe the journey of these valuables from island to island. But keda also has a more diffuse set of meanings, referring to the more or less stable social, political, and reciprocal links between men that constitute these paths. In the most abstract way, keda refers to the path (created through the exchange of these valuables) to wealth, power, and reputation for the men who handle these valuables (Campbell 1983a:203-4).

Keda is thus a polysemic concept, in which the circulation of objects, the making of memories and reputations, and the pursuit of social distinction through strategies of partnership all come together. The delicate and complex links between men and things that are central to the politics of the keda are captured in the following extract from the perspective of the island of Vakuta:

The successful keda consists of men who are able to maintain relatively stable keda partnerships through good oratorical and manipulative skills, and who

operate as a team, interpreting one another's movements. Nevertheless, many keda collapse, regularly making it necessary for men to realign themselves. Some form completely different keda, while the remnants of a broken keda may want to form another keda by drawing in new men. Yet others may never kula again because of their inability to form another keda owing to a reputation for "bad" kula activity. In reality, the population of shell valuables in any one keda is migratory and the social composition of a keda transitory. A shell's accumulation of history is retarded by continual movement between keda, while men's claims to immortality vanish as shells lose association with these men after being successfully attracted into another keda, thus taking on the identity of its new owners. (Campbell 1983:218-19.)

The path taken by these valuables is thus both reflective and constitutive of social partnerships and struggles for preeminence. But a number of other things are worth noting about the circulation of these valuables. The first is that their exchange is not easily categorized as simple reciprocal exchange, far from the spirit of trade and commerce. Though monetary valuations are absent, both the nature of the objects and a variety of sources of flexibility in the system make it possible to have the sort of calculated exchange that I maintain is at the heart of the exchange of commodities. These complex non-monetary modes of valuation allow partners to negotiate what Firth (following Cassady 1974) calls "exchange by private treaty," a situation in which something like price is arrived at by some negotiated process other than the impersonal forces of supply and demand (Firth 1983:91). Thus, despite the presence of broad conventional exchange rates, a complex qualitative calculus exists (Campbell 1983:245-6) which permits the competitive negotiation of personal estimates of value in the light of both short- and long-term individual interest (Firth 1983:101). What Firth here calls "indebtedness engineering" is a variety of the sort of calculated exchange that, by my definition, blurs the line between commodity exchange and other, more sentimental, varieties. The most important difference between the exchange of these commodities and the exchange of commodities in modern industrial economies is that the increment being sought in kula-type systems is in reputation, name, or fame, with the critical form of capital for producing this profit being people rather than other factors of production (Strathern 1983:80; Damon 1983:339-40). Pricelessness is a luxury few commodities can afford.

Perhaps even more important than the calculative aspect of kula exchanges is the fact that these recent studies make it very difficult to regard the exchange of kula valuables as occurring only at the boundaries between communities, with more giftlike exchanges occurring within these communities (Damon 1983:339). The concept of

In making a comparative analysis of such tournaments of value, it may be advisable not to follow Baudrillard's tendency to isolate them analytically from more mundane economic exchange, though the articulation of such value arenas with other economic arenas is likely to be highly variable. I shall have more to say on tournaments of value in the discussion of the relationship between knowledge and commodities later in this essay.

The kula, at any rate, represents a very complex system for the intercalibration of the biographies of persons and things. It shows us the difficulty of separating gift and commodity exchange even in preindustrial, nonmonetary systems, and it reminds us of the dangers in correlating zones of social intimacy too rigidly with distinct forms of exchange. But perhaps most important, it is the most intricate example of the politics of tournaments of value, in which the actors manipulate the cultural definitions of path and the strategic potential of diversion, so that the movement of things enhances their own ending.

Diversions, however, are not to be found only as parts of individual strategies in competitive situations, but can be institutionalized in various ways that remove or protect objects from the relevant social commodity contexts. Royal monopolies are perhaps the best-known examples of such "enclaved commodities," as Kopytoff points out in Chapter 2. One of the most interesting and extensive discussions of this type of monopolistic restriction on the flow of commodities is that of Max Gluckman (1983) in the context of royal property among the Lozi of Northern Rhodesia. In his discussion of the categories "gift," "tribute," and "kingly things," Gluckman shows how even in a low-surplus agricultural kingdom, the flow of commodities had very diverse and important implications. In his analysis of "kingly things," it becomes clear that the main function of these royal monopolies was to maintain sumptuary exclusivity (as in the royal monopoly of eland fly whisks), commercial advantage (as with elephant tusks), and the display of rank. Such royal restrictions of things from more promiscuous spheres of exchange is part of the way in which, in premodern chieftainships and empires, royalty could assure the material basis of sumptuary exclusivity. This type of process might be called decommodification from above.

But the more complex case concerns entire zones of activity and production that are devoted to producing objects of value that cannot be commoditized by anybody. The zone of art and ritual in small-scale societies is one such enclaved zone, where the spirit of the commodity enters only under conditions of massive cultural change. For an ex-

tended discussion of this phenomenon, we have William Davenport's essay on the production of objects for ritual use in the Eastern Solomons.

The phenomena discussed in Davenport's essay illuminate the commodity aspects of social life precisely because they illustrate one sort of moral and cosmological framework within which commoditization is restricted and hedged. In the funeral observances of this region, particularly the large-scale murina, much energy and expenditure are invested in making objects that play a central role in the ritual but are scrupulously placed in the category of "terminal" commodities (Kopytoff, Chapter 2), that is, objects which, because of the context, purpose, and meaning of their production, make only one journey from production to consumption. After that, though they are sometimes used in casual domestic ways, they are never permitted to reenter the commodity state. What makes them thus decommoditized is a complex understanding of value (in which the aesthetic, the ritual, and the social come together), and a specific ritual biography. We may paraphrase Davenport's observations and note that what happens here, at the heart of a very complex and calculated set of investments, payments, and credits, is a special kind of transvaluation, in which objects are placed beyond the culturally demarcated zone of commoditization. This type of transvaluation can take different forms in different societies, but it is typical that objects which represent aesthetic elaboration and objects that serve as sacra are, in many societies, not permitted to occupy the commodity state (either temporally, socially, or definitionally) for very long. In the rigid commitment of traditional Solomon Islanders to placing their most aestheticized ritual products beyond the reach of commoditization, we see one variation of a widespread tendency.

A somewhat different example of the tension between sacra and commodity exchange is to be seen in Patrick Geary's analysis of the trade in relics in early medieval Europe. The relics he describes are, of course, "found" and not "made," and the circulation of these relics reflects a very important aspect of the construction of community identity, local prestige, and central ecclesiastical control in Latin Europe in the early medieval period.

These relics belong to a particular economy of exchange and demand in which the life history of the particular relic is essential, not incidental, to its value. The verification of this history is also central to its value. Given the general approach to the difference between gift and commodity that I have taken in this essay, I would suggest that Geary may draw too sharp a contrast between them; indeed, his

whether it is groups or individuals who are involved in either kind of activity, the central contrast is that whereas enslaving seeks to protect certain things from commoditization, diversion frequently is aimed at drawing protected things into the zone of commoditization. Diversion, however, can also take the form of strategic shifts in path within a zone of commoditization.

In an extremely interesting discussion of British trade in Hawaii in the late eighteenth and early nineteenth centuries, Marshall Sahlins has shown how Hawaiian chiefs, in stretching traditional conceptions of tabu to cover new classes of trade goods (in keeping with their own cosmopolitical interests), succeeded in transforming the "divine finality" even of economic tabus into instruments of expedience (Sahlins 1981:44-5). Thus, what Sahlins calls "the pragmatics of trade" erodes and transforms the cultural bounds within which it is initially conceived. In a word, the politics of enslaving, far from being a guarantor of systemic stability, may constitute the Trojan horse of change.

The diversion of commodities from specified paths is always a sign of creativity or crisis, whether aesthetic or economic. Such crises may take a variety of forms: economic hardship, in all manner of societies, drives families to part with heirlooms, antiques, and memorabilia and to commoditize them. This is as true of kula valuables as of more modern valuables. The other form of crisis in which commodities are diverted from their proper paths, of course, is warfare and the plunder that historically has accompanied it. In such plunder, and the spoils that it generates, we see the inverse of trade. The transfer of commodities in warfare always has a special symbolic intensity, exemplified in the tendency to frame more mundane plunder in the transfer of special arms, insignia, or body parts belonging to the enemy. In the high-toned plunder that sets the frame for more mundane pillage, we see the hostile analogue to the dual layering of the mundane and more personalized circuits of exchange in other contexts (such as kula and gimwali in Melanesia). Theft, condemned in most human societies, is the humblest form of diversion of commodities from preordained paths.

But there are subtler examples of the diversion of commodities from their predestined paths. One whole area involves what has been dubbed tourist art, in which objects produced for aesthetic, ceremonial, or sumptuary use in small, face-to-face communities are transformed culturally, economically, and socially by the tastes, markets, and ideologies of larger economies (Graburn 1976). I shall have more to say on tourist art in the section of this essay on knowledge and commodities. Another, related area is that of the history and nature

of the major art and archeology collections of the Western world, whose formation represents extremely complex blends of plunder, sale, and inheritance, combined with the Western taste for the things of the past and of the other.¹⁰ In this traffic in artifacts, we can find today most of the critical cultural issues in the international flow of "authentic" (see Spooner, Chapter 7) and "singular" (see Kopytkoff, Chapter 2) commodities. The current controversies between English and American museums and governments and various other countries raise all the moral and political delicacies that come into play when things get diverted, several times over, from their minimal, conventional paths and are transferred by a variety of modes that make their history of claims and counterclaims extremely difficult to adjudicate.

The diversion of commodities from their customary paths always carries a risky and morally ambiguous aura. Whenever what Bohannan (1955) called conveyances give way to what he called conversions, the spirit of entrepreneurship and that of moral taint enter the picture simultaneously. In the case of the kula exchanges of Melanesia, the movement of commodities across spheres, though somehow out of order, is also at the heart of the strategy of the skillful and successful kula player. Inappropriate conversions from one sphere of exchange to another are frequently fortified by recourse to the excuse of economic crisis, whether it be famine or bankruptcy. If such excuses are not available or credible, accusations of inappropriate and venal motives are likely to set in. Excellent examples of the political implications of diversion are to be found in the arena of illegal or quasilegal commodity exchanges, one case of which is discussed next.

Lee Cassanelli's intriguing paper in this volume discusses the shift, in the last fifty years in Northeastern Africa, in the political economy of a quasilegal commodity called qat (catha edulis). Qat provides an excellent example of change in what may be referred to as a commodity ecumene,¹¹ that is, a transcultural network of relationships linking producers, distributors, and consumers of a particular commodity or set of commodities. What is particularly interesting, in this case, is the dramatic expansion of the scale of consumption (and of production) of qat which is clearly tied to changes in the technical infrastructure as well as the political economy of the region. Although the expansion of production appears consistent with conditions that fit with more universal patterns in the commercialization of agriculture, what is more intriguing is the expansion of demand and the response of the state – especially in Somalia – to the explosion in both the production and the consumption of qat.

The recent (1983) ban by the Somali government on the planting,

importing, and chewing of qat clearly is the most recent move in a long tradition of state ambivalence toward a commodity whose consumption is perceived as tied to unproductive, and potentially subversive, forms of sociality. In the case of the current Somali ban, it appears that qat (like cloth in Gandhi's rhetoric) is seen as a multilevel problem, one that challenges not only state control over the economy, but state authority over the social organization of leisure among the newly rich and upwardly mobile citizens of urban Somalia. We are again reminded, with this example, that rapid changes in consumption, if not inspired and regulated by those in power, are likely to appear threatening to them. Also, in the case of Somalia, we have a very good example of the tension between a rapid shift in the political economy of a regional commodity ecumene and the authority of one state in this ecumene.

Of course, the best examples of the diversion of commodities from their original nexus is to be found in the domain of fashion, domestic display, and collecting in the modern West. In the high-tech look inspired by the Bauhaus, the functionality of factories, warehouses, and workplaces are diverted to household aesthetics. The uniforms of various occupations are turned into the vocabulary of costume. In the logic of found art, the everyday commodity is framed and aestheticized. These are all examples of what we might call commoditization by diversion, where value, in the art or fashion market, is accelerated or enhanced by placing objects and things in unlikely contexts. It is the aesthetics of decontextualization (itself driven by the quest for novelty) that is at the heart of the display, in highbrow Western homes, of the tools and artifacts of the "other": the Turkmen saddlebag, Masai spear, Dinka basket.¹² In these objects, we see not only the equation of the authentic with the exotic everyday object, but also the aesthetics of diversion. Such diversion is not only an instrument of decommodification of the object, but also of the (potential) intensification of commoditization by the enhancement of value attendant upon its diversion. This enhancement of value through the diversion of commodities from their customary circuits underlies the plunder of enemy valuables in warfare, the purchase and display of "primitive" utilitarian objects, the framing of "found" objects, the making of collections of any sort.¹³ In all these examples, diversions of things combine the aesthetic impulse, the entrepreneurial link, and the touch of the morally shocking.

Nevertheless, diversions are meaningful only in relation to the paths from which they stray. Indeed, in looking at the social life of commodities in any given society or period, part of the anthropological

challenge is to define the relevant and customary paths, so that the logic of diversions can properly, and relationally, be understood. The relationship between paths and diversions is itself historical and dialectical, as Michael Thompson (1979) has skillfully shown in regard to art objects in the modern West. Diversions that become predictable are on their way to becoming new paths, paths that will in turn inspire new diversions or returns to old paths. These historical relationships are rapid and easy to see in our own society, but less visible in societies where such shifts are more gradual.

Change in the cultural construction of commodities is to be sought in the shifting relationship of paths to diversions in the lives of commodities. The diversion of commodities from their customary paths brings in the new. But diversion is frequently a function of irregular desires and novel demands, and we turn therefore to consider the problem of desire and demand.

Part of the reason why demand remains by and large a mystery is that we assume it has something to do with desire, on the one hand (by its nature assumed to be infinite and transcultural) and need on the other (by its nature assumed to be fixed). Following Baudrillard (1981), I suggest that we treat demand, hence consumption, as an aspect of the overall political economy of societies. Demand, that is, emerges as a function of a variety of social practices and classifications, rather than a mysterious emanation of human needs, a mechanical response to social manipulation (as in one model of the effects of advertising in our own society), or the narrowing down of a universal and voracious desire for objects to whatever happens to be available.

Alfred Gell's marvelous picture in Chapter 4 of the dilemmas of consumption among the Muria Gonds of central India makes many interesting and important points about the cultural complexities of consumption and the dilemmas of desire in small-scale societies undergoing rapid change. After reading his paper, it would be difficult to see the desire for goods as being bottomless or culture free, and demand as being a natural and mechanical response to the availability of goods and the money with which to purchase them. Consumption among the Gonds is closely tied to collective displays, economic egalitarianism, and sociability. This poses a problem for those Muria who, as a consequence of shifts in the tribal economy over the last century or so, have acquired considerably more wealth than the rest of their communities. The result is a pattern of what,

inverting Veblen, we might call "conspicuous parsimony," where simplicity in lifestyle and possessions is maintained against the growing pressures of increased income. When expenditures on commodities are made, they tend to revolve around traditionally acceptable commodity forms, such as brass pots, ceremonial finery, and houses, where collectively shared values are incarnated. This is not a world dominated by the ethos of limited good, as it might first appear, but one where there is no real interest in most of what the market has to offer. Group identity, sumptuary homogeneity, economic equality, and hedonistic sociality constitute a value framework within which most externally introduced goods are uninteresting or worrisome. The collective regulation of demand (and thus of consumption) is here part of a conscious strategy on the part of the wealthy to contain the potentially divisive implications of differentiation. The Muria example is a striking case of the social regulation of the desire for goods, even when the technical and logistical conditions for a consumer revolution have been met, as is the case with cloth in India, which is discussed next.

Christopher Bayly's contribution to this volume is an enormously subtle and suggestive analysis of the changing moral and political economy of cloth in India since 1700.¹⁴ It demonstrates very clearly the links between politics, value, and demand in the social history of things. In Bayly's argument, the production, exchange, and consumption of cloth constitute the material of a "political discourse" (rather as qat does in Somalia) that ties together royal demand, local production structures and social solidarities, and the fabric of political legitimacy. It is the consumption side of this political discourse that accounts for the deep penetration of English textiles into Indian markets in the nineteenth century, and not just the brute logics of utility and price. Finally, in the nationalist movement of the late nineteenth and early twentieth centuries, especially in Gandhi's rhetoric, the many strands of the political discourse on cloth are reconstituted and redeployed in what might be called a language of commodity resistance, in which older as well as more recent meanings of cloth are turned against the British imperium. Bayly's paper (which is, among other things, an extraordinarily rich application of the ideas of Werner Sombart), by taking the long view of the social life of a particular significant commodity, affords us two insights that are of considerable comparative interest: first, that the customary consumption logics of small communities are intimately tied to larger regimes of value defined by large-scale polities; and that the link between processes of "singularization" and "commoditization" (to use Kopytoff's terms) in

the social lives of things is itself dialectical and subject (in the hands of men like Gandhi) to what Clifford Geertz would call deep play.

Demand is thus the economic expression of the political logic of consumption and thus its basis must be sought in that logic. Taking my lead from Veblen, Douglas and Isherwood (1981), and Baudrillard (1968; 1975; 1981), I suggest that consumption is eminently social, relational, and active rather than private, atomic, or passive. Douglas has the advantage over Baudrillard of not restricting her views of consumption as communication to contemporary capitalist society but extending it to other societies as well. Baudrillard, for his part, places the logic of consumption under the dominion of the social logics of both production and exchange, equally. In addition, Baudrillard makes an immensely effective critique of Marx and his fellow political economists in regard to the twin concepts of "need" and "utility," both of which the latter saw as rooted in a primitive, universal, and natural substrate of basic human requirements.

My own inclination is to push Baudrillard's deconstruction of "need" and "utility" (and his relocation of them in the larger sphere of production and exchange) one step further and extend this idea to non-capitalist societies as well. What does this view of consumption entail? It means looking at consumption (and the demand that makes it possible) as a focus not only for sending social messages (as Douglas has proposed), but for receiving them as well. Demand thus conceals two different relationships between consumption and production: 1. On the one hand, demand is determined by social and economic forces; 2. on the other, it can manipulate, within limits, these social and economic forces. The important point is that from a historical point of view, these two aspects of demand can affect each other. Take royal demand, for example, as in Bayly's discussion of premodern India. Here royal demand is a message-sending or production-molding force, looked at from the internal point of view of eighteenth-century Indian society. That is, royal demand sets parameters for both taste and production within its relevant sphere of influence. But royal demand is also a message-receiving force, as is borne out in its relationship to contemporary European styles and products. Elite tastes, in general, have this "turnstile" function, selecting from exogenous possibilities and then providing models, as well as direct political controls, for internal tastes and production.

One mechanism that frequently translates political control into consumer demand is that of the "sumptuary laws" that characterize complex premodern societies, but also characterize small-scale, preindustrial, and preliterate societies. Wherever clothing, food, hous-

ing, body decoration, number of wives or slaves, or any other visible act of consumption is subject to external regulation, we can see that demand is subject to social definition and control. From this point of view, the plethora of "taboos" in primitive societies, which forbid particular kinds of marriage, food consumption, and interaction (as well as their cognate positive injunctions), can be seen as strict moral analogues to the more explicit, legalized sumptuary laws of more complex and literate societies. It is by virtue of this link that we can better understand the shrewd analogy that Douglas (1967) drew between "primitive" and "modern" rationing systems.

What modern money is to primitive media of exchange, fashion is to primitive sumptuary regulations. There are clear morphological similarities between the two, but the term fashion suggests high velocity, rapid turnover, the illusion of total access and high convertibility, the assumption of a democracy of consumers and of objects of consumption. Primitive media of exchange, like primitive sumptuary laws and taboos, on the other hand, seem rigid, slow to move, weak in their capacity to commensurate, tied to hierarchy, discrimination, and rank in social life. But, as Baudrillard (1981) and Bourdieu (1984) have shown so well, the establishments that control fashion and good taste in the contemporary West are no less effective in limiting social mobility, marking social rank and discrimination, and placing consumers in a game whose ever-shifting rules are determined by "taste makers" and their affiliated experts who dwell at the top of society.

Modern consumers are the victims of the velocity of fashion as surely as primitive consumers are the victims of the stability of sumptuary law. The demand for commodities is critically regulated by this variety of taste-making mechanisms, whose social origin is more clearly understood (both by consumers and by analysts) in our own society than in those distant from us. From the point of view of demand, the critical difference between modern, capitalist societies and those based on simpler forms of technology and labor is not that we have a thoroughly commoditized economy whereas theirs is one in which subsistence is dominant and commodity exchange has made only limited inroads, but rather that the consumption demands of persons in our own society are regulated by high-turnover criteria of "appropriateness" (fashion), in contrast to the less frequent shifts in more directly regulated sumptuary or customary systems. In both cases, however, demand is a socially regulated and generated impulse, not an artifact of individual whims or needs.

Even in modern, capitalist societies, of course, the media and the impulse to imitate (in Veblen's sense) are not the sole engines of

consumer demand. Demand can be manipulated by direct political appeals, whether in the special form of appeals to boycott lettuce grown in bad labor conditions or in the generalized forms of protectionism, either "official" or "unofficial." Again, Bayly's treatment of Gandhi's manipulation of the meaning of indigenously produced cloth is an arch-example of the direct politicization of demand. Yet this large-scale manipulation of the demand for cloth in twentieth-century India was possible only because cloth had long been, at the local level, an instrument for the sending of finely tuned social messages. Thus we can state as a general rule that those commodities whose consumption is most intricately tied up with critical social messages are likely to be least responsive to crude shifts in supply or price, but most responsive to political manipulation at the societal level.

From the social point of view, and over the span of human history, the critical agents for the articulation of the supply and demand of commodities have been not only rulers but, of course, traders. Philip Curtin's monumental recent work on cross-cultural trade in the preindustrial world suggests that earlier models, such as Polanyi's, of administered trade may have overstated state control over complex premodern economies (Curtin 1984:58). What is clear is that the relations between rulers and states varied enormously over space and time. Though studies like Curtin's are beginning to show patterns underlying this diversity, the demand component in these trade dynamics remains obscure. The very close historical links between rulers and traders (whether of complicity or antagonism) might partly stem from both parties being claimants for the key role in the social regulation of demand. The politics of demand frequently lies at the root of the tension between merchants and political elites; whereas merchants tend to be the social representatives of unfettered equivalence, new commodities, and strange tastes, political elites tend to be the custodians of restricted exchange, fixed commodity systems, and established tastes and sumptuary customs. This antagonism between "foreign" goods and local sumptuary (and therefore political) structures is probably the fundamental reason for the often remarked tendency of primitive societies to restrict trade to a limited set of commodities and to dealings with strangers rather than with kinsmen or friends. The notion that trade violates the spirit of the gift may in complex societies be only a vaguely related by-product of this more fundamental antagonism. In premodern societies, therefore, the demand for commodities sometimes reflects state-level dynamics, or, as in the kula case, the hinge function of status competition between elite males in linking internal and external systems of exchange.

This may be an appropriate point at which to note that there are important differences between the cultural biography and the social history of things. The differences have to do with two kinds of temporality, two forms of class identity, and two levels of social scale. The cultural biography perspective, formulated by Kopytkoff, is appropriate to specific things, as they move through different hands, contexts, and uses, thus accumulating a specific biography, or set of biographies. When we look at classes or types of thing, however, it is important to look at longer-term shifts (often in demand) and larger-scale dynamics that transcend the biographies of particular members of that class or type. Thus a particular relic may have a specific biography, but whole types of relic, and indeed the class of things called "relic" itself, may have a larger historical ebb and flow, in the course of which its meaning may shift significantly.

Colin Renfrew's paper on "Varna and the Emergence of Wealth in Europe" raises a series of important methodological as well as theoretical questions about commodities seen over the long run. His paper reminds us that commodities are central to some very early and fundamental shifts in human social life, specifically the shift from relatively undifferentiated hunter-gatherer societies to more complex early state societies. In the first place, to look at such processes over the very long run is necessarily to be involved in inferential models linking production with consumption. Second, to examine production processes in early human history entails looking at technological change. Here Renfrew shows us very persuasively that the decisive factors in technological innovation (which is critical to the development of new commodities) are often social and political rather than simply technical. Once this is seen, it follows, as Renfrew makes clear, that considerations of value and demand become central to the understanding of what look, at first glance, like strictly technical leaps.

Thus, in analyzing the role of gold and copper at Varna, and of similar objects of "prime value" in other prehistoric situations in Europe, Renfrew removes us from the temptations of the reflectionist view (where valuables simply reflect the high status of the people who use them) to a more dynamic constructionist view, in which it is the use of high technology objects that is critical to shifts in status structure. What is thus to be explained are changing notions of value, which in turn imply new uses of technological discoveries and new forms of political control of the products of such innovations. Renfrew's complex argument illustrates the point that changes in the social role of objects of display (themselves based on control over materials of prime value) illuminate long-term shifts in value and demand. At

the same time, his paper reminds us that the cultural role of commodities (though the central theme of this volume) cannot ultimately be divorced from questions of technology, production, and trade. Yet, though the archeological problem serves to highlight the complexity and historical depth of the relationship between values, social differentiation, and technical change, the absence of more conventional written or oral documents does make the reconstruction of value change more difficult than the reconstruction of social or technical change. Renfrew's paper has the virtue of going against the grain of what his evidence most comfortably supports.

Long-term processes involving the social role of commodities have recently been studied in three major treatises, two by historians (Braudel 1982; Curtin 1984), one by an anthropologist (Wolf 1982). Each of these studies has some distinctive virtues, but there are also some significant overlaps between them. Curtin's book is a bold, comparative study of what he calls "trade diasporas," communities of traders that moved goods across cultural boundaries throughout recorded history and up to the age of European industrial expansion. It strives to maintain a non-Eurocentric view of world trade before the industrial age, and in this it has much in common with Eric Wolf's aims in his recent book. Yet, Wolf's study, partly because of the theoretical viewpoint of the author and partly because of its concern with a much more recent chapter in the history of Europe's link to the rest of the world, is oriented far more to Europe. Curtin's and Wolf's studies do a great deal to explode the idea of commodity flows as either recent or exclusively tied to metropolitan capitalism, and they serve as important reminders of the institutional, logistical, and political backdrops against which commerce has occurred across social and cultural boundaries. But, for different reasons in each case, Curtin and Wolf are less interested in the question of demand and the related problem of the cultural construction of value. The essays in the present volume, then, complement and enrich the largely institutional, technological and economic panorama of commodity flows contained in these two studies.

Braudel, the formidable doyen of the Annales school, is another matter. In the second volume of his magisterial study of capitalism and material life from about 1500 to 1800 A.D., Braudel is not content to give us a dense and dramatic picture of the making of the modern industrial world. In this volume, whose English title is The Wheels of Commerce, Braudel is concerned, as are Curtin and Wolf (along, of course, with many economic and social historians) with the nature, structure, and dynamics of commerce in the world after 1500. Indeed,

taken together, these three studies present an astonishing picture of an extremely complex and interrelated set of what I have called "commodity ecumenes," which, starting around 1500 A.D., ties together many diverse parts of the world. Braudel does briefly discuss the demand side of this grand design. His argument concerning the relationship between supply and demand in the early capitalist world (Braudel 1982: 172-83), as always, sets things in a sweeping temporal perspective, but on the sources and consequences of changes in demand, he says little that was not anticipated by Werner Sombart, who is discussed below. Nevertheless, these three major recent treatments of the flow of commodities in the making of the world-system serve to highlight and provide context for what the essays in this volume seek to accomplish, and that is to illuminate the social and cultural dynamics of commodity flow. This tilt toward matters of value, career, and classification is, of course, intended to enrich our understanding of the idiosyncracies of things, a dimension to which previous scholarship has not paid much systematic attention.

The social history of things and their cultural biography are not entirely separate matters, for it is the social history of things, over large periods of time and at large social levels, that constrains the form, meaning, and structure of more short-term, specific, and intimate trajectories. It is also the case, though it is typically harder to document or predict, that many small shifts in the cultural biography of things may, over time, lead to shifts in the social history of things. Examples of these complex relations between small- and large-scale trajectories and short- and long-term patterns in the movement of things are not widespread in the literature, but we can begin to look at these relations with reference to the transformations of exchange systems under the impact of colonial rule (Dalton 1978:155-65; Strathern 1983), and to the transformations of Western society that have led to the emergence of the souvenir, the collectible, and the memento (Stewart 1984). In this volume, the essays by Bayly, Geary, Cassaneli, and Reddy are especially interesting discussions of the relationships between these two dimensions of the temporality of things. It is no coincidence that these scholars are all social historians, with an interest in long-term processes. The best general treatment of the relationship between demand, the circulation of valuables, and long-term shifts in commodity production appears in the work of Werner Sombart (Sombart 1967).

To Sombart we owe the major historical insight that in the period from approximately 1300 to 1800 in Europe, which he regards as the

nexus of early capitalism, the principal cause of the expansion of trade, industry, and finance capital was the demand for luxury goods, principally on the part of the nouveaux riches, the courts, and the aristocracy. He locates the source of this increased demand, in turn, in the new understanding of the sale of "free" love, sensual refinement, and the political economy of courtship during this period. This new source of demand meant that fashion became a driving force for the upper classes, satiated only by ever-increasing quantities and ever-differentiated qualities of articles for consumption. This intensification of demand, sexual and political in its origins, signaled the end of a seigneurial lifestyle at the same time as it stimulated nascent capitalist manufacture and trade.

Although Sombart's general approach to the social history of capitalism was, during and after his lifetime, legitimately criticized for a variety of empirical deficiencies and methodological idiosyncracies, it remains a powerful (though subterranean) alternative to both the Marxian and the Weberian views of the origins of occidental capitalism. In its focus on consumption and demand, it belongs to an oppositional and minority tradition, as Sombart was well aware. In this sense, Sombart is an early critic of what Jean Baudrillard calls the "mirror of production," in which much dominant theory of the political economy of the modern West has seen itself. In his emphasis on demand, in his key observations about the politics of fashion, in his placement of economic drives in the context of transformations of sexuality, and in his dialectical view of the relationship between luxury and necessity, Sombart anticipates recent semiotic approaches to economic behavior, such as those of Baudrillard, Bourdieu, Kristeva, and others.

Sombart's approach has recently been revived in an extremely interesting study of the cultural background of early capitalism by Chandra Mukerji (1983). Mukerji's argument, which converges at several points with my own, is that far from being a result of the industrial/technological revolution of the nineteenth century, a materialist culture and a new consumption oriented to products and goods from all over the world was the prerequisite for the technological revolution of industrial capitalism. In this bold critique of the Weberian hypothesis about the role of Puritan asceticism in providing the cultural context for capitalist calculation, Mukerji follows Nef (1958) and others. Her argument is a sophisticated historical account of the cultural backdrop of early capitalism in Europe. It provides fresh evidence and arguments for placing taste, demand and fashion at the heart of

a cultural account of the origins of occidental capitalism, and for the centrality of "things" to this ideology in Renaissance Europe (see also Goldthwaite 1983).

For our purposes, the importance of Sombart's model of the relationship between luxury and early capitalism lies less in the temporal and spatial specifics of his argument (which is a matter for historians of early modern Europe), than in the generalizability of the logic of his argument regarding the cultural basis of demand for at least some kinds of commodities, those that he calls luxuries.

I propose that we regard luxury goods not so much in contrast to necessities (a contrast filled with problems), but as goods whose principal use is rhetorical and social, goods that are simply incarnated signs. The necessity to which they respond is fundamentally political. Better still, since most luxury goods are used (though in special ways and at special cost), it might make more sense to regard luxury as a special "register" of consumption (by analogy to the linguistic model) than to regard them as a special class of thing. The signs of this register, in relation to commodities, are some or all of the following attributes: (1) restriction, either by price or by law, to elites; (2) complexity of acquisition, which may or may not be a function of real "scarcity"; (3) semiotic virtuosity, that is, the capacity to signal fairly complex social messages (as do pepper in cuisine, silk in dress, jewels in adornment, and relics in worship); (4) specialized knowledge as a prerequisite for their "appropriate" consumption, that is, regulation by fashion; and (5) a high degree of linkage of their consumption to body, person, and personality.

From the consumption point of view, aspects of this luxury register can accrue to any and all commodities to some extent, but some commodities, in certain contexts, come to exemplify the luxury register, and these can loosely be described as luxury goods. Looked at this way, all societies display some demand for luxury goods, and one could argue that it is only in Europe after 1800 (after the eclipse of the sumptuary laws), that this demand is freed from political regulation and left to the "free" play of the marketplace and of fashion. From this point of view, fashion and sumptuary regulation are opposite poles in the social regulation of demand, particularly for goods with high discriminatory value. In certain periods, the flow of luxury goods displays a powerful tension between these two pulls: the last centuries of the ancien régime in Europe, for example, show pulls in both directions. The first decades of colonial contact almost everywhere also display this tension between new fashions and existing sumptuary regulations. Fashion, in these contexts, is the urge to im-

itate the new powers, and this urge is often integrated, for better or worse, with traditional sumptuary imperatives. This tension, at the level of demand and consumption, is of course linked to the tensions between indigenous and introduced production systems and goods, and indigenous and introduced media of exchange. An extremely interesting case study of the complex links between trade, fashion, sumptuary law, and technology is Mukerji's discussion of the calico connection between England and India in the seventeenth century (Mukerji 1983:166-209).

The second important matter to which Sombart directs our attention is the complexity of the links between luxury goods and more mundane commodities. In the case with which he is concerned, the links principally involve the production process. Thus, in early modern Europe, what Sombart regards as primary luxury goods have as their prerequisites secondary and tertiary production processes: the manufacture of silk looms supports silk-weaving centers, which in turn support the creation of luxury furnishings and clothing; the sawmill produces wood that is critical to the production of fine cabinets; when timber is exhausted, coal comes to be in great demand for the glass industry and other luxury industries; iron foundries provide the pipes critical for the fountains of Versailles (Sombart 1967:145-66). To the degree that a growth in demand for primary luxury goods is critical to the expansion of production of second-order and third-order instruments, then the demand for luxuries has system-wide economic implications. Such is the case for complex early modern economies.

But in economies of different scale, structure, and industrial organization, the connection between luxury goods and goods from other registers of use may involve not the ripples of a complex set of production milieux and forms but, critically, the domains of exchange and consumption. Thus, to return to the kula systems of Oceania, recent analyses make it clear that the "trade" in kula valuables is related in a complex social and strategic dialectic with inputs from, and drains into, other exchange registers, which may involve marriage, death, and inheritance, purchase and sale, and so forth (see especially Weiner 1983).

Last, trade in luxuries may well provide an amicable, durable, and sentimental framework for the conduct of exchange in other goods and in other modes: here again the occurrence of gimwali or market-style exchange against the backdrop of kula is an apposite primitive example (Uberoi 1962). A very modern example of this type of relationship between trade in the luxury register and trade in less symbolically loaded registers is the commercial relationship between the

a cultural account of the origins of occidental capitalism, and for the centrality of "things" to this ideology in Renaissance Europe (see also Goldthwaite 1983).

For our purposes, the importance of Sombart's model of the relationship between luxury and early capitalism lies less in the temporal and spatial specifics of his argument (which is a matter for historians of early modern Europe), than in the generalizability of the logic of his argument regarding the cultural basis of demand for at least some kinds of commodities, those that he calls luxuries.

I propose that we regard luxury goods not so much in contrast to necessities (a contrast filled with problems), but as goods whose principal use is rhetorical and social, goods that are simply incarnated signs. The necessity to which they respond is fundamentally political. Better still, since most luxury goods are used (though in special ways and at special cost), it might make more sense to regard luxury as a special "register" of consumption (by analogy to the linguistic model) than to regard them as a special class of thing. The signs of this register, in relation to commodities, are some or all of the following attributes: (1) restriction, either by price or by law, to elites; (2) complexity of acquisition, which may or may not be a function of real "scarcity"; (3) semiotic virtuosity, that is, the capacity to signal fairly complex social messages (as do pepper in cuisine, silk in dress, jewels in adornment, and relics in worship); (4) specialized knowledge as a prerequisite for their "appropriate" consumption, that is, regulation by fashion; and (5) a high degree of linkage of their consumption to body, person, and personality.

From the consumption point of view, aspects of this luxury register can accrue to any and all commodities to some extent, but some commodities, in certain contexts, come to exemplify the luxury register, and these can loosely be described as luxury goods. Looked at this way, all societies display some demand for luxury goods, and one could argue that it is only in Europe after 1800 (after the eclipse of the sumptuary laws), that this demand is freed from political regulation and left to the "free" play of the marketplace and of fashion. From this point of view, fashion and sumptuary regulation are opposite poles in the social regulation of demand, particularly for goods with high discriminatory value. In certain periods, the flow of luxury goods displays a powerful tension between these two pulls: the last centuries of the ancien régime in Europe, for example, show pulls in both directions. The first decades of colonial contact almost everywhere also display this tension between new fashions and existing sumptuary regulations. Fashion, in these contexts, is the urge to im-

giate the new powers, and this urge is often integrated, for better or worse, with traditional sumptuary imperatives. This tension, at the level of demand and consumption, is of course linked to the tensions between indigenous and introduced production systems and goods, and indigenous and introduced media of exchange. An extremely interesting case study of the complex links between trade, fashion, sumptuary law, and technology is Mukerji's discussion of the calico connection between England and India in the seventeenth century (Mukerji 1983:166-209).

The second important matter to which Sombart directs our attention is the complexity of the links between luxury goods and more mundane commodities. In the case with which he is concerned, the links principally involve the production process. Thus, in early modern Europe, what Sombart regards as primary luxury goods have as their prerequisites secondary and tertiary production processes: the manufacture of silk looms supports silk-weaving centers, which in turn support the creation of luxury furnishings and clothing; the sawmill produces wood that is critical to the production of fine cabinets; when timber is exhausted, coal comes to be in great demand for the glass industry and other luxury industries; iron foundries provide the pipes critical for the fountains of Versailles (Sombart 1967:145-66). To the degree that a growth in demand for primary luxury goods is critical to the expansion of production of second-order and third-order instruments, then the demand for luxuries has system-wide economic implications. Such is the case for complex early modern economies.

But in economies of different scale, structure, and industrial organization, the connection between luxury goods and goods from other registers of use may involve not the ripples of a complex set of production milieux and forms but, critically, the domains of exchange and consumption. Thus, to return to the kula systems of Oceania, recent analyses make it clear that the "trade" in kula valuables is related in a complex social and strategic dialectic with inputs from, and drains into, other exchange registers, which may involve marriage, death, and inheritance, purchase and sale, and so forth (see especially Weiner 1983).

Last, trade in luxuries may well provide an amicable, durable, and sentimental framework for the conduct of exchange in other goods and in other modes: here again the occurrence of gimwali or market-style exchange against the backdrop of kula is an apposite primitive example (Uberoi 1962). A very modern example of this type of relationship between trade in the luxury register and trade in less symbolically loaded registers is the commercial relationship between the

United States and the USSR. Here, the strategic arms limitations talks can be seen as a highly competitive species of luxury trade, where the luxury in question is the guaranteed nuclear restraint of the opposite side. The ups and downs of this trade are the prerequisite for the movement of other commodities, such as foodgrains and high technology. It is precisely this type of politically mediated relationship between different registers of commodity trade that is aggressively exploited in the recent U.S. policy of "linkage," whereby Soviet intractability in one sphere of exchange is punished in another. In simpler times and societies, the equivalent of the SALT talks was to be seen in the diplomacy of gift exchange between traders and chiefs or simply chiefs and other chiefs, disturbances in which could abort trade in less loaded registers.

In all these ways, we can see that the demand for the kinds of valuables we call luxuries and what I have called the luxury register of any particular flow of commodities is intimately connected with other, more everyday, high-turnover registers in the language of commodities in social life.

This may also be the appropriate juncture at which to make a general point about the commodities dealt with in this volume, many of which have a strong luxury dimension and thus appear to constitute a sample that is bound to favor a cultural approach in a way that humbler, more mass-produced commodities might not. The fact is that the line between luxury and everyday commodities is not only a historically shifting one, but even at any given point in time what looks like a homogeneous, bulk item of extremely limited semantic range can become very different in the course of distribution and consumption. Perhaps the best example of a humble commodity whose history is filled with cultural idiosyncracies is sugar, as is shown in very different ways by Sidney Mintz (1979) and Fernand Braudel (1982: 190-4). The distinction between humble commodities and more exotic ones is thus not a difference in kind, but most often a difference in demand over time or, sometimes, a difference between loci of production and those of consumption. From the point of view of scale, style, and economic significance, Mukerji has made an eloquent argument, at least in the case of early modern Europe, for not drawing rigid boundaries between elite and mass consumption, luxury goods and humbler ones, consumer and capital goods, or the aesthetics of display as against the designs of primary production settings (Mukerji 1983: Chapter 1).

Demand is thus neither a mechanical response to the structure and level of production nor a bottomless natural appetite. It is a complex

social mechanism that mediates between short- and long-term patterns of commodity circulation. Short-term strategies of diversion (such as those discussed in the previous section) might entail small shifts in demand that can gradually transform commodity flows in the long run. Looked at from the point of view of the reproduction of patterns of commodity flow (rather than their alteration), however, long-established patterns of demand act as constraints on any given set of commodity paths. One reason such paths are inherently shaky, especially when they involve transcultural flows of commodities, is that they rest on unstable distributions of knowledge, a subject to which we now turn.

This section is concerned with the peculiarities of knowledge that accompany relatively complex, long-distance, intercultural flows of commodities, though even in more homogeneous, small-scale, and low-technology loci of commodity flow, there is always the potential for discrepancies in knowledge about commodities. But as distances increase, so the negotiation of the tension between knowledge and ignorance becomes itself a critical determinant of the flow of commodities.

Commodities represent very complex social forms and distributions of knowledge. In the first place, and crudely, such knowledge can be of two sorts: the knowledge (technical, social, aesthetic, and so forth) that goes into the production of the commodity; and the knowledge that goes into appropriately consuming the commodity. The production knowledge that is read into a commodity is quite different from the consumption knowledge that is read from the commodity. Of course, these two readings will diverge proportionately as the social, spatial, and temporal distance between producers and consumers increases. As we shall see, it may not be accurate to regard knowledge at the production locus of a commodity as exclusively technical or empirical and knowledge at the consumption end as exclusively evaluative or ideological. Knowledge at both poles has technical, mythological, and evaluative components, and the two poles are susceptible to mutual and dialectical interaction.

If we regard some commodities as having "life histories" or "careers" in a meaningful sense, then it becomes useful to look at the distribution of knowledge at various points in their careers. Such careers have the greatest uniformity at the production pole, for it is likely that at the moment of production, the commodity in question has had the least

opportunity to accumulate an idiosyncratic biography or enjoy a peculiar career. Thus the production locus of commodities is likely to be dominated by culturally standardized recipes for fabrication. Thus factories, fields, forges, mines, workshops, and most other production loci are repositories, in the first place, of technical production knowledge of a highly standardized sort. Nevertheless, even here it is worth noting that the technical knowledge required for the production of primary commodities (grains, metals, fuels, oils) is much more likely to be standardized than the knowledge required for secondary or luxury commodities, where taste, judgment, and individual experience are likely to create sharp variations in production knowledge. Nevertheless, the thrust of commoditization at the production end is toward standardization of technical (how-to) knowledge. Of course, with all commodities, whether primary or not, technical knowledge is always deeply interpenetrated with cosmological, sociological, and ritual assumptions that are likely to be widely shared. Evans-Pritchard's Azande potters (Evans-Pritchard 1937), Taussig's Colombian peasant producers (Taussig 1980), Nancy Munn's Gawan canoe makers (Munn 1977), Stephen Gudeman's Panamanian sugarcane producers (Gudeman 1984), all combine technological and cosmological layers in their production discourse. In most societies, such production knowledge is subject to some discontinuity in its social distribution, either by simple criteria of age or gender, by more complex criteria distinguishing artisan households, castes, or villages from the rest of society, or by even more complex divisions of labor setting apart entrepreneurs and workers, in role terms, from householders and consumers, as in most modern societies.

But there is another dimension of production knowledge and that is knowledge of the market, the consumer, the destination of the commodity. In small-scale, traditional societies, such knowledge is relatively direct and complete as regards internal consumption, but more erratic and incomplete as regards external demand. In precapitalist contexts, of course, the translation of external demands to local producers is the province of the trader and his agents, who provide logistical and price bridges between worlds of knowledge that may have minimal direct contact. Thus it is reasonably certain that traditional Borneo forest dwellers had relatively little idea of the uses to which the birds' nests they sold to intermediaries have played in Chinese medical and culinary practice. This paradigm of merchant bridges across large gaps in knowledge between producer and consumer characterizes the movement of most commodities throughout history, up to the present. Today, these bridges persist either because of unclos-

able cultural gaps (as between opium producers in Asia and the Middle East and addicts and dealers in New York) or because of the infinitesimal specialization of commodity production or its inverse – the distance between a particular bulk commodity (such as, say, copper) and the hundreds of transformations it will undergo before reaching the consumer. We note that such large gaps in knowledge of the ultimate market by the producer are usually conducive to high profits in trade and to the relative deprivation of the producing country or class in relation to the consumers and the trader (see Spooner, Chapter 7).

Problems involving knowledge, information, and ignorance are not restricted to the production and consumption poles of the careers of commodities, but characterize the process of circulation and exchange itself. In a powerful cultural account of the Moroccan bazaar, Clifford Geertz has placed the search for reliable information at the heart of this institution and has shown how difficult it is for actors in this system to gain reliable information either about people or about things (Geertz 1979). Much of the institutional structure and cultural form of the bazaar is double-edged, making reliable knowledge hard to get and also facilitating the search for it. It is tempting to conclude that such complex and culturally organized information mazes are a special feature of bazaar-style economies, and are absent in nonmarket, simple economies, as well as in advanced industrial ones. Yet, as Geertz himself suggests (p. 224) the bazaar as an analytical category may well apply to the used-car market (though not the new-car market) in contemporary industrial economies. We can put this point in a more general form: bazaar-style information searches are likely to characterize any exchange setting where the quality and the appropriate valuation of goods are not standardized, though the reasons for the lack of standardization, for the volatility of prices, and for the unreliable quality of specific things of a certain type may vary enormously. Indeed, systems for the exchange of kula valuables, of used cars, and of oriental rugs, though they occur in very different institutional and cultural settings, may all involve bazaar-style information economies.

But the gaps in knowledge and the difficulties of communication between producer and consumer are not really obstacles to the vigorous flow of bulk commodities intended for multiple industrial transformations before they reach the consumer. In the case of such commodities (sometimes called primary commodities), an almost infinite series of small, overlapping circles of knowledge can link original producer and terminal consumer. But this is not the case with com-

modities by destination, which are largely "fabricated," in Nancy Munn's sense, early in their careers (Munn 1977). These require more direct mechanisms for the satisfactory negotiation of price and the matching of consumer taste to producer skill, knowledge, and tradition. Perhaps the best examples of this kind of more direct communication involve the international commerce in ready-made clothes (Swallow 1982) and the tourist art trade in what Nelson Graburn (1976) has called the fourth world.

Whenever there are discontinuities in the knowledge that accompanies the movement of commodities, problems involving authenticity and expertise enter the picture. Several of the papers in this volume deal with these two issues. The first is Brian Spooner's paper on oriental carpets, which is a provocative anthropological interpretation of a problem that brings together art history, economic history, and cultural analysis. Spooner's topic – the shifting terms of the relationship between producers and consumers of oriental carpets – brings into focus a particularly striking example of a commodity linking two largely isolated worlds of meaning and function. Traded originally through a series of Asian and European entrepôts, each of which imposed economic and taste filters, today oriental carpets involve a much more direct negotiation between Western upper-middle-class tastes and Central Asian weaving organizations. But this shift involves not simply changes in the context of the negotiation of price. What is being negotiated, as Spooner pithily puts it, is authenticity. That is, as the pace of mobility and the crowding at the top of Western society become more marked, and as technology permits the multiplication of prestige objects, there is an increasingly ironic dialogue between the need for ever-shifting criteria of authenticity in the West and the economic motives of the producers and dealers. The world of dealers, further, becomes itself tied up with the politics of connoisseurship and the formalization of rug lore in the West.

In a general way, we can suggest that with luxury commodities like oriental rugs, as the distance between consumers and producers is shrunk, so the issue of exclusivity gives way to the issue of authenticity. That is, under premodern conditions, the long-distance movement of precious commodities entailed costs that made the acquisition of them in itself a marker of exclusivity and an instrument of sumptuary distinction. Where the control of such objects was not directly subject to state regulation, it was indirectly regulated by the cost of acquisition, so that they stayed within the hands of the few. As technology changes, the reproduction of these objects on a mass basis becomes possible, the dialogue between consumers and the original source becomes

more direct, and middle-class consumers become capable (legally and economically) of vying for these objects. The only way to preserve the function of these commodities in the prestige economies of the modern West is to complicate the criteria of authenticity. The very complicated competition and collaboration between "experts" from the art world, dealers, producers, scholars, and consumers is part of the political economy of taste in the contemporary West. This political economy has perhaps best been explored in France, by Baudrillard (1981) and Bourdieu (1984).

There is a particular set of issues concerning authenticity and expertise that plagues the modern West, and this set, which revolves around the issues of good taste, expert knowledge, "originality," and social distinction, is especially visible in the domain of art and art objects. In his famous essay on "The Work of Art in the Age of Mechanical Reproduction," Walter Benjamin (1968; original edition, 1936) recognized that the aura of an authentic work of art is tied up with its originality, and that this aura, which is the basis of its authenticity, is jeopardized by modern reproductive technologies. In this sense copies, forgeries, and fakes, which have a long history, do not threaten the aura of the original but seek to partake of it. In a footnote to this essay, Benjamin made the following shrewd observation: "To be sure, at the time of its origin a medieval picture of the Madonna could not yet be said to be 'authentic.' It became 'authentic' only during the succeeding centuries and perhaps most strikingly so during the last one." (Benjamin 1936:243.) In an essay on the concept of the "signature" in the modern art world, Baudrillard (1981:103) pushes this point further:

Until the nineteenth century, the copy of an original work had its own value, it was a legitimate practice. In our own time the copy is illegitimate, inauthentic: it is no longer "art." Similarly, the concept of forgery has changed – or rather, it suddenly appears with the advent of modernity. Formerly painters regularly used collaborators or "negros": one specialized in trees, another in animals. The act of painting, and so the signature as well, did not bear the same mythological insistence upon authenticity – that moral imperative to which modern art is dedicated and by which it becomes modern – which has been evident ever since the relation to illustration and hence the very meaning of the artistic object changed with the act of painting itself.

With this in mind, it is possible to place the consumption side of the processes that Spooner observes in the context of what Baudrillard sees as the emergence of the "object," that is, a thing that is no longer just a product or a commodity, but essentially a sign in a system of signs of status. Objects, in Baudrillard's view, emerge fully only in

this century in the modern West, in the context of the theoretical formulations of the Bauhaus (Baudrillard 1981:185), though it has recently been shown that the emergence of the object in European culture can be traced back at least to the Renaissance (Mukerji 1983). Fashion is the cultural medium in which objects, in Baudrillard's sense, move.

Yet problems of authenticity, expertise, and the evaluation of commodities are obviously not only twentieth-century phenomena. We have already mentioned Patrick Geary's paper in this volume, on the trade in relics in Carolingian Europe. Here there is a crucial problem with regard to authentication, and here too it is tied to the fact that relics circulate over long periods of time, through many hands, and over large distances. Here too there is a concern with fakery, an obsession with origins. But the cultural regime for authentication is quite different from the modern one. Though there is a small body of technical procedures and clerical prerogatives involved in authentication, it is by and large a matter in which popular understandings about ritual efficacy and folk criteria of authenticity play a central role. Authenticity here is not the province of experts and esoteric criteria, but of popular and public kinds of verification and confirmation.

The problem of specialized knowledge and of authenticity takes yet another form in William Reddy's fascinating case study of the shifts in the organization of expert knowledge in the textile industry in France before and after the Revolution of 1789. Focusing on two commercial dictionaries published in France, in the 1720's and in 1839, Reddy argues that though the French Revolution appeared to destroy a whole way of life overnight, this was not in fact the case. The vast edifice of everyday knowledge and practice changed slowly, uncertainly, and reluctantly. One example of this extended crisis – a period, that is, when knowledge, practice, and policy were notably out of step – was to be seen in the codified world of knowledge regarding the trade in textiles. In complex early modern systems of commodity flow, Reddy shows us, the relationship between technical knowledge, taste, and political regulation are very complex and slow to change. Ways of knowing, judging, trading, and buying are harder to change than ideologies about guilds, prices, or production. It took a very complex series of piecemeal and asynchronous shifts in politics, technology, and culture, stretching over a century, before a new epistemological framework emerged for classifying commercial products. In this new scheme, we might say that goods were reconceived as products, and the "gaze" (in Foucault's sense) of the consumer and the

trader had given way to the "gaze" of the producer. Textiles, in the first third of the nineteenth century, came to be seen in what Baudrillard calls the "mirror of production." Authenticity, in this early industrial framework, is no longer a matter of connoisseurship, but of objectively given production methods. The expertise of the dealer and the financier gives way to the expertise of industrialized production. Reddy's essay reminds us that the social history of things, even of humble things like cloth, reflects very complicated shifts in the organization of knowledge and modes of production. Such shifts have a cultural dimension that cannot be deduced from, or reduced to, changes in technology and economy.

One final example of the very complex relationship between authenticity, taste, and the politics of consumer – producer relations concerns what have been called ethnic or tourist arts. These have been subject to fairly close study by anthropologists, and there is one important collection of essays on the subject (Graburn 1976). Though the phenomena discussed under these labels include a bewildering range of objects, as Graburn notes in his introductory essay, they constitute perhaps the best example of the diversities in taste, understanding, and use between producers and consumers. At the producer end, one sees traditions of fabrication (again, following Munn), changing in response to commercial and aesthetic impositions or temptations from larger-scale, and sometimes far-away consumers. At the other end, one has souvenirs, mementos, curios, collections, exhibits, and the status contests, expertise, and commerce on which they rest. In between one has a series of commercial and aesthetic links, sometimes complex, multiple, and indirect and sometimes overt, few, and direct. In both cases, tourist art constitutes a special commodity traffic, in which the group identities of producers are tokens for the status politics of consumers.

Alfred Gell's paper in this volume contains some particularly astute observations on the kinds of complicated refractions in perception that can accompany the interaction of small traditional populations with larger-scale economies and cultural systems. Reflecting on the Muria interest in brassware produced from outside their region, Gell notes that "the Muria, a traditional people with no home-grown tradition of craft and prestige-good production, are actually much more similar to Westerners, seeking authenticity in the exotic, than they are to traditional craft-producer societies, the category to which they are erroneously believed to belong." Recent work on exhibitions and museums by anthropologists and historians (Benedict 1983; Breckenridge 1984), as well as by semioticians and literary theorists, extends

and deepens our understanding of the role of objects of the "other" in creating the souvenir, the collection, the exhibit and the trophy in the modern West (Baudrillard 1968, 1981; Stewart 1984).¹⁰ In a more general way, it might be said that as the institutional and spatial journeys of commodities grow more complex, and the alienation of producers, traders, and consumers from one another increases, culturally formed mythologies about commodity flow are likely to emerge.

Culturally constructed stories and ideologies about commodity flows are commonplace in all societies. But such stories acquire especially intense, new, and striking qualities when the spatial, cognitive, or institutional distances between production, distribution, and consumption are great. Such distancing either can be institutionalized within a single complex economy or can be a function of new kinds of links between hitherto separated societies and economies. The institutionalized divorce (in knowledge, interest, and role) between persons involved in various aspects of the flow of commodities generates specialized mythologies. I consider, in this section, three variations on such mythologies and the contexts in which they arise. (1) Mythologies produced by traders and speculators who are largely indifferent to both the production origins and the consumption destination of commodities, except insofar as they affect fluctuations in price. The best examples of this type are the commodity futures markets in complex capitalist economies, specifically the Chicago grain exchange in the early part of this century. (2) Mythologies produced by consumers (or potential consumers) alienated from the production and distribution process of key commodities. Here the best examples come from the cargo cults of Oceania. And (3) mythologies produced by workers in the production process who are completely divorced from the distribution and consumption logics of the commodities they produce. The modern tin miners of Bolivia described by Michael Taussig in The Devil and Commodity Fetishism in South America, are an excellent case in point. In the following paragraphs, I briefly discuss each of these variations, starting with the capitalist commodity markets.

The commodity sphere in the modern capitalist world-system appears at first glance to be a vast, impersonal machine, governed by large-scale movements of prices, complex institutional interests, and a totally demystified, bureaucratic, and self-regulating character. Nothing, it appears, could be further from the values, mechanisms, and ethics of commodity flows in small-scale societies. Yet this impression is false.

It should by now be clear that capitalism represents not simply a techno-economic design, but a complex cultural system with a very

special history in the modern West. This view, which has always had distinguished adherents in economic and social history (Weber 1958; Sombart 1967; Nef 1958; Braudel 1982; Lopez 1971; Thirsk 1978), has received a new impetus from anthropologists and sociologists of Euro-American culture (Baudrillard 1981; Bourdieu 1984; Douglas and Isherwood 1981; Mukerji 1983; Sahlins 1976).

The study of the cultural design of capitalism in its American form has been undertaken with enormous vigor in the last decade, and historians, anthropologists, and sociologists are beginning to put together a rich picture of the culture of capitalism in the United States (Collins 1979; DiMaggio 1982; Lears 1984; Marcus (in press); Schudson 1984). Though this larger context lies outside the scope of this discussion, it is quite clear that capitalism is itself an extremely complex cultural and historical formation, and in this formation commodities and their meanings have played a critical role. One example of the peculiar and striking cultural expressions of modern capitalism is the market in commodity futures in the United States, which developed in the middle of the nineteenth century and whose paradigmatic example is the Chicago Grain Exchange.

Trade in bulk commodities remains today an extremely important part of world trade and the world economic system (see, for example, Adams and Behrman 1982), and this large-scale commodity trade remains perhaps the central arena where the contradictions of international capitalism can be observed. Central among these contradictions is the one between the free-trade ideology of classical capitalism and the various forms of protectionism, cartels, and regulatory agreements that have evolved to restrict this freedom in the interests of various coalitions of producers (Nappi 1979). Commodity futures markets represent the institutional arena where the risks that attend the national and international flows of these commodities are negotiated by hedging on the part of some and sheer speculation on the part of others.

Markets in commodity futures revolve around a large number of transactions involving contracts to buy and sell commodities, at future dates. This trade in contracts is a paper trade, which rarely involves actual exchanges of the commodities themselves between traders. Like the stock market, these markets are speculative tournaments, in which the play of price, risk, and exchange appears totally divorced, for the spectator, from the entire process of production, distribution, sale, and consumption. One might say that speculating in commodity futures makes a dramatic separation between price and value, with the latter of no concern at all. In this sense, the logic of trade in

commodity futures is, following Marx, a kind of meta-fetishization, where not only does the commodity become a substitute for the social relations that lie behind it, but the movement of prices becomes an autonomous substitute for the flow of the commodities themselves.

Though this double degree of removal from the social relations of production and exchange makes commodity futures markets very different from other tournaments of value, such as these represented in the kula, there are some interesting and revealing parallels. In both cases, the tournament occurs in a special arena, insulated from practical economic life and subject to special rules. In both cases, what are exchanged are tokens of value that can be transformed into other media only by a complex set of steps and only in unusual circumstances. In both cases, there are specific ways in which the reproduction of the larger economy is articulated with the structure of the tournament economy.

But perhaps most important, in both cases, there is an agonistic, romantic, individualistic, and gamelike ethos that stands in contrast to the ethos of everyday economic behavior. The role of kula participation in the construction of fame and reputation for individuals in island Oceania is very clear. But the same is the case with commodity futures markets. In the second half of the nineteenth century, the "wheat pit" (the Grain Exchange) in Chicago was obviously the scene of the making and breaking of individual reputations, of intense and obsessive competitions between specific individuals, and of hubristic efforts on the part of particular men to corner the market (Dies 1925; 1975). This agonistic, obsessive, and romantic ethos has not disappeared from the commodity markets, as we are reminded by the case of the Hunt brothers in regard to silver (Marcus: in press), although the moral, institutional, and political framework that governs speculation in commodities has changed a good deal since the nineteenth century. Of course, there are many differences between the kula and the commodity futures market in scale, instrumentalities, context, and goals. But the similarities are real, and, as I suggested earlier, many societies create specialized arenas for tournaments of value in which specialized commodity tokens are traded, and such trade, through the economies of status, power, or wealth, affects more mundane commodity flows. The trade in relics, the market in commodity futures, the kula, the potlatch, and the Central Asian buzkashi (Azoy 1982) are all examples of such "tournaments of value." In each case, we need a fuller examination of the modes of articulation of these "tournament" economies with their more routine commodity contexts than is possible here.

The mythology of circulation generated in commodity markets (as well as, in other ways, in stock markets) is a mythology of rumor mixed with more reliable information: regarding commodity reserves, government regulations, seasonal shifts, consumer variables, intra-market developments (including the rumored intention or motives of other speculators), and so on. These constitute an endlessly shifting (and potentially infinite) scenario of variables that affect price. Though there have been consistent improvements in the technical basis for analyzing and successfully playing the commodities market, there remains the quasi-magical search for the formula (divinatory rather than efficacious) that will prove to be the fail-safe predictor of price shifts (Powers 1973:47). The structural basis of this mythology of circulation of commodities is the fact that it plays indefinitely with the fluctuation of prices; that it seeks to exhaust an inexhaustible series of variables that affect price; and that its concern with commodities is purely informational and semiotic and is divorced from consumption altogether. The irrational desire to corner the market in some commodity, the counterintuitive search for magical formulas to predict price changes, the controlled collective hysteria, all these are the product of this complete conversion of commodities to signs (Baudrillard 1981), which are themselves capable of yielding profit if manipulated properly. The primitive counterpart to this type of mythological and context-free construction of commodities is to be found in that anthropological staple, the cargo cults that multiplied in the stateless societies of the Pacific in this century.

Cargo cults are social movements of intense, millennial character centered on the symbolism of European goods. They have occurred mainly in the Pacific since early colonial contact, though they have precolonial antecedents and analogies in other societies. They have been subject to intensive analysis by anthropologists, who have looked at them as psychological, religious, economic, and political phenomena. Though there has been considerable variation in the anthropological interpretation of these movements, most observers agree that the emergence of cargo cults in early colonial Pacific societies has something to do with the transformation of production relations in this new context, the inability of natives to afford the new European goods they desired; the arrival of a new theological and cosmological system through the missionaries; and the resulting ambivalence toward indigenous ritual forms. The result was a series of movements, spread throughout Oceania (and later Melanesia) of uneven success, duration, and intensity, which both mimicked and protested European social and ritual forms and took either strongly oppositional or

strongly revivalistic positions in regard to their own myths and rituals of prosperity and exchange. In the symbolism of many of these movements, an important role was played by the promise by the leader/prophet of the arrival of valued European goods by plane or by ship and their "showering" upon the true believers in the movement and in the prophet.

It is difficult to doubt the contention of Worsley (1957) and others that the symbolism of the mysterious arrival of European goods has a lot to do with the distortion of indigenous exchange relations under colonial rule, the perception by the natives of the apparent contradiction between the wealth of Europeans (despite their lack of effort) and their own poverty (despite their arduous labor). It is no surprise, given their sudden subjection to a complex international economical system of which they saw only few and mysterious aspects, that their response was occasionally to seek to replicate what they regarded as the magical mode of production of these goods.

When we look at the symbolism and ritual practice of these movements, it is possible to see that they constitute not just a myth about the origins of European commodities, but an attempt to ritually replicate what were perceived as the social modalities of European life. This is the significance of the use of European military forms, speech forms, titles, and so forth, in these movements. Though often ordered in indigenous patterns, the ritual practice of cargo cults was in many cases no less than a massive effort to mime those European social forms that seemed most conducive to the production of European goods. In a kind of reverse fetishism, what was replicated was what was seen as the most potent of European social and linguistic forms in an effort to increase the likelihood of the arrival of European commodities. But Glynn Cochrane (1970) has reminded us that these cults were, however distorted, pursuits not of all European commodities, but only of those commodities that were seen as particularly conducive to the maintenance of status discontinuities in local societies. Cargo cults also represent a particular mythology of production of European finished goods by natives embroiled in the production of primary commodities for the world trade and an associated imitative and revitalistic ritual. The commodities involved in cargo, as with kula valuables, and other indigenous forms of specialized exchange, are seen as metonymic of a whole system of power, prosperity, and status. Cargo beliefs are an extreme example of the theories that are likely to proliferate when consumers are kept completely ignorant of the conditions of production and distribution of commodities and are unable to gain access to them freely. Such deprivation creates the

mythologies of the alienated consumer, just as the commodity markets of modern capitalism spawn the mythologies of the alienated trader. We turn, finally, to the third variation, the mythologies of producers at the service of demand and distribution forces outside their control and beyond their universe of knowledge.

For this type of mythology, the best account we have is Taussig's analysis of the changing symbolism of the Devil among Bolivian tin miners since the arrival of the Spaniards (Taussig 1980). Briefly, the story runs as follows. Before the arrival of the Spaniards, mining was a small-scale activity run as a state monopoly. With the arrival of the Spanish, mining became a voracious keystone of the colonial economy, the cause of massive dislocation and increased mortality among the Aymara Indian population of Bolivia. Mining always involved ritual and magic, but only after the Spanish conquest did this involve the spirit of evil, symbolized in a figure called Tio (uncle), understood in the new Christian idiom as the Devil, who was seen as the spirit owner of the mines. This devil figure came to represent all the alien forces of the new capitalist economy, which miners simultaneously feared, hated, and served, in contrast to their traditional forms of reciprocal economy. Caught between state control of production and the international commodity market, on the one hand, and the Devil on the other, they worked out a ritual that reflects the ambiguities and contradictions of an economic practice that straddles two incompatible worlds:

In effect the extended chain of exchanges in the Andes is this: peasants exchange gifts with the spirit owner; the spirit owner converts these gifts into precious metal; the miners excavate this metal, which they "find" so long as they perform rites of gift exchange with spirit; the miners' labor, which is embodied in the tin ore, is sold as a commodity to the legal owners and employers; these last sell the ore on the international commodity market. Thus, reciprocal gift exchanges end as commodity exchanges; standing between the devil and the state, the miners mediate this transformation. This circuit ensures barrenness and death instead of fertility and prosperity. It is based on the transformation of reciprocity into commodity exchange. (Taussig 1980:224).

The rites of production in the tin mines of Bolivia and their associated mythology are not a simple carryover of peasant rites of production. They reflect the tensions of a society in which commoditization has not yet become commonplace, where the fetishism of commodities, because of its incomplete hegemony, is regarded as evil and dangerous, and there is thus a paradoxical attempt to envelop the Devil in reciprocal rituals. This is not commodity fetishism in the

classic Marxian sense (where products conceal and represent social relations), but a more literal fetishism, in which the commodity, itself iconicized as the Devil, is made the pivot of a set of ritual transactions designed to offset the cosmological and physical risks of mining. In this mythology of alienated producers/extractors, the impersonal and invisible sources of control (the state) and of demand (the world commodity market) are relocated in an icon of danger and greed, social metaphors for the commodity economy. Though Taussig's account tends, like Gregory's and many others, to overstate the contrast between gift and commodity economies, his is a persuasive account of the literal fetishism of commodities that seems to accompany primary commodity production for unknown and uncontrolled markets.

In each of the examples I have discussed, the commodity futures market, cargo cults, and mining mythology, mythological understandings of the circulation of commodities are generated because of the detachment, indifference, or ignorance of participants as regards all but a single aspect of the economic trajectory of the commodity. Enclaved in either the production, speculative trade, or consumption locus of the flow of commodities, technical knowledge tends to be quickly subordinated to more idiosyncratic subcultural theories about the origins and destinations of things. These are examples of the many forms that the fetishism of commodities can take when there are sharp discontinuities in the distribution of knowledge concerning their trajectories of circulation.

There is one final point to be made about the relationship between knowledge and commodities, and it is one which reminds us that the comparison of capitalistic societies with other kinds of societies is a complicated matter. In complex capitalistic societies, it is not only the case that knowledge is segmented (even fragmented) as between producers, distributors, speculators, and consumers (and different sub-categories of each). The fact is that knowledge about commodities is itself increasingly commoditized. Such commoditization of knowledge regarding commodities is of course part of a larger problem of the political economy of culture itself (Collins 1979), in which expertise, credentialism, and high-brow aestheticism (Bourdieu 1984) all play different roles. Thus, though even in the simplest economies there is a complex traffic in things, it is only with increased social, technical, and conceptual differentiation that what we may call a traffic in criteria concerning things develops. That is, only in the latter situation does the buying and selling of expertise regarding the technical, social, or aesthetic appropriateness of commodities become widespread. Of course, such a traffic in commodity criteria is not confined to capitalist

societies, but there seems to be considerable evidence that it is in such societies that such traffic is most dense.

In contemporary capitalist economies, further, it is difficult to separate the commoditization of goods from the commoditization of services. Indeed the routine pairing of goods and services is itself a heritage of neoclassical economics. This is not to say that services (sexual, occupational, ritual, or emotional) lie wholly outside the domain of commoditization in noncapitalist societies. But it is only in complex postindustrial economies that services are a dominant, even definitive, feature of the world of commodity exchange. A thorough comparative analysis of the service dimension of commoditization, however, is something that a collection such as this one can only hope to stimulate.

But perhaps the best example of the relationship between knowledge and the control of demand is provided by the role of advertising in contemporary capitalist societies. Much has been written about this important topic, and in the United States there are signs of a revived debate about the functional effectiveness of advertising. In a widely publicized recent study, Michael Schudson (1984) has questioned the neo-Marxist analyses of the manipulation of consumers by advertising in America. He proposes that the textual and graphic images produced by the advertising machine are better regarded as a species of "capitalist realism," a form of cultural representation of the virtues of the capitalist lifestyle, rather than as techniques for seduction into specific acts of consumption. The adulation with which this argument has been greeted by the advertising profession is a source of some circumstantial doubt about the argument itself. What is probably the case is that any decisive analysis of the effects of advertising would have to proceed to see the images of advertising in tandem with changing ideas about art, design, lifestyle, and distinction, in order to unravel the role of this kind of "capitalist realism" in the social mobilization of demand (Hebdige 1983; Bourdieu 1984).

But it does seem worthwhile to make one observation about advertising that is relevant to the present argument. Whatever the effectiveness of advertising in ensuring the success of any particular product, it does seem true that contemporary modes of representation in advertising (particularly on television) share a certain strategy. The strategy consists in taking what are often perfectly ordinary, mass-produced, cheap, even shoddy, products and making them seem somehow (in Simmel's sense) desirable-yet-reachable. Perfectly ordinary goods are placed in a sort of pseudoenclaved zone, as if they were not available to anyone who can pay the price. The largely social images that create this illusion of exclusivity might be glossed as the

fetishism of the consumer rather than of the commodity. The images of sociality (belonging, sex appeal, power, distinction, health, togetherness, camaraderie) that underly much advertising focus on the transformation of the consumer to the point where the particular commodity being sold is almost an afterthought. This double inversion of the relationship between people and things might be regarded as the critical cultural move of advanced capitalism.

The relationship between knowledge and commodities has many dimensions that have not been discussed here. But the essential point for my purposes is this: as commodities travel greater distances (institutional, spatial, temporal), knowledge about them tends to become partial, contradictory, and differentiated. But such differentiation may itself (through the mechanisms of tournaments of value, authentication, or frustrated desire) lead to the intensification of demand. If we look at the world of commodities as a shifting series of local (culturally regulated) commodity paths, we can see that the politics of diversion as well as of enclosing often is tied to the possibility or fact of commodity exchanges with other, more distant, systems. At every level where a smaller system interacts with a larger one, the interplay of knowledge and ignorance serves as a turnstile, facilitating the flow of some things and hindering the movement of others. In this sense, even the largest commodity ecumenes are the product of complex interactions between local, politically mediated, systems of demand.