The past two decades have seen growing interest in the study of consumption in anthropology. The resulting work reflects a culturalist orientation, apparent in the classic texts that helped spur that growth and that have to some degree defined what counts as anthropology of consumption. This orientation construes people, objects and consumption in particular ways, and has led to interesting and worthwhile work. However, the spread of this orientation has meant foregoing other ways of seeing the place of consumption in social life. This paper draws on selected anthropological work that deals with consumption but that does not adhere to this culturalist orientation, to try to point to some of the approaches and issues that were foregone, and to raise questions about the status of that culturalist orientation. 

A recent paper by Frank Trentmann (n.d.) includes a brief review of different approaches to consumption. His review is of approaches among historians, but what he describes seems common to consumption studies generally. This is because his historians tend to approach consumption in terms of its links to other issues and processes, whether as indicators, facilitators, cause or effect.
My concern in this paper is anthropologists, and it seems that they are no different. Perhaps this is to be expected, for this is a discipline which advertises itself as being concerned with contextualising what it studies, to show how it is linked to other things. Anthropologists, then, appear to run in parallel with Trentmann’s historians. Both may spend some time trying to identify consumption as a social or historical thing, event or process, but they seem to spend much more time looking beyond it, to see what light it sheds on other issues.
That expansive view shapes this paper, which is concerned not with how consumers see themselves, but with how anthropologists see them. There is a lot of anthropological work on consumption, but my goal here is not the impossible one of reviewing and synthesising it. Rather, I want to use some of this work to raise questions about how anthropologists have thought about consumption, and particularly questions about the larger contexts in which they place consumption and consumers. So, like an anthropologist (and, following Trentmann, apparently like an historian), I will use work on the processes and mechanisms and practices of consuming to look beyond consumption.
I will do this in two uneven sections in what follows. First, and more briefly, I will sketch some of the formative works on consumption by anthropologists and the writers who have influenced them. I will use this sketch to raise questions about the approaches to social life those works contain. Then, and at greater length, I will present three case studies based on anthropological works that engage with consumption in Mexico, the English-speaking Caribbean and in Papua New Guinea. I present these cases because they locate consumption, consumers and what might be called innocent bystanders in terms of approaches and issues that seem to have slipped from view at the time that consumption rose to prominence as a topic within the discipline. Before I begin, however, I want to reflect a bit on the rise of the study of consumption.
In anthropology and elsewhere, consumption became an important topic in the 1980s. Some observations about that decade will help to explain the basis of my concerns about the sort of world constructed in the important works I will describe. While the academic study of consumption boomed, the activity itself became more problematic for many people. In the US and the UK and many other parts of the world this was, after all, the decade that saw the marked onset of neoliberal policies that shifted a lot of economic power from workers and states to the owners of capital. The consequences were striking. Media attention was focussed on the consumption boom amongst a small body of the newly enriched, the ‘yuppies’ (young upwardly-mobile professionals) and the dinks (dual-income no kids), and the industry that catered to them (e.g. Silverman 1986). However, the less visible statistical reality was disheartening for more ordinary people. In the US, for instance, the decade saw a continuation of the decline in hourly workers’ purchasing power. Households coped with this threat to their consumption by working harder: an average of 245 additional hours per year since 1973 (Schor 1991: 80-1). In more peripheral countries, things could be worse. For instance, in Mexico between 1982 and 1986, real income dropped by from 40 to 50 per cent (Heyman 1991: 176-8, 1994). In some countries, then, the 1980s was a decade of declining consumption; in many others, consumption became more uncertain for significant sections of the population.
The fervour with which anthropologists and others embraced the study of consumption in that decade would have been understandable if it had reflected this problematic state of affairs, or at least taken cognisance of it. However, as I shall show, this was not the case. The big names and the big ideas construed a world of choice among a world of goods by a world of people constrained only by the need to decide which object, among all that was on display, they wanted to buy. I find this unsettling, and what I present here is a way of expressing my unease.

I turn now to a brief discussion of four works that have reflected and shaped anthropological work on consumption: Baudrillard’s For a critique of the political economy of the sign (1981), Sahlins’s Culture and practical reason (1976), Douglas and Isherwood’s The world of goods (1978) and Bourdieu’s Distinction (1984). These works obviously do not capture the whole of work on consumption at the time, much less subsequent work. However, while these are early works, they deserve attention because they have been so influential. With a few important exceptions (most notably Mintz 1985), the framework that these works shaped remained important. And that framework is culturalist, focussing on the meanings that objects bear, meanings that are taken to explain why people consume those objects rather than others.
Baudrillard’s notion of ‘sign value’ exemplifies this. He says that objects of a certain type have meaning because they are different from objects of another type. Those types and their meanings define an overall structure of objects that maps on to a structure of society, made up of various types of people defined by their differences from other types of people. Sahlins illustrates this structural, semiological approach nicely, which is not surprising given his extensive invocation of Baudrillard. For instance, his discussion of American clothing revolves around ‘basic notions of time, place, and person as constituted in the cultural order’, and he argues that the classification of clothing in the US produces and reproduces ‘the meaningful differences between’ social units (Sahlins 1976: 181).
Baudrillard’s and Sahlins’s are the most structural and semiological of these culturalist texts. Douglas and Isherwood are more concerned with the social processes that give objects their meanings and significances, as when they point out that certain items have value because they allow households an increased flexibility in household routine and hence a greater ability to maintain desired social relationships. However, their main concern (or, rather, the aspect of their work that was most influential) is the ways that patterns of consumption reflect and recreate the structures of social life, as in their discussion of the ways that the structure of meals maps on to the structure of time.
Bourdieu’s work is the one that aspires to the most comprehensive account of consumption preferences. Distinction rests on a model of society, of social resources or capitals and of predispositions or habituses, and is as much about French society as it is about taste. The core of Bourdieu’s analysis of taste is the contrast between the sensuous orientation of those driven by necessity, particularly unskilled manual workers, and the aesthetic orientation of those who are relatively free of necessity, elites of various sorts. 
These classic works exhibit some important common themes that appear more broadly in anthropological and related work on consumption. The first is the construction of objects as bearers of meaning generated by advertisers (Schudson 1984) and consumers (Miller 1992). This has led analysts to focus mostly on items like toiletries, clothes, beverages, food stuffs and television programmes. These are fairly cheap, so that the difference between them can be treated as almost nothing but symbolic; they are inconsequential, so that the purchase on Tuesday afternoon of one soft drink rather than another, like the purchase on Saturday morning of one recording rather than another, makes no difference that anyone can detect. Another common feature of these models is their distinctive use of time. This is the evanescent time of wanting and acquiring, most visible in the more structuralist works but less salient in Douglas and Isherwood, with their (relatively neglected) concern with people’s practical strategies. But even Bourdieu’s use of life trajectories as an analytical tool has a synchronic air, because it is a time of social reproduction, in which, ultimately, nothing changes. The third common element in these works is their broadly psycho-cultural orientation to consumption. This orientation implies that all that we really need to know in order to understand consumption is the framework of desire that is in each individual's head (psychological), itself a manifestation of the collective (cultural) construction of a structure of meaning of objects of consumption and a parallel structure of people.
In its restricted time frame and its psycho-cultural orientation, these works portray the same sort of world as that in neoclassical economics, for both are interested in the same issue, people’s consumption choices in a market economy. And those consuming people are representative bearers of taste or interpreters of meaning. Indeed, for some people they are nothing but bearers of taste or interpreters of meaning: ‘The old, rigid barriers are disappearing - class and rank; blue collar and white collar; council tenant and home owner; employee and housewife. More and more we are simply consumers’ (Perry 1994: 4, quoted in Gabriel and Lang 1995: 36). These similarities are provocative, for they point out that both economists and anthropologists concerned with consumption focused on the moment of market choice, when the shopper confronts shelves piled high in a store. In this focus, these writers tend to ignore what lies outside that moment, the fact that people’s perceptions and their consumption are shaped by and shape the material, social and cultural constraints of their situations (cf. Miller’s [1987] use of Hegel’s notion of objectification); just as they tend to ignore the fact that these choices have consequences, not just for the consumers who make them, but for others as well. 
I have sketched some of the common threads in influential anthropological work on consumption. To stress again what I have aid already, this work does not define the whole of the anthropology of consumption. However, this work and its threads continue to be influential within the discipline, and continue to help define for those outside the discipline just what it is that anthropologists have to say on the topic. This predominance tends to obscure other interesting work which, precisely because it takes different approaches and situates consumption in different frames and addresses different issues, is less obviously about consumption. I want to present three cases that make use of this sort of work, each of which sees consumption less as an individual choice framed by meaning and more as a collective consequence, itself consequential, of political-economic forces. In seeing consumption this way, these cases see consumers less as choosers to be understood in terms of structures of meaning, and more as people whose acts can constrain them to choose in certain ways, and whose choices can constrain others and their actions.

The first case I want to present is directly concerned with consumers, but it focuses on people who are on the margins of capitalist society. These are people in the highlands of Sonora, in northwest Mexico adjacent to the US border. The anthropologist whose work I draw on is Joe Heyman, who has used field work, oral histories and archives to study these people’s consumption patterns as they have changed over the course of the twentieth century (Heyman 1990, 1994, 1997). This was a momentous century for Sonora’s people, for it saw the industrialisation of the border region, first with copper mines at the start of that century and then with the assembly plants, maquiladoras, that emerged in the 1960s to serve US markets.
Heyman’s work has been concerned with changes in people’s consumption as this industrialisation and their involvement with it has changed. He provides intriguing discussions of changes in the items that people consumed (e.g. beds, shoes, radios, cowboy hats, denim clothing) and some of those discussions approximate the culturalist approach that I have described. However, what I present here is his analysis of the relationship between people’s patterns or strategies of consumption and their position in the economic order. He summarises his treatment of these strategies (Heyman 1994: 179-83) as a distinction between two ideal types, which he calls ‘flow-through’ and ‘flow-conserving’ strategies.
While Heyman presents these as consumption strategies, and particularly household consumption strategies, he relates them to people’s resources, though what is important is not so much the sheer volume of their resources as it is the pattern. In relating patterns and strategies of consumption to patterns of resources, Heyman is rooting consumption in the realms of economy and political economy, rather than in the realms of culture and sign that have been the most visible anthropological approaches to the topic.
Put briefly, the flow-conserving strategy is one in which consumption of purchased items tends to be discontinuous and the ratio of purchase to self-provision varies markedly over time. On the other hand, the flow-through strategy is characterised by a steadier level of consumption of purchased items and, except in times of extraordinary hardship or prosperity, a more constant ratio of purchases to self-provision. 
I said that Heyman relates consumption strategies to resource patterns. The flow-conserving strategy characterises households where income is discontinuous, perhaps most obviously farming households. For them, income is tied to the agricultural cycle: the income they receive from this harvest has to last them until the next one. In contrast to flow-conserving households, flow-through households have continuous income flows, perhaps most obviously households reliant on wage labour. Because of their continuous income, such households have much less need to restrict their expenditures to specific times of the year. Indeed, they tend to avoid the uneven expenditure pattern of the flow-conserving households, preferring instead to have relatively continuous and predictable expenditures, balanced against their relatively continuous and predictable incomes.
These differences in strategies have a number of corollaries. One concerns the pattern of household debt. The flow-conserving households tend to accumulate debt gradually over the course of the year, and pay it off in a lump when they harvest and sell their produce. On the other hand, the flow-through households that Heyman describes are fairly poor, and if they want to buy anything substantial they do so on credit. Thus, they tend to acquire debt in a lump and pay it off gradually. Debt, then, is part of their continuous and predictable expenditures, and the need to pay off this debt makes wage work all the more important for them.
I said that the flow-conserving household has to live off the proceeds of one harvest until the next one comes in. Of course, life in the Sonora highlands was more complex than this idealisation. Few households relied only on their crops. Instead, they had other, even if less significant, economic resources: wage labour and the sale of a farm animal or items of household manufacture can generate income at other times of the year. They had another important economic strategy, which I pointed to when I said that the ratio of purchase to self-provision varied markedly over time. While these households were relatively prone to provide for themselves at all times, self-provisioning increased as debt rose over the course of the year. Worn objects were repaired rather than replaced, or were replaced by what the household made rather than what they bought; people used candles and firewood for light and heat when they were too indebted to afford paraffin; they carried water from streams rather than buy it from water-sellers.
This indicates that their consumption strategy was not linked only to their income pattern. In addition, it was linked to their relationship to economic resources. Specifically, they had to have fairly free access to things like water and wood, mud and sand, and so forth. Also, they had to have relatively free access to their own time; they had to be able to devote hours of it to collect wood and water or days of it to repair roof and wall.
Rural life in Sonora was no idyll. E. P. Thompson’s (1967) point that pre-industrial peasants work to their own regime of time should not blind us to the fact that they could work a lot, and that the work could be onerous. So, while the spreading loss of free access to material resources in Sonora may have driven people to the towns and cities, it is also certain that the restricted work regime in those places was attractive in many ways, as was the range of objects available in them. However, ensconced in town, these households found it increasingly difficult to maintain their old flow-conserving strategies. This was because the economic relations that had underpinned that strategy had disappeared. If free firewood and water were getting harder to find rural areas, they were effectively impossible to find in town. Perhaps even more important, the three or four hours of time required to get them had disappeared. The husband and older children were at work; the younger children where in school; the basic household maintenance fell more purely on the wife, even if her available time were reduced by her own wage work.
In these circumstances, people were obliged to forego the older self-provisioning that facilitated the flow-conserving strategy by allowing a periodic refusal to buy. Increasingly, shirts and shoes had to be bought rather than made. With no access to wood and no time to gather it, wood-fired cooking gave way to gas and electricity. The flow-conserving strategy gave way to the flow-through strategy. Sonora villagers became not just urban residents, but urban consumers, reliant on regular work and regular wages to pay their regular bills.
It is pertinent that the title of the paper in which Heyman describes this process in greatest detail begins with ‘The organizational logic of capitalist consumption’, and that he calls the change in strategy that I have described here ‘consumer proletarianization’ (Heyman 1994: 180). As these phrases might indicate, he is not describing the emergence of consumption, for certainly rural households consumed, and it is not clear that they consumed significantly less than urban households. Equally, he is not describing the appearance of the market in people’s lives, for rural households depended on market transactions for the selling of the goods and labour power that they produced as well as for the buying of much of what they consumed. Rather, what Heyman is describing is a change in pattern, brought about in large part by the loss of the room to manoeuvre that confronted these families most starkly when they moved to town and found jobs, though it was beginning to confront them in rural areas as access to free resources was reduced.
Heyman’s work points to some of what anthropologists missed when they embraced the culturalist approach to consumption that I described earlier in this paper. What he portrays is people whose consumption choices and strategies are shaped in basic ways by their positions in households that are themselves positioned in the political-economic order. Moreover, the points he makes do not reflect only the changing condition of a set of people in the northwest of Mexico. This much is apparent from concern in Britain about what is called ‘work-life balance’. While this phrase has a range of meanings reflecting the changing nature of work, some of its salience arises from the fact that a growing number of people and households confront a common situation. Faced with jobs that look increasingly insecure, they send more people out to work for more hours. One consequence is that their expenditure rises, as they find themselves having to buy what they used to do and make for themselves. For these households, it is not the loss of the time required to gather firewood and water, but the time required to look after small children rather than pay for a child-minder, the time required to prepare meals rather than buy them microwave-ready from a shop. Like Heyman’s maquiladora workers, these households find themselves increasingly proletarianised consumers.
Certainly, British households with two wage earners, like Sonoran households working in assembly plants in border towns, make decisions about what to buy that reflect things like sign value and cultural structures of meaning. However, Heyman’s work shows the kind of questions we can ask and the kinds of issues we confront if we attend to more than the moment of culturalist truth, the shopper facing six kinds of breakfast cereal on the supermarket shelves.

The next case I want to present resembles in some ways the culturalist approach I have described, for it is concerned with the meaning of objects in consumption. However, it locates those meanings in temporal and historical processes that are shaped by, and hence transmit, the sort of political-economic forces that the culturalist approach tends to ignore. This case is concerned with consumption in the Caribbean, capitalist territory but hardly the core of the developed world. The anthropologists whose work I draw on are Daniel Miller, who has done field work in Trinidad, and Richard Wilk, who has done field work in Belize.
As with much of the Caribbean, an important issue in these two countries is what used to be called nation-building, what now might be called the creation of a post-colonial national self-conception. While there are many aspects to building a nation, the one that concerns Miller and Wilk is what people eat and drink. Their descriptions of the link between consumption and identity, albeit national identity, help complicate the culturalist approach to consumption. They do so by showing some of the institutions and forces that shape the meaning of objects and people’s consumption decisions, institutions and forces that can be affected by the consumption decisions of outsiders. They do so as well by showing how these decisions, institutions and forces can affect legitimacy and political power.
For both Trinidad and Belize, independence and nation-building took place during a time when they confronted a growing array of global commodities, overwhelmingly produced and marketed by companies from elsewhere. In spite of the claims that creolisation is ubiquitous in countries such as these (Hannerz 1987; Miller 1992), so that global commodities acquire significant local meanings, the flood of brands marking globalisation would seem to be stony ground for developing a distinctive national identity. While the ground is not as fertile as it might be, Miller’s (1997) research in an advertising agency in Trinidad suggests that more complex processes are at work than either the flood or creolisation images imply, processes that lie beyond the view of an approach focused on people’s consumption decisions.
Miller found that the capitalist commercial logic that drove this advertising agency encouraged it to facilitate the development of a distinct Trinidadian identity, a ‘Trini way’. We would expect this for objects manufactured locally. However, what makes Miller’s discussion intriguing is that he shows how this happened clearly with truly global companies and their brands. As Miller describes the process, when a global company decides to try to expand its sales in Trinidad, it will approach a local advertising agency. However, as this is a global company, it will have its advertisements prepared centrally, and engage the Trinidadian advertising agency only to place them in effective times and places. This reduces the agency to just a buyer of time on local stations and space in local print media, which would profit the agency much less than producing the advertisements themselves. With an eye to their profits, advertising agencies argue that the Trinidadian market is distinctive, and that the advertising campaign requires a local orientation (Miller 1997: 79-83). Often enough, this strategy works; the local agency gets the creative job; their advertisements enhance an image of the Trini way, as well as the agency’s profit figures.
The advertising strategy that Miller describes indicates that the relationship between national identity and imported commodities is problematic; it takes a persuasive advertising agency to get the chance to make an international commodity resonate with the Trini way. In some sense, this is familiar territory, the complex relationship between the local and the global. But what Miller’s description points to is the way that the meaning of objects of consumption is shaped by forces that operate well beyond the local realm of cultural values: there is nothing distinctively ‘Trini way’ about the desire to increase your profits, whether you are a global manufacturing company or a local advertising agency. And as I show in the balance of my presentation of this Caribbean case, those forces and meanings impinge upon and reflect important political-economic concerns and interests.
This impinging is a recurring theme in Wilk’s writings on Belize (esp. 1995; 1999). When he first went to British Honduras (as it was then) as a student in 1973, his hosts served him tinned corned beef, white bread, tinned sardines and a 7-Up. These were suitable as a meal for a foreign visitor because they reflected what local people took to be English taste, which was high-status taste in this British colony. Local foodstuffs, on the other hand, were seen to be the foods of the poor and of the local population, two categories that overlapped significantly during the colonial era. With the gradual growth of nationalism, however, the better off in British Honduras changed their tastes, moving away from the sort of things shipped out, at least in popular imagination, from the mother country. They still, however, denigrated local foods. Rather, they adopted what was called ‘Spanish’ food, food that reflected what was thought to be eaten in Mexico and other independent countries in Central America. Only later did Belizean food emerge as an identifiable category.
This emergence was not, however, a straightforward consequence of the drive to independence or even the pressure by local firms to differentiate a local market, the process that Miller describes. Rather, Wilk says that an important factor was the development of tourism in the country. The growing number of foreigners coming on holiday wanted a taste of the local cuisine. Tamales or tinned beef with 7-Up would not do. So, restaurants catering to tourists borrowed or modified local foods, or invented dishes that echoed them. Real Belizean food was born. This cuisine was, of course, no more really Belizean than what is on offer at my neighbourhood take-away is really Chinese. Rather, its origins lay in ethnic and sub-regional cuisines, some of which were elevated to national status and some of which were ignored, in what was an historical process of trial and error by cooks seeking to attract tourists to restaurant tables. 
Like Miller’s description of multinational advertising campaigns and the Trini way, Wilk’s description of the emergence of Belizean food points to the ways that people’s understandings of the consumables that they confront are shaped by outside forces. I do not mean to imply that the outsider status of these forces is absolute, for clearly it is not: local account executives, hotelliers and restaurant owners are crucial to what happened in Trinidad and Belize. However, it remains the case that what was going on outside these countries significantly affected what was going on inside them. Further, these outside influences and the ways that they shape the local meaning of objects can constrain local people. At the level of the meaning of consumables, this is clearest in Belize, where the development of tourism meant that the consumption expectations of tourists, primarily from North America, were instrumental in generating a national cuisine.
However, when a national cuisine is created, more is at issue than the Belizean dish of the day at a tourist hotel at Ambergris Cay or Belize City. As I said, the creation of Belizean food affects different groups differently: those who have elements of their cuisine adopted are granted a national legitimacy that is denied to those whose cuisines are passed over and who remain merely local, or even invisible. Group identity and national politics, then, are affected by what goes on in these restaurant kitchens and in the imaginations of the tourists who eat there.
Miller’s work on Trinidad describes the nature of this effect, and the complexities that can lurk in a soft drink (Miller 1997: 146-9). Trinidad is divided roughly equally between those of South Asian descent (Indo-Trinidadians) and those of African descent (Afro-Trinidadians), and much of national politics in the country revolves around the tension between these two blocs: their claim to be true Trinidadians, and hence their claims to political legitimacy (Munasinghe 2001). This cleavage does not play itself out only at the level of party politics. As one might expect, it also plays itself out at the level of consumption. Thus, Miller notes that the carbonated soft-drink market is divided between two generic products. One is black sweet drinks, colas of various sorts; the other is red sweet drinks, largely indistinguishable from colas except by their colour. The black sweet drink is memorialised in the ‘rum and coke’, a common festive drink; the red sweet drink is memorialised in ‘a red and a roti’, a common snack. While all Trinidadians are familiar with both of these drinks, the black sweet drink is considered to be more Afro-Trinidadian and the red sweet drink is considered to be more Indo-Trinidadian. Thus, while Afro-Trinidadians commonly go out for a red and a roti, they are likely to see that food as characteristically Indo-Trinidadian.
However, the meanings of the red and the black, and the political cleavages and tensions they reflect, are not just a local affair. As was the case with the emergence of real Belizean food, outsiders’ tastes and assumptions and consumption preferences are important as well. To the outside world, the Caribbean drink is rum and coke; the red and a roti are invisible. In this, Indo-Trinidadians are marginal with regard to what constitutes the Caribbean in general and Trinidad in particular. Were Trinidadians left to themselves, this might be fairly unimportant stuff. However, with the growing importance of tourism, these outside identifications are fed into Trinidad. What tourists want and what they expect to see affect what local people think and do. 
This influence of outside tastes and consumption preferences is apparent with Carnival. Carnival or something like it is common in much of Europe and the New World (Miller 1994: 107-13). However, in the tourism industry it appears to have its most significant forms in New Orleans, Rio de Janeiro and Trinidad. In these three places, tourists flood in to satisfy what is, after all, their consumption preferences for the music, the parades, the dance and the drink. In Trinidad, as tourism becomes more important, the state and commercial bodies that seek to attract tourists have increasingly identified Trinidad with Carnival. And this identification seems to have worked: a lot of tourists want Carnival and see Trinidad as Carnival-land.
In all three of its most visible forms, Carnival is dark-complected to the point of being Black. In Trinidad this creates tension. As the salience of Carnival becomes greater in the public perception of the country, so the Afro-Trinidadian sector of the population has become more secure; its claims to be authentically Trinidadian, and hence legitimate, is buttressed. For Indo-Trinidadians on the other hand, whatever their actual participation, symbolically they are absent and their authenticity and legitimacy suffer in comparison. If Indo-Trinidadian complaints about the Carnivalisation of the country are any indication, tourist consumption preferences influence political legitimacy and power in Trinidad, just as they seem to do in Belize.

My last case extends the point I made about Belize and Trinidad, that outsiders’ consumption preferences can affect people’s lives. This case concerns ecotourism, particularly with a regard to a set of people who live in an area on the borders of the Gulf, Chimbu and Eastern Highlands provinces in Papua New Guinea (PNG). These people have been studied by Paige West (2000, 2001), and I draw on her work for what I say here. I will present this case at greater length than the other two. This is because the mechanisms by which these preferences are transmitted in this case are less familiar than they are in Belize and Trinidad. Hence, I want to show those mechanisms, rather than simply pointing to them (see West and Carrier forthcoming).
Being a tourist is about as close as one can get to being a pure consumer, tourism is a huge business, and ecotourism is routinely said to be the fastest growing sort. Ecotourism has many definitions, but most revolve around two points. It involves travel in order to experience attractive parts of the world defined as ‘nature’, in a way that benefits nature, or at least disrupts it as little as possible. As well, though not signalled in its name, it involves travel to experience attractive people who are exotic for the tourist, usually people construed as ‘primitive’ or ‘simple’. Again, this travel should benefit these people, or at least disrupt them as little as possible. 
In many of the tropical, Third World countries where ecotourists go, the business is seen to be extremely important economically. In part this is because of the general growth in tourism over the past fifteen or twenty years. In part also this is because changes in aid and trade policies in the closing decades of the twentieth century have meant that the economic position of many of these countries has become more precarious, as has the economic position of rural people in them. Consequently, attracting tourists, and especially ecotourists, has come to be seen as a key way for these countries to generate foreign exchange, and it has been an attractive way to generate wealth for people in rural areas.
One such set of people is the Gimi speakers in the area around Crater Mountain, in the PNG Highlands. They were attracted to the idea of ecotourism after the Crater Mountain Wildlife Management Area (hereafter simply ‘Crater Mountain’) was established by the PNG government in 1994, after sustained political agitation by a set of Australians who had spent time in that part of the country. The PNG government handed the management of Crater Mountain over to the Research and Conservation Foundation of Papua New Guinea (RCF), which was the organisation that pushed for its establishment and which had received support from the Wildlife Conservation Society, the international conservation arm of the New York Zoological Society and the Bronx Zoo.
PNG had no money to support Crater Mountain once the Wildlife Management Area was established, and the RCF was left to find its own resources. They applied for conservation funding, which came as a grant from the Biodiversity Conservation Network (BCN), about US$ 490,000 between mid-1995 and mid-1998, with a further US$ 77,000 from the Wildlife Conservation Society. Following BCN advice, the RCF’s grant application stressed the development of ecotourism and requested funds for ecotourism infrastructure.
The BCN not only provided support, it also encouraged a particular orientation to environmental conservation at Crater Mountain, an orientation that reflected the institutional matrix in which it operated. The BCN received funding from the Biodiversity Support Program, a consortium of the World Wildlife Fund, The Nature Conservancy and the World Resources Institute, funded in part by the USAID. The consortium was ‘partnered’ with two global agencies of the US government, which meant that staff from the two agencies worked with consortium staff in shaping funded projects. One of the agencies was the United States-Asia Environmental Partnership. This is run by the USAID and it promotes ‘sustainable development’ in Asia through public and private initiatives. The other was the US Commercial Service, the international arm of the US Department of Commerce. It works to help US businesses compete in the global market by promoting the export of American goods and services and by working to protect the interests of those businesses internationally. 
In the case of the BCN, ‘sustainable development’ meant encouraging ‘enterprise-oriented’ approaches to conservation. The BCN, and the US agencies behind it, proposed that if rural people were given business strategies that relied on the sustainable use of the environment for success, and if they were linked to a ‘community of stakeholders’ (effectively purchasers) outside their rural areas, then enterprise-oriented conservation would be successful. People would work to conserve their environs because they would profit by doing so. Put differently, the BCN effectively induced local people to set up programmes that would cater to consumer demand of one sort of another. The most obvious type of consumer demand was ecotourism, and local people agreed.
This agreement was a complex and significant event. To understand why, it is necessary to describe the social organisation of the Gimi-speaking people in the area. Those people are members of one or another of six clans, and though clan membership was somewhat dispersed, all significant parts of the area were considered to belong to one or another of those clans. In addition, relations between clans occasionally is antagonistic, and often is uneasy. In such a situation, as one might expect, clans keep an eye on each other, and are anxious not to see one clan gaining resources, wealth or prestige at the expense of another. There are mechanisms for defusing tension and generating a degree of cooperation, such as an ecotourism project would require. These mechanisms revolve around protracted and careful negotiations between clans. Often this is accompanied by the transfer of wealth between them or the allocation of future wealth among them in ways that reduce inequalities or redress grievances.
In such a situation, the anticipation of any activity that looks likely to generate wealth raises concerns about clan relations. The prospect of ecotourists certainly did so, which made the initial agreement to proceed with an ecotourism project a significant event. But of course, the initial agreement was a relatively easy one to secure: in the absence of concrete plans that would raise concrete concerns, the vision of a stream of wealthy tourists was an attractive one. However, things became more difficult when the decisions to be made became more concrete.
The first concrete decision was where to build the lodge where the ecotourists would stay. An obvious spot was at the small airstrip in the area. The airstrip and the area around it belonged to one particular clan, which West calls Namabu, and people were concerned that if the lodge were built on Namabu land, then this clan would get the bulk of the wealth and status that the lodge would bring. This was a particular matter of concern because the Namabu clan already had benefited a great deal from dealings with outsiders and the uses to which their land had been put. For instance, a member of Namabu was the salaried agent for the missionary air services to the village. Several RCF-owned houses were on Namabu land, and Namabu gained rental income and status from having RCF employees stay on their land. The Namabu has a storage shed at the landing strip, and charged other villagers when they stored coffee there. As well, the local school and houses for teachers were on Namabu land, which meant that Namabu children had almost no walk to school, while others had as much as a half-hour walk, and it meant that the teachers, all from other parts of the country, had formed close relationships with the Namabu.
Given the Namabu’s privileged position in the village, placing the ecotourist lodge next to the airstrip on Namabu land would cause significant tension, as the Namabu would get the bulk of the benefit. Conscious of this, the committee of elder men in charge of the issue decided it would be best to place the lodge at a hamlet about half an hour’s walk from the airstrip. Four clans had a significant number of residents at this hamlet, so that placing the lodge there would spread the anticipated benefits more equally among the local clans. It took the committee almost two years to reach this decision, but it was accepted by everyone, including the Namabu elders.
The village elders had made a decision about the location of the lodge that reflected both their desire to have ecotourism and their wish to avoid the antagonisms that would follow from placing the lodge at the airstrip. In short, they pursued the interests of local people as a whole and used their own decision-making mechanism to arrive at the decision. However, the conservation biologists and other outsiders working on Crater Mountain assumed that the decision would be a purely commercial one, shaped by the consumption preferences of outsiders. That is, they assumed that the location would be selected to attract the greatest number of ecotourists. This meant that the important issue was ease of access to the airstrip and to the projected village artefacts shop, and the attractiveness of the view of the forest from the lodge. Events proved them right.
Early in 1997 the expatriate director of the RCF was visiting the area and the decision on the location of the lodge was presented to her. She listened to the explanation for the choice and walked to the proposed location to see it. She walked back to the airstrip, where the committee of elder men were waiting, together with most of the villagers, who had turned up in celebration of her arrival. She told them that their proposed location was not acceptable. It was ‘too far from the airstrip and tourists will not want to walk that far’ and ‘the view from the airstrip is so much better’. The lodge’s location at the airstrip was fixed and the case was closed. Two years of delicate negotiations among villagers were dismissed with a five-minute discussion. 
The director’s decision caused much anger, but the committee of elders put their resentment aside and worked together to proceed with building the lodge. The promise of the project was so great: all those rich, foreign tourists who would be interested in Maimafu villagers and their natural surroundings. But the decision to proceed with the lodge raised another concrete question: who would cook and clean for the tourists, and benefit thereby? After more intense negotiations that took months, the committee decided that benefits should be spread among all the clans. As each new set of tourists arrived, a set of people from a different clan should work at the lodge and act as guides. Again, however, they were thwarted. The expatriate advisors to the RCF said that there was a better way. One person would be chosen and paid to run the lodge; several would be trained by a foreign expert in how to lead tourist treks, cook ‘bush dinners’ and interact with ecotourists, and would become paid ecotourism guides. The visible and equitable dispersing of benefits from the lodge was abandoned in favour of a commercial construction of what consumers wanted.
The situation that West describes shows again that it may be worthwhile to shift our focus away from the culturalist concern with the consumer making choices within a field of meaningful objects. Instead, as with the Caribbean cases, it can be revealing to focus on how people can be constrained by the choices made by consumers elsewhere. In this village in the Highlands of PNG, justifiable concerns about encouraging fairly equitable and amicable relations among sets of local people led leaders to cater to ecotourist consumers in particular ways. However, their decisions were overturned because they were seen as commercially undesirable.
Of course, there are no actual ecotourist consumers in this case, only their surrogates, the senior staff of the RCF. And these staff doubtless would have argued that, in acting as surrogates, they were making decisions that would increase the attractiveness of the site, increase the number of ecotourism visitors and so increase the financial benefit to local people. Similar arguments could be made by restaurateurs in Ambergris Cay producing a Belizean cuisine and by travel agents advertising Trinidad and Carnival. They may all have been right. And certainly the RCF staff were conforming to the policies of the organisations, based in the US and supported by the US government, that were supporting and advising the project. In doing so, however, they were demonstrating the ways that the past decisions of a body of Western consumers, ecotourists, created a set of expectations among the advisors that constrained the villagers of Maimafu and led to a set of decisions that looked likely to increase the inequalities between clans and increase the tension and perhaps even conflict among local people.

I said at the start that the anthropological work on consumption that is most visible is culturalist, concerned with the relationships among objects, their meanings and the identities of consumers in the moment of consumption choice. While work in the discipline has become more diverse than the relatively semiological models contained in the key writings that I described, much of that work still bears the culturalist orientation of these classic publications. Here I have presented three cases, in order to make one, simple point. There is much to be gained by looking at other anthropological work that approaches consumption in other ways and that locates objects and consumers in different worlds.
In making this point, I do not mean that the culturalist orientation has been fruitless, for it has led to interesting and intriguing work. However, it has the weakness of its strength: the intense focus that is revealing is also restricted and constricting, reflecting, perhaps, the increasingly culturalist orientation in the discipline as a whole in the closing decades of the twentieth century. That culturalist orientation not only emerged at about the same time that neoclassical economics was becoming predominant in Western societies. As well, in its focus on meaning and choice in the present moment, it echoes the common neoclassical construction of the world as market transaction. And as I said, this happened during a period when many people’s ability to choose in market transactions was becomingly markedly less secure, as was their level of consumption.
The constrictions of the culturalist view can be overcome if we turn to work that is concerned with more than the here and now of meaning and choice, whether the choice of villagers contemplating a possible marriage or the choice of consumers in a department store. Just as mainstream anthropology of consumption has taken objects and consumption and used them to get at a larger context of meaning, so other anthropologists have taken objects and consumption and used them to get at other larger contexts. However, such is the dominance of the culturalist approach that this other work often does not look like anthropology of consumption, especially to those outside the discipline. Consider the cases I have presented here: there are no structures of desire and no supermarket aisles; there are not even identities to be adopted or families to be shopped for (as in Miller 1998).
There are many reasons for this absence. An important one is that the researchers whose work I have used arrived at consumption through routes other than the conventional one. Heyman arrived at it through an interest in the political-economic forces at work in Sonora; Wilk went to Belize, then British Honduras, as an archaeologist; West went to PNG interested in anthropology of the environment; even Miller wanted to study modern capitalist society in a region that lacked the historical depth and baggage of Western Europe. Arriving at consumption via routes other than those of culturalist anthropology of consumption, they situated it in the context of questions other than those of choice, meaning and identity. In doing so, they help to show the breadth and depth of what anthropologists have to say about consumption, its corollaries and consequences. 
My own biases are, I suspect, clear. If the anthropology of consumption is defined by the culturalist approach, and a case can be made that it is, then its limits are such that it might be time to abandon it, or to recognise that it has little new to tell those who are not part of its internal debates: the constraints and restrictions seem to outweigh the insights. This does not mean that anthropologists should ignore consumption, for it is too pervasive and important for that. It does mean, however, that attending to consumption and those who consume may be most rewarding if we approach them in other ways and in terms of other issues. The most rewarding future seems likely to be one in which consumption in anthropology becomes something like what kinship used to be. That is, for a small number of specialists, a primary interest; for the rest of us, a part of the lives and processes we study and an influence on the other issues we address.

Without Frank Trentmann’s invitation, the paper would never have been written. He has my thanks. Although they did not know it, extensive conversations with Joe Heyman, Danny Miller, Rick Wilk and Paige West contributed greatly to what you see here. The opening sections of this paper draw on Carrier and Heyman (1997).
1. Adjusted for inflation, the average weekly earnings of production workers in manufacturing industries in the US was lower in 1990 than it was in 1965 (US Dept of Labor 1989: Tables 83, 85; 113, 1991: Table C-1).
2. This synchronic tendency is balanced by other anthropological work, which uses a different, epochal view of time. That view relates consumption to the emergence of capitalism in the West (e.g. Miller 1987; Mintz 1985) or the emergence of the nation state (e.g. Foster 1991; Wilk 1995).
3. In fact, the nature and position of public festivals in Trinidad is more complex than the salience and image of Carnival would indicate (see Korom 2003; Miller 1994: Chap. 3).
4. This is a perverse echo of Miller’s (1995) assertion that, for anthropologists, consumption should take the place of kinship as the key route to understanding the organisation and operation of society.

No comments: