Reading Journal

Global Networks 1, 4 (2001) 353–368. ISSN 1470–2266 353

On the trail of the global green bean:

methodological considerations in

multi-site ethnography

SUSANNE FREIDBERG

Abstract The need to demystify and ‘ground’ globalization has spurred many calls for multi-site fieldwork. This article discusses how such fieldwork was used to examine the contemporary restructuring of fresh food commodity chains between Africa and Europe, at a time of increasing European concerns about food safety and quality. Drawing on convention theory and actor-network theory for conceptual guidance, qualitative fieldwork was conducted at sites of production, import and export in ‘anglophone’ (Zambia–London) and ‘francophone’ (Burkina Faso–Paris) commodity chains. The article also discusses the challenges posed by multi-site research, especially in realms where secrecy and deception are standard tools of the trade.

The African French bean has become one of the symbols of the new global food economy, where tropical regions export high-value fresh produce to markets thousands of kilometres to the north, and where trade is regimented less and less by distance and season, and increasingly by the strict quality and safety standards of the European market. But a French bean bought in France often does not come from the same place in Africa as one bought in England, as colonial ties still dominate this trade. It has almost certainly been subjected to different forms and degrees of processing and regulation, despite ongoing European unification. It is the object, ultimately, of a whole different set of social relations of production and distribution. And all this before it even gets to the kitchen.

The magnitude of difference between what I will call the ‘anglophone’ and ‘francophone’ green bean commodity chains demonstrates particularly effectively a point already made in Goodman and Watts’ Globalising Food (Goodman and Watts 1997): namely, that processes of economic liberalization, regulatory harmonization and transnational investment do not necessarily obliterate (and may in fact accentuate) geographic variation in food provisioning practices and networks (Goodman 1997). But this analytical acknowledgement of difference and contingency poses a methodological challenge: how best to identify the sources of variation and change within transnational food commodity chains? This question holds relevance not only for the fields of economic geography and agro-food studies but also for the broader body of research on the ‘social life’ of transnational commodities (Apparurai 1986).

Certainly it helps to start by identifying the loci of power and control. Gereffi emphasizes that different key agents dominate different chains, depending on the

Susanne Freidberg

354

production processes and markets involved. Industries with high barriers to entry, like automobiles and computers, are characterized as ‘producer-driven’, while textiles, footwear, and most if not all agricultural commodity chains are ‘buyer-driven’, because they are largely controlled by the firms that control marketing, brand names and design (Gereffi and Korzeniewicz 1994; Raikes et al. 2000). In the case of fresh produce, this means major food retailers (Barrett et al. 1999; Marsden 1997).

Few would disagree that the retailers have come to play an enormously powerful role in the agro-food system. But their control over supply and marketing can explain only so much, because in certain crucial ways food is not like other commodities. To take the example of the African French bean: the differences between contemporary French and British food retailers are significant, but they do not by themselves explain why the francophone and anglophone commodity chains differ all the way from the farm (where African peasant production predominates in the former, and large-scale plantations in the latter) to the market (where the French prefer loose beans, and the British pre-packed, trimmed and microwave ready). Equally essential to the story are Africa’s distinctive regional agrarian histories, and distinctive French and British histories of colonialism, trade, and culinary innovation. Above all, the Afro-European trade in green beans and other fresh foods must be understood as a set of linked human activities and biophysical processes, all subject to a range of unpredictable perils, from infestations and wind storms to cancelled flights and market gluts. The trade therefore depends greatly on trust relations, frequent communication, and a variety of risk-management strategies which, taken together, are not easily accessible through a conventional study of a single site or sector.

This article shows how comparative multi-site ethnography can illuminate the dynamic, geographically dispersed activities and social relations that comprise transnational food commodity chains. The need to demystify and ‘ground’ globalization has spurred many calls for multi-site research (Burawoy et al. 2000; Marcus 1998; Miller 1997; Mitchell 1997). This article actually documents the experience of carrying out such a project, with all its benefits and pitfalls.

The project in question sought to contribute to the body of research investigating the relationship between contemporary concerns about food safety and quality in the industrialized world and the ongoing restructuring of certain transnational food commodity chains. More specifically, it sought to demonstrate a point neglected by this still largely Anglo-centric literature: namely, that the restructuring occurring within even a single small sector of the global food trade (like the Afro–European trade in fresh vegetables) is not homogeneous. Rather, it is geographically varied, and shaped by the ‘cultures of commerce’ of particular groups, as well as by place- specific histories of food production and distribution.

The next section outlines the research conducted in 1993–4 and 1999–2000 in Burkina Faso, Zambia, London and Paris.1 The remainder of the article discusses first the analytical concepts that framed this project, and then the challenges and dilemmas encountered in the fieldwork itself. I do not make light of the obstacles to research inside industries where investigators of any kind are considered suspect, and where deception and secrecy are standard tools of the trade. I do suggest, however, that these obstacles become more instructive, if not less formidable, in the context of a comparative study, as they show how the workings of different commodity chains are shaped by specific economic pressures, political constraints and cultural norms.

On the trail of the global green bean

355

The fields

The comparative study of transnational green beans grew out of 1993–4 doctoral field research on the social history of commercial gardening in south-western Burkina Faso. Through participant observation and interviews in gardening villages and marketplaces I became familiar with the day-to-day activities and challenges of commercial vegetable production and marketing in Burkina Faso (Freidberg 2001). In addition, I recorded the professional biographies of a number of local businessmen who, in the wake of recent trade liberalization, had just begun exporting green beans to Europe (Freidberg 1997). This biographical approach drew on the example of Gregoire and Labazée, among others, who in their work on African entrepreneurship use biographies to understand how individual entrepreneurs build careers from a combination of inherited and acquired resources (Gregoire and Labazée 1993; Labazée 1988; Labazée 1994). In my case, the businessman interviews also provided information about the logistics of agro-exporting, as well as about the French market.

In the mid-1990s, just as more and more African countries began to follow World Bank and other donors’ recommendations that they produce high value, ‘non- traditional’ food commodities such as green beans for export (Jaffee 1995), the future of these export initiatives was thrown into question by three distinct sets of events. First, the Uruguay Round of the GATT trade negotiations called for the end of Europe’s preferential trade arrangements with its former colonies, thereby putting tropical and counter-season produce exporters such as Burkina Faso into competition with the likes of Brazil and Florida. Second, two outbreaks of the deadly Ebola virus in Central and West Africa, combined with the publication of a series of influential (if perhaps alarmist) books and articles on the tropics’ ‘emerging’ pathogens, heightened the West’s popular perceptions of Africa as a disease-ridden ‘hot zone’ (Garrett 1994; Kaplan 1994; Preston 1994). Third, the mad cow crisis, combined with evidence that the rising worldwide incidence of food-borne disease was due partly (though by no means entirely) to increasing trade, helped to turn food imports into a potential political issue in Europe. Although the European Union has been divided by food crises originating from within (such as Britain’s mad cow disease, and later Belgium’s dioxin-tainted chickens) it has generally reacted in firm unison towards outside threats. Thus, recent years have seen not only the ongoing EU ban on hormone-fed US beef but also a series of bans on fish, meat and produce from parts of Africa where either disease outbreaks or lack of proper documentation has raised suspicions (though typically not evidence) of contamination.

During the same time period, UK-based researchers found that UK supermarkets’ sourcing of fresh vegetables from Kenya (a former British colony) began to rely less on small farmer cooperatives and more on large plantations, which can more easily demonstrate compliance with the supermarkets’ strict standards regarding labour, the environment, and above all on-farm sanitation and chemical use (Dolan et al. 1999). But could this shift characterize trends across the continent? In the late 1990s, it seemed unlikely to occur in France’s former colonies, and only partly because such plantations, while not uncommon in the tropical-fruit exporting regions of Côte d’Ivoire and francophone Central Africa, have been rare and generally unsuccessful in the regions suitable for growing vegetables like green beans (Mackintosh 1989). In addition, the shift in Kenya reflected the UK supermarkets’ increasingly direct and

Susanne Freidberg

356

dictatorial intervention into aspects of overseas sourcing that French retailers prefer not to be bothered with, for a number of reasons. Finally, the vast literature on these two countries’ histories of food, ways and manners2 indicates that, stereotypes aside, the British and the French have quite different ideas about what is ‘good to eat’ (Harris 1986). Given that such dietary and culinary norms have no small influence on a country’s food import patterns (for example, Britain imports from Africa far greater quantities of ‘Asian’ vegetables, such as baby corn, than does France or any other European country (Graef 1995)) they must be an integral part of any contemporary transnational food commodity chain analysis. Such was my reasoning, in any case, when I set out to compare the ‘anglophone’ and ‘francophone’ fresh vegetable trades between Africa and Europe.

The first stage of fieldwork for this study took place in the Marché Internationale de Rungis, which describes itself as the world’s largest wholesale produce market, and is the destination for most green beans from francophone Africa (nearly all the rest go to Marseilles). Located near the Orly airport on the outskirts of Paris, Rungis is a vast concrete space of warehouses, loading docks and truck parking lots, interspersed with several bars and brasseries. Most of the actual deliveries and shipping-out of produce takes place in the early morning hours, but wholesalers and especially importers do much of their business later in the day and over the phone. Drawing on information and contacts from my previous fieldwork in Burkina Faso, I interviewed nearly all the importers who worked in Africa and who were willing to talk to me (18 total) plus an assortment of wholesalers, transporters, and representatives of the fresh produce trade association. Although these office-based interviews hardly resembled conventional participant observation, the long-distance phone calls that frequently interrupted the conversations provided examples of how different importers conducted transactions and handled day-to-day crises. In addition, I had many casual conversations with tradespeople (who are overwhelmingly male) both on the wholesale floors and in one of the bars they frequented.

Six months later, I returned to Burkina Faso during the green bean ‘high season’ (January–March). Once I had met with the president of the fresh produce exporters association, all the active members of the association (ten and one women) were willing to speak to me. In these interviews they discussed their professional histories, their work as exporters, and how this work had changed in recent years. As I will discuss again later, their initial openness did not necessarily indicate any intention to reveal much. I also interviewed members and presidents of growers’ cooperatives, spoke with representatives of the agencies providing aid to the horticultural export sector, and visited several informants from my earlier research. Lastly, I frequently visited the main packhouse at the Ouagadougou airport, where a few hundred workers (mostly young women) sort and pack beans for nighttime flights. Thanks to the friendly Burkinabé woman manager there, I was able to spend as much time as I wanted on the packhouse floor, talking with the workers and learning, hands on, how they sort good beans from rejects. This kind of access was unexpected and, as it turns out, unrepeated. In the anglophone chain, such an apparent lack of concern about sanitation and public relations would have been considered dangerous and unprofessional.

For the anglophone Africa case study I originally chose Zimbabwe, which has (or had) one of the continent’s most dynamic horticultural industries. But political

On the trail of the global green bean

357

violence and land occupations in mid-2000 made on-farm visits in Zimbabwe look less feasible, if not impossible, so I instead went to Zambia. In some ways it turned out to be a better case study to compare with Burkina Faso: the two countries have nearly the same population size and per capita income (10 million and $320, respectively); both are land-locked, post-socialist, and in similarly ambiguous states of democratic transition. I do not, however, consider either country to be ‘typical’ francophone or anglophone horticultural exporters. The structure and social organization of Burkina Faso’s export sector, for example, reflects geographic and historical conditions very different from those that have shaped agrarian change in Senegal and Morocco, while mineral-rich Zambia’s historical neglect of export agriculture distinguishes it from Kenya and Zimbabwe. The labels ‘francophone’ and ‘anglophone’, in that sense, signal the importance of particular colonial experiences and places – and enduring postcolonial links to the metropole – more than they describe a standard type of exporter or trade relationship.

The biggest contrasts between Burkina Faso and Zambia lie in their agroecology (Burkina Faso has less water and a shorter growing season) and in the horticultural export sectors themselves. Whereas Burkina Faso’s exporting is in the hands of several small (in some cases one-man) operations, Zambia’s is entirely controlled by two large companies, both of which export a relatively wide range of vegetables, most of them pre-packed (namely, snowpeas, baby corn, baby squash, peppers, and of course green beans) as well as roses.3 These companies (like most in Zimbabwe) are financed by foreign corporate capital and run by people (men) of European descent; all the exporters in Burkina Faso are Burkinabé nationals, and financed by a combination of informal credit, bank loans, and creative uses of foreign aid monies. Nearly all growers in Burkina Faso (who are predominantly but not entirely male) cultivate less than a hectare, using a combination of household and hired labour. In Zambia, by contrast, production takes place on plantations of thirty to thousands of hectares, owned by the export companies themselves or by commercial outgrowers (also of European descent) and worked by piece-rate labourers. Finally, Burkina Faso’s vegetable exports have declined in volume over the past several years, while Zambia’s have boomed; Burkina’s sector is widely considered to be in crisis – and at risk of losing its hold in the European market altogether – while export horticulture is the most dynamic part of the Zambian economy.

The field research in Zambia involved interviews with the top and middle-level managers of the two companies, as well as with several of their outgrowers. I also spoke informally with some of the lower-level employees of these companies, most of whom are black Zambians, and with some of the many consultants and NGO workers promoting the country’s horticultural sector, which at the time was by far the fastest growing part of the economy. As I will discuss further in the final section, field research in Zambia posed an entirely different set of problems than in Burkina Faso, not so much because the ‘sample size’ was much smaller (two companies), but because the company managers were much more worried about negative publicity, and thus more suspicious of anyone remotely resembling a journalist. This reflects, I think, the British press’ greater propensity for ‘name and shame’ stories, especially in the food industry.

Finally, while still in Zambia I arranged interviews with the two export companies’ main ‘agents’ in London, whose companies supply fresh produce to the

Susanne Freidberg

358

UK’s top supermarkets. Although these agents play nominally the same inter- mediary role as the French importers, they are in fact much more subordinate to the demands of the supermarkets. They are also, however, extremely well positioned to talk about the power dynamics and recent trends in the fresh produce trade, and I was fortunate in that both the people I met were willing to do so. They were also able to discuss how Zambia compared with their other supply sources, namely Kenya and Zimbabwe.

In subsequent trips to London and Paris (January and July 2001) I have maintained contact with a number of importers, and conducted interviews with produce managers at two of the UK supermarkets that purchase Zambian vegetables. The ongoing research on the anglophone chain has benefited from discussions with some of the several British researchers who have already conducted studies and published work on this sector.4 This kind of informal consultation and collaboration enriches almost any kind of research, but is especially valuable in multi-site field- based projects where the fields are not only geographically dispersed but also changing rapidly. Shared knowledge, in other words, helps compensate for some of the logistical difficulties of research in the globalized economy.

The tools

Political economy analysis shaped much of the early research on the globalization of food. This literature (which has been reviewed and critiqued in detail elsewhere (Busch and Juska 1997; Goodman 1997; Goodman and Watts 1997)) treated food as a commodity more or less like any other, and saw strong parallels in the historical trajectories of industry and agriculture, in that both were being shaped by the increasing speed and complexity of corporate capital flows, and by the eroding power of the nation-state (Bonanno 1994). Much writing has since questioned these parallels, arguing that macro political economy neglects not only regional distinctiveness but also the role of two key sets of agents driving change and ‘constructing value’ in the global food system (Marsden 1997): first, the consumer and the agents that supply and regulate the consumers’ food, and second, the many kinds of biophysical nature that act upon food, directly or indirectly. One of the strengths of commodity chain analysis is that, at least in theory, it takes account both of ‘consumption’ and of ‘nature’, one way or another. In reality, the big question is which ways – which methods of fieldwork and analysis – most effectively and coherently capture the complexity and historical dynamism of the relationships between different links of a commodity chain. Multi-site ethnography offers one way, provided the focus of enquiry is well defined (Marcus 1998). This section discusses the ideas that informed the design and implementation of this project.

Consumption, quality and conventions

Within agro-food studies, both theoretical and empirical efforts to ‘acknowledge’ the consumer have had to consider the question of how much power consumers actually exercise over their food supply (Fine and Leopold 1993; Miller 1995; Warde 1997). The ‘counterculture’ of the 1960s and 1970s certainly helped bring ‘natural’ and ‘ethnic’ foods into the mainstream (Belasco 1990; Goodman and Redclift 1991), and well-publicized protests against bio-engineered foods and ‘McDonaldization’ (Bové,

On the trail of the global green bean

359

et al. 2001) have encouraged farmers in Europe and North America to switch to organic and artisanal production. But Marsden, Flynn and Harrison argue that consumer movements in general have had relatively little impact on national food policy in the UK, where the government’s disastrous handling of BSE and other food- borne risks have arguably given consumers more reason to mobilize than in most industrialized countries. Instead, they assert, the top supermarket chains have used their power over suppliers to become the most trusted and responsive gatekeepers of the national food supply. In defending the consumer interest they have also been able to mould it (Marsden et al. 2000). But this is not necessarily the case everywhere. Although large supermarkets chains now also dominate food retailing in much of continental West Europe, it is not clear that they influence consumer concerns as powerfully as in the UK. They certainly have not usurped the ‘expert’ role enjoyed by France’s specialized butchers, bakers, and top-market greengrocers (Ducrocq 1999; Larmet 1999). For the purposes of this project, what is clear is that consumers ultimately have no direct control over the importation of food across their national borders. Therefore, while consumer-based research data can provide useful contextual information (and I have used this where available)5 it seems that ethnographic enquiry should focus on the people who do have this control, and on their methods for assuring that their supply will satisfy consumer demands.

Here, French convention theory provides some useful direction. In countries such as France, where food surplus has long been a greater economic problem than scarcity, convention theory examines, among other things, how ‘quality’ helps farms and firms survive in a competitive and discerning marketplace. The term ‘quality’ has been variously defined in the contemporary agro-food literature (Murdoch et al. 1999). Bertil Sylvander uses it to distinguish between traditional food standards like size or weight, which are immediately apparent, and qualities such as purity, ‘freshness’, and taste, which may derive from particular production processes but are not always immediately obvious or universally recognized. By this definition, quality is a construct – one that consumers presumably accept and value – and its marketing depends on means of external assurance, or ‘conventions’. These include ‘institutional’ means of quality assurance (such as certification, labelling, and ‘traceability’ systems, indicating how goods are produced) and ‘transactional’ means (namely, the norms and practices through which the buyer comes to trust the seller) (Sylvander 1995).

One of the major foci of aims for this project was to investigate how different kinds of conventions shape the relationships and daily work activities of growers, exporters and importers. Both institutional and transactional means of quality assurance exist in the transnational green bean trade, but they operate quite differently in the anglophone and francophone commodity chains. In the former, prodigious paperwork and a corps of supervisors and record-keepers ensure that every bean is monitored from seed to supermarket shelf. In addition, nearly all produce is trimmed and packed in Zambia, in microwaveable plastic boxes with supermarket labels. Retailers’ own-brand labels have become a more important institutional convention in the UK than anywhere else, promising the consumer predictable quality and value (though airfreight baby vegetables are hardly cheap, at £12–15 per kilo).

Although both Zambia’s exporters and British importers emphasized the importance of knowing and trusting each other, the entire relationship rests on both

Susanne Freidberg

360

parties’ conformity to supermarket expectations, as demonstrated in their book- keeping and on-site audits. The same applies to the relationship between importer and supermarket buyer, which one importer characterizes as ‘beyond personality … it’s a totally professional relationship.’

In the francophone chain, by contrast, all beans are shipped loose in 2.5 or 5 kilo cartons, partly because the French market for pre-packed beans is very small (and supplied nearly entirely by Kenya). All cartons, as defined by EU law,6 must be labelled by grade, calibre, and national origin, but French importers rely more on the character and past performance of their African suppliers than on carton labels or other written quality assurances. After all, it is very easy to hide substandard green beans at the bottom of a carton labelled ‘premiere quality’, and in fact certain exporters are known for such tricks. With their buyers as well, French importers have traditionally depended on frequent personal contact in order to sell produce. They bemoan the anonymity and transience of most French supermarket buyers who, like their UK counterparts, increasingly hold the importers to rigid lists of standards (cahiers des charges). As one importer said about this trend:

Take, for example, tomatoes of category 1.57–67. Theoretically, if this were an industrial product, they’d all be identical. But when it comes to fruits and vegetables, these products, from the visual and gustatory aspect, can be totally different. So if you cannot explain over the telephone to a potential buyer that your product is like this or like that, he’s going to rely on a standard that means absolutely nothing. So it’s a bit frustrating to not be able to defend one’s product.

Comparative analysis of the uses and perceptions of institutional and transactional conventions reveals, first, different distributions of wealth and power. In Zambia, corporate capital and Commonwealth aid has enabled the export sector to meet retailer demands for traceability and pre-pack facilities; in Burkina Faso financing is much scarcer and more scattered. In the anglophone chain, the top five British supermarkets’ near-monopoly over the domestic food market allows them great control over not only prices, packaging and so forth, but also the routine commercial conduct of their suppliers. In the latter realm, at least, the French importers have more leeway to decide who they will do business with, and how.

Second, comparative analysis helps to clarify the real-life significance of different national and international regulatory institutions, as well as the effects of regulatory change. For example, the EU has harmonized pesticide residue laws, resulting in the likely ban of many chemicals now used by African horticultural producers (Chan and King 2000). But it is common knowledge among European produce importers that inspections at certain mainland ports of entry are more lax than others; shipments of suspect goods can be re-routed accordingly. The British government’s Pesticide Residue Committee, meanwhile, annually ‘names and shames’ all retailers caught selling residue-tainted foodstuffs. UK importers and supermarket representatives refer to this practice as one of their greatest sources of anxiety.

But the structure and regulation of markets cannot alone explain how trade relations within particular commodity chains are actually formed and maintained. In the anglophone chain, the fact that all the exporters and importers are not only

On the trail of the global green bean

361

English-speaking but also white and male makes their relationships very different, in their degree of ease, empathy and even solidarity, than relationships between the Burkinabé exporters and French importers which tend to be complicated by their respective characterizations of ‘the French’ (or ‘the white man’) and ‘the Africans’. In other words trust, as the basis of ‘transactional’ conventions, may be more difficult in trades where the participants are divided not only by distance but also racial or cultural stereotypes and memories of colonial domination. The difficulties may not be openly discussed, but they reveal themselves in casual remarks, in overheard conversations between trade partners, and above all in the comparative analysis of commodity chains rooted in different colonial histories. Indeed, one of the strongest arguments for employing multi-site ethnography in the study of transnational commodity chains is the insight that it can provide into the social and historical basis of trust relations. Finally, and returning to the role of the consumer, the comparative analysis of conventions shows how the daily conduct of commercial work is organized around different assumptions about what qualities consumers want, and how best to meet those wants.

Nature and networks

The question of what is unique about the nature of food is hardly new. Much of the early political economy of agriculture literature debated the natural obstacles to the complete industrialization of food production (Goodman et al. 1987; Mann and Dickinson 1978) while sociologists and anthropologists of consumption have written extensively on the relationship between the particularly intimate, material and in some ways perilous nature of food consumption and the dizzying variety of food taboos, anxieties, and preparation and mealtime rituals (Douglas 1984; Fischler 1993; Mennell et al. 1992).

Contemporary research analysing how nature informs food distribution is relatively scarce, though it is clearly critical to understanding what ‘globalization’, as it applies to food, really means. So far, the ‘organic’ content of food (Fine 1994) has placed certain restrictions on the kind of truly globalized ‘assembly line’ production seen in the automobile and athletic shoe industries (Dicken 1998). Instead, the past three decades have witnessed the increasing volume and diversity of cross-border and transoceanic food trades – especially in fresh fruits and vegetables – and, more broadly, the extension and reconfiguration of ‘agro-food networks’ involved not only in commerce but also research, investment, and advocacy (Lockie 2000).

Three recent analyses of such networks emphasize that they include a wide range of institutions as well as non-human ‘actants’, ranging from plants to machines to nature–culture ‘hybrids’. In Canada’s rapeseed sector, ‘the extension and modification of relationships between institutions, people, things and symbols’ led to the creation of new rapeseed varieties, suitable for human consumption (Busch and Juska 1997). In the American corn industry, the development of hybrid corn has depended on an expanding ‘collective’, whose members include (among others) parent seed lines, gene banks, land grant colleges, farmers, and even human and livestock metabolisms (Fitzsimmons and Goodman 1998). In the corporate capital-dominated world of coffee, a network of activists, farmer cooperatives, bankers and fax machines sustain an alternative coffee ‘fair trade’ (Whatmore and Thorne 1997).

Susanne Freidberg

362

All these examples show how actor–network analysis can draw nature out of the background and into the constitutive processes and relationships of transnational food commodity chains. None, however, suggest how perilously dynamic and sensitive an actor nature can be, when the commodity is something like fresh green beans, produced somewhere like the Sahel. An extra millimetre of girth, a spot, wrinkle or bend ruins the market value of a ‘fine’ or ‘extra-fine’ green bean, while an infestation or especially strong wind (Burkina Faso’s growing season coincides with the season of the Harmattan, the harsh dusty wind from the Sahara) can ruin an entire crop. But the most common ‘natural disaster’ in Burkina’s green bean business is the desiccation and decay that occurs when some combination of human and mechanical failing leaves the fragile beans immobilized somewhere en route to market, such as on the airport runway. Zambia’s fresh vegetables are vulnerable to similar hazards, but far superior cold chain facilities, better-maintained vehicles, and a much more centralized production zone all help get produce shipped with its ‘freshness’ (a key quality) intact. At the ports of entry in Europe, the tight coordination of human activity and technology remains critical to the maintenance of freshness and thus market value. Flights from sub-Saharan Africa are scheduled to arrive in the pre-dawn hours, so that the produce is sold on the wholesale docks or en route to supermarkets the same morning.

In short, efforts to stimulate, prune, preserve, or race against the clock of biophysical nature are central to the daily activities and relationships of the transnational green bean trade. Although this is arguably true for trade in all highly perishable goods, spending time in sites along a commodity chain – in fields, trucks, airports, warehouses, offices – is one of the most direct ways to gain appreciation for the role of non-human ‘actants’, simply because during that time you are subjected to at least some of their effects, namely, midday heat, flat tyres and power outages. Comparative multi-site ethnography also makes it possible to identify which natural processes and actors are place- or commodity-specific, which are continuously or occasionally present, which most preoccupy the human participants in a particular trade, and so forth. In addition, this kind of fieldwork illuminates how ‘human-nature’ interactions at different sites shape the human relationships between sites. For example, the Burkinabé exporters’ difficulty keeping produce cool in between harvest and flight time is a major source of tension in their relations with French importers who, if they receive shrivelled beans, can do little to disguise the damage in the eyes of their own clients.

The challenges

In a multi-site fieldwork project, a clear analytical framework helps clarify the objectives in each site, and thus helps ward off doubts about the depth of understanding possible from such a project. This, of course, was traditionally anthropology’s greatest suspicion of social science field methodologies that did not aim for immersion in or comprehensive knowledge of a particular society (Marcus 1998). Although few would still question the legitimacy and value of multi-site ethnography, the difficulties involved in actually doing it remain.

Some are the logistical problems familiar to conventional single-site ethnography, only multiplied. They pose more or less formidable obstacles depending on the scope

On the trail of the global green bean

363

of the project and the time, money and other resources available. For example, ideally I would have spoken fluently all the languages I might encounter in each site. Practically, this was impossible; although I was equipped with English, French and some Dioula (one of the languages spoken in southern Burkina Faso) I did not have time to learn Mooré (spoken elsewhere in Burkina Faso) or any of the several African languages spoken in Zambia’s capital, Lusaka (especially since I only decided to go to Zambia, rather than Zimbabwe, two months in advance). As it happens, the sites and social situations I wanted most to understand are, in fact, thoroughly francophone or anglophone; only occasionally did I use a translator in Burkina Faso. Local languages would have been much more indispensable, however, if this study had focused on the work and relations of production (namely, in the fields and pack houses) rather than distribution.

Another issue is time. Putting aside the practical time constraints most field researchers face (at least when the field is not near home), the subject matter of much contemporary multi-site ethnography may not even be suitable to the extended timeframe that characterized anthropological fieldwork in the past. In the case of the transnational vegetable trade, where the regulations, politics, technologies and markets are all changing rapidly, I felt that I needed to visit all the sites within a relatively compact time span (fifteen months) in order to see how they experienced roughly similar conditions. Efficiency in the field, therefore, was essential, though often frustrated either by the limitations of local phone lines or by the highly mobile schedules of the people I needed to see. In the end, however, the greatest problem was not lack of time but rather limited access, as one would expect, to places like company offices, boardrooms, and warehouses. More time will not necessarily improve access to such places, but it might open up other channels of information. In my case, ‘after hours’ conversations with people who worked for or were familiar with the companies I was interested in often yielded useful observations and leads.

The problem of access is clearly linked to positioning – that is, the question of how to identify and ally oneself as a researcher. The question, discussed widely elsewhere,7 has ethical, epistemological and often quite immediate strategic implications, to the extent that it may open or shut doors. This is certainly the case in the study of a commodity chain in which many of the subjects are pitted against each other, either as competitors or as parties to highly unequal (and perhaps unstable) exchange and contract relations. In this case, my own positioning was somewhat simplified by the fact that all the participants in the Afro-European vegetable trade are pitted against the same larger forces: the intensely competitive European market, the price-squeezing supermarkets, the ‘paranoid’ consumer, the scandal-hungry popular media. Although importers, exporters, and growers are clearly not equally equipped to confront these forces, it was not a stretch for me to describe myself to all these groups as someone concerned about the future of the entire trade. Since they share these concerns, most people were willing and in some cases eager to talk to me.

My credibility also appeared to benefit from the fact that my research involved extensive travel. In the transnational fresh produce trade, the increasing use of the Internet and other communication technologies has not diminished the value placed upon face-to-face contact and on-site experience. Even the British importers, who spend perhaps 95 per cent of their working time in air-conditioned corporate offices, described their frustrations working with supermarket buyers who are ‘fresh out of

Susanne Freidberg

364

university’ and entirely ignorant of conditions in Africa. Because I had just come from Zambia and had also spent time in francophone Africa as well as the French wholesale market, I avoided the ‘ivory tower academic’ label, and was instead considered someone who might actually possess useful information.

That said, the very scope of my research no doubt made at least some subjects nervous, because they knew I would be talking to their suppliers, clients and/or competitors. Statements promising confidentiality do not necessarily reassure people who, as in the francophone trade, are accustomed to promises being broken or who, as in Zambia, work in a community so small and intimate that no one is truly anonymous. Some members in this community also feared that I might be a corporate spy or, worse, an undercover ‘BBC agent’.

But even if they were not terribly concerned about confidentiality, many people quite understandably wanted to portray themselves positively, and thus were much more ready to discuss the weaknesses and wrongdoings of another group than their own. At best, they might concede that wrongdoings ‘happen’ in their milieu, that mistakes get made – but not by them. So, for example, exporters in Burkina Faso talked at length about the duplicity of growers and importers, but attributed their own collective reputation for shady dealings to the actions of an unscrupulous few. In Zambia, representatives of one of the export companies blamed substandard produce on the outgrowers’ ‘bad management’, while the outgrowers claimed that the firm rejected perfectly good produce whenever its market was slow. Again, the basic problem of getting the ‘whole story’ is hardly unique to multi-site research. But at least in a commodity chain study the method provides part of the solution, in that research in multiple sites allows for a kind of transnational crosscheck. Analysing the stories of actors at different points in the commodity chain does not necessarily clarify who is lying or who is ‘right’; it does, however, help to situate these stories both locally and in the transnational context of the commodity chain. In other words, it helps to make sense of contradictions and ambiguities both within and between individual accounts.

Finally, fieldwork in transnational fresh food trades cannot help but raise questions about the political and policy implications of research. In the eyes of at least some of the groups associated with contemporary anti-globalization protests, the African French bean embodies the ironies and injustices of the globalized food economy: produced in underfed countries by people earning typically less than a dollar a day, transported at great expense, and consumed by affluent city-dwellers who neglect the seasonal produce of their own countryside. This quite accurate characterization, already captured in a recent British television documentary called ‘Mangetout’, gives corporate authorities in the anglophone trade good reason to worry about press coverage. It also makes them eager to stress at least three positive points about their industry: first, export horticulture in Africa employs tens of thousands of people in areas where job options are scarce; second, most of these people are women, whose wages are crucial to family well-being; and third, labour and environmental standards in the horticultural industry (at least in countries like Zambia, Kenya, and Zimbabwe) are higher than in any other agricultural sub-sector. All these points are also quite true, and provide solid support for the industry’s argument that, overall, African countries benefit from European demand for out-of- season, exotic, ‘baby’, and otherwise frivolous fruits and vegetables.

On the trail of the global green bean

365

To the extent that multi-site ethnography facilitates understanding of, and even empathy with, the positions of diverse actors within controversial commodity chains, it complicates the politics of globalization. But complicated understandings need not be immobilizing. In the Afro-European horticultural trade, one potential arena for progressive action is the ethical trade movement and, more specifically, the British- government supported Ethical Trade Initiative (ETI), which seeks to develop codes of conduct that will ‘improve labour conditions around the world’.8 This is not by any stretch a grassroots movement; although groups such as Oxfam and Christian Aid and several trade unions are involved, some of the strongest supporters are the UK supermarkets, who seek to ward off the charges of exploitation that have plagued the athletic apparel industry, among others. Still, ETI codes of conduct concerning on- farm pesticide use, sanitation, worker housing, and medical care have been adopted by fresh produce suppliers in Kenya, Zimbabwe and Zambia. In all three countries, of course, the mostly white-owned vegetable farms have not only the financial means but also good political reasons (in the context of domestic tensions around race and land) to maintain an image of social responsibility. To the extent that codes of conduct do not account for the very different circumstances in countries with little or no corporate farming, ‘ethical’ trade risks becoming exclusionary trade.

But this possibility is not inevitable, because the ETI has become a forum for discussion – at conferences, in print, on-line – between supermarkets, suppliers, trade unionists, activists and researchers (with some overlap in the latter categories). Within the African horticultural industry, some participants in these forums have called for codes to be developed to cover more countries, more commodities (namely, tropical fruit), and more types of enterprises (namely, small farms). In addition, some have called for the financial burden of ethical trade to be redistributed, so that retailers assume some of the costs now shouldered by producers. The ETI, in other words, has become a platform for challenges not only to the scope of ethical trade but also to the entire balance of power within horticultural commodity chains. Careful qualitative fieldwork within these chains can do much to refine and bolster the arguments for change. Clearly, then, ethnographies in/of the world system (Marcus 1998) have a place in the emerging forums and networks that are simultaneously products of and challenges to globalization.

Conclusion

This article has argued that multi-site comparative ethnography offers insights into transnational trade relations that would not necessarily be revealed by either traditional ‘mono-site’ ethnography or through an analysis based purely on macro political economy. In the case of the trade in fresh, high-value vegetables between Africa and Europe, this approach helps to situate distinctive cultures of commerce in particular colonial histories, physical geographies, and constructs of food safety and quality. More broadly, it lends support to a critical analysis of globalization, by differentiating between the supposedly homogenizing forces of economic liberalization and regulatory harmonization, and the diverse ways in which the participants in transnational trade adapt to, resist or perhaps disregard formal changes in law and policy.

That said, as with any method it is important to recognize and indeed anticipate

Susanne Freidberg

366

the potential obstacles and eventual limits to knowledge. With multi-site ethnography, some of the constraints are purely practical, but others are more intrinsic to the subject matter, insofar as many activities and social relations cannot be observed in any conventional sense. These obstacles demand creativity and collaboration in the field, as well as candour and modesty in the presentation of findings. At best, the very multilateral and comparative structure of such a research project helps to open doors that might otherwise stay closed, reconcile apparently contradictory or ambiguous accounts, and identify directions for further enquiry.

Susanne Freidberg is at the Department of Geography, Dartmouth College, New Hampshire, USA.

Acknowledgements

I thank the National Science Foundation Cultural Anthropology Program for inviting me to present the original version of this article in the ‘Globalizing Methodologies’ sponsored session at the American Anthropological Association Annual Meeting, San Francisco, CA., 15–19 November 2000. I also thank the anonymous reviewers for their useful comments, and Arouna Ouedraogo and Guy Scott for their insights and assistance at crucial points of the green bean trail.

Notes

1. This research was supported by the National Science Foundation, the Fulbright IIE Program, the African Studies Rocca Family Fellowship and the Dartmouth College Rockefeller Center.

2. For example, Burnett (1997), Mennell (1985) and Teuteberg (1992). 3. There are also several smaller rose growers, operating on approximately two hectares

apiece, but many of them have gone out of business recently. 4. These include teams at the Institute of Development Studies (Dolan et al. 1999) and at the

Natural Resources Institute (Chan and King 2000). 5. For France, I have consulted consumer surveys conducted annually by the Institut National

de la Statistique et des Etudes Economiques, as well as surveys and focus group studies conducted as part of the government’s project on Les Etats Généraux de l’Alimentation, http://www.agriculture.gouv.fr/ega/index.htm. For the UK, I have referred to the Ministry of Agriculture, Fisheries and Food annual National Food Survey, as well as the secondary literature produced as part of ‘The Nation’s Diet’ Research Programme (Murcott 1998).

6. The norms applicable to the green bean are published in the Official Journal of the European Economic Community, 7 July 1962, No. 58/62.

7. For example, Coffey (1999), England (1994) and Haraway (1991). 8. ETI website, http://www.ethicaltrade.org/_html/about/faq/framesets/f_shell.shtml

References

Appadurai, A. (ed.) (1986) The social life of things: commodities in cultural perspective, Cambridge: Cambridge University Press.

Barrett, H., B. Ilbery, A. Browne, and T. Binns (1999) ‘Globalization and the changing networks of food supply: the importation of fresh horticultural produce from Kenya into the UK’, Transactions of the Institute of British Geographers, 24, 159–74.

On the trail of the global green bean

367

Belasco, W. (1990) Appetite for change, New York: Pantheon. Bonanno, A. (ed.) (1994) From Columbus to ConAgra: the globalization of agriculture and

food, Lawrence: University of Kansas Press. Bové, J., F. Dufour, and A. de Casparis (2001) The world is not for sale: farmers against junk

food, New York: Verso. Burawoy, M., J. Blum, S. George, Z. Gil, T. Gowan, L. Haney, M. Klawiter, S. Lopez, Ó.

Riain, and M. Thayer (2000) Global ethnography, Berkeley: University of California Press. Burnett, J. (1979) Plenty and want: a social history of diet in England from 1815 to the present

day, London: Scolar Press. Busch, L. and A. Juska, (1997) ‘Beyond political economy: actor-networks and the

globalization of agriculture’, Review of International Political Economy, 4, 668–708. Chan, M.-K. and B. King (2000) ‘Review of the implications of changes in EU pesticide

legislation on the production and export of fruits and vegetables from developing country suppliers’, NRI Report 2525, Natural Resources Institute, University of Greenwich.

Coffey, A. (1999) The ethnographic self: fieldwork and the representation of identity, London: Sage.

Dicken, P. (1998) Global shift: transforming the world economy, London: Peter Chapman. Dolan, C., J. Humphrey and C. Harris-Pascal (1999) ‘Horticulture commodity chains: the

impact of the UK market on the African fresh vegetable industry’, Institute of Development Studies, University of Sussex, Brighton, England.

Douglas, M. (1984) Food and the social order, New York: Russell Sage. Ducrocq, C. (1999) Distribution alimentaire: enjeux stratégiques et perspectives 1994–1998,

Paris: Les Echos. England, K. (1994) ‘Getting personal: reflexivity, positionality, and feminist research,’

Professional Geographer, 46, 80–9. Fine, B. (1994) ‘Towards a political economy of food’, Review of International Political

Economy, 1, 519–45. Fine, B., and Leopold, E. (1993) The world of consumption, New York: Routledge. Fischler, C. (1993) L’Omnivore: le gôut, la cuisine et la corps, Paris: Odile Jacob. Fitzsimmons, M. and D. Goodman (1998) ‘Incorporating nature: environmental narratives and

the reproduction of food’, in B. Braun and N. Castree (eds) Remaking reality: nature at the millennium, New York: Routledge, 194–220.

Freidberg, S. (1997) ‘Contacts, contracts and green bean schemes: liberalisation and agro- entrepreneurship in Burkina Faso’, Journal of Modern African Studies, 35, 101–28.

Freidberg, S. (2001) ‘To garden, to market: gendered meanings of work on an African urban periphery’, Gender, Place and Culture, 8, 5–24.

Garrett, L. (1994) The coming plague: newly emerging diseases in a world out of balance, New York: Farrar Straus and Giroux.

Gereffi, G., and M. Korzeniewicz (1994) Commodity chains and global capitalism, Westport, CT: Greenwood Press.

Goodman, D. (1997) ‘World-scale processes and agro-food systems: critique and research needs’, Review of International Political Economy, 4, 663–87.

Goodman, D., and M. Redclift (1991) Refashioning nature: food, ecology and culture, New York: Routledge.

Goodman, D., and M. Watts (eds) (1997) Globalising food: agrarian questions and global restructuring, New York: Routledge.

Goodman, D., B. Sorj and J. Wilkinson (1987) From farming to biotechnology: a theory of agro-industrial development, New York: Basil Blackwell.

Graef, J. (1995) ‘World market for fresh and canned baby corn’, RAP Market Information Bulletin, No. 5, http://www.iit.edu/~tulsanu/gaurav.html.

Susanne Freidberg

368

Gregoire, E., and P. Labazée (1993) Les grands commercants d’Afrique de l’Ouest, Paris: Karthala-ORSTOM.

Haraway, D. J. (1991) ‘Situated knowledges: the science question in feminism and the privilege of partial knowledge’, in Simians, cyborgs, and women: the reinvention of nature, New York: Routledge.

Harris, M. (1986) Good to eat: riddles of food and culture, London: George Allen. Jaffee, S. (1995) Marketing Africa’s high-value foods: comparative experiences of an emergent

private sector, Dubuque, IA: Kendall/Hunt. Kaplan, R. (1994) ‘The coming anarchy: how scarcity, crime, overpopulation, tribalism, and

disease are rapidly destroying the social fabric of our planet’, Atlantic Monthly, 273, 44–76. Labazée, P. (1988) Entreprises et entrepreneurs du Burkina Faso, Paris: Karthala. Labazée, P. (1994) ‘Producteurs, consommateurs et marchands du Nord ivoirien: aspects de la

construction sociale des relations d’echange’, Cahiers des sciences humaines, 30, 211–27. Larmet, G. (1999) ‘L’Organisation des pratiques d’approvisionnement alimentaire’, Paris:

INRA. Lockie, S. (2000) ‘Beyond the farm gate: production-consumption networks and agri-food

research’, Sociologia Ruralis, 40, 3–19. Mackintosh, M. (1989) Gender, class, and rural transition: agribusiness and the food crisis in

Senegal, London: Zed Books. Mann, S., and J. Dickinson (1978) ‘Obstacles to the development of capitalist agriculture’,

Journal of Peasant Studies, 5, 466–88. Marcus, G. (1998) ‘Ethnography in/of the world system: the emergence of multi-sited

ethnography’, in G. Marcus (ed.) Ethnography through thick and thin, Princeton: Princeton University Press, 79–104.

Marsden, T. (1997) ‘Creating space for food: the distinctiveness of recent agrarian development’, in D. Goodman and M. Watts (eds) Globalizing food: agrarian questions and global restructuring , New York: Routledge, 169–91.

Marsden, T., A. Flynn, and M. Harrison (2000) Consuming interests: the social provision of foods, London: UCL Press.

Mennell, S. (1985) All manners of food: eating and taste in England and France from the Middle Ages to the present, Oxford: Basil Blackwell.

Mennell, S., A. Murcott, and A.van Otterloo (1992) The sociology of food, London: Sage. Miller, D. (ed.) (1995) Acknowledging consumption: a review of new studies, New York:

Cambridge University Press. Miller, D. (1997) Capitalism: an ethnographic approach, Washington, DC: Berg. Mitchell, K. (1997) ‘Transnational discourse: bringing geography back in’, Antipode, 29, 101–14. Murcott, A. (ed.) (1998) The nation’s diet: the social science of food choice, New York:

Longman. Murdoch, J., T. Marsden and J. Banks (1999) ‘Quality, nature and embeddedness: some theoretical

considerations in the context of the food sector’, Economic Geography, 75, 107–25. Preston, R. (1994) The hot zone, New York: Random House. Raikes, P., M. F. Jensen, and S. Ponte (2000) Global commodity chain analysis and the French

filière approach: comparison and critique, Copenhagen: Center for Development Research. Sylvander, B. (1995) ‘Conventions de qualité, concurrence, et coopération: cas du “Label

Rouge” dans la filière volailles’, in G. Allaire and R. Boyer (eds) La grande transformation de l’agriculture: lectures conventionnalistes et regulationnistes, Paris: Economica, 73–96.

Teuteberg, H. (1992) European food history: a research review, New York: St Martin’s Press. Warde, A. (1997) Consumption, food and taste, London: Sage. Whatmore, S. and L. Thorne (1997) ‘Nourishing networks: alternative geographies of food’, in

D. Goodman and M. Watts (eds) Globalising food: agrarian questions and global restructuring, New York: Routledge, 287–304.