As the most powerful instrument for expressing and shaping interactions between humans, food is the primary gift and a repository of condensed social meanings. Any food system has multiple dimensions (material, sociocultural, nutritional-medical), all of which interrelate. Food derives its ‘power’ from the web of the interrelations it evokes. Besides being of academic interest, the interconnectedness of production, distribution and consumption has been acknowledged as central to the formulation of effective food policies.
Fieldworking anthropologists are well placed to research the levels and intersecting nodes at which food must be understood. They have long been interested in human diets, specifically in the sociocultural determinants of diet; changing patterns of food production and markets; and food security at community and household levels. Increasingly, anthropologists are turning their attention to the socioeconomics of hunger, famine and food aid; and agricultural development and food policy.
Anthropological studies of food draw inspiration from the pioneering research of Audrey Richards (1939), in which the social dimensions of production, preparation, distribution and consumption were outlined, along with the dynamics of commensality. Food-focused ethnographies remain a model for those studying the social and nutritional impact of economic development. D’Souza (1988) has called for similar studies — famine ethnographies — to help prevent famine and improve relief efforts. Such studies require emphasis on the interrelationships between local, national, regional and international variables.
Anthropologists take a flexible approach to food and culture, because individuals at times have to choose between contradictory norms. The semiotic approach to food, championed by "Appadurai (1981), highlights how the intellectual properties of food may be manipulated to solve this problem of choice.
Food as cultural construction
Studies of the sociocultural dimensions of food take either a cultural/structural or semiotic approach. The former treats food as ‘good to think’ within a fairly static cultural environment, as illustrated in numerous studies of binary classifications (hot—cold, wet—dry, male—female, etc.), and introduces the idea that diets can be analysed as parts of a food code. In contrast, semiotic studies show that people manipulate food to make statements about and challenge social relations. While all cultures use food to mark or build relative prestige and social status, the ways in which Hindu castes manipulate food transactions to improve relative status, and the emotions with which food preparations are charged, set Hindus apart (Messer 1984). Using food to protect or protest social positions, Hindus are experts in gastropolitics (Appadurai 1981). The semiotic approach is indebted to Levi-Strauss, who first developed the theory that food was of the order of language.
Changing patterns of production and the role of markets
The inquiry into food as idiom can be extended to the study of agriculture and agrarian change, because fields and field crops carry social meanings. For instance, the transition from upland/male farming to swampland/ female farming in Gola, Sierra Leone, has been facilitated conceptually by the ‘power’ of associated images. The shift in terrain reflects engendered concepts about danger and safety (Leach 1994).
While the literature on changing food production patterns has many references to environmental symbolism, the dominant focus is on labour relations and land tenure. Changes are set against the backdrop of global transformations. In Africa earlier this century, no one who needed land went landless. Inequalities existed, but communal groups and individuals had usufruct rights to land. The class and gender inequalities of today result from the recent introduction of cash and cash crops, through which new production relations emerged between and within households. As commodity relations destroyed bonds that secured non-free labour (e.g. that of sons), men increasingly took the major production decisions, while married women became involved as unfree labourers (Whitehead 1990).
Agrarian societies in transition are marked by widening separations in the division of labour by gender. Gender separation also pertains to new responsibilities for meeting consumption needs. Whereas men and women used to share responsibilities through sharing complementary production tasks, gender separation leaves women more and more solely responsible for household food provisioning. In addition, women must regularly supplement their agricultural activities to keep their families’ food secure. This has been noted for both Africa and Asia, especially in areas where the adoption of high-yielding crop varieties has left smallholders in heavy financial debt.
Empirical studies of agrarian change in Asia (as in Africa) highlight how macro and micro variables fuse, and reveal shared experiences as well as regionally specific ones. Javanese women, for example, share with African women exceptionally heavy labour burdens, yet, with regard to land rights, they have maintained their rights to private ownership (Stoler 1977). The literature on Asia centres on the accentuation of social distinctions under Green Revolution regimes. For India, studies of agrarian change outline a polarization between permanent (hightech) agricultural labourers and casual labourers, the latter being mainly women employed during peak labour times. The narrowing of the range of female tasks causes ‘crowding’ and devalues women’s work. Women take part in waged labour mostly as unfree members of households with restricted access to land and credit (Harriss 1977). Research in Southeast Asia confirms the rise of gender separation, crowding and the devaluation of women’s work (Stoler 1977). Here too, women take on additional paid work to achieve basic food security.
Against over-optimism about the technological benefits of modern agriculture, anthropologists argue that gains have been offset by increased social differentiation and equity losses. Gender and class intersect here. Although women bear the brunt of modern hardships, real costs fall disproportionately on women from poorer households.
In keeping with the holistic approach to food, anthropologists regularly link the concept of a dynamic, gendered agriculture with the domain of food marketing. In Bwisha, Democratic Republic of Congo, for instance, where commercialization began around independence, land scarcity and cash scarcity combine to give rise to a non-seasonal, individuated approach to farming, with household food security now being dependent on markets that prove difficult to control. This dependence stimulated gender negotiations regarding roles, rewards and duties. Since the 1985 famine, Bwisha women have regained some control over marketing and food security through a strategy whereby they introduced crops they control unambiguously (Pottier and Fairhead 1991).
Field studies of actual market performances usefully complement market analyses by economists, because the broad systemic approach by anthropologists integrates many domains of social life. Market studies often focus on how farmers and marketeers cope with the volatile nature of markets. In some cases, protection is offered by traders who enter into equilibrating relationships with farmers; in others the state attempts to increase household security through better producer prices or by opening up markets for staples not normally bought by national marketing boards. State interventions, however, regularly fail, because planners divorce market performance from conceptual and production aspects of food. Market performance must be seen in conjunction with other domains of the food system. On this basis Russell attacks the concept of equilibrating relationships, pointing out that peasant farmers in the Philippines must borrow from commercial middlemen in order to obtain credit; a practice leading to an indebtedness that obliterates the voluntary nature of personalized exchanges (Russell 1987). The need for linking markets and production relations has been heightened by the imposition of structural adjustment programmes.
Food security at community and household levels
The debate on food security generates awkward methodological questions regarding observation and measurement. The key problem is that measuring food intake is virtually impossible as it involves a specific set-up that distorts behaviour. The advice by anthropologists is to develop a resource control perspective which considers intra-household food allocations in relation to who controls which resources.
Some nutritionists, however, are less than enthusiastic about the theory that the cause of much malnutrition in women and small children is unfair distribution within households. The objection is that it is extremely difficult to establish whether such discrimination occurs. Rather than adopt the maldistribution theory, nutritionists may accept the functionalist perspective that households and individuals modify their patterns of food use and adjust to seasonal or occupational changes in food entitlement (Wheeler and Abdullah 1988). While anthropologists do not dispute that households and individuals adapt to predictable fluctuations in food supply, they note that deepening poverty makes sharing much harder than when periodic, predictable hunger occurs (Schoepf and Walu 1991).
Given the paucity of empirical studies that have explored and backed the functionalist perspective, it is suggested that both the ‘resource control’ and the ‘functionalist’ approach deserve further investigation. The value of the resource control approach, however, is being enhanced because of the damaging impact of structural adjustment measures on the poor.
Hunger, famine and food aid
The anthropological concern with combining global and local variables is reflected in studies of hunger and famine. Thus, while most food crises share key features (e.g. upheaval, social differentiation, climatic disturbance) and develop along predictable lines, each crisis must also be regarded as unique. It is for this reason that detailed famine ethnographies are encouraged (D’Souza 1988).
D’Souza argues that social scientists should go over, in some detail, the harrowing events which precede mass starvation and make the world better prepared in future (D’Souza 1988). Famine ethnographies should also look at the immediate post-famine recovery period, since the details there can be just as harrowing as the story of the famine itself (Pottier and Fairhead 1991).
Famine ethnographies focus on detailing famine coping strategies and fine-tuning Sen’s entitlement theory (Sen 1981). This theory shows that hunger and starvation result from the loss of entitlement to food, rather than from a decline in food availability. Famines therefore run through predictable stages and give rise to predictable responses. Coping mechanisms, however, may change over time, which requires continuous monitoring. Besides documenting patterns and changes, famine ethnographies must reveal the heterogeneity of responses at the community level; showing, for instance, how gender and class and agricultural cycles combine to determine which responses are open to which groups or individuals. This heterogeneity partly explains why all famines are different.
Aware of the significance of entitlement, relief organizations focus on how and when to intervene. Discussion here is marked by contradictory views on the timing of emergency aid. The first view is that farmers adjust to weather fluctuations by accumulating assets during good times and drawing down stocks in lean years. Proponents of this view believe that officials habitually misread farmers’ ‘adjustments’ as distress signals and dispatch aid when it is not needed. The counterview contends that mortgaging land or liquidating productive capital to meet current needs are signals of acute distress and detrimental to recovery, so aid must arrive well before such desperate measures are taken.
To maximize the potential of food aid, anthropologists argue that donors should do more than simply throw grain at famine victims. Instead, emergency supplies must be utilized in ways that strengthen existing food channels and bolster the mechanisms for post-famine recovery. This could mean regulating livestock prices during and after famine through purchase-and-resale policies by government or outside agencies. Such policies can be effective in mitigating distress sales and enabling herds to be rebuilt, as Sperling reveals in her Samburu study of the 1984 famine and its aftermath (Sperling 1987).
Cases of positive intervention through learning suggest that improved flows of detailed information create the potential for improved targeting and long-term recovery.
In addition to micro-level studies of hunger, famine and famine relief, anthropologists have developed an interest in the impact of broad food aid programmes. An excellent example is Doughty’s study of how the ‘Food for Peace’ programme in Peru affected the lives of its beneficiaries (Doughty 1986).
Agricultural development and food policy
Agricultural policy interventions can have devastating consequences. This is attributed, within anthropology and increasingly beyond, to the fact that food policy design and implementation remain in the hands of agricultural advisers who fail to anticipate how their interventions will affect people’s ability to access food. Local people have virtually no say in the process that determines market prices and produce availability.
The pitfalls of policy approaches that ignore the interconnectedness of the major food domains are well illustrated in the regular observation that higher incomes as a result of cash cropping do not necessarily lead to nutritional benefits. The obstacle is a set of mitigating factors, e.g. the need or tendency to spend lump payments on consumer goods other than food.
The absence of an integrated approach is also characteristic of feeding programmes that assume that if women only knew what to do, they would deploy existing resources to feed their families better (Wheeler 1986). By isolating mothers and their children as a target-group, programme implementors avoid confrontation with the full range of structural factors that impede food supply at the household level. The importance of the critique is reinforced by contributors to McMillan’s Anthropology and Food Policy (1991).
The structural constraints on food security warn against taking an overly functionalist approach to food and resource sharing. The need for caution is well illustrated by Walu Engugu, who studied the activities of women traders in Kinshasa (Schoepf and Walu 1991). These women traders, who must trade in order to earn substantial additional incomes to secure for their households even modest living standards, all gave evidence of gender struggles over income and assets.
Being ideally placed to comment on household activities and organization, which is the level at which much of the food policy debate is pitched, anthropologists are stepping up their participation in the search for more effective food policies. The key to policy success, they argue, is that policy thinking must address the three main food domains (production, distribution, consumption) within a single framework of theory and action. Dealing analytically with food allocations at household level also requires dealing with aspects of authority and power, while setting the discussion in a wider economic context.
