Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The household focus in nutritional anthropology: An overview
Ellen Messer
Center for Advanced Study in the Behavioral Sciences, Stanford,
California, USA
INTRODUCTION
This paper will discuss some of the possible effects of various types of food policies on the domestic unit, in particular, on intra-household distribution of food. Anthropological and related social-historical studies in recent years have been examining in detail characteristics and distinctions between different types of "households" and "families" and evaluating their potential utility as units for analysis. By contrast, nutritional and health surveys using the "family" or "household" as their unit of record have in most cases not dealt with the complexities of domestic organization or considered how variable family and household social relationships affect nutritional and health outcome of domestic units. Social networks beyond the local co-residential group, which also share production and consumption activities relating to food, may affect the food intakes of individual members of such families or households. Economic, kinship, and other social rules and conditions, considered in conjunction with culturally appropriate attitudes toward giving and taking food, may furthermore affect composition of co-residential groups and who eats where, what, and with whom.
To begin to sort out some of these issues, this paper will first review the most recent summaries on the household- its structures and functions-as they might be relevant to nutritional studies. It will consider then how the household level of analysis has been included or excluded in the various branches of anthropology that deal either directly or indirectly with food and nutrition. Finally, it will suggest how the literature on the household, or at the household level of analysis, can contribute to the construction of more appropriate socio-cultural variables in food and nutrition policy studies, particularly those dealing with questions of the intra-household distribution of food.
ANTHROPOLOGICAL STUDIES OF THE HOUSEHOLD
Anthropologists in each of their subdisciplines and schools have been concerned with the definition as well as the functions of the household as the level of analysis intermediate between the individual and the society/culture. Since the origins of socio-cultural anthropological studies, the rules for food production, distribution, and consumption have formed an important part of the study of the organization of human groups, whether or not anthropologists focused directly on food quest and nutrition (1-3).
Whether taking as their unit of analysis
Yet how one is to study "the household level" has been a subject of ongoing controversy for anthropologists in all schools (4-7). Yanagisako (8) concluded her review of the literature on household and family with the observation that these are "odd job" terms that are useful for description but not productive as tools for analysis and comparison. Without careful clarification, the labels tend to obfuscate rather than clarify the comparability of domestic units, domestic relationships, and domestic activities from one socio cultural group to the next. Universal definitions of families are vague and seem to rely on the mother-child bond and the constellation of vaguely defined functionally related individuals around them. For example, the family is said to correspond to the unit of reproduction and socialization, i.e., a co-residential group that shares some activities. Alternatively, household usually refers to a group of individuals (sometimes only one) who live together and also share some form(s) of activities, usually "domestic activities," related to food production and consumption or sexual reproduction and child-rearing or some combination. Although most definitions of the household are based on notions of "propinquity," "kinship," or both, there is no reason to equate family ties or co-residence with household membership. Families may not form households, and households may not be composed of families. As an example, the corporate legal unit of the family, defined in terms of its rights to land, may be quite distinct from the common set of (usually) kin who live together and share a common set of obligations based on the other requirements set by co-residence. Yet the overlapping groups of people who participate in meal-sharing, gardening, other specified activities, and co-residence may be better described as a "domestic group" than a household or family. For such cases, one may be more interested in describing "that group of people, their relationships and activities, who acknowledge a common authority in domestic matters," a "budget unit," or "a group who have a common fund of material and human resources and rules and practices for exchange within it." Nutritional studies might more productively seek to identify and analyse such functional social groupings rather than simple co-residential or kinship groupings of men, women, and children.
Ordinarily, to understand domestic organization and the composition of residential groups in any society-culture- and their implications for the nutrition of that group and its individual members-one must examine marriage rules, residence rules, and the social and biological processes leading to cycles of shared residence, work, and consumption of individuals and domestic units. How is labour recruited in different cultures? What are the rules for transmission of property between and across generations? What are the work requirements of households and individuals? What are the rules for food-sharing? What are the decision-making units of individuals at various points in the flow from food production or purchase, through preparation, distribution, and finally, consumption?
What is the sexual division of labour and what is the economic participation of women? To answer these questions it is necessary to examine the seasonal, sporadic, or permanent nature of female economic participation and to try to distinguish between household organizations that provide "backup" household labour so that women can work outside the home and those in which the organization adjusts to the demand for women's labour by changing form and/or personnel. Cultural rules for social relations and child-rearing as well as cultural attitudes towards women's inhouse versus external work may help predict such adjustments On the other hand, cultural rules and attitudes about women's working may themselves be changing in response to economic conditions. In either case, nutritional studies should be concerned with how the changing organization of the household under different economic conditions will affect how food provisioning and other household maintenance tasks are accomplished.
One should also be concerned with how particular economic demands affect the emergence of the nuclear family. In many parts of the developing world, this is part of the more general problem of how socio-economic changes affect family and household structure. Following Yanagisako (8), it is probably more useful to begin by analysing the activities characteristic of domestic relationships and then turn to the structure and composition of the group involved in realizing a particular production/consumption style. The basic nutritional anthropological question is to determine the particular arrangements by which the tasks of living (eating, clothing, personal hygiene, and household maintenance) get done. These may involve more limited or more extended domestic groups, as well as different styles of domestic leadership and management, particularly in these domestic arrangements run by women. On this last feminist point, many anthropologists have tended to simply equate "domestic" with the female domain, but it is important to distinguish to what extent women do limit themselves to a passive role in the home sphere, or take an active role in securing their and their offspring's socio-economic wellbeing. Studies of women's extended kinship networks (particularly in urban settings) as an extra part of the domestic organization are beginning to show the limits of family and household analysis for socio-economic as well as nutritional studies (9). Also, mothering may not be the most important aspect of a woman's role, and women may have far more power in economic decisionmaking than ordinarily conceived by men (8, pp. 190-196).
Wenner-Gren Conference on Households
Along these lines, many of the domestic forms and functions observed in contemporary ethnographic settings may be in the process of change. A recent Wenner-Gren conference on "Households: Changing Forms and Functions" (10) emphasized this perspective in focusing on what households do, what functions they perform, and how and why they alter through time (rather than continuing to debate on how to classify household systems in terms of genealogy and co-residence types). Participants continued to look at task-oriented, culturally defined groups in terms of kinship networks and co-residence that help define size, composition, and functions of households but also considered the processes of change in these dimensions as households adapt to socio-economic (environmental) perturbations In lieu of further refining typologies, they addressed more questions of sampling within regions and communities to ascertain the range and variations within household groups, given the evolution of the domestic cycle; also, they noted the great flexibility that exists in most household systems through which households can respond to changed conditions without changing accepted cultural form and composition.
Several perspectives contributing to the discussion were cross-cultural comparisons of "native" (emic) versus "statistical" (etic) household constructs, and of cultural values-including gender ideologies-that lead to or limit options in residence and work patterns. Critical to understanding the process of work and consumption group formation as well as individual survival strategies is the issue of power within the household-for example, do men make the rules, but women make the decisions?- and the manner in which households and individuals are linked to other units in the larger society. A key question of nutritional concern in this regard is the strategy of resort in times of food scarcity: To whom do individuals-of given age, sex, kinship affiliation-turn in times of dearth for sustenance, or for work? Was there ever a rule-governed moral economy (11) that assured people of emergency assistance but is now breaking down? If so, can the process of change in domestic groups and extended social networks be described in either native or statistical terms? In situations of demographic transition, what are the adjustments that task-oriented units make to the outmigration of males, young people, or women of childbearing age? Alternatively, how does the push of insufficient employment and consumption standards within the local unit versus the pull of the city affect migratory decisions? What are the rules by which those who are no longer co-resident continue to contribute income to the home? At what point do they define themselves as no longer household members and, therefore, not subject to mutual work and exchange obligations?
These are some of the household concepts and processes currently under study, in general, by anthropologists trying to understand the household factor in ecological, social, economic, and cultural studies that are, in today's world, mainly studies of adaptation and change. While most anthropologists have tried to describe domestic organization as a phenomenon sui generis existing at one point in time, most-at least since the 1930s-have been unable to avoid setting their analysis in historical context. Many household studies, directly or indirectly, suggest how government labour and tax policies have influenced domestic structures and functions. Some also suggest how household functions favour or interfere with individuals' participation in certain government work and health projects. In summary, anthropological studies at the household level have addressed mainly theoretical, but also practical concerns. We turn now to a consideration of household level analyses, and the insights they potentially provide for nutrition and food policy.
THE HOUSEHOLD FOCUS IN ANTHROPOLOGICAL AND RELATED STUDIES OF FOOD
To help clarify how a household focus has or could be used in food and nutrition studies, particularly those related to intrahousehold distribution of food, I review in this section how social anthropological, psychological anthropological, symbolic, ethnoscientific, and ecological studies have incorporated this level of analysis in their food-related research.
Social Anthropological Studies
British social anthropologists, working in pre-World War II colonial Africa, provided some of our earliest and most complete studies of the interrelationships between food supply, social organization, and nutrition. Although as social anthropologists they were interested in defining and analysing the basic units of socio-political organization, they found that the study of food and hunger were central to their understandings of the societies, social relations, and changing cultures disrupted by British rule. Richards, in her classic study of the Bemba of Northern Rhodesia (12), concluded that the reasons natives did not work harder (a primary concern for British mining and other economic interests) was not a question of sloth but of undernutrition. In this society, the men had been drawn away from their roles in local gardening production to the mines, and the women found it difficult to perform the heavy clearing tasks traditionally assumed by males in addition to their own cultivation and gathering roles.
Richards focused on the social relations of food production and consumption, the construction and maintenance of consumption units through common household work, and rules and practices for food-sharing. She compared the effectiveness of different household heads (women responsible for the acquisition, storage, processing, preparation, and distribution of food) at common food related tasks, noting the range and average durations of typical tasks, such as the time it took to pound a kilo of grain into flour, to gather the other ingredients, and in total to draw water, gather fuel, and prepare and serve the meal. In addition, she recorded work schedules, food classifications, and typical diets of different age and sex groups within the population. She noted how kinship relations were marked by prescribed rules of food sharing and how these regulations of hospitality were systematically thwarted in times of dearth. In summary, she noted how production and consumption units were organized around women, now that the men were less in the production picture, and how insufficient energy intake-particularly in the planting season-by these women embedded the individuals and their households and community into an ongoing cycle of insufficient food, poor nutrition and health status, and poverty.
Other British social anthropological studies, while examining the household as part of the domestic cycle of social structure and social relations, also incorporated some comment on the variable use of food resources by different households, while not concentrating directly on food and nutrition. Fortes and Fortes (13), for example, reviewed the nutritional and productive practices of the Tallensi and stated that many households (indeed, the entire community) could do better agriculturally and nutritionally if they would spend as much time cultivating their cereal gardens as their tobacco patches and if they expended grain less liberally on beer festival socializing immediately after the harvest. Goody (14), in analysing the reasons for the break-up of polygynous (multiple wife) consumption units among the La Dagaa of the Gold Coast, relied on native explanations of child feeding customs to comprehend household fissioning. Not so much jealousies or personality incompatibility among co-wives but individual mothers' needs to supervise their children's food from separate rather than common pots caused households that formerly cooked and ate together to divide into mother-child units.
More generally, studies of polygynous households, mother child residence rules, and particularly post-weaning residence rules for children provide important information for designing nutritional studies that seek to elucidate the social factors contributing to malnutrition (15) Other African studies have shown time and again that who eats with whom may not be a simple matter of residence Not all persons who reside in a common homestead necessarily eat the same foods from a common pot. Tracing the routes of food across Ivory Coast villages can indicate the lines of kinship obligation, as well as the lines of infection stemming from food-sharing. Examining at the household level which children (age, sex, birth order) eat together in a household and rules for sharing (or adult intervention in cases of inequities) can indicate further social factors in malnutrition where it is the case that youngsters grab what they can from a common pot.
Anecdotal information from social anthropologists provides some information about such eating customs that systematically neglect the very young and those unable to fend for themselves. Anecdotal reports also often indicate substantial variation in how mothers manage work in the fields and child care simultaneously. For example, one anthropologist noted distinct variations in mother-craft. Within the same culture, some mothers methodically placed and kept their young children in the shade and fed them at regular intervals with easily available vegetable and fruit resources, while others let their children linger hungrily in the sun most of the working day. More systematic reports on some of these factors might provide clues on what features of households and eating patterns, as well as women's work patterns and rules for child care, to seek in nutritional studies. Also, how available resources already cleverly used by some individuals within a culture might be taught to all to further their common nutritional well-being is an additional study area suggested by these reports (anecdotal communications provided by Timothy Weiskel for the Ivory Coast, and Jane Guyer for the Southern Cameroons).
Psychological Anthropological Studies (Cultural Food Habits)
By contrast, cultural anthropology in the United States, during the 1930s and 1940s and subsequently, focused less on social structure and more on how attitudes toward food (and other aspects of culture) were socialized in children at an early age, and affected their later social relationships, behaviour, and personalities. Studies like DuBois's classic ethnography on the Alorese (16) considered, among other things, the psycho-social nutritional dynamics between mothers and children and between older and younger children as part of the investigation of the structure and content of relations between males and females and personality formation. She focused particularly on the male-female units of mother-son and wife-husband and the sources of food anxieties, including behavioural consequences of such imaginary food shortages on those elemental social relationships. Other psychological studies also included complete descriptions of social relationships centering around food as part of their analyses of culture and personality, drawing heavily on primary data of food sharing at some specific household level (17, 18).
More focused food-habit research within the United States during the 1940s, while aimed in part at understanding the formation of food habits as a part of culture (and personality), nevertheless reported less on this household level of observation in their studies of the food patterns of different ethnic populations within the United States. In addition to improving academic understanding of food as an aspect of culture, these studies were designed to clarify how cultural rules of food classification, dietary structure, and rules for food-sharing, as well as the channels by which food reached households, influenced adequate nutrition within cultural communities and how food habits could be changed (19-23) While it is useful to go back to some of these early studies that describe how food attitudes are formed during the early years of childhood and under what circumstances of socialization they are easily amenable to change, for the most part they neglect household observations in favour of generalization based on respondent reports in their summaries of the food habits of individual ethnic groups
Subsequent literature on cultural food habits, particularly studies done of the food habits of Third World societies, have been similarly lax in reporting what goes on at the household level in terms of food-habit socialization or the contexts in which nutritional habits are engrained. Most of these surveys provide some notion of the major elements of the diet (e.g., primary and secondary carbohydrate energy source, principal vegetable protein elements that complement the starchy staple) along with simple questions of who eats them (everyone, only men, only women, only adults, only children, only women and children) and of who eats together and who gets served first; but they do not go far enough in indicating how food is distributed within the household. These observations are particularly necessary to verify such responses as, for example, that children eat from a common pot and their implications for child nutrition (24). Nor do the average figures in which such food habits are reported allow for interpretation of individual household patterns that may be nutritionally superior within the general cultural food pattern. Finally, such studies give few insights into how food habits might be changed, given shortages of certain items or poor nutritional quality of the diet of certain status groups within the society (e.g., weanling infants, adolescent girls). The short summaries of the ethnic food habits of the United States in the 1940s, which include descriptions of the social networks of the ethnic groups addressed, go farther in suggesting practical measures in this area.
Cultural Symbolic and Cognitive (Ethnoscientific) Studies
Within cultural anthropology, symbolic and cognitive (ethnoscientific) studies have also contributed to our understanding of food habits; on the one hand, to studies of the attributes of particular foods/crops that enter into their classifications and rankings as food; on the other, to our comprehension of why particular social categories (e.g., groupings based on age, sex, or economic class) in particular cultures favour or reject particular foods or classes of food. Food with its multiple referents can be used or interpreted to symbolize (mark) social group membership, relative social status and social relations, prevailing dominant subordinate relationships, or relative class or national position; symbolic anthropologists study the dimensions and resulting structures of contrasts.
Many of the studies of food symbolism have provided information on the intra-household distribution of food in the course of examining symbolic food classifications, such as "hot cold," "male-female," and "pure-impure" principles that structure acquisition, preparation, and distribution of foods in particular cultures. Such principles determine, in large part, whether access to certain classes of foods are restricted to certain social categories, such as children; menstruating, pregnant, or lactating women; or victims of particular illnesses. In each case, the food classifications, and the knowledge and attitudes that different members of the population have about them, can affect nutrient intakes and food allocations within households. Symbolic analyses of how people fight or of how they certify friendship or particular degrees of social relations with food also form parts of symbolic studies furthering information on intra-or inter-household food flows.
Less available are investigations at the household level identifying more precisely how people acquire particular food rules, or even more basic, particular tastes, taste preferences, and general ideas about edibility of food items. The extent to which such knowledge and rankings are shared within and between households or more extended kin networks demands further work. Among the areas for additional research at the household level are taste sensitivities and preferences within and between cultures, and their nutritional and health consequences. For example, Messer's study of sweetness preferences and consequent sugar intakes in a Mexican village (25) suggests that there may be substantial but describable variations in both, even within a single community. In view of modern medical concerns about high refined sugar intakes and diabetes, more careful studies of sweetness preferences and other cultural factors contributing to high sugar intakes are in order. Jerome (26), among others, has discussed some of the cross-cultural developmental issues in children's acquisitions of cultural tastes. Biological versus socio-cultural sources of preferences for higher or lower concentrations of sweetness and saltiness, as well as preferences for sourness, remain controversial topics for further nutritional investigation.
Similarly, there have been few ethnoscientific studies that explore native views specifically dealing with native food choices and with the dimensions or objective consequences of folk nutritional concepts and practices. DeWalt et al. (27) attempted to identify folk nutritional concepts and categories in one Mexican rural community study, but their food factors-arrived at by their own analysis of the preliminary data on food selections and preferences-were more their own than the people's categories. Potentially, ethnoscientific studies of diet could have informants sort elements into culturally meaningful dietary categories such as starchy staples and relishes; elicit the criteria, including costs, by which individuals, households, and larger cultural groupings choose among them; and evaluate the nutritional results of such choices.
Theophano (28) and Curtis (29), in their studies of Italian American food habits, analysed patterns of food consumption and inter- and intra household food exchange, utilizing categories of food (pasta, gravy) supplied by the people themselves. Their studies illustrate the potential usefulness of this approach for describing current food habits and tracing cultural change. Douglas (30) and Douglas and Gross (31) have also suggested structural approaches to describing food consumption and exchange patterns among households, using computer programs.
Their methods need further empirical examples, however, before they can be said to be useful.
Characterizing each of the food studies in these areas of food classification and symbolism is a need to combine interviews on food habits, classifications, and attitudes with observations of actual consumption before the studies can be said to be useful for nutritionists. Conversely, observations in nutrition surveys, for example, vitamin A deficiency epidemiologically limited to young males in South East Asia (32), can sometimes lead the cultural anthropologist to prevailing cultural symbolic rules of potential nutritional deficit for a particular segment of a population, even in cases where people report an "old rule" they claim they are no longer following. On this note, one should add that there are usually multiple factors that govern whether or not particular food symbolism dominates decisions on food allocations within the household. In Bangladesh, for example, Rizvi (33) investigated whether the cultural rule for favouring male over female children in matters of nutrition and health, as well as other cultural concerns, was standard practice across social groups and classes. She found, not surprisingly, that other factors, such as birth order, number of children, and economic position of the household, were also important in determining intra-household food distributions. Again, combinations of interview and observation provided more accurate results than simple questions about attitudes toward the two sexes.
Ecological and Biocultural Studies
Complementing analyses by ethnoscientists and symbolic anthropologists interested in how "culture comes from culture" have been studies by cultural materialists (34), ecological anthropologists (35, 36), and nutritional anthropologists (37) that have been more explicitly concerned with understanding the nutritional and environmental (including socio-economic) determinants and consequences of cultural practices (food habits) in material (scientific) terms. From the cultural materialist perspective, environmental conditions and subsistence needs are seen as directly or indirectly shaping cultural practices that maintain the material bases of society. In this view, culture-including social organization and symbolic and other aspects of cultural attitudes toward food- consciously or unconsciously serves ecological and economic ends. The units of analysis are objective measurements, sometimes nutritional such as quantities of energy (calories), protein, and other nutrients in foods. They are sometimes ecological, relating species within a local or regional ecosystem, and in other cases economic, involving financial calculations.
Most of these studies have dealt with culture, society, and the human population at the group rather than the household level. However, household analysis has been mentioned and is intrinsic to most descriptions of how subsistence systems are organized. For example, studies of the environmental resources, cultural food classifications, and patterns of food and information sharing among hunters and gatherers, such as the !Kung San of the Kalahari (38), have shown how the adaptations and food systems of even the simplest human societies involve a complex sexual division of labour and relatively precise knowledge about the locations and seasonality of upwards of eighty plants carefully classified into preferred and less preferred species. Such information is carefully taught, often within same-sex groups (39-41). The organization of hunting tasks, including formation of co-operative groups, manufacture of tools, and time planning, is another aspect of hunting and gathering economies and social life that affects kinship and co-residential groups and the food intake of individuals. Water availability and the relative resource positions of other social units from which one can expect to give and receive food, other material goods, and friendship are additional aspects of the !Kung San environment that affect individual and household level nutrition. Between the level of the individual and the social group (human population), anthropologists more often discuss social networks based on age, sex, and kinship or males and females of different ages, than households, although the shift to household analyses may come as more and more re-studies of the !Kung take place on their newly settled reserves, where they consume a processed food diet.
Ecological studies
Most ecological anthropological studies at least describe the division of labour and consumption patterns at the household level, even if the adjustments that take place at this level do not enter very much further into their analyses. Rappaport (42), for example, considered the sexual division of labour and potential conflict between pigs and people at the household (gardening) level in his classic study of the role of ritual regulation of the pig populations kept by horticultural groups in the highlands of New Guinea. He offered a model that considered women's complaints that too many pigs were digging up their sweet potato gardens as the triggering mechanism for a wholesale ritural slaughter of pigs to complete the pig population's life cycle. Ecological studies of energy flow, that is, how energy inflow (intake) matches energy outflow (expenditures) within a given biological unit (regional ecosystem, local ecosystem, household, individual), have also examined the division of labour by sex and age at the household level to understand better how local human populations get by on limited resources. Thomas's classic study of highland Andean herding patterns (43) carefully measured how children's work saved the group energy. Brush (44), in another Andean location, considered how the labour exchange patterns among households enable that socio-economy to subsist, documenting labour exchange within but also between co-residential and kinship units. None of these studies, however, has devoted equal time to a careful documentation of energy intake between and within households,
Biocultural studies
Biocultural studies, trying to understand the behavioural consequences of particular cultural health and nutrition practices, have documented how households adjust to nutritional deficiencies, such as iodine lack. Greene (45, 46) demonstrated how iodine-deficient communities in highland Ecuador managed to perpetuate themselves even with a high proportion of the population clinically impaired by varying degrees of endemic goitre through extreme cretinism; he showed that the intellectual and physical requirements of ordinary tasks for living had been so reduced that even mild cretins could perform many of them and were thus productive individuals.
Biocultural studies of psychological disturbances in certain classes of societies (e.g., women) have also dealt incidentally with intra-household distribution of food, in that individual studies attribute some of these syndromes to protein or calcium deficiencies, brought about by food rules that discriminate against women (47). However, these studies, like the symbolic studies mentioned above, have never been successful in actually measuring insufficient intakes or the clinical nutritional deficiencies in the individuals supposed to be suffering from culture bound syndromes.
ANTHROPOLOGICAL AND RELATED STUDIES OF NUTRITION AND ECONOMIC DEVELOPMENT
Among the topics that have been investigated for purposes of advancing economic development in developing nations have been the social organization of production at the household level and its relation to nutrition, health, and fertility. In particular, the role and nutritional consequences of women's work in food production, preparation, and distribution and women's time allocation to household maintenance and child care versus cash employment have been dealt with in any number of policy studies (48). At the household and community levels, the relationships between nutrition, productive behaviour, and reproductive behaviour have been reviewed by public health economists in a growing literature on women's and children's work, time allocation, and possible implications for nutrition (49) and fertility decision-making (50). Some studies seem to show that the extra income of the mother may be the main economic factor accounting for lower levels of malnutrition within populations where some mothers work as traders (51). Other studies suggest that the mother's time away from household responsibilities, in the absence of adequate supplementary domestic and child-care arrangements, may account for the poorer nutritional status of the child. This disadvantageous nutritional outcome may stem from economic demands on the mother's time, which leave her inadequate time to prepare a balanced diet to meet children's nutrient needs (49, 52). Alternatively, the child may suffer hunger, exposure, and illness while the mother works, and therefore be at greater nutritional risk than the child of a woman who is not working (52). As indicated above, the particular skill of the mother at managing work and child care, along with the cultural mechanisms for providing surrogate mothers of greater or lesser quality, are also significant factors in nutritional and health outcomes, and there may be significant intracultural variation in these matters.
Factors Determining Intra-household Food Distribution
The related questions of intra-household distribution of food under different working conditions for male, female, and juvenile household members have also been considered. Gross and Underwood (54), in their classic study of energy flow among sisal workers in Brazil, showed that male wage earners received preference in the allocation of calories within the household. They were fed first, in sufficient quantity to sustain their work, often at the expense of children and women, who received inadequate calories, eating what was left after the male's energy needs had been met. The impact of women's work and income on the nutrient intakes of household members, either because women allocate more of their income directly to the food budget or because they make the intra household distribution of food more equitable when they contribute to the food budget, demands further study.
Recent evaluations of the significance of agricultural development projects for women's lives and the nutrition of households indicate that the socio-cultural factors determining whether and where women work in cash crop agriculture and the effects of such labour-force participation on their household diets are complex. Most agricultural cash crop projects seem to favour male over female enterprises, even where the reasons this is so have to do with the complex of socio-cultural rules governing women's access to land, labour, and technology. For each project, one must examine in great detail what activities (including food producing activities) women forego by participation: Do they benefit or lose in terms of total income (what was the opportunity cost of work foregone, including income in kind)? Even if income is greater under the development project, do women control their earnings to the same extent that they did under the previous household labour arrangements? Finally, are women able to choose to allocate more income to food of higher nutritional quality under the new economic conditions?
Participation of Women in Development Projects
All of these issues-health considerations of child care notwithstanding-will affect whether women choose to participate and, on balance with their offspring, benefit nutritionally from a development project designed to improve production and, by implication, their income and well being. Any arrangement that decreases the amount of control women have over the returns for their labour (e.g., where they work in someone else's field rather than their own to produce the same crop income) donates some of the value of their labour (product) to another owner/manager. Similarly, if purchased food is always higher in price than food produced at home, then there must be compensation in women's income proportional to the relative food price differences to make up the difference in food available under the new versus the previous system. Particularly in farming systems where women work some fields while their husbands control others, projects must be extremely sensitive to the possible lowering in value of food that women can produce or buy with income earned at opportunity cost to home food production and recognize how such considerations interfere with women's desire to participate in development schemes and whether they should participate.
Also along these lines, one should try to anticipate seasonal effects on diet of particular patterns of labour force participation. Are women used to earning a small amount per day or week, say, through food processing? If so, plans to automate food processing will interfere with their daily economic earnings, and most likely, with the nutritional wellbeing of their offspring who count on their earnings for food. Are the crops that women garden ordinarily sold in small amounts; again, to meet immediate food budget needs? If so, then any scheme that concentrates earnings from crop harvests, even if the earnings are superior, will interfere with ordinary nutritional budgets in the absence of other kinds of steps to inculcate nutritional and economic planning. This is not to say that expenditure patterns and the division of labour in economic production and household decision-making, particularly in the area of the food budget, do not change over time. However, before a project is initiated, there has to be a careful study of the resources that various members of the household control at the outset of the project and how the project might interfere with that household division of resources and subsequent decision-making patterns (see Jones [55] for a relevant case study). While the study does not provide adequate information on actual effects on food budgets as they affect nutrient intake, it does provide information on the complex of factors involved in the sexual division of labour )
In general, studies of women's participation in the labour force have shown the need to consider the varying ways in which income from different sources is acquired and used and the importance of seasonal and other fluctuations (55). However, the nutritional implications of sexual division of labour and decision making related to income allocation are rarely drawn except by Cosminsky and Scrimshaw (57).
Socio-cultural Factors in Malnutrition
A final area that incorporates household analysis are studies, usually by nutritionists, on epidemiological factors in malnutrition. In particular, they have tried to isolate factors in residence rules, infant-feeding customs, and mothering (patterns of child care) that seem to contribute to higher or lower frequencies of malnutrition in children of specific age groups within the same culture or culture area. Household factors, material or ideological, that contribute to the decision to breast-feed or not (58) and household and extended family support networks that help a mother with nursing and other aspects of child care are among the household variables described in such instances. Feeding customs following or in addition to breastfeeding, the mother's use of time-how much time she has to spend with the child-and how well she supervises the child's nutrient intake and health also affect the child's nutritional status and health. See Popkin (49) for a review of this literature.
Weaning, child health, and behaviour
The time immediately following weaning is usually particularly important. Physical separation from the mother, as well as inadequate food, may contribute to cultural patterns of failure to thrive (59). The social organization of food provisioning and child care enter into the equation of whether the child thrives. Important issues are who feeds the child, how often, and in what quantities, what types of food are given, and who meets the child's emotional needs. How the mother manages her time, her money, and her children are aspects of mothercraft that contribute to some children's thriving and others failing to thrive in apparently the same environments (60-64).
The quantity and quality of parent-child interactions within household units are additional social factors affecting the child's nutritional status and well being, although the relative contributions of the extra stimulation versus the extra nutrient intake are not clear. For example, intensive observations in a Mexican intervention study indicated that infants who received nutritional supplementation were substantially more active and interactive with their environments (including with their mothers and fathers) than their unsupplemented counterparts (65), although the effects of the extra stimulation in addition to the food provided by the programme have contributed to this outcome.
Local understanding of and responses to moderate nutritional disorders and clinical nutritional disease also affect nutritional and behavioural outcome. In many cases, child-feeding customs and food classifications contribute to clinical deficiencies like protein-energy malnutrition and avitaminosis-A (15). Also, general sanitation conditions and particular sources of infections within the household or larger compound where a child traverses have impact on his nutritional well-being. Interactions among all economic, socio-cultural, and disease factors further condition how the child fares in nutrition, health, and other interrelated aspects of behavioural development (65). While there have been numerous studies attempting to link levels of childhood malnutrition with particular socio economic factors, usually in household environments, simply quantifying the socio-economic factors within households may not be sufficient to understand the relationships between social factors and malnutrition. Rather, it is the way the various resources are organized, how adults manage their time in relation to child rearing, and how a child learns to fend for himself within such environments that determine the nutritional outcome, measured in physical and functional terms.
DISCUSSION
Each of these anthropological or related studies suggests what investigation at the social level between the socio cultural community or subcommunity and the individual can show us about individual and group nutrition. Prior to initiating a nutrition survey or nutrition intervention, one should begin by defining the group and its members-by co residence, kinship ties, tasks, food exchange, or some combination of relevant factors. By next investigating how such units conform to structural rules-and as Arnould and Netting point out (10), it may be the "flexibility" rules that are the most significant-one can chart the different patterns in social organization and resource use within cultural groups. From these patterns, one can begin to discuss how they facilitate the interpretation of the dynamics of eating patterns, eating habits (including socialization), and other aspects of household functioning that either favour or interfere with nutrition and health.
Studies of eating and budgetary groups are essential for understanding social aetiological factors in malnutrition. Along with child rearing units, they provide a context in which to understand the impact of socio-cultural food ideologies-and nutrition education-on child nutrition. It may be, given the social organization, that the individual mother's ideas are not entirely responsible for the food the toddler eats and the health environment in which he/she toddles, although most nutrition programmes are aimed at improving the mother's nutritional capabilities. Studies of the organization of resources according to eating units, food budget units, and child rearing units are also critical to ascertain where and how a mother's work for income benefits or hampers her child's food intake, nutrition, and general health and educational wellbeing.
To summarize, this discussion suggests that nutritional anthropologists and other medical and social scientists examining social factors in nutrition revise their concept of the household as a focus in nutritional studies, and give greater emphasis to eating, budget, and child-rearing groups. On the basis of current household studies, anthropologists can improve our ability to examine kinship and co" residential factors that contribute to the formation of these groups as they relate to other functional activity groups within a society and indicate how such patterns contribute positively or negatively to nutritional outcome. The task remains to incorporate such perspectives on domestic organization into the social factors standardly conceived by nutritionists and policy planners trying to ascertain current nutritional conditions and aetiological factors in the social fabric and how such conditions contributing to malnutrition might be modified in particular communities or social groups.
Additionally, social historical data on forms and functions of household units can help policy planners anticipate the adjustments that might be made at the household level to particular kinds of economic development initiatives, given structure, function, and flexibility/formation rules of a given society at a certain point in time. Socio-economic changes taking place in both rural and urban environments indicate a pressing need to understand how men, Women, and children allocate their productive time and arrange social obligations to ensure subsistence. Policy planners also need to know the contexts in which people in different cultures learn the nutritional values of different foods and how to care for themselves and others in order to design nutritionally effective policies.
REFERENCES
1. F. Bell, "The Place of Food in the Social Life of Central Polynesia," Oceania, 2:117-135 (1931-1932).
2. R. Firth, Primitive Economics of the New Zealand Maori (George Routledge and Sons, London, 19291.
3. R. Firth, We, the Tikopia: A Sociological Study of Kinship in Primitive Polynesia (London,1936).
4. D. Bender, "A Refinement of the Concept of Household: Families, Co-residence, and Domestic Functions," Amer. Anthro., 69: 493504 11967).
5, J. Goody, "The Evolution of the Family," Household and Family in Past Time (Cambridge University Press, Cambridge, UK, 1972).
6. R. Rapp, "Family and Class in Contemporary America: Notes toward an Understanding of Ideology," Science and Society, 42: 278-300 11978).
7. R. Sanjek, "The Organization of Households in Adabraka: Toward a Wider Comparative Perspective," Comp. Studies in Society and His. 24: 57-103 (1982).
8. S. Yanagisako, "Family and Household: The Analysis of Domestic Groups," Ann. Rev. of Anthro., 8: 161 -206 (1979).
9. S. Yanagisako, "Women Centered Kin Networks in Urban Bilateral Kirnship," Amer. Ethnol., 4: 207-226 (1977).
10. E.J. Arnould and McC. Netting, "Households: Changing Form and Function," Current Anthro., 23:571-575 (1982).
11. J. Scott, The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia (Yale University Press, New Haven, Conn., USA, 1976).
12. A. F. Richards, Land, Labour, and Diet in Northern Rhodesia: An Economic Study of the Bemba Tribe (G. Routledge and Sons, London, 1939).
13. M. Fortes and S. L. Fortes, "Food in the Domestic Economy of the Tallensi,'' Africa, 9: 237-276 (1936).
14. J. Goody, "The Fission of Domestic Groups among the Lo Dagaba," in J. Goody, ea., The Developmental Cycle in Domestic Groups (The University Press, Cambridge, UK, 1958), pp. 53-91.
15. A. Burgess and R. F. Dean, eds. Malnutrition and Food Habits (Macmillan, New York, 1962).
16. C. Dubois, "Food and Hunger in Alor," in Leslie Speir, A. Irving Hallowel, and Stanley S. Newman, eds., Language, Culture, and Personality: Essays in Memory of Edward Sapir (Sapir Memorial Publication Fund, Menasha, Wis., USA, 1941), pp. 272-281.
17. R.F. Fortune, Sorcerers of Dobu (Dutton, New York, 1932, 1963).
18. A.R. Holmberg, Nomads of the Long Bow (Smithsonian institution Press, Washington, D.C., 1950).
19. M. Mead, "The Social Psychology of Food Habits," in A. Burgess and R.F. Dean, eds., Malnutrition and Food Habits (Macmillan, New York, 1962), p. 77.
20. M. Mead, "Food Habits Research: Problems of the 1960's," National Academy of Sciences, National Research Council Publication no. 1225 (Washington, D.C., 1964).
21. Committee on Food Habits, "The Problem of Changing Food Habits," National Research Council Bulletin no. 109 (Washington, D.C., 1943).
22. Committee on Food Habits, "Manual for the Study of Food Habits," National Research Council Bulletin no. 111 (Washington, D.C., 1945).
23. E, Montgomery and J. Bennett, "Anthropological Studies of Food and Nutrition: The 1940's and 1970's,'' in W. Goldschmidt, ea., The Uses of Anthropology, Special Publication of the American Anthropological Association 11 (1979).
24. M. Vemury, Rural Food Habits in Six Developing Countries: A CARE Study on Environmental, Social, and Cultural Influence on Food Consumption Patterns (CARE, New York,1981).
25. E. Messer, "Some Like It Hot" (paper presented at the 78th annual meeting of the American Anthropological Association, Cincinnati, Ohio, USA, Nov. 1979).
26. N. Jerome, "Diet and Acculturation: The Case of Black American Immigrants," in N. Jerome, R. Kandel, and G. Pelto, eds., Nutritional Anthropology (Redgrave, Pleasantville, N.Y., USA, 1979).
27. K. M. DeWalt, P. B. Kelly, and G. H. Pelto. "Nutritional Correlates of Economic Microdifferentiation in a Highland Mexican Community," in N. Jerome, R. Kandel, and G. Pelto, eds., Nutritional Anthropology (Redgrave, Pleasantville, N.Y., USA, 1979).
28. J. Theophano, "It's Really Tomato Sauce but We Call It Gravy: A Study of Food and Women's Work among Italian American Families" (Ph.D. dissertation, University of Pennsylvania, Philadelphia, Pa., USA, 1982).
29. K. Curtis, "I Can Never Go Anywhere Empty Handed: Food Exchange and Reciprocity in an Italian-American Community" (Ph.D. dissertation, Temple University, Philadelphia, Pa., USA, 1983)
30. M. Douglas, Purity and Danger (Penguin Books, Baltimore, Md., USA,1966).
31. M. Douglas and J. Gross, "Food and Culture: Measuring the Intricacy of Rule Systems," Soc. Sci. Info., 20: 1-35 (1981).
32. M. S. Van Veen, "Some Ecological Considerations of Nutritional Problems on Java," Ecol. Food and Nutr., 1 :25-38 (1971).
33. N. Rizvi, "Socio-economic and Cultural Factors of Intrahousehold Food Distribution in Rural Bangladesh" (paper presented at the 80th annual meeting of the American Anthropological Association, Los Angeles, 1982).
34. M. Harris, The Rise of Anthropological Theory (Thomas Y. Crowell, New York, 1968).
35. J. Steward, Theory of Culture Change (University of Illinois Press, Urbana, III., USA, 1955).
36. M. Vemury, Rural Food Habits in Six Developing Countries: A CARE Study on Environmental, Social, and Cultural Influence on Food Consumption Patterns (CARE, New York, 1981).
37. N. Jerome, R. Kandel, and G. Pelto, eds., Nutritional Anthropology (Redgrave, Pleasantville, N.Y., USA, 1979).
38. R. Lee and I. DeVore, Man the Hunter (Aldine Press, Chicago, 1968).
39. R. Lee, "What Hunters Do for a Living; or, How to Make Out on Scarce Resources," in R. Lee and 1. DeVore, eds., Man the Hunter (Aldine Press, Chicago, 19681, pp. 30-48.
40. R. Lee, "! Kung Bushman Subsistence: An Input-Output Analysis," in A. P. Vayda, ea., Environment and Cultural Behavior (Natural History Press, Garden City, N.Y., USA, 1969).
41. P. Draper, "! Kung Women: Contrasts in Sexual Egalitarianism in Foraging and Sedentary Contexts," in R. Reiter, ea., Toward an Anthropology of Women (Monthly Review Press, New York, 19751, pp. 77-109.
42. R. A. Rapport, Pigs for the Ancestors (Yale University Press, New Haven, Conn., USA, 19671.
43. B. Thomas, "Human Adaptation to a High Andean Energy Flow System," Occasional Papers in Anthro. no, 7 (Department of Anthropology, Pennsylvania State University, University Park, Pa., USA, 1973).
44. S. Brush, Mountain, Field and Family: The Economy and Human Ecology of an Andean Valley (University of Pennsylvania Press, Philadelphia, Pa., USA, 19771.
45. L. Greene, Nutrition and Behavior in High/and Ecuador (University Microfilms, Ann Arbor, Mich., USA, 19761.
46. L. Greene, ea., Ma/nutrition, Behavior, and Social Organization (Academic Press, New York, 19771.
47. A. Kehoe and D. Giletti, "Women's Preponderance in Possession Cults: The Calcium Deficiency Hypothesis Extended," Amer. Anthro. 83: 549-561 (1981).
48. A. Cowan, ea., The International Conference on Women and Food, 3 vols. (Consortium for International Development, US Agency for International Development, Washington, D.C., 1978).
49. B. Popkin, "Time Allocation of the Mother and Child Nutrition," Ecol. Food Nutr., 9: 1-14 (19801.
50. R. Evenson, "Food Policy and the New Home Economics," Food Policy, 3: 180-193 (19811.
51. R. Tripp. "Farmers and Traders: Some Economic Determinants of Nutritional Status in Northern Ghana," J. Trop. Pediatr., 27:15-22 (1981).
52. B. Popkin and F. Solon, "Income, Time and the Working Mother and Nutriture," J, Trop. Pediatr. Environ. Child Health, 6: 156-166 (1976).
53. S. Kumar, "Role of Household Economy in Determining Child Nutrition at Low Income Levels: A Case Study in Kerala." Department of Agricultural Economics, Cornell University, Occasional Paper no. 95 (Ithaca, N.Y., USA, 1977).
54. D. Gross and B. Underwood, "Technological Change and Caloric Costs: Sisal Agriculture in Northeastern Brazil," M. Anthro. 7: 724-740 (1971).
55. C. Jones, "The Impact of the SEMRY I Irrigated Rice Production Project on the Organization of Production and Consumption at the Intrahousehold Level" (paper prepared for the US Agency for International Development, Washington, D.C., 1983, Xeroxed).
56. J. Guyer, "Household Budgets and Women's Incomes," African Studies Conference, Boston University, Working Papers, no. 28 (1980).
57. S. Cosminsky and M. Scrimshaw, "Sex Roles and Subsistence: A Comparative Analysis of Three Central American Communities," in C. Loveland and F. Loveland, eds., Sex Boles and Social Change in Native Lower Central American Societies (University of Illinois Press, Urbanana, III., USA, 1981).
58. G. Pelto, "Perspectives on Infant Feeding: Decision-Making and Ecology," Food Nutr. Bull, 3 (3): 16-29 (1981).
59. R. Dean, "Malnutrition in Its Setting," in A. Burgess and R. Dean, eds., Malnutrition and Food Habits (Macmillan, New York, 1962), pp. 19-29.
60. E. Pollitt, "Behavior of Infant in Causation of Nutritional Marasmus." Amer. J. Clin. Nutr., 26: 264-270 (1973).
61. J. Cravioto and E. DeLicardie, "Microenvironmental Factors in Severe Protein Calorie Malnutrition," in Nevin Scrimshaw and M. Behar, eds., Nutrition and Agricultural Development (Plenum Press, New York, 1976).
62. M. A. Kerr, J. Bogues, and D. Kerr, "Psychological Function of Mothers of Malnourished Children," Pediatrics, 62:778-784 (1978).
63. D. Franklin and I. Valdes, "Desnutrición Infantil y Su Relación con el Tiempo y las Habilidades de las Madres," Separate de Cuadernos de Economía, 16 (49) 343-358 (1979).
64. E. Waldmann. "The Ecology of the Nutrition of the Bepedi, Sekhukuniland," in Food, Ecology and Culture (Gordon and Branch Science Publishers, New York, 1980).
65. A. Chávez and C. Martinez, Nutrición y Desarollo Infantil (Nueva Editorial Interamericana, Mexico City, 1979).
66. M. Herrera, J. Christiansen, N. Ortiz, J. Clement, L. Vuori, D. Waber, B. De Paredes, and M. Wagner, "Effects of Nutritional Supplementation and Early Education on Physical and Cognitive Development," in Life-Span Developmental Psychology Intervention (Academic Press, New York, 1980), pp. 149-184.