Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

A comprehensive interview approach to food patterning

Three studies of meal-format rules
Actual meal data

Cussler and DeGive's (1952) classic study of food patterning in a county in Georgia is extraordinarily rich and comprehensive. The study involved a large number of families each of whom was interviewed for 13 days. Interviews were divided into 13 topical areas, and each area was given a day. The interviews were loose and open-ended, allowing informants to talk freely. This open-ended feature is evident in the lengthy sample transcript provided.

Topics included: general food practices, nutritional knowledge and beliefs, communication, economic v. creative influences, the influence of class and race on food habits, attitudes toward government programmes, socialization, effects of cultural contact, and effects of the war. For example, the detailed description of the suggested questions for one topic was as follows (pp. 195-196).

The Influence of the Social Structure on Food Habits

What foods are considered ``White folks' food? What foods are considered ``Negroes' ''food? What foods are considered "fit only for pigs"? What foods are considered country" foods, town" foods (to designate one's rise in status from rural to urban life?) What attitudes toward such diets are held by the respective groups? Are certain foods eaten by the respective groups out of a sense of obligation to their status? Is there shame attached to certain foods and practices, for instance, fat meat and communal use of the dipper in the water bucket?

Is pride felt about certain other foods and practices? For instance, does the farmer consider his food habits healthier? Is he proud of his country hams," home-grown wheat. etc.?

What is the status pattern within the family? Who gets served first, and who gets choice pieces of food- father, children. youngest child, oldest son, etc?

How much and what kind of neighborliness is practiced with regards to foods and feeding activities?

Do neighbors help with the plowing, harvesting, planting and caring for a garden, preserving of foods, hog-killings, syrup-making, getting the grain to the mill, etc.? Do they lend machinery for these activities willingly? Is there much borrowing of foodstuffs, and are they lent willingly? Do neighbors give away or swap surpluses often? What recipients are preferred by them Whites, Negroes, hard-working people, shiftless people, destitute people, widows, etc.?

What is the landlord's attitude toward his tenants" food habits and nutritional status? Is he concerned, does he try to help, is he indifferent, does he actively oppose better nutrition for them?

The voluminous data (27 pages were used to display one family's interview transcript) provided a variety of different analyses. Typical daily menus (item lists) were described for three classes - wage labourer, tenant/share-cropper, and landowner). Seasonal deviations as well as class and racial deviations were described.

Two problems are evident in this study. First, the data were presented with either too little generalization or else too much. There is only a simple listing of responses, and as we can see from the data presentation example that follows in table 2, the answers appear idiosyncratic and unrelated. There is no indication that these are group-shared ideas about food practices. Second, when generalized composites are presented (table 3), we are not given any idea of their frequency, representativeness, or variability. Another concern is that this study was based entirely on stated ideals. There was no verification that these stated rules were actually behavioural practices. However, the study demonstrates the kinds of insights possible when people are given opportunities to talk. This kind of interview with fewer informants followed by an informed survey would have been more efficient and revealing.

Indirect Techniques in Studying Food Choices

Thomas Fitzgerald (1976) studied food choices in a village of 600 in North Carolina. using a "Food Choice Game," which involved Q-sorts among 100 picture cards, each depicting a local food item. This technique was used in conjunction with other more ethnographic observations and interviews. Foods were sorted according to three dimensions: like most/least, healthy/unhealthy, and frequently used/infrequently used. Fitzgerald states that using such a game is less threatening to informants than asking about real behaviour. However, he does not mention the frequent discrepancy between responses to such a non-contextualized task and actual behaviour.

Fitzgerald's concern is typical of those who feel that food behaviour is a sensitive area and one that requires indirect research strategies. He states: ''In the study of food behaviours, however, direct observation is not always possible or always desirable" (p.71). His two-year study was limited in the actual behaviour observed to 'occasional dropping-in unannounced" or such public occasions as picnics.

Table 2. Mode of presenting excerpts from open-ended interviews

Sickness and convalescence Infancy Childhood
  "All my children are breast-fed breast-fed till they had all their teeth and could eat anything I could."  
  "I gave my babies 'pot likker,' 'specially when they were sick."  
  "Babies should be started on a little bacon (fat meat with a streak of lean) when they are weaned.
I raised 10 children that way."
  "I fed my baby on cow milk, karo syrup, puff wheat, and oatmeal - started him eating before he was weaned."  
Adolescence Hard work Old age
"High-school children eat too ables and corn bread and milk Colas." "My husband thinks vegetables have to be keerful and eat would keep up his strength." "Should eat like a baby. They many sweets and Coca- light foods, like bread and als. Beans is heavy. Aunt CilIa (over 80) can eat her cabbage only when it's cooked in water and the water drained off and butter on 'em."
"My boy would always eat a big iced-tea glass full of milk and bread and fill his plate full of vegetables. He usually had two glasses of milk and two platefuls." "My husband has to have a corn bread, oatmeal and cere-

good substantial breakfast -

he is a great lover of meat."

"I always have tea for supper, and I know it's not good for the children."   "Old people is just like babies - can't eat dried peas and beans."
"Chil'un don't like much breakfast, but eat a big supper."   "Old people ought have a diet that would give them their vitamins."
    "Old people should eat very restricted diet and cut down on the amount of different things."

Source: Cussler and DeGive, 1952, p. 252.

Areas Requiring Direct Observational Techniques

While many informative studies of items rely on interviews and survey instruments about ideal rather than real behaviour or about recalled rather than actual behaviour, when we move to the next two levels (1) recipes and cuisine rules and (2) meal formats and cycles - it becomes essential to incorporate observations of real food behaviour. The understanding of relationships between food items in rules of food preparation and those between dishes in meals is not always consciously realized by an informant. One can cook a complex dish but not be able to explicate the principles and procedures out of context. As we will see, most of the studies of cuisine rules and meal formats rely heavily on the observation of food preparation and eating activities in context.

Table 3. Mode of presenting generalized data from open-ended interviews

Wage labourer Tenant and share-cropper Landowner
Breakfast Breakfast Breakfast
Eggs or oatmeal or grits Cornflakes and cream-gravy, Fruit or fruit juice or tomato
Hot biscuits or eggs and bacon (or ham juice
Coffee or cocoa in winter), or fat meat, Cereal and milk or grits
(Peaches in season) ham, or sausage Eggs and bacon or ham
  Hot biscuits and butter Toast or hot biscuits and
  Coffee or cocoa butter
    Coffee or cocoa
Dinner Dinner Dinner
Eggs or meat (fat meat String beans Chicken or ham or beef or
usually, sometimes Lettuce and onions canned fish
sausage, ham, beef, fish, or Shoulder-bone Turnip greens or spinach
chicken) Corn bread Potatoes or rice
Sweet or Irish potatoes or rice Biscuit Hot biscuits - whole wheat
Cabbage or "greens," cooked Butter one day, white the next
with fat meat Milk or buttermilk or iced tea Butter
Corn bread Canned peaches or cane Canned peaches
Milk or buttermilk syrup or black molasses Buttermilk or milk or iced tea
Supper Supper Supper
Vegetables left over from Winter: Vegetables, usually warmed
dinner Corn bread or biscuits over from dinner
Bread (hot biscuits or corn Milk or cocoa Milk or buttermilk or hot
bread or biscuits left over Sometimes meat or hot cocoa
from breakfast or dinner) soup Biscuits and butter
Honey or syrup Summer: Canned pineapple or pears or
  Left-overs from dinner peaches
  Iced tea or milk Honey

Source: Cussler and DeGive, 1952, p. 256.

Recipes and Cuisine Rules

Recipes and cuisine rules are at a level of organization in a food system that stands between the food item and the meal format/meal cycle. They contain rules for putting together items into dishes that are the basic components of meals.

Cuisine rules include rules for combinations or segregations of items, but they also refer to foodpreparation procedures, cooking media used, heat application, and flavouring principles. The distinction between cuisine rules and meal formats often is not clear-cut. Frequently, a meal format is defined by a particular type of dish. However, the meal format is more than the dish, since it includes the rules for selection of "side dishes" and courses, as well as the temporal order and spatial placement of these dishes.

Cultural constructs that prescribe which foods "go together" and which do not, and how types of food should be processed, cooked, and flavoured strongly, influence the way in which new items are incorporated. They also partly determine why old items may be difficult to displace. Selecting the type of dish very often determines the specific foodstuffs to be eaten.

There has been little analysis of this level, and the relevant studies have used varying concepts and methodology. The examples discussed here are studies of subsets of cuisine rules in China and India and among Italian-Americans.

Essential for any useful comparative studies of cuisine rules is a comprehensive lexicon of cuisine practices. Elizabeth Rozin (1982) has provided such a list of manipulative techniques that create a useful point of departure:

1. Processes that involve physical changes in size, shape, or mass.

(a) Particulation (making smaller, cutting, chopping, etc.).
(b) Incorporation - adding substances and mixing, stirring, beating, and whipping.
(c) Separation or extraction, e.g., oils from seeds, fat rendering, physical manipulation (pressing,) or using agents to separate.

2. Manipulating water content:

(a) Brief marination.
(b) Water-curing (salt liquid, pickling).
(c) Soaking and leaching

(d) Dry curing

(e) Freezing.

3. Manipulation of chemical changes:

(a) Heat

(b) Fermentation.

4. Flavouring.

Every cuisine uses a distinct complex of spices or group of spice complexes for its dishes. The structure of a cuisine can include a limited number of flavours used ubiquitously for all dishes or a range of distinct flavour complexes used to mark different events.

The cultural significance of these cuisine rules for processing and flavouring food items is the very fact that they assign meaning to certain food items. Some dishes- a range of food items manipulated in distinct ways- are associated with particular occasions or social contexts and cannot be used for others. Such cultural conceptions of dishes and their assignment to particular circumstances has a direct influence on nutrient intake. Initiating dietary changes may require manipulating the rules for when a dish may be used rather than introducing new items. Recipe systems must also be taken into account. Items may be interchangeable and recipe procedures rigid in some systems, while in others preparation techniques may be flexible and item choice may be very limited.

Recipes are the encoded rule systems specifying the relationships between items in dishes. Obtaining repertories of recipes has been attempted through intensive participant observation of food preparation such as that used by the Andersons (1969) in China and by Goode, Curtis, and Theophano (1984) among Italian-Americans. Recipe repertories can also be collected through interviews with key informants, although such collections tend to be biased toward exotic, unusual dishes that are infrequently prepared. Surveys can be used to generate a list of community-wide, frequently prepared dishes and recipes, for these can be collected from many households and analysed for their underlying structural rules. Ultimately, types of similarly constructed dishes can be identified.

Written recipes are often misleading, especially when they have been recorded by outsiders who translate procedures into their own system of categories and concepts. They also often leave out important information. There is no substitution for direct observation and for listening to the native's account of how and why something is being done. For example, the preparation of Italian-American gravy (which is of paramount importance in Italian-American households), when written in American recipe format, is comparable to an Anglo-American's version of spaghetti sauce. However, the two are totally different in taste, thickness, and visual appearance, in the first place because there are some actual differences in preparation. For example, in the Italian-American version, the spices are treated differently, the cooking period is longer, and most importantly a wide variety of gravy meats are used, including ribs with bones, not merely ground beef as in the Anglo-American version. The presence of the bones in the dish from the beginning adds to taste and thickness.

However, the most significant differences are those in the meaning and importance given to the dish in the Italian-American community. Because gravy has a central place in the food system and is imbued with significance, its preparation is undertaken with care and includes aspects of ritualistic performance. Each housewife makes a recognizably different recipe signalling household and lineage differences. They are careful to incorporate recipe features or to use utensils passed down from mothers or mothers-in-law. To watch and listen to an ItalianAmerican housewife preparing gravy clearly demonstrates that, from the native's point of view (the "emic" view), this dish is unique and bears no resemblance to an Anglo-American version of a commercially produced bottled sauce. Even if the actual ingredients (nutrients) and chemical procedures used were identical (in "etic" terms), the two dishes would never "taste" the same (Goode, Curtis, and Theophano, 1984).

In our study of the Italian-American food system, "one-pot dishes" (wet, spicy mixtures, small pieces of meat) are served with bread or pasta. We found two basic types of one-pot dishes: "gravy" and soup/stew. The non-gravy one-pots are disappearing from the dish repetoire. Gravy remains a significant dish, which is contrasted with "platters" (dry, segregated presentations of meat, vegetables and starch).

The distinction between gravy and platters and their assignment to particular food events is a major aspect of the Italian-American food system. The data collected through interviews led us to understand that the concept of gravy dishes was very important, that such dishes were contrasted to platters and that a weekly pattern of alternation existed between gravy and platter dishes. These clues were followed up through discussion during the participant observation phase.

Gravy was used in most special-event formats. Gravy dishes were mixed with platter elements for two kinds of feast formats and used with whole roast courses for Sunday dinner and calendrical holidays. Moreover, there was a hierarchy of gravy types. Boiled macaroni with gravy was an everyday dish while baked macaroni/gravy dishes (ravioli, lasagna) were restricted to festive occasions.

"Gravy" was found to be significant in defining women's role competence, relating husbands and wives, designating the meaning of events and creating affective bonds within the family. Thus, the intricacy of its preparation and the time and labour involved made it a central component in understanding the food system and its relationship to the social order - family roles, family tradition and ethnic identity. This dish exerted a conservative force, a kind of "gravitational" hold on procedures and even utensils. However, the platter was an "open" recipe, inviting new items and new methods of preparation.

Italian cuisine is not unique in its use of stylistic rules. Khare's observations (1976) of food preparation in households in North India emphasized the tremendous importance of foodpreparation rules for distinguishing "pure" and "impure" foods and for assigning them to different social times and contexts. Kacha foods are water-boiled or roasted grains that are completely segregated in space and time from their less ritually pure counterparts. They require ritual precaution in cooking; the space in which they are cooked is restricted to prevent contamination and the rules for who may receive and share such foods are restrictive. These highly ritualized foods are not for special occasions, but are for the everyday domestic meals of the households. Pakka foods use more expensive ingredients, such as ghee or clarified butter, and are fatbasted or fried. They are distributed and shared across household lines for feast occasions, and far fewer precautions are taken to keep them segregated and "pure."

Strict recipe categories and rules for appropriate use reflect the importance which such cultural constructs may have in influencing the frequency and periodicity of intake of particular foods, such as fat in this case. Kacha and pakka categories also call for different specific grains.

Khare's methodological strategy included both interviewing and participant observation. To supplement interview data he observed 151 households on a few occasions each, rather than a few households on a continuous basis. He observed a total of 173 meals and 151 ceremonial meals in his sample households, which included nine different caste groups and both rural and urban households.

Khare states that interviewing prior to observation influences the observed behaviour because families tend to conform to their previous statements. On the other hand, if observations are followed by interviews, the statements are bent to justify the performance already witnessed. Khare discusses the following way to overcome the discrepancies between elicited information about norms and actual observed behaviour. If you observe several households regularly until you are no longer regarded as an intrusive observer, you can check the relationship between elicited ideals and actual mealtimes in these households. The relative merits of one-shot or occasional observations of a large number of households as opposed to the intensive and continuous observations of a few households must be considered for different research problems.

The work of Chang (1977) and the Andersons (1969, 1972, 1977) in China and South-East Asia demonstrates that the historical and spatial continuity of Chinese cuisine is located not in the specific ingredients but in the stylistic rules that govern the form of dish composition - the fan/shih distinction between rice and other foods underlies the basic composition of most dishes. Meat and vegetables in specific sauces are a general category that is potentially eclectic and open to the inclusion of new food varieties.

The Andersons compiled their data by observing a number of cooks engaged in food preparation in several regions in China. They used continuous, first-hand observation and questioning in the kitchens of functioning households. They observed methods of particulation, cooking media, and types of heat application and spices. Their observations revealed that the Chinese macro-cultural level had certain commonalities, i.e. they found that all meats and vegetables are particulated in the same way for the shih accompaniments. Regional cuisines are distinguished not so much by the items used or by the way they are processed as by the flavour of spice combinations and cooking media used. The locus of cultural distinctiveness in this system lies in the rules for combining and segregating foods and the proportions of ingredients, not in the items themselves. The implication of the Chinese pattern of cuisine rules is that new food items can enter the system easily as long as they are treated analogously to old items.

The Andersons focused largely on the level of recipes and dishes. They do not provide information about the range of dishes in a repertory or on how dishes are organized into meal structures and cycles. For example, how does the fan/shih dish differ throughout the weekly cycle? How are festive meals distinguished from everyday meals? They also do not provide information about class differences.

Meal Format and Meal Cycles

Menu structures serve to mark differences in social occasions. They are used to punctuate activity and ritual cycles as well as to underscore and honour different levels of social participation.

Formats are significant for intake because format rules often specify the dishes and items that must be present. The community's repertory of formats and the rules for their assignment to occasions is important in food patterning.

Some food systems have relatively undifferentiated meal formats with daily meals all resembling each other in structure and content, but perhaps differing in scale. For example, the daily meals (as opposed to snacks) of many agricultural systems consist of several events each day that differ from each other only in the size of portions and the number of items in, or accompanying, particular dishes. The way the food is organized in time and space and distributed to family members does not vary. However, a major distinction usually exists between ordinary meals and festivals. Even here, the literature seems to be divided into studies of either ordinary meals or feasts, with little attempt to see the relationship between them.

And yet several studies point out that such relationships exist. Feasts often elaborate, invert, or delete aspects of daily meal structure. If one assesses a long period of time, there are discernible patterns in the way a group celebrates work and leisure cycles as well as seasonal, ritual, and life-cycle events. Each cycle tends to be internally differentiated by food events and specific rules for each food event. Minimally, all food systems have some meal-format differentiation, since food events are always used to punctuate the annual cycle and the lifecycle and since food exchange contributes significantly to creating social cohesion.

Meal formats are important in understanding change in intake resulting from contact with new environments and groups of people and the acquisition of new activity patterns and new sources of food. Often. such external changes lead to using feast formats more frequently or to creating or borrowing new formats and new dishes and items from other groups or from the mass media.

Change can also occur in meal cycles which are patterns of sequencing meal events to punctuate the flow of time and activities. Rotenberg (1981) has illustrated how changes in work activities have been reflected in the decrease in the number of meals in Vienna. Such changes in meal cycles, through reduction or addition of formats, have obvious impacts on food intake.

In Vienna, throughout the nineteenth century, the daily cycle included breakfast (upon waking), a mid-morning break (called "fork-breakfast"), lunch, tea, and supper. Breakfast included bread, butter, and a beverage, while ' fork-breakfast" included soup, sausage, and beer. The midday meal was the largest meal; it was hot and consisted of three courses. Tea was a brief bread or pastry and beverage, and the evening meal was cold - cold cuts, bread, cheese, and beverages. The structure of the meals was alike for all classes, but the content differed. Changes in work location, occupational structure, and a compressed workday have led to a reduction of the number of meals to three. Breakfast has remained the same, but lunch has been transformed into supper. The mid-morning and mid-afternoon breaks have been reduced. People still eat one hot and one cold meal a day. However, whether the hot meal is at midday or in the evening depends on work and household situations. The evening meal is almost uniformly the large meal today.

The meal type, as opposed to the item or dish, becomes an important way to introduce variety into a food system. Meal types differ not only in their content (items) but in their content rules as well. Some forms permit only narrow and specific content (items), while others are open. There cannot be an appropriate American Thanksgiving without a turkey, but the roast meat on Sunday or at Christmas or Easter can be of several varieties. Meal formats also include decorative aspects and patterns for serving that concern status order, amount allocation, decorum, etiquette, and the degree of control or autonomy in the process. For example, the use of sauces the eater can select and apply and the use of techniques for individual cooking at the table (fondues, Mongolian hotpots) allow for more autonomy than there is in a meal that is entirely dished out away from the table.

Contents - Previous - Next