"three squares a day" usually refers to the American idiom that means eating three full, hearty meals (as opposed to a quick snack, or just a hamburger and fries) every day. A square meal implies that you eat a full serving of meat, vegetables and perhaps dessert.
A similar idiom is "three hots and a cot," which often refers to why some indigent people purposely get themselves arrested – in prison, they'll be served three hot meals per day, and be provided with a place to sleep.
The observation that in many societies, food consumption is organized around some concept of the meal led much early sociological research to focus on the nature and meaning of meal structure. The pioneering work of Mary Douglas offers theoretical grounding of the study of food and eating in localized empirical studies of dining.
In her article "Deciphering a Meal" (1975), she identifies two contrasted food categories—meals and drinks. Meals are structured and named events (for example, lunch, dinner) whereas drinks are not. Meals are taken against a background of rituals and assumptions that include the use of at least one mouth-entering utensil per head; a table; a seating order; and cultural restrictions on the pursuit of alternative activities (such as reading) while seated at table.
A meal also incorporates a series of contrasts: hot and cold, bland and spiced, liquid and semi-liquid. Both meals and drinks reflect the quality of social relationships. Drinks are generally available to strangers, acquaintances, and family. Meals, by way of contrast, are reserved for family, close friends, and honored guests.
Douglas's key empirical study with colleague Michael Nicod relates meal structure to meal content. Nicod recorded over various periods the dining patterns of four English working-class families, whose diet of the time centered on two staple carbohydrates—potatoes and cereals, in contrast to upper-and middle-class diets, which made greater use of a range of cereals, beans, and roots.
Focusing on the type and cycle of meals within the domestic economy, Nicod identified three types of meal: Meal A, a major meal, served around 6:00 P.M. on weekdays and in the early afternoon on weekends; Meal B, a minor meal taken at 9:00 P.M. or 10:00 P.M. on weekdays and 5:00 P.M. on weekends; and Meal C, consisting of a biscuit and a hot drink. This last meal was a flexible component available at any time in the daily dietary cycle but most often taken both in late afternoon, on the return home of the principal wage earner, and before retiring for the night.
Meal Content and Cycles
Of the three types, Meal A is accorded the greatest analytic importance by Douglas and Nicod. A strong correspondence between the weekday evening meal and the Sunday meal was apparent. In both instances, the first course was the main course, always hot and savory and based on a tripartite structure of potato, centerpiece (meat, fish, eggs, with one or more additional vegetables), and dressing—usually gravy.
The second course repeated these rules of combination except that it was sweet. The staple took a cereal form (pastry, sponge), the centerpiece was often fruit, and the dressing custard or cream. On Sundays and other special occasions, the second course was often followed by a third consisting of a hot drink and biscuit(s).
This third course maintained similarities in rules of combination in that a biscuit has a cereal staple form enclosing a fruit or cream-type filling. In one respect, however, the pattern differed, in that liquids and solids were totally separated, in contrast to other courses, and the structure was reversed in so far as the hot drink appeared in a cup or similar receptacle, whereas the cold biscuit was on a plate.
Thus, according to Douglas and Nicod, meals possess the following elements. First, they have rules of nonreversibility in the archetypal meal. Second, the order of food runs from savory to sweet and from hot to cold in terms of the principal food items consumed. Third, quantity decreases with each course as formal patterning of foodstuffs increases. Regarding the latter, Douglas sees the first course as fairly amorphous, but as a meal progresses, this gives way to increasing geometric precision and structure (Douglas, 1982; Wood, 2000).
The "Cooked Dinner"
Whereas Douglas places equal emphasis on meal structure and meal content, it is the latter that has driven many subsequent and empirical studies of meal taking. Murcott's (1982) study of thirty-seven pregnant women in South Wales is almost as important as Douglas's work for the elaboration of a research tradition.
Murcott found that the "cooked dinner," an elaborated model of Douglas's principal Meal A type, comprising meat, potatoes, at least one additional vegetable, and gravy, was regarded as a "proper meal," and perceived as essential to family feelings of health and well-being. Structurally, the cooked dinner was thought of as a meal in itself, was heavy and large rather than small and light, and hot, never cold. Thus, although a succession of courses was permissible, the cooked dinner as centerpiece could, in contrast to sweet-based items, stand in its own right as a meal. Fresh meat was a priority, and potatoes were always specified and itemized separately from other vegetables. Certain meats had common circulation—beef, lamb, pork, chicken—whereas others, notably turkey, were reserved for special occasions.
Fish was not regarded as an acceptable substitute for meat in the cooked dinner form. While potatoes were invariably a constant (roast on Sunday, usually boiled at other times), slightly more flexibility was evident in the choice of additional vegetables. Even here, however, certain rules appeared to operate. First, additional vegetables were almost invariably green and from "above ground" (typically peas, beans, sprouts, cabbage, and occasionally broccoli and cauliflower). Second, additional vegetables were prepared only in addition to these and were generally from "below ground" (for example, carrots and parsnips).
Together with meat and potatoes, the final ingredient necessary to the structural integrity of the cooked dinner was gravy, last in the cooking and serving sequence, and poured onto the plate after other items had been assembled, an action Murcott sees as linking and transforming items in the cooked dinner into a coordinated whole. Cyclically, the importance of the cooked dinner to Murcott's sample was emphasized in the fact that it was eaten on only three or four days of the week (including, invariably, on a Sunday) and thus had relative scarcity in the family dietary system.
The "Cooked Dinner" in America
The British work of Douglas and others is echoed in various American studies, which generally support a view of meals as central to domestic dining systems. In a 1942 study of food habits in Southern Illinois, researchers found that food consumption centered on three staples: potatoes, beans, and pork.
The authors term this the "core diet" and note that around the core was, first, a secondary core, consisting of many foodstuffs that had recently become available for purchase from local stores; and, second, a peripheral diet of infrequently used foods outside of the core and secondary core (Bennett et al., 1943).
All three of these concepts are used in similar form by Norge Jerome, who charts content variations found in meals and snacks for all of "normal" weekday meals, "Sunday dinner" and festive meals (for example, Thanksgiving). Jerome argues that the dietary order consists of core and staple items; secondary core items, which are added to or substituted for items in the core as circumstances and contexts vary; and peripheral dietary items, which are those items used infrequently, including ceremonial foods.
In a much more complex series of studies of Italian-American diet in Philadelphia, Judith Goode and various colleagues relate variations in the selection of different meal formats to meal cycles (food consumption patterns over time), community values, and activity patterns of households, building up a many-layered picture of the interrelationships between the role of food in people's lives and other aspects of the social order (Brown, pp. 66–68).