Introduction: The fate of meals?

1Food is essential for human survival and for social life. Food builds our physical bodies and is intimately bound up with cultural classifications, world views and cosmologies, and thus with identity (Fischler, 1988). Food and eating are proclaimed to lie at the very core of sociality: they signify “togetherness” (Mennell et al., 1992). The meal is a marker for social relationships and ritualised sharing of food (Douglas, 1975; Sjögren-de Beauchaine, 1988). In Simmel’s classical analysis (1910) the meal is described as a mediator of socialisation. The need for food is something all humans have in common. “This is precisely what makes the gathering together for a shared meal possible in the first place, and the transcendence of the mere naturalism of eating develops out of the socialisation mediated in this way” (Simmel, 1910 in Frisby & Featherstone, 1997: 135). The act of eating at the same table – commensality – is an essential feature of social life (Mäkelä, 2000; Sobal, 2000). Commensality means coordination, reciprocity and redistribution, and in daily life meals are habitual events that bring social groups together (Fischler, 2011).

2Mary Douglas (e.g. 1975) analyzed how food and social relations are interconnected. While meals belong to the realm of family and close friends, drinks may be shared with other people, too. Analyses inspired by Douglas (1975, see also Douglas and Nicod, 1974) have highlighted the structure of meals – meal formats – as ordered systems relating to other ordered systems, and to gendered roles and responsibilities in households (Murcott, 1982, 1983 and 1986). In the Nordic countries, similar studies have highlighted the meanings of food and meals in various population groups, and shown, that here too, food and meals are important media in social life for building groups and social relations (Bugge & Døving, 2000; Ekström, 1990; Fürst, 1985; Holm, 1996; Mäkelä, 1996; Haukenes, 2007; Sylow & Holm, 2009; Roos & Wandel, 2004).

3It has been a repeatedly posed question both in media and the scientific literature about modernisation whether traditional regular meal patterns and meal formats are being disrupted (Murcott, 1995 and 1997). Judging from media discussion and modern urban life, we get the impression that ordinary daily eating has changed considerably during the last decades. In the academic literature, we see hypotheses about individualisation, the dissolution of family meals, and globalisation as key elements of modernisation (Burnett, 1989; Mintz, 1996; Ritzer, 1993). Increased snacking and consumption of fast food are depicted to epitomize these tendencies. Interestingly, various attempts to grasp the proclaimed new patterns of eating have often included developing apposite names to the recognized change. Market researchers created the term “grazing” (Caplan, 1997). The grazing hypothesis contends that food is eaten in less patterned ways than before with regard to time, place and contents (Senauer et al., 1991). Jean-Pierre Poulain (2002) has called it “vagabond feeding” or “nibbling”. On a more general level it has been argued that in a state of “gastro-anomy” (Fischler, 1988) previous cultural norms for what should be eaten when and with whom disappear, and regular meals become increasingly rare and replaced by irregular eating patterns. Also in the Nordic countries this discussion has been prominent which is witnessed by a series of local concepts describing modern gastro-anomy. In Sweden, the term “breakfastification” (frukostisering), (Ekström, 1990: 73) denotes the alleged de-structuration of meal eating, and in Denmark and Norway the concept “eating food on the go” (spisning i forbifarten, see Holm, 2003; mat i farten, see Bugge, Lillebø & Lavik, 2009) emerged.

5In the following we first present the theoretical background for our study. We then outline the methodological challenges of a comparative survey and our solutions to them. This is followed by important results from the study focusing on the character – both similarities and differences – of eating patterns in the Nordic countries in the late 1990s. We end by summarizing some main findings and their significance as well as briefly presenting our new study which will give us a possibility to analyse changes in Nordic eating patterns between 1997 and 2012.

Theoretical background

6Food consumption is not a social institution in the same way as the family or the market; it is better framed as a set of coordinated practices, in which the practices as well as the overall configuration are subject to institutionalisation (Kjærnes et al., 2007). Everyday food habits are neither explicit, individual acts of decision-making, as assumed in cognitive approaches (see e.g. Conner & Armitage, 2002), nor are they mere unconscious, pre-determined acts (Gronow & Warde, 2001). Habits normalise practice. They are the “way things are done” (by ”Us” if not by ”Them”). Such “normal” practices describe how things are usually done, but also how things should be done.

7However, food consumption is not only guided by social norms and tradition. Eating does take place within the habitual and ordinary practices of everyday life and the ways in which these practices are organised in time and space influence the patterns of eating. Eating contributes to ordering our days into segments: morning, midday, afternoon, and evening. The order and rhythm of eating – the meal pattern – form intersections between the public sphere of production and the private sphere of reproduction, of family and recreation (Aymard et al., 1996). The focus of interest is then the organisation of schedules, the particular modes in which food preparation and meals interchange with work and other activities, as part of cyclical calendars as well as throughout the day. These schedules are influenced by societal change.

8A classical study by Rotenberg (1981) shows how eating patterns in Vienna shifted from the early 1900s up until the 1980s. During this period, meals and snacks shifted depending on the organisation of work, not only in terms of time and contents of meals, but even with regard to who ate together and where the meal took place. The five-meal pattern, typical of European commercial cities around 1900, included three daily meals at home with family members (morning, midday and evening), one daily meal at the workplace with colleagues and one afternoon meal at a café with personal friends and old school-mates (at least a common practice of middle class men). This eating pattern reflected a form of societal organisation in which adults tended to work within or close by the family residence, allowing women to cook a large midday meal while working in and around the home, and men to come home in the middle of the day. By the 1930s industrialisation had implied a reorganisation of everyday life, with industrial workplaces located in the periphery of the cities, larger parts of the workforce now engaged in wage-labour rather than family owned businesses, and work-days subjected to strict definitions of time for breaks, which were now shorter. A three-meal pattern, typical for industrial societies had now emerged, where the family meal was relocated to the end of the working day, and socialising with friends to take place during weekends. Similar general shifts in meal patterns were identified in a Finnish study (Prättälä & Helminen, 1990). The three-meal pattern was observed among urban industrial workers already in the 1920s, but a more massive shift in this direction did not take place until the 1960s and 1970s.

9Following from this, in order to understand trends of eating we need to know the broad social context and its influence on the situations in which people eat. Today’s society is often conceptualised as a post-industrial society. Individualisation, de-traditionalisation and what appears to be freedom of choice are often claimed to lie at the core of modernity (Giddens, 1994), and flexible and more individualised work life and lifestyle are characteristic of post-industrial societies. The question remains, therefore, whether a shift away from this three-meal pattern can be identified, indicating a new post-industrial eating pattern. Or, is it the case, as has been repeatedly suggested, that eating today is less structured, and that general patterns are difficult to find?

Studying Nordic eating patterns: eating events and the eating system

10When it comes to the conceptualisation and everyday concepts of eating, there are significant differences between the Nordic countries. Three of the four countries (Denmark, Norway and Sweden – plus the Swedish-speaking parts of Finland) form a language community, and they share some parts of their vocabulary for meals. However, even though the vocabulary may be shared, the concepts have different meanings, and their social and cultural implications are often quite diverse. While the term ”frukost” means breakfast in Sweden and Norway it signifies lunch in Denmark, and while the term ”middag” (meaning a cooked dinner) is ordinary in Norway and Denmark it cannot be taken for granted that the evening meal is called a ”dinner” in Finland and Sweden. The very term ”måltid” (meal) has different connotations, too. In Denmark and Norway eating open sandwiches with a topping such as cheese, liver paste or ham is generally conceptualised as a meal, whereas in Finland and Sweden the term ”meal” tends to be associated with cooked, hot meals. Which meal would be seen as the main meal of the day, would likely differ between the countries as well.

11The variations in the meanings of vocabulary and the differences in the social and cultural implications of meal concepts entailed some difficulties for our study of how eating is patterned in the Nordic countries. Analyses at a population level with quantitative methodology call for strictly structured design for the study. We needed to construct a questionnaire which would allow us to capture the variations in how eating is organised in the four countries while at the same time posing the same questions in all countries. On account of this, we decided to avoid culturally laden concepts such as ”meal”, ”lunch” and ”dinner” and instead work on the basis of what we termed a deconstruction – reconstruction process.

12Starting out with ideas resembling the 24-hour recall method used in dietary surveys, we first applied a deconstruction approach and asked the respondents about every occasion of eating something during the day before the interview (for practical reasons, just a beverage or only eating chewing gum, sweets, etc., were excluded). In our analyses these occasions were characterised as ”eating events”. Thus, we broadened the concept of a ”food event” (Douglas & Nicod, 1974) to use as our starting point a very inclusive concept of ”eating event” which allowed us also to focus on the social situations in which eating took place. This way we hoped to avoid making false generalisations on the basis of cultural, national or ideological prejudices or “taken-for-grantedness.”

Nordic eating in 1997

14The data were collected in April 1997 using computer assisted telephone interviews (CATI). We interviewed representative samples of the populations in all four countries aged 15 years and above (in total, N = 4 823). The questionnaire consisted of loops where questions on the time, the structure, the contents, and the context of eating events, were recorded chronologically (Mäkelä et al., 1997). Interviewing was evenly distributed throughout the week (excluding Sundays for practical reasons) so that the records covered all days of the week except Saturdays (Kjærnes et al., 2001). A record of one day of eating allowed us to ask in detail about the specific situations of everyday life in which food is eaten. But one day would not take account of the variability of an individual’s diet and the data therefore did not allow analyses of patterns and routines at the individual level. The situational focus meant that the analytical unit was for the most part the eating event rather than the individual.

15In the analysis, considerable recoding of the data was performed in order to construct variables in the eating system. These were then analysed according to frequency and character of eating events and according to time intervals (Kjærnes et al., 2001).

The eating pattern

16The results showed that eating in the four countries took place from early morning until late at night indicating some flexibility in eating patterns (Gronow & Jääskeläinen, 2001). Still, in all countries commonly shared eating hours could easily be identified. Figure 1 shows the daily rhythm of eating in the four countries in terms of frequencies of three types of eating events: breakfast (the first eating event), cold eating events, and hot eating events. The frequencies have been added on top of each other to show the total proportions of respondents having had something to eat during one-hour intervals throughout the day.

17Figure 1 demonstrates that eating patterns in the four Nordic countries were similar in some respects and different in others. On one hand, there were rhythms that were relatively uniform and distinct for each country; on the other hand, for the first part of the day the rhythm appears to be similar across the countries. Eating hours were most uniform in Denmark (i.e. with the highest peaks), least so in Finland and Sweden. During the first part of the day, the timing of the main peaks was fairly identical between the four countries. After around 3 pm the rhythms started to divert and there were clear differences in the time of the evening meal. Norwegians had an earlier evening meal than those in the other countries (at 4-5 pm), whereas the largest ”dinner” peaks were found at 5-6 pm in Finland and at 6-7 pm in Denmark and Sweden. On Sundays, eating was generally much more heterogeneous, compared to the weekday patterns. Eating started later during the day and the events were spread out more evenly over the whole day (Gronow & Jääskeläinen, 2001). These findings indicate that eating is strongly socially coordinated, first and foremost with reference to the organisation of the working day. This is not just a matter of time constraints, but represents a strong institutionalisation of eating even for people outside the workforce.

18The peak hours shown in Figure 1 reflect conventional Nordic cultural models of three main meals: breakfast, lunch and dinner. The smaller peaks in between indicate more flexible eating in between main meals. The types of eating events that dominated at different hours were quite uniform within the countries, especially during peak hours. Some were also similar across the countries. A large majority of the respondents had a hot meal in the late afternoon/evening. However, for lunch time eating Figure 1 shows important national differences. Cold lunches dominated completely in Denmark and Norway, whereas large groups in Finland and Sweden, respectively, had a hot meal between 11 am and 1 pm. Accordingly, it was relatively common in Finland and in Sweden to have two hot meals – and a small percentage even had three. A similar pattern did not exist in Denmark and Norway. These differences demonstrate two distinct meal patterns in the Nordic region: the ”western” pattern (Denmark and Norway) with one hot meal, and the ”eastern” pattern (Finland and Sweden) with two hot meals. In the two ”western” countries sandwiches take the place of a second cooked meal, but these eating events are also strongly coordinated, with shared conventions with regard to the format and content of the meal.

The meal format

19In all four countries, people reported a range of eating events that showed both complexity and variation. Yet, certain meal formats dominated at the various hours of the day. In the morning the large majority had eaten a simple meal with a limited selection of foods and beverages. Bread was the main component in all countries, typically eaten with butter or margarine, jam, cheese, salami or liver paste. Coffee or tea, milk, and sometimes juice were common drinks. Additional eggs or foods like ham or bacon were relatively rare. Foods eaten with a spoon also occurred in Finland (typically porridge), Denmark (cereals and/or yoghurt), and Sweden (all of the above), but not in Norway. In Norway, more than 80%, and in the other countries, more than 60% of the respondents reported that their breakfast only consisted of one food item or dish (Mäkelä et al., 2001).

20As indicated in Figure 1, the midday meal varied between the countries. As noted above, whereas in Denmark and Norway this was typically a cold meal consisting of open sandwiches, this was most often a hot cooked meal in Finland and Sweden. It is likely that this distinction is a reflection of differences in welfare policies between the four countries: in Finland and Sweden, many adults can buy a lunch for a reduced price in their workplace lunch restaurant, whereas in Denmark and Norway, no such systems are predominant. Hot lunches reflect systems with institutionalised provision of meals and cold lunches provisioning systems based on food individually brought from home.

21In the evening a hot meal was typically eaten. Most weekday hot meals had a uniform and relatively simple structure, generally consisting only of one dish. Starters were very rare (below 10% on any day), while 25-30% of the cooked events included a dessert (typically ice cream or berries and fruit). In each of the countries different food items dominated the hot meals. Steaks and pork chops were frequent in Denmark, whereas fish was typical in Norway. Boiled vegetables such as carrots, beans, and peas dominated in Denmark and Norway, and raw vegetables such as lettuce, tomatoes and grated root crops were common in Finland and Sweden. Potatoes were the most important staple in all four countries, supplemented by bread in Finland and Sweden. This two-staple pattern was hardly found in Denmark and Norway. Hot dishes with one component were more common in Norway (such as lapskaus – a beef and potato casserole – or pizza), whereas Swedes had more meals with four or five components. In all countries, there were more components on hot dishes eaten on Sundays than on those eaten on week days. The typical four-component hot dinner plate (a centre, a staple, vegetables and trimming) which has earlier been described as the ”proper meal” format for cooked dinners in United Kingdom (Douglas & Nicod, 1974; Murcott, 1982) was found in all countries, but appeared not to be very frequent. Less than 20% of hot meals in all countries, and in Finland less than 10% represented this combination (Mäkelä, 2001).

22Some of the differences in what was eaten appear to reflect national economic conditions. Thus, the fact that Norwegians eat more fish and Danes more meat is likely to be influenced by the characteristics of the food producing systems in these nations (Norway being a major fish, and Denmark a major pork, producer and also exporter), but some differences may depend on the differences in welfare systems. The finding that Finns and Swedes eat more uncooked vegetables than Danes and Norwegians may be due to fact that the hot lunches in Finland and Sweden that have been supported by welfare services are subjected to regulatory norms suggesting bread and salads, for example, to be served with the meals for health reasons, whereas meals in Denmark and Norway have traditionally been regarded as entirely private matters.

The social context of eating

23In general, incomes in the Nordic countries are high compared to other countries, large proportions of the populations go to work or school during the days, many participate in sports and cultural life, the age for retirement is high, and many live alone. It could therefore be expected that much eating would take place outside the private homes, in the public sphere of restaurants, cafés, etc. In the 1997 study, however, the majority of eating events took place at home, eating at the workplace coming second. Eating in other people’s homes and eating out in cafés or restaurants constituted only minor proportions of everyday eating. Frequent eating at restaurants, cafés, bars and fast-food outlets, often seen as a characteristic of modern life, did not show in the data. Eating in places like fast food outlets or in the street was negligible.

24At almost all times of the day, more people had eaten at home than in any other place (Holm, 2001b). In the middle of the day, most eating took place at workplaces or schools. Most of this workplace eating appeared as organised social events, usually in a canteen or separate eating room. However, in Denmark more people ate at their working place (the desk, the building site) indicating somewhat less institutionalisation of workplace eating in this country. Eating at a café or restaurant during work hours was generally rare, though slightly more frequent in Sweden.

25The Nordic eating system found in 1997 was a mixture of events taking place alone, with other household members, with colleagues (lunch), and with friends. A considerable proportion of breakfast eating took place alone, even in multiple person households. Those who lived with a family usually shared the evening meal with their family members. Eating in the company of friends appeared to play a very small role, suggesting that such events were not a frequent everyday habit. The tendency to eat with friends was somewhat more common during the weekend than on weekdays and most often it would take place in one’s own or other people’s homes. While individual eating took place throughout the day, the socially shared peak meal hours in each country were, to a large degree, social events. Moreover, on Sundays, when meals were usually not structured by working hours, family eating dominated, even for a large proportion of those who lived alone. The data thus showed that even though a considerable part of eating took place alone, eating was also still a social matter in the sense that at least some of the events during the day took place in the company of other people. A pattern of eating primarily in solitude was found first of all among respondents who lived alone and more so among the elderly than in younger age groups.

26According to qualitative studies, family meals have great significance and meaning as symbols and operators of family cohesion, in the Nordic as well as in other countries (Bugge and Døving, 2000; DeVault, 1991; Ekström, 1990; Iversen & Holm, 1999; Mäkelä, 1996). Our analyses demonstrated that family eating was still common in 1997. Further, it turned out that on the day before the interview more than half (between 54% and 64%) of those living in multiple person households had at least one hot meal, where the entire household ate together, and ate the same food (Holm, 2001a). This was the case both in households with and without children.

27Most of the eating at home took place at a kitchen or dinner table (Holm, 2001b). A relatively frequent alternative was a coffee table/sofa. Television sets are usually placed within view from the sofa in the Nordic countries, and evening snacks seemed to be brought to the coffee table, sometimes even the evening meal (most often in Norway, least so in Finland). This was confirmed when inquiring what went on while eating, where about one fifth of eating events at home took place while watching TV. Pizza with the family in front of the TV has been described as a shared family ritual at the end of the week (see e.g. Bugge & Døving, 2000).

Stability and change

28In her report on “The Problem of Changing Food Habits” from 1941, Margaret Mead describes the cultural anthropological conception of food habits: “Food habits are seen as the culturally standardised set of behaviors in regard to food manifested by individuals who have been reared within a given cultural tradition. These behaviors are seen as systematically interrelated with other standardised behaviors in the same culture” (Mead, 2008:18). Seen in this perspective, food habits in the Nordic countries must be expected to change as other practices and institutions change: the rhythm of daily routine tasks such as work, sleep, leisure activities, and the size and composition of households are important examples of such practices and institutions which may be expected to change rather slowly. But public welfare policies and commercial services may shift more rapidly, and new ideas and discourses including and addressing the moralities and meanings associated to food may change norms and conventions in society. Thus, what is eaten, when, where, with, and prepared by whom may change, but the role of food as a mediator of social interaction and a key signifier of social groups and identity is likely to remain stable.

Conclusion: A modern eating system – or a phase of transition?

29In the 1997 study, our interest was on the influence of modern everyday life on the rhythms, structures, and the social organisation of food and eating. The study demonstrated that the eating system in the Nordic countries was varied and multifaceted with regard to all these three aspects of the eating system (Kjærnes, 2001). While flexibility and individual eating were evidently part of the picture in 1997, social coordination of rhythms and social interaction around eating were prevailing traits. Our results thus questioned the proposition that modern eating is characterized by “destructuration”, individualisation and globalisation. A disappearance of traditional features and dissolution of the sociality of eating could not be identified as dominant features. A more mixed picture emerged, confirming that eating was very much a social activity characterised by nationally different, but distinct, shared rhythms of eating. Eating patterns were strongly influenced by the ways in which Nordic welfare societies are organized in terms of working life, food distribution systems, and the family institution.

30Considerable and extensive changes have taken place in the Nordic societies during the post-war period, the large increase in female participation in the workforce being one such important change, and the publicly financed institutionalized welfare systems, another. In this perspective, the observed family and home centered eating might be interpreted as representing a “cultural lag”.

31Even so, it is not easy to foresee the direction of modernisation tendencies. For example, the meal formats appeared to be relatively simple, but from this study alone we cannot judge whether this was the case even in earlier periods, or, whether this relates to female wage-labour and signifies a shift in the division of labour between private households and industry. Will food industry and services increasingly take over every-day cooking? At the same time long-term tendencies of growing incomes, more varied supplies of fresh as well as processed food, and a highly pervasive gastronomic discourse might lead to more complexity of everyday eating rather than increasing simplicity and “convenience”. While we know from other sources (Groth et al., 2001; Roos et al., 1996) that there are considerable social inequalities in nutritional aspects of dietary habits within the countries, the meal patterns and formats observed in our study were socially quite homogeneous within the countries.

33By basing the new study on corresponding representative samples of the Nordic populations and by using a largely identical questionnaire, the new study allows us to make comparison across time and to analyse whether or not the patterns of uniformity and flexibility found in the 1997 study is a testimony of the composite character of modern post-industrial eating patterns or a reflection of an ongoing change. The new study will also allow us to evaluate whether the changing economic conditions and welfare systems have affected socio-economic discrepancies of eating. Last but not least, the new study enables addressing the potential impact on eating patterns of contemporary debates on, e.g., healthy eating, sustainability, and climate change.