The Class and Culture of Norwegian Culinary Straw Men: A Response to Flemmen,Hjellbrekke and Jarness’ ‘Class,Culture and Culinary Tastes: Cultural Distinctions and Social Class Divisions in Contemporary Norway’

Abstract

This research note offers a critical-constructive discussion of the article ‘Class, Culture and Culinary Tastes: Cultural Distinctions and Social Class Divisions in Contemporary Norway’, written by Flemmen, Hjellbrekke and Jarness (FHJ) (Sociology, 2018(1)). Concerns are raised about the methods and the use of the data. A robustness analysis with alternative data and/or alternative methods is suggested. Conceptually, the analysis of FHJ is considered not to engage adequately with a more qualitative body of historical and ethnological literature, as well as the impact of Norwegian agricultural policy. To describe and understand the evolution of social meaning and social patterns of the consumption of ‘traditional’ Norwegian foodstuffs, a qualitative approach could have contributed constructively. Overall, wider implications for Bourdieu-inspired analyses of cultural consumption are addressed.

Keywords

Class consumption food culture food history food policy methodology multiple correspondence analysis Norway principal component analysis traditional food

Introduction

Is the combination of the playful, omnivorous consumption of both exotic and traditional food something new for the Norwegian middle and upper classes? And is it possible to understand this class and food cultural evolution by relying on quantitative data and multiple correspondence analysis? We have read the inspiring 2018 article ‘Class, Culture and Culinary Tastes: Cultural Distinctions and Social Class Divisions in Contemporary Norway’ written by Flemmen, Hjellbrekke and Jarness (FHJ), which deals positively with these issues. Their study of Norwegian class cultures and differences in ‘original taste’ has several interesting aspects. However, parts of the analysis and results seem to be based on the class and culture of two Norwegian culinary ‘straw men’, which makes it problematic. Our attention is aimed at some methodological aspects and the claim that ‘traditional peasant food of pre-capitalist Norway can be refashioned as a badge of distinction in the 21st century’ (Abstract and p. 145).

Research Questions and Answers

Let us first recap the research questions and how they are framed. The authors position their research mainly according to two research gaps. The first is that remarkably little attention has been paid to food and eating practices in contemporary class analysis. The second is that little is known about cultural class distinctions in the egalitarian Scandinavian societies and that Norway is a particularly interesting case. It is beyond the scope of this article to address the framing conditions further. Instead, we first concentrate on the main research questions:

To investigate the potentially classed nature of taste, two explorative research questions are addressed: what are the main lines of division in tastes for food and drink in contemporary Norway; and, how are these differences related to social class divisions? (p. 129)

The authors treat these explorative questions by employing multiple correspondence analysis (MCA). They conclude that:

the divisions in one’s tastes for food and beverages are clearly connected to social class divisions: upper- and upper-middle-class households endowed with large amounts of economic and cultural capital consume healthily and eclectically, whereas the opposite is true of the lowest regions of the class structure. (p. 145)

However, as we will elucidate, their conclusions are more wide-ranging than their empirical analysis can justify. A reflection upon the method and data transformation, therefore, seems a logical place to begin.

Methodology

The data set is the Norwegian Consumer Expenditure Survey (CES) for 2012. The survey was conducted by Statistics Norway (SSB, 2013) and is intended to be representative of the Norwegian population. A total of 3363 Norwegian households recorded all their expenses in a two-week period in 2012. For this annual survey, households are distributed in all 26 two-week periods of the year and throughout the country. Expenditures and quantities are registered for 250 pre-selected food and non-alcoholic beverage items. Food consumed at restaurants, bars, cafeterias and fast-food outlets is not included. For each household and each item, the purchase is multiplied by 26 to get a yearly purchase figure. All the recorded expenditures are in Norwegian kroner (NOK) and are continuous variables.

There are several debatable uses of the data from CES in the analysis. What is recorded in CES is not individual consumption but rather consumption at the household level. As FHJ emphasize, Norway is an egalitarian society where there is little economical difference between a person with a university education and a person without a university education. This means that there will be many cases in which university-educated individuals marry or cohabitate with a person without a university education. In addition, individuals without higher education may have incomes that exceed those of university-educated individuals. How is it possible to classify a household in regard to educational level when the education of only one person in the household is recorded?

The classification regarding household income is also worth mentioning, as it is used without correction. There is a difference in purchasing power on the individual level between a single-person household with an income of one million NOK and a household comprising a mother, father and three children with a similar total income. In addition, there will be a difference between a young couple with children and a large mortgage and an older couple with no mortgage and whose children have left the household. If we take into account that most Norwegians own their own homes, we may assume that, in most cases, purchasing power is stronger among older individuals than among younger individuals. An alternative way to measure purchasing power would be the total consumption per consumer unit of non-durable goods and services.

The data sample is restricted to include only those households in which the main income earner is between 24 and 76 years of age. CES comprises households in which the main earner is between 18 and 86 years of age, so there should be no reason to restrict the data sample. The data sample is further restricted by excluding households who have several instances of zero purchases recorded. Since the period for which a household keeps a record of expenditures is only two weeks, there may be several small households that purchase less variety than larger households. Why, then, is there a need to use sample weights to correct for non-responses when the authors delete households from their own needs? An alternative strategy is to use imputation to correct for non-responses.

The authors apply multiple correspondence analysis (MCA) to analyse tastes (food purchases) among different social classes in Norway. To be able to use MCA, they categorize the variables and use from two to five categories, depending on the distribution of the original variable (p. 135). This means that they use two, three, four or five categories that they claim are the most appropriate. Transforming a continuous variable to a categorical one can be done in a number of ways, depending on where the interval limits are set. Would the analysis change if other limits were used? An alternative method of analysing the CES data is principal component analysis (PCA), which is constructed to analyse continuous variables. It is robust against non-normality and makes it possible to handle non-linearity when applied to continuous variables (Jolliffe, 2002). The only problem regarding CES and PCA is the zero purchases in the CES data, but this can be handled in many ways, for example, by imputation (Rubin, 1987).

Conceptual, Food Cultural and Contextual Sensitivity

The relationship between class, playfulness and a category that is inconsequently designated as traditional foodstuffs, traditional peasant foods, peasant diets and Norwegian husmannskost is of main concern in the concluding discussion. However, one can ask whether FHJ are setting straw men up against each other when they argue against what they refer to as an established truth: that contemporary upper- and middle-class people ‘play’ with a taste for the cosmopolitan and the exotic. Contrasting to such a description of distinctive consumption of today’s Norwegian upper- and middle-class households, FHJ claim that these classes additionally eat traditional foods as a new symbol of distinction and argue that this calls for a need for a relative understanding of taste. Through this assertion, they posit a class-based food cultural evolution, where the upper and middle classes did not consume traditional products in playful, omnivorous ways in earlier times. However, these are complex and context-dependent questions, and the analysis lacks a clarification of the food cultural concepts that are subject to the same operationalization through specific products. Smoked salmon, meatballs, oatmeal and swede are described by the authors as ‘traditional items, once staples of the Norwegian peasant diet’ (p. 135), and that:

When the traditional is re-appropriated by the upper classes and assembled with foreign foodstuffs and dishes, the very meaning of ‘the traditional’ is removed from its original social context – and thus conveys new meanings. The upper-class consumption of items like swede, flatbread and gamalost in contemporary Norwegian society is sociologically different from the peasant consumption of the same items a century or even decades ago. (p. 145)

Let us first consider the concepts, and then the products, and finally, we offer some contextual remarks.

Concepts

FHJ ‘use the term “traditional” to distinguish locally produced foodstuffs once staples of the Norwegian peasant diet from more recent imported and globalised foodstuffs made available to Norwegian consumers’ (note 8). They also argue that the social meaning of food is continuously renegotiated and that objects need to be explained inductively and not decided a priori (p. 146). However, in their analysis, ‘tradition’ is a priori understood as the opposite of ‘exotic’ or as ‘authentic’, inspired by Johnston and Baumann’s (2014) typology based on the analysis of the American gourmet foodscape. Using MCA, FHJ do not offer the possibility of leaving it to the individual to express what ‘traditional’ connotes, although they acknowledge that there are ‘lay notions of what constitutes traditional food products in different contexts’ (note 8). This is also the case in Norway (Amilien and Hegnes, 2013), and different understandings of tradition are also open to degrees of ‘playfulness’. What if traditional food is understood differently between and within classes depending on a French, American or Norwegian context? Furthermore, what if their own understanding is representing the middle-class researchers’ understanding? Operating within a paradigm of reflexive sociology, these questions would be expected.

In addition to ‘tradition’, FHJ use the concepts ‘peasant’ and ‘husmann’ (cottar) without reflecting upon their meanings in reference to different classes. Notaker (2000: 78–79) writes that differences between classes are mirrored in their diets, and that this division can be found in the countryside as well as between cities and countryside. The population in rural areas was also complex and consisted of different classes with different foods or diets. ‘Peasant’ and ‘husmann’ (cottar) are examples of such social groups or classes. Whereas the peasant usually owned his land, the cottar often rented land from a farmer or landlord. The status of the cottar also differed between different Norwegian regions.

The last conceptual distinction that FHJ present, more or less implicitly, is related to foodstuff and diet. It may be mentioned that, whereas foodstuff may be understood as mostly related to production, diet is more often related to consumption.

‘Traditional’, ‘peasant’ and ‘husmann’ make sense as relative sensitizing concepts (Blumer, 1954) when they are used exploratorily in combination with ‘foodstuff’ and ‘diet’. However, when they are operationalized by pre-selected specific products, the analysis of the evolution of classed consumption becomes blurred.

Products

Descriptions of formerly class-based food cultural practices related to the so-called traditional products that FHJ refer to exist, although they do not refer to this volume of literature. It consists mostly of historical and ethnological sources related to Norwegian foods, and several of them describe the products as consumed by both lower and higher classes, more or less omnivorously and playfully (for an example in English, see Amilien, 2007).

In contrast to FHJ, Notaker (2000) gives several examples indicating that salmon in general, including smoked salmon, was eaten by the upper class. In Døving (1997), salmon is described as a food item that was traditionally characterized as a gourmet food consumed by the upper class. However, with fish farming (which started in the 1970s in Norway) and decreasing prices, salmon has now become an everyday food for the ordinary Norwegian.

Meatballs have a history similar to that of salmon, and Notaker (2006) claims that meatballs became popular throughout the population between 1850 and 1950 and that the breakthrough for this dish was unthinkable without the introduction of the mincer. Young women learned to use this new tool at schools of home economics, and they spread their knowledge. Notaker (2006: 14, our translation) writes: ‘If it was not snobbish to cook meatballs, it was in any case associated with a certain prestige.’

Porridge and gruel were made with flour, but towards the end of the 19th century the porridge from rolled oats, earlier a festive dish, became more accessible because of the expansion of modern mills. In a cookbook from 1912 this is acknowledged with these words: ‘Oatmeal porridge is more nutritious than rice and barley porridge. Recently it has also become daily fare, especially for breakfast, with rich and poor’ (Løken, 1912: 38–39, our translation).

Døving and Lien (2000) write about the low status of swede in the late 1980s, noting that it is virtually absent in contemporary Norwegian food discourse. They also claim that swede’s position in the Norwegian dietary pattern is, first and foremost, as a side to traditional dishes such as salted and smoked meats (2000: 110). With its favourable storage potential, swede was not a product that was eaten historically only in rural areas; it was also consumed in cities. In the context of the First World War, it was well utilized, and in the interwar period, it was referred to as the ‘northern orange’. During this period, the bourgeoisie ‘played’ with the swede and cooked it in different ways (Notaker, 2000: 292). Old cheese (gamalost) has also enjoyed a certain prestige and has been consumed by both farmers and the upper class (Fosså, 2000: 149). In addition, it was eaten by foreigners and preferred over other cheeses (Notaker, 2000: 70).

Notaker (2000: 106) writes that when countryfolk picked berries, it was most often to sell them to members of the elite who had access to sugar and used them for drinks, desserts and jams. Many sold cloudberries to the cities, where they were also exported abroad. Furthermore, Notaker writes that senior civil servants both belonging to the elite in the city as well as in the countryside ate cloudberries in the same way as strawberries, with milk and cream. Cloudberries were considered exotic by city people, as well as those in other countries.

The Norwegian word brød (bread) signified until the 20th century flatbread (crispbread) and was considered a ‘national’ bread (Notaker, 2000: 255). Oats and barley were the main grains cultivated in Norway, and because they had the necessary gluten to be used in leavened bread, flatbread, lefse, porridge and gruel became the basic nourishment for the common man in Norway, not exclusively by peasants, but also in the towns and also by people of higher social status. The most prolific and popular cookbook author in the first part of the 20th century has a recipe for flatbread in her 1914 upper-class cookbook (Erken, 1914).

Whereas the products FHJ make use of in their class-based analysis may be understood as traditional in some ways, it is more challenging to relate them to their description of the evolution of class-based consumption. Some of the products were consumed mainly by lower classes and some by upper and middle classes as well, but products such as salmon and cloudberries were historically consumed by the upper classes, contrary to what FHJ argue.

Context

In regard to more contextual interest, FHJ claim that ‘local food traditions may be more relevant in Norway than in other European countries’ (p. 145). However, Norway is more often used as an example for the opposite argument (Amilien et al., 2008), and it is common to draw a line between the food cultures in the northern and southern parts of Europe (Montanari, 1994). The southern food culture is usually described with a stronger understanding of local food traditions (Barjolle and Sylvander, 2000).

FHJ also leave out the important impact of changes in the political context. Towards the end of the 1980s, global and European trade policies were undergoing major changes. This affected Norway and Norwegian agricultural policies; Norwegian food products now faced new competition from foreign imports. To counter the competition, Norwegian authorities and other key stakeholders began building what some described as ‘mental border protection’. Simply put, this strategy aimed to convince Norwegian consumers to choose Norwegian products, similar to what DeSoucey (2010) describes as gastronationalism. In the Norwegian Agricultural Authority’s strategy work at this time, it was a clear goal to base ‘mental border protection’ on the Norwegian competitive advantage. Among such advantages explicitly mentioned were local food traditions and old Norwegian food traditions and food culture and the like (Landbruksdepartementet (Ministry of Agriculture), 1992). From this point onwards, Norwegian governmental authorities have promoted a range of different food qualities and the sale of high-quality foods through different initiatives, such as labelling schemes. This triggered a new way of thinking about and looking at food that emerged in Norway towards the end of the 1990s and the beginning of the 2000s. The Norwegian turn to new qualities coincides with a growing focus on food labelling in Europe over the past two to three decades, but with a different motivation. Whereas the Norwegian ‘quality turn’ is characterized by a top–down initiative, the European indicates a bottom–up process made by consumers, retailers and producers away from standardized products and towards alternative qualities (Goodman, 2003), such as traditional foods. This difference implies that what is referred to as the ‘quality turn’ may, in different contexts, cultures and countries, consist of entirely different ways and social mechanisms for turning to and constructing alternative qualities (Hegnes, 2012).

Concluding Comments

FHJ show that there are lines of division in regard to tastes for food and drink in contemporary Norway, and that these are related to social class. However, the methodological approach is debatable, and their conceptual approach and description of a class-based food cultural evolution, based on the operationalization through specific food products, is more problematic. They do not refer to data concerning playfulness either before or now. Furthermore, they do not present their own data about the previously class-based consumption of the products to which they refer, nor do they use secondary sources to support their arguments. Nevertheless, they draw conclusions as if they had such data, causing the construction of the class and culture of both ancient and contemporary Norwegian culinary straw men.

We do not know the consequences of restricting the data sample and using the sample weights on the restricted data sample. It all depends on the underlying distribution of the variables and the way the authors constructed their categorization. For example, the purchase distribution for a food item in CES consists of one discrete part: purchase or not purchase, and one continuous part: the quantity purchased conditional on positive purchase. When zero observations for that item are deleted the distribution will shift to the right and will no longer be representative for the original distribution. This will have an impact on the construction of the categorical variables and the results.

A robustness analysis to determine whether the results and conclusion of the study are valid would have been beneficial. One way of doing that is to use an alternative Norwegian data set that may be more suitable for analysis with MCA. The Norwegian Monitor database (NMD) is one good option. What is registered in the NMD is the frequency of consumption (at home and away from home) and not just household expenditure on food consumed at home, as in CES. The consumption data are discrete (ordinal), and there are many more varieties of food items from which to choose. The unit of the data is the individual rather than the household, and individual income, household income and education are registered. Thus, there is no need for any ad hoc transformation of the data to use MCA. Another way of doing a robustness analysis is using PCA after imputing the zero observations. This can be done with CES and the same variables used in the article. A third way of conducting a robustness analysis is constructing regression models taking account of both the discrete and continuous distribution of the expenditure on food items. Then, among other variables, income and education may be included as predictors, and hypotheses of effects of classes on consumption may be tested statistically. This can be done using CES.

Whereas Bourdieu (1984) and Johnston and Baumann (2014) make use of both qualitative and quantitative data in their renowned analyses, FHJ depend solely on MCA. MCA may be used as a tool to explore omnivorous consumption, but how can MCA be used to uncover playfulness? Playful tastes for the ‘cosmopolitan’ and the ‘exotic’ are understood as the opposite of traditional foods. One may ask if this is a constructive distinction in the Norwegian context, and further, what does it actually comprise or mean? Why cannot national traditional foods also be cosmopolitan and exotic foods? This is also interesting because they emphasize ‘the value of using MCA to map the connections between taste divisions and social divisions, as the structure of taste divisions is mapped inductively from the data’ (p. 146, emphasis in original). Although MCA maps inductively from data, one can ask if the data are inductively collected in the first place. Consumption and a new focus on local and traditional foods do not expand only among the upper classes as an element of class distinction; they are also a result of gastronationalism (DeSoucey, 2010) initiated by Norwegian authorities to protect the national market in relation to increased international competition.

FHJ’s dedicated use of the class perspective and MCA seems to unintendedly serve as blinders for other theoretical, methodological, conceptual and contextual inspiration that could have contributed to an overall more robust analysis. If studies of the evolution of cultural distinction and class lack substantial, conceptual, historical and political perspectives the results will provide simplified descriptions of complex and contextual relations and mechanisms. Some of FHJ’s results should be considered as new hypotheses rather than convincing descriptions of a class-based food cultural evolution. However, the hypotheses are interesting and should be further studied.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Atle Wehn Hegnes is Researcher at the Department of Economics and Society, Norwegian Institute of Bioeconomy Research.

Geir Wæhler Gustavsen is Researcher at the Department of Economics and Society, Norwegian Institute of Bioeconomy Research.

References

Amilien

(2007) Turtle soup and water porridge: Some social and cultural perspectives on food habits in the city of Oslo, 1860–2000. In: Atkins

PPJ

Lummel

Oddy

(eds) Food and the City in Europe since 1800. Hampshire: Ashgate.

Amilien

Hegnes

(2013) The dimensions of ‘traditional food’ in reflexive modernity: Norway as a case study. Journal of the Science of Food and Agriculture 93(14): 3455–3463.

Amilien

Schjøll

Vramo

(2008) Forbrukernes forståelse av lokal mat [Consumers’ conceptions of local food]. Rapport 1. Torshov: The National Institute for Consumer Research.

Barjolle

Sylvander

(2000) PDO and PGI products: Market, supply chains and institutions. Protected designations of origin and protected geographical indications in Europe: Regulation or Policy? FAIR – CT 95 – 0306, Final Report, European Commission.

Blumer

(1954) What is wrong with social theory? American Sociological Review 19(1): 3–10.

Bourdieu

(1984) Distinction: A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press.

DeSoucey

(2010) Gastronationalism: Food traditions and authenticity politics in the European Union. American Sociological Review 75(3): 432–455.

Døving

(1997) Fisk: En studie av holdninger, vurderinger og forbruk av fisk i Norge [Fish: A study of the mechanism behind the fish consumption pattern]. Rapport 12. Lysaker: The National Institute for Consumer Research.

Døving

Lien

(2000) Myten om den perfekte kålrot [The myth of the perfect swede]. Norsk Antropologisk Tidsskrift 11(2): 108–127.

10.

Erken

(1914) Stor kokebok for større og mindre husholdninger [Big Cookery Book for Bigger and Smaller Households]. Kristiania: Aschehoug.

11.

Flemmen

Hjellbrekke

Jarness

(2018) Class, culture and culinary tastes: Cultural distinctions and social class divisions in contemporary Norway. Sociology 52(1): 128–149.

12.

Fosså

(2000) How old is old cheese? Gamalost in coffin-shaped boxes and eccentric jars. In: Walker

(ed.) Milk beyond the Dairy. Proceedings of the Oxford Symposium on Food and Cookery 1999. Totnes: Prospect Books.

13.

Goodman

(2003) The quality ‘turn’ and alternative food practices: Reflections and agenda. Journal of Rural Studies 19(1): 1–7.

14.

Hegnes

(2012) Introducing and practicing PDO and PGI in Norway: Turning to protected quality through translations of meaning and transformations of materiality. Anthropology of Food. Available at: http://journals.openedition.org/aof/7210.

15.

Johnston

Baumann

(2014) Foodies: Democracy and Distinction in the Gourmet Foodscape. New York: Routledge.

16.

Jolliffe

(2002) Principal Component Analysis. New York: Springer.

17.

Landbruksdepartementet [Ministry of Agriculture] (1992) Konkurransestrategier for norsk mat (Ren Mat-strategien) [Competitive Strategies for Norwegian Food (The Clean Food Strategy)]. Oslo: Ministry of Agriculture.

18.

Løken

(1912) Matstel og husstel for almindelige husholdninger [Cooking and Housework for Common Households]. Kristiania: Aschehoug.

19.

Montanari

(1994) The Culture of Food. Oxford: Blackwell.

20.

Notaker

(2000) Ganens makt. Norsk kokekunst og matkultur gjennom tusen år [The Power of the Palate: Norwegian Culinary Art through Thousand Years]. Oslo: Aschehoug.

21.

Notaker

(2006) Mat og måltid [Food and Meal]. Oslo: Aschehoug.

22.

Rubin

(1987) Multiple Imputation for Nonresponse in Surveys. New York: Wiley.

23.

SSB (2013) Survey of Consumer Expenditure 2012. Oslo: Statistics Norway.