CULTURAL AND MORAL CLASS BOUNDARIES IN A NORDIC CONTEXT: Findings from a city in Denmark

With the work presented in Distinction (1984 [1979]), Pierre Bourdieu set the stage for an ongoing interest in the workings of highbrow culture as a marker of ‘class’, and more generally in cultural capital as a basis of exclusion from the job market and from accessing high-status groups (Lamont and Lareau 1988). The most general proposition, demonstrated by numerous empirical examples throughout Distinction, is that of homology between spaces; tastes and lifestyles across various areas of practice are ordered in oppositions according to similar logics, such as form vs. function, quantity vs. quality, distinguished vs. vulgar (Bourdieu 1984 [1979]: 1–7). Corresponding to these patterns is a social hierarchy or a space of social positions, ordered, as a first principle, according to the overall volume of both cultural and economic capital and, as a second principle, according to the relative amount that individuals or groups possess of each of the two forms of capitals (Bourdieu 1984 [1979]: 175ff.). In the decades following Distinction, debates have arisen about the nature of cultural hierarchies, but also about the most relevant forms of symbolic boundary-drawing in different contexts. Of particular interest for the purpose of this article is the argument put forward by Michéle Lamont, who states that Pierre Bourdieu is underestimating the importance of ‘moral signals’ and wrongly asserts that ‘differences directly translate into hierarchalization’ (Lamont 1992: 181). Studying various forms of boundary drawing is important if we are to understand processes by which some are excluded from desired goods, groups, positions, etc., and the findings presented here confirm the significance of moral signals when it comes to drawing symbolic boundaries between different social positions. However, the workings of different repertoires for drawing boundaries may not be as different in effect as suggested by Lamont.

In regards to the nature of cultural capital in the USA, some researchers have found evidence that possession of legitimate culture as indicated by participation in high-culture arts may function as cultural capital in much the same way as in France (DiMaggio 1982; DiMaggio and Mohr 1985; DiMaggio and Mukhtar 2004). Others have argued that cultural consumption may well function as a marker of status, but that it is expressed – not by an exclusive taste for highbrow art forms –but increasingly by an omnivorous combining of highbrow genres with middle or lowbrow genres (Peterson and Simkus 1992; Peterson and Kern 1996). The same pattern was found in a study on musical tastes conducted by Bryson (1996). Bryson found that political tolerance is associated with musical tolerance, and that musical exclusiveness decreases with educational level. As argued elsewhere, there are few in the current study who merit the title ‘culturally omnivores’, and generally those with highbrow tastes are likely to reject lowbrow tastes (Prieur et al.2008). The material presented in this article points instead to a ‘global’ mode of orientation as one of the important forms of distinction among those in privileged positions.

When it comes to the sources that people make use of for constructing symbolic boundaries towards others Norbert Elias argued that the boundaries establishing the differences between the included, ‘we’, and the outsiders, ‘them’ often is endowed with or build upon perceptions of moral and human worth, with some groups being stigmatised as less worthy than others (Elias and Scotson 1994). A similar argument is put forward by the Canadian sociologist Michèle Lamont who argues that Bourdieu with his emphasis on the workings of cultural capital tend to neglect other important sources for drawing symbolic boundaries. Lamont (1992: 4) makes a distinction between three different types of symbolic boundaries: moral boundaries, drawn on the basis of moral character, such as honesty, work ethics, personal integrity and consideration for others; socioeconomic boundaries, drawn on the basis factors such as wealth, power, status or class, and cultural boundaries, drawn on the basis of education, intelligence, manners, taste and command of high culture. Lamont suggests that the strength of these boundaries may vary in different contexts. Based on a study of the French and American upper-middle classes, she shows that Americans are less concerned with signals of high cultural status than the French, and that the French are less likely to emphasise materialism (189). Also, Lamont argues that symbolic boundaries may work independently, not only of each other, but also of objective socioeconomic boundaries (1992: 188). Therefore she objects to a framework that assumes that tastes and preferences are signs of distinction and social position, but do so, without specifying the conditions under which such differences create inequalities. She states that Bourdieu wrongly asserts that ‘differentiation leads directly hierarchalization’ (1992: 182). In contrast to Lamont, the approach I take in this article, will be to maintain a Bourdieu-inspired framework that studies the interrelationships between different repertoires for potentially accenting socio-economic aspects as a relation between social and symbolic space (see also Weininger 2005: 84). This does not entail a neglect of the specific meanings that people apply to cultural and moral differences, since the latter is explored through in-depth interview.

In a French context, the Bourdieusian approach, and in particular the notion of ‘class habitus’, has been criticised for not taking into account the dispositional and contextual plurality arising from the heterogeneity of individual life courses as well as from variations in the contexts of cultural consumption (Lahire 2008). This critique finds some resonance in a major UK-based project on cultural capital and social exclusion reported in Culture, Class and Distinction (Bennett et al.2009). Based on a multiple correspondence analysis (MCA) covering cultural activities from seven different areas, the study demonstrates the existence of systematic patterns of cultural tastes and practices across various fields of study, and that the most powerful dimension of cultural difference reflects what Bourdieu termed the total volume of capital (Bennett et al.2009: 43 ff.). However, concluding on the various studies involved, they stress that in some fields of practice other cleavages, such as age and gender, are even more important than those stemming from class habitus (251 ff.). Furthermore, it is argued that no unifying national canon of legitimate culture can be detected in the UK. Rather, the dominant expressions of cultural capital are found to be an adoption of various more omnivorous orientations. The latter are contrasted with the ‘fixed’ or ‘static’ tastes which, by implication, are those of the working class (254–5). Other British studies are more reluctant to support a homology thesis based on class. Chan and Goldthorpe (2005, 2007) find that although both education and status may be linked to patterns of cultural consumption, traditional occupationally based measures of class are not likely to be so. It should be noted, though, that none of these studies is able to discriminate between economic and cultural fractions, as was a central point for Bourdieu in Distinction.

In a Nordic context, highbrow culture as a marker of social class is confirmed by quantitative studies using techniques similar to those adopted by Bourdieu (Rosenlund 2000; Prieur et al.2008; Skjøtt-Larsen 2008). Other studies support the findings of Peterson and Kern (1996), arguing that patterns of cultural consumption may still be stratified by class or education, but that people in privileged positions are more likely to exhibit ‘omnivorous’ or ‘eclectic’ patterns of cultural consumption rather than limiting their taste to highbrow consumption (Purhonen et al.2010; Jæger and Katz-Gerro 2010); that other factors such as age and gender may be more important than class and educational level (Bihagen and Katz-Gerro 2000; Purhonen et al.2009), or that other resources, such as organisational capital, may be relatively more important (Broady 1998; Danielsen 1998). Finally, a number of qualitative studies question the significance of legitimate culture as capital, and find that egalitarian and anti-elitist moral sentiments are likely to work against legitimate culture as a means of exclusion in the Nordic countries (Hjelseth 2005; Skarpenes 2007; Lien et al.2000). Most at odds with the Bourdieusian notion of high culture as a means of distinction and exclusion is probably Skarpenes (2007). In a Norwegian context, he argues that it is considered illegitimate to draw boundaries towards others based on knowledge and cultural hierarchies, and that moral notions of ‘being good’ and ‘being tolerant’ are valued higher by the highly educated than cultural or socioeconomic success.

The main questions to be addressed here are:

• •

  Do cultural consumption and politico-moral attitudes serve as markers of class positions?

• •

  Which of the two sources appears to be the most pertinent when it comes to drawing symbolic boundaries between ‘them’ and ‘us’? How are they related to each other and to social stratification?

• •

  To what extent are privileged groups able to extend their cultural and moral hierarchies beyond their own social positions?

Based on a survey conducted in the Danish municipality of Aalborg (n = 892) and by using MCA, a space of social positions is constructed. In order to access whether both cultural tastes and politico-moral stances may serve as markers of class positions, indicators of these are included as ‘supplementary points’ in the space of social positions. One of the novelties this approach is that cultural practices and politico-moral stances are both included within the same framework, but also that In-depth interviews with respondents from the survey are introduced in order to elaborate on the specific ways that people in different social positions draw symbolic boundaries towards others. The concluding discussion will focus on the content of and the relation between different forms of social differentiation and different forms of boundaries, on the most important forms of distinction and drawing of symbolic boundaries, and on the extend to which boundaries are being drawn from less-privileged social positions.

The quantitative analysis is based on a survey collected in the Danish municipality of Aalborg in 2004. Aalborg is the main city in the northern part of Denmark and has a population of 162,000 (2004). One thousand six hundred individuals aged between 18 and 75 were randomly selected among the residents of the municipality of Aalborg. Out of these, 1174 were successfully interviewed (73.4%). Data were gathered with the specific aim of constructing a space of social positions and a space of lifestyles along the lines of Bourdieu's social theory. In order to focus on current social position, students were omitted along with respondents below the age of 25 and older than 67 years. Thus the analysis is based on answers from 892 respondents.

The qualitative analysis is based on interviews with 16 respondents selected from the survey among respondents agreeing to participate in an in-depth interview. They have been chosen to represent different positions within the space of social positions (see Figure 3). The interviews include issues of personal biography, work experience, experience of social differences, political attitudes and attitudes towards other social groups.

Using MCA¹ on the survey data allows me to analyse interrelations between many variables at a time, and to make use of multidimensional spatial constructions in the interpretation. It is the same method that Bourdieu used in Distinction, which allows me to apply a comparable approach. MCA is based on individuals×categorised variables tables. It is a formal-geometric approach to multivariate statistics, and the interpretation is based on graphical as well as numerical output (Le Roux and Rouanet 2004: 14). The first step in the analysis will be an interpretation of a cloud of active categories of a number of axes. Answers that frequently occur together will be situated closely to each other, while those that never or only rarely occur together will be situated far from each other. Based on these patterns, I will interpret the sociological principle at stake along each of the axes. I then insert indicator variable categories into the map. These ‘supplementary points’ do not affect the established axes. Following the recommendations of Le Roux and Rouanet (2004: 116), all categories included exceed 5% of the total number of respondents. In some cases, this has led to a re-coding of the categories, collapsing infrequent answers with the ‘nearest’ category.

I set out by constructing a two-dimensional space of social positions based on economic capital and cultural capital. The questions have been formulated to resemble the choices made by Bourdieu (1984: 126–31). Regarding cultural capital, the main differences in relation to the choices made by Bourdieu is to include an indicator of the size the of book collection held by the respondent (objectified cultural capital), along with an indicator of the educational level of the respondent's best friend. The latter based on the assumption that one's choice of friends in this regard may serve as an additional indicator of one's cultural capital (see Bourdieu 1987: 5). The respondent's occupation is only included as a supplementary variable. Eleven questions with a total of 38 categories were chosen from the survey. These are listed thematically in Table 1.

The total variance of the model is summarised along a number of axes, each summing up a descending part of the total variance. The first three eigenvalues are well separated. After the third axis, eigenvalues tend to fall steadily. The first vertical axis sums up 56% of the total variance. The second horizontal axis sums up an additional 19% of the variance. The third axis sums up another 8%. Due to the limited space and the high level of variance summarised by the first axes, the analysis will be limited to the two most important ones, summing up 75% of the variance.²

Figure 1 presents answers to the 11 questions included in the construction of the two-dimensional space of social positions. Tables 2 and 3 display the variables contributing the most to the variation of the first two axes, in the order of their contribution. Further, the orientations and the contribution values of the categories contributing above average (100/38 = 2.6%) to the variation of each of the first two axes are listed.

The first axis is a general one with contributions from indicators of both economic capital (38%) and cultural capital (50%). The vocational indicators have contributed 12%. Fifteen categories have contributed above average to the variation of the axis. At the bottom of Figure 1 are indicators of low levels of economic capital (income_DKK < 220,000; car value_DKK < 50,000 and residence_rented). Indicators of low levels of cultural capital are also placed at the bottom half (educ_primary school; father's educ_primary school; friend's educ_short and books_not many) along with having an unskilled father. At the top side of Figure 1 are indicators of high levels of economic capital (income_DKK >600,000; car value_DKK >200,000 and residence owner?), indicators of high cultural capital (educ_higher tek/nat/adm; educ_higher tech/nat/adm; educ_higher hum/soc educ_higher hum/soc; father's educ_bachelor and friend's educ_long), along with indications of an upbringing with high levels of both cultural and economic capital (father's voc_white collar). I interpret this axis to be one of capital volume. Moving from the bottom to the top of Figure 1, there is an increase in the total amount of resources, cultural and/or economic in nature.

The second axis also sums up contributions from indicators of both economic capital (53%) and cultural capital (34%). The vocational indicators have contributed 14%. Sixteen categories have contributed above average. To the left in Figure 1, there are indicators of high levels of cultural capital (educ_higher hum/soc, father's educ_master and friend's educ_long). Regarding economic capital, however, it is indicators of low economic capital that are situated to the left (income_DKK < 220,000; car value_DKK < 50,000 and residence_rented). This goes along with indicators of a privileged upbringing (father's voc_white collar). On the right-hand side of Figure 1 are indictors of relatively low levels of cultural capital (educ_vocational and father's educ_primary) and high levels of economic capital (car value DKK >200,000; savings etc._yes; residence_owner and value of summerhouse DKK >800,000). This goes along with vocational indicators implying an upbringing most likely to be associated with low levels of cultural capital (father's voc_unskilled and father's voc_farming). Because of the reversed relationship between economic and cultural capital, I interpret this axis to represent the composition of capital. Moving from the left towards the right of Figure 1, there is a relative increase in economic capital compared to cultural capital. Furthermore, this is a separation according to the specific types of cultural capital, with higher education within the humanities and sociology on the left-hand side, and practical, vocational training on the right.

An indicator of respondent's vocation has also been inserted as supplementary points in Figure 1. It confirms and makes it possible to expand the interpretation of the first two axes. At the top are vocations characterised by high levels of both income and educational attainment. At the bottom are vocations characterised by low levels of both income and educational level. Along the second axis, the vocations on the far left-hand side are within teaching and healthcare, both demanding high educational levels. On the right-hand side, the latter are opposed by craft and managerial occupations. The group of managers has the highest median income and the group of workers within craft has higher median income than the other vocations at the lower end of Figure 1. Hence, the opposition between fractions with relatively high levels of cultural capital vs. fractions with relatively high levels of economic capital is also an opposition between, on the one hand, teaching and health care occupations (predominantly female and public sector) and on the other hand, craft and managerial occupations (predominantly male and private sector).

In the following section, indicators of cultural preferences and practices and of politico-moral attitudes and practices will be inserted into the space of social positions (Figure 1). The lifestyle components indicated in Figure 2 consist of 49 variable categories. Twenty-six categories are derived from questions covering cultural preferences and practices from the following areas: use of the city (city), regular reading of newspapers (paper), use of the Internet (Internet), preferences in art (art), type of food served to guests (dinner) and preferences in television programmes (TV). Twenty-one categories cover politico-moral attitudes. The questions are termed politico-moral because some are related to conflicts over the distribution of economic capital, which I will refer to as ‘old politics’ (OP). Others are moral in character, but have proven to have had a growing impact on voting behaviour in the last decades (Borre 2007). I refer to these as ‘new politics’ (NP). ‘Voting’ at the last general election was included to cover political practice. Finally, two categories were added to indicate locality and nationality as sources of identity. All the categories are inserted as supplementary points and do not contribute to the content of the axes (Le Roux and Rouanet 2004: 233).³

Each of the four quadrants will be described separately, followed by examples from the qualitative interviews with persons from the relevant quadrant. Figure 3 gives the spatial distribution of the interviewees.

In the top left quadrant of Figure 2, we are likely to encounter individuals who are privileged regarding their overall volume of capital, but mainly through their possession of cultural capital. Starting with one of the most legitimate forms of cultural practices, having visited an arts exhibition within the last year (art: attend) is more pronounced among the privileged groups than among the underprivileged. Regarding preferences in art, inhabitants of the privileged cultural quadrant are more likely than others to dismiss figurative art forms, by rejecting that ‘you can see what the picture is supposed to be’ is important (art: prefer figurative_5/5). Also, there is a preference for installation art (art: like installations_1/5), which was included in the survey as an example of the more avant-garde art forms. Regarding use of the city, the highbrow choices of art museum and going to the theatre (city: theatre_1/3) are the most characteristic ones. When having people round for dinner, the menu is likely to be inspired by exotic foods rather than the traditional local cuisine, and the main concern is not plenitude (plenty of food_3–4/5, dinner: new and exotic_1/4, dinner: traditional_5/5). The paper most frequently read is the social-liberal Politiken with extended coverage of cultural themes and activities. Regarding television, there is a preference for international news rather than local news, and for news on the US presidential campaign rather than news on the Danish Royal Family (TV_intern., TV_US pres). The latter, global rather than local orientation, is also evident from the propensity not to take pride in or to be neutral about local or national belonging (proud of being Danish_3/3, Proud of being Aa_2/3). In general, the cultural preferences within the privileged cultural quadrant are characterised by high levels of participation in the most legitimate art forms. There is little interest in figurative art but a good deal in avant-garde art forms like installation art. Finally, there is an international or global orientation rather than a national or local orientation across several fields of practice.

Politico-moral attitudes in the privileged cultural quadrant are characterised by political self-confidence (politics complicated_5/5), ethnic tolerance (np: hire danes_5/5) and a solidarity that transcends national borders (np: increase aid_1/5). Also, there is both the means and the willingness to pay extra for organic food (np: organic_1/5). Characteristic choices in terms of political parties are The Reed-Green Alliance belonging to the far left (RGA), and The Social-Liberal Party (SLP) belonging to the centre-left.

From the patterns detectable in Figure 2, we would expect the major cultural opposition to be between the privileged cultural quadrant and the underprivileged groups at the bottom in Figure 2. Cultural distinction as a mode of drawing symbolic boundaries was not widespread in the qualitative material. However, it clearly exists as a mode of drawing symbolic boundaries for Carsten, a high-school teacher in his thirties. In a conversation about leisure activities performed with his colleagues, he explains that they could easily be found at a handball match ‘with a pint in our hands’, and he emphasises that there is ‘no longer anything degrading about that’. Later, however, Carsten talks about the importance of having the right cultural capital for entering certain social circles, and how this is a challenge to some of his friends who are not educated within the humanities:

One of the recurring issues in the interview was football. Carsten does not disapprove of sports as such. He mentions professional cyclists and golfers as people he admires for their individual achievements. It is the collective aspects of football, both on and outside the field, which repels him. Carsten ends the conversation about football supporters by relating their level of engagement to his own when it comes to music and states: ‘… I'm probably more moved by music’.

That Carsten uses football and the fan culture as a means of dis-identifying with the working class fits well with the findings of Fiske (1992), who argues that fan culture is often associated with cultural forms that the dominant value systems denigrates. A similar argument can be found in Hjelseth (2005). Carsten's affinity for the classical legitimate culture and avant-garde culture is elaborated on throughout the interview. He speaks with confidence of classical music and opera, but he also emphasises experiments with the latest music and performance. He is not a cultural snob in the sense that he only consumes highbrow culture (Peterson and Kern 1996). Rather, he resembles the highbrows depicted by Jarness (2009), consuming many non-elite cultural products, but doing it with various degrees of distance and irony. Furthermore Carsten's cultural consumption and orientations are not limited by the local or the national cultural scenes. He frequents the theatres and musical scenes in Copenhagen as well ‘Die Oper’ in Berlin, and states: ‘we no longer have geographical limitations. If we want a cultural experience, that is good, then we will take that trip’.

In most of the interviews from the privileged quadrants, tolerance has been an important theme. Avoiding discrimination based on ethnicity, gender and other social categories is a core value among the privileged groups, and especially within the cultural quadrant. However, excluding others on grounds of their intolerance seems not just acceptable, but to be an important statement for some. Asked if there are any types of people he would rather not be sitting next to at a wedding party or a similar social occasion, Jens, a schoolteacher in a managerial position in his late forties, replies:

During the interview, Jens does not draw any explicit socioeconomic or cultural boundaries between himself and the working class. In fact, he emphasises that he likes to watch football and makes a point of this being ‘culture as well’. However, Jens is quite willing to draw the line when it comes to racism and discrimination – attitudes which he associates mainly with underprivileged social conditions.

Cultural practices are at work for some when symbolic boundaries are drawn. However, the centrality of the politico-moral issues in the qualitative interviews does lend some support to the arguments made by Lamont (1992), Skarpenes (2007) and Bennett et al. (2009: 66) when they stress the more limited role played by cultural capital as a means of drawing symbolic boundaries or expressing superiority.

In the economic quadrant at the top right of Figure 2, we are likely to encounter individuals who are privileged regarding their overall volume of capital, but mainly through their possession of economic capital. Regarding cultural preferences and practices, there is an interest in anything to do with money or activities related to purchasing power. The most characteristic paper read here is Borsen, which is a business and financial newspaper. The most distinct use of the Internet is for shopping and banking (Internet_shopping etc.). Regarding the politico-moral stances, there is an emphasis on growth at the cost of greater income inequality (op: growth above equality_1/5). Also, there is a wish to join the European currency (op: Euro_1/5). Regarding political practice, the most pronounced preferences are the two parties in government at the time, the Liberal Party (party_LP) and the Conservative People's Party (party_CPP). Hence, the politico-moral stances in the privileged economic quadrant are characterised by a liberal/conservative political orientation and an orientation towards economic growth rather than equality.

In general, few boundaries were made in the interviews with respondents from the privileged economic quadrant towards other social positions. An example of this is Thomas, co-owner of a small firm in the information and communications industry in Aalborg. Asked if he could point to any social groups having more privileges than they deserve, he explains:

Another example from the privileged economic quadrant is Torben, an engineer working in the information and communications sector. His job involves some travelling abroad. Torben grew up in a small village in a rural region, and he is conscious not to let material status symbols create a distance between himself and his parents and his background. Throughout the interview, however, he does express a sense of detachment, and he is likely to distance himself from traditionalism, immobility and intolerance:

Later in the interview, he describes some of his childhood friends who cancelled a joint trip to London as having ‘their feet somewhat more rooted in the local soil …’. In these quotes, distance is created by depicting his childhood friends as too geographically ‘rooted’ in the local community and ‘stuck’ in the sense of not aspiring to try anything new. In other sequences, he is also critical of the traditionalism or conventionalism he has experienced. The description of his childhood friends seems to concur with the ‘fixed’ and ‘static’ tastes by which the various dominant tastes were contrasted in the study by Bennett et al. (2009). Furthermore, this opposition seems exemplary of the more general opposition between the cosmopolitan and local modes of orientation found throughout the study, which will be elaborated upon in the concluding discussion.

The last position to be described could be termed the underprivileged position. Based on the pattern of the cloud of individuals (Figure 3), the underprivileged quadrants seem to be less dispersed according to the principle of composition of capital. Also, in the qualitative interviews, there are many similarities and expressions of solidarity and shared identities between the groups inhabiting the two quadrants at the bottom in Figure 2. For reasons of convenience, however, the quantitative findings from the two quadrants of the map will be described in turn.

In the bottom left quadrant of Figure 2, we are likely to encounter individuals lacking in overall resources, and with less economic capital compared to cultural capital. The cultural activities and preferences, as well as the politico-moral stances, are marked by the scarcity of economic resources, and are the ones most directly contrasting those of the privileged economic quadrant. When using the Internet, it is manly for chatting (Internet_ Chat). Turning to the moral and political stances, people are likely to describe themselves as less focused on material comfort than most people (op: less materialistic_1/5). Also, in opposition to the privileged economic pole, there is a reluctance to exchange the national currency with the Euro (op: Euro_5/5). In contrast to the highbrow choice of ‘Politiken’ as well as the economically privileged choice of ‘Borsen’, the most distinct newspaper across both quadrants at the underprivileged pole is the tabloid paper ‘Ekstra Bladet’ (paper: EB). The political party most markedly related to the bottom left quadrant is the Socialist People's Party (party_SPP). The Social Democrats (party_SD) is popular across both quadrants of the underprivileged positions. Finally, at the bottom left quadrant are those who did not vote (party_ did not vote).

In the bottom right quadrant of Figure 2, we are likely to encounter individuals lacking in overall resources, and more so cultural capital. Male working-class occupations within craft or machine operation are frequent here. It is in this section we find the strongest opposition to the privileged cultural pole. Participation in art and the cultural activities included in the survey is less frequent at the underprivileged pole (art_ don't attend and city_theatre_3/3). There is a taste for figurative art where ‘you can see what the picture is supposed to be’ (art_figurative_1/5). Regarding the Internet, the most distinct practice is not to use it (Internet: Don't use), and in particular not to use it for e-mailing or information-seeking, which is low across the two underprivileged quadrants. There is an orientation towards the traditional Danish cuisine (dinner: traditional_1/5) rather than the new and exotic (dinner exotic_4/4). As for television, the preference is for local rather than international news (TV_loc) and for news on the Royal.

Family rather than the US presidential election (TV_Royal Family). Also, there is a preference for relaxing entertainment genres rather than the more demanding documentary genre (TV_ entertainment).

Turning to the moral and political stances, it is also in the underprivileged quadrants and in particular in this ‘blue collar’ quadrant that we find the main opposition to the tolerant and cosmopolitan segments at the privileged cultural pole. First of all, there seems to be some alienation towards the political debate, expressed by indicating that politics is complicated (politics comp.). There is an orientation towards traditional family values, indicated by a stronger tendency to think that children suffer when women give as high a priority to their career as many do today (np: women's work_1/5), and by a greater tendency to accept the more physical/traditional forms of punishment in childrearing (np: smack children_1/5). Also, there is a greater tendency to make distinctions based on some of the categories dismissed by the cultural elite as irrelevant and prejudicial. This is displayed by the above-mentioned greater acceptance of gender differences vis-à-vis the family and the labour market, but also in a greater propensity to differentiate between native Danes and immigrants when it comes to hiring labour (np: hire Danes_1/5). Though it is not the most economically deprived quadrant, the willingness to pay extra for organic food is the lowest (np: organic_5/5). The most defining choice of political party is the right-wing populist party The Danish People's Party (DPP), but, as noted above, votes for the Social Democrats are also frequent in this quadrant.

Cultural participation is less frequent within the underprivileged quadrants, and in general the talks on cultural preferences and consumption did not take up much time. The most interesting and surprising finding was an indignation about having to pay for highbrow cultural services, which are not consumed in this part of social space. Poul, an electrician in his late forties, stated his aversion to the tax-funding of highbrow culture as the main reason why he no longer supported the Social Democrats:

The ability of the cultural elite to extend their cultural hierarchies to other social groups may possibly exist in relation to the middle classes, but it appears to be limited at the underprivileged pole. The highbrow cultural hierarchy is not necessarily met by an autonomous cultural hierarchy of the underprivileged.⁴ Rather, it is met by a politicised unwillingness to fund the activities of the cultural elite. The same opposition is found by Bennett et al. (2009: 201–5) in Britain. They argue that political critique is aimed at state policy, rather than at an identified bourgeois class.

Based on the quantitative analysis, it is among the less-privileged groups that we are most likely to find those who agree to discriminate between native Danes and immigrants regarding jobs. Also, we are likely to find those who take pride in their locality and nationality. Asked if some groups are better off than they ought to be, Janne, an unskilled factory worker in her forties, expands on her views on immigrants:

Janne expressed more hostility towards immigrants than was the case with the other informants, and she was supportive of the Danish People's Party. However, she makes a good example of the moral characteristics ascribed to some immigrants as unwilling to ‘settle in’ and to contribute to the welfare state by taking paid work. Other informants use the same type argument to express more open attitudes towards immigrants ‘as long as they accept the conditions here’ as expressed by a health care assistant (Bente). In both cases the willingness to ‘settle in’ is valued.

The first conclusion is a confirmation of the homology thesis advocated by Pierre Bourdieu (1984 [1979]). Inserting indicators of cultural preferences and politico-moral stances into a space of social positions makes distinct and recognisable patterns emerge. Regarding cultural preferences, the main opposition is between the privileged cultural quadrant and the underprivileged quadrants. At the privileged cultural quadrant, participation in and preferences for the legitimate and avant-garde art forms go together with preferences for new and exotic dishes and for a serious and liberal choice of newspaper. The privileged economic quadrant mainly differentiates itself from this pattern by an interest in money and private enterprise. In matters of politico-moral stances, the main opposition is again between the privileged cultural quadrant and the underprivileged quadrants. Here the culturally privileged quadrant is characterised by tolerance of ethnic minorities, political self-confidence, and a global orientation. The privileged economic quadrant differs from the cultural quadrant by an orientation towards economic growth rather than economic equality. Regarding preferences in political parties, there are indicators of a rightwing alliance between the economic fractions, and more so from privileged positions facing a leftwing alliance comprised by the cultural fractions.

The second conclusion to be drawn is that both cultural preferences and politico-moral stances are at work when symbolic boundaries are drawn. However, whilst examples of symbolic boundary-drawing based on cultural distinctions do exist, it is more pronounced, even within the fractions richest in cultural capital, to draw boundaries on the basis of politico-moral criteria. This is in accordance with Lamont (1992), who found that moral criteria may work as a source of symbolic boundary-making. It also gives some support to the qualitative studies from other Nordic contexts arguing that egalitarian and anti-elitist moral sentiments are likely to work against legitimate culture as a means of exclusion (Hjelseth 2005; Lien et al.2000; Skarpenes 2007). However, in relation to both of these arguments it is important to note that cultural and moral boundaries seem to be working together. Rather than one hierarchy replacing the other politico-moral attitudes seem to be structured according to the same logic as cultural preferences and practices. They are ordered according to the logic of the space of social positions and they are used for drawing symbolic boundaries towards the same social positions. Therefore it seems more reasonable to emphasise the importance of politico-moral evaluations within the framework developed by Bourdieu in Distinction rather than to dismiss the framework altogether.

The global vs. the local: One of the distinctions cutting across both the cultural and the politico-moral attitudes and modes of drawing boundaries is a divide between what may be termed the ‘global’ and the ‘local’. In the literature, the global-local distinction takes many forms. One is related to the movement of people, goods or information across geographical space. This would include travelling and being connected/using global networks of communication like the Internet (Larsen et al.2006: 46–49). In the current study, such a position is supported by the fact that not to use the Internet is a strong marker of the less-privileged positions, but also more generally in the attitudes towards travelling in the qualitative interviews. Others, Like Hannerz (1990), make a distinction between the cosmopolitan and the local as first and foremost a mode of orientation. The cosmopolitan orientation includes both a willingness to engage with different cultures and having the knowledge to do so. The locals, according to Hannerz, are characterised by an engagement in a single, often locally anchored culture. Bauman (1998), in his elaboration on the concept of ‘glocalization’, is more alert to the two-sided process of, on the one hand, the globalisation of wealth, power and the privileged and, on the other, the localisation of others. Finally the ‘global’ and ‘elite’ connotations often associated with cosmopolitanism is challenged by Lamont and Aksartova (2002). They do so by introducing the term ‘ordinary cosmopolitanisms’ to refer to the cultural repertoires of universalism by which black and white workers in the United States and white and North African workers in France manage to bridge racial differences in their everyday life, and on a local, rather than, global scale. Evidence of such practices among the less privileged can be found in expressions of acceptance of immigrants who are ‘willing to accept the conditions’, while boundary practices are organised around ‘those who won't settle in’.

This study clearly supports the claim that there is an opposition between ‘global’ and ‘local’ modes of orientation. This is found in diverse areas such as the food served to guests or in the preferred type of news. Regarding politico-moral attitudes, the divide is evident in attitudes towards immigrants as well as in attitudes regarding foreign aid. Finally, the privileged groups, and in particular those privileged by cultural capital, are least likely to take pride in their nationality or in their locality. In the qualitative interviews, this opposition is also emphasised in different ways. One expression is given by the male engineer who distances himself from his childhood friends on the basis of geographic and cultural mobility. This seems to concur with Larsen et al. (2006) as well as with the findings of Bennett et al. (2009), who found the working class to be depicted as ‘stuck’ and opposed to various privileged variants of being omnivorous. Finally, the global-local divide may be identified by a stronger tendency to discriminate between different nationalities within the less-privileged sections of the social space as well as in the shying of intolerant and racist attitudes which was a recurring theme in many of the interviews with those from privileged positions.

Finally, boundaries are not only drawn from the privileged towards the less-privileged positions. Antipathies towards the cultural elite are expressed through aversions against state subsidisation of highbrow culture. One possible interpretation is that the ability of the cultural elite to extend its own value hierarchy to the underprivileged groups is questioned. There are anti-elitist sentiments among underprivileged positions and a challenge of what is considered good taste, not only in the fine arts but also in moral values and in political practices (voting). What is valued in the less-privileged positions are various notions of being ‘ordinary’, of being able to ‘settle in’ and to be part of and contribute to a community and a shared culture (which from a global-cosmopolitan perspective may be conceived of as local, traditional and exclusionary or intolerant).

The research has been financed by a grant from The Danish Social Science Research Council and a contribution from Aalborg University. I wish to thank the anonymous reviewers for their helpful advice. Also I wish to thank my colleagues from the COMPAS-project; Annick Prieur, Lennart Rosenlund, Stine Thidemann Faber, as well as Gitte Harrits. Although I am sole responsible for this article, many of the findings draw on collective efforts. Other publications from the study quantitatively examines patterns of cultural preferences (Prieur et al. 2008) and political attitudes (Harrits et al. 2010). The social space construction presented here is inspired by but not identical to the earlier versions.

Jakob Skjøtt-Larsen is associate professor in sociology at Aalborg University, Denmark. His earlier work deals mainly with cultural sociology, changing class structures and trust.