Viewpoints and points of view: situating symbolic boundary drawing in social space

This article demonstrates how Pierre Bourdieu's concept of social space and Michèle Lamont's concept of symbolic boundaries can be fruitfully combined in cultural-stratification research. Focusing on the case of Stavanger, Norway, the analysis also shows how Multiple Correspondence Analysis is compatible with other research techniques, such as qualitative in-depth interviews. The approach adopted provides a practical application of Bourdieu's double reading of social relations. It combines the first, objective moment of situating 46 individuals subjected to qualitative interviews in the local social space (i.e. a system of relations between individuals’ possessions of cultural and economic capital) with the second, subjective moment of mapping the interviewees’ evaluations and classifications of other people's lifestyles. It is shown how intertwinements of various discursive repertoires of evaluation (cultural-aesthetical, moral-political and socio-economic) work in both contradictive and reinforcing ways to construct symbolic boundaries between classes and class fractions. The findings draw attention to both the capital volume and the capital composition dimensions of social space in that symbolic boundary drawing takes on different forms along these dimensions.

Scholarly debates about class, status and lifestyles were revitalised in the wake of the advances made by Bourdieu (1984, 1985, 1989). Bourdieu asserted that class relations express themselves through socially structured and symbolically significant lifestyle differences in contemporary societies. According to his homology model of cultural stratification, the system of class relations (defined as a social topography reflecting a differential distribution of capitals) has structural affinities with the system of status relations (defined as a social topography of lifestyles properties, such as dress, speech, bodily dispositions, aesthetical preferences and moral-political stances). This homologous relationship, moreover, is linked to the monopolisation of advantages and opportunities in social life, as well as exclusion in terms of group formation.

Since Bourdieu's initial advances based on French data from the 1960s and 1970s, an expanding body of research has investigated the nature of lifestyle differences in other empirical cases. The most heatedly debated question concerns the social distribution of lifestyle properties, particularly those related to cultural consumption and aesthetical tastes. The crucial question here is whether lifestyles vary by class or whether this distributional principle is diminishing and/or being replaced by other principles. This is the heart of the so-called cultural omnivore debate (see, e.g. Chan and Goldthorpe 2010; Hazir and Warde 2016; Peterson 2005). A related debate has focused on processes pertaining to group formation and social exclusion. One of the main concerns here is whether and how social actors’ evaluations and judgements of other people's lifestyles constitute boundaries between groups of social actors (see, e.g. Friedman and Kuipers 2013; Lamont 1992; Skjøtt-Larsen 2012).

In this primarily theoretical–methodological paper, I will argue that these analytically distinct aspects of the class–status nexus tend to be conflated in unfortunate ways in contemporary cultural-stratification research. Based on a methodological one-sidedness, scholars have typically tended to investigate empirically only one of these aspects while making unsubstantiated claims about the other. Aiming to counteract this unfortunate tendency, I employ an integrated approach which accommodates objections raised to ‘objectivist’ cultural-stratification research preoccupied with mapping the social distribution of lifestyle properties, as well as objections raised to ‘subjectivist’ research preoccupied with the subjective-discursive aspect of symbolic boundary drawing. I discuss how an integrated empirical approach can be achieved by returning to Bourdieu's double reading of social relations, i.e. by incorporating both subjective and objective aspects of the class–status nexus.

I use empirical data from a study of cultural stratification in the city of Stavanger, Norway, to demonstrate the practical application of this approach by combining quantitative and qualitative methods. First, a local social space is constructed on the basis of survey data by means of Multiple Correspondence Analysis (MCA). Second, the positions of 46 individuals invited to participate in qualitative interviews are then located within the constructed space. Finally, the interviewees’ symbolic boundary drawing vis-à-vis other people's lifestyles is analysed by taking into account their location in the social space. Thus, the analysis maps the relationship between the interviewees’ points of view with regards to other people's lifestyles on the one hand and, on the other, the vantage points in the social space from which the points of view are taken.

In order to break with the tendency to push social research into the ditch of either ‘objectivism’ or ‘subjectivism’, Bourdieu paved a middle way to study the interconnectedness between social structures and mental structures (see, e.g. Bourdieu 1996: 1–6). Bourdieu endeavoured to map empirically the correspondences between ‘objectivity of the first order’ – the system of relations between individuals’ possession of scarce resources – and ‘objectivity of the second order’ – the system of classification and the mental and bodily schemata that function as symbolic templates for the practical activities of social actors (Wacquant 1992: 7).

Bourdieu's (1984, 1985, 1989) account of the class–status nexus rests upon three key concepts: the social space, the symbolic space and (class) habitus. The multidimensional social space objectifies the system of relationships between social positions. The structure of the social space is shaped by the social distribution of various forms of capital and the relative strength between them. The symbolic space depicts distributional oppositions between individuals’ properties in terms of cultural practices and lifestyles. Bourdieu asserts that the symbolic space is homologous to the social space, meaning that the two are structured in similar ways, and that clusters of positions in the social space roughly correspond to distinct lifestyles. The third component in Bourdieu's model is the notion of habitus which is held to be the mediating factor between the social space and the symbolic space. Habitus refers to individuals’ generative schemes of perception and appreciation guiding practice and is thought to consist of durable dispositions inscribed in body and mind.

Although Bourdieu explicitly warns against confusing ‘classes on paper’ with ‘real classes’, and thus presuming erroneously that social groups will automatically emerge, for instance from conflicting interests on the labour market, he nevertheless holds that proximate positions in social space represent ‘probable classes’ (Bourdieu 1985: 725ff). Through affinities in the structures of the habitus, people are regarded as ‘disposed to get closer, as well as being easier to bring together, to mobilise’. However, for ‘classes on paper’ to become ‘realised classes’ they have to be constructed as unified collectives through symbolic representation. Class struggles are in fact classification struggles between opposing groups of social actors and this pertains to the categories of perception of the social world. Open class struggles are, however, a rarity. Instead, social space ‘tends to function symbolically as a space of lifestyles or as a set of Stände, of groups characterised by different lifestyles’ (Bourdieu 1985: 730). Although the formation of class-structured status groups based on social estimations of lifestyles is not a primary concern in Bourdieu's empirical inquiries, he holds that lifestyle differentiation and social exclusion in terms of group formation should be viewed as analytically distinct processes and that their interconnectedness in the empirical world should be explored.

This crucial point is often overlooked in contemporary cultural-stratification research. An unfortunate analytical conflation is particularly evident in the debate about the cultural omnivore where researchers have discussed whether Bourdieu's advances are outdated due to a rise in broad and eclectic aesthetical orientations in the upper strata of society. What cultural omnivorousness actually means, how it should be measured and what a rise in it entails with regards to group formation and exclusion processes is, however, highly contested (see, e.g. Flemmen et al. 2017; Hazir and Warde 2016; Jarness 2017; Peterson 2005; Robette and Roueff 2014). One way of interpreting the apparent rise in omnivorous tastes is that the more social actors situated in the upper regions of social space tend to do or like the same things as people situated elsewhere in this space, the less distinctive their lifestyles and tastes – and, accordingly, the less the exclusionary impact of lifestyles. This is what DiMaggio and Mukhtar (2004) have referred to as ‘the meltdown scenario’.

Whereas some authors argue that the rise of omnivorousness implies a rebuttal of the homology model (e.g. Chan and Goldthorpe 2010), others argue that the development of new hybrid consumption styles and aesthetical orientations is conceivable within a Bourdieusian framework. Recently, cultural-stratification research has been focused on the ways in which certain transgressive cultural practices, preferences and tastes constitute a form of cultural capital, i.e. a type of aesthetical disposition that entails privileges and advantages in social life. According to this view, the rise of the cultural omnivore does not indicate the erosion of cultural capital as such but instead heralds the rise of ‘new forms of distinctions’ and the increasing impact of an ‘emerging’ or ‘cosmopolitan’ form of cultural capital (Bennett et al. 2009; Prieur and Savage 2013).

Both the ‘meltdown argument’ and the ‘emerging cultural capital argument’ are, however, underpinned by problematic assumptions about the relationship between lifestyle differences and processes of exclusion and the monopolisation of advantages and privileges. The former presupposes that eclectic or hybrid cultural preferences imply openness, not only to a broad variety of cultural goods and activities but also to other cultural consumers. The latter presupposes that class-structured lifestyle differences entail social exclusion and monopolisation, as implied in the capital metaphor. Such inferences are hardly convincing, however, as there is nothing contradictory about the idea that people might display preferences for a wide range of cultural goods, while harshly judging other people's preferences and tastes (Lahire 2008). Conversely, even though different tastes are systematically distributed according to the structures of social space, it does not necessarily follow from this that estimations of honour or prestige are involved, or that certain tastes function as embodied forms of cultural and/or symbolic capital.

One promising approach geared towards exploring the insufficiently studied link between lifestyle differences and processes of exclusion and monopolisation has been advocated by Lamont (1992, 2000; Lamont and Molnar 2002). In particular, Lamont has problematised the assumption that lifestyle differences lead directly to hierarchisation and group formation. Instead, through qualitative interviews, she has mapped empirically the various repertoires of evaluation used by people to demarcate themselves symbolically from others. Such judgements, it is argued, lead people to seek or avoid interaction with others. The studies have revolved around the notion of symbolic boundaries, referring to the everyday conceptual distinctions made by social actors to categorise people, practices, tastes, attitudes and manners. Of particular interest with regards to the focal point of this paper is the way in which Lamont (1992) has documented the discursive repertoires of evaluation drawn upon by upper-middle-class men in France and the United States. The analytical strategy involves mapping the relative salience of various subtypes of symbolic boundaries. Lamont extracts three subtypes from her interview data: cultural, socio-economic and moral boundaries.¹ Since American middle-class men draw stronger moral boundaries than cultural boundaries compared to their French counterparts, Lamont holds, this finding rebuts Bourdieu's universalistic advances concerning the exclusionary impact of class-structured differences in cultural tastes and aesthetical orientations (Lamont 1992: 181–8).

Although certainly promising with regards to bridging certain gaps in cultural-stratification research, this symbolic boundary approach does, however, involve some serious problems of its own. As discussed elsewhere (Jarness 2017; Jarness and Friedman 2017; see also Holt 1997), these include: the idea that the existence of symbolic boundaries constitutes a necessary precondition for the existence of objectified social boundaries; the a priori analytical precedence of discursive consciousness to the neglect of practical consciousness; the idea that the relative strength of subtypes of symbolic boundaries can be tapped simply by measuring their occurrences in qualitative interview data; and, an underdeveloped concept of class structure, particularly with regards to structural horizontal divisions between fractions within the upper and middle classes. Nevertheless, a modified version of Lamont's approach could potentially provide a deeper understanding of whether and how lifestyle differences are manifested discursively in the ways in which people evaluate and classify others. In the following analysis I aim to do just that, while drawing on insights from the Bourdieusian tradition.

The analysis expands on a series of studies of the city of Stavanger conducted by Lennart Rosenlund (2009, 2015). Using MCA and other techniques in novel ways, Rosenlund has deployed and extended Bourdieu's theoretical–methodological framework in empirically mapping the historical evolution of the local class structure, as well as the social distribution of lifestyle properties. Rosenlund's studies point to structural affinities between the local social space and the local symbolic space that resonate well with the homology model advanced by Bourdieu.

Expanding on Rosenlund's (2015) construction of a local social space based on data from the Stavanger2008 European Capital of Culture survey, 46 Stavangerians were strategically recruited for in-depth qualitative interviews. The interviewees were recruited so as to ensure a dispersion of cases structurally situated along the two most important axes of the multidimensional social space: the vertical capital-volume dimension and the horizontal capital-composition dimension. Due to anonymity issues, it was not possible to recruit a subsample from the original sample of survey respondents. An alternative strategy was employed by using the Bourdieu-inspired class scheme the Bourdieu-inspired Oslo register data class scheme (ORDC) (Hansen et al. 2009) for recruitment purposes, and then locating the interviewees in the social space constructed on the basis of the Stavanger2008 survey at a later stage. The details of this procedure will be outlined below. The sample includes 25 men and 21 women aged 25–67 (avg. 39.8). All the interviewees are ethnic Norwegians residing and working within the borders of the municipality of Stavanger.

I conducted the interviews myself between September 2009 and May 2010 while the survey was being carried out. I used a semi-structured list of questions constructed so as to invite the interviewees to classify and evaluate their own and other people's lifestyles, with a particular emphasis on both material and non-material consumption styles. The list of questions was inspired by that of Lamont (1992, 2000) but I also added specific questions adjusted to the local context (e.g. specific places, establishments and urban areas in Stavanger).

Drawing on Bourdieu's (1984) notion of a field of cultural consumption I analyse classifications, evaluations and judgements uttered by interviewees as position-takings within this field. I thus view interviewees’ classificatory practices as being implicated in a more or less misrecognised game in which the definitions of what constitutes socially esteemed lifestyles are one of the fundamental stakes. The analytical strategy involves two basic steps. First, the 46 interviewees are located in the local social space by means of MCA. Second, different types of boundary drawing manifested in the qualitative interviews (i.e. the discursive demarcations manifested in the ways in which interviewees talk about other people's practices, tastes and manners) are analysed in terms of the interviewees’ situatedness in the social space. In focusing on the discursive level, the analysis is limited to explicit, antagonistic position-takings and not the taken-for-grantedness of doxa (see Bourdieu 1977: 159–71).

The local social space in Stavanger is constructed by means of Specific MCA which is a statistical technique that provides a geometric model of categorical data. It is used to detect and visually represent underlying structures in a data set (Le Roux and Rouanet 2010). It proceeds by calculating chi-squared distances between individuals (rows) and categories (columns) in a data matrix. The geometrical distances between rows and columns are maximised in order to reveal the underlying dimensions best suited to describing the central oppositions or distributional differences in the data.

When interpreting the results from an MCA, one focuses on two clouds. The cloud of categories facilitates a visual assessment of the coordinates of all the categories used in the construction of the space. An inspection of the categories’ contributions to the axes, as well as the distances between the categories, is of great importance to the interpretation of the substantive sociological meaning of the oppositions within the space. In Rosenlund's construction of the local social space in Stavanger, the cloud of categories can reasonably be interpreted as indicating a differential distribution of cultural and economic capital following a dual distributional logic. The variables used in the construction of the space include: household income, the value of one's house, the value of a second home, the value of one's car, one's education level, education type, occupation, occupational sector, father's education, father's occupation, the number of books at home, the frequency of theatre visits last year (for more details on the construction of the space, see Jarness 2013; Rosenlund 2015).

The method used to strategically recruit individuals for qualitative interviews and locating them in the social space is inspired by the Danish COMPAS project (Faber et al. 2012; Skjøtt-Larsen 2012), where an innovative combination of MCA and qualitative interviews was employed.² In the case of Stavanger, it was not possible to recruit a subsample from the original sample of survey respondents due to anonymity issues. Therefore, an alternative strategy was employed. First, I used the ORDC class scheme (Hansen et al. 2009) – this mimics Bourdieu's model of the social space by categorising occupations vertically (according to assumed capital volume) and horizontally (according to assumed capital composition). I then picked a number of occupations typical of each category in this scheme. I then searched for potential interviewees through the online Yellow Pages and Google, using ‘Stavanger’ and the various occupational titles as search queries. Contact was established by email or telephone, either personally or through secretaries. Second, at the end of the interview sessions, the interviewees were subjected to parts of the original survey questionnaire to obtain information on their scores for the variables used in the construction of the social space. Third, information about the interviewees was coded and then added to the original survey data set. Finally, by way of Specific MCA the local social space was reconstructed with the 46 interviewees projected onto the space as supplementary cases. Figure 2 depicts how the individuals subjected to qualitative interviews are dispersed in the space, relative to those participating in the original survey. The interviewees are well dispersed along both the capital-volume and the capital-composition dimensions of the space.

In what ways do the interviewees draw symbolic boundaries between themselves (i.e. ‘people like us’) and those with whom they would rather not associate (i.e. ‘the others’)? In the following I shall briefly outline how the boundary-drawing dynamic is manifested in somewhat different ways along the two most important dimensions of the social space: the vertical dimension, distinguishing between people endowed with high and low volumes of capital, and the horizontal dimension, distinguishing between people endowed with a preponderance of either cultural or economic capital.

Interviewees located in the upper regions of the social space typically demarcate themselves from people referred to as ‘lower-class types’, ‘the working class’, ‘the common people’ and such like. Although these interviewees are reluctant to claim superiority over such people, there is no doubt that they are viewed as somewhat different from themselves and the types of people with whom they associate socially. Several evaluative criteria are activated when such people are depicted: they supposedly lack knowledge, education and aesthetic sensibilities; they are vulgar in their material consumption choices; they live unhealthy lives; and they exhibit dubious moral-political values.

Many interviewees in the upper regions of the social space – particularly those endowed with a preponderance of cultural capital – report a distinct commitment to their consumption of cultural goods, such as certain forms of literature, theatre and visual art. People who do not share their enthusiasm are often regarded as rather ‘dull’. Demarcations are typically expressed against the ways in which ‘common people’ embrace ‘popular’ and ‘folkish’ forms of entertainment. Evaluative criteria related to knowledgeability and aesthetics are also intertwined with politics and the ability to reflect critically on societal processes:

Assuming that ‘idiots’ will shun him for his interest in political and aesthetical issues, this university-educated journalist, Thomas, rather self-ironically identifies with the ‘cultural elite’. This term is widely used in colloquial Norwegian and in public debates and it has been increasingly deployed to mean something negative, unpleasant and undesirable (Haarr and Krogstad 2011; Ljunggren 2017). Interestingly, Thomas links such anti-elitist sentiments to the Progress Party (Fremskrittspartiet), the third-largest party in the Norwegian Parliament and currently a part of the ruling coalition government with the Conservative Party. Its politics are typically described as ‘right-wing’ or ‘populist’ by interviewees located in the cultural fraction, most of whom adhere to parties to the left of the Norwegian multiparty spectrum. Accused of possessing dubious morals (e.g. ‘racism’, ‘homophobia’, ‘egoism’), as well as aesthetical flaws (e.g. ‘vulgarity’, ‘flashiness’), people adhering to the politics of the Progress Party are often deemed ‘unintelligent’, ‘feeble-minded’ and ‘dumb’. This highlights that different repertoires of evaluation – in this case cultural-aesthetical and moral-political ones – are intertwined and work in mutually constituting and reinforcing ways.

Interviewees in the upper regions of the social space – particularly those endowed with a preponderance of economic capital – are also opinionated about other people's consumption styles related to the material consumption of goods. These interviewees are far more concerned with consumer brands than the other interviewees in the sample and they typically shun brands connoting something ‘cheap’ and ‘generic’. Such classificatory distinctions are also reflected in their demarcations against people perceived as exhibiting ‘poor’ consumption styles. Explicit boundary drawing is particularly salient when it comes to looks, appearances and clothing styles:

Two typical figures of distaste recur in the data; one figure is depicted as ‘trashy’, ‘vulgar’, ‘loud’ and ‘cheap’, whereas the second figure is not defined so much by the presence of bad taste as a lack of good taste. Both are typically associated with ‘lower’ and ‘simple’ types. As we can see above, the wealthy lawyer, Pia, even depicts such people as belonging to a ‘lower class’. Derogatory classifications of other people's consumption styles are also related to a perceived conformity, whether in terms of clothing style, interior design or one's choice of motor vehicle. Valorising ‘individuality’ and ‘exclusivity’ themselves, the conformity of others seems to puzzle many interviewees located in the upper regions of the social space.

So far we have seen examples of how interviewees located in the upper regions of the social space view with suspicion the lifestyles of ‘lower-class types’. But what about the interviewees located in the lower regions? First of all, it has to be borne in mind that these interviewees express a distinctly practical orientation to culture, implying their shunning cultural goods which are ‘too complicated’, as well as ‘overly expensive’ or ‘unnecessarily fancy’ material goods. Unless goods provide them with immediate satisfaction in terms of entertainment, comfort or practical knowledge, they do not see any point in craving them. Accordingly, people who are perceived as craving such goods are viewed with suspicion:

Although most of these interviewees report that they too might enjoy having goods such as cars and boats, what is perceived to be ‘excessive luxury’ and ‘squandering’ is clearly frowned upon and linked to dubious morals. However, most of the interviewees located in the lower regions of the social space report that they do not have a problem with consumers of ‘complicated’ cultural goods or ‘flashy’ material goods as long as such people do not ‘look down on them’ for not wanting to conform to their lifestyles and tastes. This is the case with the secretary, Eva, who expresses scepticism of what she perceives to be the dubious tendency of the rich to denigrate people who do not care about expensive clothing brands: The contention that education, money, the possession of goods or the mastery of certain lifestyles ‘do not count’ in the calculation of human worth is a recurring theme expressed by these interviewees. Instead, moral qualities such as being kind, caring, respectful and thoughtful seem to be the most important – and in some cases the only – criteria used when judging other people's lifestyles. If others breach these moral standards, sharp boundaries are drawn, especially if those who breach them are perceived to be rich and powerful.

As summarised in Table 1, there are clear signs of mutual disapproval between interviewees located in the upper and lower regions of the social space. When it comes to the evaluative repertoires drawn upon, those located in the upper region are much more prone to judge others based on cultural-aesthetic and socio-economic criteria than those in the lower region. They do, however, share the tendency to judge others based on moral criteria but – crucially – in different ways: those in the upper region denigrate the ‘lower-class types’ for practising activities and exhibiting attitudes that are perceived as morally dubious (e.g. racism, homophobia and promiscuity). Those in the lower regions do not so much denigrate the ‘upper-class types’ for their intrinsically immoral lifestyles, as much as they denigrate them for being uppish and for thinking that they are better than others because of what they are or what they do.

Although almost all of the interviewees located in the upper regions of the social space acknowledge their privileges and advantages vis-à-vis others, they typically conceive of themselves as ‘normal’ and ‘down-to-earth’ since they, according to themselves, are ‘humble’, ‘respectful’ and ‘accommodating’ towards those who are ‘less fortunate’. When this code of modesty is regarded as breached, the interviewees in the upper regions of the social space typically draw on moral evaluative repertoires to demarcate themselves from such ‘self-assertive’ individuals and lifestyles. These demarcations are similar to those expressed by interviewees in the lower regions; however, they are in fact much more articulated and specific in comparison.

Although not a perfectly clear-cut division, there is a strong tendency for interviewees who are endowed with a preponderance of cultural capital to express their aversion to the self-assertiveness of ‘the rich’ and ‘the economic elite’, whereas interviewees endowed with a preponderance of economic capital express aversion to the self-assertiveness of ‘the snobs’ and ‘the cultural elite’. Interestingly, they draw strikingly similar moral boundaries against other people's supposed elitism and self-assertiveness. Yet it is quite clear that the practices, tastes and lifestyles that elicit this kind of distaste are quite different. Amongst those located in the cultural fraction of the upper regions of the social space, the ultimate target for their aversion is people who lead expensive lifestyles rendered possible by the booming oil industry:

Several interviewees reported feeling marginalised in an environment perceived as inhabited and dominated by people conspicuously showing off lifestyles, tastes and values alien to their own. These interviewees regard Stavanger as pervaded by ‘egoistic’ and ‘individualistic’ values, represented by right-wing political attitudes, a striving for material luxury and the conspicuous exhibition of expensive status symbols. The loudness and flashiness of the rich are typically linked to character flaws such as leading a ‘shallow inner life’. An important moral distinction between being ‘stylish’ and simply being ‘vain’ seems to be particularly important. The accusation of vanity is also related to the sphere of cultural consumption. The recurring theme is that the rich do not really like the sorts of cultural goods and activities as they themselves do:

This quote highlights the importance of separating between the whats and hows of cultural consumption: a given cultural good can be perceived, appreciated and appropriated in different ways, meaning that taste cannot be directly inferred from the objects people do or do not prefer (Bourdieu 1984; Holt 1997; Jarness 2015). The hows of cultural consumption also seem important to the interviewees: there are several examples of boundary drawing related to the ‘wrong’ way of appreciating the ‘right’ cultural goods and this seems to be a hallmark of the undistinguished. In this particular case, Sara regards those who attend performances by the Stavanger Symphony Orchestra without a proper aesthetic appreciation as ‘pathetic’ and ‘ridiculous’. This runs counter to Daenekindt and Roose's (2017: 40–1) assertion that the hows of cultural consumption are not accessible to observers and that they thus cannot function as status markers.

Those endowed with relatively more economic than cultural capital, on the other hand, tend to demonstrate similar contempt for the pretentious lifestyles of those whom they conceive of as ‘the cultural elite’. What seems to elicit most resentment is the fact that both cultural producers and the consumers of certain cultural forms receive state funding. For those unfamiliar with or uninterested in such culture, it is simply incomprehensible why the state hands out money to artists and their faithful ‘acolytes’:

Notwithstanding the various connotations the term ‘cultural elite’ may have for different people, the feelings of resentment towards those who are conceived of as consumers or producers of high culture belonging to certain social cliques are fairly widespread. Interestingly, the boundary drawing directed against what is perceived as the ‘snobbish’ lifestyles of the cultural elite is intertwined with morality and politics in similar ways as the boundary drawing directed against the rich. Cultural preferences perceived as ‘strange’ or ‘weird’ are interpreted as a pretentious and dubious quest for specialness. A supposedly obsolete and left-wing counterculture combined with intellectual snobbery seems to elicit particular resentment: Implicitly, ordinariness is deemed to be more honourable than the pretentious efforts of people who show off their intellectual abilities. Just as we saw how interviewees in the cultural fraction perceive the rich as flashy and pretentious, the interviewees in the economic fraction perceive the cultural elite as endowed with equally dubious motives: their cultural style is associated with showing off in ways deemed ‘alien’ and even ‘loathsome’.

As summarised in Table 2, interviewees in the cultural and the economic fractions of the upper regions of the social space express a clear sense of mutual antipathy. Interestingly, they draw on very similar evaluative repertoires in symbolically constructing otherness: ‘they’ are morally dubious types of people due to their self-assertiveness and elitist attitudes and ‘we’ despise them for that. However, the lifestyles referred to when making such moral judgements are clearly different. Those located in the economic fraction infer a lack of moral qualities from a supposed elitism implied by cultural consumption and cultivating esoteric knowledge, while those located in the cultural fraction infer a lack of moral qualities from a supposed elitism implied by exaggerated material consumption.

This analysis has demonstrated how Pierre Bourdieu's concept of social space and Michèle Lamont's concept of symbolic boundaries can be combined in cultural-stratification research. We have seen how symbolic boundaries are systematically drawn along both the capital volume and the capital composition dimensions of the local social space. This suggests that classes and class fractions existing ‘on paper’ are tied to subjectively constructed symbolic boundaries between groups of social actors. In others words, the findings support the Bourdieusian notion that classes and class fractions tend to take the form of more or less amorphous status groups in contemporary society.

The interviewees draw on cultural-aesthetical repertoires of evaluation in distinguishing themselves from ‘vulgar philistines’; they draw on socio-economic repertoires in distinguishing themselves from the undistinguished material consumption styles of the less economically fortunate; and, they draw on moral-political repertoires in distinguishing themselves from the bad manners of self-assertive elites. The extent to which each individual interviewee draws on these repertoires varies. But on the whole, moral-political repertoires are by far the most frequently drawn upon. This resonates well with other Norwegian studies pointing to particularly strong egalitarian sentiments in Norwegian society (Gullestad 1992; Jarness and Friedman 2017; Ljunggren 2017). But are we thus to conclude that egalitarian sentiments and moral boundaries somehow counteract or outweigh the exclusionary impact of other types of symbolic boundaries, as Lamont (1992) suggests in her comparison of the French and the American cases?

It is undoubtedly fruitful to distinguish between various subtypes of symbolic boundaries on an analytical level. However, this analysis has demonstrated that there is a need for a more complex analytical strategy than simply measuring the relative salience of various repertoires of evaluation in terms of their numeric occurrences in qualitative interview data. Although moral repertoires are undoubtedly the most salient in the present empirical case, the kind of argument put forward by Lamont is problematic for several reasons. First of all, moral boundaries are typically expressed as demarcations against what is perceived as ‘unnatural’ or suchlike. Implicitly, the taken-for-grantedness of normality is used as a yardstick against which others are judged. However, whatever is conceived of as normal, natural and ordinary varies systematically in that interviewees located in different regions of the social space have quite different conceptions of ordinariness. So although interviewees located in the cultural fraction of the upper region of the social space apparently draw on similar moral repertoires as interviewees located in the economic fraction when distinguishing themselves from self-assertive others, the actual practices, tastes and lifestyles against which they demarcate themselves are clearly different. In other words, drawing on similar repertoires of evaluation should not be interpreted as some kind of group affiliation without taking into account the practices, tastes and lifestyles against which interviewees demarcate themselves. Not taking this aspect into account in the present analysis would have effectively masked the crucial intraclass tensions along the capital composition dimension of the social space.

Second, the various types of repertoires depicted by Lamont very seldom appear in any pure form. Moral repertoires of evaluation are typically tightly intertwined with other types of repertoires. For instance, people adhering to the politics of the Progress Party are despised by many interviewees and they are regarded as both ‘vulgar philistines’ and ‘racist homophobes’. Undistinguished cultural tastes and immoral political preferences are seen as two sides of the same coin. In this case, cultural-aesthetical and moral-political repertoires work in mutually constitutive and reinforcing ways, rather than in competing and contradictive ways. Thus, whether and how different repertoires work in either contradictive or reinforcing ways should be an empirical question to be explored, not something that should be built a priori into an analytical strategy.

Finally, the very same interviewees draw on different combinations of discursive repertoires when classifying different people, depending on who and what the classified people are perceived to be. It seems the interviewees are more prone to draw on moral repertoires when talking about people above themselves in a subjectively perceived social hierarchy, compared to how they classify people perceived as below them. This seems to be the case across the social space. Although it is a fruitful analytical endeavour to explore each interviewee's habitus (or ‘mental maps’ as Lamont calls it) in terms of the evaluative repertoires drawn upon, the relationship between the boundary drawers and the targets of their judgements must be taken into account.

These crucial points should not, however, be read as an attempt to rebut the symbolic boundary approach to cultural-stratification research. They should instead be read as a contribution to developing further this promising approach in order to bridge the gaps between objectivist and subjectivist analytical frameworks. By considering the situatedness of symbolic boundary drawing in social space, as well as the intertwinements of different repertoires of evaluation, the hitherto insufficiently studied link between lifestyle differences and processes of social exclusion and group formation can be explored in more fruitful ways. Herein lies the potential for mapping whether and how practices and tastes function as a form of capital in various contexts, as well as the potential for critically assessing a core assumption underpinning research on cultural omnivorousness, namely that inferences about social actors’ ‘snobbish exclusion’ of or ‘openness’ towards others can be drawn simply by mapping their cultural consumption preferences. This analysis has demonstrated that the ways in which social actors classify, evaluate and judge other people's lifestyles deserve an analytical dimension of their own.

I would like to thank Lennart Rosenlund for helping me to project my respondents onto his model of the social space, as well as Magne Flemmen, Johs. Hjellbrekke and two anonymous European Societies reviewers for their insightful comments on earlier drafts of this article.

No potential conflict of interest was reported by the author.

Vegard Jarness is Postdoctoral Fellow at the Department of Sociology, University of Bergen. His research interests include social inequality, class and cultural stratification. Recent publications have appeared in Sociology, British Journal of Sociology, Poetics and The Routledge Companion to Bourdieu's ‘Distinction’.