Abstract
The debate over the rise of eclecticism, more particularly Peterson’s ‘omnivore thesis’, has received much attention over recent years. For Lahire, eclecticism reflects less an increasing individual openness to a variety of cultural styles than intra-individual dissonances. Based on a large survey into cultural practices in the Wallonia-Brussels Federation (Belgium – N = 2021), this article compares these two notions with the aim of understanding the complexity of the structural relations that organise cultural tastes and practices into ‘cultural profiles’. I argue that omnivorousness cannot be reduced to dissonance but instead both notions characterise different configurations of cultural choices. More importantly, I show that identifying omnivorous and dissonant patterns matters less than understanding how these patterns emerge from tensions between existing and emerging cultural hierarchies at the individual and social levels.
Keywords
Introduction
In cultural sociology, Bourdieu’s Distinction (1979) has been the starting point for an impressive volume of research, including research that emerges as a critique of his work or at least some aspects of it. Among the latter, two theories are particularly noticeable: Peterson’s ‘omnivorousness’ (e.g. Peterson and Kern, 1996) and Lahire’s later ‘dissonance’ (2006). While discussions about Peterson’s work have dominated debates on patterns of cultural choice, especially in the Anglo-Saxon world, Lahire’s theory has turned out to be very influential but to a greater extent in the French-speaking world. Although both ‘omnivorousness’ and ‘dissonance’ could seem at first sight to share many similarities, Lahire actually used his concept to offer a critical view on omnivorousness as resulting from widespread intra-individual dissonance rather than from greater tolerance mainly restricted to upper social categories. Yet, despite Lahire’s comments providing much food for thought for anyone interested in the interpretation of omnivorousness (e.g. Bellavance et al., 2006; Rimmer, 2011), the links between the two notions have been largely left unexplored. Relying only on music tastes, Daenekindt and Roose’s (2014) article constitutes a welcome exception, arguing that dissonance and omnivorousness are conceptually distinct. Yet it does not explore empirically the differences and assumes that omnivores are the ‘most dissonant’ because they cross highbrow and lowbrow boundaries.
This article aims to compare the empirical realities of dissonance and omnivorousness in order to improve our understanding of what these two notions entail. It poses fairly simple but nonetheless important questions: Can omnivorousness be understood through dissonance? What is the link between the two? Can one be reduced to the other? This article demonstrates that one cannot simply allow dissonance to explain the phenomenon of omnivorousness, and that if we systematically compare their empirical manifestations, they appear as two relatively distinct configurations of cultural profiles. Additionally and perhaps more pertinently, responding to such questions requires that we ask ourselves about the way that ‘cultural profiles’ (a term borrowed from Lahire, 2006) are constituted. Inspired by a Bourdieusian approach, these profiles are defined here as a structured ensemble of cultural practices, of leisure activities and tastes which both classify and can be classified. Responding to the call by sociologists such as Savage et al. (2005), this article considers that it is less useful to know who is omnivorous (or dissonant) than to understand ‘how cultural forms are organized in hierarchical (or non-hierarchical) ways’ (p. 8), and how these organisations can form omnivorous and dissonant patterns. I will pay attention to whether dissonance and omnivorousness – if observable – entail similar relations to the distinction between high and low cultures, and, more specifically, whether they are associated with the maintainance or collapse of cultural boundaries.
I start by discussing both Peterson’s and Lahire’s theories. Then I explain the way I have operationalised and measured both concepts, and also the notion of (dominant) cultural legitimacy, which arguably underpins them. Following this, I provide empirical evidence that omnivorousness and dissonance are related to each other but nevertheless constitute different phenomena. At the end of the article, I discuss the implications of my findings and show how the now traditional approach in cultural sociology of identifying cultural omnivores has drawn researchers’ attention away from understanding how cultural patterns emerge from tensions between existing and emerging cultural hierarchies at the individual and social levels.
The Snob and the Omnivore
The figure of the ‘snob’ is linked sociologically to Bourdieu’s (1979) study of mechanisms of distinction, even though Bourdieu himself never used the term (Bennett et al., 2009: 12). According to Bourdieu (1979), an individual’s social position depends on her or his level of cultural, economic and social capital that defines her or his lifestyle, considered as a set of practices, consumed products and tastes that can be classified and which in turn classify people. He advances the principle of homology guaranteed by habitus, which tends to make lifestyle space correspond to social space. In this view, high culture is linked to the upper classes, those richest in terms of economic and, more importantly, cultural capital. They seek to distinguish themselves from the middle and lower classes and their middlebrow and lowbrow orientations.
Since the 1990s, the Bourdieusian perspective has suffered from almost constant attack by supporters of ‘omnivorousness’ (Peterson, 2005; Peterson and Kern, 1996; Peterson and Simkus, 1992). The idea is simple: the distinction which opposes high culture to low culture has been progressively losing its relevance in favour of one differentiating omnivores and univores. The very principle of homology is called into question by the fact that while upper social strata have higher chances of liking high culture, they also enjoy more popular cultural forms. The lower strata are thus considered as ‘univores’, investing themselves in a narrower range of activities. Interestingly, one could interpret this movement from snob to omnivore as the progressive disappearance of a certain kind of univore, defined by a disposition towards high culture. More precisely, in this article, I will follow Karademir Hazır and Warde’s encompassing definition: cultural omnivorousness refers to the appearance of new ‘repertoires of cultural practices […], which are marked both by an increased breadth of cultural tastes and participation and by a willingness to transgress previously entrenched boundaries between hierarchically ranked cultural items or genres’ (2016: 77). The rise of the omnivore is due, according to Peterson (2005), to social structural and cultural changes, such as social mobility (van Eijck, 1999) and an aestheticisation of popular culture (notably by the media).
At first sight, this growing eclecticism interpreted as omnivorousness by Peterson marks a profound transformation of the link between social stratification and cultural classification, but not necessarily its collapse. Research has indeed shown omnivorousness to be a feature of younger, highly educated people (López-Sintas and García-Álvarez, 2002; Peterson and Kern, 1996). However, while the statistically positive link between educational level and social position on the one hand, and omnivorousness on the other, has never to my knowledge been refuted, the nature of the relation between the latter and age appears essential but more complex. Peterson more recently revealed that younger cohorts were nowadays less likely to enjoy high culture (Peterson, 2005; Peterson and Rossman, 2008; Rossman and Peterson, 2005), suggesting a sort of de-intellectualisation or a popularisation of cultural referents. Drawing on the terminology of Warde and Gayo-Cal (2009), one can say that younger people are increasingly likely to be omnivores by volume (in the number of tastes and activities favoured) rather than by composition (a variety of tastes and practices characterised by diverse legitimacy). Warde and Gayo-Cal (2009) have actually shown that it was not necessarily younger people who were the most omnivorous but the middle aged. This appears to be supported by Donnat’s ‘Branchés’ (2004) (which could be translated as ‘connected’), a group in middle age characterised by broad cultural resources. Those of middle age have had greater exposure to diverse social contexts and have taken on board a greater diversity of cultural referents.
That certain socio-demographic characteristics are producing divergent results marks an underlying problem with the notion of omnivorousness; an interpretative plasticity due in part to many different operationalisations (Karademir Hazır and Warde, 2016; Robette and Roueff, 2014). Despite this problem, the existing link with categories of age, social class and education suggests that omnivorousness constitutes a ‘boundary-drawing mechanism’ (Lizardo and Skiles, 2013) that is both vertical and horizontal, as younger people’s appreciation of emerging cultural forms (Prieur and Savage, 2013) has become central for grasping the diversity of cultural profiles. There can thus be various ways of being ‘open to diversity’, to paraphrase Ollivier (2008; see also Bellavance, 2008). Mixing political, cultural and social dimensions, Fridman and Ollivier (2004) even speak of ‘an ostentatious openness to diversity’, showing that tolerance is a part of the character of those with a wide range of social, economic and cultural resources. Their approach has the benefit of indicating that omnivorousness involves more than just a range of tastes, but can also be seen as a ‘discriminating attitude’ (Warde et al., 2008). It re-affirms the link between omnivorousness and cultural capital and distinction (Bryson, 1996).
More generally, although it is tempting to see omnivorousness as evidence of the collapse of cultural hierarchies, empirical research has usually interpreted it as a new form of distinction typical of social elites (Coulangeon and Lemel, 2007). In any case, it tends to suggest that omnivores draw new social boundaries and acquire some distinctiveness by crossing previously established cultural boundaries. Yet Lahire offers another perspective on this supposedly growing eclecticism which he perceives to be less distinctive. If it entails some (cultural) boundary-crossing, dissonance takes place in a much messier fashion and both between and within individuals’ profiles.
Lahire’s Dissonant Spirit
Lahire (2006) proposes a sociological programme focused on intra-individual variations or ‘dissonances’ within cultural profiles. Bourdieusian sociology emerged, according to Lahire, from a belief in the superiority of high ‘legitimate’ culture, understood as the dominant cultural order, set against low and ‘illegitimate’ forms of culture. Lahire remarks that the overly Manichaean character of this vision rests on a questionable presupposition: that all individuals in society (consciously or otherwise) comply with this legitimate/high versus illegitimate/low classification. He suggests that with the emergence of concurrent spaces of socialisation (clubs, groups, media, etc.), the argument for an over-riding classification of cultural legitimacy is today falling into obsolescence.
For Lahire, when attention is turned towards intra-individual variations, we are able to see within each individual the ‘trace’ and ‘symptom’ of the plurality of cultural offers and of social groups (2006: 61) with which they have been confronted. At the individual level, very different cultural universes become co-habitants, rather than rivals. This results in a decreased belief in culture traditionally defined as legitimate, following the emergence of scientific, commercial and entertainment cultures (Lahire, 2006: 250) and a related ‘plurality of orders of cultural legitimacy’ (Lahire, 2008: 172). This plunges the individual into a far more complex socialisation matrix than has previously been the case. These tensions between different forms of culture which encompass their own sets of values and their own principles of legitimacy are therefore responsible for intra-personal variations. Bourdieu’s presumption of coherence in individual lifestyles is not tenable according to Lahire. Cultural dissonances hence describe individual sets of cultural practices and preferences that are heterogeneous in terms of their degree of legitimacy both within and across cultural domains. The recognition of a multitude of cultural orders does not, however, go along with the collapse of boundaries. Power relations operate between these different orders and they are not equal, whether that be at a societal or individual level. Not everything has equal value: ‘the symbolic war of everyone against everyone does not mean that everybody has equal means to do battle and impose their views’ (Lahire, 2006: 672). 1 The distinction between high and low culture, even if it is somewhat tarnished, still remains key in relationships of domination and in cultural hierarchies, even more so as schools have never stopped transmitting it (Daenekindt and Roose, 2015).
The choice of the word ‘dissonance’ comes from a desire to show that individuals do indeed maintain the different orders as separate, classing them, putting them in hierarchies, giving them a value. For Lahire, even if dissonance is the most frequent state statistically, it is not, however, to be considered as ‘normal’ (2006: 673). Dissonance is not experienced as a new configuration of people’s relationships to culture, but as the result of tensions between more and less socially valued cultural orders. In that sense, dissonance shares a (cultural) ‘boundary-crossing character’ with omnivorousness (Daenekindt and Roose, 2014: 83). However, these tensions do not only characterise relationships between groups but also internal divisions within individuals. It is as if individuals were reproducing – in their biographical constructions – the power struggles between the different cultural orders, which leads to internal conflicts. Lahire speaks of a ‘plurality of legitimate cultural orders’ (2006: 60), but eventually introduces a hierarchy inside the self which reflects the highly stratified society we live in.
Dissonance does not symbolise ‘a subjectively assumed hybridisation’ (Lahire, 2006: 673) or a socially valued eclecticism. Therefore, omnivorousness is not seen by Lahire as ‘one of the possible forms of dissonance’, as suggested by Daenekindt and Roose (2014: 83). On the contrary, Lahire quite clearly means to distance himself from Peterson’s perspective, to which he appears close at first sight. Lahire addresses a series of criticisms against Peterson’s notion of omnivorousness and eventually questions its very existence in contemporary social life. He tends to reject what he calls ‘the zoological metaphor’ (p. 255) in favour of intra-individual dissonances for a variety of methodological and theoretical reasons (see also Lahire, 2008). Some of these are largely well founded if we only take into account Peterson’s work, but they are clearly less so if we look at the entire body of work on omnivorousness. This is especially accurate for the two main methodological criticisms Lahire addresses to Peterson. First, Lahire’s critique of Peterson’s unique interest in musical preferences does not hold in light of research findings exploring other domains and practices (López-Sintas and García-Álvarez, 2002, 2004; van Rees et al. 1999; Vander Stichele and Laermans, 2006). Second, one major critique Lahire addresses to Peterson’s statistical work is that it focuses on the aggregate level. Upper social strata, instead of being omnivorous, could indeed fragment into a multitude of homogenous sub-clusters, but other statistical works have shown the occurrence of omnivorous patterns at the individual level (e.g. Vander Stichele and Laermans, 2006).
In my view, Lahire’s theoretical criticisms are more damaging. They are all articulated around the idea that Peterson over-interprets the implications of his empirical findings. Lahire argues that Peterson ‘yields to a certain ideological trend’ (Lahire, 2006: 257), which wholeheartedly endorses the model of cultural eclecticism as a ‘“new” aesthetic’ of consumption (Lahire, 2008: 182). First, the shift towards such a new model could not even be empirically proven by Peterson, according to Lahire. Second, the explanations provided by Peterson to account for the emergence of omnivorous patterns are ‘idealistic’ (Lahire, 2006: 258), especially when Peterson associates it with greater tolerance. That being said, Lahire’s vision of omnivorousness is somewhat too reductive. Peterson and Kern clearly stipulate that omnivorousness did not signify ‘that the omnivore likes everything indiscriminately’ (1996: 904). Unsurprisingly then, Lahire argues that omnivorousness is particularly flattering for elites, who appear to be cosmopolitan when compared with lower social strata, considered to be univores. This is somewhat unfair to Peterson, who wanted to see omnivorousness as a general tendency affecting all social strata, even if the upper ones more significantly. However, it is also true that empirical research has continued to provide support for the elitist character of eclecticism (Bryson, 1996; see also Coulangeon and Lemel, 2007). Despite this body of evidence, Lahire dismisses this elitist view of eclecticism by offering another interpretation of it: ‘by opening the field of investigation as I have done, we realise that every social class is involved, to one degree or another, in this intra-individual variation from one cultural register to another’ (2008: 183).
Therefore, for Lahire, the explanation lies elsewhere, in the intra-individual variation: ‘An important part of the variety of practices is explicable as much by the diversity of contexts, conditions and reasons in/for which consumers are driven to act as by personal eclecticism’ (2006: 257). He aims this way to reframe the debate towards the sociology of an individual as being at the crossroads of multiple contexts of socialisation. As a consequence, individuals do not invest themselves in everything in an undifferentiated manner. There are certain circumstances of life that push them to do certain activities rather than others (going out with children, with friends, etc.), which are largely ignored by Peterson’s quantitative analysis.
In summary, Lahire does not believe that omnivorousness is a new form of discriminant cultural orientation that would push privileged people to select cultural referents in the cultural repertoires of different social groups (characterised by different degrees of legitimacy) and mix them with their own in a very specific and distinctive way. If omnivorousness is anything at all for Lahire, it would be the symptom of widespread intra-individual dissonances due to a diversity of contexts of socialisation that Peterson has mistakenly over-interpreted as a new aesthetic paradigm. And this would explain why intra-individual variations are the most common feature within all social strata. Lahire is not far from seeing omnivorousness as a conceptual and methodological artifice that reflects not much more than incoherence in the cultural profiles. In the next sections, after having introduced the data, I will undertake an empirical test of this claim and will argue that Lahire is too quick to dismiss the validity of Peterson’s theory. Omnivorousness cannot be reduced to dissonance, but instead each notion characterises different configurations of cultural practices and preferences. However, Lahire’s approach is also essential for avoiding some of the pitfalls that the dominant focus on omnivores in cultural sociology has led to. It especially forces us to come back to a refined understanding of the tensions between existing and emerging cultural hierarchies at the individual and social levels, and of the ways in which these underlie people’s cultural profiles.
Measuring Legitimacy, Dissonance and Omnivorousness
This article relies on data from a general survey of the cultural practices and consumption of the Francophone population living in Wallonia and Brussels (‘Enquête générale sur les pratiques et la consommation culturelles’). This data was gathered in 2007 at the request of the Observatory of Cultural Policy of the Wallonia-Brussels Federation. It comprises a representative sample of the Francophone population aged from 16 years upwards living in that part of Belgium (N = 2021; response rate: 32%). The data is exceptional by the variety of fields examined. Amongst other things the questions related to a variety of outings, ‘flâneries’ (shopping, antique fairs, flea markets, etc.), outdoor activities (walks, picnics, etc.), entertaining activities (restaurants, cafes, dance clubs, karaoke, etc.), cultural outings (theatres, museums, art galleries, etc.), reading, music, internet, television and cinema tastes (see Guérin, 2009). One of the greatest strengths of this survey is its scope in terms of the number of cultural areas and the cultural genres within them investigated.
Legitimacy
Legitimacy is central to the measurement of both the concepts of dissonance and omnivorousness since, as I explained, they both entail crossing cultural boundaries between high/legitimate and low/illegitimate culture. It was calculated across seven fields: socio-cultural outings; at-home socio-cultural leisure activities; reading, music, and cinema tastes; magazine and televisual preferences. To measure the degree of legitimacy, I used Warde and Gayo-Cal’s method (2009), which evaluates, in general and for three different generations, the ratio between the percentages of people with a diploma of higher education and of those people without a diploma at all or with a diploma from primary education at the most. Based on these ratios, the authors distinguish legitimate, common and unauthorised tastes and practices. This is a refined method for calculating legitimacy, taking into account age and education, two essential factors in lifestyle modelling (van Eijck and Bargeman, 2004). This operationalisation implies linking legitimacy to the tastes of the highly educated. This definition of legitimacy is a purely empirical definition that serves solely for the purpose of analysis.
When it comes to the choice of generations, I refer to the work of Donnat (1994, 2009). He identifies four generations which have established different relations to culture. Briefly, the first generation is constituted of people who became adults before 1968. They have been less affected by changes as far as information technology is concerned, other than television. The second generation gathers together those individuals born in the 1950s and 1960s. These are the ‘baby boomers’, who experienced and participated in the development of the music and leisure industries as well as associated technologies. The third and fourth generation distinguish those younger than 30 and those older than 30 years old (but younger than 41). The latter are characterised by a less acute familiarity with the ‘new screen culture’ and by less extensive free time, related to the arrival of children. I have not differentiated the fourth from the third generation because it does little but accentuate the tendencies already present in the third.
As an illustration, consider Table 1 for book tastes (see the note below it for its interpretation). Based on preferences from the population of readers 2 (those who read at least one book during the 12 months prior to the survey), Table 1 ranks people’s tastes as legitimate, common and not legitimate. A common taste would mean that it does not have a dominant (‘high’ or ‘low’) distinctive feature across the different educational groups and generations. The tables indicating the legitimacy of the other practices and fields of taste are available in the online supplemental material (see the link at the end of the paper). The proportion of ‘legitimate’ styles of literature is relatively large (more than 50%). As an example, classic literature is principally overrepresented amongst the youngest highly educated generation and decreases in legitimacy through the two other generations. Its legitimacy is thus common. In comparison, historical fiction, bestsellers, books awarded literary prizes, biographies, political essays and historical books 3 are, amongst others, recognised as legitimate. Amongst the ‘not legitimate’ genres, we find science fiction novels, romantic novels, real-life stories, cookery books and DIY books.
Degree of legitimacy of book genres.
Note: Based on the responses of people who had read at least one book during the 12 months prior to the survey, in other words based on 66% of the sample.
Legitimacy calculation based on Warde and Gayo-Cal’s (2009) method. Column (a) shows the percentage of readers who most often read the literary style mentioned at the beginning of the line. Columns (b) to (e) evaluate, in general and for three different generations, the ratio between the percentages of people with a diploma of higher education and of those people with a diploma from primary education at the most. For instance, for classic literature, 20.7% of people with a diploma of higher education read most often classic literature compared to 16.3% of people with a diploma from primary education at the most. The ratio of 20.7 to 16.3 gives 1.3 in column (b) for all ages. Column (f) indicates to what extent this ratio might indicate a particular profile: legitimate (2 ⩽ L), common (C between 1 and 1.99), not legitimate (1 > N). When two out of three columns are identical, the profile is considered to belong to the dominant ratio. For instance, the taste for political and philosophical essays is legitimate across all generations (L/) and, although it is common for the oldest generation (above 65 years), it is legitimate across the two first generations (< 41 and 41–65 years). Therefore, the acronym ‘L/LLC’ is used and designates this genre as legitimate. When the three columns indicate a different value, the profile is considered to be shared. For instance, although the taste for adventure novels is common across all generations (C/), it is not legitimate for people aged 40 years or less, legitimate for people aged between 41 and 65 years, and common for people aged at least 66 years (C/NLC).
When there is no one in a cell of the crosstab between education level and reading genres, the value 1 has been arbitrarily attributed to the cell as it does not make sense to divide by 0 and that 0 divided by any value always gives 0. When this is the case, the ratios have been put in italics to distinguish them from the others.
Note that this category may not include those who only read comics. Graphic novels and comics were spontaneously mentioned by a significant proportion of those who had read at least one book and hence included as a category in the book preferences.
Some of these findings may be surprising. For instance, one might have expected that classic literature and poetry would be legitimate. This can be explained with reference to three reasons. First, the classification is based on the answers of those who actually perform the activity (here reading). Such an approach is useful both in identifying people who do not express tastes in different cultural areas and also, most importantly, in avoiding the problem that the degree of legitimacy of genres would be primarily defined by the differences between the least and the most educated in terms of their propensity to engage in this activity. Otherwise there is a risk that most reading genres would be legitimate, as the highly educated read more than the rest of the population, and that the analysis would somewhat artificially increase the number of legitimate genres and mask important fine-grained differences in the status allocated to reading preferences among readers. I have opted here for a more strict approach in the definition of cultural legitimacy. Second, the analysis has the advantage of evaluating legitimacy across three different generations, reflecting changes over time in the status of cultural genres. Age is indeed a powerful source of heterogeneity in cultural patterns (Lizardo and Skiles, 2015), and a genre may create cultural distinction in one generation and not in another, reflecting partly different contexts of socialisation. Third, some genres, such as poetry, are not appreciated by a group significantly large enough actually to be culturally distinctive.
In any case, cultural legitimacy is here tested empirically, instead of relying on existing knowledge or common presuppositions. This approach differs by its level of refinement from Peterson and Lahire’s operationalisations. Peterson and Kern (1996) defined musical genres (opera and classical music) as highbrow when those who liked them were regular attenders at opera, ballet, plays, and art galleries. Lowbrow and middlebrow genres were chosen based on existing literature. Lahire ‘took into account, on the one hand, the composition of their publics and, on the other hand, scholarly (through extensive existing research) and ordinary knowledge of cultural differences within the social world’ (2008: 168). By comparison, the analysis presented in this article took nothing for granted.
Dissonance
The measurement of dissonance is carried out in two stages. First, for each domain of tastes and practices, the population is divided into four categories: profile with a legitimate tendency (at least 50% of legitimate preferences and less than 50% of illegitimate preferences), profile with an illegitimate tendency (at least 50% of illegitimate preferences and less than 50% of legitimate preferences), mixed profiles and no preference expressed. Next, based on these categorisations realised for each domain of tastes and practices, the general degree of dissonance by individual was evaluated by taking the ratios of legitimate and illegitimate domains against the total number of domains of tastes expressed. When one of the two ratios is higher than 50%, meaning that more than half of the domains are either legitimate, or illegitimate, the individual can be considered as consonant. If this is not the case, I have assessed whether there is a tendency visible behind the dissonance. If between 41% and 50% of the domains have a degree of clear (il)legitimacy, and the remaining proportion is not associated with an opposite distinctive pattern, we can establish a legitimate or illegitimate tendency. In the other cases, the individuals were considered as pure dissonants. That being said, if people did not express their tastes for at least three cultural areas, they have been included in the category ‘without expression’.
Table 2 shows the breakdown of the population sample according to dissonance. I have also distinguished the most consonant with at least 85% of their tastes and practices being either legitimate or illegitimate. We can immediately see that these individuals are extremely rare. In the rest of this article this distinction will no longer be made. These results, like those of Lahire (2006), clearly show the predominance of dissonants, representing nearly 60% of the sample.
Distribution of individuals according to their dissonance and consonance.
Where values do not total exactly 100% this is due to the fact that the data are rounded and weighted.
Omnivorousness
To measure omnivorousness, I used the difference between ‘omnivorousness by volume’ and ‘omnivorousness by composition’ established by Warde and Gayo-Cal (2009), since it can help further describe patterns of cultural choice and of boundary-crossing, and how they relate to mechanisms of dissonance or omnivorousness. Omnivorousness by volume is calculated through the average number of tastes and practices without accounting for their level of legitimacy. Omnivorousness by composition is calculated here by the number of ‘legitimate’ and ‘non-legitimate’ tastes/practices.
The Characteristics of Dissonance and Omnivorousness
While there are certain parallels to be made between dissonance and omnivorousness in terms of their relations with socio-economic characteristics, notable differences exist between these two phenomena from this point of view as well. First, as Table 3 reveals, the degree of consonance or dissonance does not appear to be linearly linked to age (even if there is a significant relationship between the different profiles and the age categories). Dissonants are not necessarily the youngest nor the consonants the oldest (or vice versa). It is difficult to establish a linear relation between these two variables. In effect, two categories appear to be the oldest, the legitimate consonants and the dissonants with a legitimate tendency (whom I will call the ‘legitimate dissonants’ from now on). This suggests a positive link between age and legitimate culture rather than dissonance. Almost 70% of these two categories are at least 41 years old. The ‘pure’ dissonants and the dissonants with an illegitimate tendency (or the ‘illegitimate dissonants’) display slightly younger traits: around 50% of them are aged 40 years or less. In addition, while both legitimate consonants and legitimate dissonants present relatively similar age profiles, the illegitimate consonants tend to be different from the illegitimate dissonants. Illegitimate consonants have a contrasting age profile: 19% are younger than 30 years old and a similar proportion is older than 65 years. A look at Table 4 enables us further to understand their profile. Table 4 indeed reveals that those who have the greatest number of illegitimate tastes and practices are the youngest. This indicates that the illegitimate consonants include important proportions of younger and older persons who appreciate illegitimate culture but not to the same extent (the older being characterised by a narrower volume of illegitimate tastes and practices).
Age and dissonance.
Note: Cramer’s V = 0.093; p ⩽ 0.01.
Where values do not total exactly 100% this is due to the fact that the data are rounded and weighted.
Average number of tastes and practices by age (standard deviations in brackets).
Note: Anova based on a test (Brown–Forsythe) not presuming the homogeneity of variances, F* = 11.473 (df2 1444.313) (a), 9.303 (df2 1632.898) (b), 15.382 (df2 1459.350) (c), 51.079 (df2 1485.084) (d), 12.168 (df2 1511.947) (e); df1 = 4, p ⩽ 0.01. For column (f), the homogeneity of variances is respected: F = 60.211, df = 4, p ⩽ 0.01. One case was excluded from all the analysis of omnivorousness because the values were extreme.
In contrast, age is more clearly associated with the degree of omnivorousness (Table 4). The oldest (older than 65 years) are those who least appreciate a wide variety of tastes and practices whatever their degree of legitimacy (respectively on average 7.52 and 13.49). In comparison, the youngest – aged 21 years or less – have a large number of tastes and practices (respectively on average 10.14 and 22.66). However, they tend to be omnivorous by composition in terms of practices, but ‘only’ omnivorous by volume when it comes to their tastes: indeed the youngest like to do many different activities characterised by diverse degrees of legitimacy, but, if they appreciate a lot of cultural genres, these genres tend to be more homogeneous in terms of legitimacy (either common or illegitimate). Out of 10.14 genres appreciated on average, 4.33 are illegitimate and only 1.54 legitimate. These findings seem to be in line with Peterson’s claim about a reduction in the link between omnivorousness and legitimate culture amongst the younger generations (Peterson, 2005; Rossman and Peterson, 2005).
At this point let us move on and assess the role of the level of education. It can be expected that our two concepts would be associated with education, given that the measure of legitimacy underlying them is based on it. Nevertheless, the question is less about the strength of the link than a comparison of this link in the cases of dissonance and of omnivorousness. To put it in another way, I seek to know whether the most dissonant and the most omnivorous are defined by similar levels of education.
While legitimate consonants include the greatest proportion of university graduates, and illegitimate consonants the greatest proportion of people with at most a diploma of primary education (see Table 5), those who like and practice the highest number of illegitimate activities/tastes do not necessarily have the lowest level of education (see Table 6). People with a higher education degree tend to be more omnivorous by volume and by composition (especially those with a non-university higher education degree), both in terms of their activities and tastes. 4 When one pays attention to educational levels, dissonance does not match omnivorousness. As I have just explained, the most consonant are both those with the lowest and the highest levels of education, yet the least educated are the least omnivorous and the most educated the most omnivorous. People who have completed secondary education are overrepresented among the dissonants; yet they are not the most omnivorous by volume, and most importantly do not have the most contrasting patterns of cultural choice in terms of the legitimacy of their tastes and practices.
Educational level and dissonance.
Note: Cramer’s V = 0.232; p ⩽ 0.01.
Where values do not total exactly 100% this is due to the fact that the data are rounded and weighted.
Average number of tastes and practices by educational level (standard deviations in brackets).
Note: Anova based on a test (Brown–Forsythe) not presuming the homogeneity of variances, F* = 50.974 (df2 1319.193) (a), 114.587 (df2 953.755) (b), 110.886 (df2 1442.003) (e), 12.786 (df2 1672.479) (f); df1 = 4, p ⩽ 0.01. For columns (c) and (d) the homogeneity of variances is respected: F = 12.623 (c), 93.400 (d); df = 4, p ⩽ 0.01.
If one pays attention to socio-professional categories, apart from pensioners and students for whom their profile is, in part, related to their age, one must note the very large proportion of professionals and higher executives amongst the legitimate consonants (more than 10%, while they do not represent more than 3% in the other categories of dissonance). These professionals and higher executives are most strongly associated with legitimate culture and proportionally enjoy less illegitimate genres and do fewer illegitimate activities. Nevertheless, they are characterised by omnivorousness by volume. Their omnivorousness by composition is principally due to their appreciation of common genres. In this sense, there is no collapsing of cultural hierarchies: attachment to legitimate culture makes people rather selective when it comes to indulging themselves in other, less legitimate cultural activities and genres (in line with Bryson’s (1996) ‘patterned tolerance’).
On the other hand, non-qualified workers are overrepresented amongst both illegitimate consonants and illegitimate dissonants. They like a narrower selection of genres and have a less diverse set of activities than the other groups, this being mainly the result of a distanced relationship with legitimate culture. No socio-professional category is particularly overrepresented within the pure dissonants, apart from, and not in a very pronounced fashion, students. If one looks at the proportion of dissonance within each socio-professional category, it is clear that dissonance is the most frequent scenario in each group. A similar assessment allowed Lahire to conclude that dissonance was a ‘trans-class condition’ (2006: 205), due to its strong representation amongst the population in general. Therefore dissonance does not appear particularly distinctive, while – and my results confirm this – omnivorousness can be thought of as producing forms of distinction.
The Link Between Omnivorousness and Dissonance
As discussed earlier, Lahire does not believe that omnivorousness translates into new distinctive cultural patterns but rather into widespread intra-individual dissonances. However, the previous section has already demonstrated that dissonance says nothing about the breadth of tastes and practices, and hence could hardly be related to omnivorousness. Dissonance simply refers to the composition of a given cultural profile across different domains of taste (reading, music, etc.), and to its coherence in terms of cultural legitimacy.
To explore further their differences, Table 7 shows the association between dissonance and omnivorousness. I investigate here whether dissonants are really the most omnivorous, but also whether they appreciate high culture, given that the literature on the topic – as suggested above – has constantly shown a link between omnivorousness and high culture (even if the youngest cohorts seem to move away from some traditional forms of high culture – DiMaggio and Mukhtar, 2004; Peterson, 2005).
Average number of tastes and practices according to the degree of dissonance (standard deviations in brackets).
Note: Anova based on a test (Brown–Forsythe) not presuming the homogeneity of variances, F* = 38.430 (df2 667.692) (a), 322.381 (df2 451.023) (b), 110.707 (df2 870.119) (c), 160.549 (df2 898.044) (e); df1 = 4, p ⩽ 0.01. For columns (d) and (f), the homogeneity of variances is respected: F = 59.660 (d), 22.236 (f); df = 4, p ⩽ 0.01.
The first result is that pure dissonants are not the most omnivorous, whether that be in terms of tastes or of practices. Second, the most omnivorous in terms of volume are 1) legitimate dissonants and 2) legitimate consonants. It is not therefore adequate purely and simply to assimilate dissonance with omnivorousness. They do not reflect similar patterns of tastes and activities: even though they may be conceptually close, they do not refer to the same processes. In addition, despite the fact that young people have fewer legitimate tastes (this is not true for activities), legitimate culture and omnivorousness are clearly associated. While omnivorousness can thus be considered to some extent elitist, given its links with legitimate culture and upper social strata, the same cannot necessarily be said for dissonance.
As a consequence, individual openness, which is arguably translated into omnivorousness, appears to be related to profiles with a breadth of legitimate tastes and practices and upper socio-professional categories. People with high levels of educational and economic capital, among whom one could find Lahire’s ‘elites’, are associated with omnivorousness by volume and, to a certain extent, omnivorousness by composition (mainly due to their appreciation of common cultural genres and activities), but not with pure dissonance. The link between consonance and univorousness again tends to indicate that they cannot simply be assimilated with each other: the legitimate consonants who could be seen as ‘snobs’ are not necessarily more univorous than others. On the contrary, univores appear to be found amongst the illegitimate consonants who do not only reject legitimate tastes but also many of the tastes with a weaker legitimacy. Looking at the results, this univorousness seems related to lower social positions, but also appears to be the result of a generational effect and thus of socialisation: it is the oldest who like the smallest number of different tastes and who have the smallest number of different practices. Conversely, intermediate age categories, whose members were doubtless socialised at times when different values were successively being advocated, appear to be closer to omnivorousness by composition.
Conclusion: Beyond Omnivorousness and Dissonance
This article shows the clear parallels that exist between the two notions of dissonance and omnivorousness, but also their divergences. Dissonance and omnivorousness are related to each other but clearly correspond to different cultural phenomena. Systematic cross-tabulations with socio-demographic variables suggest that the emergence of dissonant and omnivorous individual patterns of cultural choice is most likely due to different relationships with existing cultural hierarchies, and therefore to different mechanisms of formation. Omnivorousness is linked to upper social strata which have at their disposal greater cultural and economic resources for navigating between different cultural genres, as Fridman and Ollivier (2004) have made clear. Similarly Donnat (2004) also explains how his ‘Branchés’, the most educated and eclectic in the French population, have accumulated cultural knowledge that enables them to be at ease with different cultural registers. Dissonance does not imply the possession of such resources; it implies incoherence in the cultural profile. It is important to underline this phenomenon and to study it, but it does not offer information about the variety or the range of tastes and practices and, more importantly, about how they relate to each other.
More than that, omnivorous or dissonant characteristics are not sufficient to define the cultural profiles of individuals, given the persistent role of legitimate culture in these profiles (whether it be negative or positive). Therefore, while there is a great temptation to interpret omnivorousness as simply a sign of dissonance, my results show that pure dissonants are not the most omnivorous and that consonants are not necessarily the most univorous. The most omnivorous are, first of all, legitimate dissonants and, second, legitimate consonants, demonstrating a clear relation between legitimate culture and breadth of taste. The position of these legitimate consonants, who correspond to Bourdieu’s snobs (liking primarily legitimate genres), suggests that cultural distinction is still central in the constitution of cultural profiles, as evidenced by its role in the development of an omnivorous set of tastes and practices.
Concentrating on individuals, Lahire offers a theoretical framework for the idea of eclecticism which stresses the importance of taking account of the different contexts within which individuals are socialised. For him, in contrast to Peterson, individual omnivorousness would not reflect a greater tolerance which would have come into being following a series of structural and cultural changes. Lahire’s approach is essential to the extent that dissonance offers a way of thinking about individual cultural profiles as the battleground of conflicts between cultural orders with different legitimacy. Nevertheless, doubtless too focused on the individual, Lahire stops there and does not help us increase our understanding of the link between these cultural orders, which he actually does not define except to say that they classify cultural objects and activities according to certain values (entertaining, serious, etc.). In a certain sense, the concept of dissonance seems to suggest that different behaviours and practices should not be seen as forming structured ensembles or profiles. It is on this point that Lahire’s theory appears the most vulnerable.
Bellavance (2008) helps us better understand the possible articulation between these ‘orders’. He shows how different hierarchical principles can coexist and be juxtaposed, without abolishing symbolic barriers between artistic and cultural genres. He proposes different principles, such as the distinction between the old (traditional, established, conformist) and the new (modern, rebellious, avant-garde). Above all, according to Bellavance, the high/low distinction which defines a dominant cultural order for Lahire may be statistically less significant, following the rise of eclecticism, but has remained essential in the understanding of the organisation of tastes and practices in cultural profiles. Therefore, if cultural profiles can be perceived as dissonant bricolage between different cultural orders, it is also important to understand how these orders relate to each other and, to put it differently, how people come to like some serious cultural genres but also more entertaining lowbrow ones.
People draw on these different orders (high/low, old/new, etc.) to construct their cultural profiles (Hanquinet, 2013), but in a meaningful way that neither the ideas of a greater tolerance or individual dissonance seem fully to grasp. Therefore, instead of spending time in identifying and labelling people as dissonant or omnivores, it now seems to be more interesting to seek to describe people’s patterns of cultural choice and to assess how these relate to still central cultural hierarchies articulated around values such as low/high, entertaining/serious, emerging/established, and so on. Both dissonance and omnivorousness are simply the outcome, or the symptom if one prefers, of this mixing of genres and, hence, values, which is far from being random but rather is highly structured and dependent on the types of resources people can rely on. The tradition in (mainly quantitative) cultural sociology of identifying omnivorous cultural patterns has been misleading, first because it has drawn attention away from describing people’s relationships to existing but also emerging cultural hierarchies, and second, because omnivorousness is not necessarily a new form of distinction as has been argued previously, but instead emerges from the relations and tensions between different cultural orders of distinction. Lahire’s conception of dissonance has the merit of making us reconsider the importance of this dominant conceptualisation of omnivorousness, but suffers from a similar problem of vagueness, even though he gives us many ways of further grasping how tastes and practices are organised at the individual and social levels.
Footnotes
Acknowledgements
I would like to thank Louise Callier and Jean-Louis Genard from the Université Libre de Bruxelles for their help on this project. I am also thankful to the Observatoire des Politiques Culturelles de la Fédération Wallonie-Bruxelles (Belgium), and especially Michel Guérin, for having funded and initiated the survey on cultural participation in the Wallonia-Brussels Federation. I am grateful to Michaela Benson, Céline Teney, Paul Chappell and Alan Warde for their comments on early versions of the article. The first version of this text was translated by Tim Neal.
Funding
This research was funded by the Observatoire des Politiques Culturelles de la Fédération Wallonie-Bruxelles.
