Class Position and Musical Tastes: A Sing-Off between the Cultural Omnivorism and Bourdieusian Homology Frameworks GERRY VEENSTRA University of British Columbia
The longstanding debate between the homology and omnivorism approaches to the class bases of cultural tastes and practices rages on in cultural sociology. The homology thesis claims that class positions throughout the class hierarchy are accompanied by specified cultural tastes and specialized modes of appreciating them while the cultural omnivorism thesis contends that elites are (increasingly) characterized by a breadth of cultural tastes of any and all kinds. This study tests the applicability of these theses to musical tastes in Canada through the application of multiple correspondence analysis, latent class analysis, and logistic regression modeling to original telephone survey data (n = 1,595) from Toronto and Vancouver. I find that musical omnivorism, an appreciation for diverse musical styles, is not dispersed along class lines. Instead I find a homology between class position and musical tastes that designates blues, choral, classical, jazz, musical theater, opera, pop, reggae, rock, and world/international as relatively highbrow and country, disco, easy listening, golden oldies, heavy metal, and rap as relatively lowbrow. Of the highbrow tastes, all but jazz are disliked by lower class people, and of the lowbrow tastes, country, easy listening, and golden oldies are concurrently disliked by higher class people. Consistent with the homology thesis, it appears that class position is aligned with specific musical likes and dislikes. Le vieux d´ebat entre les approches de l’homologie et de l’omnivorisme ˆ et des pratiques culturels fait rage dans aux bases des classes des gouts la sociologie culturelle. La th`ese de l’homologie pr´etend que les positions des classes a` travers la hi´erarchie des classes sont accompagn´ees par des ˆ culturels sp´ecifi´es et des modes sp´ecialis´es permettant leur gouts appr´eciation. La th`ese de l’omnivorisme culturel, en revanche, soutient ˆ que les e´ lites sont (de plus en plus) caract´eris´ees par un e´ ventail de gouts Gerry Veenstra, Department of Sociology, University of British Columbia, 6303 N. W. Marine Drive, Vancouver, BC, Canada V6T 1Z1. E-mail: [email protected]
C 2015 Canadian Sociological Association/La Soci´ et´e canadienne de sociologie
Class Position and Musical Tastes
135
culturels de toutes sortes. Cette e´ tude exp´erimente l’applicabilit´e de ces ˆ musicaux au Canada a` travers l’application de l’analyse th`eses aux gouts des correspondances multiples, de l’analyse des classes latentes et du mod`ele de r´egression logistique aux donn´ees d’origine de l’´etude collect´ees par t´el´ephone (n = 1,595) de Toronto et de Vancouver. Je me rends compte que l’omnivorisme musical, une appr´eciation de styles musicaux divers, n’est pas dispers´e le long des classes sociales. Bien au contraire, je trouve qu’il existe une homologie entre la position des ˆ musicaux qui d´esigne les blues, la chorale, la musique classes et les gouts ˆ classique, le jazz, le th´eatre musical, l’op´era, la pop, le reggae, le rock et la musique du monde/internationale comme des styles des classes au niveau intellectuel relativement e´ lev´e. Cette homologie d´esigne la country, la disco, la musique d’ambiance, les anciens succ`es, le heavy metal et le rap comme des styles des classes au niveau intellectuel ˆ des classes au niveau intellectuel e´ lev´e, tous relativement bas. Des gouts les styles a` l’exception du jazz ne sont pas appr´eci´es de la basse classe. ˆ des classes au niveau intellectuel bas, le country, la De mˆeme, des gouts musique d’ambiance et les anciens succ`es ne sont pas appr´eci´es de la haute classe. Selon la th`ese de l’homologie, il apparaˆıt que la position des classes est fonction des pr´ef´erences et des aversions musicales sp´ecifiques.
THE HOMOLOGY AND CULTURAL omnivorism frameworks have been at loggerheads for more than 20 years (e.g., Atkinson 2011; Bennett et al. 2009; Coulangeon and Lemel 2007; Gebesmair 1998; Goldberg 2011; Lizardo and Skiles 2012; Peterson and Kern 1996; Peterson and Simkus 1992; Rimmer 2012; Warde, Wright, and Gayo-Cal 2007). The homology thesis claims that class positions throughout the class hierarchy are accompanied by specified cultural tastes and specialized modes of appreciating them while the cultural omnivorism thesis contends that elites are (increasingly) characterized by a breadth of cultural tastes of any and all kinds. This study provides new evidence on the relationship between class position and cultural tastes by analyzing data from a novel telephone survey on music and class conducted in Toronto and Vancouver, Canada. Specifically, it seeks to establish whether elites and lower class people in urban English-speaking Canada are characterized by distinct sets of musical tastes or distinguished from another primarily by the degree to which they manifest omnivorous musical tastes.
THE BOURDIEUSIAN HOMOLOGY FRAMEWORK The homology framework, belonging to a long line of cultural inquiry that includes Thorstein Veblen’s The Theory of the Leisure Class ([1899] 1994) and Herbert Gans’ Popular Culture and High Culture (1974), finds its fullest expression in Pierre Bourdieu’s magnum opus, Distinction: A Social Critique of the Judgement of Taste (Bourdieu [1979] 1984). There the French sociologist maintained that a multitude of diverse cultural tastes
136
CRS/RCS, 52.2 2015
and practices in 1960s France were fundamentally manifestations of class habitus. For Bourdieu, class positions are distinguished by their differing amounts of economic capital, essentially monetary wealth, and cultural capital, valued cultural resources that include educational credentials, which locate them in different parts of a multidimensional social space (a society-wide “field”). Specifically, the sum total of the two forms of capital positions agents on the primary axis of social space, that which distinguishes the upper, middle, and lower classes, while relative composition of the capitals situates agents along the secondary axis of social space which distinguishes dominating and dominated sections of the classes. The cultural tastes of the wealthy and highly educated members of the upper class, comprising highbrow culture, represent the “legitimate,” “sophisticated,” and “cultured” tastes and practices of society. Members of this class have the power to delimit highbrow tastes and appropriate modes of appreciating them and can use their familiarity and facility with these cultural forms to maintain and reinforce boundaries between themselves and others. Many highbrow tastes are especially enshrined in the better educated but less-wealthy segment of the upper class, the home of the intelligentsia, the dominated portion of upper-class space. Lowbrow or “popular” tastes and practices, embraced by the less-wealthy and lesseducated members of society, are the antithesis of highbrow culture; they serve as a negative reference for the tastes of the dominant class (Swartz 1997). Finally, middlebrow culture reflects the (imperfect) attempts of the members of the middle class, the petit bourgeoisie, to embrace highbrow culture and to distinguish themselves from the lower class. For Bourdieu, then, there is a homology, a kind of isomorphic relation or one-to-one correspondence, between the multidimensional space of positions and the multidimensional space of cultural tastes. A key principle underlying the homology of class positions and cultural tastes, according to Bourdieu, is the opposition between the “tastes of luxury (or freedom) and the tastes of necessity” (Bourdieu [1979] 1984:198). The tastes of luxury are the tastes of people born into a habitus which is defined by distance from necessity; these people possess freedoms of thought and action that are facilitated by possession of capital. In particular, freedom from necessity facilitates development of the “aesthetic gaze,” a mode of consumption which stresses appreciation for the form of a cultural object rather than its function and tends to pass aesthetic rather than ethical judgments on it. Accordingly, order, etiquette, and restraint are virtues in upper-class space. The denial of lower, coarse, vulgar, venal, servile—in a word, natural— enjoyment, which constitutes the sacred sphere of culture, implies an affirmation of the superiority of those who can be satisfied with the sublimated, refined, disinterested, gratuitous, distinguished pleasures forever closed to the profane. (Bourdieu [1979] 1984:7)
Class Position and Musical Tastes
137
The middle class, striving for distinction but lacking the capital and habitus needed to fully appropriate upper-class lifestyles, in turn valorizes “asceticism, rigor, legalism” (Bourdieu [1979] 1984:331). In contrast with the bourgeois habitus which conveys an “ethos of ease, a confident relation to the world and the self,” the petit bourgeois habitus presents an “ethos of restriction through pretension” (Bourdieu [1979] 1984:339). The tastes of necessity, finally, most common in the lower class, derive from the need to produce labor at the lowest cost; people with low levels of (especially economic) capital tend to valorize tastes of necessity out of material necessity. Enjoyment and fun, impropriety, and fulfilling material needs are virtues in lower class space because the functions rather than the forms of cultural objects are of central importance here. Bourdieu asserted that “nothing more clearly affirms one’s ‘class’, nothing more infallibly classifies, than tastes in music” (Bourdieu [1979] 1984:18). Upper-class people, free from the strictures of necessity, tend to have a taste for difficult, abstract music, especially difficult classical music. For example, appreciation for Bach’s The Well-Tempered Clavier was prevalent among the most highly educated members of Bourdieu’s survey sample from the 1960s. “As legitimate culture, classical music gathers around it the highest values of aesthetic formalism associated with Kantian ‘distinterestedness’” (Prior 2013:183). The French upper class also rejected with disgust “the most popular and most ‘vulgar’ singers, such as Les Compagnons de la Chanson, Mireille Mathieu, Adamo or Sheila” (Bourdieu [1979] 1984:60). Middle-class space in turn tends to foster appreciation for popularized forms of legitimate music (Gershwin’s Rhapsody in Blue was especially popular among engineers and technicians in Bourdieu’s data) whereas the members of the lower class tend to consume music with simple, repetitive structures (such as the popular waltzes of Strauss) which are typically imposed on them by experts and artists in the field of musical production (Atkinson 2011). From the standpoint of Bourdieu’s homology framework, then, musical tastes and distastes, seemingly personally idiosyncratic in nature, are in fact reflections of class-based habitus. This represents what I think of as the strong form of the homology thesis. It describes a multidimensional space of positions prescribed by a specified complex configuration of capitals, namely, the sum total of economic capital and cultural capital on the primary axis and the relative composition of these forms of capital on the secondary axis. It is a relational visioning of entwined spaces of positions and cultural tastes that requires relational statistical techniques, such as multiple correspondence analysis (MCA) or latent class analysis (LCA), to illuminate the character of the spaces in quantitative data. In the weak form of the homology thesis, the space of positions and the spaces to which it is homologous are not configured in as complicated a manner. For instance, while evidence suggests that the sum total of economic capital and cultural capital
138
CRS/RCS, 52.2 2015
structures the primary dimension of the social spaces of contemporary France (Coulangeon and Lemel 2007), Canada (Veenstra 2010), Finland (Kahma and Toikka 2012), and the United Kingdom (Gayo-Cal et al. 2006; Le Roux et al. 2008) as well as France of the 1960s (Bourdieu [1979] 1984), the relative composition of these capitals does not appear to contribute to structuring the social spaces of contemporary France (Coulangeon and Lemel 2007), Finland (Kahma and Toikka 2012), and the United Kingdom (Gayo-Cal et al. 2006; Le Roux et al. 2008). If one capital or the sum of several capitals effectively delineates class positions then the weak form of the homology thesis is evident and nonrelational techniques such as regression analysis are presumably up to the task of uncovering a homology between positions and tastes. Precisely which capitals delineate class positions and the manner in which they do so in a given context is an empirical question.
CULTURAL OMNIVORISM Most incarnations of the cultural omnivorism perspective in the sociology of culture similarly identify social class as a fundamental basis of cultural tastes and practices. They depart from the homology framework, however, by claiming that breadth of cultural tastes has supplanted possession of specified highbrow/lowbrow tastes as the notable cultural delimiter of class boundaries. It is argued that, while the Bourdieusian homology storyline may have held in past eras, elites in contemporary societies are now better characterized as cultural omnivores than as highbrow aficionados. These omnivores consume and practice culture speaks of many different kinds, from hip hop to heavy metal to classical music, from grungy, muddy team sports such as rugby to ascetic, individual pursuits of body and mind such as yoga and tai chi, and so forth (Veenstra 2010). They seek variety and eclecticism and are intrinsically inclusive and tolerant in their tastes, in contrast with univores, who, displaying taste for a narrow range of activities or objects, are intrinsically exclusivist and intolerant. Members of the upper class are therefore distinguished from members of other classes by the breadth of their cultural repertoires rather than their possession of any specified highbrow cultural tastes or practices. Members of the lower class are in turn distinguished from others by the relatively narrow focus of their cultural repertoires, by their propensities to be univores, not by their allegiances to any specified lowbrow cultural tastes or practices. It is the desire and ability to consume a multitude of diverse cultural forms, including diverse musical styles, that now distinguishes higher class people from lower class people in contemporary societies. If the “highbrow snob” has been replaced by the cultural omnivore and the “lowbrow slob” has been replaced by the univore then the principles, processes, and mechanisms underlying the stratification of cultural tastes have likely changed as well. In particular, the distinction between tastes
Class Position and Musical Tastes
139
of freedom and necessity may no longer be preeminent. Erickson (1996) proposes that “cultural inequality is not so much a hierarchy of tastes (from soap opera to classical opera) as it is a hierarchy of knowledge (from those who know little about soap opera or opera to those who can take part in conversation about both)” (p. 219). Emmison (2003) argues that “cultural mobility entails the display of cultural competence in a plurality of domains with concomitant social rewards accruing to those demonstrating these capacities” (p. 213). Omnivores can move easily among cultural realms (Emmison 2003) and in business might use whatever form of cultural knowledge is necessary to make a good impression in job interviews (Garnett, Guppy, and Veenstra 2008) or build social networks to get a better job (Erickson 1996). In short, that which is now the most useful for getting ahead, namely, a breadth of cultural knowledge and familiarity with multiple cultural forms, is what is now classed. This is presumably because breadth of knowledge/familiarity is more readily achieved by upper-class people but also because people who have it are relatively likely to achieve upper-class standing. Various explanations for the historical change in the fundamental nature of the class bases of culture from homology to omnivorism have been proposed. For example, the increasingly specialized nature of occupations and greater mobility between occupations may have affected people’s cultural repertoires, since occupations tend to foster their own cultures and people who are required to communicate across occupations or move into a new occupation need to be conversant with a wide range of cultural forms (Erickson 2008). Growing income inequality may have contributed to a growing inequality in the ability of people to participate in culture while educational inflation may have produced relatively more highly educated people in upper-end occupations who are conversant with a wider range of cultural forms garnered through their experiences with educational systems (Erickson 2008). Increasing amounts of social mobility, especially from lower to higher strata, may contribute to increasing sociocultural heterogeneity (van Eijck 2001). Peterson (2005) suggests that snobbish exclusion “was an effective marker of status in a relatively homogenous and circumscribed WASP class [but that] omnivorous inclusion seems better adapted to an increasingly global world culture managed by those who make their way, in part, by showing respect for the cultural expressions of others” (p. 273). Increasingly globalized media industries, social and geographic mobility, and varied networks may also be at the root of the open and varied “cosmopolitan habitus” seemingly prevalent among elites (Atkinson 2011). There are several useful ways of distinguishing strong and weak forms of the omnivorism thesis. As with the homology thesis, relational (strong) and nonrelational (weak) renderings of the omnivorism thesis can be envisioned. For instance, one can imagine a social space delimited by sum total and relative composition of capitals where the upper-class section is
140
CRS/RCS, 52.2 2015
bursting with diverse musical tastes and the lower class section is bereft of them, a multidimensional space of positions homologous to a top-heavy space of (omnivorous) tastes. Alternatively, breadth of tastes might simply be associated with one or both of economic capital and cultural capital in linear-causal regression models. The distinction between “omnivorism by volume” and “omnivorism by composition” (Warde et al. 2007) can also designate strong and weak forms of the omnivorism thesis. In regard to the former, omnivorism is defined by indiscriminate liking of any and all cultural forms, a strong statement of the omnivorism thesis. In regard to the latter, however, “some distinctive status orientation is entailed in the patterns of cultural preferences involved” (Warde et al. 2007:145). For example, the definition of musical omnivore employed by Peterson and Simkus (1992) required that they like traditionally highbrow classical musical and opera. Achterberg and Houtman (2005) operationalized musical omnivorousness as the status distance between the most highbrow music and the most lowbrow music chosen by a person. Explicitly incorporating dislikes into her operationalization, Bryson (1996) depicted musical omnivores as people who consumed many different musical genres while simultaneously rejecting the music genres with the least educated fans, these being country, heavy metal, and rap music. In these formulations, musical omnivorism is prescribed in the breadth of its manifestation. Finally, strong and weak forms of the omnivorism thesis can be distinguished by the degree to which breadth of taste has supplanted specified highbrow tastes among elites. Some scholars suggest that highbrow tastes are a thing of the past and the omnivore now reigns supreme in a “post-Bourdieu era” of cultural sociology (Vander Stichele and Laermans 2006). Others contend that the omnivorism framework represents an evolution rather than a repudiation of the homology framework where the former has not (yet) overthrown the latter (e.g., Emmison 2003; Garc´ıa´ Alvarez, Katz-Gerro, and Lop´ez-Sintas 2007; Lop´ez-Sintas and Zerva 2005; Peterson 2005; Peterson and Kern 1996). They suggest that elites can in fact be split into two groups: exclusive highbrows with a univorous and exclusive highbrow taste; and inclusive highbrows with an omnivorous taste inclusive of middlebrow and lowbrow tastes (Emmison 2003; ´ Garc´ıa-Alvarez et al. 2007; Peterson 2005).1 Lowbrow taste, a popular taste that nearly everyone likes, is no longer associated with any specific class ´ (Garc´ıa-Alvarez et al. 2007). In short, the strong form of the omnivorism thesis does away with highbrow snobs entirely while the weak form accommodates the existence of highbrow snobs and omnivorous elites, with
1.
Research in this vein suggests that exclusive highbrows tend to be older and inclusive highbrows tend to be younger (Ollivier 2008). This age- or cohort-related factor speaks to the notion that aggregate changes in elite tastes from highbrow to omnivore are generationally prescribed, with younger omnivore elites replacing aging highbrow exclusivist elites.
Class Position and Musical Tastes
141
their relative weights in numbers and influence a matter of contextual and historical specificity. Multiple formulations of cultural omnivorism and various underlying principles and reasons for historical changes in the class bases of cultural tastes notwithstanding, the body of empirical research purportedly verifying the existence of class-delimited cultural omnivores in various international contexts is now voluminous (Peterson 2005). Cultural omnivorism has become de rigueur in some circles. Other scholars uphold the contemporary viability of the homology perspective, questioning the evidence for and utility in the notion of a class-delimited musical omnivore (e.g., Atkinson 2011; Rimmer 2012; Savage 2006; Warde et al. 2007). This study contributes to the ongoing tussle between the Bourdieusian homology and cultural omnivorism perspectives by investigating relationships between musical likes and dislikes and indicators of class position in original survey data from two large Canadian cities (Toronto and Vancouver).
ANALYTICAL PLAN The analysis proceeds in four stages. First, I investigate whether diverse musical tastes are widely dispersed, as the homology approach might predict, or tightly clustered, as might be consistent with musical omnivorism. To accomplish this, I apply MCA and LCA to a set of 21 musical likes and dislikes, illuminating contours of a musical field that distinguishes between people who like most musical genres, people who dislike most musical genres, and people who are ambivalent about most musical genres. Second, I investigate the degree to which and how this patterning of musical tastes is structured by various markers of inequality, namely, city of residence, age, gender, racial identity, immigrant status, economic capital, and cultural capital. In particular, I seek to determine whether this Canadian musical field is structured by the sum total of economic capital and cultural capital and the relative composition of economic and cultural capitals as the field of cultural tastes was structured in France of the 1960s. I find that the patterning of musical tastes previously identified is partly reflective of age, racial identity, and immigrant status but mostly unreflective of possession of economic capital and cultural capital. Third, I determine whether a measure of breadth of musical likes is predicted by one or more of the markers of inequality in an Ordinary Least Squares (OLS) regression model. I find that women, middle-aged people, and native-born Canadians are relatively likely to be musical omnivores but wealthy and/or highly educated elites are not. Finally, I investigate whether individual musical likes and dislikes are associated with the markers of inequality in binary logistic regression models. Consistent with the weak form of Bourdieu’s homology thesis, I find that elites
142
CRS/RCS, 52.2 2015
and lower class people are linked with distinct sets of musical likes and dislikes.
METHODS Survey Sample Between January and June of 2009, the Survey Research Centre (SRC) at the University of Victoria in British Columbia conducted telephone interviews with 732 adults living in the city of Toronto and 863 adults living in the Vancouver Census Metropolitan Area. The SRC used random-digit dialing techniques to obtain residential telephone numbers, a next-birthday strategy to select one resident per household aged 19 or older to interview, and a computer-aided telephone interviewing system to conduct the interviews. The introductory script from the callers informed prospective interviewees that all adults aged 19 and older were eligible for this study, that there were potential risks of an emotional kind from questions pertaining to racial and ethnic identity, experiences of discrimination in everyday life and sexual orientation, and that the ultimate goal of the study was to help to improve the health of Canadians. No incentives, monetary or otherwise, were provided for participation. In total, callers spoke with 17,060 people and successfully recruited 1,595 interviewees, representing a cooperation rate of 9.3 percent. Demographic characteristics of the survey samples and the cities from which they were obtained are described in Table 1. Comparison of the city samples to the 2006 Census by gender, age, racial identity, immigrant status, and educational attainment indicates that both city samples are biased toward women, older people, Whites, native-born Canadians, and people with a university degree. Comparison of cases with complete data (n = 1,503) with those with missing data (n = 92) indicates that the missing cases are relatively likely to be non-White immigrants. Musical Tastes Unlike most research of its kind (but see Bennett et al. 2009; Bryson 1996, 1997; Carrabine and Longhurst 1999; Savage 2006; Sonnett 2004; Tampubolon 2008b), this study focuses on musical likes and dislikes. The cultural omnivore is identified by her liking of diverse cultural forms. From the homology perspective, however, the tastes of elites and lower class people are as or more likely to be manifested as distastes than as tastes (Bourdieu [1979] 1984). My respondents were asked about their likes and dislikes in regard to 21 musical genres. “For each of the following types of music, please tell me whether you like or dislike or perhaps feel neutrally about each type: classical music, hip hop, choral music, folk music, rap, opera, country music, pop, jazz, easy listening, reggae, rock, heavy
Class Position and Musical Tastes
143 Table 1
Characteristics of the City Survey Samples and Their Populations Toronto
Variable
Categories
Gender
Male Female Aged 19–34 Aged 35–44 Aged 45–54 Aged 55–64 Aged 65 and older Asian Black South Asian White Other Born in Canada Immigrated to Canada Less than high school High school Community college or technical school diploma Bachelor’s degree or higher Less than $40,000 $40,000–59,999 $60,000–79,999 $80,000–99,999 $100,000–149,999 $150,000 or more Missing
Age
Racial identity
Immigrant status Educational attainment
Household income
Survey sample N = 732 %
Vancouver
Census 2006 %
Survey sample N = 863 %
Census 2006 %
35.4 64.6 20.4 18.5 23.6 19.6 17.8 3.0 6.4 4.1 79.3 7.1 65.3 34.7
47.3 52.7 29.7 21.0 18.3 13.0 17.9 13.3 8.4 12.1 53.0 13.2 47.8 52.2
32.8 67.2 12.5 16.4 23.7 26.9 20.5 6.1 0.2 5.1 82.7 5.9 72.3 27.7
48.2 51.8 28.6 21.2 20.3 14.1 15.8 23.5 1.0 10.9 54.7 9.9 56.5 43.5
5.2
20.4
5.2
17.5
24.7 15.3
24.3 25.8
35.4 18.6
26.9 30.5
54.8
29.5
40.7
25.0
15.4 12.3 14.3 10.1 14.3 16.9 16.5
15.4 11.5 11.6 10.4 18.0 14.3 18.9
metal, musical theater, gospel, blues, new age, big band, golden oldies, world/international, disco.” Respondents were also asked “You’ve mentioned that you like . Which one of these is your absolute favourite?” and “You’ve mentioned that you dislike . Which one of these do you dislike the most?” Distributions of these musical tastes variables for the 1,543 respondents who provided information for them are described in Table 2.
Classical Rock Pop Blues Golden oldies Jazz Easy listening Musical theater Folk music Big band World/international Reggae Gospel Country music Choral music Disco Opera New age Hip hop Rap Heavy metal
74.3 67.6 65.1 63.4 62.1 61.7 59.1 57.1 55.5 52.6 52.0 46.5 44.6 43.6 42.5 41.0 39.8 30.7 27.1 16.6 15.0
Like (%)
8.0 16.1 13.1 15.8 15.1 17.0 17.3 19.4 19.2 22.0 13.5 28.4 27.9 27.4 25.8 33.0 30.5 35.2 44.6 60.7 66.3
Dislike (%)
17.7 16.4 21.8 20.8 22.7 21.2 23.6 23.5 25.3 25.4 34.5 25.2 27.6 29.0 31.8 26.1 29.7 34.1 28.3 22.7 18.7
Neutral (%)
Rock Classical Pop Jazz Golden oldies Country Easy listening Blues Folk music Hip hop Opera Gospel World/international Musical theater Reggae Big band Disco Choral music Heavy metal New age Rap
Most liked
20.1 15.2 8.1 7.8 6.4 5.8 5.5 4.7 4.1 3.0 3.0 2.9 2.5 2.1 1.8 1.7 1.7 1.2 1.1 0.8 0.5
%
Heavy metal Rap Country Hip hop Opera Gospel Disco New age Easy listening Reggae Musical theater Choral music Jazz Rock Folk music Classical Golden oldies Pop Big band World/international Blues
Most disliked
Musical Likes and Dislikes of Toronto and Vancouver Respondents
Table 2
33.7 26.5 7.0 6.0 5.0 2.8 2.5 2.3 2.0 1.7 1.6 1.5 1.2 1.2 1.0 0.9 0.9 0.8 0.7 0.4 0.3
%
144 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
145
RESULTS Musical Field In a relational musical field, musical tastes attain their meaning primarily in relation to one another. Table 3 describes bivariate relationships between the 21 musical likes/dislikes variables. Although most of the associations are positive, some of the stronger positive associations, such as those between rap and hip hop, choral and gospel, choral and folk, opera and classical, reggae and rock, and jazz and blues, may reflect blurred or overlapping boundaries between genres. Negative associations between musical tastes are few and far between, the strongest of which are between hip hop and choral, hip hop and folk, heavy metal and choral, heavy metal and easy listening, and heavy metal and golden oldies. Bourdieu ([1979] 1984) used correspondence analysis to craft his depictions of the fields of capitals, lifestyles, and habitus in France. To create a visual rendering of a relational musical field in this study, I applied the Burt method approach to MCA to the 21 musical likes/dislikes variables and plotted the primary dimensions derived from it in a multidimensional space. The first two dimensions extracted by the MCA explained 78.3 percent of the total variability, with 53.2 percent of the variability attributable to Dimension 1 and 25.2 percent of the variability attributable to Dimension 2. The variable categories are plotted visually in Figure 1, with Dimension 1 forming the vertical axis and Dimension 2 forming the horizontal axis. The predominant pattern of musical tastes in Figure 1 is the presence of distinct clusters of categories that illuminate three collectives of people: the members of Quadrant 1 who like most, if not all, musical genres (“inclusive” musical omnivores); people located in the lower half of the field who dislike most genres; and the members of Quadrant 4 who are ambivalent about or unfamiliar with many of the genres. Next, I applied LCA to the set of 21 musical tastes. The loadings of the variable categories for the three classes produced by the LCA are presented in Table 4. Class 1 closely resembles Quadrant 1 of Figure 1, with high loadings for liking all genres excepting hip hop, rap, and heavy metal (“nearly inclusive” musical omnivores). Class 2 strongly overlaps with Quadrant 3 of Figure 1, with high loadings for disliking all genres excepting classical and easy listening but also for liking classical, easy listening, golden oldies, pop, and rock. Class 3 strongly overlaps with Quadrant 4, with high loadings for ambivalent attitudes toward all but classical and rock and for liking classical, pop, jazz, rock, and blues. The LCA therefore illuminates characteristics of the clustering of musical tastes that are perhaps not readily apparent in the MCA, namely, that liking classical, liking pop, liking rock, disliking rap, and disliking heavy metal effectively traverse all three classes and that liking hip hop, liking rap, liking heavy metal, disliking classical, and disliking easy listening do not meaningfully fit in
0.106 0.108 0.233
hip hop
classical
0.203 0.131 0.073 0.266 0.167 0.163 0.100 0.107 0.144 0.158
0.164 0.149 0.485
0.062 0.208 0.109 0.172 0.136 0.192 0.142 0.238 0.169
0.073
0.086 0.253 0.203 0.076 0.319 0.058 0.068 0.204
choral
0.172 0.098 0.150 0.112 0.155 0.272 0.324 0.184 0.168 0.244 0.191 0.208 0.128
0.303 0.110 0.295 0.163
folk
0.215 0.234 0.236 0.139 0.237 0.254 0.226 0.123
0.106 0.166 0.225 0.109 0.185 0.109 0.130 0.078
rap
0.101 0.068 0.139 0.090 0.074 0.246 0.172 0.242 0.080 0.093 0.105 0.136 0.132 0.113 0.099 0.168 opera
0.143 0.104 0.169 0.080 0.116 0.080 0.091 0.224 0.182 0.165 0.122 0.185 0.141 0.218 0.135 country
0.198 0.109 0.218 0.123 0.199 0.094 0.190 0.219 0.121 0.087 0.176 0.243 0.091 0.146 pop
0.171 0.220 0.184 0.333 0.096 0.174 0.130 0.192 0.188 0.145 0.172 0.110 0.260 jazz
0.116 0.200 0.137 0.091 0.192 0.188 0.427 0.153 0.266 0.149 0.179 0.122 easy listening
0.094 0.113 0.131 0.236 0.129 0.089 0.176 0.171 0.277 0.112 0.196 reggae
0.313 0.219 0.130 0.202 0.255 0.237 0.188 0.152 0.221 0.234 rock
0.268 0.107 0.091 0.218 0.187 0.154 0.143 0.118 0.224 heavy metal
0.102 0.110 0.123 0.167 0.120 0.125 0.112 0.116 musical theater
0.290 0.182 0.184 0.299 0.281 0.201 0.224
gospel
0.235 0.163 0.246 0.222 0.217 0.161
Bivariate Associations between Musical Likes/Dislikes
blues
0.220 0.271 0.184 0.211 0.184
new age
0.208 0.150 0.247 0.231
big band
0.402 0.231 0.217
golden oldies
0.201 0.213
Notes: Each cell contains a Cramer’s V value. Cramer’s V ranges from a low of zero, representing no relationship, to a high of one, representing a perfect relationship. Values in bold are positive relationships wherein people who liked the one genre tended to like the other and/or people who disliked the one genre tended to dislike the other. Values in italics are negative relationships wherein people who liked the one genre were relatively unlikely to like the other and/or people who disliked the one genre were relatively unlikely to dislike the other. Values in neither bold or italics are not so straightforwardly described and empty cells in the table indicate relationships that were not statistically significant (p > 0.05).
hip hop choral folk rap opera country pop jazz easy listening reggae rock heavy metal musical theater gospel blues new age big band golden oldies world/international disco
Table 3
0.210 world/ int.
146 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
147 Figure 1
Two-dimensional mapping derived from a multiple correspondence analysis of musical likes and dislikes Quadrant 1
Quadrant 4
neutral rap neutral heavy metal neutral hip hop like new age most neutral reggae neutral disco neutral golden oldies neutral gospel neutral choralneutral opera neutral new age neutral big band dislike new age most neutral country neutral folk neutral musical theater neutral rock like new age like musical theater most like jazz most neutral easy listening like world/international most like folk most like blues most like big band neutral blues pop dislike blues most neutral like musical theater dislike world/international most like reggae like big band most neutral world/international like disco like jazz like blues neutral jazz gospel dislike easy listening most like world/international likelike like pop most like pop like like rock neutral classical classical like folk like opera choral like rock most dislike heavy dislike musical theater most like most heavy metal like golden oldies like easy listening most metal dislike jazz most like hip hoplike easy listening like country dislike choral most like golden oldies most Dimension 2 dislike folk most dislike golden oldies most like rap dislike gospel most like opera most dislike rap most like disco most dislike big band most dislike heavy metal like classical most dislike country most dislike hip hop most like gospel most dislike rap dislike disco most dislike reggae most like country most like choral most dislike hip hop like reggae most dislike opera most dislike country dislike easy listening dislike disco dislike opera dislike new age like hip hop most like heavy metal most
dislike gospel dislike choral dislike folk
dislike pop most
dislike reggae dislike rock
dislike jazz dislike musical theater dislike big band like rap most dislike golden oldies
dislike pop
dislike world/international dislike blues dislike classical
dislike rock most
Legend musical like/dislike (active) most liked genre (supplementary)
dislike classical most
Quadrant 3
most disliked genre (supplementary)
Dimension 1
Quadrant 2
any of them. That is, the predominant pattern evident in the MCA of musical tastes, that distinguishing between liking musical genres in general, disliking musical genres in general or feeling ambivalent about musical genres in general, is not monolithic in its manifestation. To further investigate this point, the most liked and most disliked musical genres were overlaid on the MCA plot as supplementary variables (Figure 1). When pushed to declare their strongest musical allegiances, the inclusive musical omnivores of Quadrant 1 are relatively likely to favor big band, blues, folk, jazz, musical theater, and new age over all others and to dislike easy listening, heavy metal, new age, and world/international most of all. The denizens of Quadrant 3 are relatively likely to favor heavy metal, hip hop, reggae, and rap over all others and to dislike classical, opera, and rock most of all whereas the inhabitants of Quadrant 4 are relatively likely to favor pop, rock, and world/international and to dislike blues and musical theater most of all. Again, tastes in music are not perfectly captured by the predominant pattern evident in Figure 1.
Class 2 0.615 0.232 0.320 0.444
0.309 0.420 0.531 0.420 0.592 0.301 0.523
0.378 0.295 0.402 0.203 0.322 0.514 0.394 0.284
Class 1 0.899 0.279 0.667 0.827
0.580 0.633 0.812 0.864 0.770 0.605 0.766
0.855 0.742 0.890 0.464 0.857 0.900 0.731 0.622
0.430 0.262 0.569 0.243 0.361 0.426 0.412 0.296
0.282 0.258 0.592 0.536 0.418 0.456 0.704
Class 3 0.688 0.288 0.257 0.373 classical - dislike hip hop - dilike choral - dislike folk - dislike rap - dislike opera - dislike country - dislike pop - dislike jazz - dislike easy listening - dislike reggae - dislike rock - dislike heavy metal - dislike musical theater - dislike gospel - dislike blues - dislike new age - dislike big band - dislike golden oldies - dislike world/international - dislike disco - dislike 0.262
0.700
0.573
0.402
Class 1
Class 2 0.210 0.680 0.528 0.410 0.786 0.550 0.424 0.335 0.411 0.280 0.621 0.374 0.814 0.487 0.574 0.453 0.652 0.551 0.382 0.376 0.615 0.251
0.204
0.232
0.517
0.515 0.278 0.298
0.313 0.208
Class 3
Notes: To facilitate interpretation of the classes, values greater than 0.500 are in bold and values less than 0.200 have been deleted .
classical - like hip hop - like choral - like folk - like rap - like opera - like country - like pop - like jazz - like easy listening - like reggae - like rock - like heavy metal - like musical theater - like gospel - like blues - like new age - like big band - like golden oldies - like world/international - like disco - like
classical - neutral hip hop - neutral choral - neutral folk - neutral rap - neutral opera - neutral country - neutral pop - neutral jazz - neutral easy listening - neutral reggae - neutral rock - neutral heavy metal - neutral musical theater - neutral gospel - neutral blues - neutral new age - neutral big band - neutral golden oldies - neutral world/international - neutral disco - neutral
0.231
0.274
0.219
0.230 0.266 0.237
0.319 0.228
Class 1
0.230
Class 2
Loadings on Three Classes from the Latent Class Analysis of Musical Likes/Dislikes
Table 4
Class 3 0.273 0.399 0.534 0.456 0.346 0.440 0.444 0.358 0.365 0.422 0.409 0.217 0.333 0.454 0.505 0.368 0.553 0.513 0.469 0.536 0.453
148 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
149 Figure 2
Sociodemographic variable categories overlaid on Figure 1 Quadrant 1
Quadrant 4
$150 or more
Black
Asian
19 - 34
45 - 54 $60 - 79,999 55 - 64 postgraduate born in Canada university female 35 - 44 high school White $100 - 149,999 immigrated > 20 years $80 - 99,999 < $40,000 male $40 - 59,999 community college or technical school 65 and older other racial identity
Dimension 2
less than high school
immigrated 20 years ago Immigrated ࣘ 20 years ago Born in Canada (reference) Education Less than high school High school Community college or technical school University Postgraduate degree (reference) Household income
The longstanding debate between the homology and omnivorism approaches to the class bases of cultural tastes and practices rages on in cultural sociology. The homology thesis claims that class positions throughout the class hierarchy are accompanied by specified cultural tastes and specialized modes of appreciating them while the cultural omnivorism thesis contends that elites are (increasingly) characterized by a breadth of cultural tastes of any and all kinds. This study tests the applicability of these theses to musical tastes in Canada through the application of multiple correspondence analysis, latent class analysis, and logistic regression modeling to original telephone survey data (n = 1,595) from Toronto and Vancouver. I find that musical omnivorism, an appreciation for diverse musical styles, is not dispersed along class lines. Instead I find a homology between class position and musical tastes that designates blues, choral, classical, jazz, musical theater, opera, pop, reggae, rock, and world/international as relatively highbrow and country, disco, easy listening, golden oldies, heavy metal, and rap as relatively lowbrow. Of the highbrow tastes, all but jazz are disliked by lower class people, and of the lowbrow tastes, country, easy listening, and golden oldies are concurrently disliked by higher class people. Consistent with the homology thesis, it appears that class position is aligned with specific musical likes and dislikes. Le vieux d´ebat entre les approches de l’homologie et de l’omnivorisme ˆ et des pratiques culturels fait rage dans aux bases des classes des gouts la sociologie culturelle. La th`ese de l’homologie pr´etend que les positions des classes a` travers la hi´erarchie des classes sont accompagn´ees par des ˆ culturels sp´ecifi´es et des modes sp´ecialis´es permettant leur gouts appr´eciation. La th`ese de l’omnivorisme culturel, en revanche, soutient ˆ que les e´ lites sont (de plus en plus) caract´eris´ees par un e´ ventail de gouts Gerry Veenstra, Department of Sociology, University of British Columbia, 6303 N. W. Marine Drive, Vancouver, BC, Canada V6T 1Z1. E-mail: [email protected]
C 2015 Canadian Sociological Association/La Soci´ et´e canadienne de sociologie
Class Position and Musical Tastes
135
culturels de toutes sortes. Cette e´ tude exp´erimente l’applicabilit´e de ces ˆ musicaux au Canada a` travers l’application de l’analyse th`eses aux gouts des correspondances multiples, de l’analyse des classes latentes et du mod`ele de r´egression logistique aux donn´ees d’origine de l’´etude collect´ees par t´el´ephone (n = 1,595) de Toronto et de Vancouver. Je me rends compte que l’omnivorisme musical, une appr´eciation de styles musicaux divers, n’est pas dispers´e le long des classes sociales. Bien au contraire, je trouve qu’il existe une homologie entre la position des ˆ musicaux qui d´esigne les blues, la chorale, la musique classes et les gouts ˆ classique, le jazz, le th´eatre musical, l’op´era, la pop, le reggae, le rock et la musique du monde/internationale comme des styles des classes au niveau intellectuel relativement e´ lev´e. Cette homologie d´esigne la country, la disco, la musique d’ambiance, les anciens succ`es, le heavy metal et le rap comme des styles des classes au niveau intellectuel ˆ des classes au niveau intellectuel e´ lev´e, tous relativement bas. Des gouts les styles a` l’exception du jazz ne sont pas appr´eci´es de la basse classe. ˆ des classes au niveau intellectuel bas, le country, la De mˆeme, des gouts musique d’ambiance et les anciens succ`es ne sont pas appr´eci´es de la haute classe. Selon la th`ese de l’homologie, il apparaˆıt que la position des classes est fonction des pr´ef´erences et des aversions musicales sp´ecifiques.
THE HOMOLOGY AND CULTURAL omnivorism frameworks have been at loggerheads for more than 20 years (e.g., Atkinson 2011; Bennett et al. 2009; Coulangeon and Lemel 2007; Gebesmair 1998; Goldberg 2011; Lizardo and Skiles 2012; Peterson and Kern 1996; Peterson and Simkus 1992; Rimmer 2012; Warde, Wright, and Gayo-Cal 2007). The homology thesis claims that class positions throughout the class hierarchy are accompanied by specified cultural tastes and specialized modes of appreciating them while the cultural omnivorism thesis contends that elites are (increasingly) characterized by a breadth of cultural tastes of any and all kinds. This study provides new evidence on the relationship between class position and cultural tastes by analyzing data from a novel telephone survey on music and class conducted in Toronto and Vancouver, Canada. Specifically, it seeks to establish whether elites and lower class people in urban English-speaking Canada are characterized by distinct sets of musical tastes or distinguished from another primarily by the degree to which they manifest omnivorous musical tastes.
THE BOURDIEUSIAN HOMOLOGY FRAMEWORK The homology framework, belonging to a long line of cultural inquiry that includes Thorstein Veblen’s The Theory of the Leisure Class ([1899] 1994) and Herbert Gans’ Popular Culture and High Culture (1974), finds its fullest expression in Pierre Bourdieu’s magnum opus, Distinction: A Social Critique of the Judgement of Taste (Bourdieu [1979] 1984). There the French sociologist maintained that a multitude of diverse cultural tastes
136
CRS/RCS, 52.2 2015
and practices in 1960s France were fundamentally manifestations of class habitus. For Bourdieu, class positions are distinguished by their differing amounts of economic capital, essentially monetary wealth, and cultural capital, valued cultural resources that include educational credentials, which locate them in different parts of a multidimensional social space (a society-wide “field”). Specifically, the sum total of the two forms of capital positions agents on the primary axis of social space, that which distinguishes the upper, middle, and lower classes, while relative composition of the capitals situates agents along the secondary axis of social space which distinguishes dominating and dominated sections of the classes. The cultural tastes of the wealthy and highly educated members of the upper class, comprising highbrow culture, represent the “legitimate,” “sophisticated,” and “cultured” tastes and practices of society. Members of this class have the power to delimit highbrow tastes and appropriate modes of appreciating them and can use their familiarity and facility with these cultural forms to maintain and reinforce boundaries between themselves and others. Many highbrow tastes are especially enshrined in the better educated but less-wealthy segment of the upper class, the home of the intelligentsia, the dominated portion of upper-class space. Lowbrow or “popular” tastes and practices, embraced by the less-wealthy and lesseducated members of society, are the antithesis of highbrow culture; they serve as a negative reference for the tastes of the dominant class (Swartz 1997). Finally, middlebrow culture reflects the (imperfect) attempts of the members of the middle class, the petit bourgeoisie, to embrace highbrow culture and to distinguish themselves from the lower class. For Bourdieu, then, there is a homology, a kind of isomorphic relation or one-to-one correspondence, between the multidimensional space of positions and the multidimensional space of cultural tastes. A key principle underlying the homology of class positions and cultural tastes, according to Bourdieu, is the opposition between the “tastes of luxury (or freedom) and the tastes of necessity” (Bourdieu [1979] 1984:198). The tastes of luxury are the tastes of people born into a habitus which is defined by distance from necessity; these people possess freedoms of thought and action that are facilitated by possession of capital. In particular, freedom from necessity facilitates development of the “aesthetic gaze,” a mode of consumption which stresses appreciation for the form of a cultural object rather than its function and tends to pass aesthetic rather than ethical judgments on it. Accordingly, order, etiquette, and restraint are virtues in upper-class space. The denial of lower, coarse, vulgar, venal, servile—in a word, natural— enjoyment, which constitutes the sacred sphere of culture, implies an affirmation of the superiority of those who can be satisfied with the sublimated, refined, disinterested, gratuitous, distinguished pleasures forever closed to the profane. (Bourdieu [1979] 1984:7)
Class Position and Musical Tastes
137
The middle class, striving for distinction but lacking the capital and habitus needed to fully appropriate upper-class lifestyles, in turn valorizes “asceticism, rigor, legalism” (Bourdieu [1979] 1984:331). In contrast with the bourgeois habitus which conveys an “ethos of ease, a confident relation to the world and the self,” the petit bourgeois habitus presents an “ethos of restriction through pretension” (Bourdieu [1979] 1984:339). The tastes of necessity, finally, most common in the lower class, derive from the need to produce labor at the lowest cost; people with low levels of (especially economic) capital tend to valorize tastes of necessity out of material necessity. Enjoyment and fun, impropriety, and fulfilling material needs are virtues in lower class space because the functions rather than the forms of cultural objects are of central importance here. Bourdieu asserted that “nothing more clearly affirms one’s ‘class’, nothing more infallibly classifies, than tastes in music” (Bourdieu [1979] 1984:18). Upper-class people, free from the strictures of necessity, tend to have a taste for difficult, abstract music, especially difficult classical music. For example, appreciation for Bach’s The Well-Tempered Clavier was prevalent among the most highly educated members of Bourdieu’s survey sample from the 1960s. “As legitimate culture, classical music gathers around it the highest values of aesthetic formalism associated with Kantian ‘distinterestedness’” (Prior 2013:183). The French upper class also rejected with disgust “the most popular and most ‘vulgar’ singers, such as Les Compagnons de la Chanson, Mireille Mathieu, Adamo or Sheila” (Bourdieu [1979] 1984:60). Middle-class space in turn tends to foster appreciation for popularized forms of legitimate music (Gershwin’s Rhapsody in Blue was especially popular among engineers and technicians in Bourdieu’s data) whereas the members of the lower class tend to consume music with simple, repetitive structures (such as the popular waltzes of Strauss) which are typically imposed on them by experts and artists in the field of musical production (Atkinson 2011). From the standpoint of Bourdieu’s homology framework, then, musical tastes and distastes, seemingly personally idiosyncratic in nature, are in fact reflections of class-based habitus. This represents what I think of as the strong form of the homology thesis. It describes a multidimensional space of positions prescribed by a specified complex configuration of capitals, namely, the sum total of economic capital and cultural capital on the primary axis and the relative composition of these forms of capital on the secondary axis. It is a relational visioning of entwined spaces of positions and cultural tastes that requires relational statistical techniques, such as multiple correspondence analysis (MCA) or latent class analysis (LCA), to illuminate the character of the spaces in quantitative data. In the weak form of the homology thesis, the space of positions and the spaces to which it is homologous are not configured in as complicated a manner. For instance, while evidence suggests that the sum total of economic capital and cultural capital
138
CRS/RCS, 52.2 2015
structures the primary dimension of the social spaces of contemporary France (Coulangeon and Lemel 2007), Canada (Veenstra 2010), Finland (Kahma and Toikka 2012), and the United Kingdom (Gayo-Cal et al. 2006; Le Roux et al. 2008) as well as France of the 1960s (Bourdieu [1979] 1984), the relative composition of these capitals does not appear to contribute to structuring the social spaces of contemporary France (Coulangeon and Lemel 2007), Finland (Kahma and Toikka 2012), and the United Kingdom (Gayo-Cal et al. 2006; Le Roux et al. 2008). If one capital or the sum of several capitals effectively delineates class positions then the weak form of the homology thesis is evident and nonrelational techniques such as regression analysis are presumably up to the task of uncovering a homology between positions and tastes. Precisely which capitals delineate class positions and the manner in which they do so in a given context is an empirical question.
CULTURAL OMNIVORISM Most incarnations of the cultural omnivorism perspective in the sociology of culture similarly identify social class as a fundamental basis of cultural tastes and practices. They depart from the homology framework, however, by claiming that breadth of cultural tastes has supplanted possession of specified highbrow/lowbrow tastes as the notable cultural delimiter of class boundaries. It is argued that, while the Bourdieusian homology storyline may have held in past eras, elites in contemporary societies are now better characterized as cultural omnivores than as highbrow aficionados. These omnivores consume and practice culture speaks of many different kinds, from hip hop to heavy metal to classical music, from grungy, muddy team sports such as rugby to ascetic, individual pursuits of body and mind such as yoga and tai chi, and so forth (Veenstra 2010). They seek variety and eclecticism and are intrinsically inclusive and tolerant in their tastes, in contrast with univores, who, displaying taste for a narrow range of activities or objects, are intrinsically exclusivist and intolerant. Members of the upper class are therefore distinguished from members of other classes by the breadth of their cultural repertoires rather than their possession of any specified highbrow cultural tastes or practices. Members of the lower class are in turn distinguished from others by the relatively narrow focus of their cultural repertoires, by their propensities to be univores, not by their allegiances to any specified lowbrow cultural tastes or practices. It is the desire and ability to consume a multitude of diverse cultural forms, including diverse musical styles, that now distinguishes higher class people from lower class people in contemporary societies. If the “highbrow snob” has been replaced by the cultural omnivore and the “lowbrow slob” has been replaced by the univore then the principles, processes, and mechanisms underlying the stratification of cultural tastes have likely changed as well. In particular, the distinction between tastes
Class Position and Musical Tastes
139
of freedom and necessity may no longer be preeminent. Erickson (1996) proposes that “cultural inequality is not so much a hierarchy of tastes (from soap opera to classical opera) as it is a hierarchy of knowledge (from those who know little about soap opera or opera to those who can take part in conversation about both)” (p. 219). Emmison (2003) argues that “cultural mobility entails the display of cultural competence in a plurality of domains with concomitant social rewards accruing to those demonstrating these capacities” (p. 213). Omnivores can move easily among cultural realms (Emmison 2003) and in business might use whatever form of cultural knowledge is necessary to make a good impression in job interviews (Garnett, Guppy, and Veenstra 2008) or build social networks to get a better job (Erickson 1996). In short, that which is now the most useful for getting ahead, namely, a breadth of cultural knowledge and familiarity with multiple cultural forms, is what is now classed. This is presumably because breadth of knowledge/familiarity is more readily achieved by upper-class people but also because people who have it are relatively likely to achieve upper-class standing. Various explanations for the historical change in the fundamental nature of the class bases of culture from homology to omnivorism have been proposed. For example, the increasingly specialized nature of occupations and greater mobility between occupations may have affected people’s cultural repertoires, since occupations tend to foster their own cultures and people who are required to communicate across occupations or move into a new occupation need to be conversant with a wide range of cultural forms (Erickson 2008). Growing income inequality may have contributed to a growing inequality in the ability of people to participate in culture while educational inflation may have produced relatively more highly educated people in upper-end occupations who are conversant with a wider range of cultural forms garnered through their experiences with educational systems (Erickson 2008). Increasing amounts of social mobility, especially from lower to higher strata, may contribute to increasing sociocultural heterogeneity (van Eijck 2001). Peterson (2005) suggests that snobbish exclusion “was an effective marker of status in a relatively homogenous and circumscribed WASP class [but that] omnivorous inclusion seems better adapted to an increasingly global world culture managed by those who make their way, in part, by showing respect for the cultural expressions of others” (p. 273). Increasingly globalized media industries, social and geographic mobility, and varied networks may also be at the root of the open and varied “cosmopolitan habitus” seemingly prevalent among elites (Atkinson 2011). There are several useful ways of distinguishing strong and weak forms of the omnivorism thesis. As with the homology thesis, relational (strong) and nonrelational (weak) renderings of the omnivorism thesis can be envisioned. For instance, one can imagine a social space delimited by sum total and relative composition of capitals where the upper-class section is
140
CRS/RCS, 52.2 2015
bursting with diverse musical tastes and the lower class section is bereft of them, a multidimensional space of positions homologous to a top-heavy space of (omnivorous) tastes. Alternatively, breadth of tastes might simply be associated with one or both of economic capital and cultural capital in linear-causal regression models. The distinction between “omnivorism by volume” and “omnivorism by composition” (Warde et al. 2007) can also designate strong and weak forms of the omnivorism thesis. In regard to the former, omnivorism is defined by indiscriminate liking of any and all cultural forms, a strong statement of the omnivorism thesis. In regard to the latter, however, “some distinctive status orientation is entailed in the patterns of cultural preferences involved” (Warde et al. 2007:145). For example, the definition of musical omnivore employed by Peterson and Simkus (1992) required that they like traditionally highbrow classical musical and opera. Achterberg and Houtman (2005) operationalized musical omnivorousness as the status distance between the most highbrow music and the most lowbrow music chosen by a person. Explicitly incorporating dislikes into her operationalization, Bryson (1996) depicted musical omnivores as people who consumed many different musical genres while simultaneously rejecting the music genres with the least educated fans, these being country, heavy metal, and rap music. In these formulations, musical omnivorism is prescribed in the breadth of its manifestation. Finally, strong and weak forms of the omnivorism thesis can be distinguished by the degree to which breadth of taste has supplanted specified highbrow tastes among elites. Some scholars suggest that highbrow tastes are a thing of the past and the omnivore now reigns supreme in a “post-Bourdieu era” of cultural sociology (Vander Stichele and Laermans 2006). Others contend that the omnivorism framework represents an evolution rather than a repudiation of the homology framework where the former has not (yet) overthrown the latter (e.g., Emmison 2003; Garc´ıa´ Alvarez, Katz-Gerro, and Lop´ez-Sintas 2007; Lop´ez-Sintas and Zerva 2005; Peterson 2005; Peterson and Kern 1996). They suggest that elites can in fact be split into two groups: exclusive highbrows with a univorous and exclusive highbrow taste; and inclusive highbrows with an omnivorous taste inclusive of middlebrow and lowbrow tastes (Emmison 2003; ´ Garc´ıa-Alvarez et al. 2007; Peterson 2005).1 Lowbrow taste, a popular taste that nearly everyone likes, is no longer associated with any specific class ´ (Garc´ıa-Alvarez et al. 2007). In short, the strong form of the omnivorism thesis does away with highbrow snobs entirely while the weak form accommodates the existence of highbrow snobs and omnivorous elites, with
1.
Research in this vein suggests that exclusive highbrows tend to be older and inclusive highbrows tend to be younger (Ollivier 2008). This age- or cohort-related factor speaks to the notion that aggregate changes in elite tastes from highbrow to omnivore are generationally prescribed, with younger omnivore elites replacing aging highbrow exclusivist elites.
Class Position and Musical Tastes
141
their relative weights in numbers and influence a matter of contextual and historical specificity. Multiple formulations of cultural omnivorism and various underlying principles and reasons for historical changes in the class bases of cultural tastes notwithstanding, the body of empirical research purportedly verifying the existence of class-delimited cultural omnivores in various international contexts is now voluminous (Peterson 2005). Cultural omnivorism has become de rigueur in some circles. Other scholars uphold the contemporary viability of the homology perspective, questioning the evidence for and utility in the notion of a class-delimited musical omnivore (e.g., Atkinson 2011; Rimmer 2012; Savage 2006; Warde et al. 2007). This study contributes to the ongoing tussle between the Bourdieusian homology and cultural omnivorism perspectives by investigating relationships between musical likes and dislikes and indicators of class position in original survey data from two large Canadian cities (Toronto and Vancouver).
ANALYTICAL PLAN The analysis proceeds in four stages. First, I investigate whether diverse musical tastes are widely dispersed, as the homology approach might predict, or tightly clustered, as might be consistent with musical omnivorism. To accomplish this, I apply MCA and LCA to a set of 21 musical likes and dislikes, illuminating contours of a musical field that distinguishes between people who like most musical genres, people who dislike most musical genres, and people who are ambivalent about most musical genres. Second, I investigate the degree to which and how this patterning of musical tastes is structured by various markers of inequality, namely, city of residence, age, gender, racial identity, immigrant status, economic capital, and cultural capital. In particular, I seek to determine whether this Canadian musical field is structured by the sum total of economic capital and cultural capital and the relative composition of economic and cultural capitals as the field of cultural tastes was structured in France of the 1960s. I find that the patterning of musical tastes previously identified is partly reflective of age, racial identity, and immigrant status but mostly unreflective of possession of economic capital and cultural capital. Third, I determine whether a measure of breadth of musical likes is predicted by one or more of the markers of inequality in an Ordinary Least Squares (OLS) regression model. I find that women, middle-aged people, and native-born Canadians are relatively likely to be musical omnivores but wealthy and/or highly educated elites are not. Finally, I investigate whether individual musical likes and dislikes are associated with the markers of inequality in binary logistic regression models. Consistent with the weak form of Bourdieu’s homology thesis, I find that elites
142
CRS/RCS, 52.2 2015
and lower class people are linked with distinct sets of musical likes and dislikes.
METHODS Survey Sample Between January and June of 2009, the Survey Research Centre (SRC) at the University of Victoria in British Columbia conducted telephone interviews with 732 adults living in the city of Toronto and 863 adults living in the Vancouver Census Metropolitan Area. The SRC used random-digit dialing techniques to obtain residential telephone numbers, a next-birthday strategy to select one resident per household aged 19 or older to interview, and a computer-aided telephone interviewing system to conduct the interviews. The introductory script from the callers informed prospective interviewees that all adults aged 19 and older were eligible for this study, that there were potential risks of an emotional kind from questions pertaining to racial and ethnic identity, experiences of discrimination in everyday life and sexual orientation, and that the ultimate goal of the study was to help to improve the health of Canadians. No incentives, monetary or otherwise, were provided for participation. In total, callers spoke with 17,060 people and successfully recruited 1,595 interviewees, representing a cooperation rate of 9.3 percent. Demographic characteristics of the survey samples and the cities from which they were obtained are described in Table 1. Comparison of the city samples to the 2006 Census by gender, age, racial identity, immigrant status, and educational attainment indicates that both city samples are biased toward women, older people, Whites, native-born Canadians, and people with a university degree. Comparison of cases with complete data (n = 1,503) with those with missing data (n = 92) indicates that the missing cases are relatively likely to be non-White immigrants. Musical Tastes Unlike most research of its kind (but see Bennett et al. 2009; Bryson 1996, 1997; Carrabine and Longhurst 1999; Savage 2006; Sonnett 2004; Tampubolon 2008b), this study focuses on musical likes and dislikes. The cultural omnivore is identified by her liking of diverse cultural forms. From the homology perspective, however, the tastes of elites and lower class people are as or more likely to be manifested as distastes than as tastes (Bourdieu [1979] 1984). My respondents were asked about their likes and dislikes in regard to 21 musical genres. “For each of the following types of music, please tell me whether you like or dislike or perhaps feel neutrally about each type: classical music, hip hop, choral music, folk music, rap, opera, country music, pop, jazz, easy listening, reggae, rock, heavy
Class Position and Musical Tastes
143 Table 1
Characteristics of the City Survey Samples and Their Populations Toronto
Variable
Categories
Gender
Male Female Aged 19–34 Aged 35–44 Aged 45–54 Aged 55–64 Aged 65 and older Asian Black South Asian White Other Born in Canada Immigrated to Canada Less than high school High school Community college or technical school diploma Bachelor’s degree or higher Less than $40,000 $40,000–59,999 $60,000–79,999 $80,000–99,999 $100,000–149,999 $150,000 or more Missing
Age
Racial identity
Immigrant status Educational attainment
Household income
Survey sample N = 732 %
Vancouver
Census 2006 %
Survey sample N = 863 %
Census 2006 %
35.4 64.6 20.4 18.5 23.6 19.6 17.8 3.0 6.4 4.1 79.3 7.1 65.3 34.7
47.3 52.7 29.7 21.0 18.3 13.0 17.9 13.3 8.4 12.1 53.0 13.2 47.8 52.2
32.8 67.2 12.5 16.4 23.7 26.9 20.5 6.1 0.2 5.1 82.7 5.9 72.3 27.7
48.2 51.8 28.6 21.2 20.3 14.1 15.8 23.5 1.0 10.9 54.7 9.9 56.5 43.5
5.2
20.4
5.2
17.5
24.7 15.3
24.3 25.8
35.4 18.6
26.9 30.5
54.8
29.5
40.7
25.0
15.4 12.3 14.3 10.1 14.3 16.9 16.5
15.4 11.5 11.6 10.4 18.0 14.3 18.9
metal, musical theater, gospel, blues, new age, big band, golden oldies, world/international, disco.” Respondents were also asked “You’ve mentioned that you like . Which one of these is your absolute favourite?” and “You’ve mentioned that you dislike . Which one of these do you dislike the most?” Distributions of these musical tastes variables for the 1,543 respondents who provided information for them are described in Table 2.
Classical Rock Pop Blues Golden oldies Jazz Easy listening Musical theater Folk music Big band World/international Reggae Gospel Country music Choral music Disco Opera New age Hip hop Rap Heavy metal
74.3 67.6 65.1 63.4 62.1 61.7 59.1 57.1 55.5 52.6 52.0 46.5 44.6 43.6 42.5 41.0 39.8 30.7 27.1 16.6 15.0
Like (%)
8.0 16.1 13.1 15.8 15.1 17.0 17.3 19.4 19.2 22.0 13.5 28.4 27.9 27.4 25.8 33.0 30.5 35.2 44.6 60.7 66.3
Dislike (%)
17.7 16.4 21.8 20.8 22.7 21.2 23.6 23.5 25.3 25.4 34.5 25.2 27.6 29.0 31.8 26.1 29.7 34.1 28.3 22.7 18.7
Neutral (%)
Rock Classical Pop Jazz Golden oldies Country Easy listening Blues Folk music Hip hop Opera Gospel World/international Musical theater Reggae Big band Disco Choral music Heavy metal New age Rap
Most liked
20.1 15.2 8.1 7.8 6.4 5.8 5.5 4.7 4.1 3.0 3.0 2.9 2.5 2.1 1.8 1.7 1.7 1.2 1.1 0.8 0.5
%
Heavy metal Rap Country Hip hop Opera Gospel Disco New age Easy listening Reggae Musical theater Choral music Jazz Rock Folk music Classical Golden oldies Pop Big band World/international Blues
Most disliked
Musical Likes and Dislikes of Toronto and Vancouver Respondents
Table 2
33.7 26.5 7.0 6.0 5.0 2.8 2.5 2.3 2.0 1.7 1.6 1.5 1.2 1.2 1.0 0.9 0.9 0.8 0.7 0.4 0.3
%
144 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
145
RESULTS Musical Field In a relational musical field, musical tastes attain their meaning primarily in relation to one another. Table 3 describes bivariate relationships between the 21 musical likes/dislikes variables. Although most of the associations are positive, some of the stronger positive associations, such as those between rap and hip hop, choral and gospel, choral and folk, opera and classical, reggae and rock, and jazz and blues, may reflect blurred or overlapping boundaries between genres. Negative associations between musical tastes are few and far between, the strongest of which are between hip hop and choral, hip hop and folk, heavy metal and choral, heavy metal and easy listening, and heavy metal and golden oldies. Bourdieu ([1979] 1984) used correspondence analysis to craft his depictions of the fields of capitals, lifestyles, and habitus in France. To create a visual rendering of a relational musical field in this study, I applied the Burt method approach to MCA to the 21 musical likes/dislikes variables and plotted the primary dimensions derived from it in a multidimensional space. The first two dimensions extracted by the MCA explained 78.3 percent of the total variability, with 53.2 percent of the variability attributable to Dimension 1 and 25.2 percent of the variability attributable to Dimension 2. The variable categories are plotted visually in Figure 1, with Dimension 1 forming the vertical axis and Dimension 2 forming the horizontal axis. The predominant pattern of musical tastes in Figure 1 is the presence of distinct clusters of categories that illuminate three collectives of people: the members of Quadrant 1 who like most, if not all, musical genres (“inclusive” musical omnivores); people located in the lower half of the field who dislike most genres; and the members of Quadrant 4 who are ambivalent about or unfamiliar with many of the genres. Next, I applied LCA to the set of 21 musical tastes. The loadings of the variable categories for the three classes produced by the LCA are presented in Table 4. Class 1 closely resembles Quadrant 1 of Figure 1, with high loadings for liking all genres excepting hip hop, rap, and heavy metal (“nearly inclusive” musical omnivores). Class 2 strongly overlaps with Quadrant 3 of Figure 1, with high loadings for disliking all genres excepting classical and easy listening but also for liking classical, easy listening, golden oldies, pop, and rock. Class 3 strongly overlaps with Quadrant 4, with high loadings for ambivalent attitudes toward all but classical and rock and for liking classical, pop, jazz, rock, and blues. The LCA therefore illuminates characteristics of the clustering of musical tastes that are perhaps not readily apparent in the MCA, namely, that liking classical, liking pop, liking rock, disliking rap, and disliking heavy metal effectively traverse all three classes and that liking hip hop, liking rap, liking heavy metal, disliking classical, and disliking easy listening do not meaningfully fit in
0.106 0.108 0.233
hip hop
classical
0.203 0.131 0.073 0.266 0.167 0.163 0.100 0.107 0.144 0.158
0.164 0.149 0.485
0.062 0.208 0.109 0.172 0.136 0.192 0.142 0.238 0.169
0.073
0.086 0.253 0.203 0.076 0.319 0.058 0.068 0.204
choral
0.172 0.098 0.150 0.112 0.155 0.272 0.324 0.184 0.168 0.244 0.191 0.208 0.128
0.303 0.110 0.295 0.163
folk
0.215 0.234 0.236 0.139 0.237 0.254 0.226 0.123
0.106 0.166 0.225 0.109 0.185 0.109 0.130 0.078
rap
0.101 0.068 0.139 0.090 0.074 0.246 0.172 0.242 0.080 0.093 0.105 0.136 0.132 0.113 0.099 0.168 opera
0.143 0.104 0.169 0.080 0.116 0.080 0.091 0.224 0.182 0.165 0.122 0.185 0.141 0.218 0.135 country
0.198 0.109 0.218 0.123 0.199 0.094 0.190 0.219 0.121 0.087 0.176 0.243 0.091 0.146 pop
0.171 0.220 0.184 0.333 0.096 0.174 0.130 0.192 0.188 0.145 0.172 0.110 0.260 jazz
0.116 0.200 0.137 0.091 0.192 0.188 0.427 0.153 0.266 0.149 0.179 0.122 easy listening
0.094 0.113 0.131 0.236 0.129 0.089 0.176 0.171 0.277 0.112 0.196 reggae
0.313 0.219 0.130 0.202 0.255 0.237 0.188 0.152 0.221 0.234 rock
0.268 0.107 0.091 0.218 0.187 0.154 0.143 0.118 0.224 heavy metal
0.102 0.110 0.123 0.167 0.120 0.125 0.112 0.116 musical theater
0.290 0.182 0.184 0.299 0.281 0.201 0.224
gospel
0.235 0.163 0.246 0.222 0.217 0.161
Bivariate Associations between Musical Likes/Dislikes
blues
0.220 0.271 0.184 0.211 0.184
new age
0.208 0.150 0.247 0.231
big band
0.402 0.231 0.217
golden oldies
0.201 0.213
Notes: Each cell contains a Cramer’s V value. Cramer’s V ranges from a low of zero, representing no relationship, to a high of one, representing a perfect relationship. Values in bold are positive relationships wherein people who liked the one genre tended to like the other and/or people who disliked the one genre tended to dislike the other. Values in italics are negative relationships wherein people who liked the one genre were relatively unlikely to like the other and/or people who disliked the one genre were relatively unlikely to dislike the other. Values in neither bold or italics are not so straightforwardly described and empty cells in the table indicate relationships that were not statistically significant (p > 0.05).
hip hop choral folk rap opera country pop jazz easy listening reggae rock heavy metal musical theater gospel blues new age big band golden oldies world/international disco
Table 3
0.210 world/ int.
146 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
147 Figure 1
Two-dimensional mapping derived from a multiple correspondence analysis of musical likes and dislikes Quadrant 1
Quadrant 4
neutral rap neutral heavy metal neutral hip hop like new age most neutral reggae neutral disco neutral golden oldies neutral gospel neutral choralneutral opera neutral new age neutral big band dislike new age most neutral country neutral folk neutral musical theater neutral rock like new age like musical theater most like jazz most neutral easy listening like world/international most like folk most like blues most like big band neutral blues pop dislike blues most neutral like musical theater dislike world/international most like reggae like big band most neutral world/international like disco like jazz like blues neutral jazz gospel dislike easy listening most like world/international likelike like pop most like pop like like rock neutral classical classical like folk like opera choral like rock most dislike heavy dislike musical theater most like most heavy metal like golden oldies like easy listening most metal dislike jazz most like hip hoplike easy listening like country dislike choral most like golden oldies most Dimension 2 dislike folk most dislike golden oldies most like rap dislike gospel most like opera most dislike rap most like disco most dislike big band most dislike heavy metal like classical most dislike country most dislike hip hop most like gospel most dislike rap dislike disco most dislike reggae most like country most like choral most dislike hip hop like reggae most dislike opera most dislike country dislike easy listening dislike disco dislike opera dislike new age like hip hop most like heavy metal most
dislike gospel dislike choral dislike folk
dislike pop most
dislike reggae dislike rock
dislike jazz dislike musical theater dislike big band like rap most dislike golden oldies
dislike pop
dislike world/international dislike blues dislike classical
dislike rock most
Legend musical like/dislike (active) most liked genre (supplementary)
dislike classical most
Quadrant 3
most disliked genre (supplementary)
Dimension 1
Quadrant 2
any of them. That is, the predominant pattern evident in the MCA of musical tastes, that distinguishing between liking musical genres in general, disliking musical genres in general or feeling ambivalent about musical genres in general, is not monolithic in its manifestation. To further investigate this point, the most liked and most disliked musical genres were overlaid on the MCA plot as supplementary variables (Figure 1). When pushed to declare their strongest musical allegiances, the inclusive musical omnivores of Quadrant 1 are relatively likely to favor big band, blues, folk, jazz, musical theater, and new age over all others and to dislike easy listening, heavy metal, new age, and world/international most of all. The denizens of Quadrant 3 are relatively likely to favor heavy metal, hip hop, reggae, and rap over all others and to dislike classical, opera, and rock most of all whereas the inhabitants of Quadrant 4 are relatively likely to favor pop, rock, and world/international and to dislike blues and musical theater most of all. Again, tastes in music are not perfectly captured by the predominant pattern evident in Figure 1.
Class 2 0.615 0.232 0.320 0.444
0.309 0.420 0.531 0.420 0.592 0.301 0.523
0.378 0.295 0.402 0.203 0.322 0.514 0.394 0.284
Class 1 0.899 0.279 0.667 0.827
0.580 0.633 0.812 0.864 0.770 0.605 0.766
0.855 0.742 0.890 0.464 0.857 0.900 0.731 0.622
0.430 0.262 0.569 0.243 0.361 0.426 0.412 0.296
0.282 0.258 0.592 0.536 0.418 0.456 0.704
Class 3 0.688 0.288 0.257 0.373 classical - dislike hip hop - dilike choral - dislike folk - dislike rap - dislike opera - dislike country - dislike pop - dislike jazz - dislike easy listening - dislike reggae - dislike rock - dislike heavy metal - dislike musical theater - dislike gospel - dislike blues - dislike new age - dislike big band - dislike golden oldies - dislike world/international - dislike disco - dislike 0.262
0.700
0.573
0.402
Class 1
Class 2 0.210 0.680 0.528 0.410 0.786 0.550 0.424 0.335 0.411 0.280 0.621 0.374 0.814 0.487 0.574 0.453 0.652 0.551 0.382 0.376 0.615 0.251
0.204
0.232
0.517
0.515 0.278 0.298
0.313 0.208
Class 3
Notes: To facilitate interpretation of the classes, values greater than 0.500 are in bold and values less than 0.200 have been deleted .
classical - like hip hop - like choral - like folk - like rap - like opera - like country - like pop - like jazz - like easy listening - like reggae - like rock - like heavy metal - like musical theater - like gospel - like blues - like new age - like big band - like golden oldies - like world/international - like disco - like
classical - neutral hip hop - neutral choral - neutral folk - neutral rap - neutral opera - neutral country - neutral pop - neutral jazz - neutral easy listening - neutral reggae - neutral rock - neutral heavy metal - neutral musical theater - neutral gospel - neutral blues - neutral new age - neutral big band - neutral golden oldies - neutral world/international - neutral disco - neutral
0.231
0.274
0.219
0.230 0.266 0.237
0.319 0.228
Class 1
0.230
Class 2
Loadings on Three Classes from the Latent Class Analysis of Musical Likes/Dislikes
Table 4
Class 3 0.273 0.399 0.534 0.456 0.346 0.440 0.444 0.358 0.365 0.422 0.409 0.217 0.333 0.454 0.505 0.368 0.553 0.513 0.469 0.536 0.453
148 CRS/RCS, 52.2 2015
Class Position and Musical Tastes
149 Figure 2
Sociodemographic variable categories overlaid on Figure 1 Quadrant 1
Quadrant 4
$150 or more
Black
Asian
19 - 34
45 - 54 $60 - 79,999 55 - 64 postgraduate born in Canada university female 35 - 44 high school White $100 - 149,999 immigrated > 20 years $80 - 99,999 < $40,000 male $40 - 59,999 community college or technical school 65 and older other racial identity
Dimension 2
less than high school
immigrated 20 years ago Immigrated ࣘ 20 years ago Born in Canada (reference) Education Less than high school High school Community college or technical school University Postgraduate degree (reference) Household income
