“ANYTHING BUT HEAVY METAL”: SYMBOLIC EXCLUSION AND MUSICAL DISLIKES*

“ANYTHING BUT HEAVY METAL”: SYMBOLIC EXCLUSION AND MUSICAL DISLIKES*

Bethany Bryson Princeton University

I provide quantitative evidence of a cultural phenomenon. Using data on musical dislikes from the 1993 General Social Survey, I link literatures on taste, racism, and democratic liberalism by showing that people use cultural taste to reinforce symbolic boundaries between themselves and categories of people they dislike. Contrary to Bourdieu’s (1984) prediction, musical ex- clusiveness decreases with education. Also, political tolerance is associated with musical tolerance, even controlling for educational attainment, and rac- ism increases the probability of disliking genres whose fans are dispropor- tionately non-White. Tolerant musical taste, however, is found to have a spe- cific pattern of exclusiveness. Those genres whose fans have the least educa- tion-gospel, country, rap, and heavy metal-are also those most likely to be rejected by the musically tolerant. Broad familiarity with music genres is also significantly related to education. I suggest, therefore, that cultural tol- erance constitutes multicultural capital as it is unevenly distributed in the population and evidences class-based exclusion.

R esearch on cultural taste assumes that patterns of taste reflect and influence

social structure and economic inequality. Taste is theorized to act as a basis for exclu- sion (Weber [1968] 1978; Parkin 1979; Bour- dieu 1984; Lamont and Lareau 1988) in which one group prevents other groups from gaining access to valuable resources, such as educational credentials (Milner 1972; Bour- dieu and Passeron 1977; DiMaggio 1982), business contacts (Kanter 1977), or marital

* Direct all correspondence to Bethany Bryson, 2-N-i Green Hall, Princeton University, Prince- ton, NJ 08544 (bpbryson@princeton.edu). Ear- lier versions of this paper were presented at the annual meetings of the American Sociological Association (Los Angeles 1994) and of the South- ern Sociological Society (Atlanta 1995). I owe special thanks to Paul DiMaggio and Michele Lamont as well as to Courtney Bender, Frank Dobbin, Timothy Dowd, Richard Peterson, Michael Relish, Bruce Western, Charles Willie, and four ASR reviewers for their careful readings and insightful comments on earlier drafts. Thanks also go to members of the GSS Culture Module Workshop at Princeton University whose discus- sion and support helped this paper take shape. [Reviewers acknowledged by the author are Diane L. Barthel, Joseph Kotarba, Lynn Smith- Lovin, and Deena Weinstein. -ED. ]

partners (Kalmijn 1994). According to theo- rists of elite culture, high-status individuals see other cultural forms as crude (Bourdieu 1984), vulgar (Veblen [1889] 1953), or dis- honorable (Weber [1968] 1978). These theo- rists suggest that as occupational prestige in- creases, so does the cultivation of cultural “distinction”-a process involving rejection of other, nonelite, cultural patterns. Despite the centrality of negative cultural evaluations in these theories, survey researchers have studied taste only as cultural preference. To the extent that the failure to hold a particular cultural preference is not the same as dislik- ing the item, the methodological convention of studying preferences has prevented us from testing a central assumption in the field of culture and inequality: Do high-status people actually dislike low-status culture?

Another important body of sociological lit- erature posits that education increases toler- ance of political and religious nonconformity (Adorno et al. 1950; Stouffer 1955; Davis 1975; Nunn, Crockett, and Williams 1978; Lipset 1981), racial integration (Greeley and Sheatsley 1971; Hyman and Wright 1979), and many normative violations (Davis 1982). Although Adorno, Stouffer, Davis, and oth- ers focus on cultural, ideological, and reli-

884 AAerican Sociological Review, 1996, Vol. 61 (October:884-899)

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
“ANYTHING BUT.. .”: SYMBOLIC EXCLUSION AND MUSICAL DISLIKES 885

gious differences, they argue that these dif- ferences reflect political rather than cultural orientations. The sociology of culture has not considered the implications of these findings for theories of status.

I show how research on racism and demo- cratic liberalism can generate fruitful re- search questions for the study of cultural taste and social inequality by means of a single linking proposition: Individuals use cultural taste to reinforce symbolic bound- aries between themselves and categories of people they dislike. To test the resulting hy- potheses, I use the number of music genres respondents dislike-rather than prefer-as a measure of cultural exclusion.

The first survey data on cultural dislikes are available from the 1993 General Social Survey, which used Likert scaling for a set of questions on musical taste. First, I use these data to test the proposition that high- status people are more culturally exclusive than other people. Next, I test whether demo- cratic liberalism is associated with decreased exclusiveness in musical taste. Third, I test my hypothesis that negative attitudes toward African Americans shape dislikes in musical taste. And finally, I identify the exclusionary boundaries of “broad taste.”

BACKGROUND

Music as a Symbolic Resource

Music has long been considered an important part of social life. Its symbolic and ritual powers are used to explain both social cohe- sion and cultural resistance (Willis 1977; Hebdige 1979; Rose 1994). Furthermore, music is an important cultural and communi- cative medium. For instance, Cerulo (1995) describes how national anthems represent identity and communicate a nation’s position in the world system. Likewise, Weinstein (199 1) demonstrates that heavy metal music generates community and solidarity among fans while sending an unmistakable message to its detractors.

Music contains a complex set of dimen- sions, sounds, lyrics, visual cues, social rela- tions, and physical acts. (DeNora 1991; Dowd 1992). Music also permits many lev- els of engagement, from humming to oneself to screaming above the music with 30,000

fans. Given its symbolic and social potency, it is no wonder that music is such an impor- tant part of human society, that nearly every nation has an anthem, that most religious cer- emonies involve music, and that singing is so frequently a part of political rallies. The im- portance of music to group identity and so- cial differentiation, then, suggests that musi- cal taste provides a good test for questions about symbolic boundaries. Therefore, I use musical taste to examine a more general theory of cultural exclusion.

High-Status Exclusiveness

Most sociologists of culture agree that some forms of cultural consumption serve as mark- ers of social status (Weber [1968] 1978). For instance, knowledge of fine arts, literature, and upper-class etiquette signals wealth and prestige. Such knowledge may also serve as a passkey for entrance into elite social life. Bourdieu (Bourdieu 1984; Bourdieu and Passeron 1977) calls this passkey cultural capital because it is cultural knowledge that can be translated into real economic gains, for example, by allowing access to elite so- cial networks and clubs where business deals often are made (Kanter 1977).

By restricting access to resources, social status can be translated into market position and political status. This process can be seen as the result of two interrelated levels of ex- clusion. First, social exclusion is a process of social selection that is based on a previ- ously determined set of cultural criteria and is exercised by people with high levels of in- come, education, and occupational prestige (Bourdieu and Passeron 1977). Social exclu- sion occurs at the level of social relations and is the sort of “social closure” that Weber ([1968] 1978: 342, 933, 935) addresses as the monopolization of resources and inclu- sion in social intercourse.

The second level, symbolic exclusion, is the source of those “previously determined cultural criteria.” Whereas social exclusion refers to the monopolization of human inter- actions, symbolic exclusion depicts the sub- jective process that orders those social inter- actions-taste. This process, then, is a form of “boundary-work” (Gieryn 1983; Lamont 1992) that continuously recreates the posi- tive, negative, and neutral attitudes toward

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
886 AMERICAN SOCIOLOGICAL REVIEW

cultural cues and that define these cues as more or less acceptable in various situations. The present study focuses on symbolic ex- clusion. The analytical distinction between social exclusion and symbolic exclusion highlights an important empirical difference between behavior and attitudes. Note, how- ever, that symbolic systems are social and that social exclusion can occur without physical interaction.

Music is one type of cue that can be used to construct symbolic boundaries between groups or individuals. Therefore, I analyze musical exclusion as a type of symbolic ex- clusion and operationalize it as dislike for various music genres. I use the terms musi- cal tolerance or cultural tolerance to refer to the absence of dislike for a cultural cue or music genre. Musical tolerance, then, is operationalized as the complement of musi- cal exclusiveness-not its opposite.’

The crux of symbolic exclusion is dislike, and according to Bourdieu, the exercise of dislike and exclusion is more important to high-status individuals than to others:

Tastes (i.e., manifested preferences) are the practical affirmation of an inevitable differ- ence. It is no accident that, when they have to be justified, they are asserted purely negatively, by the refusal of other tastes. In matters of taste more than anywhere else, all determination is negation; and tastes are perhaps first and fore- most distastes…. The most intolerable thing for those who regard themselves as the possess- ors of legitimate [“highbrow”] culture is the sacrilegious reuniting of tastes which taste dic- tates shall be separated (Bourdieu 1984:56-57).

For Bourdieu, the relationship between the symbolic level and the social level is recipro- cal (Bourdieu and Wacquant 1992). While

they shape each other, other material and sub- jective factors intervene to prevent the two levels from being perfectly aligned.2 Sym- bolic exclusion and social exclusion are as- sumed to work in a manner similar to another pair of terms more familiar to American soci- ologists-prejudice and discrimination.

Bourdieu’s (1984) main exposition on what I have called “symbolic exclusion” argues that knowledge about fine arts is a status cue while popular taste is rejected. “The higher the level of education, the greater is the pro- portion of respondents who, when asked whether a series of objects would make beau- tiful photographs, refuse the ordinary objects of popular admiration . .. as ‘vulgar’ or ‘ugly”‘ (Bourdieu 1984:35). Bourdieu’s per- spective, then, expects high-status individu- als to be the most culturally exclusive. That is, they distinguish themselves with an exclu- sive culture that rejects the cultural patterns and tastes of other groups.

Educated Tolerance

When the well-documented finding that edu- cation increases political tolerance (Adorno et al. 1950; Stouffer 1955; Davis 1975; Nunn, Crockett, and Williams 1978; Lipset 1981) is extended to cultural tolerance, the predicted effect of education is the opposite of that expected by theories of high-status exclusiveness.

To the extent that political tolerance is a belief that civil liberties should be extended to nonconformist groups (Stouffer 1955), po- litical intolerance is a measure of symbolic exclusion. That is, political tolerance refers to the willingness to include specified groups within the boundary of “citizen”–or “us” as opposed to “them” (Gamson 1995). In the realm of public opinion, then, the term “po- litical tolerance” can be seen as a general re- luctance to symbolically exclude noncon- formists from the category “citizen.”

To link these two literatures, I propose that dislike of social groups is associated with dislike of music genres. My specific expec-

I Sullivan, Pierson, and Marcus (1979) propose an “alternative conception” of tolerance in which dislike is a prerequisite of tolerance. However, I use measures based on Stouffer (1955) that do not include the dislike requirement. In symbolic ex- clusion, there is no point beyond dislike (though dislike may vary in intensity). Thus, cultural/mu- sical tolerance refers only to the absence of dis- like. Furthermore, the concept of “indifference” challenges the binary opposition of “inclusion” and “exclusion.” Here, tolerance is the comple- ment of exclusiveness because it includes its op- posite, inclusiveness, plus a residual category, in- difference.

2 For simplicity, I have omitted an important mediating variable in my description of Bourdieu’s theory. See discussions of habitus in Bourdieu (1977) and Bourdieu and Passeron (1977).

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
“ANYTHING BUT . .”: SYMBOLIC EXCLUSION AND MUSICAL DISLIKES 887

tations are twofold. First, political intoler- ance-the general tendency to exclude social groups symbolically-should be positively related to musical exclusiveness-the gen- eral tendency to exclude music genres sym- bolically. Second, because I see both politi- cal intolerance and musical exclusiveness as forms of symbolic exclusion, contra Bour- dieu’s prediction, I expect education to re- duce musical exclusiveness, just as it reduces political intolerance, and income and occu- pational prestige are expected to have little or no effect on musical exclusiveness when the impact of education is held constant (Davis 1975).

Symbolic Racism

Kinder and Sears (1981) propose a two-stage description of racism and public opinion. Termed symbolic racism, the model suggests, first, that racism shapes cultural (value) ori- entations and, second, that racism and the re- sulting set of orientations together may ex- plain public opinion about interracial issues. Whites’ stereotypes about African Ameri- cans, which can be considered symbolic ex- clusion, may be good predictors of Whites’ discomfort with residential integration, which can be considered an estimate of so- cial exclusion (Farley et al. 1994).

I provide a theoretical foundation and an empirical test for the relationship between racism and cultural orientations that has been named symbolic racism. Here, “stereotypes” are understood as symbolic boundaries be- tween social groups that reinforce simple dislike. These “stereotypes” or cultural dif- ferentiations are, furthermore, extended from the realm of values (usually relating to work, family, and economics) to the field of musi- cal taste. Thus, racism is expected to predict dislike for the types of music that are dispro- portionately liked by Hispanic Americans or African Americans.

Patterned Tolerance

Recent research on political tolerance raises new questions about the reason for and uni- versality of education’s liberalizing effect (Phelan et al. 1995). Jackman and Muha (1984) critique the earlier assertions of Stou- ffer (1955), Davis (1975) and others that

education increases democratic liberalism through simple enlightenment. Jackman and Muha claim that highly educated people have only a superficial commitment to the rheto- ric of democratic liberalism, and oppose real social changes if the changes threaten their status. Jackman and Muha show that the strong effects of education on abstract beliefs about the importance of racial equality are not present for attitudes about concrete ac- tions intended to foster racial inequality. Their work suggests that the political toler- ance displayed by educated respondents is, in fact, only a carefully cultivated status symbol.

In a new formulation of the superficial ide- ology explanation, Schuman and Bobo (1988) show that opposition to neighborhood racial integration may be based on perceived class differences between Whites and Afri- can Americans rather than a lack of commit- ment to racial equality. In abstract form and when the class status of an African American family is at least equal to that of the respon- dent, racial integration is approved, but in concrete form, respondents often see residen- tial integration as the entrance of lower-class families into middle-class neighborhoods. Thus, respondents displayed a commitment to democratic liberalism with respect to ra- cial integration but continued to resist class integration.

If my proposition that dislike of a social group is evidenced by dislike of that group’s perceived culture is correct, Schuman and Bobo’s (1988) findings suggest that the ap- parently tolerant tastes of educated respon- dents may mask a systematic dislike of mu- sic genres whose audiences have lower than average levels of education. This prediction has important implications for our under- standing of the wide-ranging tastes of highly educated cultural “omnivores” (Peterson 1992; Peterson and Simkus 1992; Peterson and Kern 1996, in this ASR issue). That is, rather than being indiscriminately broad, om- nivorous taste may include high-status types of music that are popular among non-Whites, especially “world music” (Peterson 1990) genres like reggae and Latin music, while ex- cluding low-status genres like gospel and country regardless of their association with race or ethnicity. (See DiMaggio and Peterson 1975 for a discussion of country

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
888 AMERICAN SOCIOLOGICAL REVIEW

music’s status and audience.) Identifying boundaries around broad taste would allow us to more confidently interpret Peterson and Kern’s (1996) findings as a specific pattern of taste, rather than as evidence against the existence of high-status culture (Halle 1993).

A tendency for patterns of broad taste to exclude low-status genres would suggest that cultural breadth, or tolerance, could itself be a source of cultural capital. Unlike the refined form of cultural capital that Bourdieu (1984) documented in France, however, this contem- porary American emphasis on breadth and tolerance would be more accurately described as multicultural capital-the social prestige afforded by familiarity with a range of cul- tural styles that is both broad and predictably exclusive.3 I add the term “multi” to “cultural capital” in order to specify a content of cul- tural capital, not to modify its meaning. That is, multicultural capital should not be in- cluded in a list of “types” of capital (e.g., so- cial capital, cultural capital, and economic capital). However, the term could be used in an as yet nonexistent list of types of cultural capital (e.g., multicultural capital, high-cul- tural capital, counter-cultural capital, techno- cultural capital, etc.). (See Lamont 1992 and Erikson 1991 for work in this direction.)

This specific pattern of broad taste can be considered a form of cultural capital to the extent that it meets three criteria (Lamont and Lareau 1988). First, cultural tolerance and openness are widely recognized as symbols of social status among upper-middle-class Americans (Lamont 1992), and that recogni- tion is evident, though less pervasive, in the working class (Lamont forthcoming). Sec- ond, familiarity with this cultural style must, nevertheless, be at least somewhat restricted. Using Bourdieu’s (1984, chap. 8) methodol- ogy, then, the frequency of “don’t know” re- sponses to questions about musical taste is expected to decrease with education. The third characteristic of cultural capital is that it can serve as the basis of social exclusion. In this case, the potential for exclusion would

be evidenced by a class-based distribution of cultural tolerance, on one hand, and a pre- dictable pattern of symbolic exclusion (more dislike of low-status genres), on the other.

HYPOTHESES

High Status Exclusiveness

HI: People with high levels of education, in- come, and occupational prestige dislike more types of music than do people with low levels of education, income, and prestige.

Educated Tolerance

H2: People with high levels of education dis- like fewer types of music than do people with medium and low education, con- trolling for income and occupational prestige.

H3: People who are reluctant to extend civil liberties to stigmatized groups dislike more types of music than do people with more tolerant political attitudes.

Symbolic Racism

H4: Whites who have high racism scores dis- like the types of music that are dispro- portionately liked by people of color more than do people who report less rac- ist attitudes.

Patterned Tolerance

H5: People who dislike few music genres will dislike those types of music that are liked by people with low levels of edu- cation more than other types of music, when education is controlled.

H6: People who have high levels of educa- tion are less likely to report that they are unfamiliar with any music genre.

MEASURES

Dependent Variables

The General Social Survey (GSS) is a nearly annual survey of noninstitutionalized adults in the United States conducted by the Na- tional Opinion Research Center using a strati- fied random sampling method. The 1993 GSS

3″Multiculturalism” in Canada refers to a gov- ernment policy on intercultural relations. In the United States, however, the word invokes a cel- ebration of cultural difference and is historically associated with “cultural pluralism” and “diver- sity” (Feuer 1991). I use the more general mean- ing.

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
“ANYTHING BUT.. .”: SYMBOLIC EXCLUSION AND MUSICAL DISLIKES 889

includes a set of questions about culture, in- cluding musical tastes as well as leisure ac- tivities and values (Davis and Smith 1993; Marsden and Swingle 1994). These new data make information on musical dislikes avail- able for the first time. Like other surveys of taste and participation in the arts, this survey presented respondents with a list of musical categories, but rather than having them choose their favorite or mark all they like, the GSS asked all 1,606 respondents to evaluate each of 18 music genres on a five-point Likert scale ranging from “like very much” to “dis- like very much” (see Table 3). Using these data, I derive a measure of musical exclusive- ness by counting the “dislike” and “dislike very much” responses given by each respon- dent. This method highlights the interesting responses and avoids the flattening effect of averages. If a respondent gave pop music a 1 and classical music a 5, the average score would be 3-the same as a response of 3 for each genre. By counting only negative re- sponses, the exclusiveness of the score is pre- served. “Don’t know” responses are treated as missing and those respondents are elimi- nated from the analysis, leaving 912 valid cases. The exclusiveness scale has a mean of 5.78 and a standard deviation of 3.76, with a possible range of 0 to 18.

To test the effects of racism on musical taste, I divide the exclusiveness scale into two subscales to separate the genres that might be perceived as “belonging to” or rep- resenting racial or ethnic minorities. One scale is a count of the “dislike” responses to the six genres that are disproportionately liked by Black or Hispanic respondents with a significance level of p < .001.4 The remain- ing 12 genres form the other scale. In no case was the percentage of non-Hispanic Whites who reported liking a particular genre less than 60 percent. For example, only 26 per- cent of rap music “fans” are Black (com- pared to 11 percent in the sample) and 9 per- cent are Hispanic American (compared to 4

percent of the sample). Eighteen percent of gospel music fans are Black, while that num- ber is 16 percent for each of reggae, blues, and jazz. Thirteen percent of Latin music fans are Hispanic.

Intergroup Affect

The political intolerance scale is based on a count of intolerant responses to 15 dichoto- mous questions on the respondent’s willing- ness to allow a person from five different groups to (1) “make a speech in your com- munity,” (2) “teach in a college or univer- sity,” and (3) have a book in “your public li- brary.” The hypothetical person in each of these five groups is described one who (a) is “against churches and religion,” (b) “believes that Blacks are genetically inferior,” (c) “ad- mits he is a Communist,” (d) “advocates do- ing away with elections and letting the mili- tary run the country,” and (e) “admits he is a homosexual.” These questions are based on Stouffer’s (1955) scale, but they are asked of only two-thirds of the sample. The scale ranges from 0 to 15, has a mean of 5.24, a standard deviation of 4.72, and an alpha (measuring reliability) of .9163.

I create the racism scale by collecting all the 1993 GSS questions about racial atti- tudes. Next, I remove questions with ex- tremely small variances and compose a list that maximizes valid responses. (To form a scale, all items must be asked of the same set of respondents, yet most of the racism ques- tions are asked only of certain portions of the sample.) Finally, the responses (all dichoto- mous) are coded in the same direction, and factor analysis (available on request from the author) suggested the removal of one item. The remaining questions are: (1) “Would you yourself have any objection to sending your children to a school where more than half of the children are Black?” (2) “In general, do you favor or oppose the busing of Black and White school children from one school dis- trict to another?” (3) “On average Blacks have worse jobs, income, and housing than White people. Do you think these differences are . . mainly due to discrimination?” (4) “… because most Blacks don’t have the chance for education that it takes to rise out of poverty?” (5) “. . . because most Blacks just don’t have the motivation or will power

4 While I use the term “African American” in my theoretical discussion, 34 of the 137 respon- dents coded as “Black” based on skin color list a primary ethnic identity other than African or American. (“Black” is not a coded option on this question.) Most of these report either a West In- dian or Native American Indian family origin.

This content downloaded from 128.200.138.25 on Wed, 29 Oct 2014 11:54:54 AM All use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp
890 AMERICAN SOCIOLOGICAL REVIEW

to pull themselves up out of poverty?” The racism scale has a range of 0 to 5, a mean of 2.65, a standard deviation of 1.56, and an al- pha of .54.

ANALYSIS