Need discussion wrote

Need discussion wrote
Social Class, Cultural Repertoires, and Popular Culture: The Case of Film Author(s): Lisa A. Barnett and Michael Patrick Allen Source: Sociological Forum , Mar., 2000 , Vol. 15, No. 1 (Mar., 2000), pp. 145-163 Published by: Springer Stable URL: https://www.jstor.org/stable/3070340 REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/3070340?seq=1&cid=pdf- reference#references_tab_contents You may need to log in to JSTOR to access the linked references. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected] Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms Wiley and Springer are collaborating with JSTOR to digitize, preserve and extend access to Sociological Forum This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Sociological Forum, Vol. 15, No. 1, 2000 Social Class, Cultural Repertoires, and Popular Culture: The Case of Film Lisa A. Barnett1’2 and Michael Patrick Allen3 This research examines two different conceptions of the relationship between social class and familiarity with popular culture in the United States. Specifi- cally, it focuses on differences between members of the upper-middle class and members of the lower-middle class in terms of theirfilm viewingpractices. The data for this analysis was obtained from a survey of 364 individuals randomly selected from two neighborhoods in a medium-sized city, one predominantly upper-middle class and the other predominantly lower-middle class. Members of the upper-middle class view more “art” films, as well as more “classic” films and “blockbuster”films, than members of the lower-middle class. These differences are largely attributable to the fact that members of the upper-middle class view morefilms both in theaters and on videocassettes than members of the lower-middle class. Moreover, these differences are reduced, but not en- tirely eliminated, by the fact that members of the lower-middle class view more films on television than members of the upper-middle class. Finally, these dif- ferences in thefilm-viewingpractices of the members of these two social classes, as identified by their neighborhood of residence, obtain even controlling for a series of demographic and socioeconomic background variables. KEY WORDS: culture; film; social class; cultural repertoires. STATEMENT OF THE PROBLEM Over the past two decades, sociologists have sought to explicate the relationship between culture and social class (Gans, 1974yf 0 X F K R I W K L s The authors would like to thank the anonymous reviewers for their helpful comments on earlier versions of this manuscript. 1Department of Sociology, 1120 First Ave. N. E., Coe College, Cedar Rapids, Iowa 52402. 2To whom correspondence should be addressed. 3Department of Sociology, Washington State University, Pullman, Washington 99164. 145 0884-8971/00/0300-0145$18.00/0 ? 2000 Plenum Publishing Corporation This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen work has taken as its point of departure the concept of “cultural capital” first proposed by Bourdieu (1973yf , Q S U R S R V L Q J W K L V F R Q F H S W % R X U G L H u argued that cultural capital, defined originally in terms of a familiarity with and an appreciation of legitimate or “high” culture, contributes to the maintenance of boundaries between the members of different social classes. Specifically, he claimed that a familiarity with “high” culture serves as a basis for distinguishing members of the dominant class from members of subordinate classes. Although the introduction of the concept of cultural capital has led to a recognition of the importance of culture in the process of social stratification, recent empirical research (Lamont, 1992; Peterson, 1992yf K D V F D O O H G L Q W R T X H V W L R Q W K H D S S O L F D E L O L W R I W K L V W K H R U D W O H D V W L Q L W s initial form, to the United States. In particular, this research has suggested that familiarity with various forms of “popular” culture may also serve as a basis for social exclusion in the United States. Indeed, recent research (Lamont, 1992; Peterson, 1992yf V X J J H V W V W K D W P H P E H U V R I W K H X S S H U P L G G O e class possess more extensive “cultural repertoires,” defined in terms of a familiarity with different forms of both “high” and “popular” culture, than members of the lower-middle class. Although the research to date on the importance of popular culture in the cultural repertoires of Americans is suggestive, it is far from conclusive. Moreover, recent research has ignored one of the most significant forms of popular culture: film. This neglect of film as a component of cultural capital in the United States is somewhat surprising, especially in view of the fact that Bourdieu presented empirical research to suggest that viewing art films, as opposed to mainstream films, was associated with different levels of education in France (1984:271yf 8 V L Q J R U L J L Q D O V X U Y H G D W D W K L s research seeks to explicate the importance of film viewing practices in the cultural repertoires of upper-middle class and lower-middle class Ameri- cans. In so doing, it investigates the relationship of the concept of cultural repertoires, defined in terms of a familiarity with both high and popular culture, to the concept of cultural capital, defined in terms of familiarity with high culture. Indeed, this research argues that these two conceptions of the relationship between social class and cultural practices can be seen as complementary, rather than contradictory, at least within the Ameri- can context. CULTURAL CAPITAL AND CULTURAL REPERTOIRES The first theory that sought to explicate the relationship between social class and culture in any detail was developed in France by Bourdieu (1973, 1984yf $ F F R U G L Q J W R W K L V S H U V S H F W L Y H F X O W X U H L V L P S O L F D W H G Z L W K V R F L D O F O D V s 146 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture inasmuch as differences in cultural practices contribute to the maintenance of social boundaries between social classes. At the heart of this theory is the distinction between “high” or “legitimate” culture and “mass” or “popular” culture. Moreover, this theory asserts that members of the domi- nant class and the institutions they control possess the power to “conse- crate” certain cultural practices, thus certifying them as “legitimate.” These cultural practices then serve to distinguish and thereby segregate members of the dominant class from the members of the subordinate classes. At the core of this theory is the concept of “cultural capital.” The precise definition of this concept has evolved somewhat over time. However, Lamont and Lareau (1988:156yf F R Q F O X G H W K D W F X O W X U D O F D S L W D O F D Q E H V W E H V H H Q L Q W H U P s of “widely shared, high status cultural signals (attitudes, preferences, formal knowledge, behaviors, goods, and credentialsyf X V H G I R U V R F L D O D Q G F X O W X U D l exclusion.” Thus, cultural capital is a function of cultural competency, defined as the mastery of valued cultural codes. According to this theory, one cannot fully appreciate any cultural object, be it a painting or a sym- phony, unless one is competent in the use of the symbolic codes appropriate to that object. As Bourdieu (1984:2yf D V V H U W V D Z R U N R I D U W K D V P H D Q L Q g and interest only for someone who possesses the cultural competence, that is, the code, into which it is encoded.” Based on the research presented by Bourdieu (1984yf L W V H H P V U H D G L O y apparent that high culture constitutes an essential component of cultural capital in France. Moreover, the earliest applications of this theory to the United States indicate that cultural capital, in the form of high culture, is strongly implicated in the process of social stratification. For example, DiMaggio and others (DiMaggio, 1982; DiMaggio and Mohr, 1985yf G H P R Q – strate the relevance of cultural capital to both educational attainment and marital selection. Specifically, DiMaggio (1982yf P H D V X U H V F X O W X U D O F D S L W D l in terms of the involvement of students in art, literature, and music. Al- though he acknowledges that these are not necessarily ideal indicators of cultural capital, DiMaggio (1982:191yf D V V H U W V W K D W K L J K F X O W X U D O P H D V X U H s represent the best alternative for several reasons,” including the relative popularity of these cultural forms and their relevance to a “common cultural currency among American elites.” However, more recent research on cultural capital has raised questions about the centrality of high culture to the creation and maintenance of class boundaries in the United States. In particular, several studies (Lamont, 1992; Peterson and Kern, 1996yf V X J J H V W W K D W K L J K F X O W X U H F R P S U L V H V R Q O y one aspect of the broader cultural repertoires of Americans. In particular, these studies point to the importance of popular culture as a basis for class distinctions and social exclusion in the United States. For example, Lamont (1992yf I R X Q G G L V W L Q F W O G L I I H U H Q W R U L H Q W D W L R Q V W R S R S X O D U F X O W X U H D P R Q g 147 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen upper-middle class men in France and the United States. As she (Lamont, 1992:114yf R E V H U Y H V P L Q W H U Y L H Z V D Q G R W K H U V W X G L H V V X J J H V W W K D W $ P H U L – cans have broader cultural repertoires than the French and are less exclusive in their attitude toward mainstream culture.” More specifically, Lamont (1992:238yf D V V H U W V W K D W $ P H U L F D Q V R I W H Q V W U H V V W K H E U R D G H Q L Q J R I K R U L ] R Q s through culture” whereas the French typically stress “refinement.” Other researchers have reached similar conclusions. For example, Press (1994:225yf asserts that Americans use familiarity with popular culture to “gain status and make valuable contacts and alliances.” Specifically, she (Press, 1994:238yf F O D L P V W K D W I D F L O L W Z L W K O R Z R U P D V V F X O W X U D O I R U P V I X Q F W L R Q s in important ways as markers of status, class, prestige, and even social mobility in our culture.” These arguments are consistent with the research of Peterson (1992yf R Q F X O W X U D O R P Q L Y R U H V ) R U H [ D P S O H 3 H W H U V R Q D Q d Kern (1996yf S U R Y L G H H Y L G H Q F H W K D W P H P E H U V R I W K H X S S H U P L G G O H F O D V V L n the United States have become less “snobbish” and more “omnivorous” in their musical tastes in recent years. According to their research (Peterson and Kern, 1996:904yf R Q H R I W K H F X O W X U D O F K D U D F W H U L V W L F V R I W K H X S S H U P L G G O e class is “an openness to appreciating everything” from Broadway musicals to country music. The observation that upper-middle class Americans possess relatively broad “cultural repertoires” that include various forms of popular culture, and that these repertoires serve as a basis for social exclusion, can be explicated in terms of cultural competencies and cultural codes. Although Bourdieu explicitly recognizes that certain cultural competencies are re- quired to decode high culture texts, he largely ignores the possibility that different but often equally complex cultural competencies may be required to decode popular culture texts. However, Fiske argues (1987:19yf W K D t “popular cultural capital requires a set of cultural competencies.” He notes, for example, that fans of soap operas are very familiar with the “cultural codes” specific to this particular television genre. In a similar vein, Blewitt (1993yf D U J X H V W K D W W K H F X O W X U D O F R P S H W H Q F U H T X L U H G W R G H F R G H D P D L Q V W U H D m Hollywood film is often as complex as that required to decode an art film. Although the cultural repertoire and the cultural capital conceptions of the relationship between social class and popular culture may seem at odds with one another, it can be argued that these two conceptions actually complement, rather than contradict, each other. This is especially true if cultural capital, defined as the mastery of valued cultural codes, is consid- ered as a generic concept that takes different forms in different contexts. Accordingly, cultural capital in France may well take the form of a specific familiarity with the comparatively narrow range of cultural codes associated with high culture. Conversely, cultural capital in the United States can take the form of a cultural repertoire, consisting of a general familiarity with a 148 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture fairly broad range of cultural codes that encompass both high culture and popular culture. This interpretation is entirely consistent with both the early research (DiMaggio, 1982; DiMaggio and Mohr, 1985yf W K D W I R X Q d class differences in cultural capital, defined in terms of high culture, and with more recent research (Lamont, 1992; Peterson and Kern, 1996yf W K D t found class differences in cultural repertories that encompass both high culture and popular culture. This is the point of the observation by Gripsrud (1989:199yf W K D W W K H G R X E O H D F F H V V W R W K H F R G H V D Q G S U D F W L F H V R I E R W K K L J h and low culture is a class privilege.” Despite their seminal insights, the pioneer studies of the appropriation of popular culture by members of the American middle class have their limitations. For example, Lamont (1992yf H [ D P L Q H V W K H G L I I H U H Q F H V E H W Z H H n upper-middle class men in France and the United States in terms of their familiarity with both high culture and popular culture, but neglects to provide any explicit comparisons between upper-middle class and lower- middle class individuals in either nation. Moreover, her analysis is based on a limited number of qualitative interviews. Similarly, Peterson and Kern (1996yf S U R Y L G H D P R U H T X D Q W L W D W L Y H D Q D O V L V R I W K H F R Q F H S W R I F X O W X U D O U H S H U – toires, but their results are limited by their reliance on a secondary analysis of existing survey data. In particular, they are limited to studying stated preferences for broadly defined musical genres. It is interesting to note that no one to date has examined the relation- ship between social class, cultural repertoires, and film-viewing practices. This omission is especially glaring in view of the fact that film is arguably one of the dominant forms of popular culture in the United States. As one celebrated American novelist, essayist, and screenwriter (Vidal, 1992:2yf puts it, “movies are the lingua franca of the twentieth century.” Indeed, film- viewing practices provide an excellent arena for examining the relationship between social class, cultural repertoires, and popular culture. Films are generally considered to be a form of popular culture. However, they are also considered to be an important art form. As a matter of fact, film is a medium whose content can span the spectrum from introspective art to escapist entertainment (Monaco, 1981:7-20yf $ V D J H Q H U D O U X O H % R X U G L H u is skeptical of the artistic merits of certain “art” films because they “give the impression of bringing legitimate culture within the reach of all, by combining two normally exclusive characteristics, immediate accessibility and the outward signs of cultural legitimacy.” Nevertheless, he examines the class differences in the audiences for different types of films. In particular, Bourdieu concludes (1984:271yf W K D W W H D F K H U V D Q G S U R I H V V L R Q D O V Z K R S R V V H V s more cultural capital than economic capital, generally prefer art films “that demand a large cultural investment” whereas business executives, who possess more economic capital than cultural capital, tend to prefer main- 149 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen stream films, defined as those “spectacular feature films, overtly designed to entertain.” This study compares members of the upper-middle class and members of the lower-middle class in terms of their film-viewing practices. Many films, especially “blockbuster” films, appeal to a very large and relatively undifferentiated audience. However, other types of film, especially art films, appeal primarily to a more highly educated audience that possesses the cultural competencies required to decode these films. The original formula- tion of cultural capital in terms of high culture implies that members of the upper-middle class should demonstrate more familiarity with art films than members of the lower-middle class. Conversely, the more recent for- mulation of cultural capital in terms of cultural repertoires implies that members of the upper-middle class should demonstrate more familiarity with all types of films, including “classic” films and blockbuster films as well as art films, than members of the lower-middle class. Of course, any differences in film-viewing practices may be attributable, at least in part, to differences in income, inasmuch as viewing a film in a theater requires an expenditure that is not entirely inconsequential to many members of the lower-middle class. Therefore, we hypothesize that members of the upper-middle class view films in theaters more often than members of the lower-middle class. However, the availability of films of all types on videocassettes and both network and cable television probably reduces somewhat the differences in film-viewing practices between the members of different social classes. At the time of this survey, 86yb R I D O O K R P H V L n the United States had videocassette recorders and 64yb R I D O O K R P H V V X E – scribed to cable television. As a result, we hypothesize that differences in film-viewing practices between members of the upper-middle class and members of the lower-middle class are less pronounced with respect to viewing films on video or on television than they are with respect to viewing films in theaters. Nevertheless, we hypothesize that the differences in the film-viewing practices of upper-middle class individuals and lower-middle class individuals persist across all modes of film viewing. SAMPLE AND MEASURES The theory we have proposed asserts that members of the upper- middle class and the lower-middle class differ in terms of their film-viewing practices. In order to test this theory, we must identify individuals who are members of these social classes. Bourdieu (1984:101yf G H I L Q H V R E M H F W L Y e class” in terms of “homogeneous conditions of existence imposing homoge- neous conditionings and producing homogeneous systems of dispositions 150 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture capable of generating similar practices.” At the same time, he (Bourdieu, 1984:103yf U H M H F W V D Q D W W H P S W W R L G H Q W L I W K H L Q G H S H Q G H Q W H I I H F W V R I G L I I H U – ent dimensions of social class, such as income and education, on cultural practices. Instead, the significance of social class (Bourdieu, 1984:106yf O L H s in “the structure of relations, between all the pertinent properties . .. and to the effects they exert on practices.” Indeed, at one point, Bourdieu (1984:124yf V X J J H V W V W K D W W R D F F R X Q W P R U H I X O O I R U W K H G L I I H U H Q F H V L Q O L I H – style between the different class fractions-especially as regards culture- one would have to take account of their distribution in a socially ranked geographical space.” In light of these suggestions, it can be argued that neighborhood of residence provides an appropriate measure of social class for this research. As a result of residential segregation, neighborhoods within large urban areas tend to be highly homogeneous with respect to the income, education, and occupational prestige of their residents (White, 1987yf 2 I F R X U V H E R W h economic considerations, such as the distance to work and the affordability of housing, and structural constraints, such as the availability of financing, are important factors in the decision to reside in a particular neighborhood (Kain and Quigley, 1975yf 1 H Y H U W K H O H V V S X U H O V R F L D O F R Q V L G H U D W L R Q V V X F h as the status associated with a certain neighborhood or the aesthetic charac- teristics of the area, are also implicated in the decision to reside in a particular neighborhood. For example, Butler and his associates (1964:141yf find that status aspirations often affect the choice of a particular residential neighborhood. As a result, neighborhood of residence probably serves as an unobtrusive measure of subjective class identification. Indeed, some of the earliest empirical research in the field of social stratification identified both “dwelling place” and “house type” as primary dimensions of social class in American society (Warner, 1949yf ) R U W K H V H U H D V R Q V L W F D Q E e argued that neighborhood provides a comprehensive measure of social class inasmuch as it corresponds to the “homogeneous conditions of existence” that are most likely to yield “homogeneous systems of dispositions” (Bour- dieu, 1984:101yf . In order to obtain a sample of individuals from the upper-middle class and a comparable sample of individuals from the lower-middle class, we identified two neighborhoods within a medium-sized city in the Northwest. These neighborhoods were identified using census tract information be- cause census tracts are small geographical areas that are relatively homoge- neous in terms of their demographic characteristics. Indeed, census tracts were designed from the outset to correspond closely to the concept of a “natural area.” Out of the almost one hundred census tracts in this city, we identified two that are similar with respect to certain demographic characteristics but are quite dissimilar with respect to relevant socioeco- 151 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen nomic characteristics. Both neighborhoods are residential areas that consist primarily of single-family dwellings with only a limited number of apartment buildings. The Lower Park neighborhood, which is inhabited primarily by lower-middle class families, is located near a major urban arterial and a business district that includes a small shopping mall. It is comprised of relatively modest homes, most of which were constructed in the 1950s. By contrast, the Upper Hill neighborhood, which is inhabited primarily by upper-middle class families, is located at some distance from any major arterials or business districts. It is comprised of comparatively large homes, some of them built in the 1930s and others of them built in the 1970s. It might be added that the boundaries of these two neighborhoods and their social characteristics are clearly understood by most of the residents of this city (Suttles, 1972yf . The demographic and socioeconomic characteristics of these two cen- sus tracts and the city at large are presented in Table I. To begin with, it is apparent that residents of Upper Hill and Lower Park are generally similar to the residents of the city at large in terms of their age and race. Both of these neighborhoods are predominantly white, as is the city at large. However, the residents of Lower Park are less educated and earn somewhat less than other residents of the city. Conversely, the residents of Upper Hill are much more educated and earn substantially more than other residents of the city. Moreover, the houses in the Lower Park are somewhat less expensive than those in the city at large whereas the houses in Upper Hill are much more expensive. Table I also presents coefficients of variability for several variables. These coefficients are simply standard deviations that have been divided by their respective means for purposes of comparison (Blalock, 1960:73-74yf , Q J H Q H U D O W K H / R Z H U 3 D U N D Q G 8 S S H r Table I. Demographic Characteristics and Coefficients of Variability for Sample Neighbor- hoods and Total City, 1990″ Characteristic and Lower Upper Total variability Park Hill city Demographic characteristics Median age (yearsyf 4 Percent white 95.8 97.4 96.5 Percent college graduate 8.9 54.6 21.0 Median family income (dollarsyf 8 Median house value (dollarsyf 0 Coefficients of variability Age 0.616 0.608 0.642 Education 0.167 0.169 0.196 Family income 0.764 0.767 0.860 House value 0.333 0.665 0.615 aCoefficients of variability are standard deviations divided by their means. 152 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture Hill neighborhoods are more homogeneous than the city at large in terms of their demographic and socioeconomic characteristics. The sole exception involves the house values in Upper Hill, which contains a mix of both older homes and modern homes. Although specific distributions are not presented, it might be noted that most of the residents in Lower Park are engaged in manual-clerical occupations whereas most of the residents of Upper Hill are engaged in managerial-professional occupations. A short, self-administered mail questionnaire, accompanied by a stamped return envelope, was sent to 420 individuals randomly drawn from each of these two census tracts. These individuals were identified using census maps of the Lower Park and Upper Hill neighborhoods and a reverse telephone directory that listed the names of the residents by street address. Inasmuch as this reverse directory included the names of all adult residents at each address, we attempted to balance the gender composition of the sample by alternating between male and female residents. Those members of the sample who were no longer at the address provided by this reverse directory were replaced by neighbors. The response rate to the first mailing was 25yb $ V H F R Q G P D L O L Q J Z D V V H Q W W R Q R Q U H V S R Q G H Q W V D I H Z Z H H N V O D W H U , accompanied by an offer of nominal compensation. As a result of this second mailing, we were able to achieve a 43yb U H V S R Q V H U D W H T X H V W L R Q – naires from Lower Park and 199 questionnaires from Upper Hill. In order to assess the degree of bias in these samples due to nonresponse, a 10yb sample of nonrespondents from each neighborhood was surveyed by tele- phone several weeks after the second mailing in order to ascertain their age, gender, and education. In general, respondents had slightly more edu- cation than nonrespondents in both samples. Overall, we concluded that these minor differences in the characteristics of the respondents and nonre- spondents are not likely introduce any significant bias into our analyses. In order to maximize the response rate, the mail questionnaire sent to each member of the sample was limited in length to four pages. It included a series of both forced-choice and open-ended questions about the television and film-viewing practices of respondents as well as their social, economic, and demographic characteristics. The family income of the respondent was measured by a six-point scale that divided annual family income into six equal categories ranging from less than $20,000 to $100,000 or more. Similarly, the education of the respondent was measured by a six- point scale that divided educational achievement into six categories ranging from some high school education to graduate school education. Finally, the occupational prestige of the respondent was measured by a eight- point scale based on the seven category schema developed by Erikson and Goldthorpe (1992yf + R Z H Y H U W K L V V F K H P D Z D V P R G L I L H G E D G G L Q J D Q H L J K W h category for individuals who were either retired or unemployed. 153 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen In order to measure film-viewing practices, respondents were asked whether or not they had seen each of sixteen films and, if so, whether they had seen that film in a theater, on videocassette, or on television. In order to aid in identification, each film was listed by title, year of release, and the name of its principal star. These sixteen films were selected from among the most popular and critically acclaimed films produced since 1940 (Ray, 1985; Sklar, 1994yf $ W O H D V W W Z R I L O P V Z H U H F K R V H Q I U R P H D F K G H F D G H I U R m 1940 to 1990 and four films were chosen from the first five years of this decade. In general, these films were selected to represent the diversity of films produced during this period. However, particular films were also selected to represent one of three basic types of films: classic films, art films, and blockbuster films. The validity of employing these particular films to measure each of these three types of film was examined using exploratory factor analysis. Specifically, the correlations among the frequencies with which respondents had viewed each of these sixteen films were analyzed using principal component factor analysis. The results of this analysis, after an orthogonal rotation of the three principal factors, are presented in Table II. This analysis reveals that most of these films are associated with one of three different factors. Indeed, for each of these three factors, it was possible to identify four films that were highly correlated with this factor but not Table II. Principal Component Factor Analysis with Varimax Rotation of the Correlations Among Film-Viewing Practices of Respondents with Respect to 16 Films Film (yearyf ) D F W R U ) D F W R U ) D F W R U 3 Classic films High Noon (1950yf 6 Casablanca (1942yf 5 Citizen Kane (1941yf 6 Lawrence of Arabia (1962yf 2 Art films The Piano (1993yf 7 A Room with a View (1985yf 6 Schindler’s List (1994yf 3 The Last Emperor (1987yf 3 Blockbuster films Jurassic Park (1994yf 8 Born on the Fourth of July (1986yf 9 The Godfather (1972yf 5 Pulp Fiction (1994yf 2 Other films Psycho (1960yf 9 The Graduate (1967yf 5 Chinatown (1974yf 2 Rebel without a Cause (1955yf 9 Eigenvalue 5.52 1.74 1.19 Variance explained 34.5 10.9 7.5 154 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture with the other two factors. These results confirm that individuals who viewed a film of one type were more likely to have viewed other films of that same type than they were to have viewed films of another type. The first factor, comprised of four films, including High Noon, corresponds to films that can be identified as classic films. Similarly, the second factor, comprised of four films including The Piano, corresponds to films that can be identified as art films. Finally, the third factor, comprised of four films including Jurassic Park, corresponds to films that can be identified as block- buster films. The remaining four films were excluded from further analysis because they were highly correlated with both the classic and block- buster factors. The results of this factor analysis were used to construct separate scales to measure each of these three types of film viewing practices. Specifically, three separate four-item summative scales, with scores ranging from zero to four, were constructed to measure these different types of film-viewing practices. For example, the scale measuring the practice of viewing art films was constructed on the basis of whether not or a respondent had seen each of four art films: The Piano, A Room with a View, The Last Emperor, and Schindler’s List. Although these four films might not qualify as art films by the criteria employed by film scholars, they are clearly films that sacrificed a measure of commercial appeal in the pursuit of social, political, and artistic concerns. Similar four-item summative scales were constructed to measure viewing practices with respect to classic films and blockbuster films, respectively. These summative scales are highly reliable in terms of their internal consistency with alpha coefficients ranging from 0.759 for the classic film scale to 0.620 for the blockbuster film scale. In addition, separate summative scales were constructed to measure whether or not respondents saw each of these same films in theaters, on video, or on television. Conse- quently, it is possible to analyze film-viewing practices both in terms of the types of films viewed and in terms of the manner in which each type of film was viewed. Finally, it is possible to combine these three summative scales into a single twelve-item summative scale, with scores ranging from zero to twelve, to measure overall film-viewing practices. In each case, the scores on these scales can be interpreted unambiguously as the absolute number of films of each type viewed by each respondent. RESULTS If the theory we have proposed is correct, we expect to find class differences in the cultural repertoires of individuals with respect to their film-viewing practices. In particular, we hypothesize that members of the 155 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen upper-middle class have broader cultural repertoires and thus have viewed more films in general than members of the lower-middle class. In order to evaluate this hypothesis, we examine differences in the number of films viewed by the members of our two samples. The mean number of films viewed by the residents of both Lower Park and Upper Hill are presented in the first section of Table III. These results indicate that upper-middle class individuals have viewed significantly more films of all types than lower- middle individuals. For example, the residents of Lower Park had viewed only 5.43 of the twelve films in our index of all films whereas residents of Upper Hill had viewed 7.09 of these films. This difference implies that residents of Upper Hill have viewed 30.6yb P R U H I L O P V L Q J H Q H U D O W K D n residents of Lower Park. The statistical significance of the difference be- tween these means is established using a simple regression model with a dummy variable for neighborhood to perform one-way analysis of variance. Similarly, we hypothesize that upper-middle class individuals have viewed more art films than lower-middle class individuals. The mean num- Table III. Differences in Average Film Viewing Practices by Type of Film and Mode of Viewinga Type of film and Lower Upper Mean mode of viewing Park Hill difference All films 5.43 7.09 1.66h Type of film Art 1.03 1.90 0.87′ Classic 2.54 2.91 0.37′ Blockbuster 1.88 2.33 0.45′ Mode of viewing Theater 2.09 3.72 1.63′ Video 1.48 2.17 0.69d Television 2.34 2.28 -0.07 Art films Theater 0.30 1.04 0.74b Theater + video 0.78 1.80 1.02″ Theater + video + television 1.03 1.90 0.87′ Classic films Theater 1.06 1.31 0.25 Theater + video 1.45 1.88 0.43′ Theater + video + television 2.54 2.91 0.37c Blockbuster films Theater 0.79 1.42 0.63b Theater + video 1.44 2.26 0.82h Theater + video + television 1.88 2.33 0.45h “The mean for all films does not include films seen more than once across different modes of viewing. bp < 0.001. Cp < 0.01. dp < 0.05. 156 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture ber of films of each type viewed by the residents of both Lower Park and Upper Hill are presented in the second section of Table III. There are significant differences in the film-viewing practices of lower-middle class individuals and upper-middle class individuals with respect to each of the three types of film. As expected, residents of Upper Hill have viewed more art films than residents of Lower Park. For example, the residents of Upper Hill had viewed 1.90 of the four films in our index of art films whereas residents of Lower Park had viewed only 1.03 of these films. Indeed, the difference of 0.87 films implies that residents of Upper Hill have viewed 84.5yb P R U H D U W I L O P V W K D Q U H V L G H Q W V R I / R Z H U 3 D U N $ O W K R X J K W K H D U H V W L O l statistically significant, the differences between the residents of Upper Hill and Lower Park are somewhat less pronounced with respect to classic films. This pattern may reflect the fact that classic films are generally more accessible than art films. We also expect to find social class differences in the mode of viewing films. Specifically, we hypothesize that upper-middle class individuals have viewed more films in theaters than lower-middle class individuals but that lower-middle class individuals have viewed more films on videocassette and television than upper-middle class individuals. The means for both Lower Park and Upper Hill residents in terms of the number of films viewed in theaters, on videocassette, and on television are presented in the third section of Table III. These results indicate that members of the upper- middle class have viewed significantly more films in theaters than members of the lower-middle class. For example, residents of Upper Hill have viewed 3.72 of the twelve films in our index of all films in theaters whereas residents of Lower Park have viewed only 2.09 of these films in theaters. This differ- ence of 1.63 films implies that residents of Upper Hill have viewed 77.9yb more films in theaters than residents of Lower Park. However, contrary to our expectations, residents of Upper Hill have also viewed more films on videocassette than residents of Lower Park. For example, residents of Upper Hill had viewed 2.17 of the twelve films in our index of all films on videocassette whereas residents of Lower Park have viewed only 1.48 of these films on videocassette. This difference of 0.69 films implies that resi- dents of Upper Hill have viewed 46.6yb P R U H I L O P V R Q Y L G H R F D V V H W W H V W K D n residents of Lower Park. Moreover, there is no significant difference be- tween members of these two neighborhoods with respect to viewing the twelve films in our index of all films on television. Finally, we hypothesize that the differences in film viewing practices between the members of these two social classes with respect to different types of films are mitigated by the ready availability of these different types of film on videocassettes and television. The results presented in the last three sections of Table III enable us to examine the effect of film viewing 157 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen on both videocassette and television in reducing the class differences in film-viewing practices. In particular, the differences between upper-middle class individuals and lower-middle class individuals, in terms of viewing art films, are actually exacerbated by the inclusion of films viewed on videocassette to those already viewed in theaters. For example, residents of Upper Hill have viewed 0.74 more art films in theaters than residents of Lower Park. When the number of art films viewed in theaters is combined with the number of art films viewed on videocassette, residents of Upper Hill have viewed 1.02 more art films than residents of Lower Park. However, the class differences in film-viewing practices with respect to art films are mitigated somewhat with the inclusion of films viewed on television to those already viewed in theaters and on videocassette. This general pattern obtains for both blockbuster films and classic films as well. With the sole exception of classic films, residents of Upper Hill have viewed more films of all types in theaters than members of Lower Park. In short, the differences in the film-viewing practices of the lower-middle class individuals and upper- middle class individuals are exacerbated, rather than mitigated, by the inclusion of films viewed on videocassette. However, these class differences in film-viewing practices are mitigated somewhat, but not entirely elimi- nated, by the inclusion of films viewed on television. Last but not least, it is possible to determine whether these differences in film-viewing practices obtained using neighborhood as a measure of social class persist after controlling for a series of background variables. These background variables include socioeconomic characteristics associ- ated with social class, such as education, income, and occupation, and other demographic characteristics, such as age and gender. As expected, the simple correlations between these socioeconomic variables and each of the film-viewing practices within each sample are generally positive. Multiple regression analyses of the effects of social class, as measured by neighbor- hood of residence, and the other background variables on film viewing practices by type of film are presented in Table IV. In these analyses, the effects of social class are examined using a dummy variable for those individuals who reside on Upper Hill rather than in Lower Park. These analyses reveal that social class, as measured by neighborhood, remains a significant predictor of film-viewing practices, even controlling for other background variables, for all types of film except classic films. In short, even though neighborhood of residence is highly correlated with such socio- economic variables as education (0.473yf L Q F R P H f, and occupation (0.238yf L W F O H D U O K D V L Q G H S H Q G H Q W H I I H F W V R Q F H U W D L Q I L O P Y L H Z L Q J S U D F W L F H V . Indeed, the coefficients of determination presented at the bottom of Table IV indicate that, especially with respect to "all" films and art films, the neighborhood variable alone explains a large proportion of the explained 158 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture Table IV. Multiple Regression Analyses of the Effects of Social Class (Neighbor- hoodyf D Q G % D F N J U R X Q G 9 D U L D E O H V R Q ' L I I H U H Q W 7 S H V R I ) L O P 9 L H Z L Q J 3 U D F W L F H s Independent All Art Classic Blockbuster variables films films films films Upper Hill 1.212a 0.628a 0.306 0.364b (0.395yf f (0.180yf f Income 0.185 0.115a 0.024 0.031 (0.122yf f (0.055yf f Education -0.008 0.001 0.054 -0.078 (0.141yf f (0.064yf f Occupation 0.119b 0.054 0.047 0.014 (0.069yf f "p < 0.01. bp < 0.05. variance in film viewing. These results clearly indicate that neighborhood of residence captures some dimension of social class relevant to the forma- tion of cultural repertoires that is not captured by the standard socioeco- nomic variables of income, education, and occupation. At the same time, these results indicate that other socioeconomic and demographic variables have significant effects on particular film-viewing practices even controlling for social class, as measured by neighborhood of residence. Indeed, the coefficients of determination presented at the bottom of Table IV indicate that, especially with respect to classic films and blockbuster films, socioeconomic and demographic variables explain a large proportion of the explained variance in film viewing. To begin with, occupation has a significant, albeit a somewhat modest, positive effect on the number of all films viewed controlling for the effect of social class. Similarly, income and being female have significant positive effects on the number of art films viewed. Conversely, age and being female have signifi- cant negative effects on the number of blockbuster films viewed. In both cases, the effects of being female are relatively modest. Finally, age has a significant positive effect on the number of "classic" films viewed. However, as we shall see, the effects of age on viewing both blockbuster and classic films are probably attributable to the effects of age on mode of film viewing. Multiple regression analyses of the effects of neighborhood of resi- dence and the other background variables on film viewing practices by mode of viewing are presented in Table V. Once again, these analyses reveal that social class, as measured by neighborhood of residence, is a significant predictor of viewing films in theaters, even controlling for other background variables. Indeed, the coefficients of determination presented at the bottom of Table V indicate that, especially with respect to films viewed in theaters, the neighborhood variable alone explains a substantial 159 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen Table V. Multiple Regression Analyses of the Effects of Social Class (Neighborhoodyf D Q d Background Variables on Different Modes of Film-Viewing Practicesa Independent All Viewed in Viewed on Viewed on variables films theaters video television Upper Hill 1.212b 1.428b 0.434 -0.059 (0.395yf f (0.276yf f Income 0.185 0.294b -0.030 -0.124 (0.122yf f (0.085yf f Education -0.008 -0.138 0.051 -0.072 (0.141yf f (0.098yf f Occupation 0.119c 0.163c 0.037 -0.017 (0.069yf f (0.048yf f Age -0.006 0.056b -0.044b -0.011 (0.011yf f (0.008yf f Female -0.141 0.580C 0.120 -1.075b (0.357yf f (0.250yf f Constant 5.176b -2.129' 3.741 3.771h (0.985yf f (0.689yf f R2 (Upper Hillyf 1 R2 (full modelyf 7 F ratio 5.897 14.985 9.013 2.840 Probability 0.001 0.001 0.001 0.012 aStandard errors of parameter estimates are given in parentheses. bp < 0.01. cp < 0.05. proportion of the explained variance in the mode of film viewing. Income, occupation, and age also have significant positive effects on the number of films viewed in theaters. The only significant predictor of viewing films on videocassette is age. Specifically, younger adults have viewed more films on videocassette than older adults. The effect of age on mode of viewing can be explained, at least in part, by the fact that young adults often have young children at home and, as a result, may find it inconvenient to view films in theaters. Finally, being female has a significant negative effect on the number of films viewed on television. In general, the fact that older adults view more films in theaters is offset by the fact that they have viewed fewer films on videocassette. Also, the fact that women view more films in theaters is offset by the fact that they have viewed fewer films on television. CONCLUSIONS These results suggest that there are significant differences in the cul- tural repertoires of members of the upper-middle class and members of the lower-middle class in terms of their film viewing practices. Consistent with our theoretical expectations, upper-middle class individuals have 160 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture viewed more films of all types than lower-middle class individuals. As expected, these differences in the film-viewing practices of the members of these two social classes are most pronounced with respect to art films. Also consistent with our theoretical expectations, members of the upper- middle class have viewed more of these films in theaters than members of the lower-middle class. However, contrary to our theoretical expectations, members of the upper-middle class are also more likely to have viewed films on videocassette than members of the lower-middle class. Finally, consistent with our expectations, these differences in film viewing practices of the members of these two social classes are reduced somewhat by the fact that lower-middle class individuals are marginally more likely to have viewed films of all types on television than upper-middle class individuals. In general, this research confirms the observations by several researchers (Lamont, 1991; Press, 1994; Peterson and Kern, 1996yf W K D W P H P E H U V R f the upper-middle class possess more extensive cultural repertoires than members of the lower-middle class. It also confirms that "high" culture, in the form of art films, is an important component of these cultural repertoires (Bourdieu, 1984yf . In addition, this research has introduced a number of theoretical and methodological innovations that may prove useful in future research. To begin with, it has focused attention on the importance of films in the cultural repertoires of individuals. In so doing, this research has introduced a relatively unambiguous technique for measuring the film-viewing practices of individuals with respect to different types of film. Although the specific films used to measure different film-viewing practices are certainly subject to modification, future research on film-viewing practices cannot ignore the distinction between art films, classic films, and blockbuster films. Simi- larly, this research has pointed to the importance of considering different modes of viewing films of each type. It has also demonstrated the validity of using carefully constructed neighborhood samples as a means of identifying members of different social classes. Indeed, this research clearly suggests that neighborhood of residence measures an important dimension of social class, relevant to the formation of cultural repertoires, that goes beyond the aggregate effects of such socioeconomic variables as income, education, and occupation. In so doing, these findings confirm the observation that neighborhoods can represent "symbolic communities" whose members share certain cultural practices (Hunter, 1974yf . In reaching these conclusions, this research suggests a number of possi- ble directions for future research on the relationship between social class, cultural repertoires, and popular culture. One of the most obvious questions raised by this research is the manner in which cultural repertoires serve as a basis for social exclusion. Further research is required to determine how 161 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Barnett and Allen individuals employ their cultural repertoires in social interaction as a means of maintaining class boundaries. Another obvious question raised by this research is the manner in which individuals with different cultural reper- toires interpret various forms of popular culture. It may well be that exten- sive cultural repertoires provide individuals with the means of developing a more analytical perspective toward popular culture. Very little is known about the ability of individuals from different social classes to adopt an "aesthetic disposition" toward various forms of popular culture (Bourdieu, 1984yf , I I L O P L V L Q G H H G R Q H R I W K H G R P L Q D Q W I R U P V R I S R S X O D U F X O W X U H L n modern society, we must seek to understand the manner in which people appropriate films and incorporate them into their daily lives. REFERENCES Blalock, Hubert M., Jr. 1960 Social Statistics. New York: McGraw-Hill. Blewitt, John 1993 "Film, ideology, and Bourdieu's cri- tique of public taste." British Journal of Aesthetics 33:367-372. Bourdieu, Pierre 1973 "Cultural reproduction and social re- production." In Richard Brown (ed.yf , Knowledge, Education, and Cultural Change: 71-112. London: Tavistock. 1984 Distinction: A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press. Butler, Edgar W., George Sabagh, and Mau- rice D. Van Arsdol, Jr. 1964 "Demographic and social psychologi- cal factors in residential mobility." So- ciology and Social Research 48:139- 154. DiMaggio, Paul 1982 "Cultural capital and school success: The impact of status-culture participa- tion on the grades of U.S. high-school students." American Sociological Re- view 47:189-201. DiMaggio, Paul and John Mohr 1985 "Cultural capital, educational attain- ment, and marital selection." Ameri- can Journal of Sociology 90:1231- 1261. Erikson, Robert and John H. Goldthorpe 1992 The Constant Flux: A Study of Class Mobility in Industrial Societies. Ox- ford: Clarendon Press. Fiske, John 1987 Television Culture. London: Rout- ledge. Gans, Herbert 1974 Popular Culture and High Culture. New York: Basic Books. Gripsrud, Jostein 1989 "'High culture' revisited." Cultural Studies 3:194-207. Hunter, Albert 1974 Symbolic Communities: The Persis- tence and Change of Chicago's Local Communities. Chicago: University of Chicago Press. Kain, John F. and John M. Quigley 1975 "Measuring the value of housing qual- ity." In John Kain (ed.yf ( V V D V R Q 8 U - ban Spatial Structure: 261-276. Cam- bridge, MA: Ballinger Publishing. Lamont, Michele 1992 Money, Morals and Manners: The Cul- ture of the French and the American Upper-Middle Class. Chicago: Univer- sity of Chicago Press. Lamont, Michele and Annette Lareau 1988 "Cultural capital: Allusions, gaps and glissandos in recent theoretical devel- opments." Sociological Theory 6:153- 168. Monaco, James 1981 How to Read a Film: The Art, Tech- nology, Language, History, and The- ory of Film and Media. New York: Oxford University Press. Peterson, Richard A. 1992 "Understanding audience segmenta- 162 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms Social Class, Cultural Repertoires, and Culture tion: From elite and mass to omnivore and univore." Poetics 21:243-258. Peterson, Richard A. and Roger M. Kern 1996 "Changing highbrow taste: From snob to omnivore." American Sociological Review 61:900-907. Press, Andrea L. 1994 "The sociology of cultural reception: Notes toward an emerging paradigm." In Diana Crane (ed.yf 7 K H 6 R F L R O R J y of Culture: 221-245. Oxford: Black- well. Ray, Robert B. 1985 A Certain Tendency of the Hollywood Cinema: 1930-1980. Princeton, NJ: Princeton University Press. Sklar, Robert 1994 Movie-Made America: A Cultural History of American Movies. New York: Random House. Suttles, Gerald D. 1972 The Social Construction of Communi- ties. Chicago: University of Chicago Press. Vidal, Gore 1992 Screening History. Cambridge, MA: Harvard University Press. Warner, W. Lloyd 1949 Social Class in America: A Manual of Procedure for the Measurement of So- cial Status. Chicago: Science Re- search Associates. White, Michael J. 1987 American Neighborhoods and Resi- dential Differentiation. New York: Russell Sage Foundation. 163 This content downloaded from 66.77.17.54 on Wed, 29 Jun 2022 15:57:02 UTC All use subject to https://about.jstor.org/terms