Published on February 7, 2008
The stratification of Lifestyles in France. Global omnivouresness and taste eclecticism in different domains.Philippe Coulangeon & Yannick Lemel: The stratification of Lifestyles in France. Global omnivouresness and taste eclecticism in different domains. Philippe Coulangeon & Yannick Lemel Brno Meeting (May 24-27, 2007) of the ISA Research Committee 28 on Social Stratification and Mobility Introduction: Introduction The social stratification of taste and cultural practice in contemporary western societies has been extensively investigated since Bourdieu’s seminal work Distinction: A Social Critique of the Judgment of Taste (1984). Most of the current studies in this field appear to take the form of both empirical and theoretical reappraisals of and challenges to Bourdieu’s hypotheses. Among the most common challenging alternative theories, the omnivore/univore thesis clearly appears as the most collectively accepted one (Peterson and Simkus, 1992). However, the extent to which Peterson’s results really challenge Bourdieu’s theoretical construct remains questionable. Moreover the meaning of “omnivorousness” would appear to be quite controversial. We’d like to use a recent survey conducted in France on a very large range of cultural practices and leisure activities in order to evaluate the present scope of Bourdieu’s model and the criticism it has received. Summary: Summary Bourdieu, the Benchmark Survey & Data The Research Questions In Search of Life-styles Are the life-styles independent from social positions ? Life styles & Class positions. The homology thesis. The Research Questions. The Answers Conclusion & Evaluation Bourdieu’s Basics: Bourdieu’s Basics Bourdieu’s view of cultural consumption and lifestyles is basically structured by two concepts, i.e. structural homology and habitus. Habitus are cultural structures that exist in people’s bodies and minds and shape a wide variety of their behaviours, beliefs and thoughts. The effect of habitus is largely due to informal mechanisms acquired during the stage of primary socialisation. It does not involve a genuine learning process. Structural homology is the assumption that social class structure is linked to the structure of aesthetic preferences through an isomorphic relation (one-to-one correspondence). More specifically: People’s tastes are seen as channelled by their position within the class structure, which is defined by the volume of capital and its "composition" . The sphere of tastes is organised hierarchically in line with a “highbrow/lowbrow” opposition, so that the dominant tastes and practices are always the dominants’ tastes and practices regardless of the cultural or artistic domain under consideration. This is our topic in this presentation We will not speak about habitus Bourdieu’s criticisms: Bourdieu’s criticisms ► Some radically individualistic arguments deny the social dimension of taste and lifestyles: Featherstone, 1987; Bauman, 1988; Giddens, 1991; Beck, 1992. ► The “omnivore/univore” hypothesis judges that the main social distinction today is a matter of cultural diversity rather than “highbrow or lowbrow” culture. The “omnivore/univore” hypothesis was originally proposed by Di Maggio (Di Maggio, 1987). It was systematised in a seminal article by Richard Peterson and Peter Simkus on the musical tastes of contemporary Americans in 1992 . However, many elements of this theoretical construct remain quite unclear. More precisely, is the “omnivore” style a privilege of the “dominants” only, i.e. the only ones to be characterised by a large cultural diversity? Or is this latter a matter of degree, as some of Peterson’s recent writings could suggest? Survey & Data: Survey & Data Survey on Sport and Cultural Practices , Insee, 2003 Random sample of individuals living in ordinary household in France All population: N= 5626, Some analyses restricted to people employed or retired: N =3691 Questions on both activities (what people do) and tastes (what people prefer) in different domains : Activity domains: sports, outdoor recreation, reading, “cultural” activities, artistic hobbies, etc: 44 scales see Preferred genres in two domains: reading books: see listening to music: see Individual variables: Age, Gender Education : in years required for the highest diploma obtained, Social status: scale calculated with Goodman model II applied to occupational homogamy data Income: Log of household income/number of persons living in the household Class position: EGP class schema and a Bourdieusian « social space »: see Research Questions: Research Questions The questions we will try to answer are the following: 1. Can we identify lifestyles? The notion of “lifestyle” does not have a truly canonical definition. In all cases, however, the notion leads to the idea of a presumed coherence of a minimum number of activities and intentions. We will use clustering techniques to try to find such configurations. In this process, an important place will be given to the correspondence analysis methods for they are the techniques used by P. Bourdieu . 2. If yes, which principles of organisation possibly underlie them : highbrow/lowbrow, univore/omnivore, others ? 3. If yes, are these lifestyles independent from people’s social characteristics? 4. If these lifestyles are not independent from people’s social characteristics, can they be put in correspondence with “class” positions? 5. If yes, they could be put in correspondence, is this an univocal one? These are obviously not questions to which there will be definite and conclusive answers. Our analysis will be along the lines of the « stylised facts » inquiry rather than hypothesis tests. Products/Domains and Brands/Genres. : Products/Domains and Brands/Genres. In consumption analysis, there is an obvious distinction between “products” – e.g. refrigerators – and “brands” – Brandt or Liebert. This distinction is a routine one for the economists (with the idea of separable utility) and for the marketers who know that they will have to struggle in order to impose their brand to the detriment of others, without even being able to increase the market shares of their product. The behaviour determinants could be different at both analysis levels. In the analysis of cultural consumption, this idea that the interpretative principles could not be of the same nature according to the level of analysis – product or brand – is not so common, the analysis tending to focus on the “brand” level which is supposed to be more relevant. Nevertheless, it must be all the more examined since the explanatory theories – univocal structural homology from a Bourdieusian perspective or that of omnivorousness – probably do not have the same meanings and do not refer to the same interpretation according to the level they are applied to. We will try to do both kinds of analysis. Activities as well as genres: “Reading books” versus “Attending movies” as well as “Thrillers” versus “Classics novels”. In Search of Lifestyles: In Search of Lifestyles The basic sequence will be as follows: MCAs or CAs will enable to synthesise the range of activities through a limited number of dimensions; Hybrid clustering techniques will then be used to group the individuals on the basis of their similarities by reference to these dimensions. This sequence will be applied to the 44 domain scales (“products”) and the two set of genres (“brands”) separately, thus identifying three set of clusters of “life-styles”, “reading styles” and “music-styles” respectively. All in all, the conclusions seem very reliable . > The clusters are quite independent from the precise clustering techniques used. > with qualifications, the number of activities/genre is the main principle differentiating the various clusters. Example: readings genre > The correlations between the three classifications are high. Reading Books. Clustering Analysis: Reading Books. Clustering Analysis “Mixed classification” on factor coordinates of MCA (2), with a preliminary research of stable groups by the K-means clustering method; then aggregation from class centres by Ward’s method, resulting aggregation in 5 groups. The genres are classified according to the number of clusters to the definition of which they contribute the most. The Similarity Between the Various Styles: The Similarity Between the Various Styles The correlations between the three classifications are very high. Probably, the underlying dimension is linked to the fact that the number of activities is the main principle differentiating the various clusters. From that point of view, considering the whole set of «products» or only «brands» among certain «products » leads to the same conclusions and to the same structuring principle of “omnivorousness”. The correspondence is nonetheless not bi-univocal. It is particularly marked at the extremes, i.e. at the clusters gathering the least practicing persons or, on the contrary, the ones who practice the most, without being, here again, bi-univocal. Bivariate correspondence analysis Inertia explained by the first component: 0,99 % 0,91% 0,80% Are the Lifestyles Independent From Social Positions? : Are the Lifestyles Independent From Social Positions? It is not conceivable to present the detailed results of multivariate analyses . We will only illustrate the fact that the differences in lifestyles between people of different social characteristics are significant and that they importantly contribute to render the diversity of practices as appear through multinomial logistics explaining life-styles by an usual set of background variables. The result can be observed both when dealing with clusters built from sets of activities or from reading and listening styles. These results are enough to invalidate the individualisation thesis, at least in its most extreme forms. Comparing people’s lifestyles and their social positioning . Summary: Comparing people’s lifestyles and their social positioning . Summary The comparison of people’s lifestyles and their social positioning leads to quite similar conclusions, whether we examine social position defined in terms of Bourdieusian social space or approached in the EGP class scheme. The conclusions are the following. The groups of people sharing a lifestyle clearly distinguish themselves from one another in the space of activities (mechanical result of their construction mode), but far less through their social positioning, even though there is a certain correlation. The homology thesis – in the strictest sense of the word suggesting a biunivocal correspondence – cannot be retained. Life-styles in the Space of Activities and in a Social Space: Life-styles in the Space of Activities and in a Social Space Space of activities Social space Each ellipse is calculated so as to include 50% of the population embracing the «lifestyle» considered Lifestyles very clearly distinguish themselves from one another in the space of activities. They are associated with places in the social space that much less distinguish themselves, even though the hierarchical organisation by volume of “global capital” is clear. There is no differentiation related to the “composition of the capital. Reading Styles & Life-styles in a Social Space: Reading Styles & Life-styles in a Social Space Each ellipse is calculated so as to include 50% of the population embracing the «style» considered Reading Styles & Music Styles and in a Social Space: Reading Styles & Music Styles and in a Social Space Each ellipse is calculated so as to include 50% of the population embracing the «style» considered Overall conclusion. Assessing the alternative thesis: Overall conclusion. Assessing the alternative thesis There is no support for the thesis according to which an individualisation would lead to the disappearance of all social components of lifestyles. As for the principles of activity organisation, the idea seems to be confirmed that cultural diversity rather than a highbrow/lowbrow opposition is structuring. But here this cultural diversity can be analysed like a gradient rather than like the specific lifestyle of a social category. Assuming that our operationalisations of Bourdieu’s concepts are satisfactory, there seems to be no confirmation of the idea of a strong structural homology. > The dimension of global capital is very clearly effective to differentiate lifestyles but the overlapping is important and a one-to-one correspondence between life-styles and social positions is not visible . > The dimension of “capital composition ” is not effective to differentiate lifestyles. > Symmetrically the dimension of the space of practices that is not the immediate translation of the activity volume, cannot be retraced by position in a social space like the first and main dimension. Evaluation. : Evaluation. >> The analysis was carried out in two steps: lifestyle research and then examination of the homology between lifestyles and class positions. This way of proceeding is obviously questionable when referring to Bourdieu’s approach which poses a dialectic relation between both elements. In many regards, “classes” are also groups of status in the Weberian sense. If one adopts a strictly structuralist point of view, the universes of practices and positions both enable us to locate the same underlying structure. As both universes are in homology, it wouldn’t matter which one was used. Things are different, however, if we embrace a less strict perspective by considering that activities as well as status elements are just as many proxies, each helping us get information about the membership class. >> Even if, one can assume that what is measured when people are questioned about the genres they listen/read to most often is much closer to individuals’ real preferences than what is measured when respondents are surveyed abstractly on their tastes when answering questions about their preferences, estimating taste on the basis of practices could be viewed critically (Hugues and Peterson, 1983; Peterson and Simkus, 1992). ANNEXS: ANNEXS The domain scales (1): The domain scales (1) The domain scales (2): The domain scales (2) Reading Genres: Reading Genres Question on the book genres the respondents have read in the past year. The persons could choose up to 11 genres among the 11 proposed: > thriller > romance > classics > other novel genre > historical books > books on politics, economics, religion and human sciences > scientific and technical books > cookery, decoration and do-it-yourself books > art and photography books > sports books > other Musical Genres: Musical Genres Question on the musical genres the respondents most frequently listen to. The wording of the question allows the quotation of a maximum of three genres from the following list: > Pop music > International pop > Techno music > World music > Rap > Rock > Jazz > Classical music, opera Why a special place to Multiple Correspondence Analysis ?: Why a special place to Multiple Correspondence Analysis ? Multiple correspondence analysis is a descriptive technique rather similar to standard factor analysis except that it deals with nominal variables. Multiple correspondence analysis decomposes the overall chi-square statistic/inertia of contingency tables in successive ‘dimensions’ as is done by factor analysis. It is then possible to select lower-dimensional representation, which allows the reconstruction of most of the inertia matrix of variables. The distances between the individual/variable points in the associated dimensional displays are such that two points that are close to each other are similar with regard to the pattern of relative frequencies across the variable/individual. We will give a special place to Multiple Correspondence Analysis for they are the techniques used by P. Bourdieu and they are particularly adapted to the data available to us. A Bourdieusian social space: A Bourdieusian social space CA, employed & ex-employed persons three variables: education (in years completed), social status scale (based on Goodman model III on homogamy data) Log of household income/head Confirms the Bourdieusian idea of bidimensionality of social space. Suggests the construction of two scales: global capital (E+SS+I) composition of capital (E+SS)/2-I Proxies for a Bourdieusian social space: Proxies for a Bourdieusian social space CA, employed & ex-employed persons three variables: education (in years completed), social status scale (based on Goodman model III on homogamy data) Log of household income/head Confirms the Bourdieusian idea of bidimensionality of social space. Suggests the construction of two scales: global capital (E+SS+I) composition of capital (E+SS)/2-I
Livegrace Geronda is on Facebook. Join Facebook to connect with Livegrace Geronda and others you may know. Facebook gives people the power to share and...