The, SotioCogij ôfCcctíuve Bnergfrig fktweßcai fers^eeůfi/ei
V/~d?i4<7!  Craw- íeď.
10
Methodological Dilemmas in the Sociology of Art
Anne Bowler
To date, the important contribution of sociology to the study of culture and the arts has been to demonstrate the necessity of understanding the work of art and role of the artist in their social, political, and historical contexts. Empirical studies of the social and institutional matrices within which aesthetic objects are materially produced have pointed to the implicitly collective nature of artistic production (Becker 1974, 1982). Such studies have problematized traditional art-historical and literary-critical conceptions of the artist as isolated genius. Similarly, sociological research has shown the crucial role of economic and organizational factors in structuring the emergence of new artistic genres and styles (White and White 1965; Zolberg 1983; Crane 1987). Such work speaks directly to current debates on canon formation, posing a serious challenge to unreflexive classifications of timeless "great works" and demonstrating the ways in which the category of the "great work" is itself a socially and historically contested terrain. Investigations into the composition of audiences and forms of audience response have revealed the interdependence of access to culture with economic, political, and social position (Bourdieu 1984; Gans 1974,1985). Finally, there now exists a wide body of work on the social uses of art in the reproduction of systems of stratification and class power (see, for example, Bourdieu 1984; DiMaggio 1982). In sum, long before concepts like hegemony and "the Other" became the fashionable rallying cry in literature departments across the country, sociologists had turned their attention to the analysis of the inextricable connection between art, ideology and power.
Too often, however, sociological analyses concerned solely with the organization of systems of cultural production, the social role of the artist as tastemaker or the class structure of audiences have
248
ANNE BOWLER
resulted in mechanistic conceptions of the relationship between cultural forms and social processes. Describing the connecting links between art and society has been a largely one-sided project for sociologists, for whom cultural forms and practices continue to appear as the manifestation, measure or byproduct of some presumably more basic social factor, e.g., institutional strain, group solidarity, stratification, etc.1 As Goldfarb (n.d.: 15-16) has written, "the arts' place in society is situated, and their day-to-day functioning is explained, but their real distinctiveness, their broader cultural and political significance in the historical developmenl of society is ignored."
In his now paradigmatic essay, "Art as collective action," Howard Becker (1974) noted, not without irony, the tendency of sociological studies of art to write of the organizations and institutions of artistic production without reference to the social actors or activities through which those very organizations and institutions came into being. Today, we might make a structurally similar observation, noting the number of studies grouped together under the heading "sociology of art" which confine themselves to the analysis of the social conditions of artistic production and reception without reference to the ostensible object of analysis: the work of art. The implications of this are significant. First, it undermines the subdisdpline's self-description as the sociology of art.2 Second, it calls into question the sociological claim that the work of art is, in the final analysis, a social product. For if the work of art is both socially located and materially produced, why does it continue to be so systematically excluded from the domain of sociological inquiry? Third, the relationship between cultural forms and social processes remains obscure.3 Specifically, sociology cannot account for the ways in which cultural forms and practices do not simply reflect an already given social world but, rather, play a constitutive role in the construction of that world (Wolff 1992: 707).
Taken together, these shortcomings underscore the need for new methodological strategies in the sociology of art capable of grasping the complex interplay of aesthetic, social, economic, and political factors. Toward this end, this chapter presents two arguments: first, for the autonomy of artistic works and practices as objects of inquiry in their own right; second, for the importance of attention to questions of meaning. These arguments pose a direct challenge to both the traditional doctrine of aesthetic neutrality in sociology which
DILEMMAS IN THE SOCIOLOGY OF ART             249
mandates that not only questions of aesthetic judgment but the work of art itself remain outside social analysis and other sociological approaches to art which attempt to bracket questions of meaning in the attempt to place the subdiscipline on a presumably more objective footing. The last section of the chapter presents examples of the application of these arguments from some recent work in the sociology of art, highlighting specific methodological strategies and showing their fruitfulness for future research.
The Autonomy of Art
Two of the more persistent problems in the sociology of art may be formulated by the following pair of questions: What is the relationship of art to society? And how should this relationship be studied? Implicit in each of these questions is a tacit assumption that art is not reducible to society in some simple, uncomplicated way, that the nature of the relationship of art to society requires clarification, and that art is a legitimate area of sociological study. In other words, each of these questions pivots around some concept of the autonomy of art, however relative, provisional or contingent one might want this concept to be and, further, points to the significance of the concept of autonomy for sociological analysis.4
Thus invoked, however, it is necessary to provide a cautionary note about a concept with a long and troubled past. Part of this derives from a confusion over multiple and competing definitions of the term, not unlike the concept of culture. More problematic, however, is its common association with the radical doctrine of aesthetic autonomy espoused most dramatically by various late-nineteenth century aestheticist movements and perpetuated today, in various forms, by art historians and literary critics who continue to bracket social and political questions in favor of the formal properties of the allegedly free-floating work.5 As stated at the beginning of this chapter, it has been the very real and significant contribution of sociology precisely to demystify this concept of autonomy, what Eagleton (1990) has called the ideology of aesthetic autonomy.
For this reason, it may be argued that the term should be abandoned altogether. Nevertheless, the continued importance of the concept becomes clear when we consider two other definitions of
250
ANNE BOWLER
autonomy advanced first in the aesthetic philosophy of the Frankfurt School (in particular, the work of Adorno and Marcuse) and, more recently, in the philosophical and sociological work of Bürger, Habermas, and Goldfarb: a historical/institutional definition which traces the historical development and structure of the autonomous institution of art in modern society and a methodological definition concerned with the autonomy of art as an object of inquiry in its own right.6 What follows is a brief sketch of the first definition, which points out some of its strengths for how we might begin to conceptualize the institutional status of art in modern society. However, the bulk of my discussion is directed at the second definition, where I present an argument for its importance for the contemporary sociology of art.
The autonomous status of art in the historical/institutional sense refers to the institutional framework for the production and reception of aesthetic works in modern bourgeois society (Bürger 1984). It is widely acknowledged that the autonomous institution of art may be understood as part of a historical process of cultural differentiation which began during the Renaissance and reached definitive form by the late eighteenth century when it received its most systematic philosophical elucidation in the work of Kant. Historically, we find the origins of the autonomous institution of art first in the decline of the religious/cultic functions of art with the development of artistic production for court and patron. Nevertheless, it is only with the emergence of the modern capitalist market that, as Weber (1979: 342) observed, "art becomes a cosmos of more and more consciously grasped independent values which exist in their own right." The emergence of a differentiated sphere of art is thus coterminous with the processes of societal rationalization and capitalist modernization from which the two modern systems of state and economy arise (Habermas 1984).
The contribution of this concept of autonomy is twofold. First, it allows for the distinction to be made between two interrelated but distinct dimensions of this historical development: (1) the institutional framework of art in modern society; that is, the development of systems of production and reception mediated by the mechanisms of a commercial, capitalist market; and (2) the doctrine or ideology of aesthetic autonomy as the necessary opposition of art and society. Second, it reveals the fundamentally contradictory character of the autonomous status of art in modem society. For the freedom
DILEMMAS IN THE SOCIOLOGY OF ART            251
of artistic production from its traditional religious and courtly functions is inextricably tied to the status of art as a commodity. Consequently, the "liberation" of art from traditional religious and courtly modes of power involves the reinscription of the aesthetic in the abstract modes of power of the modern market.
While the historical/institutional definition of autonomy does not solve the problem of the relation of art to society once and for all, it does, however, provide us with more solid historical grounds on which to understand the relation of art to other spheres in modern society. Specifically, it allows for a conceptualization of art as a sphere always connected with but not simply reducible to other social spheres. Similarly, it allows for the complex, historically changing structure of this relationship, something that neither traditional Marxist formulations of art-as-ideology or traditional art-historical conceptions of the free-floating, transcendent object have been able to grasp. Particularly important is the fact that it gives the question of the relationship of art and society itself an important place in the analysis. The methodological definition of autonomy addresses these issues in more detail.
If the concept of autonomy in the historical / institutional sense delineated above has been recognized as an important contribution to the analysis of art, the methodological definition of the autonomy of art, by which I mean the analysis of artistic works and practices as objects of inquiry in their own right, has had a more problematic history (Goldfarb 1985, 1989: 204-5). The primary source of resistance can be located in the doctrine of aesthetic neutrality which has dominated sociological approaches to art and culture (Bird 1979; Zolberg 1990:44-5). As Bird (1979) has cogently noted, there are two interrelated but analytically distinct aspects of this doctrine: first, the insistence that sociologists resist questions of aesthetic judgment; and secondly, that the work of art itself remain outside the domain of sociological analysis. Bird (1979: 30) states, "The sociologist must confine himself or herself to the objective facts of production and consumption ... found in the social relations governing the production of art: 'the socialization and careers, the social positions and roles' of artists, 'the distribution and reward systems', [and] 'tastemakers and publics.'" Thus, we can see that the very subject matter of the subdiscipline carries with it an implicit problem. Engagement with the aesthetic object threatens to implicate the social scientist at any moment in matters
252
ANNE BOWLER
involving questions of meaning and judgment (Goldfarb 1985: 3; Wolff 1989:10). To do so places the scholar, particularly the young scholar not yet established in the field, at risk of the charge of not being properly "sociological."
The methodological argument for the autonomy of art poses a direct challenge to the principle of aesthetic neutrality. Artistic works and practices are neither a reflection of society nor a secondary byproduct of some presumably more basic or "objective" social mechanisms. At the core of this argument is the decisive rejection of a base/superstructure model of art and society which has plagued both successive developments in Marxist aesthetics and, in different forms, sociological approaches to the study of art within which the work of art continues to appear as the manifestation of some other social processes (Williams 1989: 165-6). Particularly influential in this regard has been Williams's thoroughgoing critique of the concepts of base and superstructure as well as poststructuralist theories of discourse and power which point to the active role of language and other sign systems in society (Barthes 1972a, b; Foucault 1980,1984; Williams 1977). As Wolff (1992: 707) has more recently written, "far from reflecting the already-given world, ... cultural forms participate in the production of that world."
While it is easy to assert, in theory, one's rejection of a base/ superstructure model of culture, the extent to which the work of art is absent from analysis means that the model is, in practice, implicitly upheld. Thus, central to the project outlined here is the development of a sociological approach which can account for what Williams, following Bakhtin, has called the "specificity" and Goldfarb has termed the "distinctiveness" of aesthetic forms and practices (Williams 1977, 1989; Goldfarb 1985, 1989). In this view, sociological approaches which continue to privilege generalizabiliti as a methodological criterion for the study of art run the risk o smoothing over precisely those contradictions and differences be tween cultural objects (as well as between cultural and other socia factors) from which significant sociological insight may be gleaned. What this means, in practice, is coming to terms with questions o genre, form, content, narrative, representation, aesthetic conventioi and intertextuality - questions which can only be addressed bj direct engagement with the work of art.
Two points need to be made clear at this juncture. First, thi methodological argument for the autonomy of artistic works anc
DILEMMAS IN THE SOCIOLOGY OF ART             253
practices is not an attempt to simply reverse the traditional causal framework of base and superstructure nor does it privilege aesthetic and cultural factors over other social forces.8 To the contrary, the central thrust of Williams's critique of the base/superstructure model of society is to show their mutual interdependence. Second, the argument for the inclusion of artistic works and practices in sociological analysis is not based on the assumption that the meaning or significance of a particular work resides solely in the artist's intentions or is somehow "embedded" in the object and thus simply needs to be unearthed or revealed. Rather, the central contribution of what has come to be called the "cultural studies" approach following the work of Williams has been to demonstrate the ways in which the meaning or significance of cultural works and practices may be altered by changes in social location and historical context.9 The task, as Williams (1989) has argued, lies in a focus on the elucidation of the specific and historically changing relations between cultural works and practices with social institutions and processes.
In sum, like the institutional/historical definition of aesthetic autonomy, the methodological definition is useful for the sociology of art. Specifically, it allows for the conceptualization of artistic production as a sphere always connected with but not reducible to other social processes. Similarly, it allows for the analysis of aesthetic works and practices without recourse to the myth of the transcendent object or artist-as-genius. Finally, this approach positions the relationship of artistic works and practices with social processes at the center of analysis. For the autonomy of art in both of these definitions does not imply that art and society are somehow "separate" in some absolute sense but that the autonomy of art as either a differentiated sphere or an object not reducible to some other social factor íŕseZ/becomes an important focus of the analysis.
The Problem of Meaning
From the foregoing, it might be easy to conclude that sociologists who continue to exclude the work of art from analysis simply suffer from an out-moded allegiance to a base/superstructure model of society. Arguably, in some instances, this may be the case. But the problem is not nearly so simple. Until recently, sociologists have
254
ANNE BOWLER
paid relatively little attention to the study of art, designating it a subject matter more properly relegated to the fields of philosophy, art history and literature (Zolberg 1990: 29-52). While sociologists have begun to turn their attention to the study of culture and the arts, what distinguishes the sociological approach to art from that of the humanities is the sociologist's focus on the institutions in which aesthetic objects are produced and received, leaving questions of meaning to art historians and literary critics.
In part, this may be the result of a subdiscipline which has only recently begun to develop. As Crane (1987: 148) has observed, "systematic analysis of visual materials by social scientists has rarely been done and few guidelines exist for a sociological examination of aesthetic and expressive content in art objects." More is at stake here, however, than the growing pains of a relatively new field of inquiry. For many sociologists, explicit disavowal of questions of meaning appears as a necessary condition for remaining faithful to, if not the positivist tradition in sociology, a commitment to "rigorous" social science. Wuthnow (1987), for example, noting the strong association of interest in cultural issues with "the branch of sociology that emphasizes its humanistic elements rather than its scientific aspirations," states, "Culture remains, by many indications, vaguely conceptualized, vaguely approached methodologically, and vaguely associated with value judgments and other sorts of observer bias" (pp. 5-6). More recently, a statement of this position can be found in DiMaggio's (1991) call for "an analytic sociology of culture, distinct from criticism and textual interpretation" (p. 153).
It is not my intention here to inveigh against the evils of positivism. Moreover, I would like to explicitly state my own commitment to a rigorous sociology of art. The question is, what does "a rigorous sociology of art'/ mean? My own objective is to move beyond the notion of some simple dichotomy between the empirical and theoretical, including the presumed binary opposition between institutional and interpretive approaches to the study of culture.10 Nevertheless, the impulse toward a simple refusal to engage questions of meaning and interpretation has to be acknowledged as problematic. First, it is nothing new to point out that the very act of choosing what kind of art to study entails an evaluative component which assigns significance to the objects selected for analysis, something few sociologists of culture today would deny. Beyond
DILEMMAS IN THE SOCIOLOGY OF ART            255
this initial step, however, the traditional strategy adopted by sociologists to eliminate an evaluative component from their research has been to work with existing systems of classification. But as Wolff (1992) and Bird (1979) have demonstrated, far from guaranteeing the objectivity of the analysis, such a strategy actually tends to confirm and reinforce existing aesthetic hierarchies. Finally, although sociologists have, for the most part, abandoned the crude notion that art simply reflects society, the refusal to engage questions of meaning nevertheless tacitly begs a form of residual reflectionism through the assumption that the "objective" facts lie in the organization of production and consumption rather than the meaning(s) of the artistic works and practices in question.
It will be noted that sociologists have not altogether ignored the problem of meaning in the pursuit of methodological strategies appropriate to the analysis of culture and the arts. One of the more serious attempts to address this problem may be found in Wuthnow's (1987) Meaning and Moral Order. According to Wuthnow, sociology has been hindered by a subjectivist approach to culture that privileges the problem of meaning in cultural analysis. But because meaning, presumed to reside in the psychological states of individuals, is ultimately inaccessible to the social scientist, sociologists should be advised to go "beyond the problem of meaning" and confine themselves to the observable aspects of culture that can be studied objectively (Wuthnow 1987: 60-5). As Wuthnow (1987: 335) states, "Even if cultural analysis is regarded as an interpretive science, the need remains to put its claims on as solid an empirical footing as possible."
Three underlying assumptions in Wuthnow's argument warrant brief examination. The first of these is Wuthnow's repeated association of the sociological interest in meaning with a "subjective" approach to the study of culture. As Griswold (1987a: 3) has observed, there is no reason why meaning has to be conceptualized solely at the level of the individual. Rather, meaning is constructed as an ongoing process in the complex and often changing intersection of a plurality of factors. Meaning is no more located in the psyche of a single individual than it is somehow eternally "embedded" in the work of art, as traditional art historians and literary critics would have it.
A second assumption in Wuthnow's argument is that abandoning the problem of meaning is the necessary prerequisite toward
256
ANNE BOWLER
the establishment of an objective approach to culture. According to Wuthnow, sociology's "subjectivist" approach to the study of culture has its roots in the discipline's underlying adherence to the dualism of subject and object which associates culture with the subjective while society or social structure are viewed as objective realities (Wuthnow 1987: 23-8). Ostensibly, it is Wuthnow's intention to get beyond this subject/object dualism. The strategy Wuthnow proposes is to abandon the problem of meaning, the central concern of "subjectivist" approaches, and thus render the sociology of culture "objective" (1987: 60-5, 333). But as Calhoun (1992a) has incisively observed, this does not dispense with the subject/object dualism. Eliminating the dualism of subject and object would necessitate rejecting the definition of meaning as subjective. Instead, by relegating meaning to the subjective, Wuthnow merely reinscribes the dualism of subject and object on another level.
The third assumption in Wuthnow's argument, that a "post-structuralist" approach will provide sociologists with a methodological foundation for the objective analysis of culture, is particularly problematic.11 According to Wuthnow (1987: 51-3, 60) poststructuralist approaches to cultural analysis are characterized by a "shift away from the problem of meaning" in favor of a focus on the formal relations between cultural symbols and sign systems. This is fundamentally misleading. The problem of meaning sits at the core of the poststructuralist project.12 Arguably, the central contribution of the wide body of often quite disparate work done under the heading of poststructuralism has been to show that meaning neither emanates from the experience of a single knowing subject nor is fixed in some absolute, ahistorical sense. As historian Joan Scott (1988: 5), who has argued for the usefulness of poststructuralist theory for the analysis of gender, has written: "Instead of attributing a transparent and shared meaning to cultural concepts, poststructuralists insist that meanings are not fixed in a culture's lexicon but are rather dynamic, always potentially in flux. Their study therefore calls for attention to the conflictual processes that establish meanings, to the ways in which such concepts as gender acquire the appearance of fixity." In any case, the problem of meaning does not disappear simply by declaring it outside the purview of analysis. For it is precisely the insight of post-structuralist theory to show the crucial role discourses, language
1J1LEMJV1AS IN THE SOCIOLOGY OF ART             257
and representation play in the construction of social worlds (see, for example, Foucault 1979, 1980).
Implications for Research
Thus far, I have argued for the autonomy of artistic works and practices as objects of inquiry in their own right and the importance of attention to questions of meaning. In presenting these arguments, I have drawn on insights from work within the field of sociology as well as other work outside the field properly defined; namely, cultural studies and poststructuralist theories of discourse and representation. Two important points need to be made clear here. First, neither argument represents an attempt to reverse the causal framework of base and superstructure, as previously noted, or assign culture the status of an independent variable. An important insight associated with early work in cultural studies, specifically with reference to the research that came out of the Birmingham Centre in the 1970s, explored the various ways in which audiences may make critical use of popular culture.13 Such studies directly contributed to the increased awareness of the importance of cultural analysis that has taken place in the social sciences since the early 1980s. They also provided a useful and necessary corrective to the antipathy for popular culture held by traditional Critical Theory as well as the generally "high cultural" focus of earlier work in the sociology of art. Nevertheless, the utility of the contemporary cultural studies approach as it has developed since the late 1970s remains limited for the sociology of art for two reasons. The first has to do with the simple fact that, to date, work in cultural studies has been almost exclusively on popular culture. The value of a cultural studies approach to Abstract Expressionism, for example, is not yet clear. A second, potentially more serious problem derives from the unfortunate tendency which can be noted in the wide body of work that now calls itself "cultural studies" to automatically privilege cultural factors as sources of opposition to systems of stratification and social control. It has become something of a maxim in such work that culture (specifically, popular culture) equals resistance.14 Second, my argument for the analysis of aesthetic works should
258                                 ANNE BOWLER
not be viewed as an attempt to privilege the analysis of aesthetic factors over other social processes. The contribution of poststructuralist theories of discourse and representation, as I have suggested, has been to demonstrate the ways in which systems of representation play a constitutive role in social relations. Poststructuralist approaches to gender, for example, have traced the ways in which historically shifting cultural definitions of femininity have informed both institutional barriers to women in public life as well as a succession of strategies in feminist practice (Parker and Pollock 1981; Riley 1988; Scott 1988). Similarly, studies in the area of race and ethnicity have examined the role of cultural images and stereotypes in the construction of specific social groups as subjects of systems of power and social control (Gates 1986; Pratt 1986). Power, as these analyses demonstrate, rests not just on the organization or control of certain material factors but also the social meanings given to these factors (Weedon 1987: 2). Nevertheless, the wholesale adoption of poststructuralist theory for the sociology of art remains deeply problematic. The formal emphasis of poststructuralist analyses of texts often operates to minimize the significance of social factors and thus simply turns the sociological tendency to ignore the aesthetic dimension on its head. In its most extreme forms, social and material factors disappear altogether. Society simply becomes another "text."15
What is needed, therefore, is the development of a sociology of art capable of surmounting the traditional impasse that has existed between institutional and interpretive approaches to the study of culture and the arts. In practice, this means an approach capable of simultaneous attention to aesthetic issues and social structure. One area of scholarly inquiry which has begun to develop such an integrative approach to the study of culture in recent years has been the feminist analysis of art.16 The creation of a substantial body of work in this area since the early 1970s has articulated two central problems which provide concrete illustration of the methodological arguments advanced in this chapter: first, the problem of explaining the exclusion of women's artistic production from the modernist canon and theoretical literature on modernity; and second, the related problem of the representation of women in the modernist work of art. The last section of this, chapter therefore turns to empirical examples of these problems demonstrating that they can only be adequately addressed in a framework which
DILEMMAS IN THE SOCIOLOGY OF ART              259
«.ombines attention to institutional factors with the analysis of works i »f art and questions of meaning.
Institutional Analysis and the Work of Art
'rior to the late nineteenth century, institutional barriers to women's participation in artistic production took the form of the exclusion ■ if women from membership in the prestigious and influential icademies. In practice, this meant the restriction of women from a system of academic training and privilege during a period sigrüfi-•■ant for the professionalization of art and the rationalization of its methods of study (Parker and Pollock 1981: 27-8, 87). Most significant, by many accounts, was the exclusion of women from the life-class. For the study of the nude constituted not only the most jrivileged course within the academic curriculum but was considered to be the very cornerstone of the education and training )f great artists (Chadwick 1988: 167; Parker and Pollock 1981: S3-5, 87).
By the late nineteenth century, however, academy membership vas no longer the central issue. Alternatives to academic training and exhibition combined with the declining influence of the academies in the face of modernist challenges to tradition meant a shift in the complex relations among the institutions of artistic production and reception, the artistic career and the nature and definition of the work of art. How then may we account for the exclusion of the work of female artists from the modernist canon? The case of Mary Cassatt (1844-1926), the subject of an important study by feminist art historian Griselda Pollock (1980), illustrates the necessity of bringing the work of art into the analysis.
Despite the social and cultural norms that defined artistic work as an unsuitable career for the middle-class, respectable woman of the late nineteenth century, existing documentation strongly supports the conclusion that the professional success of Mary Cassatt was based on those institutional criteria that Lang and Lang (1990) have identified as necessary for the creation and survival of artistic reputation. In the late 1870s, she became an active member of the French Impressionists, the only American ever to exhibit with the group.17 In addition to Degas, her work was admired by Pissaro, Gauguin and the writer and critic Huysmans (Harris and Nochlin
260
ANNE BOWLER
1976: 237, 239; Rubinstein 1982: 134). Critical acclaim for her work did not end with her association with the Impressionists, however. Cassatt went on to work and exhibit to critical praise in France up until the last decade of her life (Bullard 1972; Harris and Nochlin 1976: 237-41; Matthews 1984; Seldin 1987). The first biography of Cassatt, financed by the American collector James Stillman and written by Achille Segard, appeared in 1913.18 Later in this century, preservation of Cassatt's artistic reputation became the project of art historian and curator Adelyn Breeskin, who compiled a complete record of Cassatt's oeuvre.19
In light of these factors, how can we explain the marginalization of Cassatt in the modernist canon? Until recently, references to Cassatt have been conspicuously absent from leading art history survey texts.20 As late as 1973, Cassatt merited only brief discussion in a major history of Impressionism (see Rewald 1973). And it was not until 1970 that a retrospective of the artist's work at a major national museum took place.21 As Pollock's study suggests, the answer to this puzzle lies in an analysis of the modernist work of art.
A central point of agreement in otherwise competing definitions of aesthetic modernism lies in the recognition of artists' increasing rejection of traditional narrative modes of representation in favor of the attempt to create a new, universal language of form. But this emphasis on form did not mean the absence of content. Rather, underlying the increased attention given to formal concerns was the idea that new techniques in art were needed to represent the new "contents" of a rapidly changing, specifically modern world: technology, the meaning of progress, and the changed consciousness of time and space. Two of the central themes of early modernisl painting and literature, war and the public life of the city, illustrate this point. Particularly important, as Pollock notes, is the representation of public space: the fascination with the city streets, cafes, and arcades of Baudelaire's "painter of modern life" to which the "respectable" woman had limited access. In this context, the wori< of women artists, like Cassatt, whose canvases shared the stylistic orientation and formal characteristics of other innovative artists oi the day but who consistently depicted interior, domestic spaces fell outside the definition of what "counted" as modernist (Pollock 1988: 50-90; Wolff 1990: 34-66).
It will be noted, of course, that Cassatfs work was not exclusive!)
DILEMMAS IN THE SOCIOLOGY OF ART             261
confined to the private spaces of the bourgeois home.22 Moreover, male colleagues of Cassatt, such as Degas and Renoir, also painted domestic scenes.23 But there is a notable asymmetry at work here, as Pollock (1988: 50-90) demonstrates. The public spaces painted by Cassatt are confined to the settings and subjects of polite society: elegant, bourgeois families in the park, debutantes at the theater, etc. The public spaces represented by Degas and Renoir are not so circumscribed. In addition to the scenes of bourgeois recreation in settings like parks, gardens, and the theater favored by the Impressionists, their canvases also included backstage scenes of dancers, courtesans, mistresses and kept women in settings like the cafe, cabarets, or brothels.24 This asymmetry is particularly significant when we consider that it is as a painter of maternal scenes of mother and child that Cassatt is most often characterized despite the fact that such canvases constitute less than one-third of her total oeuvre (Breeskin 1981). Most importantly, it is as a painter of the maternal that Cassatt has been criticized and derided by modern art historians.25
The Significance of Meaning in Content and Form
The marginalization of women artists in art history has not, of course, meant the absence of the representation of women in art. In fact, through the course of the nineteenth century, a period in which women continued to be largely excluded from academic training, women, in particular the female nude, became ever more present in painting as objects of representation (Parker and Pollock 1981: 115-16). This points to the importance of coming to terms with the meaning of women's increased presence in painting as objects of representation at the same time that women's artistic production continued to be excluded from the modernist canon. Traditionally, sociologists concerned with the problem of meaning in art have focused on content. Content analysis, developed in the 1920s for the study of political propaganda and subsequently extended to the analysis of the mass media and popular culture, has provided valuable insight into the character and substance of cultural forms of communication in the modern world (McCormack 1982). And, as McCormack (1992) has recently argued, content analysis continues to be a useful tool for the study of culture.
u
íl
262
ANNE BOWLER
Nevertheless, the limits of content analysis become clear when we consider the case of a series of exhibitions in late nineteenth-century Britain analyzed in a study by Wolff (1990: 12-33).
Two of the more prominent themes revealed by a content analysis of some of the more well-known works of British painting and literature in the mid to late nineteenth century center around the cult of domesticity and trope of the "fallen woman." Works depicting the sanctity of family life and the moral charge of women as mothers and wives received praise from the influential criticism of Ruskin (Wolff 1990: 13-14).26 A series of exhibitions organized by social reformers for the poor of London's East End in the 1880s emphasized Old Master portraits of the Madonna and Child. Existing documentation records sizable numbers of viewers: from 10,000 in 1881 to 76,000 in 1892 (Wolff 1990: 21-2). At the same time, portraits of the "fallen woman" in painting, literature and the theater were neither uncommon nor unpopular. One canvas, Past and Present by Augustus Egg, depicting the fate of the unfaithful wife as a homeless prostitute, for example, drew both crowds of viewers and shocked reactions from the press when shown at the Royal Academy in 1858 (Wolff 1990: 26).
Confined to the analysis of content, it would be tempting to conclude that, taken together, the themes of domestic life and the fallen woman formed a seamless web of moral meaning in the cultural life of nineteenth-century Britain: the exaltation of the bourgeois family on the one hand and dire warnings of the consequences for sexual transgression on the other. This conclusion, however, is complicated by a formal convention characteristic of some of the more successful painters of the fallen woman and female sexuality more generally. As Wolff observes, an artisťs adoption of a neoclassical mode of representation allowed for the portrayal of female sexuality in ways which open up the possibility of a different reading. Nude figures set in Ancient Greece or Rome rendered the female body an object of exoticism displaced from the moral dangers of modern life. Adopting the stylistic conventions of neoclassicism transformed the body into an expression of the classically derived formal values of harmony, balance and order. Babylonian Marriage Market (1875) by Edwin Long, for example, which sold at Christie's in 1882 for a sale-room record of over six thousand pounds, featured a number of scantily clad young women being sold off in order of beauty to potential husbands. Praised by
DILEMMAS IN THE SOCIOLOGY OF ART            263
Ruskin for its "great merit," it received favorable critical responses in the same journals which had objected to Egg's Past and Present (Wolff 1990:27). The representation of women constitutes the focus (or content) of each painting. But what in one context appears to have signified moral instruction in another context appears to have signified sensual pleasure removed from contemporary mores. In neither case is the ultimate "meaning" of the painting certain. We cannot know for sure how individual viewers responded to either canvas. Rather, the goal of the analysis is to map out a field of possible meanings available within the network of various factors, both social and aesthetic.27
Conclusion
This chapter has presented two major arguments relevant to the development of methodological strategies adequate for a sociology of art: the importance of the autonomy of art as an object of analysis in its own right and the need for sociologists to take seriously the question of meaning. In practice, this does not imply the abandonment or rejection of a "scientific" approach to the study of art. It does, however, call for a "shift in the center" which no longer privileges the analysis of social over aesthetic factors or claims an objective status for itself by abdicating questions of meaning. What this entails is the legitimization within sociology of a number of diverse analytical and interpretive methods whose importance cannot be decided by reference to traditional scientific standards alone. The potential strength of the sociology of art lies not in attempting to make the subdiscipline adhere to a single theoretical framework or set of methodological principles. We can no longer be satisfied with the artificial separation that has existed between the study of society and the study of art. This is the challenge for the sociology of art.
Notes
An earlier version of this paper was presented at the 87th Annual Meeting of the American Sociological Association, August 1992. For their helpful comments and suggestions, I would like to thank Vera Zolberg, Jeff Goldfarb and David Weisberg.
264
ANNE BOWLER
1    Griswold (1987a), Zolberg (1990), and Wolff (1992) have all made this observation.
2    Becker (1982) acknowledged this point in the preface to Art Worlds with the following statement: "it might be reasonable to say that what I have done here is not the sociology of art at all, but rather the sociology of occupations applied to artistic work" (p. xi).
3    An eloquent statement of this problem has been made in Raymond Williams (1989: 165): "What at last came through, theoretically, in the significant new keywords of 'culture' and 'society', was the now familiar model: of the arts on the one hand, the social structure on the other, with the assumption of significant relations between them."
4    Recent evidence,of this can be seen in Alexander's (1990) introduction to a volume of essays on culture and society which positions the autonomy of culture at the center of analysis.
5    See Bürger (1984) for an analysis of the radical doctrine of aesthetic autonomy advanced by aestheticist movements of the nineteenth century. For a current example of the attempt by art historians and literary critics, to bracket social and political questions, see my analysis (Bowler 1991) of Italian futurism and fascism.
6    There is another definition of autonomy central to both the early and later work of the Frankfurt School: the critical capacity of the autonomous work to resist domination (Adorno 1984, 1988; Goldfarb 1982) and the defense of the differentiation of art as an autonomous sphere as part of the "incomplete project of modernity" (Habermas 1981).
7    In addition to Williams, this is a point that is increasingly made in a number of different fields, including the social sciences. For an excellent example in anthropology, see Abu-Lughod (1993).
8    See Swidler (1986) on this issue.
9    See, for example, Williams's (1973) analysis of the changing significance of artistic-literary representations of urban and rural experience in England from the late sixteenth century to the mid-twentieth century. More generally, the construction of meaning in art and literature as a dynamic interaction between specific historical conditions of production and reception (audience, the individual reader, critical response, etc.) has been the center of work by Bakhtin (1981) and reception theory, most notably, Jauss (1982). More recently, this point has been taken up by American sociologists like Griswold (1987a) who combine institutional and interpretive approaches to culture.
10    Griswold (1987a, b) and Wolff (1982, 1990, 1992) are exemplary of this attempt to forge links between institutional and interpretive analyses.
11    It will be noted that Wutiinow's use of the term poststructuralist is an idiosyncratic one within which he groups together the work of Douglas», Foucault and Habermas (Wuthnow 1987: 50). Legitimate objections may be raised about this system of classification. What is at issue here, however.
DILEMMAS IN THE SOCIOLOGY OF ART               265
is Wutiinow's definition of the salient characteristics of a poststructuralist approach to culture.
12    Wuthnow acknowledges that poststructuralist theory addresses the problem of meaning in so far as it analyzes the ways in which systems of meaning work. Nevertheless, Wuthnow (1987) repeatedly insists that poststructuralist approaches involve the "de-emphasis" on the problem of meaning (p. 53), and a "shift away from the problem of meaning" (p. 60). More problematic perhaps is the fact that poststructuralism rejects the distinction Wuthnow makes here between how meaning is constructed and meaning per se.
13    The most enduring examples of this type of study are probably the early analyses of British youth subcultures undertaken by Hall and lefferson (1976) and Hebdige (1979). For an overview of the theoretical and empirical issues central to the cultural studies approach which came out of the Birmingham Centre, see the excellent discussion by Hall (1992).
14    A recent example of this tendency appears in a volume on women viewers' responses to popular television (Brown 1990). Invoking Bakhtin's concept of the carnivalesque to interpret women's responses to soap opera does not, by itself, form a convincing foundation from which to conclude that the "feminine discourse" of soap opera watchers constitutes the subversion of dominant, patriarchal social norms (Brown, 1990: 183-98).
15    Derrida and De Man are perhaps the most famous examples of this tendency.
16    With respect to painting, see the important studies by Parker and Pollock (1981), Pollock (1980,1988), Nead (1988, 1992) and Wolff (1990). See also the collections of essays on feminist art criticism edited by Betterton (1987) and Raven et al. (1988).
17    It is likely that Cassatt's status as the only American member of the Impressionist group increased her visibility in French art circles. In the catalog accompanying the second one-person exhibition of her work at the Paris gallery of her dealer Durand-Ruel in 1893, the critic André Mellario wrote "Cassatt is perhaps, along with Whistler, the only artist of eminent talent, personal and distinguished, that America possesses" (quoted in Bullard 1972: 17).
18    On Stillman's role, see Bullard (1972: 19).
19    See the catalogs published by Breeskin in 1970 and 1979. Breeskin's work is important in the context of Lang and Lang's (1990) findings on factors relevant to the survival of artistic reputation. According to Lang and Lang's study, survival of artisitic reputation is dependent, in part, on the existence of survivors willing and able to act as mediators to an artist's posterity (p. 285). It is also important to note that through her relationship with several important American collectors, Cassatt played an instrumental role in the introduction of Impressionist painting to an American audience (see Bullard 1972: 18; Petersen and Wilson 1976: 89; Seldin 1987).