The Strong Program in Cultural Sociology

By Jeffrey C. Alexander and Philip Smith

This essay has appeared in The Handbook of Sociological Theory, edited by Jonathan Turner (New York: Kluwer, 2001), and in Alexander’s The Meanings of Social Life (New York: Oxford, 2004).

1. Introduction

Throughout the world, culture has been doggedly pushing its way onto the center stage of debates not only in sociological theory and research but also throughout the human sciences. As with any profound intellectual shift, this has been a process characterized by leads and lags. In the UK, for example, culture has been making headway since the early 1970s. In the USA, the tide began to turn unmistakably only in the mid-1980s. In continental Europe, it is possible to argue that culture never really went away. Despite this ongoing revival of interest, however, there is anything but consensus among sociologists specializing in the area about just what the concept means and how it relates to the discipline as traditionally understood. These differences of opinion can be usefully explained only partly as empirical reflections of geographical, sociopolitical, or national traditions. More importantly, they are manifestations of deeper contradictions relating to axiomatic and foundational logics in the theory of culture. Pivotal to all these disputes is the issue of ‘cultural autonomy’ (Alexander l990, Smith 1998a). In this paper, we employ the concept of cultural autonomy to explore and evaluate the competing understandings of culture currently available to social theory. We suggest that fundamental flaws characterize most of these models, and we argue for an alternative approach that can be broadly understood as a kind of structural hermeneutics.

Lévi-Strauss (1974) famously wrote that the study of culture should be like the study of geology. According to this dictum, analysis should account for surface variation in terms of deeper generative principles, just as geomorphology explains the distribution of plants, the shape of hills, and the drainage patterns followed by rivers in terms of underlying geology. In this essay, we intend to apply this principle to the enterprise of contemporary cultural sociology in a way that is both reflexive and diagnostic. Our aim is not so much to review the field and document its diversity — although we will indeed conduct such a review — as to engage in a seismographic enterprise that will trace a fault line running right through it. Understanding this fault line and its theoretical implications allows us not only to reduce complexity, but also to transcend the kind of purely taxonomic mode of discourse that so often plagues handbook articles of the present kind. This seismographic principle will provide a powerful tool for getting to the heart of current controversies and understanding the slippages and instabilities that undermine so much of the territory of cultural inquiry. Contra Lévi-Strauss, however, we do not see our structural enquiry as a disinterested scientific exercise. Our discourse here is openly polemical, our language slightly colored. Rather than affecting neutrality, we are going to propose one particular style of theory as offering the best way forward for cultural sociology.

Back to top

2. The Fault Line and its Consequences

The fault line at the heart of current debates lies between “cultural sociology” and the ’sociology of culture.’ To believe in the possibility of a “cultural sociology” is to subscribe to the idea that every action, no matter how instrumental, reflexive or coerced vis-a-vis its external environments (Alexander 1988), is embedded to some extent in a horizon of affect and meaning. This internal environment is one towards which the actor can never be fully instrumental or reflexive. It is, rather, an ideal resource that partially enables and partially constrains action, providing for both routine and creativity, and allowing for the reproduction and transformation of structure (Sewell 1992). Similarly, a belief in the possibility of a “cultural sociology” implies that institutions, no matter how impersonal or technocratic, have an ideal foundation that fundamentally shapes their organization and goals and provides the structured context for debates over their legitimation. When described in the folk idiom of positivism, one could say that the more traditional “sociology of culture” approach treats culture as a dependent variable, whereas in “cultural sociology” it is an ‘independent variable’ that possesses a relative autonomy in shaping actions and institutions, providing inputs every bit as vital as more material or instrumental forces.

Viewed from a distance, the “sociology of culture” offers the same kind of landscape as “cultural sociology.” There is a common conceptual repertoire of terms like values, codes and discourses. Both traditions argue that culture is something important in society, something that repays careful sociological study. Both speak of the recent ‘cultural turn’ as a pivotal moment in social theory. But these resemblances are only superficial. At the structural level we find deep antinomies. To speak of the “sociology of culture” is to suggest that culture is something to be explained — by something else entirely separated from the domain of meaning itself. To speak of the sociology of culture is to suggest that explanatory power lies in the study of the “hard” variables of social structure, such that structured sets of meanings become superstructures and ideologies driven by more these more “real” and tangible social forces. In this approach, culture becomes defined as a “soft,” not really independent variable: it is more or less confined to participating in the reproduction of social relations.

A notion that has emerged from the extraordinary new field of science studies is the sociologically inspired idea of the “strong program” (e.g. Bloor 1976; Latour and Woolgar 1986). The argument here is that scientific ideas are cultural and linguistic conventions as much as they are simply the results of other, more ‘objective’ actions and procedures. Rather than only “findings” that hold up a mirror to nature (Rorty l979), science is understood as a collective representation, a language game that reflects a prior pattern of sense-making activity. In the context of the sociology of science, the concept of the strong program, in other words, suggests a radical uncoupling of cognitive content from natural determination. We would like to suggest that a strong program might also be emerging in the sociological study of culture. Such an initiative argues for a sharp analytical uncoupling of culture from social structure — which is what we mean by cultural autonomy (Alexander l988, Kane 1992). As compared with the sociology of culture, cultural sociology depends on establishing this autonomy, and it is only via such a strong program that sociologists can illuminate the powerful role that culture plays in shaping social life. By contrast, the “sociology of culture” offers a ‘weak program’ in which culture is a feeble and ambivalent variable. Borrowing from Basil Bernstein (1971), we might say that the strong program is powered by an elaborated theoretical code, whereas the weak program is limited by a restricted code that reflects the inhibitions and habitus of traditional, institutionally oriented social science.

Commitment to a cultural-sociological theory that recognizes cultural autonomy is the single most important quality of a strong program. There are, however, two other defining characteristics that must drive any such approach, characteristics that can be described as methodological. One is the commitment to hermeneutically reconstructing social texts in a rich and persuasive way. What is needed here is a Geertzian ‘thick description’ of the codes, narratives and symbols that create the textured webs of social meaning. The contrast here is to the ‘thin description’ that typically characterizes studies inspired by the weak program, in which meaning is either simply read off from social structure or reduced to abstracted descriptions of reified values, norms, ideology, or fetishism. The weak program fails to fill these empty vessels with the rich wine of symbolic significance. The philosophical principles for this hermeneutic position were articulated by Dilthey (1962), and it seems to us that his powerful methodological injunction to look at the ‘inner meaning’ of social structures has never been surpassed. Rather than inventing a new approach, the deservedly influential cultural analyses of Clifford Geertz can be seen as providing the most powerful contemporary application of Dilthey’s ideas.

In methodological terms, the achievement of thick description requires the bracketing out of wider, non-symbolic social relations. This bracketing out, analogous to Husserl’s phenomenological reduction, allows the reconstruction of the pure cultural text, the theoretical and philosophical rationale for which Ricoeur (1971) supplied in his important argument for the necessary linkage between hermeneutics and semiotics. This reconstruction can be thought of as creating, or mapping out, the ‘culture structures’ (Rambo and Chan 1990) that form one dimension of social life. It is the notion of the culture structure as a social text that allows the well-developed conceptual resources of literary studies — from Aristotle to such contemporary figures as Frye (l957) and Brooks (l985) — to be brought into social science. Only after the analytical bracketing demanded by hermeneutics has been completed — after the internal pattern of meaning has been reconstructed — should social science move from analytic to concrete autonomy (Kane 1992). Only after having created the analytically autonomous culture object does it become possible to discover in what ways culture intersects with other social forces, such as power and instrumental reason in the concrete social world.

This brings us to the third characteristic of a strong program. Far from being ambiguous or shy about specifying just how culture makes a difference, far from speaking in terms of abstract systemic logics as causal processes (à la Lévi-Strauss), we suggest that a strong program tries to anchor causality in proximate actors and agencies, specifying in detail just how culture interferes with and directs what really happens. By contrast, as E.P. Thompson (1978) demonstrated, weak programs typically hedge and stutter on this issue. They tend to develop elaborate and abstract terminological (de)fences that provide the illusion of specifying concrete mechanisms as well as the illusion of having solved intractable dilemmas of freedom and determination. As they say in the fashion business, however, the quality is in the detail. We would argue that it is only by resolving issues of detail — who says what, why, and to what effect — that cultural analysis can become plausible according to the criteria of a social science. We do not believe, in other words, that hard headed and skeptical demands for causal clarity should be confined to empiricists or to those who are obsessively concerned with power and social structure. These criteria also apply to a cultural sociology.

The idea of a strong program carries with it the suggestions of an agenda. In what follows we discuss this agenda. We look first at the history of social theory, showing how this agenda failed to emerge until the 1960s. We go on to explore several contemporary traditions in the social scientific analysis of culture. We suggest that, despite appearances, each comprises a weak program, failing to meet in one way or another the defining criteria we have set forth here. We conclude by pointing to an emerging tradition of cultural sociology, most of it American, which in our view establishes the parameters of a strong program.

Back to top

3. Culture in Social Theory: From the Classics to the 1960s

For most of its history, sociology, both as theory and method, has suffered from a numbness towards meaning. Culturally unmusical scholars have depicted human action as insipidly or brutally instrumental, as if it were constructed without reference to the internal environments of actions that are established by the moral structures of sacred-good and profane-evil (Brooks 1985) and by the narrative teleologies that create chronology (White 1987) and define dramatic meaning (Frye 1957). Caught up in the ongoing crises of modernity, the classical founders of the discipline believed that epochal historical transformations had emptied the world of meaning. Capitalism, industrialization, secularization, rationalization, anomie and egoism — these core processes were held to create confused and dominated individuals, to shatter the possibilities of a meaningful telos, to eliminate the ordering power of the sacred and profane. Only occasionally does a glimmer of a strong program come through in this classical period. Weber’s religious sociology, and most particularly his essay ‘Religious Rejections of the World and Their Directions’ (l958; cf. Alexander 1988) suggested that the quest for salvation was a universal cultural need whose various solutions had forcefully shaped organizational and motivational dynamics in world civilizations. Durkheim’s later sociology, as articulated in critical passages from The Elementary Forms of Religious Life (l968) and in a posthumously recovered course of lectures (Alexander l982), suggested that even contemporary social life had an ineluctable spiritual-cum-symbolic component. While plagued by the weak program symptom of causal ambivalence, the young Marx’s writings on species-being also forcefully pointed to the manner in which non-material forces tied humans together in common projects and destinies. This early suggestion that alienation is not only the reflection of material relationships adumbrated the critical chapter in Capital ‘The Fetishism of Commodities and the Secret Thereof’ which has so often served as an unstable bridge from structural to cultural Marxism in the present day.

The communist and fascist revolutionary upheavals that marked the first half of this century were premised on the same kind of widespread fear that modernity had eroded the possibility of meaning-full sociality. Communist and fascist thinkers attempted to alchemize what they saw as the barren codes of bourgeois civil society into new, re-sacralized forms that could accommodate technology and reason within wider, encompassing spheres of meaning (Smith 1998c). In the calm that descended on the postwar period, Talcott Parsons and his colleagues, motivated by entirely different ideological ambitions, also began to think that modernity did not have to be understood in such a corrosive way. Beginning from an analytical rather than eschatological premise, Parsons theorised that ‘values’ had to be central to actions and institutions if a society was to be able to function as a coherent enterprise. The result was a theory that seemed to many of Parsons’ modern contemporaries to exhibit an idealizing culturalist bias (Lockwood 1992). We ourselves would suggest an opposite reading.

From a strong program viewpoint, Parsonian functionalism can be taken as insufficiently cultural, as denuded of musicality. In the absence of a musical moment where the social text is reconstructed in its pure form, Parsons’ work lacks a powerful hermeneutic dimension. While Parsons theorized that values were important, he did not explain the nature of values themselves. Instead of engaging in the social imaginary, diving into the febrile codes and narratives that make up a social text, he and his functionalist colleagues observed action from the outside and induced the existence of guiding valuations using categorical frameworks supposedly generated by functional necessity. Without a counter-weight of thick description, we are left with a position in which culture has autonomy only in an abstract and analytic sense. When we turn to the empirical world, we find that functionalist logic ties up cultural form with social function and institutional dynamics to such an extent that it is difficult to imagine where culture’s autonomy might lie in any concrete setting. The result was an ingenious systems theory that remains too hermeneutically feeble, too distant on the issue of autonomy to offer much to a strong program. Flawed as the functionalist project was, the alternatives were far worse.

The world in the 1960s was a place of conflict and turmoil. When the Cold War turned hot, macro social theory shifted towards the analysis of power from a one-sided and anti-cultural stance. Thinkers with an interest in macro-historical process approached meaning through its contexts, treating it as a product of some supposedly more ‘real’ social force — when they spoke of it at all. For scholars like Barrington Moore and C. Wright-Mills and later followers such as Charles Tilly, Randall Collins and Michael Mann culture must be thought of in terms of self-interested ideologies, group process and networks rather than in terms of texts. Meanwhile, during the same period, micro-sociology emphasized the radical reflexivity of actors. For such writers as Blumer, Goffman and Garfinkel, culture forms an external environment in relation to which actors formulate lines of action that are ‘accountable’ or give off a good ‘impression’. We find precious little indication in this tradition of the power of the symbolic to shape interactions from within, as normative precepts or narratives that carry an internalized moral force.

Yet during this same period of the 1960s, at the very moment when the halfway cultural approach of functionalism was disappearing from American sociology, theories that spoke forcefully of a social text began to have enormous influence in France. Through creative misreadings of the structural linguistics of Saussure and Jacobson, and bearing a (carefully hidden) influence from the late Durkheim and Marcel Mauss, thinkers like Lévi-Strauss, Roland Barthes and the early Michel Foucault created a revolution in the human sciences by insisting on the textuality of institutions and the discursive nature of human action. When viewed from a contemporary strong program perspective, such approaches remain too abstracted; they also typically fail to specify agency and causal dynamics. In these failings they resemble Parsons’ functionalism. Nevertheless, in providing hermeneutic and theoretical resources to establish the autonomy of culture, they constituted a turning point for the construction of a strong program. In the next section, we discuss how this project has been derailed by a succession of weak programs that continue to dominate research on culture and society today.

Back to top

4. Weak Programs in Contemporary Cultural Theory

One of the first research traditions to apply French nouvelle vague theorising outside of the hothouse Parisian environment was the Centre for Contemporary Cultural Studies, also known as the Birmingham School. The master-stroke of the school was to meld ideas about cultural texts onto the Neo-Marxist understanding that Gramsci established about the role played by cultural hegemony in maintaining social relations. This allowed exciting new ideas about how culture worked to be applied in a flexible way to a variety of settings — all the while without letting go of comforting old ideas about class domination. The result was a “sociology of culture” analysis which tied cultural forms to social structure as manifestations of ‘hegemony’ (if the analyst didn’t like what they saw) or ‘resistance’ (if they did). At its best, this mode of sociology could be brilliantly illuminating. Paul Willis’s (1977) ethnographic study of working class school kids was outstanding in its reconstruction of the zeitgeist of the ‘lads’. Stuart Hall et al.’s (1978) classic study of the moral panic over mugging in 1970s Britain, Policing the Crisis, managed in its early pages to decode the discourse of urban decay and racism that underpinned an authoritarian crackdown. In these ways, Birmingham work approached a “strong program” in its ability to recreate social texts and lived meanings. Where it fails, however, is in the area of cultural autonomy (Sherwood, et al., 1993). Notwithstanding attempts to move beyond the classical Marxist position, neo-Gramscian theorizing exhibits the tell-tale weak program ambiguities over the role of culture that plague the luminous Prison Notebooks themselves. Terms like ‘articulation’ and ‘anchoring’ suggest contingency in the play of culture. But this contingency is often reduced to instrumental reason (in the case of elites ‘articulating’ a discourse for hegemonic purposes) or to some kind of ambiguous systemic or structural causation (in the case of discourses being ‘anchored’ in relations of power).

Failure to grasp the nettle of cultural autonomy and quit the “sociology of culture”-driven project of ‘Western Marxism’ (Anderson 1979) contributed to a fateful ambiguity over the mechanisms through which culture links with social structure and action. There is no clearer example of this latter process than in Policing the Crisis itself. After building up a detailed picture of the mugging panic and its symbolic resonances, the book lurches into a sequence of insistent claims that the moral panic is linked to the economic logic of capitalism and its proximate demise; that it functions to legitimate law and order politics on streets that harbor latent revolutionary tendencies. Yet the concrete mechanisms through which the incipient crisis of capitalism (has it arrived yet?) are translated into the concrete decisions of judges, parliamentarians, newspaper editors and police officers on the beat are never spelled out. The result is a theory which, despite a critical edge and superior hermeneutic capabilities to classical functionalism, curiously resembles Parsons in its tendency to invoke abstracted influences and processes as adequate explanation for empirical social actions.

In this respect, by contrast to the Birmingham School the work of Pierre Bourdieu has real merits. While many Birmingham-style analyses seem to lack any clear application of method, Bourdieu’s oeuvre is resolutely grounded in middle range empirical research projects of both a qualitative and quantitative nature. His inferences and claims are more modest and less manifestly tendentious. In his best work, moreover, such as the description of a Kabyle house or a French peasant dance (Bourdieu 1962, 1977), Bourdieu’s thick description abilities show that he has the musicality to recognize and decode cultural texts that is at least equal to that of the Birmingham ethnographers. Despite these qualities, Bourdieu’s research can also best be described as a weak program dedicated to the sociology of culture rather than cultural sociology. Once they have penetrated the thickets of terminological ambiguity that always mark out a weak program, commentators agree that in Bourdieu’s framework culture has a role in ensuring the reproduction of inequality rather than permitting innovation (Alexander 1995, Honneth 1986, Sewell 1992). As a result culture, working through habitus, operates more as a dependent variable than an independent one. It is a gearbox, not an engine. When it comes to specifying exactly how the process of reproduction takes place, Bourdieu is vague. Habitus produces a sense of style, ease and taste. Yet to know just how these influence stratification something more would be needed – a detailed study f concrete social settings where decisions are made and social reproduction ensured (cf. Lamont 1992). We need to know more about the thinking of gatekeepers in job interviews and publishing houses, the impact of classroom dynamics on learning or the logic of the citation process. Without this ‘missing link,’ we are left with a theory that points to circumstantial homologies but cannot produce a smoking gun.

Bourdieu’s understanding of the links of culture to power also falls short of demanding strong program ideals. For Bourdieu, stratification systems make use of status cultures in competition with each other in various fields. The semantic content of these cultures has little to do with how society is organized. Meaning has no wider impact. While Weber, for example, argued that forms of eschatology have determinate outputs on the way that social life is patterned, for Bourdieu cultural content is arbitrary and without import. In his formulation there will always be systems of stratification defined by class, and all that is important for dominant groups is to have their cultural codes embraced as legitimate. In the final analysis, what we have here is a Veblenesque vision in which culture provides a strategic resource for actors, an external environment of action, rather than a text that shapes the world in an immanent fashion. People use culture, but they don’t seem to really to care about it.

Michel Foucault’s works, and the poststructural and postmodern theoretical program they have initiated, provides the third weak program we discuss here. Despite its brilliance, what we find here, yet again, is a body of work wrought with the tortured contradictions that indicate a failure to grasp the nettle of a strong program. On the one hand, Foucault’s (1970, 1972) major theoretical texts, The Archaeology of Knowledge and The Order of Things, provide important groundwork for a strong program with their assertion that discourses operate in arbitrary ways to classify the world and shape knowledge formation. His empirical applications of this theory should also be praised for assembling rich historical data in a way which approximates to the reconstruction of a social text. So far, so good. Unfortunately, there is another hand at work. The crux of the issue is Foucault’s genealogical method; his insistence that power and knowledge are fused in power/knowledge. The result is a reductionist line of reasoning akin to functionalism (Brenner 1994), where discourses are homologous with institutions, flows of power, and technologies. Contingency of those texts. This binding of discourse to social structure, in other words, leaves no room for understanding how an autonomous cultural realm hinders or assists actors in judgement, critique or in the provision of transcendental goals that texture social life. Foucault’s world is one where Nietzsche’s prison house of language finds its material expression with such force that no room is left for cultural autonomy, and by implication, the autonomy of action. Responding to this sort of criticism, Foucault attempted to theorise self and resistance in his later work. But he did so in an ad hoc way, seeing acts of resistance as random dysfunctions (Brenner 1994: 698) or unexplained self-assertions. These late texts do not work through the ways that cultural frames might permit ‘outsiders’ to produce and sustain opposition to power.

In the currently most influential stream of work to come out of the Foucaultian stable, we can see that the latent tension between the Foucault of the Archaeology and Foucault’s genealogical avatar has been resolved decisively in favor of an anti-cultural mode of theory. The proliferating body of work on ‘governmentality’ centers on the control of populations (Miller and Rose 1990; Rose 1993), but does so through an elaboration of the role of administrative techniques and expert systems. To be sure, there is acknowledgment that ‘language’ is important, that government has a ‘discursive character’. This sounds promising, but on closer inspection we find that ‘language’ and ‘discourse’ boil down to dry modes of technical communication (graphs, statistics, reports etc.) that operate as technologies to allow ‘evaluation, calculation, intervention’ at a distance by institutions and bureaucracies (Miller and Rose 1990: 7). There is little work here to recapture the more textual nature of political and administrative discourses. No effort is made to go beyond a ‘thin description’ and identify the broader symbolic patterns, the hot, affective criteria through which policies of control and coordination are appraised by citizens and elites alike. Here the project of governmentality falls short of the standards set by Hall et al., which at least managed to conjure up the emotive spirit of populism in Heath era Britain.

Research on the ‘production and reception of culture’ marks the fourth weak program we will identify. Unlike those we have just discussed, it is one that lacks theoretical bravura and charismatic leadership. For the most part, it is characterized by the unsung virtues of intellectual modesty, diligence, clarity and a studious attention to questions of method. Its numerous proponents make sensible, middle range empirical studies of the circumstances in which ‘culture’ is produced and consumed (for overview see Crane 1992). For this reason, it has become particularly powerful in the United States, where these kinds of properties assimilate best to professional norms within sociology. The great strength of this approach is that it offers explicit causal links between culture and social structure, thus avoiding the pitfalls of indeterminacy and obfuscation that have plagued more theoretically ambitious understandings. Unfortunately, this intellectual honesty usually serves only to broadcast a reductionist impulse that remains latent in the other approaches we have examined. The insistent aim of study after study (e.g., Blau 1989, Peterson l985) seems to be to explain away culture as the product of sponsoring institutions, elites or interests. The quest for profit, power, prestige, or ideological control sits at the core of cultural production. Reception, meanwhile, is relentlessly determined by social location. Audience ethnographies, for example, are undertaken to document the decisive impact of class, race and gender on the ways that television programs are understood. Here we find the “sociology of culture” writ large. The aim of analysis is not so much to uncover the impact of meaning on social life and identity formation, but rather to see how social life and identities constrain potential meanings.

While the sociological credentials of such an undertaking are to be applauded, something more is needed if the autonomy of culture is to be recognized — viz. a robust understanding of the codes that are at play in the cultural objects under consideration. Only when these are taken into account can cultural ‘products’ be seen to have internal cultural inputs and constraints. However, in the production of culture approach such efforts at hermeneutic understanding are rare. All too often meaning remains a sort of black box, with analytical attention centered on the circumstances of cultural production and reception. When meanings and discourses are explored, it is usually in order to talk through some kind of fit between cultural content and the social needs and actions of specific producing and receiving groups. Wendy Griswold (1983), for example, shows how the trickster figure was transformed with the emergence of Restoration drama. In the medieval morality play, the figure of ‘vice’ was evil. He was later to morph into the attractive, quick thinking ‘gallant’. The new character was one that could appeal to an audience of young, disinherited men who had migrated to the city and had to depend on their wits for social advancement. Similarly, Robert Wuthnow (1989) argues that the ideologies of the Reformation germinated and took root as an appropriate response to a particular set of social circumstances. He persuasively demonstrates that new binary oppositions emerged in theological discourse — for example those between a corrupt Catholicism and a pure Protestantism. These refracted the politics and social dislocations underlying religious and secular struggles in 16th century Europe.

We have some concerns about singling such work out for criticism, for they are among the best of the genre and approximate to the sort of thick description we advocate. There can be little doubt that Griswold and Wuthnow correctly understand a need to study meaning in cultural analysis. However, they fail to systematically connect its exploration with the problematic of cultural autonomy. For all their attention to cultural messages and historical continuities, they do little to reduce our fear that there is an underlying reductionism in such analysis. The overall effect is to understand meanings as infinitely malleable in response to social settings. A more satisfying approach to Griswold’s data, for example, would recognize the dramatic narratives as inevitably structured by constraining cultural codes relating to plot and character, for it is the combinations between these make any kind of drama a possibility. Similarly, Wuthnow should have been much more sensitive to the understanding of binary opposition advocated by Saussure: it is a precondition of discourse rather than merely a description of its historically specific form. And so to our reading, such efforts as Griswold’s and Wuthnow’s represent narrowly lost opportunities for a decisive demonstration cultural autonomy as a product of culture-structure. In the final section of this chapter, we look for signs of a structuralist hermeneutics that can perhaps better accomplish this theoretical goal.

5. Steps Towards a Strong Program

All things considered, the sociological investigation of culture remains dominated by weak programs characterized by some combination of hermeneutic inadequacy, ambivalence over cultural autonomy, and poorly specified, abstract mechanisms for grounding culture in concrete social process. In this final section, we wish to discuss recent trends in cultural sociology where there are signs that a bona fide strong program might finally be emerging.

A first step in the construction of a strong program is the hermeneutic project of ‘thick description’ itself, which we have already invoked in a positive way. Drawing on Paul Ricoeur and Kenneth Burke, Clifford Geertz (1973) has worked harder than any other person to show that culture is a rich and complex text, with a subtle patterning influence on social life. The result is a compelling vision of culture as webs of significance that guide action. Yet while superior to the other approaches we have considered, this position too has its flaws. Nobody could accuse Geertz of hermeneutic inadequacy or of neglecting cultural autonomy, yet on close inspection his enormously influential concept of thick description seems rather elusive. The precise mechanisms through which webs of meaning influence action on the ground are rarely specified with any clarity. Culture seems to take on the qualities of a transcendental actor (Alexander 1987). So in terms of the third criterion of a strong program that we have specified — causal specificity — the program initiated by Geertz runs into trouble. One reason is the later Geertz’s reluctance to connect his interpretive analyses to any kind of general theory. There is a relentless emphasis on the way that the local explains the local. He insists that societies, like texts, contain their own explanation. Writing the local, as a consequence, comes into play as a substitute for theory construction. The focus here is on a novelistic recapitulation of details, with the aim of analysis being to accumulate these and fashion a model of the cultural text within a particular setting. Such a rhetorical turn has made it difficult to draw a line between anthropology and literature, or even travel writing. This in turn has made Geertz’s project vulnerable to takeover bids. Most notably, during the 1980s the idea that society could be read like a text was taken over by poststructural writers who argued that culture was little more than contending texts or ‘representations’ (Clifford 1988) and that ethnography was either allegory, fantasy or biography. The aim of analysis now shifted to the exposition of professional representations and the techniques and power relations behind them. The resulting program has been one that has told us a good deal about academic writing, ethnographic museum displays and so on. It helps us to understand the discursive conditions of cultural production but has almost given up on the task of explaining ordinary social life or the possibility of a general understanding. Not surprisingly, Geertz enthusiastically devoted himself to the new cause, writing an eloquent text on the tropes through which anthoropologists construct their ethnographic authority (Geertz 1988). As the text replaces the tribe as the object of analysis, cultural theory begins to look more and more like critical narcissism and less and less like the explanatory discipline that Dilthey so vividly imagined.

Inadequate as it may be, the work of Geertz provides a springboard for a strong program in cultural analysis. It indicates the need for the explication of meaning to be at the center of the intellectual agenda and offers a vigorous affirmation of cultural autonomy. What is missing, however, is a theory of culture that has autonomy built into the very fabric of meaning as well as a more robust understanding of social structure and institutional dynamics. We suggest, following Saussure, that a more structural approach towards culture helps with the first point. In addition, it initiates the movement towards general theory that Geertz avoids. In short, it can recognize the autonomy and the centrality of meaning, but does not develop a hermeneutics of the particular at the expense of a hermeneutics of the universal. We return to the promise of such a structural hermeneutics below.

As the eighties turned into the nineties, we saw the revival of ‘culture’ in American sociology and the declining prestige of anti-cultural forms of macro and micro thought. This strand of work, with its developing strong program characteristics, offers the best hope for a truly cultural sociology finally to emerge as a major research tradition. To be sure a number of weak programs organized around the “sociology of culture” remain powerful, perhaps dominant, in the U.S. context. One thinks in particular of studies of the production, consumption and distribution of culture which (as we have seen) focus on organizational and institutional contexts rather than content and meanings (e.g. Blau 1989; Peterson 1985). One also thinks of work inspired by the western Marxist tradition that attempts to link cultural change to the workings of capital, especially in the context of urban form (e.g. Davis 1992; Gottdeiner 1995). The neo-institutionalists (see DiMaggio and Powell 1991) see culture as significant, but only as a legitimating constraint, only as an external environment of action, not as a lived text as Geertz might (see Friedland and Alford 1991). And of, course, there are numerous U.S.-based apostles of British Cultural Studies (e.g. Fiske 1987; Grossberg, Nelson and Treichler 1991), who combine virtuoso hermeneutic readings with thin, stratification-oriented forms of quasi-materialist reduction. Yet, it is equally important to recognize that there has emerged a current of work that gives to meaning-full and autonomous texts a much more central place (for a sample see Smith 1998b). These contemporary sociologists are the ‘children’ of an earlier generation of culturalist thinkers — Geertz, Bellah (1970; cf. Alexander and Sherwood forthcoming), Turner (1974) and Sahlins (l976) foremost among them — who wrote against the grain of sixties and seventies reductionism and attempted to demonstrate the textuality of social life and the necessary autonomy of cultural forms. In contemporary scholarship, we are seeing efforts to align these two axioms of a strong program with the third imperative of identifying concrete mechanisms through which culture does its work.

Responses to the question of transmission mechanisms have been decisively shaped, in a positive direction, by the American pragmatist and empiricist traditions. The influence of structural linguistics on European scholarship sanctioned a kind of cultural theory that paid little attention to the relationship between culture and action (unless tempered by the dangerously ‘humanist’ discourses of existentialism or phenomenology). Simultaneously, the philosophical formation of writers like Althusser and Foucault permitted a dense and tortured kind of writing, where issues of causality and autonomy could be circled around in endless, elusive spirals of words. By contrast, American pragmatism has provided the seedbed for a discourse where clarity is rewarded; where it is believed that complex language games can be reduced to simpler statements; where it is argued that actors have to play some role in translating cultural structures into concrete actions and institutions. While the influence of pragmatism has reached American cultural sociologists in a diffuse way, its most direct inheritance can be seen in the work of Ann Swidler (1986), William Sewell (1992), Mustafa Emirbayer and his collaborators (e.g. Emirbayer and Goodwin 1996; Emirbayer and Mische 1998), and Gary Alan Fine (1987), where efforts are made to relate culture to action without recourse to the materialistic reductionism of Bourdieu’s praxis theory.

Other forces have also played a role in shaping the emerging strong program in American cultural sociology. Because these are more closely related than the pragmatists to our argument that a structuralist hermeneutics is the best way forward, we will expand on them here. Pivotal to all such work is an effort to understand culture not just as a text (à la Geertz) but rather as a text that is underpinned by signs and symbols that are in patterned relationships to each other. Writing in the first decades of the 20th century, Durkheim and his students such as Hertz and Mauss understood that culture was a classification system consisting of binary oppositions. At the same time Saussure was developing his structural linguistics, arguing that meanings were generated by means of patterned relationships between concepts and sounds. A few decades later, Lévi-Strauss was to pull these linguistic and sociological approaches to classification together in his pioneering studies of myth, kinship, and totemism. The great virtue of this synthesis was that it provided a powerful way for understanding the autonomy of culture. Because meanings are arbitrary and are generated from within the sign system, they enjoy a certain autonomy from social determination — just as the language of a country cannot be predicted from the knowledge that it is capitalist or socialist, industrial or agrarian. Culture now becomes a structure as objective as any more material social fact.

With the themeatics of the ‘autonomy of culture’ taking center stage in the 1980s, there was a vigorous appreciation of the work of the late-Durkheim, with his insistence on the cultural as well as ‘functional’ origins of solidarity (for a review of this literature see Emirbayer 1996, Smith and Alexander 1996). The felicitous, but not altogether accidental, congruence between Durkheim’s opposition of the sacred and the profane and structuralist theories of sign-systems enabled insights from French theory to be translated into a distinctively sociological discourse and tradition, much of it concerned with the impact of cultural codes and codings. Numerous studies of boundary maintenance, for example, reflect this trend (for a sample see Lamont and Fournier 1993), and it is instructive to contrast them with more reductionist weak program alternatives about processes of ‘Othering’. Emerging from this tradition has been a focus on the binary opposition as a key tool for asserting the autonomy of cultural forms (see Alexander and Smith 1993, Smith l991, Edles l998, Magnuson l997).

Further inspirations for structural hermeneutics within a strong program for cultural theory have come from anthropology. The new breed of symbolic anthropologists — in addition to Geertz, most notably Mary Douglas (l966), Victor Turner (l974) and Marshall Sahlins (l976, l981) — took on board the message of structuralism, but tried to move it in new directions.

Postmodernisms and post-structuralisms have also played their role, but in an optimistic guise. The knot between power and knowledge that has stunted European weak programs has been loosened by American postmodern theorists like Steven Seidman (1988). For postmodern pragmatistic philosophers like Richard Rorty (e.g. 1989), language tends to be seen as a creative force for the social imaginary rather than as Nietzsche’s prison house. As a result, discourses and actors are provided with greater autonomy from power in the construction of identities.

These trends are well known. But there is also an interdisciplinary dark horse to which we wish to draw attention. In philosophy and literary studies, there has been growing interest in narrative and genre theory. Cultural sociologists such as Robin Wagner-Pacifici (l986, l994, 2000, with Barry Schwartz 1991), Margaret Somers (1995), Wendy Griswold (1983), Ronald Jacobs (1996, 2000), Agnes Ku (1999), William Gibson (l994) and the authors of this article are now reading literary theorists like Northrup Frye, Peter Brooks, and Fredric Jameson, historians like Hayden White and Aristotelian philosophers like Ricoeur and MacIntyre (cf. Lara l998). The appeal of this theory lies partially in its affinity for a textual understanding of social life. The emphasis on teleology carries with it some of the interpretive power of the classical hermeneutic model. This impulse towards reading culture as a text is complemented, in such narrative work, by an interest in developing formal models that can be applied across different comparative and historical cases. In other words, narrative forms such as the morality play or melodrama, tragedy and comedy can be understood as ‘types’ that carry with them particular implications for social life. The morality play, for example, does not seem to be conducive to compromise (Wagner-Pacifici 1986, l994). Tragedy can give rise to fatalism (Jacobs 1996) and withdrawal from civic engagement, but it can also promote moral responsibility (Alexander forthcoming, Eyerman forthcoming). Comedy and romance, by contrast, generate optimism and social inclusion (Jacobs and Smith 1997, Smith 1994). Irony provides a potent tool for the critique of authority and reflexivity about dominant cultural codes, opening space for difference and cultural innovation (Jacobs and Smith 1997, Smith 1996).

A further bonus for this narrative approach is that cultural autonomy is assured (e.g., in the analytic sense, see Kane 1992). If one takes a structuralist approach to narrative (Barthes l970), textual forms are seen as interwoven repertoires of characters, plot lines and moral evaluations whose relationships can be specified in terms of formal models. Narrative theory, like semiotics, thus operates as a bridge between the kind of hermeneutic inquiry advocated by Geertz and the impulse towards general cultural theory. As Northrop Frye recognized, when approached in a structural way narrative allows for the construction of models that can be applied across cases and contexts but at the same time provides a tool for interrogating particularities.

It is important to emphasize that while meaning-full texts are central in this American strand of a strong program, wider social contexts are not by any means necessarily ignored. In fact, the objective structures and visceral struggles that characterize the real social world are every bit as important as in work from the weak programs. Notable contributions have been made to areas such as censorship and exclusion (Beisel 1993), race (Jacobs 1996), sexuality (Seidman 1988), violence (Gibson 1994; Smith 1991, 1996; Wagner-Pacifici 1994), and failed socio-historical projects for radical transformation (Alexander l995, forthcoming). These contexts are treated, however, not as forces unto themselves that ultimately determine the content and significance of cultural texts. Rather, they are seen as institutions and processes that refract cultural texts in a meaning-full way. They are arenas in which cultural forces combine or clash with material conditions and rational interests to produce particular outcomes (Ku l999, Smith l996). And beyond this they are seen as cultural metatexts themselves, as concrete embodiments of wider ideal currents.

Back to top

6. Conclusions

We have suggested here that structuralism and hermeneutics can be made into fine bedfellows. The former offers possibilities for general theory construction, prediction and assertions of the autonomy of culture. The latter allows analysis to capture the texture and temper of social life. When complemented by attention to institutions and actors as causal intermediaries, we have the foundations of a robust cultural sociology. The argument we have made here for an emerging strong program has been slightly polemical in tone. This does not mean we disparage efforts to look at culture in other ways. If sociology is to remain healthy as a discipline, it should be able to support a theoretical pluralism and lively debate. There are important research questions, in fields from demography to stratification to economic and political life, to which weak programs can be expected make significant contributions. But it is equally important to make room for a genuinely cultural sociology. A first step toward this end is to speak out against false idols, to avoid the mistake of confusing reductionist sociology of culture approaches with a genuine strong program. Only in this way can the full promise of a cultural sociology be realized during the coming century.

Back to top

References