Mike Savage. Handbook of Sociology. Editor: Craig Calhoun, Chris Rojek, Bryan Turner. Sage Publications. 2005.
Introduction
Compared to other disciplines, the Catholicism of sociology is striking. Sociologists spend relatively little time policing their boundaries and worrying about what ‘real’ sociology is (see Urry 1981; Abbott, 2001). Yet, although this openness opens up all sorts of possibilities for the subject as a kind of mobile discipline, it leads to difficulties in defining what the discipline-specific expertise of sociology actually is. What exactly do sociologists know that those in other disciplines do not? What are sociologists particularly good at? This question is not always easy for sociologists to address: its methods are not distinct to it, most of which are better entrenched in other disciplines. Documentary analysis is lionized by historians, ethnography by anthropologists, and survey methods are used throughout the social sciences. Most of its methods for analysing survey data are shared with other disciplines: there is no sociological equivalent to econometrics. Sociological theory blurs into more amorphous forms of social theory. Attempts to develop distinctively sociological theory, evident in the classic writing of Weber, Durkheim, Simmel and Parsons, have given way to hybrid theory, as likely to draw on literary and cultural theory and elements of philosophy, rather than any particular sociological canon. It is difficult to find any substantive area of sociological inquiry that is not more strongly anchored in the expertise of other disciplines: consider family and kin (anthropology), urban (geography, planning), work and organization (economics, management), politics (political science), economic (economics) and so on.
One exception to this fluidity, I argue in this chapter, is that way that the study of stratification has acted as a unifying force within sociological analysis, as a distinct area where the discipline of sociology claims distinctive preeminence. Classical sociological theorists, notably Marx, Weber, Pareto and, to a lesser degree, Simmel and Durkheim, saw social relationships as fundamentally unequal. Many sociological methods were pioneered as ways of understanding social inequality. Community studies, developed in the UK by Booth and Rowntree, and by Hunter in the United States, were preoccupied by the issue of community power, and the geography of social inequality. After the Second World War, the development of sample surveys was linked to studies of class and mobility (in the UK: Glass, 1954; Goldthorpe and Lockwood, 1968/69; Goldthorpe, 1980; and in the United States: Blau and Duncan, 1967). Methods of survey analysis were initially used to study inequality. Blau and Duncan (1967) used regression and path analysis techniques as a means of understanding the determinants of status attainment in the United States, Goldthorpe (1980) and his associates (Erikson and Goldthorpe, 1992) used log-linear modelling as a means of exploring trends in class mobility. Bourdieu (1984) used correspondence analysis to examine the affinities between class and cultural taste (Bourdieu, 1984). Bott (1956) launched the tradition of social network analysis as a means of understanding class differences in family relationships. And so on.
To be sure, an interest in stratification is not unique to sociology. However, if we adopt Abbott’s (1988) insights into the competitive nature of expertise, with different professional groups struggling against others to maintain jurisdiction over a particular kind of work, we can note that in other disciplines, inequality was seen either as derived from, or as caused by, more fundamental axial concerns. Historians were interested in how class relations generated social change (famously Thompson, 1963 and the myriad debate it inspired), political scientists in how inequalities might affect the polity (Barrington Moore, 1966), and economists in how income differentials were related to the valuation of human capital by market processes (Becker, 1964). In these cases, stratification was relevant to the domain interests of the discipline concerned, but did not stake out its central concern. Sociology became distinctive in putting stratification—its nature, ramifications, causes and consequences—at the heart of its intellectual endeavour. Apparently disparate features of middle and late twentieth century sociology begin to appear more unified and coherent when this point is recognized.
However, although stratification played a core role in defining the discipline’s central concerns for much of the twentieth century, this now looks increasingly fraught. Even though the study of stratification is technically sophisticated, it is also seen as old-fashioned, parochial and self-referential, unable to come to terms with fundamental features of contemporary social change (e.g. Bauman, 1982; Pahl, 1989; Lash and Urry, 1994; Pakulski and Walters, 1996; Crompton, 1998). This challenge is serious and far-reaching, and poses serious questions about the value of sociological perspectives. The second and larger part of this chapter examines this challenge and considers responses to it.
This chapter therefore has four parts. The first part briefly shows how the study of stratification came to play a central role in defining sociological concerns during the middle and later twentieth century through its concern with social relationships. The second part shows how this sociology of social relationships has been undercut over the past two decades in the context of deep-rooted theoretical problems in the study of stratification. This section shows how stratification sociology has increasingly been beset by two powerful critical objections: problems in sustaining an explanatory theory of exploitation, and questions about inter-relating stratification with social and cultural change. Sociologists traditionally saw inequality as produced by relationships of exploitation between social groups, but that this relational approach to stratification has been undercut through criticisms of theories of exploitation. The result is that distinctive sociological perspectives have given way to market-based, economic approaches, and that the marginalization of the sociology of stratification poses distinct problems for the vitality of sociology as a discipline.
The third part of the chapter takes particular stock of debates about exploitation. Although there was considerable virtuosity from the 1980s in developing new approaches to the analysis of exploitation, I argue that they all ran into the problem that axes of inequality could proliferate in a way that seemed to provide a warrant for an anarchic sociology. In the fourth part, I therefore consider how we can think freshly about sociological conceptions of stratification. My main argument here is that we can sidestep debates about exploitation through focusing on the processes by which social resources, assets and capacities are accumulated. By switching our attention in these terms we are better able to align the interests of stratification sociology with broader concerns regarding socio-cultural change that are central to the contemporary sociological imagination. I develop my arguments through a critical engagement with the sociological theory of Pierre Bourdieu, which I think has the most potential here.
The Emergence of the Stratification Paradigm within Sociology
Wagner (2001) has shown how early attempts to define sociology as the ‘science of society’, as championed by nineteenth-century positivists and the classical theorists and institutional leaders of sociology, had largely failed by the 1920s. This positivist vision of sociology required the other social sciences to be subordinate to sociology (since the economy, political system, etc. were seen as parts of a wider society). Such was the entrenched power and institutional strength of (particularly) economics and politics as disciplines in both European and American universities that this was never likely to happen. In addition, an animating force in early sociology saw it as part of a social reform movement, linked to Christian pro-gressivism in the United States (Abbott, 1999) and Fabianism and social planning in the UK (Yeo, 1996), that ultimately required a sociology attuned to practical political issues and concerns, not a sociology of evolutionary social laws and relations. Thus, even the structural functionalism of Parsons, which shored up this classic positivist conception of sociology in the most prestigious American universities into the second part of the twentieth century, rested on shallow foundations. From the 1950s, even as the subject expanded dramatically, sociology fractured, with withering critiques from interactionists and ethno-methodologists (e.g. Garfinkel, 1967), and from those with more radical and practical conceptions of the discipline, such as Marxists and feminists.
In the face of this splintering, sociological champions expounded the virtues of their discipline not in terms of its ability to analyse ‘society’ (however that might be understood), but rather in terms of the peculiar sensibilities they offered to critical inquiry, a project which included their ability to redefine and illuminate issues developed in other disciplines (see for instance C. Wright Mills, 1959; Berger and Luckmann, 1963; Bauman, 1990). Perhaps the single most important attempt to situate sociology’s distinctive critical sensibility was in its sensibility towards stratification, which allowed the subject to be re-positioned as a critical discipline attuned to the fundamentally inegal-itarian nature of modern social relations (Lockwood, 1964; Rex, 1968).
Thus it was that stratification sociology emerged, largely un-announced, as an alternative means of reconciling a discipline in need of some kind of unifying framework. An example is Nisbet’s (1959) influential attempt to redefine the sociological tradition by invoking its five core unit ideas: status, authority, alienation, community and the sacred. Although outwardly concerned to show the conserva-tivism of the discipline, two of these unit ideas—status and authority—were directly concerned with inequality: two more—alienation and community—were related to it, and only the unit idea of the sacred was largely irrelevant to it. This attempt to redefine the sociological project was compatible with the emerging view that sociology was a critical discipline that questioned established and received ways of thinking. A critical perspective involves a deep scrutiny of social beliefs and perceptions that tend to generate views of the social order as holistic communities of shared interests. Whilst the ‘sociology as science of society’ school shared such views, seeing society as essentially a version of ‘community’ (see Therborn, 1976), a commitment to a critical perspective allowed sociologists to define their own distinctive perspective compared to other (supposedly less critical) disciplines.
Sociologists were also able to relate their concerns to political action by anchoring their concerns with stratification to political campaigns. In the UK, links between class politics and sociological concerns could be found in debates about citizenship (Marshall, 1950), cultural change (Hoggart, 1956; Williams, 1958), education policy (Marsden and Jackson, 1962; Halsey et al., 1980). Arguments about social change tended to be couched in terms of discussions about the changing (or unchanging) nature of class (see Savage, 2000 for elaboration). In the American context, sociologists examined the implications of sociological approaches for evaluating the idea that American society was relatively classless (e.g. Blau and Duncan, 1967). Sociology became increasingly seen as composed of writers with distinct claims to analysing inequality (see famously Giddens, 1971, who placed Marx and Weber alongside Durkheim as a key sociologist). For Marx, the extraction of surplus product meant that some social groups could live from the labour of subordinate groups, and this process of surplus extraction coloured all aspects of social and cultural life. Weber’s emphasis, in part based on his Nietzschean awareness of the omnipresence of conflicting values, was on the generative role of power, as individual actors attempt to impose their goals and values, and in the process how resources of class, status and command allow some actors to pursue their objectives over others. Much subsequent social theorizing follows in the footsteps of these two thinkers.
Stratification provided a key conduit linking disparate sociological sub-disciplines. Over the past 30 years there has been increased recognition of the plurality of axes of stratification. Issues of gender and of race and ethnicity have been increasingly recognized to be significant forces generating inequality. In these cases, the initial entry of these areas as legitimate within sociology was linked to the recognition—contested in some cases—that they defined a further axis of inequality irreducible to others. Thus the study of gender (rather than the ‘family’) was introduced into sociology as a result of feminist concerns over structural inequality between men and women (e.g. Oakley, 1973). A similar story can be told for race and ethnic inequalities. Stratification provided one way of defining sociology’s domain concerns in a situation of chronic identity crisis (for a recent version of this argument see Turner and Rojek’s (2001) rejection of a ‘decorative sociology’ in favour of one focused on the topics of ‘scarcity and solidarity’). Stratification defined a set of central concerns linking theory, method and sub-areas that allowed sociologists to engage in some kind of dialogue with a degree of shared understanding. To some extent, the chronic disputes on these issues—between micro and macro perspectives, between Marxists, Weberians and others, between those championing class, gender and race as the main axis of inequality-allowed this situation to persist. Because these disagreements were recognized as ones characterizing different sociological positions, they helped nurture a shared set of concerns and understandings, and which united sociologists against other disciplines.
The Eclipse of the Sociology of Stratification
Over the past two decades, however, this core role for stratification sociology has been called into question. The sub-discipline of stratification itself became increasingly insulated from sociology as a whole. As I note elsewhere (Savage, 1994), for much of the post-war period there were no textbooks on the sociology of stratification in the UK, in large part because it could not be easily differentiated from the discipline as a whole. However, during the 1990s a rush of textbooks and edited collections diagnosed the state of the sociology of stratification (Saunders, 1990; Edgell, 1992; Breen and Rotman, 1995; Lee and Turner, 1996; Pakulski and Walters, 1996; Scott, 1996; Crompton, 1998; Devine, 1998; Savage, 2000; Roberts, 2001). The verdicts regarding the prognoses for stratification varied between supporters, critics and revisionists, but what was most important was the scope that these discussions had for generating insularity in the sub-discipline.
There were clearly deep-rooted concerns. One was the emergence of a new way of defining sociological expertise, which saw the discipline as a kind of academic avant-garde, especially well placed to diagnose current forms of social change. This drew upon elements of classical sociology, notably the concerns of early sociologists to situate their discipline in the context of industrialization and the transition to modernity (Nisbet, 1959; Kumar, 1978), as well as the process- and context-oriented sociology associated with the French Le Play tradition or the American Chicago School (see Savage and Warde, 1993; Abbott, 1999). In a period of significant social change, it seemed incumbent on sociologists to propound new syntheses explaining why older social forms were giving way to new ones, and what the main dimensions and contours of change were. The revival of this genre could be traced back at least to Bell’s (1973) account of post-industrialism, though his subtitle ‘A venture in social forecasting’ indicates that he saw this as a rather novel enterprise for a sociologist. However, during the 1980s this kind of sociology mushroomed, nurtured by concerns regarding post-modernism, the collapse of state socialism, new forms of flexible production, globalization, and so forth (Lash and Urry, 1987, 1994; Bauman, 1992; Giddens, 1990; Beck, 1992). During the 1980s it seemed almost mandatory for any (British) sociologist worth his or her salt to come up with some new account of epochal social change. The appeal of sociology was here redefined as its ability to quickly interpret current changes, using whatever new and challenging ideas were being developed in the humanities or (more rarely) other social sciences. In the process, sociological theory became almost completely detached from the day-to-day analyses of empirical sociologists, who became more likely to look for alternative kinds of social theory originating in other disciplines: the advocacy of rational choice theory amongst a number of quantitative sociologists being a case in point. Therefore, although during the 1980s and 1990s empirical stratification sociologists in Europe and the US continued to innovate (e.g. Sorensen, 1986; Erikson and Goldthorpe, 1992; Wright, 1997), their research failed to attract much interest outside the specialist research community.
This excitement about the potential of using sociology as a diagnostic of social change took place at exactly the same time that serious intellectual problems were becoming more manifest in stratification sociology’s attempts to understand social change. After all, it had been one of the main claims of stratification sociology that it offered a way of understanding change in terms of processes of class formation (see Abrams, 1981; Savage et al., 1992). This problematic, loosely rooted in the work of Marx and Weber, related social change to the demographic, social, cultural and political organization of social groups, and thereby offered a historically specific account of how inequalities could generate the formation of social groups with shared identities, who would then engage in forms of action and affect historical change. Whilst there were more (generally Marxist) or less (generally Weberian) determinist versions of this approach, some common concerns could easily be seen, and they informed the work of historical writers such as Anderson (1974), Barrington Moore (1966) and the stratification sociology of Goldthorpe (1980). However, this approach depended on showing how social actions, identities and class relations were mutually inter-related. In a period of muted overt class conflict and marked weakness of social solidarities based on class, this tradition offered little purchase on the analysis of contemporary social trends (see Bauman, 1982 for an especially thoughtful account here). In addition, insofar as class formation did appear important it could be interpreted as backward-looking (e.g. Calhoun, 1982) and as failing to have a handle on contemporary social change. However subtle the story might be, it looked difficult to talk of class formation, at least in the classical sense, in any part of the globe at the end of the twentieth century. This partly explains why some stratification sociologists changed their focus completely and emphasized the lack of any association between class positions and group identity. Defenders of class analysis, such as Goldthorpe and his associates in the UK, did not claim that class consciousness and identity were marked, but borrowed the tools of rational choice theory to show how class effects could be produced by rational (and not necessarily class-aware) individual actors (Goldthorpe, 2000; and see the discussion in Savage, 2000). Alternatively, Bourdieu’s cultural sociology was deployed to show how cultural and symbolic capital led to the ‘dis-identification of class’ (see, for example, Skeggs, 1997; Charlesworth, 2000).
However, at the same time as class formation appeared weak and relatively unimportant, sociologists and other social theorists became increasingly interested in the issue of ‘identity’ (Calhoun, 1994; Rutherford, 1994), though now usually seen as independent of class relations. A major claim here was the argument that identity politics could not be related to prior determinations, but that the ‘struggle for recognition’ (Honneth, 1985) was an autonomous area of social conflict which could not be related in any obvious way to social inequalities. Indeed, some writers such as Nancy Fraser (1995) emphasized the irreconcilability of the politics of distribution and recognition. With these developments, one of the main ‘promises’ of stratification sociology, its potential for highlighting the relationship between structure and agency, was undercut.
These problems were compounded by a fundamental problem for classical sociological approaches to class. Classical sociological theory had emphasized that social inequalities were fundamentally produced by divisive social relationships. The advantages of some groups could only be understood as related to the disadvantages of others. This relational perspective took different forms, ranging from Marx’s labour theory of value, functionalist accounts of the social value of inequality, Weberian views regarding the omnipresence of competition and its differentiation into class, status and party, and so on. These all offered distinctively sociological ways of understanding inequality, which did not reduce inequalities to purely economic processes, but anchored them in social relationships. However, over the past 30 years all have been largely discredited—or at the least have been subject to withering intellectual critiques, leaving sociologists little alternative but to fall back on the kind of market-based explanations of inequality that have been developed within economics. Let us consider this issue further in the next section.
Defining Inequality: The Limits of Relational Approaches
Social inequality exists in all known societies. These inequalities have numerous dimensions. They include economic inequalities pertaining to the uneven distribution of material conditions of life; status or cultural inequalities relating to the cultural approbation of different kinds of practices and people; and political inequalities relating to the rights and privileges different groups command through various claims to citizenship and entitlement. There are numerous axes of inequality, setting genders, classes, ethnicities, generations, residential groups, sexualities and those with different bodily abilities and aptitudes against each other. The problems of the sociology of stratification cannot therefore be attributed to the fact that social inequality is hard to spot. Rather, the difficulties lie in being able to provide a convincing sociological theory specifying how social relations systematically generate inequality. This is evident if we look at three influential sociological accounts of social inequality, all of which commanded support in the mid-twentieth century. First, the functionalist view can be traced back to Durkheim, and was elaborated by Parsons and in particular by Davis and Moore (1945). This related inequality to the need to reward society’s most functionally important occupations. It was thus claimed that social inequalities were generally legitimate and based on shared social values and norms. Economic inequalities could be seen as generally socially sanctioned. One problem with this argument is that there is no obvious popular support for income inequalities and public opinion does not on the whole endorse major income inequalities (see Savage, 2000). Insofar as there is acceptance of inequality, this is generally a blanket endorsement of meritocratic values rather than a specific recognition that particular occupations are more socially valued (e.g. Kluegel et al, 1986). A key element of contemporary acceptance of economic inequalities is the widespread (though not universal) belief that markets are acceptable means for distributing incomes. Brint’s (1993) demonstration that over the twentieth century affluent American professionals became ever less likely to justify their relatively strong economic rewards on the basis of their social role, and more likely to justify it in terms of their market position, is a case in point. In the light of issues such as this, the functionalist argument has really become a version of economists’ arguments about the effectiveness of market mechanisms, and contains little distinctively sociological content.
Secondly, Marxist approaches saw market processes as derived from more fundamental inequalities in the ownership of the means of production. This offered a robust critique of market-based reasoning, but Marxist perspectives have been deeply troubled by their reliance on the labour theory of value. The argument here was that in capitalist society value is expropriated from workers by the owners of the means of production because workers are not paid an amount equivalent to that which they have invested in the commodities they have produced: in this way surplus value is extracted from the workers. The labour theory of value has been subject to intense critique, much of it, perhaps paradoxically, from Marxists. In part, these criticisms have been based around observations of the difficulty of defining the place of professionals, managers and routine white-collar workers within the surplus extraction process, and hence the difficulty of knowing who exactly extracts the surplus value in an era of impersonal corporate capitalism. But these difficulties are related to more fundamental criticisms from Elster (1985), Wright (1985) and Cohen (1989) who use logic drawn from analytical philosophy and economics to show the difficulty of putting the labour theory of value on acceptable ‘micro-foundations’. Cohen, for instance, uses game theory to claim that the labour theory of value cannot hold, because the price of commodities varies over time in response to changing demand. Thus, the price of commodities cannot be seen as linked to the value embedded in them. The traditional Marxist defence to this objection has been to note that Marx never saw the price of commodities as the direct result of the labour value embedded in them (the ‘transformation problem’), but once this point is granted, it means that the market cannot be seen as strongly conditioned by the inequalities of capitalist production processes. This thereby undermines the main virtue of Marxist approaches. Because market forces become analytically separable from the extraction of surplus value, market-based accounts of economic inequality are admitted through the back door, so to speak. This therefore leads to a similar outcome to that of functionalist theory. Recent Marxists who have attempted to resuscitate class analysis, such as Wright (1985, 1997) have predominantly used economists’ concepts of capital, asset and rent. I discuss the implications of their arguments further below.
Weberian approaches to stratification recognized the pertinence of market processes more than other sociological accounts (see Scott, 1996). Deep in Weber’s thinking is an awareness of the omnipresence of competition as fundamental to human social relationships. At the same time, Weber subtly recognized the power of status and party to interfere with ‘pure’ market mechanisms in generating social inequalities. But the significance of Weber’s work in stratification theory has perhaps been over-rated: its popularity has traditionally relied on defining it as a sensible alternative to Marxism, and thereby it implicitly depends on maintaining Marxism as a viable current of work (see Savage, 2000). With the decline of Marxism, however, the reliance of Weberian approaches on a market-based account of economic inequalities became more manifest. Weberians tend to take market-based inequalities as given, with attention directed more to the intersection between economic, status and political inequalities in the generation of social groups. Weberian work could thus easily be incorporated into a descriptive listing of various facets of inequality with relatively little analysis of their inter-relationship. There might thus be seen to be an elective affinity between Weberian approaches to stratification and the kind of variable-centred approach to sociological analysis dissected by Abbott (2002).
These problems explain why the contemporary sociology of stratification has become largely descriptive (Sorensen, 2000). Crompton (1998) demonstrates the steady rise to influence of what she terms ‘employment aggregate’ approaches to stratification since the Second World War. Employment aggregate approaches to class analysis (Crompton, 1998) conflate class with the division of labour (through their use of a class schema) and seek to demonstrate the links between these employment-based classes and a range of dependent outcomes, such as social mobility prospects (Erikson and Goldthorpe, 1992), voting behaviour (Evans, 1999), health outcomes (Bartley, 2004) and so on. Some sociologists (e.g. Grusky and Sorensen, 1998) argue that the logical consequence of this approach is to decompose classes into more specific occupational groups (hence ultimately making the concept of class redundant). Furthermore, even though this body of work is predominantly descriptive, it could still claim to be important by providing an account of why stratification is socially relevant—a kind of rich description of the extent and pervasiveness of social inequalities in contemporary societies. Furthermore, sociologists in this tradition have not established any close correlation between class position and people’s sense of class awareness and identity (Evans, 1992; Breen and Rotman, 1995; Goldthorpe, 2000). This means that there is no clear link between class position, class consciousness and class action in the way that traditional sociologists within the class formation tradition expected. Insofar as class is important, its importance seems to rely on mechanisms that do not depend on recognition and agency.
We thus see stratification theory having two major problems: the lack of a clear sociological theory of exploitation, and the lack of a theory of identity and agency. Instead, contemporary stratification makes do with a kind of rich description of social inequalities. The problems here have best been exposed by Sorensen (2000), who argues that much stratification sociology fails to provide an explanatory framework for stratification theory, and relies on the tautological claim that advantages are generated through being in a position of advantage, which fails to address the issue of how such advantages are structurally generated. Fundamental to the sociological enterprise, he maintains, is the attempt to show how exploitative processes produce the kinds of inequalities that exist. Without this demonstration, it cannot be established that stratification is a core concern of sociology. Sorensen correctly sees the project of an explanatory stratification theory as fundamental to the sociological enterprise. In the next section I consider how these problems might be addressed by drawing on new developments in economic sociology and Bourdieu’s reflexive sociology. I begin by considering how Sorensen’s own observations highlight some crucial issues for the study of stratification.
Assets, Exploitation, and Accumulation
Like Wright (1985, 1997) and Tilly (1998), Sorensen attempts to renew sociological approaches to inequality by critically reworking the concepts of economic theorists to show how structural inequality is an inherent part of market processes. Rather than attempting to develop a sociological theory of inequality that does not rely on market processes, he shows how it is more useful to develop a distinctively sociological account of markets in order to show that markets depend on social processes. This involves taking the fight to the terrain traditionally occupied by economists. Indeed, over the past 20 years a fertile area of economic sociology has developed, especially in the United States, which shows sociologists no longer taking the market as a given but showing instead how it is socially constructed and embedded (White, 1981; Granovetter, 1985; DiMaggio and Powell, 1991; Roy, 1997; Callon (ed.), 1999; Fligstein, 2001; White, 2003). Strikingly, little of this literature directly attempts to demonstrate how social stratification is embedded within markets, since their concerns tend to be more with the construction of markets as institutions or fields (Fligstein, 2001), though there are some gestures in this direction. Roy (1997), for instance, shows how the rise of large American corporations in the early twentieth century cannot be attributed to their greater efficiency but has to be seen as linked to the power resources of dominant elites in American society.
In order to consider more fully how sociologists can excavate a new place for stratification sociology within a market-based account of social inequality, a distinctive sociology of assets and capital is required. The terms capital and assets have increasingly been extended to a wider array of social practices—for instance, to cover human capital, social capital, cultural capital, symbolic capital and so forth. Some writers, such as Fine (2000), are critical of this development, seeing recent debates over social capital as testimony to the hegemony of economics in the social sciences, and as an indication of the marginalization of sociological and political economy approaches. However, Fine’s account misplaces the sociological potential of these intellectual exchanges, and is also open to the criticism that the status of political economy approaches without the labour theory of value is anyway problematic. The work of Roemer (1982) and Wright (1985) on the concept of assets, and Sorensen (2000) on rents, can in fact be seen to perform important intellectual work because it allows a way of conceptualizing class that does not rely on the labour theory of value (see, for instance, Roemer, 1982: 47f, and Cohen, 1989). Rather, class is derived from exchange relations (conceived of as exchanges between individuals in a game-theoretical sense), using concepts from neoclassical economic theory to help model such exchanges. Within such a perspective assets are processes that prevent free markets (in labour, property etc.) operating and which lead to structural inequality, and hence markets can be seen as socially constructed devices.
Therefore, it is possible to see a certain radical potential in the deployment of ‘asset’ theory. For, whilst these concepts work within the terrain of economic theory, their intent is to demonstrate how market processes are necessarily structured by the causal powers of assets that systematically advantage some agents over others. Roemer (1982), for instance, emphasizes that in reality free markets cannot ever operate, and hence that assets, and thereby class inequality, are integral to exchange relationships. Once this step is taken, then the potential exists to fully disrupt market-logic’. This is especially true if the further step is taken of embedding asset theory within Bourdieu’s concept of the field (e.g. Fligstein, 2001), in which the market can be seen as an abstract social construction (hence allowing the monetarization of exchange relations) dependent on certain exclusions and institutional foundations. Here we can see the critical potential for the concepts of asset and capital to be related to the way that they offer a distinctive means of conceptualizing stratification through the critical realist emphasis (Archer et al, 1998) on the generative properties of specific entities (i.e. capitals or assets) in particular kinds of contingent circumstances. It is important to recognize the difference between this realist approach and other forms of class analysis. The differences from other forms of class analysis are importantly highlighted by Sorensen (2000), though similar arguments can be found in Wright (1985, and to some extent 1997). However, if assets are to be defined as a generative causal entity, then how exactly do we recognize an asset, given the realist emphasis that it cannot be directly observed? Realists have tended to skirt round this issue, preferring discussions about the general differences between abstract and concrete research rather than a concise account of which assets are of prime social importance. To put this issue another way, stratification theory needs a way to distinguish assets, as generative entities with causal powers, from contingent correlates of social advantage or disadvantage. The confusion of sociologists on the number and range of assets is itself indicative: Wright distinguishes organization, skill and property assets. Bourdieu distinguishes cultural capital, social capital, economic capital and symbolic capital.
The usual way of handling this issue is to see an asset as intrinsically exploitative and relational. This is the approach famously taken by Wright (1985), who identifies three assets, based on property (owners exploit those without property), skill (those with skill exploit those without) and organization (superordinates exploit subordinates). He can thus distinguish between unequal outcomes generated by one or more of these three exploitative assets, and those caused contingently. For instance, health is undoubtedly a correlate of economic well-being. Since it could be claimed that someone’s good health does not entail another’s bad health (and hence that someone in bad health would not necessarily be better off by changing the health of the person in good health), then it could be claimed that ‘health assets’ do not exist. This might be true even though there is a clear association between health and life chances. This would make it possible to distinguish the generative powers of relational assets from outcomes. However, it is actually rather difficult to draw such a clear distinction. Health is a relative state. The employment prospects, for instance, of those in bad health would be improved if those in good health were not in the labour market. A similar example is that of left-handedness. It is known that being left-handed is associated with having a higher mortality rate than right-handers, and this association can probably be explained by the difficulties of lefthanders in dealing with right-handed ‘technologies’. In this case, one would have to see ‘handedness’ as a form of exploitative asset. The problem is therefore that using this game-theoretical logic, assets can rapidly be multiplied so that they become diffuse and diverse to the extent that any kind of advantage can be redefined in such terms. It is therefore interesting to see that most sociologists tend to fall back on the established sociological canon, with three assets broadly approximating to class, status and party. The problem, however, is that there is no credible justification for this within the parameters of the theoretical framework they use. In addition, this restriction has had the unfortunate effect of leading to ‘recycling of old wine in new bottles’, in which only the established, sociological orthodoxies of class, status and power (reworked as property, skill/culture and organization) become defined as assets of class inequality.
In more recent years, Roemer (1988) and Wright (1997) have reacted to the can of worms opened up by claiming that only assets linked to the labour process are ‘real’ assets. This is a means of reasserting the conventional Marxist emphasis on labour (defined as employment) whilst not using the labour theory of value. For Roemer (1988: 5), ‘a class is a group of people who all relate to the labour process in the same way’. Sorensen and Wright’s advocacy of rent theory points in a similar direction. However, this approach leads back to the descriptive ‘employment aggregate approach’, whereas the value of the asset approach was supposed to lie in providing an explanatory alternative to this. It also fails to address the crucial issue that the prime focus on the labour process needs to be justified, not posited through a priori logic.
Recent stratification theory has therefore reached something of a dead end. One way out of this impasse is to use the concept of asset in a way similar to that Bourdieu used the concept of capital in connection with that of the field. In realist terms this involves thinking not only about capitals as generative, causal, mechanisms, but also about fields as the environments in which capitals can be effective. Substantively, this leads us to focus on the accumulatory and convertible potential of capitals (or assets) in different kinds of fields. Stratification can thus be seen as involving the cumulative stacking of advantages over time by those who can draw on assets. In Bourdieu’s phrasing,
Capital, which in its objectified or embodied forms, takes time to accumulate and which, as a potential capacity to produce profits and to reproduce itself in identical or expanded form, contains a tendency to persist in its being, is a force inscribed in the objectivity of things so that everything is not equally possible or impossible. (Bourdieu, 1997: 46)
In Marx’s terms, in the capital-labour relationship, it is the daily exchange of labour power for wages, and the routine accumulation of capital, that define the nature of this specific relationship as one pertinent to stratification. Rather than focusing on the abstract, cross-sectional, exchange between social parties, we should look at the cumulative potential of assets to reproduce themselves and to accumulate. This allows us to pay greater attention to the way that certain assets can be accumulated in various kinds of devices and practices, or can convert themselves into other sources of advantage. This approach draws its inspiration from Marx’s definition of capital as lying in its ability to accumulate, through the M-C-M1 cycle. The advantage of this approach is that it does not require a labour theory of value (though it is compatible with it).
This emphasis on how inequalities are reproduced, accumulated or challenged does of course address much work carried out within the sociology of stratification, for instance in the field of social mobility, much of which is concerned with the extent to which advantages can be transmitted inter-generationally and preserved over the life course (Erikson and Goldthorpe, 1992). Devine’s recent (2004) qualitative study of how middle class professionals in the United States and the UK were supported by their parents, and also use various strategies to support their own children, is a case in point. Here, the focus on transmission over time allows a more fluid approach to stratification which does not demand a relational theory of class to be useful. Bourdieu is theoretically useful for developing this orientation. Drawing on standard sociological arguments, Bourdieu sees those who have capital as exploiting those who do not. Bourdieu’s main innovation can be seen in relating his theory of capital to a theory of the field. Fields are also relational. A field is ‘a separate social universe having its own laws of functioning independent of those of politics and the economy’ (Bourdieu, 1993a: 162). Because a field has ‘its own laws of functioning, its specific relations of force, its dominants and its dominated …’ (Bourdieu, 1993a: 163), it defines the stakes around which actors relate in particular contexts (see further, Bourdieu, 1993b). This relationality of the field allows sociologists to relax their assumptions about capitals without losing the central argument that stratification in general is a relational process.
Bourdieu’s concept of capital runs parallel to Marx’s emphasis on capital as accumulated labour and thereby capital as an inherently relational process of exploitation. Thus cultural capital is defined as dependent on intellectual abstraction that is in opposition to the necessities that lie at the heart of working class and peasant life. Cultural capital stands in direct opposition to the immediacy and practicality of popular culture. Bourdieu (1984) establishes this case in part by showing differences in taste and disposition between classes using surveys of French taste in the 1960s. There is now an extensive debate about whether Bourdieu overemphasizes such differences (Lamont, 1992; Longhurst and Savage, 1996; Bennett et al, 1999), which we need not go into here. What does need reiterating is that even when such differences can be seen, this does not demonstrate that they are relational. A further part of Bourdieu’s argument is his rather weakly developed historical insight that the emergence of the Kantian aesthetic can be linked as a reaction to popular culture during the nineteenth century. The elaboration of a skhole (Bourdieu, 1999) based on abstraction and logic involves the distancing of culture from the immediacy and necessity of everyday life, and one can therefore see how the emergence of an academic intellectual elite depended on drawing direct contrast with lay, practical knowledge. Here Bourdieu draws on Durkheimian and Weberian themes regarding the differentiation of modern social relations; he focuses on the slow development of various kinds of intellectual fields, as each becomes more autonomous from economic and political relations.
This process by which fields emerge can rather be seen as one akin to primitive capital accumulation, in Marx’s terms. For Marx, primitive capital accumulation depends on creating the initial stock of capital necessary to set the capitalist system in operation, which he saw as critically dependent on the enclosure movement in Britain to expropriate wealth which could then be used for investment. The primitive accumulation of cultural capital, in like manner, involves the formation of an intellectual cadre in key areas of social life who champion and advance the scholastic approach. It is possible to link such a process to that of professionalization. Larson (1977) has famously shown, for the British case, how the middle years of the nineteenth century were especially important in allowing professions to define their skills as being based in science and academic learning (rather than the ‘craft’ knowledge and practices that they had hitherto relied on), and this process involved the direct de-legitimation of lay knowledge. The case in medicine, where doctors used the monopoly provided by the 1858 Act to make the lay knowledge of wise women, herbalists, apothecaries and the like illegal, is especially well known, but similar processes whereby the gap between ‘craft’ skills and ‘academic’ expertise was established through the denigration of craft skills by abstract reason can be found elsewhere.
Once the initial stock of capital has been created and the field has emerged that provides the institutional and social context in which further capital can be invested and accumulated, exploitation becomes less visible as it no longer has to be exercised through forcible expropriation. It accumulates through the routine workings of the capitalist economy. If this argument were applied to cultural capital, once the academic and educational system that allows cultural capital to be accumulated and transmitted is set in place, it no longer needs to be overtly and directly contrasted with craft knowledge or popular culture. Whilst Bourdieu’s concern to make such implicit contrasts between cultural capital and its plebian ‘other’ visible is laudatory (see especially 1984) at one level, it also misses the point in that it is precisely the fact that cultural capital does not need to overtly denigrate popular culture in its mature operation that is crucial to its social effectiveness and legitimacy. This point allows current debates about cultural fragmentation to be put in a different context. Thus the observation that many members of the cultured middle class are increasingly likely to ‘mix and match’ their tastes and are happy to employ certain parts of a popular aesthetic (Petersen and Kern, 1996) can be taken as evidence not that cultural capital does not exist, but rather that mature cultural capital is not contaminated by interaction with a popular culture against which it no longer needs to define itself overtly (see also Bryson, 1996; Warde et al, 1999).
Thus, the relationships embedded in ‘mature’ capital can be seen as existing in a relatively weak sense, in which the routinization of the organization of capital accumulation serves to make exploitation itself invisible. This is, of course, precisely what Marx saw as fundamental to industrial capitalism, which explains why he began his account in Capital with a chapter on commodity fetishism. In this respect my argument is an extension of Marx’s arguments about capital and labour to other social processes. It becomes a matter of analytical nit-picking as to whether the exploited groups within these arrangements might be better off in different sets of relationships, since this in no way tells us anything about the possibilities for actually changing such relationships. Returning to our counterfactuals, we can admit that there are a plethora of exploitative relationships. Left-handers are indeed exploited by having to live in a world organized around the implicit needs, values and norms of righthanders: indeed such is the routinization of this form of exploitation that it is hardly ever remarked on. However, and this is the point, what makes capital sociologically interesting is not that this is a form of exploitation—for this is so ubiquitous—but that the social consequences of different kinds of capital depend on relating them to fields that allow the accumulation and convertibilty of capital. Although right-handers could be said to exploit, it is not clear that their advantages accumulate over time, or can be directly converted into advantages in other fields. Those with cultural capital, for instance, are able to use their qualifications to help secure better jobs. Those who are right-handed cannot secure better jobs by virtue of their right-handedness. There may simply be relatively few social implications of right-handedness for accumulation and convertibility. There is no field of handedness in which left- and right-handers play games with each other, and hence there is little investment in handedness as a salient social process.
The reason why ‘handedness’ is not a class asset is related to the way that it cannot readily be abstracted so that it can be accumulated. The abstraction of capital from embodied social interactions is critical to its potential to accumulate: it is for this reason that economic capital, abstracted through money, is easier to accumulate than any other. The more abstract money is, the more it aids exchange and conversion: consider, for instance, the greater ease of conversion of same currency compared to other currencies. Whereas economic capital is entirely disembodied and depends on being stored in abstract systems, other forms of capital can never be disembodied. Social capital cannot be detached from the actual personal ties between people, and even though its potential for accumulation is still considerable through reliance on weak ties and ties through third parties, it is less easy to accumulate than economic capital. Insofar as cultural capital can be accumulated it depends on being objectified through being instantiated in institutional processes and a canon of ‘high culture’.
The counter-point is that the convertibility of capitals depends on agents concretely moving between fields. Convertibility depends on particularizing’ social relationships. The process of accumulating financial money can largely be left to the routine workings of financial markets once a mature economic field is established. To convert economic capital to cultural capital, an agent needs to make a particular contingent transaction that involves stepping outside the routines of that field. It therefore recognizes a place for agency and identity rather different to that in traditional stratification theory. Traditional stratification theory expects social groups to ‘form’ cultural and social identities on the basis of their group membership, but the accumulation and convertibility of capital does not require identities of this kind. Identities can be seen as a strategy of the disadvantaged within any field to attempt to gain tactical advances by drawing particular attention to themselves in ways that allow them to stake claims that go beyond the rules of the game as defined within that field. For instance, claiming rights to specific treatment in employment relations through membership of a specific disadvantaged social group claims that the ‘normal’ rules of the game for hiring employees should not apply. Whilst such interventions may improve the situation for such groups within the field, they can be seen to be limited in that by defining relevant groups as needing special treatment they confirm their inability to ‘really’ play the game.
Thus whereas both Nancy Fraser (1995) and Miriam Fraser (1999) see recognition politics as involving vital political stakes, there are clearly difficulties for some groups to make claims for recognition (see also Skeggs, 1997). Recognition is a limited tactic of the powerless. The powerful do not wish to draw attention to themselves: their ability to engage in conversion strategies depends on them not having identities that fix them to a particular field. A very wealthy man who wants to convert some of his fortune into political influence is advised not to trumpet his extreme wealth, as it will define him as someone seeking undue influence. And so on. Identities are not to be understood as the outcome of a social position, which attach people to a social group membership, but can be better seen as ‘tactical’ moves in a field seeking a limited adjustment to the rules of that game. This approach allows us to handle the fragmentation, mutability and complexity of contemporary identities in ways that are entirely explicable within the conceptual apparatus of habitus, capital and field.
These observations suggest the value of relating identities to the network sociology of Harrison White (1992), Mark Granovetter (1985), Mische (2002). In White’s simple but important observation, identities are based on contingencies, an awareness that one does not fit routine. Identities need not arise from routine processes of capital accumulation, investment and conversion, but instead develop in the interstices of social relations where particular groups think that they cannot compete in a given field as a result of contingent circumstances that affect them and not others. This allows a way of relating stratification theory to issues of identity and subjectivity that do not rely on the problematic baggage of a class formation approach which assumes that identities are linked to an awareness of group membership and identification.
Conclusions
In this chapter I have explored how the sociology of stratification can be re-positioned so that it is better able to relate to the central concerns of the discipline as a whole. My basic argument is that we should focus less on the relational nature of stratification, and more on processes of accumulation and convertibility. This approach allows us to take the issue of temporality seriously by posing questions such as how are the advantages of particular social groups sustained; under what situations can they be challenged; what kinds of institutional and cultural contexts allow advantages to be routinely accumulated so that they are not seen as contentious, whereas in others they might be called into question. How can advantages in some situations be converted into advantages in others? These are middle-range’ questions which are open to empirical investigation and which would renew sociology’s critical mission.
We can draw much from Bourdieu, but only if this is approached critically. Rather than focusing specifically on Bourdieu’s concept of capital, as most stratification theorists have, we are better advised to reflect on the importance of the field (and on the interplay between field and capital) to bring out the radical potential of his perspective. This allows us to avoid being distracted by problems in sustaining a distinctive relational theory of exploitation which has problematized much stratification sociology. We have seen how this led to defensiveness and introspective inquiry, and empirical stratification researchers losing touch with theoretical debates. Amongst theorists it has led to arcane and esoteric debates about the micro-theoretical foundations of exploitation, and it has made it more difficult for stratification theorists to boldly show how their arguments matter. Using Bourdieu’s distinction between capital and fields, we can run with the idea that there are multiple sources of exploitation without this leading to anarchy, since attention can be focused on how intersections between capitals and fields allow some kinds of exploitation to be socially and historically more engrained than others. This allows the prospect for more fertile debate between theorists and empirical researchers than has been the case in recent years.
We are also in a position to examine class formation in a more developed way. In recent years stratification theorists have found the lack of association between class position and class identities problematic. Why does inequality matter if people do not appear to identify with their position? Some stratification theorists have dealt with this problem by operating with largely structural accounts of class process, whilst other cultural theorists have dealt with it by emphasizing the social indeterminacy of identities. In this chapter we can see how a ‘tactical’ account of identity might be developed that has strong overlaps with network approaches. This is perfectly at home with the recognition of the flexible, fragmented and contextual nature of contemporary identities and allows a way of engaging with current debates in this area. Finally, this interest in accumulation and temporality allows the sociology of stratification to connect with current debates about socio-cultural change in ways that can challenge simple accounts of change and have the potential to offer more nuanced insights into persistence and the remaking of privilege.