Robert van Krieken
University
of Sydney
Published in: Sociological Review 38(1) 1991: 1-25
The Poverty of Social Control: explaining power in the historical sociology of the welfare state
Abstract
The concept 'social control' has been criticised from a variety of quarters in recent years, particularly by historians and historical sociologists. However, it remains in common usage in sociological studies of welfare, deviance and social control. This paper shows, first, how this reliance on the concept of social control is rooted in a wider-ranging argument in social and political theory concerning the liberal-democratic fusion between the state and civil society, and that the lack of resolution of this argument is the foundation of the persistence of the concept social control in other areas of social inquiry, despite its repeated 'falsification'. Second, the paper highlights the main arguments against the use of 'social control' in explaining social order, in particular the misunderstanding of class, culture and power which its use encourages, and the paper will conclude with a discussion of alternative ways of conceptualising the operation of power in contemporary societies.
In principle there is much to recommend this orientation towards the study of the state and social welfare. It is beyond question that the acceptance of 'normality', whether explicit or not, and the lack of perceptiveness of the politics of social structure which characterised mainstream sociology, was well left behind. It is a view of the contemporary state which has roots in a broader-ranging argument in political and social theory, that the liberal distinction between the state and civil society has been breaking down since the advent of parliamentary democracy. Jean Cohen and Andrew Arato (1989) have recently pointed out that as diverse a range of writers as Carl Schmitt (1985), Niklas Luhmann (1982), Michel Foucault (1977b; 1979) and Jürgen Habermas (1984; 1987; 1989), albeit from quite different perspectives and in very different ways, all question the continuing viability of the notion of state and civil society as separate spheres of interaction. Cohen and Arato highlight the argument contained in Schmitt's (1985) critique of parliamentary democracy, that one of the consequences of increasing democratic pressure on the state to respond to an ever-larger range of constitutive interests has been to increase the degree of state intervention into society to such an extent that the distinction no longer applies, and we should now be speaking of a Sozialstaat (Cohen & Arato, 1989: 486-7). Luhmann (1982) argues from another standpoint that the tripartite division of society into state, economy and civil society does not make sense of modern social systems, and that they are in fact differentiated along quite different lines, as interaction, organisation and societal systems.
Michel Foucault, on the other hand, reasoned that the similarity of the logic structuring social interaction in the different spheres of society, the logic of disciplinary power, has produced a dense web of interrelated power networks which undermine the meaning of the distinction between state and civil society. The latter, for Foucault, has become yet another arena for the operation of disciplinary power relations, which construct the self-disciplined, individualised subjectivity upon which state power centres, rather than being a sphere of life in any way autonomous from the state (Foucault, 1979; Cohen & Arato, 1989: 486-91). Similarly Habermas (1989), put very crudely, sees civil society as having been eroded by the welfare state, through processes such as juridification and bureaucratisation, leading him eventually to argue for the reinvigoration of the public sphere and an alternative modernisation of civil society, in opposition to the internal colonisation of the lifeworld (Habermas, 1987: 332-73).
However, when one attempts to utilise such a view of the relationship between the state and civil society in a particular arena of empirical sociological research, one quickly encounters some serious problems, both conceptual and empirical. My aim is this essay is to outline the ways in which the background understanding of a fusion between state and civil society continues to be translated into a more or less submerged social control paradigm in one field of sociological enquiry, the historical sociology of welfare, despite the sophistication of the theoretical debates around power, agency, structure, and constraint (Betts, 1986; Burns, 1986; Clegg, 1989; Crow, 1989; Lloyd, 1989; Wrong, 1979). The view of social welfare as either an agency of social control or a vehicle for disciplinary strategies and techniques appears to retain the status of a central myth in the historical sociology of welfare. Like all myths it organises our perception and understanding in a very particular and, I will argue, misleading way, producing an understanding of power which disables us from grasping either the bases of social stability or the sources of social change.
Social Control as an organising concept
There are significant differences between various writers on the historical
sociology of welfare in their understandings of the operation of power,
and it is clearly impossible to identify a coherent 'social control paradigm'
that a unified body of theorists all share. There is, for example, worlds
of difference between an orthodox marxist approach, one inspired by the
work of Antonio Gramsci (1971), a feminist understanding, and the approach
to power and discipline which emerges from Michel Foucault and those influenced
by his work. Even within the work of one writer there may be contradictions
in their approach to social control. Nonetheless, one can identify two conceptual
tendencies which run throughout the literature on the historical sociology
of the welfare state, and most commentators will display at least one of
them.
The first of these tendencies is that of state- centredness, towards the idea that the state generally, and social welfare in particular, is the primary, if not the only locus of power and social control and, conversely, that the welfare state's impact on civil society is primarily that of social control. For example, Harris and Webb recently introduced their book on juvenile justice by describing the juvenile justice system as 'an element in a steady expansion of state control over the working-class young' (Harris & Webb, 1987: 1). Social welfare and the welfare state are perceived as reinforcing 'the prevailing ideologies of individualism, family life and the work ethic', and social workers as 'exercising a variety of forms of social control on behalf of the state - social psychological, moral ideological, and material - in the process of regulating access to welfare services and benefits' (p. 220). One loses count of the number of discussions of an aspect of the relationship between the state and family life which will declare in passing that since the turn of century the state has extended its control by prescribing middle-class child-rearing practices, creating bureaucratic apparatuses such as welfare agencies and juvenile courts, and through the enforcement of compulsory education and restricted labour legislation (Tiffin, 1983: 131). Through social policy, we are often told, 'the state manages the politics of life to shape the social to accord with the tasks and exigencies faced by the state' (Hewitt, 1983: 67).
Despite his references to Foucault, who argued against seeing the state as the locus of power (Foucault, 1980), Philippe Meyer (1983) has put one of the strongest versions of the state-centred argument, and his book is an example of how the social control paradigm can persist within a broader theoretical framework which in fact contradicts it. In Meyer's account the subject of social control of family life is 'the State', which he describes as proceeding 'like a gardener': pruning, clipping, landscaping different lifestyles (p. 1). There is no coyness about agency, with Meyer declaring, for example, that 'the State's social gardening operation was chiefly carried out by magistrates, the army and the police' (p.4), as well as social workers, whose activity is compared to Taylorism in their management of everyday life. The family is described as having 'its very functions delegated by the State, and able to exercise them only under State supervision', thus reduced to 'escorting its progeny to the institutions which have established a monopoly over apprenticeship, health, sport etc' (p.12).
The notion that child welfare operates to protect children in any way, or that the Children's Court might dispense any sort of justice is dismissed as a fraud, a facade for the real exercise: the State's regulation of everyday life, the function of which 'is simultaneously to intimidate, to improve, to tame and to model the family' (p.38). Meyer revives the notion of a former 'community' lost to the regulated order of the modern state, leading to 'the total impoverishment of sociability' (p.117). Meyer consistently uses characterisations like 'the progressive destruction of society by the State', and concludes, in terms of which the nineteenth-century defenders of laissez-faire, individualism and Church authority against state expansion would have been proud, as follows:
In monopolising the organisation of communal life in order to establish officially what is good for society, the State is involved in a constant work of pauperising sociability - of exterminating society. The elements that provisionally led to the constitution of the family, have now been superseded by the atomisation of that unit, at once so weak and yet still too impenetrable.
The only way to halt this State demolition of social life is to develop forms of self-help and community initiative. Any attempt to produce,
...a communal life outside the field of the institutions, any attempt to regain initiative, any dissidence against the universe of protection, every effort to enable society to reclaim the role of actor and do its own gardening, constitute so many manifestations of resistance to the possible destruction of society by the State, so many instances of opposition to this outrageous possibility becoming an inevitable doom. (p.121- 2)
The picture here is of the State as the Holocaust: destruction, impoverishment, repression and certainly social control are what Meyer sees as characterising the State's relation to family life in this view of child welfare, perhaps the starkest statement of a social control approach.
Second, there is a tendency towards an ojectification of those seen as being at the receiving end of processes of discipline, management, administration and policing. In other words, in the literature which appears to have resolved the problems of state-centredness, the state is no longer seen as the centre of power (Melossi, 1990), and power is conceptualised in terms that go beyond domination and repression, there still remains a persistent tendency to see most individuals, but particularly the 'powerless' as the objects of management, administration and intervention, their actions, beliefs, behaviour and thoughts policed, administered, regulated and colonised by teachers, social workers, doctors and psychiatrists (Lasch, 1977; Donzelot, 1979). Echoing Nicholas Kittrie's (1971) critique of 'the therapeutic state', Castel, Castel and Lovell (1982), for example, rally us to defend 'the last remaining territory not yet fallen under the sway of the old guardians of law and order and the new engineers of the mind', and have a very bleak view indeed of France's future as an 'advanced liberal society', a society destined to be characterised by an organisation of everyday life in which techniques of manipulation 'become coextensive with all aspects of social life', a 'padded world watched over night and day by squads of skilled specialists', skilled at 'manipulating people to accept the constraints of society' (p. 320). The contribution that the objects of social control and social regulation, individual or collective subjects, make to the process is regarded here as being primarily that of resistance (Foucault, 1977a, 1977b; Barbalet, 1985). Resistance to control and domination, the opposition of the working class, women, children and youth to the attempts of the state and welfare professionals to regulate their lives, is what introduces the primary element of fluidity and human agency as opposed to the iron requirements of structure. This was how the arguments of both action sociologists and the Foucault of Discipline and Punish were incorporated as an antidote to the functionalist tendencies of structuralist radical sociology, to acknowledge that systems of domination and regulation do not work perfectly, that people are capable of opposing techniques of social control and refusing to acquiesce to its demands.
On occasion one also encounters arguments which make use of Foucault's later work on 'government' (1979), where he emphasised that power organises and promotes particular aspects of individual freedom (Donzelot, 1979: 94; Patton, 1989: 270-71; Melossi, 1990: 174-80). A power relationship, unlike a relationship of violence, is based on a two-way relationship, so 'that "the other" (the one over whom power is exercised) be thoroughly recognised and maintained to the very end as a person who acts'; in this later work Foucault emphasised that power 'incites, it seduces, it makes easier or more difficult', it is always 'a way of acting upon an acting subject or acting subjects by virtue of their acting or being capable of action' (Foucault, 1982: 220). However, ultimately there is still an assumption that individual actions and capacities have been organised rather than they have any independent organising effect of their own. Foucault still spoke of his project in terms of a concern with how 'human beings are made subjects' (1982: 208), and put the active constitution of the self by acting subjects into the particular category of 'technologies of the self', leaving 'technologies of power' defined as those which 'determine the conduct of individuals and submit them to certain ends or domination, an objectivising of the subject' (1988: 18). The general spirit this concentration on the objectification of subjectivity is well-illustrated by the rather lovely and telling example chosen by Dario Melossi to close his recent study of social control:
There is, in the Museum of Anthropology in Mexico City, a statuette of a man carrying on his back an idol, in human form. Because the man is so bent under the weight of the idol, his face is hardly discernible. One can see little more than the beaming face of the idol and the hunched body of the man who is carrying it. The power of the idol lies entirely with the man, but he does not know it. He thinks it is an honor to carry such an important and powerful idol (Melossi, 1990: 185).
The important characteristic of this illustration is that while the man carrying the idol is said to constitute the power of the idol, he is also regarded as being unaware of that power, and thus ultimately at the mercy of the idol, its object, because of that lack of knowledge.
These arguments and theoretical perspectives have a somewhat problematic relationship to the historical evidence on the development of social welfare, in part because of their heavy reliance on either the writings of welfare reformers and bureaucrats themselves, or on the apparent objectives of welfare policies, legislation, and organisational structures (e.g., Dingwall et al, 1984), producing a focus on intent rather than effect. It is not surprising that welfare reformers often expressed an interest in maintaining public order, eliminating pauperism, encouraging industriousness, and all at the minimum cost to the state, but the conclusion is then often drawn that social welfare was and is clearly a matter of social control, without examining whether and to what extent such effects were in fact achieved, or whether welfare had anything to do with them. Piven and Cloward's (1972) well-known argument about the expansion of social welfare in times of unrest, for example, is in fact an observation of how legislators and bureaucrats perceived the relationship between welfare and disorder, leaving untouched the more difficult question of whether expanded welfare services played any role in the settling of social unrest. The social control arguments work best as part of a history of the strategies, ambitions and plans of those who hoped to direct, manage and coordinate social life, and if this was all that was being said one could raise no objection. However, by either remaining silent on this question of outcome or by failing to avoid formulations which imply the coexistence of outcome with ambition, much of the sociology of welfare literature goes well beyond that very particular project to create the impression that we are being told about 'how things actually were' as well as 'how some hoped things would be'.
The history of the welfare state can frequently be read as a story where the character of the relationship between the controlled and controllers, between people and idols, goes beyond resistance to include not just acceptance, but positive encouragement, and not simply an organised or governed encouragement. One where the impetus for the expansion of state intervention into civil society includes the active desire among the controlled for a more active state, for an expanding welfare bureaucracy, and where such an expansion of state intervention might be said to have diminished rather than increased their subjection to control and domination. A close examination of the historical evidence also raises the question of whether the activities of the state make much difference to social change at all, and whether the kind of social regulation sociologists criticise are better understood as based in the economic relationships and structures of market capitalism, or perhaps in the particular cultural patterns of Western societies. Social historians have been arguing along these lines for some time now (Fox, 1976; Muraskin, 1976; Thompson, 1981; Cohen & Scull,, 1983; McGovern, 1986), and it is perhaps one of the main sources of the irritation historians feel towards sociologists, but their arguments have failed to penetrate very deeply into the historical sociology of the welfare state.
Social Control: the theory and the evidence
The first, most obvious problem with conceptualising social welfare as part
of an expanding system of state-centred social control is that the extent
and effectiveness of social welfare's contribution to social order, in comparison
to other ordering and regulating mechanisms, is more often assumed than
demonstrated, and there are good grounds for considerable skepticism in
this regard. Gareth Stedman Jones (1977) has made the point forcefully in
his critique of the social control approach to the history and sociology
of leisure, that 'the greatest "social control" - if one wants
to use that word - available to capitalism is the wage relationship itself
- the fact that in order to live and reproduce the worker must perpetually
resell his or her labour power' (p. 169), and this point is also generally
applicable to the state-centred approaches to the study of social control
and welfare. In his study of the history of child and youth welfare in Germany
up to the fall of the Weimar Republic, Detlev Peukert (1986) argues that
'the bureaucratic creations of the formal sector can only exist when they
are based on functioning informal sectors, the reserves of which they give
shape to' (p. 313). The continued development of the capitalist system of
wage labour was thus heavily dependent on 'the reservoirs of the black economy,
domestic labour, familial socialisation, and self-help (p. 313). Peukert
also argues for a sensitivity to the limits of social discipline,
as the continued repetition of demands for more and better social control
up to 1932 appears to indicate that although social pedagogy attempted to
replace crumbling systems of community self- regulation, the attempt was
marked more by lack of success than effective implementation (p. 312; for
similar Australian evidence, see van Krieken 1990).
It is possible to deny the consensual view of functionalist sociology, in which most of us simply became socialised to accept the dominant social norms, leaving only deviance and abnormality outside those norms, without having to argue that there is no value consensus at all, or that it must have been put in place through a process of state- controlled social engineering. Ideological consensus is maintained by processes other than state intervention, whether it be through the media, welfare, the law, or educational. Thomas Mathieson, for example, regards the media's construction of our perceptions and self-perceptions of need and desire as the most central feature of social control in contemporary capitalism (Mathieson, 1987). Andrea Abele and Marlene Stein-Hilbers have argued in similar terms that institutional forms of social control are based on informal, pre-institutional social control, such as the media and what they refer to as 'individual and collective everyday knowledge' (Abele & Stein-Hilbers, 1978: 173), and Stuart Henry (1983) highlights the dependence of state power on the intermediary institutions and organisations which structure everyday life, especially the workplace, emphasising the relative autonomy of this sphere of 'private justice'. In Peter Hall's comparative analysis of state and economy in Britain and France, he also points out that 'the institutional networks affecting state action extend well into society, in such a way as to expose the state again to societal influences', so that the state should be regarded as 'a network of institutions, deeply embedded within a constellation of ancillary institutions' (Hall, 1986: 17).
When one examines the ways in which working-class families responded to social welfare, as well as to other forms of state intervention, they were far from passive. This has become a point of extensive debate in relation to the history of education, where recent research is increasingly pointing to the active involvement at a grass-roots level in the expansion of state education, and to the very specific and limited nature of the opposition to it. Most frequently the observation is that parents resisted education where their child's labour was seen as essential for the family's economic survival, and not as a principled opposition to the whole idea of state-controlled education of their offspring (Wimshurst, 1981; Kaestle, 1983: 69). As William Reese (1986) has put it in relation to the history of schooling:
Struggles for playgrounds, free breakfasts, social centers, and other welfare reforms as well as for fair community representation on the school board were an integral part of larger community activism by working-class, socialist, and neighbourhood interests hostile to business rule and centralised decision-making....school innovation and reform were produced by interaction, resistance, adaptation, and accommodation, with the power of capital clearly in a dominant though never unchallenged position. Liberals, Socialists, and business efficiency advocates all agreed that schools would be important social institutions in the twentieth century, and for that reason many of the period's most radical thinkers understood that public schools like other political institutions were contested ground (p.239).
The operation of child welfare in Australia also indicates that parents were often the initiators of the judicial processes which took their children within the orbit of state action (van Krieken, 1989), and the same point has been made in the work of Jennifer Davis (1984) for London, Michael Katz (1978), Barbara Brenzel (1983) and Linda Gordon (1988) for the USA, David Crew (1986: 250) for Weimar Germany, and Jeroen Dekker (1985) for the Netherlands.
There were clearly coercive aspects to the involvement of welfare professionals in family life, and this is what Greta Jones (1986) emphasises in her argument against Jacques Donzelot's account of welfare clients operating in alliance with social work experts (Donzelot, 1979: xxii). Jones sees no ambiguity at all, arguing that experts were not invited into family life, but invited themselves, sometimes forcing their way into the home. Rather than any kind of alliance or complicity, Jones points to evidence of 'a subterranean revolt among families at the intrusion of the expert in which women took part equally with men' (Jones, 1986: 37-8). In any case, as Lasch points out, for Donzelot too women clients were always the subordinate partner in the alliance (Lasch, 1980: 29).
However, this is only part of the story. There were also times when state intervention was invited, even if the hosts disliked the ultimate outcome. In cases of family violence, for example, Linda Gordon has found that clients frequently initiated intervention; even when the stakes were as high as losing ones children, most complaints came from parents or relatives mobilising their own standards of childrearing (1986: 469). Although the Massachussetts Society for the Prevention of Cruelty to Children did become known as 'the Cruelty', a favourite indicator of working-class fear of state intervention (Katz, 1986: 109), this did not prevent parents, relatives and neighbours from contacted the organisation. As many as sixty percent of the complaints to the MSPCC in its first ten years came from family members, requests for help which 'came not only from victims but also from mothers distressed that they were not able to raise their children according to their own standards of good parenting' (Gordon, 1986: 472) Gordon also argues that the process of intervention into family life by state agencies must be seen as precisely that: a process with ambiguous outcomes, and one in which clients actively participated, to often have a significant impact on the final outcome (1988: 289-99). Once one recalls that families are not homogenous units, that they consist of power relations with weaker and stronger members, it becomes easier to see that the impact of state intervention into family life was frequently to modify those power relations, 'usually in the interest of the weaker family members' (Gordon, 1986: 472). Both Gordon's work and that of Margo Horn (1984) has demonstrated that clients often manipulated their social workers to direct the development of their case, and a variety of historical investigations have come to similar conclusions: David Crew argues that the activities of the German Familienfrsorgeamt between 1918 and 1933 'created a 'structured space' within which social workers and clients tried to affect one another...and individuals and families developed 'strategies for survival'...which were sometimes directly at odds with the intentions of the social workers themselves' (Crew, 1986: 246). Kerreen Reiger has said the same of the Australian evidence: women clients were 'active participants, either accepting, rejecting, or modifying the experts' decrees', depending on how the advice they received fitted with their own concerns (Reiger, 1985: 215-7).
Quite apart from the question of the responses of those directly affected by welfare intervention, a key characteristic of the historical sociology of welfare literature is the notion that the goals of state intervention were and remain intrinsically alien to the working class, the interests of women, the concerns of ethnic minorities, and so on. The most common observation is that particular ideologies are reinforced or reproduced by and through social welfare, with the assumption that those ideologies would otherwise not arise at all, or in a significantly different form. Steven Humphries' book on industrial schools and reformatories thus describes them as 'institutions of class control', utilised by 'the middle class and the state to contain those elements of working-class youth culture that most threatened their continued domination'. This was to be achieved by inculcating in working-class youth 'habits thought appropriate for the respectable working class, such as obedience, discipline, honesty, cleanliness and sobriety' (Humphries, 1981: 211, 214). In her book on women and femininity in Australia, Jill Matthews says of child welfare legislation that it was 'directed towards controlling the children of families which did not conform to a particular cultural standard derived from the ideals of the population reformers' (Matthews, 1984: 86).
However, the values and norms at issue here, honesty, sobriety, punctuality, refinement of manners and morals, self- discipline, cleanliness, thrift and the work-ethic were not simply middle-class or bourgeois values, they were also an integral part of a broader working-class culture. The whole point about these being values 'thought appropriate' for the respectable working class was that the respectable working class themselves adhered to them, making that class as such the source of the social control at issue here as the middle- class or the state. The historical evidence is unambiguous on this point: in his study of Sunday Schools in England, Thomas Laqueur concludes that the congruence of the values of honesty, punctuality, hard work and so on with both the requirements of the industrial system and the political interests of the bourgeoisie, did not in itself make them 'middle-class values'. Sunday Schools were able to transmit these values precisely because they were those of 'the working-class men and women who taught in and supported the schools' (1976: 239). F.M.L. Thompson (1988) comes to a similar interpretation, both of Sunday Schools and other middle-class inspired attempts at moral reform, and develops a view of nineteenth century society similar to Trygve Tholfsen, who argued that nineteenth century working-class and middle- class cultures should be seen as overlapping with each other (Tholfsen, 1976: 18; see also Faler, 1974). As Janet McCalman puts it for her discussion of working-class culture in the Melbourne suburb of Richmond, the respectable working class 'had little sympathy for the dirty, the unchaste and the drunken - these were self-disciplines that cost nothing' (McCalman, 1984: 20).
Class, Culture and Power
Behind the problem of the centrality of state intervention and the interpretation
of historical evidence lies another, perhaps more basic and difficult one
of how we should conceptualise the relations between (a) class, gender and
race divisions, (b) the market definition of desire and need, and (c) culture,
the arena of everyday life, morality, behavioural norms, family ideology
and relationships. By culture I mean basically what Raymond Williams means
by it: institutions, manners, habits of thought, intentions, way of life
(Williams, 1960: 327). The point at issue is whether we should agree that
there are fundamental and clearly identifiable oppositions between working-class
and bourgeois culture, between men and women's attitudes to family life,
and so on, within which the state and social welfare play a significant,
formative role.
It has probably been the moral concern with oppression and injustice, together with an intellectual determination to discover an explanation which gets to the heart of how power operates within and through social welfare, which has made it difficult to resist the stubborn seductiveness of a social control paradigm. Resistance has been made even more difficult by the fact that middle-class intellectuals working in connection with social work and social policy probably found it easier to simply reverse the valencies placed on their work, from positively valued 'help' to negatively valued 'social control', than to entertain the notion that what they did might make little difference in relation to broader based social institutions, structure, and processes of change. Better to be a renegade agent of the state, 'in and against the state', than risk being an irrelevant band-aid. As Frances Fox Piven has said of socialist-feminist accounts of the welfare state, the determination to argue the view that the state exercises control over women, replacing the declining patriarchal family with an equally patriarchal relationship to the state is 'often much stronger than the evidence for it' (Piven, 1984: 15).
There is a fine line between being sensitive to the injuries of class, gender and race, both hidden and obvious, and simply assuming the impotence and subordination on the part of those we see as controlled and domination, either unable to identify their sources of social power, or limiting them to resistance. If one looks closely at the workings of the state in an area such as child welfare, one does not see the bourgeoisie, professional welfare experts, or the state bureaucracy assuming control of the reproduction of working- class culture, but a much denser and untidier mixture. One sees parents and children being driven to state agencies by the exigencies of poverty and single parenthood in a world hostile to dependency and a labour market hostile to women; children being treated harshly and regimented in institutions, but often with the support and tacit consent of their own parents, relatives and neighbours; families being coerced into particular patterns of behaviour by bureaucrats, but bureaucrats with working-class backgrounds mobilising moral categories familiar and acceptable to their working-class neighbours.
The perception of the formation of culture which informs the social control paradigm in historical sociological studies of social welfare owes much to Marx's famous remarks on ideology, that the idea of the ruling class are always the ruling ideas, that the class controlling material production also controls ideological production, so that generally 'the ideas of those who lack the means of mental production are subject to it' (Marx & Engels, 1976: 59). Working-class culture is in this formulation doomed to be overshadowed by the ruling class and the state, and at best confined to opposition and resistance. If there is any agreement between bourgeois and working-class culture, it occurs because the working class have been socialised, coerced, or gently persuaded into it. Within the logic of dominating/dominated, the only possibility for a genuine working-class culture lies in the revolutionary subject and the occurrence of a socialist revolution which overthrows capitalism's basic relations of dominance. Even in the more sophisticated Marxist analyses of ideology influenced by the work of Antonio Gramsci, his concept of hegemony slips and slides from the original meaning of a coalescence of different ways of life into a single one, to the notion of a dominant ideology imposed on the working class. This is most clearly manifested in the persistent use of terms like 'bourgeois hegemony' or 'middle-class values'. This logic also persists in the literature which argues against Marxism's primary focus on class, to also examine gender and race, for the basic assumption of the ideological and cultural powerlessness of the controlled remains essentially the same.
However, Marx also made some other, equally famous remarks on the formation of consciousness which provide a different understanding of the culture of the dominated. Consciousness, he argued, 'can never by anything than conscious existence, and the existence of men is their actual life process.' His materialist starting point is that consciousness be taken as 'the real living individuals themselves, and consciousness is considered solely as their consciousness' (Marx & Engels, 1976: 37). Gramsci's understanding of hegemony includes the winning of the consent of the dominated, and this often involved their quite active participation in the construction of the 'dominant' ideology. Richard Gray has argued for England that hegemony should not be seen as a 're-moulding' of the respectable strata of the working class in the bourgeois image, as whatever integration took place was 'a two-edged affair', in the sense that the aristocracy and the middle- class also had to alter their norms and behaviour in response to a moral critique which cut across class boundaries (Gray, 1977: 87-8). The particular form taken by features of everyday life such as family and work morality are thus never simply 'bourgeois' or 'middle-class', their concepts shaped as much by working-class as by bourgeois experience, politics and culture. As Steven Spitzer has put it recently, in throwing out the 'bathwater' of functionalist socialisation models, it is important to avoid abandoning the 'baby' that connects social control to a wider context of social organisation, a connection which 'is mediated as much by the consciousness, desires and goals of those who are to be controlled as the interests and agendas of the controllers' (Spitzer, 1987: 57). To argue that the state 'constructed' a respectable family morality within the working class as a whole, that the cultural and ideological divisions between the rough and the respectable working class were 'major achievements of the forces of discipline and moralisation' (Garland, 1985: 39), or that 'the family is a product of liberal family reform' (Minson, 1985: 184) is empirically false, and to the extent that analyses of the history and sociology of the welfare state are in danger of promoting these arguments, explicitly or implicitly, they should be approached with circumspection. Working-class familial respectability was a relatively workable cultural response to difficult and changing social conditions, and pre-dated organised attempts to reform working-class familial relations and morals on any sort of mass, and therefore effective, scale (e.g. Houlbrooke, 1984; Macfarlane 1986). The notion that state agencies imposed a certain familial ideology on a small proportion of deviant families is more plausible, but still requires close empirical examination, given that we know much more about the goals of state action than we do about its effects. If we perceive the ideology underlying social welfare as bourgeois or middle- class, and by definition alien to working-class culture, we come uncomfortably close to suggesting that the working class as a whole wanted, and perhaps still want, to be dirty, drunk, poor, unemployed, unmarried, and to bash and abandon each other to destitution. One can of course argue that these labels were often attached unfairly, and this was often what welfare clients objected to in the way that social welfare operated, but the notion of a just social control remains quite some distance away from the current social control paradigm.
The historical critiques of theories of social control suggest that it is essential to develop a sensitivity to the ways in which particular patterns of family life and social interaction have developed in relation to a given economic and cultural environment, independently of, as well as in response to, state action and attempts at social and psychological regulation (an excellent example here is Reinhard Sieder's (1986) essay). The kind of explanatory logic which responds to this requirement is not one based on concepts like control, regulation, repression, domination or administration, but one of asymmetrical negotiations, alliances and compromises occurring within structured fields of power relations (De Swaan, 1988: 248; Rose, 1987: 71-3). One cannot properly realise the notion that power is not a thing which some people, groups or institutions possess, but rather a social relationship, without also jettisoning the zero-sum logic which lies at the heart of notions not just of social control, but also of policing, regulation and colonisation, and perhaps ultimately the very notion of the dissolution of the distinctions between state and civil society. Those who are disadvantaged in a power relation continue to act and have agency, even domination is a process of interaction, as Georg Simmel argued in 1908:
Within a relationship of subordination, the exclusion of all spontaneity whatever is actually rarer than is suggested by such widely used popular expressions as 'coercion', 'having no choice', 'absolute necessity', etc. Even in the most oppressive and cruel cases of subordination there is still a considerable measure of personal freedom...Interaction...exists even in those cases of superordination and subordination...where according to popular notions the 'coercion' by one party deprives the other of every spontaneity, and therefore of every real 'effect' of contribution to the process of interaction (Simmel, 1971: 97- 8).
More recently, Anthony Giddens echoed this point in an argument for the notion of a 'dialectic of control', when he wrote that 'all relations of autonomy and dependence are reciprocal; however wide the asymmetrical distribution of resources involved, all power relations express autonomy and dependence 'in both directions'' (Giddens, 1982: 39; also Hall, 1986: 265). Unless the agency of the powerless (less powerful?) is treated as autonomous to some extent, even and especially when it consolidates rather than resists existing power relations, we have returned to the logic of dominated/without power versus dominant/with power, thus unable to explain coherently the persistence of social order generally, and in particular the role that state action played, or did not play, in the constitution of society.
The problem has been neatly captured by Bruno Latour (1986) in his distinction between a 'diffusion' and a 'translation' model of power. In the diffusion model, essentially that underlying the social control paradigm, a person or group is endowed with power, which enables them to diffuse their will - in the form of what Latour calls a 'token', orders, claims or artefacts - throughout the surrounding social space, which is regarded as medium with greater or lesser resistance, such as poor conviction, inertia, opposing interests, and so on. The occurrence of a phenomenon is explained in terms of the power of those who possess it, with the objects of that power only acting to either transmit it (through obedience) or to resist it. Latour's argument against this model is that the 'medium' of this transmission of power - the individuals upon which it acts - are not a medium at all, but constitute the very stuff of the action taking place, they are all doing something essential for the existence and maintenance of power, and they all 'shape it according to their different projects' (p.268). As he puts it:
'Power' is always the illusion people get when they are obeyed; thinking in terms of the diffusion model, they imagine that others behave because of the master's clout without ever suspecting the many different reasons others have for obeying and doing something else; more exactly, people who are 'obeyed' discover what their power is really made of when they start to lose it. They realise, but too late, that it was 'made of the wills of all the others' (pp. 268- 8).
In this sense power is not the cause of collective action, but its consequence, or simply an aspect of it, and the causes must be sought in something more complex, namely the collision and combination of a constellation of projects and strategies, even if unequally resourced and practically effective (see Callon, 1986 and Law, 1986 for discussions of the translation model of power).
Conclusion
In order to take seriously the problems associated with 'social control'
as an organising sociological concept, it is important to go beyond heaping
scorn on it towards making some more fundamental changes to our conceptual
apparatuses. The problems lie not simply in the term itself, but in a particular
orientation towards the state, civil society, and human action. The lack
of fit between the two conceptual tendencies I have identified - state-centredness
and objectification - and the more densely-textured reality of state-society
relations demand, first, a much more complex understanding of ideology,
culture and social change, one which does not see the state or the professional
middle class as the centre of power or the focal point of all processes
of stability and change. In their anxiety to 'bring the state back in' (Evans,
Rueschemeyer & Skocpol, 1985), many writers seem to have developed a
peculiar amnesia about the fact that the social world does not revolve around
the state, and to agree with this point at a theoretical level is no guarantee
against the Phoenix-like persistence of the social control paradigm in the
substance of the analysis. Lenin would probably not have agreed, Stalin
certainly not, but we can do worse than take more seriously than we have
Marx's argument that 'only political superstition believes at the present
time that civil life must be held together by the State, when in reality
the State is upheld by civil life' (Marx, 1975: 121).
Second, it also remains necessary to base specific studies of state action on a more sophisticated understanding of the contribution of the dominated, managed, policed, socially controlled and disciplined to the construction and development of power relations. It is possible to 'de-centre' the operation of power while still seeing human subjects as the objects of power, capable only of resistance, to slip from the analysis of the social construction of subjectivity to a denial of that subjectivity. The closer one looks at the operation of power, the clearer it becomes that we need to include not only the desires and aspirations of everyday subjects, but also the possible realisation of those aspirations. State agencies and policies are themselves produced and constantly re-produced within a framework of broader social structures and process, within a complex constellation of human actions, desires, efforts and projects, and we are ill-equipped to understand either the stability of existing social relationships or how and why they might change while we remain attached to a view of human beings as the objects, whether passive, resistant or seduced, of control, social engineering, management and discipline.
Acknowledgments
Much of this paper was written during a period of study leave supported by the University of Sydney's Special Studies Program. Earlier versions were presented at the Sociological Association of Australia and New Zealand's Conference in 1988, and at the International Sociological Association's Congress in 1990. I would like to thank those who have offered comments, criticisms and support, especially Michael Muetzelfeldt, Heinz Sünker and the two anonymous reviewers.