Civil society: Bringing the family back in

ABSTRACT This article explores the complex and contradictory positioning of the family within civil society literature. In some accounts, the family is seen as the cornerstone of civil society. In others, the family is positioned firmly outside – even antithetical to – civil society. This paradox arises from the ways in which civil society is variously defined through a series of binary oppositions – in relation to each of which the family sits uneasily. And while feminist critiques have tried to bring women back into view, they too tend to marginalize the family. In addition, the normative nature of these oppositions has meant that while civil society tends to be seen as the property of the political ‘left’, the family is often associated with the political ‘right’. The article argues that we need to move beyond oppositional definitions of civil society and assumptions about the family if we are to understand the multiple ways in which the family is implicated as not only the ‘reproducer’ of particular resources and dispositions but as a principal source and focus of civil society engagement and activism.


Introduction
In attempting to chart the social history of the family, Davidoff, Doolittle, Fink, and Holden (1999, p. 51) point to the simple conundrum that 'the family is everywhere and the family is nowhere'. Much the same can be said of the family's paradoxical position in relation to civil society. For despite the family's central role in social, cultural (and biological) reproduction, it is largely absent from the majority of contemporary literature on civil society (Ginsborg, 2008;Howell, 2006). And while there has been increasing recognition of the gendered dimension of the concepts and practices associated with civil society, a quick scan through the Contents and Index pages of most edited collections on civil society (e.g., Eberly, 2000a) indicates only fleeting, if any, coverage of the family. Similarly, there is a marked absence of references to civil society in edited collections on the family (e.g., Allen, 1999).
In this article, we argue that, if we want to increase our understanding of civil society and how it is changing, we need to give the family a far more central role than it has had hitherto. The family, we suggest, is significant for our understanding of civil society along a number of dimensions. Firstly, the family is an important (if not the most important) agent in cultivating those dispositions and capabilities that foster civic engagement. Secondly, the family can be seen as one of the keys providers of mutual welfare. Thirdly, the family is in itself a site and focus of civil activism. And finally, civil society works to support and reconfigure families in important but overlooked ways. In short, changes in the family change civil society, and changes in civil society change families.
However, if we are to grasp the complex and mutually-constitutive relationship between the family and civil society we need to move beyond ideologically driven and oppositional representations both of the family and of civil society. The authors begin by outlining the shifting and contradictory ways in which the family has been positioned in conceptualisations of civil society, as well as feminist critiques of how these conceptualisations have 'relegated' women to the domestic sphere.
While these feminist critiques have subsequently contributed to a more widespread recognition of women's role in (and exclusion from) civil society, the position of the family remains under-developed. In part this is because of the conjuncture of two converging and oppositional attributes associated with civil society and the family. Civil society, certainly within European-inspired social theorisations, is commonly seen to be the property of the political 'left'. The family, on the other hand, is commonly seen to be the property of the political 'right'. The family is therefore not only 'outside' the ambit of civil society, there is often the tacit assumption that strong family ties may even operate against the development of a 'healthy' public sphere.
However, these assumptions are based on normative definitions of civil society and of the family rather than empirical evidence. We argue that the family needs to be 'brought back in' and consider how this might be done. In particular, we look at some emerging attempts to develop an analytical framework, and particularly that of Mikiko Eto (2012), who extends Iris Young's associative spheres to include the family. The authors argue that while Eto's framework marks a step in the right direction, it does not go far enough. The family does not only 'socialize' its members into different orientations towards civil society, it is in itself a site of civil society engagement and activism.
Where is the Family in Relation to Civil Society?
The positioning of the family within conceptualisations of civil society is both shifting and contradictory. In many accounts, and especially (but not exclusively) in American political sociology which tends to privilege the non-governmental aspects of civil society, the family is seen as an essential component. For example, Eberly (2000b, p. 3) describes civil society as 'consisting of families, neighbourhoods, voluntary associations, and an endless variety of civic enterprises'. Indeed, Eberly and Streeter (2002, p. 3) claim that during the twentieth Century these networks have been subject to systematic destruction from 'statist ideologies'. Similarly, Carter (1999), in his account of the decline of civility in the US, contrasts the qualities of the family with the 'legalism' of the welfare state which he believes has led to the 'erosion' of American society. Carter (1999, p. 230) upholds the virtues of the families, which, he argues, are the first 'civil societies' marked out by respect and caring for others. Cohen and Arato (1994, p. 724 n81), coming from a very different perspective, state that 'the family is a core institution in and of civil society'. Like Carter, they see it as the 'first association of 'civil society', which 'if conceived of in egalitarian terms, could have provided an experience of horizontal solidarity, collective identity, and equal participation to the autonomous individuals comprising it' (p. 631 n48). On this side of the Atlantic, somewhat exceptionally, the British sociologists Halsey and Young (1997, p. 785) also argue that the family needs to be recognized as the cornerstone of associative activity and mutualism. They argue that the family should be seen as: … a small collective of a special kind, the emphasis is on co-operation rather than competition, and on long-term commitment rather than choice … It teaches people the most precious ability of all, the ability to transcend self-interest and regard the interests of others as in some way their own: the kind of altruism which is at the heart of the collective consciousness and which holds all societies together.
However, in general, conceptualisations inspired more by European social and political thoughtparticularly that of Hegel and Habermas, have positioned the family outside of civil society. In these accounts the public dimension of civil societyrather than its non-governmental attributesis privileged. Civil society here is the space for political participation, debate and opinion formation within the public sphere. Although there is no consensus between theorists on the potential of civil society to effect significant social change, they all concur that the familyand particularly the modern nuclear familyis not part of civil society. Indeed, the family is often seen as the cradle of self-interest rather than selfless mutualism and can even be portrayed as inimical to civil society engagement.
There are a number of factors that appear to contribute to this paradoxical positioning of the family. One is the way in which conceptions of civil society are constructed around or between different sets of binary oppositions: the 'public' and 'private'; the 'state' and the 'market'; and where it is characterized by 'associative' rather than 'coercive' relations and practices that are 'voluntary' not 'obligatory'. The family sits uneasily in relation to each of these oppositions.
For instance, in relation to definitions based on the distinction been the 'public' and the 'private', civil society is nearly always associated with the 'public'. The concept of the 'public sphere', and particularly as expounded by Habermas (1989), embodies a discursive space that is (ideally) open and accessible to all and where issues of public interest are raised in order to pursue the common good. Private interests are inimical to the pursuit of the public good and need to be bracketed out. Within this conception of civil society, the family can only be located within the private sphere. Far from contributing to the public sphere, the self-interest associated with families may even be seen as working against the development of a strong civil society. Pateman (1980, p. 24) in her critique of the way in which women and families are positioned within much social theory shows how: … the family itself is a threat to civil life. Love and justice are antagonistic virtues; the demands of love and of family bonds are particularistic and so in direct conflict with justice which demands that private interest is subordinated to the public (universal) good.
It follows that pursuing the public good will mean the diminution of the family. Indeed, as Donzelot (1979, p. 5) points out, utopian socialists believed that 'its disappearance was programmed for a dawning socialism, and its partial breakup, its crises, were considered so many signs heralding the latter's arrival'.
However, this division between the public and the private has been severely critiqued, particularly by feminists (e.g., Fraser, 1997;Landes, 1988), who point out its gendered assumptions. Far from being 'open' and 'accessible' to all, the public sphere has historically been based on exclusions. As Phillips (2002, p. 72) points out: 'civil society is often presented in terms that make it seem like a place where women are not'. Not only do these accounts often fail to recognize the significant contribution that women have made to the development of civil society, the bracketing of 'private' interests into the domestic domain denies the political dimensions of the gendered division of labour and unequal power relations within households and families. As feminists have long argued, it is important to recognize that the personal is political.
The family also has an ambiguous relationship with definitions of civil society based on the distinction between 'the state' and 'the market'. The conception of the family as selfinterested and essentially 'private' emphasizes its economic function. Indeed, traditional libertarian approaches see 'the family as the protector of private property, of the bourgeois ethic of accumulation, as well as the guarantor of a barrier against the encroachments of the state' (Donzelot, 1979, p. 5). Certainly, the family is often presented as a unit of consumption that is deeply embedded in the neoliberal 'project'. Abbott and Wallace (1992, p. 16), for example, argue that the frequent emphasis within conservative political regimes on 'family values' and 'pro-family policies' is 'essential for the maintenance of capitalism'.
However, in contrast, others emphasize the increasing incorporation of the family into the state. Donzelot's (1979) historical account of the 'policing' of families shows how governments have increasingly intervened to mould families into specific functions. He identifies a succession of regulatory interventionsparticularly around childrenthat have been put in place in order to 'tame' the family. He argues that the protection of children 'allowed for the destruction of the family as an island of resistance'. (Donzelot, 1979, pp. 93-94). Here, it could be argued that while the family was embedded in civil society in the past, it is no longer.
As Cohen and Arato (1994, p. 533) point out in relation to its paradoxical position, the family is alternatively seen to be subjected to either the economic pressures of the market or subsumed within the bureaucratic pressures of the state. Part of the difficulty arises not only from the slippery conceptualization of civil society, but because of the changing nature of the family itself. As Howell (2006, p. 46) argues, in considering the location of the family in relation to civil society, it is important to consider 'the cultural specificities of the scope and the social, economic and political significance of the family and household'.
Just as the family is ambiguously positioned in relation to definitions of civil society based on public/private and state/market oppositions, it also is ambiguously positioned in relation to the definitions that emphasize the associative (as opposed to coercive) dimension of civil society. As we have seen, those who argue that the family is located firmly in civil society highlight its associative qualitiesemphasizing its 'horizontal ties' and collective responsibility. However, others reject this representation of the family and focus instead on the way in which families are based on 'vertical ties' and unequal relations. Walzer (2002, p. 35), for example, argues that civil society should include 'all social groups that are or can be understood as voluntary and non-coercive, thus excepting only the family, whose members are not volunteers' [our emphasis].
And while feminist perspectives argue for greater recognition of women in civil society, they do not generally include the family. As Fraser (1989, p. 120) argues with reference to male-headed nuclear families: Such families can be understood as normatively secured rather than communicatively achieved action contexts, that is as contexts where actions are (sometimes) mediated by consensus and shared values but where such consensus is suspect because it is prereflective or because it is achieved through dialogue vitiated by unfairness, coercion, or inequality. [Our emphasis] However, just as the family's relationship with the state and the market has changed over time, so too have the internal dynamics and composition of the family. Barber (1998, p. 54) for instance, claims that particular kinds of contemporary families might legitimately be considered part of civil society, particularly those 'which are open and egalitarian in the long term because they assure equality among the various roles within them … and which eventually produce autonomous adults'. Distinctions are also made between different kinds of family. Eto (2012, pp. 113-114), for example, argues that families should not necessarily be 'barred' from civil societybecause they are becoming increasingly diverse: 'A heterosexual couple with biological children is not the only form a family can take … same-sex couples with adopted children, and remarried couples with children from a former marriage' are 'closest to associations' and may therefore be included.
This leads us to the rather odd situation where some families (democratically-organised ones, those based on same-sex partnerships) are to be included within the ambit of civil society, whereas others (and in particular male-headed nuclear families) are not. This is a rather unsophisticated assumption because it assumes that male-headed nuclear families will inevitably have coercive practicespractices which will apparently be entirely absent in alternative families.
The discussion of whether families should be considered part of civil society on the basis of their internal composition reveals the need to take on board, as Howell (2006) argues, the contextual specificities of social arrangements. The discussion also highlights the very normative nature of the concept of civil society itself.

The Normative Dimensions of Civil Society and the Family
The problem with trying to clarify the relationship between the family and civil society is that they are both highly-loaded conceptsprone to competing claims as to their value and purpose and for which there is frequent slippage between the 'is' and the 'ought'. As Zimmer and Freise (2007, p. 21) outline, civil society is a highly normative concept directed towards a 'utopian programme' that seeks to increase democratic engagement and bring about greater social justice. And while the study of civil society clearly has an empirical dimension, this is often infused with particular ideological and political preferences. As Barber (1998, p. 12) points out, civil society 'tells us something about how we actually do behave even as it suggests an ideal of how we ought to behave'.
The normative dimensions of civil society are evident in Table 1, which outlines how the implied positive attributes which are often associated with it are contrasted with the far less positive attributes of the family.
However, the positive attributes associated with civil society may be more imagined than real. As Fraser (1997) and others have argued, private interests cannot just be bracketed outbecause the public sphere has been constituted by such interests. And neither can inequalities. As Phillips (2002) points out, members of associations will bring with them their existing inequalities of resources, information and power. Indeed, she argues that the lack of regulation within civil society associations may make them 'more coercive and less protective of individual inequalities and freedom than the much-despised institutions of the state' (2002, p. 81). Certainly, there have been a number of scandals about sexism and misogyny in traditional left-wing political associations (e.g., Penny, 2013).
Moreover, there is no hard and fast dividing line between what counts as an issue of public concern and what is 'only' a private issue. For example, the decision as to whether a public protest is interpreted as an expression of 'nimbyism'and therefore in the interest of some private (selfish) agenda as opposed to one of public activism (selfless) is probably more about whether we approve (or otherwise) of the causerather than any essential difference in the nature of the activity. Similarly, what counts as democratic engagement and greater social justice is not always clear-cut. Although these terms are commonly associated with movements on the left, those on the right can also make legitimate claims to be striving for greater representation and justice.
The evaluative nature of judgements about the family and civil society is sharply illustrated in accounts of the tensions between 'women's groups' and 'pro-family' groups. Glasius (2004), in her research on the way in which delegates from 'women's' and 'profamily' movements interacted with state delegates in the negotiations of the International Criminal Court, concedes that both of these movements inhabit 'civil society'. However, there is clear partisanship in her representation of the two camps. On one side are the 'social justice and women's rights activists' and on the other are the 'conservatives, antiabortionists, and religious fundamentalists'. Moreover, she argues that knowing more about the pro-family movement 'will be helpful to the transnational women's rights activists who continually have to confront them' (p. 224). It is possible that both movements are concerned with issues of social justice and women's rights, but they have very different starting points.
In general, there is a tendency within much of the civil society literature to assume that the family is the property of the political right and that civil society is the territory of the political left. But this kind of positioning is extremely problematicand ignores the extent to which 'pro-family' activism can have its roots in more left-leaning causes. As we shall return to later, while 'pro-family' movements may generally be seen as holding conservative values, this is not necessarily the case. For example, in the US, the FamilyValues@Work coalition is a network of over 2000 grassroots organizations that lobbies for 'family friendly' workplace policies such as paid sick days and family leave insurance.
The difficult question of the how the family should be understood in relation to civil society cannot be resolved through reconceptualisation alone (or by case-by-case judgements of individual families) because the concepts of the family and society are generally too ideologically-loaded for the issue to be resolved theoretically. Rather the relationship needs to be addressed empirically. That will mean setting aside many of the assumptions about the family, about justice and about civil society and making the family the central object of scrutiny.

Bringing the Family Back in
While feminist critiques of conventional understandings of civil society have emphasized the significance of bringing women out of the confines of the domestic sphere and into the public sphere, they have generally been brought in without their family 'baggage'.
In this section, we explore the various ways in which we can bring the family back in to our understandings of civil society, and in particular we examine the model developed by Mikiko Eto (2012). Eto attempts to overcome the public-private dichotomy, which largely compartmentalizes women into the private sector, by drawing on and expanding Iris Young's concept of civil society. Young (2000) rejects the simple 'spatial' division into 'spheres' by proposing three kinds of 'associative activities': private, civic and political. Private associations, which include families, social clubs, and private gatherings, are concerned with 'basic matters of life, death, need and pleasure' (p. 160). Civic associations include a broader range of voluntary associations and differ from private associations in that they are open to outsiders. Political associations are self-explanatory, and obviously include political parties as well as a range of lobbying organizations. In addition to these three kinds of activities, Young argues for a distinction to be made between associations that are 'self-organizing' and those which are 'public'. Generally, the more 'private' the activity, the more it is likely to be self-organizing. The more civic and political, the more likely it is to operate in 'public'.
Eto takes on board Young's rejection of a simple spatial division but questions Young's continuing location of family as a private activity. Eto modifies the model (Figure 1) so that it now includes a series of associative activities that link family and state institutions. Each of the five overlapping elements interacts with its neighbouring element. In this way, Eto seeks to provide 'an expanded conception of civil society which is not isolated from everyday experience or from the influence of state political institutions and attempts to explain how the consciousness, expectations, and demands emerging from citizens' everyday life are transformed into political associations, specifically social movements and citizen's interest groups, and how their activities then interact with state political institutions' (2012, p. 114).
In conceptualizing the ways in which the family, as part of citizens' everyday life, connects to the public sphere, Eto argues that the family acts 'as an important gateway to civil society activities'. In particular, she claims that 'the family holds the potential to be an active player in creating associative activities through its role as a prospective base in civil society'. She goes on to argue that 'the family provides individuals with the basis for developing their social awareness, and the family could therefore contribute to nurturing active citizens in civil society '. (p. 114) Without wishing to deny this crucial role, the authors argue that Eto's reconceptualisation has a number of difficulties. Firstly, although her model tries to soften the divide between the private and public spheres, she still locates the family at the margins of civil society. It is 'independent from, but open to, civil society as the basis of prospective associative activities' (Eto, 2012, p. 117). The grounds for largely excluding the family are again based on the condition of voluntary association: Many if not all marriages are the product of the free will of the individuals; children, however, have no free will in choosing which family they are born into. The family whose relationships lie in kinship ties, seeks to defend private concerns, often limited to family members, rather than public interests shared by communities or the wider society.
As argued, this sharp division between free will and coercion, between private concerns and public interests, is not really sustainable. Despite Eto's assertion that there are 'few families who regularly discuss serious social problems … at their dining tables' (p. 113), she herself identifies instances where families play a role in social activism when their concerns coincide with public issues (pp. 118-119). There is also some inconsistency and romanticism in Eto's representation of the family. Despite its coercive properties, she claims that: 'the essence of family ties is love … civil society associations are sustained by "solidarity"' (p. 113). Whether these descriptions of both family ties and civil society are empirically accurate must be open to question.
However, despite these idealisations and minor inconsistences, Eto's model marks an important step in the right direction. In particular, it shows that the family is intimately linked to a range of private and civic associations. This article wants to give the family and one which recognizes the family as not only the 'gateway' to prospective civil society activities, but a site of civil society engagement itself.
This article proposes that the family has a more significant role in relation to civil society than simply socialization. In constructing the analytical framework (Figure 2), this article has drawn from Foley and Edwards' (1998) distinction between different functions of civil society: socialization, welfare and representation. However, this study has extended their model to emphasize that the family is not only the source of dispositions and skills that may promote civic engagement, but is itself the site of civic activism. This study also wants to highlight within its framework the mutuality of the relationship between civil society and the family. Just as the family supports various forms of welfare and associational activity, so too do civil society activities support the welfare of diverse kinds of family arrangements.

The Family and Socialization into Civil Society Engagement
Foley and Edwards (1998, p. 5) claim that one of the main functions of civil society is ' … building citizenship skills and the attitudes crucial for motivating citizens to use their skills'. Clearly, the family has an important role to play, perhaps the most important role, in building these skills. At the most fundamental level, the family matters in relation to civil society because of its role in socializing the young. As Howell (2006, p. 47) points out 'the family shapes norms and practices in the sphere of civil society'. It serves as the 'gateway' (Eto, 2012, p. 114) to future civil society activity through the formation of attitudes, of skills and of affiliations with civil society organizations.
In relation to the family's cultivation of particular dispositions, Young (1997) provides us with the example of the feminist and social activist bell hooks, who attributes her own capacity for critical engagement to the family environment in which she grew up. But this kind of anecdotal evidence is supported by extensive large scale analysis of the intergenerational transmission of attitudes towards a range public issues. For example, Sherkat and Blocker (1994) found strong continuity between the political orientations of many civil rights activists in the US and their parents. Similarly, and in recognition of the fact that civil society activism does not always have to be of the 'left', Coffé and Voorpostel's (2010)'s analysis of voting patterns of the radical right in Switzerland finds a relationship between parents' political preferences and attitudes towards immigration and those of their children. Quintelier's (2015) study of Belgian families found high levels of intergenerational transmission of political participation intentionand particularly among families of relatively high socio-economic status. In terms of skills, a substantial body of evidence shows that families are significant for passing down a range of social assetsoften conceived as different kinds of 'capitals' (e.g., cultural and social)that are important for all kinds of social engagement. Of particular relevance here is Bengtson, Biblarz, and Roberts's (2002) longitudinal analysis which highlights the intergenerational transmission of self-esteem and self-confidenceattributes which are likely to be significant conditions for civil society engagement. We also know that there is a strong intergenerational dimension to membership of civil society organizations. Affiliations to religious organizations are clearly family-driven, but so too is membership of trade unions (Bryson & Davies, 2017).
While these various studies indicate the significance of the family in the transmission of dispositions and resources that are crucial for civil society engagement, we need to undertake further research on whether particular kinds of family arrangements influence particular kinds of dispositions and skills. To what extent is socio-economic status and the availability of particular familial assets and capitals a necessary or sufficient condition of civil society engagement? And do the increasingly 'diverse' family arrangements heralded by Eto contribute to the development of distinctive sensibilities and skills?
The Family and Welfare The second principal function of civil society identified by Foley and Edwards (1998, p. 5) is ' … to heal the sick, counsel the afflicted, support the penniless'. And in this area too, the family is significant. As Howell (2006, p. 49) points out 'it is in the household that people have their first experience of association. It is here that they develop a sense of empathy towards others, trust in 'strangers', a sense of citizenship and responsibility towards those beyond their immediate household or family unit' In addition to providing welfare for family members, the family is a significant unit for the provision of welfare within wider communities. For example, Power and Willmot's (2007) Neighbourhood Study based in East London and the North of England highlighted the significance of between-family and between-household networks in the provision of support. The household may also be increasingly engaged in the direct provision of forms of welfare that have hitherto been the responsibility of the state. The growth of home education, for example, and the related rise of support networks and organizations (Fensham-Smith, 2017), such as Education Otherwise, represent a trend to take education away from government institutions.
In relation to more formal forms of welfare provision such as volunteering, research from the U.S. (Hodgkinson & Weitzman, 1992) has shown that individuals are twice as likely to volunteer if their parents volunteered. Using panel data, Mustillo, Wilson, and Lynch (2004, p. 538) argue that 'volunteering runs in families, and that family units are important careers of the volunteer tradition'. In terms of 'supporting the penniless', there is also a strong intergenerational dimension to charitable giving. Wilhelm, Brown, Rooney, and Steinberg's (2008) analysis of panel study data in the US found that a strong correlation between parents and children in the amount of giving. While this was particularly marked in relation to religious giving, there was also a positive association for secular giving, for example for poverty relief. They conclude that generosity emerges, at least in part, from the influence of parental charitable behaviour.
Again, though, the relationship between socialization and welfare needs further investigation. Is it, as Moen, Erickson, and Dempster-McClain (1997, p. 283) argue, that the inheritance of social, cultural and economic resources is more important than the inheritance of values? And do different kinds of family arrangements lead to different kinds of engagement in welfare activities?

The Family and Representation
Finally, the third dimension of civil society ' … gives identity and voice to the distinct interests … stimulates public debate and presses government for action on a thousand and one matters of public interest'. (Foley & Edwards, 1998, p. 5). It is perhaps in the area of political activism that the relationship between the family and civil society is most overlooked. There is, though, more than one aspect to this. This article want to argue that the family is not only the cradle of dispositions and the provider of welfare within civil societyit is in itself a site and focus of activism. As mentioned earlier, social movements that are 'pro-family' are often seen to be essentially conservative in character and, some might argue, seek to take decision-making away from the public sphere. However, they are many other forms of activism in relation to the family that are being fought over in publicand not just those that can be classified as conservative and religiously-affiliated. The ways in which the family is changing have meant that 'pro-family' movements are not simply a struggle between leftist and rightist movements. For example, increasing rates of marital breakdown have seen the growth of organizations campaigning for the welfare and interests of single parents. In the UK, Gingerbread is committed to 'supporting, advising and campaigning with single mums and dads to help them meet their family's needs' (https://gingerbread.org.uk/). Similarly, the US-based Parents Without Partners pursues the interests of single parents whether they be 'male or female, custodial or non-custodial, separated, divorced, widowed or never married' (http://www.parentswithoutpartners.org/).
The family is also at the heart of men's movementsand importantly not just men's movements but fathers' movements. Most famously, in the UK, Fathers4Justice (http:// www.fathers-4-justice.org/) has been running a high profile campaign and acts of civil disobedience order to change family law, which it believes unfairly discriminates against men in gaining access to their children after divorce. Over the Atlantic, the Fathers Rights Movement describes itself as a 'collective movement' -'passionate about empowering fathers to stand up for their rights and to educate the public and family court system about the importance of fathers in society, as well as bring greater awareness to the imbalance and injustice that effects the rights of fathers' (http://fathersrightsmovement.us/about/). Demographic changes, the increasing employment of women and incidence of marital breakdown have also brought about new forms of activism from grandparents. Organizations such as Grandparents Plus (http://www.grandparentsplus.org.uk/) in the UK and the National Association for Grandparenting (http://www.grandsmatter.org/) in the USA have, like the fathers' movements, been increasingly campaigning for their legal rights and recognition in the event of the dissolution and reconfiguration of family arrangements.

The Mutuality of the Family-civil Society Relationship
The emergence and actions of these various campaigning movements reveal not only the way that the family is implicated in supporting civil society, but also the way in which civil society is implicated in supporting the family. Just as families may well strengthen rather than weaken civil society, so too may civil society organizations seek to strengthen rather weaken the family. As Figure 2 tries to indicate, the flows of influence are two-directional.
Families may be primary socialization institutions, but civil society also works to support families in these educative processes through myriad activities and networksfrom selfhelp groups such as Mumsnet (www.mumsnet.com/) to supplementary and Sunday schools. In terms of welfare, while families certainly provide different forms of welfare outside the immediate remit of the family, civil society organizations reciprocally provide welfare for families. For example, in the UK, Family Action (www.familyaction.org.uk) offers a range of community-based services and financial support to 'strengthen families and communities'. 'Save the Family' (www.savethefamily.org.uk/) supports parents who have had their children 'taken into care' in rebuilding their lives so that the family can be reunited. The mutuality of the relationship can be seen most clearly in the social movements that campaign for family rights (mentioned above).
Here the boundary between the family and civil society is very blurred indeed.
Seen in this light, it is very hard to see the relationship between the family and civil society as one of conflict and opposition. The relationship between the family and civil society is not only close but to some extent mutually constitutive. Moreover, these developments require us to examine not only how the changing nature families has reconfigured civil society, but also how movements in civil society have had a bearing on changes in the family.

Conclusion
This article began by outlining the paradoxical positioning of the family in relation to civil society. In some accounts, the family is the cornerstone of civil society. In others, the family is not only outside but even antithetical to civil society. This article have argued that this paradox arises from the ways in which civil society is defined through a series of binary oppositionsin relation to each of which the family sits uneasily. In addition, the normative nature of these definitions has meant that while civil society tends to be seen as the property of the political 'left', the family is usually associated with the political 'right'.
Like Howell (2006), article argue that the relationship between the family and civil society needs more than the largely superficial consideration it has had so far. But it appears that the any more in-depth consideration is a matter of empirical enquiry rather than theorization alone. Certainly it is important to move beyond idealized notions of either the family or of civil society. The nature of the relationship between families and civil society, and the conditions and consequences of this relationship are likely to be complex and shifting. They will require us to chart not only how changes in the family change civil society, but how changes in civil society have changed families. After all, as Bengtson et al. (2002, p. 167) conclude in their analysis of How Families Still Matter: 'The family is the fulcrum balancing change and continuity over time in human society'.

Disclosure Statement
No potential conflict of interest was reported by the authors.

Funding
This article is based on research supported by the Wales Institute of Social & Economic Research, Data & Methods (WISERD). WISERD is a collaborative venture between the Universities of