ARTICLE SUBVERSIVE INTENT: A SOCIAL THEORY OF GENDER Zarina Maharaj Introduction Gender inequality is die structured (institutionally specified) inequality of access to material and non-material social resources between the sexes, generat- ing male privilege and domination and female subordination in society. This recognition that the social power of men over women is a culturally constructed one, insidiously operating as it does through society's structures, gives rise to a critical question, one of theory and practice, for feminists driven to find ways of changing the status quo: how do we create the conditions for gaining equality of access to social resources? What are the social issues around which we should be mobilising for struggle to achieve this goal of empowerment? These questions of feminist politics must be informed by sound theory about the structural dynamics of a society in general and its prevailing gender order in particular. Such a social theory of gender is currently in the making. In being underpinned by the notion of social structure as historically constituted by social practice, this theory aims to account for gender relations in a way that is challenging and transforming mainstream social and political theory. Feminist theories rooted in such practice-based structural approaches to understanding gender have in this sense made 'feminism' academically respectable and women's subordination amenable to theory-based political strategy and action for change. This article outlines and discusses the main conceptual underpinnings of such a practice-based structural approach to accounting for the experiences and interaction of women and men as gendered beings. The power of such a practice-based approach emerges in the framework it provides for the analysis of the social relations of gender in any socio-historical context, 'development' or otherwise. Clearly, in an article this size, such an outline can only but be brief. The main purpose, though, is to capture the sense of this theory, and the impact of its transformative potential 40 TRANSFORMATION 24 (1994) MAHARAJ ARTICLE Outline of a Systematic Framework for the Social Analysis of Gender The 'historicity of structure': RW Connell vs A Giddens The notion that social structure is historically composed has implications for the possibility of different ways of structuring gender, reflecting the dominance of different social interests. It is in fact only in terms of such a notion that a political agenda for changing the status quo makes any sense. A social theory of gender with subversive intent must therefore be underpinned by such a concept of social structure. It will be argued in this section that the notion of structure espoused by Anthony Giddens, which has enjoyed enormous currency, does not meet the requirements of a social theory of gender whose aim is transformative, whereas that of RW Connell does. According to Connell, 'structure' is more than another term for 'pattern' and refers to the intractability of the social world... It reflects the experience of being up against something, of limits on freedom... The concept of social structure expresses the con- straints that lie in a given form of social organisation... these constraints on social practice operate through a complex interplay of powers and through an array of social institutions. Accordingly, attempts to decode a social structure generally begin by analysing institutions (1987:92). This conception of social structure, as the pattern of constraint on practice inherent in a set of social relations, is not new. Gramsci, Williams, Said and Foucault, for example, share the belief that 'a collective culture sets limits and exerts pressures on thought and action' (Cocks, 1989:40-2) through what Gramsci calls 'hegemonic forms of cultural organisation' (Connell's 'structures') or what 'discourse theory' would call 'discursive forms' or 'discur- sive structures'. The gendered division of labour, for example, counts as a social structure precisely because, operating as it does through institutional mechanisms like the differential skilling and training of women and men, it forecloses a whole range of job options to women: it limits or constrains their economic and other social practices in significant ways. Skilling and training is just one of the institutional mechanisms by which the gendered division of labour is made a powerful structure of social constraint By constraining practice through institutions, it would appear that structure is not immediately present in social life but underlies the surface complexity of interactions and institutions. But this fails to capture the concept of practice as TRANSFORMATION 24 (1994) 41 ARTICLE MAHARAJ the substance of social structure (Connell, 1987:93). The idea of a sharp separa- tion between underlying structure and surface practice must be overcome, a more active connection between structure and practice must be made. Connell's example of how such an active connection can be made refers to a work on kinship which describes a matrifocal kinship structure in a working class London family: the mother is the core figure and mother-daughter relations are such that they pop in and out of each other's houses up to 12 times a day, exchanging services such as care in sickness and negotiating about other family relationships, including the daughter's marriage. Here is an example of a 'struc- ture being shown in its very process of constitution, constantly being made and remade in a very active social practice... The notion of "structure" here is not abstracted from practice...' (1987:93). Giddens could not agree more with Connell about the idea of an active presence of structure in practice and an active constitution of structure by practice. In fact, he long ago (1979) formalised this idea theoretically in his concept of the 'duality of structure', explaining this concept of duality in his more recent work (1986) as the 'double involvement' of institutions and individuals. He says, in Chapter 1: To speak of institutionalised forms of social conduct is to refer to modes of belief and behaviour that occur and recur, or as the terminology of modem social theory would have it, are socially reproduced across long spans of time and space... societies only exist insofar as they are created and recreated in our own actions as human beings... We have to grasp what I would call 'the double involvement' of individuals and institutions: we create society at the same time as we are created by it. Institutions, I have said, are patterns of social activity reproduced across time and space... It is very important indeed to stress this point., (my emphasis). More formally, Giddens' 'duality of structure' refers to 'the essential recursive- ness of social life as constituted in social practices: structure is both medium and outcome of the reproduction of practices. Structure enters simultaneously into the constitution of... social practices and "exists" in the generating moments of this constitution' (Giddens, 1975:5, my emphasis). Clearly, Giddens is making it a logical, definitional requirement of 'structure' that the practice that constitutes it is socially reproduced. Connell sharply differs here. 'By making the link of structure and practice a logical matter, a requirement of social analysis in general, Giddens closes off the possibility that its form might change in history. This is the possibility raised... explicitly by the practical politics of liberation movements; its significance for the analysis of gender is evident' (ConneD, 1987:94). 42 TRANSFORMATION 24 (1994) MAHARAJ ARTICLE The point that Giddens has missed is that, being constituted by everyday practice, structure is vulnerable to major changes in practice. In this sense, 'practice can be turned against what constrains it; so structure can be deliberately the object of practice' (Connell, 1987:95). Sexual minorities, for example, are currently challenging the cultural hegemony of heterosexuality by challenging its structures. Gay marriages, or the raising of children by gay parents, or gay discourses themselves, for example, pose such a challenge. It is this vulnerability of structure to practice that is what makes us agents of history. As structures become modified by human practice so the experiences and options for people these emergent structures generate, change; the cultural 'limits and pressures' that bound people's practices change, what counts as 'common sense' changes. In this sense 'practice cannot escape structure, cannot float free of its circumstances... It is always obliged to reckon with the constraints that are the precipitate of history. For example, Victorian women rejecting marriage were not free to adopt any other sexual life they pleased. Often the only practicable alternative was chastity' (Connell, 1987:95). Giddens' model, then, needs an opening towards history. It needs to recognise that, rather than being a logical requirement of structure that social reproduction occurs, it is simply a possible empirical outcome. But it is an important one, and the cyclical practice which produces it is what is meant by an institution. In this sense 'institutionalisation' is the creation of conditions that make cyclical prac- tice probable. It is in the interests of dominant social groups to create the conditions for cyclical practice (Connell, 1987:141). Giddens' 'theory of structuration', then, is incompatible with a thoroughgoing historicity in social analysis. In his terms, a politics of transformation becomes irrational. Given that much of structuration theory is about finding ways of releasing the 'transforma- tive capacity' of agents, this criticism amounts to an undermining of much of Giddens' work. Then what of the politics of gender transformation? No framework for the social analysis of gender which is not founded on structure as historically composed can claim to understand the world in order to change it. The accolade of 'praxis' applies only to theories that recognise us, people, as the shakers and makers of our history. In this sense, in die terms of many of the major frameworks for the social analysisof gender that emerged in the 1970s, political action for change was also irrational. Of liberal, radical, Marxist and socialist feminisms, liberal feminism 'was perhaps the least enamoured of social structural explanation, tending to emphasize the power of prejudice, irrationality and discrimination. Women's oppression was typically conceived in terms of female socialization into a limited range of roles and assumptions, and the way these social roles were then TRANSFORMATION 24 (1994) 43 ARTICLE MAHARAJ reinforced by a cultural tradition that persisted in viewing women as very different from men' (Barrett and Phillips, 1992:3). Such implicit and explicit individualism was contested by the other feminisms. Marxist feminists argued, for example, that the key problems lay in a system that actively benefited from women's oppression. Their analysis '...stressed exploitation rather than sexist prejudice, the structure rather than the individuals who operated in it, and more specifically the material benefits that capitalism derived from women's position and role... (R)adical feminists stressed not capital but men... as the ones who got the good deal... in the ensuing arguments,... (these) feminists were concerned... with what to pinpoint as the crucial source of women's oppression' (Barrett and Phillips, 1992:3). Socialist feminism, by contrast, recognises not only class but also race, gender, age, religion, etc as social features structuring women's oppression. Unlike Marxist feminism, which seeks to understand women's position in society from a class-based perspective, socialist feminism sees not only class but also race and gender (and other structural features of society) as conditioning women's experience. Moreover, race, class and gender are seen as autonomous, though intertwined, structural features through which power rela- tions are generated to shape the subordinate status of women. In this sense, unlike the other feminisms of the 1970s, socialist feminists were not involved in disagreements about what to pinpoint as 'the crucial source' of women's oppres- sion. Such disagreements revolved around the deeper question of whether the main determinant of gender inequalities was to be found in direct power relations between men and women (the assumption of radical feminists) or somewhere else. As the previous passage indicates, Marxist and liberal feminist theories lacked this focus on power. Liberals focussed rather on custom as the determinant of women's oppression, while Marxists focussed on class relations, the capitalist system or the 'relations of production' (understood in class terms) as underlying women's oppression (Connell, 1987:41-2). So does it make sense in the terms of these three feminisms to talk of a feminist politics of transformation? lake radical feminism, which focusses on the social categories of men and women 'as units rather than on the processes by which these categories are constituted' (Connell, 1987:54). Different brands of radical feminism propose different theories of the power relations between these categories, including for example innate male dominance and aggression. With all the brilliance of radical feminism's insight that direct conflicts of interest and power relations between men and women are the key to women's oppression, a programme of political action to change these relations makes no sense in the terms of this theory, since human agency in structure does not feature here. Again, in the terms of Marxist feminism, a feminist political strategy for change has no TRANSFORMATION 24 (1994) MAHARAJ ARTICLE place: implementing the proletariat's strategy to overthrow the ruling class will automatically bring about the emancipation of women. As for liberal feminism, with its emphasis on female role socialization, women's liberation will flow from a politics of reform of our expected roles. Gayle Rubin, recognising the need for a structural analysis of the power relationships by which women are subordinated to men (she focusses on the institution of kinship as the basis of gender inequality in her attempt to explain gender relations as a social structure) sums up some of the shortcomings outlined in the previous paragraph in her paper 'The traffic in women': If innate male aggression and dominance are at the root of female oppression, then the feminist programme would logically require either the extermination of the offending sex, or else a eugenics project to modify its character. If sexism is a by-product of capitalism's relentless appetite for profit, then sexism would wither away in the advent of a socialist revolution. If the world historical defeat of women occurred at the hands of an armed patriarchal revolt, then it is time for Amazon guerrillas to start training in the Adirondacks (1975:157-8). As already noted, 1970s feminisms were largely concerned with what to pinpoint as the crucial source of women's oppression. 'The diversity of (their) answers helped conceal the consensus in the(ir) questions; yet behind all the sharp disagreements over what was primary or secondary, feminists united in the importance they attached to establishing the fundamentals of social causation... This consensus has since broken up...' (Barrett and Phillips, 1992:4). One reason for this has been the impact on feminist thinking of post-modernist ideas that developed as a reaction to the 'belief in reason and rationality... in the possibility of grand schemes of social reform' (Barrett and Phillips, 1992:4), based on a rationalist model for understanding the world revealed in the thinking of philosophers like Hegel, Descartes, Leibniz and Spinoza. The realisation that the search for 'the cause' of women's oppression was leading up a blind alley was based for one on the realisation of a deeper, more general misconception of social reality as in some sense monolithic. Post-modernist feminisms of the 1990s are characterised by their rejection of the tenets of rationalist world views wherever and however these raise their heads in the context of thinking about women in society. Connell, a post-modern socialist-feminist (for a discussion of the distinction between this paradigm and that of post-modernism see Maharaj, 1993), attempts to account for gender relations in terms of historically specific social structures, dismissing as misleading unanswerable questions about ultimate origins, root causes or final analyses, questions rooted in essentialist assumptions. His attempt TRANSFORMATION 24 (1994) 45 ARTICLE MAHAR/W poses, instead, the answerable question, albeit a very difficult one, of how gender relations are organised as a going concern; a question which, in terms of identifying oppressive structures conceived as historically mutable, offers us the hope at least of fighting our way out of our current gender orders. The 'hoKstk' approach to the structures of women's subordination This approach sees women's specific experiences as generated by intersecting structures which may derive from any social realm, be it the realm of culture, economics, politics,religionor ideology. What the generating structures are and from which realms they derive depends on the specific experience under analysis. Women's experiences in Hispanic societies, for example, derive as much from culture (the 'macho* of Latin American men, for example) as it does from the Catholic religion, the class position of the women being analysed, men- ethnicity, their age, the status of their economies in global terms. Ideologies of gender, race, class, ethnicity, and other relevant ideologies of superiority or systems of social stratification intersect with each other in specific ways in specific contexts to generate specific gendered experiences. With regard to the "Third World' Beneriaand Roldan, the Latin American GAD feminists, have this to say: the assumption (is false) that capitalist penetration into Third World countries has a dynamic of its own, independent of its socioeconomic and historical context. In our view, each develop- ment process has to be understood in conjunction with pre-existing patterns of accumulation and relations of subordination/domina- tion that have conditioned and are in turn conditioned by that process (1987:7, my emphasis). The differing experiences of women in the Third world derive not only from gender-related factors but from a pattern of growth that systematically generates acute class differences and social hierarchies (Beneria and Sen, 1992:3). This holistic view of the structures of women's oppression, long ago espoused in the IDS classic 'Of marriage and the market' (Young et al, 1981), stands in marked contrast to reductionist views which seek to identify oppressive struc- tures in one specific realm. Marxist-feminists, for example, posit structures in the economic realm as the cause of the very different experiences of women both within and across societies. As already indicated, mis form of cultural essen- tialism shares with other early feminisms a belief, now rejected, in a 'cause' of female oppression. Heidi Hattmann in 'The unhappy marriage of Marxism and feminism' recognised that something was amiss in this assumption, and postu- lated an account of women's oppression in terms of the dual, semi-autonomous structures of class and patriarchy. But even here, 'patriarchy' is used ahistorically, TRANSFORMATION 24 (1994) MAHARAJ ART|CLE and the attempt to study the interplay between it and class missed the fundamen- tal point that 'real life does not present itself in a dualistic manner but as an integrated whole, where multiple relations of domination!subordination - based on age, race, ethnicity, nationality, sexual preference - interact dialectically with class and gender relations' (Beneria and Roldan, 1987:10, my emphasis). These multiple relations reflect the 'multiple axes on which power in society inevitably turns... This principle of power's fragmentation leaves us no reason to suppose that all of those axes are reducible to one or logically primary or a cause of others... there is (no) single centre to the life of social power' (Cocks, 1989:50). These statements constitute a direct challenge to the racism and ethnocentric assumptions of white, middle-class feminists, sealing the fate once and for all of the original sex and class debate. From the point of view of setting limits to human experience, structures from any realm may be relevant What is more, within any realm, a structure may be ideological or material in nature. Women's relatively low salaries, for example, a material economic structure, becomes part of the ideology of women and work (which reinforces and is reinforced by other material structures of women's work). In the sense, then, of structuring (women's) experience, ideological and material structures are on an ontological par with each other. 'Cultural factors' is a shorthand for these jointly, and captures the idea of hegemonic power being transmitted through culture, a notion which traditional political theory has failed to capture. It is clear from the foregoing that Connell's attempt to produce a systematic, formal theoretical framework with transformative potential for gender relations has taken openly and freely from the best insights of other theorists. What makes his contribution original is the way he has combined these insights with his own to realise such a theory. 'Pattern of constraint* as 'structural inventory' The idea of structures working together, intersecting with each other within a specific configuration of social relations to constrain and shape experiences into what they are, has been made explicit and formalized in the notion of 'structural inventory'. The 'pattern of constraint on practice inherent in a set of social relations' is made specific through the idea of a set of structures in a specific configuration with each other generating specific experiences by setting limits to, boundaries around, social practice. The shape of the boundary constitutes the pattern of constraint What is bounded is the experience. In the examples above, specific patriarchal relations intersecting with specific pre-existing modes of accumula- tion generate specific boundaries of experience, set specific limits to social TRANSFORMATION 24 (1994) 47 ARTICLE MAHARAJ practice (including thought). As Connell's following definition shows, 'structural inventory' operationalises the abstract 'pattern of constraint', turning it into a formal and explicit tool of social analysis, one which social theorists like Beneria have in any case been using, as he himself indicates: ...structural inventories push towards a(n)... exploration of a given situation, addressing all its levels and dimensions. There is nothing arcane about this. Any historian reviewing the background to a particular event, any politician scrutinising the current state of play or balance of forces, is compiling a structural inventory. Any attempt to grasp the current moment in sexual politics, to define where we have got to, any attempt to characterize the gender relations of another culture, likewise involves a structural inven- tory (Connell, 1987:98). Note that this definition by example as it stands is not really helpful as it does not make at all clear and explicit the importance of specifying the particular configuration of the structural features of a situation in analysing how those features shape that situation. His emphasis is on 'compiling a list' of structural features, a necessary but not sufficient condition of structural analysis. His examples, though, implicitly assume such a configuration. Where the 'situation' under (inventory) analysis is a gendered experience ie Whitehead's 'substratum' requiring explanation (1979), the list of relevant structural features always includes, according to Connell, specific structures of labour, power and cathexis (1987:99) to be discussed in the following section. Where the gendered experience under analysis occurs in a specific institution like the home, workplace, the school, the street, he calls the relevant structural inventory its 'gender regime'. The 'gender order' of a society is a historically constructed pattern of power relations between men and women and definitions of masculinity and femininity. It is, as I understand it, a concept which is meant to capture the gendered dimension of all social experience. Connell, therefore, uses the term 'gender order' to refer, rather abstractly at this point, to the structural inventory generating gendered experience at the level of an entire society (1987:98-9). What he is driving at is the dynamic relationships between the institutions of society in shaping gendered experiences across society. The state, the family, and the institutions of capitalist industry, for example, are structural features acting together in concert to produce the gendered ex- periences of the labour market. Women's part-time employment, or the 'reserve army of labour' are cases in point. But such action in concert is not necessarily deliberate, nor always harmonious, For example, the emotional relationships of the family and the demands of a state at war create unavoidable conflicts. Another 48 TRANSFORMATION 24 (1994) MAHARAJ ARTICLE example of 'institutional abrasion' is the terms of the relationship during times of economic recession among state, family and labour market: during times of increasing unemployment deliberate cuts in welfare benefits occur, increasing the economic disadvantages of women and accelerating the feminisation of poverty (Connell, 1987:134-6). To say that structures of labour, power and cathexis are the major structural features of any gender regime and of any gender order is to specify a framework for the structural analysis of women's experience of oppression in any institution of any society at any time. In this sense, Connell's framework for the social analysis of gender amounts to a meta-theoretical framework: it suggests iden- tifying the culturally specific structures of labour, power and cathexis at play in order to understand and analyse the gender relations in any institution in any socio-historical context. Structures of labour, power and cathexis This section identifies examples of these structures and discusses their interplay in order to illuminate the framework for the social analysis of gender referred to in the previous paragraph. • Labour There are substantially different social structures which condition the relations between men and women in quite different ways. One has to do with the division of labour: the organisation of housework and childcare, the division between unpaid and paid work, the segregation of labour markets and creation of 'men's jobs' and 'women's jobs', discrimination in training and promotion, unequal wages and unequal exchange (Connell, 1987:96). These are specific structures which are useful to regard as elements in a category of related structures constituting the gendered division of production, consumption and exchange in society, ie the gendered division of labour • Power Another has to do with structures of authority, control and coercion as they affect women. Examples are the hierarchies of the state and business which virtually exclude women, institutional and interpersonal violence against women, sexual regulation and surveillance, domestic authority and the contest- ation of such authority (Connell, 1987:96). Again, it is useful to see these as elements of a category of related structures, labelled for convenience the 'gendered division of power'. • Cathexis A third social structure has to do with the recognition that sexuality is socially constructed; that sexed bodies are perceived 'ethnomethodologically', through TRANSFORMATION 24 (1994) ARTICLE MAHARAJ our particular conceptual lenses that derive from our society's specific con- figuration of social structures. This means that the bodily dimension of sexuality does not exist before, or outside the social practices in which relationships between people are formed and carried on (Connell, 1987:111). Since such structures reflect and reinforce dominant interests, Foucault claims that 'every- thing above and beyond the brute raw body that appears to be either some natural expression or extension of it., is in actuality a marking on the body made by power relations in a "political field" (Cocks, 1989:56). The structure of sexuality recognises 'that the body... is an artefact of specific configurations of power...' (Cocks, 1989:56). Sexuality, in other words, 'is enacted or conducted, it is not expressed'(Connell, 1987:111). 'Cathexis', in Connell's terms, refers to the structure that constrains and so shapes people's emotional attachments to each other. It refers both to the hegemonic 'limits' placed on practices that constitute emotionally charged social relationships in which the bodily dimension features and to the social practices which challenge such hegemony. Gay practices in this sense constitute a cathec- tic structure. Heterosexual practices another. These are related to another struc- ture of cathexis, to do with sexual desire. "The social patterning of desire is most obvious as a set of prohibitions... expressed in law... prohibiting sexual relation- ships between certain people... in our culture objects of desire are generally defined by the dichotomy and opposition of feminine and masculine; and sexual practice is mainly organised in couple relationships' (Connell, 1987:112). Feminist arguments on sexuality as outlined in the previous paragraph, by challenging the 'naturalness' of hegemonic sexuality through emphasing its social construction, constitute in their own way yet another type of cathectic practice, that of cathectic 'praxis'. 'Cathexis' refers, then, to the category of structures to do with sexuality. The practices which constitute these structures follow a social logic of their own, and are unaccountable in terms of the structures of the division of labour and power. To say that these structures are different is not to say they are separate: in practice they are inextricably interwoven. Indeed, in any social interaction between people they are present together, as coalesced ideas internalised in the minds of the interactors, ideas which (ethnomethodologically) influence the nature of those interactions, giving them their particular 'vibes'. (It is in these very terms that one can make sense of Foucault's remark that 'the individual is the fine target of hegemonic power... that individuals are the capillaries through which power diffuses itself through culture' (Cocks, 1989:44-45)). Structures lived as practices are embedded in our minds as the ideas, and our hearts as the feelings, which constrain our practices; since power operates through structures (power does not exist apart from the structures through which it operates), and 50 TRANSFORMATION 24 (1994) MAHARAJ ARTICLE structures operate through us, as our practices, then we are indeed the targets and capillaries diffusing discursive power through culture. Distinguishing the struc- tures of labour, power and cathexis analytically is done merely to explain the logic of structural analysis. The three major elements in the structural inventory of gendered experience in any specific institution can be found from among specific structures in each of the three categories of labour, power and cathexis outlined above. The particular experience under analysis, say wife beating, will have a context including at one level the race, class and nationality of the couple involved, and at another the economic contribution each member makes to that family, the sexual and social esteem in which they hold each other, etc. These levels complement each other in the analysis, providing a context that suggests the structures from the three categories that are likely to be at play in this situation. But the guidelines for structural analysis offered by this framework stop here. Precisely how the suggested structures interweave to generate that experience is the next stage of the problem of analysis. It requires for its solution a creativity in thinking for which there are no guidelines. In his chapter 'Gender regimes and the gender order' Connell engages in just such creative analysis of gendered experience in specific Western institutions, rooting his analysis in the framework of structural inventory with labour, power and cathexis (constituted in Western terms) as structural features. Using this framework, we too could begin to engage in creatively analysing gendered experience in institutions, in whatever socio-his- torical context The constitution of social categories The structures of labour, power and cathexis are all implicated in any society's ideas of 'masculinity' and 'femininity'. These structures ideologically construct ' women'and 'men' in terms of certain work-related characteristics, a certain type of sexuality and a certain possession or lack of authoritative decision-making capacity of the sort necessary to control the levers of power in political and other institutions. These structures differ, then, in their effects in the shaping of 'masculinity' and 'femininity'. Structures being historically mutable, so are 'men' and 'women' who are constantly being produced by changing social formations. As the opening statement of this article implies, current inequalities of gender power through unequal access to social resources are embedded in the very structures which define 'men' and 'women'. 'Gender... so conceived gives rise to feminist politics that focus on "long-run" gender interests and goals to do away with male domination... Since, however, gender is constructed simultaneously with a multiplicity of relations - such as class, race and ethnicity - each historical TRANSFORMATION 24 (1994) 51 ARTICLE MAHARAJ analysis may show that women perceive long-run gender interests differently and according to their own life experience' (Beneria and Roldan, 1987:12). This passage raises extremely interesting issues about social interests: how they are differently constituted according to the different cultural constructions of 'men' and 'women'; yet within these historicised categories, the very facts of inequality and oppression provide a motive for collective action, the motive (or 'objective interest') in doing away with male domination; how these differ from interests that are articulated by processes of political mobilisation that define collective goals and strategies relevant to the socio-historical context; who articulates these interests and how. But a discussion of these issues would form the subject of another article. The five elements discussed above constitute the main conceptual underpin- nings of a practice-based theoretical framework for analysing and therefore understanding the social relations of gender. Conclusion The transformative potential of this theory, explicitly recognising and sys- tematically rooting itself in the historicity of social structure, distinguishes it from all previous attempts at theorising the social relations of gender. But what practical guidelines does it offer for realising this potential, for changing the current gender order? Can we draw on any theoretical links suggested by this theory between structural analysis and the politics of our liberation? And if so, what would this mean in practice for our activism? Connell suggests that to identify arenas of struggle which will open up new historical possibilities for the gender ordering of societies, it is necessary first to identify the major structural features of the gender orders of those societies. In his view the major structural features of First World capitalist societies, for example, are institutionalised heterosexuality and the invalidation or repression of homosexuality; heavily masculinised core institutions such as the state; and the gendered separation of domestic life from the money economy and the political world. These patterns together sustain the overall subordination of women by men. Identifying the dynamics which have the potential to transform these features amounts then to identifying the conditions for changing in fun- damental ways the conditions of future social practice (Connell, 1987:159). Given the role of the state in constructing the 'ideal' family form and hence domestic and public patriarchy then any dynamics which will weaken the institutional order of family-plus-state to sustain the legitimacy of men's power must count as progressive. Challenges to the legitimacy of the state posed by women's demands for fair and equal treatment before the law on the basis of equal citizenship, such as demands for equal pay and equal opportunities in 52 TRANSFORMATION 24 (1994) MAHARAJ ARTICLE education, is a source of such family-plus-state institutional weakening: respond- ing to such demands to maintain its legitimacy involves the state in strategies which inevitably weakens domestic patriarchy. Examples of such strategies include state funding of women's education on a scale comparable with men's, the training of police for intervention in domestic violence, the framing of laws which give women greater control over their reproductive capacity, changing the provisions about property, taxation and pensions which treat a married woman in her own right, etc. These all undermine the taken-for-grantedness of male authority in the home on which the reproduction of power inequalities rests. But this should not be taken to mean that in attempting to maintain its legitimacy in the face of challenge, the state deliberately sets out to undermine domestic patriarchy (Connell, 1987:159-60). As Connell is quick to point out, the result is not an automatic disruption of the institutionalised order of power it is an increasing vulnerability to challenge. Whether and how such challenges develop is another matter. The 'crisis of the family* outlined above is just one type of challenge to the gender order of rich capitalist countries, a 'crisis tendency' opening up new historical possibilities. The emergence of alternative patterns of sexuality on a significant scale from hegemonic heterosexuality would amount to another such tendency. According to Connell, there is evidence for this possibility being realised (Connell, 1987:161). Similarly, the definition of a married woman's interests as being essentially those of her husband and children is the hegemonic pattern: the definition of her interests as those of a group of exploited women in a factory, say, is subversive. As already noted, it is a further question whether these possibilities are realised, whether new groupings are formed to take these challenges further. But these examples of crisis tendencies (another important one surrounds the problems of childcare, women's employment and fathering, but there are many, many more) point to a rational link between structural analysis and women's liberation politics, a link which provides the framework for guiding political action: creating or identifying crisis tendencies amounts to identifying arenas of political struggle, where conditions for structural change are emerging; political activism is about expanding then exploiting those conditions. How? By working to construct majority groupings around the crisis tendencies which make radical majorities conceivable in the first place. Majorities matter if the process of social change is to come under conscious human control... (S)tructures cannot be levered into new shapes without mutations of grassroots practice. But majorities do not fell from heaven.They have to be constructed... The lion in the path is the calculus of interests... In a gender order TRANSFORMATION 24 (1994) 53 ARTICLE MAHARAJ where men are advantaged and women are disadvantaged, major structural reform is, on the face of it, against men's interests... Whether the gender order's tendencies towards crisis have gone far enough to provide a basis for majorities committed to major structural reform is perhaps the key strategic question radical politics now faces (Connell, 1987:285-6). To gain such insights as this theory provides is to my mind the right and duty of every political activist engaged in gender studies. These insights provide the rationale for our activism, equipping us both to defend our belief that social change is in principle possible through our efforts, and to make us realise just what we are up against in the analysis and practice involved in trying to bring about such change. REFERENCES Barrett, M and A Phillip* (edi) (1992) - Destabilising Theory (Polity Press). Beneria, L and M RoMan (1987) -The Crossroads cf Class and Gender (University of Chicago Press). Beneria, L and G Sen (1992) - 'Class and gender inequalities and women's role in economic development'. Feminist Studies, 8(7). Cocks, J (1989) - The Oppositional Imagination (Routledge). Cornell, RW (1987) - Gender and Power (Polity Press). Giddens, A (1975) - Central Problems in Social Theory (Macmillan). Giddens, A (1986) - Sociology: a brief but critical introduction (2nd edition). Haranann, H (1981) - 'Hie unhappy marriage of Marxism and feminism: towards a more progressive union', in H Haitmann - The Unhappy Marriage of Marxism and Feminism: a debate on class and patriarchy (Pluto Press). Lukes, S (1974) - Power: a radical view (Macmillan). Maharaj, Z (1993) - 'Feminism and the changing boundaries of knowledge: some theoretical issues' (unpublished paper). Rubin, G (1975) - "The traffic in women' in R Rayne and M Reites (eds) - Toward an Anthropology of Women (Monthly Review Press). WhitebeadjA (1979) - 'Some preliminary thoughts on women's subordination', IDS Bulletin, Young, K, C Wolkowitz and R McCullagh (eds) (1981) - Of Marriage and the Market (CSE 54 TRANSFORMATION 24 (1994)