At several points in this analysis of institutions the importance of context has become obvious, especially the context provided by other institutions. To compile a complete structural inventory it is therefore necessary to go beyond a collation of gender regimes to the relationships between them.

In some cases this relationship is additive or complementary. The patterns surrounding women’s part-time employment are a familiar example. The conventional division of labour in working – class families in Western cities assigns most childcare and house­work to the wife-and-mother; and femininity is constructed in a way that defines the work of caring for other family members as womanly. The labour market constructed by capitalist industry and the state offers some low-paid, low-status part-time jobs; and curiously enough most of the people recruited to these part-time jobs are married women. This pattern of recruitment is justified by employers on the grounds that married women only want part – time work because of their domestic responsibilities and only need low pay because theirs is a ‘second wage5. At home the much heavier domestic work of women is justified by husbands because their wives can only get part-time jobs.

The dovetailing is neat, and it is anything but accidental. The pattern has developed particularly in the 1970s and 1980s, and in the context of the recession represents a practical accommodation between the institutions involved. The dovetailing of structures is produced by a meeting of strategies: the profit-maximizing strateg­ies of employers in a slack labour market and the household work strategies (as Pahl calls them) of the employees.

If this kind of fit were the normal case, we would indeed have the tightly integrated system presupposed by categoricalism. But the gender regimes of interacting institutions are rarely so harmonious. I know of no starker example than ‘The Blood Vote5, a famous poem and poster used in the World War I campaign against conscription in Australia. Two stanzas run:

Why is your face so white, Mother?

Why do you choke for breath?

0 I have dreamt in the night, my son,

That I doomed a man to death.

1 hear his widow cry in the night,

I hear his children weep;

And always within my sight, О God!

The dead man’s blood doth leap.

The conflict dramatized here, between the emotional relationships of the family and the demands of a state at war, is a common theme in pacifist campaigns, including the current campaign against nuclear weapons. A well-known poster reads: ‘What to do in the event of nuclear war: Kiss your children goodbye.5

A more complicated pattern of institutional abrasion surrounds the reversals of state policy about welfare. There has long been a conflict between the goal of redistribution implicit in welfare policy and the goal of stabilization implicit in the machinery of repression and ideological control. Once internal to the state, a classic ‘steering problem’ in the language of Jurgen Habermas, this has been externalized by the recession as a conflict over the terms of the relationship among state, family and labour market. The welfare conservatism of the post-war boom, predicated on full employment, managed class and gender tensions with a gradual extension of welfare measures. But the state was never accepted as wholly benign. A constituency developed for the New Right swing towards welfare cuts, labour market discipline and selective direct repression.

Given the feminization of poverty the welfare cuts have deepened the economic disadvantages of women, while, as we have seen, the military and police are the preserves of men. The changing balance of advantage in gender relations is perhaps reflected in opinion-poll data showing a much heavier support for Reagan among men than women in the United States presidential election of 1980. But there is no simple translation of sexual politics into votes. One of the first governments to use the New Right rhetoric, the Fraser government in Australia, was elected in 1975 with heavier support from women than men; and even in the 1983 election that ousted it, women stayed with the conservative parties more than men did.

A third pattern of connection between institutions sets them in parallel, so to speak, as the domain of a common strategy or movement. For instance the Equal Employment Opportunity campaigns have moved from one organization to another, attempt­ing to bring them into line with the policy, using experience gained in one to move things along in the next. At another level, ‘coming out’ as gay has to be done in a whole series of settings: workplace, family, friendship networks, etc. As Wendy Clark observes, the emotional patterns to be dealt with vary: coming out to your parents is different. But the logic of the process still links the institutions.

What is common to the three patterns is the fact of politics, the social struggle around the terms of relationship between insti­tutions. The state/family connection has been particularly inflamed. It has given rise to political programs that range from Alexandra Kollontai’s desire to use the revolutionary state in the Soviet Union to dismantle the patriarchal family, to the Vatican strategy of using the Italian and Irish states to prop it up. The more generalized programs we see here are a sign of the formation of more generalized interests than the analysis of any one institution could account for. Here we come to a second major step in the constitution of the gender order.

The basic point is that the groupings which are the major actors in sexual politics on the large scale are constructed historically. It may still sound strange to say this of categories like ‘women’ and ‘men’. But it makes sense when the ‘construction’ itself is more closely defined. It means giving a particular content to a social category, establishing particular contrasts with and distances from other social categories, and constituting an interest around which identity and action can be organized. To recall the argument of chapter 4, the biological categories of female and male determine only a limited range of practices (bearing, suckling, etc.) which define the participants as occupants of a range of parallel situations – in Sartre’s technical term, ‘a series’. Groups which are actors in sexual politics are constituted by moves in which seriality is negated and a collective practice undertaken. Such moves are necessarily social processes, action in history.

That is very abstract, but cases in point have already been discussed. The categories ‘male’ and ‘female’ are not categories of social life and sexual politics; the categories ‘men’ and ‘women’ are. The two pairs overlap but the second pair is far richer and more complexly determined than the first. The category ‘men’ for instance has a specific cultural content in a given time and place. Its meaning in social action is not the same in Bali in the 1980s, London in the 1980s and London in the 1680s. One of the key differences between these two dates in London was the expulsion of homosexual practice from the conventional definition of mascu­linity. The creation of a social category of ‘homosexuals’ has been mentioned several times; less noticed is the simultaneous creation of a category of ‘heterosexuals’. Since heterosexuality only has meaning through a gender opposition, this actually requires two categories, heterosexual men and heterosexual women. Their newly purged polarity became the major axis of gender relations in the twentieth century.

The construction of a social category is not quite the same thing as the constitution of a social interest. ‘Twins’, for instance, is a well-recognized category, but it is not easy to define any interest that all or most twins share. An interest is defined by the possibility of advantage or disadvantage in some collective practice. The groupings of sexual politics are constituted as interests by the facts of inequality and oppression. The interests are articulated by processes of mobilization that define collective goals and strategies. Interests are not necessarily articulated. For instance it is clear from David Lane and Felicity O’Dell’s The Soviet Industrial Worker that it is politically impossible for women workers in the Soviet Union to mobilize as a distinct group outside the Party and the unions, both of which are controlled by men. Yet their unequal social and economic situation provides a good motive for collective action.

Interests can be constituted on very different bases, which may cut across each other. Marriage and kinship involve collective practices in which advantage may be gained for one family over another family, and people may well see this as the primary definition of their interests. On the other hand ‘men’ and ‘women’ are collectivities of a more generalized kind which have conflicting interests defined by the inequalities of income, power and so on already documented.

Both of these cross-cutting sets of interests are real, and both can become the basis of a very active politics. Feminism is a mobilization on the latter basis. Mobilization on the former is illustrated by Pierre Bourdieu’s account of Kabyle society in Algeria, with lineages manoeuvering for advantage through mar­riages as much as through land deals. The tendency in social theory has been to pick one basis of interest and regard the rest as secondary. Thus Bourdieu simply subsumes the interests of Kabyle women in those of their menfolks’ lineages. Conversely Christine Delphy ignores the material interest ruling-class wives have in their husband’s careers by assimilating their situation to that of peasant and working-class women. This is not a point to be settled by postulation. The way of defining interest that is ascendant at a given time and place is an empirical question. Indeed a great deal of sexual politics is precisely about trying to make a latent interest salient in practice.

Interests are immediately defined by existing inequalities, but they can also be defined in a longer-term sense. Gender relations are historical, they can be remade in new patterns, and the new patterns will advantage and disadvantage particular groups. So there are interests in historical transformations. Barbara Ehrenreich’s The Hearts of Men is a notable attempt to define the interest heterosexual men had in the remaking of sexuality and the family in postwar North American society. It is not easy to do this, since the public articulation of sexual issues reflects interests only in oblique and censored ways. But it is clear that a pattern of interest can be found.

The conflict of interest on this society-wide scale, the formation and dissolution of general categories and the ordering of relation­ships between institutions, together amount to a macro-politics of gender. This is analytically distinct from, though at all points linked to, the face-to-face issues that are the usual sense of ‘sexual politics’. The ‘gender order’, defined abstractly at the end of chapter 5, can be defined dynamically as the current state of play in this macro-politics.

Its processes include the creation and contestation of hegemony in definitions of sexuality and sexual character (see chapter 8) and the articulation of interests and organization of political forces around them (see chapter 12). At stake are the institutional resources that bear on gender relations, such as the state powers discussed earlier in this chapter; cultural definitions of gender (see chapter 11); and through both of them, the definition of historical possibilities in gender relations. The historical dynamic in this macro-politics is the crux of the social analysis of gender, and the most difficult of all issues to grasp. It will be addressed, in a preliminary way, in the following chapter.