In common-sense understanding gender is a property of individual people. When biological determinism is abandoned, gender is still usually seen in terms of socially produced individual character. It is a considerable leap to think of gender as being also a property of collectivities, institutions and historical processes. This view is nevertheless required by evidence and experience of the kind just traversed. There are gender phenomena of major importance which simply cannot be grasped as properties of individuals, however much properties of individuals are implicated in them. It may be useful to get clearer the exact sense in which we can speak of gender as collective, and social practices as gender- structured.

Chapter 4 argued that the social relations of gender are not determined by biological difference but deal with it; there is a practical engagement rather than a reduction. It is this engagement that defines gender at the social level, demarcating gender – structured practice from other practice. ‘Gender’ means practice organized in terms of, or in relation to, the reproductive division of people into male and female. It should immediately be dear that this does not demand an overriding social dichotomy. Gender practice might be organized in terms of three, or twenty, social categories. Indeed our society recognizes a fair variety – girls, old men, lesbians, husbands and so on. It should also be clear why a dichotomy of women and men is likely to be an important part of any gender order.

Gender then, is a linking concept. It is about the linking of other fields of social practice to the nodal practices of engendering, childbirth and parenting. The definition just offered leaves wholly open the question of how extensive and how tight those links are, and what their social geometry is. There are times and places where the links are more extensive and compelling, where (to change the metaphor) a greater percentage of the social landscape is covered by gender relations; and times and places where they are less. This is a basic reason why Gayle Rubin’s concept of a ‘sex/gender system’ in all societies cannot be sustained.

Gender in this conception is a process rather than a thing. Our language, especially its general categories, invites us to reify. But it should be clear that the ‘linking concept’ is about the making of the links, the process of organizing social life in a particular way. If we could use the word ‘gender’ as a verb (I gender, you gender, she genders…) it would be better for our understanding. The Marxist-feminist literature of the late 1970s that talked awkwardly about ‘gendered subjectivity’ was groping in this direct­ion, and discussions of ‘gendered language’ are already there.

The ‘process’ here is strictly social, and gender a phenomenon within sociality. It has its own weight and solidity, on a quite different basis from that of biological process, and it is that weight and solidity that sociology attempts to capture in the concept of ‘institution’.

The notion of institution classically signifies custom, routine and repetition. Anthony Giddens in The Constitution of Society follows this by defining institutions as practices with ‘the greatest space – time extension’ within societies, or ‘the more enduring features of social life’. But practice does not have the long duration Giddens attributes to institutions; practice is of the moment. What persists is the organization or structure of practice, its effects on subsequent practice. This can either depart from, or reproduce, the initial situation; that is to say, practice can be divergent or cyclical. As argued in chapter 3, it is not a logical requirement that social reproduction occurs; that is simply a possible empirical outcome. But it is an important one, and the cyclical practice that produces it is what is meant by an ‘institution’. The process of ‘institutionalization’ then is the creation of conditions that make cyclical practice probable.

Speaking of a ‘cycle’ is intended as an alternative to the notion of a continuum of practice, the idea that structure is a matter of sameness in practice. This is implicit in Giddens, explicit in Adrienne Rich’s concept of the ‘lesbian continuum’ as a transhistori – cal reality. Structure may involve a cycle of opposites. Anne Parsons’s psychoanalytic study of the non-Oedipal ‘nuclear com­plex’ in Naples is a particularly clear example, where a particular pattern of masculinity is socially reproduced through the relation with the mother, and femininity through the relation with the father. The opposites may also be at the collective level. The discussion of power earlier in this chapter noted situations where a local power pattern is out of kilter with the more global one, for instance women having authority within a household. This does not undermine the conception of a power structure, merely the idea that a structure must be homogeneous.

Putting these bits together, gender is institutionalized to the extent that the network of links to the reproduction system is formed by cyclical practices. It is stabilized to the extent that the groups constituted in the network have interests in the conditions for cyclical rather than divergent practice.