In the final analysis, categoricalism can recognize power but deletes from its analysis the element of practical politics: choice, doubt, strategy, planning, error and transformation. The ‘battle of the sexes’ comedy in traditional popular culture is a vision of sexual politics with just this kind of deletion. From one year to the next husbands err, wives nag, mothe, rs-in-law carp, girls flirt, boys will be boys and so on for ever amen. Whatever people attempt, nothing ever changes. In more sophisticated categorical­ism practical politics is marginalized by logic rather than by humour. In texts like Gyn/Ecology a particular line in sexual politics is seen as a logical necessity springing from the inner nature of patriarchy, rather than as a choice of strategy.

Choice always implies doubt, and the possibility of going wrong. To give full weight to the practical character of politics, while holding to the recognition of power that is categoricalism’s advance over role theory, is the most general requirement of the next steps in the theory of gender.

This requires a form of social theory that gives some grip on the interweaving of personal life and social structure without collapsing towards voluntarism and pluralism on one side, or categoricalism and biological determinism on the other. In modern writing about gender this has been done best in fiction and autobiography. In books like Doris Lessing’s The Golden Notebook, Anja Meulenbelt’s The Shame is Over, Patrick White’s The Twyborn Affair, Nadine Gordimer’s Burger’s Daughter, there is a strong sense of the constraining power of gender relations (and other structures like class and race), a sense of something that people fetch up against. Yet this ‘something’ is neither abstract nor simple, being real in other people and their actions, with all their complexities, ambiguities and contradictions. And this reality is constantly being worked on and – in ways pleasant and unpleasant – transformed.

In very general terms, we know how to build this kind of understanding into social theory, since parallel problems have arisen in other helds. In class analysis they arose in a debate between ‘structure’ a la Althusser and ‘history’ a, la Thompson. In theoretical sociology they arose in the polemic of symbolic interactionism and ethnomethodology against systems theory and functionalism a la Parsons; and more recently in the debates about Levi-Strauss and structuralism. The outlines of a solution can be found in the work of Sartre, Kosfk, Bourdieu, Giddens and other theorists who have focused on the interconnections of structure and practice. In principle, categoricalism can be resolved by a theory of practice, focusing on what people do by way of constituting the social relations they live in. In principle, voluntarism can be overcome by an attention to the structure of social relations as a condition of all practices.

Though no theory of gender has been formally couched in these terms, some feminist and gay liberation theorists, and a larger number of field researchers, have made a beginning with this kind of analysis. Their work is not recognized as a ‘school’, and their politics are very diverse. So it is worth reciting some examples to indicate the kinds of bases that already exist for a development of theory in this direction.

One of the first was Juliet Mitchell. The second section of her now somewhat neglected book Woman’s Estate described the social position of women in terms of four ‘structures’, in each of which a particular form of oppression is generated. This idea has important implications for the concept of structure that will be discussed in chapter 5. Partly influenced by Mitchell’s work, the American anthropologist Gayle Rubin developed a formal comparative analysis of the ‘system of relationships’ by which women are subordinated to men. Though her discussion of the ‘sex-gender system’ leans toward an abstract structuralism, her 1975 paper ‘The Traffic in Women’ shows more clearly than any other study what a systematic social theory of gender might be like.

In another famous essay, ‘Compulsory Heterosexuality and Lesbian Existence’, Adrienne Rich points to the importance of the social relationships constructed by women among themselves in contradistinction to their connections with men. Her concept of a ‘lesbian continuum’ drifts towards categorical theory but can be approached historically. This is one of the issues taken up by Jill Matthews in her study of‘the historical construction of femininity’ in twentieth-century Australia. Good and Mad Women uses the records of psychiatric incarceration to study the impact of changing ideals of femininity in the lives of particular women. She stresses the historicity of femininity (and by implication of masculinity) as lived experience, not just imposed regulation; and links face – to-face household relationships to the large-scale patterns of demographic, economic and cultural change. On a larger scale again, David Fernbach in The Spiral Path offers a social and relational analysis of what is commonly seen as pre-social desire (or antisocial behaviour). Focusing on homosexual relationships among men, he sets the modern emergence of homosexual identity in the context of the history of gender relations stretching back to the neolithic age. This is speculative in many ways, but certainly a good deal more like a real history than the myth-making about ultimate ‘origins’.

Instead of unanswerable questions about ultimate origins, root causes or final analyses, these studies pose the question of how gender relations are organized as a going concern. They imply that structure is not pre-given but historically composed. That implies the possibility of different ways of structuring gender, reflecting the dominance of different social interests. It also implies different degrees to which the structuring is coherent or consistent, reflecting changing levels of contestation and resistance. Sexual politics is embedded in the structure of gender relations at the most basic level. Structures develop crisis tendencies which materialize in radical sexual politics. These questions are the point of departure for the chapters of Part II, which attempts to develop a practice-based approach to the structure of gender relations.

But this is not the whole scope of the theory of practice, which also points to the historicity of gender at the level of personal life. The idea that forms of sexuality are socially constructed has emerged in the work of radical historians and in discourse analysis and interactionist sociology, not to mention the work of Marcuse. Femininity and masculinity as character structures have to be seen as historically mutable. There is nothing to prevent several forms of sexual character emerging in the same society at the same time. Multiple femininities and masculinities are, I would suggest, a central fact about gender and the way its structures are lived. These questions are taken up in Part III, which develops from several starting-points a practice-based approach to personality.

Before launching out, however, there is one more issue to address. The main reason why it has been difficult to grasp the historicity of gender relations is the persistent assumption that a transhistorical structure is built into gender by the sexual dichot­omy of bodies. This is the assumption that sex role theory finally falls back on, and most kinds of categoricalism too. A social theory is pointless, or at best peripheral, if it is true that the basic determinants are biological. The relationship between the body and social practice is thus a crucial issue for the theory of gender, which needs to be clarified before the structure of social relations can be addressed.


Extrinsic Theories

(pp. 41-7). Delphy (1977) p. 25 issued an unanswerable challenge to ‘class-first’ dogmatism; which rolls on regardless. On Chinese family policy see Stacey (1979) and Croll (1983). The major statements of social reproduction theories of gender are Mitchell (1975); Kuhn and Wolpe (1978); Centre for Contemporary Cultural Studies Women’s Studies Group (1978) and Burton (1985). My criticisms here are partly based on a general critique of reproduction theory, reprinted in Connell (1983), chapter 8. The difficulties in British Marxist – feminist theory are carefully set out by Barrett (1980). For the practical campaigns of socialist feminists in that country see Cootc and Campbell (1982), and in Australia Court (1986) and Stevens (forthcoming). The most thoughtful criticism of ‘dual systems’ concepts I have seen is by I. Young (1981). I have written in more detail on the link between patriarchy and capitalism in Connell (1983), chapters 3, 4 and 5.

Sex Role Theory

(pp. 47-54). Bern (1974) and Steinmann and Fox (1974) illustrate the blurring of sex role and sexual character; Wesley and Wesley (1977) the blurring with sex differences. Carrigan, Connell and Lee (1985) sketch the history of sex role ideas. My argument in this section is based on a general critique of role theory in Connell (1983) chapter 10, and its application to sex roles in Carrigan, Connell and Lee; and the analyses of sex role concepts by Franzway and Lowe (1978) and Edwards (1983). Nye et al. (1976) is the source of the list of family roles quoted. The ‘agencies of socialization’ argument is discussed in more detail in chapter 9 below. Pleck (1981) sees role theory as alternative to psychodynamics. David and Brannon (1976) illustrate the tacit equation of male and female roles. Edwards (1983) provides a comprehensive criticism of ‘deviance’ notions in sex role theory. For the theme of change in sex roles see Lipman-Blumen and Tickamyer (1975) and Pleck (1976). One outgrowth was a curious literature purporting to be a history of sex roles, e. g. Branca (1978). My argument on role theory’s failure to meet real criteria of historicity was formulated in discussions with Tim Carrigan and John Lee.

Categorical Theory

(pp. 54-61). Johnston’s (1973) witty argument for separatism helped establish lesbianism as a political question; the quote is on p. 276. Willis (1984) gives an excellent history of the emergence of categorical views of gender in the women’s liberation movement. The categorical treatment of ‘relations of production’ reaches a peak in Bland ct al. (1978). Money (1970) is an example of biological-detcrminist investigation of deviance. Raymond (1979) is a notable example of theory being pushed towards biological essentialism by a categorical political logic. Categoricalism as rhetoric is well illustrated in the work of Spender (1982). Baldock and Cass (1983) and Campbell (1984) are among many feminist studies undermining the normative model of the family. On masculinity and aggression see Farrell (1974); Dinnerstein (1976) and Kelly (1984); Connell (1985a) expands the argument here. Quotation on false universalism from Eisenstein (1984) p. 132. On the complexity of interests in gay liberation see Gay Left Collective (1980) and Thompson (1985), chapter 3.

Practice-based Theory

(p. 61-4). On images of the ‘battle of the sexes’ see Orwell’s famous 1941 essay on the picture postcards of Donald McGill. The connection between strategy and doubt is clearer in grass-roots radical politics than in radical theory: compare Alinsky (1972). Thompson’s history (1968) is more valuable as a model than his polemic against structuralism (1978), though that has some fine moments. I have attempted a critical evaluation of Sartre’s and Bourdieu’s theories of practice in Connell (1983), chapters 5 and 8. Lefebvre’s (1976) pp. 73—91 criticism of the search for a ‘generative nucleus’ is important in opening up a fully historical perspective on structures of power.