Accounts of gender that do give a major place to power and conflicts of interest have generally expressed this awareness through a particular form of theory. It has no familiar name, partly because its logic cuts across such familiar divides as the conflict between cultural feminism and socialist feminism. I will call the approach ‘categorical’.

Its major features are, first, a close identification of opposed interests in sexual politics with specific categories of people. Jill Johnston’s definition of men as ‘the natural enemy of women’ is a pungent example. Second, the focus of argument is on the category as a unit, rather than on the processes by which the category is constituted, or on its elements or constituents. Third, the social order as a whole is pictured in terms of a few major categories – usually two – related to each other by power and conflict of interest. If sex role theory tends to dissolve into individualism, categoricalism resolutely stays with the big picture and paints it with a broad brush.

In sex-role theory a common terminology blurs logically distinct concepts; here the same basic idea is expressed in many different terms. Early women’s liberation theorists borrowed from political economy and anthropology. Roxanne Dunbar argued that women were a lower ‘caste’, while Shulamith Firestone wrote of‘sex class’, modelling her argument self-consciously on Marx. Academic feminists borrowed terminology from academic social science. Alice Schlegel and Janet Chafetz wrote of ‘sexual stratification’, while Myra Strober registered ‘the birth of a new science, dimorphics’. The term seems to have been coined as a joke; the fact that it was taken seriously shows how desperate was the need for a theoretical framework. Similarly the discussion of ‘patriarchy’ in radical feminism from the mid-1970s was commonly based on a categorical theory of gender, as in Susan Brownmiller’s well-known argument that rape is ‘a conscious process of intimidation by which all men keep all women in a state of fear’.

Chafetz, who has gone furthest in formalizing this approach as social theory, is particularly clear about its major presupposition. Women and men can be treated as ‘internally undifferentiated general categories’. Analysis takes the categories for granted while it explores the relationship between them. Categorical theories of gender differ from each other mainly in the accounts they give of this relationship.

In one line of thought it is essentially a relation of direct domination. Dunbar and Firestone were among the pioneers of this approach. Newer versions such as Mary Daly’s picture of global patriarchy have focused on men’s violence towards women as its essence. The treatment of pornography and rape in cultural- feminist analyses by authors like Susan Griffin and Andrea Dworkin is closely connected. Pornography is regarded as an expression of the violence in male sexuality and a means of domination over women; rape as an act of patriarchal violence rather than sexual desire.

The academic literature on ‘sexual stratification’ generally takes a more abstract and open-ended approach, assuming no more than that the relationship between the categories is unequal. This has been the framework of a great deal of empirical research (some cited in chapter 1) mapping the unequal material resources of women and men and their unequal life-chances. As theory this is slender. But it has raised questions about the correlates or conditions of different levels of inequality between the sexes. Chafetz for instance conducted a cross-cultural survey to see what general conditions (economic development, environment, religion, etc.) were linked to high or low levels of inequality in ‘the overall status of women and men’.

Here Chafetz verges on an extrinsic approach to gender. Categoricalism in fact provides the model of gender for most of the extrinsic theories discussed above. Analyses of the sexual division of labour, for instance, have usually set up the gender categories as a simple line of demarcation in economic life, adding complexity by mapping the twists and turns of this line in different societies. Only a minority have concerned themselves with what Margaret Power called ‘the making of a woman’s occupation’ (my italics), a question which, by making the process of constructing categories a central issue, leads away from the abstract logic of categoricalism.

Similarly, much of the discussion of ‘relations of production’ has little to say about practice at the point of production. Notionally about social relations, actually it uses such concepts mainly to demarcate categories. Ultimately the person and personal practice can be eliminated from the equation altogether, as happened in French structuralism. Theoretical attention is focused on the social place or category into which the individual is inserted. Along these lines one can arrive at a strongly categorical theory of gender which is not biological-determinist. The ‘places’ in Juliet Mitchell’s and Gayle Rubin’s structuralist models are socially defined and the male-female opposition they stress is socially constructed. Much the same is true of the semiotic analyses of gender that appeared from the late 1970s on.

In other cases the social basis of categoricalism is a simplified normative model of the family. This is true of most of the Marxist – feminist literature on ‘domestic labour’. It is also true of Christine Delphy’s more original analysis of patriarchy as an economic system. Here the categories are constructed by the social institution of marriage, and the core of the relation between them is the husbands’ appropriation of a surplus from the wives’ unpaid labour. Yet another example is Chodorow’s psychology of femininity, consciously based on an attempt to find a social rather than a biological underpinning for the psychoanalytic account of children’s emotional development. Here the sexual division of labour in childcare is the core of the relationship between the categories.

For all the sophistication these authors show in developing social frameworks, the overall map of gender they produce is not too different from one based on a simple biological dichotomy. Categorical thinking about gender is most obvious when the categories can be presumed to be biological and the relationship between them a collective or standardized one. Thus Brownmiller’s ‘rape… all men… all women’, or Dworkin’s ‘pornography: men possessing women’. It is worth noting that biological reductionism does not necessarily produce categoricalism. In some of the literature on transsexualism, for instance, researchers are interested in biological bases for deviance from conventional categories. But theories of biological bases or ‘biogrammars’ usually lead to categoricalism in one of its strongest forms. For most authors assume (wrongly) that reproductive biology divides humans neatly into two distinct categories.

Categoricalism came from a number of sources: structuralism, biologism and the sheer rhetorical appeal of large plain categories for mobilizing political action. It has remained important because it met the need for a clear-cut alternative to liberal feminism and role theory. For some problems, treating gender in terms of ‘internally undifferentiated general categories’ is perfectly adequate as a first approximation. The descriptive literature on sex inequalit­ies in income, education, occupation and health is certainly successful in these terms.

The trouble starts when the first approximation becomes the end of the analysis; when the categories* ‘women’ and ‘men’ are taken as absolutes, in no need of further examination or finer differentiation. For there are problem areas where this will not work at all, and others where the approach rapidly becomes misleading. Analysis couched in terms of a normative standard family is perhaps the commonest example. A good deal of feminist research on welfare has gone towards exploding the assumptions that underpin so much official welfare and economic policy: that everyone (or nearly everyone) lives in a nuclear family, that all women have (or should have) a man supporting them, that having children presupposes having a husband.

Another form of categoricalism focuses on a representative individual. The treatment of ‘male sexuality’ in much of the literature on violence against women is a case in point. So is the argument that explains pollution, indiscriminate exploitation of resources and the threat of nuclear war, by the personal aggressive­ness and ruthlessness of the typical man.

The insight underlying this argument is certainly correct. A power-hungry and emotionally blunted masculinity is part of the social machine that is wrecking the environment, that has for instance devastated half of the beautiful island of Tasmania by wood-chipping and hydroelectric development. Historical research like that of Brian Easlea has traced the themes of masculinity, control and power in the growth of the sciences and technologies that culminated in the nuclear bomb. But to theorize this as the direct outcome of masculinity is to miss the point of the social. machinery that makes a given form of masculinity environmentally destructive. (In other periods of history aggressive masculinity did not result in a radically degraded environment.) It misses the social arrangements that give a particular kind of masculinity a hegemonic position in sexual politics and that marginalize others. And in many arguments it misses the social processes that construct this kind of masculinity in the first place.

The analysis of gender through a representative individual is one case of what Hester Eisenstein calls ‘false universalism’. She poses the issue in these terms:

To some extent, this habit of thought grew inevitably from the need to establish gender as a legitimate intellectual category. But too often it gave rise to analysis that, in spite of its narrow base of white, middle-class experience, purported to speak about and on behalf of all women, black or white, poor or rich.

Theory framed in this way has a strong tendency to lump together different periods of history and different parts of the world. Texts like Mary Daly’s Gyn/Ecology and Kathleen Barry’s Female Sexual Slavery recite patriarchal atrocities from wife-burning in India, genital mutilation in Africa, foot-binding in China, to pornography in the United States. These are offered on the assumption that each is an example of the same underlying structure. In such theories the world-wide dimension of gender becomes a universal common structure of patriarchy.

The conviction that every country and every period shows the same structure has led Western feminists arguing on these lines into classically ethnocentric positions. As critics like Kalpana Ram have pointed out, an account of suttee based on racist Western sources that present Indian women as passively suffering atrocity, and which takes no account of their resistances, mobilization, or purposes, does justice neither to the issue nor to Indian women. It is possible to combine a clear recognition of the exploitation and subordination of women, and the fundamental character of the social changes needed to correct it, with an equally strong recognition of the specific ways in which subordination is embedded in different cultures, the different forms it takes and the different strategies therefore required. Such a case is argued for instance by the Draft Report of the 1979 workshop on feminist ideology and structures held in Bangkok by the United Nations Asian and Pacific Centre for Women and Development. But this leads away from categorical theory.

It is possible to incorporate class, race or nationality into a categorical theory of gender, if these structures are treated in a categorical way too. It involves cross-classification on several variables at once – a very familiar move in quantitative social science. The basic operation is logical multiplication, and the result is a grid on which people can be located. A simple two dimensional cross-classification is central to Tolson’s work on masculinity, for instance:


Men Women

Middle Class


Working Class

A three dimensional cross-classification underlies recent discussions of the triple oppression of non-Anglo working women:


Men Women


Any number of dimensions can be added. They become harder to picture on paper, but can easily be handled by computers.

Yet the more sophisticated the cross-classification becomes, the more firmly is the analysis embedded in a static logic of categories. And this intensifies the difficulty categorical theory has in dealing with divisions that arise within the field of gender itself; that is, divisions that have to do with the constitution of the gender categories.

The most notable problem is the politics of sexual object-choice. Heterosexism and homophobia must be regarded as one of the key patterns in gender relations. It is very difficult to come to grips with heterosexual dominance while using a categorical model of gender. One might set up a new cross-classification, with ‘homo/ hetero’ crossing ‘female/male’. But there is no obvious reason in categorical analysis why this division should matter in the first place. Cross-classification confuses the issue, since that operation logically equates women’s homosexuality with men’s homosexu­ality. There is every reason to think there are important differences between them, not only in the forms of their expression but in the ways they are initially constituted. As the chequered experience of lesbians in the gay liberation movement shows, the solidarity of women and men against oppression from straight society cannot be taken for granted. That is to say, the logical cross-classification does not correspond in any simple way to the constitution of a social interest.

Categorical theory often stresses conflict of interest, but has difficulty with the way interests are constituted and the ways people contest the structures that define interests. Its recognition of social conflict is highly schematized. Over a range of issues, categoricalism underplays the turbulence and contradictoriness within the social process of gender.

The political consequences are important. Of the two main varieties of categorical theory, the academic or stratificationist leads to a politics of access, for instance trying to raise the number of women in political leadership positions, in the law or in business management. It does not generate any particular reasons to question the social arrangements that create those positions. To this extent its practical consequences are little different from the strategy of role reform in liberal feminism.

The women’s liberation movement certainly has called conven­tional power arrangements into question, and so do the more radical categorical theories of gender. Their tendency, however, is to pose this as an all-or-nothing issue. Rather like the ‘big bang’ theory of revolution implicit in Marxist structuralism, categoricalism projects a distant future and a distant past where there is no oppression, but tends to assimilate everything in the grim present to manifestations of male power and female subordination. The effect is to offer women a metaphysical solidarity (‘all women an omnipresent enemy (‘all men and a strong implication that struggle in existing relationships is pointless, since the structure and the categories are universal. Since most women do not have the conditions in their own lives for substantial withdrawal from relationships with men, the practical result is an unresolvable dilemma within feminism. Tension around themes of purity and guilt has been tangible in the movement for a number of years.