The lack of alternative housing is one of the reasons women may stay in, or return to, a violent marriage. Domestic violence is a very large part of all violence. It is difficult to get reliable statistics, but a good many pointers exist. A study of four police stations in different suburbs of Sydney in 1977-8 showed that calls to* domestic assaults exceeded calls to other assaults – in some areas by a factor of three. (The notorious reluctance of police to intervene in domestic violence has some basis in the risk to the police themselves: in the United States about 20 per cent of police deaths in the course of duty occur in such cases:) A widely quoted study in Scotland found that 25 per cent of violent crimes reported to the police were assaults against wives and girlfriends. A national sample survey in the United States in the mid-1970s reported that 28 per cent of couples acknowledged at least one violent incident and estimated that the true figure was closer to 50 or 60 per cent. In a recent phone-in in Victoria 22 per cent of women had experienced domestic violence but did not report it. While violence by women against men is far from negligible, the majority of serious domestic assaults are by men against women.

This is part of a larger pattern of violence against women. Rapes reported to the police in Australia rose from around 600 a year in 1972-3 to over 1,200 in 1981-2. In the United States the figure rose from 47,000 in 1972 to 79,000 in 1983. There is reason to think that these figures are still very much understated and that most rapes are not reported to the police. An alternative approach is to conduct a survey of the general population asking people if they have been victims of various crimes. One such survey, done on a very large scale by the Australian Bureau of Statistics in 1983, estimated 26,700 sexual assaults on adult women during a year, including 8,600 rapes or attempted rapes. (Attacks on teenagers would have to be added to this figure,) This gives a rate of five sexual assaults for every 1,000 adult women, which may not seem very many until it is seen in its context of very widespread intimidation at a lower level, from wolf-whistles and shouted jokes to sexual harassment in workplaces. A number of the mothers we interviewed in Delia Prince’s suburb will not let their daughters go out on the neighbourhood streets at night because of the fear these assaults generate.

Homosexual men also have reason to fear assault in public places – and one of the main groups they have to fear are the police. There are no official statistics on these matters, but police violence on a considerable scale has been carefully documented for New York from I960 through a twenty-year period in which the growing political strength of gays led to a decline in the attacks. The murder of Dr George Duncan in Adelaide in 1972 was almost certainly the result of an entertainment for off-duty police which consisted of bashing homosexual men and throwing them in the River Torrens; Dr Duncan could not swim.

Other groups also are involved in attacks on homosexuals. A 1985 phone-in conducted by Gay Hotline in Sydney (sponsored jointly by the police) over two days received calls describing fifty – three episodes of assault on homosexual men. In almost all these cases the attackers were groups of teenage boys or very young men, a fact that raises disturbing questions about the role of violence and homophobia in the construction of masculinity.

These attacks occur in a context of general hostility towards homosexuals that ranges from discrimination in housing to legal danger. Antagonism to lesbians is less explicit than to gay men – there is a tendency to make lesbians socially invisible. Nevertheless it becomes very clear in such cases as a lesbian mother fighting a custody battle in the courts. Lesbianism itself is liable to be offered as evidence that a woman is unfit to have custody of a child. Gay men face the criminal law because of their sexuality. In most of Australia it is still illegal for adult men to commit ‘the abominable crime of buggery’ or ‘indecent assault upon a male person’. In 1985 the state of Queensland even legislated to prevent hotels from serving beer to ‘sexual deviants or perverts’, meaning gays, along with ‘drug dealers’ and ‘child molesters’. This degree of legislative stigmatization is exceptional; the attitudes underlying it are not.

Statistics on the operation of these laws are patchy but an indication is given by a study of Sydney court cases by the NSW Bureau of Crime Statistics and Research. This showed 100 to 200 prosecutions a year in the 1940s, rising to 400 to 500 prosecutions a year from the mid-1950s to the late 1960s. Great peaks and troughs (117 cases in 1966, 1,204 in 1968, a lull in the early 1970s) reflect changes in the intensity of policing. In the mid – 1970s, when Sydney was well on the way to recognition as one of the ‘gay capitals’ of the world, the rate was rising again. In 1975 NSW courts processed 300 homosexual prosecutions under the Crimes Act and Summary Offences Act (plus others, their number uncertain, under the Child Welfare Act).

Police surveillance in pursuit of these laws merges into police harassment. ‘Decoy’ squads on homosexual beats produce a stream of individual prosecutions; there were said to be two such squads operating in Sydney in the mid-1980s, one in Adelaide. Much more publicity is gained by large-scale police raids on homosexual bars and baths, such as the 1977 raid on the Truxx bar in Montreal (146 arrested), the 1981 raid on bath houses in Toronto (304 arrested), or the 1983 raids on Club 80 in Sydney (over 100 arrested, though few charged). These patterns of intervention in turn merge into the violence already discussed.

While there are clear patterns of domestic assault and sexual assault against women and of violence against gays, there are other patterns of violence which affect men more drastically than women and heterosexual men as well as gays. The 1983 Australian national survey already mentioned estimated 113,000 assaults in a year against women and 278,000 assaults against men. In the same year there were 294 reported homicides in Australia: 41 per cent of the victims were women and 59 per cent were men. In the United States the 1978 rate of assaults per 1,000 people was 12 for women and 22 for men. In 1981 the United States had 23,600 homicides: 21 per cent were of women and 79 per cent were of men. In that year the rate of death by homicide per 100,000 people was 3 for white women, 13 for black women, 10 for white men and 65 for black men.

Those charged with or convicted of violent crime are overwhelm­ingly men. For instance, in NSW in 1983 some 519 people were convicted in the higher criminal courts of homicides, assaults and like crimes; 93 per cent of them were men. A study of homicide in NSW over a 50-year period found that between 80 per cent and 85 per cent of those charged were men. In the United States in 1983, 87 per cent of the people arrested for murder were men. In the same year men were 87 per cent of those arrested for

Violence, Prejudice, the State

The dominance of the military by men was even more complete in the past. The United States armed forces, for instance, were 99 per cent men in I960, 1965 and 1970. The percentage of women began to creep up in the early 1970s, with a formal policy of expanding women’s role in the military starting in 1973. In most armies, nevertheless, women are forbidden to have a combat role. In anti-aircraft units in Britain during World War II women were allowed to do every job, including aiming the guns at the German planes – but not to pull the trigger.

What of those who control the machinery of state, who direct

Table 4 Women at upper levels of the state


Date Body



(1) United States

1984 Congress


(2) Japan



(3) Soviet Union

1985 Supreme Soviet


(4) India

1985 (federal)


(5) Australia

1985 (federal and state)


(6) Colombia




(1) United States

1985 state and federal courts


(2) Britain

1984 high, circuit and county



(3) Soviet Union

1984 People’s courts


(4) India



(5) Australia

1985 all federal courts



(1) United States

early 1980s


(2) Japan



(3) Italy



(4) Australia



the police, the bureaucracies, the soldiers? It is notoriously difficult to locate actual power-holders with any precision, but a reasonable first approximation is possible, by looking at three groups from whom some key power-holders are certainly drawn: judges, generals and members of national parliaments (see table 4). The statistics on parliaments are relatively easy to get, and are shown for countries chosen from the six groups defined earlier. Figures forjudges and generals are scrappy but illuminating. As a change, the figures given in this table are the percentages of women in the groups named.

The Soviet Union looks rather good on these figures. But in the communist systems power resides more in the Party than the parliament – and in the Central Committee of the Communist Party of the Soviet Union, the participation of women in 1981 was 4 per cent.

At the end of an excellent review of the world-wide evidence on women in politics ten years ago, Kathleen Newland argued that a global change in the political position of women was under way. Legal equality (for example the right to vote) was now established and there was extensive participation in grass-roots politics. ‘Yet women have only rarely surfaced among the leaders in politics. They are nearly absent from the positions where policy is hammered out, where decisions are made, where real power resides.’ Ten years later it seems that not a great deal has changed. The levers of state power are still in men’s hands.

The facts just recited are a very small fragment of the evidence now available on the social positions of women and men and on relationships surrounding different forms of sexuality. Though some of the details will be relevant to later steps of the argument, just one conclusion needs to be drawn here. The patterns of gender and sex that appear through this evidence are not just an important feature of human life; they are specifically social. They involve inequalities of income, the working of institutions, the distribution of power, the division of labour and other distinctively social facts. Whether these social facts have non-social causes will be considered in chapter 4.

Two assumptions or working hypotheses underpin the argument that follows. First, that the facts set out in this chapter are connected with each other; that is, we are dealing not with a shapeless heap of data but with a social structure, an organized field of human practice and social relations. This hypothesis is needed to get the analysis under way, though the evidence cited already suggests that it is sound, given the way similar patterns reappear across different fields of social life. A general concern of the book, and the specific topic of Part II, is how this structure can be understood.

There is no convenient name for this subject. Terms like ‘sexual politics’ and ‘patriarchy’ are useful in pointing to parts of it but not to the whole. An American conference in the mid-1970s managed to invent a whole new science of gender and call it ‘dimorphics’; luckily no more was heard of that. Gayle Rubin wrote of‘the sex/gender system’, which is better but begs questions about what exactly a ‘system’ is. Kate Young and her co-authors speak of ‘the social relations of gender’, the most precise term but an awkward phrase. ‘Gender relations’ is a reasonable abbreviation and I will use it as a name for the whole domain.

The second assumpdon is that the two ‘levels’ of fact discussed in this chapter – personal life and collective social arrangements – are linked in a fundamental and constitutive way. It makes no sense to theorize one without also theorizing the other. I introduced this chapter with a case study, not to humanize the statistics, but because of this basic point about method. The large-scale structures of gender relations are constituted by practices such as those Delia Prince and her family are engaged in. At the same time their practice cannot float free; it must respond to, and is constrained by, the circumstances which those structures constitute.

This is an abstract point in theory, but it is also a tangible reality. A final point about Delia will illustrate this. Delia told us she wanted to be a veterinarian. Her parents knew of this and were willing to bear the cost of her training, not a small matter for them. When we returned to that school several years later to report back and follow up, we learnt how things had turned out. Delia had left school at sixteen, as she had predicted; and she had found a job. It was a notable one both in terms of the general sexual division of labour and the specific history of her parents’ marriage. She was not a veterinarian but a veterinary nurse.

There would be value in thinking through a life history in greater depth, as will be argued in Part III; but it is also important to know that conclusions go beyond the one case. Accordingly I will leave the Princes at this point and move on to other sources in later chapters. The approach to personal life that their interviews helped to develop nevertheless remains basic to the treatment of historical, psychoanalytic and other evidence.

It has been difficult, and remains difficult, to get the field of gender relations into focus. Strong emotions are involved. For many people it is threatening even to see these patterns as social. It is comforting to think the patterns are ‘natural’ and that one’s own femininity or masculinity is therefore proof against challenge. Western intellectuals, by and large, have helped this evasion. The grand systems of philosophy and social analysis, from Thomism through Marxism to functionalism and systems theory, have taken the gender arrangements of the day pretty much for granted. Conventional politics marginalizes ‘women’s issues’ equally firmly.

Nevertheless feminism, gay liberation and the research they have stimulated, have now made a case that cannot be ignored. A vast field has been opened up and existing theory and practice must be reconstructed in the light of it. The habit of mind that treats class, or race, or North-South global relationships as if gender did not matter, is obsolete – and dangerous. For the facts of gender do not go away. Aid programs to Third World countries, .by ignoring gender in principle, in fact give resources to men rather than to women. Industrial and nationalist militancy that ignores questions of gender reinforces men’s violence and the patterns of masculinity that lie behind it. The question of human survival, in the face of a global arms race and widespread environmental destruction, requires us to understand a play of social forces in which gender has a major part.



(p. 1). On how easy it is to fall into stereotypes about ‘typical girls’, see Griffin’s book of that name (1985).

A Teenager and her Family

(pp. 1-6). This study is based on interviews from the research described in Connell, Ashenden, Kessler and Dowsett (1982). Sec two papers with a focus on gender and education: Connell et al. (1981) and Kessler et al. (1985).