Society, like Miriam’s family, tacitly accepts some illegitimate forms of male support. A divorced woman with a ten year old daughter worked as a nurse, and had a male friend, an elderly army officer. His car was often parked outside her house, and people assumed she was his mistress. The neighbours were indulgent: “She has to support her daughter.” A prostitute in another block had trouble with the neighbours over the number of men visiting her house. A compromise was reached: the men would be more discrete when they came and went. “She has to earn a living.” None of these are socially acceptable ways for women to provide for themselves and their children, but they are actions that are compatible with a woman’s role.
Most women are not abused and most men are not abusive. If I agreed with the division of male and female spheres of influence approved by many Pakistanis, I would interpret my material as demonstrating the wickedness or inadequacy of some men. Instead, I am arguing that this is a structural division of power in which women are subordinate to men. If all daughters were indulged, wives cherished and mothers honoured, power would still be differentially distributed.
The problem of bias does not lie in asserting that gender relations in Pakistan are based on inequality of power between men and women. I think that is, indeed, obvious and non-controversial. Conservative, traditional male Pakistanis as well as feminists would agree that women are dependent on male protection. They would disagree in evaluating the results of that dependency. Abuse of power in an unequal relationship can be attributed to the inequality of the relationship or individual abuse of the system. Consider child abuse in the United States: I read horror stories in the newspapers, and I my first reaction is to consider this the result of mad or bad individuals, not structural inequality.
It is important, I think, to examine the “typical”. Statistics can give a useful picture of this. Statistics are a comforting thing, rather like a dress that makes you look ten pounds thinner. But sometimes the dress is inappropriate: too haute coutoure, too bon marché, too worn out. Statistics, like sinners, anecdotes and right-hand margins, must be justified. Statistics forces the social scientist to construct an hypothesis about his/her data and specify the conditions which will support this hypothesis. If data is collected and processessed in a way congruent with the requirements for statistical analysis, your conclusions may certainly be wrong but it does keep you honest. In addition, error is more readily apparent to you and others. The results of a statistical analysis can suggest anomalies and areas for investigation. At the same time, the collection of information for statistical analysis, and the interpretation of the results, requires considerable prior knowledge of the subject of investigation and interpretation (if it is at all interesting) is always based on a non-statistical model of what is going on. (Consider the strong association between milk consumption and the absence of cancer...)
I have quantitative data collected in a census and surveys of the area, information on over fifteen thousand individuals.1 Census information is organized as individual records that can be converted to household records. The socio-economic survey, conducted at the same time, is organized as household records of 462 households. The socio-economic survey is a 20% sample of the census. One of the questions in the census was “Did X [each named individual] work last week?” From normative statements about gender roles, we would assume that men work and women are supported financially by men.
Ninety two percent of males over the age of fourteen worked in paid employment in the week preceding the census. Eight percent of females over the age of fourteen worked in paid employment in the week preceding the census. Age and sex are the most important variables co-relating with paid employment in the data: grown up males work; generally speaking, children and adult females don’t. Normative values are translated into statistically significant results. Men are supposed to support women while women work in the household and, for almost everyone, this is what happens. A further examination of employment looking at age is interesting:
Table 1: People employed last week by sex and age.
Age 0-20 21-40 41-60 61-80 80+ Total
Male
Employed 104 404 157 32 3 700
Not Employed 781 43 16 21 10 871
Per cent (%) Employed 11.7% 90.3% 90.7% 60.3% 3%
Female
Employed 13 32 14 5 0 64
Not Employed 836 358 129 34 3 1360
%Employed 1.5% 8.9% 9.7% 12.8% 0%
Age is self-reported. People in Lahore, particularly older people, are not very concerned with accuracy of chronological age. There is a tendency to report age approximately. There is also a tendency to report age normativally: women should be younger than their husbands, women should be married before the age of 23, etc. Reported age sometimes reflects a regularization of age-related status rather than chronology. I think the table is interesting, however. The percentage of employed men between the ages of twenty-one and sixty is relatively constant. At sixty-one, the percentage drops, reflecting retirement, ill-health, or inability to find work. The same is not true for female employment: the percentage increases until the final age category, 80+. I suspect this increase reflects loss of a male earner in the household and the presence of other adult women in the house to perform domestic duties. Consider the following table:
Table 2: Number of women
in household working 0 1 2 3 4 TOTAL
---
No of Married 1 324 34 4 1 1 364
Couples in 2 61 12 2 0 0 75
household 3 17 3 1 0 0 21
4 2 0 0 0 0 2
---
TOTAL 404 49 7 1 1 462
The labour of men and women in Lahore is complementary. Household work is not optional. Child care is necessary, clothes must be washed and ironed, (without, usually, a washing machine), food is bought daily and prepared. An additional adult woman in the household makes paid employment, for a woman, more feasible. In households with one married couple, 11% of women work outside the home. In households with more than one married couple, 24% of women do so.
There are a number of factors influencing whether a woman has paid
employment or not. I flip through individual household records, and
can see factors which seem to influence women working. In identically
structured households, however, there are no working women. I think
women are employed if there is no male earner, the household is relatively
educated and the woman is between school and marriage, there are other
women in the household to perform household duties, husband and wife
are employed jointly as servants or in running a shop, the woman can
earn while remaining at home, the family income is so low that every
available household member earns if at all possible...and, of course,
there are also women who simply like working and whose male relatives
are agreeable, indifferent, or unable to prevent it
Summary and Conclusions
Supporters of the traditional position of women in Pakistan argue
that masculine and feminine roles are complementary: the women’s
sphere within the house, the man’s without. Both men and women
do important, and necessary, work. Further, segregation of the sexes
has a second consequence. Some paid employment, primarily that involved
in serving female clients, is reserved for women. This employment
allows a small number of women financial independence. The majority
of women, however, are dependent on the protection and support of
related males.
The constraint imposed by their gender role is both external, in the possibilities available to women and men, and internal, in the possibilities that women and men see for themselves. The personal situations of the women that I have described were a result of their dependency on male relatives. This dependency is seen as appropriate. When the results were unsatisfactory, they did not reject the role of dependent but looked instead for a more satisfactory patron. Continuing dependency was always the most realistic prediction of the result of whatever choices these women made.
Both women and men have commonly accepted, popularly perceived rights. A woman’s personal autonomy is circumscribed in exchange for physical security and support. Men control women and are expected, in exchange, to care for these women. When a woman rejects--or is unable to fulfil--these expectations, she is nagged, cast off, beaten or even killed. When a man rejects--or is unable to fulfil--these expectations, his wife, mother or sister suffers. A man may suffer social disapproval, a woman may starve.
The men related to these women that I have described have little enough power. Miriam’s husband cannot feed her, Kauser Bibi’s husband earns too little to educate his children, Mumtaz’s youngest son could not initially care for her and her daughters. Each of the men concerned, however, has the power to negotiate directly with the external world. They may, like their women, starve, but it is socially expected that they work, steal or cheat to avoid doing so. Women, on the other hand, are expected to perform these actions more indirectly, preferably through men. Women are not only formally, but practically, dependent on the good will and sense of duty of men. Economic power is not the only source of power; moral obligation, ties of affection, community standards, helps to constrain and determine action. (I suspect these are, actually, more powerful than simple economic power: we aren’t all dope dealers, and we do have children.) Each of the women I have described did receive quite considerable support from male relatives.
The particular structure of male/female roles in Lahore does not mean that women have no rights, or that they have miserable lives. It does mean that they have limited power to enforce any rights that they theoretically have, and limited resources to change a life that is unsatisfactory.
Culturally, gender roles are described by normative statements which prescribe separation of the sexes and the dependency of women on men. This separation of the sexes is only one of a set of possible (and, I think, useful) analytic binary divisions: two others which are significant in these cases are the separation of kin and non-kin and the separation of public and private spheres. There are other normative statements which say that kin should be supported against non-kin and that mothers will make any sacrifice for their children. These normative statements are not internally consistent. Actors (and observers) can appeal to different rules at different times to explain actions. This gives women a greater flexibility than normative statements about gender roles would suggest possible.
Culture and social organization are not isomorphic. Male, non-kin and the public sphere are culturally organized into one set of relations while female, kin and the private sphere form another. Social organization elaborates institutions which regulate relations between members of the first set--between men, between non-kin groups, in the public sphere--while placing less emphasis on regulating relations within the kin group in the private sphere. Since women’s role is defined as private and kin-based, women have little institutional enforcement of culturally prescribed rights.
The most pessimistic assessment of the position of women in Pakistan that I have heard did not come from a Pakistani feminist or a foreign observer but from an uneducated old hag at Eid: we had gathered in the courtyard in the heat of the summer sun to see the sacrifice of a shoulder high and beautiful silky-haired goat. The goat’s throat was cut, it sank to its knees on the pavement, the blood sprayed the courtyard. We women retired to a cool and dim room. The old woman sat on a cherpai and rolled a bidi:
“Goats and women. Goats and women. One day in their lives, decorated, with fine cloth. Then the sacrifice. Goats and women, they are just the same.”