Historically, we see non-linear shifts of control and autonomy in motherhood and mothering. Despite many arguments about the degradation of mothering, it does not seem that women’s control of reproduction and sexuality is today less than it was at the turn of the century. There is a need for more adequate assessments of control and autonomy within feminist discourses. Rather than treating changes in terms of devaluation, my proposal is to focus on contradictory non-linear processes in which gains and losses appear, and where positive gains in control of women’s reproduction, sexualities and mothering can also be considered.

At the beginning of the twentieth century, powerful social pressures dictated that women should expect to have children and that they should only have them within marriage. Marriage and motherhood were supposed to be synonymous, and they were regarded as the best achievements for women of both working and middle classes. An unmarried working-class woman could hardly hope to earn even a subsistence wage, and most middle-class women were likewise forced to marry for a living. Although, in general, the centre of the married woman’s world was her children and husband (Lewis 1984:3), the experience of being a woman differed according to the realities of class and bodily experience (Giles 1995:2). Having time to read books and paying a nurse to look after her child or children was quite different from having to do the laundry while her children tugged at her skirt.

In keeping with the established need for a woman to have a man to keep her and her child or children, legal and social norms were that lone mothers, whether widows, deserted wives or unmarried, ought to keep themselves (Lewis 1984:56). However, there was ambiguity within state welfare policies on how to treat such women: as mothers or as workers? The most common solution was to take a woman with children but without a man into the workhouse instead of providing her with conditions to earn her own living (Lewis 1984:62). The granting of outdoor relief was accompanied by greater social control, with deductions being made for ‘improper’ behaviour (Thane 1978).

Lone mothers were grouped together in these matters, but unmarried mothers were a special focus. Their behaviour was considered immoral and bastardy laws were harsh. Lewis (1984:11) argues that these laws expressed the state’s desire to reaffirm moral values, particularly regarding female sexuality, as well as to curtail social expenditures on this group.

The options for the single pregnant woman were few. In the late nineteenth century, many kept their condition secret or committed infanticide (Horn 1990:156-7). Some single mothers were assumed to be insane (Spensky 1992:108). Before adoption was made legal in 1926, informal adoption was another possibility and some babies were sold through advertisements (Lewis 1984:64).6 Humphries and Gordon (1993:169) found that middle-class mothers advertised in the columns of the Exchange and Mart for the adoption of children of ‘gentle birth’, an aspect that made children more attractive to adoptive parents. Some women, however, dared to undergo conventional disapproval and entered ‘bachelor motherhood’ as a political statement against the control of women’s sexuality (Rover 1970:132-9). They were a small number, but these feminists’ advocacy of ‘free unions’ as an alternative to marriage had some impact on views of women’s subordination in the early twentieth century (Bland 1986).

Women’s sexuality was understood as reproduction, so motherhood and mothering were frequently treated in terms of women’s and children’s health problems. A key problem with marital sex was fear of pregnancy and excessive childbearing.

Fertility rates were high, but falling, especially among the middle class. According to Titmus (1976:95), the average working-class woman, who married during the 1890s in her teens or early twenties, spent fifteen years in pregnancy and birth, with about ten pregnancies. Humphries and Gordon (1993:5) show that only five in ten pregnancies would result in surviving children, with three pregnancies ending in miscarriage, and two babies dying during birth or infancy. Middle-class women had on average about three children in the 1900s, but also numerous miscarriages and pre-natal deaths.

Infant mortality was also high. In the peak of the 1890s only about three-quarters of births survived (Caunce and Honeyman 1993:8; Ross 1993). Government statements often stressed women’s responsibilities for high infant-mortality rates. Better mothering, it was argued, would prevent infant mortality. Full-time motherhood was therefore stressed, but policies to allow this were absent. Such an ideal reflected an old concern of segments of the middle classes. Hall (1980) argues that, in the mid-nineteenth century, working – class parental attitudes were seen as decadent and inadequate in face of the Victorian ideology of domesticity. Hypocritically, women workers and servants were lectured on the damage done by neglecting their families and children, and middle-class commentators were shocked by the standards of housewifery of factory operatives. Yet women needed to work as domestic servants or in factories to make a living, and struggled to take good care of their families and homes.

The twentieth century has seen significant shifts in these fundamental parameters of mothering, sexuality and reproduction. By the inter-war years, maternal and infant mortality had declined. So had family size. Mothers’ lives were no longer dominated by constant pregnancy and childbirth. Birth control was more effectively practised (Humphries and Gordon 1993). Lewis (1984:20) remarks that the birth control methods used in the 1950s were already available in the 1870s. Yet access was difficult. Lewis argues, like Humphries and Gordon, that the greater control of fertility towards the middle of this century is possibly related to negotiation of relationships and more closely shared views of both husband and wife regarding their life choices. Ross (1993) maintains that contraception was almost exclusively a woman’s issue. She argues that in the 1920s abortion (illegal in Britain until 1967) contributed significantly to the declining birth rate.

The ties between sexuality and reproduction were gradually loosened. But this did not imply more flexible codes of sexual morality. Bland (1986:141) shows that following the acceptance of sexual desire in women before the First World War, during the war itself, ‘there were extraordinary restrictions put on the movements of all women and surveillance of their sexual behaviour’. Women’s control of their sexualities only re-emerged as a theme from the late 1960s in the context of more effective and wider availability of contraception and abortion (Walby 1990:169).

In the late 1940s, conservative familialism also took the form of pronatalist concerns. Fabian social philosophy focused on the mother. Policies to make motherhood attractive, however, were ambivalent. Some created nurseries, play centres and laundries to ease the burden on mothers or to facilitate employment. Others gave free access to contraceptive advice and obstetric improvements, and others even planned for part-time jobs to be made available, or for the introduction of family allowances (see Riley 1983:155-75).

Given the centrality of the ideology of constant maternal care for children’s adequate development, even lone mothers were given allowances to stay at home after the Second World War (Lewis 1992a). (Previously only widows received pensions, introduced in 1925.) Benefits were given, and still are, until the child reaches 16 years of age. Any earnings were, and in 1996 still are, deducted pound for pound above a certain small disregard. Sexual ‘fidelity’ was, and still is, required. The assumption is that a sexual relationship with a man would make him responsible for keeping the woman and her child or children, and result in a withdrawal of state support.

It has been argued that these allowances and benefits were granted because illegitimacy had increased considerably during the war as a result of social dislocation and death. However, after a brief war-time peak of illegitimate births that reached 7 per cent of live births in 1943 and 10 per cent in 1945, the illegitimacy rate quickly dropped to 5 per cent in 1950, staying fairly constant throughout the 1950s. In 1855, the illegitimacy ratio had been 7.8 per cent (Blaikie 1994:2). In 1961 the ratio for Britain was still only 5.8 per cent, rising to 9 per cent in 1976 (Lewis 1992a:45). In the mid-1990s it is almost three times higher. In face of this, the introduction of allowances for lone mothers after the Second World War appears to be linked more to the short-term shock of high rates of war-time illegitimacy and the centrality of the defence of the mother in Fabian philosophy. It was also a necessary element of the new Beveridge system of social insurance based on the breadwinner-dependent model of care (Gardiner 1996).

Personal and family life in the 1950s remained conservative. In the sociological literature of the time the importance of the family as a socializing agency was stressed, and well-defined sex roles for men as breadwinners and women as carers were envisaged as the ideal family arrangement (Parsons and Bales 1956; Willmott and Young 1960). On sex matters, there are indications that women and men did not regard good sex as an important issue in marriage, although fidelity was very important (Lewis 1992a:48).

Yet views about sex and marriage were changing. Divorce and the breakup of parents and their children became common. In 1950, 7.1 per cent of marriages ended in divorce compared to 1.6 per cent in 1937 (Humphries and Gordon 1992:219). Lewis (1992a: 50-3) argues that the views of the Royal Commission on Divorce appointed in 1951 reflected post-war anxieties about the family. Housing shortages and the falling age of marriage were seen as the main reasons for increased divorce, but women’s ‘emancipation’ was also a major concern.

Growing rates of divorce and lone motherhood have commonly been associated with greater sexual freedom for women (see Smart, Chapter 2, and Roseneil and Mann, Chapter 11, in this volume). Cohabitation has over time become quite common. While only a tiny proportion of the population cohabited before 1960, in the late 1980s more than 20 per cent of people in their mid-twenties were cohabiting (though cohabitation often precedes marriage, particularly following the birth of a child; Gershuny and Brice 1994).7

Rather than simply reflecting changing sexualities in society, however, contemporary lone motherhood appears to be part of a more general continuing process of redefinition of gender identities: the increasing autonomy of mothers from fathers. In this process, while particular constructions have been maintained, motherhood and mothering have been dramatically redefined over the last 100 years. If motherhood still has ideological and legal ties of a more traditional kind, economic, social and technological changes have transformed mothering.

Declining fertility rates have enabled women to increase control over reproductive sexuality and to expand their control over mothering work. There are conflicting interpretations of some of these developments. For instance, Brookes (1986:166) argues, in her study on reproduction from the mid-nineteenth century to the Second World War, that ‘scientific’ management of contraception and the birth process led to a devaluation of women’s traditional knowledge and skills. Medical expertise led to a decline of women’s control over their own bodies and a consequent loss of autonomy (see Ehrenreich and English 1979). Yet, contradicting her own findings, Brookes also shows that women increasingly did make choices about when and how often they would bear children, a finding that is never related to her general assessment of degradation of mothering conditions.

Another aspect relating to autonomy and control is that motherhood and mothering can increasingly be done without men. Non-traditional ways of becoming mothers have gained prominence, notably artificial insemination, the ‘one night stand’, or adoption. Also larger numbers of women have chosen not to become mothers. The proportion of children (below 16 years old) in the total population of Britain has fallen from one-third in 1901 to one-fifth in the mid-1990s. Fertility rates in Europe have declined from 2.6 births per woman in 1960 to 1.54 in 1990 (OPCS in the Guardian, 11 April 1995).

Children’s demands are also different. Perhaps the roles of mothers and fathers are diminishing in face of the growing importance of peers, teachers, mass media, television, video games and stories. Yet as Davidoff (1995:11) remarks, contemporary social and political controversy over the role and responsibilities of family relationships in the 1990s has focused almost exclusively on mothers (and marginally on fathers) despite the existence of so many other ‘significant relationships’ in children’s lives. What is classified as ‘significant’ is subject to change over time. How does this apply to mothers? How has employment been ‘significant’ in the social constructions of mothers?