INTRODUCTION

This article explores the changing place of ‘motherhood’ in the lives of girls and young women in Ethiopia from a generational, life course perspective. It employs the concept of ‘motherhood in childhood’¹ to refer to pregnancy, motherhood and mothering among girls under age 18, and as a lens through which to examine changing discourses of child wellbeing in the country.² The analysis centres on multi-generational narratives from young women, their mothers and grandmothers, as part of Young Lives, a fifteen-year study that has traced the life trajectories of a group of girls growing up in poverty.³ These narratives make it possible to examine the social value given to, and roles of marriage and motherhood in, life trajectories towards womanhood, and to highlight changes in these values and practices across three generations (Thomson, 2008). In past generations, marriage and motherhood in childhood was the norm but has increasingly come to be seen as incompatible with expectations for modern childhood. In earlier generations, girls across Ethiopia were often betrothed during the early phase of middle childhood, or even before then, while marriage was generally arranged before puberty (Pankhurst et al., 2016). The current expectation that girls will delay marriage and motherhood until after the teen years represents a crucial revision in the ‘temporal norms’ of girlhood (Freeman, 2010, cited in Thomson and Baraitser, 2018: 68). Influenced by international advocates and organisations operating in Ethiopia, whose concern has been to promote children’s and women’s rights and endorsed by government policy, these changes reflect a growing popular discourse of female empowerment that claims significant expansion of ‘choice’ for girls. Nevertheless, closer inspection of girls’ lived experiences, including their experiences of marriage and motherhood, indicates a more complex, uneven, picture. The concept of motherhood in childhood in this article is used as a heuristic device to think critically about the polyvocality of young mothers who straddle and navigate multiple social categories of girlhood and motherhood, these having been shown to be fluid, rather than fixed, categories in girls’ transitions to adulthood (Johnson-Hanks, 2002). Our evidence suggests that girls contend with multiple, sometimes contradictory, messages regarding the kinds of respectable life paths they should pursue, and their choices continue to be constrained by many of the challenges related to poverty and to living in a patriarchal society that earlier generations of girls and young women faced.

Ethiopia offers an apt context for exploring these intergenerational and life course dynamics owing to its position as one of the world’s fastest growing economies; recent improvements in access to public services; and the popularity of ‘empowerment’ programmes for girls and women. Despite these important developments, there is a dearth of first-hand accounts of women’s changing experiences of marriage and motherhood and their relation to notions of childhood and adulthood. The Government has committed to end child marriage by 2025, and as a signatory to the Convention on the Rights of the Child and the African Charter on the Rights and Welfare of the Child, sets the minimum legal age for marriage at 18 years; unions before this age are termed ‘child marriages’ in line with international policy.⁴ The Government defines child marriage as a ‘harmful traditional practice’ and emphasises the victimhood of affected girls. With the aim of eliminating the practice through attitudinal and behavioural change and supported by international actors, it has prohibited child marriage in law and mounted awareness-raising campaigns.⁵ Overall rates of child marriage have declined and maternal age at first-time birth has risen across Ethiopia, although the rate of change has varied by region (CSA and ICF International, 2012; Mekonnen et al., 2018; Teklu et al., 2013).

However, girls continue to marry and become parents during childhood and indeed are far more likely to do so than are boys. On the assumption that education is both a deterrent and a pathway to social mobility, measures aimed at eliminating marriage and motherhood in childhood are complemented by extensive efforts to promote girls’ education.

‘Motherhood in childhood’ offers a window into the changing roles that girls play in rearranging the constraints under which they marry and begin child bearing (LeVine, 2011), for which we draw on a range of view points, including scholarship addressing critical perspectives on children’s agency in global context (Bordonaro and Payne, 2012; Lancy, 2012), the standpoints of marginalised mothers (hooks, 2007), and feminist concerns around women’s empowerment in developing countries (Cornwall and Edwards, 2010) and across the life course (Kabeer, 2018).

CURRENT STUDY

The article draws on survey and qualitative data generated with young women, their mothers, grandmothers and guardians as part of Young Lives, a comparative, longitudinal study of childhood poverty. ⁶ Young Lives has been documenting the life trajectories of two age cohorts of children and young people (girls and boys) in Ethiopia since 2002. ⁷ In addition to five survey rounds with 3,000 young people and their households across twenty communities, a nested subsample of 60 girls and boys and their families participated in four waves of qualitative research (from the age of 12 to 20) in five localities, plus a follow-up study on young marriage and parenthood in 2015. Repeat biographical interviews and group discussions with multiple generations explored everyday experiences of poverty, risk and childhood. ‘Motherhood’ was not a core aspect of the original research, but intimate relationships, and marital and maternal transitions began to feature in interviews as the cohorts reached their teen years. A wide yet strategic net is cast across the dataset to capture experiences of first-time motherhood, focusing on life course narratives and biographical ‘timeline’ drawings created with girls and their mothers, grandmothers and guardians.⁸ Thematic and biographical analyses are combined within and across cases and for generational pairs of daughters and mothers. Survey data locate the qualitative findings in wider patterns of the cohort. More than one in three of the young women had married by age 22 (nearly half of them by age 18).⁹ One in ten had given birth by age 18, rising to over 1 in 4 by age 22 (Woldehanna et al, 2018). The number of young women who bore children outside of marriage or co-habitation was extremely low (Briones and Porter, 2019).

HOW PATRIARCHY SHAPES WHAT IT MEANS TO GROW UP A GIRL

Despite the decline in marriage and motherhood in childhood, comparison of the experiences of the generation of females born in 1994 with those of previous generations of women makes clear that patriarchal forces have remained very strong across the years, these forces consistently challenging potentially progressive aspects of government policy. This is despite significant changes through the generations in what it means to grow up a girl. In essence, power is distributed unequally across generations and genders and gender roles are relatively prescriptive, particularly in rural areas, so that girls and women are assigned reproductive roles and responsibilities that are lower in status than the productive roles undertaken by boys and men (Heissler and Porter, 2013). Generally, girls and young women lack the authority to act independently, with the expectation that elder kin will influence the age at which they marry and the choice of partner, these unions creating economic and social alliances between families that are formalised through bride price, dowry and other transactions (Boyden et al., 2013; Chuta, 2017).

There are notable spatial variations in relation to marriage practices and forms of marriage payments and their implications for age of marriage and in structural societal differences between elites and wealthy families and commoners and the importance of alliances sometimes promoting child marriage. In Goody’s typology (1976) comparing systems of Eurasian systems of dowry and African systems of bridewealth, Ethiopia being at the crossroads, exhibited both types, with dowry common in the northern part of Tigray and bridewealth in much of the south. There were also intermediary forms such as that commonly found in the central Amhara areas with an ideology of matching endowments from the groom’s and the bride’s kin. Recent Ethiopian literature raises some doubts about Goody’s typology, such as Heran Sereke-Brhan’s (2002, 2005) historical work on nineteenth century elite marriages that revealed the importance of marriages in forming alliances among the nobility and forging links across ethnic groups, and studies of earlier historical records of women’s roles providing evidence of the degree of autonomy and agency exercised by certain categories of women (Belcher, 2015). The prevalence of early-teen and even pre-teen marriage was historically much more common in the north where dowry types of marriage were the norm. Rather than delaying girls’ marriage while parents searched for a suitable match, this led to competition for arranging marriages early, with the nobility and more wealthy households seeking to arrange early betrothals. Among the Oromo, marriage by abduction or elopement was often considered a legitimate form of marriage particularly for poorer men unable to pay the bridewealth but requiring subsequent lesser compensatory reconciliatory payments (Cerulli, 1956; Holcomb, 1973). The ethnographic literature is not very explicit on early marriage, in part since age is often not considered a diacritical factor, though there are frequent references to girls marrying earlier than boys (see Boyden et al., 2013). There is scant ethnographic evidence on the age of marriage in southern Ethiopia, with more concern about prescribed and proscribed partners.

Older women in this study said that they had lacked agency in marital decision-making and several highlighted their suffering in these situations, and yet they often did so while also endorsing the ideal of marriage and motherhood in childhood. Haftey’s grandmother (Tigray region), who was married at age 9, was keen to arrange her granddaughter’s marriage when the girl was 16, despite recalling—with a mixture of pain and humour—the suffering her own generation experienced when married off young:

“They just covered our faces, made us sit us on the mule and sent us to the men. Then, we cried when we went there [laughing]… we tried to escape, we’d hide ourselves and they’d beat us… You are called a bride and you just sit… you know that you have to sleep with a man, then you cry when you have pain [laughing].”

The pressure on women to conform is reinforced by the understanding, shared by women of all ages and across all research sites, that attaining social adulthood is contingent upon marriage and motherhood, this conferring enormous symbolic value on both transitions. Alternative sources of social affirmation were especially limited for older generations, as one mother emphasised, “during [my] time, a girl could do nothing better than marry. There was no other role for us.” But even younger women reserved the status of ‘woman’ for those who mere married; one group in Addis Ababa did not regard unmarried females, no matter their age, as women. Others confirmed this, “in Ethiopian culture, it is difficult for a girl to live unmarried,” so “a female has to get married and have children”. Marriage also held significant instrumental value and was often cited as a solution to poverty, livelihood insecurity and social marginalisation among women, especially when the husbands were educated and/or older than their partners (cf Archambault, 2011). ¹⁰ For many women in the older generations, motherhood too was instrumentalised, insofar as having many children was taken as an insurance against infant mortality: “When you have a baby, you don’t know what’s going to happen to them. Some of them will be sick, but if [my daughter] gives birth to six children that’s going to be the safest as it can compensate for the loss” (Older mother).

Given the centrality of marriage and motherhood to womanhood, parents are sometimes complicit in entrapping girls into marrying against their will. Chaltu’s older sister died one year into her marriage; under the local custom of membeto (substitution for a deceased sister) the husband requested that his in-laws provide another wife. He was given another older sister, then aged 16. But she escaped before the wedding ceremony and Chaltu was offered as a substitute. When Chaltu objected, she was told that marriage is a ‘chance’ rather than a choice and that she should accept her fate. Eventually, her parents orchestrated her abduction by the man’s family and she was married, aged 14, to a man of 40. Although Chaltu’s case could be regarded as somewhat exceptional for the younger generation born in 1994, not least because of her age, there was no clear consensus among the younger women as to what age is ‘too young’ or ‘too old’ to marry. That said, unmarried girls occupy evidently an ambiguous status and if they remain single for too long risk being ridiculed as arjitalech (‘she has become old’) or qumo qer (unmarriageable, literally ‘remaining standing’). “If a girl is late to get married, everybody says shocking words to her as if she is born to be dependent on her parents. Other people ask her why she doesn’t get married and why she doesn’t go out of the house.” Unmarried girls may be ascribed behavioural attributes that are disapproved, as explained by young mothers in a rural village: “The community has a positive attitude about married girls because they are perceived as organised and integrated persons. If a girl is not married she is perceived as disorganised… going about, here and there, and is disrespected”. They are thought to be liable to engaging in pre-marital sex, risking pregnancies that are neither planned nor socially endorsed, thereby tarnishing personal and familial reputations. Delaying marriage for too long is also thought to limit women’s marriage options to divorcees, widowers, or polygamous arrangements.

As in the past, marriage and motherhood remain sources of respect and social recognition for the younger generation of women, and in some localities were said to also facilitate access to land and married women’s groups. Accordingly, gendered expectations of this generation continued to centre on mastery of appropriate reproductive skills, as in previous generations. Valued moral attributes include responsibility, respect, obedience and modesty, and hard work and practical skills in the domestic arena are also prized (Boyden, 2012; Poluha, 2004). Teaching girls to learn domestic skills is considered one of the primary responsibilities of older women, as is the imperative to raise the young to be morally and socially acceptable to others (Ruddick, 2007/1983). A woman from Addis Ababa explained; “all girls at this time and in the past have domestic responsibility” and “if the girl fails to learn these things it reflects failure of the mother so they keep their daughters busy with housework”. According to a group of urban older mothers, the window of opportunity to inculcate these skills must not be missed: “Unless a girl is shaped and let to know about all the household activities before she sees her menstruation, it becomes difficult to shape her personality after her period.”

Mothering is also a valued form of labour (hooks 2007/1984), described by some as ‘motherwork’ (Collins 2007/1991, 2007/1993) to place emphasis on those aspects of mothering that have to do with ensuring the physical and social survival of the young in contexts of economic and social inequality. In Ethiopia girls typically begin preparing for ‘motherwork’ at around age 5, and by age 12 many regularly undertake domestic chores unsupervised, some also assuming responsibility for tending to ill parents and grandparents, occasionally taking time away from schooling. The skills young women are expected to acquire are very precise, and often focus on cooking, as one mother indicated with regard to her daughter:

[P]reparing injera (flatbread), tlews (baking), brews good suwa (local drink) for her husband, making good gogo (local bread). She has to be a woman who is ready with boiled and cooked food when the husband comes home after farming and making good business in the market.

For some of the girls in the younger generation the pressure to fulfil domestic responsibilities in the natal household and the desire for independence were key motivations for marriage: “We served our family so far, so we need to marry early and build up our own family.’ Haymanot (who married at age 15) compared herself to her unmarried friends: “They are still dependent on their parents, but I live independently… [L]iving an independent life is good, dependency is not good.” Similarly, Tirhas felt that she had gained social status through marriage: “I could have a comfortable life with my parents but people respect you when you are married. It is not common for a matured girl to live with her parents.” However, for some, marriage simply marked a shift from a childhood relationship constrained by parental authority to a marital one restricted by spousal control. That young women bear a significant domestic burden following marriage is supported by the survey data, which show that young married women, aged 19, were spending up to 8 hours a day on unpaid work and care (Crivello and Espinoza, 2018: 148). Such patterns reflected the naturalisation of rigid gender norms that assigned men the role of ‘breadwinners’ and women, carers. Young mothers in one community argued: “the man has the decision-making power and the wife has to be polite and obedient.” The man “administers the family” and “the wife is administered by her husband.” Their narratives also contrasted the substantial weight of maternal responsibility with men’s abnegation of their obligations, some claiming that men were ‘useless’ at childcare. When asked if their husbands helped, one group replied: “Nothing! [laughter]… How can a man help?”

EXPECTATIONS AROUND FIRST-TIME MOTHERHOOD

Given this context, the issue was not so much whether girls would marry or bear children, but when. Once married, young women faced considerable pressure to conceive and failure to do so invited the castigation of husbands (“we didn’t marry a wife to eat together”) and in-laws (“she is infertile and she came just to feed herself”), with motherhood also cited as necessary in preventing divorce. Many respondents regarded motherhood as a vehicle for personal fulfilment, as giving life meaning and ensuring contentment. One mother envisioned her daughter’s ‘bright future’ through the image of “her mahzel [traditional sling for carrying infants] and lig’e [the colostrum]” and commented “I just want to see her have children…” Similarly, a ‘comfortable life’ for a young woman was said to entail “having two to three cows, seeing her hair oiled with butter, drinking milk, carrying her child, remembering her mother...” Young mothers echoed: “Having a child is above everything. We get happiness for giving love and affection to our children”; “children are jewels to women”; “it is impossible to live without children”; and “I would like to have four or five children to make them fill the house… my mother has eight children and when we come together we fill the house to create a joyful family.”

Motherhood was also thought to be central to a girl’s sense of personhood and self-worth, “a mother has to live for her baby once she gives birth”. Some women perceived their physical presence to be expressed through that of their babies. One had considered avoiding breastfeeding in a bid to retain her youthful appearance, but changed her mind on the grounds that “after giving birth, my beauty is my baby.” The conflation of a woman’s identity with her maternal role was institutionalised in some communities by naming new parents after their first born child. Many young mothers experienced this re-naming as a social affirmation rather than a loss of identity. Similarly, successful motherwork was conceived as a social practice rather than an innate personal attribute. For Chaltu, married at age 14, the measure of a good mother was not the quality of the dyadic mother-child relationship, but rather: “To become a good mother, a woman has to create good relationships with the community members. She has to perform different activities for her household and for the community members… Now I am a matured woman.” Some argued that first-time motherhood was crucial for a woman’s social maturation. Tirhas, who gave birth at age 15, reasoned, “[I] became mature especially after I gave birth. At the beginning, I was acting like a child. I got angry about little things and he [baby’s father] used to just tolerate me.”

Despite these abiding norms around motherhood, ideas about first-time motherhood are certainly changing and a sizeable proportion of women in the older generation thought that marriage and motherhood were incompatible with childhood. Only 14 per cent of surveyed parents/guardians expected their daughters (then aged 12) to marry before age 20, and most expected they would not have their first child until around age 21 to 30 (71 per cent). That said, the picture is very complex. For example, even though a group of young rural mothers identified 20 as the normative age for first time motherhood in their community, they had all given birth by age 16. In fact, the most recent survey (n= 366) indicated that by age 20, 1 out of 3 young women in rural areas had given birth before age 18 (n=59), while none of those in urban areas had given birth before age 18 (n=12). Only 3 girls in the sample (all from rural areas) had given birth below age 15.

In qualitative narratives, some women expressed doubt about the suitability of motherhood for very young girls. For example, reflecting on her own experience, a woman from Addis Ababa indicated:

“As I was too young when I gave birth to my first child, I had little responsibility… My mother was responsible for caring for the child… I went out to play … Regarding my second child (I gave birth six months ago), I realized the beauty of having a child… Now my mother is not with me. I provide full support to the child. There is quite a difference in the experience between my first and second time of giving birth.”

Mulu’s mother also had a very mixed experience of marriage and seemed in no hurry to marry Mulu off. She was married off at age 10 to a 20 year-old man: “Our marriage was through our parents… He saw me and asked his parents if he could marry me… my parents agreed.” She divorced him after three years, remarried at age 14, divorcing again at age 20 and remarrying at age 21. She had six children altogether. In contrast, at age 22 and intent on completing her education before starting a family, Mulu remained unmarried and was enrolled at university. For some of the older women, raising the age of first-time motherhood was important for breaking intergenerational cycles of suffering. Two mothers noted, “I have suffered a lot and I don’t want her and her own children to suffer like me” and “I had nine children but I suffered a lot. So I don’t want her to suffer like me.”

THE PART PLAYED BY SCHOOLING IN REVISING GENDER SCRIPTS

Even though motherhood and marriage continue to be core to female trajectories through the early life-cycle, the teenage years of girls born in 1994 look quite different from those of previous generations. The expansion of formal schooling—especially at secondary level—signalled a major intergenerational shift in the Young Lives sample (c.f. LeVine 2011). Many of the older women had grown up in communities without schools. Others were excluded because girls’ education was not valued: “if a woman tried to go to school she would be discouraged by being told that women’s education means nothing” (Older women’s group discussion). One woman explained that her parents were illiterate and saw no purpose in educating girls, so tore up her school notebooks. Another said: “My parents were not aware of the benefits of education. They knew only marriage and house work.”

By contrast, all three generations of women held schooling to be fundamental to the lives of girls born in 1994; this generation was far more likely to go to school and to remain there through their teens. Whereas 21 per cent of the parental generation never went to school, this applied to only 4 per cent of the younger generation (males and females), and a quarter of whom reached post-secondary level, compared to just 6 per cent of their parents. As Table 1 shows, differences in school completion between girls (n=433) and their mothers/guardians (n=431) were especially stark; the widest gaps being between urban young women and rural older women.

Regarded as the most important route out of poverty and towards better lives, education was cited as central to a ‘good childhood’ for the younger generation; 99 per cent of those surveyed at the age of twelve (n=980) saw education as being ‘essential for my future life’. When surveyed three years later, 93 per cent (girls and boys) agreed with the idea that if they studied hard at school, they would be rewarded with a better job in the future (Tafere, 2014: 9). Girls’ parents hoped that education would bring their daughters work in the future: “[B]eginning from her early childhood, I wanted my daughter to finish a university education and get a good job so that she pays back what I have invested in her. … She is the best of all my children” (Mother of 17 year-old). Fatuma’s mother, in Addis Ababa, explained her hopes for her married daughter, “I wish for her to become an educated and self-sufficient women. I do not want her to stay at home but I want her to be in a modern family. She has to work as well as her husband.”

So, while gender scripts for the younger generation continued to be moulded by powerful patriarchal norms, increased access to education appears to have heralded important changes in ideals around girlhood and female adulthood. First, education was recognised as broadening girls’ horizons. Women in the older generations frequently characterised the younger generation as being more ‘aware’, ‘smarter’ and ‘wiser’ than they are. Haymanot’s mother described her daughter thus:

Haymanot is very wise and can work the work of a man and that of a woman … her knowledge is great. The mind of the children at this time is smarter than ours during our childhood … we were like angotat (unripe fruit) and that is keeping us back. Now, the government is educating everyone… the children are very smart even while they are inside the womb… [T]hey are more aware of things than we are…

Second, schooling was perceived by some in the older generation as having given girls new knowledge and skills, and also greater confidence—attributes that can be important for building a sense of selfhood and self-respect and for the realisation of personal agency. Yordi’s mother had confidence in her 15 year-old daughter succeeding in school “because she has self-esteem [ras-metemamen].” It was argued that in the past “all life decisions were in the hands of parents”, whereas young women ‘today’ decide for themselves, acting ‘in their own interest’. Similarly, women across all generations agreed that “now is a better life for girls because they have freedom”, with the younger generation perceiving girlhood in the past as a time when there were “no schools, girls were unaware of their rights, they were passive.” Outcomes of increased self-assurance among younger women were said to include more open communication between mothers and daughters on topics considered sensitive, such as menstruation, more equitable spousal relations and decision-making, and greater gender equality.

Third, education seems to have been an important factor in changing views on the appropriate age for and social meaning of marriage and motherhood. In a 2010 study, Erulkar et al. found that 72 per cent of girls with no education had been married, in contrast with 22 per cent of girls with nine or more years of education. Schools in Ethiopia have been used as a significant channel for the communication of government policies aimed at ‘modernising’ society and eliminating ‘harmful traditional practices’. One woman referred to this directly: “because of education, people have understood that [child marriage] is a harmful practice”. The younger generation of girls were taught about children’s rights at school, and in girls’ clubs and empowerment programmes, including their right not to marry and/or become mothers at a young age (Marcus and Brodbeck, 2015).¹¹ For most of the unmarried girls born in 1994 the ideal gender script for female childhood involved completing their education, taking up employment and saving money prior to marrying, this ideal script reflecting the intrinsic value placed on education as a good in itself, rather than the actual likelihood that education will be a route to formal employment and prosperity.

Fourth, the knowledge and self-assurance gained by girls through education nonetheless seems to have given some the confidence to resist being married off by their families during their teen years. Several cases were cited in which girls sought the aid of their teachers in convincing parents to abandon plans to arrange their marriage and allow them to remain at school. One mother admitted: “Now, if I marry off my daughter without her interest [consent], she will refuse and oblige me to pay back any bride wealth I take.” In another case, having raised her orphaned granddaughter from a young age, Haftey’s grandmother tried to arrange the girl’s marriage when she was aged 16. The grandmother reasoned, “I wanted her to get married while I am still alive”. But she was forced to abandon her matchmaking effort because Haftey would have none of it. The grandmother claimed: “When I tell my … granddaughter to marry, she gets angry and threatens to report me to the authorities” arguing (while laughing) that Haftey “wanted to jail me”.

Marriage has now become surrounded by a discourse of ‘choice’ wherein girls are understood to formulate and act upon their own interests. One older mother commented: “During our age, the parents simply gave their daughters to husbands in their early childhood but now everything depends on the choices of the girls”. Young women’s accounts corroborated this view: “If she wants to marry she can marry at the age when she considers that she is ready for marriage. Marriage is based on the interest of the girl. The parents have little involvement in the marriage arrangement… [S]he decides on her fate.” Many of the younger generation questioned the relevance of arranged marriages, perceiving them to be ‘old fashioned’.

The generational differences described so far point to a revised gender script whereby the linear life trajectory marked by the feminised milestones of marriage and motherhood has been interrupted in the younger generation by aspirations to complete school, and to earn money, prior to becoming an adult. This view was clearly articulated by a group of young mothers in Addis Ababa, “Age should not be the major factor. Female children should start to work some jobs before they decide to marry.” They saw very young mothers as being susceptible to poor self-image and ill-being: “[Y]ou will be overloaded with thinking. You become responsible…dirty. You feel inferior”. Similarly, speaking about her fifteen year-old daughter, Beletch’s mother said: “If marriage is concerned with an age, she is old enough to marry even now. My only concern for her now and in the future is that she will learn and finish her school… hurrying to marriage is not important.” On the other hand, Netsa’s mother reckoned that her seventeen year-old daughter should marry, “When she is matured… [L]ater is better … [I] don’t want her to have a child before marriage and before she has her own job.” Another mother maintained, “I cannot say this age or that age, we parents cannot force our children anymore. The time she wants to marry is all about her decision.”

Quite a few of those in the younger generation aspired to delay their first pregnancy, space births and have smaller families. Continuing education, saving money and securing housing were all considerations: “I believe that in the future I will no longer be poor. I will use family planning and have fewer children [than my parents]. My parents have five children, which may be a reason why we are poor. I want to continue my education to university” (unmarried girl). A number of the young women were able to challenge social pressure around fertility by using modern contraception, an option that had become more feasible following government efforts to promote family planning through community health extension workers. Some girls in Addis Ababa were able to access contraception prior to marrying, although access remains constrained for unmarried girls and women. A few married women resorted to using birth control clandestinely, without informing their husbands, with the risk of discovery and marital discord:

“I didn’t want to get pregnant. I was taking pills during the first year because I wanted to save money before having a child. On the second year of the marriage, he figured out that I was taking the pill when I went to the health centre. He then forbade me from going to the health centre. He really wanted me to get pregnant. But I was very mad because I didn’t want a child then.” (Almaz, married age 17)

In rural areas, fertility decisions continue to be negotiated in a wider family context, including in relationships where young women had had a say in marriage decisions (Chuta, 2017).

AMBIGUOUS AGENCY

Education and associated notions of children’s rights, individual agency and gender equality may be central features of modern childhood in Ethiopia, but there are challenges to achieving these goals in contexts where girls’ lives and trajectories continue to be moulded by poverty and discriminatory gender norms. Ethiopian anthropologist Tatek Abebe has described young people’s entrapment between the ‘disparate worlds’ of ‘eroded tradition’ and an ‘unfulfilled modern life’, this generating uncertainty in life choices (Abebe 2008, 23). In the Young Lives sample, the interaction between old and new values, and between norms, aspirations and material realities, meant that the younger generation of women confronted competing - and sometimes contradictory - expectations concerning ‘respectable’ female life paths towards adulthood, with direct impact on their everyday lives (c.f. Camfield and Tafere, 2011). Although seemingly they had greater choice in their pathways to adulthood than did previous generations, their capacity to make independent decisions and to follow them through in the transition from childhood to adulthood and beyond remained limited (Cornwall and Edwards, 2010: 2). For many, the challenges to exercising personal choice grew following the birth of their first child. This points to the continued efforts on the part of older generations to control the reproductive capacity of young women.

The term ‘ambiguous agency’ has been used to refer to:

examples of agency amongst children and youth which is in stark contrast to established and normative conceptions about childhood and moral and social ideals about the kind of behaviour young people should demonstrate, the activities they should be engaged in, and the spaces and places deemed appropriate for them to inhabit. (Bordonaro and Payne, 2012: 366; see also Durham, 2000: 116)

In this vein, ambiguous agency is used here to characterise the ambiguities, dilemmas, contradictions and risks confronted by females in the younger generation in their efforts to shape their life transitions, and evidenced in their marital decisions. Despite the rhetoric around changed norms, values and practices, in the Round 5 survey only 60 per cent of married women in this generation (then aged 22) reported having been involved in deciding who they married and the majority felt they had married too young (Briones and Porter, 2019). Similarly, increased expectations of female independence associated with school education had not reduced the pressure to marry, as Haftey’s grandmother highlighted; “The idea is that it is better to prioritise education but not to live an unmarried life”. Indeed, girls who rated themselves poorly as students, had fallen behind with their studies, or left school early saw marriage as the obvious alternative option. Tirhas received her first marriage proposal at age 12, but her parents turned the offer down because she was still at school. However, she failed her eighth grade exam so had to drop out and feeling she could no longer decline proposals, married at around age 13; “I did not have the guts to confront [my family] because I failed my education and my fate was getting married.” By contrast, in some contexts, the perception that education had failed to deliver on its promise to transform girls’ lives elevated marriage as their next best chance in life; “Education is dead and girls do not want to die with education. They want to establish a family now through marriage. Many girls stay idle after completing grade 10… Girls should not die together with education.”

Despite the continuing control by older generations over marital decisions, one of the most dramatic changes reported across the generations has been the rise in ‘voluntary’, or couple-initiated unions - also termed ‘elopement’ or ‘voluntary abduction’ - leading to (urban) cohabitation or informal unions. Generally organised by the couple without parental involvement, some respondents argued that a preference for informal unions reflects the diminished authority of older generations over younger generations and an increase in female agency. Certainly, older women bemoaned the perceived decline in respect for adult authority and knowledge and the weakening of intergenerational relations associated with the practice. Sometimes informal unions are seen as a solution to unplanned pregnancy,¹² and they often have special appeal for poorer young men who struggle to accumulate the bridewealth required for a formal marriage. And there have been some perverse consequences. For a start, enjoying greater autonomy in life decisions does not seem to have led to a rise in the age of marriage. In communities where young people are driving these decisions, the notion of underage marriage as a ‘harmful traditional practice’ does not seem to resonate. Strikingly, in some areas young people were said to be choosing to marry even younger, “[T]he time is fast …The youth are in a hurry to get married and to get divorced.” In fact, given that in some young women’s estimation gaining autonomy from one’s natal family was an important goal, ‘overage’ marriage had come to be viewed as a greater social problem than ‘child’ marriage. “It is a shame to become an old woman without marriage.”

Ayu’s story reveals the increasing limits and possibilities around young women’s decision-making as they advance through the life cycle, illustrating ambiguous agency. Although when she was age 12 Ayu hoped to have an arranged marriage around age 22, she had agreed to an informal marriage at age 15, despite such unions having lower status in her community. She had left school two years earlier, having only reached the second grade, and was happy to start earning money in agriculture. She met her husband at work and although she wanted to marry him, felt bypassed in the decision when he sent elders to ask her parents’ permission to marry. She refused his proposal, explaining: “It was my right to decide my own fate,” and that “he should have asked me first and go with my agreement before sending elders to my parents. My parents have no right to decide on my behalf. I have the right and capacity to decide about my future life.” Ayu later consented to marry this man through a voluntary union; she thought this arrangement would give her greater say in her future. By age 21 she had two children, but married life had not given her the scope for decision-making she had hoped for. Her husband had not allowed her to continue working, she had lost touch with friends on moving house and her mother had made her give birth to her first child at her mother’s house, despite her expressed preference to have the baby at a clinic. Nonetheless, Ayu described herself as happy and satisfied and felt her husband supported her plan to open up a small shop when the children were older.

Since informal unions such as Ayu’s are seldom endorsed by clan or family, they carry significant social risk.¹³ Some of the young women involved expressed regret at having bypassed arranged marriage and several felt they had been pressured into marrying in this way. Buzuneh married at age fourteen out of self-protection, to counter rumours that she had had sex with her then eighteen year-old boyfriend. She initiated the marriage through litu/seena, a local form of voluntary marriage where the girl enters the groom’s parental home without permission, prompting his family to accept the marriage. In hindsight, she reflected that her marriage had been forced; having married to avoid the shame of a premarital relationship she experienced the shame of having married too young. Young men’s promises of economic support were often broken: “We get married because the boys disturb us when they think that we are matured enough. They just follow us and promise that they would help us to live a better life… if they promised that they would buy us good clothes and live a better life, we say, ‘ok’ [laughter]”.

Motherhood seems to heighten the challenges to young women’s agency, especially for those in informal marriages. Concern was expressed about the likelihood of women being abandoned by partners following pregnancy and childbirth; having ignored traditional marriage practice, following abandonment women in informal unions are not necessarily able to count on familial support. There was scepticism about whether girls empowered to make decisions for themselves grasped the potential adverse consequences of this newfound agency. Indeed, many young women were uncertain in their maternal roles, and seemed to oscillate between co-existing identities --- ‘child’ and ‘daughter’ and ‘mother’, ‘wife’ and ‘woman’, these ‘both/and orientations’ (Collins, 1990) suggesting the multiple standpoints and shifting agency of young mothers (Harding, 2004). In this way, motherhood in childhood rendered motherhood and childhood fluid rather than fixed categories (cf Johnson-Hanks, 2002). Some of those who were married and/or mothers did not see themselves as adults. For example, despite being age 21 and having given birth three years previously, Asmeret insisted, “I am a young girl because my age does not make me a grown woman. I am a mother to my child, but by my age, I am still young.” Nor did Tirhas (aged 18) consider herself a grown woman, although she had married at age 13 and gave birth age 15: “It is not a matter of age, but of attitude.” On the other hand, 19 year-old Haymanot was resentful that she looked ‘older’ than her actual age. Even though she had two children, she stated; ‘I see myself as a youth (menesay), but people think that I am an adult woman (abay sebeyti).’ Her mother, meanwhile, referred to Haymanot as ‘still a child’.

Assessments by some of the mothers in the younger generation of what it was like to give birth at a young age revealed a mix of joy, regret, gain and loss. Many felt they were not ready to become mothers so young, as one noted: “I gave birth to my first child at the age of 13… I did not know about motherhood, and I could not provide care for my child. My mother took the responsibility to care for my child. I couldn’t even give breast milk for my child. I acted as a child even after I gave birth….” Yeshi, who had her first child at around age 15, felt guilty for weaning her baby too soon: “I really feel bad… but I was a naïve person with a childish personality”. They gradually developed the capacity to provide maternal care and through this, new ways of thinking and relating to others (Ruddick 2007: 97). Meanwhile, some of the women saw motherhood as having exacerbated their poverty: “If your life is poor, you get disappointed to get pregnant and have a baby. You wonder what sin you did to be punished by poverty”. Financial hardship clouded the potential joys of mothering, as childcare responsibilities constrained young mothers’ efforts to get ahead (Ruddick, 2007: 98); “Now I am tied at home because I have to care for the children… We had a plan to improve our livelihood… however, I became busy with children and my husband became the only breadwinner for the family.”

CONCLUSION

This article has drawn on narratives of marriage and motherhood from three generations of Ethiopian women to reflect on changes in the expectations and experiences of female childhood. It argues that marriage and motherhood continue to be both highly valued sources of feminine identity and prerequisites for transitioning to full adult status for the younger generation of women. But in the context of rapid social and economic change the age and social meaning of these transitions are increasingly contested. In alignment with norms promoted by international actors, Government policy represents motherhood and marriage as incompatible with modern ideals of female childhood. Marriage and childbirth before the age of 18 are perceived to run counter to the goals of development and render teenage mothers victims of child rights violations. The evidence from Young Lives suggests that these norms are increasingly internalised by girls and their families since the timing and social significance of marriage and motherhood are changing, with the younger generation of women spending longer periods in school and delaying transitions to adulthood. These developments mark a changing gender script for young women, mapping a normative trajectory from childhood schooling, to adult employment, followed by marriage and motherhood.

Even though the revised gender script is accepted by many of the women in all three generations as the one to strive for, the majority of those in the younger generation were unable to achieve it. The lived experiences of the younger generation of women reveal complex, highly nuanced and at times contradictory expressions of female agency across the early life course. Marriage and motherhood continue to be a childhood reality despite policy approaches aimed to translate conceptualisations of modern childhood from the international level to local contexts. The notion of ambiguous agency highlights the contradictions confronted by the younger generation of women as they contend with the discourses and realities of marriage and motherhood. The expansion of schooling has heralded important shifts in Ethiopian girls’ roles and imagined futures, and imbued many in this generation with a sense of expanded choice in their school, work and reproductive trajectories. Many of them were exposed to emergent social values related to children’s rights and gender equity, cultivating for some a sense of agency and self-interest that played out in marital decision-making. Examples are given of notable changes in marital practices wherein teenage girls and young couples in some communities are deciding for themselves when and who to marry. But such expressions of agency are also shown to come with social risks, especially following motherhood. Although able to influence their marital choices, over time and particularly once they had children, these young women were increasingly subject to the influence of patriarchy.

The Young Lives findings resonate with insights from feminist scholars of development who, first, highlight the role of context in making sense of young women’s choices and ‘empowerment’ (Cornwall and Edwards, 2010; Folbre, 1994). For example, attention to the everyday circumstances and material and cultural logics of motherhood in childhood would caution against simple attributions of either victimhood or agency to teenage mothers (Ansell et al., 2018). This requires moving beyond adult-centric approaches that tend to equate motherhood with women and that position children as dependent objects of adult care, these approaches reducing motherhood in childhood to problem, or even ‘deviant’, childhoods (Caputo, 2018). Second, a temporal perspective is attuned to life cycle dynamics, and can trace the non-linear pathways between choice – action - outcome that are too often oversimplified in mainstream development discourse (Cornwall and Edwards, 2010). One example of this is the case of informal unions arranged by young couples lowering rather than raising the age of marriage. Third, the rhetoric of ‘choice’ and agency that is often used in Ethiopia to characterise the current generation of girls and young women is only partially borne out in experience. Though it may appear that girls nowadays have many more possible life paths to pursue, their actual choices remain heavily constrained by poverty and by restrictive gender norms; given vested familial interests in women’s reproductive capacity, seemingly these restrictions rise as they become mothers. This reality requires “more attention to both the structural inequalities that women face, and the lived realities of challenging these” (Chopra and Muller, 2016: 8). Addressing structural inequalities increases the likelihood that gains made early in the life course in terms of encouraging girls’ expanded horizons and self-determination can be sustained through the transition to adulthood. Finally, the young women in this study became mothers at different ages within the second decade of their lives, a life phase everywhere “shown to be a critical time for examining the proximate interactions of individuals with the cultural models and norms transmitted to them in the re-creation of new generational norms and practices” (LeVine, 2011: 426). Young mothers have a vital role to play in influencing these processes of social change and in reshaping the norms of family life and social relations.

ACKNOWLEDGEMENTS

The authors thank the daughters, mothers and grandmothers who participate in Young Lives, for giving their time and for sharing details about their lives. We thank Dr Yisak Tafere who led data collection and the wonderful team of researchers in Ethiopia. Thanks also to Kristine Briones for excellent research assistance on the survey. We are also grateful to Dr Charlotte Beyer and for the suggestions provided by anonymous reviewers. Young Lives has been core-funded by the UK Department for International Development (DFID). The qualitative data used in this article were supported by a combination of funding from DFID and the William and Flora Hewlett Foundation. Analysis of the data and writing were supported by the International Development Research Centre (IDRC) in Canada as part of the Young Marriage and Parenthood Study (YMAPS) involving Young Lives in Ethiopia. The views expressed are those of the authors.

1. See UNFPA 2013 and UNFPA 2015 where we first encountered this term.↩

2. This definition of childhood, in which age 18 is the upper threshold, is the one endorsed by the UN Convention on the Rights of the Child and the Ethiopian Government. Until recently, social definitions of childhood in Ethiopia did not refer to chronological age. But the government definition is employed widely in law, policy and practice, and therefore is increasingly influential in shaping popular understandings and discourse.↩

3. The terms ‘young women’, ‘young mothers’, and ‘younger generation’ are used to distinguish between the young women in the core sample and the two older generations of women, termed ‘older women’, ‘older mothers’, ‘grandmothers’, ‘previous generations’.↩

4. We acknowledge national legal differences and socio-cultural variation in the definition of a child.↩

5. Growth and Transformation Plan II (2015/16-2019/20), page 56.↩

6. www.younglives.org.uk↩

7. We focus on the older cohort, born in 1994.↩

8. Main data sources: qualitative longitudinal case study research (2007, 2008, 2011, 2014); child marriage and parenthood sub-study (2015); and a sub-study on children’s work (2013).↩

9. In stark contrast 7% of young men had married by age 18 and 2% had fathered a child by age 22.↩

10. In the Round 5 survey, the average age gap between young women and their husbands was seven years (Briones and Porter, 2019).↩

11. The Government reported, ‘10,652 children’s councils and clubs were strengthened and 2.96 million children were given awareness about their rights’ (Growth and Transformation Plan II (2015/16-2019/20), page 57).↩

12. It is unclear from our evidence whether there has actually been an increase in premarital sex among the younger generation compared to earlier generations.↩

13. Some of these unions are subsequently formalised through a customary ‘reconciliation’ process involving the transfer of gifts and cash to the bride’s family.↩