Women bargaining with patriarchy in coastal Kenya: contradictions, creative agency and food provisioning

Abstract

Gender analysts have long recognised that challenging existing patriarchal structures involves risks for women, who may lose both long-term support and protection from kin. However, understanding the specific ways in which they ‘bargain with patriarchy’ in particular contexts is relatively poorly understood. We focus on a Mijikenda fishing community in coastal Kenya to explore contradictions in gendered power relations and how women deploy these to reinterpret gendered practices without directly challenging local patriarchal structures. We argue that a more complex understanding of women’s creative agency can reveal both the value to women of culturally-specific gendered roles and responsibilities and the importance of subtle changes that they are able to negotiate in these. With reference to food provisioning, the analysis contributes to more nuanced understandings of gendered household food security and women’s creative approaches to maintaining long-term security in their lives.

Keywords:

• Agency
• gender relations
• Kenya
• patriarchy
• small-scale fisheries

Introduction

Gender analysis of the household has focused extensively on inequality, particularly between men and women in marriage, but has historically devoted less attention to the ways in which women exercise power within the household and the benefit that they derive from this. In this article, we contribute to a critical scholarship that explores the contradictions inherent in gendered power relations of the household, the resulting complexities of women’s relational interests in patriarchal structures, and women’s creative agency in reshaping these in their interest. We do so through an empirical investigation of a Mijikenda fishing community in coastal Kenya with particular reference to food provisioning. However, before turning to our case study, we rehearse the theoretical concerns that form our starting point and why these matter for gendered understandings of household food security.

Although academic attention to the ‘patriarchal bargain’ is mostly concerned with the way in which women’s interests are bound up with the interest of their hierarchically-positioned husbands, and to a lesser extent their other male kin, patriarchal family relations also undisputedly wrap up men’s interests with those of their wives, and their other female kin. So, whilst considerable analytical interest has focused on the way in which inequality in patriarchal structures constrains women’s agency, somewhat less attention has been paid to how interdependence can be generative of power for women (Stark 2016). Following Habermas’s (1976) analysis of the way in which the incomplete structure of class is manifest in contradictions that offer opportunities for subordinate groups to negotiate, Connell (2009) conceptualises the interdependent nature of gender relations as showing similar contradictions that can be strategically exploited by women. Crucially, whilst the domestic arena is a major site for the reproduction of gender inequalities, it is also ironically a significant site in which women can exercise considerable power, including in relation to men (Robson 2006; Jackson 2007; Hanrahan 2015; Stark 2016).

As Jackson notes, ‘bargaining with patriarchy’ (Kandiyoti 1998) is often taken to suggest a simple ‘bartering of power and personhood for material security and protection in a world where these are essential for survival’ (2007:124). In this interpretation, the scope for agency is restricted to a defined room to manoeuvre within existing material circumstances and established gender norms. This is problematic since the very notion of agency ‘necessarily involves a partial transcendence of its material conditions of emergence’ (McNay 2000:4). A stronger conceptualisation recognises material and ideological constraints but also acknowledges ‘those creative or productive aspects immanent to agency in order to explain how, when faced with complexity and difference, individuals may respond in unanticipated and innovative ways which may hinder, reinforce or catalyse social change’ (McNay 2000:4). As Jackson notes in relation to marriage ‘[c]onjugality also offers possibilities for women to manipulate discourses of respectability, manage ironic performances of compliance, and engage in cultural inversions and mimicry of the gender order (Boddy 1989)’ (2007:124). Thus, making space for a more meaningful conceptualisation of agency as being potentially creative, invites more detailed empirical attention to ‘bargaining with patriarchy’.

The locus of much debate about ‘bargaining with patriarchy’ is implicitly or explicitly the conjugal dyad of man and wife and considerable attention has been given to culturally-specific norms about the entitlements and obligations of husbands and wives (Jackson 2007:126). Whilst their symbolic significance undoubtedly extends beyond the married state, a focus on conjugal bargaining neglects the significance of gendered bargaining with others within and beyond the kin group during marriage and the way in which the nature of bargaining with patriarchy changes across the life course as individual women marry, divorce, or choose to remain single. We therefore suggest that there is value to considering ‘bargaining with patriarchy’ in a broader framework that attends to the creative agency that women (and men) deploy at different times in their lives.

Such an approach is consistent with the ‘performative’ view (Butler 1990) of critical social theory that sees everyday practices as constituents of gendered power. Everyday practices include productive activities, marriage practices, rituals and social activities, as well as the words used in these practices, shaping symbolized meanings and gendered emotions (Connell 2009). Whilst everyday practices broadly reify men’s subordination of women, they also provide opportunity for women to employ their creative agency - often in hidden or implicit ways – to reinterpret, manipulate and indeed change actual power relations (Butler 1990). Women’s exercise of power is often hard to see not only because of their strategic intent to be unthreatening but also because their subjectivities influence their emotions of pleasure, shame and guilt and they benefit from the protection and security they gain through existing gender relations (Kabeer 2000). Wifehood and motherhood shape women’s perceptions and expectations of conjugal relationships (Whitehead 1981) and it is difficult to disentangle affect and interest (Molyneux 1998) for women as agentic subjects as well as for researchers. Although methodologically challenging, it is undoubtedly problematic that exploration of women’s interests in marriage has ‘tended to focus narrowly on female altruism as the complete denial of self-interest with less attention given to the potential (and even strategic) synergies between the two’ (Brickell and Chant 2010:150). We argue that these latter possibilities need further scrutiny if we are to understand better how women exercise creative agency in everyday gendered practices.

Despite its often extremely subtle nature, women’s power can drive changes in gender relations, including those within the family. Indeed, studies of women’s power to change gender relations through everyday practices are long standing in anthropology. Moore (1986) shows how Endo women deployed the rituals of female circumcision rites as a source of female power to negotiate for their material needs with their husbands and male kin, whilst Parkin and Nyamwaya (1987:16 cited Jackson 2007:123) note that ‘though women have often been depicted as … the subordinate bearers of men’s children, they have also been shown … as the primary agents of change in marriage and the family’. These and other ethnographic studies (Francis 1998; Feldman 2001) suggest that theorisation of ‘bargaining with patriarchy’ needs to move beyond seeing the architecture of patriarchy as a ‘given’ or non-negotiable constraint and instead to understand in a more complex way women’s roles in constructing, reifying and reshaping that architecture.

Food provisioning is a core site for negotiations of everyday gender practices. Food provisioning has a central bearing on women’s security in small-scale fishing communities in the developing world, such as the Mijikenda fishing community in coastal Kenya, where our study was based. However, weak understandings about the nature of gender bargaining are evident in the particular framing of women’s empowerment as key to enhancing household food security (Brickell and Chant 2010:153; Jackson 2007:109; IFPRI 2002). The vision of women’s empowerment that is promoted by development policy revolves, in large part, around increasing women’s bargaining power within the household by promoting their economic independence from it. By default, this vision risks undervaluing the extent to which the household delivers security, including food security, and the creative ways in which women wield their bargaining power to develop that security over the longer term (Cornwall et al. 2007; Jackson 2007; O'Laughlin 2007).

Ethnographic fieldwork was conducted from May 2011 to March 2012 in a Mijikenda fishing village in South Kilifi District, Coastal Kenya. This study was not related to a development intervention. Our focus in this article is on the experience of four women whose varied family background and economic conditions reflect realities experienced by other women in the study village and with whom the first author developed a strong rapport. They are Najima, a currently-married woman raising young children, two divorced women, Tumaini, who lives alone with her children, and Zawadi, who has live-out partnership with a married man, and an older woman, Saumu. Accordingly, these four allow us to explore bargaining with patriarchy not only within and around marriage but also beyond and after it. We discuss our findings thematically beginning with the role of marital relationships in security and food provisioning, moving on to consider gendered bargaining around food provisioning after divorce or separation and looking finally at how gender bargaining is experienced by older women.

Below, we build on these theoretical concerns to explore women’s creative agency in bargaining with patriarchy through everyday gender practices around food provisioning in a small-scale fishing community. The next section introduces our research context and briefly describes the methods used. Following this, we turn to our findings which investigate how differently-positioned women perceive, experience and negotiate with contradictory gender relations around providing food for their families across the course of their lives. In doing so, we illuminate the creativity of women’s approaches to bargaining with patriarchy with the aim of extending mainstream understandings of gender bargaining with patriarchy in non-industrial settings in sub-Saharan Africa. The article concludes by stressing the importance of analysing the complex ways in which women exercise power in specific contexts if we are to understand how household food security is gendered.

The research context and methods

The Mijikenda ethnic group live along the coast and its immediate hinterland from the border of Tanzania to north of Malindi, and have a shared Mijikenda identity with significant patronage networks through livelihoods and marriage (Wills 1993). Although there is an absence of contemporary ethnography on the Mijikenda, important studies from the 60’s to the 80’s offer a good understanding of gender and social relations. The study focuses on Mijikenda who emigrated from the immediate hinterland to the coast between the 1930s and 1950s and initiated a cash economy based on commodities such as palm products, fish and cashew nuts. They eventually expanded their livelihood opportunities through intermarriage with those who have resources on the shore. This is what Herlehy (1984) describes as ‘blood-brotherhood’ relations, highlighting the Mijikenda’s specific strategy of using a blood-connection as a means of securing a livelihood. They have extended their kin relations by arranging their daughters’ marriages with men from different areas and ethnic groups as a means of coping with seasonal food shortages and expanding their trading activities (Herlehy 1984). Their businesses are conducted within local networks from production to consumption. Parkin (1972) calls this economic system a ‘redistributional’ or ‘internal’ economy, the profits from which are redistributed within the community in various ways, in contrast to the capitalist or external economy in which profits are reinvested in an individual’s economic activities, delinking their livelihood from the community.

Even today for example, in the study site, the trading businesses of fisheries and palm wine favour the Mijikenda women, many of whom have male relatives who are palm-tree owners or tappers, or fishermen. Based on such blood-brotherhood relations (Herlehy 1984), Mijikenda people still partly sustain the ‘redistributional economy’ (Parkin 1972) and see security in their lives as being about having udugu (kin or supportive relationships). Both cash and food are exchanged among the udugu members and this plays a significant role in maintaining security. The Mijikenda who come from large families consider their situation as not poor. The term ‘poor’ (maskini) is associated with the size of the family rather than income. Migrant families, single mothers and divorcees, who come from distant places, are considered as very poor and food insecure, as they do not have enough support from relatives in times of need. In such a situation, njaa (shortage of food) may continue for longer periods and occur more frequently than for the people from big families. In this article, we consider Mijikenda people’s feelings of security based on their ideology of udugu and explore how Mijikenda people achieve it and what this means for their strategies around accessing and assuring a supply of food.

Data were collected by the first author who lived in the village and interacted with many villagers over the course of the fieldwork. The first author speaks Kiswahili fluently and therefore all data were collected without interpreters and recorded in field notes.

Ethnographic fieldwork on the topic of gender requires consideration of reflexivity (England 1994:244) around how the positionality of the researchers shaped informant’s behaviour (Callaway 1992) and in turn infleunce the researcher’s interpretations.

In this study, the first author was initially categorized by the host community as mzungu, which means a white or a non-black person. The word implies ascendancy over the local people, due not only to the skin colour but also to the characteristics of mzungu in general as holding economic and educational status, as representing donor agencies, and with using English. Being seen in this way could have limited access to local people, especially poor men and all women, but speaking Swahili, living as a villager, behaving modestly, dressing acceptably and showing particular respect to all men and older women questioned her initial identification as mzungu.

Gradually she came to be seen more as an ‘unmarried girl’ (msichana) and as such she needed to follow the gender rules and norms in the community in order to access men and married women. She first built acceptance by the male leaders in the village and their families, talking with them about the life histories of her grandfathers, fishing in her country, rice wine, Buddhist funerals and so on, before attempting to build close relationships with married and unmarried women. Her positionality changed during interactions over the course of the fieldwork and some respondents eventually viewed her as a friend or a daughter.

Throughout her respondents representations of themselves responded to their understanding of her identity and relationship to them. This influenced the ways in which they ‘storified’ their experiences (Reissman 1994:114) revealing both the gendered ways in which they want to be understood and what their expectations were for her behaviour (Caplan 1993). Reflecting on this process provides insights for interpretation: being alert to elements of self-justification, post-hoc rationalisation and selective self-representations is revealing of respondent’s subjectivities. Beyond this, the researcher’s own subjective experiences, emotional responses in the field, and intellectual preferences also impinged on how she made sense of her respondent's narratives (Callaway 1992). Co-authoring involved challenging initial interpretations through consideration of alternatives on the basis of respondents’ testimonies and other aspects of empirical evidence. Nevertheless, our final analysis remains an interpretive exercise that privileges our sense-making and, like all similar accounts, is necessarily subjective, provisional and open to re-interpretation.

Marital relationships, security and food provisioning

Marriage is central to security in Mijikenda society and is a universal expectation for all women (and men), almost all of whom will marry at least once in their late teens or early 20s, by which their status changes from a girl to a woman. The local definition of marriage includes established sexual partnerships whether or not couples have been officially married. Marriage differs from temporary or commercial sexual relationships in the sense that women in marriages are responsible for food provisioning and childbearing. We focus here on two key reproductive roles – child-bearing and food-provisioning – and examine the way in which contradictions in gendered power relations around these roles can be used creatively by women in their negotiations around security.

Having many children, in particular sons, is a symbol of wealth through which men demonstrate their power and prestige. Having many children can therefore be a source of power for women and it partly shapes women’s desire to have many children (Bledsoe 2002). In this context therefore, fulfilling domestic responsibilities including child-rearing and food-provisioning can be a powerful negotiating tool through which women can strengthen their bargaining power without losing their husband’s favour or his provisioning. Similarly, whilst women’s domestic roles and responsibilities in cooking are a marker of male dominance as bread-winner, they are also of value to women both intrinsically and as resources for negotiation. In Kumasi, Ghana, Clark (1989) observes that young Asante women often prioritise cooking the evening meal for their spouse over earning cash through their own trading at the market. Despite the hard work involved, the quality of the meal, in terms of not only the type of dish but also the time and energy spent on cooking, expresses the extent of her satisfaction with her spouse. Although cooking is a significant burden for young women, the women receive substantial returns in the long-term that lead to favourable conjugal relations, thereby securing long-term material and economic support.

Whilst women’s reproductive responsibilities are at the heart of gendered power relations in marriage, married women with young children put the contradictions around their responsibility for reproductive roles to work in creative ways when they are bargaining with patriarchy, as Najima’s experience illustrates.

After a very difficult childhood, Najima’s fortunes changed significantly with her marriage. Najima’s father died when she was about 7 years old and she was raised by her mother at her natal home. She was sent to Mombasa to work as a house-girl when she was about ten years old, but ran away because she was sexually abused by the male head of the house. Arriving back at her mother’s natal home, she found that her mother had gone to live in Mombasa, so her uncle arranged for her to live and work as a cook in a small café in the nearest town. There, she eventually developed a close relationship with a palm-wine tapper and they married. She considers her marriage to have changed her life, describing her first day in her husband’s homestead as unforgettable. In her words: ‘I was so happy and excited. I had my own house and farm. My life was finally secured on that day.’ Now 30 years old, she has established herself as the wife of a palm-wine tapper and the mother of four children, and this social status forms the foundation of her security. Her husband’s social connections enable her to access fishermen from whom she buys fish on credit to sell on. The profit is small, but she controls it. Below we examine in more depth how Najima’s bargaining with patriarchy has been integral to gaining this security for herself and her family through marriage and wider patriarchal networks.

Najima and her husband live in the temporary residential area for migrants and her husband spends most of his time with other tappers and friends drinking palm wine in the public space under the tree in front of his house, and therefore her behaviour is always being monitored by his colleagues and friends. A friend of her husband’s said ‘Najima is an ideal wife. All Mijikenda wives should be like her’. Another said ‘He finally got a baby boy recently, which reassured him. If his wife has two or three more sons, he’ll be a real strong man’. Talking about their wife or partner is a main theme when men drink and, therefore, Najima’s good reputation is important in maintaining her husband’s social life with other men, and for this he depends heavily on Najima being a good wife and her reproductive capacity. It is also important that Najima is well accepted by his palm tree owners and his colleagues, so that she continues to be allowed to use the nearest borehole and to collect palm leaves and husks for firewood at the campsite and negotiate her fish trading relationships. Locally-recognised masculinity shapes men’s ideals of what a woman should be, and Najima tries to be the ideal wife as a means of strengthening her marriage as well as the access to resources that this brings her.

Najima usually fulfils her husband’s requests even if she has to compromise her fish business and her childcare. For example, one day in February 2012 while she was frying ray (fish) to sell she was disturbed by her son, who cried four times. She knew she could not ask her husband to hold his baby while his colleagues were around, as it could embarrass him and damage her reputation in the campsite community. Each time the baby cried she stopped frying and suckled him, changed his nappy or tried to please him. Meanwhile she found that the frying pan had a small hole and oil was leaking out which had caused all the firewood to burn up, filling her kitchen with smoke. Her son cried increasingly loudly.

In the midst of her struggle with the frying and her crying son, her husband called her to cook lunch for his colleagues and friends. Tellingly, Najima looked pleased to be asked and said loudly ‘Sawa’ (OK). She left the ray and immediately started cooking for the men, prioritising her husband’s request over her own small business and childcare. She collected firewood again, went to fetch water, looked for a big aluminium pan, washed it and boiled the water. A strong wind made it difficult for her to control the fire and it took another hour and a half to cook the meal for her husband and his friends. While she cooked, she explained to the first author that she was a better cook than the woman next door. This suggests that she may perceive his request as an opportunity to show what a good wife she is and thus strengthen her husband’s opinion of her and that of the wider community.

Her extra cooking work delayed the rest of her work and she and her children skipped lunch. She reasoned that ‘I do not have to cook lunch for us. My husband has already eaten’ and only cooks lunch when her husband is around. After frying the fish, she cooked the evening meal, which took another hour. During supper, her husband held his baby boy affectionately, looking at him constantly. After supper, he said quietly to Najima ‘Thanks for the meal’, and she replied ‘You’re welcome, and thank you too’. He handed the baby boy to Najima and went to tap trees. After her husband left, Najima suckled their baby and said ‘I am really happy living with him at the camp because he has more emotional attachment to my children. He is much more supportive now’.

Her routine activities are not mechanical but deliberately organized in favour of her husband and his friends rather than her children or her fish-processing work. Through this behaviour she secures their well-being by developing her relationship with her husband as well as ensuring his long-term support, since kin relationships (udugu) are valued intrinsically for their ability to provide for long-term, all-round security. Her account confirms her creativity in bargaining around patriarchal relations: her strategic behaviour has generated udugu and this in turn means that she and her son are more secure, althouth this long-term security is obtained at the cost such as her and her children’s nutrition and her labour and time burderns. The value that she attaches to this relates closely to her earlier painful experience when, after her father died, her mother moved away and she found herself without the security that supportive social relations can provide. Although she must work very hard, she feels secure in her marriage and proud of what she sees as her achievement in building and maintaining it. Her case illuminates how women’s bargaining power related to food security is embedded in, rather than isolated from, broader patriarchal social relations. Significantly, this bargaining is not only about trading tensions between her individual interests and those of her husband’s and/or the patriarchal community, Najima sees her bargaining as generating more space for good outcomes for her and her family. Although they remain very poor, it is simplistic to see this as coping: to do so would be to undervalue Najima’s creative agency, the udugu she has generated, and their contribution to shoring up her household’s food security over the longer term.

Negotiations after divorce and separation

Bargaining with patriarchy is not confined to conjugal relations. The breakdown of a marriage does not necessarily mean the end of bargaining, nor a failure to bargain effectively on a woman’s part. Indeed, bargaining with patriarchal relations beyond the marital dyad is likely to be reconfigured but still significant to women’s security in the event of marital breakdown (Seeley 2012).

Although marital relationships in this context are considered to be a foundation of security by the Mijikenda, this does not necessarily mean that women have no option for divorcing. Although we do not have data on divorce and separation in the study site (where marriage may not be formalised), divorce and separation were not uncommon. What is important for women in particular is whether they have an alternative form of support (udugu) or not, and women work to maintain or develop new udugu relations. Here, we examine how women’s creative agency around food provisioning is reorganized through divorce and separation in the current context by exploring the experiences of Tumaini and Zawadi.

Tumaini, 24, was taken to the homestead of her future husband in the hinterland of Kilifi when she was around 8 years old and was married immediately after her first menstrual period. After five years, she divorced him to escape domestic violence and moved alone to the coast, where her two sisters lived. There she met a boyfriend who built her a house, helped her to gain access to fish trading, fathered a baby girl and later left her. Since then she has lived with a daughter by her ex-husband and the child she had with a boyfriend plus a nephew; the two sons she had with her ex-husband live with him. She still lives in the house that her ex-boyfriend built for her and engages in fish trading and is proud of raising her children alone. However, her family has no cash and little food in the kitchen compared to other respondents who had a partner. The following account is revealing of how she secures food provisioning in these circumstances.

During the third week of March 2012, Tumaini stopped her fish trading to prepare her farm for the rainy season, but she went to her field for only one day and then rested at home for six days saying that she was too tired to work. She ran out of money after three days and had no food in her kitchen. On the fourth day, she sent her children to her elder sister’s house for supper while she ate at a friend’s house. Her elder sister is a widow supported by her sons and Tumaini feels free to visit her and her daughters. On the morning of the fifth day, her regular customer repaid the KSh 20 (this was equivalent to US$1 at the time of fieldwork 2011/2012) that he owed her for the fish he had bought on credit the previous week. She and her children ate KSh 20-worth of fried potato fritters for breakfast. In the afternoon, by chance she met her youngest daughter’s grandmother on the way back from the field, who gave her 2 kg of beans and two coconuts. In the evening, she cooked all the beans with the coconuts. Although this could have fed her family for two days, she gave half to her elder sister’s family and a bowlful to her neighbour, who came to ask for food, whilst she and her children ate the remainder that evening. On the sixth day, her neighbour brought some chapattis and Tumaini and her children ate them for breakfast. She explained that the neighbour had borrowed her pan to make chapattis for her boyfriend and so she returned it with some of the chapattis. In the evening, she sent her daughter to her friend’s house to ask for a bowl of maize flour and cooked ugali (maize porridge) with local wild green vegetables that she picked from her younger sister’s house. The following morning, she started to trade fish again, purchasing fish on credit from ‘her’ fisherman, a brother of her ex-partner.

It is clear that Tumaini has an extensive support network that helps to protect her family from going hungry. This mutual support system is not based simply on her kinship: she asks her widowed elder sister and three unmarried friends for food. The three friends have partners, and the four of them support each other in times of need. For example, Tumaini helped them during the high fishing season between November and December 2011: she gave maize flour, cooked beans and coconuts to these friends and her elder sister at least seven times within a month, and it seemed that she spent more of her earnings on them than on her children. Although Tumaini has a very close relationship with her younger sister, she rarely asks her for help. Her reasoning is that her younger sister is married, and asking a married woman for food is not socially acceptable. As an example, Tumaini frequently offered the first author a meal when she visited but Najima rarely did so, waiting until her husband offered because she did not feel free to offer food to a friend without his permission. Consequently, the first author was careful to visit when husbands or partners were around, to ensure their acceptance, and observed that when she visited the homes of male respondents, their wives looked after her as their husband’s guest. This reveals contradictory power relationships in cooking and eating: although wives are primarily responsible for cooking, they may not have autonomy in the distribution of the meals they cook. Significantly, while Tumaini does not ask her younger sister for food, she gives her food when her husband fails to provide cash for the family. Tumaini explains that by doing so, she benefits from her sister’s husband in ways such as using his connections in her fish trading business, or to repair her house or solve problems with her neighbours. Thus, her support network is highly gendered and she uses subtle everyday practices to bargain with patriarchal relations in ways that support her security in food provisioning.

In this way, Tumaini ensures that her family’s food security is relatively stable, regardless of her income on any particular day. Her fish trading is not her sole means of providing meals. Her family’s food provisioning is closely associated with her interdependent relationships with friends’ partners, her older sister’s son, her younger sister’s husband and male kin and her ex-husband’s family. By engaging with, and seeking food from, friends/relatives Tumaini demonstrates her ability to bargain through patriarchal structures, whether they are with women or men. She is creative in her strategies to develop and maintain relationships from which she derives both intrinsic and material value. For Tumaini, security lies not in assets or food but in relationships. Having been married off at eight by her father and suffered from domestic violence by her husband, patriarchal dominance is part of her life. Food provisioning is achieved through building and investing in diverse relationships which are different from the oppressive forms of patriarchy that she has experienced (Jackson 2007).

Now we turn to Zawadi. Her husband was a migrant worker from Tanzania who died in a traffic accident in 2008 when they had four children. Because she had no relationship with her husband’s kin, she was not under pressure to follow cultural custom by remarrying his brother or forfeiting her children to his family. She decided not to remarry, because new husbands do not customarily accept responsibility for children from a previous relationship and she did not want to be separated from them. Where children (particularly sons) are not in the custody of paternal kin, women who remarry often send their children (or just their daughters) to live with their own mothers because their new husbands often refuse to take responsibility for another man’s children. However, since her mother had died, Zawadi did not have the option of sending the children to her mother. Therefore, she explains, she chose a (live-out) partner who could contribute as a provider but without taking parental authority over her children. The man she chose is a married migrant worker whose family lives in the immediate hinterland. He works at the slaughterhouse in a nearby town and spends most of his time at the coast. He usually stays at her local bar where she serves him with food and drinks.

Zawadi’s relationship with him is beyond that of a sexual partner. She spends his cash not only on providing her children’s daily needs but also on purchasing low-value meat from him, and uses her earnings from selling it to employ three of his relatives, one as a transporter to pick up her wine from her tapper and the other two as bar girls in her small bar. This helps her to cope with both her childcare responsibilities and her palm wine business; more importantly, it makes a difference to her perception of their relationship. She sees herself as helping him rather than depending on him. She said that she did not realize that women could live without a partner until she experienced it after her husband’s death. She, however, emphasised that men are different; men always need at least one woman. She is supporting her partner who is away from his family in the hinterland and she is proud of employing his relatives through her earnings.

Zawadi’s relationship with her partner is favourable to developing her economic activities and raising her social status. Employing somebody is a sign of success in business. Furthermore, the new practice that a woman of childrearing age employs a young man contributes to changing gender patterns in the community (Connell 2009). Although she depends on her partner financially, her social position has risen. She sees men’s support positively as the result of her successful negotiations rather than her dependency. This is indicative of a process of change in gender relations that is creating new gender practices and meanings. Words and practices are redefined by individuals who interpret their meanings differently (Butler 1990), and women exercise agency implicitly by reinterpreting meanings in their own interests (Moore 1986; Kabeer 2000). In this case, Zawadi’s access to food through non-marital relationships is not a symbol of her powerlessness or subordination. This is consistent with the findings of other studies in sub-Saharan Africa (Foley and Drame 2013; Iversen 2005; Kaufman and Stavrou 2004; Longfield et al. 2004; Wojcicki 2002).

Also, the cases of Tumaini and Zawadi show that the social provisioning of food is embedded in gendered redistribution processes in which women necessarily employ different strategies with members of their kinship network and friends. This finding corroborates critiques of economic understandings of the ‘household’ in which members’ strategies are understood to be largely determined by the legal marriage framework (Carr 2005; Guyer and Peters 1987; Moore 1994; Nasirumbi et al. 2014). A focus only on the formal framework of the household obscures not only women’s actual material and economic support but also their negotiation strategies for and investment in maintaining their extensive relationships.

Older women’s interpretations and negotiations

Having described negotiations over food security of three young women with different family backgrounds, we now discuss older women’s strategies for ensuring the daily meals and their role in food provisioning and childcare for their families. Looking at older women’s lives illustrates how the negotiation strategies of women may evolve over the life course.

Older women do not necessarily view their own experience of aging negatively and many find that their social position in the family and wider society may improve with age. The literature suggests that older women’s increased social position derives from various sources such as the local value of seniority and shifts in women’s family position and relationships (Cheater 1986; Oyewumi 2002; Paulme 2011) and post-menopausal women’s roles and influence in rituals, marriage and reproductive health (Berglund 1989; Udvardy 1992). For example, in her study on women’s reproductive role and aging in Gambia, Bledsoe (2002) shows that women’s perceptions of their aging are closely associated with their reproductive outcomes, with women who have many children perceiving growing older as an honourable achievement (2002:256). In general, women rely strongly on their children, grandchildren and kin relationships in later life and see this as recognition of their investment in bringing children up, whilst men’s strategies in later life are generally more partner-oriented (e.g. Cliggett 2001; Oppong 2006). In Ghana, Tsai and Senah (2013) found that strong reciprocal relationships with kin are closely associated with older people’s perception of security and well-being, and that this was especially the case for women. Importantly for older women, Weinreb’s (2002) research in rural Malawi found that in patrilineal societies, male kin are more important than female kin as sources of support. In short, gendered kin relationships, particularly those with children and grandchildren and those with male kin, are important for women’s security as they age.

Through the case history of Saumu we explore her perceptions of, and strategies for, maintaining the supply of daily meals and securing her long-term wellbeing. We highlight the shifts that have occurred throughout her life in her relationships and the changes in negotiation strategies.

Saumu is in her late 50s and has four children by her ex-husband and another four by an ex-partner. Saumu divorced in 1982 to escape sexual abuse from her father-in-law and returned to her original homestead with her four children. She then had a boyfriend through whom she engaged in an informal trading. Her relationship with her boyfriend continued for a long time and she had another four children by him. Although her male partner helped her develop an economic activity, she did not marry him and she emphasised that she did not receive any specific support from him for the children. She wanted all her sons to establish their families in her homestead so that her later life would be secure.

During the course of the fieldwork for this study, Saumu lived with her two school-aged daughters. Her two older sons had their own families in the same compound, and her three older married daughters lived within walking distance of her house. She had a total of 15 grandchildren. She had succeeded in her palm wine trading business in her 20s and 30s by cultivating extensive relationships with her partner and male friends and retired from this work before she turned 50. She explained: ‘My eldest son was ready to support me and I became a grandmother at that time. I thought I had reached the age when I could rest and relax’. She was very proud of having eight children and receiving support from her older children.

Her main household food expenses were provided by her eldest son, while her brother, who lived in a neighbouring village, contributed several sacks of maize a year and the school fees for her younger children, including the youngest son at boarding school. The brother left home when he married, moving to his wife’s father’s spacious plot on the shore, but maintains his house in his original homestead where Saumu lives. Although Saumu has taken de facto control over the homestead, she sees her brother’s support as essential to maintaining that control and recognises that his financial contributions to her are strategic for him in ensuring his long-term responsibility for the homestead, which protected her from being seen as a lonely divorcee with no male supporters.

Her morning often started with waiting for her palm tapper to tap her trees. Her daughters woke up much earlier than her to do all domestic work before going to school. Saumu would sit on her bench and look after her sons’ five children, who lived next door, while their wives were busy with domestic work including fetching water three times, carrying a total of 90 litres. Saumu said that she had worked like that when she was young. She was proud that her current status released her from domestic work. Her married daughters visited her with their young children almost every day and helped with domestic work fetching water, washing her clothes and cooking. Saumu explained that her older daughter did not want to stay with her parents-in-law during the daytime when her husband was not around, and that the other married daughter wanted her own income to spend as she pleased. Saumu’s relationship with her daughters is not unusual in coastal Mijikenda villages, where women tend to develop reciprocal relationships with their natal family rather than their parents-in-law.

Saumu’s case shows that gendered roles and responsibilities remain central to older women’s bargaining power. Saumu’s strategy in negotiating with her providers (namely, her eldest son and daughter-in-law) and also her daughters, is very different from Najima’s strategy with her provider (her husband), illustrating how women’s position can change over time and the different negotiation tools that they use at different life stages. Looking after her grandchildren helps to reduce the younger women’s domestic burden.

Saumu’s own perception of aging was positive as she saw her status as an outcome of her efforts earlier in her life (Bledsoe 2002), and in fact she had substantial support from her family. Her case helps us to understand young women’s negotiation strategies, their desire to have many children and their attitudes to cooking and child-rearing. Labour exchange in domestic work between mother and daughter and mother and daughter-in-law play a significant role as a source of power for women to bargain within the family and wider kin-group to ensure the long-term security. This is an area that past and current research on gender has relatively neglected. There has been a tendency to analyse child care responsibilities, as with other of women’s reproductive roles, largely in terms of their burden of labour and their constraint on women’s ability to earn independent incomes. This study supports some of the new critical work on ‘care’, which recognises that giving care can be intrinsically satisfying and can be a way of gaining or wielding power (Robson 2006; Huijsmans 2013; Hanrahan, 2015). These findings question the vision of women’s empowerment in development policy, which emphasises women’s economic independence. This conventional view overlooks the agency of women within patriarchy and neglects their strategies for building food security. As a result there is a risk of undervaluing the importance of young women’s roles in domestic work for their bargaining power and security in their later lives.

Conclusion

This study has explored ‘bargaining with patriarchy’ with the theoretical focus not only on conjugality but also on broader kin relations at various life course stages. We have looked closely at the contradictions inherent in gendered power relations and women’s creative agency which may bring about subtle changes in personal power relations rather than radical changes in gendered structures. The conceptualisation of agency as being potentially creative, has allowed us to draw more detailed empirical attention to the ways in which women bargain with patriarchy across the life stages.

We have highlighted the resulting complexities of women’s relational interests in patriarchal structures. Mijikenda women deliberately organize gendered everyday activities such as cooking, child-rearing, budget allocation and economic activities, in ways that provided them with opportunities to strengthen their long-term support relationships. Najima prioritized cooking for her husband over her own trading and childrearing activities; Saumu retired in her late 40s when her income from trading in palm wine was still very high to care for her grandchildren; Tumaini shared her food with her sisters’ and friends’ families even when she had no surplus. These women’s strategies cannot be explained by orthodox approaches to household food security in which individuals’ direct access to economic activities are central (Doss 2017; Kawarazuka et al. 2017). In this context, women’s capacity to maintain their own and their children’s everyday security, including their daily meals, are not simply related to their own assets and income or that of their husbands (Karl 2009; Garcia and Wanner 2017, Doss et al. 2017), rather it is built on their gender roles and relationships within patriarchal structures although the negotiations take place at the cost of their and their children’s everyday diets and labour as shown by the cases of Najima and Tumaini. This corroborates the literature on the domestic arena as a significant site in which women can exercise considerable power, including in relation to men (Robson 2006; Jackson 2007; Hanrahan 2015; Stark 2016).

We have also shown that Mijikenda women develop a wide range of social relationships beyond the immediate family (with both men and women) in which they can also exert significant power. Looking only at intra-household relationships is not sufficient to understand their strategies for long-term security or what these mean for their food-provisioning. Whether within or beyond the family, whether with women or men, these wider relationships are of course also part and parcel of the patriarchal structure (Meinzen-Dick et al. 2014; Wencélius et al. 2016). We have shown that, here too, women are ‘bargaining with patriarchy’ and creatively negotiating for better options and outcomes for themselves and their families in the long-term and for everyday survival (Ali 2014).

So, far from patriarchy being non-negotiable, its inherent contradictions paradoxically create opportunities for women to negotiate within and beyond conjugal relationships (Beck 2017). We have highlighted these contradictions around food provisioning, the creativity that women can deploy in exploiting them, and the resulting negotiations, particularly around food-provisioning. As such, our study contributes a food-provisioning focus for the scholarship on ‘doing gender’ (Huijsmans 2013) and the ways in which everyday practices are constituents of gendered power (Butler 1990). A more nuanced understanding of the creative or productive aspects of agency enables us to evidence the ways in which women bring about subtle changes in gender relations in a specific context. In doing so, we contribute to a better understanding of actual women’s interests in, and experiences of, culturally-specific patriarchal relations.

Notes on contributors

Nozomi Kawarazuka is an Associate Scientist (Gender) at the International Potato Center, CGIAR. She has a PhD in Development Studies from University of East Anglia, UK. Her research focuses on social aspects of food security and nutrition, gendered processes of agricultural innovation and adoption of new technologies in Southeast Asia and sub-Saharan Africa.

Catherine Locke is an independent academic associated with the School of International Development at the University of East Anglia in the UK. She has over 20 years of research experience focusing on issues of gender and social development in Asia and Sub-Saharan Africa. Her research interests include gendered livelihoods, gendered family lives, social reproduction and migration.

Janet Seeley is a social anthropologist by training. She is Professor of Anthropology and Health at the London School of Hygiene and Tropical Medicine. She has actively engaged in research on the social aspects of health since 1987, working in sub-Saharan Africa and Asia. Other current research interests include chronic poverty and social protection, internal migration, livelihoods and gender.

Acknowledgements

We thank our respondents and their families for their warm welcome, cooperation and support. We are grateful to four anonymous reviewers of this journal for their valuable comments. We also thank the Kenya Marine and Fisheries Research Institute and the Wildlife Conservation Society in Mombasa for their institutional support during the fieldwork. This study was partly funded by School of International Development, University of East Anglia, UK and CGIAR Research Program on roots, tubers and bananas (RTB).

Disclosure statement

No potential conflict of interest was reported by the author(s).