Exploring young women’s constructions of love and strategies to navigate violent relationships in South African informal settlements

Abstract

This paper explores young women’s experiences and constructions of love-relationships and intimate partner violence in South Africa, and the role of agency in women’s decisions to remain in or leave violent love-relationships. Understanding why young women stay in or leave violent love-relationships is key to developing nuanced understandings of agency and informing intimate partner violence prevention interventions. Data were collected from 15 young women in informal settlements in eThekwini Municipality, South Africa, via in-depth interviews, photovoice and participant observation, and were analysed inductively. While women’s love-relationships were frequently violent, they often stayed in them for long periods, usually because the relationships met other important aspirations, including the desire for respect and dignity. Nonetheless, many women left when they no longer felt loved and respected, which they believed was shown by men’s behaviours, specifically: indiscreet affairs; not spending time together; not spending money on her and any child(ren); and public (and humiliating) violence. Emotional and economic support from families also influenced women’s decisions about leaving or staying. These data demonstrate that agency goes beyond definitive acts of leaving violent relationships; rather it is fluid, contested and contextual, with many factors influencing young women’s goals and actions.

Résumé

Cet article examine, chez les jeunes Sud-Africaines, l’expérience et les constructions des relations amoureuses et de la violence exercée par les partenaires intimes, et le rôle de la capacité d’agir dans les décisions de rester dans des relations amoureuses violentes. La compréhension des raisons pour lesquelles les jeunes femmes restent dans les relations amoureuses violentes, ou s’en retirent, est essentielle à l’élaboration de compréhensions nuancées de la capacité d’agir et à l’orientation des actions de prévention de la violence exercée par les partenaires intimes. Les données ont été collectées auprès de quinze jeunes femmes vivant dans des zones d’habitation informelles de la municipalité de eThekwini, en Afrique du Sud, à travers des entretiens en profondeur, le photovoice et l’observation participative, puis analysées de manière inductive. Alors que les relations amoureuses des jeunes femmes étaient souvent violentes, il était fréquent que ces femmes y restent impliquées pendant de longues périodes, habituellement parce que ces relations répondaient à d’autres aspirations importantes, parmi lesquelles le désir d’être respecté et la dignité. Néanmoins, de nombreuses femmes partaient quand elles ne se sentaient plus aimées et respectées, ce qui, selon elles, était démontré par le comportement des hommes, de manière précise: liaisons indiscrètes; pas de moments passés ensemble; pas de dépenses pour elles ou les enfants; et violences en public (et humiliantes). Le soutien émotionnel et économique apporté par les familles avait lui aussi une influence sur les décisions des femmes de partir ou de rester. Ces données démontrent que la capacité d’agir dépasse les actes définitifs de retrait des relations violentes; qu’elle est au contraire plutôt fluide, contestée et contextuelle, avec beaucoup de facteurs d’influence des objectifs et des actions des jeunes femmes.

Resumen

En este artículo analizamos las experiencias y construcciones de mujeres jóvenes de Sudáfrica en torno a las relaciones de amor y la violencia de pareja, y el papel de la capacidad de acción en las decisiones de las mujeres de abandonar o permanecer en relaciones de amor violentas. Comprender los motivos que tienen las mujeres para permanecer o abandonar las relaciones de amor violentas es fundamental para poder entender mejor los matices de la capacidad de acción y crear programas de prevención contra la violencia de género. Para este estudio se recabaron datos de quince mujeres jóvenes en enclaves informales del municipio de eThekwini (Sudáfrica) mediante entrevistas exhaustivas, así como la técnica de fotovoz y observación de participantes, que se analizaron mediante un enfoque inductivo. Si bien las mujeres tenían con frecuencia relaciones de amor violentas solían conservarlas durante largos periodos de tiempo, normalmente porque en las relaciones encontraban otras aspiraciones importantes, tales como el deseo de obtener respeto y dignidad. Sin embargo, muchas mujeres las abandonaban cuando dejaban de sentirse amadas y respetadas, un factor que según ellas se mostraba en la conducta de los hombres, por ejemplo con respecto a: historias indiscretas de amor; no pasar tiempo juntos; no gastarse dinero en ellas o sus hijos; y la violencia pública (y humillante). El apoyo emocional y económico de las familias también influía en las decisiones de las mujeres de abandonar o quedarse en este tipo de relaciones. Estos datos demuestran que la capacidad de acción va más allá de las acciones definitivas de abandonar relaciones violentas; más bien es una medida fluida, cuestionada y contextual, con muchos factores que influyen en los objetivos y las acciones de las mujeres jóvenes.

Keywords:

• Intimate partner violence
• love-relationships
• agency
• leaving
• South Africa

Introduction

South Africa has high levels of intimate partner violence (IPV), which may take physical, sexual, emotional and economic forms. Data from informal settlements in eThekwini Municipality, South Africa, showed that 65% of women reported physical and/or sexual violence in the past year (Gibbs et al. 2018). Population-based estimates for Gauteng, South Africa, from 2010, showed a past-year prevalence of physical IPV in adult women of 13% (Machisa et al. 2011). Despite high levels of IPV, women tend to tolerate the violence and remain in love-relationships for long periods of time (Jewkes and Morrell 2012; Shai 2018). Yet this picture is complicated as some women do leave – a few early in the relationship, and others after a longer period of time. This paper uses agency as a theoretical construct to explore why some women stay in violent love-relationships and others leave, usually after a period of tolerating the violence; and how agency, or lack thereof, influences these decisions. This work offers insights for a more nuanced understanding of women’s actions and choices in violent love-relationships and is critical for deepening our understanding of women’s lives and agency.

Kabeer (1999) has defined agency as the ability to identify one’s goals and act upon them, goals and actions which will differ for women in different contexts. She argues that women’s lack of agency and power, often the result of structural barriers including gender inequalities, harmful social norms and poverty, undermines their ability to define their goals and act upon them. A growing body of research outlines how women’s lack of agency places them at increased risk of experiencing IPV (Jewkes et al. 2010; Mannell, Jackson and Umutoni 2015; McCleary-Sills et al. 2016).

Women’s agency in the context of IPV has been extensively explored; however, as noted by Madhok, Phillips and Wilson (2013, 7) discussion of agency is often unproblematised and ‘under-theorised’. Madhok, Phillips and Wilson (2013) and Mannell, Jackson and Umutoni (2015), among others, discuss these limitations, noting that many authors frame women who stay in violent relationships as failing to show agency, arguing that women only display agency if they leave, or report, violent relationships. In addition, these discussions often fail to recognise the intersections with race, class and sexuality and how these impact upon, and limit, women’s agency.

There is a growing body of research to counter this limited approach to agency and IPV (Mahmood 2001; Madhok, Phillips and Wilson 2013; Mannell, Jackson and Umutoni 2015; Campbell and Mannell 2016). These researchers critique the narrow view that agency must involve women ‘leaving or reporting violent relationships’, calling for scholars to recognise women’s particular social and economic contexts, especially in heightened poverty and patriarchy, and to acknowledge women’s individual goals and actions. This work encourages greater recognition of the small acts of survival that women take, including: seeking emotional support from friends and family, or adjusting behaviours to reduce her partner’s violent behaviour, such as not going out or texting friends. Madhok, Phillips and Wilson (2013) also express concern that much recent agency discourse places excessive responsibility on the individual woman to overcome her circumstances on her own, without recognising the broader structural factors limiting her choices. This growing body of research suggests that a more nuanced understanding of why and how women respond to IPV, and exit relationships, will enable a deeper understanding of the women’s lives and the role of agency therein.

Scholars exploring IPV and agency in South Africa highlight similar concerns with narrowly defining agency and not recognising contextual factors impacting on choices. First, structural challenges, especially where women are unemployed, living in poverty and surrounded by patriarchy, severely limit women’s goals and actions (Dunn and Powell-Williams 2007; Jewkes and Morrell 2012). Jewkes and Morrell (2012) also noted that in some cases young women give the appearance of having agency as they may behave in flirtatious ways and are often able to decide whether to begin a relationship; however, once they are in a love-relationship their choices became very constrained by patriarchy, age-hierarchy and poverty. In South Africa, when amplified in informal settlements, these structural realities may have a significant impact on women’s agency, which may then inform and constrain women’s responses to violent love-relationships. Second, dominant social norms in South Africa contribute to the construction of love-relationships and legitimise high levels of controlling behaviour and violence by men and tolerance of this from women (Jewkes and Morrell 2012; Stern, Buikema and Cooper 2016).

Discussion of agency often ignores the role of love in shaping women’s choices and actions. By way of exception, Hunter’s (2010) work in KwaZulu-Natal, South Africa, offers insight into the role that love may play in women’s decision-making. He refers to two types of love: ‘provider-love’ rooted in men supporting women, either through lobolo or by giving money and gifts; and ‘romantic-love’, displayed through men showing affection and kindness and spending time together with their partner. Hunter argues that many women may look for both types of love, and the desire for both influences whether or not they stay in love-relationships. Other authors have found similar expectations concerning love-relationships (Leclerc-Madlala 2008; Wood, Maforah and Jewkes 1998; Jewkes and Morrell 2012). Jewkes and Morrell’s (2012) work in the Eastern Cape, for example, highlighted how young women believed gifts held more than a monetary value, as they reflected a woman’s desirability and a man’s love. Women’s actions within love-relationships were significantly influenced by the desire to be loved, respected and supported, and by constructions of love-relationships.

Further exploration of why young women stay in, or leave, violent love-relationships, and whether they want to disrupt these relationships, is key to developing a more nuanced understanding of agency. It may also help in the design of evidence-based programmes and interventions. By exploring the lives of young women living in poverty in informal settlements in South Africa, this paper interrogates whether, why and how young women exit violent love-relationships and the role of agency therein. It also considers whether women who remain in violent love-relationships are making critical choices, and exercising agency, within their relationships.

Methodology

Data were collected between May 2016 and February 2017 from a group of women from two communities participating in a randomised control trial (RCT) of a Stepping Stones, Creating Futures intervention that aimed to reduce IPV and HIV-risk (Gibbs et al. 2017). Qualitative methods were utilised for their value in meaningfully involving marginalised groups, and their ability to provide insight into women’s social world, offering an in-depth understanding of social context, perspectives and experiences (DiCicco‐Bloom and Crabtree 2006; Sprague 2018).

Setting

The study took place in two urban informal settlements (slums) in eThekwini Municipality. The two communities from which women were drawn were broadly similar, allowing for comparability across themes. Both were very poor, with high rates of unemployment – a reported 39% of young people were unemployed in 2011 in the eThekwini Municipality (Statistics SA, 2011 census) – and low levels of education, over-crowding and limited or no services or formal housing. Despite being similar, the communities differed slightly from one another: the first had slightly better housing and taps in the yards. The second community had more dilapidated housing and shacks, and only communal taps.

Methods

Semi-structured, in-depth interviews (IDIs) were undertaken with 15 women. The interviews were supplemented by photovoice methodology¹ in one community (involving eight of the women) and light-touch participant observation in the other community (seven of the women).

Participants

The women were recruited through convenience sampling. Eligibility for participation in the study included being unemployed, aged between 18–30 and not being in school or college (linked to the livelihoods focus of the larger RCT intervention). Interviews were conducted before or just as the intervention that took place as part of the larger RCT study began. Just over half had children, none were in formal employment and only two had completed secondary school. Women were asked about their relationships in open-ended questions, in isiZulu, and most stated that they were in love-relationships at the time of being interviewed. Three were living with their partners, two others moved between their natal home and their partner’s and one was married. To ensure confidentiality we use pseudonyms throughout this paper, and identifying information has been removed.

Data collection

Each woman participated in two interviews (a total of 30 interviews). The first involved open-ended questions about her life focused on building trust, and the second, usually conducted within two weeks, focused on relationships, power and decision-making. Interview topics included: background information on their lives, relationships, experiences of IPV, livelihoods, power and aspirations. Each interview lasted between 45 and 90 minutes.

All 15 women participated in either the photovoice or the participant observation process. Data collected through these different forms of observation were used to enhance our understanding of the community dynamics and the women’s lives. The interviewer (NN) kept notes of her observations in the field. For women involved in the photoshop element of the study, three 3-hour workshops were held, with two periods in between for taking photos. In all, the women produced 24 photo-posters of their lives. All the interviews and photovoice sessions were conducted by the second author in isiZulu, and translated and transcribed into English.

The research was guided by the WHO (2001) guidance for research in this area (World Health Organization 2001). In cases where participants became distressed or reported being emotionally affected by the research, they were offered access to a counsellor and/or referred to an appropriate service provider. Among the women involved in the qualitative research, one woman was referred to a service provider when she requested additional support following a discussion with the research team; she reported that the process of discussion helped her to reflect on her situation, and she wished to seek further counselling.

Data analysis

Data were analysed inductively (Braun and Clarke 2006; Silverman 2001) by the first two authors, with a focus on women’s key experiences, beliefs and goals. IDI transcripts were read repeatedly to inform the identification of themes, and the first two authors met regularly to reflect on the emerging analysis. Transcripts and photo-posters from the photovoice sessions, and participant observation notes were then reviewed and supplemented the initial analysis. As a result of the inductive coding, a series of initial themes were identified centred on love-relationships, staying in relationships and IPV. A second level of coding then generated sub-themes and connections between these themes, which collectively form the key narrative for this paper; the sub-themes were selected for frequency and the relative importance that the women placed on them.

Several strategies were employed to ensure rigour in data collection and analysis. Data triangulation through the IDIs, photovoice and participant observation enabled comparisons and differences between themes to be identified using different data sources. The first two authors also met regularly to reflect on the emerging themes and analysis, important for reducing bias and deepening the analysis. Altogether, almost one year was spent in the field, allowing an in-depth contextual understanding of the communities. Carefully designed question guides allowed space for reflection and sought to avoid leading responses. Finally, reflexivity was used to check on the integrity of the data collection and to avoid and correct for, among other things, social desirability effects.

Ethical approval

Ethical approval for the study was received from the South African Medical Research Council and from the University of KwaZulu-Natal. Participants signed informed consent forms, for both the IDIs and the photovoice sessions.

Findings

Women in this study experienced frequent violence in their love-relationships; however, their constructions of love and love-relationships were decisive in informing whether they left or remained in them. They identified four clear behaviours that they believed indicated love and respect, and when these were present many exercised their agency and chose to stay, irrespective of whether or not they experienced violence. The first behaviour was the appearance of fidelity maintained by discretion; while infidelity was tolerated by the women, discretion was essential. Second, partners needed to provide financially for their partner and child(ren). Third, partners were expected to openly acknowledge the relationship and spend time with his partner. Finally, women did not want to be humiliated or shamed publicly; this included through public fighting, visible injuries and indiscreet affairs.

Experiences of IPV

All but three women in this study discussed multiple experiences of physical, emotional, sexual and/or economic violence within their love-relationships. Significantly, less IPV was reported in short-term, casual relationships. Most of the violence described by the women involved being hit, slapped and shouted at; these incidents often took place in the context of the male partner’s jealousy or drunkenness. In four cases, women were locked in their partner’s shack alone, either overnight or for many hours. Many of the women reported experiencing emotional violence, either alongside other forms of violence or as the primary form of violence. Emotional IPV included frequent shouting, the use of abusive language and being ignored for long periods. In at least two cases, the violence began or escalated during pregnancy. Women occasionally spoke with friends or close family members about the IPV they had experienced, but more commonly it was seldom spoken about and never reported to police or service providers.

Constructions of love

Findings show that women’s views about of love significantly shaped their willingness to tolerate violence in love-relationships, and their willingness, or reluctance, to leave a violent relationship. For many, remaining in a violent relationship was an agentic choice. These constructions were informed by gender norms and inequalities, and related expectations placed on women. While most of the women would have preferred not to experience violence, they tolerated or downplayed the violence they experienced. Many did not view being shouted at, slapped or locked up as violence, or saw it as relatively minor violence.

Following an argument, Langa’s partner had locked her up alone for eight hours in his single room shack while he went out with friends. During the photovoice discussion she commented that although she did not like what occurred it was not bad enough to end the relationship over. She, like others, did not view these events as a significant transgression of acceptable behaviour within a love-relationship.

Even in cases where the women found the violence to be inappropriate, it was often tolerated within a love-relationship, seldom leading to outrage or decisions to leave. This was illustrated in photovoice discussion sessions where talk about the violence women experienced was often light-hearted. While the women expressed not liking the violence, their talk seldom included outrage or condemnation of the men’s behaviour.

During a photovoice discussion, Noluvuyo described how late one night, when drunk, she had demanded money from her partner for their baby and how she had been beaten:

Noluvuyo: And he tells me ‘you are undermining me, get out of here’, I say ‘okay I am leaving, you are being selfish’ … I left. But I forgot my phone, so I called out to him ‘baby please bring me my phone’, he told me ‘you are crazy’, I told him ‘Sabelo I will kick down this door’, he said ‘okay fine kick it down’. I told him okay, I will count to 3 and so I began counting and he was laughing, I took the risk and kicked down his door. He got up from his bed. I stood there looking at him and I thought if I run now, he will give me the biggest slap that would scratch my face. He got up and dragged me by my feet …

Zoleka: *Laughing aloud throughout the narration*

Sthelo: Oh my God, he is so skinny, and he was dragging you, you are like a sack of I don’t know what.

Interviewer: He dragged you and then?

Noluvuyo: He beat me up. He hit me, but it was fine because I was laughing as he was hitting me. (*laughing as she speaks*)

Noluvuyo’s amused re-telling of the incident and the laughter it generated may well reflect her and the other young women’s coping mechanisms. The use of laughter in such a context may also be an attempt to minimise the bad situation when among friends, as well as being a signal of anxiety and distress from the other women listening to the story.

Women ‘preferred’ certain forms of violence, which they viewed as less significant than others. A number said they would rather be hit then shouted at, as shouting may go on for a while, whereas a slap is over quickly. When discussing how she and her partner resolved conflict, Thobile, who had never been hit but whose partner frequently shouted, said she found his shouting frustrating and never-ending, saying she would have preferred to be slapped, believing this would have ended the conflict quicker:

No man, I just don’t like a person who shouts rather slap me instead of shouting at me … A person who hits you is better because they let it go quickly. A person who shouts doesn’t let it go.

In addition to women not seeing the violence they experienced as significant, in some cases, these ‘minor incidents’ were understood to be a sign of, or driven by, love. During participant observation the interviewer was talking with Sthelo and her mother, who explained that Sthelo’s partner was only possessive, jealous and violent towards her because he loved her so much.

Findings highlighted how violence and love were intertwined for young women, with many talking about violent partners being ‘good partners’ and tolerating violence as an inevitable part of a love-relationship. Enhle, when describing a ‘good boyfriend’, said that her ex-partner who beat her was, aside from being violent, a good partner especially as he provided for their child financially. ‘One person I can say was a good boyfriend in my life was my ex (baby’s father) but the problem is that he used to hit me … but he was a good boyfriend’.

The severity of the violence inflicted was also important in determining its acceptability. Ntombi noted that a partner who loves you can hit you, but not too severely:

Interviewer: So, if he loves you, he can hit you, but not injure you?

Ntombi: No, he can hit me, maybe just one slap.

These constructions of love influenced women’s agentic decisions to stay or leave violent relationships, with women often choosing to remain if they felt loved and respected. Nonetheless, several women did leave their love-relationships after a long period: five left following her partner’s indiscreet affairs; two left when her partner stopped supporting the child(ren); and three left when the physical violence became publicly humiliating. However, these actions were fluid, and a number of the women ended and then re-started their relationships later. Furthermore, two women who ended their relationships towards the end of the interview period mentioned considering returning to the relationships if circumstances improved.

Discreet infidelity

Discretion concerning concurrent relationships, by both the man and his other partner(s), was perceived as an indication that he loved and respected his primary partner. Zoleka, who had a close and supportive family and had been with her partner for some years, believed that while multiple relationships were wrong, they had to be tolerated as inevitable. However, like many of the women she had expectations around how they should occur:

Zoleka: … you can’t prevent that because men do that, they always have girlfriends, always looking for girlfriends, you can’t change that. It’s just something you have to understand, and it’s fine as long as your partner respects you.

Interviewer: What does that mean, ‘as long as he respects you’?

Zoleka: Let’s say your boyfriend has a girlfriend, he has to make sure that he covers his tracks, he has to make sure that you don’t see or hear anything about it … Like with phones, he must delete things when they are done talking or texting.

Interviewer: So, a guy who respects you in that way is a good boyfriend?

Zoleka: Yes, the respect is important.

Zoleka finally ended this relationship when the affair her boyfriend was having became public and humiliating for her.

Sthelo’s relationship became violent when, after a few years together, her partner began hitting her and locking her in his shack. However, she only ended it several years later when he began openly dating her neighbour. Importantly, on this occasion Sthelo had the emotional and economic support of her mother, who she lived and shared confidences with. This support played a significant role in enabling her to leave. For her, the public humiliation and loss of dignity when her partner publicly dated her neighbour was unacceptable. This indiscretion was a clear indication he no longer loved her:

I just realised that this person does not love me, he is using me, you can’t say you love me but come and date another girl in my neighbourhood. Him dating someone around here is a sign that he didn’t love me, because how could he do that …

Provider-love and poverty

The second reason to leave was ‘provider-love’ and expectations that love should be demonstrated, in part at least, by a man spending money on his partner and child(ren). Most of the women had expectations that their partner should provide in some way including money for food, cosmetics, hair-care or mobile phone airtime. This was not necessarily linked to economic need but rather was a sign of love and respect. When asked what makes a good partner Ntombi noted:

He has to give me money. A man must ask whether I still have cosmetics, and give me money to do my hair, that is a good man. I shouldn’t be the one who is asking him for money, no. He must just give money, yes that is a good man and we could have a good relationship.

Enhle had been involved with the father of her daughter for several years and had experienced a lot of violence from him. Yet she stayed with him because he was a loving father and supported their baby. After he had a child with another woman, however, he stopped supporting or seeing Enhle’s child: ‘He loved his [new] child. Since the new baby-mama things, he didn’t care much for my child’. Enhle had initially tolerated the violence, but when her partner stopped supporting her and seeing their baby, she left him. However, she concluded the interview by reflecting that she might consider returning to him, since he has now left the mother of his second child.

Three women stayed with their violent partners because of poverty and a lack of alternative ways to survive. In Ndoni’s case, she and her siblings were very poor, and she had no family support. She remained with the father of her three children for several years despite frequent violence, because he occasionally supported her and the children. However, he did not provide consistent support, and she eventually left when he stopped supporting their children regularly:

Interviewer: Would you have stayed if he beat you but helped you and looked after you and the kids?

Ndoni: I would have, I would have stayed. I am a very patient person and I can persevere through relationships, most of my relationships are long-term.

Noluvuyo, who had been with her partner for four years, was unhappy and had attempted suicide twice prior to the time we spoke to her. She tolerated her partner’s frequent shouting and hitting because he supported their child financially. She had no other source of support and a poor relationship with her family.

Sebenzile had been with her partner for 15 years; they had one child and shared a very traditional patriarchal relationship. She reported that while her partner never hit her, he was often verbally abusive. She stayed with him because she wanted her son to have an involved father and felt that if she left him, he would no longer support his child.

If I were to leave, my child would not have a father, he won’t have us as a family together … But if we stay together as we are now, it’s nice … it’s nice because he (the father) also gets to see what the child needs. Men don’t look after children if they don’t live with them.

Publicly acknowledging the relationship

Women’s expectations that a loving partner should openly acknowledge their relationship emerged as the third theme. Violence could be tolerated if the woman received the social status that came with having a visible, loving partner. Thembeka reflected on the importance of her partner spending time with her.

Interviewer: What are some of the things that he can do that would make you happy?

Thembeka: I like it when he spends time with me. And is understanding, he must understand you and do things that make you happy and feel loved and welcomed. But the most important thing is time.

Enhle too stressed the importance of being publicly recognised as the girlfriend and felt strongly that her ex-partner had been better than Zodwa’s (another woman in the study), because while both their partners hit them, her partner publicly acknowledged her as his girlfriend, showing he respected and loved her, while Zodwa’s did not.

Enhle: At least Andile acknowledged me publicly as his girlfriend. Zodwa’s boyfriend hasn’t acknowledged her, she is hidden but he still beats her. It doesn’t look right. Andile also showed me that he loved me, and he told people about me, he told my friends that he loved me.

Interviewer: That is the difference?

Enhle: Yes. Because Zodwa’s man beats her and nobody knows that she is his girlfriend. I think it’s better if he beats you but everyone else knows that you are together publicly.

Public humiliation and shame

A fourth theme concerned public humiliation and shame. Women believed that a loving partner should not publicly humiliate and shame them. To avoid shame, women attempted to keep any violence they experienced private, and were more accepting of it if it was private. For example, they did not want to be hit or shouted at in public, nor did they want visible injuries.

When Noluvuyo challenged her partner late one night she realised he would hit her. However, she did not leave as she thought that if she tried to run, he ‘would scratch my face’. Instead, she remained at his shack despite the likelihood of violence, to avoid visible injuries. Two other women, who described being locked in their partner’s room alone for hours, reported they did not try to escape or call for help because of the public humiliation of other people knowing they had been locked up. When Zoleka, who was shut up in her partner’s shack while he went drinking following an argument, was asked during the photovoice discussion why she did not climb out the window Noluvuyo quickly replied: ‘imagine if she [Zoleka] tried escaping via the window and bumped into his mother’. There was a very strong sense that such events needed to remain private to avoid humiliation.

Two women ended their love-relationships when the violence they experienced became public. Olwethu ended her frequently violent relationship, after four years, following a particularly explosive fight in the street. Similarly, Ntombi was regularly abused for years by the father of her child and experienced frequent hitting, kicking and hair-pulling. She accepted this violence as she believed he loved her. However, when the violence escalated and became public, showing that he no longer respected nor loved her, she left him.

I wasn’t seeing the love any more. He would hit me all the time. How can you say you love me when all you do is give me blue eyes? You say you love me but, you give me blue eyes. You don’t love me, you are making me a fool and humiliating me in front of people. Because if you love someone slap them and then leave them alone. He would make sure that I was injured and had blue eyes, or my eye couldn’t open. I thought no this man does not love me, he is making me a fool.

Family support

Family support, both emotional and economic, proved to be an important consideration when women were assessing their options for leaving violent relationships. Both Ndoni and Noluvuyo referred to the fact that they could not turn to their family for economic support, which limited their options and agency. Conversely, both Zoleka and Sthelo retained very close connections with their families, and both referred to these relationships as factors that supported them to leave when they felt publicly shamed.

Discussion

In contexts of poverty and patriarchy such as this setting, women have limited options to gain respect and dignity; one of the few options available to them is through love-relationships. While these relationships provided emotional and economic benefits, they often included violence. This violence was frequently accepted and seen as a trade-off for being loved and respected. That said, to date the role of love and respect in women’s agency and decision-making, especially concerning whether to remain in a violent relationship, has been inadequately explored in the literature (Mannell, Jackson and Umutoni 2015; Campbell and Mannell 2016).

The narratives elicited in this study show that the young women viewed respect, economic support and in some cases even violence as expressions of love. But when love was no longer expressed in these meaningful (and useful) ways, women exercised their agency and left. Our findings support the growing body of work exploring the complexity of women’s agency within contexts of IPV (Jewkes and Morrell 2012; Madhok, Phillips and Wilson 2013; Campbell and Mannell 2016) and contribute a critical perspective which emphasises how women’s constructions of love may inform their agentic decisions.

In contrast to scholars who argue that women only display agency at the point of leaving a relationship (Scheffer Lindgren and Renck 2008), we concur with those who argue that agency, especially in the context of love-relationships, may be fluid, contested and enacted in many ways (Jewkes and Morrell 2012; Mannell, Jackson and Umutoni 2015). Nearly all the women in this study displayed agentic behaviours both when they made choices to stay in violent love-relationships and when they left them. Despite many of the women not liking the violence, they chose to stay so long as the relationship met their desire for love, dignity and respect, leaving only when it stopped doing so.

Meanings and constructions of love and the desire to be loved and respected played a critical role in whether women remain in relationships, violent or not, supporting work by Wood, Maforah and Jewkes (1998), Hunter (2010) and Jewkes and Morrell (2012). While structural factors such as poverty, unemployment and limited education constrain women’s choices and increase their dependence on their partners, desire for love and love-relationships were key factors determining staying or leaving. These constructions of love-relationships, as Hunter (2010) and Jewkes and Morrell (2012) have noted, are bound up in both romance and economics, informed by the structural realities of patriarchy, harmful social norms and poverty, with most women seeking both ‘romance-love’ and ‘provider-love’ in their relationships.

Women in this study identified four factors that indicated love and respect: discretion around male infidelity, ‘provider-love’, public acknowledgement of the relationship and not being publicly humiliated or shamed. These factors informed women’s ‘goals and actions’ and agentic decision-making. When any of these ‘markers of love’ were violated, the women believed they were no longer loved. With few exceptions, at this point the love-relationship ended. However, despite leaving, ex-partners sometimes remained on the side, with some of the women speaking of possible reconciliations alluding to the ongoing complexity of such relationships even after a moment of closure.

Emotional and economic support from the natal family and an independent income influenced four women’s decisions about whether they remained in a violent relationship, highlighting how emotional support from the family is important in enabling women to make critical choices. In two of these cases, emotional family support rather than money enabled women to leave their partners when they no longer felt loved. Conversely, two women who had poor relationships with their families could not turn to them for emotional or economic support, which limited their options to leave.

Women’s agentic behaviour within violent love-relationships is better understood from a deeper understanding of women’s individual goals and why they hold these, and individual perspectives concerning violence, love and respect. Nonetheless, as Kabeer (1999) has argued, the process of identifying goals is in and of itself constrained by women’s lack of power and agency, which is in turn linked to structural barriers. More research is required to deepen our understanding of the interplay between these different factors and hence inform effective violence prevention responses.

Limitations

There were of course limitations to this study. While we worked with participants for almost a year, this was a relatively short period over which to build meaningful relationships with the women. A longer-term relationship might have allowed for a richer understanding and a more nuanced interpretation of the data. Despite this, we drew on a range of different data sources, each of which provided insight into women’s lives, enabling a deeper interpretation of the data and thereby increasing the validity of the findings. Convenience sampling, however, posed limitations on the generalisability of the study findings.

Conclusion

Our analysis sheds light on how constructions of love, IPV and agency intersect. Findings suggest that women may be better placed to reflect upon and resist IPV if they are involved in processes that involve reflection on what it means to be a woman in a love-relationship in a context of heightened poverty. Reflecting also on locally prescribed forms of masculinity and femininity and their implications for the form that sexual relationships take would also be helpful, as would be reflection on how best to break the ties of economic dependence on men. Such individual-level interventions need to be complemented by initiatives to transform the broader patriarchal and socio-economic context.

Women’s narratives in this study support the growing body of research calling for a more nuanced understanding of agency that recognises that women who remain in violent love-relationships have often made agentic decisions to do so. Such behaviour questions the dominant orthodoxy that women only show agency in the context of violent relationships when they leave a violent partner or report him to the police. Women’s agency and decisions to remain in, or leave, violent love-relationships are complex and informed by many factors including, as this study has shown, constructions and expectations of love-relationships, the desire for respect and dignity, as well as poverty and patriarchy.

Declaration of interest

The authors declared no competing interests exist.

Acknowledgements

Thanks go to the participants for their time and input as well as to the larger study team.

Additional information

Funding

The study was funded by the UK Government’s Department for International Development through the ‘What Works to Prevent Violence? A Global Programme to Prevent Violence against Women and Girls’ research programme managed by the South African Medical Research Council.

Notes

1 This participatory technique uses photographs and associated narratives to represent women’s lived experience, and to imagine futures (Wang and Burris 1997). Harley (2012) and others have noted that the technique is useful in empowering research participants and generating critical self-reflection, both individually while planning and taking the photographs, and later through facilitated group discussion.