Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Soc Sci Med. Author manuscript; available in PMC Jun 1, 2013.
Published in final edited form as:
PMCID: PMC3217103

Sexuality and the limits of agency among South African teenage women: theorising femininities and their connections to HIV risk practices


In South Africa, both HIV and gender-based violence are highly prevalent. Gender inequalities give men considerable relational power over young women, particularly in circumstances of poverty and where sex is materially rewarded. Young women are often described as victims of men, but this inadequately explains women’s observed sexual agency. This paper takes a different approach. We use qualitative interviews and ethnographic observation among 16 young women from the rural Eastern Cape to explore ways young women construct their femininities and exercise agency. The data were collected as part of an evaluation of Stepping Stones, which is a participatory behavioural intervention for HIV prevention that seeks to be gender transformative. Agency was most notable in particular stages of the dating ‘game’, especially relationship initiation. Constructions of desirable men differed but generally reflected a wish to avoid violence, and a search for mutual respect, sexual pleasure, romance, modernity, status and money. Agency was constrained once relationships were consented to, as men expected to control their partners, using violent and non-violent methods. Women knew this and many accepted this treatment, although often expressing ambivalence. Many of the women expressed highly acquiescent femininities, with power surrendered to men, as a ‘choice’ that made their lives in cultural terms more meaningful. In marked contrast to this was a ‘modern’ femininity, centred around a desire to be ‘free’. A visible third position, notably emerging after the Stepping Stones intervention, rested not on a feminist challenge to patriarchy, but on an accommodation with men’s power whilst seeking to negotiate greater respect and non-violence within relations with men. These multiple and dynamic femininities open up possibilities for change. They demonstrate the need to engage with women, both as victims of patriarchy and active supporters of the gender order. The multiplicity of women’s hopes and desires and circumstances of emotional and relational fulfilment provides potential for interventions with women that acknowledge existing gender inequalities, validate women’s agency, reduce violence and prevent HIV.

Keywords: sexuality, femininity, gender, South Africa, sexual agency, gender inequality, intimate partner violence, HIV prevention


An abiding concern in Sub-Saharan Africa is the unequal impact of HIV on young women. Studies have highlighted age-differentials in young women’s relationships and the problems of coerced and transactional sex, and observed that young women are unable to negotiate safe sex because of the way gender inequality plays out in the realm of intimacy (Jewkes & Morrell 2010; Campbell, Baty, Ghandour, Stockman, Francisco & Wagman 2008). In understanding the vulnerability of women, important theoretical work has explored gender relations.

In South Africa, utilizing Raewyn Connell’s theories (1987, 1995) about gender and power and specifically her concept of hegemonic masculinity, scholars have shown how particular understandings of masculinity legitimate unequal and often violent relationships with women (e.g. Wood & Jewkes 2001; Jewkes, Dunkle, Koss, Levin, Nduna & Jama, 2006). Hegemony connotes the existence of agreed values and practices, not necessarily associated with repression and violence. It can operate differentially at a global, regional or local level (Connell & Messerschmidt 2005). The concept of hegemonic masculinity has been used to make sense of the ideals, values and practices of men. A corresponding literature on how women are involved in shaping the contours of heterosexual relationships, and their acquiescence or resistance to prevailing gendered power relations, is less well understood. The assumption, whether accepted or problematised, is primarily that they are on the receiving end of patriarchal power and almost defenceless when it comes to negotiating heterosexual relations (Campbell, Baty, Ghandour et al 2008; Gavey 2005).

There is a broad international literature on femininity(ies), crossing a number of academic disciplines, which has been differentially helpful in contributing to our understanding of women’s roles and gendered positions in heterosexual relationships. Femininity has been described by Holland as an ‘elusive’ concept, seen variously as a normative order (i.e. a set of psychological traits such as being nurturing), a performance or a process of interaction (Holland 2004:8). The plurality of femininities has been well illustrated by discussion of lesbian femininities, emerging youth femininities, such as Anita Harris’s ‘future girl’ with her desire and determination to take charge of her life, Holland’s work on alternative femininities of women with tattoos and body piercing, and Gavey’s emphasis on cultural location of femininities (Gavey 2005, Harris 2004, Holland 2004, Jeffreys 1996, Butler 1990).

Most analyses of femininity share with Connell’s (1987, 1995) framework an emphasis on the social construction of gender identities but have not explicitly sought to build on her attempts to link multiple gender identities with hierarchies of power. Connell has predominantly written about men and masculinities, yet because of the relational nature of gender, this theoretical work is critically important in understanding women and femininities. Indeed, much of her analysis of the construction of men’s gendered identities is clearly relevant for the study of femininities. For example, she describes the existence of multiple, mutable masculine positions and identities, arranged hierarchically with respect to each other (some are viewed as legitimate whilst others are censored), and superior to women. She argues that these are reproduced against a backdrop of gendered histories, culture (including gendered value systems (ideals)) and material circumstances. She asserts that these are important in the overall framing of masculine behaviours and attitudes without determining or prescribing these, an argument that is echoed by theorists of femininity who draw on Foucault, such as Gavey (2005) or Renold (2005). Connell argues for the existence in a particular setting of a hegemonic masculinity, which is a cultural ideal of manhood that gains its legitimacy from acceptance that is shared between those who embody and benefit from the ideal and those subordinated through it (after Gramsci, 1971). In many contexts, being dominant and in control of women are important aspects of hegemonic masculinity, and violent behaviours are justified in pursuit and demonstration of this (e.g. Wood & Jewkes 2001). Numerous authors in South Africa have argued that it is precisely such a configuration of masculinity that underpins the high levels of violence against women and resulted in, for example, 42% of men disclosing perpetration of intimate partner violence and 28% disclosing rape (Jewkes, Sikweyiya, Morrell & Dunkle 2009; Abrahams, Jewkes, Laubscher & Hoffman, 2006; Morrell, 2001).

Acknowledging a multiplicity of femininities, many authors refer to a culturally dominant ‘traditional’ femininity, such as that evocatively described by Holland’s informants as ‘fluffy femininity’ (Holland 2004; Harris 2004, Gavey 2005). Connell describes this as an ‘emphasized femininity‘ (Connell, 1987), and importantly asserts that it is characterized by compliance with women’s subordination and accommodation of the interests and desires of men. Women who endorse, or enact, an emphasized femininity are complicit with the unequal structuring of gender relations and at least tacitly accept their subordination. As with masculinity, there are social rewards and sanctions which flow from adoption of particular femininities, enforced through local moral discourses as well as the threat and use of violence, which constrain the element of ‘choice’ for women.

Within health literature, epidemiology and qualitative research has tended to treat women as a homogeneous group and as victims of (all) men. In part this is an effect of methodologies that deploy the category ‘woman’ or ‘female’ in a blunt and uncritical manner. Use in this way prevents reflection on ‘what women want’, ‘how do they try to get it’ and how these influence ‘what they get’. Notably this obscures the impact of these factors on their relationships with men and their exposure to HIV risk. Yet there is an earlier literature which sought to identify and analyse women’s sexual agency and emotional power but this literature has not found its way into the mainstream of analysis of HIV risk (Vance, 1989). Carol Vance’s landmark volume examined the contradictory location of women’s sexuality and how its expression in pursuit of pleasure in contexts of danger resulted in women being placed at risk by the very actions they took to affirm femininity.

Women negotiate their sexuality under conditions of patriarchal inequality but are not simply passive, even when young (Campbell, 2000; Wojcecki & Malala 2001). This is most visible in the literature on transactional sex, which shows some women having multiple sexual partners as a way of accessing services, goods and cash as well as prestige, in a context where having ‘rich’ boyfriends is the mark of success (Hunter 2002; Leclerc-Madlala, 2004). Other research shows young women as able to express themselves in heterosexual relations and as actively positioned with respect to condom use, both in facilitating and opposing their use (O’Sullivan, Harrison, Morrell, Monroe-Wise & Kubeka 2006; Campbell 2000), and similarly positioned with respect to contraception and the pursuit of motherhood (e.g. Wood & Jewkes 2006).

In pointing to women’s agency in South Africa we do not underestimate the constraints they experience, or the risks. Although there are diverse messages positioning women and girls sexually, there are strong cultural roots to messages girls receive, that they should be passive, innocent and will be held accountable for how they are treated by men (Bhana, 2007; Jewkes, Penn-Kekana & Rose-Junius 2005). Rape is widespread in South Africa, including forced first sex, and attempts to negotiate the terms of sexual intercourse are often understood by men as defiance or a lack of respect and result in assault (Varga, 2004; Wood & Jewkes, 1998). Poverty is widespread and many women are locked into relations of dependency on violent men because of poverty (Campbell 2000). Yet it is important that we engage in a more complex theoretical exploration of women, gender relations and their agency in constructing femininities if we are to understand, and as a public health project, influence the shape of, their sexual practices.

This paper draws on the narratives of sexually-active teenage women from a very poor area of South Africa and explores their agency in relationships. We attempt to steer a course between a focus on the individual and the need to generalize about young women. Acknowledging the gendered complexity of identity, we identify common traits, patterns or processes that enable analysis that reaches beyond the diversity of individuals. We note how values might be aspirational rather than lived, but pay close attention to the way in which the young women negotiate the values that they encounter in their particular social settings. In so doing we seek to acknowledge and make social sense of the multiplicity and fluidity of identities, and contradictions between aspirations and actions (c.f. Wetherell 1996).


The research was conducted in the rural Eastern Cape Province, an area with high levels of poverty and unemployment (over 50% among women). The participants lived in three locations, a deep rural village, an emerging town and the major town of Mthatha. The table provides a sketch of the three sites and some details of the sixteen research participants. The women interviewed were all participants in the evaluation of the HIV prevention behavioural intervention Stepping Stones (Jewkes, Nduna & Jama, 2002; Jewkes, Nduna, Levin, Jama, Dunkle, Puren et al 2008), which included a qualitative component (see Jewkes, Wood & Duvvury 2010). This was an intervention that spanned about 50 hours and used participatory methods to engage participants in a process of learning, predominantly using critical reflection. Gender was a key theme running through out Stepping Stones. The participants were chosen purposively as a group who showed diversity in sexual and social experience and home backgrounds, from among a wider group of the research trial participants at three schools who had volunteered to take part in the qualitative study. Our research strategy involved forming a detailed acquaintance initially with participants and following them over time.

Unless otherwise stated, this paper is based on an analysis of data collected before the Stepping Stones intervention. At this time, in 2003, two young, female Xhosa-speaking interviewers formed an acquaintance with the participants and their families, before conducting formal taped in-depth interviews. In the village, on the suggestion of the local community, one interviewer spent a week living in the home of one of the participants and generally spending time with all five of them whilst collecting data and recording it in notes. In the two urban settings, the interviewer also spent time with participants chatting and visiting their homes. After establishing an acquaintance and discussing many aspects of their lives informally, between 1-2 taped individual in-depth interviews of about an hour were conducted with each participant, in these they spoke of their homes, friends, boyfriends and other aspects of their lives. The interviews flowed out of the relationships that developed between the interviewers and interviewed who were little separated by age, language, culture or experience. They were thus very candid and intimate. The interviews were not considered as isolated from the broader ethnographic milieu and so there was no reason to expect the interviewees to have engaged in dissimulation or concealment. A total of 17 interviews were conducted and together with extensive field notes constitute the data base of this article.

Nine to twelve months after the initial interview (in 2004), the interviewers returned to visit the participants in the village and Mthatha, all of whom had then had the Stepping Stones intervention. Acquaintance was renewed, informal interviews and participant observation continued, and one taped interview per participant was conducted (giving 10 further in-depth interviews). This was 5-10 months after the end of the Stepping Stones workshops and interviews chiefly asked about Stepping Stones (findings described in Jewkes et al 2010), their lives and relationships. We refer to the findings of these interviews towards the end of this paper.

All participants signed informed consent and were assured confidentiality. The study had approval from the University of Pretoria research ethics committee. Interviews were taped, transcribed and translated from isiXhosa. The data were coded by both authors and analysed using content analysis and analytic induction. Both authors have a 30 year acquaintance with the area in which the study was located, and they drew on this in data analysis. Through combining data from interviews, in-depth acquaintance over some months and observations, we assembled, what Connell describes as “rich evidence about impersonal and collective processes as well as about subjectivity” (1995 p.89).

What did the women want?

All the women interviewed currently had one or more boyfriends, but these men occupied different positions in their lives. For many, obtaining and keeping boyfriends was a central pursuit. Boyfriends were their main preoccupation, and although they attended school, helped at home and saw their girlfriends, these did not capture their imagination, engage their full attention, nor were they the imagined vehicles for realising dreams. A couple of the women presented a notable contrast to this. Sandi was a very successful school athlete, and Pinky pronouncedly positioned herself as a dutiful daughter. Pinky had moved to town in order to cook and clean for her grandparents, aunt and cousin, duties she took very seriously. For these women, studying was woven as a major thread into their future aspirations, boyfriends made them feel womanly, but were somewhat peripheral. For two women (Beauty and Xoliswa) boyfriends were central to their lives, but not in a carefree way. They were cohabiting (or equivalent), financially dependent on older men, and their lives were dominated and constrained by their position as partnered women.

Most women still lived with their families and dating was their main source of entertainment and excitement. For the great majority, desirability to men was central to their constructions of successful womanhood (as they are to dominant constructions of young African manhood c.f. Wood & Jewkes 2001) and performing, testing and affirming this were key preoccupations. Flirting was fun, an entertainment in a context of limited alternatives. Having partners was a demonstration of desirability, attractiveness and critical for assessments of feminine ‘success’. The local reality was that these were usually sexual partners, but sex itself was not universally described as essential for feeling feminine. Indeed relating to men did not always involve sex, or did so only occasionally, but some women had sex frequently, with one or more men.

Desirable men

In discussing a desirable partner, age, a desire to avoid violence, and find mutual respect, sexual performance, romance, modernity and money, intersected, and often competed. Women’s constructions of ‘desirable men’ were a reflective embodiment of their ideas of what made them feel feminine and having good sex was important here. The status and looks of a male partner reflected on the woman herself. In partner choice, women had considerable agency. Generally women did not initiate relationships, although we were told that ‘other’ (modern) women in Mthatha would do so, but they acted to invite proposals. They often took considerable care over deciding who they would or would not ‘admit’. Both dreams and reputations were at stake. A quick decision might make them look too easy and, since women did not want to end relationships too often for fear of getting a bad reputation, they had to make sure they chose well, particularly avoiding violent men. Protecting reputation also involved not having too many partners from one location or else, as Sandi cautioned, “everybody from that area will know how I am in bed”.

Age was a consideration. Generally the women dated men who were older, mostly only by 2-3 years, but three had partners 10-15 years older. Some women explained that older men would make less desirable partners as they would beat, insist on sex and be controlling. On the other hand others felt they were more respectful and sexually experienced whereas teenagers could be unpredictable and dangerous. Teenagers might leave them after sex and were violent - “boys rape” and might ‘streamline’ them (gang rape, described as where boyfriends “invite their friends and take you by force”) (Wood 2005).

Another consideration was a desire for mutual respect. Women were concerned that they should be able to respect their partner and (appropriately) submit to him in order to feel good and feminine. It gave them a sense of dignity, and many indicated that they wanted to be respected in return. In order to get a man they could respect, age was again a concern. As Thobeka explained:

it is important not to have a relationship with someone your same age, because there is no dignity and you will disregard him. If he says no, you just ignore that because he is a child”

Sexual performance was very important, as most women wanted good sex. There was a considerable amount of discussion of sexual performance and desire in the interviews, reflecting a traditional openness about sex (c.f. Jewkes, Penn-Kekana & Rose-Junius 2005). Again related to age, Nokuzola (17) explained:

N: If you were used to sleep with the young boy, when you have met someone who is older by three to four years, you just know it yourself that ‘No! Now I have met an adult’.

Interviewer: You said there is a difference between a circumcised man and a boy, so I want to know that difference. Some other people usually say the difference is in size…

N: I meant that.

Other women feared that a larger penis could damage them. Popular discourses of vaginal damage, ‘coldness’ and wetness were used to dissuade young women from favouring older partners, with some success.

Women also wanted romance, to be wooed, and this was often equated with modernity and a sense of style. Here ‘old fashioned porridge’ was pitted against ‘cheese and polony’ (a cheap processed meat). Modern men who dressed well and used cell phones were favoured. They could send text messages (SMS) in English (rather than isiXhosa), and were less controlling. Nokuzola explained that she was swept off her feet by the modern cell phone seduction of her new partner (Slu):

N: He used to send me some messages in the morning, during the day and when I was about to sleep, he sent me some nice messages, hey I was also swept away and I admitted him.

I: What kind of messages for instance?

N: Maybe he says have a nice day, you see, hey he would send me another one late, he would write them and tell me that I should have some nice dreams and think about him because he is also thinking about me, things like that, you see.

This contrasted with her main partner (Lungi), a ‘traditional person’, manifest in his controlling behaviour. This Nokuzola recognised, protested about, but did not resist, his jealousy being evidence of passion:

Slu at least, if I am walking together with Slu and meet with my friends he waits for me to chat with them in the meantime stands there at the side while Lungi even drags me and says ‘hey lets go, lets go, I want to take you home’, always he wants to take me home, he doesn’t want to see me in town, maybe just hanging around town, and I don’t have to be spotted by him chatting with my friends or do something.

Money and status were also important. Discussing why ‘all the women’ in Mthatha lusted after a teacher at her school, Lindiwe (15) reflected that his style and money (coyly hinted at, but not mentioned) were important. She was similarly reluctant to point to the other obvious conclusion, that women perceive their status (or fantasies of their status) to reflect their partner’s power (c.f. ‘incorporated wives’ (Callan & Ardener 1984)):

all I know is that he is in demand, maybe it is because he has a two-door BMW and a white Toyota which is ND [registration number from another Province] maybe that is why he is in demand, but I do not think that is it .. I think [the car] is not something special, in fact what I see in him, it is like the way he smiles, it is attractive really (laughing)

Other women were quite open about the role of money, and linked it to demonstration of mutual respect. Whilst Phindi (19) asserted ‘I can love him even if he has no money’ and, like other women, emphasised the importance of romance and faithfulness, she seemed to view gifts of money as an entitlement:

…I wish I would have a relationship with someone who would only be in love with me, he should not drink and smoke, or I want a person who loves me [Laughs] who will always think about me and I will think about him. I don’t like someone who has multiple partners. I wish I should just have a person whom I know that I am in love with, but I wish he should give me money when there is something that I need…I would see that this person does not love me if he gives me nothing, but he sees that I am suffering, no I could see that he does not love me. …I wish he should not spend a long time with out seeing me and if there is something I need from him knowing that he has it, he should give it to me.

Most of the women were in relationships that were described as primarily motivated by ‘love’. Yet providing gifts or favours to girlfriends was important in courtship and particularly valued when resources at home were tight. Nokuzola valued her violent, gangster boyfriend, because through him she could have things without ‘begging and begging’. Some of the women were very proud of what they were given by men. Beyond monetary value, it reflected their desirability, boyfriends’ worth and status, especially when gifts came from men who they did not have sex with, or at least not yet. Describing the courtship by her taxi-driver boyfriend, Phumla explained:

when going to school he used to say I should wait for him at the stop, and I waited for him and he gives me a ride, he would take me to town first without making me pay, I used to do that quite often

Others had more ambiguous feelings towards the money they were given. Lindiwe was receiving R150 (US$20) from a boyfriend with a regularity akin to being salaried, but she denied that she expected the money. Ntsiki explained she was cross that a man pursuing her had paid R100 (US$ 12) for a school trip as she did not want people (including him) thinking she was his girlfriend, when she had said he must wait a year for her to finish school.

Transactional sex was not just about respect and power for all women. Beauty and Xoliswa were trapped in transactional sexual relationships with abusive older men. Utter desperation had led Beauty when she was 14 into a relationship with a married man who still supported her and her child with meagre financial contributions (“he gives me the amount he thinks is good for me, but he never gave me an amount of R200 (US$28). It is usually less than that amount”).

Agency in relationships

Whilst women appeared to have considerable agency at the point of choosing partners, once the choice was made, their power was greatly circumscribed, and in many respects surrendered.

In many cases boyfriends made all decisions and this was presented as ‘the way it should be’:

Interviewer: Who decides on when to have sex or not?

Thobeka: It is my boyfriend, because I was not well informed about sexual relations.

I: And now that you know about it, who is supposed to take a lead in decision making on what to do in a relationship?

T: It is supposed to be him… he is the one who says what to do, when and how.. I can say I can’t decide on anything.

In a minority of cases there were accounts of more respectful sexual decision-making. As Nosipho explained:

…When he wants to meet with me, he sends Cebo [his older brother] to tell me …and …the time he would come. I accompany him, or if I like I go to his home. When we arrive, we chat and enjoy. If I am going to sleep, I sleep over at his place and if I will not, he accompanies me home. If I sleep over, we sleep together and have sex, and after we have finished he accompanies me home.

Women often had little ability to get more attention or sex from a partner, or obtain better sex or get condom use. This did not always stop them trying, for example Lindiwe spoke about the frequent quarrels she had with her younger boyfriend over him always insisting on being ‘on top’ and the fact that sex was too quick. Phumla spoke about her futile attempt to get her boyfriend (named Thads) who cohabited with someone else to use a condom:

‘Thads man (mfondini) let us use a condom’, he would say he won’t dare use a condom, I then said ‘Please Thads’ and he said he doesn’t want to and if I don’t want to have sex with him I should just leave it like that and I said ‘let us leave it like that’, he said ‘you are mad and I will just beat you’, yooh I realized that I won’t let myself be beaten for that.

Several of the women were ‘bored’ in their relationships, expressing exasperation over the emotional toll taken by their boyfriends’ other partners, violence and emotional coldness. Most women could end relationships, and seemed to do so in a formal sense more than men, but only in certain circumstances. When their boyfriends had other serious partners – cohabiting or a new marriage – a relationship might split. Loyiso left her partner after he made several other women pregnant. Excessive quarrelling was another reason for break up, but these did not necessarily lead to the end of relationships. Most notable was how few relationships were actually ever ended. More usually relationships fluidly drifted out of ‘being’ and retained a latent potential of subsequently being revived. Whilst it was apparent that at times women could leave relationships fairly easily, especially those with khwapheni (non-primary partners), they also spent much time fantasising about exercising their agency in ending relationships with their different men. Given that they so rarely did this, it seemed that these fantasies were a substitute for real power, perhaps also a reluctance to end relationships because of their importance for self- and peer-esteem, or possibly out of fear of violence.

A desire for monogamy

Women’s positions on sex, monogamy and partner numbers differed. Several women said they wanted relationships that were mutually monogamous, and were often quite judgemental of peers with multiple partners. Yet with unreliable boyfriends, they described being obliged to have multiple partners in order to avoid being single. A few viewed multiple partners as a form of resistance to male control, even revenge for male infidelity (Ayanda: ‘I will never stand on one leg. I am not a pole’).

Like if you have a relationship and you love your boyfriend, you don’t have that mind of accepting another boyfriend. The reason why you accept somebody else is when you see the action of your first boyfriend. If your boyfriend is doing wrong, you decide to move on so that in case he does a thing, you know that you are balanced [on two legs]. (Thobeka)

Whilst most women engaged very actively in picking up boyfriends, including when away on holiday, they did not always have sex with them. A (male imposed) discourse that ‘real love’ should be proved through sex was often mentioned, but it seemed that sex was often quite infrequent and in two interviews Loyiso and Lindiwe claimed to be currently ‘abstaining’. Women who were concerned about their reputation (not being seen as ‘loose’) could successfully draw on a gendered discourse of (a need to avoid) being ‘finished’ by too much sex (before marriage) to negotiate either infrequent sex or secondary abstinence (also discussed in Wood, Lambert & Jewkes 2007).

Patriarchy, poverty, and cultural constraints on women’s agency

The activities of flirting, partner selection and managing multiple boyfriends could suggest substantial agency and degrees of freedom that did not in fact exist. Within relationships women’s agency was highly constrained by the structural dimensions of their lives, including an overarching narrative of patriarchy, age hierarchy and the socio-economic context of severe poverty of the Eastern Cape. A strong generational element to local hierarchies gave older men particular patriarchal power (over other (younger) men and over women), and older women power over younger ones. In essence, men were valued more highly than women. This was very simply expressed by one of the Mthatha women when explaining how the neighbourhood organised to support funerals when young people died without burial societies, i.e. insurance for funeral expenses. She said: “If a girl has died we donate R2, if one of the boys passed away we donate R5.” She, like most of the women interviewed, saw this as reflecting a natural order, and thus as entirely legitimate.

The sexual economy was also shaped by the practices of mothers knowingly giving sons pocket money to spend on their girlfriends, whereas daughters got less or none, and needed a boyfriend to help to pay for cosmetics and hair dressing. Whilst levels of unemployment were high for both men and women, men had better prospects of earning than women. This further influenced the unequal basis of sexual relations with women, and also added to the power of those men (for example, taxi drivers or gangsters) who did have money.

“They defeat us with strength and speech [bayasoyisa ngamanla nangentetho]”

This quote from Thami captures the reality of relationships and also provides the reason why the agency of young women was so limited. Men expected to control women and used both violent and non-violent methods. Many women were afraid of their boyfriends even if they had not been beaten by them, reflecting awareness of the power and violent potential of men. Men quite simply expected women to obey them (‘show respect’) and were adept at arguing that their control was good for women. They had a notable repertoire of non-violent controlling methods which largely rendered beating ‘unnecessary’. Examples of these were ‘begging’ (often for sex), being ‘offended’, witheringly using ridicule (for example, responding with ‘I don’t like that money song’ when Beauty asked for child maintenance), ‘looking strongly’ at a girlfriend, all of which women found compelling ‘because I love him’ or because they were unable to otherwise assert themselves. In addition boyfriends had a notable repertoire of manipulative strategies, particularly directed at getting sex as they wanted it. For example an appeal to allow him to get “vitamins” through sex without a condom, and an assertion that “you are going to kill me” if they did not have sex.

Xoliswa, who lived with her 28 year old boyfriend and was pregnant at the time of the interview, described a complex array of manipulation and controlling behaviour and extreme violence, which she tolerated. Entrenched in a relationship, evidently having realised the goal of having a ‘desirable’ financially secure man to provide for her, she had lost her agency:

Interviewer…You said Lindani is jealous…

Xoliswa…Yes he usually kept me in his room. I was very light in complexion. Hands and feet were light as of someone who is from the hospital bed. When I went home mom said why am I so light, is my boyfriend locking me up in his room? I said yes, he does not want me to go anywhere. He expects me to stay in the room until he comes back from work.

I…Why did your boyfriend not allow you to go out when he was not around?

X…He says he loves me, he does not want me to be seen by other men. He does not want me to have friends especially friends of my age. He said it would be better to be a friend to someone who is older, not young women because they will teach me and influence me with bad things.

I…Hey your boyfriend is cruel.

X…Yes, he is a quiet, cruel guy. If he does not like something, he takes what ever is in front of him and hits me. …I do not like to stay here but I am just staying because he is taking responsibility of my schooling… and I also became pregnant. Things like that, I want him to take care of me.

This was a particularly cruel twist of fate, as she had left a relationship with a kind, loving man who also financially supported her, but was less financially secure, for this man.

When physical violence was used against them, many women and their families tolerated it, although there was less tolerance if a weapon was used and it left ‘marks or bruises’ (c.f. Wood, Lambert & Jewkes 2007)). Phumla’s (late) father had beaten her sister’s boyfriend after he stabbed her, but that was an unusual occurrence.

Trickery and some forcing into sex were sometimes explained away as a sign of love and were not reasons to end a relationship. Some women argued strongly that beating was right in some circumstances (usually of infidelity), as Sandi explains:

Interviewer…What if he beats you for a reason, how do you feel?

Sandi…I realized that I am wrong, I was supposed to take that thing [the beating]….It is me who did him wrong… it shows that he loves me when he beats me because he wants me to refrain from doing bad things and do good things.

Part of the tolerance of violence was related to the general commonness of its use in social life. Phumla spoke at length about the quickness of people around her to settle things with a fight and instances of her mother using considerable force again her and her sister at home. When she became pregnant, her friend urged her to go to her boyfriend’s home to stake her claim over him. Initially reluctant, because she was scared of being stabbed by the woman he lived with, she found instead that her boyfriend chased the other woman away “hit her with a hammer on the head. … she escaped through the window, and ran away naked”. This had no impact on the determination of either woman, as later she heard in the community “Norma [the other woman] was talking about me saying ‘No matter what I do Mike [the boyfriend] belongs to her’”

Some women did resist the control of men, but usually did not leave them or even threaten to do so. There was no evidence that women challenged patriarchy overall, and to a large extent women saw male control as legitimate and reflective of a natural order.

Conservative femininities

The femininities, understood as expressions of identity among women which combine, inconsistently and sometimes contradictorily, aspirations and practice, showed considerable diversity. Yet most women were notable for their conservatism and acquiescence. Unequal gender relations which marked domestic arrangements were accepted, as was the expectation that women should publicly be passive, obedient and respectful in relations with men. Among the conservative group were the most vulnerable women who offered no resistance to male domination, those who seemed unquestioningly to accept patriarchy as well as women who supported a patriarchy, whilst personally seeking (with variable determination and success) some respect and power in relationships with male intimates. Conservatism was often not reflected so much in rigid adherence to rules of elders and boyfriends, as in discretion and moderation in their transgression.

Being conservative and acquiescent could also bind women to support highly oppressive traditional practices. Thobeka, who proudly told us she had no girlfriends because women ‘gossip’ (a discourse used by boys to undermine girl’s solidarity and sharing of knowledge about boy’s reputations), eagerly described her care in choosing a bride for her brother to marry cruelly by ukuthwala (wife abduction). Yet it also overlapped with more modern assertions of sexual independence and a sexual agency. Getting very drunk and being sexually available to men was not seen by Thobeka as ‘untraditional’. She revelled in tales of her extreme drunken behaviour and an occasion when her intoxicated sister invited several boys who hung out at the village shop to have sex with her. She was a proudly conservative, acquiescent woman who strongly supported the knowingly-harsh, gender order.

Acquiescence to patriarchy was cross-generationally reinforced, even by more educated mothers. Nokuzola was vulnerable. She’d been pregnant at 14 and her taxi-driving, gangster partner had once organised a friend to abduct her and beat her up. Her mother, who was a teacher, had at one stage wanted the relationship to end, but that was because she thought him uncouth. He would drive into the family yard and hoot his horn for her ‘as if she was his wife’. This brash behaviour caused annoyance, but apparently not his violence and gangsterism as she accepted him when he stopped this.

Modernity and aspirations of control

In urban areas, several of the women presented themselves as women in control of their lives. This was positioned as a ‘modern’ femininity (c.f Barlow, Dong, Poiger, Ramamurthy, Thomas & Weinbaum, 2005), but was clearly part reality, part fantasy. The modern girl was presented as a young woman, wooed by men with cell phones, steering a course through the pitfalls of teenage sexuality - pregnancy, HIV, heart-break, and embarrassment caused by immature boyfriends - to complete school and have a career. In a clear rejection of the conservatism and acquiescence, this femininity was framed around replicating some of the practices and assertions of empowered men.

Lindiwe (15) described herself as a ‘free person’. She attended a social whirl of parties and inhabited a highly sexualised world with her two boyfriends and many other male friends. ‘Playing’ the field made her feel in control and feminine. She presented herself as one step ahead of her partners, as she cheated on her (high status) main boyfriend and controlled her young boyfriend, who regularly gave her money and was ‘afraid’ of her.

Yet our final interview with her showed she clearly did not fully occupy an empowered modern girl femininity that she had so strongly earlier presented and sought. A cost of presenting an invincible ‘modern girl’ identity as a 15 year old in Mthatha was seen in loss of space for expressing and responding to emotional vulnerability. She couldn’t stop her teacher boyfriend from having other women, including her cousin, so she affected an uncaring nonchalance. Instead of being a victim, she became a fellow ‘player’, giving as good as she got, lusting after men, and when a man hit her, she hit him back. But her bravado was unconvincing. She clearly cared a great deal and was wracked with jealousy and at the end of the second interview although saying she was ‘abstaining’ she admitted that she felt powerless, hurt and neglected by him and was losing weight and was pregnant.

Femininities in transition: Challenging male violence, striving for respect

A third femininity showed signs of an emerging ‘feminist’ consciousness which incorporated non-acceptability of violence and control and a demand for respect. This was most clearly articulated in the final set of interviews, and in part reflected impact of the Stepping Stones intervention, but elements of it were also visible in some initial interviews (an observation which is not entirely surprising given the national efforts to promote gender equity in South Africa since 1994, see Hassim, 2003, Morrell, Jewkes & Lindegger, in press). This femininity emerged out of, rather than being constructed in opposition to, the more conservative cultural model of gender relations.

Several women aspired to have relationships where there was mutual ‘control’ (respect) and where ‘each and everyone should listen to another’. Expressing views shared by others, Ayanda explained:

“If I say… he should obey what I say, if I say “no”, it should be “no”. “Yes” should be “yes”. I should not say I do not like something and he does it”.

Loyiso, who had been further empowered by the Stepping Stones workshop, ultimately stood her ground against her manipulative, womanising, controlling boyfriend, and before they split she told him: “what is not going to happen is for you to control me… you will only control yourself”. She eventually split up with him and was the only informant to leave an abusive partner during the research.


This paper has explored the intimate heterosexual relationships of a group of young women and described how these were shaped by their ideas, goals, and aspirations related to who they were as women, i.e. their femininities. The importance they gave to having boyfriends, their constructions of male desirability, and the quality of these relationships are critical for understanding the contours of gender power in their relationships and thus for understanding the origins of their HIV risk and risk practices.

We have observed that whilst constrained by patriarchy, poverty and limited family support, when initiating dating, women exercise agency. Women most notably exercised choice in partner selection, but this flowed from their ideas of appropriate gender relations and for the most part involved choosing a partner to whom they felt able to submit. Once in a relationship, women were caught in the powerful matrix of heterosexual masculinities that constrained their agency. Women generally did not seek to realize their goals in their relationships and often drew on ideas of the legitimacy of male superiority to help them accept their lot. In this respect they demonstrated complicity constructing forms of femininity which accept male domination (c.f. Connell 1986). Notwithstanding this, there was a diversity of femininities and evidence of their dynamic nature. Whilst some women accepted their lot, and to different degrees embraced it, others adopted a ‘modern girl’ femininity, constructed in opposition to the dominant conservative cultural model. Others, in response to Stepping Stones, showed evidence of an emerging feminist consciousness, albeit still blended with a more traditional femininity. Differences between femininities translated into differences in hopes for, and experiences of, relations with male partners, yet none of the femininities were presented as posing a substantial challenge to the prevailing gender order.

In observing these three femininities, we support the utility of Connell’s analysis in the study of women but with some qualification (Connell 1986). As with masculinities, these femininities were dynamic in nature. There was evidence of tension and competition between them, and many women perceived their conservative femininity to be both socially expected and rewarded, yet no femininity was ‘hegemonic’ because there was no single set of values endorsed by all the women. Nor was one position considered to be ideal. Each position had its own values and legitimating discourses. Critically, there was not an obvious hierarchy among the femininities which is such a key factor in analyses of masculinities. The femininities were effectively arranged laterally to one another – each having their own scope and influence that distinguished them from the other femininities. All the femininities were subordinate to the power of men. Thus none could be ‘hegemonic’, according to Gramsci’s use of hegemony (Gramsci 1971, discussed further in Hearn 2004). Women who adopted more empowered femininities did not specifically challenge the power or legitimacy of the conservatives, although they did view their positions as ‘better’ than more complicit identities (c.f Holland 2005; Harris 2004).

Much of the health literature on sexuality frames women’s practices instrumentally and often reduces young African women’s agendas to the pursuit of material reward. Yet we have shown a framework of meaning underlying these practices constructed in women’s emotional worlds. For example, when women seek older partners, they do not see themselves as relinquishing power for consumerist reward (c.f. Leclerc-Madlala 2004, Hunter 2002), but as pursuing a strategy that they hope will give them more power (of another type), respect and pleasure in the relationship than they would get with a younger boyfriend. Whilst most women wanted attention and romance, and indulged fantasies of love and companionship, they often could not get it and ended up with a compromise that was sometimes radically different. This has important implications for HIV prevention as interventions need to be based on a nuanced understanding of motivations for behaviours.

The sexual practices of young women that have been identified as critically important in their HIV risk, including having multiple and older partners, were rooted in and flowed from their gender identities, which were mostly shaped by their inherent socio-economic vulnerability, the constraints of patriarchy and their submissiveness to the controlling practices of men. Whilst Lindiwe as a modern girl was not submissive, in her quest for equity she adopted practices associated with men and masculinity, both sexually and emotionally, and in so doing denied her own emotional vulnerability. Echoing the reflections of Vance (1989), this was ultimately equally risky as being passive. The group of women who expressed a desire for mutual respect in relationships offered a glimmer of both more gender equitable relationships, ones formed around a more substantial engagement in each other’s lives, shared hopes and dreams and greater communication around sex, albeit these were most visible in interviews conducted after an intervention. Many of the women spoke positively of these relationships and they generally carried less sexual risk. Feminist consciousness for these women involved not the overthrow of the established gender order, but a selective blending of traditional and modernist ideas on gender in a way that gave them considerably more power in their day-to-day dealings within intimate relationships. This construction is highly resonant with African femininities discussed by Mikell (1997), where women sought to affirm their identities and exercise greater public power without challenging or rejecting the overall structure of corporate control of the families, lineages and localities to which they belong. In cases where women strove for more gender equitable relationships, the long term prospects of some protection from HIV infection were enhanced. Research on the intersections of gender inequity, violence and HIV have shown the critical importance of gender inequity in HIV risk (Jewkes, Dunkle, Nduna & Shai 2010; Dunkle, Jewkes, Brown, Gray, McIntyre, Harlow, 2004).

Implications for intervention

The focus of women’s empowerment work has largely been on restraining men, educating women and economically empowering them. This has failed to acknowledge the sexual and emotional agendas of women, particularly the extent to which women accept a surrender of power in order to meet cultural expectations of ‘good’ women. Women expressing conservative femininity are examples of this gender pact. Our conservative femininity should not be read as a fixed category (or ‘type’), as the relationships in which gender is constituted are dynamic and these constructions are part of a systematic and collective process (Connell 1995). It is also not be read as a subject position that simply describes the thoughts and emotions of an individual. We recognise that in the process of talking the informants do express a subject position, yet in our analysis we have sought to develop a way of understanding the collective construction of gender identities. Through combining interviews and cultural emersion, we understand our conservative femininities to reflect a prevailing cultural model of ‘good’ young womanhood, adopted for the most part not so much through choice as through processes of learning how to be a ‘normal’ woman.

Recognizing a multiplicity of femininities and the complicity of women with men’s sexual power poses particular challenges for gender health interventions. Rather than focusing on admonishing the taking of risk and instrumental pursuits in relationships, sexual health promotion programmes may be more successful if they provide space for young women to discuss their sexual desires and hopes for emotional and relational fulfilment. Whilst education needs to address the politics of gender power, women need to be supported and brought to a tangible understanding of the possibilities and potential for them to assert control in sexual and relationship domains of their lives. In this respect, both skills building (communication and other skills) and tangible economic empowerment are important. The Image study (Pronyk, Hargreaves, Kim, Morison, Phetla, Watts et al 2007) showed that combining an economic empowerment initiative (revolving micro-loans) with a gender intervention and community action on gender could enable partner violence to be reduced even among poor and disempowered women. This was an example of a tangible empowerment intervention providing space in which women could reframe the micro-dynamics of domestic power.

Schools are a key setting for gender interventions. In South Africa the new Curriculum and Assessment Policy Statement (CAPS) for Life Orientation includes a focus on gender and violence (Department of Education 2011). Our findings point to the importance of exposing young women (and men) to a variety of ways of being women, and raising consciousness on the dynamics and operations of (gendered) power as well as focusing on women’s right to live without violence. Furthermore, through teaching communication skills and enabling class discussion about sexuality and intimacy, it is possible to promote the idea that irrespective of the context of relationships, and whether or not women are able or desire to challenge the dominance of men in society more broadly, there is scope for negotiating the terms of intimacy within relationships.

Table 1
Sketches of research participants.

Research highlights

  • young women construct their femininities and exercise agency in relationships in a context constrained by patriarchy and poverty
  • women’s agency chiefly concerned partner choice and relationship initiation, as men partners expected to control women, and women knew this
  • many women expressed highly acquiescent femininities, with power surrendered to men, in a choice that was culturally rewarded
  • other women tried to accommodate men’s power whilst negotiating greater respect; a third femininity aspired to was a more empowered ‘modern girl’
  • multiple and dynamic femininities provide possibilities for change, essential for promoting health and reducing violence


The study was supported by the National Institute for Mental Health (grant no.MH 64882-01). Kate Wood and Nata Duvvury were co-investigators on the study for which the data used in this paper was originally collected, that is the qualitative evaluation of Stepping Stones. This data was collected by Khanyisile Bakam, Siyabulela Sinkoyi, Lindiwe Farlane and Yandisa Sikweyiya. We thank our participants for giving us a window on their lives. Thank you also to the Robert Wood Johnson Foundation Health and Society Scholars Program Working Group on Gender and Health at Columbia University for editorial feedback.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Abrahams N, Jewkes R, Laubscher R, Hoffman M. Intimate partner violence: prevalence and risk factors for men in Cape Town, South Africa. Violence and Victims. 2006;21(2):247–264. [PubMed]
  • Barlow TE, Dong MY, Poiger UG, Ramamurthy P, Thomas LM, Weinbaum AE. The Modern Girl around the World: A Research Agenda and Preliminary Findings. Gender & History. 2005;17:245–294.
  • Bhana D. Childhood sexuality and rights in the context of HIV/AIDS. Culture, Health & Sexuality. 2007;9:309–324. [PubMed]
  • Butler J. Feminism and the subversion of identity. Routledge; London: 1990. Gender Trouble.
  • Callan H, Ardener S, editors. The Incorporated Wife. Croom Helm/Centre for Cross Cultural Research on Women; Beckenham/Sydney/Dover, NH: 1984.
  • Campbell C. Selling sex in the time of AIDS: the psycho-social context of condom use by sex workers on a Southern African mine. Social Science & Medicine. 2000;50:479–94. [PubMed]
  • Campbell JC, Baty ML, Ghandour RM, Stockman JK, Francisco L, Wagman The intersection of intimate partner violence against women and HIV/AIDS: a review. International Journal of Injury Control and Safety Promotion. 2008;15:221–231. [PMC free article] [PubMed]
  • Connell R. Gender and Power: Society, the Person and Sexual Politics. University of California Press; Palo Alta, Calif.: 1987.
  • Connell R. Masculinities. Polity; Cambridge: 1995.
  • Connell RW, Messerschmidt JW. Hegemonic masculinity - Rethinking the concept. Gender and Society. 2005;19(6):829–859.
  • Department of Basic Education . Final draft. Department of Basic Education; Pretoria: 2011. Curriculum and Assessment Policy Statement (CAPS). Life Orientation Grades 7-9.
  • Dunkle KL, Jewkes RK, Brown HC, Gray GE, McIntryre JA, Harlow SD. Gender- Based Violence, Relationship Power and Risk of Prevalent HIV Infection among Women Attending Antenatal Clinics in Soweto, South Africa. The Lancet. 2004;363:1415–1421. [PubMed]
  • Gavey N. Hove. Routledge; Sussex: 2005. Just sex? the cultural scaffolding of rape.
  • Gramsci A. Selections from the Prison Notebook. Lawrence & Wishart; London: 1971.
  • Harris A. Future Girl: young women in the Twenty-first century. Routledge; New York: 2004.
  • Hassim S. The Gender Pact and Democratic Consolidation: Institutionalizing Gender Equality in the South African State. Feminist Studies. 2003;29(3):504–528.
  • Hearn J. From hegemonic masculinity to the hegemony of men. Feminist Theory. 2004;5:49–72.
  • Holland S. Alternative femininities, Body, age, identity. Berg; Oxford:
  • Hunter M. The Materiality of Everyday Sex: thinking beyond ‘prostitution’ African Studies. 2002;61:99–120.
  • Jeffreys S. Heterosexuality and the desire for gender. In: Richardson D, editor. Theorising Heterosexuality. Telling it straight. Oxford University Press; Oxford: 1996.
  • Jewkes R, Nduna M, Jama PN. Stepping Stones, South African adaptation. Second Edition Medical Research Council; Pretoria: 2002.
  • Jewkes R, Penn-Kekana L, Rose-Junius H. “If they rape me, I can’t blame them”: reflections on the social context of child sexual abuse in South Africa and Namibia. Social Science and Medicine. 2005;61:1809–1820. [PubMed]
  • Jewkes R, Nduna M, Levin J, Jama N, Dunkle K, Puren A, Duvvury N. Impact of Stepping Stones on HIV, HSV-2 and sexual behaviour in rural South Africa: cluster randomised controlled trial. British Medical Journal. 2008;337:a506. [PMC free article] [PubMed]
  • Jewkes R, Sikweyiya Y, Morrell R, Dunkle K. Technical Report. Medical Research Council; Pretoria: 2009. Understanding men’s health and use of violence: interface of rape and HIV in South Africa.
  • Jewkes R, Dunkle K, Koss MP, Levin J, Nduna M, Jama N, Sikweyiya Y. Rape perpetration by young, rural South African men: prevalence, patterns and risk factors. Social Science and Medicine. 2006;63:2949–61. [PubMed]
  • Jewkes R, Wood K, Duvvury N. “I woke up after I joined Stepping Stones”: meanings of a HIV behavioural intervention in rural South African young people’s lives. Health Education Research. 2010 [PMC free article] [PubMed]
  • Jewkes R, Dunkle K, Nduna M, Shai N. Intimate partner violence, relationship gender power inequity, and incidence of HIV infection in young women in South Africa: a cohort study. The Lancet. 2010;367:41–48. [PubMed]
  • Leclerc-Madlala S. Transactional sex and the pursuit of modernity. Social Dynamics. 2004;29:1–21.
  • Mikell G, editor. African feminism: the politics of survival in Sub-Saharan Africa. University of Pennsylvania Press; Philadelphia: 1997.
  • Morrell R, editor. Changing men in Southern Africa. University of Natal Press; Pietermaritzburg: 2001.
  • Morrell R, Jewkes R, Lindegger G. Hegemonic Masculinity/ies in South Africa: Culture, Power and Gender Politics. Men and Masculinities. in press.
  • O’Sullivan LF, Harrison A, Morrell R, Monroe-Wise A, Kubeka M. Shifting Sexualities: Gender Dynamics in the Primary Sexual Relationships of Young Rural South African Women and Men. Culture, Health and Sexuality. 2006;8:99–113. [PubMed]
  • Pronyk P, Hargreaves JR, Kim JC, Morison LA, Phetla G, Watts C, Busza J, Porter JD. Effect of a structural intervention for the prevention of intimate partner violence and HIV in rural South Africa: a cluster randomised trial. The Lancet. 2006;368:1973–83. [PubMed]
  • Renold E. Girls, Boys, And Junior Sexualities: Exploring Children’s Gender and Sexual Relations in the primary school. Routledge; London: 2005.
  • Vance C, editor. Pleasure and Danger: Exploring Female Sexuality. Pandora; London: 1989.
  • Varga CA. How Gender Roles Influence Sexual and Reproductive Health Among South African Adolescents. Studies in Family Planning. 2004;34:160–172. [PubMed]
  • Wojcecki J, Malala J. Condom use, power and HIV/AIDS risk: sex-workers bargain for survival in Hillbrow/Joubert Park/Berea, Johannesburg. Social Science & Medicine. 2001;53:99–121. [PubMed]
  • Wood K, Maforah F, Jewkes R. ‘He forced me to love him’: putting violence on the adolescent sexual health agenda. Social Science and Medicine. 1998;47:233–242. [PubMed]
  • Wood K, Jewkes R. ‘Dangerous’ love: Reflections on violence among Xhosa township youth. In: Morrell R, editor. Changing men in Southern Africa. University of Natal Press; Pietermaritzburg: 2001.
  • Wood K, Jewkes R. Blood Blockages and Scolding Nurses: Barriers to Adolescent Contraceptive Use in South Africa. Reproductive Health Matters. 2006;14(7):109–118. [PubMed]
  • Wood K, Lambert H, Jewkes R. ‘Showing roughness in a beautiful way’: talk about love, coercion and rape in South African youth culture. Medical Anthropology Quarterly. 1987;21:277–300. [PubMed]
  • Wetherell M, editor. Identities, Groups and social issues. Sage/Open University Press; London/Thousand Oaks: 1996.
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...