From First Love to Marriage and Maturity: A Life-Course Perspective on HIV Risk among Young Swazi Adults

Abstract

This paper uses a life-course approach to explore the sexual partnerships and HIV-related risk of men and women in Swaziland throughout their adolescence, their 20s and 30s. Twenty-eight Swazi men and women between the ages of 20 and 39 discussed their life histories in 117 in-depth interviews, with an average follow-up of 9 months. Many participants reported painful childhood experiences, including a lack of positive role models for couple relationships. Women’s first sexual partnerships often involved coercion or force and resulted in pregnancy and abandonment by partners, leaving women economically vulnerable. Most men and women reported a desire to marry and associated marriage with respectability and monogamy. Men typically did not feel ready to marry until their 30s, while women often married only after years in tumultuous relationships. A high degree of relationship instability and change was observed over the study period, with half of participants reporting concurrency within their primary relationship. Participants’ narratives revealed significant sources and circumstances of risk, particularly multiple and concurrent sexual partnerships, violence, and lack of mutual trust within relationships, as well as social ideals which may provide opportunities for effective HIV prevention.

Introduction

A person’s risk of HIV infection varies by age, gender, and life circumstances, but these variations in risk over time and through various life stages are not always contextualized or well understood. HIV risk factors are typically assessed over brief periods or at single points in time, without consideration of how those factors may have developed over time, or the impact of previous events in a person’s life-course. Among men and women in sub-Saharan Africa, certain periods of risk, such as the very high HIV incidence experienced by girls and young women, have received considerable attention (Dellar et al. 2015). Less is known about other periods of risk, such as the high incidence seen among middle-aged men and divorced and widowed women.

A life-course approach situates an individual’s identity and agency over time within structural, social, and cultural contexts (Giele and Elder 1998). This perspective considers both trajectories, which measure the ‘long-term processes and broader patterns of events in an individual’s experience in specific life spheres over time’, and transitions, or ‘brief events that mark chronological movement from one state to another’ (Donnelly et al. 2001, p. 160). In relation to HIV, a life-course perspective can integrate knowledge about specific transitions or longer-term trajectories that put people at risk for HIV infection with investigation of why these events occur and how they are related to both sociocultural norms and earlier events in an individual’s own life history.

A small number of studies have utilised a life-course approach to HIV risk in Africa. Early life experiences such as economic deprivation, death of parents, and abuse were found to shape the life choices and empowerment of South African sex workers (Campbell 2000). In Uganda, significant differences in life transitions (including sexual debut, marrying one’s first sexual partner, and being widowed or divorced) were noted between women who were and were not HIV infected (Boileau et al. 2009). HIV testing is often sought at the time of significant life events such as marriage or pregnancy (Maman et al. 2001) or when a sexual relationship is commenced or ended (Lupton et al. 1995), suggesting connections between HIV-related behaviours and key life transitions. Adolescence is a key life transition, and anthropologists such as Hunter (2010) and Dilger (2003) have provided richly contextualised perspectives on how young Africans navigate established cultural norms and new forces of modernisation and globalisation as they enter adulthood and initiate sexual relationships.

Other research, while not explicitly utilising a life-course approach, has found robust links between traumatic childhood experiences and risky or violent behaviours later in life. Orphans have been found to have younger age of first sex and to experience worse psychological health and increased sexual risk behaviours (Thurman et al. 2006). South African men who experienced childhood trauma are more likely to perpetrate rape (Jewkes et al. 2006), and men and women who have experienced past abuse are more likely to practice risky sexual behaviours, such as transactional sex (Dunkle et al. 2004, 2007).

In Swaziland, HIV incidence is highest for women in their late teens through 30s, and for men in their late 20s through early 30s (Ministry of Health 2012). HIV transmission occurs primarily within heterosexual partnerships (National Emergency Response Council on HIV/AIDS [NERCHA] 2009), and being in a sexual relationship without being married or living together is associated with a two-fold increase of HIV incidence among women (Ministry of Health 2012). Although data on sexual concurrency among Swazi adults are lacking, high prevalence of multiple sexual partnerships within short periods of time (Cockcroft et al. 2010) suggests significant sexual concurrency. Sexual concurrency has been theorised to exponentially increase the spread of HIV within a population (Morris and Kretzschmar 1997), and may be a critical factor in southern Africa’s very high prevalence HIV epidemics (Halperin and Epstein 2007). We use a life-course approach to examine the sexual partnerships and HIV risk of Swazi men and women throughout their adolescence, 20s, and 30s, when HIV risk is the highest.

Methods

We conducted 117 in-depth, life-course interviews with 28 Swazi men and women between the ages of 20 and 39 (58 interviews with 14 men and 59 interviews with 14 women). Participants were interviewed three to five times between July 2013 and August 2014, with interviews lasting up to 90 minutes and each participant being interviewed for an average of three hours in total. Methods as well as findings from initial rounds of interviews are described in detail elsewhere (Ruark et al. 2014).

Participants were recruited from a shopping centre in central Mbabane. We selected this venue because it was frequented by Swazis of diverse backgrounds and socio-economic strata. Participants were eligible to participate if they were between the ages of 20 and 39, were sexually experienced, and were willing to talk openly about their lives, particularly sexual partnerships. The study purposively sampled men and women who were diverse in terms of age, socio-economic status, education level, and marital status. Men and women who were willing to participate were interviewed in a private location three to five times over a period of 1 to 13 months (average 9 months), with the time between interviews depending on the participant’s preferences and availability. Some participants required more interviews to fully discuss their sexual partnerships, but no participant left the study prematurely. Interviews were performed by same-gender Swazi researchers between the ages of 25 and 35 in siSwati or in a mixture of siSwati and English, after obtaining written informed consent. Participants were offered cell phone airtime worth approximately $2 as an incentive at each interview.

Interviews explored participants’ sexual partnership histories in detail, with a focus on transitions and trajectories. Participants were asked to describe how individual sexual partnerships had begun and ended, who initiated these transitions, and the circumstances involved when one or both partners had concurrent sexual partnerships. Partnership trajectories were examined through questions about motivations, aspirations, and expectations for sexual partnerships.

All interviews were audio recorded, transcribed verbatim, and translated from siSwati into English. Data were coded using NVivo 10, and also analysed collaboratively by AR, NM, and LD, focusing on issues of translation, cultural context, and interpretation of areas of inconsistency, contradiction, and omission in participant narratives. Each participant’s interviews were ‘restoried’ (Creswell 2007, p. 56) into a coherent life-course narrative that included a chronology of major life events and a brief narrative of each unique sexual partnership. Following a narrative analysis approach (Kohler Riessman 2008), we treated each participant’s narrative analytically as a unit and attended to questions of how participants chose to represent themselves, how their accounts changed over time, and how interactions between interviewers and participants shaped participants’ narratives.

The Institutional Review Board of The Miriam Hospital (Providence, RI, USA) and the Scientific and Ethics Committee of the Swaziland Ministry of Health approved the study.

Findings

Participant narratives

Four detailed participant narratives are presented in an online supplement. These narratives provide a more in-depth description of men’s and women’s relationships, decisions, and circumstances of HIV risk over time, as well as the structural, cultural, and social factors that shaped their personal histories.

Relationship trajectories

Early childhood experiences

Participants’ narratives provided evidence that childhood experiences may shape expectations and trajectories of adult sexual relationships in significant ways, and a number of participants discussed the importance of relationship role models. A woman named Jabu (29 years old, cohabiting) was very self-aware about the impact of her childhood experiences, stating,

Life wasn’t easy when I was growing up. My parents used to fight when drunk… It had a negative impact in my life because I also ended up drinking and smoking… We are not a united and loving family. I can say this is the reason we decide to live with our boyfriends, [while] not married to them, as we are looking for love and closure.

While some participants reported close and loving relationships between members of their families, many participants reported painful childhood experiences, including not knowing their fathers, abandonment by mothers or fathers, and witnessing or experiencing violence and abuse. Futhi¹ (21 years old, partnered) remarked, ‘I never experienced my mother’s love,’ and said the only thing she had ever learned from her father was ‘violence.’ Mary (26 years old, partnered) explained,

I don’t know what it’s like to be with your parents, to get their love… I live that life where I have to think for myself because it feels like my parents are dead even though they’re still alive. I just need them to play their role and help me here and there but they don’t do that.

Participants’ adult lives often closely replicated the circumstances of their own childhoods, particularly for women. For example, among the seven women who had ever been married, six had parents who had married, while among the seven never-married women, six had parents who had never married (Table 1). A number of participants who had unstable romantic relationships and children from different fathers had lacked stability in their childhoods and parents who themselves had children from different partners. Women who had watched their mothers stay with partners who had concurrent sexual partners were likely themselves to tolerate concurrency from their partners. A similar parallel was observed regarding partner abuse.

Table 1.

Selected life experiences of participants

		Women (N = 14)	Men (N = 14)
Knew biological father		11	9
Lived with both parents throughout childhood		5	2
Expected to marry or be with 1st sexual partner “forever”		10	6
Married 1st sexual partner		0	3
Described 1st sexual encounter as “forced” or “rape”		3	0
Ever had a partner who was 10 or more years older		6	1
Ever had children		12	8
	Had child with 1st sexual partner	6	3
	Had first child while a teenager	5	0
	Had children by more than one partner	6	2
Ever married		7	5
	Parents ever married	6	2
	Parents never married	1	3
Never married		7	9
	Parent ever married	1	5
	Parent never married	6	4
Concurrent sexual partnership, ever		8	10
Concurrent sexual partnership, during study		4	6
Partner had concurrent sexual partner*, ever		14	8
Partner had concurrent sexual partner*, during study		6	1

^*While in a sexual partnership with the participant

Many participants reported growing up with a lack of positive models of couple relationships. Many had also experienced their parents’ separation. Half of women and men had never had married parents, while only seven participants had lived with both biological parents throughout their childhood. Some participants stated explicitly that they did not want to repeat their parents’ mistakes and had found positive role models elsewhere. For example, two men had chosen to conduct courtships and marry under the supervision of their pastors, including abiding by strict rules about not having sex until marriage. However, most participants thought that the quality of relationships among both their peers and elders was poor. John (35 years old, cohabiting) said that bad relationships affected the ‘nation as a whole.’ Musa (male, 34 years old, married) stated, ‘A bad relationship in the community affects everyone. If, for example, I always abuse my wife, what will the young boys copy from me?’

First sexual encounters

Women were often pursued by older sexual partners while still adolescents, and often reported feeling coerced, persuaded, or tricked into having sex against their will. Women also commonly reported being wooed with money and gifts, although some women reported that they had refused to have sex with a man even after receiving gifts or money. Very few women represented their first sexual encounters as being a positive experience, and three women called their first sex ‘forced’ or ‘rape.’ Notably, ten of the fourteen women in this study said that they hoped or expected to marry their first partner or be with them forever, and many felt hurt or used when their hopes were disappointed. Not one woman ultimately married her first sexual partner.

Men, in contrast, admitted that they were strongly motivated during adolescence to have sex. They described engaging in relatively short-term and casual partnerships with age-mates. While women often described their first sexual encounters in great detail and with strong emotion, male participants in this study typically had little to say about their first sexual partners. Nonetheless, six men reported intentions to marry their first sexual partner, and three had.

Pregnancy and parenthood

Reports of condom use varied widely, from no use to consistent use, although reported consistent use was in some cases belied by a pregnancy. Busi (37, partnered) expressed a common sentiment when she said, “we didn’t use any condoms… I told myself that I won’t be in any relationship again so there was no reason of doing it.” Busi married this partner but the relationship was not sexually exclusive, so that her assumption of a lack of risk was misplaced. Several women recounted that they stopped using condoms after the first sexual encounter, due to feeling trust for their partners. It is therefore not surprising that six women reported that their first sexual partnerships led to pregnancy, and five women became mothers while still teenagers.

For all women, having a child had long-term consequences and changed the trajectories of their future sexual relationships. Women became economically vulnerable and often economically dependent on their children’s fathers, and on future male partners. No woman in this study ever explicitly stated that she had entered a sexual relationship because of poverty or to provide for her children. Yet difficulty in financially providing for their children, the importance of financial support from male partners, and the inadequate support provided by their children’s fathers were a constant refrain throughout women’s stories. Women were also reluctant to separate from the fathers of their children due to the perceived stigma of having children from different fathers, the difficulty of finding a new partner who would accept another man’s children, and their desire for their children’s fathers to be involved in their children’s lives. Yet six women had had children with two or more men, and six were not in a relationship with the father of any of their children at the start of the study.

Having children did not influence the course of men’s sexual relationships as it did for women. While most men with children represented themselves as responsible fathers who provided for their children, not one had his child live with him after a separation from the child’s mother, and women’s accounts indicated that men could often hide the fact that they had children until well into a new relationship. Furthermore, the stigma associated with having children out of wedlock appeared to be directed solely towards women. For example, Busi reported that her long-time partner’s family was refusing to allow their marriage because she had children from previous relationships, while her partner’s children from another relationship were not seen as a problem. Some men did perceive that fatherhood had changed them for the better, and some reported deciding to be sexually monogamous after they had a child together. Samuel (38 years old, cohabiting) recounted,

I used to spend a lot of time and money with friends, going out drinking and maybe spending on girls. Since I met my partner, I have had to be responsible. The first month she got pregnant she insisted that I should open a savings account for my child.

Marriage and maturity

Marriage was a nearly universal desire among participants in this study, although only three men and three women were married at the start of the study. Men and women associated marriage with stability, respectability, maturity, and sexual ‘faithfulness’ (not having sex with non-marital partners). While most women began to dream of marriage from their first sexual relationship, some reached their goal only after years of enduring a tumultuous relationship and a partner’s repeated concurrent sexual partnerships. Other women chose to marry and stay married, despite a lack of satisfaction with their marriages, because they valued the social status of marriage. Lindiwe (female, 33 years old, married) stated, ‘I wanted someone who would make me a human being at the end of it all and not be like other girls who just wander without marriage.’ Yet other women, particularly those well into their 30s, seemed to think that their aspirations to marry might never be fulfilled.

Men possessed greater power to marry, if they wished, and also the ability to delay marriage for some years without hindering their chances. Notably, all of the unmarried men in this study expressed a desire to marry their current main partners, although most also gave reasons why they planned to delay marriage for some years. Men thought they should first become more financially stable (in part because of the obligation to pay lobola, or brideprice), and also believed that once they married they should stop having outside partners, a step which most men in their 20s expressed they were not yet ready to take. With one exception, every unmarried man in his 20s reported he had other sexual partners in addition to the partner he professed to love and intend to wed. Some men expressed considerable shame and remorse, and seemed caught between the desire to be respectable and monogamous and the desire to enjoy themselves through having multiple partners.

In contrast, men in their 30s generally represented themselves as having already achieved monogamy and respectability. (No participant in this study was in a polygamous marriage, and polygamy was mentioned only a few times by participants.) Like Mandla (37 years old, cohabiting), who called his conduct during his 20s ‘not the proper life’, men were reluctant to talk about the sexual partnerships of their 20s. Bheki (36 years old, cohabiting) said, ‘Look at how old I am. I could not be running after women at this age… I have to settle down, get my wife and take her home.’ In fact, he was the only man in his 30s who did admit, very obliquely, that he was still having concurrent partners.

Most participants in this study expressed optimism about the trajectory of their current relationships, often despite painful past experiences. They expected to marry, to raise children with a partner, and in the words of Futhi, ‘to be a happy family.’ However, for some participants, middle age also brought more sober outlooks. Busi was sure that if her current relationship failed she would not have another relationship, confessing, ‘I am really tired, and men, they are all the same. We have problems, but you decide to stay and you tell yourself that this time I will make it work as I am tired.’

Relationship transitions

Many of the relationships described by participants in this study were notable for their lack of clear-cut transitions. It was often not apparent from participant’s accounts when sexual partnerships had begun or ended, how many times participants had separated from and reconciled with sexual partners, or whether a participant was still involved in a certain sexual partnership at the time of the interview. Some of this lack of clarity was may have been due to participants’ reluctance to fully disclose the personal details of their lives. Yet in some cases the participants themselves seemed unsure of the current status of a relationship, particularly when a partner lived some distance away and was not seen on a regular basis.

The longitudinal nature of this study, with participants being followed up for an average of 9 months, allowed us to observe a number of relationship transitions (summarised in Table 2). Over the course of the study, two men and two women married their partners, while two men and four women permanently separated from their partners. Most men and women were still with their original partners at the end of the study, although a number of these relationships experienced separations or concurrent sexual partnerships over the study period.

Table 2.

Changes in relationship status over study period

Men (N = 14)	Women (N = 14)
Change in relationship status
2 men married their partners	2 women married their partners
2 men permanently separated from girlfriends 1 entered multiple new partnerships	4 women permanently separated from boyfriends or cohabitating partners 3 entered new partnership
	2 women separated from, then reconciled with partners 1 had other partners during separation 2 hoped to marry partner
No change in relationship status during study follow-up
7 men dating or cohabiting with same partner 4 dating, 3 cohabiting 4 had a concurrent partner 7 intended to marry partner	2 women dating or cohabiting with same partner 2 dating and periodically cohabiting with partner 2 reported partner had concurrent partner 2 hoped to marry partner
3 men remained married	3 women remained married
	1 woman remained single

Initiation of sexual partnerships

Participants in this study described 98 distinct sexual partnerships as part of their life histories (women 51 and men 47). The circumstances in which sexual partnerships began were varied, ranging from the traditional (a young widower pursuing a hard-working farmer’s daughter at the encouragement of his father and brothers), to the decidedly modern (a young woman starting her most recent romance through the social network Facebook). Men were usually, although not always, the initiators, and almost every woman had stories of men ‘proposing love.’ Couples often initiated a sexual relationship after brief acquaintances, in some cases on the same day they met. Many participants reported that sex just ‘happened’, with no discussion of the relationship or issues such as condom use, HIV status or testing, or other sexual partners. Participants, especially women, were often surprised to discover some time into the relationship that their partners had other sexual partners.

Ending of sexual partnerships

Men and women described the ending of sexual partnerships quite differently, with men describing more cases in which they did not expect the relationship to continue beyond a few sexual encounters, whereas women described more cases in which one partner had chosen to end the partnership. The most common reason given by men and women for their own decision to end a partnership was that their partner had a concurrent sexual partner. Notably, not a single participant recounted a relationship ending because he or she had a concurrent sexual partner, which may imply a selective telling of sexual histories. Furthermore, the only reason that men gave for ending a sexual partnership was that a female partner had a concurrent sexual partner. In contrast, women discussed ending relationships for reasons such as a partner abusing alcohol, being physically violent and abusive, lying, and not providing financially.

Concurrent Partnerships

At the beginning of the study, 8 of 14 men, and 10 of 14 women, reported having engaged in concurrent sexual partnerships themselves, 8 of 14 men and all 14 women reported having ever had a partner who had another partner at the same time, and 3 men and 5 women reported that they currently had concurrent partners. Data collected over the course of the study suggested that these initial figures may have been underestimates. Six men reported having concurrent partners during the study period, while four women and one man discovered during the study that a partner had a concurrent sexual partner. Altogether, 14 of 28 participants reported during the study period either that they had a concurrent partnership, or that a partner did. In some cases, both partners had concurrent partners during the study period.

Discussion

Many participants in this study followed the patterns of their birth families, even when they wished to break generational practices of alcohol use, violence, and abuse—behaviours which could place them at risk of HIV. The links between witnessing and experiencing child abuse and experiencing or perpetrating violence as an adult are well established (Jewkes et al. 2002), and intimate partner violence is associated with higher risk of HIV infection (Dunkle et al. 2004; Dude 2009). Similarly, children of alcohol abusers are more likely to abuse alcohol themselves (Kerr et al. 2012) and alcohol use is a well-known risk factor for HIV acquisition for both men and women (Fisher et al. 2007).

Also notable were a few participants who had chosen not to follow in their parents’ footsteps, such as men who married and were involved in their children’s lives despite being raised without married parents or present fathers. Some women stated that had chosen partners who were not violent and abusive, as their fathers had been. Men and women who chose to direct their lives in what they felt were positive ways often spoke of the importance of role models, particularly couples with happy marriages, and communities such as religious congregations, which supported and reinforced their chosen values.

The narratives presented by participants in this study suggest a process of social learning, in which people learn from one another through observation, imitation, and modelling (Bandura 1977). The pathway between childhood experiences and adult trajectories is far more complex than simple cause-effect, and likely impacted on by a multitude of socioeconomic and cultural factors (including multi-generational poverty and structural disadvantage).

While some participants in this study reported close and loving relationships with their families, others reported relationships that were detached, abusive, lacking in love, and the source of lifelong pain. Secure attachment to caregivers during infancy and childhood is crucial and a lack of secure attachment has lifelong adverse effects on later intimate and sexual relationships (Mikulincer et al. 2002). Furthermore, a 2007 national survey found that a very high proportion of Swazi girls experience violence and abuse, with 33% experiencing sexual violence, 25% experiencing physical violence, and 30% experiencing emotional abuse before the age of 18 (Reza et al. 2007). Growing up without both parents, as did most participants in this study, may also be considered an adverse childhood event, as it is associated with poorer sexual and reproductive health, including earlier sexual debut and higher risk of adolescent pregnancy (Blum and Mmari 2005). One-quarter of Swazi children have lost one parent, 4% have lost both parents, and only 22% live with both parents (CSO & Macro International Inc. 2008). Previous research has also found that children’s health suffers when the quality of their parents’ relationship is poor (Surkan and Poteat 2011).

A significant number of women in this study reported being coerced, forced, or raped in their first sexual encounter. A previous study in Swaziland found that 15% of Swazi girls and young women were persuaded or tricked at their first sexual encounter, while 5% were forced or raped (Reza et al. 2007), while another study found significant coercion at first sexual intercourse for both boys and girls (Buseh et al. 2004). This study also aligns with previous findings that while adolescent boys and young men tend to have sex with age-mates, many girls and young women have older sexual partners. According to the 2006/7 Swaziland Demographic and Health Survey (DHS), 7% of girls ages 15–19 who had higher-risk sex in the past year did so with a partner who was 10 or more years older (CSO & Macro International Inc. 2008). Girls and women are at increased risk of HIV in age-disparate partnerships (Kelly et al. 2003) although they often do not perceive themselves to be at risk (Leclerc-Madlala 2008).

Girls and young women have a very high per-act risk of HIV infection (Pettifor et al. 2007), partly due to biological factors (Glynn et al. 2001), and are also put at risk through lack of condom use (Clark et al. 2006). While their lack of condom use is often attributed to lack of power (Luke 2005), the accounts of women in this study suggest another dynamic of risk. If a girl or young woman believes her relationship to be one of love, trust, and lifelong commitment, she may not question her partner when he tells her there is no need for them to use condoms, and may herself see no need.

In this study, both men and women reported continuing existing sexual partnerships until a new partnership had become established, rather than ending a partnership and then beginning the search for a new sexual partner. The distinction between these two sexual partnering norms may seem slight, but small differentials in the gap or overlap between sexual partnerships may be epidemiologically crucial (Morris and Kretzschmar 1997). A gap between sexual partners that is longer than the period of acute HIV infection (typically, a few months) should ensure that neither partner will enter the new sexual partnership recently infected and highly infectious (Mah and Halperin 2008). Conversely, even small periods of sexual concurrency may magnify the risk of HIV transmission, as people may be exposed to new sexual partners with recent infections and high viral loads, and HIV can spread rapidly through sexual networks (Mah and Shelton 2011).

Previous research in southern Africa has noted the paradox of low marriages rates in the presence of widespread aspirations to marry (Hosegood et al. 2009; Harrison and O'Sullivan 2010; Hunter 2010). The fact that Southern Africans marry later than do other Africans, and thus potentially have more partners before marriage, has been implicated in the high HIV prevalence in the region (Bongaarts 2007). Swazi census data show a decline in marriage since the 1980s (Ndlangamandla 2007). A comparison of DHS data from across the continent shows that the proportion of adult men and women who have never been married is several times higher in Swaziland than in most other countries in the East and Southern Africa region (ICF International 2014).

For most study participants, parenthood preceded marriage. While 85% of Swazi women and 52% of Swazi men in their 20s and 30s have children, only half this number are married (CSO & Macro International Inc. 2008). Furthermore, for this age cohort, 65% of mothers had their first child before the age of 20, and 40% of fathers had fathered children with multiple women (CSO & Macro International Inc. 2008). While many women in this research saw marriage as a primary objective of their romantic relationships, having children before marriage often threatened this goal. Men often hid the existence of children from sexual partners, and by women’s accounts often failed to adequately support their children financially. Previous research in Swaziland with female sex workers found that many felt economically vulnerable, and in some cases entered sex work, after separating from the fathers of their children (Fielding-Miller et al. 2014).

In this study, marriage was described as being associated with maturity, commitment and sexual fidelity. Given that men generally described marriage as an event to be delayed until they were ready to commit to a partner, marriage may be more a product of intentions to be faithful, rather than vice-versa. Men in their 20s declared intentions to marry, but often admitted that they were not faithful to their primary partners. In contrast, men in their 30s, almost without exception, represented themselves as having reached an age of respectability, maturity, and monogamy.

Participants’ narratives about their behaviours and aspirations may be viewed as reflecting cultural scripts, which Leclerc-Madlala (2009, p. 105) reminds us often have more to do with ‘assumptions and expectations’ rather than ‘people’s actual behavior’. Harrison and O’Sullivan (2010, p. 995) identify a cultural script of being ‘entitled freedom to ‘look around’ before settling down’ among men in KwaZulu-Natal. In an ethnographic study of marriage and HIV risk in Uganda, Parikh (Parikh 2007, p. 1206) comments that young men who find themselves without the economic resources to marry may engage in multiple and concurrent sexual partnerships as an ‘alternative route to masculinity’. Both of these scripts aptly describe the high-risk sexual behaviour reported by men in this study, but this study also found evidence of men at the age of ‘maturity’ following a cultural script of monogamy, or at least claiming to do so.

Participants thus described cultural scripts both for high-risk and low-risk sexual behaviour, a paradox that may be viewed as deriving from the co-existence in a society of ‘ideal culture’ (the cultural norms that people publicly affirm) and ‘real culture’ (the norms which they actually follow) (Chambliss and Eglitis 2013). The disjuncture between the two has been associated with increased HIV risk, such as when an ideal culture of trust and sexual exclusivity leads to lack of condom use in spite of a real culture of concurrent sexual partnerships (Ackermann 2003). In this study, participants described an ideal culture of marriage and monogamy, although it was clear from their accounts that the real culture of their social environment was characterized by low rates of marriage and lack of sexual exclusivity between partners. Some participants were quite dissatisfied with these cultural realities, a dissatisfaction which might provide leverage for effective HIV prevention. HIV prevention activities might promote partner reduction and sexual exclusivity by building on existing aspirations to ideal culture, and encouraging young Swazi adults to envision and pursue cultural norms which would more closely align real culture with ideal culture.

While this study elicited detailed narratives from participants, we cannot verify to what degree participants were describing the objectively true circumstances of their lives, and to what degree they were creating and presenting to the researchers idealised identities and life histories. In particular, men in their 30s who claimed to be monogamous may have been expressing social ideals rather than describing their actual behaviour, and social desirability bias may have been heightened by the age of the male interviewer, who was in his 20s and thus in a younger age cohort. Yet many accounts from both men and women seemed too divergent from social ideals, too detailed, and too full of colour and texture to be mere inventions, and women generally appeared more frank than men. It seems more likely that participants omitted, rather than invented, passages in their histories. In grappling with questions of truth and representation among life histories of South African sex workers, Campbell asserts that ‘objective veracity’ is not really the point, but rather how ‘people reconstruct and account for their life choices’ because this reflects ‘social identities’ (Campbell 2000, p. 489). A research challenge is to accurately describe both the actual behaviour and the social reality of individuals whose mental and physical health is at risk.

Conclusion

A life-course perspective on HIV risk among young Swazi adults reveals multiple circumstances of risk, as well as the presence of cultural scripts and norms of ‘real culture’ which support high-risk behaviours such as multiple and concurrent sexual partnerships. More research is needed as to how the specific transitions and trajectories of sexual partnerships influence risk, and how individuals might be encouraged to rewrite risky scripts. In addition, a deeper understanding of the disjunctures between actual behaviours and what is socially desired might inform interventions that seek to align these two realities in ways that decrease HIV risk.