Social Science & Medicine 52 (2001) 1613–1627 ‘I think condoms are good but, aai, I hate those things’: condom use among adolescents and young people in a Southern African township Catherine MacPhaila,*, Catherine Campbellb b a CSIR Mining Technology, PO Box 91230 Auckland Park, 2006, South Africa Department of Social Psychology, London School of Economics and Political Science, Houghton Street, London. WC2A 2AE, UK Abstract Levels of heterosexually transmitted HIV infection are high amongst South African youth, with one recent survey reporting levels of 18.9% amongst 17–20 year olds and 43.1% amongst 21–25 year olds. In these groups levels of knowledge about HIV are high, but perceived vulnerability and reported condom use are low. Much existing research into youth HIV in developing countries relies on survey measures which use individual knowledge, attitudes and reported behaviour as variables in seeking to explain HIV transmission amongst this group. This paper reports on a focus group study that seeks to complement existing individual-level quantitative findings with qualitative findings highlighting community and social factors that hinder condom use amongst youth in the township of Khutsong, near Carletonville. Study informants comprised 44 young women and men in the 13–25 year age group. Data analysis highlighted six factors hindering condom use: lack of perceived risk; peer norms; condom availability; adult attitudes to condoms and sex; gendered power relations and the economic context of adolescent sexuality. Informants did not constitute a homogenous group in terms of their understandings of sexuality. While there was clear evidence for the existence of dominant social norms which place young peoples’ sexual health at risk, there was also evidence that many young people are self-consciously critical of the norms that govern their sexual behaviour, despite going along with them, and that they are aware of the way in which peer and gender pressures place their health at risk. There was also evidence that a minority of youth actively challenge dominant norms and behave in counter-normative and healthenhancing ways. The actively contested nature of dominant sexual norms provides a fertile starting point for peer education programmes that seek to provide the context for the collective negotiation of alternative sexual norms that do not endanger young peoples’ sexual health. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Adolescents; Peer education; HIV; Behaviour change; Sexuality; Sexual health; Condom use; South Africa Introduction Against the background of the growing HIV epidemic in South Africa, it is increasingly being argued that preventive interventions in this region may be most effective if directed at young people below the age of 16 years. Thus, for example, a recent survey in the South African township of Khutsong, our community of *Corresponding author. Fax: 27-11-482-3267. E-mail addresses: cmacphai@csir.co.za (C. MacPhail), c.campbell@lse.ac.uk (C. Campbell). interest in this paper, indicated that HIV infection was almost non-existent in the 13–16 year age group, followed by a sharp increase in the 16–18 year age group (18.9%), with the peak infection rates of 43.1% for the community as a whole being experienced by the 21–25 year age group (Williams, Campbell, & MacPhail, 1999). This paper reports on a study that aims to increase our understandings of the influences on adolescent sexuality within a broader interest in HIV-prevention in southern Africa. In using the terms adolescent and young person we refer to a group aged between 13 and 0277-9536/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 7 7 - 9 5 3 6 ( 0 0 ) 0 0 2 7 2 - 0 1614 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 25 years. Through the presentation of findings from focus groups with young township residents, we seek to highlight a range of factors that militate against condom use by this group of young people, despite high levels of knowledge about HIV infection and of the sexual health-enhancing benefits of condoms. Such factors include individual-level perceptions of health and vulnerability, community-level factors such as peer and parental pressure, and wider social influences including the social construction of male and female sexuality and gendered power relations, as well as economic constraints } all of which we will argue serve to inhibit condom use by young people, and to place them at risk of HIV infection. However, mindful of the fact that young people do not always constitute a homogenous grouping, and that there will always be a range of variations in the extent to which young peoples’ behaviour serves to reproduce or resist dominant social norms, we also seek to highlight examples of young people with counter-normative behaviours and views. Much research into adolescent sexuality has treated adolescents in a stereotyped and one-dimensional way (Aggleton, 1991; Aggleton & Campbell, 2000) with inadequate attention to young people whose views and behaviours challenge dominant stereotypes. One of the aims of sexual health promotion is to provide the context for the renegotiation of dominant high-risk behavioural norms by young people, and for the collective establishment of new norms of behaviour. It is therefore vital that research focuses not only on the way in which dominant norms place young peoples’ sexual health at risk, but also on the ways in which particular young people resist these norms, sometimes leading to alternative and less risky sexual behaviours and practices. Young people in South Africa have received much of their knowledge of sexual health promotion from government mass media campaigns (Friedland et al., 1991). More recently there has been a move towards participatory peer education approaches. However, such approaches appear to have made limited impact on the epidemic, which continues to rise. The rising epidemic is particularly focused among young people, with females being especially influenced (see Table 1). Understandings of the influences on sexual behaviour and the mechanics of sexual behaviour change are still limited, particularly in the southern African context. Due to these inadequacies in our knowledge, we have limited tools for understanding what is driving the epidemic amongst young people. Greater knowledge of the influences driving the epidemic will allow for a better understanding of the factors helping or hindering existing programmes and could potentially help with improving existing intervention programmes. In a review of the academic literature on the evaluation of HIV prevention programmes in developing countries, MacPhail (1998) highlights the dominance of survey approaches, generally based on the KAPB model of sexual behaviour. Information gleaned from KAPB surveys does not enable developers of intervention programmes to consider the contexts in which knowledge is gained and sexuality negotiated (Joffe, 1996), therefore giving only a partial picture of the complex factors shaping sexuality. In addition, the success of HIV intervention programmes is frequently evaluated using these narrow KAPB variables despite the realisation that increased knowledge does not impact on future behaviour (Elliot, Crump, McGuire, & Bagshaw, 1999). Furthermore, while programmes have begun to incorporate participatory methods in intervention, the community change that they aim to generate remains unevaluated due to the strong adherence to KAPB surveys which lack the potential for measurement of this kind (MacPhail & Campbell, 1999). Kippax and Crawford (1993) have criticised the concept of ‘sexual behaviour’ used in these studies. They argue that sexuality is too complex a phenomenon to be conceptualised in terms of decontextualised and quantifiable individual behaviours (e.g. condom use, anal sex) of the type measured in KAPB studies. Critics such as Holland and collegues argue for a more complex and contextualised definition of sexuality. Holland, Ramazanoglu, Scott, Sharpe, and Thomson’s (1990, p. 339) definition of sexuality forms the basis for the research reported below. By sexuality we mean not only sexual practices, but also what people know and believe about sex, particularly what they think is natural, proper and desirable. Sexuality also includes people’s sexual identities in all their cultural and historical variety. This assumes that while sexuality cannot be divorced from the body, it is also socially constructed. It is within this context of understanding the societal, normative and cultural contexts in which individuallevel phenomena such as knowledge, attitudes and behaviour are negotiated or constructed that the current research is located. Much of previous research, particularly in developing countries, has concentrated on the phenomenon of sexuality at the level of the individual, while neglecting societal, normative and cultural contexts. Focusing on the individual-level assumes that sexual behaviour is the result of rational decision-making based on knowledge. In reality, the complex nature of sexuality means that adolescents conduct their sexual lives through experiences and beliefs that have been generated through their membership of particular societies and communities. A wider view of the other levels of influence therefore C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 needs to be utilised (Campbell, 1997; Campbell & Williams, 1998). Attempts to study the sexual behaviour of young people have concentrated on these individuallevel indicators that have been easily measured through the use of quantitative methods. However, the use of quantitative methods in the evaluation of HIV prevention programmes provides answers limited to whether an intervention has worked rather than promoting an understanding of why the intervention has worked. In addition, in many instances when working within the field of HIV, researchers are still attempting to understand rather than to measure. Quantitative methods do not allow researchers to consider the processes at work but rather limit them to the final outcome. In this paper we address both issues through moving beyond the individual and using qualitative approaches. Literature review In this section we refer to the literature on adolescent sexuality in both developed and developing countries in the context of our interest in the way in which the social construction of sexuality might predispose young people to poor sexual health. While the literature on adolescent sexuality in developing countries tends to be quantitative and limited in focus, this is not the case in developed countries, where much more work has been done on the broader context of sexual behaviour. Four themes that dominate in this literature (particularly within the US, UK and Australia), have formed the starting point for our expanded research agenda in the southern African context. These themes encompass female sexuality (in particular conflict between sexual feelings and social norms), gendered power imbalances, features of male sexuality, and peer norms and values. The sexuality of young people has enjoyed less attention in literature emanating from developing countries although the limited work conducted in South Africa will be discussed below. Despite some notable exceptions (see Holland, Ramazanoglu, Scott, Sharpe, & Thomson, 1990, 1991; Holland, Ramazanoglu, Sharpe, & Thomson, 1992, 1994a) this literature often tends to refer to adolescents as a homogenous group and to make sweeping generalisations about their sexuality. Aggleton (1997) criticises this literature for failing to take account of wide variations in the sexuality of young people, arguing that such generalisations have played a key role in undermining the success of sexual health promotion among youth. He indicates that simplistically generalised views of adolescent sexuality held by adults, have influenced policy and practice so that young people do not receive the knowledge and services they require. This point is taken up in our discussion of our own research findings. 1615 Literature from developed countries Research on the broader context of sexual behaviour has been particularly concentrated on the contradiction between social norms of female sexuality and the sexual feelings of young women, highlighting the danger in which young women place their sexual health when adhering to social norms (Holland, Ramazanoglu, Sharpe, & Thomson, 1994b). A high regard for the preservation of reputation means that young women adhere to social definitions of sexual encounters as initiated by men, against female resistance. Women, therefore, often do not have condoms available and make few efforts to gain knowledge of their partners’ sexual histories, as this would be tantamount to admitting to themselves and society that they plan to engage in sex. In addition, women often avoid carrying condoms due to the negative reputations and labels associated with women who actively seek sex (Holland et al., 1990; Hillier, Harrison, & Warr, 1998). Social pressures encourage young women not to engage in sex but those that do are expected to do so in the confines of ‘serious’ and ‘trusting’ relationships (Holland et al., 1990). This emphasis on ‘serious’ relationships encourages premature trust of partners and therefore the non-use of condoms. (Holland et al., 1991; Ingham, Woodcock, & Stenner, 1991). Literature on the sexuality of young people in developed countries has highlighted imbalances in gender power that prevent young women from negotiating safe sexual encounters (Holland et al., 1992). Social constructions of masculinity that promote the idea of men ‘needing’ sex further constrain women’s negotiation potential by limiting opportunities for women to either refuse sex or negotiate safe sex (Wilton & Aggleton, 1991; Moore & Rosenthal, 1992; Ramazanoglu & Holland, 1993). Holland et al. (1991) indicate that many of the young women interviewed for their study in Manchester and London had experienced sexual initiation through coercion and force. While society does not accept male violence as a matter of course, the social construction of male and female sexualities in Western culture frequently blur the distinction between male violence, coercion or rape and ‘normal’ heterosexual sex. Here we must emphasise that not all men or women conform to these norms. As we will seek to highlight in our research findings below, a minority of young men and women define their sexuality outside of the norms constructed by society and in so doing, challenge traditional/normative social constructions of relationships (Holland et al., 1990, 1992). There is also a growing research literature concerned with male sexuality. Masculine sexuality is manifest in society’s classification of ‘normal’ men as being associated with multiple partners and power over women. Tension develops between the emotional vulnerabilities 1616 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 of young men and the behaviour that they are expected to adopt in order to be accepted as masculine in society (Holland et al., 1994a). The need for men to engage in multiple sexual relationships combined with internalised negative attitudes towards condoms place their sexual health at risk (Holland et al., 1990; Wight, 1994). Health interventions have frequently encouraged young people to use condoms or to ‘know’ their partners. Among all young people, but young men in particular, there is the perception that they can filter out partners dangerous to their health. Partners can therefore be categorised as ‘clean’ or ‘unclean’ based on their social interactions and appearance so that decisions about making use of condoms can be made (Waldby, Kippax, & Crawford, 1993). All groups are particularly influenced by the norms of their peers. This is especially true of young people and has been well documented in the literature regarding sexual behaviour (see Reed & Weinberg, 1984). Studies with American college students have shown that discussions of safe sex within friendships were a strong predictor of practising safer sex (Lear, 1995). Gender differences in the impact of peers were however found, with females being more likely to morally pressure their friends than their male equivalents. Although these examples indicate that peer norms assist in the adoption of safe sexual behaviour, Fisher, Misovich, and Fisher (1992) have indicated that in the majority of cases peer norms encourage risk. They indicate that peer norms function to promote unsafe sexual behaviour and to encourage concern about sexual health to be viewed in a negative light. Peer education stems from the belief that well-liked and respected peers may be able to encourage others towards behaviours that promote HIV prevention rather than the high-risk behaviours usually associated with peer norms (Serovich & Greene, 1997). Literature on adolescent sexuality in South Africa This surge of research concerning the various dimensions of adolescent sexuality in developed countries has not been matched in developing countries. To date, the majority of work in developing countries has looked at narrow variables of knowledge, attitudes and behaviours with the assumption that sexuality and sexual behaviour fall within the ambit of rational decisionmaking by individuals (Fife-Shaw, 1997). In most instances empirical research findings have concentrated on age at first sex, number of partners, awareness of condoms and knowledge about HIV (Flisher, Ziervogel, Chalton, Leger, & Robertson, 1993; Richter, SwartKruger, & Barnes, 1994; Swart-Kruger & Richter, 1997). While this information is frequently used to indicate the relative success of HIV-intervention programmes, the multi-dimensional context in which adolescent sexuality is constructed and negotiated is frequently not addressed (MacPhail, 1998). Historically, the South African literature on issues relating to teenage sexuality has concentrated on adolescent pregnancy and contraception use (Kau, 1988). More recently, however, there has been a trend towards an expanded focus to consider the broader social and community contexts in which young people negotiate their sexuality. Richter and Swart-Kruger (1995) indicate the importance of the social and economic contexts in which street children negotiate their sexuality. They show that for individuals such as street children there are limited opportunities to exercise control over their sexual encounters. In certain respects, research findings on condom use by South African youth are similar to research findings with youth in developed countries. Thus for example research by Preston-Whyte and Zondi (1991) and Abdool Karim, Abdool Karim, Preston-Whyte, and Sankar, (1992) highlight the way in which the use of condoms is seen to militate against young men’s notions of masculinity and pleasure. In addition, Abdool Karim et al. (1992) have illustrated that young people continue to view the use of condoms as only necessary among those already infected with STDs or HIV. However, in other instances the factors constraining the use of condoms among South African youth have been quite different to what has been documented in developed countries. PrestonWhyte and Zondi (1991) and Abdool Karim et al. (1992) point to the importance of fertility for young men and therefore young men’s opposition to condom use purely due to their contraceptive value. In relation to our particular interest in adolescent sexuality in the context of power relationships, the most promising literature in the South African context has been the emergence of a distinct body of South African literature on the incidence of violence and coercion experienced by young women and the impact that this may have on the adoption of safe sex (Wood & Jewkes, 1997; Wood, Maforah, & Jewkes, 1998). However, despite such positive developments in South African research, there has been a tendency to focus entirely on stereotypical norms of gender and sexuality without considering the ways in which some young people transform or contradict stereotypes. We have already referred to Aggleton’s (1991, 1992) critique of the tendency among researchers to consider young people as a homogenous group without taking intra-group differences into account. Variations in the social and cultural environments of young people make for heterogeneous behaviours and beliefs that are manifest both between populations, such as between South Africa and developed countries, and within populations, such as youth residing in our study area. If a key dimension of HIV-prevention programmes involves peers working together to develop the 1617 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 confidence and solidarity to assert their rights to sexual health and non-violent relationships (Campbell, Mzaidume, & Williams, 1998), there is a need to look further than the socially defined norms that often hinder the development of such confidence and solidarity. In addition to considering the ways in which young people reproduce stereotypical norms and relationships, there is the need to investigate counter-stereotypical ways in which particular young people might already be developing strategies for resisting stereotypical gender norms, and for reshaping their sexual relationships in more health enhancing ways (Campbell, 1995). For this reason one key interest in our analysis was to examine not only dominant representations of adolescent sexuality, but also the ways in which these representations might be deconstructed and reconstructed in ways that promote safe sex behaviour. Methods Context The present study was conducted in the township of Khutsong that lies about an hour to the south west of Johannesburg, South Africa. Khutsong is the township associated with the mining town of Carletonville where a community-based HIV prevention programme is currently taking place. The intervention includes syndromic management of STDs, condom distribution and peer education among particularly vulnerable groups. Levels of HIV in Khutsong are high, particularly among the younger population groups. There are also significant differences in the prevalence of HIV between the sexes and various age groups (see Table 1). A recent survey in the Khutsong community (Willams, Gilgen, Campbell, Taljaard, & MacPhail, 2000) revealed that levels of knowledge amongst our group of interest were high, with 89.9% stating that HIV transmission could be prevented by the use of condoms; 81.4% stating that remaining faithful to your partner could prevent infection; and 86.7% indicating that using clean needles could prevent HIV transmission. Of the Table 1 Rates of HIV infection among young people in Khutsong Sex Age group HIV positive (%) Male 13–16 17–20 20–21 13–16 17–20 21–25 1.2 3.3 22.4 7.6 29.9 58.0 Female years years years years years years Table 2 Sexually active adolescents in Khutsong Sex Age group Sexually active (%) Male 13–16 17–20 20–21 13–16 17–20 21–25 18.3 81.4 98.8 19.1 87.4 100 Female years years years years years years entire group sampled, 65.2% of males and 76% of females reported being sexually active when responding to a question asking if they had ever had penetrative sexual intercourse. A further breakdown of these figures is available in Table 2. Of those who had sex with regular partners, 69% said that they never used condoms, 16.7% sometimes and 14.3% always. Among casual partners the figures were slightly different with 59.3% never using condoms, 7.6% using them sometimes and 33% using them all the time. There were also slight differences in the gender distribution of condom use, with males using condoms consistently more often. In the majority of cases, the younger age groups made less use of condoms than their older contemporaries did. These findings are consistent with a large research literature that suggests that knowledge of sexual health risks is not necessarily a good predictor of condom use (Bertrand et al., 1991; Fergusson, Lynskey, & Harwood, 1994). It was against the background of the poor knowledge–behaviour relationship highlighted by the survey that the present study sought to examine the broader contextual factors influencing the use or non-use of condoms by young people in Khutsong. Subjects Research informants consisted of 44 young people, aged between 13 and 25 years, half males and half females. Informants were recruited through a convenience snowball sample with initial contacts being established through part-time staff of a large HIV-prevention programme with which the authors of the paper are associated. At the time of the study (early 1999), none of the informants had any particular relationship to the project, or any particular interest or involvement in HIV prevention. Data collection took the form of eight focus group discussions with between six and eight participants in each group. The focus groups were single sex and were further broken down into three age categories, namely 13–16 year olds, 17–20 year olds and 21–25 year olds. 1618 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 Focus groups Our choice of focus groups over individual interviews was determined by three considerations: the first conceptual, the second pragmatic and the third related to our particular empirical research agenda. Conceptually we chose focus groups because of our interest in moving away from the conceptualisation of ‘sexual behaviour’ as the product of individual decisions, in favour of our preference for the concept of ‘sexuality’ as a socially negotiated phenomenon, strongly influenced by peer norms. Lunt and Livingstone (1996) describe the focus group as a microcosm of ‘the thinking society’, capable of revealing the processes whereby social norms are collectively shaped through debate and argument. As opposed to in-depth interviews, which highlight the views and attitudes of single individuals, focus groups reveal the way in which particular individuals’ opinions are accommodated or assimilated within an evolving group process. Individual inputs weave and clash through the process of dialogue and argument between individual participants as peers ask one another questions, exchange anecdotes and comment on one another’s’ experiences and points of view. In many ways, the focus group also serves as a microcosm of the processes underlying successful peer education. Ideally peer educational settings provide a forum where peers can weigh up the pro’s and con’s of a range of behavioural possibilities, generating their own questions, answering these questions in their own terminology and in the light of their own priorities (Campbell & Mzaidume, 2000). From a pragmatic point of view, Kitzinger (1995) argues that focus groups are particularly appropriate for facilitating the discussion of taboo topics because the less inhibited members of the group often break the ice for shyer participants. This was certainly the case in our study in a southern African context where a range of obstacles hinder young people in discussing sex with adults or with young people of the opposite gender. In the feedback sessions at the end of each of our single sex focus groups, participants repeatedly stressed how unusual and positive they had found the experience of being able to argue openly about sex in a supportive context. Many said they had felt free speaking of issues they had never discussed before, and which they might have felt more reluctant to raise in a one-to-one setting or a mixed sex situation. Finally our choice of focus groups over individual interviews was determined by our empirical research agenda, namely our interest in uncovering both normative and counter-normative discourses regarding sexuality and relationships. Such competing discourses are more likely to be expressed in the context of debate and argument inherent in the focus group approach, than in the less controversial setting of an individual interview. Discussion topics After a 15-min ‘ice-breaker’ exercise, discussion time was equally divided between the four questions outlined below. These questions were intended to elicit information about the context in which adolescent sexuality is negotiated in Khutsong. 1. Why do people have relationships and are there different reasons for males and for females? Why do people therefore have sex and are there differences in male and female reasons? 2. Why do some people use condoms and some people do not? 3. Is AIDS a problem in this community? Among whom? Do others see that it is a problem? 4. Are there people who go against the norms of masculinity and femininity? What is the community reaction to them? Debates were lively with participants frequently interrupting one another and arguing over points of disagreement. Each focus group included a short feedback session. Without exception, people said that they had enjoyed participating in the groups, had found the discussions interesting and informative, and had not found it embarrassing to talk about sex or relationships. Several informants suggested that the researchers should run more groups of this nature to generate broader debate and discussion among young community members about what were too frequently regarded as taboo topics. The focus groups concluded with free question sessions and condom demonstrations and distribution. Focus groups were conducted by the first author and a co-interviewer who was fluent in Tswana and Xhosa (the first languages of the focus group informants). Nearly all the informants were fluent in English, however, and chose to express themselves in this medium, with translation skills only being needed on the rare occasions where someone could not think of a particular word or phrase. The discussions were taperecorded and later transcribed, and where necessary translated into English. Method of analysis Discussions were analysed by means of a two-stage interpretative thematic analysis, involving the detailed reading and re-reading of the discussions in order to generate explanations of why it is that young people continue to knowingly engage in potentially lifethreatening sexual behaviour. The first stage of the data analysis involved sorting the focus group transcripts into C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 broad content categories. Because of the informal nature of the focus groups, much effort went into systematising the data. After an initial sorting process, the second stage of immersion in the data sought to generate an account of factors mediating between high levels of knowledge about HIV and low levels of condom use among the participants and their peer groups. This stage of analysis pointed to six broad groupings of factors that might be said to mediate the relationship between AIDS-related knowledge and condom use among young people in Khutsong. These factors fell on a continuum ranging from the individual level of analysis, most frequently dealt with in the literature on behaviour change, to the societal and community levels that have been relatively ignored in literature stemming from developing countries. Each of these is discussed in turn below: (i) the extent to which young people have internalised the threat of HIV infection and see themselves as personally vulnerable; (ii) the influence of peer norms regarding sexual activity and the use of condoms; (iii) the availability of condoms in the Carletonville area; (iv) adults approval or disapproval of sex and condoms; (v) power in heterosexual gendered relationships and; (vi) economic constraints limiting opportunities for young people in Carletonville. In analysing the material there were two goals. First, to establish broad areas of consensus amongst focus group respondents on various topics, and second, to draw attention to areas of controversy or debate. Given the broader context of our interest in HIV prevention, we are interested not only in the way in which health damaging norms and social and sexual relationships are reproduced by the majority, but also in the possibility that these might be transformed. Hence our particular interest in those young people who challenge normative views and behaviours. In each section of our data presentation we seek to highlight both areas of broad consensus as well as debate. It is in the space opened up by such debates that the possibility of sexual behaviour change lies. Presentation and discussion of findings Perceived risk of HIV infection A requirement for translating knowledge into behaviour change is a feeling of personal vulnerability to HIV infection. HIV has been characterised as a disease of ‘others’ from the earliest reports of infection (Joffe, 1997). The primary group infected by the disease was homosexual men; a group already marginalized and made to seem ‘other.’ Within developing countries many of the initial cases of HIV were discovered among populations of sex workers. These women, like homosexual men, form a group different to the general 1619 population that can be ‘blamed’ and allow the threat of HIV infection to be externalised from those who would not classify themselves as either homosexual or as a sex worker. A baseline survey in the Khutsong community already referred to in this paper indicated that among young people in Khutsong there is little perception of their own risk despite high levels of infection. Almost 70% of young men said that there was no chance of their becoming infected or that they didn’t know whether or not they were personally vulnerabe, indicating that they didn’t connect their own behaviour with HIV risk messages. The rates for females were very similar. Contrast this to a mere 30% among both sexes who felt that there was either a moderate chance (12.9%) or good chance (17.6%) of their own infection (Williams et al., 2000). High levels of knowledge documented in the baseline survey were born out within the focus group discussions, as all participants knew about HIV either from school, the media, their peers or parents. There was, however, some confusion about methods of transmission and participants expressed fears about the possibility of transmission through kissing and through unexpected skin contact with infected blood. Questions asked at the end of each session illustrated that for the most part informants had good knowledge about HIV and were interested in furthering their knowledge. There were conflicting views about the incidence of HIV in their community and the extent to which themselves and other people in the community are vulnerable to infection. Some felt that HIV and AIDS were not very common in Khutsong and Carletonville. High levels of denial in communities not yet experiencing AIDS-related deaths, but with high levels of HIV infection, have been documented in other countries (see Ray, Latif, Machekano, & Katzenstein, 1998). I don’t think that AIDS and HIV are so common here. I think the main problem is STDs (13-16M)1 Others felt that the disease was a problem in their community and that personal experience with AIDS had forced people to acknowledge that it really exists. I think that now days most people have seen other people having AIDS. So I think that is the main point, they think that AIDS is real because in 1994, at the beginning people didn’t think AIDS was there. They say ‘Ag AIDS is nothing.’ But now I think they are . . . , they have seen some people with AIDS (17-20F) 1 The bracketed numbers refer to the age group and sex of the participant. 1620 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 Among the focus group participants it was more common for young women than men to state that HIV was a problem in their community. While acknowledging that HIV was a threat in their community, the younger men tended to view HIV as a disease associated with rape, commercial sex or excessive alcohol consumption. Throughout the discussion they made references to how easy it would be to be unwittingly infected by a woman who had previously been raped or who had not informed them that she was a sex worker. There was the implication that in these cases it would be the responsibility of the female in question to insist on condoms. Among male and female participants there was the notion that condoms are generally unnecessary in ‘steady’ relationships but that they should be used in casual encounters. Some indicated that they would make use of condoms to prevent pregnancy with regular partners but that condom use is most important for preventing getting a disease in casual relationships. AIDS is spreading because of the prostitutes. Even if she is not your girlfriend and you tell her that you want sex she will take you to her house. Sometimes you will find that there are a few guys and she will sleep with all of you. She will not even be thinking of condoms at the time (13-16M) Informants indicated that within regular relationships trust mitigates against using condoms. Young women argued that for a steady partner to insist on condom use is seen as indicating a lack of respect and trust that could destroy one’s reputation within the peer group. The discussions among young men indicated that although they are aware of HIV and have the knowledge to instigate condom use the majority has chosen to externalise the threat of HIV, making it the responsibility of others or an occurrence characterised by ‘abnormal’ sexual encounters. In the context of high levels of knowledge about HIV, the participants disagreed on the extent to which HIV infection has affected their community. The majority of participants did not mention personal vulnerability, although when expressly asked young women mentioned that they thought there was a moderate chance of their own infection. This contrasts with the attitudes of young men who externalised the threat of HIV infection to other groups, indicating that their own infection would be a mistake as a result of other’s actions. What friends do: peer norms Published literature has pointed to the influential roles that the attitudes and beliefs of peers have on other members of the peer group (DiClemente, 1990). South African and international authors have highlighted the way in which young people tend to internalise the frequently negative attitudes their peers express about condoms. In the South African context this includes attaching negative attributes to the contraceptive value of condoms and the notion that condoms are only necessary for those already infected with STDs or HIV (Abdool Karim et al., 1992; Akande, 1997). While many of the participants highlighted the importance of using condoms and their intention to use them in relationships, their general discussion indicated the manner in which negative peer norms impinge on consistent use of condoms. It’s if you have two girlfriends, your steady and your secret lover. You can never use a condom with your steady but you can use one with your secret lover because you don’t know if she has a disease. (2125M) If a boy wants to use a condom she will say it is because he disrespects her, because he wants to use ‘a plastic’ (13-16F) Young men appear to be particularly influenced by the dominant views of their peers and there were frequent references made to the ways in which young men using condoms were jeered at and belittled by their friends. Many of the participants stated that they had been accused of being stupid after using condoms and had decided that they would not use them again. In a context where young men stand by during their friend’s sexual encounters to warn of approaching adults, there is little chance of using condoms without being noticed. Young men who attempt to withstand dominant discourses of masculinity by avoiding sex are also subjected to taunting and teasing. Guys were asking me how I could not have had sex with such a nice girl. They said I was stupid. They said I didn’t know anything about sex. That’s why any girlfriend that I get, I want to make sure that I have sex with her (13-16M) Yet, despite stories about the taunting they suffer at the hands of their peer groups, there were participants in the discussions who had chosen not to adhere to peer norms by refusing to have sex. In most instances these young men were highly knowledgeable about HIV and were members of church groups that forbade sex before marriage. The participants also spoke with respect of young men from Lesotho with strong family ties that were known to be celibate. Young women appeared to feel the effects of peer norms to a lesser degree than their male counterparts. While admitting that there were instances when young women could be called ‘boring’ if C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 1621 they chose to have protected sex, the majority made no mention of peer norms ‘forcing’ them to have unsafe sex. There were, however, frequent references to the way in which the young men that they had relationships with would not allow them to use condoms. Campbell (1997) has discussed the importance frequently attached by men to the notion of flesh-to-flesh sex and many male participants raised this point. Young women, reporting on their male partners made jokes about these beliefs about flesh to flesh sex that prevent the use of condoms. As highlighted by the international literature there were some young women who resisted stereotypical notions of females and condom use. One participant in particular said that she enforced the use of condoms in her relationships and that to ensure condoms were always available she carried them herself. authority clinics. Male participants said that they also made use of retail outlets and Lovers Plus condoms when they had money but that there were not a reliable source due to lack of funds. I refuse. I don’t want to be doing that without a condom. I can say no thanks. I also think that the best thing is to always have condoms in your pocket because you don’t know what time you are going to have sex (17-20F) I don’t always have a condom when I need one. Sometimes you don’t know when sex is going to happen because he just asks you to come. Then you need a condom and it’s not there (17-20F) Much of the information gained during the focus group discussions mirrored what has been said of young people in developed countries. There was the tendency for Khutsong youth, particularly males, to make distinctions between partners who require condoms and those who don’t. Trust was given as the dominant reason for not using condoms but was never based on a negative HIV test or discussion about sexual histories. Rather, appearance and reputation determined trustworthiness. A contrast within the group and in opposition to developed country adolescents was the extent to which young women were beginning to ignore peer norms and influences. This was apparent in their determination to both carry and use condoms. Condom availability Within the Carletonville area, as in other areas of South Africa, free condoms are supplied in the government Department of Health and dispensed by local authority clinics in the township. The development of the Mothusimpilo Project in Carletonville has seen an improvement in the supply by ensuring that quantities are sufficient and that distribution takes place regularly. In addition to free condoms, a social marketing programme provides Lovers Plus condoms at a highly subsidised rate through retailers in the district. There are also commercial brands available at retail outlets throughout the town. Participants in the discussions made use of a variety of sources of condom supply ranging through friends, their schools and large retail stores. The majority, however, use the free condoms that are supplied through the local I get my condoms from the clinic sometimes. Not every time but most of the time. I have also bought Lovers Plus and the flavoured ones from Clicks (1316) Despite the relatively good availability of condoms in the Carletonville district participants still mentioned instances in which they had to have unprotected sex because they did not have access to condoms. Indeed, at the end of all focus groups there was great appreciation for the condoms distributed by the first author. Availability of condoms is particularly problematic for young women on two levels. First, the negative attitudes of nursing staff at the local clinics prevent them from accessing this free source of condoms. Many of the young women mentioned that they no longer went to local clinics after having had unpleasant experiences with the staff. While they continue to access health care through private doctors, their access to condoms is decreased, as they are not as freely available in doctor’s consulting rooms. I won’t go to the clinics. The nurses shout at you. They get angry when you take condoms and sometimes when you have relatives who are nurses they ask ‘What are you doing with condoms? Do you have a boyfriend? I am going to tell your mother.’ (21-25F) Second, social norms encroach on the extent to which young women are prepared to carry condoms with them. Participants mentioned that gossip is a constant source of conflict in the township and that women carrying condoms risked being labelled a ‘bitch’ or promiscuous. Male participants confirmed these notions by stating that they wouldn’t trust young women who carried condoms. Yah, it will worry a guy if a girl carries lots of condoms. He will worry that when he’s not there, what is she using them for? It means that I’m not alone in having sex with her (21-25M) Female participants recognised this problem but were also aware that young women who are able to overcome 1622 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 social distrust of female condom carrying are protecting themselves in terms of pregnancy and disease. Indeed, many of the older participants said that they ignore social norms about carrying condoms and replace them with their own norms. The community says that she likes sex because she’s carrying a condom in her bag. I think that girl is taking care of herself because she doesn’t want to be affected by STDs, AIDS and unwanted pregnancy (17-20F) Some young men agreed that young women carrying condoms did not always indicate that they are exceptionally sexually active. They argued that in the same way that they were frequently handed large numbers of condoms, young women were also exposed to condom distribution. In addition, some agreed with their female counterparts in saying that young women should protect themselves. Although attempts have been made by service providers to increase the access that young people have to condoms, there remain barriers. For young women in Khutsong condoms remain relatively inaccessible due to the attitudes of nursing sisters and clinic staff. As has been indicated in the international literature the spectre of young women carrying condoms remains problematic, but many of these young women felt that they would be prepared to risk their reputations for the sake of safety. However, it remains questionable whether these young women are able to consistently translate their determination into actual behaviour. The majority of young men supported dominant definitions of femininity by admitting concern about the reputations of young women with condoms. There were however, a small minority who agreed with female viewpoints. Adult views on sex and condoms Throughout the discussions there were oblique references made to the role that adults play in the sexuality of young people in Khutsong. The small size of Khutsong has resulted in close interrelationships between many of the adults. Parents inform one another of the actual and suspected sexual activity of their children in attempts to limit their behaviour. While there is concern among adults about the spread of HIV and STDs among the youth of the township, adults do not condone the use of condoms but rather prefer to encourage abstinence through punishment and gossip. Indeed, during feedback sessions at the end of the discussions many of the participants indicated that this was a rare opportunity to discuss sexuality with an adult who would not punish them and expressed a desire for their friends and siblings to also have this opportunity. The problem is that they just beat you for having sex. Others will just condemn you and spread the rumours around that you misbehave. They don’t give you any advice (13-16M) Male participants stated that parents’ disapproval of youth sex was often the reason that they didn’t use condoms at all. Most lived at home with their families and indicated that their opportunities to have sex were constrained by their parents. When the opportunity arises to have sex, many don’t bother with condoms as they are considered a waste of precious time during which adults are absent from home. In addition, adults in the community are seen as setting a bad example by giving young people conflicting messages about sex. Participants indicated that after heavy drinking many adults indulge in relatively public sex that acts as encouragement to the younger generation. Attitudes of adults to the sexual behaviour of adolescents mirror that of developed countries. Although some participants indicated that they could speak to their parents about sex, condom use was infrequently discussed. In agreement with the international literature, adults in Khutsong appear to be unable to view adolescent sexual behaviour as anything other than dangerous and irresponsible. Most participants suggested that adults in the community preferred warning them off sex through punishment and gossip, rather than encouraging them to use condoms when having sex. ‘We don’t call it rape, they’re our boyfriends’: gendered power relations The imbalance in power between male and female partners in heterosexual relationships holds sway over the ability of young women to either refuse sex or negotiate the use of condoms. This is particularly true for young women in South Africa where high levels of physical and sexual coercion and violence, triggered by attempts to discuss condoms or AIDS have been documented (Varga & Makubalo 1996; Wood & Jewkes, 1997; Wood et al., 1998). Khutsong is no exception to the gender imbalance found in South Africa. Male participants spoke of tricking young women into having sex, lying about using condoms and coercing women into having sex with groups of their friends. In addition, young men spoke of punishing women who had too many partners by beating them to teach them a lesson. Female participants indicated that the majority of men engage in relationships to satisfy their sexual needs and that C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 women are powerless to demand relationships on other terms. I also think that it is because usually men have ninety percent and women have ten percent of power. Men are the head of the family so that causes trouble because they can abuse women (13-16F) If sex is not willingly provided, many men in the community feel that they can insist on it as being a necessary part of a relationship and as proof of their girlfriend’s love. Violence and coercion are often used on unwilling sexual partners. They find you on the street and they force you to go home with them so that they can have sex with you. It is rape but we don’t call it rape because they are our boyfriends (21-25F) Where the imbalance in power reduces the female voice in the negotiation of sex, negative attitudes towards condoms internalised by some young men dominate relationships. In instances such as these the chances that women are able to insist on condoms are very small. Running counter to stereotypes about male dominance in relationships, a minority of male participants belonging to religious organisations heatedly defended the rights of women in sexual relationships. Church member’s continuing resistance of stereotypes of masculinity may provide some reasons for the protective effect of church membership indicated in Khutsong (Campbell & Williams, 2000). I agree with these guys about sex with your girlfriend. There must be an agreement. The problem is that others do it forcefully (13-16M) Only one woman in the focus group discussions indicated how she resists stereotypes of masculinity, femininity and power in relationships. She recounted a story about being forced into having sex by a friend during which time she physically attacked him, forcing him to flee in terror. Other participants expressed admiration for her abilities and indicated that they would feel powerless to act in such a situation. Issues of sexual violence and coercion were similar among Khutsong youth and youth in developed countries. Many young women experience violence in the course of their sexual relationships that lessens their abilities to either refuse sex or negotiate condom use. Very few young women have developed methods of increasing their power within relationships and the majority of men echoed common male perspectives. The limited ability of women to increase their power and therefore their ability to negotiate within sexual relationships mirrors findings in other countries. The small 1623 amount of work concerned with young men, limits the extent to which similarities and differences between South African men and men from developed countries can be explored. However, it would appear that there are more similarities than variances. Economic context of adolescent sexuality The use and non-use of condoms by young people in Khutsong cannot be divorced from the economic context in which they live their lives. At the simplest level, participants indicated that poverty in the Khutsong community was a powerful agent in preventing young people from purchasing condoms. Concern was expressed about the ability of the government to continue providing free condoms and the consequences this would have for the sexual health of the community. Diseases will not end if they sell condoms to us. People here are very poor; if someone gets five rand they spend it on bread and candles, not condoms (2125F) Efforts are being made by the Society for Family Health (SFH) to encourage people to purchase condoms but in many cases there is no available money for what is perceived as a luxury. Participants in the discussions were very interested in the female condom recently launched by SFH but admitted that at nine rand for a pack of two they were unaffordable. In addition, a complex relationship between poverty, gender relations and male ‘need’ for sex was identified as resulting in little condom use. During discussions about the nature of relationships male and female groups identified sex as a driving force for males to engage in relationships and money as one of the dominant reasons for females to have relationships. Webb (1997) has outlined the commercialisation of youth sex in the southern African context, claiming that women frequently engage in sexual relationships with the expectation of monetary remuneration. Participants spoke of young women and girls in Khutsong who engage in sexual relationships in exchange for lifts home from school, gifts and subsistence cash. There was the feeling that in situations of economic dependence such as these, young women would be unable to demand safe sex with their partners. There is this school called X, it is a bit far. Sometimes the students ask for lifts and struggle to get them. Others will sleep with the guys who give them lifts (13-16M) Little mention has been made of the economic context of adolescent sexuality in developed countries. Among the South African youth reported on here the 1624 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 importance of poverty and the limited abilities of young women to achieve economic independence impact strongly on condom use. Among economically marginalized communities in developed countries the problems associated with having to pay for condoms may find some resonance. However, the commercialisation of sexual relationships has received little or no attention in the international literature. Of importance in the context of Khutsong is the realisation that relationships of this nature are the exception, rather than the rule. While many young women expect gifts in the course of their relationships, relationships are not viewed as a source of income. Conclusion Despite increasing levels of HIV infection amongst young people in South Africa, with levels of HIVinfection in Khutsong at 43. 1% in the 21–25 year age group, there are few published studies of factors influencing the sexuality of young South Africans. Those studies that do exist often focus on individuallevel explanations of sexual behaviour, based on information gleaned from KAPB surveys, despite the well-documented limitations of KAPB models of sexual behaviour. In an attempt to begin to address this gap, our open-ended focus group study has highlighted a range of determinants of sexuality that range beyond the individual-level factors such as the low levels of perceived vulnerability that characterised our focus group respondents. Our findings also draw attention to a wider range of influences on sexual behaviour including factors such as peer norms and pressures; negative and unsupportive adult attitudes to youth sexuality; restricted availability of condoms; and broader social issues related to the social construction of gender and to economic constraints on young people. In conclusion we comment briefly on the implications of our findings to Aggleton’s (1991) advice that researchers and practitioners in the field of health promotion should take account of the differences in young people’s sexuality across various social, cultural and geographic spaces. In particular we comment on the implications of our findings for two particular dimensions of difference amongst adolescents: between-group difference and within-group difference. Two types of between-group differences are relevant to our interests: differences between adolescent sexuality in South Africa and developed countries such as US, the UK and Australia and differences between our findings in the Khutsong community and research findings on youth sexuality in other parts of South Africa. In relation to within-group difference, we comment on differences in sexuality within our particular group of adolescents in Khutsong. As we have seen, the focus group discussions were frequently characterised by fierce arguments and debate indicating that youth within Khutsong do not constitute a one-dimensionally homogenous group in relation to sexual norms and practices. Each of these forms of difference is discussed in turn. Focus group discussions with young people in Khutsong point to many significant parallels in the social and peer-influenced contexts in which they negotiate their sexuality } which are similar to those experienced by young people in developed countries. Thus for example, in many respects, the factors influencing adolescent sexuality in the Khutsong context are no different to those documented in the cities of London and Manchester by Holland and her associates (1990, 1991, 1994a,b). Thus, young people in Khutsong classify new relationships as ‘serious’ so as to justify their sexual behaviour and incorporate issues of trust that prevent them from using condoms in relationships in which they actually know very little about their partners. In addition, young men in particular rely on appearance and reputation to make decisions about certain women being ‘safe’ and therefore not requiring condoms for sexual intercourse. Young women in both contexts referred to factors limiting their access to condoms. They also referred to social pressures against the carrying of condoms in settings where young women’s reputations were frequently destroyed by the gossip of their peers and adults. Within these areas of similarity between Khutsong and developed countries there are instances in which negative influences appear to impact more substantially on South African youth. For young women the influence of male violence and coercion on their ability to negotiate condom use or refuse sex seems more important than in developed countries. This is not surprising given that levels of rape in South Africa estimated at one rape every 35 s are considerably higher than is the case in countries such as the UK and Australia (Robertson, 1998). Notions of masculinity that include the ideals of flesh to flesh sex with numerous partners are particularly well developed in South Africa and prevent young women from adequately protecting their health. While Holland et al. (1991) have documented male reluctance to use condoms; there are no beliefs that the use of condoms may be dangerous to male health as has been found in South Africa (Campbell, 1997). With regard to differences between our Khutsong research findings and investigations of sexuality in other areas of South Africa, very little research of this kind has previously been conducted in this country. There are however some differences to what have previously been documented. Preston-Whyte and Zondi (1991) and Abdool Karim et al. (1992) have indicated that the contraceptive value of condoms alone is responsible for their negative connotations among young South Africans. C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 They further indicate that for young people proof of fertility is an important factor in the move from child– adulthood. Our findings differed here. The focus group participants in Khutsong indicated that loss of educational opportunities and economic hardship were reasons that they avoided teenage pregnancy rather than welcoming it. There has also been the indication that negative views associated with condoms are male in origin and internalised by young women. Women in Khutsong determinedly defended condoms in their discussions, indicating that they had not internalised dominant male views but that their limited condom use went against their desires. In most instances their failure to use condoms could be attributed to the negative views of their male partners and their inability to question male views due to fear of violence. We can only speculate about the reasons for differences shown between youth in Khutsong and research on young people elsewhere in South Africa (Preston-Whyte & Zondi, 1991; Abdool Karim et al., 1992). Differences could be the result of historical changes that have taken place in young people’s sexuality in the almost ten years since previous work was conducted. In addition, the geographical location of the research differs widely, with previous work having been conducted in rural and traditional KwaZulu/Natal while Khutsong is situated in the more modernised urban Gauteng province. In terms of our broader interest in HIV prevention initiatives that incorporate peer education and community participation, the instances in which young people challenge dominant norms are of particular interest (given that a key aim of peer education is to provide the context for the collective renegotiation of dominant norms of behaviour that might be placing young people’s sexual health at risk). Among young people in Khutsong there were distinct differences between the views expressed by males and females. Male participants were more likely to have internalised the views of their peers and the social definitions of masculinity and femininity. Female participants confirmed this in their discussions of young men as the people most likely to be the source of gossip about reputation and sexual activity. In contrast, young women appeared to have developed strong relationships in which they defended one another’s rights and abilities to use condoms. Older female participants (21–25 years) in particular frequently indicated that social pressures no longer concerned them as much as they had in the past and that their sexual health was now of more importance than it had been in their teenage years. Within the male and female groups there were however, also differences. This was particularly apparent in the male groups in which church members continually challenged their male contemporaries. The areas in which they challenged social norms included male domination over women, the 1625 idea that males ‘need’ sex and the idea that young women carrying condoms could be classified as sexually promiscuous. Interestingly, in the light of discussion about peers jeering at non-conformers, these men were accepted and respected by all group participants. In the light of our interest in factors shaping youth sexuality, our analysis of our Khutsong focus groups has pointed to a range of factors that place young people’s sexual health at risk. However, this work has also highlighted the fact that young people’s sexuality is a contested and complex process. In terms of HIV, many of our research findings about young people’s experiences in negotiating their sexuality are consistent with survey evidence for low levels of condom use in this group. However, the focus groups highlighted areas of debate and difference in the views of our informants, which could provide space for peer education programmes that lead to the possibility of behaviour change. For intervention programmes, young people who challenge stereotypical norms and beliefs provide a fertile starting point for debates about the possibility of developing new behavioural norms. New norms and values negotiated by peer groups in this way provide health enhancing environments in which healthy sexual behaviour is more likely to be maintained. Acknowledgements This research was conducted as part of the evaluation of the Mothusimpilo Project in Carletonville. Thanks to Prof. Brian Williams, who initiated the project and gave valuable insight during the writing of the paper. Thanks to Zodwa Mzaidume of the intervention team at Mothusimpilo, Gaph Phatedi and Mrs Dijhejane for recruiting informants. David Molebatsi, Palesa Nxumalo, Patricia Zita, Adam Mphephoka and Prudence Ngoako assisted in translating and transcribing discussions. The work reported on in this paper was funded through the Department for International Development (UK) as part of the Carletonville-Mothusimpilo intervention project. Thanks to the Khutsong community for making community venues available for focus group discussions. Finally, thanks to the young people who participated in the discussions with such interest and enthusiasm. References Abdool Karim, S. S., Abdool Karim, Q., Preston-Whyte, E., & Sankar, N. (1992). Reasons for lack of condom use among high school students. South African Medical Journal, 82, 107–110. 1626 C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 Aggleton, P. (1991). When will they learn? Young people, health promotion and HIV/AIDS social research. AIDS Care, 3, 259–264. Aggleton, P. (1992). Young people, HIV/AIDS and social research. AIDS Care, 4, 243. Aggleton, P. (1997). Success in HIV Prevention. Horsham: AVERT. Aggleton, P., & Campbell, C. (2000) Working with young people } towards an agenda for sexual health. Sex and relationship therapy, 15, 283–296. Akande, A. (1997). Black South African adolescents’ attitudes towards AIDS precautions. School Psychology International, 18, 325–341. Bertrand, J., Makani, B., Hassig, S., Niwembo, K., Djunghu, B., Muanda, M., & Chirhamolekwa, C. (1991). AIDSrelated knowledge, sexual behaviour and condom use among men and women in Kinshasa, Zaire. American Journal of Public Health, 81, 53–58. Campbell, C. M. (1995). The social identity of township youth: Social identity theory and gender (Part 2). South African Journal of Psychology, 25, 160–167. Campbell, C. M. (1997). Migrancy, masculinity and AIDS: The psycho-social context of HIV-transmission on the South African gold mines. Social Science and Medicine, 45, 273–283. Campbell, C. M., & Williams, B. G. (1998). Evaluating HIVprevention programmes: Conceptual challenges. Psychology in Society, 24, 57–68. Campbell, C., Mzaidume, Y., & Williams, B. (1998). Gender as an obstacle to condom use: HIV prevention amongst commercial sex workers in a mining community. Agenda, 39, 50–57. Campbell, C., & Willams, B (2000). Is social capital a useful conceptual tool for exploring community level influences on HIV? An exploratory case study from South Africa. Submitted for publication. Campbell, C., & Mzaidume, Y. (2000) Grassroots participation, peer education and HIV-prevention amongst sex workers in South Africa, submitted for publication. DiClemente, R. (1990). The emergence of adolescents as a risk group for Human Immunodeficiency virus infection. Journal of Adolescent Research, 5, 7–17. Elliot, N., Crump, J., McGuire, A., & Bagshaw, S. (1999). Knowledge, attitudes and behaviour toward HIV infection among family planning clinic attendees: Changes between 1991 and 1997. New Zealand Medical Journal, 113, 121–123. Fergusson, D., Lynskey, M., & Harwood, L. (1994). AIDS knowledge and condom use in a birth cohort of 16 year olds. New Zealand Medical Journal, 107, 480–483. Fife-Shaw, C. (1997). Commentary on Joffe (1996) AIDS research and prevention: A social representation approach. British Journal of Medical Psychology, 70, 65–73. Fisher, J., Misovich, S., & Fisher, W. (1992). Impact of perceived social norms on adolescents’ AIDS-risk behaviour and prevention. In R. DiClemente, Adolescents and AIDS: A generation in jeopardy (pp. 117–136). Newbury Park: Sage. Flisher, J., Ziervogel, C., Chalton, D., Leger, P., & Robertson, B. (1993). Risk-taking behaviour of Cape Peninsular highschool students: Part VIII. Sexual behaviour. South African Medical Journal, 83, 495–497. Friedland, R., Jankelowitz, S., de Beer, M., de Klerk, C., Khoury, V., Csizmadia, T., Padayachee, G., & Levy, S. (1991). Perceptions and knowledge about the acquired immunodeficiency syndrome among students in university residences. South African Journal of Medicine, 79, 149–154. Hillier, L., Harrison, L., & Warr, D. (1998). When you carry condoms all the boys think that you want it: Negotiating competing discourses about safe sex. Journal of Adolescence, 21, 15–29. Holland, J., Ramazanoglu, C., Scott, S., Sharpe, S., & Thomson, R. (1990). Sex, gender and power: Young women’s sexuality in the shadow of AIDS. Sociology of Health and Illness, 12, 336–350. Holland, J., Ramazanoglu, C., Scott, S., Sharpe, S., & Thomson, R. (1991). Between embarrassment and trust: Young women and the diversity of condom use. In P. Aggelton, P. Davies, & G. Hart, AIDS: Responses, interventions and care (pp. 127–148). London: Falmer Press. Holland, J., Ramazanoglu, C., Sharpe, S., & Thomson, R. (1992). Pleasure, pressure and power: Some contradictions of gendered sexuality. The Sociological Review, 40, 645–674. Holland, J., Ramazanoglu, C., Sharpe, S., & Thomson, R. (1994a). Achieving masculine sexuality: young men’s strategies for managing vulnerability. In L. Doyal, J. Naidoo, & T. Wilton, AIDS: Setting a feminist agenda (pp. 122–148). Southport: Taylor & Francis. Holland, J., Ramazanoglu, C., Sharpe, S., & Thomson, R. (1994b). Desire, risk and control: the body as a site of contestation. In L. Doyal, J. Naidoo, & T. Wilton, AIDS: Setting a feminist agenda (pp. 61–79). Southport: Taylor & Francis. Ingham, R., Woodcock, A., & Stenner, K. (1991). Getting to know you . . . young people’s knowledge of their partners at fist intercourse. Journal of Community and Applied Social Psychology, 1, 117–132. Joffe, H. (1996). AIDS research and prevention: A social representational approach. British Journal of Medical Psychology, 69, 169–190. Joffe, H. (1997). The relationship between representational and materialist perspectives: AIDS and the ‘other’. In L. Yardley, Material discourses of health and illness (pp. 132– 149). London: Routledge. Kau, M. (1988). Sexual behaviour and contraceptive use by adolescent pupils in the Republic of Bophuthatswana. Curationis, 11, 9–11. Kippax, S., & Crawford, J. (1993). Flaws in the theory of reasoned action. In D. Terry, C. Gallois, & M. McCamish, The theory of reasoned action: Its application to AIDSpreventive behaviour (pp. 253–269). Oxford: Pergamon. Kitzinger, J. (1995). Qualitative research: Introducing focus groups. British Medical Journal, 311, 299–302. Lear, D. (1995). Sexual communication in the age of AIDS: The construction of risk and trust among young adults. Social Science and Medicine, 41, 1311–1323. Lunt, P., & Livingston, S. (1996). Rethinking the focus group in media and communications research. Journal of Communication, 46, 79–98. MacPhail, C. (1998). Adolescents and HIV in developing countries: New research directions. Psychology in Society, 24, 69–87. C. MacPhail, C. Campbell / Social Science & Medicine 52 (2001) 1613–1627 MacPhail, C., & Campbell, C. (1999). Evaluating HIV/STD interventions in developing countries: Do current indicators do justice to advances in intervention approaches? South African Journal of Psychology, 29, 149–165. Moore, S., & Rosenthal, D. (1992). The social context of adolescent sexuality: Safe sex implications. Journal of Adolescence,, 15, 415–435. Preston-Whyte, E., & Zondi, M. (1991). Adolescent sexuality and its implications for teenage pregnancy and AIDS. CME, 9, 1389–1394. Ramazanoglu, C., & Holland, J. (1993). Women’s sexuality and men’s appropriation of desire. In C. (e. ), Ramazanoglu, Up against Foucault: Explorations of some tensions between Foucault and Feminism (pp. 239–264). London: Routledge. Ray, S., Latif, A., Machekano, R., & Katzenstein, D. (1998). Sexual behaviour and risk assessment of HIV seroconvertors among urban male factory workers in Zimbabwe. Social Science and Medicine, 47, 1431–1443. Reed, D., & Weinberg, M. (1984). Premarital coitus: Developing and establishing sexual scripts. Social Psychology Quarterly, 47, 129–138. Richter, L., Swart-Kruger, J., & Barnes, J. (1994) Street children and youth in South Africa: AIDS-related knowledge, attitudes and behaviour. In: B. Grannon (ed.), Children and youth at risk: HIV/AIDS issues, residential care and community perspectives. Cape Town: NACCW. Richter, L., & Swart-Kruger, J. (1995). AIDS-risk among street children and youth: Implications for intervention. South African Journal of Psychology, 25, 28–31. Robertson, M. (1998) An overview of rape in South Africa. Paper of the Centre for the Study of Violence and Reconciliation, Johannesburg. Serovich, J., & Greene, K. (1997). Predictors of adolescent sexual risk taking behaviours which put them at risk for contracting HIV. Journal of Youth and Adolescence, 26, 429– 444. 1627 Swart-Kruger, J., & Richter, L. (1997). AIDS-related knowledge, attitudes and behaviour among South African street youth: Reflections on power, sexuality and autonomous self. Social Science and Medicine, 45, 957–966. Varga, C., & Mkubalo, L. (1996). Sexual (non)negotiation among black African teenagers in Durban. Agenda, 28, 31–38. Waldby, C., Kippax, S., & Crawford, J. (1993). Cordon Sanitaire: Clean and unclean women in the AIDS discourse of young heterosexual men. In P. Aggleton, AIDS: Facing the second decade (pp. 29–39). London: Falmer Press. Wight, D. (1994). ‘Boys’ thoughts and talk in a working class locality of Glasgow. Sociological Review, 42, 703–737. Williams, B., Campbell, C., & MacPhail, C. (1999) Managing HIV/AIDS in South Africa: Lessons from industrial settings. Council for Scientific and Industrial Research, Johannesburg. Williams, B., Gilgen, D., Campbell, C., Taljaard, D., & MacPhail, C (2000). The natural history of HIV/AIDS in South Africa: a biomedical and social survey. Council for Scientific and Industrial Research, Johannesburg. Wilton, T., & Aggleton, P. (1991). Condoms, coercion and control: heterosexuality and the limits to HIV/AIDS education. In P. Aggelton, P. Davies, & G. Hart, AIDS: Responses, interventions and care (pp. 149–156). London: Falmer Press. Webb, D. (1997). HIV and AIDS in Africa. Cape Town: David Philip. Wood, K., & Jewkes, R. (1997). Violence, rape and sexual coercion: Everyday love in a South African township. Gender and Development, 5, 41–46. Wood, K., Maforah, F., & Jewkes, R. (1998). He forced me to love him: Putting violence on adolescent sexual health agendas. Social Science and Medicine, 47, 233–242.