Of “Bad Behaviour” and “Dangerous Sex” Moral Responses to the Threat of HIV/AIDS among Children and Youth in Rural Burundi

food security placed Burundi second to last in the Hunger Index for Africa in 2010. 15 These glooming statistics indicate that most children and youths in Burundi are growing up and coming of age in difficult circumstances. As will be elaborated in the context section of this article, the provinces of Makamba and Cibitoke, where the research participants live, were directly affected by the recent civil war, and the post-conflict situation is still taking its toll on the welfare of the population in these areas where resources are scarce and the security situation remains unstable. 6,16–18 Against this backdrop of harsh socio-economic realities, this article seeks to analyse the influence of an HIV/AIDS prevention project focusing on the attitudes of children and youth who were targeted as agents of social change in the approach adopted for the intervention. The article begins with an introduction to the methodology of the study, followed by a description of the data analysis. The results of the study are then presented under the headings of: a) attitudes and role models; b) early sexual debut and pregnancy; c) multiple partners and sexual assault; and e) bad behaviours and the Abstract Burundi has been tormented by armed conflict for decades. In the midst of reconstructing rural communities, young people are concerned with their future and the need to avoid “bad behaviours” in order to have a better future. This article is based on findings from an interview-based survey conducted among nine to 18 year-old boys and girls who live in rural Burundi and participated in an HIV prevention and empowerment project run by an international NGO. The findings show how the children and young people perceive sexual relations and the threat of HIV/AIDS, and how they attempt to deal with these by assuming a moral high ground and becoming role models in their community. The article argues that the research participants draw on a binary moral discourse of good and bad behaviour in which sex is dangerous and should be avoided. From their accounts, it is clear that the message they have adopted of sexual abstinence demands a continuous effort on their part to avoid a whole range of temptations and pressures in their daily lives. Moreover, the findings presented in the article reveal that the threat of HIV/ AIDS is just one among many concerns. In the local context, burning issues pertaining to sex, including issues of sexual assault, transactional sexual relations, early pregnancies and unwanted pregnancies, loomed large and were the primary concern of young people. The article reaches the conclusion that the HIV prevention project did not address these crucial issues, and that the abstinence message the participants appropriated is not a sufficient measure to combat the rising HIV prevalence among youth in rural Burundi.


Introduction
In Burundi, nearly 90 per cent of the population lives in rural areas. 1 A disturbing trend in the country's HIV/AIDS prevalence is that while it has recently been decreasing in urban areas, it is increasing in the rural areas and children and youths aged 12 to 20 are the most affected. 2,3 [iTo explain the reasons for the spread of HIV prevalence in rural Burundi is beyond the subject of this article, but new evidence challenges the common assumption that conflict (as well as post-conflict situations) and also refugee populations contribute significantly to the spread of HIV infection in Sub-Saharan Africa. It shows that there is a need to question and perhaps revise the assumption that conflict and refugee populations necessarily increase HIV/AIDS prevalence [49,50,61,62]] iThis article examines how the threat of HIV infection is perceived by children and youth living in rural Burundi and how they respond to this threat. The article is based on findings from an interview-based survey conducted among nine to18 year-old boys and girls who live in the provinces of Makamba and Cibitoke. They are not infected with HIV, but they are at risk and have become increasingly aware of the risk through their participation in an HIV prevention and youth empowerment project run by the international development agency, ADRAii [iiADRA is an acronym for Adventist Development and Relief Agency]. Evidence from Southern Africa indicates a strong connection between poverty and vulnerability to HIV infection. 4 Economic factors may reinforce unsafe sexual practices, which have also been reported from the Burundian context where nearly 70 per cent of the population lives below the poverty line. [5][6][7][8][9][10] Burundi has been tormented by armed conflict for four decades and a brief glance at the rankings of Burundi in international indexes reveals that the living standards of the population rank among the lowest in the world. It is listed among the four countries in the world with the lowest GDP per capita. 11,12 In the Human Development Index as of 2011, Burundi was ranked as the third poorest country in the world. 13 Foreign aid has come to represent a disturbing 42 per cent of its national income, which is the second highest rate in Sub-Saharan Africa. 14 Food, medicine, and electricity are all in short supply, and especially challenges related to food security placed Burundi second to last in the Hunger Index for Africa in 2010. 15 These glooming statistics indicate that most children and youths in Burundi are growing up and coming of age in difficult circumstances. As will be elaborated in the context section of this article, the provinces of Makamba and Cibitoke, where the research participants live, were directly affected by the recent civil war, and the post-conflict situation is still taking its toll on the welfare of the population in these areas where resources are scarce and the security situation remains unstable. 6,[16][17][18] Against this backdrop of harsh socioeconomic realities, this article seeks to analyse the influence of an HIV/AIDS prevention project focusing on the attitudes of children and youth who were targeted as agents of social change in the approach adopted for the intervention. The article begins with an introduction to the methodology of the study, followed by a description of the data analysis. The results of the study are then presented under the headings of: a) attitudes and role models; b) early sexual debut and pregnancy; c) multiple partners and sexual assault; and e) bad behaviours and the street. The article ends with an elaborated discussion of the presented results in view of other research findings from Sub-Saharan Africa, followed by a conclusion.

Methods and ethics
The present study is based on individual interviews and group interviews with a total of 284 research participants aged nine to 18years among whom by far the majority (253 participants to be exact) were between 12 and 16years old. Nearly an equal number of boys and girls were interviewed. The interviews were conducted in September 2011 during field trips to the Makamba and Cibitoke provinces in partnership with ADRA. The research participants were interviewed by 20 interviewers in teams of two about their participation in the ADRA-funded project, ' It aimed to do so by improving their knowledge and understanding of HIV prevention and transmission and by enhancing their capacity to deal with the effects of HIV/AIDS in their communities by training them to voice their concerns. The students in the so-called "Listener's Clubs" (LCs); an activity pursued under the 'Breaking the Silence' project, met once a week to discuss issues brought up in weekly radio programmes funded by ADRA, or concerns that they brought forward themselves. They participated in the radio programmes and communal events to raise awareness of HIV/AIDS and children's rights and shared this knowledge with their peers, parents and other adults. Teachers from each of the schools where the LCs took place received training in HIV/AIDS awareness, children's rights and how to facilitate and support the LC participants in their activities. Parents were also trained as dialogue facilitators and were encouraged to be supportive of their children's participation in the project. The participants were all enrolled in public schools at the time of the study and were generally active members of the LCs. They were all from poor families; yet given their school attendance and voluntary participation in the LC activities, their families must be considered relatively resourceful within the local context. The interviews were carried out on the premises of the schools that catered for the LC activities. Permission to interview the LC members was granted from the heads of the schools preceding the field trips and from teachers prior to the interviews. The project was welcomed and approved by all relevant local authorities and stakeholders. Whereas institutionalized ethical review boards have become a standard feature of research projects in many places worldwide, both Burundi and Denmark do not have such standardised schemes in place for qualitative social science research undertakings; i.e. there is no committee to which the research design can be submitted for ethical approval. This implies that the researchers have to account for the ethical standards of the research by revealing insights into the actual undertaking of their work. The design and implementation of the interview based survey were overseen by a senior researcher at the University of Copenhagen, Denmark. In the field, informed consent was obtained from the participants, which means that the researcher asked participants for permission to include them in the study after having provided information about the research objectives. 19 At the beginning of each interview, the interviewers therefore took time to introduce themselves, describe the research objectives, and outlined the principles of voluntary participation and confidentiality prior to conducting the actual interview. This is what Fluehr-Lobban has dubbed informed consent without forms (1994).The interviewers conducted semi-structured interviews. They used the same question guide in which the questions were grouped according to themes such as the students' participation in the LCs, challenges they have faced and what they perceived the impact of their participation in the LCs to be. In addition to these interviews, conversations with local ADRA staff members during the field trips also provided a perspective on their views of some of the issues that came up in the interviews with the LC participants. In the interest of anonymity, the names of students were not recorded, and therefore the only characteristics used to situate students in the process of writing up the material have been those of age and sex. The interviews were transcribed immediately after the interview sessions ended, and upon concluding the field trips the interview transcripts were translated from Kirundi into English. The translator participated in the field trip to Makamba and was present during many of the interviews and thus had a good sense of the content and context of the transcripts. The translations were cross-checked by the first and second authors and occasional ambiguities in the translation were discussed and adjusted. Subtleties of cultural expressions are likely to have been lost in the translation and given the sensitivity of the interview content, some accounts may not have fully revealed the depths of the participants' personal experiences. 20

Data analysis
The transcripts of the translated interviews were carefully read through and coded according to recurring themes. These themes were identified through attention to patterns regarding repeated phrases and prevalent issues. The statements were then categorised and compared according to province, gender and age and summarised in a matrix of themes, quotes and cases that enabled an inductive approach to be used in the write-up of the material. Whereas the children differed in chronological age and maturity, their statements were remarkably similar and consistent. One reason for this discursive homogeneity found in the material could be related to the fact that most of the children made their statements based on knowledge and attitudes with very little reference to actual experience as mentioned above. In terms of gender, the participants expressed fairly similar ideas about moral norms; but these ideas also entailed gendered expectations of what constitutes good and bad behaviour; i.e. in relation to the use of pornography and ways of dressing. In order to contextualise the collected data, reports and surveys from the 'Breaking the Silence' project were read along with transcripts translated into English from radio programmes produced by ADRA, which discussed issues of sexual abuse in local communities and HIV/AIDS. Reports produced by non-governmental and UN organizations concerning the situation of children and youth in Burundi were also consulted through a desk study, as was epidemiological data on HIV and AIDS prevalence in Burundi and the Sub-Saharan African region. Furthermore, literature and news updates on the political situation in Burundi were consulted. For the data analysis, the empirical findings were brought into

Burundi as a post-conflict setting?
Burundi has been marred by a series of armed conflicts since the mid-1960s with outbreaks of civil war in 1965,1972,1988, and again in 1993, three months after the first multiparty elections were held. 21,22 This latest conflict lasted nearly 13years and resulted in mass atrocities and ethnic cleansing, displacing more than one million people within the country and across the borders, especially in Tanzania and the Democratic Republic of Congo (DRC). 23,24 In spite of the fact that peace and power-sharing agreements were signed, and that the remaining rebel group National Liberation Forces (FNL) began its transition to a political party in 2009, sporadic fighting has continued between government forces and FNL. 18,25 In 2015, Burundi's president, Pierre Nkurunziza, prompted widespread protests by announcing that he was seeking a third term in office, which sparked a new wave of violence [65]. Hence frequent violence and political instability go hand in hand in Burundi and as a consequence, the maintenance of peace remains fragile, which makes the use of the designation "post-conflict" rather ambivalent. In this vein, it may be more adequate to perceive of Burundi as a nation where crisis amounts to chronicity rather than to a separated period of time. 23 The children and youth interviewed for this survey live in the Makamba province in the Southeast of Burundi bordering Tanzania and in the province of Cibitoke in the Northwest bordering the DRC. The two provinces are directly affected by the current instability. In Cibitoke, the presence of FNL forces continue to cause unrest. So does the unstable security situation in the DRC, which contributes to a continuous flow of Congolese refugees, and the recruitment of youths for the various rebel movements in that region also continues. [25][26][27] Cibitoke was one of the central areas for the violent conflict during the civil war and has remained neglected by the government, which is further exacerbated by the fact that only a few international organizations are working in Cibitoke due to the security situation. 16 This is in contrast to the situation in Makamba where security is more stable and where many international organizations have been working for manyyears. 16,28 The presence of international organizations is partly meant to facilitate the repatriation and reintegration of hundreds of thousands of war displaced Burundians who are returning from Tanzania afteryears, or even a decade or more, in exile, which adds pressure to already fragile communities. 28,29

Attitudes and role models
A baseline survey conducted by ADRA among 300 children and youths attending primary school in the Cibitoke and Makamba provinces showed that 97 per cent of boys and 94 per cent of girls had heard of AIDS and knew that it was a serious condition. 28 In the present qualitative study, the research participants expressed a sense of empowerment vis-a vis the question of how one should deal with the risk of HIV and AIDS. In all likelihood, the findings of the present study are informed by the insights that the children and youths have gained through their participation in the activities and interventions of the 'Breaking the Silence' project. They emphasized that through their participation, they have acquired knowledge, skills and confidence, while they also expressed a sense of responsibility towards their peers to share this new knowledge. In fact, the majority of the research participants actively aimed to position themselves as role models in their community and hereby attempted to act upon this knowledge by changing their own behaviour as well as that of others. This is demonstrated in the two statements below: I have gained a lot of knowledge that I wouldn't have gained, had I not been a LC [Listeners Club] member. I know how to protect myself but also others from having our rights violated. I also feel like I have matured with the project and I feel I have a responsibility towards my fellow students and myself. I need to share my knowledge and teach others. (Female, 11).I act as a good role model. When people see me they will know that I'm a LC member and they can say that I have good manners and I go to school. It can help change the attitude of others, even though I'm not talking. (Female,15).
This willingness to mobilize both peers and adults in their community and to be agents of change comes across as a rather striking feature in many of the interviews. This is revealed by the fact that the LC members made constant reference to the concept of the role model, which implies that one should adhere to certain attitudes and behaviours. LC members used a variety of expressions to collectively refer to the types of attitudes and actions that should be avoided such as "bad behaviour", "bad manners", "bad attitudes" or simply to be "misbehaving", which was the most common expression. In the following sections, we explore the attitudes and actions that were considered to be bad according to the research participants. We do so in order to gain an understanding of how they perceive the risk of HIV/AIDS and how to avoid it. Moreover, it provides an overview of other issues that are of profound concern to them.

Early sexual debut and pregnancy
First of all, the notion of misbehaving referred to having sex or to other connotations of the sexual relations that young people and their peers engage in. In the LCs, it seemed that participants were taught that they 'should wait with having sex' and according to a 14-yearold female participant, both the LC and the church: "help teaching that having sex is a sin". Participants were also taught to "dress properly" and according to some participants, when girls "dress inappropriately" it "induces" boys to want to have sex with girls. This issue was mentioned by male and female participants alike. It seemed, however, that only girls were blamed for "not dressing well", which refers to "transparent clothes or clothes that are too short such as not covering the knees". To be sexually active was associated with the risk of contracting HIV as evidenced by the fact that participants often connected the two explicitly the way it is expressed in the following interview excerpt: We advise others how they can prevent HIV/AIDS. We go to advice others who are misbehaving because they might get HIV/AIDS. Misbehaving is when someone is having sex. (Male, 15) The risk of HIV was linked to pregnancy and several participants related how they had advised a pregnant girl to get tested for HIV. When two 15-year-old male participants explained to the interviewer how they felt that they made a difference, they exemplified this with situations in which they had influenced friends as well as a pregnant adult woman to get tested: I like teaching friends. There is even one of my friends I taught and she got tested when she was pregnant. At my home when my father died, my mother started going out with another man, who was said to be HIV positive. When I saw she was pregnant, I advised her to go and get tested so that she can protect her baby from HIV/AIDS. LC activities also changed my behaviour, now I'm a good example to others. The extent to which some LC members alluded to a sense of responsibility for their peers with regard to HIV/ AIDS is remarkable. The statement presented below attests to this and to a situation where LC membership produces a kind of moral superiority that sets LC members apart from their peers: We are different and we have to help protect other children by informing them about their rights and teaching them how to prevent HIV/AIDS. Many children are not aware of how dangerous HIV/AIDS is, so it is up to us to inform them, because we have access to the information (Female 11). Concerns voiced over sexual relations often pertained specifically to the issue of having an early sexual debut, and especially among female participants, early pregnancy was a major issue. In this regard the notion of misbehaving was linked to unwanted pregnancy, which is reflected in the three statements presented below: I feel very sad about what is happening to young girls. Some get pregnant, others marry young and leave school. It's a big problem and I would like for it to change. One of my neighbors was misbehaving, going out to meet men and coming home late. The mother supported her but the father was not happy. So we went to talk to her and advise her. She didn't listen and ended up getting pregnant (Female 14). There are those we advise and they accept. The advice we are likely to give is that people who have sex might get pregnant or get HIV/AIDS (Female 13). When LC members were asked why others had left the group, the two main reasons presented were pregnancy and poverty. In their statements, LC members made connections between poverty and early pregnancies and recounted how girls would either opt to get pregnant and then get married or the other way around, which bears testimony to early pregnancy as a sort of marriage strategy. Girls who attend school and become pregnant are legally entitled to return to school after giving birth, but many young mothers never return. A common reason is of course that they have to take care of their newborn child, but we were also informed of cases where girls were not allowed to come back to school because they were understood to be bad examples to other students, or were sent to another school further away, rendering them unable to attend. Such penalties were not applied in the cases of very young fathers, which reflects a gender divide.

Multiple partners and sexual assault
Throughout the interviews, nearly all participants talked about adultery, which was categorised as bad behaviour. This was most clearly demonstrated by the frequent use by both male and female participants of the phrase "to commit adultery". To have sex with multiple partners was also connected to the risk of becoming infected with HIV, as explained by a 14-year-old boy: One of my friends, a boy, was having sex with a lot of girls, so I told him it was no good. You can get a disease. The boy has now stopped. It was not only perceived to be wrong to have multiple sexual partners; it was also important to take into account the characteristics of the person with whom one was having sex. In particular, this pertained to the girls who should not have sex with older men or with men who offered them money. These points are aptly captured in the following interview extracts: One time the ADRA staff came and asked us which students were having sex and where they did it. We told them that they had sex with the motorcyclist, hairdressers and bicycle taxi riders. The girls got very angry and told us that they would beat us up, so we got scared (Male 12). They advise us to stop misbehaving. Teachers also advise girls to stop going out with soldiers (Female 15). A lot of things have changed. Girls could go into a salon and have their hair cut for free, but the fourth time she would have to sleep with the guy. The guys on motorbikes used to be driving around and when they saw a girl, they would stop and offer her a lift. Many girls would not go where they were supposed to, but stop somewhere else and have sex (Male 14). None of the participants referred to their own experiences from a first person perspective, but they frequently talked in the third person perspective about others who, for instance, had sex for money or who had suffered sexual abuse. It remains uncertain whether some of the participants may in fact have been talking about themselves, which ADRA staff underscored could well be the case. This observation may not only pertain to this particular topic, but is probably true of the other issues discussed in the interviews as well. From conversations with staff, it was evident that sexual assaults are common occurrences in the communities where this survey was conducted, which was a source of great distress to the staff. They moreover informed us that LC participants had suffered sexual abuse which in some cases had resulted in unwanted pregnancies. This was also reported in the baseline survey conducted by ADRA: "There is no doubt that sexual abuse of children is present in both Makamba and Cibitoke. Almost all pupils could tell stories about older men approaching younger girls either forcing them to have sex or tempting them with money, sweets or clothes". 28 It seems that the assaults are typically committed by members of the extended family or teachers or other local authorities, which is also documented elsewhere in Burundi. 5,30 A study by Amnesty International indicates that more than 63 per cent of rapes in Burundi are committed by neighbors or close relatives (2007:11). One interviewee from the Cibitoke province mentioned that "one teacher has been sleeping with students" and another interviewee from the Makamba province elaborated upon a similar occurrence about a teacher who impregnated a student with dire consequences for the girl: A P6 teacher got a student pregnant. He was forced to marry the girl but left her with his family. Now he is teaching P2. The girl is staying with her family in law, but they are mistreating her and accusing her of all sorts of things. He is refusing to live with her even though he paid her family. The bad attitudes had stopped, but started again this summer and now they are increasing (Female 11). Two male participants also mentioned sexual assault when they explained their tasks as LC members: There are people who get drunk at night and they force others to have sex. We keep advising them (Male 12) I advised one guy who use to rape children and he changed (Male 16) Sexual atrocities loomed large in the armed conflicts of Burundi over the past two decades, and high levels of sexual violence continue to be reported in the country, although sexual violence often go underreported in local communities. 5,[30][31][32][33] This issue also is widely documented in other countries in Sub-Saharan African that have recently suffered from armed conflict, such as Sierra Leone and Côte d'Ivoire, where prosecution of sexual violence is rare although appropriate legal frameworks are in place. 30,33,34 Local ADRA staff expressed great concern and frustration over the ease with which the violators can bribe the juridical system as well as the families to avoid trial and subsequent prison sentences for their crimes. The staff mentioned cases in which girls as young as threeyears old had been raped. In spite of the fact that the perpetrator had been caught, he had apparently succeeded in bribing the police, the judges and the family and therefore continued to walk around freely in the community. As indicated by the account of the student who was raped by her teacher, it is a common practice to use an informal dispute resolution system whereby the victim's family negotiates with the perpetrator for some form of compensation. In other instances, they agree that she marries the perpetrator as a means of providing her with both social and economic support. 33 What emerges from the material presented in this section is that sex and sexuality are only introduced and talked about as something negative and dangerous. Any positive connotations that sex could have are completely omitted, which implies that sexual abstinence is the sole solution for a young person in order to avoid such dangers. This gloomy perspective that undermines the participants as sexual beings is further underlined by the local perception of space where the street is understood to be closely linked to the dangers of sex and bad behaviour.

Bad behaviours and the street
A final aspect of misbehavior that we wish to outline here are the activities associated with the public space of the street that is seen to be marked by an absence of respect and morality. Throughout the interviews, the participants talked about how hanging out in the street can lead to a lot of "bad things" and for some of the participants this was connected to the influence of "bad friends" and socializing with "bad groups". Instead of spending time in the street, both boys and girls should attend school and "help parents in home activities" as it was often stated. Some participants referred to their own experience of hanging out in the street such as one 14-year old male participant who explained: "I used to misbehave. I could go in the streets after school, now I go straight home". Others talked about how they try to influence peers who are not coming to school such as these two 13-year-old boys: Many children spend the day in the streets, not going to school, but we continue advising them to go back to school. We advise other children, who spend their time in the streets, because it might have bad consequences for them. Before I used to beat up my young brothers and sisters, but now I stopped. In the neighborhood children used to spend a lot of time in the streets, but I advised them and they stopped. In the accounts of the participants, the street is equated with bad behaviour such as drinking or making money as opposed to going to school. This is well captured in the following statement made by a 15-year-old female participant who was tempted to leave school before she joined the LC due to financial hardship at home: I was kind of influenced by the students who left school to go for money. The LC encourages me to focus on school and forget all other things. The street offers young girls an opportunity to gain access to petty cash through sexual relations with older men. This is evident in the disheartening statement by another 15-year-old girl: At one point in my life I thought about becoming a prostitute because I'm an orphan, so I needed the money. If I hadn't joined the LC, I would have left school a long time ago. In this manner, the street has connotations of bad morals and dangerous sexuality. For the participants this also encompassed the screening of "bad movies" in local shops. As a concerned participant explained: We advise other children, who spend their time in the streets, because it might have bad consequences for them. We also try to explain to children that they shouldn't watch bad movies. (Male 13) Some participants who brought up these "bad movies about people having sex" mentioned that they had stopped watching them after joining the LC, such as this 14-year-old boy: "I used to watch bad movies but I stopped because of the LC teachings". The issue of pornographic movies was by and large brought up by male participants. Apparently, the screening of porn movies takes place after school hours in small shops where a VCR or DVD player and television serve as a small cinema. According to ADRA staff, the pornographic movies are illicit and mainly come from Europe. There is no minimum age stipulating who can watch (or not watch) these movies, and to do so has become common among youths in rural Burundi as young as 10years old. Research from Western Kenya documents a similar trend where pornographic videos are screened in small shops for a minute fee with minors present in the audience. 35,36 Likewise, Geissler and Prince report from Kenya that pornography is shown in battery-driven VCR-cinemas where: "youths of both sexes and many older people are now familiar with these kinds of images, which generally are referred to as showing 'bad things', [...], inverting the values of Christian morality and creating a new place for sex in the moral imagination." (2007:139). In keeping with the observations from the Kenyan context, we were informed that young people in Burundi generally watch these movies without their parents' approval and even steal money from the parents to be able to go, which was also discussed in a radio programme produced by ADRA with LC participants: Journalist: from where do young children get the money to pay the entrance in those cinemas?
Pupil: They may work for it or they ask parents by lying that there is something they go to buy whereas they go to watch the film. Some of the LC participants also condemned the parents for not knowing what their children are up to or for not disapproving of their children's behaviour. In a different radio programme, the same issue was brought up where the parents' point of view came to the fore: Parent: For a girl it is not accepted to go in cinemas, but parents don't see what they do because when they go they don't warn you. She moves for a moment and she tells you she goes to visit a friend whereas she goes to watch the films. But it is not normally allowed for girls to go in the cinemas. It is only because parents don't see what to do, as she is not a goat that we can tie up. Even when you tie up the goat it breaks the rope. The discourse among the participants was that when boys and girls watch these movies, it led them to "get involved in sexual intercourse", and for young girls to have sex with older men. Such accounts of girls engaging in different types of relationships with older men whom they meet at the cinemas are also documented from Western Kenya. 35 The cinemas were thus perceived to be places of promiscuity by the participants who take a clear moral stand against them. Yet the movies also seemed to generate a lot of curiosity and appeared as a temptation for the participants to abstain from. When asked what would make participants leave the LC, one male participant of 14years even mentioned that: "If I start feeling like going to watch bad movies then I would quit". Based on such statements it is fair to infer that the street is equated with a range of "bad behaviours". In order to prevent engaging in such behaviours, LC members saw it as crucial to avoid spending time in the street. For the male participant quoted above, failing to do so meant that he would lose his status as a role model. Actually, many LC members did not manage to "stop with their bad manners like having sex" and ended up leaving the group and "started hanging out in the streets instead". There were also accounts of peers who had left only to re-join on a later occasion, as well as stories of how participants encouraged those who had left to bring an end to their bad behaviour and return to the LC. When the discussion in the interviews turned to what would happen should the 'Breaking the Silence' project come to an end, some said that they had "learned enough" and would be able to continue with their "good behaviour" and others again stated that the extent to which they would be able to continue with the good manners would depend more heavily on "external influences". The majority of participants said that they or others would go back to or continue their "bad behaviour", or they claimed that their "bad behaviour would increase". Some had an even gloomier outlook on the future; such as a 15-year-old male participant who believed that the LC participants would no longer have a say in the community if ADRA left the project and therefore predicted that: "Many would go back to adultery and bad groups, they would no longer take time for the [LC] activities, and there would be many bad consequences".

Discussion
It has within contemporary childhood research and child-focused programming become common to view children as social actors who can provide a unique perspective on the social world about matters of concern to them. 37,38 However, as argued by Bhana (2009Bhana ( , 2010 in the body of work around sexuality and HIV and AIDS there has been little focus upon younger children "partly due to children's assumed incompetence around abstract sexual health issues" (2009:597). The findings from the interview-based survey in Burundi contribute with new knowledge concerning both children and youth and, in alignment with Bhana's argument, their acute knowledge "has developed through direct social and personal experience of communities affected and infected with HIV" (2009:601). Our findings show how children and youths are concerned with HIV/ AIDS and how they express a wish to avoid getting infected, but they also reveal that HIV/ AIDS is just one among many concerns.

Binary morality
First and foremost, failing to adhere to the restricted notions of "good behaviour" is seen to spoil their chances of achieving a better future. The efforts to promote abstinence were often coined in a language linked to visions of the future. A good example is the following statement made by a 12-year-old girl: "We are preparing for our future by being good students". This was also mentioned by other LC members, such as two male participants of 14years who expressed that they were going to have a better future by being role models and not getting involved in any sexual relationships because it could be "dangerous". While the abstinence message may serve them well at this particular point in time, they are, nevertheless, likely to become sexually active in the near future irrespective of the abstinence message. In their efforts to adhere to the advice of abstinence, the LC members end up subscribing to what 39 have coined a binary morality of right and wrong where sex is by definition dangerous and wrong. From a Southern African context, the social historians Delius and Glaser have argued that it was Christianity that infused sexuality with silence and shame, which they see as a possible contributor to the stigma associated with HIV/AIDS today (2005:33). Likewise, Bhana et al. reminds us that the emergence of Islam and Christianity in Southern Africa introduced views about sex that were moralistic and controlling (2007:132), and 40 has showed how it is particularly pronounced in Pentecostal Christianity and its public discourse on sexual morality. In his study among committed church goers in a Cape Town township, he found a clear discrepancy between the abstinence ideal and the young people's practices, and asserts that "premarital abstinence stands out as a possible but highly exceptional sexual style" (2011:680). We found that the LC members viewed sexuality as a moral category to be controlled and the ability to control it as a determinant of the future. Their appropriation of this sexual abstinence discourse was evidence of the ability to speak in a language governed by a binary morality, but they were not equipping themselves and others to take the necessary safety measures when it came to the practice of the sex that they eventually would have. The binary moral approach can be seen to help the LC members navigate in their everyday life where making choices between good and bad behaviour is an ongoing process of continuous effort. There is a real danger, however, that this critical potential of renegotiation gets lost if the knowledge-sharing becomes too steeped in a binary morality of right versus wrong. This issue has also been found in other studies from on youth in Sub-Saharan Africa. [41][42][43] In a study by among youth in rural KwaZulu/Natal, South Africa, the importance attached to "good behaviour" relating to sexual conduct was also highlighted. The engagement in sexual relations during the teenageyears and the entertainment of multiple partners were seen as the essence of misbehaving and therefore transgressed the accepted community norms (2008:181). Harrison found that this "moral framework" (2008:185) did not inhibit sexual relations among peers to take place, but rather it often made relationships hidden and unsafe sex a common practice. According to Harrison, abstinence is an ideology that places the youth at risk to the dangers they wish to avoid and by idealising abstinence, the absence of a constructive and positive discourse surrounding sexuality and prevention fail to prepare young people for sexual life. Research from rural Uganda on youth and sexual relations also found that the ideal of abstaining from sex before marriage caused much of this behaviour to take place in secret although here, it was more because of "adult disapproval and moral approbation" than an attitude widely shared among peers. 41,42 It is argued that a narrow focus on abstinence compromises the ability to respond effectively to the young's needs and is likely to increase their vulnerability to HIV and other sexual threats. Young people are agents in their own lives, and they are being agentic whether they engage in sexual relation to satisfy emotional or material needs. This should be the starting point from which to design youth-centred HIV prevention and Bell and Aggleton advocates for youth-led HIV prevention that does not discourage sex. 41 In addition, marital sex may also be risky in terms of acquiring HIV. This is widely documented among married heterosexual couples, and makes knowledge of safe sex practices crucial both before and after marriage. 44,45

Peer education
Our findings show that the participants attach a high value to the opportunity to influence their peers and even adults in the surrounding community. From a general perspective, peer education is promoted because it is believed that through peer-to-peer interaction, health messages can be delivered in an interactive environment in which the involvement of children and young people themselves should help bring about 'a critical renegotiation of young people's gender and sexual identities. 3 Viewed specifically from the 'Breaking the Silence' project, the participants clearly distinguished between right and wrong and thereby relied on a "binary morality" as discussed above. Peer educators everywhere are by and large equipped with the global 'safe/r sex' model that promotes abstinence, faithfulness and condom use, which is usually known as the ABC-model ('Abstain, Be faithful or Condomoze'). 9,46 In the present programme, however, the A in ABC had taken centre stage, while the B and C have not loomed large. The current analysis therefore suggests that LC membership may be more successful at producing moral superiority on the part of LC members than it is at inspiring safe sex practices among all. In simple terms, an equation takes place whereby HIV prevention becomes synonymous with sexual abstinence. Focusing too much on abstinence may thus be tricky because it evokes a passive imagery where abstinence sounds like an absence of actions, whereas it demands of the individual young person that he or she works hard on him-or herself to abstain from a range of actions and temptations as documented by the accounts presented in this article. In fact, the temptations seemed to be all around -caused by girls dressing improperly, men offering favours and money, and local cinemas screening porn videos. In this vein, the street becomes a symbol of the moral wrong as opposed to the places of the moral right; namely the school, the home and the church. In his book on Burundian refugees in a refugee camp in Tanzania, Turner shows how changes among the youth in the camp was also perceived to be a result of camp life resembling that of urban life. Especially the marketplace in the camp represented the virtues and vices of the city that threatened morality and Burundian customs. For some, it was a place they had to stay away from because it was full of temptations such as alcohol, "prostitution and promiscuity" (2010:74). There is a striking similarity between the discourse invoked by Burundians outside of Burundi in the refugee camps and the discourse found among the participants in the provinces of Cibitoke and Makamba as well as in the response to avoid the 'dangerous' public spaces. This might indicate that such a response to changes in sexuality and lifestyle more broadly among youth is easily perceived as a problem of morality that is caused by external changes to rural life. Yet, the rural and resource deprived setting in which the participants live seems to make the task of abstaining from sex to be much more than a question of morals. It is a setting where sex offers young girls the opportunity to get gifts, favours and money while pregnancy can lead to marriage, which may in some cases actually ensure that a girl is financially provided for 47 show in their study from rural Malawi that young women's engagement in transactional sexual relations is not only taken up by those who are desperately poor, but those who aspire to social mobility, economic independence, or simply a life enhanced by soap and lotions (2007:157). They also argue that transactional sex is not solely -and perhaps not even primarily -about sex or money, but has to do with establishing, maintaining, and sustaining ties that bind a man and a woman in a social relationship of unequal interdependence at the time of the exchange, a relationship that may be reactivated in the future. Finally, the ties that derive from the exchange of sex for money are just one form of patron-client interactions that are pervasive throughout Sub-Saharan Africa and that have proved to be resilient over time (ibid). The moral superiority assumed by the LC members can be seen to obscure the realities of their lives and the multifaceted changes of contemporary society. It also ignores the various ways in which children and young people carve out a path through precarious life conditions where it does not seem helpful to reduce all matters to a binary morality of abstinence versus sexual engagement.

Being a risk and being at risk
While we acknowledge that the abstinence message may hold out a path to the immediate future for children and some young people, the difficult socio-economic circumstances and the prevalent issues of sexual abuse and assault mean that it may be short sighted to promote abstinence as a safe-sex measure in the long run. To advocate a moral discourse of abstinence and bodily control is problematic given that the children and youths in reality are not always in control of their bodies. The school and home are conceived as spaces that are safe from bad behaviour, but as some of the accounts reveal these places also pose dangers such as from teachers or family members. In this regard, children and young people do not necessarily pose a risk to themselves. They are also at risk of becoming the victims of assault and exploitative behaviour committed by adults, which is equally important to address and should be dealt with by accountable stakeholders such as parents and community leaders. This aspect of sexual relations, where some are unequal and abusive, comes across as a very urgent and pertinent issue to address. This issue is also examined by 48 in a study from Durban, South Africa, exploring the meanings that young adults give to gender and sexuality in the context of HIV/AIDS. They argue that gender power relations therefore need to be addressed in HIV prevention and education to make it more effective, including the gender dynamics and power relations at school and in class. 44,45 In view of the widespread occurrence of unequal sexual relations and downright abuse in Burundi, it is not advisable to focus so narrowly on abstinence and to present it as the singular moral path and stepping stone to the future. We therefore take issue with what one might deem an over-emphasis on abstinence in the 'Breaking the Silence' project. We find that the B and C (be faithful or condomise) of international best practice are missing as an answer to how young people can envision a future where they can have an active sex life without becoming infected with HIV. Ideally, HIV prevention projects should adopt a more comprehensive approach and go beyond the abstinence message and technical aspects of how HIV can be contracted, avoided and tested. In calling for a more holistic approach to prevention programmes, 49 makes the case for an open approach to sex education in which sexual relations can both entail "positive aspects" and "negative outcomes" and where there is no blaming and shaming of youth in relation to their sexual feelings and desires. It should present a balanced view of sex as both enjoyable and risky, and youth should be introduced to "the whole spectrum of discourses, from disease to desire" (2010:315,317). While we agree that HIV prevention should be taught as part of a nuanced approach to sexual education, we also acknowledge the constraints of implementing such HIV prevention projects in contemporary rural Burundi. The constraints are especially prudent for faith-based organisations that work in a context of strong religious structures that influence the project agenda and prescribe what can be disclosed and discussed. In such environment, it would be wise to first attempt to negotiate the language of the binary moral discourse in order to encourage stakeholders to use a less stigmatising vocabulary in relation to sexual behaviour. In doing so, it would be useful to not only to address the ways in which children and young people pose a risk to themselves, but also the ways in which they are at risk. Taking into consideration that the 'Breaking the Silence' project included parents and teachers, the potential to reach the relevant stakeholders with this message was not fully utilised.

Conclusion
The empirical findings from this study show how the participants in the HIV prevention project make an effort to change their behaviour and be good examples to others by abstaining from premarital sex, spending time in the street, from watching porn movies, and from engaging in a range of other activities that are considered to be bad. We have suggested that LC membership produced a kind of moral superiority that set LC members apart from their peers and in this way, they ended up subscribing to a discourse characterised by a binary morality in which sex, and other behaviours that can lead to sex, were deemed to be dangerous and wrong. We have also raised the concern that the installation of a language of binary morality, in which sex is understood to be only dangerous, is neither likely to translate into safe sex practices for the participants themselves in the long run, nor for their peers in the short run. The interview participants were curious about sex, and among the boys this curiosity was especially revealed in relation to the illicit porn screenings. Among girls, financial benefits from sexual relations were highlighted. Discussing our findings with research on youth and sexuality in Sub-Saharan Africa, it becomes clear that although the abstinence message is employed as a solution across the African continent, young men and women are still likely to engage in sexual relations, and they have a variety of reasons and motivations for doing so. There is every reason to keep questioning the different versions of morality that are being promoted in international HIV/AIDS programming. When programmers and local and national stakeholders sensitise children to avoid dangerous behaviours, they may inherently install new notions of what sexuality is and can be. It is not simply a matter of carving out a trajectory towards a new and bright future; it may also be a powerful, restricting and limiting discourse that disregards the realities under which children and young people in Burundi and elsewhere are coming of age. Judging from the accounts offered by the participants in this study, the realities that are disregarded both entail their sexual desires and curiosities as well as their more or less voluntary and sometimes outright abusive sexual relations with adults.