Strategies of safety

When threats of violence become everyday life among South African adolescents

Marie Rosenkrantz Lindegaard, Ann-Karina Henriksen


In contemporary South Africa threats of violence are highly present, and the fear of violence manifests itself in various ways. The article presents an analytical framework for studying how Capetonians live with violence and establish a sense of safety as a part of everyday practice. It is argued that such studies must remain sensitive to the ways in which experiences with violence and practices of safety are embodied, as well as to the ways in which this embodied knowledge is reworked continually, shaping and being shaped by everyday practice. With particular attention to Capetonian adolescents, it is shown how differentiated bodies are ascribed meaning through bodily signs of class, gender and race. Bodily signs shape experiences, representations and practices of violence and safety. Violence does not need to take place to have an effect, the fear of violence alone conditions everyday practice. The study of safety places emphasis on the effects of violence and how violence influences processes of subjectivity.

South African adolescents grow up in one of the most violent societies in the world (UNDP 2002). In a survey, conducted in 2003 among 550 high school students, who differed in terms of class, gender and race, they were asked to mention two of the major worries of their everyday lives: five percent mentioned violence and crime; the vast majority referred to school work and friendships (Lindegaard & Henriksen 2004b).           

When we asked people how they experienced living with violence, they would often reply that they did not know what we were talking about - they did not experience any violence! If we turned the question around and asked what people did to stay safe, their reply would be long and detailed. Focusing on safety we brought attention to what people were doing to avoid violence, rather than how they experienced being exposed to it. Staying safe was explained as following ‘the rules’, which seemed to reflect social norms more than actual risks of violence. The way Capetonian adolescents practiced safety differed across class, gender and race.          

Our shift in perspective from violence to safety enables a different approach to violence. Establishing safety is not exclusively for people living with an extreme presence of violence. In places where violence rarely occurs, most people also have a feeling of what is safe and unsafe. It is not that we live our lives being aware of these risks, but we usually know when we put ourselves in potential danger. This mechanism – not being aware but anyway knowing - also applies to Capetonian adolescents. In the following sections we will show how theories of embodiment are useful to explain how people stay safe without being aware of it, and also how strategies of safety are embedded in social hierarchies and struggles about power.

Violence and uneven structures of power

The article is based on six months of fieldwork in Cape Town in the first part of 2003. We worked with adolescent boys and girls in the age between 15 and 30.[note 1]Adolescence is defined in terms of relationships rather than age. Around the age of 15, Capetonians move around on their own and are thus exposed to issues of danger and safety. They depend on their parents, and girls also on their boyfriends, who usually provide means of safety such as transport and protection. Women in their late twenties and early thirties, who live alone, also feel dependent on other people for protection, which makes their strategies of safety similar to younger boys and girls. In our analysis, adolescence is related to the way Capetonians live with the fear of violence.
Most of them characterised themselves as coloureds, some as whites and a few as blacks.[note 2]We did not come across any adolescents, who characterised themselves as ‘Indian’. There has existed a range of different racial categories in the history of South Africa. However, probing into this will be behind the scope of this article. Worth remarking is that in the official discourse there only exist two racial categories: African and white.
The majority lived in the poorest part of the city referred to as Cape Flats, which includes residential areas characterised by both shacks and more middle class[note 3]‘Class’ is used as conceptualised by Bourdieu (1984). When we refer to ‘middle class’ it is not defined in terms of economic resources but rather in terms of power. Middle class behaviour is characterised by a need to create distinction and thus emphasise superiority in relation to lower class.
houses with several rooms and small gardens; some lived in former [note 4]Former’ and ‘previously’ are used to emphasise that residential areas, schools and people no longer are officially categorised through certain racial constructions. Due to slow changes of race and class, former ‘black’, ‘coloured’ and ‘white’ areas are nevertheless still characterised by being inhabited by people formerly belonging to these categories.
white areas with large houses, private cars and alarm systems. We did the most intensive study among high school students (age 15 - 19) in Heideveld, which is a low-income, coloured area on the Cape Flats, and compared this with a smaller study among high school students in Pinelands, which is an affluent, predominantly white residential area close to the city centre.           

The presence of violence - both experienced and imagined - shapes the lives of Capetonians across race and class, which is why we worked with adolescents mixed in terms of race, class and gender. South African adolescents are frequently exposed to incidences like mugging, rape, domestic violence, shootouts, stabbing, armed robbery, housebreak-inn, car theft, etc. either on a personal level or through stories told by neighbours, friends, family or in the media. South Africa is renowned for its high crime rates with nearly 1.6 million annual cases of reported theft; over 50,000 official reports of murder and 54,000 officially reported cases of rape (SAPS 2004). However behind such statistics lies a highly uneven distribution of violence. Statistics on violence in the Western Cape, where Cape Town is situated, show that most violent crimes and homicides take place on the Cape Flats and most petty and property crime in the more affluent parts of the city previously referred to as white. Most crimes are committed by men; younger people are more exposed to violence than older people; men are mainly victims of violence outside of home; women are mainly exposed to violence at home; most of the violence women are exposed to can be defined as sexual (statistics quoted in Gibson 2003: 45ff & Jensen 1999: 77ff). Capetonians’ ways of living with all this violence reflect the distribution of violence. However they also reflect uneven structures of power between men and women, across racial groups and between different classes. As the analysis unfolds it will become clear that safety is also a lens to study social change in post apartheid South Africa, since it is through the cross cutting presence of violence that we can gain insight into how categories of race, class and gender are being reproduced, challenged and constructed anew.

A sense of safety in a violent setting

Most perceptions of what is considered safe and unsafe among Capetonian adolescents depend on time and place. Across race, class and gender adolescents generally agree that empty areas such as parks and streets are unsafe especially at night. Areas around train stations are mostly considered unsafe because people are often mugged there. Certain streets can be walked until dusk; it is safe to draw money from the cash machine at 4 PM, risky at 5 PM and unthinkable at 7 PM. Even the choice of clothes is to some extent informed by notions of safety, as a male roommate explained:

I never leave home in these sandals; I always wear my tackies [slang for trainers] so I can run when I have to.

Among adolescents it was common practice to phone their acquaintances when leaving or arriving at home. People with a car considered it common sense to drive around with locked doors, just as it is normal to run a red light in certain parts of the city at night. Many everyday practices like these were structured by fear of violence, but most of the time Capetonians were not aware about what they did to stay safe. Talking about violence was widespread but when asked directly it was common for participants to emphasise that it actually was not such a big issue. As a woman in her mid-twenties said:

I hope you go back to Europe and tell them that violence is actually not such a big issue here in South Africa.

In the following we will try to conceptualise the connection between talking a lot about violence and still insist that it is not really an issue in daily life.

Symbolic orders based on storytelling

Jensen (1999) argues that Capetonians cope with the threat of violence by creating systems of “symbolic orders”, which serve to predict the occurrence of violence, and symbolically contain it in time and space. Always driving around with the doors locked and dividing the city into go and no-go areas and streets were examples of practices, which form part of this order. These orders are ‘symbolic’ because they are based on representations of violence, which in Jensen’s analysis refers to exchanging stories about it (see also Feldman 1991). The recent violent incidences such as a break-in in the neighbourhood or a rape case presented on the news are everyday conversation subjects at dinner parties or in the changing room at the gym (Lindegaard & Henriksen 2004a). Jensen argues that these stories about violence make Capetonians aware of it in an explicit sense. Being aware brings about a feeling of control and thus a feeling of being able to predict and avoid violence. When stories of violence circulate in Cape Town, they serve the double purpose of informing about violence and making sense of violence. Jackson (2002) explores the link between violence and storytelling: “In making and telling stories we rework reality in order to make it bearable” (Jackson 2002: 17).

Most of the time symbolic orders are tacit and embodied knowledge, where avoiding certain streets or not driving alone at night become common sense and naturalised and thus rarely reflected upon. It is only when violence happens in the wrong place that people become aware of their symbolic order. Once the order has been renegotiated it resumes to being tacit knowledge. Exchanging stories is a main resource for the establishment, maintenance, negotiation and redefinition of symbolic orders.

Including bodily experiences and embodied knowledge

Jensen’s analytical framework (1999) helps us to understand people’s need to talk about violence and classify places and times as safe or dangerous. It provides a tool to analyse general patterns of creating safety among Capetonians. However, it does not provide insight into how social categories such as race, class and gender shape discourses and experiences of violence and safety. The main problem in Jensen’s approach seems to lie in his emphasis on discourses rather than lived experiences, and in his neglect to investigate the power structures underlying the production of discourses. If we wish to conceptualise how people can talk about violence yet not experience it as an issue in their daily lives, it is pertinent to supplement the discursive approach with an analytical focus on processes of embodiment. Distinguishing safe from unsafe is not an objective process, which solely reflects real dangers. It is also reflects power structures, and it is deeply embedded in constructions of subjectivity and negotiations of social categories and divisions. Thus when Capetonians define certain behaviour as dangerous, we need to ask: Dangerous to whom and in what way?

Representations and experiences of violence and safety are tied up with notions of femininity and masculinity, racial stereotypes and notions of class in adolescents’ everyday practices. Strategies of safety become part of their attempts to fit into established categories of race, class and gender, but also to transcend these categories and establish new subject positions. Symbolic orders, which are not only based on storytelling but also on bodily experiences and tacit knowledge, we refer to as strategies of safety.

Strategies of safety

We use the word ‘strategy’ in a way similar to how Bourdieu (1990) uses it: an act of awareness, which is nevertheless rarely well thought-out and reflected upon. Bourdieu acknowledges the potential confusion inherent in the word strategy, with its strong connotations to rational choice theory. However, strategy is used: “to designate the objectively oriented lines of action which social agents continually construct in and through practice.” (Bourdieu & Wacquant 1992: 129). Strategies of safety are rational in the sense that they make perfect sense to the agent but they are not necessarily articulated or well thought out. Staying inside is a strategy to avoid random violence, but while staying inside, people rarely consider that they do so to avoid random violence. What is at stake is mostly habitual practice informed by tacit knowledge of danger and safety.

The participants in our study struggled with articulating and making sense of experiences of bodily expression, which informed their strategies of safety but could not be accounted for in a straightforward way. Sometimes it was easily articulated, like a girl who said: “I don’t go to the park, because people get mugged”. She knows this because she has heard stories about it. But at other times people experienced a certain feeling; their legs started shaking or suddenly they started running down the street instead of walking as before. As a young woman, who had been mugged recently, described it:

When I walk down the street, my body reacts when I see a coloured man. It’s like, in my mind I’m not a racist, but my body is differently informed; it remembers.

A similar struggle can be identified within research on violence, when certain experiences are referred to as “instincts” (Feldman 1991: 77) or “being paranoid” (Taussig 2004: 173). By calling it instincts you indirectly indicate that it is an irrational biologically determined experience; it is different from knowing. The body takes over and decides for the mind. By naming it paranoia, Taussig doubts whether a guy followed by the security police in Columbia ‘really’ is in danger or just has become paranoid. The bodily expressions of fear are thus perceived as being based on something real or unreal. Both Feldman and Taussig make a distinction between knowledge of the mind and knowledge of the body. Knowing what, where and when it is safe or unsafe requires knowledge of both mind and body, because this is how knowledge is generated. We accordingly need to acknowledge “bodily information” (Winkler 1994: 259) and to recognise the body as socially informed, as “a field of perception and practice” (Csordas 1994: 34). Rather than approaching the body in a Foucauldian sense as a representation of the world as Feldman and Taussig do, the body must be perceived as being-in-the-world: “Representation is fundamentally nominal, and hence we can speak of ‘a representation’. Being-in-the-world is fundamentally conditional, and hence we must speak of ‘existence’ and ‘lived experience’”(Csordas 1994: 10). When a young woman in our study was asked why she crossed a street at a certain moment or why she had three knives in the drawer next to her bed, she replied: “because I am a little paranoid”. Paranoia was often the word used for such behaviour, which could not be explained verbally. Our argument is that such practices have to be acknowledged analytically.

The focus on representation tends to overlook that much of everyday practice happens without much articulation or awareness. Bourdieu uses the metaphor of the football player to account for how agents engage in practice. The player exists through playing: “And he does so ‘on the spot’, ‘in the twinkling of an eye’, ‘in the heat of the moment’, that is, in conditions which exclude distance, perspective, detachment and reflexion” (Bourdieu 1990: 82). Playing the game is entirely different from watching the game from the side. On the side, you can analyse, reflect and make informed and thought through evaluations, but on the field you have a ‘feel for the game’, which informs your dispositions - your strategies. Practice unfolds in time and the temporal structure of any practice is constitutive for its meaning. Bourdieu argues that many analytical misrepresentations “flow from the tendency to confuse the actor’s point of view with the spectator’s point of view” (Ibid.). In this light the ‘knowing subject’ becomes an analytical misrepresentation, because it perceives subjects as making daily decisions, which are informed and well thought through.

Knowledge about danger and safety is generated through discursive representations, as Jensen (1999) suggests, but also through bodily knowledge, as when you, for no specific reason, feel unsafe walking down the street; if you suddenly check whether your door is locked as you pull up at a red light. Most of the time Capetonian adolescents stay safe through practices, which they are not necessarily explicitly aware of and therefore do not give consideration to. By incorporating bodily knowledge in the analysis we become able to use strategies of safety as a lens to understand other social processes such as doing class, gender and race through strategies of safety. We refer to ‘doing’ because we wish to emphasise that people are not black or white/ female/male, low class/middle class. They are aspiring these categories or social positions through their practices (Butler 2000); they are in the process of becoming and strategies of safety reflect what positions they aspire to achieve or maintain.

Negotiating social positions through strategies of safety

So far we have argued that the notion of embodiment is crucial for understanding how people live with violence and establish a sense of safety. We have argued that this body of knowledge is under continual construction, being shaped by and in turn shaping everyday practices. We further need to analyse what agents do with this embodied knowledge, how it becomes part of their daily practices of negotiating social positions of race, class and gender. Strategies of safety are embodied but they are also motivated. A girl, in our study, who had a sexually mature body and lived in a poor coloured neighbourhood, knew she could not walk the street like other girls. Sexual violence frequently occurred in this area and girls with mature bodies experienced to be particular vulnerable to sexual harassment (Lindegaard & Henriksen 2004b). The girl with the mature body did not consciously think that she was different from other girls, yet she chose to stay inside. She had often been sexually harassed, which had made her uncomfortable and afraid. She described her body as becoming stiff when she was exposed to such innuendo and comments she was unsure whether she should run away or pretend she did not care. These very bodily experiences informed her practices of staying inside. If strategies of safety were equally necessary for all people, such daily negotiations might not be needed. Adolescents are positioned differently according to class, gender and race but also in terms of bodily build, family reputation, education and residential area. Their power to negotiate their strategies of safety reflects these different positions.

In order to capture how these processes are negotiated we will perceive the body as marked by signs, which become visible/invisible depending on context (Butler 2000). A woman becomes female by being ascribed the bodily sign of femaleness and it is the context, which provides the possibility of interpreting her behaviour as female. In some situations her sign of femaleness will be almost invisible and other signs like whiteness or middle class will be more exposed. Certain strategies of safety are available to the body depending on the interpretation of its signs. Feminine or masculine signs, black or white signs, signs of physical power, sexual openness, signs of social class and signs of vulnerability, motivate certain strategies of safety. We introduce this approach because it captures how embodiment is a process, where social structures are applied but also negotiated (Moore 1994). Approaching race, class and gender as bodily signs, which change meaning depending on contexts, helps us capture how adolescents communicate certain subject positions through their strategies of safety. Class, gender and race are not pre-existing categories, which determine strategies of safety; class, gender and race are constructed through strategies of safety; adolescents are doing class, gender and race when they strategise safety.

The division of the following into three separate sections about class, gender and race is misleading in the sense that the three categories come into play simultaneously, gaining significance “in and through each other” (McClintock 1995). However, we try to argue that some situations emphasise the femininity of a certain body, others make that characteristic less relevant. When a black man walks the streets of Cape Town in the evening, his blackness is emphasised. Lower class is stressed when people live in a house with an unlocked door; when they stay safe through social networks rather than technologies of safety such as alarms and high fences. Driving a car and having an alarm system emphasises whiteness. Adolescents confirm their relationship to their parents as a relationship of dependency by being in need of their protection. Young people’s position as receiving protection emphasises their lack of power in relation to their parents. Staying at home, being in need of protection and expressing vulnerability communicates femininity. The ability to be outside the house and in the streets, being mobile, providing protection and reacting with violence in violent situations communicates masculinity.

Doing class through strategies of safety

What was most striking in the beginning of our fieldwork was how economic resources played a role in relation to strategies of safety. Capetonian adolescents with little economic resources stayed safe by being informed about violence through their neighbours and friends living in the same residential area. They knew who was involved in gangsterism or other crime, and thus who might cause violent incidents. Staying safe meant avoiding certain houses, cars and people. People, who could afford security systems, had the fact that they lived in former white residential areas in common. They were not necessarily white but they shared the ability to afford a house in these areas.

People, who stayed safe through their social networks, mainly lived in former black and coloured residential areas. Because race and class are changing very slowly, living in what was formerly defined as a black or coloured area currently still implies that residents are perceived according to apartheid categories of being black and coloured. Strategies of safety can therefore be used as a lens to study social change in the sense that people’s ways to stay safe communicate affiliation of class. When black and coloured people invest in safety measures such as alarm systems and security buttons, they simultaneously communicate that they are changing their previously defined position as lower class. These changes can also be observed through practices of taste such as in people’s clothes, music, food, sport, films; ways of greeting, speaking, walking, having relationships, expressing sexuality, etc. In the following we will give a few examples of how strategies of safety can be used to study affiliation of class.

Middle class’ strategies of safety and the possibility of safely fitting in

Brian is a coloured boy, who lives in Pinelands and attends Pinelands high school, which is a former white model C [note 5]During apartheid schools for white learners were called Model C schools, coloured schools were called Model B and schools for blacks were called Model A. Today the terminology is still used but now in combination with ‘former’ or ‘previous’.
school nowadays highly racially mixed. After school he drives home with his mum and his presence outside is limited to the secure compound where he lives. He is driven everywhere and his house is highly secure. Emma is a white girl; she lives in the same area as Brian, attends the same school, and employs more or less the same strategies of safety. The difference between the two is how well their middle class strategies of safety fit with their other bodily signs such as male/female and coloured/white. When Emma is employing these strategies of safety her bodily signs of whiteness and middle class are emphasised. For Brian this is different. His body is coloured and the combination of being both coloured and a male in contemporary Cape Town involves strong connotations to being criminal and lower class (Lindegaard & Henriksen 2004b). Coloureds are stereotypically seen as behaving like gangsters, which means hanging out in the street; speaking loud using hand language and slang; wearing brand name clothes; using violence and being aggressive and provoking (Jensen 2000). Brian’s middle class strategies of safety counter this image. He stays out of the street; his appearance is quiet; he emphasises that he is strongly against violence and he does not like brand name clothes. His behaviour is not consciously calculating but rather based on tacit knowledge about how he safely fits both into groups of adolescents at school and in his residential area. He has experienced that people sometimes misinterpret him and take him for being stereotypically coloured:

I never use hand language like Kyle [a coloured classmate]. Then people would get the wrong impression. I am not like other coloureds.

Due to the signs ‘read onto’ Brian’s body, his position as middle class is more ambiguous than Emma’s and he therefore has to work harder on his appearance to convince his surroundings about this position. If Brian walked the street his middle class position would be threatened. In the following example a woman, white and middle class, manages to employ ‘lower class’ strategies of safety without being seen as lower class.

Middle class with ‘lower class’ strategies of safety

Sara is a white university student, who lives with her boyfriend in a former white area. After a violent robbery she decided that she wanted to start walking the street and live in a house without any security measures. As she expressed it:

We cannot continue living in our own small fortresses. People become paranoid when they stay inside all the time. We need to get out of our houses and get to know our neighbours. That is the best form of protection.

Sara proposes strategies of safety, which in fact are very similar to those employed by most black and coloured people who cannot afford anything else. The point here is that she does not risk her middle class position when she strategises safety in this way. If Brian would be doing the same thing he would not be able to maintain his middle class position. Due to her bodily signs of whiteness, Sara can walk the street in her neighbourhood without the risk of being perceived as poor or potentially criminal; people might think she is weird or a hippy and if she is exposed to violence most people will conclude that she “asked for it” and had brought it upon herself. Strategies of safety re-articulated with bodily signs such as being white and female, thus communicate what kind of middle class person Sara supposedly is. As her example also shows, deliberately choosing her own strategies can be problematic and might not only put one’s life in danger but also exclude one from being perceived as belonging to certain social groups (white women, who stay inside). Emma cannot afford to behave like Sara, which might have something to do with the fact that she still lives with her parents. She feels obliged to behave like a ‘good girl’, and that means staying inside, which thus is a means to please her parents’ expectations. Strategies of safety can thus be used to communicate constructions of gender, which is what we attend to in the following part.

Doing gender through strategies of safety

We identified certain gendered strategies of safety and gendered discourses of danger and safety across a range of different social contexts. Young women often expressed greater vulnerability than men and they generally felt unable to defend themselves in a violent situation, which was often explained by their weak physique compared to men. Women therefore stayed inside more often than men and needed male protection both in and outside the house to stay safe. Young men, on the other hand, expressed being able to defend themselves and often felt responsible for the safety of female friends and family members. Domestication seemed to be a feminine strategy of safety, whereas men experienced the need to protect themselves both in and outside the house. However, the relationship between gender and strategies of safety was twofold: women stayed at home because they felt vulnerable, but also because this strategy makes you a woman. It is a practice, which feminises, just as fighting back, providing protection and being present in public space are practices, which masculinise.

Strategies of safety, and by implication subject positions, are not equally accessible to all boys and girls. The ways adolescents manoeuvred and took up particular positions were influenced by a range of factors such as relations to family and friends, economic resources, racial markers and bodily signs of physical strength and vulnerability. Some boys employed ‘feminine’ strategies of safety without being perceived as girls, and some girls were able to move around ‘like boys’ without being excluded from being feminine.

Boys with ‘feminine’ strategies of safety

Staying at home as a strategy of safety and expressing fear of violence was not uncommon among boys. However, the consequences for such behaviour were highly dependent on the person’s bodily signs. Most coloured boys had to balance on the fine line between being perceived as moffies (derogatory for gay/weak man) or gangsters (criminals, macho ideal), which will be exemplified by the following two boys.

Wahid is 17 years old and lives in the coloured residential area of about officially 18.000 inhabitants, Heideveld (Jensen 2001). When asked how he avoids violence in the area (on average two boys were killed per week during our fieldwork), he replied in the following way:

I stay in the house, but like (…) I find pleasure in studying. If I study, I stay out of problems (…) like being killed and that stuff. But sometimes, when I walk to the library to do a project or so, I sometimes feel threatened, that is why I always ask someone to walk with me. Walking with someone will not stop the gangsters from killing us, but I will feel safer if I am building up a conversation with someone, so I normally walk with someone. And that is mainly during the day. When I walk alone at night, I usually run! I do not walk [laughs]. I am afraid of what might happen.

Wahid’s main strategy of safety is to stay at home as much as possible and walk with someone if he goes outside. These strategies resemble the way women stay safe, and Wahid also admits that some people see him as a moffie. Wahid’s bodily signs of being male nevertheless communicate violence, and situated in public, he is therefore at risk of being involved in violence. In Heideveld adolescent boys must either join the gang residing in their area or effectively convince the gang that they do not challenge their claim to public space. By displaying fear of violence, bodily expressed by running and by staying at home as much as possible, Wahid communicates being closed to violence. It is not all boys who need to be as feminised as Wahid in order to be safe, which the following boy, Haroon, is an example of.

Haroon is 19 years old, has left high school recently and is presently doing a degree at the Cape Technicon.[note 6]The Cape Technicon is a higher learning facility in Cape Town, providing education on BA level.
He is the oldest brother of four and lives with his parents in Heideveld. His response to avoiding violence is very different to Wahid’s:

I don’t really think about it, like I go more or less where I want to during the day and use the car if I need to go out at night. People know me in the area. They know I am a good guy. They would never touch me knowing my family, it’s the same with my sisters, not that they walk around alone, but people know us.

Haroon did not express fear of being caught up in gangsterism as Wahid did. In fact one of Haroon’s close friends from primary school was a high profile member of the biggest gang in the area at the time (the Americans); his friend was a frequent guest at Haroon’s house, which would usually be considered risky due to possible association with one of the gangs and thereby eventually forced involvement in gang fights. Haroon explains that he stays safe because people know him in the area. Haroon’s different strategies of safety explicate that he is differently positioned than Wahid, and the two do not take up similar subject positions with equal ease. Haroon lives in the part of Heideveld with one-storey, privately owned houses and gardens. His family is widely respected in Heideveld, especially due to both of his parents being politically involved in the community, and being respected and active in the Muslim congregation. Haroon himself is also a practising Muslim. In comparison Wahid lives in an apartment in the part of Heideveld consisting of communally owned courts. His father no longer lives with the family; his mother works hard with her catering business, and his uncle is involved in gangster-related activities. His two cousins dropped out of school at the age of 16 because they fell pregnant with men they were not married to. Wahid wants to be a lawyer, which is why he works so hard at his studies. Wahid and Haroon both have bodily signs of being coloured and male, they live in the same residential area and have attended the same school. However, their family relations and living standards give them different possibilities of staying safe and also different possibilities of being masculinised.

Haroon managed to balance the fine line between being a gangster and being a moffie; Wahid had to work harder on his appearance to avoid falling in as one or the other. Fulfilling masculine ideals of being protectors was difficult for both Wahid and Haroon due to the stereotypical notions of coloured men. When a coloured man uses violence as protection, it can readily be misinterpreted as aggressive use of violence; he therefore has difficulties being protective without being perceived as a criminal violator. A white boy’s description of a reaction to an attempted break-inn shows that being positioned as white gives more possibilities of using violence without being perceived as a criminal.

Peter is 17 years old and lives in Pinelands. He just returned home from school, and he is alone in the house when he hears sounds of people in the back yard. He explains that he immediately suspected it were burglars, and therefore he got up from the couch and called out loud. Then he phoned the neighbours and security guards; went through the house assisted by his dogs and a baseball bat, met up with the neighbours outside the house where they searched the area for the burglars who were now on the run. His response stressed getting angry rather than afraid, and reacting rather than becoming passive. In comparison, Peter described his mother’s reaction as: “She is paranoid for weeks”, which resulted in weeks with drawn curtains, hermetically closed windows, constant check ups for locked doors and the installation of an improved alarm system. His story communicated that he perceived himself as the successful protector of the home. It was a story about masculinity. Unlike the coloured boys participating in our research, he did not at any point consider it necessary to stress that this incidence did not make him prone to violence. Without exception the coloured boys participating in our research all emphasised that their abilities as protectors did not mean they were also violators. Fighting a burglar in a coloured area usually means fighting a gangster, and such fights eventually make you ‘turn out a gangster’. However, not fighting a burglar is considered feminine and turns you into a moffie.

Living with the threat of violence is a difficult task for coloured boys; coloured men are discursively connected to violence, which makes it difficult for them to be both masculine and non-violent. Peter’s bodily signs of being middle class and white give him the power to use violence, without being considered a criminal, but he can also be a soft guy, without necessarily being seen as a moffie. Wahid cannot do gender is these different ways without being categorised as either a moffie or gangster. Haroon can move around more freely than Wahid but if he uses violence like Peter he would immediately fall into the disgrace of being a criminal. Strategies of safety reflect social hierarchies and power and they communicate how boys are positioned and masculinised differently. Among girls the question was not whether they could use violence or not; their major concern was whether they could maintain a position as presumably sexually inexperienced and thus respectable ‘good girls’.

Girls with ‘masculine’ strategies of safety

Hanging out in the street was among girls considered dangerous, partly because of violence, partly because “it shows that you are open to all kinds of stuff”, as a girl expressed it. Being ‘open’ implies being sexually accessible. Girls stayed inside not only to stay safe, but also to communicate that they were ‘respectable’ and desexualised girls. The notion of respectability is central for the strategies of safety expressed by adolescent girls in our study. However, respectability was not a position equally accessible to all girls, and some girls were positioned to negotiate the extent of being respectable without jeopardising their mobility and safety.

Yasmina is 17 years old and lives in Kensington, which is a middle class, coloured residential area. Her parents divorced a few years ago, so now she lives with her mother and younger brother. She never walks alone and her mum drives her where she needs to go. She explains that as a Muslim girl she is not supposed to go anywhere without male company, but since her mother trusts her, she was often allowed to go out with a group of friends. Protection against violence implies being protected and closed off from dangers in the street but it also means being closed off from sexual penetration. South Africa, and in particular Cape Town, has some of the highest figures of sexual violence in the world (SAPS 2004) and the female body is constructed as vulnerable and open to both sexual and violent penetration. Women are therefore in need of protection, which can close them off from outside penetration: “Women are often raised to believe that their selfhood and self-worth lies in the control of physical access to their bodies” (Dorkin 1979 in Gibson 2003: 44). Good girls are not accessible: they are closed off partly through embodied notions of respectability and through the help of protectors (Lindisfarne 1994: 88). This relates to the expression of being a ‘virgin to crime’, which was widely used among the participants in our study, to designate people who have not been exposed to violence. The metaphor confers notions of being untouched and pure, of victimisation as essentially feminine and notions of being ‘closed’ to crime as a virgin is ‘closed’ to sexual penetration.

Yasmina perceived safety as a family issue: her mother provides for her safety by being “over-protective”. She was never out late and was not allowed to attend private house parties. She considered it as “a matter of being accommodating, we must all help to make my mother look like she is doing well raising us on her own”. Yasmina was acutely aware that her behaviour reflected on her family, which was particularly sensitive in relation to having boyfriends. Yasmina had a boyfriend and suspected that her mother knew, but they never discussed it. However, in relation to speaking about the boyfriend she stressed, that she had strong views on premarital sex:

Morals and I’m too young, I don’t need that [premarital sex], I have a lot of dignity and pride, why would I want to go that far and spoil my life (…) like depending on how far I went. I would definitely not have sex. But like seeing an ex-boyfriend walking with my husband and knowing that I went so far with him (…) uh, I think it is very sacred.

Yasmina taps into the discourse of respectability, where sexuality is perceived as contaminating for the respectability and moral value of a woman. In contemporary South Africa coloured and black women are unrespectable by default; they need to make themselves respectable through certain behaviour, whereas white women are respectable per se (Salo 2003). White women can more readily behave in a sleg way (wear short skirts, have boyfriends and pre-marital sex, walk around alone) and still be ordentlik (good women, housewives, caring) due to their position as whites. Yasmina has a lot of “dignity and pride” and her view on sexuality seems to be qualifying for receiving trust from her mother. Among adolescent girls parental trust was often contingent on parents being convinced that their daughters were not sexually active and that the immorality of premarital sex was embodied to the extent of self-control and self-surveillance among the girls. As the following example shows, not all girls are positioned to negotiate this amount of mobility and safety by investing in respectability as Yasmina did.

Shireen is 16 years old and lives in Heideveld. She stays in one of the poorest parts also referred to as the courts, in a two-room apartment, with her mother, grandpa, three sisters and brother. She rarely sees her father, who she says uses drugs and lives by the station, which is the more dangerous part of Heideveld. She explains that she spends a lot of time at home, unlike previously:

We were a group of friends hanging out at each other’s houses. We were outside, hanging out on the corners and stuff. We were just sitting there with guys at first that made us bad in Heideveld, because when you stand on the corner, most people start saying bad things about you.

Shireen struggled with rumours and other people’s perception of her. Staying inside is partly a strategy of safety but also a way of communicating that she is a respectable girl. Hanging out in the street not only communicates openness to violence but also sexual openness. Considering these rumours designating Shireen as a woman of the street, there is no space for a boyfriend. Shireen does not have the power to communicate that she is both respectable and sexually open:

My mum brags about me not having a boyfriend, always staying home, never having friends over and she buys stuff to keep me happy, like clothes and jewellery, like they are rewarding me for not having a boyfriend, but dressing me up for one [laughs].

Her mother assists in communicating that Shireen is sexually closed, since her reputation reflects on the reputation of the entire family (and vice versa). The family is dependent on their networks in times of scarce resources but also for staying informed about violent activities; they are networks of safety in a broad sense. Shireen is struggling to prove the respectability of her and her family, but she feels unable to control her appearance:

They can see if you are a virgin - like if your bum is loose, then you are not a virgin. Tight bums and small breasts are virgins […] they [the guys] will say you are not a virgin, even the girls say that. The girls will just watch you from behind, and say you are not a virgin anymore, but I don’t know how they know that.

A loose bum and big breasts, essentially a matured body, symbolises that a girl is sexually open. Shireen has a feeling that her body is made unrespectable; that her body determines her respectability and exposes her as immoral.Virginity is not only about behaviour but also about bodily expression. Shireen’s description of the bodily markers of virginity can be captured by what Bourdieu refers to as “hexis”:

Bodily hexis […] is assumed to express the ‘deep being’, the true ‘nature’ of the ‘person’, in accordance with the postulate of the correspondence between the ‘physical’ and the ‘moral’ which gives rise to the practical or rationalised knowledge whereby ‘psychological’ or ‘moral’ properties are associated with bodily or physiognomic indices (Bourdieu 2001: 64).

Bourdieu’s concept of hexis implies that no matter how Shireen behaves, her bodily hexis and thus her embodied accumulated experiences (habitus) will expose her as being a bad girl. In Bourdieu’s conceptual framework, social structures, such as practices of class, sexuality, gender, race, family relations, which structure her behaviour, are incorporated to an extent, which makes it impossible for her to do anything but behave the way they determine. What Bourdieu seems to oversee is that Shireen also has embodied the reactions she gets from her behaviour and her behaviour is therefore also adapted to these reactions. By staying out of the street she negotiates her bodily signs and this adapted behaviour makes her less vulnerable and open to sexual and violent penetration. Sometimes she anyway has to leave the house, which exposes her vulnerability. This is particularly influential in relation to sexual violence:

I do not know about other girls but I do not like when there are taxis driving past and all of them say that your are a cheap girl, that you like what the boys are doing to you. I can see girls do not like it. Most of them are laughing while they do it […] Like sometimes they push you in a corner and touch you, keep your hands tight and I do not like that. It happens many times, like a guy living in Catrina Court, he likes pushing the girls against the wall. I get angry and push but he is too strong for me […] Afterwards he just leaves you and laughs. There is nothing you can do or say. There is a lot of sexual harassment like: “your bum is right for me” and when you shout at them, they all just laugh. None of them say: “that child is too young”. They all just laugh. Everyone just laughs and then the friends join them.

Shireen’s concern is expressed in terms of what people will say about her, rather than her own feelings of being sexually harassed. As mentioned earlier, Shireen used to hang out in the street, “which made her bad in Heideveld”. This reputation makes her negotiating power weak in relation to convincing her surroundings about her respectability. She has in other words difficulties resisting the disgrace of sexual violence; instead of seeing her as a victim of harassment her surroundings will, due to her position, see it as a sign of sexual promiscuity. She is struggling against being recognised as sexual and unrespectable and this limits her options. Skeggs contends that: “the middle class have far more alternatives to how they can be.” (1997: 91), which becomes clear through a comparison to Yasmina, who is less domesticated than Shireen and who has a boyfriend, yet remains respectable. Certain clothes or accents cannot camouflage Shireen’s being, her presumed sexuality. She lacks the power needed for co-articulating her racially and morally constructed body.

The extent to which girls are able to legitimate their position as respectable, can be ranked according to certain social categories: white girls are always respectable, coloured and black girls with good family relations, economic resources and good education are much stronger positioned than girls from poor backgrounds. Strategies of safety are reflected by these hierarchies but the hierarchies are also maintained through the strategies. Some girls need to stay at home to emphasise that they are respectable; when stronger positioned girls move around like boys they emphasise that they have the power to behave like sleg women but anyway are respectable. These negotiations are complicated and do not only depend on skin colour, class and gender but also on build, maturity, clothes, accent, family relations, friends and how certain situations develop. Strategies of safety have to be situated within the social, as the social is also a part of the strategies.

Doing race through strategies of safety

Within recent years the intersection of race and class, as it was constructed and maintained through apartheid ideology and law, has been changing, and connotations of race and class are being renegotiated. Discourses of violence and safety are part and parcel of identity politics, which take place in contemporary South Africa in relation to race and class. Living in a house with alarm systems, driving a locked car and being afraid of walking the streets, are ways of ‘doing whiteness’. Domestication is part and parcel of doing femininity. Safety through social networks emphasises colouredness or blackness. But in between these rigid categorisations there are people crossing social borders and communicating alternative subjectivities through their strategies of safety. Such persons illuminate social categories precisely because they do not fit readily into given social categories. In fact they deliberately attempt to define themselves differently and even redefine social categories through their practices of safety and discourses of violence.

During our fieldwork we came across a range of different people, who could be defined as “border subjects” (Staunæs 2004). We met a young black girl living in the more affluent part of Cape Flats and attending a high (model C) school in Pinelands. She felt unsafe when using public transport:

My mother takes me everywhere I need to go, because I don’t really feel safe on the taxi or trains with all those people and stuff. But when I say this to people where I come from, some call me a whitie, like they say I’m a coconut.

A ‘coconut’ is a derogatory term for a person who is black on the outside, but white on the inside. It is usually applied to people who challenge the intersection between race and class, where persons of colour are no longer also lower class. Her strategy of safety did not match her perceived racial category, and she was therefore defined as belonging to the ambivalent definition of a coconut. We did not come across any common slang word for whites behaving ‘black’, which might have something to do with whiteness being a category of power (Frankenberg 1993). When whites behave in ways inappropriate to what is considered ‘white’, they do not put the reputation of whites as a whole at risk. In the following example a group of white women challenge what could be defined as ‘white’ strategies of safety, which nevertheless does not make them less white.           

Among a group of white young feminists living in Observatory, politics of safety became politics of identity. They often discussed safety and violence; whether they should take up self-defence, carry a weapon or mobilise their community to get acquainted in order to enable them to watch out for each other. By expressing distress with their domestication, the empty streets and alienated neighbours of their residential area they simultaneously rebelled against being positioned as fragile, middle class, white women. Their discussions of alternative strategies of safety were at the same time attempts to communicate new meanings of a range of social categories such as race, class and gender. At a dinner party with this group of women we met a young coloured man. As everyone was leaving in separate cars, he started walking down the street - alone and on foot. He refused to take a lift and to our concern about his safety he laughed and asked whether we really thought anyone would hurt him: “a coloured man almost two meters tall? If anything they are afraid of me!” His strategy of safety emphasised masculinity and colouredness, which may have been particularly important to him after a dinner party with twelve middle class, white women in a previously white residential area.           

Some boys in our study were embodying both white and black strategies of safety depending on the situation they were in. When Lucas, who was a black, 17 year old boy living in an area previously defined as black and who attends a school in previously white area, was hanging out with friends in his residential area, he would also walk around in the street in the evening. He knew the gangs ruling the area and he was himself a part of a boy gang. They behaved macho like the gangsters, but were different from them because they did not commit any crimes. When he, on the other hand, hung out with class mates, who were living in an area previously defined as white, he would embody their strategies of safety in the sense that he would only hang out in the street during the day. He was usually transported by car and he underplayed his gang affiliation in such situations. These changes of both strategies of safety and masculinity were also present in his accent. Among boys in his residential area he spoke Xhosa or English with a strong Cape Flats accent; among his friends in the more affluent area he would switch into British English (model C school) accent. His accent thus symbolised his shifts of identity, which he claimed was not conscious or calculating but rather natural shifts just happening when he felt it was necessary.  

Strategies of safety are used to emphasise social categories but also to challenge the meaning of them; they are used to define senses of belonging and thus to disturb or create order in categories of class, gender and race.

Perspectives on safety

Safety is not only about creating order in relation to violence; it is also about inclusions and exclusions of social groups; it is a lens to explore how people position themselves in relation to each other. Through the lens of safety it becomes possible to identify the myriad of ways in which violence, experienced or imagined, affects everyday practice and conditions social relations. It becomes clear that the power of violence not only lies in the act in itself but also in the disturbing effect, which the fear of violence has on witnesses’ ability to establish a sense of safety. Studying violence through the lens of safety adds a dynamic element to the analysis of victims and perpetrators, because strategies of safety are about manoeuvring and negotiation rather than breakdown and victimisation.

Our research has emphasised that the study of violence and safety needs to stay sensitive to processes of embodiment and the way bodily signs shape lived experiences. Perceptions of safety and danger are developed through storytelling and stories are ascribed meaning through the body. Positions such as being black, female, highly educated, hard working, having a matured body, strong family relations etc. influence how adolescents take up certain positions as part and parcel of their strategies of safety. Across race and class in South Africa, it is perceived as sensible behaviour to keep girls inside as much as possible, which does not apply to boys. Coloured boys need to employ different strategies of safety than white boys due to racial prejudices about them being inherently violent. The presence of violence is therefore not only a potential risk for life; in South Africa it also reproduces patriarchal gender roles and racism. Our analysis shows how social categories of apartheid are reproduced through concerns of safety, but also how agents manoeuvre around these categories, challenge them and attempt to establish new social categories and divisions.

Research in the Netherlands and Great Britain has shown that adolescent girls are more structured in their movements around the city than boys: boys hang out in the street; girls have appointments and are less present outside (Youth Spaces 2003; Valentine et al 1998:17). Is that a consequence of certain gender ideals or is it related to forms of violence? Strategies of safety reflect positions of power. Safety as an analytical lens is therefore powerful to explore social boundaries and the maintenance of hierarchies.


Marie Rosenkrantz Lindegaard ( and Ann-Karina Henriksen ( both have a master degree in anthropology from University of Copenhagen, Denmark. Marie Rosenkrantz Lindegaard is now doing a PhD at the Amsterdam School for Social Science Research, University of Amsterdam, also being a research fellow at Leiden University Medical Centre. She is working with masculinity and violence among South African adolescents. Ann-Karina Henriksen is on maternity leave. We would like to thank the co-researchers, the participants in the symposium on violence, health and human rights in Amsterdam 2004 and the reviewers for giving highly valuable comments on the paper.


Bourdieu, P.
1984          Distinction: A social critique of the judgement of taste. London: Routledge.
1990          The logic of practice. Cambridge: Polity Press.
2001          Masculine domination. London: Polity Press.
Bourdieu, P. & L. Wacquant, 
1992          An invitation to reflexive sociology. Chicago: University of Chicago Press.
Butler, J.
2000              Agencies of style for a liminal subject. In: P. Gilroy et al. (eds.) Without guarantees. In honour of Stuart Hall. London: Verso, pp. 30-37.
Csordas, T.
1994          Introduction: The body as representation and being in the world. In: T. Csordas (ed.) Embodiment and experience: The existential ground for culture and self. Cambridge: Cambridge University Press, pp. 1-26.
Feldman, A.
1991          Formations of violence. Chicago: University of Chicago Press.
Frankenberg, R.
1993          White women, race matters. The social construction of whiteness. Minnesota: University of Minnesota Press.
Gibson, D.
2003          Rape, vulnerability and doubt. Medische Antropologie 15 (1): 43-64.
Jackson, M.
2002          The politics of storytelling. Copenhagen: Museum Tusculanum Forlag.
Jensen, S.
1999          Discourses of violence: Coping with violence on the Cape Flats. Social Dynamics 25(2): 75-97.
2000          Of drug dealers and street gangs: Power, mobility, and violence on the Cape Flats. Focaal 36: 24-46.
2001          Claiming community. Negotiating crime: State formation, neighbourhood and gangs in a Capetonian township. Roskilde: Roskilde University Press. Published PhD thesis.
Lindegaard, M.R. & A. Henriksen,
 2004a        Living with violence. A study on sexuality, safety and violence among adolescents in Cape Town. Tidskriftet Antropologi 46: 51-61.
2004b        Vulnerable women and men at risk. Safety and gendered experiences of violence among adolescents in Cape Town. Copenhagen: Institute of Anthropology, University of Copenhagen. Published MA Thesis.
Lindisfarne, N.
1994          Variant masculinities, variant virginities: rethinking ‘honour and shame’. In: A. Cornwall & N. Lindisfarne (eds.) Dislocating masculinities. Comparative ethnographies. London: Routledge, pp. 82-96.
McClintock, A.
1995          Imperial leather. Race, sex and gender in the colonial context. London: Routledge.
Moore, H.
1994              A passion for difference. Oxford: Polity Press.
Salo, E.
2003          Negotiating gender and personhood in the new South Africa: Adolescent women and gangsters in Manenberg township on the Cape Flats. European Journal of Cultural Studies 6(3): 345-65.
2004       Crime statistics:
Skeggs, B.
1997          Formations of class and gender: Becoming respectable. London: Sage.
Staunæs, D.
2004          Gender, ethnicity and school life [Køn, etnicitet og skoleliv]. Frederiksberg: Forlaget Samfundslitteratur.
Taussig, M.
2004          Talking terror. In: N. Scheper-Hughes & P. Bourgois (eds.) Violence in war and peace. Cornwall: Blackwell Publishing, pp. 171-74.
2002          Human Development Report, South Africa (
Valentine, G. et al.
1998          Cool places: Geographies of youth cultures. London: Routledge.
Winkler, C.
1994          Rape traumas: Contexts of meaning. In: T. Csordas (ed.) Embodiment and experience: The existential ground for culture and self. Cambridge: Cambridge University Press, pp. 248-68.
Youth spaces 
2003          Unpublished report from urban geography. Amsterdam: C.M. Kan-Instituut, University of Amsterdam.