The Public Accounts of a ‘Private’ Act: Domestic Violence in the Eyes of Mamelodi, a South African Township

Domestic violence against women is a serious social and public health problem facing women in South Africa and other countries. This social malaise in South Africa is often seen as “private” – committed within the home space, or “imperceptible” to the public. In other words, domestic violence is framed as a phenomenon that takes place exclusively behind the closed doors of the home. This study contends with this notion of “imperceptibility” in the domestic violence discourse. Using qualitative data from Mamelodi, a black township in Pretoria, South Africa, this paper argues that the notion of imperceptibility is reductionist and helps to perpetuate the act. The commission of domestic violence does not only take place in the privacy of the home; it is also committed in public, and it is visible to the community where it takes place.

vulnerable members of society (South Africa, 1998, p. 1). Since the introduction of this Act in 1998, domestic violence has continued to increase in South Africa. Form 1 of the Domestic Violence Act 116 of 1998 explains the complainant's rights and the steps a woman may take to protect herself, her children and/or other members of the shared household (South Africa, 1998, p. 1). This legislation (The Domestic Violence Act 116 of 1998) is based on the assumption that violence against women happens only at home. Hence, victims may apply the necessary steps for their protection from "domestic" violence. This notion which constructs domestic violence as occurring only within the confines of "the home" perpetuates and reinforces domestic violence in South Africa.
The conceptualisation of domestic violence as a "phenomenon behind closed doors" is common in South Africa. For instance, Bassadien and Hochfeld (2005, p. 5) noted that domestic violence in South Africa is seen as imperceptible because its commission occurs within the private sphere of the home to which women are relegated. In another research study on domestic violence in South Africa, Themistocleous (2008, p. 47) idealised the home as a haven for security and happiness, but unfortunately violence at home is also part of the experiences of millions of South Africans. Furthermore, Themistocleous (2008, pp. 47-48) identified certain factors that influence the commission of domestic violence within the home space in South Africa: a. A combination of emotional intensity and personal intimacy are the characteristics of family life. Family ties are normally charged with strong emotions, often mixing love and hate.
b. Much violence within the family is tolerated and even approved of.
c. There is a measure of social tolerance or approval of violence between spouses, almost as if marriage and parenthood provide a "licence to hit". A cultural acceptability of this form of domestic violence is expressed in the saying: "A woman, a horse and a hickory tree; the more you beat them the better they be." While the notion of "imperceptibility" seems to pervade the discourse on domestic violence in South Africa, it has equally been noted that the borders of this social problem have expanded to public places, such as shopping areas, markets, bars or taverns, the streets, and so on (Bassadien & Hochfeld, 2005, p. 5). This article agrees with the findings of Bassadien and Hochfeld (2005) that domestic violence in South Africa occurs beyond the walls of the family homes. However, while Bassadien and Hochfield (2005, p. 5) had focused on the wider South African society, in our study, we focused specifically on Mamelodi, a black township in Pretoria.

The Identity of the "Culture of Violence"
South Africa is alleged to have a "culture of violence"a label which dates back to the years of apartheid (Bowman, 2002, p. 857). Part of the blame for the high rate of domestic violence against women is that violence has become an "acceptable" way of resolving domestic squabbles even in the post-apartheid era (Bowman, 2002, p. 857). South African scholars also refer to domestic violence as a desire to exert power and control over women, which falls under the rubric of a "culturally" entrenched pattern in traditional communities (Bowman, 2002, p. 858). South Africa has communities where patriarchy is culturally entrenched, and where men exert power and control over women; this has arguably led to increased domestic violence in South Africa.
Although South Africa is no longer governed by the apartheid regime, violence in the post-apartheid era has mainly a gender coloration and has increased in frequency (European Commission, 2010). Gender in this context is defined as a social relation characterised by power inequalities that hierarchically produce, organise and evaluate masculinities and femininities through the contested but controlling practices of individuals, organisations and societies (Feree, 2010, p. 424;European Commission, 2010). Some statistics will illustrate the trajectory of domestic violence in South Africa that demonstrates the so-called "culture of violence". For instance, between 2008 and 2010 in Gauteng Province, about 15 307 cases of domestic violence were reported, out of which 12 093 cases involved females as victims (Machisa, 2011). Thorpe (2013, p. 1) views domestic violence as particularly hard to measure because the police in South Africa do not keep separate statistics on assault cases perpetrated by husbands or boyfriends. She adds that domestic violence statistics are 53 Caritus Mazibuko & I Umejesi -The public accounts almost impossible to access because domestic violence is not in itself a crime category, even though, according to the National Instructions (7/1999) relating to the implementation of the Domestic Violence Act 116 of 1998, all domestic violence incidents must be recorded in a Domestic Violence register in police stations (Thorpe, 2013, p. 1). In KwaZulu-Natal Province, about 7 244 cases of domestic assaults were reported in the first half of 2015 (Crime Stats SA, 2015). An earlier report painted an even grimmer picture in KwaZulu-Natal Province. The number of cases of domestic violence from September 2010 to December 2010, a four-month period, was 4 371 (Matthews, 2013, p. 5). Bendall's (2010, pp. 100-103) analysis shows that women in South Africa are predominantly under the control of men and simply accept their position as the victim; as a result, it is impossible to quantify the full extent of the problem of domestic violence and the statistics tend to underestimate the full extent as many victims do not come forward. This implies that there may be many unreported cases of domestic violence, especially in the rural communities.

Domestic Violence and 'The Locale'
In a study undertaken by Anderson (2010, pp. 731-732), it was found that there was a connection between violence and the place where people lived especially, where there exist entrenched structures of inequality. Domestic violence has been attributed to the association between concentrated poverty, weak external agents of social control, and weak social organisation in locations where most residents were poor, and possibly less able to create and maintain a community crime prevention drive (Miles-Doan, 1998;Van der Ende, Yount, Dynes, & Sibley, 2012, p. 1143. This is framed around the entrenched normative notion of the poverty-violence nexus. Location as an indicator of domestic violence is illustrated by a study conducted by Peberdy (in Molatlhwa, 2013) in various communities or townships in Gauteng Province in South Africa, namely Evaton, Sebokeng, Sharpeville, Boipatong, Bophelong and Tshepiso. Peberdy's findings (in Molatlhwa, 2013) revealed that a woman living in one of the above townships was likely to be beaten by her partner for the following reasons: going out without telling him, neglecting her duties of taking care of children, being argumentative, refusing him sex, burning food, or being unfaithful. This implies that women's rights are violated for many reasons. These findings were part of a survey which investigated male attitudes to domestic violence at local government level (Molatlhwa, 2013; see also Michalski, 2004, pp. 652-653). Miles-Doan (1998, p. 625) suggests that violence depends to a certain extent on interaction with others who live nearby and of whom individuals are most aware, and are therefore, most likely to be influenced by, and by the social context to which they are most frequently exposed. Mead's analysis (in Dlamini, 2004, p. 46) is that individuals initiate and direct their own action while at the same time being influenced by the attitudes and expectations of others in the form of the generalised other. This means that people tend to conform to the norms and values of the community in which they live, in order to feel a sense of belonging. The individual and the community are regarded as inseparable and the individual can only become a human being in a social context, a context in which she or he develops a sense of self, which is a prerequisite for thought (Dlamini, 2004, p. 46). Community and individuals are inseparable: without individuals a community cannot exist since a community is a result of human activities and individuals are the products of that community (Dlamini, 2004, p. 47).
Domestic violence is related to women's status, and in societies characterised by inequality, violence against women manifests as an intrinsic, pervasive facet of gender relations. From this vantage point, social disadvantage is regarded as a stress factor that reinforces traditional structures of male dominance that support violent behaviour arguably by men (Courtenay, 2000;Jewkes, 2002). This is the case in the South African context, where domestic violence is prevalent and tolerated to the extent that it has come to be widely accepted and perceived almost as normal. Kim andMotsei (2002, p. 1245) found that men's attitudes and beliefs regarding physical abuse were described using terms such as "discipline" or "punishment". Men felt they were "justified" in beating women when they did not listen or when they stood up for their rights Kim andMotsei (2002, p. 1246 Campbell's (1990) analysis of urban South African families refers to the dominance of patriarchal structures as a "township ideology", within which women do not regard themselves as being oppressed. From this vantage point, a patriarchal community normalises authoritarian male rule and conditions people to accept group oppression as the natural order (hooks, 2000, p. 38). Campbell (1990, p. 14) further points out that "men's experience of racism and economic deprivation often causes a reactionary backlash within the familyto the detriment of womenrather than opening up the space for resistance to race and class oppression". Hattery (2009, p. 19) found that there is no specific socioeconomic profile of perpetrators. In South Africa men from all races, ethnicities, ages, levels of education and occupations can commit domestic violence. Tshoaedi (2008) argues that Campbell's (1990) notion of "township ideology" regards men as the undisputed heads of households and figures of authority, and African women as oblivious of their gender oppression and as lacking the potential to develop a gender consciousness or to fight against gender inequalities within their families. For Campbell (1990), the frustration experienced by men in the public domain is expressed in the private domain with women. Men take out their frustration they experience at work on their women partners at home. This implies that in the intersections of race, class and gender, women may be subjected to compounded oppression. These women are aware of their lower status and power. In more detail, Campbell (1990, p. 15) notes: "The commitment of working-class township fathers to the traditional role of father as ultimate authority in the family must be seen within the context of the contradiction between their position in the work-place and their childhood socialisation within a patriarchal social order. This socialisation was saturated with the old-fashioned values of a man as commanding unquestioning obedience from his family and as a proud, fearless respected force within the wider community. " We may argue that in a stratified, unequal community, gender relations were complicated and rigidified to further entrench patriarchal oppression (Tshoaedi, 2008, p. 13). However, since South Africa is not a homogenous country, the so-called "township ideology" varies from township to township, since people in different townships conform to different cultures, norms and values (Tshoaedi, 2008, p. 13). It is also important to note that violence is not correlated to "location". There is no tangible evidence yet that supports the locale theory, hence the location where people live does not generate violence. Violence, especially domestic violence, is an outcome of the cumulative social, economic, psychological and relationship tensions which the society and individuals navigate daily (Jewkes, 2002). The notion of a "poverty-locale-crime nexus" stigmatises the poor and validates the fraught nature-nurture theory. In other words, it presupposes that it is in the nature of people born in poor neighbourhoods to be criminal.
The locale theory does not correlate with the reality in a country such as South Africa where unemployment, poverty, inequality, and the "township" were historically and institutionally constructed by the centuries of racial oppression mediated by the White supremacist, colonial and apartheid governments (Simpson, 2000). It was this institution which "localised" people in squalor, deprived them of basic freedoms and human rights, which created the conditions that bred crime, not the patches of geography where people live that generated crime (Modiba et al., 2011, p. 872).

Social Learning Theory: a Conceptual Framework
In order to explain the prevalence of domestic violence in South Africa, but especially, among township dwellers, it is important to consider the theory of social learning. Social learning theory is a behavioural approach based on the work of Albert Bandura (1977). It looks at a person's behaviour resulting from his or her cognitive processes and exposure to certain patterns of behaviour. Among others, it also considers that gender roles are learned rather than inheriteda major departure from the naturenurture theory which presupposes that roles are "ascribed" by nature. It regards domestic violence as the result of "learned behaviour" and is based on the intergenerational transmission of violence theory (McCue, 2008, p. 13). Bandura (in Kurst-Swanger & Petcosky, 2003, p. 43) states that violence is learned, either directly or indirectly, and is reinforced in childhood and continued into adulthood as a response to stress or a method of conflict resolution.
According to the theory, an individual learns to become a member of a community through the process of socialisation. For LaViolette and Barnett (2000, p. 16), learning applies not only to observable behaviours but also to cognitions (thoughts) and attitudes. Learning is strengthened through reward and punishment as well as through observation of others' behaviour, and ideally, an individual learns to discard non-productive behaviour and to retain healthy functional behaviours and beliefs through social reinforcement.
According to Kubeka (2008, p. 285), social learning theory enables us to explore the aggressive behaviour associated with domestic violence. Kubeka (2008, p. 285) argues that there is a focus on the manner in which behaviour is acquired and retained through the observation of significant "others", namely, the community to which one is associated from birth. In explaining the social reality of domestic violence, Kubeka (2008, p. 285) states that: from childhood and imprinted by parents thus the dangerous unintended lessons learnt are (a) those who love you the most are also those who hit you, (b) those you love are those you hit, (c) violence can be and should be used to secure good ends, and (d) violence is permissible when other things do not work.
Social learning theory asserts that gender roles were learned through the reinforcements of social behaviour, whether positive or negative, that children receive for engaging in gender "appropriate" and "inappropriate" behaviour (Wharton, 2011, p. 38). Social learning theorists argue that gender-typical behaviour (including hitting one's wife/partner) was a learned behaviour created by the mechanism of reinforcement of social behaviour (Wharton, 2011, p. 39). Experiences of violence in childhood may teach children that violence is normal in certain settings. Men use violence and women learn to tolerate it, or at least to tolerate aggressive behaviour because of lack of resources or options (Jewkes, 2002(Jewkes, , p. 1426. In Social learning theory, domestic violence is a goal-oriented mechanism that maintains an imbalance of power between a man and a woman in a violent relationship (LaViolette & Barnett, 2000, p. 20). From a sociological perspective, LaViolette and Barnett (2000, pp. 20-21) argue that abusers can be viewed as the extreme end of a continuum of controls meant to reinforce male dominance over women in the unequal power relationship between men and women. However, Hundley (2012) argues that people in general possess more power than they realise and all that is needed is to apply it. For Hundley (2012), power is learnt rather than inherited; therefore, women can exert power in these violent domestic relationships and are capable of using it, for example, by leaving the abusive relationships.
This paper draws on the concept of social learning to explain a "social malaise", now seen as a part of the "culture of violence" in South Africa's poor black townships, such as Mamelodi. As highlighted above, social behaviour is learned from the milieu rather than "inborn" in people. Violence as a social phenomenon is a consequence of societal failures to rein in factors inherent in society that are capable of impacting acceptable behavioural patterns. While not being structuralist, we recognise that in a society that has witnessed generations of socioeconomic alienation, high rates of alcoholism and joblessness, extraneous social stresses are bound to affect patterns of behaviour in which frustrations create tensions capable of altering intimate relationships (Mazibuko & Umejesi, 2015).

Methodology
This article draws on a study on domestic violence in Mamelodi Township, a poor black neighbourhood in Pretoria, South Africa. As researchers, we keenly observed the different living conditions in this township. It is imperative to mention that since the post-apartheid era, Mamelodi has developed a stratified residential system depicting the emerging social classes of the new South Africa (see also Ramafamba & Mears, 2012, p. 1564. Hence, the housing range from well-built brick houses for the emerging middle class to small informal dwellings made of sheet metal (shacks) and affectionately called "Mandela Area" for the poor. Mamelodi is divided into two main sections -Mamelodi West and Mamelodi East. The Pienaars River separates the East and West.
Before requesting ethical clearance from the Ethics Committee of the University of South Africa, we consulted with professional workers 59 Caritus Mazibuko & I Umejesi -The public accounts about the best steps to consider for referral counselling and guidance for those research participants in need of that kind of service. We collected referral information on other issues and distributed it to non-government organisations in Mamelodi Township. Contact details for additional information, debriefing or counselling were provided to all the women participating in the research.
The study made use of a qualitative research method and utilised in-depth interviews and observation for data collection. In this article, the interview responses of research participants were drawn and analysed thematically. The social worker at a non-governmental organisation in Mamelodi introduced us to some of the research participants. We then relied further on referrals.
In this study, we interviewed women participants from the West and East of Mamelodi Township. These interviewees were purposely selected in the hope of understanding the diverse narratives on domestic violence. Racially, the 27 interviewees were all black South Africans -Mamelodi is historically a black enclave in Pretoria. It is important to point out that the interviewees were survivors and victims of abusive relationships. In this paper, we use the word "survivor" for those who had left their abusive relationships, and "victims" for those who are presently in abusive relationships.
The interviews took place in private on the premises of Ithemba Community Centre (this was the pseudonym used in the study). Other pseudonyms were used to refer to the female participants. The interview questions probed the experiences of the study participants with domestic violence in their community, and the ways in which their community responded to domestic violence. The interviews were audio-taped with the consent of the research participants and were then transcribed. All of these women participants were educated. Their qualifications ranged from matric certificates to university postgraduate certificates. All the participants were employed at the time of the field work in 2013-2014. Included is the table with information on the women in the sample, their ages, number of children, number of interviews, education level, marital status, work positions, and their place of origin. South Africa has nine provinces; however, the participants' places of origin were five provinces, namely GP-Gauteng Province (GP); KwaZulu-Natal (KZN); Eastern Cape (EC); Mpumalanga (MP); and Limpopo (LP).

Limitation of the Study
The study was limited by various factors which include the fear of retribution by women participants. Although it was not difficult to find abused women, we found it difficult to glean information from them. Various interviewees expressed the "fear of being caught" by their current or estranged partners. As already noted, Mamelodi is a poor township with high unemployment rate; hence, most women depend on their male partners for sustenance. This unequal relationship between male and female partners engenders patriarchal control of the dependent women; hence, the researchers struggled to find women who would grant in-depth interviews. There was also the sociological inhibition of some women,

Caritus Mazibuko & I Umejesi -The public accounts
namely the belief that women should not reveal family secrets to outsiders. It is believed that such revelation by a woman constituted "bad behaviour" and gossiping. In a close community, such as Mamelodi, no one wanted to be labelled a gossiper and suffer the social consequences that accompanied it. We observed that while women are caught up in this restrictive norm, it is not generally applicable to men who have been found to carelessly share relationship secrets with friends in the shebeens (local drinking parlours).

"Men Beat Up Their Women in Mamelodi": Witnesses' Accounts of Domestic Violence
The field accounts of the participants are crucial for this paper. Hence, much of the questions asked focused on the narratives of those women who had witnessed domestic abuse in Mamelodi as survivors, victims or observers. A participant, Thembi (aged 63, a university graduate and primary school principal in Mamelodi), who resides in the middle-class area of Mamelodi, was eager to participate in the interview. She asked that we (the two researchers and herself) step out of her office, walk around the school and be orientated with Mamelodi township. She then said that poverty-stricken locations such as Mandela Area are the ones with a high occurrence of domestic violence. She remarked that: In these mkhukhus [local tag for informal settlements/shacks], that is where I normally see people fighting, men beating up women. Yes, men beat up their women in Mamelodi. They are not all married to these people. Many just cohabitat in these mkhukhus.
The use of the word "normally" was often used by different research participants used when commenting on witnessing domestic violence. It was a way of describing the frequency of this phenomenon. It suggested that such behaviour was expected of certain men, especially those who live in the informal section of the township that was widely regarded as the "wretched of Mamelodi". In other words, the poorest of the poor.
There was a common view among the research participants that we should have focused the study mainly in the area called Mandela Area, as mentioned earlierthe informal/shack section. According to Nomonde (aged 59, mother of three, a local nurse): "If you want to see women being beaten on the street, just drive through Mandela Area. You won't even have to ask anyone to participate in your study; you will just see it for yourself". We asked this respondent, how often did her community intervene to stop this act? Her answer was: "No, the community does not intervene. We as professional nurses simply stitch them if they have open wounds. We also advise these women to seek counselling and report the case to the police," she quipped, smiling.
We as the researchers decided to drive to the Mandela Area and observe the people who lived in this area. It is the poorest area of Mamelodi. The people are living in shacks, and there are visibly run-down tuck shops, with the highest rate of unemployment. For instance, it was a common sight to see healthy young men and women during working hours roaming the streets, simply because they did not have jobs.
Because of this pervasive poverty level in Mamelodi, certain families have found different coping strategies. One of such strategies is that parents openly encourage their daughters to go into a relationship with an older man, which they think will be of economic benefits to their families. In one of the introduction meetings with the social worker, she said certain daughters who attended counselling sessions had confessed to her that their mothers would encourage them to have relationships with older and wealthier men.
Of course, these men take advantage of these relationships to exploit and abuse the girls. Another participant who confirmed this was Nontombi (aged 57, unmarried mother of three, a graduate working in one of the government offices). She said: Living in poverty has influenced older women to even encourage their daughters to pursue relationships with sugar daddies [older men with money] in an attempt to tackle poverty… poverty has crippled our minds with the assumption that money is everything.
As a mother of three daughters, Nontombi went on to say that she would never accept any of her daughters having intimate relationships with older men. According to Nontombi, these older men own and control these young women. These young women become properties of these older men, and in most instances, these older men are married.
Other research participants pointed out the asymmetric power relationship that underscore men's attitude towards women. Men often have this attitude of "she must come back, no matter what happened". There were cases of women who remained submissive in their intimate relationships in the face of recurrent abuses. For instance, Sibongile (aged 59, widowed mother of three, registered professional nurse) explained: There was an incident of a woman who came into my house in her underwear, running away from her husband who was beating her. She ran to my home for safety and rescue. She slept in my home that particular night and the next morning she woke up and said it was fine; then she went back to her house.
The above statement indicates the lack of respect and dignity for women in abusive relationships. The decision to return to an abusive husband or boyfriend often gives away any respect the woman had for herself. Noluthando (aged 34, divorced postgraduate, a trainee manager) explained: He would break up with me, even in public, and I would run back to him and plead with him to take me back; pathetic, right? Noluthando was the financial backbone in the marriage, since her husband was unemployed; yet she received no respect from him. The motivation behind her returning to her husband could not have been because of poverty or financial security -she wanted her marriage to "survive", albeit in an abusive relationship. Lerato (aged 42, divorced mother of two, a graduate teacher in a primary school in Mamelodi), also related an incident regarding a dehumanising experience of a woman who lived next-door to her: There was this man who was married to this particular woman. Upon beating her some days ago, he brought another woman home. His wife then had to sleep on the floor while he slept with his new woman on the bed in the same bedroom. When the community found out about their sleeping arrangements, they became furious and attacked the man. They came and broke the windows and spray-painted them.
While collective violence against perpetrators may not solve this problem, collective rejection of this malaise will surely send strong message to these offenders. Research participants such as Anele (aged 27, a government employee), Noluthando, Themba, (aged 33, married, mother of two, a graduate teacher), Zethu (aged 38, a high school graduate, married, mother of three, working as a cleaning service), Mercy (aged 40, a primary school teacher) and Lerato shared the view that infidelity committed by men is common in their community. The extent that some men engage in infidelity is viewed by other members of the community as unacceptable behaviour. Sometimes the community (which included men and women) took matters into their own hands, as evidenced in Lerato's story above. For Thapiso (aged 25, an unmarried university graduate doing community service in Mamelodi) The double standards were highlighted by the fact that women were not allowed to engage in infidelity like men, but remained faithful to them, even when they were treated so badly by men.
One explanation of this "double standard", as Thapiso puts it, is the feeling of insecurity by certain men. We found a common narrative among the women we interviewedthat they had heard men publicly warning their female partners to desist from looking at other men or greeting other men. Such warnings, if ignored, attracted severe consequences. Mulalo (aged 23, holder of a postgraduate qualification and a social work volunteer) did not listen to her partner's warning. She explained: He came to the party and saw me talking to a guy I grew up with. He called me roughly, angry even. He said I am disgusting him and embarrassing him in front of everyone in the party. He asked me: ukuthi yiwona ke amadoda ami engilala nawo? [So these are the men I am in sexual relationships with?] He began beating me up, saying I am cheating on him with them. He dragged me outside the party hall in the full glare of everyone, saying he is taking me to his flat. He threw me inside his car and drove off. I fought him inside the car until he stopped the car in Mandela Area. He beat me seriously when we arrived at his flat. After all the beatings he says he is taking me back to the party. Imagine the humiliation. I was so scared of him, but we went back.
Mulalo told us that her partner had informed her on various occasions that she should "respect" him before other men. It is intriguing how having a chat with a guy you grew up with amounts to disrespect to a male partner. It is clear therefore, that the definition of respect in such abusive relationships is women must not talk to or have non-intimate male friends.
The other research participants, who did not have personal experience of domestic violence, explained that they heard stories and noises as indications that a woman was being beaten by her partner in the home, and this was enough evidence to suggest that domestic violence exists in their community. For instance, Nomsa gave the account of what she often experiences in her neighbourhood: You see domestic violence is not as normal or constant as it used to be in the past or when I was growing up. Domestic violence was just a normal thing back in those days, especially during sunset. You could hear a wife being chased around the house and beaten by her husband. Nowadays physical abuse has been replaced by emotional abuse because there is no physical proof with the emotional abuse. So, in the present times, men shout at their women, even in the middle of the night, or threaten to shoot them dead. It is common here. I hear these threats every night.
Nomsa is 57 years old and she was speaking of recurring events that took place in the 1980s, before the introduction of the Domestic Violence Act 116 of 1998 in South Africa. Nomsa's childhood was exposed to domestic violence in her community and she explained how there was a general misconception among children that domestic violence was normal in intimate relationships; hence, the community never took any measures to stop domestic violence. In her community then, domestic violence was acceptable. She further explained that: When girls entered into relationships, some even preferred boys who actually beat them, as signs of love and possessiveness. Many young men in those days lived up to that expectation.

Domestic Violence and Community Response
The intractable problem of domestic violence has often been framed by certain cultural practices which are underlined by privacy. There is a view that a fight between intimate partners should be "left to the partners" to deal with, even though the fight may have taken place in public. According to Busisiwe (aged 53, a university postgraduate, mother of two, and company manager), dealing with domestic violence at community level has proven to be challenging since domestic violence is still regarded as a private issue that is best discussed by family members only. The word "private" does not suggest that domestic violence happens privately; rather it implies that it is between intimate partners.
Busisiwe noted that even if you hear or even see incidents of domestic violence, you cannot get involved, because you do not want to be perceived as umuntu othanda izindaba zabantu (nosey or something similar). Themba stated that in her opinion, the lesson learnt is to not get involved between two people (a couple); they will end up hating you because they may think you were trying to break up their relationship or something. Best you stay out of the lovers' quarrel.
Although the research participants acknowledged that domestic violence was perceived as a private issue in their community, they suggested how they thought their community could become involved. Research participants such as Mercy, Noluthando and Themba had similar views: if the community witnessed domestic violence, then they should report it to the police. However, there was a contradiction in what they were suggesting because some of these research participants never reported their own cases to the police. The evidence of not reporting the case to the police can be seen in Themba's interview response: Themba and Busisiwe's views highlighted "fear" of the neighbourhood as a bulwark against reporting domestic violence cases to the police, or community members becoming involved. Similarly, this same fear plays a significant role in preventing other research participants from reporting their cases to the police. For example, Noluthando noted that her ex-boyfriend, who used to abuse her, worked as a police officer and also a part-time law student. According to Noluthando, my exboyfriend represented the law and I feared that reporting him would expose the police system. Noluthando explained that she had witnessed on several occasions where policemen just got verbally reprimanded after beating their partners instead of due punishment. We asked the respondent how she thought the law should deal with perpetrators of domestic violence. Themba's response reflects the general opinions of other participants: As women we need to stand together if we see a man beating up a woman. We should, rather than report the case, let us take the matter into our own hands. We must beat this man because what he is doing is wrong. Since, even if we take the case to the police, they will say it is not domestic violence. The police can only register the case if they see blood and if there is no blood as evidence of assault, the police will not accept it.
Themba's expectation of the women in her community was that they should all take a stand against domestic violence, albeit confrontational. Although what she was suggesting women to do is a crime in South Africa, we found an inherent vagueness of what should be really classified as criminal behavior. According to Themba, if there is no blood rushing through the woman's face, then it is not yet criminal. In other words, the police do not take domestic violence seriously, unless there is evidence of bodily injury. Domestic violence is not in itself a crime category and there are no separate statistics on assault cases perpetuated by husbands or boyfriends (Mazibuko, 2017). We know that Themba was speaking figuratively, but her statement shows the levity with which law enforcement agencies, especially the police, deal with cases of intimate partner abuse.
Themba explained that she had witnessed incidents of domestic violence where the cases were dropped by the police because there was no physical evidence. It remains a huge challenge when cases of domestic violence are not reported to the police because this sends a message to perpetrators that their behaviour will have no consequences. It also sends a message to women victims or survivors of domestic violence that they have no protection, thereby reinforcing their sense of vulnerability.

Discussion and Conclusion
This study has highlighted one major fact about domestic violence in South Africa's poor black townshipsthat while residents in townships such as Mamelodi see domestic violence as a social malaise in their community, they do not think it is an act hidden under the privacy of the home. Domestic violence as explained by the interviewees occurs both at home and in public. There is overwhelming evidence of men beating their female partners in public places, chasing their partners down the street and fighting in places such as party houses. Some interview participants also noted that while some couples may not have engaged in public fights in the streets, the screams and cries of beaten women, and the sights of injured women who are treated in public health centres make domestic violence very visible.
It is therefore crucial to reassess the notion of domestic violence as a phenomenon noticed only in the confined space of the home. Bassadien and Hochfeld (2005, p. 5) argued that the home space has mediated the "imperceptibility" of domestic violence and that it is rarely witnessed. The reality, as this paper has shown, is that domestic violence is never really imperceptible, even when it is committed in the home. The screams of victims that awaken the neighbourhood at night and the sights of the injured in public, all point to the fact that this act is not exclusively a "home issue" or a "private matter".

Caritus Mazibuko & I Umejesi -The public accounts
It must be acknowledged that South Africa has a very high crime rate, domestic violence being one of the major flanks of the statistics (Machisa, 2011). This comes at a great cost to the health and lives of women (Thorpe, 2013;Mazibuko & Umejesi, 2015). Although South Africa has some of the best legal and institutional frameworks on gender rights in Africa, such as the Domestic Violence Act 116 of 1998, issues such as police inefficiency, protracted court cases and cultural impediments stand in the way of prompt punishment for offenders. It is our opinion that institutional improvements such as the establishment of community rights advocacy groups to monitor and prosecute these abusers will act as a deterrent against this deepening social malaise in South African society. The public must be empowered through the revision of existing frameworks. The Domestic Violence Act 116 of 1998 should have guidelines on how witnesses of domestic violence should be protected, if they were to report it to the police. In addition, the congested courts hinder speedy prosecution of offenders. We therefore suggest the establishment of "family and relationship courts" similar to traffic courts, where offences are dealt with more urgently. We hope these courts will expedite the trial and conviction of offenders.
Finally, this study challenges the notion of "privacy" and "imperceptibility" in the domestic violence discourse in the South African society, especially in poor black townships. By perceiving domestic violence as confined to the home space, the society inadvertently reinforces the continuation of this malaise. Such a perception of domestic violence fails to challenge established social norms that continue to trivialise this major social crisis. It is important to note the central place of 'family income' and 'control' in the narratives of abused women. Future research should therefore focus on the place of women's financial contributions or lack of it, to their household income in the committal and perpetuation of domestic violence among poor black people especially in the townships. This is important since most households hinge their sustenance on the financial contributions of men. Other areas that could provoke future research include the educational and employment statuses of male abusers. There is often the anecdotal suggestion that uneducated and unemployed black men have the greater propensity to engage in domestic violence than educated and employed black men. This opinion creates stereotypes of poor black men in the townships that aligns social status to behavioural patterns. This narrative needs to be probed scientifically.