Violence against women occurs in all societies in some form or another. The home is sometimes the most dangerous place for abused women, because they often know and live with the perpetrators (Fedler & Tanzer, 2000). Fathalla (2005) and Fedler and Tanzer (2000) assert that to many people staying out of harm's way is a matter of locking doors and windows and avoiding dangerous places, whereas for others, it is not possible to escape because the threat of violence is behind the doors, well hidden from public view.

According to Fedler and Tanzer (2000), violence in some cases is supported by various cultural and traditional practices within the society. Mullender and Morley (1994), cited by Park and Khan (2000), maintain that culture is universally used by many men in order to maintain their power over women, and to justify misogyny and chauvinism. A study conducted by Moletsane and DeLancey (2003), found that violence against women is an additional aspect of life that increases the likelihood of a woman becoming HIV positive. The perpetrators of domestic violence have control and advantages in the domestic arena. Moletsane and DeLancey (2003) also found that customs, practices and beliefs encourage multiple sexual partners and risky behaviour, and in most cases, men have the final decision as to whether or not to use protection. In their study, Park, Peters and De Sa (2000) found that more than 80% of all HIV infected women get the virus from a male sex partner. The victims of domestic violence who become HIV positive, suffer psychological trauma which, according to Zappulla (1997), is associated with extreme anxiety over possible rejection and judgment by family members, friends and the community.

The year 2004 was a significant one in South African history. It marked the end of legislated apartheid and saw the introduction of South Africa's first democratic elections, as well as the implementation of a range of changes aimed at eradicating apartheid's injustices and excesses. But for South African women, freedom from racial oppression and violence has not led to freedom from gender oppression and violence (Vetten, 2000).

The international agency Interpol, maintains that South Africa reports some of the highest figures for rape in the world (Crime Information Management Centre, 1996). Vogelman and Eagle (1991) describe violence against women as endemic in South Africa. The problem has become so intractable, that the National Crime Prevention Strategy (NCPS) has declared crimes of violence against women to be "priority crimes" (Department of Safety and Security, 1996, p. 37). As a psychologist, who provides intervention to abused women on a regular basis, this author was motivated to conduct this study to explore their experiences, which are a reflection of the high rate of violence against women in South Africa.

The purpose of this research study is to explore and investigate the experiences of abused women living in shelters. The study also aims to construct the experiences of the participants into informed knowledge.

Pelser et al., (2005) maintain that perceptions of violence differ vastly between cultures and communities. Research on violence committed against intimate partners has shown that some people, including women, regard such violence as acceptable if it does not injure or leave a mark on the victim (Vetten, 2000). Abrahams and Jewkes (2005) found that many women in their study thought wife beating was justifiable under certain circumstances, including refusal to have sex. These attitudes are symbolic of a general pattern of acceptance of victimisation as part of being female. A culture of silence prevails which is a serious concern. In a survey conducted in three provinces in South Africa, more than a third of the participants agreed that a man beating a wife was a sign of love (Abrahams & Jewkes, 2005); hence some abuse cases are not reported in South Africa. It is evident that in South Africa women need to be informed about their human rights. According to Saur et al., (2005), domestic violence is a vicious circle which can be addressed and tackled by profound awareness of human rights, the right to bodily integrity, and by profound empowered strategies on an economic and educational level.

Due to the seriousness of domestic violence in South Africa, the Domestic Violence Act (Act 116) was introduced in 1998. The aim of the Act is to protect the victims of domestic violence by creating obligations on law enforcement bodies to protect the victims as far as possible. The Act sets out a broad range of behaviours that constitute domestic violence; including physical, sexual, verbal, emotional and psychological abuse, stalking, intimidation, harassment, malicious damage to property, unauthorized access to the complainant's property, as well as other forms of controlling behaviour which may cause harm to safety, health or well being of the complainant (Dissel & Ngubeni, 2003). The Domestic Violence Act has its own challenges. According to Vetten (2005), while South Africa has eleven official languages, the application forms for the protection order are available in only two languages, namely; English and Afrikaans. Over and above language, the written completion and reading of the application forms challenges women with varying degrees of literacy. Further, the forms are not available in Braille, and sign language interpreters for Deaf women are not readily available at courts.

World Health Organization (in Fathalla, 2005) mentions that; "Violence pervades the lives of many people around the world, and touches all of us in some way." A study by the World Health Organisation (2006) confirms that violence against women is pervasive and that violence by intimate partners, rather than other perpetrators, is a dominant occurrence in women's lives. The study also revealed that for women with partners, the lifetime prevalence of physical or sexual abuse, or both, by an intimate partner ranged from 15% to 71% in the different sites investigated. The study conducted by the World Health Organization (2006), across all countries, reports that between 20% and 75% of women had experienced one or more acts of emotional abuse.

A World Bank report (in Fathalla, 2005) estimates that rape and domestic violence account for about 5% of the total disease burden in women aged 15~44 years in developing countries. By damaging women's physical, mental and emotional capacity to care for family, violence against women also affects the health of other family members, particularly young children.

The various manifestations of violence against women in South Africa have been influenced by social, cultural and political history; by colonisation and apartheid, in particular and by the resultant race, class and gender divisions (Vetten, 2000). Vetten (2000) further mentions that in order to understand violence against women in South Africa, one must have some sense of South Africa's particular colonial and apartheid history and the current transition to democratic government. Throughout the history of South Africa, race has divided the people of the country. Notions of racial superiority and racial separation have led to the creation of institutions such as migrant labour and domestic labour. In addition people have had to face forceful removals, states of emergency and the artificial creation of so-called "homelands." These racist acts and institutions have either directly or indirectly resulted in the break-up of traditional African family structures. Together with the effects of urbanisation, the codification of customary laws and modernisation, there have been many challenges in terms of men's traditional authority and patriarchy.

Social hierarchies that privilege the traditional authority of men over women remain fairly intact, especially in rural areas. Within this context, violence is used to maintain the patriarchal social hierarchy, as well as to compete for scarce resources (Van der Waal, 1996). The European colonisation of Africa resulted in the development of a dual legal system within South Africa, namely: an European system of law applicable to the colonisers and a "subordinated and regulated version of indigenous law for the colonized" (Seidman, 1978, pp. 29-34 in Currie, 1994). However, the interaction between colonial rule and African customs produced another type of indigenous law (Chanock, 1987 cited by Nhlapo, 1991). The customary law was born of an alliance between colonial administrators and indigenous elites. African male elders, seeing their power dwindling, attempted to maintain it by manipulating customs such as lobola (dowry) and kinship (Chanock, 1987 in Nhlapo, 1991).

According to Archer and Lloyd (2002), there is a general perception among some communities in South Africa that men are superior and women are inferior. Archer and Lloyd (2002) further state that one paradoxical part of this belief is that women are viewed as being nicer, more morally upright human beings. This unequal power balance between men and women can have an effect on the psychological well-being of women. If women and/or girls are continually told that they are inferior and unworthy, they may develop an internalised oppression (Hunter & Forden, 2002). This internalised oppression can lead to self-destructive behaviour such as difficulty with assertiveness, as well as a tendency to settle for less in life. Kaplan and Sadock (1988) maintain that behavioural, cultural, intra-psychic and interpersonal factors contribute towards abuse against women. The abusers' aggression is bullying behaviour designed to humiliate their wives in order to build up their own low self-esteem. In view of the high prevalence of intimate partner violence in South Africa, it is not surprising that much of this violence is witnessed by children in the home (Abraham et al., 2005). This certainly has an impact on children.

The reaction of children who witness violence against their mothers takes a range of forms (Park & Khan, 2000). They overhear the violence as it occurs and witness the beatings. Some, particularly, older children, attempt to intervene to protect their mothers, whereas others are encouraged to support or participate in the abuse and degradation of the mothers. Even when children are never themselves physically harmed by the abuser, they are indelibly impacted by the violence. The greatest benefit for children in a shelter, after safety has been provided, is to have contact with other children in a similar position. Muller and Morley (1994), in Park & Khan (2000), maintain that in a refuge children can have their experiences heard, validated and understood by other children and adults who have been through similar things and who understand how they feel.

Family violence can have the following effects on children according to Park and Khan (2000): emotional, somatic or physical, social and behavioural effects.

Children from violent homes often feel tense and anxious. They might feel guilty because the abusive partners often use their children's behaviour as an excuse to start a fight. The children then feel it is their fault and that they should be able to stop the violence or protect their mothers. They might develop withdrawal behaviour and lack of concentration, which can affect their scholastic performance. Due to lack of attention or affection as a result of mother's stress, they might resort to inappropriate behaviour, such as throwing temper tantrums or swearing, in order to seek attention. The longer children live in an abusive home, the more likely it becomes that they will exhibit their fears and feelings through illnesses, such as headaches, stomach aches and asthma. Children sometimes model their parents' behaviour. Children from abusive homes might grow up learning that violence is used to address conflict and to solve problems. Often they do not learn alternative ways of solving problems and may use violence to get what they want. This pattern is called the cycle of violence or intergenerational violence theory (Park & Khan, 2000).

Due to the increase of violence against women in South Africa, shelters for abused women were established, as described in this section. While shelters for destitute women or pregnant unmarried teens have been in existence for decades, shelters for abused women are a relatively new phenomenon in South Africa. According to Park, Peters and De Sa (2000), the first shelter in South Africa specifically designed for abused women, was opened in 1984 by People Opposing Women Abuse (POWA). Currently, there are approximately 25 shelters for abused women in South Africa. Many of these shelters have a therapeutic component and offer both individual and group counseling. They employ experts such as psychologists and social workers to work with and rehabilitate the victims for life after the shelter.

Shelters are defined by Gelles (1976), Bowker and Maurer (1985), and Dobash (1992) (all cited by Park, Peters & De Sa, 2000) as places of safety where abused women and their children can seek refuge when their lives are endangered by interpersonal violence. Shelters for abused women play an important empowering role in providing social, psychological and emotional support for women in abusive relationships. The safety of the shelters provides women with opportunities to plan for a safer future. Beyond meeting such practical needs of abused women and their children, shelters play an important strategic role, namely working towards eliminating violence against women.

According to Batliwa (1994), women's empowerment is only vaguely understood by most abusers, who view such a concept as yielding absolute power to women. Batliwa (1994) further mention that empowerment is a process whereby the powerless or less powerful members of a society gain greater access and control over material and knowledge resources, challenge the ideologies of discrimination and subordination, and transform the structures and institutions which uphold the unequal distribution of resources and decision-making power.

Shelter services in South Africa are categorised as short-term or long-term shelters. Short-term shelters are intended to deal with emergency crisis situations and are equipped to shelter women for two to six weeks. Long-term shelters usually extend their services from three to nine months. While making use of short-term shelter services, women are encouraged to make plans for a safer future without delay. On the other hand, the long-term shelter service allows women sufficient time to deal with the practical and emotional issues related to coping with or leaving the abusive situation, normalises the situation for both women and children, provides counseling, and allows the abused women to establish meaningful and supportive relationships with other residents and shelter staff (Park, Peter & De Sa, 2000).

This study is designed using qualitative and narrative research methods. According to Lieblich, Tuval-Mashiach and Zilber (1998), narrative research refers to a study that analyses narrative materials. The data can be collected in the form of a story or a life story recounted in an interview. Lieblich, Tuval-Mashiach and Zilber (1998) further maintain that the mission of psychology is to explore and understand the inner world of individuals. Lieblich, Tuval-Mashiach and Zilber (1998) further mention that one of the clearest channels for learning about the inner world is through verbal accounts and stories presented by individual narrators about their lives and their experienced reality. According to Zapulla (1997), narrative keeps us alive; it preserves our experiences of human existence and provides a dynamic account of mankind's history. This study employed the interpretivist paradigm, which according to Burton and Bartlet (2005), attempts to understand the social world as it is from the perspective of individual experience.

The interview questions were divided into the following categories: causes of abuse, types of abuse, reason for staying in abusive relationship, referral to shelter, and lessons learned from their experiences. The findings are summarised in Table 1.

The findings indicated in Table 1 are discussed in the section below.

Perceptions of violence and abuse differ vastly between cultures and communities (Pelser et al., 2005). Some of the participants in this study considered abuse to refer only to domestic violence and abuse such as beatings that result in wounds, bruises and scars. They did not regard verbal, emotional/psychological and financial/economic abuse as being abuse, because there is no evidence such as wounds, bruises and scars. They regarded this type of abuse as normal reactions from their spouses. Some of the participants mentioned that when they were growing up, their mothers were shouted at, ridiculed by their fathers, and most of the men in their communities had extra-marital affairs. The themes are discussed below.

The sample in this study is very small; therefore the findings cannot be generalised to all abused women in South Africa. This study focused only on the abused women themselves, and not on their abusers. It is recommended that a similar study be conducted with a broader sample of abused women in all shelters, and on men who abuse women.

According to this study, it is evident that cultural practices and perceptions that women are weak still exist in some communities in South Africa. Women should not allow themselves to be victims of this school of thought. Even though the shelter system is not viewed as a permanent solution for victims of domestic violence, it gives them a second chance in life, and to be able to regain their proper place in society. I agree with Vetten (2000) that while shelters provide counseling and support to the abused women in South Africa; they will not in the long term provide preventative solution to the problem. However, other strategies such as education and training, in combination with support are needed. It is recommended that more shelters be established in order to empower victims of domestic violence. Public awareness of the problem of violence against women is needed.

It is recommended that men who are abusers be educated about the rights of individuals and rehabilitated. Government should assist in the establishment of more shelters, as well as in education projects intended to educate women about abuse and their human rights.