‘Selves’ in contradiction: Power and powerlessness in South African shelter residents’ narratives of leaving abusive heterosexual relationships

Abstract

South Africa has one of the most advanced constitutions in the world. Several progressive laws that promise the protection of women, including the Domestic Violence Act, and a range of state-funded bodies have been established to promote women’s rights. Despite these signs of transition to democracy in the post-apartheid era, violence against women remains problematically high. The dominant perspective in both South African and international literature on the high rate of violence against women has been that of women’s ‘powerlessness’. This article goes beyond approaches that emphasise women’s victimhood. It explores women’s agency from the perspective of the narratives of 16 women in two shelters in Cape Town. Drawing from Scott’s (1990) concept of power and resistance, and using a feminist poststructuralist analytic lens, the article provides insight into the complexity of women’s subjectivities ‘post-abuse’. It highlights women’s shifting sense of power in relation to their abusers, and how this imbued women with a sense of agency as seen through their retrospective accounts of their motivations to leave the abusive relationships.

Keywords

Woman abuse narrative subjectivities intimate partner violence violence against women agency South Africa

Introduction

South Africa has one of the highest rates of woman abuse in the world. In 1999, the country had the highest rate of intimate femicide reported anywhere in the world (8.8 per 100,000 women over the age of 14 years old) (Matthews et al., 2008), and, according to Fuller (2010), a rate of rape of women that is more than twice that of the United States, with Black and Coloured (or ‘mixed-race’) women 4.7 times more likely to be raped than White women (Anderson, 2000). With few exceptions (such as Boonzaier, 2008; Shefer, 2004), much of the focus on this problem in South Africa has been on women’s victimhood and powerlessness at the hands of men, a perspective that draws largely on theories such as Walker’s (1979a, 1979b) ‘battered woman syndrome’, Dutton and Painter’s (1993) ‘traumatic bonding’, and Herman’s (1992) notion of ‘captivity’. While we acknowledge the traumatic effects of woman abuse and the problem of patriarchy, particularly in marginalised townships¹ where women are most vulnerable, we argue, along with Profitt (1994, 2000), that this victim perspective often glosses over women’s capacity to enact resistance and agency in their lives even within the context of imbalanced social relations of power. Thus, there is a need for research that addresses women’s resistance as it exists in dialogue with the material and social realities of women’s oppression.

This paper addresses the issue of women’s resistance by examining women shelter residents’ narratives of their desire to leave abusive heterosexual relationships. We utilise Scott’s (1990) concept of power and resistance to shed light on the complexities of meaning that women give to their own sense of power and powerlessness and the fluctuations of identity that occur throughout their tellings. By looking at how women retrospectively construct their desire to leave these relationships, we gain insight into the different subjectivities that women adopt during this telling and what these subjectivities mean in terms of women’s sense of agency in the world.

Agency and the stages of disengaging from abusive relationships

Much feminist literature in the field of intimate partner violence has attempted to move away from static notions of women’s powerlessness and victimhood. Studies that have focused on women’s agency while leaving an abuser have described the process of leaving as consisting of many stages that begin before physical exit from the relationship (Anderson and Saunders, 2003). Burke et al. (2001), for example, outlined a process of leaving that consists of certain behavioural changes seen through the stages of pre-contemplation, contemplation, preparation, action and maintenance. Studies that have addressed the earlier stages of leaving have highlighted certain turning points (see Khaw and Hardesty, 2007) and defining moments (for example, Taylor, 2002) that motivate a woman’s decision to leave. Other studies have foregrounded women’s positive self-reconstructions that emerged throughout the phases of disengaging from an abusive partner, particularly women’s active ‘reclaiming of the self’ during which they move from a ‘victim’ to a ‘survivor’ identity (see Enander and Holmberg, 2008; Landenburger, 1998; Patzel, 2001; Wuest and Meritt-Gray, 1999, 2001).

The move away from static notions of women’s victimhood has led to more attention being focused on women’s power within abusive intimate relationships. Inversely, then, this feminist work has often highlighted aspects of women’s agency and resistance while downplaying the challenges and structural constraints of women’s lives (Boonzaier and de la Rey, 2003). Ignoring such challenges and extracting the abusive relationship from the broader socio-cultural and material environment within which it occurs may result in studies losing a certain richness of understanding of women’s lived realities and their complex responses to abuse (Yoshioka, 2008).

‘Victim’ and ‘survivor’: A critical perspective

As noted above, many authors have conceptualised abused women’s agency as processes by which women move from the position of victim to the position of survivor who negotiates separation from an abuser. This idea of the transition from ‘victim’ to ‘survivor’ is problematic, however, given that this trajectory may not be a smooth one. The situation for economically disadvantaged women is complex as they may be controlled by their intimate partners, but also lack access to material resources, and therefore be economically dependent upon the men who abuse them. These women are unlikely to call the police and attempt to have an abusive partner jailed. Often women do not manage to leave the abusive situation and, if they do, many return to their abusers at the end of a shelter stay as they have nowhere else to go. The act of leaving, while often saving a woman from the current abuser, may drive her into poverty, homelessness or further abuse at the hands of other men. Thus, the distinction between the notion of ‘survivor’ and ‘victim’ is not as clear cut as some (e.g. Baker, 1997; Lempert, 1996; Wuest and Meritt-Gray, 2001) have suggested. Moreover, the process of moving from a ‘victim’ to a ‘survivor’ identity is unlikely to be a simple linear progression. Women who manage to leave abusive relationships do not necessarily assume and then maintain a survivor identity. Rather, women may constantly fluctuate between positioning themselves as either ‘victim’ or ‘survivor’ at different points: positionings that emerge out of the complex material and contextual realities of their lives (Boonzaier and de la Rey, 2003).

The process of leaving an abusive partner has also been recognised as a complex, non-linear journey that consists of many contradictions. Anderson and Saunders (2003), for example, have shown that women continue to grapple with the meaning of the abuse long after the physical point of leaving the relationship. The act of leaving in itself does not constitute a final move to the identity of survivor (see Enander, 2011). Evident in the literature are multiple and nuanced understandings of women’s resistance as well as an outline of certain shifts in cognition that occur during the process of disengaging from abuse. Such studies are particularly relevant to our current inquiry, which explores women’s retrospective accounts of their motivations to leave their abuser.

Researchers such as Enander (2010) have conceptualised cognition as determined by discourses that shape emotion and action and, as such, have a significant impact on the leaving process. Enander (2010) argued that women oscillate between constructions of their abuser as ‘good’ (Jekyll) and ‘bad’ (Hyde) and that these multifaceted descriptions shape women’s positioning of agency in relation to the abuser. In a more recent article, Enander (2011) showed that greater amounts of physical violence result in greater cognitive emotional dissonance for abused women. She suggested that after repeated acts of violence, a woman who thought of her partner as good may begin to experience conflict between the perception of her partner as ‘good’ and her actual experience of abuse. This cognitive dissonance, or cognitive emotional dissonance as Enander termed it, leads to a gradual shift from positive emotions for the partner to ‘cold’ emotions – a shift that results in women maintaining separation from abusers.

Margerita Hydén (2005) also highlighted the complexity of women’s resistance through her examination of women’s agency ‘post-abuse’. She found that, over time, women spoke from different subject positions, which varied from ‘the wounded’ to ‘active agent’. Significantly, Hydén’s focus on multiple positions in women’s talk has allowed her to problematise mainstream meanings of fear as a signifier of women’s powerlessness (see Hydén, 1999). Complementing the literature on the complexity of leaving, Towns and Adams (2009) argued that women experience certain dilemmas between intellectual ideologies and the ideologies of their lived experiences. These tensions stimulate ‘internal debates’ (p.750) that are particularly prominent when a woman leaves an abusive relationship and may eventually lead to active resistance on the woman’s part.

Conceptual framework: Self, identity and transcripts of power

In his book Domination and the Arts of Resistance, Scott (1990) explores the secret transcripts of power that slaves, labourers and prisoners display behind the scenes – discourses that are largely hidden from their oppressors. His work on public and hidden transcripts of power provides a valuable added dimension for understanding women’s narratives. The concept of ‘transcripts’ refers to scripts that shape an individual’s self-construction. The public transcript is the script by which subordinate groups conform to social expectations of how they should behave towards their ‘superiors’. Scott argued that underlying these public transcripts are hidden transcripts through which the oppressed can harbour a sense of themselves that is generated in response to their oppression and that could be understood as a form of resistance against oppression.

Scott’s (1990) conceptualisation is based on the notion that power is both performative and relational. In this sense, people draw on public and private transcripts in particular ways that are descriptive of their sense of power in certain interactional settings. This approach concurs with a Foucauldian notion of power as multidimensional and relational. According to this view there is some form of resistance at every point of power (Foucault, 1982). Scott’s notion of power serves as a useful analytic frame with which to explore the way in which individuals fluctuate between resistance and subjection. This approach supplements Foucault’s concept of power by providing a means with which to explore the nuanced interplay between accounts of self as powerful and powerless and how these constructions co-manifest at various moments.

This poststructuralist approach moves beyond a rational, unitary self, recognising a precarious, contradictory and ever-changing self that is constructed through language and in different interactional contexts (Davies and Harré, 1990; Davies et al., 2006). In this sense, an individual constructs and reconstructs a self at different socio-historical moments according to the scripts and audiences that are present. Various storylines offer different, often contradictory, subject positions from which an individual draws at different moments to position themselves and others in certain ways (and to negotiate new positions) (Davies and Harré, 1990). In this analysis, we adopt an understanding of subjectivity as the conscious and unconscious thoughts and emotions of a person, their sense of self and their ways of relating to others in the world (see Weedon, 1987).

In this paper, we draw on Scott’s framework to further the insights in the field of women’s agency and research into the process of leaving abusive relationships. More specifically, we look at how women retrospectively narrate the process that lead to their desire to leave the abusive relationship and the nuances of power that manifested in these accounts. Our analysis addresses the intersection of the public transcript (the ‘battered woman’ script) with the hidden transcript of women’s power, a transcript which, according to Scott, represents a kind of ‘ideological dissent’ (p.190). In this paper we conceptualise resistance as an active self-construction strategy. We focus on the ways that women draw on some scripts and resist others through their retrospective accounts of their motivations to leave an abusive partner. As we will show, the co-manifestation of transcripts of power and powerlessness in women’s accounts of leaving have implications for thinking about the complex psychological processes involved in leaving and women’s agency and subjectivities ‘post-abuse’.

Feminist poststructuralist work in the field of intimate partner violence (such as Hydén, 1999, 2005) and the work of Enander (2010, 2011) has contributed significantly to the field by providing insight into the manifestations of both the agentic/powerless transcripts that may co-exist in a single woman’s account as she continually fluctuates between positioning herself as powerful and powerless in relation to her abuser. This paper adds to this work by exploring the intersection of the transcript of power and the transcript of powerlessness. Ultimately, we unpack the meaning that is made through these particular moments of intersection in terms of our research question: How do the women construct their desire to leave abusive relationships?

Methodology

Research context

The research project on which this paper is based was a critical feminist inquiry into understanding women’s experiences of disengaging from an abusive partner. In-depth, semi-structured interviews were conducted with 16 women shelter residents who had experienced physical and psychological abuse at the hands of an intimate partner. The first author initiated contact with the shelters and conducted the interviews. The interviews took place in private rooms on the premises of the shelters. The interview questions covered challenges that women experienced while disengaging from an abusive relationship, person changes/growth that they experienced at this stage of their life and the role that the shelter played in facilitating this process of disengaging. These interviews were audio taped, with the consent of the participants, and were then transcribed verbatim.

Women were recruited from two shelters located in townships near the greater Cape Town area. These townships are characterised by high levels of crime, substance abuse, unemployment and general economic deprivation. All of the women were living in circumstances of severe economic deprivation, threads of which were woven into the stories about themselves and their worlds. Racially, 15 of the participants could be identified as ‘coloured’² and one as ‘white’. All were of South African nationality. The age range of the participants was 24 to 57 years old, with a mean age of 35.4 years. The average education level was grade 10, and only five of these women completed their high school education. All of the women had at least one child staying with them at the shelter. At the time of the interviews, the average length of shelter stay of the participants was 4.6 months, with a maximum stay of 36 months and a minimum stay of three weeks.

It must be acknowledged that these women’s narratives of violence are constructed from the vantage point of the shelters after they had already left the abusive relationship. They construct their decision to leave in retrospect, narrating their past from their present lives at the shelter.³ The shelter may facilitate the process of naming abuse and shifting away from minimising abuse, and it may facilitate particular interpretations of abusive experiences that draw on the broader societal context of power imbalances. It may also make available certain scripts of empowerment by which the women can articulate and negotiate their sense of power in the world. Here the women are connected with others who have experienced similar abuses, and this may inform their consciousness surrounding the issue. From a feminist poststructuralist view, the women are not looking back and relaying an absolute ‘truth’, but, rather, are constructing a narrative truth that is informed by scripts that are available to them at a specific moment.

Data analysis

The data analysis involved drawing themes that provided insight into the main thematic focus, namely women’s constructions of their desire to leave abusive relationships. We identified themes that spoke to the public transcript of abused women: pathology and powerlessness in relation to men. We also identified themes that represented the hidden transcript: women’s scripts of their sense of power (and ideological dissent) in relation to male abusers. Extracts of the women’s narratives were chosen according to those that most aptly conveyed a sense of the interplay between the public and the hidden transcript, as we have interpreted these concepts. Themes were extracted from brief segments of the women’s narratives rather than extended biographical accounts. Themes were then explicated across cases, providing a holistic view of the women’s voices. These methods of analysis are consistent with thematic narrative analysis as outlined by Riessman (2008). Attention was also paid to the language that the women used in the telling of their stories. Dominant transcripts were identified in women’s narratives, particularly with regards to scripts of ‘abused women’ and ‘active agents’. This involved an examination of the women’s use of language and the ways in which their articulations of self were structured by the surrounding social system (Collins, 2003).

Ethical considerations

A number of ethical concerns were kept at the forefront of the study, namely, voluntary and informed consent, confidentiality, anonymity and non-maleficence. Informed consent was obtained and women’s rights to refuse to disclose certain information, as well as their rights to withdraw from the study were respected, along with the provision that no negative consequences would arise from non-participation or from withdrawal. The interviews were conducted in an ethically sensitive manner and women were informed about the availability of counsellors at the respective shelters, should the interviews have raised any areas of concern. In this publication, women are identified through pseudonyms, with no further identifying details being provided, including the names and locations of the shelters.

Analysis: Constructing the desire to leave

In what follows we analyse the women’s narratives of their desire to leave abusive heterosexual relationships. The analysis revealed that the women drew on the cultural ‘battered woman transcript’ and at times in their narratives also resisted this transcript and invested in more powerful ‘alternate selves’. These oscillations of self were signified by narrations that adhered to hegemonic constructions of femininity and the ‘abused woman’ on the one hand and to alternate constructions of self as ‘powerful moral agent’ on the other. The former hegemonic constructions stand as Scott’s (1990) public transcript that provided a cultural script of abused women as powerless (and pathological), while the latter refers to women’s hidden transcript of power in relation to men.

The ‘battered woman’ script

The women in this study worked with the ‘battered woman’ script in complex and varied ways. This cultural script depicts abused women as being weak, pathological and passive. They drew on this script by identifying the unequal power dynamics of their relationship and, in doing so, constructed themselves as a ‘powerless’ and ‘fearful’ victim who held little power in relation to their abuser. The narratives of self as ‘powerless victim’ encompassed constructions of self as ‘abused woman’ that were juxtaposed with constructions of self outside, or beyond, the abuse.

Because a man the way they abuse you it makes you feel scared, you get weak, you turn up to do everything they want you to do. You get to be also everything they want you to be. It’s like a role you playing now. It’s like a game you playing for them now and the only way to end is you, you make it stop.

(Jolynn)

Above, Jolynn highlights her sense of a loss of identity in relation to her controlling abuser (you get to be also everything they want you to be). She uses the pronoun ‘you’ to construct women in general as being ‘weak’ and ‘fearful’ in relation to men; thus she draws on dominant scripts of masculinity and femininity that appear to shape her account of her lived experience. She, however, articulates submissiveness as a role that women adopt and not as an inherent part of women’s nature. She contrasts the submissive ‘game’ that women play with women’s ‘true’ self that may be strong and capable. It is this self that she refers to when she says, ‘you make it stop’. Jolynn is highlighting her own sense of women’s capacity for action. The positioning of self as agent blurs the boundary of the powerful/powerless duality and suggests the complexity of women’s experiences of self in relation to their male abusers. In this excerpt, Jolynn highlights the contradiction between a ‘capable self’ and a ‘submissive self’ and makes a call to the potential agency of abused women – seen by her description of women’s capacity to make the abuse stop. Here she invests in an agentic self that has the responsibility and the capacity to end the ‘game’ with the abuser (‘the only way to end it is you’). We interpret the contradiction between the agentic self and the passive abused self as representing a juncture for Jolynn: a point at which leaving became a possibility. It is this juncture that she retrospectively narrates as her motivation to take action by leaving.

The women also positioned themselves as ‘fearful victims’ in relation to their abusers. They did this by narrating certain insights of ‘realising the danger’ during which they constructed an increased awareness of the danger that the abuser posed to their lives. This was illustrated most notably via women’s descriptions of particularly violent incidents of physical abuse. Altoise painted a picture of her ‘fearful self’ by constructing her lack of physical power in relation to her abuser.

He take me one night here by the arms. He’s a big man. He take me there was a blue mark and things like that and he throw me against the wall there. There was a knock here, my eye was blue, he slapped me, you know. Then I thought to myself, ‘Oh well, this is the last straw. I have to run for my life. This guy is gonna kill me’.

(Altoise)

Altoise positions herself as a helpless victim who is vulnerable to intense physical domination at her partner’s hands. By focusing on his strength and capacity to exert power over her, she constructs herself as having limited power in relation to the abuser. She articulates a retrospective account of the seriousness of the abuse and displays the source of her ‘fear’. This narrated moment may be seen as representing what various scholars have defined as a turning point in the relationship, a moment at which a woman’s cognition shifts to become orientated towards leaving (see Khaw and Hardesty, 2007; Patzel, 2001). Here Altoise highlights the point at which she realised that she had to leave her abuser. Within the context of socio-cultural pressures that blame women for leaving and breaking up the family unit (see Towns and Adams, 2009), Altoise may draw on this incident to construct a plausible ‘explanation’ for why she left her abuser. Her account could be read as conveying her need to respond to this brutality in a way that will save her life (‘I have to run for my life’). She constructs herself as ‘victim’, thus connoting fleeing, helplessness and lack of choice. Of course, ‘run for my life’ could also be interpreted in terms of Altoise narrating her active choice to engage in activity to preserve her life (and thus articulating a capable, powerful self). Her representation of her sense of responsibility to herself that is signified by the need to protect her life is, however, situated against the backdrop of her account of self as vulnerable in relation to her abuser. The construction of a ‘fearful self’ that co-exists with her sense of increased power in relation to her abuser may create a kind of dissonance of self (‘vulnerable’ woman versus a ‘capable’ woman who can change this vulnerability). This dissonance, as we interpret it, is manifest in Altoise’s retrospective account of her ‘realisation’ that she had to act to save her life by leaving.

Below, Rochelle constructed her ‘fear of death’ as being the impetus for her to leave, regardless of the hardships that being a single mother may bring.

I just had to draw the line. It was like a wake up call for me with the last assault that um just because I was close to death I had to draw the line somewhere. September last year was just my line and that is how I ended up here (at the shelter), I had nowhere else to go.

(Rochelle)

Rochelle depicts her victimhood in terms of the severe assaults that she had limited physical power to shield. She speaks of ‘drawing the line’; this indicates her action of setting boundaries that delineated when the abuse was becoming too severe to bear (when she was almost killed). Here she draws on her moral power in outlining the level at which the abuse was unacceptable. Rochelle also articulates her sense of responsibility to set these boundaries as she says, ‘I had to draw the line somewhere’. The decision to leave can accord Rochelle more power as she is removed from the immediate danger of the abuse. This is, however, situated against the backdrop of her lack of socio-economic opportunities. The situation of women like Rochelle who are dependent on their abusers for finances and shelter is consistent with Scott’s (1990) account of the benefits that subordinate groups often receive if they maintain the silence of the hidden transcript. In this case Rochelle invests in her responsibility to protect herself over and above her identity of married/cohabiting woman. Staying with her abuser can be beneficial when it offers Rochelle more safety than going out into the world by herself. When staying in the relationship accords her more danger, she is then justified in choosing to leave. She constructs the danger as being severe enough to leave even though she had ‘nowhere to go’.⁴ This articulation of the danger serves to ‘explain’ how Rochelle’s investment in her identity of ‘wife’ was decreased. Rochelle instead invests in a more positive identity of ‘strong woman’ and ‘moral agent’ who must protect herself by leaving the abusive situation.

Lillian draws on the ‘battered woman’ script in her description of an abused, depressed self that lacks self-esteem. She suggests, however, that this self may be only part of the picture.

I didn’t realise that I was being abused by him until I started to have effects of depression and you know getting at that point where my self esteem was so low it looked even to everybody that I looked stupid. I looked like I had no confidence. It’s like my brains were closed up in a box.

(Lillian)

Lillian highlights the effects that the abuse had on her self confidence that led to her ‘looking stupid’ in front of others. Here she positions herself in relation to discourses of abused women as ‘pathological’, representing a social self that is less powerful in relation to her abuser. Negative positions such as these are reflective of discourses of women in general as ‘mad’/‘bad’ or as ‘good virgin’/‘bad whore’ (Macdonald, 1995), which offer women limited (often polarised) subject positions. Lillian stresses that ‘it looked’ as if she was stupid and ‘it looked’ as if she had no confidence. Thus, she sets up a disparity between how others position her due to the effects of the abuse and the fact that she does not position herself in this way. Lillian presents herself as clever (articulation of her sense of power) but the influence that the abuse had on her sense of self had an impact on her behaviour and interactions with others. This is signified by her statement that her brains were, ‘closed up in a box’. Thus, she sets up a contradiction between her sense of self and her sense of how others position her as ‘abused woman’.

Our findings concur with Enander (2011) in the sense that dissonance is a crucial aspect of the women’s accounts of their desire to leave the abuser. The availability of scripts of women’s power (that contradict the battered woman script) may facilitate such dissonance of self that may in turn result in changes in the way that women narrate a sense of their power in relation to others. The intersections of the public and hidden transcript created inconsistencies in the women’s narratives of self in abusive relationship. These inconsistencies are heightened as they invested in identities that accord them a greater sense of power in the world. The analysis showed that women complexly managed ‘moral selves’ as they explored and worked through certain contradictions of self that emerged along the way. Many women drew on these contradictions as a basis to ‘explain’ why they left their relationships and such ‘explanations’ seemed to be an important element for the women in their accounts of leaving. Overall, the women drew on the ‘battered woman transcript’ in their constructions of themselves as ‘powerless’ and ‘fearful’ victims in relation to their abusive partners. This ‘abused self’ was contrasted alongside alternate, more powerful subjectivities that the women drew upon at specific moments. The interaction between these particular scripts generated certain contradictions of self that seemed to play an important role in women’s constructions of their desire to leave.

Managing a moral self

Studies have shown that an increased awareness of the dynamics of the abusive situation leads to a surge of emotions that stand as a resource to leaving (Hydén, 1994; Kirkwood, 1993). In the current study the women narrated self as ‘good person’ and ‘moral agent’. They also highlighted contradictions between a more socially valued ‘good self’/mother and the ‘bad self’ – a pathological subjectivity that occurred as a result of the abuse and who intends to harm the abuser because of feelings of desperation and anger. In what follows, we explore the ways in which the women narrated their desire to leave as being the result of a complex ‘identity decision’ through which they invested in identities that accorded them more power at certain moments.

Self as a ‘good person’ and ‘moral agent’

Some women in the study spoke about managing a moral self. These narratives centred on the contradictions of self that their intent to harm the abuser generated. The women examined these contradictions, drawing on their moral power and responsibility. They invested in more socially valued subjectivities – such as non-violent identities – that may be less threatening to their sense of themselves as good woman and mother.

Mia spoke of the contradictions that her thoughts of killing the abuser generated in terms of her sense of herself as a moral agent in the world.

I’m not that type of person and that’s how I feel sometimes when he was drunk and beating me up. And then he sleep and then I feel maybe I must kill you now but my heart was never strong enough to do it. It was just not me. So I said to myself, ‘before I become a person that do it (murder), I must get out’.

(Mia)

In the extract above Mia narrates her thoughts of killing her abuser as being the result of violent episodes of drunken battering on her abuser’s part. Here she constructs an ‘abused self’ whose thoughts of murder are a consequence of her partner’s abusive behaviour, a response that she is powerless to control. She contrasts this with her sense of her ‘true’ nature (‘my heart was never strong enough to do it’). On the one hand, Mia’s narrative can be interpreted as being in line with the public transcript as she represents herself as being both negatively affected by the abuse and passive in the sense that she is not able to take this kind of action against the abuser. We interpret her account, however, in the context of the extreme nature of murder and the negative identity that is associated with a person who commits such an act. Mia draws on a ‘positive self’ that makes an active choice not to enact harm on her abuser. This is consistent with the hidden transcript through which she articulates an agentic self. Here she strengthens her identity and sense of herself as someone who is capable of evaluating a point of action and as capable of taking action by leaving. She invests in a self that is compatible with her own values and that may accord her more social power than the negative identity of murderer. She attempts to show that if she committed violence against her abuser, then her sense of self as moral agent would have been threatened. Interestingly, her narrative suggests that it is the prioritising of moral concerns that enabled her to leave her abuser.

Leaving occurs in the face of multiple social expectations, including the value of a father figure (see Peled and Gill, 2011; Shefer et al., 2008) and idealised notions of romantic heterosexual love (Jackson, 2001; Towns and Adams, 2000). Towns and Adams (2009) have highlighted certain ideological dilemmas that women face at the point of separation, such as the tension between collectivism and individualism. Relatedly, the women in the current study engaged in dialogue about the abuser’s dangerousness in order to justify their decision of aligning to the individualistic ideologies of leaving. In the example above, Mia explicitly mentions the physical threat that her partner posed to her. She aligns with individualistic ideologies first by wishing her abuser dead and thus positioning herself in an imaginary way as separate from him. Second, in her final statement that she ‘must get out’ she positions herself as a possible survivor who is capable of leaving the abuse. Both of these examples strongly contradict collectivist ideologies of romantic love (see Jackson, 2001). At the beginning of the excerpt, Mia outlines the specific context (‘drunk and beating me up’) within which her thoughts of killing her abuser and then her thoughts of leaving him emerged. This description serves as a backdrop setting (an exceptional context of brutal force against her), offering Mia a means with which to justify the positioning of self in line with these individualistic ideologies of separation.

Contradictions of self as mother

A concern for the effects that the abuse has on children has been found to be a significant turning point that motivates many women to leave a relationship (e.g. Khaw and Hardesty, 2007; Taylor, 2002). In the current investigation, women’s identities of mother seemed to play a pivotal role in their ‘managing a moral self’ by leaving.

I was having so much anger against him. I was starting to have these thoughts I must do something to this man when he’s sleeping. I can throw him with boiling water or something and when I’m alone (I) used to think, ‘hey this is wrong. I am going to end up in jail if I’m going to do something to him and my children will suffer. Let me rather leave him’.

(Maleka)

I tell for myself today is the day I’m going to do something to this man. And I take the tablets and the Ratex (rat poison) in my hand. That’s the day I decided I want to kill him … but you know, when I was standing there and I see my daughter’s face and I think who’s going to look after her if I (am) going to do something like that.

(Beryl)

In the extracts above Maleka and Beryl highlight the contradiction between their intention of harming the abuser (their abused, angry and, perhaps, desperate self) and their responsibility as mothers. Maleka constructs her intention for violence as being a disempowering effect that is caused by the abuse. Beryl narrates her intention to harm her abuser as being an active decision (‘that’s the day that I decided I want to kill him’) that is not merely a passive response to abuse. Beryl highlights her capacity to exert a negative change on her abuser’s life. This is signified through her intention to poison him. This construction of her own capacity to enact harm on her abuser suggests her increased sense of power in relation to him. Beryl chooses not to enact agency in this way because the end result may mean that she would not be able to take up the position of ‘good mother’. If, as in Beryl’s account, the construction of ‘angry abused wife’ gives her a sense of power, then investment in a particular identity becomes complex as the identity of mother also accords her power. Here Beryl works with the two different powerful storylines and invests in an identity that is more morally acceptable and, as such, may accord Beryl an increased sense of self-worth.

Both Maleka and Beryl construct their desire to leave as being the result of a rational choice in which they engaged in the strategy of ‘managing moral selves’ and not acting in a violent manner. Their moral stance is closely tied to what the consequences of harming the abuser will mean in terms of the impact this will have on their children’s lives, and the possible threat to their subjectivities as mothers. If they end up in jail, their identities of mother would be shattered as they would then be unable to care for their children. They construct their desire to leave as being fuelled by their identities as mothers, identities that are imbued with cultural meanings that relate to their moral responsibilities in the world (Davies and Harré, 1990). Their accounts show that women do not only draw on scripts of ‘emphasised femininity’ to ‘explain’ why they stay in abusive situations (Boonzaier and de la Rey, 2003), they also position themselves within these scripts to ‘explain’ why they leave. Ultimately, the women highlight contradictions between their bad ‘abused self’ who intends to commit violence and the ‘good self’/mother. They draw on identities of responsible, moral persons whose active choices may lead to a more favourable outcome in terms of facilitating their capacity to continue looking after their children. Such subjectivities may accord them more social value than the negative identity of ‘angry abused wife’ who commits harm.

Conclusion

The women in this study articulated the hidden transcript through ‘realising the danger’, ‘managing a moral self’ and drawing on ‘alternate selves’ – all of which constitute agentic storylines by which the women narrated a sense of their power in relation to the abuser. In this paper we show that women shelter residents construct their desire to leave their abusive relationship as involving a complex engagement between their powerless/pathological ‘abused woman’ self and their ‘powerful moral agent’ self. Their accounts show a fluctuating sense of their power in relation to the abuser. The women present themselves in agentic ways that contradict meanings around the pathological and powerless victim prescribed by the ‘battered women’ script. The women work through the contradictions between these two scripts and take on identities that accord them more power at specific moments as well as identities that are consistent with their sense of self at these moments. Ultimately, they invest in positive forms of subjectivity – selves that are capable of evaluating their situation and that have the control to choose to act in ways that will have positive consequences for their sense of self as culturally appropriate moral subject. These positionings may result in positive changes for their sense of self-worth and the ways in which they experience the world. In addition to the representation of self through language, this paper provides insight into the women’s lived experience as ‘abused women’.⁵ Constructions of self, manifest via the women’s use of language, also provide insight into the experience of ‘being subjected’ that occurs as a result of the ways in which the women are positioned as ‘abused women’, as well as the different ways in which they position themselves (Davies and Harré, 1990).

The analysis shows that constructing a self as vulnerable, ‘powerless victim’ in relation to the abuser is a crucial element in the women’s accounts of their desire to leave. Our findings are consistent with past research that has shown that different subject positions are taken up at various moments to achieve varied purposes (see Boonzaier and de la Rey, 2003). The women in the current study position themselves as ‘fearful’ and ‘powerless’. These constructions are narrated within the context of descriptions of the serious danger that the abuse posed to themselves as well as their children. A common theme is the women’s identities of mother and their concern for their children. What we hope this analysis has shown is not that women are trapped by one gendered subject position replacing another, but rather that they draw on subject positions that have more social value and that may accord them an increased sense of self-worth. By depicting staying in the relationship as more dangerous than leaving, the women could justify their decisions to leave and maintain positive social identities, even in the face of the imperative of femininity that positions women as being responsible for maintaining the family system.

The women’s investment in ‘powerful selves’ occurs at a specific moment in South Africa in which women are gaining increasing power in the public sphere. The broader context of poverty does, however, also colour their narrations of self (Brock, 1993, as cited in Profitt, 2000). The women’s situations of poverty mean that when they leave they are making a choice that could result in negative material consequences. Many describe having nowhere to go when they leave the abuser and do not have the finances to support their children alone. In this study, the women look back on their situations of abuse and, in doing so, narratively structure certain dissonances of self in relationship. This discord, or contradiction, between these two ‘selves’ seems to form the basis by which the women ‘explained’ why they left their relationships. The women paint a picture of the severe danger of the abusive environment and pair this with their sense of moral power and responsibility. This narration of the threat that the abuse poses to their sense of self and, in some cases, their physical selves, or lives, serves to equate the partners’ abuse as being as serious as the violation that they could experience outside the home (at the hands of other men or the violence of poverty). Here the women articulate that their investment in their identity of ‘wife’ decreased as this position no longer holds a sense of power and safety.

Our work contributes to the research that shows leaving as a process. It offers a unique angle for exploring the psychological aspects of abused women’s agency, one that interprets resistance as an active process of self-construction that is beyond mere visible action. We shed some light on Scott’s notion of power and resistance when applied to the issue of woman abuse. The women’s working through contradictions of self stands as Scott’s (1990) ‘rupture’ of the ‘normal world’ public transcript of battered women as powerless victims. It is at the site of contradiction that the public and the hidden transcript intersect. This contradiction of selves, according to our interpretation, constitutes a significant part of the women’s retrospective accounts of their desire to leave their abusive partners.

Women do not draw on discourses of power and resistance in any simple way that replaces dominant discourses; rather a complex negotiation exists between the two. Our utilisation of Scott’s concept of power relations provides us with a framework with which to analyse the ways in which constructions of self as powerful and powerless co-manifest at particular moments in the women’s retrospective accounts of leaving. In this way we offer a novel perspective on the complex psychological process of leaving an abusive heterosexual relationship and offer new knowledge about women’s subjectivities ‘post-abuse’. We also contribute fresh insights into the multifaceted issue of women’s agency in light of their narratives of their desire to leave these relationships within the broader South African context of poverty, deprivation and limited opportunities for marginalised women who may want to leave an abusive partner.

Footnotes

Acknowledgements

We would like to thank the women shelter residents who took part in the larger study on which this paper is based. These women gave up their time and spoke openly about their experiences.

Funding

The author(s) received no financial support for the research and/or authorship of this article.

Notes

Author Biographies

Samantha van Schalkwyk is a PhD candidate in Psychology and affiliate of Trauma, Forgiveness, & Reconciliation Studies at the University of the Free State. Her research interests include gender-based violence, community-based research methodologies, and the psychosocial aspects of women's agency within the African context.

Floretta Boonzaier is Senior Lecturer in the Department of Psychology at the University of Cape Town. Her primary research areas include explorations of the psychological and social landscape of gender-based violence in the South African context, including the intersections of raced, classed and gendered subjectivities.

Pumla Gobodo-Madikizela is a clinical psychologist and Senior Research Professor in Trauma, Forgiveness, & Reconciliation Studies at the University of the Free State. Her research has focused on the trauma of violence and understanding individual and collective strategies of working through trauma, and attempts at healing in the aftermath of trauma (including its intergenerational effects). Her interest in this area is on the reparative aspects of dialogue encounters between victims/survivors and perpetrators of violent abuse.

References

Abraham

(2000) Isolation as a form of marital violence: The South Asian immigrant experience. Journal of Social Distress and the Homeless 9(3): 221–236.

Anderson

(2000) Rape in South Africa. Georgetown Journal of Gender & Law 1: 789–821.

Anderson

Saunders

(2003) Leaving an abusive partner: An empirical view of predictors, the process of leaving, and psychological well-being. Trauma, Violence, and Abuse 4: 163–191.

Baker

(1997) “And I went back”: Battered women’s negotiation of choice. Journal of Contemporary Ethnography 26(1): 55–75.

Boonzaier

(2008) “If a man says you must sit, then you must sit”: The relational construction of woman abuse: Gender, subjectivity and violence. Feminism & Psychology 18(2): 183–202.

Boonzaier

de la Rey

(2003) “He’s a man and I’m a woman”: Cultural constructions of masculinity and femininity in South African women’s narratives of violence. Violence Against Women 9(8): 1003–1029.

Burke

Carlson Gielen

Mcdonnell

(2001) The process of ending abuse in intimate relationships: A qualitative exploration of the transtheoretical model. Violence Against Women 7(10): 1144–1163.

Collins

(2003) Social psychology and research methods. In: Ratele

Duncan

(eds) Social psychology: Identities and Relationships, Lansdowne, South Africa: University of Cape Town Press, pp. 23–41.

Davies

Harré

(1990) Positioning: The discursive production of selves. Journal for the Theory of Social Behaviour 20(1): 43–63.

10.

Davies

Browne

Gannon

(2006) Constituting the feminist subject in poststructuralist discourse. Feminism & Psychology 16(1): 87–103.

11.

Dutton

Painter

(1993) Emotional attachments in abusive relationships: A test of traumatic bonding theory. Violence and Victims 8(2): 105–120.

12.

Enander

(2010) Jekyl or Hyde or “Who is this guy?”-Battered women’s interpretations of their abusive partners as a mirror of opposite discourses. Women’s Studies International Forum 33: 81–90.

13.

Enander

(2011) Leaving Jekyll and Hyde: Emotion work in the context of intimate partner violence. Feminism & Psychology 21(1): 29–48.

14.

Enander

Holmberg

(2008) Why does she leave? The leaving process(es) of battered women. Health Care for Women International 29: 200–226.

15.

Foucault

(1982) Afterword: The subject and power. In: Dreyfus

Rabinow

(eds) Michel Foucault: Beyond Structuralism and Hermeneutics, Chicago: University of Chicago Press, pp. 208–226.

16.

Fuller

(2010) Mandela’s children. National Geographic 217: 80–108.

17.

Herman

(1992) Trauma and Recovery: From Domestic Abuse to Political Terror, New York, NY: Basic Books.

18.

Hydén

(1994) Woman Battering as Marital Act: The Construction of a Violent Marriage, Oslo: Scandinavian University Press.

19.

Hydén

(1999) The world of the fearful: Battered women’s narratives of leaving abusive husbands. Feminism & Psychology 9(4): 449–469.

20.

Hydén

(2005) “I must have been an idiot to let it go on”: Agency and positioning in battered women’s narratives of leaving. Feminism & Psychology 15(2): 169–188.

21.

Jackson

(2001) Happily never after: Young women’s stories of abuse in heterosexual love relationships. Feminism & Psychology 11: 305–321.

22.

Khaw

Hardesty

(2007) Theorizing the process of leaving: Turning points and trajectories in the stages of change. Family Relations 56: 413–425.

23.

Kirkwood

(1993) Leaving Abusive Partners: From the Scars of Survival to the Wisdom for Change, Newbury Park, CA: Sage.

24.

Landenburger

(1998) The dynamics of leaving and recovering from an abusive relationship. JOGNN 27(6): 700–706.

25.

Lempert

(1996) Women’s strategies for survival: Developing agency in abusive relationships. Journal of Family Violence 11(3): 269–289.

26.

Ludsin

Vetten

(2005) Spiral of Entrapment: Abused Women in Conflict with the Law, Johannesburg, South Africa: Jacana Media.

27.

Macdonald

(1995) Representing Women: Myths of Femininity in the Popular Media, London: Edward Arnold.

28.

Matthews

Abrahams

Jewkes

(2008) Intimate femicide-suicide in South Africa: A cross sectional study. Bulletin of the World Health Organisation 86(7): 552–558.

29.

Patzel

(2001) Women’s use of resources in leaving abusive relationships: A naturalistic inquiry. Issues in Mental Health Nursing 22(8): 729–747.

30.

Peled

Gil

(2011) The mothering perceptions of women abused by their partner. Violence Against Women 17(4): 457–479.

31.

Profitt

(1994) “Battered women” as “victims” and “survivors”: Creating space for resistance. Canadian Social Work Review 13(1): 23–38.

32.

Profit

(2000) Women Survivors, Psychological Trauma, and the Politics of Resistance, New York, NY: Haworth.

33.

Riessman

(2008) Narrative Methods for the Human Sciences, Thousand Oaks, CA: Sage.

34.

Scott

(1990) Domination and the Arts of Resistance, New Haven, CT, USA: Yale University Press.

35.

Shefer

(2004) Psychology and the regulation of gender. In: Hook

(eds) Critical Psychology, Landsdowne, South Africa: University of Cape Town Press, pp. 187–209.

36.

Shefer

Crawford

Strebel

(2008) Gender, power and resistance to change among two communities in the Western Cape, South Africa. Feminism & Psychology 18: 157–181.

37.

Taylor

(2002) “The straw that broke the camel’s back”: African American women’s strategies for disengaging from abusive relationships. Women & Therapy 25(3/4): 79–94.

38.

Towns

Adams

(2000) “If I really loved him enough, he would be okay”: Women’s accounts of male partner violence. Violence Against Women 6: 558–585.

39.

Towns

Adams

(2009) Staying quiet or getting out: Some ideological dilemmas faced by women who experience violence from male partners. British Journal of Social Psychology 48(4): 735–754.

40.

Walker

(1979a) How battering happens and how to stop it. In: Moore

(eds) Battered Women, Beverly Hills, CA: Sage, pp. 59–78.

41.

Walker

(1979b) The Battered Woman, New York, NY: Harper & Row.

42.

Weedon

(1987) Feminist Practice and Poststructuralist Theory, New York, NY: Basil Blackwell Ltd.

43.

Wuest

Merritt-Gray

(1999) Not going back: Sustaining the separation in the process of leaving abusive relationships. Violence Against Women 5(2): 110–133.

44.

Wuest

Merritt-Gray

(2001) Beyond survival: Reclaiming self after leaving an abusive partner. Canadian Journal of Nursing 32(4): 79–94.

45.

Yoshioka

(2008) The impact of the cultural context on the experience of domestic violence. In: Keeling

Mason

(eds) Domestic Violence: A Multi-Professional Approach for Healthcare Practitioners, Berkshire, UK: Open University Press, pp. 81–90.