An Exploration of Men?s Subjective Experiences of their Violence Toward their Intimate Partners URSULA LAU 9600790Y A research report submitted to the Faculty of Humanities, School of Human and Community Development, in partial fulfilment of the requirements for the degree of M.A. Research Psychology Supervisor: Garth Stevens ii DEDICATION In memory of my mother, Angela (1946-2007) iii To the men who have profoundly influenced my life: My father, William and My fianc?, Dylan iv DECLARATION I declare that An Exploration of Men?s Subjective Experiences of their Violence Toward their Intimate Partners is my own work and that all the sources cited have been indicated and acknowledged by means of complete references. ________________________URSULA LAU _______day of _________ 2007 v ACKNOWLEDGEMENTS To those who have guided me on this journey: Heavenly Father, for opening doors for me in this research and granting me the wisdom and insight to learn, appreciate and understand the experience of another. The research participants, for your courage in telling your stories, for trusting me with your experiences and for allowing me to enter your world. The co-ordinators of participating organisations, for your generosity, time and wonderful spirit in affording me access to the participants. Garth Stevens, for being my supervisor and mentor. Your insights have added richly to this work and your emotional support and encouragement have enabled me to persevere. Dylan, for your love, strength, hope and inspiration during my periods of frustration and disappointment as well my ?mirror? for joy. Angela, for teaching me compassion, humility, warmth, persistence and courage. Your exemplary character has been the source of my inspiration in my work and my studies. Mohamed Seedat, for your interest in my work, for your patience, encouragement and support for my learning. Claudia, Clive, Cindy and Michelle, for your quiet encouragement and unconditional love. Lorna, Anita, Brandon, Bonita, Carla, Terence, Samantha and Eno, for your enduring support and loyal friendship. Steve, for your guidance and encouragement through my young adult years. Nadira, for introducing me to feminist psychoanalysis. I would also like to express gratitude to DAAD (German Academic Exchange Service) and the WITS Postgraduate Merit Award for their scholarship support for the duration of this research. vi ABSTRACT The research served a dual purpose: (i) to explore men?s subjective experiences of their violence toward their intimate partners and, (ii) to examine how men talk about their violence in an attempt to establish credibility in their accounts. The first emphasised the subjective and emotional bases of individual experience and the second contextualised these descriptions within a broader societal framework. Highlighting the shortcomings of a quantitative research paradigm, the research utilised a qualitative framework which privileged first-person descriptions as the primary sources of subjective meaning. Although oriented toward a phenomenological approach, the research drew upon elements of psychoanalysis and discursive psychology. Twelve men were recruited from three organisations in Johannesburg. Via in-depth semi- structured interviews, men?s most vivid incident(s) of violence were explored. Thematic analysis revealed two levels of meaning: men?s descriptions of their violence (narrative content) and, processes by which they talked about their violence (narrative form). On the subjective dimension, seemingly contradictory experiences of violence were evident, clustering around five central themes: (i) violence as ?being out of control?, (ii) violence as ?having control? over another, (iii) the continuum of love and violence, (iv) violence versus emotionality and (v) the violent self as ?not me?. In feminist-psychoanalytic terms, men?s emotional dependence on their partners was denied or repressed. Violence represented a negation or devaluation of the feminine where male vulnerability and powerlessness, once exposed, became intolerable to bear. The ability to integrate and tolerate contradictory aspects of self (i.e. ?emotional? and ?rational?) was a decisive step towards healing and becoming the ?changed man?. On the discursive level, through ?talk?, men negotiated an identity of ?changed man? that provided distance from the ?violent self?. Attention to the narrative as a persuasive tool revealed ways in which the men attempted to establish credibility in their accounts of violence ? achieved by socially positioning themselves in relation to their violence, agreeing to talk and employing impression management ?strategies?, such as dissociations, justifications and confessions. Reconciling the two levels of analyses, the tension between dominant gendered discourses on masculinity that men relied on (i.e. that which fosters masculine ?toughness?, whilst diminishing ?weakness? or emotionality), and the psychological interior of their actual experiences was evident. A ?multiplicity approach? that accords significance to both societal constructions of gender and their impact on men?s behaviour, whilst giving expression to the psychological reality of men?s experiences could prove beneficial in fostering change. Keywords: men?s violence, subjective experiences, accounts of violence, phenomenology, masculinities, feminist-psychoanalysis, multiplicity. vii TABLE OF CONTENTS DEDICATION ii DECLARATION iv ACKNOWLEDGEMENTS v ABSTRACT vi TABLE OF CONTENTS vii CHAPTER 1: INTRODUCTION 1 Part I: Framing the Research 1 1.1 STATEMENT OF THE PROBLEM 1 1.2 CONTEXTUALISING THE RESEARCH 2 1.3 THEORETICAL FRAMEWORK 4 1.4 PURPOSE 5 1.5 METHODOLOGICAL APPROACH 5 Part II: Outline of Chapters 5 1.6 CHAPTER 2 5 1.7 CHAPTER 3 6 1.8 CHAPTER 4 6 1.9 CHAPTER 5 6 CHAPTER 2: LITERATURE REVIEW 8 Part I: Exploring Theoretical Perspectives 8 2.1 INTRODUCTION 8 2.2 BIOLOGICAL AND PSYCHOBIOLOGICAL PERSPECTIVES 8 2.3 PSYCHOLOGICAL PERSPECTIVES 10 2.4 SOCIOLOGICAL PERSPECTIVES 15 2.5 FEMINIST/SOCIO-POLITICAL PERSPECTIVES 16 Part II: Locating Male Violence within Gender and Masculinities 17 2.6 WHAT IS MASCULINITY? 17 2.7 GENDER AND SOCIAL CONSTRUCTIONISM 18 2.8 A GENDER RELATIONS APPROACH 19 2.9 A RETURN TO PSYCHOANALYSIS 22 2.10 A MULTIPLICITY APPROACH 25 2.11 CONCLUSION 26 Part III: Overview of Male Violence Research 26 2.12 INTRODUCTION 26 2.13 REVIEW OF QUANTITATIVE RESEARCH ON MALE VIOLENCE 27 2.14 REVIEW OF QUALITATIVE RESEARCH ON MALE VIOLENCE 30 2.15 CONCLUSION 41 viii CHAPTER 3: RESEARCH DESIGN AND METHODOLOGY 43 3.1 INTRODUCTION 43 3.2 THE DISCURSIVE VERSUS THE PHENOMENOLOGICAL 43 3.3 ?INTERPRETIVE RESEARCH?: LINKING THE ?EXPERIENCE-NEAR? TO THE ?EXPERIENCE-DISTANT? 44 3.4 METHODS APPROACH 45 3.5 DESIGNING THE INTERVIEW SCHEDULE 46 3.6 NEGOTIATING ACCESS 49 3.7 RECRUITING PARTICIPANTS 51 3.8 PARTICIPANT CHARACTERISTICS 52 3.9 ETHICAL CONSIDERATIONS 53 3.10 PILOTING 55 3.11 DATA ANALYSIS 56 3.12 CONCLUSION 61 CHAPTER 4: RESULTS 62 4.1 INTRODUCTION 62 Part I: Men?s Subjective Experiences of Violence 62 4.2 THEME 1: VIOLENCE AS ?LOSING CONTROL? 63 4.3 THEME 2: VIOLENCE AS ?HAVING CONTROL? OVER ANOTHER 66 4.4 THEME 3: THE CONTINUUM OF LOVE AND VIOLENCE 74 4.5 THEME 4: VIOLENCE VERSUS EMOTIONALITY 79 4.6 THEME 5: THE VIOLENT SELF AS ?NOT ME? 86 Part II: Establishing Credibility in Accounts of Violence 93 4.7 TALKING ?VIOLENCE? IN ITS CONTEXT 94 4.8 REFLEXIVITY 112 4.9 CONCLUSION 117 CHAPTER 5: CONCLUSION 118 5.1 INTRODUCTION 118 5.2 THE INTERIORITY OF THE VIOLENT EXPERIENCE 118 5.3 DISSOCIATIONS, JUSTIFICATIONS AND CONFESSIONS 124 5.4 RECOMMENDATIONS 126 5.5 LIMITATIONS 129 5.6 DIRECTIONS FOR FUTURE RESEARCH 132 5.7 SIGNIFICANCE OF THE RESEARCH 133 REFERENCES 134 APPENDIX A: ORGANISATION INFORMATION SHEET 150 APPENDIX B: PARTICIPANT INFORMATION SHEET 151 APPENDIX C: PARTICIPANT CONTACT AND SCREENING FORM 152 APPENDIX D: CONSENT FORM (INTERVIEW) 153 APPENDIX E: CONSENT FORM (RECORDING) 154 APPENDIX F: INTERVIEW GUIDE 155 APPENDIX G: TRANSCRIPTION NOTATION 157 APPENDIX H: INTERVIEW TRANSCRIPTS 158 CHAPTER 1: INTRODUCTION Part I: Framing the Research 1.1 STATEMENT OF THE PROBLEM Addressing the problem of intimate partner violence (IPV) at its source ? the men who perpetrate harmful acts against their partners (Wood, 2004) ? may yield more fruitful results in curbing the alarmingly high statistics of women abuse in South Africa. Interventions have typically been one-sided and victim-focused in approach, subsequently ostracising the rehabilitative needs of the perpetrators. International research into male violence against intimates, the majority emerging from a quantitative orientation, has generally been framed from socio-political and psychological ideologies (Carden, 1994). Qualitative studies have deepened understandings of violence by examining men?s personal accounts of violence. Those, framed from what Gadd (2000) terms a ?discourses of violence? approach, have identified the performative aspects of men?s talk, namely the verbal ?strategies? and ?devices? employed to account for their violence and how these are reflective of broader societal and gender ideologies that perpetuate woman abuse (Anderson & Umberson, 2001; Boonzaier & de la Rey, 2004; Cavanagh, Dobash, Dobash & Lewis, 2001; Fuller, 2001; Hearn, 1998; Hyden, 1994; Hyden & McCarthy, 1994; Ptacek, 1988; 1997; Scully & Marolla, 1984; Stamp & Sabourin, 1995; Totten, 2003). Whilst attention to the processes of men?s talk about deviance (Hearn, 1998) is crucial in presenting a critical slant to interpretation, the subjective meaning of experience in these studies tends to be minimised, diminished or accorded less significance (Goldner, Penn, Sheinberg & Walker, 1990). Studies oriented toward phenomenological and psychoanalytic-interpretive frameworks respectively, have shed light on various facets of the violent experience (the physiological, emotional and psychic dimensions) offering insight into the interiority of violence (Eisikovits & Buchbinder, 1997; Gadd, 2000; 2003; 2004; 2006; Reitz, 1999; Thomas, 2003). Whilst ?discourses of violence? studies tend to negate the essence of subjective experience on the emotional and psychic dimension, phenomenological studies tend to place less emphasis on men?s choice and personal accountability in relation to their violent acts. Incorporating both the emotional and performative elements, the present research reflects a dual purpose: To give voice to men?s subjective experiences of violence toward their intimate partners on a phenomenological level, and to highlight aspects of their talk about violence and its connections to social, cultural or gender ideologies. 2 1.2 CONTEXTUALISING THE RESEARCH According to Moffett (2007, p. 129), South African statistics for gender-based violence1 are notoriously high, particularly so ?for a country not at war?. The Unilever Institute of Strategic Marketing at the University of Cape Town (2004, as cited in Moffett, 2006) revealed that one in three women is likely to be raped in her lifetime. In a national study, the Medical Research Council revealed that one in four women in the general South African population has experienced physical violence at some point in their lives (Jewkes, Penn-Kekana, Levin, Ratsaka and Schreiber, 2001). Recently, a national study on female homicide further indicated that a woman is killed by her intimate partner every six hours (Mathews, Abrahams, Martin, Vetten, van der Merwe & Jewkes, 2004). Efforts at curbing the alarmingly high rates of women abuse have been zealously undertaken by various non-governmental organisations, rendering frontline assistance (e.g. sheltering, counselling and legal services to domestic violence victims), as well as facilitating preventative measures (advocacy and lobbying, empowerment initiatives, education and public awareness campaigns) (Vetten, 2005). However, as Padayachee (2005) argues, far-reaching interventions along with impressive achievements in areas of legal reform have failed to curb the rising statistics on women abuse. Research aimed at addressing knowledge gaps regarding the epidemic of woman abuse have also been undertaken. Studies framed from a public health perspective have identified poverty, alcohol abuse, having witnessed parental violence and societal norms (i.e. perceptions of abuse as acceptable) as potential risk factors for abuse (Abrahams, Jewkes, Hoffman & Laubscher, 2004; Abrahams, Jewkes, Laubscher & Hoffman, 2006; Jewkes, 2001; Jewkes, 2002; Swart, Seedat, Stevens & Ricardo, 2002). With stress on primary preventive programmes to address gender relations and societal/gender norms (Abrahams et al., 2004; Abrahams et al., 2006; Jewkes, 2002), rehabilitative goals are directed towards challenging men?s attitudes and gender stereotypes said to underlie woman abuse. Change in this sense is conceptualised as an attitude-behavioural shift (Schmidt, Kolodinsky, Carsten, Schmidt, Larson & MacLachlan, 2007). The continued prevalence of domestic violence, despite multi-level interventions signals that research and intervention efforts may have been too narrowly focused on female victims of abuse. Responsibility is typically vested with the victim, usually the woman, to extract herself from an abusive situation (Abrahams, Jewkes & Laubsher, 1999; Padayachee, 2005). In some instances, empowerment initiatives encouraging women to 1 Gender-based violence includes wide-ranging acts of violence against individuals on the basis of their gender identity. These include domestic violence, intimate partner violence, abuse of the girl child, human trafficking and violence instigated by homophobic reactions (Moffet, 2006, p. 131). 3 seek protection orders as measures of personal safety may incite further enactment of violence by their abusive partners. Focusing research and intervention initiatives on the needs of victims alone, therefore, clearly does not effectively deal with the complexities of domestic violence, particularly when the abusive behaviour on the part of the perpetrator has not successfully been resolved (Ko Ling, 2001). Such one-sided interventions, which seemingly aim to protect and empower victims of abuse may paradoxically alienate and ostracise the abuser (Wexler, 1999). Designating woman abuse as a ?woman?s problem? has also translated to less research conducted on the men who abuse (Abrahams et al., 1999). Curtailing the woman abuse epidemic therefore involves tackling the problem at its ?primary source? ? the men who enact abuse towards their intimate partners (Wood, 2004, p. 556). Recently, efforts to understand male perpetration of IPV is evidenced in the use of men as research subjects or participants (Abrahams & Jewkes, 2005; Abrahams et al., 1999; Boonzaier & de la Rey, 2004; Wood & Jewkes, 2001). In a study across six municipalities in Cape Town, Abrahams et al. (1999) reported a staggering 43.5% of male workers from a sample of 1394 men had abused their female partners, physical, sexually or both. Justifying violence against women, frequent partner conflict and conflict related to the man?s infidelity were associated with abuse in the past year. Associations between IPV and alcohol/drug use, early experiences of violence and use of violence in other settings was also evident. Given that some were involved in criminal activities, the authors reasoned that violence against women is a reflection of a generally violent South African society. In a subsequent study, Abrahams and Jewkes (2005) identified that witnessing the abuse of one?s mother during childhood was associated with violent adult behaviour (e.g. IPV, community violence, antisocial behaviour leading to arrest and possessing an illegal firearm leading to arrest). While contributions to male battery research have been extensive internationally, this area of research in South Africa appears to be in its infancy. National studies (Jewkes et al., 2001; Vetten, 1996; Mathews et al., 2004), although provide valuable insight into the scope of domestic violence, nevertheless fail to shed light on the intricacies, dynamics and complexities of IPV (Goodrum, Umberson & Anderson, 2001). Based on a need for a more thorough understanding of abusive men (Gelles, 1999), research exploring domestic violence from the subjective perspective of the perpetrator becomes pertinent. To some extent, these concerns are gradually being addressed via qualitative research. Utilising a gender framework, Boonzaier and de la Rey (2004) revealed that both men and women relied on varying gendered discourses (e.g. femininity as subordination, masculinity as authority) to make sense of violence in their relationships and, at times, positioned 4 themselves within and outside these dominant constructions. Men?s challenges during a time of economic uncertainty, unemployment and poverty reflected a struggle to achieve ?successful? masculinity (i.e. through socially prescribed roles of ?breadwinner? and ?provider?) (Boonzaier, 2005). Feelings of ?emasculation?, powerlessness and lack of tolerance for being controlled underpinned men?s use of violence, and was most pronounced where women were perceived as having a ?superior? status (i.e. earning capacity and educational level) to their men. Exploring connections between violence, sexual relationships and masculinity, Wood and Jewkes (2001) found, among Xhosa township youth, that masculine peer approval and status was defined in terms of men?s sexual conquests, their partners? sexual desirability and their ability to maintain control over them. Limited as these studies are, they nevertheless provide insight into the dynamics of violence from a subjective lens. As Julia Wood (2004) suggests, change may be precipitated via an understanding of the abusers? perceptions of themselves and their violent actions. Rather than impose various systems of meanings onto men that reflect minimal understanding of their needs and concerns, the goals of treatment should be ?tailor-made? and aligned with their views and perspectives (Ko Ling, 2001). Echoing this view, Hearn (1998) notes that ?in order to stop men?s violence towards known women, it is probably useful to understand how men understand violence? (p. 60). A thorough and detailed exploration of the ?in-the-moment? violent encounter would give rise to deeper understandings of abusive behaviour. On a subjective level, access to emotions and concomitant thoughts that fuel violence becomes possible. Following Goodrum et al. (2001), insight into violent men?s perceptions of their actions becomes an initiatory point from which rehabilitation proceeds. 1.3 THEORETICAL FRAMEWORK Understanding men?s violence toward their intimate partners (hereafter referred to as ?men?s violence?) has been explored via numerous theoretical avenues. These wide- ranging approaches traverse the disciplinary boundaries to reveal various orientations: biology, psychobiology, psychoanalysis, social learning theory, systemic theory, sociology and socio-political/feminist theory (Carden, 1994; Hearn, 1998). Miller?s claim (as cited in Harway & O?Neil, 1999, p. 9) that multidisciplinary explanations are essential in linking ?developmental and biological characteristics, personalities, sub-cultural variations, and economics, social, political, and community dimensions? becomes pertinent in this regard. In line with this rationale, a ?multiplicity approach? proposed by Goldner (1999; Goldner et al., 1990) is adopted. Such a stance assumes that one level of explanation does not supersede another but is allowed to co-exist alongside the other. Whilst diverse theoretical contributions are not dismissed, the present research was 5 situated between two seemingly divergent theoretical orientations, namely a masculinities framework and psychoanalysis. In so doing, both sociological and psychological dimensions are incorporated into a critical social psychological framework without diminishing their respective contributions (Ptacek, 1988). 1.4 PURPOSE The research aims were twofold: (i) to provide an in-depth description of the subjective experiences of heterosexual men who perpetrate(d) violence against their intimate partners and (ii) to engage in a critical analysis of the ways in which men produce credible accounts in their talk about their violence. Attending to both subjective experiences (in the former) and aspects of performativity (in the latter), the research attempted to understand the emotional dimensions of IPV, at the same time, being critical of the ways in which men attempt to favourably present themselves by negotiating a non-violent identity. In line with a multiplicity approach, this ?double agenda? (Goldner, 1999, p. 330), emphasises the psychological dimension, the ?power of feelings? (Chodorow, 1999, p. 6) that are often depicted as overwhelming in an experience of violence, and the moral discourses, which stresses choice and personal accountability. 1.5 METHODOLOGICAL APPROACH Adopting a qualitative research framework provided the basis upon which an exploration of thoughts and feelings, the foundation of experience (Seale, Gobo, Gubrium & Silverman, 2004) was possible. Although located broadly within an interpretive paradigm, the research traversed the boundaries of phenomenological (or ?experience-near?) and social constructionist (or ?distanciated?) approaches (Kelly, 1999). Connecting subjective experience with a broader interpretation reflecting theory, society, language, culture or history, understanding of experience from inside and outside the context is attained. Based on these considerations, the next section details areas of coverage within each chapter. Part II: Outline of Chapters 1.6 CHAPTER 2 Locating the present research necessitates a critical engagement with the existing theoretical frameworks from which explanations of men?s violence are derived. Addressing this aim, Chapter Two is divided into two sections. The first locates the various theoretical approaches in terms of their individualist and/or social emphases. In so doing, an attempt is made to position the present research within a masculinities and feminist psychoanalytic framework, thereby bridging the divide between intra-individual, interpersonal and social approaches in explaining men?s violence. Part II presents an 6 overview of male violence research reflecting the diverse theoretical orientations. Focus is directed from quantitative approaches towards a qualitative paradigm that draws upon the dimension of experience. 1.7 CHAPTER 3 Building upon the study?s theoretical rationale presented in the previous chapter, Chapter Three details its methodological and epistemological foundations. Although emphasising men?s subjective experiences, it nevertheless attempts to contextualise these within the broader social dimension. Cognisant of the tensions between these ?insider? or ?first- person? approaches (e.g. phenomenology) and ?outsider? or ?third-person? interpretations (e.g. social constructionism) (Kelly, 1999; 2006), the chapter attempts to reconcile these divergent aspects by situating the research broadly within an interpretive paradigm, embracing both the ?experience-near? and ?experience-distant? pillars of the ?interpretive continuum? (Kelly, 1999, p. 399). Based on these epistemological considerations, the chapter details the process of devising the in-depth interview schedule as a data collection tool and the decision-making steps undertaken in the thematic analysis of data, both in terms of narrative content and narrative form. Ethical considerations pertaining to recruitment, interviewing, analysis and writing up of the research is also detailed in this chapter. 1.8 CHAPTER 4 Based on the rationale and processes set out for employing thematic analysis in the previous chapter, Chapter Four is divided into two parts. ?Thick? descriptive analyses of men?s subjective experiences of their violence is detailed in the form of content themes in Part I. Whilst links to social discourses are made evident in men?s descriptions, the emotional bases of experiences are highlighted from a psychoanalytic lens. Part II connects subjective descriptions to the social context by highlighting the processes by which men talk about their violence. Through the act of talking, men constructed fluid identities, evolving from ?past abuser? to ?changed man?. Personal and epistemological reflexivity are also considered in relation to the ?rules of engagement? and the power dynamics between the researcher and participants. 1.9 CHAPTER 5 Locating the results in relation to broader theoretical debates, the final chapter attempts to highlight contributions of the present research in understanding male violence. Seeming contradictions in men?s experience of violence are addressed (i.e. ?losing control? and ?having control?, love and violence as a continuum of experience, and men?s power and men?s vulnerability), revealing the tension between social and cultural inscriptions of masculinity and the reality of men?s psychological experience (Goldner et 7 al., 1990). Implications for intervention are explored within these parameters as well as directions for future scholarship. Taking into account the principles of rigour in qualitative research, this final chapter also considers the limitations of the research. 8 CHAPTER 2: LITERATURE REVIEW Part I: Exploring Theoretical Perspectives 2.1 INTRODUCTION Understanding the phenomenon of male violence has been explored from diverse, multi- disciplinary theoretical positions. Carden (1994) situates psychological and feminist perspectives on opposing ends of a continuum of etiological models of violence. Psychological models of male violence have stressed individual, intrapsychic and developmental factors that give shape to the genesis of the abusive personality (Bowlby, 1984; Dutton & Bodnarchuk, 2005; Dutton, Saunders, Starzomski & Bartholomew, 1994; Fonagy, 1999; 2001; Fonagy, Moran & Target, 1993; Kohut,1972; Kohut & Wolf, 1978; Wexler, 1999; Zosky, 1999). Feminist or socio-political models, on the other hand, place emphasis on gender relations and power as part of a wider social system of oppression which sanction men?s domination of women (Dobash & Dobash, 1979; Bograd, 1988a; Hearn, 1999; Harway & O?Neil, 1999; Marin & Russo, 1999; Petrik, Olson & Subotnik, 1994; Pratto & Walker, 2004; Russell, 1995; Smith, 1990; Yllo, 2005). The aim of this section is to critically review explanations of male violence, as they focus on the individual dimensions to a consideration of the broader socio-political contributions. In so doing, the seeming disparate psychological and feminist models will be integrated with a view to offering what Goldner (1999; Goldner et al., 1990) terms a ?multiplicity? approach. A masculinities framework is proposed, which highlights masculinity as a product of dominant social discourses (Connell, 2001), as well as masculinity as an emotional and intrapsychic creation (Chodorow, 1995). Framed from a gendered perspective, feminist psychoanalysis provides insight into men?s violence by bridging the gap between the inner world of the individual and his outer world reflected in his relations with broader society (Craib, 1987). 2.2 BIOLOGICAL AND PSYCHOBIOLOGICAL PERSPECTIVES In the area of brain anatomy, associations between structural brain damage and behavioural expressions of violence have been found. In particular, the limbic system and frontal lobe areas have been implicated. Aggression and impulsive violence has been associated with a hyper-responsive amygdala, one of the structures situated in the limbic system that regulates emotions (Davidson, Putnam & Larson, 2000). Cortical lesion studies among individuals with damage to the frontal cortex, in particular, have 9 identified behavioural patterns, namely argumentativeness, failure to heed to consequences of actions, impulsivity and violence amongst others (Raine & Scerbo, 1991). In particular, injury in the orbito-frontal cortex has been linked to antisocial behaviour, such as aggression, impulsivity, social disinhibition and lack of empathy (Dinn & Harris, 2000; Grafman, Schwab, Warden, Pridgen, Brown & Salazar, 1996). Apart from criticisms levelled at the possible bi-directional nature of influences of brain dysfunction and violence, biological theories tend to neglect the social, psychological and situational influences that also influence violent behaviour (Raine & Scerbo, 1991). Genetic and endocrine theories are two other major fields of biological research which suggest that violence is underpinned by biological influences (Greene, 1999; Hearn, 1998). Genetic theories draw upon studies on the heritability of violent traits and the effect of genetic abnormalities on violent behaviour. Twin studies have shown some support for the heritability of ?criminal behaviour?, but not that of violent tendencies specifically (Volavka, as cited in Greene, 1999). The influence of abnormal chromosomal patterns on hyper-aggressive behaviour, moreover, is less evident. Men from prison populations presenting with chromosomal anomalies were also found to have lowered scores on intelligence and achievement tests relative to the general population (Tedeschi & Felson, as cited in Greene, 1999). To deduce that violent behaviour results from abnormal chromosomal patterns, therefore, would be to ignore these potentially confounding factors (Manning, as cited in Hearn, 1998). In endocrine theories, the role of testosterone (represented as androgen, the male hormone) has been implicated in men?s aggressive behaviour (Whitehead & Barrett, 2001). Men convicted of violent crimes evidenced higher testosterone levels compared to non-violent criminals (Booth & Osgood, as cited in Kanazawa, 2006). To isolate the role of hormones as an explanation for male violence, however, ignores the influence of social variables, which may also have a moderating influence. Given that testosterone levels are elevated in the face of challenges or competitive situations, the link between violence and testosterone is not clearly evident (Mazur and Booth, 1997; 1998). Moreover, as articulated by Clare (as cited in Whitehead & Barrett, 2001, p. 16): ?aggressive behaviour might cause higher levels of testosterone? rather than the inverse. Biological and psychobiological approaches to understanding male violence tend to focus on intra-individual contributing factors. Situating understanding ?in the biological body?, these perspectives are founded on ?Nature? and ?the natural? in explaining violence in men (Hearn, 1998, p. 17 & 19). In as much, by failing to account for social and cultural determinants or influences of male violence., such a stance neglects issues of morality 10 and accountability for violence (Hearn, 1998). Such approaches, moreover, support an evolutionary psychology that fails to consider diversity amongst men (and women). In suggesting that men are ?inclined? to violence by virtue of biological factors, moreover, such approaches justify male dominance and marginalisation of women (Whitehead & Barrett, 2001). Psychological perspectives, arguably, suffer from similar criticisms. However, while some approaches are narrowly focused on individual processes (Kohut, & Wolf, 1978; Fonagy, 1999; 2001), others have attempted to transcend the individual focus to incorporate interpersonal and social influences (e.g. social learning and systemic theory (Anderson & Schlossberg, 1999; Carden, 1994; Harway & O?Neil, 1999; Pagelow, as cited in Kalmuss, 1984; Zosky, 1999). 2.3 PSYCHOLOGICAL PERSPECTIVES Psychological theories can be grouped into psychoanalytic, social learning and systemic theories, each placing emphasis on different explanations for men?s violent behaviour (Carden, 1994). 2.3.1 Psychoanalytic Theory According to classical psychoanalysis, violent behaviour is seen as internal to the individual. With its emphasis on internal dynamics, violence is conceptualised as ?psychologically expressive?, as opposed to the feminist notion that violence is a form of instrumental control (Hearn, 1998, p.21). Freud (as cited in Fonagy et al., 1993) in particular emphasised that aggression was the ?anti-life? force that impedes ego development. Kernberg (as cited in Fonagy et al., 1993), proposed that feeling states of hatred, rage and disgust are synthesised into the aggressive drive structure. For these theorists, aggression is an innate aspect of man?s condition (Fonagy et al., 1993). Other theorists from the object relations and self-psychology schools, provide a more interpersonal emphasis, situating aggression as a reactive-protective response arising from external influences as opposed to innate drives (Kohut, 1972; Tesser, 1991). Premised on the assumption that early relationships provide the psychological ?template? upon which subsequent relationships are modelled (Zosky, 1999), various formulations have been proposed to understand male violence, for instance, arrested childhood development and narcissism (Kohut & Wolf, 1978; Wexler, 1999), disorganised attachment patterns and traumatic childhood (Dutton et al.,1994; Fonagy, 1999; Fonagy et al., 1993), and personality disorders (Dutton, 1995; Kernberg, as cited in Zosky, 1999; Lackhar, as cited in Zosky, 1999). 11 Kohut (1972) conceptualises violent behaviour as a response to potential injury to the self, one which threatens the cohesion of the self. Violence arises from a narcissistic injury originating from early deprivations of essential parental mirroring responses (i.e. approval and recognition), or experiences of constant criticism or ridicule. Failing to develop an internal sense of self-worth and competence, validation and approval is sought from external sources. In adult intimate relationships, the partner provides the mirroring function for his emotional needs. The need for emotional intimacy, however, conflicts with fears of being wounded. Perceived failures to obtain ?mirroring? responses from the intimate partner reawakens a fragile/fragmented self, provoking violent responses. Aggression (?narcissistic rage?), a manifestation of self-fragmentation in its extreme form, may also exist alongside other behaviours (i.e. gambling, substance abuse, reckless sexual behaviour) (Zosky, 1999). Focusing on early relationships, attachment theories also accord significance to the interpersonal domain. Unconscious representations of self and other (termed the ?internal working model?) are moulded by the infant?s subjective experiences in relation to caregivers, and becomes progressively influential in shaping relational patterns with others in later life (Bowlby, as cited in Worley, Walsh & Lewis, 2004). Disruptions (e.g. loss or separation from the caregiver) impede the bonding process, culminating for the infant in intense anxiety and anger (Carden, 1994; Fonagy, 1999; Fonagy et al., 1993). Threats of separation are met with powerful emotional reactions of terror, grief and rage which, if continued into childhood/adolescence, result in a proneness to extreme anger in adulthood (Bowlby, 1984; Dutton, 1995; Dutton et al., 1994). Male violence towards intimates is an attack on another who threatens alienation. It is an expression of protest behaviour or ?intimacy anger? (Dutton et al., 1994, p. 1367) in a desperate attempt to preserve the intimate bond (Denzin, 1984b). The connection between internal mental processes and the social environment is therefore clearly evident. Where psychoanalysis and social learning theory intersect, this connection becomes more apparent. Fonagy (1999) argues that, more than resulting from disrupted attachments, violent behaviour centres on the failure to ?mentalise?, in other words, to understand the mental and affective states of oneself and another. The decisive role of trauma in the psychogenesis of a violent personality (Carlson, 2005; Cohen, Brown, Smailes & Bernstein, as cited in Fonagy, 1999; Corvo, 2006; Dutton, 1995; Johnson; Rosenbaum & Leisring, 2003; Rosenbaum & O?Leary, 1981 as cited in Rosenbaum & Leisring, 2003) has been documented. Trauma enacted by an attachment figure impedes the developmental process and the capacity to ?mentalise?. As a coping response, the child compromises between an unsafe environment and the need for physical nurturance 12 from his caregiver. By avoiding the mental states and feelings of the other, he is ?distanced? from his caregiver?s intentions to harm or humiliate him. Consequently, his ?reflective capacities? are destroyed and ultimately his sense of self (p. 5). Subsequent attachment relationships are disorganised in an anxious attempt to seek physical closeness while maintaining mental detachment. Individualist psychological explanations, emphasising intrapsychic developmental processes, construe male violence as a form of pathological response to unmet psychological needs. However, as Jukes (1999) notes, the problem of male violence is highly prevalent, universal, and all too pervasive to be merely a manifestation of pathology. An integrated psychoanalytic-social learning model brings to the fore both individual and social explanations in understanding the ?intergenerational transmission of trauma? (Kalmuss, 1984; Hearn, 1998). Based on this formulation, men who directly experienced abuse or witnessed the abuse of a mother are likely to enact violence against their partners in adulthood (Hotaling & Sugarman, as cited in Hearn, 1998; Rosenbaum & O?Leary, 1981; Rosenbaum & Leisring, 2003). In (unconsciously) identifying with the abusive parent (usually the father), violent behaviour is modelled in subsequent relationships (Dutton, as cited in Lawson, 2001). Witnessing violence of a parent potentially damages trust in the victimised parent?s ability to provide a safe environment, subsequently threatening the security of the attachment bond. Dutton (1995) alludes to the child?s emotion of shame at his own impotence in the experience of abuse. This induced state of impotence results from suppressed rage or the inability to express emotion in the presence of the perpetrator. Consequently, such shaming experiences produce expressions of rage and violence in adulthood. Abusive behaviour is thus conceptualised as arising from ?arrested development? in childhood and is learned response to trauma (Dutton, 1995; Wexler, 1999). A social learning perspective also underpins moral developmental theory, although emphasis is placed on cognitive developmental processes (Gondolf, 1987). 2.3.2 Moral Developmental Theory Rooted in Piaget?s (as cited in Gondolf, 1987) cognitive developmental theories, Kohlberg?s (1987) theory of moral judgement proposes a three-level process which corresponds to six stages of moral development through which the individual advances. The ability to reason and make moral judgements is refined throughout this progression, such that behaviour is more consistent and humane at the highest stage. At the egocentric level, the self is seen as undifferentiated from the world, and where self-needs are given priority. Appropriate social engagements with the world lead to the realisation that others? exist with their own needs and feelings. This results in an orientation towards 13 the other (or an advance toward the concrete operational level). Abstract reasoning at the formal operational level is the ultimate achievement whereby principles are adopted that foster one?s favourable contribution to society. Moral reasoning and judgments are made based on the corresponding six stages: (i) physical consequences of obedience and punishment, (ii) na?ve egotism (or being able to respond occasionally to others? needs), (iii) orientation towards approval from others, (iv) orientation towards showing respect for authority, (v) recognising the law and elements of contract, and (vi) an orientation towards conscience, mutual respect and trust (Kohlberg, 1981). Advancement through the various stages may be impeded for certain reasons, such as the lack of stimulation by the social environment, or mental growth that is not fully accommodating of new forms of social interaction, or perhaps the delayed moral development of significant others who guide the individual (Kohlberg & Turiel, as cited in Gondolf, 1987). According to Gondolf (1987), the batterer is typically fixated at the egocentric (or ?me- orientated?) level of moral development. His own needs are reinforced at the expense of his partner who is the ?object? of his control. Failure to acknowledge the wrongness of his actions, along with tendencies to minimise or justify the abuse characterise this stage of defiance. Like psychoanalytic-social learning theories, Kohlberg?s moral developmental theory is cognisant of the socialisation processes that enhance or impede development. Shedding an alternative light on socialisation, however, it tends towards a more behaviouristic approach to social learning theory (Cohn & White, 1986). Violent behaviour is seen to be taught through dysfunctional behaviours in the family of origin (Gondolf, 1987). 2.3.3 Social Learning Theory As an alternative approach to the psychoanalytic-social learning model, social learning theory posits a vicarious learning of violence. Directing focus on ?violence as external sense data that are observable and reproduced, replicated or imitated over time? (Hearn, 1998, p. 24) therefore shifts focus from psychological explanations of male violence towards the role of social and cultural factors. While both psychoanalytic and social learning perspectives emphasise that childhood experiences form the basis of adult personality, the latter approach addresses the contexts of learning violence beyond the confines of the family unit. Aggressive or violent behaviour between family members not only communicates the appropriateness of such behaviour within the family, but also serves as a model for learning violent behaviour (Pagelow, as cited in Kalmuss, 1984). Outside the home, school and the media also provide the contexts in which violence is learned ? whether this is directly consequent to aggressive acts being reinforced or through vicarious observation of others being rewarded for behaving violently (Hearn, 1998). In demonstrating the recursive nature of violence (i.e. its transcendence beyond 14 the individual and its transmission across time to create an ?environment?, ?culture? or ?system? of violence) (Anderson & Schlossberg, 1999; Hearn, 1998), social learning theories overcome criticisms of psychoanalytic explanations that isolate individual personality from the social world (Hearn, 1998). Systems theory explores the interpersonal domain of the intimate relationship, while at the same time locates it within the situational and social context which give shape to the conflictual patterns of interaction (Anderson & Schlossberg, 1999; Harway & O?Neil, 1999). 2.3.4 Systems Theory Systems theory proposes a bi-dimensional understanding of male abusive behaviour, particularly taking into account how contextual factors (social, cultural, familial and individual levels) are moderated by the dynamics and patterns of interaction that occur within the family system (Goldner et al., 1990). With reference to these ?transactional sequences?, wife abuse is understood as ?a relationship issue with both parties participating (although not necessarily participating equally) in the violent sequence? (Neidig, as cited in Carden, 1994, p. 556). Although the systemic conceptualisation provides an interpersonal perspective on social and relational contexts, the uniqueness of historical and situational variables, as well as the patterns of interaction, is underscored (Anderson & Schlossberg, 1999). Socioeconomic factors, such as educational level, occupational status and level of income provide the contextual background for male abusive behaviour (Hotaling & Sugarman, as cited in Anderson & Schlossberg, 1999; Beasley & Stoltenberg, 1992). Interactional patterns associated with battering, according to Anderson & Schlossberg (1999), reflect rigid relational styles, involving hostile and angry behaviours, deficiencies in communication abilities, negative verbal interactions such as criticism, blame and threats. Where the couple?s poor problem solving and communication skills are placed alongside contextual factors (e.g. the economically disempowered husband), violence becomes a strategic attempt to deal with difficulties (Babcock, Waltz, Jacobson & Gottman, as cited in Anderson & Schlossberg, 1999). An important aspect of systems theory is the premise of neutrality adopted in relation to both partners of the relational unit (Scheel & Ivey, 1988, p. 316). This ?non-blaming? stance has been criticised severely by feminism, arguing that such an approach absolves perpetrator accountability and implicates the female partner as a co-instigator in the violence process (Hansen & Harway, 1993). Despite criticisms the systemic perspective moves away from an individual focus to examining the responses of both parties to a relationship and brings to spotlight the context in which violence occurs (Anderson & Schlossberg, 1999). The 15 multilayered context (biological, psychological and societal levels in the analysis of violence) is similarly advocated by sociological perspectives (Harway & O?Neil, 1999). 2.4 SOCIOLOGICAL PERSPECTIVES Although the role of individual psychological processes are not refuted in the analysis of male violence, sociological perspectives tend to minimise these factors in favour of explaining violence as ?an outcome of myriad characteristics of the social order, such as a lack of community, economic inequality, and particularly the characteristics of family as a social institution? (Loeske, Gelles & Cavanaugh, 2005, p.2). 2.4.1 Structural theories Stressing the relationships between poverty and class disparities (as indicated by income, educational level or occupation) (Evans, 2005), structural approaches consider violence as the outcome of social structures reflective of conditions within a society (Rupesinghe, 1994). Whilst violence is seen to affect all social classes, it is individuals from lower socio-economic backgrounds that present with the greatest risk (World Health Organisation, as cited in Evans, 2005). More specifically, as Evans (2005) details, domestic violence in low income groups occurs not only with a higher frequency, but also with greater severity. 2.4.2 Cultural theories The ?culture of violence? theory presupposes that violence is unevenly distributed in society. Conceptualised as a learned response, violence is regarded as a feature of marginalised subcultures (such as lower socioeconomic groups or ?ethnic? groups in ghetto areas) (Witt, 1987). Applied to ?culture? in a broader sense, this approach suggests that violence is reinforced by a society that approves of and justifies it, such that violence is the normed response to stressful situations (Owen & Strauss, 1975). Cultural norms or rules suggest a precedent for violence most notable in the sanctioning of violence as discipline, the notion of ownership of one?s spouse or children or the reinforcement of the privacy of family (e.g. keeping it ?behind closed doors?) (Witt, 1987). 2.4.3 Conflict theories Conflict theory proposes that violence is directly influenced by economic values of a culture. With its Marxist roots, emphasis is placed on class differences and uneven social relations, the struggle over scarce resources and the competition for wealth and power, resulting in the exploitation of the rich by the poor. Violence is consequently the result of frustration and desperation experienced by the lower classes (Brown, 1998). Applied to family violence, resource theory was proposed as a theoretical approach postulating a link between income and violence (Gelles, 1985). Male violence is conceptualised as the 16 ultimate recourse to control in the absence of other means, such as through the use of finances and material goods. By deduction, men of lower socioeconomic status are at higher ?risk? for utilising violence as a form of control (Anderson, as cited in Loeske, 2005). Thus, elements such as unemployment or financial stress, serve to place strain on the family and, more particularly on the man who may experience failure in fulfilling his socialised provider role (Witt, 1987). With its emphasis on social structures, social forces and social processes, sociological explanations offer an alternative understanding of male violence by distancing itself from individual processes (Loeske et al., 2005). Nevertheless, as Hearn (1998) argues, while social structural explanations of violence underscore the importance of the social context, they nevertheless ignore a fundamental aspect of societal organisation ? a gendered system of oppression. As with systems perspectives, similar criticism is levelled at sociological perspectives for their failure to address the power dimensions underlying traditional gender roles, as well as women?s position relative to men in a patriarchal society (Anderson & Schlossberg, 1999; Zosky, 1999). 2.5 FEMINIST/SOCIO-POLITICAL PERSPECTIVES The socio-political view, adopted by feminism (Carden, 1994), emphasises the intimate connection between domestic violence and patriarchal structures, such as social, political and cultural systems which govern language, beliefs, values, norms, skills, habits, customs, laws and institutions (Harway & O?Neil, 1999; Marin & Russo, 1999; Russell, 1995). Underscoring the gendered aspects of male violence (Hearn, 1998), the feminist position implies that patriarchy comprises of ?a structure, in which men have more power and privilege over women, and an ideology that legitimises this arrangement? (Smith, 1990, p. 257) [emphasis added]. Male violence toward women is conceptualised as the use of power and control to assert values of male privilege, entitlement and domination over women (Hearn, 1999; Marin & Russo, 1999; Petrik et al., 1994). As Dobash and Dobash (1979, p. 24) substantiate, men who abuse subscribe to western cultural ideals of male aggressiveness and female subordination, such that physical force is used to enforce their dominance. What is proposed here is a perspective which locates men?s violence toward women on a societal level by highlighting the ?power, position and practices of men? (Whitehead & Barrett, 2001). Embedded within feminist theory is a gendered perspective, also known as ?the daughter? of feminist theory (Handrahan, 1999) which explores how categories of gender and sex pervade every aspect of human life. Emphasis is placed on shared cultural ideologies (e.g. gender stereotypes) that legitimise men and women?s disparate positions of power (Pratto & Walker, 2004). Given the focus of the present research, a masculinities framework has been adopted. The following 17 section will highlight the relevance that masculinity theory as a gendered perspective has for understanding men?s violence toward their intimate partners. In line with a ?multiplicity approach? (Goldner, 1999), an attempt is made to link a masculinities perspective with a feminist psychoanalytic one, allowing exploration of both societal and intrapsychic processes that sustain and underpin male violence. Part II: Locating Male Violence within Gender and Masculinities Masculinity theory, derived from its ?feminist parentage? (Whitehead & Barrett, 2001, p. 3) and emerging from the broader gender theory situates men and masculinities and their relationship to power at the centre of enquiry. Accordingly, ?men? represents not a homogenous category, but by virtue of disparate positions within a society, have differential standings relative to power (Connell, 2000). 2.6 WHAT IS MASCULINITY? Although somewhat difficult to define given its multiple representation and transformation by various discourses, Whitehead and Barrett (2001) offer the following definition of masculinities: ?those behaviours, languages and practices, existing in specific cultural and organisational locations, which are commonly associated with males and thus culturally defined as not feminine? (p. 15-16). Connell (2000; 2001; 2005) traces different attempts to define masculinity. Essentialist definitions emphasise a core feature(s) of masculinity that presupposes a fundamental and universal quality or characteristic that is seen as persistent and internal (e.g. men being ?risk-takers? or ?aggressive?) (Bohan, 1997; Fuss, 1989). In gendered terms, men are assumed to be ?born different? from women or to have distinct ?natural? qualities (Davis & Gergen, 1997, p.4). However, as Cornwall and Lindisfarne (1994) suggest, the male/female dichotomy fails to explain how individuals are gendered differently depending on contexts. ?Multiple masculinities?, a term proposed by Connell (2000; 2001; 2005), emphasises the differing positions (sometimes oppressive) which men occupy within society by virtue of their sexual orientation, culture, ?race?, age, religion, etc. Positivist social science, embracing tenets of essentialism, provides an understanding of masculinity which accordingly reflects a so-called ?neutral? or ?objective? definition. Preoccupied with statistics, this stance posits that the differences between men and women are based on ?facts? (Bohan, 2002; Connell, 2000; 2001; 2005). Borrowing from the psychoanalytic notion that contradictions are contained within personality, however, Connell (2001) is quick to observe that ?to define masculinity as what-men-empirically- are is to rule out the usage in which we call some women ?masculine? and some men 18 ?feminine?? (p. 32). Normative conceptualisations of masculinity similarly hark back to essentialism in prescribing masculine behaviours for men. Sex role theory, for instance, highlights the socialisation aspect of masculinity, pointing towards the role of family and stereotyped representations of men in popular media in contributing to the male ?gender personality? (Connell, 2001; Leach, 1994). Although moving away from a ?biological determinism? (Bohan, 2002) towards a more social definition of masculinity, sex role theory assumes a complementary relationship between masculinity and femininity (i.e. based on gender differences) and offers limited scope for an analysis of power relations as it exists in societal structures (Leach, 1994). A semiotic approach, according to Connell (2000, 2001; 2005) gears toward a more abstract and symbolic contrast of masculinity and femininity. In other words, masculinity is defined as ?not femininity? or, following Lacanian psychoanalysis, it is signified by the phallus (?the name of the father? imposed through language and the law), whereas femininity is marked by lack or absence or incompleteness (Hook, 2006). Beyond the level of discourse, however, this approach offers a view of masculinity that does not fully encapsulate masculinity in its diverse forms. What is proposed by Connell (2001) is that masculinity needs to be considered in a range of contexts, what he terms places: ?gendered places in production and consumption, places in institutions and in natural environments, places in social and military struggles? (p. 33). Central to the present research is the place of negotiation between men and women, the relationships through which ?gendered lives? are enacted. Masculinity, therefore, is also, to use Connell?s (2001) terminology, ?a place in gender relations? (p. 33). 2.7 GENDER AND SOCIAL CONSTRUCTIONISM Shifting away from essentialist notions of masculinity/femininity, a social constructionist approach turns the spotlight on ?gender performances? in social life (Maracek, Crawford & Popp, 2004, p. 192). As Connell (2005, p. 6) notes, gender, rather than being fixed, is ?done? or ?accomplished?. ?Being a man? is not fixed. Its meaning emerges in human interactions, in social activities and language practices. Meanings are thus ?co-constructed? (Maracek et al., 2004). By implication, the unstable and incoherent nature of individuals? accounts reflect the shifting positions, situations, relationships and contexts that make up the continuity of social interaction. Social constructionism, therefore, is sceptical of universal claims about gender and sex, positivist approaches that reduce masculinity/femininity to a set of scores and evolutionary approaches that separate biology from language and culture (Maracek et al., 2004). This shift, therefore, exemplifies a move away from the individual focus to the social dimension to consider the role of power in the construction of gender (Bohan, 1997). In this sense, the ?social mind? has been advocated by social constructionism to reflect the ?indissolubility of psyche and culture? (Maracek et al., 2004, p. 198). 19 2.8 A GENDER RELATIONS APPROACH A gender relations approach, according to Connell (2000), lends insight into the various dimensions of gender, the relation between (men and women?s) bodies and society or what he terms the ?configuration of gender? (p. 24). The dimensions of gender refer to the patterns in social relations, in other words, gender as a social structure (Connell, 2000; 2001). Gender relations then constitutes a dominant structure of all documented societies. To belabour this point, Beall, Eagly and Sternberg (2004) assert that gender is the first organising principle upon which all subsequent interactions between people take place. Connell (2001) asserts that these processes are organised through the ?reproductive arena?. In other words, as a social practice, gender constantly makes reference to ?bodies and what bodies do? (p. 34), but in the sense of historical processes relating to the body rather than the notion of the body as ?biological base?. As Whitehead and Barrett (2001) note, these reflect social and cultural expectations of behaviour as opposed to biological determinants. To understand how these processes apply in gender practice is to consider how discourse (imposed by culture, society or social group) informs how gender is organised (e.g. how masculinities are constructed in the media, a culture or community) (Whitehead, 2001), or how institutions such as state, workplace, or school are ?configured? in terms of gender (e.g. the state as a masculine institution) (Connell, 2001). Relating back to the body, behaviour in everyday life is characterised by human bodily structures and processes of reproduction (i.e. sex difference/similarity, sexual arousal, intercourse, childbirth) (Connell, 2000). A gender relations framework, therefore, provides a useful basis from which to consider and understand men?s violence toward their intimate partners. To elucidate the ideas raised by Connell (2000; 2001; 2005), it is necessary at this juncture to derive knowledge regarding how gender is structured and how such an arrangement potentially supports a pattern of men?s violence against women. 2.8.1 Power relations and hegemony Related to the notion of patriarchy, is what Gramsci (as cited in Gadd, 2003, p. 334) terms a ?hegemonic masculinity?. Emphasising Marxist notions, Antonio Gramsci (as cited in Hearn, 2004) explained hegemony in terms of the power of the dominant economic ruling class to control society via an interrelated political conglomerate comprising the state, the law, capitalists and intellectuals. Recently, the term has been related to a gender system, rather than an economic system as proposed by Gramsci (Hearn, 2004). Hegemonic ideologies serve to legitimate the interests of those holding power, at the same time marginalising the interests of the less powerful (Wetherell & Edley, 1999). This understanding is adopted in the current research, with emphasis on ?hegemonic 20 masculinity? in relation to its intimate connection with the subordination of women (Connell, as cited in Wetherell & Edley, 1999). To quote Hearn?s (2004) restatement of Connell (1995), it is ?a form of masculinity or configuration of gender practice which is in contrast to other less dominant or subordinated forms of masculinity ? complicit, subordinated, marginalised? (p. 55). Such subordinated forms of masculinity comprise, for instance, the effeminate or homosexual male (Carrigan, Connell & Lee, 1985). Evolving slightly in terms of its current usage, ?hegemonic masculinity? as defined by Connell (as cited in Hearn, 2004) reintroduces the Gramscian understanding of the relationship between cultural ideal and institutional power: ?the configuration of gender practice which embodies the currently accepted answer to the problem of legitimacy of patriarchy, which guarantees (or is taken to guarantee) the dominant position of men and the subordination of women? (p. 77). Such a dynamic legitimates patriarchy?s reinforcement of men as dominant and women as subordinate by virtue of its ?claim to authority?. Whilst violence or physical aggression does not necessarily constitute hegemony, it nevertheless reinforces its authority. To illustrate, the reasoning of ?male as rational? and ?female as emotional? reflects a well established patriarchal ideology (Seidler, 1994). Masculinity, then, assumes a hegemonic status in its claim to have reasoning power, attaining legitimacy in representing societal interests as a whole, whether in macro institutions of state or in micro structures of the family (Connell, 2005). ?Hegemonic masculinity? thus presents a powerful concept in shedding light on how male domination and female subordination is reinforced on various societal levels. What emerges here is that power may be understood in three ways: (i) as a physical force, (ii) as relational and positional (i.e. men holding positions of power), or (iii) as exercised in relation to discourse (Whitehead & Barrett, 2001). Taking up this last point, male identity and its validation through discourse needs to be considered. In line with the interests of the present research, the point is how masculinity discourses reinforce power relations between men and women as they are manifest in intimate partner violence. 2.8.2 Masculinity and discourse Whitehead and Barret (2001, p. 15) conceptualise masculinities as fluid, plural and historically shaped by dominant ideologies or discourses of ?masculinism?. The concept of discourse is simply defined by Foucault (as cited in Whitehead, 2001, p. 352): The subject constitutes himself in an active fashion, by the practices of self, which are nevertheless not something that the individual invents by himself. They are patterns that he finds in his culture and which are proposed, suggested and imposed on him by his culture, his society and his social group. 21 Echoing Foucault, Connell (2005) points out that masculinity, as a discourse, rather than an isolated or individual construction, is the ?collective work of a group? (p.168), whether it be a culture, a community or peer group. Hekman (as cited in Whitehead & Barrett, 2001) asserts that dominant gender discourses provide validation for a ?malelist? perspective of the world, one that is reified as reality, thus taken as ?truths? in relation to interactions with others. As such, discourses then, more than ways of speaking, operate as powerful mediums which convey particular knowledges, validating some forms whilst nullifying others (Whitehead & Barrett, 2001). Adopting particular discourse such as ?man as strong/woman as weak?, ?man as dominant/woman as weak?, man as disciplinarian/woman as undisciplined?, ?male sex drive? (Boonzaier, 2001; Hollway, 1984; Whitehead & Barrett, 2001, p. 17; 21; Wood, 2004) may not only reinforce gender inequalities, but also legitimate men?s violence against women. Greig (2001) outlines four main themes of discourses which reflect an intersection between men?s violence and constructions of masculinity. The first relates to male emotional suppression, the second to the role of culture/society in both emotional suppression and socialisation of violence (i.e. constructions of ?tough? and ?brave? men who are able to maintain physical and emotional stoicism) (Umberson, Anderson, Williams & Chen, 2003), and the third to the insecurity attached to the masculine identity. As articulated by Heise (as cited in Greig, 2001, p.5), these masculine constructions advance and perpetuate male abusive behaviour: ?Men in many cultures wage daily battle to prove to themselves and others that they qualify for inclusion in the esteemed category ?male?, is to be reduced to the status of woman or, worse, to be ?queer??. Reflecting what Kaufman (1994) terms the ?paradox of men?s power?, contradictory states of both power and powerlessness manifest in the lives of men. In adopting the notion of power as a form of domination and control, men have alienated more nurturing aspects of themselves, as well as internalising the fear of failure. According to Kauffman (1994), in these lies the root of powerlessness. These insights have precipitated a shift in focus from sole preoccupation with patriarchy to a more gendered emphasis, specifically gender socialisation and gender identity (Greig, 2001). The notion of patriarchy, however, is by no means negated. What discourses of masculinity/ies have revealed is that men are not merely, in Greig?s (2001) terms, ?agents of patriarchy?. Rather, they are moulded by ?gender pressures? that could result in violence. An examination of these discourses, therefore, provides insight into the influence of patriarchy on individual life and, in some ways, apportions blame for men?s violence onto various societal levels rather than directing it solely onto men (Greig, 2001). 22 Whilst such a framework offers valuable insights into the location of men?s violence on the broader level of patriarchy and discourse, it nevertheless disregards the more personal and subjective aspect of gender. For instance, Chodorow (1995; 1999, p. 115) notes that any gender-related category ? be it man, woman, father, sister, brother, femininity, masculinity ? develops meaning not merely from language alone, but also from ?personally experienced emotion and fantasy in relation to a person connected to that label?. Whilst socio-political perspectives transcend an intra-individual focus of male violence by looking toward structured power and oppression in society, their ?singular focus on power and inequality? (Goldner, 1999, p. 326), however, not only ?de- psychologise[s] battering? (Rosenbaum & Leisring, 2003, p. 8), but also overlooks the interrelations between these various factors. More importantly, premised on the notion that men are socialised into exerting power and control, socio-political models fail to account for why men in the majority are not violent toward their partners (Zosky, 1999). The macro-societal focus therefore disregards the intrapsychic and relational dynamics of violent relationships, whilst classical psychoanalytic approaches are lacking in reference to broader societal influences. Whilst a feminist account of gender, according to Flax (1993, p. 51), differs from a psychoanalytic one in terms of its significant weighting of ?the role of relations of domination in the production and maintenance of gender?, Chodorow (1999) nevertheless locates the experiences of gender domination in emotional and intrapsychic life and in the developmental context of primary relationships. The bidirectional influence of feminism and psychoanalysis on each other, therefore, seem to inform theories of gender (Harcourt, 1994). The psychology of men and women, proposed by Chodorow (as cited in Bell, 2004, p. 153) is ?an intertwined conflictual whole, as part of a totality of social and psychological relations?. Relating back to Carden?s (1994, p. 551) ?perceptual continuum of the etiology of wife abuse? (with feminist and psychological models on polar ends), an integration of these seemingly contradictory perspectives may therefore provide a link between understanding the inner world of the male batterer, as well as, its relation to the broader societal context. In marrying the two perspectives, feminist psychoanalysis is also located within a gender theoretical framework. 2.9 A RETURN TO PSYCHOANALYSIS The gap between the inner world of the individual and the outer world reflected in relations with broader society is bridged in Craib?s (1987; 1990, 2001) integration of psychoanalysis and social theory with a specific focus on male dominance and masculine identity. He proposes that object relations theory offers ?the possibility of looking at intersubjectivity as both a social-historical and interpersonal psychological 23 phenomenon? (p. 174). The notion of intersubjectivity, defined as the recognition of the self as being interconnected with another, but also as distinct and separate (Benjamin, 1990) is explicated in the merging of psychoanalysis and feminism which provides an account of male dominance and violence. Termed by Carrigan et al. (1985, p. 590) as the ?psychodynamics of masculinity?, the connection between the social structure and psyche becomes apparent. Linking these to Connell?s notions of gender, he proposes that masculinity/femininity, rather than ascribed to ?essences?, represent the ?cathexic components of experience that men and women negotiate subjectively and intersubjectively? (Gadd, 2003, p. 334). Chodorow (1989) locates the experiences of gender domination in emotional and intrapsychic life and in the developmental context of primary relationships. The social context for the development of the self and the acquiring of gender is the captured in Chodorow?s concept of ?assymetrical parenting? in which both boys and girls learn to define themselves based on their relationship to ?a single, psychologically gendered woman? (Goldner et al., 1990, p. 581). The primacy of the mother figure and the relative absence of the father has significance for the development of distinct identities for both sexes (Chodorow, 1978). Gender identity for the boy child is tied up in the process of differentiation or separation-individuation (i.e. the gradual perception that the self/subject is distinct from the other/object). In contrast to the girl whose separation from the mother is inhibited due to her unique identification with her, the boy is propelled into separation from his primary caregiver towards an identification with the father representing the public domain of status and power. The primacy of the mother figure during the early developmental stages coupled with the unavailability of the father figure ? a concrete figure of identification ? the boy comes to learn masculinity by learning to be ?not- feminine?. Without the concrete male figure, an ?image of masculinity? is internalised, rather than the male person as a whole. Masculinity for the boy is therefore perceived as essential to his self-identity (Craib, 1990). To be ?not-feminine? therefore means the repudiation of all feminine aspects, including strong emotions of empathy and dependency. Dominance is thus derived from the denial of dependency and the need to prove his masculinity for fear of losing it (Hagemann-White & Micus, 1999). Borrowing from Chodorow?s (1989) assertion of the primacy of the mothering in the gender development, Benjamin (1990) explores the dynamics of power in relationships, by tracing the origin of dominance and submission between the sexes to this early relationship (Bell, 2004). The mother is regarded as the ?bestower of recognition? (Benjamin, 1980, p. 144), a fundamental human need connected to survival and intimacy with another. Attaining ?mutual recognition?, in other words, recognising that another is 24 similar but separate and distinct from oneself (i.e. ??you? who are ?mine? are also different, new, outside of me?) (Benjamin, 1990, p. 15) is the basis of her intersubjective theory. Developmentally, the conflict lies between the child?s need to attain an autonomous identity and the need for recognition by another. This ?paradox of recognition? is encapsulated in Benajmin?s (1990, p. 33) assertion that ?at the moment of realising our own independence, we are dependent upon another to recognise it?. Successful development, therefore, is based on maintaining ?constant tension? between the need for self-assertion and the need for recognition from another. The breakdown of mutual recognition and attunement arises when the infant: (i) feels that his actions are ineffective in influencing his mother?s actions, consequently resulting in feelings of powerlessness, or (ii) in his incessant attacks of fury, subjugates the mother, such that her existence is annihilated and is no longer able to grant him the recognition he seeks (Benjamin, 1980). The breakdown of tension between self-assertion and the need for recognition creates the ?negative cycle of recognition?, in which the search for recognition becomes a power struggle and the basis for domination (Benjamin, 1990, p.28). Benjamin (1990) conceptualises gender identity development as the struggle for recognition. The girl who attempts to seek identification with the father (who is experienced as an exciting and autonomous individual in the outside world) is rejected on the basis that he denies recognition of himself in the girl, and by so doing, preserves his masculine identity. Relegated to the realm of the mother, the girl learns to suppress her desire for autonomy and learns to fulfil it vicariously by pleasing another by being the ?object of desire?. In contrast, the boy?s ?attempt to seek identificatory recognition? from the father is reciprocated on the basis of his recognition of himself in the boy child At the same time, the alienation of the mother represents a critical move towards separation- individuation. Becoming autonomous and masculine, therefore, necessitates both paternal recognition and severing the maternal bond of emotional connectedness (Benjamin, 1990). Identification with the father enables him to deny feelings of dependency, whilst conflict between his need for maternal connection versus separation (i.e. ?becoming independent without the experience of loss?) (p. 104) is resolved by the ?splitting? of contradictory strivings to each parent. As such, separation-individuation becomes organised around the structure of gender. The coexistence of recognition and self-assertion, characterising the intersubjective dynamic of infancy, is replaced by an ?objectifying attitude? in which the woman is consigned to the place of ?the other? (Benjamin, as cited in Hagemann-White & Micus, 1999). In terms of male identity development, the process of repudiating femininity creates difficulties for the psychological basis for masculinity, such that the experiences of 25 ?feminine? are interpreted as threats to the defence structure of the male identity (Goldner et al., 1990). For the male, differentiation of male and female aspects is therefore transformed into domination. Benjamin?s (as cited in Hagemann-White & Micus, 1999, p. 5) concept of ?rational violence? rests upon this notion of dominance and is defined as ?seeking recognition without giving recognition?, the need for connectedness, but the denial of all feelings associated with it. Dominance and violence is therefore conceptualised as an attempt to reassert gender difference and male power when confronted with the terror of being too alike his female counterpart in his feelings of fear and his need for protection and dependency (Goldner et al., 1990). As an explanation for male violence, feminist psychoanalytic analyses imply that the construction of gender difference, more than psychological process or a social role, is ?a universal principle of cultural life that manifests itself in the individual psyche, the metaphysical framework, and the ideologies of a society? (Young, as cited in Goldner et al., 1990, p. 580). With emphasis on emotional aspects of experience, this perspective is distinguished from purely structural and socio-cultural theories. The latter, with their macro-societal in focus, have been criticised as being void of actor identity, context and ?moment-by-moment? sequences of events and therefore lack applicability and meaning to actual episodes of violence. Classical psychoanalytic conceptualisations, on the other hand, are less cognisant of oppressive social structures that exert influence on individual behaviour. A critical study of men?s violence that takes cognisance of both subjective experience and social reality, therefore may require shifting between various perspectives (Scheff & Retzinger, 1991). 2.10 A MULTIPLICITY APPROACH Goldner and her colleagues (Goldner, 1999; Goldner et al., 1990) have advocated that in order to gain an in-depth understanding of male violence towards intimates, it becomes necessary to embrace a ?multiplicity approach?. In a ?both/and? stance, multiple perspectives are able ?to coexist and to retain their unique characteristics? (Goldner, 1999, p.329). Attention is thereby drawn to how each of these different positions, rather than allowed to elevate one above the other, are permitted to enrich each other in explaining a specific instance of violence (Goldner, 1999; O?Neil, as cited in James, Sneddon & Brown, 2002). This ?both-and? position accommodates for the interplay of seemingly contradictory discourses and encourages, to quote Goldner (1999) ?a language that?speak[s] to the psychological aspects of moral conflicts and the moral aspects of psychological conflicts? (p. 328). Psychoanalytic theories which place emphasis on intrapsychic determinants of behaviour, therefore can be placed alongside socio-political models to provide a more 26 detailed understanding of male-on-female violence. Such a perspective embraces the intra- individual and socio-political viewpoints in understanding the male violent experience. 2.11 CONCLUSION A review of various explanatory frameworks underpinning the study of male violence offers critical inquiry into contributing factors to the phenomenon on dimensions that reflect the biological, psychobiological, psychological, sociological, socio-political or feminist perspectives. By locating the present research within a gendered framework of masculinities and feminist psychoanalysis, an attempt is made to bridge the gap between individualist and socially-derived orientations. A ?multiplicity? approach is adopted, one which permits various levels of explanation to exist without dissolving the other (Goldner, 1999, p. 329). Combining discourses in this way allows for a more nuanced analysis of male violence. The following section provides an overview of research conducted and framed within dominant models of understanding, offering a critique of the various explanatory frameworks and ideologies to explain male violence. Acknowledging the limitations of quantitative approaches which have sought to measure male violence via predisposing factors, etiological variables and personality characteristics, the review directs focus towards research that explores men?s violent experiences reflecting both discursive and phenomenological aspects of experience. Part III: Overview of Male Violence Research 2.12 INTRODUCTION Since the early 1970s, substantial interest in the area of violence against women gave rise to research on predominantly feminist interventions (Anderson & Umberson, 2001; Saunders & Hamill, 2003) targeted mainly at female victims of battering. Given that minimal means and services existed prior to that time which catered to the safety and legal needs of women, programming efforts were invested in women victims rather than male perpetrators. The feminist perspective, as the dominant theoretical approach at the time, informed practical interventions in areas of advocacy, clinical work and research (Gelles, 1999). Internationally, research studies focusing on the male batterer emerged during the 1980s along with the proliferation of domestic violence treatment programmes (Gondolf, 1997). Quantitative studies, subsequently, focused primarily on profiling abusive men according to personality, behavioural and attitudinal variables (Reitz, 1999). Highlighting the shortcomings of this approach in understanding the personal meanings men ascribe to their violent behaviour, qualitative studies emerged to consider the discursive and subjective elements of men?s violent experiences. In reviewing these studies, their dominant theoretical frameworks are highlighted. 27 2.13 REVIEW OF QUANTITATIVE RESEARCH ON MALE VIOLENCE With its socio-political emphasis, feminist research has explored the link between patriarchal ideology and women abuse (Bograd, 1988a; 1991; Coleman & Straus, 1986; Dobash & Dobash, 1979; Pagelow, as cited in Carden, 1994; Smith, 1990; Walker, 1979; Yllo, 1984; Yllo & Straus, as cited in Smith, 1990). Smith (1990), for instance, found that husbands, who espoused patriarchal attitudes and beliefs in a domestic context were more likely to have assaulted their wives at some point in the relationship. The more prevalent these norms were, according to Yllo and Straus (as cited in Smith, 1990), the greater the incidence of wife assault. Coleman and Straus (1986) similarly concluded that conflict between husband and wife regarding the legitimacy of patriarchal norms led to husband-on-wife violence. In a meta-analytic review, Sugarman and Frankel (1996), however, found partial support for the patriarchal ideology and domestically-violent men when assessed on the individual level. While men?s attitude toward domestic violence use predicted assault, many inconsistencies were found between men on measures of ?traditional? gender attitudes and gender schemas. Within this framework, others have alluded to cultural notions of masculinity, specifically men?s powerlessness and the need to control as factors related to intimate violence (Jakupcak, Tull & Roemer, 2005; Moore & Stuart, 2005; Petrik et al., 1994). Placing emphasis on cultural influences and gender socialisation of masculinity, Jakupcak et al. (2005), demonstrated that men?s fear of emotions predicted the outward expression of anger, hostility and aggression. Integrating feminist and resource theories, other studies have stressed associations between the lack of resources, use of violence and socio-cultural patriarchal norms. Having fewer resources relative to their partners, men were more likely to resort to violence as a compensatory response to a lack of control, powerlessness and perceptions of failure in fulfilling the family provider (Anderson as cited in Melzer, 2002; Yllo, 2005). Melzer (2002) argued that men in occupations that sanction the use of violence (e.g. police work) are more likely to exercise violence towards their partners. This suggests that the masculine identity is founded upon contradictory aspects of aggression ? prosocial behaviour affording protection to the vulnerable (i.e. women/children), on the one hand, and unsanctioned antisocial behaviour (Howard and Hollander, as cited in Melzer, 2002). The link between male gender role stress (the experience of distress in situations that threaten masculine identity) (Copenhaver, Lash & Eisler, 2000), and intimate partner violence was evident among substance-abusing men. Alcohol was used to deal with male insecurities regarding male role expectations, which ultimately increased the propensity for abusive behaviours. Whilst providing insight into the association between institutionalised patriarchal systems, culturally defined male gender roles and IPV, these studies, cohering around themes of male 28 insecurity, powerlessness and the need to control, tend to support a reductionist view that men are ?oversocialised? creatures who simply abuse their power in defence of patriarchy (Dutton, 1988; Goldner, 1999; Hastings & Hamberger, 1988). Reducing intricate dynamics to ?singular constructs? of power and control, therefore, downplays the complexity of emotions in abusive relationships (Goldner, 1999; Goldner et al., 1990). Rosenbaum and Leisring (2003) contend that ?power and control is more than a strategy, it is a human need? which warrants exploration of background factors, and the context of childhood trauma that engenders a state of ?powerlessness and dyscontrol? (p. 8). Psychological research has emphasised that these feelings, arising from childhood abuse or trauma, are the bases of various psychiatric conditions, such as depression, anxiety disorders and post-traumatic stress disorder. Extensive research has established the association between experiencing and/or witnessing abuse in childhood and abusive behaviour in adulthood (Carmen, Reiker & Mills, as cited in Dutton, 1995; Dutton, 1995; Dutton & Golant, 1995; O?Hearn & Margolin, 2000; Rosenbaum & Leisring, 2003; Rosenbaum & O?Leary, 1981). According to Dutton (1995), these studies acknowledge, apart from the learning of violent behaviour via direct modelling, that dysfunctional defence mechanisms of hypervigilance, splitting, projection and denial develop in response to victimisation during childhood. A traumatic response to abuse, as Dutton and Golant (1995) suggest, involves the suppression of rage due to fear of severe punishment. The failure of ?fight/flight? responses in reducing ?aversive arousal? results in a dissociation of self from the experience of violation (i.e. all emotions are shut off in face of the perpetrator). In this state of learned helplessness, feelings of shame are induced and experienced in relation to his own impotence. Physical abuse experiences coupled with shaming, provide the foundation for the development of the abusive personality. From an attachment theory perspective, Dutton et al. (1994) found that anxious attachment (chronic anxiety related to fear of rejection and abandonment) was related to various abusive personality attributes, namely anger, jealousy, borderline personality organisation and trauma symptoms. Accordingly, chronically frustrated attachment needs create an anger response to fear of separation (?intimacy anger?) and difficulties with emotional regulation, which are both risk factors for increased abusive behaviour. Associations between the perception of danger in intimacy and violent behaviour in intimate relationships has been demonstrated among male batterers. The need to control in such relationships was seen to produce responses of extreme anxiety and anger (Dutton and Strachan, 1987). While more psychodynamic-orientated perspectives have emphasised defence mechanisms in response to a childhood trauma and disrupted attachments, research framed from a social 29 learning perspective has emphasised violence as a learned response to direct or indirect abuse (Hertzberger, as cited in Putallaz, Costanzo, Grimes & Sherman, 1998; Kalmuss, 1984; MacEwen & Barling, 1988; Rosenbaum and O?Leary, 1981;). Rosenbaum and O?Leary (1981) found that abusive husbands were more likely to have experienced abuse as a child and more likely to have witnessed abuse between parents. Investigating the intergenerational transmission of violence, Kalmuss (1984), concluded that being hit by a parent and witnessing violence between parents are both significantly associated with marital aggression in adulthood, the latter is a stronger predictor than the former. Also implicit in social learning theory is the role of cognitions (i.e. attitudes, beliefs and expectations) in driving abusive behaviour, (Carden, 1994). O?Hearn & Margolin?s (2000) found a strong correlation between abuse in the family of origin and actual physical and emotional aggression toward their intimate partners among men who condoned marital violence. Holtworth-Munroe and Hutchinson (1993) suggested that maritally violent men, compared to nonviolent men, displayed greater tendencies to attribute negative intent, selfish motivation and to blame their wives. Apart from social learning, attachment and attributional theories, prolific research has centred on batterer psychological profiles (Carden, 1994). Utilising a variety of assessment instruments, extensive research has aimed at identifying commonalities among men based on specific personality variables, which has highlighted elevated levels of psychopathology (e.g. Beasley & Stoltenberg, 1992; Bersani, Chen, Pendleton & Denton, 1992; Flournoy & Wilson, 1991; Dutton & Starzomski, 1993). According to Hastings and Hamberger (1988), batterers manifest significantly higher levels of dysphoria, anxiety, depression and somatic complaints compared to nonbatterers, and were typically characterised as moody, sullen, hypersensitive and overreactive to rejection. Elevated borderline symptomatology and tendencies toward negativism and passive-aggression were also evident. Despite consensus on elevated psychopathology, subsequent studies have failed to replicate these findings (Beasley & Stoltenberg, 1992; Dutton & Starzomski, 1993; Flournoy & Wilson, 1991; Gondolf, 1988; Gondolf, 1996). More recent research, moreover, has contested whether borderline personality traits and post-traumatic stress disorder are typical of the abusive personality (Dutton, 2003; Gondolf, 1999; Gondolf, 2001). Holtworth-Munroe and Stuart (1994) proposed a typology which categorises batterers into family only, dysphoric/borderline, and generally violent/antisocial subtypes. Subsequent research has yielded mixed support for these findings (e.g. Delsol, Margolin & John, 2003; Waltz, Babcock, Jacobson & Gottman, 2000). Other researchers have attempted to delineate batterer subtypes based on observational, physiological and self-report measures 30 (Gottman, Jacobson, Rushe, Shortt, Babcock, Taillade & Waltz, 1995; Jacobson & Gottman, 1998). Type I (termed ?cobras?), exhibiting more antisocial personality traits, evidence an external appearance of arousal despite deceleration of heart rate during the violent encounter, compared to Type 2 (termed ?pitbulls?) who are internally aroused and have an accelerated heart rate during the act of aggression. Compared to Type 2 batterers, Type 1 abusers displayed more severe violence toward their partners, were more emotionally abusive and more likely to report antisocial personality traits. Subsequent studies, however, have failed to replicate these findings (e.g. Babcock, Green, Webb & Graham, 2004; Meehan, Holtworth-Munroe & Herron, as cited in Babcock et al., 2004). The failure of replication in these studies challenges the notion of a ?unitary batterer profile? (Hamberger & Hastings, as cited in Gondolf, 1988, p. 188). Research on abuser personality characteristics has typically adopted a psychiatric conceptualisation, linking aggression to various forms of personality disorder classifications (Dutton, 1988). Evaluation of court- referred abusers in treatment are based largely on clinical assessment instruments (such as the MCMI, MMPI, etc.), yielding elevated scores on various psychopathology scales, which are not necessarily generalisable to a community sample of batterers (e.g. Flournoy & Wilson, 1991; Hamberger, Lohr, Bonge & Tolin, 1996; Hastings & Hamberger, 1988). These clinical observations, moreover, serve to promulgate the stereotypical views of batterers as either ?sadistic psychopaths? or ?victims of abuse or neglect? (Gondolf, 1988). Attempts to understand male violence have resulted in a proliferation of research framed from various theoretical frameworks, each competing for a dominant voice. In maintaining its objective stance, quantitative studies tend to impose their own systems of meaning on the ?subjects? of study. The extensive employment of surveys, such as the Conflict Tactics Scale, and clinical assessment instruments, such as the MCMI and MMPI afford little opportunity for the expression of unique experiences (Goodrum et al., 2001). In the neglect of ?meanings, contexts or consequences of individual acts? (Yllo, as cited in Goodrum et al., 2001, p. 222), the subjective voice, which is central to understanding, is silenced. Addressing these criticisms, qualitative research has turned the lens towards the subjective angle to consider men?s meaning systems in relation to violence. 2.14 REVIEW OF QUALITATIVE RESEARCH ON MALE VIOLENCE Although studies in the qualitative paradigm have also revolved around dominant theories of violence, emphasis is directed to understanding the abusive behaviour from the batterer?s perspective. In so doing, insight is gained into the ?subjective experience of violence, the complexities of violent relationships and the multifaceted realities of the informants? (Esikovits & Peled, as cited in Rosen, Stith, Few, Daly & Tritt, 2005, p. 323). These can be 31 categorised into those seeking to understand the performative aspects of men?s talk about their violence, what Gadd (2000) terms the ?discourses of violence? approach, and those that investigate the interiority of men?s violent experiences (?phenomenological? or ?psychoanalytic-interpretive? approaches) (Hollway & Jefferson, 2000), each reflecting divergent research paradigms. Whilst both attend to subjective meaning, the former focuses on how men?s talk functions to neutralise their acts and how these are linked to ideologies on a broader societal context (Hearn, 1998). The latter is premised on ascribing subjective meanings to events in an attempt to understand and construct participants? ?life experiences? (Denzin, as cited in Eisikovits & Buchbinder, 1997). 2.14.1 ?Discourses of violence? Studies framed from the ?discourses of violence? approach theorises men?s violence largely as an instrumental strategy of control premised on the notion that the individual is ?inscribed by self-interested discourses of violence and masculinity? (Gadd, 2000, p. 430). Hearn (1998) proposes that men?s talk about violence (regardless of the audience) is embedded with neutralisation strategies that deny, excuse, justify or rationalise acts of violence. He interviewed men who had been arrested, men in programmes for men, men on probation, men from prisons, men in contact with welfare agencies, as well as men with no agency affiliations. Interlinking two elements of the ?text of violence?: men?s description of violence and men?s accounting for violence, Hearn (1998, p. 84) argued for the intimate connection between these seemingly distinct analytical concepts. Men?s descriptions of violence reflected ways of accounting for their violence. For instance, the men described their violence in various forms, ranging from violent thoughts, verbal, emotional and psychological violence, anger as a euphemistic reference to physical violence, threats to harm, violence to objects and pets and direct physical exertion (tugging hair, backhanding, kicking, punching, using weapons, stabbing and committing murder). Embedded in these descriptions were implicit and explicit strategies for accounting for violence. By creating a rationale, the account was given some kind of credibility. Hearn (1998) outlines one or more elements that constitute an account: In giving an account of his violence the man may (a) recognise and name his violence; (b) refer to his intention to do harm; (c) refer to his production of harm; (d) accept blame and/or responsibility for his violence; (e) explain or attempt to explain those elements (p. 105). In this respect, some of the men in Hearn?s (1998) study derived meaning for their violence by narrating the ?double self?. The act of ?distancing? was a strategy employed by the men who generally presented two selves: the violent self who committed the violent act (generally located in the past) and the nonviolent talking self being interviewed. The separation of the 32 selves, according to Hearn (1998), is a crucial element of explaining away, justifying and rationalising the violence (e.g. ?It was out of character?). Men?s descriptions were therefore laden with attempts to distance, justify, rationalise or explain away their violence. Repudiations of violence, for instance, reflected a discourse of denial or detachment (e.g. ?I?m not the violent type?) (p. 111). Quasi-repudiations reflected strategies of minimisation (e.g. ?I don?t see slapping as really violent?) (p. 114). The men also made use of excuses (acceptance of blame but not responsibility) and justifications (acceptance of responsibility but not blame). According to Hearn (1998), these constitute the moral discourses that function to assign blame or responsibility to himself, his partner, his past etc. (e.g. ?I seem to explode? (p. 123)? or ?she provoked something you know? (p. 132).2 Locating these strategies within the broader societal context, Hearn (1998) demonstrated how discourses of justification, for instance, are typically premised on men?s right to use violence in order to correct women?s behaviour in the context of the intimate partner relationship. This reflects men?s social positioning within the heteropatriarchal context, one in which men are esteemed as the ?head of the home?, authoritarian and disciplinarian and women as caregiver and nurturer. Men?s accounts, therefore, reflect how men ?say and show? their societal power (Hearn, 1998). Other studies have also attempted to demonstrate how accounts of violence reflect societal and cultural ideologies, problematising the link between men?s attempts to control women through the use of violence and codes of masculinity, which are entrenched in various societies (e.g. Anderson and Umberson, 2001; Boonzaier & de la Rey, 2004; Buchbinder & Eisikovits, 2004; Cavanagh et al., 2001; Eisikovits, Goldblatt & Winstok, 1999; Fuller, 2001; Hyden, 1994; Ptacek, 1988; Stamp & Sabourin, 1995). In a Peruvian study, Fuller (2001), found that men legitimated their violence on grounds of restoring order ?disrupted by female insubordination? (p. 27), evidenced in the woman?s failure to attend to marital and household duties or her complaints against her partner for failing to execute his family responsibilities. Male aggression was premised on the difficulty in attaining a balance between two contradictory male roles: the representation of a marital bond founded upon mutual reciprocity (i.e. women provide mothering and household duties and men support their family materially) and a culture of masculinity (which places premium on authority and male friendship bonding). Wood (2004, p. 571) similarly found ?dueling narratives of manhood? among incarcerated men ? ?men as protectors of women? and ?men as entitled to control women?. Although the cultural code of chivalry is expressed abstractly (i.e. ?It?s 2 Concepts of ?distancing?, ?excuses? and ?justifications? are dealt with at depth in ?Chapter 4: Results? in relation to the present study. 33 wrong to hurt a woman?), on a concrete level, participants subscribed to patriarchal views and maintained feelings of entitlement in relation to hurting their intimate partners. Anderson and Umberson (2001) examined the construction of gender based on men?s accounts of domestic violence. Men attempted to absolve themselves of blame by trivialising the violent incident of constructing their violence as ?rational response to extreme provocation? (p. 362). Participants gendered their constructions of violence in diverse and contradictory ways, in certain instances adorning the masculine persona of strength, power and rationality (whilst emphasizing their partners? irrationality and vulnerability), in others, construing themselves as vulnerable and powerless in the face of the criminal justice system. According to Butler (as cited in Anderson & Umberson, 2001, p. 359), such shifting representations are a reflection of the ?relational construction of gender and the instability of masculine subjectivity?. Perceptions of victimhood have been revealed by men in other studies, notably feeling fear, humiliation, being alienated and misunderstood in relation to helping systems (e.g. police intervention) for domestic violence (Buchbinder & Eisikovits, 2004; Shoham, as cited in Buchbinder & Eisikovits, 2004). Men have attributed violence as a recourse to feelings of disempowerment and lack of control, emanating from social forces and within the context of their intimate relationship. In Stamp and Sabourin?s study (1995), for instance, men attributed their violence to their wives? behaviour or personality, their jealous reactions or perceptions of their partners being abusive or controlling. Consistent with previous studies, account types reflected excuses, justifications, minimisation and denial, which provided the means by which the men exercised control over the meaning and representation of their acts of violence (Eisikovits & Winstok, 2002; Holtzworth- Munroe, 1992; Holtworth-Munroe & Hutchinson, 1993; Jukes, 1999). Underlying the motive for excuses and justifications is the attempt to negotiate a non-deviant identity. As Jukes (1999) notes, linked to the denial of responsibility or the minimisation of wrongness is the aim of ?avoid[ing] the guilt and anxiety associated with integrating his abusiveness into his self-image? (p. 49). In Buchbinder and Eisikovits? (2004) study, this was revealed in a self-transforming process whereby men redefined their social identity of ?criminal? (one enforced upon as a result of police intervention for their abuse) by narrating an identity of a law-abiding citizen, or a ?human?in crisis?, thus in need of ?treatment? rather than ?arrest? (p. 455). The performative aspects of talk, their context and implications were also highlighted in Hyden and McCarthy?s (1994) study. Disclaiming the violent act was represented on two levels: the silent narrator and that of disowning responsibility. To quote Hyden and McCarthy (1994) through disclaiming, ?one simply appears as the 34 victim or witness of happenings whose origins and explanations lie entirely outside of one?s own sphere of influence ? outside of one?s self? (p. 556). Recognising the social condemnation of woman abuse, the disclaimer becomes a means of ?smoothing out? a dented image as reflected in one participant?s words: ?I?m afraid I can?t remember anything about the actual event?I must have got a black-out, because I don?t remember anything until I found my wife bleeding on the floor?. Perpetrators of violence also employed minimising descriptions and language reflecting the mutuality of participation. In the former instance, the violent act is somewhat diminished and neutralised (e.g. ?I pushed her around a bit?). In the latter, men?s descriptions implied reciprocity of violence between perpetrator and victim (e.g. use of the term ?fight? as opposed to ?assault? dissolves the relations of dominance and subordination). The ?mutuality of participation? (p. 552) is further emphasised in other studies examining the accounts of violence by cohabiting partners (Boonzaier & de la Rey, 2004; Cavanagh et al., 2001; Eisikovits et al., 1999). Eisikovits et al. (1999) found that some couples negotiated a joint narrative, one containing ?a script of interpersonally agreed upon ways of living with violence? (p. 609). Particular social and cultural norms were woven into accounts that provided legitimacy and functionality for violence. In these respects, violence was considered acceptable and expected as the key to violence is diminished and shifted towards the normative context (e.g. ?There are no families without some kind of violence?) (p. 614). Couples? narrative constructions of violence as ?loss of control? (i.e. having a short-temper, stressful life or having ?nerves?), moreover, functioned to preserve the intimate union. Whilst violence is acknowledged, it was construed as an event beyond the man?s control. By ascribing it to external sources, the man is spared the responsibility for his violent acts and the woman acquires the illusion of control in a situation of powerlessness. As Eisikovits et al. (1999, p. 610) notes, ?violence is thus transformed into a feature of problematic families in general, rather than of deviant behaviour in particular?. Cavanagh et al. (2001) revealed how the social structural inequalities between men and women are mirrored on a micro-interactional level, namely in the interpersonal interactions in intimate relationships. Men?s use of denial, blame and minimisation, referred to as ?techniques of neutralisation? (Sykes & Martza, as cited in Cavanagh et al., 2001, p. 700), were seen as dynamic and functional, having a meaning and purpose. In the same way that violence is instrumental and purposeful, so too were men?s responses to violence. ?Not remembering? the violence, a form of denial, for instance, was interpreted as diminishing women?s experiences to that of ?unreality? and the centrality of violence to that of ?silence?. Subscribing to notions of a ?culture of violence? or ?cycle of violence?, men justified their actions on the basis of witnessing or experiencing violence in their communities. 35 In a South African study, Boonzaier and de la Rey (2004) found a similar negotiation of meaning between partners. Discourses of blame, minimisation and justification, though common in men?s accounts, were also supported by women. Men resisted labelling themselves as ?abusers?, at times, situating themselves as ?non-violent?. Women, similarly, refrained from assigning similar constructions to their partners, opting to represent the violence as something beyond the man?s control or as a response to provocation. Both partners drew upon hegemonic gendered discourses that prescribed ?masculinity as authority? and ?femininity as subordination? (p. 452), at the same time positioning themselves as weaving in and out of these discourses. Although subscribing to these hegemonic constructions, as in Anderson and Umberson?s (2001) study, the men also constructed themselves as emasculated in relation to their ?masculinised? partners who exhibited qualities of dominance and control in the relationship. Although concurring with men?s accounts of violence, women also challenged the traditional gender hierarchy by drawing upon discourses of empowerment and resisting passive femininity. What these shifting of positions in talking about violence reveal once again is the performativity of gendered positions. Anderson and Umberson (2001) note that through the physical act of violence, men temporarily restore the culturally inscribed gender binaries (?masculine aggression? and ?feminine weakness?) that are seen to be subverted by their assertive partners. Through the interpretive act, the assignment of meaning by talking about their violence, men were also able to exert control over their partners by holding them responsible for the violence (e.g. through use of justifications). To quote Anderson and Umberson (2001): ?by gendering violence, these batterers not only performed masculinity but reproduced gender as dominance? (p. 375). These studies framed from the ?discourses of violence? approach have offered insights into the ways in which men attribute meaning to their violence through the act of talk, and specifically, how these discourses are linked to broader gender and societal ideologies. Hyden and McCarthy (1994) emphasise that the context of talk, namely the interview, is itself a form of discourse that is shaped by questioning and responding. As a form of social activity, meaning is generated, negotiated and co-constructed between the interviewer and respondent (Riessman, 1993). As in other social contexts, the management of impressions, meanings and responses are pronounced. The strategies of accounting for violence has been termed ?remedial work? (Goffman, as cited in Cavanagh et al., 2001), which is intended to ?transform morally offensive behaviour into ?what can be seen as acceptable?? (p. 700). What is inferred is that the speaker engages in a management of meaning and responses of others in a purposeful and deliberate manner, denoting a conscious awareness (Cavanagh et al., 2001). 36 The ?discourses of violence? approach, however, ?presumes a male subject inscribed by self-interested discourses of violence and masculinity? (Gadd, 2000, p. 430). What is lacking in such a perspective, according to Gadd (2000), is the acknowledgement of ?psychic complexity? (p.430), one which attends to psychological and emotional processes of violent men, without denouncing the influence of socio-structural influences. Wood (2004), moreover, has argued that ?justifications? and ?excuses?, the predominant classificatory systems of meanings and motives of men, are merely labels that have been assigned based on ?outsider?s views?, namely those of researchers and clinicians. Advocating an actor-focused interpretation, she proposes that men?s accounts of intimate violence be angled from the ?insider?s view? ? the batterer?s view of himself and his actions from his subjective viewpoint. Echoing this view, Flick (1998) articulates that in foregrounding the performative aspects of talk, subjective meanings are less visible. 2.14.2 Exploring the interiority of violence The ?insider?s view? referred to in Wood?s (2004) study reflects an understanding of men?s violence from the perspective of men themselves. Taking this further, the ?insider?s view? will denote studies that give cognisance to the emotional or psychological aspects of men?s violence. Utilising phenomenology as a research paradigm, studies have attempted to explore men?s experiences of violence on the level of emotions (Denzin, 1984a; Denzin 1984b; Eisikovits & Buchbinder, 1997; Reitz, 1999; Thomas, 2003). Placing emphasis on the human experience and the first-person perspective, through a ?bracketing? process, these studies have sought to lay aside researchers? pre-existing assumptions about domestic violence whilst privileging batterers descriptions of their experiences (Reitz, 1999). Denzin (1984a) maintains that in order to understand violence, it is necessary to view ?emotionality and the self [as being] the core of violence? (p. 169). Based on the premise that emotionality is self-feelings directed toward self or others, he argues that it is only from the perspective of the ?self-reflective, feeling, violent individual? that violence can be understood (Athens, as cited in Denzin, 1984a, p.169). According to Denzin (1984b, p. 483-4), violence involves negative symbolic interaction between intimates and is a situated, interpersonal, emotional and cognitive activity. This entails an ?emotional enactment wherein one person?s emotional and cognitive definitions of an interpersonal situation are articulated and inflicted, symbolically (verbally) and physically, on another?. It is ?situated? in an interpersonal space which is seen as being ?out of control? (Denzin, 1984a, p.167). Using case-study snippets, he demonstrated that violence in intimate relationships is organised around emotions of anger, loss, rage, hostility, fright and fear, as the individual makes efforts to regain through violent action what has been lost or threatened ? whether it be self-esteem, personal safety (in response to bodily threat), self-control or control over another. 37 Thomas? (2003) study on men?s anger narratives revealed misconceptions about men?s anger. Different to what previous studies have claimed, Thomas (2003) found that men?s anger, rather than representing a reward of power over others, was experienced as an emotional state of discomfort. Emphasising its potentially destructive nature, anger had its basis in the physiological experience, conceptualised as a powerful force that takes control. Distinctions were made in this regard between ?wrong anger?, such as anger resulting in a loss of control, and ?legitimate? or ?justified? anger, defined as anger that is effective in stating its point and proportionate to the offence (p. 167). Also employing the phenomenological method, Reitz (1999) unveiled two levels of themes in the structure of men?s violent experiences. Contextual themes, the location of self in relation to other in relationships, revealed dualities of experience captured in being ?big or little?, ?good or bad? and ?winning and losing? (p. 151). Violence reflected the desire to assert the more favourable aspect of these dualities. Where men felt small, helpless and childlike in relation to their partners, the violent experience was a display of ?metaphorical bigness? (p. 153). However, the instability of experience was also acknowledged as the men moved in and out of dualities. Whilst feeling little and childlike in certain instances, in others the adult role was assumed during which the partner was perceived as ?bad? and needing punishment. Winning and losing metaphors revealed perceptions of the domestic bond as being conflictual and adversarial. Focal themes denoted the experience of violence that were represented on a continuum of being in control, being out of control, pressure and explosion. The change from pressure to explosion was associated with experiences on a bodily level (muscle tension in chest, arms or stomach). Where pressure and exploding occurred in a state of being ?in control?, there was an awareness of self and the element of choice, for instance, on how severe the form of violence would take. In a state of ?out of control?, however, men had a ?not-me? experience centred on lack of awareness and choice (i.e. as a stranger to the person they believed themselves to be, such as a ?crazed animal? or ?Mr Hyde?). Eisikovits and Buchbinder?s phenomenological study (1997) reflected men?s use of war metaphors to denote experiences of conflict and violence. Intimacy in this context was construed as threatening, and violence was depicted as a natural response to being attacked by their partners in their weak spots. In addition to the ?war? played out externally, the participants also made reference to an internal war fought on the level of the body, experienced as a loss of control. Anger was conceptualised as a liquid that fills up the body in response to arguments that invade the inner space as they are forced to be taken in. Metaphors of ?swallowing?, ?piling up? and eventual ?blowing up? conveyed the uncontrollability in relation to containing extreme emotions. The state of inner conflict, what Eisikovits and Buchbinder?s (1997) terms a ?civil war? (p. 491) between containing and 38 releasing anger, culminates in a splitting of self ? one who is violent and the other who tries to overpower or contain the violence. ?Being in control? and ?being out of control? as a prevalent theme in men?s violence studies were also identified in other studies, although not specifically grounded in a phenomenological orientation (e.g. Hyden, 1994; James, Sneddon & Brown, 2002). In a study by James et al (2002), some men perceived their violence as instrumental (i.e. employed to achieve an end), whilst others experienced it as a form of expressive release outside of their control. Violence was conceptualised as reflecting either tyrannical or exploder violence respectively, and similar to Reitz (1999), were assumed to lie on a continuum varying according to the extent of control and intentionality. Emotions of anxiety, anger, humiliation and shame were identified as precipitants in both ?tyrant? (instrumental) and ?exploder? (expressive) violence. Exploder-violence was more likely to occur in ?symmetrical? relationships (i.e. women would complement their partners in participation in escalating conflicts), while tyrant-type violence is reflected in dominant-submission relationships. Studies on escalation dynamics have also shed some light in uncovering the emotions underlying states of loss of control. Escalation has generally been defined as ?an emotionally charged process involving intense emotions such as anxiety, anger, helplessness, humiliation, shame, guilt, jealousy, hostility, low self-esteem, and a sense of loss? (Winstok et al., 2002). Based on transcribed psychotherapy interviews, Lewis traced various emotional sequences and found that instances of angry escalation was evoked by preceding feelings of shame caused by actual or perceived rejection. Where shame is evoked but unacknowledged (i.e. denied), anger is aroused at the other who is perceived to devalue the self; and the other is therefore regarded as the source of hostility. The self therefore feels rageful ?at its inferior place ?in the eyes? of the other? (Lewis, as cited in Retzinger, 1991, p. 49). Escalation is therefore the likely outcome. Lansky (1987), using a detailed case study as a backdrop, demonstrated that violence is a defensive operation against overwhelming feelings of shame based on real or fantasised attacks by the partner. Situated within a ?collusive system? of character pathology of both partners to the relationship, one partner?s ?pathologic propensity? to inflict shame interacts with the other?s heightened proneness to shame, thus fuelling tension to a point of escalation. Based on shame and anger studies, Retzinger (1991) proposes a theory of escalation which centres on alienation, shame and conflict. The emotion of shame signals a disruption (i.e. threat or damage) to the affectionate bonding system. Such a threat or damage can take the form of unrequited love, loss of face, disapproval or devaluation. Rather than acknowledge the shame (and subsequently awareness of possible dissolution of the bond), the perpetrator feels alienated. Blame is thus 39 projected and the other is perceived as the source of attack. Resultant anger therefore serves as a protest against the threat and as a mechanism to safeguard the loss of face. Adopting what has been termed a psychoanalytic-interpretive framework (Hollway and Jefferson, 2000), Gadd (2000; 2003; 2004) has attended to the psychic dimensions of men?s violent experiences. Drawing upon Hollway and Jefferson?s (2000) notion of the ?defended subject?, Gadd (2003) suggests that men?s violence toward intimates may reflect ?contradictory experiences of masculinities?, particularly the ?tensions between inner mental life and social expectations? (p. 337). To paraphrase Kaufman (as cited in Gadd, 2003), pain and alienation may be unconscious motivations for violence, in a manner that undesirable emotions are repressed, but manifest as control over others. The notion of the ?defended subject? thus assumes that the individual is a ?fragmented subject, anxiously managing [emphasis added] (repressing, splitting and projecting) thoughts, feelings and memories that threaten the integrity of the self? (Gadd, 2000, p. 430). The notion of management, rather than a conscious and deliberate strategy of negotiating a non-deviant identity as presupposed in a ?discourses of violence? approach, is one who invests, at times unconsciously, in discourses due to some reward or satisfaction (Hollway & Jefferson, 2000). Framed in this manner, Gadd (2000) reasoned that one man?s inability to recall his actions during a violent episode reflected not a deliberate attempt at concealing his actions, but an unconscious and learned strategy of forgetting arising from unresolved interpersonal conflicts from the past. His partner who misconstrued his intention to assist rather than criticise her was reminiscent of childhood experiences of being misrepresented and misunderstood. Underlying his violence were feelings of being misrecognised and feelings of persecution which gravitated towards humiliation and rage. Gadd (2000) theorises that such feelings of disenfranchisement had influence on the formation of his adult masculinity, such that violence served as a temporary means to fend off persecutory feelings. Framed in this manner, masculinity from a psychoanalytic- interpretive stance appears more precarious, subtle and layered than masculinity in a sociological sense. To quote Gadd (2000), the ?very observable traits and behaviours that we so often take to be indicative of masculinity are, in fact, the observable manifestation of men?s sometimes (but not always) unconscious attempts to fend off psychic threats to their sense of vulnerability? (p. 445). 2.14.3 Reconciling divergent approaches: A multiplicity approach What these studies have highlighted is the complexity of the violent experience. What becomes apparent is the significance of understanding men?s violence beyond the level of discourse. As studies in the preceding section have illustrated, a ?discourses of violence? approach offers fruitful contribution to the study of men?s violence in theorising the link 40 between men?s social power, discourses of masculinity and men?s strategies for accounting and talking about violence (Gadd, 2000; Hearn, 1998). Studies geared toward a more insider perspective, however, have attempted to bring to the fore the core of the violent experience. Phenomenological studies, in particular, have consistently emphasised that the emotions of violence also operate on a physiological level (Eisikovits & Buchbinder, 1997; Reitz, 1999; Thomas, 2003). In understanding gendered experience on the level of the body, Gadd (2000), quoting Connell (1995, p. 51-53), points out that biographically, they are more resistant to deconstruction than a sociological analyses would allow: ?The sweat cannot be excluded?Bodily experience is often central to our memories of our own lives, and thus in our understanding of who and what we are?. Studies adopting a psychoanalytic-interpretive approach have made the psychic dimension of experience more visible (Gadd, 2000). However, whilst the phenomenological approach is premised on the assumption that individuals accurately ?tell it like it is? given in the context of an empathic and safe encounter with the interviewer (Terre Blanche & Kelly, 1999), psychoanalytic-based approaches attempt to access the latent structures of subjective meaning that are not accessible on the surface (Flick, 1998). Subjective meanings are thus derived from both surface structures (accessible to the participant) and deep structures that are ?hidden? (Flick, 1998). What is proposed is that by ?articulating the many emotional truths of men?s ?experience??the disparity between what violent men feel, say and do? is revealed (Gadd, 2000, p.431). The phenomenon of men?s violence could be more fully appreciated by accessing the insider?s seemingly contradictory experiences of being ?in control? and ?out of control?, as well as unveiling the emotional complexities of experience. Sole emphasis on the insider?s perspective of men?s violence, however, fails to challenge men?s moral accountability in terms of making apparent their ways of accounting for and talking about their violence. To re-introduce Hyden and McCarthy?s (1994) notion of the interview itself as a form of discourse, two levels of responsibility become apparent as they relate to the micro context of the interview and the macro context of society. On one level, the principle of ethics becomes operational, requiring the adoption of a non-judgemental listening stance and incorporation of all views, both the dominant and the marginal, the mainstream and the deviant. On another level, morality is also operative as it is reflects the ideal order of a given society. To paraphrase Hyden and McCarthy (1994), morality ?re-presents itself within the dominant discourses of a society as to how citizens should behave in line with certain religious, social and legal dictates? (p. 545). In the realm of violence research, the issue is both an ethical and moral one. To attend to the subjective dimension via the phenomenological orientation is to address the first level, whilst locating men?s subjective experience within broader social reality fulfils the second. This ?double agenda?, as Goldner (1999, p. 330) terms it, emphasises the simultaneous operation of two discourses ? a 41 psychological one which gives voice to the ?power of feelings? (Chodorow, 1999, p. 6) which, at times, reveals itself as disembodied and overwhelming, and a moral one, which stresses choice and personal accountability. In this light, men?s talk about their violence invites expression of meaning via a psychological dimension, and at the same time, holds him ?fully accountable for its destructiveness? (Goldner, 1999, p. 330). Based on this review, it is apparent that a more comprehensive understanding of men and their violence is needed. A thorough and detailed exploration of the ?in-the-moment? violent encounter would allow for a deeper understanding of the emotions, thoughts and behaviours that underlie the experience of violence. It would also serve to highlight, from the abuser?s perspective, the interactional dynamics between himself and his partner, which provide the ?fertile ground? upon which the violent encounter is enacted. As Denzin (1984b, p.486) notes, violence takes place within an ?interactional framework of superordinate and subordinate relationships?. Subsequently, research into domestic violence would need to acknowledge that the self in relation to his situation is defined, not in isolation, but in interaction with others and society (Goodrum et al., 2001). In order to fulfil these criteria, the proposed research will adopt Goldner?s (1999; Goldner et al., 1990) ?multiplicity? approach in its attempt to explore the ?inside? experiences of male perpetrators of intimate partner violence, at the same time illustrating their links to the broader societal context. 2.15 CONCLUSION Various conceptualisations of intimate partner violence have yielded a proliferation of research reflecting somewhat divergent theoretical understandings. This section has attempted to provide an overview of this research framed within these perspectives. The study of men?s violence toward intimate partners, traditionally dominated by quantitative research paradigm, have relied on observation and measurement of attitudes, behaviour, personality characteristics, conflict-interaction styles and learning histories (Reitz, 1999). Given that data are aggregated and reduced to numerical scores, studies in this paradigm have been criticised for their tendency to decontextualise specific instances. In addition, they fail to yield the richness and depth, the context and language and the personal meaning reflected in subjective accounts. The present research attempted to address these limitations by adopting a qualitative research paradigm. Within this framework, studies categorised under the ?discourses of violence? approach (emphasising the ideological functions of men?s talk about violence) were distinguished from those exploring the interiority of violence and illuminating the emotional reality of men?s violent experiences. Goldner (1999) notes, rather than conceive one perspective as nullifying or excluding an alternative, it is possible to adopt theoretical positions that are seemingly contradictory. 42 Simply put, a psychological framework does not excuse violent behaviour. On the other hand, adopting a feminist perspective does not exclude the psychological roots of behaviour. The stance of multiplicity, to quote Goldner (1999) embraces the political metaphor of the ?cultural mosaic?: The notion of multiplicity?connotes a process that not only allows but actually invites multiple perspectives to coexist and to retain their unique characteristics. The habit of mind creates a context in which each perspective acts as a check on the other. Rather than elevate one discourse above another, this posture is analogous to the physics metaphor of wave and particle, where first one, and then another way of seeing dominates. (p. 329) Upholding this view of multiplicity, the present research is framed from a masculinities and feminist psychoanalytic framework, thereby providing a crucial linkage between subjective and social realities. As this section has revealed, in order to access the insider perspective of violence, it is necessary to adopt a research paradigm which allows for an exploration of the phenomenon as defined and understood by participants themselves. The following chapter will provide a rationale for locating the present research within a qualitative paradigm. In so doing, an attempt is made to connect this to various decision making processes on the level of methodology and data analysis. 43 CHAPTER 3: RESEARCH DESIGN AND METHODOLOGY 3.1 INTRODUCTION The aim of the present research was to explore men?s subjective experiences of their violence toward their intimate partners. Different from a quantitative approach which seeks to isolate and control known variables, a qualitative framework adopted by the present research provided access to the men?s social realities in terms of understanding both subjective meaning and performative aspects of language (Terre Blanche, Kelly & Durrheim, 2006). An understanding of their social world, by necessity, involves insight into the social structure, namely patterns of their social roles and relationships. However, as articulated by (Seale et al, 2004, p. 3), it is experience, with its foundations in feelings and emotions that is situated at ?the deepest level of social reality?. It is this, in the ?final analysis [that connects] us to other living creatures?. Building on from the theoretical foundations discussed in the previous chapter, this chapter attempts to locate the present research within an epistemological framework. In so doing, methodological considerations are outlined as they pertain to the data collection method, the instruments used as well as the form of data analysis chosen. Importantly, ethical considerations are addressed as they pertain to various aspects of the research process. 3.2 THE DISCURSIVE VERSUS THE PHENOMENOLOGICAL Embedded within a social constructionist paradigm, discursive approaches maintain that language, rather than being neutral or transparent, assists in constructing reality (Terre Blanche & Durrheim, 1999). Reflected in the work of Wetherell and Edley (1999), discursive psychology seeks to investigate how subjectivity is constituted in practices of discourse. As Wetherell and Edley (1999, p. 337) note, ?what it means to be a person, the formulation of an internal life, an identity and a way of being in the world develop as external public dialogue moves inside to form the ?voices of the mind??. Discursive practices are embedded in all social practices, taking the form of shared means of ?self- accounting, vocabularies of motive, culturally recognisable emotional performances and available stories for making sense? (Wetherell & Edley, 1999, p. 338). Focus on the study of men?s violence in this respect coheres around the relationship between men?s violence and social positioning of power, for instance, men adopting patriarchal ideologies that foster dominance over women (Gadd, 2003). Carrigan et al. (as cited in Gadd, 2003, p. 334) notes, however, that such a relationship needs to be theorised in such a manner that does not reduce masculinit(ies) to ?more or less unrelieved villainy? or does not designate ?all men as agents of the patriarchy in more or less the same degree?. Such a schematic view moreover fails to capture the complexities of emotional experience and 44 appears inadequate in ?produc[ing] an authentic inner world? (Jefferson, as cited in Gadd, 2003, p. 337). According to Connell (as cited in Gadd, 2004, p. 337), a psychodiscursive approach does not fully capture a ?multidimensional understanding of gender?. In order to reveal the layers of subjective meaning, a phenomenological approach privileges the subjective perspective, according this with greater significance than objective reality (Rossman & Rallis, 2003). Emphasis is placed on the importance of appreciating the individual?s subjective experiences or the ?first-person? accounts of human experience in order to truly understand behaviour (Thorne, 1996). This is premised on various ontological principles: (i) people are viewed as agents who shape and create their everyday world (Rossman & Rallis, 2003), (ii) people are regarded as the source of their thoughts, feelings and experiences and are considered real (Terre Blanche & Durrheim, 1999) and, (iii) the social world is understood from the world of the first-person, or subjective experience (Schwandt, 1994). The interpretive paradigm embraces the characteristic features of phenomenology and hermeneutics. Hermeneutics, which is the work of interpretation, is ?the methodological meaning of phenomenological description? (Heidegger, as cited in Denzin, 1984a, p. 8). Through a detailed examination of a text, all that is known about a phenomenon is ?revealed? and ?engulfed? through interpretation (Denzin, 1984a, p.8; Neuman, 1994). Such understanding takes place at ?the everyday and bracketed levels of meaning? (Denzin, 1984a, p. 8). In other words, inquiry commences with the ordinary, everyday human understanding of others (Rose, Beeby & Parker, 1995) but also involves bracketing ? the suspension of one?s own pre-existing frame of reference, thereby allowing the data to ?talk for itself? (Kelly, 1999; Terre Blanche & Kelly, 1999). Adopting such a position implies a willingness to be reflective, self-critical, thorough, while at the same time assuming that that which is spoken by the individual is ?sensible? and ?normal?, not ?insane? or ?deviant? (Chatterdon, as cited in Bograd, 1988b). The interviewing context, therefore, becomes a facilitating environment of openness and trust within which the individual is able to express him or herself in an authentic manner (Terre Blanche & Durrheim, 1999). The research attempted to reconcile the two seemingly divergent approaches by traversing the boundaries between phenomenology (premised on the interpretive approach) and discourse (premised on social constructionism). 3.3 ?INTERPRETIVE RESEARCH?: LINKING THE ?EXPERIENCE-NEAR? TO THE ?EXPERIENCE-DISTANT? Subsumed in the term ?romantic hermeneutics? (Kelly, 1999), the interpretive paradigm gives priority to the world of the first-person, or subjective experience (Schwandt, 1994). 45 Upholding the interpretive method of verstehen ? translated literally as ?to understand? (Schwandt, 1994), a position of empathic understanding is adopted. To understand men?s violence ?as it is lived in its context? (Kelly, 1999, p. 405) requires then the framing of experience using participants? own terms of reference. However, to bridge the gap between lived experience and social reality, the research employed Kelly?s (1999; 2006, p. 346) label ?interpretive research? to denote both interpretive and constructionist research. Whilst phenomenology as an interpretive method is located on the ?experience-near? (or contextually derived) side of this interpretive continuum, social constructionism is situated on the ?experience-distant? (or theoretically led) pillar (Kelly, 1999). While much emphasis is placed on exploring participants? subjective meaning, the proposed research attempts to connect the ?intricacies of the subjective experience of contexts? (Kelly, 1999 p. 405) to the broader social reality. Given the limitations of the phenomenological (or the ?experience-near?) orientation in terms of understanding patterns across time and situation, it becomes necessary to consider not only the context of experience, but also its location in relation to the overall structure of understanding. The research therefore places accent on the tension between description, the phenomenological ?insider accounts? and interpretation, the constructionist ?outsider perspectives? (Kelly, 2006). The ?cooperation? of the two orientations, according to Kelly (2006, p. 348), is the crux of qualitative practice. The ?hermeneutic circle?, a term christened by Gadamer (as cited in Stahl, n.d.) reflects this understanding of the circularity of the text; that is, in the interpretation of the text, the meanings of structures should be considered in relation to the meaning of the whole (Kelly, 1999), or what Denzin (1984a, p. 8) calls the ?interpreted totality?. While the focus of interpretation is on the uniqueness and distinctiveness of thoughts, feelings and experiences of men who are violent towards their intimate partners, their relation to broad principles reflecting ?a coherent clustering or ordering of themes of experience? (Kelly, 1999, p. 407) are also considered. This in turn would generate understanding not only on the individual level but also in terms of its intersections with broader societal/social issues. 3.4 METHODS APPROACH In upholding the epistemological stance (ways of gaining knowledge about the social world) of interpretivism, the research relied on in-depth interviews with the participants (Snape & Spencer, 2003). Generated data collection methods, such as in-depth interviews and group discussions, allow the participants a ?direct and explicit opportunity to convey their own meanings and interpretations through the explanations they provide? 46 in a spontaneous manner or when prompted by the researcher (Lewis, 2003, p. 57). According to Vangelisti, Crumley & Baker (as cited in Wood, 2004), interviews allow the story of relationships to emerge from which the listener gains access, not only to the teller?s experiences, but more critically, insight into his ?attitudes, feelings, thoughts, meanings and reasoning? (p. 560). To quote Hall (2007, p. 33), ?narratives help us regulate how we experience and express emotion?. Due to the sensitive nature of the research, aspects relating to confidentiality were better contained in a one-on-one setting (when compared to focus group discussions) (Lewis, 2003). Apart from the emphasis on subjective meaning and the exploration of a complex phenomenon such as violence, the in-depth interview, according to Mishler (as cited in Banister, Burman, Parker & Tindall, 1994) is also about empowering disadvantaged groups in a manner that validates their views and gives voice to their experiences. In the present research, in-depth interviews presented opportunities for clarification and elicited detailed understandings of participants? emotions, thoughts and motivations, in this manner, revealing their underlying realities (Terre Blanche & Kelly, 1999; Terre Blanche & Durrheim, 1999). Apart from ?bracketing? of biases and prejudices, ?intuiting? as a phenomenological principle was also applied during the interview. This mode of awareness involves ?sensing? and being attuned to slight nuances to what is being conveyed by the participant (Rose et al.,1995). 3.5 DESIGNING THE INTERVIEW SCHEDULE 3.5.1 The defining features of in-depth interviews The in-depth interview, although often described as a ?conversation with a purpose? (Webb & Webb, as cited in Legard, Keegan & Ward, 2003, p. 138) is nevertheless an exchange between the participant and researcher, each having different roles and objectives. The present research adopts the ?travel metaphor? perspective on in-depth interviewing. Aligned with the constructivist research model, knowledge rather than being given, is ?created and negotiated?. The researcher and participant embark on a journey to develop the meanings of stories, which are subsequently interpreted by the researcher. To quote Kvale (as cited in Legard et al., 2003, p. 139) the traveller (i.e. the researcher) ?asks questions that lead the subjects to tell their own stories of their lived world?. According to Legard et al (2003), the in-depth interview is defined by its key features, namely, structure combined with flexibility (set topics are covered but allow room for probes and exploration); interaction (free talk is encouraged with the interviewer picking up on points in the participants? responses); depth (achieved via probing or follow-up questions to explore, penetrate and seek explanation of the participant?s meaning), as well as its generative nature (new knowledge or thoughts on the topic arise, 47 either self-directed or guided by the researcher). To bring these features together requires face-to-face interaction between participant and researcher. The design of the interview schedule, therefore, needs to take cognisance of these defining aspects. (See Appendix F for interview guide). 3.5.2 Addressing ?structure? The interview schedule was designed as a non-scheduled standardised form (all questions are asked, but in a different manner or sequence). The semi-structured approach facilitated the exploration of set topics, at the same time, allowing room for unstructured questions. In this way, comparisons across individual accounts were possible, while affording participants the opportunity for personal and unique expression (Williamson, Karp & Dalphin, 1977). Classified as a ?topical interview? (Rubin & Rubin, as cited in Arthur & Nazroo, 2003, p. 110), the researcher assumes an active role in guiding the discussion around core issues, while being open to participants? framing and structuring of responses (Marshall & Rossman, 1995). Related to structure is the ordering of sub-topics within the interview guide (Arthur & Nazroo, 2003). Following an outline of the research objectives and issues of confidentiality, Section I aimed to elicit information regarding the participant?s personal context (e.g. age, education level, relationship status, and nature of relationship with spouse). These opening topics ?set the scene? for the participant, simultaneously, allowing the researcher to gain insight into his present circumstances (Legard et al., 2003). For the participant, these neutral topics set the tone for an unthreatening atmosphere. For the researcher, this allowed some anticipation of areas for probing at a later stage during the interview. Section II aimed to ensure a gradual, non-threatening transition from exploring participants? meanings (or definitions) of violence to more personal, contextual information pertaining to the violent encounter. Introducing conceptual questions in the initial stages of the interview allows participants? initial reflections on a particular term to be ?unclouded? by subsequent discussions (Arthur and Nazroo, 2003). Section III transitioned both participant and interviewer toward a much deeper and intense level of engagement. Turning attention to the most vivid experience of violence toward his partner, the participant is led to recall such an experience in rich detail, followed by his motivations for his behaviour. This follows the rationale that people find it easier to talk about an experience or behaviour prior to revealing its underlying feelings, motivations or attitudes (Arthur & Nazroo, 2003). Section IV elicits the reflective stance of the participant by exploring at depth his feelings and thoughts at the time of a particular incident of violence. Continuing in this mode, Section V allows him to look back on the incident in order to develop meaning from it. The ?winding down? phase of the interview was initiated when he was taken back to the present situation and asked to reflect on his relationship as it is in the 48 present, as well as his motivations for joining a men?s programme or workshop. Section VI thus attempted to draw the interview to a close by ending on a more positive note. As Arthur and Nazroo (2003) note, participants are able to move away from feelings of distress, anger and frustration possibly generated from the interview, towards more neutral emotions. Although this logical ordering of sections enabled a smooth transition from one sub-topic to the next, the nature of ?free talk? in the actual interviews accommodated for the flexibility of ?weaving in and out? of sections. 3.5.3 Achieving Depth Follow-up questions below each sub-topic acted as probes or prompts to encourage elaboration. These were, however, employed merely as a guide to amplify, explore or elicit further explanation or clarification (Legard et al., 2003). In practice, it was found that probing was better facilitated by attending to the participants? responses and using the content as a basis for further exploration. The elements of interaction and generative nature revealed themselves in the actual interview, rather than ?pre-designed? in the interview guide. 3.5.4 Encouraging ?free talk? and ?new thoughts? With a total of six sections, the interview guide focused on brevity and conciseness. In addition to achieving more in-depth data collection, shorter guides, according to Arthur and Nazroo (2003), facilitates better engagement with the participant rather than creating over-reliance on the guide for its explicit detail. In order to ensure ?active interviewing?, short phrases and simple statements were utilised in framing of questions in favour or wording entire questions. This enabled the researcher to be more attuned to the language used by the participants themselves (Arthur & Nazroo, 2003, p. 123; Legard et al, 2003). Notably ?participant? was used in place of ?you? (e.g. ?how participant describes current relationship with partner/spouse?) to facilitate spontaneity in framing questions in the interview as opposed to reading them directly from the interview guide. In short, considerations related to length and phrasing of questions were designed to elicit a more natural and engaging interaction. In addressing the features of in-depth interviews, the interview schedule was a guide for the discussion of several topics, albeit ?not [emphasis in original] as an exact prescription of coverage? (Arthur & Nazroo, 2003, p. 115). Different to a fixed structure of quantitative approaches (e.g. semi-structured questionnaire), its structure encourages rather than stifles the processes of reflection and facilitates the discovery of new ideas and themes. More significantly, as Arthur and Nazroo (2003) assert, the interview guide serves as a public document that makes the research objectives and processes transparent, particularly so as participants? actual transcripts remain hidden from view. 49 3.6 NEGOTIATING ACCESS Access to participants was negotiated at various levels, from establishing rapport with various co-ordinators of the men?s programmes to gaining and maintaining the trust of the men individually. For one organisation, the researcher?s relationship with the organisation had to be ?re-established? following the resignation of the co-ordinator of the men?s group. Not only was he the key ?gatekeeper? to gaining access, he also represented the ?glue? that had moulded the group towards a cohesive unit for several years. His sudden departure from the organisation, which came to the researcher?s knowledge much later, resulted in a shift towards finding new ways of accessing participants. New links had to be established with a newly appointed co-ordinator who was less well-acquainted with the men. The negotiation stage became a back-and-forth process as researcher was informed by the new co-ordinator about the difficulties of making contact with, what appeared to be, a disconnected group that had lost interest in attending weekly meetings. After several postponements, a day was set aside for the researcher to meet potentially interested participants. Prior to this meeting, however, the researcher was put in contact with another ?gatekeeper?, one man who was both a potential participant and a particularly influential member in the group. Together with the new co-ordinator, he had arranged for the meeting to take place. On the appointed day, the men trickled in slowly, a total of nine men, typically making appearances in small groups (apart from one man who showed up alone). This acquaintance with the men was in some respects a ?trust-building exercise?, one which allowed the participants to become familiar with the researcher and the research. This was an important ?pre- interview phase? during which the researcher felt was necessary to establish rapport, and was the basis upon which telephonic contact with individual participants was made for setting up appointments. ?Gaining access? once again featured as a ?theme? following this phase. Of the nine participants recruited for the research, one did not have a direct contact number through which a meeting time could be negotiated. Although he had initially requested that he be contacted via a fellow participant, this did not prove feasible. After several ?no-shows? for scheduled appointments, he was subsequently excluded from the research. For these participants, negotiating access seemed to be tied up with issues of establishing trust. The opportunity to establish such a phase, however, presented itself differently in recruiting men from a second organisation. This organisation represented men with vast demographic differences from the those of the first group (e.g. race, education, socioeconomic status, area of residence). Initially the two sites were selected as they ?represented? the diversity of men with a history of abusive behaviour, in this way, allowing for a reasonable amount of demographic variation (Dobbert, as cited in Marshall 50 & Rossman, 1995). Although generalisability of results is not a key criterion for soundness in qualitative studies, the aim was to access diverse participants by eliciting participation from two divergent sites, thereby allowing fulfilment of the criterion of qualitative ?diversity? (Ritchie, Lewis & Elam, 2003). Only one individual, however, willingly volunteered to tell his story. In this instance, all logistical arrangements prior to the interview was arranged telephonically. Building face-to-face rapport, briefing on the research expectations and providing informed consent were all discussed on the same day as the interview. (See Appendix B, C, D and E for letter of participation and documents related to informed consent). A rather different dynamic of gaining access and establishing trust was encountered with a third organisation. This men?s programme, rather than operating from a ?support group? focus, relied largely on recruiting men for particular workshops held. Rather than pre- existing as a cohesive unit, the majority of men had only been acquainted with other participants for the first time at the workshops. More significantly, these men had no prior contact with the co-coordinators/facilitators of the men?s programme prior to the workshop. It was at this juncture, the second day of a three-day workshop, that the researcher was invited to share with the group the aims of the research for the purpose of eliciting their participation. Volunteering to be interviewed in this context therefore was not merely about participation. For the participants it carried a significant weight of trust, not only in the researcher and the research process, but also in the workshop itself and in the workshop facilitators. By granting the researcher a space in the workshop, the facilitators had tacitly conveyed to the men that it was psychologically safe to participate in the research. In this respect, negotiating access was about gaining and maintaining the trust of individuals on various levels. The reluctance to participate was evident. Initially, one man openly volunteered to be interviewed. As the day progressed, and as trust developed collectively, other men volunteered their participation. As with the preceding organisations, trust had to be reaffirmed on an individual basis by detailing issues pertaining to ethics. In all, twelve men from the three organisations in and around Johannesburg were interviewed. The final sample subsequently comprised a group of men that were of similar socioeconomic status and did not reflect the diversity criterion as initially intended. The groups, however, held slightly different approaches to intervention with perpetrators of domestic violence. One organisation offered various workshops specifically for perpetrators of domestic violence and organised community visits to promote non- violence. Another group similarly offered workshops to men, although these dealt with a range of issues in addition to domestic violence, such as HIV/AIDS. For both groups 51 intervention was based on an educational approach and was targeted on a group level. Nevertheless, for both organisations, individual counselling was also available. The other group specifically focused on male abusers and focused intervention in the form of a support group whereby the men would meet to share their difficulties and find solutions. These apparent differences that emerged between the men?s groups were not completely ?by design?. Although recruiting of participants to represent diversity in terms of ethnicity and socio-economic status was the initial aim, the researcher was led along different pathways to discover that diversity was reflected in other areas aside from socioeconomic status and ethnicity. The mode of intervention adopted by the various organisations presented a good mix of participants with qualitatively different subjective experiences. 3.7 RECRUITING PARTICIPANTS The purpose of qualitative research is ?to gain an understanding of the nature and form of phenomena, to unpack meanings, to develop explanations or to generate ideas, concepts and theories? (Ritchie et al., p. 82). Selecting participants therefore is based on inclusion of ?constituencies, events, processes? and characteristics (p. 82) that shed light on the phenomena under scrutiny. Taking cognisance of the key features of qualitative sampling outlined by Ritchie et al. (2003), the aim initially was to recruit men who were (i) fluent in English, (ii) were involved in an intimate relationship with their partners for at least 6 months (prior to participation), (iii) had enacted one or more acts of physical violence toward their partners over the past 12 months (prior to participation), and (iv) had voluntarily joined the men?s programme or support group of the particular organisation, whether through self-referral or via referral by family members or friends or other source. The criteria were set out to achieve some basis of comparison based on content of responses across participants. Fluency in English was a requirement in order to preserve the element of first-person accounts and to avoid disruptions to rapport should a translator intervene in the interview process. As a common medium of exchange, it also facilitated the researcher?s interactions with the participants. In addition, certain criteria for exclusion were proposed. Previous research has indicated differences in attributional styles of accounting for violence between self-referred men versus men who are court-mandated to undergo treatment (Dutton, as cited in Holtworth- Munroe, 1992). In addition, circumstances and experiences unique to a prison population would fall outside the scope of the proposed study (Ritchie et al, 2003). In this respect, participants on parole were not included in the study. Those who had legal proceedings against them at the time of the study were also not be included in order to guard against the potential use of research material for court purposes. 52 Whilst much structure was directed into planning for the recruitment of participants, in reality the process of seeking participants was dependent more on the circumstances encountered than on following predetermined procedures. A review of the ?screening form?3 that was completed during the recruitment phase (Appendix C) indicated that the majority of men failed to fulfil all the narrowly defined criteria. Rather than abandon or exclude men who were willing to be interviewed, the time frames specified in criteria (ii) and (iii) were removed. As was illustrated with negotiating access, recruiting interested participants for the research was dependent on the researcher ?taking the back seat?, being guided by the circumstances surrounding the research, rather than dictating how the research should proceed. The eventual sample reflected a convenience sampling approach, one which relies on ease of access to participants rather than based on a clear sampling strategy (Ritchie et al., 2003). The recruitment of participants was, to a large extent, dependent on the assistance offered by relevant organisations in and around Johannesburg who had men?s programmes that addressed domestic violence. Richie et al. (2003) outlines two requirements of qualitative sampling. ?Symbolic representation?, denotes that the sampling unit (in this case, the participants) should be representative of the features that typify the phenomenon being investigated; that is, it should ?represent? or ?symbolise? the character of men?s violence toward their intimate partners. ?Diversity? ensures that a wide spectrum of features associated with the phenomenon is included in the sample. These principles form the basis of qualitative sampling, particularly in relation to purposive and theoretical sampling. 3.8 PARTICIPANT CHARACTERISTICS Having enacted violence against their partners was the common ground that brought the participants to the present research. Despite the commonality of experience, the uniqueness of each participant was revealed during the researcher?s one-on-one contact with the men. Eleven of the twelve participants were Black and one White. Apparently conscious of the social ?lines of difference? (Shefer, 1999, p. 158, cited in Boonzaier, 2001) between themselves and the researcher, most participants made reference to ?culture? as a defining feature of their identities. Although half the men were in their twenties, the youngest participant was eighteen years old and the eldest, seventy. Relationship status was also revealed as central to men?s definitions of themselves and their experiences. For instance, one participant noted: ?I?m single but I?m not lonely?. Another commented on being ?involved in a?relationship, but [with] boundaries?. Of the 3 The screening form was initially devised as a means of selecting participants meeting the criteria for the study, but was subsequently used as an information base to facilitate comparisons across participants. 53 twelve men, three were married, eight were involved in intimate relationships and one was single. Those involved with an intimate partner described their relationships, variously as ?a good relationship?, ?healthy?, ?happy?, as one being ?in love?, as one where ?we treat each other in the right way?, as one which involves ?trying to build on? each other, ?good [with] bumpers along the way?, ?sometimes?healthy, sometimes unhealthy?, ?not as good?, and as ?sometimes?fight[ing] each other?. Two men referred to themselves as students. One participant who was unemployed described himself as ?a loafer?, whilst another, also unemployed, revealed that he spent most of his time ?sitting at home?. One retired participant volunteered most of his time in support groups assisting abusive men to change. The men who were employed were engaged in wide-ranging occupations involving, for instance, housekeeping, handiwork, selling, community liaison work and entertaining. For the majority of these men, grade 11 was the highest educational level attained, with few holding a matric (grade 12) qualification. Apart from the three participants recruited during a two-day workshop and one participant who had attended personal counselling, the men were involved in a men?s programme for at least eighteen months. The men positioned themselves at different points relative to their violence (Hearn, 1998). Five men explicitly stated that they were either no longer abusive to their partners, or had started a new relationship and were not abusive in their current relationship. The remaining men either admitted to being violent in their current relationship, or volunteered information pertaining to their last violent enactment on the screening form. Inconsistent information was obtained from one participant regarding his history of abuse (i.e. mismatch of information on the screening form and interview). Descriptions of violence by these men reflected forms of abuse ranging in severity. These were characterised in expressions such as: ?smashed? chair, ?shouting?, ?slapped?, ?klapped?, ?smacked her all over her body?, ?punched?, ?grabbed?, ?pulled?, ?pushed?, ?grab her hair?, ?beat? with a ?feather stick?, ?mop stick? or ?bricks?, ?really beat her with the bottle?, used ?fist?, ?kicked?, ?fetch? her and ?hit? her, ?smacked? on the face, ?beat [her] to have a blue eye?, ?raped? and ?force her to have sex?. 3.9 ETHICAL CONSIDERATIONS Intrinsic to negotiating researcher-participant relationships is a consideration of ethical principles to ensure safety for both parties. Following Lewis? (2003, p. 66-71) guidelines, this section addresses the principles of informed consent, anonymity and confidentiality, and mechanisms to safeguard harm. 54 3.9.1 Informed consent Letters of invitation to the prospective participants detailed the rationale, objectives, expectations and ethical concerns relating to participation in the research. Voluntary participation was stressed and that participation or non-participation would not jeopardise their group membership in any way. The men were also fully informed that participation required a 60 to 90 minute in-depth interview that focused on a vivid experience of violent enactment toward his intimate partner. Formal signed consent was obtained on the day of the interview. Individuals were informed that they are at liberty to refrain from answering questions that were too sensitive in nature, or even withdraw from the study at any time, without being prejudiced. With additional signed consent, the interview was audio-taped (See Appendix D & E). 3.9.2 Anonymity and confidentiality Anonymity was guaranteed, based on the principle that the identity of the participants would not be made known outside the research context (Lewis, 2003). Confidentiality, on the other hand, ensures that no identifying information of participants will be presented in any form in reports, comments or presentations, whether through direct references to the individual in reports (such as by name) and via indirect means (such as contextual information) (Lewis, 2003). The participants were assured that only the researcher would have access to the interview material (transcripts and recordings). Whilst certain extracts from texts would be quoted verbatim, participants were assured that no associations between quoted texts and their identities will be made. In cases where contextual information could have potentially identified specific participants, these were omitted in the transcriptions. The participants were informed that recordings from the interview would be destroyed following transcription. Research findings, following examination of the research report, will be also be made available to participants and the organisations. Participants were also informed about the possibility of publishing the findings in a journal at a future date. 3.9.3 Protecting participants from harm Given the sensitive nature of the research, the researcher attempted to guard against possible harm to participants. Prior to commencing interviews, the participants were cautioned that sensitive issues and possible feelings of discomfort may arise during and after the interview. As a safeguard against possible re-enactment of violence towards their partners, the participants were required to have a debriefing session with a social worker or counsellor on the premises immediately following the interview. Contact details of counselling services were also provided in the invitation letter to participants and the men were informed that the researcher was unable to render counselling to them. Participants were cautioned that the researcher would be ethically bound to report any information 55 revealed in the interview context that could be potentially harmful to self or others so that efforts can be made to intervene. Arrangements were made for the availability of a social worker during interviews and for debriefing purposes. During interviews, the researcher was attentive to participants? manifestations of discomfort (i.e. expressed verbally or observed via physical gestures). Where participants displayed discomfort, the researcher highlighted this to ascertain whether participants felt comfortable to proceed with the interview (e.g. ?Do you feel comfortable talking about this??). Process checks were also undertaken in the concluding phase of the interview (e.g. ?How do you feel telling me about this??). 3.9.4 Protecting the researcher from harm Research participants were men with a history of violence toward their intimate partners. To ensure researcher safety, precautions were taken to ensure that all interviews were conducted on the premises of the organisations. Staff and social workers were present on site on the dates on which interviews took place. The researcher?s contact details were not provided, but participants were able to make contact with her via her university email or through the co-ordinators of the various men?s programmes. 3.10 PILOTING Assessment of the interview guide on aspects of clarity, scope, depth and content was a critical phase of the research (Arthur & Nazroo, 2003). This provides an indication as to whether it is adequately flexible to comprehensive and coherent accounts of violence (Arthur & Nazroo, 2003). Following a pilot on the first two participants, changes were made to the interview schedule. ?Level of education? was inserted in the first section in order to illustrate men?s present circumstances (Anderson & Umberson, 2001; Cavanagh et al., 2001). Prompts related to relationship length with present partner, commencement of abuse, and previous history of violence in relationships were also included. Although the intention was to elicit participants? memory of a violent incident restricted to the past year (to reduce recall bias), it was subsequently decided to allow participants to talk of an incident that was most vivid or meaningful to him. Various studies have revealed that events steeped in emotion are more often remembered and with greater clarity and detail (Christianson & Loftus, as cited in Mentis, n.d.). Lastly, the participant?s recall of an event was followed by a more general follow-up prompt which aimed to stimulate his thoughts on other reasons for his violence, apart from what was previously mentioned. This allowed him to generate alternative meanings to his violence (Rossman & Rallis, 2003). The prompt referring to emotions expressed during violence was redundant and was removed. Whilst most revisions were put into effect following the piloting phase, with the experience, skill, spontaneity and insights acquired following each interview, other 56 prompts subsequently were added or excluded. For instance, to ease the transition from the ?re-experience? of an emotional event to present reality, some participants were asked to reflect on the actual process of telling their story (i.e. ?how do you feel telling me about this experience??). In some respects, this facilitated returning their focus to the present, what Labov (as cited in Riessman, 1993, p. 18) terms the ?coda?. Unlike quantitative research, the pilot interviews were not excluded from data analysis. Arthur and Nazroo (2003) note that unless significant changes are made (e.g. revising research aims), the pilot data derived from qualitative research forms part of the data analysis. The interviews were conducted over a total period of three weeks. These ranged from 16 to 68 minutes. The transcription process, which entailed the verbatim recording of participants? words into readable text, yielded approximately 163 pages (single-spaced), revealed the essence of qualitative research. As Scheff (1997) notes, via transcription, the researcher is able to convey participants? words and gestures more accurately. 3.11 DATA ANALYSIS Data analysis is ?the process of bringing order, structure, and meaning to the mass of collected data? (Rossman & Rallis, 2003, p. 278). It is likened to the creative processes of immersion (in relation to words and impressions of the data), incubation (allowing the data to generate mentally), gaining insight (identifying the meaning entrenched in the data) and interpretation (making sense of the findings and drawing conclusions) (Bargar & Duncan, as cited in Rossman & Rallis, 2003). The research employed two levels of analysis: thematic analysis of content, and analysis of narrative structure. 3.11.1 Thematic Analysis In line with principles of interpretive analysis, the texts were interpreted from the stance of empathic understanding. Upholding the interpretive-hermeneutic process, analysis results in a ?thick description? of men?s experiences of their violence, as well as accounting for the researcher?s role in constructing this description (Rossman & Rallis, 2003; Terre Blanche & Kelly, 1999). The emphasis on ?thick description? provides the basis upon which meanings, values and ?rules? by which people order their daily lives are derived. In this sense an ideographic statement is made (Guba & Lincoln, 1983; Neuman, 1994). Although various definitions of thematic analysis abound, it is generally defined as a method of data analysis that provides a coherent means of data organisation into thematic headings which reflect both the research question and the preoccupations of the participants as these are revealed in the (interview) data (Banister et al., 1994). The approach offers an analysis that is accessible and is not attached to 57 any pre-existing theoretical framework. As such, it can be framed from both an essentialist/realist method (that reports on the subjective reality of participants, their experiences and meanings), or a constructionist method that locates the meanings, experiences, events and realities of participants within various societal discourses (Braun & Clarke, 2006). Embracing elements of both positions, the present study utilised thematic analysis as a ?contextualist method? (p. 81), one which acknowledged subjective meaning of experiences, at the same time, taking cognisance of how the societal context is imposed on subjective reality (Braun & Clarke, 2006). 3.11.2 Locating the research epistemology Prior to conducting thematic analysis, various decision-making processes refined the strategy in approaching the data. Braun and Clarke?s (2006) definition of a theme as a ?patterned response or meaning within the data set? formed the basis of theme identification (p. 82). Different to a ?category? which is usually explicit, the theme may comprise a phrase or sentence that describes more subtle processes (Rossman & Rallis, 2003). Theme prevalence was ?measured? in terms of the theme appearing across the (entire) data set. In other words, the majority of participants had to allude to an idea before it constituted a theme. Apart from repetitions, participants? use of metaphors were also identified (Ryan & Bernard, 2003). Although no specific analysis of metaphors was undertaken, they were used as evidence for a particular theme. Secondly, an inductive or ?bottom up? approach to analysis was conducted. This involves ?a process of coding the data without fitting it into pre-existing coding frames, or the researcher?s analytic preconceptions? (Braun & Clark, 2006, p. 83). This data-driven process allowed themes to emerge that moved beyond the specific research questions asked (Braun & Clark, 2006). This bottom-up approach suggests that general rules or classes are inferred from specific instances. Through coding ? ?the process of defining what the data are all about? (Charmaz, as cited in Ezzy, 2002, p. 86) ? themes, categories or concepts are ?allowed? to emerge from the data. Thirdly, the level of analysis in the present research whilst drawing upon the semantic content (i.e. the surface meaning or descriptive level), also progresses beyond what is said to examining the latent content (i.e. the underlying assumptions or interpretive meaning). In so doing, effort is made to ?theorise the significance of the patterns and their broader meanings and implications? (Braun & Clarke, 2006, p. 84). In this sense, the research epistemology is situated between the two extremes of essentialism/realism and constructionism. As noted earlier, the interpretive framework privileges the individual?s subjective experiences which are accorded as real and as emanating from the self (Terre 58 Blanche & Kelly, 1999). Based on the assumption that language directly and passively reflects meaning and experience, this position asserts that motivations, experience and meaning can be theorised in an uncomplicated manner (Braun & Clarke, 2006; Burr, 1995). Constructionism, on the other hand, is concerned with ?the broader patterns of social meaning encoded in language? (Terre Blanche & Durrheim, 1999, p. 149). Rather than seeing language as the pathway to ?underlying realities?, for constructionism, language itself is the object of study. Being non-neutral in nature, it serves to construct reality. As such social behaviour and practices are seen as fundamentally constituted in language (i.e. when we talk, we actively construct our world) (Burr, 1995; Terre Blanche & Durrheim, 1999, p. 149). In bridging the gap between essentialism/realism and constructionism, the present research allowed for meaning to be theorised on the level of both subjective meaning and the sociocultural contexts that give shape to individual accounts (Braun & Clarke, 2006). 3.11.3 Conducting thematic analysis Thematic analysis was based on guidelines offered by Braun and Clarke (2006), who formulate a step-by-step procedure following their critical review of research utilising thematic analysis as a data analytic tool. In addition to Braun and Clarke (2006), however, recommendations offered by Boyatzis (1998), Coffey and Atkinson (1996), Marshall and Rossman (1995), Rossman and Rallis (2003) and Ryan and Bernard (2003) were also considered: i. Data Management Following the interviews, as well as during transcription, information pertaining to interview length, date and participants? identifying information was recorded. ii. Data Familiarisation Data immersion, a process of familiarisation with the ?breadth? and ?depth? of the data (Braun & Clarke, 2006; Rossman & Rallis, 2003) was undertaken during the transcription phase. Particular points of interest and potential patterns were highlighted to mark attention for subsequent analysis. During transcription, effort was made to ensure that the final product mirrored as closely as possible the original verbal account (Braun & Clarke, 2006). Whilst many transcription methods exist (e.g. as in conversation analysis with its complex notation) (Elliot, 2005), Banister et al.?s (1994) guidelines were adapted and modified as a midway between complex and more simplistic transcription (See Appendix G for transcript notation). Participants? speech characteristics, such as pauses, emphases and indistinct words, were recorded. Following Boonzaier (2001), notation was also provided to reflect pronunciation variants. 59 iii. Generating Codes Following transcription of recorded interviews, the data was organised through reading and re-reading of texts (Rossman & Rallis, 2003; Marshall & Rossman, 1995) during which points of interest were jotted down (Braun & Clarke, 2006). Using the research and interview questions, preliminary categories were proposed to provide direction for subsequent data gathering and analysis, but were subsequently regrouped to reflect the actual data emerging from the analysis. Whilst categories direct the gathering of data, thematic analysis emerges from ?the deep familiarity with the data that comes from categorising? (Rossman & Rallis, 2003, p.282). Coding of the data, which is defined as ?formal representation of categorising and thematic analysis? (Rossman & Rallis, 2003, p.285) was facilitated using ATLAS.ti, a qualitative analysis software tool. Whilst the manual coding process would follow an identical procedure (i.e. reading through each line of the transcript, identifying and naming codes to summarise the data), the software package was utilised for its ease of data organisation. Once the coding process was complete, a list of codes and attached quotations was stored by the computer and printed. This strategy economised on the ?cut and sort? technique which involves the physical sorting of numerous quotations into piles (Ryan & Bernard, 2003). Coding, according to Rossman & Rallis (2003), constitutes the ?evidence? of a category or theme, and is thus narrower than the theme. In this manner, codes created were matched with specific data extracts. Following Braun & Clarke (2006), data extracts were often coded several times according to their fit with potentially different themes. iv. Themes Search Themes were extracted by sifting through the data for organising principles underlying the texts, in other words, recurring ideas, analogies, patterns of belief and actions which signal a deeper level of understanding (Rossman & Rallis, 2003; Rossman & Marshall, 1995). Based on a list of codes, themes were generated via two processes: a) various codes were grouped into potential themes, b) coded extracts were collated according to the identified themes (Braun & Clarke, 2006). v. Themes Review Once a set of initial themes were devised, the data extracts were reviewed once again to assess the degree of fit and coherence between the themes, codes and their supporting quotations (Braun & Clarke, 2006). Certain themes were broken down into sub-themes. For instance, the theme violence as a ?loss of control? was subsequently divided into two sub-themes: i) ?my emotions took control? and ?I wasn?t thinking straight?. On another level, themes review also involved assessing the validity of individual themes in relation to the data set as a whole. At this stage, other relevant information that was not included in initial coding was incorporated into the final themes. 60 vi. Labelling the themes Braun & Clarke (2006) and Boyatzis (1998) note that the label assigned to a theme should capture the essence of its content. This involved moving back and forth between categories, codes and themes. For each data extract, specific information was extracted that reflected the ?story? of each theme. As demonstrated by each phase of analysis, a complex and mass amount of data was gradually simplified into manageable chunks. The following phase of analysis involved data interpretation, namely obtaining meaning and insight into the words and acts of the participants and their meaning for both the participants and researcher (Marshall & Rossman, 1995). As Terre Blanche & Kelly (1999) state, the purpose of interpretive analysis is not merely the collection of ?snippets? of ?real life?, but also involves contextualising these descriptions into some kind of perspective. Whilst the present research focused on understanding the subjective experiences of abusive men on an individual level, it also aimed to contextualise individual understanding in relation to performative aspects of talk and their interrelations with much broader societal ?themes?. 3.11.4 Narrative content and narrative form Reflecting the dual aims of the research (i.e. to explore both subjective experiences of men?s violence and the performative elements of their talk regarding violence), thematic analysis addressed both the content and form of men?s narratives respectively. In relation to content the aim was to elicit a ?thick description? of their experiences. Based on a phenomenological framework, themes relating to various facets of the phenomenon of violence were identified. This involved attending to men?s descriptions of their violence, which typically reflected various levels of experience, namely on the level of the body, thoughts, emotions and on the psychic dimension. In order to locate men?s subjective experiences in broader terms, however, it was necessary to draw upon analyses that reflected the societal context in which their experiences were embedded. Utilising a narrative approach, attention was directed to the form of men?s accounts. As Coffey and Atkinson (1996) propose, a narrative approach provides insight into ?the ways in which social actors produce, represent, and contextualise experience and personal knowledge? (p. 54), which is not possible with merely the analysis of content. The present research therefore also examined the ?storied qualities? of the data with a view to understanding how the participants conveyed meanings through language. Such an approach was adopted to examine the data based on their status as accounts (Coffey & Atkinson, 1996). According to Besley (2005), the act of verbalisation serves a performative function. Stated differently, individuals in recounting events of the past perform certain 61 kinds of ?speech acts? such that they position themselves within particular reference points based on their own or others? evaluations of them (Coffey & Atkinson, 1996). During this process of narration, the self identity is both constructed and discovered (Ezzy, 1998). In the context of the present research, the aim was to examine how men accounted for and talked about their violence toward their intimate partners (Riessman, 1993). 3.12 CONCLUSION In line with the ?experience-near? pillar of the interpretive continuum (Kelly, 1999), the present research attempted to explore men?s experiences of their violence based on the ?insider perspective?. At the same time, a distanciated perspective was adopted in order to contextualise these experiences within a broader social context. Based on these epistemological frames, this chapter served to substantiate the use of the in-depth interview as a methodological tool in eliciting subjective accounts. As an analytical tool, thematic analysis was employed to reveal two levels of analysis. The next chapter details the results of the present research through a careful analysis of narrative content (i.e. men?s descriptions of the violent experience) and narrative form (i.e. the process by which men talked about their violence). Thus, the tension between personal or subjective experience and socially constructed meanings (derived through discourse) are depicted. 62 CHAPTER 4: RESULTS 4.1 INTRODUCTION The data analysis is presented in two parts. Part I will focus on the content of the participants? talk (what was said). Insight into the men?s subjective experiences of their violence is obtained via attention to the emotional and psychological dimensions. Attention to content, therefore, is in line with the ?experience-near? end of the interpretive continuum introduced in the previous chapter (i.e. the ?lived experiences? of their violence in its context) (Kelly, 1999). However, an exclusive focus on content or subjective experiences suggests that men?s talk about their violence is taken at face value. Whilst attention to the subjective elements illuminate the psychological reality of experience, it is necessary to shift the lens toward the ?experience-distant? pillar of the interpretive continuum to understand the process by which the participants engaged in their talk of violence. Part II addresses the question of how men related their narratives of violence. Attention to these aspects introduces a critical reflexive gaze on the researcher?s influence on how the men ?shaped? their talk. As the listener, she mediates the space between the immediate social interview context and a broader societal one. Part I: Men?s Subjective Experiences of Violence In men?s narratives, there was a conflation of anger, temper and violence. Three men made reference to ?heat? as a metaphor for anger: ?feeling heat? (P2) and ?hot? (P6), and for ?feeling? violence: ?White hot (.). Red (.). Blind (4). Desperation (3). Out of control?? (P9). Physiologically this was experienced as being ?nar? or ?sick?. For instance, P7 noted: ?when you feel like maybe you are, you are sick, you can?t even/?/ understand anything/.../until you do this [violence]?. In other instances, anger was experienced as hyperventilation: ?Sometimes find that/?/when I breathe, I breathe faster? (P8) and rapid heartbeat: ?when I become very angry and (.) my heart will pound very fast? (P3). In describing anger on the physiological level, the notion of being ?out of control? was experienced on the level of the body. The interrelations between themes were highly evident, supporting the phenomenological supposition that each depicts various facets of a particular phenomenon (Thomas, 2003). Five main themes were identified and participants? definitions of self from past ?abuser? or ?batterer? to perceptions of their present selves as ?changed men?. 63 4.2 THEME 1: VIOLENCE AS ?LOSING CONTROL? 4.2.1 ?My emotions took control? Violence was depicted as a volatile state during which men?s emotions were capable of driving behaviour in ways that were unpredictable and uncontrollable. For ten participants, ?losing control? was an emotionally overpowering experience, typified by states of anger, rage or temper. P6, for instance, remarked: ?Anger was controlling me?. Construed as having its own life force or energy, anger was seen as contained within one?s self, latent, waiting to be set off by something external, as revealed by P5: ?Anger is waiting for somebody to trigger it, that?s how I see it in me. Don?t trigger that, it?s there, it?s closed, I?m happy now, but if you push me to a stage where I?ll react somehow?. For P9, violence was described as an ?explosion of temper?. The build-up and release of anger, for P5, was a ?point of no return?: I feel like, I?ve unleashed something that was eating me (.) out of, violence and so, I was, unleashing them because that?s why I was hurting, emotionally, so because I don?t share my anger with some people, I?m building it up, the time it comes out (.) I go ballistic ? there?s no turning back. P7 alluded to the dissipation of anger following a violent eruption: ?It help me because after beating her, I was not angry anymore?. Other participants recounted that in the moment of anger, they were desensitised to the obvious suffering of their partners, or to the potential negative consequences of their actions to themselves: You do things by emotional?okay?so you forget of saying (.) that person got a pain or what??cause of you are emotional. You got anger to do with you, so by the time of the accident, I just wanted to beat him [his wife], whether she..she scream or what, I forgot. (P4) Similarly, P11 expressed the short-sighted nature of his anger: ?You can kick someone when you are angry (.) you are doing bad things, you don?t care if people see you or don?t see you ? just doing your thing?. Other men made reference to the potentially lethal consequences of their violence when their emotions took control. P8 articulated: Sometimes you find that I don?t only fight with my hands. Sometimes I use weapons, dangerous weapons, whereby I didn?t know that maybe it can damage something in your body, you know, find that maybe I?m too angry, can?t control my feelings and that I take a bottle and I?I hit you in your head, find that you got scars, you know. In a similar vein, P3 explained: ?The temper that I had at the time, as if I had a gun I would, I would just take out the gun and shoot??. P10 also recounted: ?but what I feel at the moment, I feel like eh uh [laughs] a lot of things come into my mind, I don?t know, if I had a gun, maybe I was going to use it, I don?t know. If I had a gun that time to use it.? 64 Using metaphorical language, the element of uncontrollability of anger or emotions is likened to a ?drunken state?. P2, for instance, commented: ??like when you drunk and you do things and you tell people that I was drunk ? yeah. It?s a similar type of thing, but now you?re sober?. Referring specifically to the role of alcohol in an episode of violence, P9 described this uncontrollable state as one of ?drunken rage?. In a vivid illustration of uncontrollable rage, P5 depicted the self metaphorically as a ?half man/half beast?. Revealing a dissociative split of two selves, the violent self is seen to emerge from within, taking control of the ?real self?. This construction of dual self as Hearn (1998) would suggest, presents violence as an exceptional event, as something out of the ordinary, and that the behaviour is somewhat ?out of character?: P5: I feel like, I?ve unleashed something that was eating me (.) out of, violence Ursula: Mm hm. Can you identify or put a name on what that thing was that was eating you? P5: Mmm (2) ei, most I think it?s sphinx I don?t know, maybe that..that sphinx, sphinx or what it called? Ursula: Mm hm? P5: yeah, the thing of a (.) head of a lion, body of a man ? something like that (.), like it was a mixture of, you know things like?I can?t explain them? Less dramatically, P2 remarked: ?I was not me in a way /?/ Nah (.) I was not feeling like me. You know I felt so small, I was so small for myself, you know (.) you know (.), felt like, eish I?m nobody now?. P7 affirmed that the self that emerged during episodes of violence was not really him: ?No, I, I?m not that person. I only act violent at that time. Maybe I can be angry for only two minutes, five minutes, then after than I start, I start (.) blaming myself?. 4.2.2 ?I was not thinking straight? Different to ?losing control? as an emotionally volatile state, men?s loss of control was also equated to having temporarily lost the faculties of reasoning or logic, resulting in a reduced capacity to make sound judgements about situations (Hearn, 1998). Nine men recounted irrationality in their thinking during the violent episodes, typically depicted by haziness, confusion and distorted judgments. P5 summarised this point: ?When you are violent, you are not thinking straight? and questioned the logic that ?thinking? and ?violence? could co-exist: ?I (.) I don?t think I was even thinking (2). I..I can?t explain it. I can?t explain it (.) because really if you are thinking, you wouldn?t do whatever I was doing (.)?. The association between irrational thought and impulsive action was also made apparent by P11: 65 I didn?t think very straight. When I saw him, she ran away. She was running away, she knows what she has done (5) and I fetch him [girlfriend], and I hit him [girlfriend] (7) and I didn?t care that she must die or what (3) at that time. Echoing this view, P3 made reference to violent actions constrained by the confused and panic-stricken mind, depicted as having a life of its own and having power to direct all acts: At the time, you find that my mind (.) is clouded?em (.). I?m angry and my mind is clouded. That means I?ll pick up a stone or (.) pick up something (.). So the danger with that is that you find that sometimes I..I am a kind of person who will do something that I will regret. The mind, in its disorientated, confused and bewildered state is further revealed by P2: [Sighs] ai (.) like there were so many things you know. It?s when it?s like when you ever been in a television room whereby security guards check televisions, like you see so many views of places, different kinds of places so that?s the situations that I was in, you know. I saw so many pictures in my mind. I had so many different kind of minds (.) you know. During this dissociative state, the violent self once again was experienced as overpowering the ?real self? (Hearn, 1998): ?you know, so the mind was just grabbing and doing you know. I was not me in a way?. The role of alcohol in producing disorientation was highlighted, particularly with reference to the intensifying effect that drink had on their mood and ability to make rational judgments. In particular, P8 noted how drinking strengthened his will to carry out potentially lethal acts: ?At that time, I used to (.) I was like (.) drinking a lot, you know /?/ Cause when I think of eh (.) I thought of many things, you know (.) I?ll kill her..I?ll kill that girl (2) things like that??. In a similar vein, P6 attributed his poor judgement regarding his partner?s fidelity to his excessive drinking: Oh, I used to drink a lot (.). One day, if I remember it was December, we both went to trip, have a drinks, drinks and then dance. I found her eh, like busy with my friend talking together?so I took it in a (.) in a bad way that she didn?t want me, which wasn?t, and I started beating her up?Then (.) by the following day, morning, I realise no that I am wrong. ?Not thinking?, reflecting the irrationality during the violent act, was contrasted with ?sober? reflection and hindsight after the event. These contrasts permeated men?s descriptions of their violence. For instance, ?jumping to conclusions?, a phrase used by P3, conveyed the short-sightedness of his actions which was juxtaposed with his more rational and reflective appraisal of the event: 66 Maybe it?s because she..she will take long time to answer when I ask question (.). She takes some times and (.) that will make me to conclude ? I was very quick at concluding in fact?(3) As I said, during the process (of beating) I would feel (2) nothing because em maybe at the time you?ll find that I?m angry and (.) I?m doing what I think I should be doing? (.). But after some times, I would think that (.) usually I will think about myself than (.) what I did. I would think about the situation vis-?-vis?/?/ the problem that I think was a problem (.) eh trying to check if (.) my reaction was up to standard to the problem (.) that I..I think I had. But most of the time, I would find that (.) my reaction was worse than the problem. As the excerpts have illustrated, ?losing control? was manifest as the breakdown of rational judgement and depicted as a dissociative experience during which the irrational self had no control over his actions. However, P3?s reflective statement: ?I?m doing what I think I should be doing? revealed an element of intentionality alongside the irrationality of violence. Juxtaposed with the ?losing control? narrative was one which portrayed violence as a deliberate attempt at having control over their partners. 4.3 THEME 2: VIOLENCE AS ?HAVING CONTROL? OVER ANOTHER In contrast to experiences of losing control, men also experienced their violence as an intentional and conscious exertion of control. ?Having control? over one?s partner was manifest in three forms: having ownership, instilling respect and enforcing discipline. 4.3.1 ?I thought I owned her? Reflecting a discourse of ?men?s social ownership of women?, elements of possessiveness and sexual entitlement in men?s intimate relationships were evident (Ptacek, 1997). Seven participants expressed ownership over their partners. Violence was a typical response to feeling threatened by their partner?s associations with others, in some instances, fearing they would be influenced negatively by friends. Underlying these concerns were suspicions of her infidelity, whether actual or assumed. P2 recounted that not finding his partner at home fuelled suspicions of infidelity and precipitated violence: So she came back the same?the next morning (.) you know. So you never know where they were, you know, all night and stuff. ?Cause they will tell you the next day, ?Okay I was going to see my other aunt?, okay and after the aunt says ?No no?. ?Okay let me tell you the truth, ?cause I was scared ?cause you gonna beat me up..blah blah blah?. Then you ask them, ?when did I beat you up the first time?? Like then it tends to be like violence now ?cause you get pissed off, you know. Someone is lying and you?re trying to be kind. P3 related feeling ?insecure? about his partner?s whereabouts. Her physical attractiveness was a cause for concern and led to suspicions of cheating: ?she was beautiful [laughs], you know um (.) and eh (.) I know that every time when I don?t find her (.) eh (.) my conclusion is that is with a boy somewhere /?/ I remember I myself 67 beating her, for that reason that I don?t find you? ?where were you???. The majority of men subscribed to discourses denoting women as male ?property? (P3) that sanctioned controlling behaviours through violence. P6 remarked: ?Ja, because we in a relationship /?/ She?s mine, everything?s mine, all is mine (.), nobody can tell me even her, can?t tell me?. Linking ownership to control, P2 stated: ?I thought I was owning, you know when you in a relationship, you owning each other and stuff, I was used to like talking to her like, ?Okay stop this?. If I say stop this, she?s gonna stop ?like control each other?. Beyond the discursive level, these claims to ownership are also explicated on the psychic dimension. Male identity development for Chodorow (1978) is premised on the notion of ?difference?. Whilst forging identification with the paternal figure who is seen as similar, the boy rejects recognition of himself in the mother who is regarded as different. Not only different, the woman is considered ?other?, as less-than-human ? able to be owned but not recognised, able to have her but not become like her (Beauvoir, as cited in Benjamin, 1980; Chodorow, as cited in Benjamin, 1980). In this manner, ?rational violence? becomes operative ? rational in the sense that it is voluntary and plays out in issues of possession and control. It is seeking recognition that is motivated by a need for intimacy but denies this and its associated feelings (Hagemann-White & Marcus, 1999). The ?moment of control?, as Hagemann-White and Marcus (1999) point out, becomes ?a calculated transgression in which the will of the object is denied, the recognition of the self is forced and the end point is determined only by the subject? (p. 5). This was evidenced on men?s descriptions of control on the level of women?s sexuality. A refusal of sex, interpreted as ?evidence? of her infidelity, for instance, incited the cause for violence. P7 portrayed this understanding: ?It become violent because you going to start up beating her?accusing her that she had other boyfriends, so that?s why she doesn?t want sleep with you?. Reflecting the theme of ownership over one?s partner, is the assertion of active on passive, domination and submission (Benjamin, 1990), as revealed in the statement by P7: ?Because ah she doesn?t want to sleep with me?and I want to sleep with her, so I start up beating her?. The denial of mutual recognition, the acknowledgment of another as separate from oneself, having feelings, desires and wishes of her own (Benjamin, 1990) was further evident in P1?s statement: ?She was always screaming (.) screaming, telling me that I?m hurting her but I would continue?I would continue, because the only thing that I was more about concentrating was my feelings, not about her feelings?. Underlying the use of force were fears of rejection and suspicions of infidelity. As Hagemann-White and Mircus (1999) articulate: ?the insistence of force, control and 68 omnipotence is often the only way to approach the feminine without feeling immediately threatened? (p. 5). P6?s response to his partner?s refusal of sex underscores this point. Her rejection was interpreted as evidence that she didn?t ?love me? (P6) and was met with feelings of loneliness: ?I felt like I?m alone, I don?t have a girlfriend?. Sexual coercion, beatings, followed by his ultimate ?disowning? of her indicates the extent of his control and his inability to tolerate her self-assertions: My girlfriend came to visit me?then I said ?can we sleep and have sex?, you know. She said ?no?. She said ?no, but we in a relationship, no, you mustn?t do like this?. I force her. He didn?t want, then I left. I left her, I beat her up and said, ?go? (.) and she did go /?/ She refused and so I beat her up, and say, ?go, I no longer, I no longer want to see your face. The theme of social ownership is most vividly illustrated by P5 whose description of violence converges in an interplay of jealousy, possession and ?obsession?: If I can see her being hurt, maybe if I can give her a scratch, she?ll just concentrate on me because maybe nobody will like her with those scars. I was thinking bad things to do to her, so that maybe she, she only be involved with me because I am the one who did that to her, other people won?t see her eh much prettier (.) except for me. Ja, those were the things I was thinking in my mind, just if I can do that, maybe I?ll set my territory. On the subjective dimension, this excerpt vividly depicts how control clearly becomes an issue. Unable to tolerate the anxiety arising from potential abandonment by his partner, he forces recognition onto her who is seen as the object, and in the process, he silences her will (Hagemann-White & Micus, 1999). Relating to Hegel?s (1952) ?master-slave dialectic? (i.e. for the master to be recognised, he is dependent on the slave to recognise him as such) (as cited in Benjamin, 1980), his need for recognition becomes the ?vehicle of domination? (p. 151). The dilemma of dependency is resolved in his possession and control of another. In this instance, the act of ?marking his territory? is not only a means to secure her as his property. In a literal way, he translates ownership into a tangible or physical one. In Benjamin?s (1980) words, the ?psychic pain of separation is captured in the physical pain inflicted upon the violated? (p. 161). She becomes the literal object of his affections and he retains his possession of her. By defacing her appearance, she is no longer available for others to lay claim to and is forced to recognise only him. Ownership, possessiveness and suspicions of infidelity were intimately linked with men?s use of violence. P7 feared that his partner?s associations with a girl friend would lead to infidelity: ?I thought that she?s [partner?s friend] going to give her boys.? About maybe one o?clock?in the morning, they came, having some bottle of beers and, they were drunk?So knock on the door, so I opened the door. I ask her where she, 69 where she, where she was [?] so I started telling her that ?you see that girl of yours?, okay and we argue, argue, I end up beating her and really beat her with the bottle. In this excerpt, violence denoting ownership and violence as an expression of disapproval of another?s behaviour is blurred (Hearn, 1995). This introduces another distinct, yet related theme: ?As a man, I needed to correct her for what she did wrong?. 4.3.2 ?I had to show her who is in control? The ?head of the household? discourse often cited in accounts of abusive men (Dobash & Dobash, 1979; Fuller, 2001; Moffett, 2007; Ptacek, 1988; Winstok et al., 2002; Wood, 2004) was also a prevalent theme in the research. Closely related to the theme of having ownership of one?s partner, these participants subscribed to the view that men have certain rights, privileges and entitlements over women. Where these entitlements were unjustly withheld, violence was the outcome. Rich (as cited in Ptacek, 1988; 1997; Winstok et al., 2002) proposed the notion of ?husband-right? (and by extension ?boyfriend- right?) which ?men are presumed to enjoy simply because of their gender: the ?right? to the priority of male over female needs, to sexual and emotional services from women, to women?s undivided attention in any and all situations? (pp. 219-220). Whilst the notion underlying this theme is about beating one?s partner as a means to control or regulate her behaviour, two seemingly distinct yet interrelated sub-themes emerged: exercising violence (i) as a form of punishment, and (ii) to demand respect. (i) She needed to show me respect as a man Closely mirroring this theme is P4?s assertion regarding his authority status in the home: ?The thought was that I want to satisfy myself that she must obey the law, of this house?. Discourses of ?man as authoritarian, governor and lawmaker? who has full dominion over his woman (Winstok et al., 2002) are revealed in his depiction of his wife as an ?obstinate mule? that requires coercion to obey its master. His account was noticeably a gendered one as he emphatically drew upon the ?good wife? discourse, holding up as ?ideal? the virtues of obedience, submissiveness and faithfulness (Ptacek, 1988; 1997): So (.) let me try this way, maybe she will listen what I?m meaning, so (.) if doesn?t, you know, like for instance, we said ?you cannot take a (.) horse, you cannot force the horse to drink the water?but you can take it to the river?. So I said, let me take this person and show him where I wanted to drive him. If doesn?t want, it?s where I can leave, but let me take him on this road, you know this road of beating that this is wrong. I don?t need it but if you wanted me to keep on beating, that means the person is not suitable for me?because if you are the woman being, you listen when (.) something is happening to you once, not more than once. 70 Similar to P4, P6 made reference to forcing his partner to listen by instilling fear or threat of force in response to challenges to his authority as a man: If I say ?one?, she can?t say ?two??When I say a thing, I mean it, should understand ? kind of making her afraid of me, never cheat on me /?/ she listen to me because (.) some days I used to beat her up, I used to beat her up, (as) they say do your homework, I do it. As illustrated in this excerpt, control is tied up with the notion of using force as a reasoning power (Benjamin, 1980). When analysed in terms of ?motives? or ?strategies?, the men appeared to justify their violence in response to their partner?s failure to listen. For instance, P9 related feelings of frustration at his wife for her failure to listen to and acknowledge him: ?On occasions, I would grab her hair and it was always a typical case of (.) frustration ? that word again, you know, listen to me, don?t interrupt me. I want you to listen and try and say to you, ?why do you do this, why do you do that?.? P2 expressed similar frustrations and resorted to physical violence in an attempt to get his partner to ?pay attention?. The use of aggression to mould behaviour was made apparent: Okay, it?s just klapping and klapping, trying to hold like ?listen? you know. Because someone is lying you know and you can see, you can hear. You are not talking the same thing and you are trying to wake them up you know: ?Wake up!? /?/ So eish I was so pissed off, that I slapped her (.) many times in her face, in her head, trying to make her listen. The failure to listen was perceived by some men as a direct challenge to their socially designated status of male authority. Interpreted from a psychoanalytic perspective, however, reveals the subjective elements of being ?driven to violence?. Violence, as noted earlier, appears to be in Benjamin?s (1980, p. 162) terms the ?struggle for recognition? reflecting the dynamics of domination and submission in a framework of a gender hierarchy. The failure to achieve mutual recognition, the breakdown in tension between asserting independence of self and acknowledging the selfhood of another, was manifest by P9 who exploded when his wife had angrily asserted herself, voicing her disapproval at him for being kept waiting for a period of time: She was annoyed, very annoyed at being kept waiting for so long, and she remonstrated with me, and said, ?you know, how dare you keep me?This is not the first time and I really object? and she was really angry [?] I just (.) blew up and I backhanded her in the car. I just lashed out with the back of my hand. In denying her voice, he relegates her to object status: ?It was, here was this person controlling me, telling me what to do (.), how dare she, uh, who is she to speak to me like this /?/ Besides, ?who are you to talk to me that way, you?re just alone, you?re just my wife. Who are you to talk to me that way???. P7, made reference to beating his girlfriend for challenging his authority (i.e. failing to come home at the appointed time, for being 71 drunk and raising her voice at him): ?I?ve told her that, I?ve told her that she must come back at seven. It started because she was drunk, it started (.) she raising her voice to me, so because I was so angry, I started beating her?. For P8, beating his partner was a method of demanding respect: ?I thought, okay fine (.) maybe when I do this, she?ll..she?ll eh start to respect me?. In a more dramatic account, P4 highlighted how violence bordering on threat of death had crossed his mind following his suspicions that she had been unfaithful: ??so I say, let me show this woman that I can kill him alone, but she?that was, you know the concept I was having?. His wife?s refusal to leave their home upon his insistence takes the dynamics of violence and control on another level ? threatening to kill: ?Okay you wanted, want to see I could kill you inside my house??. So that was the accident I nearly (.) do, but because I was so angry but I control myself and say, ?No, let me see, eh maybe after three days, if she can repeat what have happened?. For these men, control once again reflects Benjamin?s (1980) notion of forcing recognition onto the object and denying her autonomy and subjectivity. In effect, he is saying: ?I am the recognised and nurtured one, not the recogniser and nurturer? (p.147). In time, as she loses her right to an independent existence, she is no longer able to recognise him. In a Hegelian sense, in becoming dependent on his slave, he eventually loses his own subjectivity to him ? the slave is gradually unable to recognise him (Benjamin, 1980). In reaching this point, rationality and control is bestowed to one partner, as the other gradually relinquishes herself to him (Benjamin, 1980). On this psychic level, violence in the case of P4 became a perpetual seeking for recognition, over time becoming ineffectual. Threatening his partner with death was introducing a new level of resistance in a dialectic of control. Placing violence against women within the context of gangsterism and township violence during apartheid, P1 drew upon the discourse of male dominance/female submission in justifying acts of sexual coercion: ?So when you call a lady, say, ?come to me?. If she refuse, you will kick her (.) and she wouldn?t just go past that territory?. Beating one?s partner was a means by which a man?s ?rightful? or ?legitimate? authority was reinstated. For these men, their partners? challenge to their male authority was interpreted as undermining their manhood, as P10 asserted: But sometimes our girlfriends makes us feel less ? I mean I?m a man /?/ (2)/?/ In general, where a man feel less uh..or feel like he, he?s violated, I?m not sure, I?m not saying when they are violated they should act in this way, but, violence is bound to happen. 72 Three participants reported gaining a sense of satisfaction in the act of dominating or subordinating their partners. For P5, insult to his manhood was revoked the instant that he exerted his physical strength over his partner: I can say (.) I was satisfied somehow. Now she can know who the man is, whose in control. Because if I don?t beat her up, it?s like (2) I?m a sissy or something?so?that?s why this thing, that?s why it?s happening even now. Men are in control in the relationship, so I wanted to show that ?whose in charge here??. I didn?t think about her woman?s feelings. In a similar vein, P7 remarked: ?I was so proud /?/mm I was proud because she was screaming, so she was not fighting back.? P6 also revealed: ?Aah I was enjoying like beating up a woman. It?s my culture?. P1, relating feelings of pride in having ?conquered? a woman sexually, reflected on the importance of showing dominance as a man: ??so this is how we might end up (.) having sex forceful with the ladies because you say no I don?t want to be beat by the lady, I?m a man.? P3 also recounted how beating his partner distilled in him a sense of ?manly pride? giving him something to talk about with friends: I would say it?s some sort of satisfaction?/?/um (2) because sometimes I would talk to that about that with my friends /?/You know how ?she was not there yesterday, and ei how I beat her (.), she won?t do that again?, eh, I?ll talk about that to my friends? These excerpts illuminate the instrumentality of violence as a means of demanding respect from partners, particularly in instances where male authority was actively challenged. However, the feelings of satisfaction, enjoyment and pleasure that these men alluded to during the act of domination is suggestive of something more than the mere positioning of self in terms of a patriarchal male discourse. The erotic element of domination was also illustrated in these accounts. However, as Benjamin (1980) notes, real masochism is not desire for pain for its own right, but the ?proof of servitude? (p. 156). By extension, the satisfaction experienced by the men appeared to lie in the want of domination rather than in the suffering of another. P1, for instance, recounted: ?And I couldn?t stop, because the problem with us, the more the girl screams, is the more we enjoy, that?s what we are telling ourselves, okay that?s what we are [inaudible] I?m far stronger than her?. In his act of sexual violence, he reaffirmed that he is different and separate from her ?through his power to negate her? (Benjamin, 1980, p. 158). In Hegelian terms, it is the master who ?can know pleasure? and the slave who ?must experience pain? (as cited in Benjamin, 1980, p. 161). Beating one?s partner was justified on the grounds of disciplining. Through punishment, violence became a means by which the men attempted to command respect from their partners. Interpreted psychoanalytically, women were punished for asserting their subjectivity: their desire to be known and their desire to act and be separate from their men (Benjamin, 1980). 73 (ii) She needed to be punished Nine participants made reference to beating their partners in order to discipline or punish them for behaviour which they deemed as inappropriate. For P2, using violence was a means to get his partner to ?listen?. On another level, he justified his ?klapping? her on the basis of her ?lying?. In addition to trying to get her to ?listen? or acknowledge his authority over her, his use of force suggested punishment for dishonesty. He amplified on the notion that violence was a last resort, failing other attempts, to get his partner to ?behave?. You see your girlfriend talking to the same man /?/you feel the other heat now. It?s not gonna be the same like before. Like before maybe you used to talk about things ? this time you not gonna be like talking and stuff, you gonna be using force? The idea of correcting ?bad behaviour? was apparent in P2?s metaphor of ?punishing the bad child?: So it was one situation like that where I thought maybe I felt like she is my kid now, she needs to get a hiding (.) you know?. Employing the reward/punishment metaphor, P5 remarked: ??so I beat her up for a lie?: Eh, by that time neh, I felt that okay, I beat her up, she wouldn?t do this again. If she do this, this is what she?s going to get (2). So by beating her, I thought maybe I think she wouldn?t do this things again, because she know what?s the reward. Echoing this view, P8 emphasised that beating was about teaching or communicating what was ?out of bounds? in a relationship: ?I did it so eh..eh (.) she will know that no I don?t like that (.) She won?t do that, she won?t do it again?. Apart from the parent/child analogy, another more implicit juxtaposition of opposites is apparent. The man was constructed as the disciplinarian, enforcer and punisher of transgressions who legitimated his punishing ? he is right and she is wrong. P12?s justificatory account highlighted this understanding: ?That I?m doing the right thing ?cause (.) I love the girl and the girl don?t love me so, I had to do what?s right for me (.) by beating the girl?. Behaving ?incorrectly? was grounds for anger: ?Ei! ? the anger (2) ? it?s that the girl is doing things the wrong way, you see?. P10 also pointed out that violence was a means to communicate that his partner was at fault: ?But I think violence I..we use it, I use it because I wanted to express, what she did was wrong?. Similarly, P8 drew upon this analogy: ?She was talking to me, she didn?t eh..eh..eh respond correctly?She didn?t..she didn?t answer what I was asking, you know. She was asking, she was also asking me questions instead of..instead of giving me answers?. On a psychoanalytic level, for these men punishment as a form of rational control, was ensuring that the boundaries between the masculine and feminine were upheld. Women?s exclusion from rational individualism, as Benjamin (1980, p. 150) notes, is to 74 be ?either an object of it? (i.e. ?I thought I owned her?) or ?a threat to it? (i.e. ?She needed to be punished?, ?She needed to show me respect as a man?). 4.4 THEME 3: THE CONTINUUM OF LOVE AND VIOLENCE More than half the participants compared the emotionality of anger to that of love, both seen as part of the repertoire of human response. P2 summed up this notion: The situation controls itself sometimes, ?cause of like you know love, it?s inside, it?s emotionally, you know. It?s not something you can buy from the shop ? airtime of love or something, airtime of getting pissed off [laughs]...naturally, you know. Drawing on the metaphor of being ?in love?/?love sick?, the emotions of violence was therefore normalised and accepted as part of the everyday and natural course of life. Love (and synonymously violence) as irrational is made evident by P3 who conflates love with violence: ?I think it was..it was love (.) um (.) I?m not sure whether I?m using the correct term in fact, because love can?t make you do that, but I think it was..it was my loving her?em (4) and to some extent, insecurity?em (4) trying to (.) hold onto her??. In this way, the irrationality of the violent act was emphasised. P5 reconciled the irrationality of violence with ?loving too much? ? ?too much? to a point that it bordered on obsession: ?The thing is that I like her too much, that is why I say maybe obsession sometimes?can do the things that (.) bad like, obsessed, the other thing?. Interpreted in this manner, being consumed with love becomes synonymous with obsession. Both have the potential to drive one to a loss of faculties of reason, a state of ?losing control?. Implicit in these descriptions is the link between violence and perceived abandonment. The ?fear of losing one?s partner? highlighted the vulnerable emotions underlying the act of control. P7 substantiated on this point: ?I don?t want to lose her, so that?s why I gets (2) acting violent to her? and affirms: ?I?ve told myself that ah (2) I?ll, I?ll lose her because of this jealous of mine, because she always tell me that /?/ I don?t want to lose her?. P3, similarly, commented on his reasons for violence: ?Because I don?t want her to have another boyfriend?. Jealousy as a motive for violence is ?not something you can brag about? but must remain hidden from view, as P2 articulated: Jealousy is what you, you know you are jealousy of this women, you know you love this woman. You don?t wanna lose her and stuff?And then you think maybe you know, get pissed off, you start to do other things like beating her up, and stuff and then you know, but it?s not quite good, you know. Jealousy, as an ambivalent emotion, contains both elements of love and hate. However, whereas ?good enough jealousy? is the ability to hold in balance contradictory feelings towards the love object (Hinshelwood, as cited in Yates, 2000), destructive jealousy emanates from the inability to tolerate such emotional ambivalence. The coexistence of 75 love and violence, once again, enters into the equation. According to psychoanalytic thinking, these states of emotional ambivalence, reminiscent of oedipal strivings of both desire for and rejection of the maternal love object, is a central feature of love (Izzard, as cited in Borochowitz & Eisikovits, 2002). As an emotion, romantic love is associated with a dependence and vulnerability toward the love object. As Neu (as cited in Borochowitz & Eisikovits, 2002), however, points out, these emotions that typify the experience of love also threatens the need for autonomy, and subsequently results in the potential for negative emotions of jealousy, disappointment, fear of abandonment, anger and hatred. Attachment theory also alludes to this contradictory state of love and violence, where violence becomes a desperate act, an attack on another who threatens abandonment in an attempt to maintain the affectional bond (Denzin, 1984b; Dutton et al., 1994). Unconsciously motivated by pain and perceptions of abandonment, these typically ?unruly emotions? that are suppressed and manifest alternatively as violent control (Kauffman, as cited in Gadd, 2003, p. 335). Love and violence follows a cyclical pattern which has its basis in early object relations. Violence therefore is the end state of a composite of contradictory and powerful emotions: possessiveness, idealisation, disappointment, anger and rage. In this context, love simultaneously coexists with violence (Borochowitz & Eisikovits, 2002). Jukes (1999), similarly, labels the claim of ?loving too much? as that of idealisation, either of the woman or the relationship. Notably, it is the ?desire to have a perfect relationship with the woman never behaving in a way which threatens the idealisation? (p. 71). Some men drew upon idealised notions of love, reflecting a ?true love? discourse. P2, notably defined the intimate relationship as one ?that is gonna be there for me forever?: In a relationship, it?s like team work, like working together to do, doing things together you know (.) in the sake of love?you know, one bonding thing. You don?t have to like pull that side, that side. You have to pull to one side, one direction one destination. Similarly, P5 evoking the metaphor of a relationship as a marriage covenant, defined honesty and trust as its solid foundation: ?A relationship is based on, eh honesty (2) trusting each other everything, so when you break one of those, it?s like marriage, breaking the vows??. Love, but more particularly loving ?sincerely? was seen as a form of investment, by which a ?return? of some form is expected. The investment, moreover, is a significant emotional one, as Participant recounted: ?it was the relationship of a lifetime??. The expected return is to be ?loved back?, which translates to an expectation of loyalty and fidelity from one?s partner, as P5 reflected: ?When she start doing cheating and all those stuff, I say, I put all my my time, my energy, my thoughts to this person?and this is how she repays me /?/ there were a lot of things you know, that created that anger??. 76 The failure of unrequited love and resultant feelings of anger was further portrayed in the following exchange with P12: I loved the girl, so (.) she didn?t love me as I love her /?/ The girl was like cheating to me, because I wasn?t cheating to the girl and the girl was cheating, so, my heart was sore because I was straight with him. He wasn?t straight to me. Then I used to then, beat her? However, the complete investment of self into another person without accounting for risk results in a tragic outcome. P2 eloquently noted: ?I never left any space for disappointment, you know I never thought that I was gonna have moments like that and stuff, you know. I thought it?s gonna be like true love ? one way and stuff?. Echoing P2, P5?s ?space for disappointment? is an attempt at self-preservation following broken trust: In a relationship, we got all this honesty what what neh. If you start (.) eh to break one of those neh?/?/ it?s like you are editing this person?oh this person is capable of breaking the rules, meanwhile when she does this, she will?it?s like, okay in your?you create some, space for disappointment [laughs]. Shattering the idea of ?true love? or perfect love led to feelings of disappointment, disillusionment which was not anticipated, nor foreseen. However, not only was the relationship idealised, P2 perception of his partner prior to her betrayal also reflected the perfect ideal: ?And as soon as I saw her, I thought of like this is my angel coming you know??. For P10, his partner reflected not only ideal of innocence, but also his emotional dependence through ?trust? and ?protection?: ?Yes, the one you trust, you..you tell yourself, this one, she?s a guardian angel. Before you know it (.) she?s gone from you /?/Ai, it?s unexplainable, the pain goes to the ro..goes to the heart straight??. Dobash and Dobash (1979) similarly noted that men expected their women to live up to a rigid ideal of a perfect woman, ?a plaster statue of a saint? (p. 98) as one man put it. The angelic status ascribed to these women in the present research moreover conveys both the qualities of a ?pre-fallen? state, one of innocence and purity, and at the same time she is ?guardian protector? of his vulnerable state, his feelings that he has invested into her that she is to watch over and keep safe. In their fallen state, following actual or perceived infidelity, women were denigrated to the status of ?prostitute? (P12) or ?bad woman? (P2). On a psychic level, the ?all good, all bad? split, according to Segal (as cited in Gadd, 2003), exposes the ?flip side? of masculine power, in other words, their feelings of powerlessness at their dependence on women, emotionally and sexually. Reflecting on the subjective dimension, Pleck (as cited in Wexler, 1999) ascribes to women two forms of power upon which men are dependent: first, their emotionally 77 expressive power, or capacity for emotional expression; and second, their masculine- validating power. The first may relate to men?s need to seek connection with emotionality through a source other than their own. As Goldner et al (1990) notes, the woman bears the intolerable feminine feelings for both herself and the man. The recognition that these feelings are his own within prompts a violent response. Women are in this sense the ?guardian protectors? of men?s emotional life, in one P10?s words, his ?secrets?: ?you can imagine someone close to you, you tell him your secret, tells you your secret, stuff like that and to get, you get betrayed, I mean (.) no one knows the reaction of that is?? The second relates to men?s need to be reassured of their ?fundamental masculinity and masculine self-worth? (p. 130). Bordo (as cited in Whitehead & Barrett, 2001) notes that the image of woman as sweet, gentle, angelic is necessary to support a masculine one as strong, confident and calm. In this manner, masculinity is dependent on femininity. The partner?s infidelity therefore becomes a rude awakening that propels her ?fall from grace? in his eyes. By losing one?s partner, particularly to another man, he is in a sense made to feel ?less of a man?. Simply put, her assumed infidelity is an act of invalidating his manhood, whilst she acknowledges another?s. What the following excerpts also illustrate is that male domination, to quote Benjamin (1980), is embedded in ?a struggle for recognition between men in which women are mere objects or tokens: the prize? (p. 159). Let?s say I?ve got a girlfriend and then my girlfriend talks to someone that I think is a pimp or a player, so someone that you scared of, ?cause sometimes you tend undermine yourself in a manhood situation, sometime you know, thinking that guy is the one who takes all the girls and stuff and you see your girlfriend talking to the same man?you feel the other heat now. It?s not gonna be the same like before. Like before maybe you used to talk about things ? this time you not gonna be like talking and stuff, you gonna be using force. (P2) I was happy ?cause she made me laugh neh and the thought of someone having her, he?s taking that happiness away from me, so when I?m not happy, obvious that I?ll be angry?Having her was happiness and not having her was?eish, sadness, ja. Worst part, taken by some person, ai, no (.) and the other thing [?] for me it was a long-term relationship, you know. I?ve put more time, it?s like when somebody took her neh, that three years of my life, there?s no meaning for me. (P5) While men were dependent on women for their emotionally expressive and masculine- validating power, they also sought for protection from such dependency. Asserting a male identity required that they repudiate their ?sameness, dependency and closeness with another person?, at the same time, ?attempting to avoid consequent feelings of aloneness? (Benjamin, 1980, p. 150). At the same time, however, P5?s dependency was evident in his ascribing a meaningless existence without her. Threatened with 78 abandonment, and unable to tolerate separation and loneliness, violence was the means of regaining recognition. The tension between achieving independence and recognition is thus apparent (Benjamin 1990). For P10, the greatest vulnerability lies in the emotional investment of oneself into ?the one? you love the most: Because with the other lady, I mean it was no, a good relationship, I trusted her, she trusted me until, what happened, so you see, you are?I can say from my experience is, you are likely to get hurt by the, by the person who you think is the, is the best one?the one that you trust, that is the one that is going to, to put you down. (P10) Similarly, P5 made reference to the link between ?true love?, betrayal and anger: ?Just, you know, when I?m angry, I?m angry?by then, especially by someone that I love (.) sincerely?. The shock of betrayal and the cruel realisation of unrequited ?love? was also reflected by P12: ?Ai, it was not expected. I thought the girl love me (.), so (.)?she used me, so my heart was just, broken /?/ so, my heart was sore because I was straight with [her]. [S]he wasn?t straight to me?. Alluding to the interconnection between love and violence, P10?s earlier statement conveyed the unpredictability of the response to betrayal: ?no one knows the reaction of that is??. P12 echoed this view, emphasising that ?beat[ing] the girl? was a way to ?make my heart right?. However, love as an emotional investment was distinct from love as sex. Whilst violence was a response to having lost or being abandoned, violence was also interpreted as an expression of love. The contradictions between love and violence was further emphasised by P6 statement: ?I beat her up, I beat her up. She loves me. Eh, if I beat her up, I show that I love her? and P8?s remark: ?I also told myself that means those girls, if ever we beat them, we show them that we love them??. P5 affirmed that such a view was supported by women: ?Some girls they think if you beat them neh, you express your love, that you still care, that?s why you beat?. P3?s earlier statement emphasised the contradiction of such a position, which is almost irreconcilable: ?love can?t make you do that [beat someone]?. P7 highlighted the incongruent state of affairs further by introducing the association between love, sex and violence: ?Because I understand that, you cannot beat the person and after go sleep with her, go make love with her?. Love, violence and sexuality were intertwined in men?s narratives. Man?s sexuality as a fulfilment of masculinity (Connell, 2005) manifests as control over women?s sexuality (Ptacek, 1999; Seidler, 1994; Wood & Jewkes, 1997). This view was affirmed by P1: ??if you are a man, you know the man, your manhood will be determined by how..how often do you have sex. If you don?t have sex often, then you are not man enough??. By extension, masculine self-worth was dependent on women?s validation through sex. Sex was equated with love, as P1 revealed: ?We still have a strong belief that a love will only 79 be determined by sex (.), not by anything. They say we are just measuring love and sex??. Wood and Jewkes? (1997, p. 42; 2001, p.134) similarly pointed to this equation: sex is the ?purpose? of love. For the male township youth in her study, ?to love? meant that their partners would be expected to have penetrative sex, often in exchange for money, material goods or food. As P1 noted, this act of initiation sealed the relationship and confirmed their status as ?complete lovers?: So we still tell ourselves that even now that..that mentality still existing that if you are in love with a lady and you didn?t have sex with that lady then (.) you are not yet complete lovers?They say you must have sex with her, they say then she?s ?registered?. That?s why after having sex, okay you?re register okay, she?s mine?? Withholding sex was therefore interpreted as a withdrawal of love and, to paraphrase Benjamin?s (1990), the failure of recognition (i.e. powerlessness associated with the realisation that his actions do not influence her). Through sex, men sought for an intimate emotional connection with their partners that was disguised as masculine virility. The need for love was clothed as ?sex?, as P6 revealed: Ursula: So how did you feel when she said ?no?? P6: I felt like I?m alone, I don?t have a girlfriend Ursula: Okay (2) so you felt that she didn?t really? P6: love me The failure to obtain recognition, for these men, resulted in forcing recognition. Through coercive sex, manhood is reinstated. 4.5 THEME 4: VIOLENCE VERSUS EMOTIONALITY A masculine identity involves a disconnection between emotions and personal circumstances. Stated differently, emotions are denied, hidden or repressed in response to daily stress and relationships (Seidler, 1994; Umberson et al., 2003; Walton, Coyle & Lyons, 2004; Whitehead, 2001). Violence is therefore conceptualised by men as ?an expression of extreme and cumulative emotional upset? (p. 244). Simply put, they lose control of their emotions as these overtake them. Similarly, batterers? depictions of violence in Anderson & Umberson?s study (2001) made appeal to the act as a natural masculine reaction in response to provocation or frustration. In the present research, men?s descriptions of violence cohered around two interrelated yet distinct sub-themes: (i) violence in response to another?s emotionality and (ii) violence as a defence against own emotionality 4.5.1 Violence in response to another?s emotionality For some participants, violence was a reaction to their partner?s emotionality. P4?s violence, for instance, was ?driven? by his partner?s ?irrationality?: ?All the time when I 80 wanted to (.) to resolve the problem, she come?she become emotional (.) so that emotional push me to be emotional??. Being ?emotional? in P4?s sense was being violent. P1 similarly remarked: ?In most cases women are so emotional. I mean, they like to scream, yell at you, but so sometimes I get upset. I?m a human being, get upset end up having responding violently?? Implicitly gendered in this explanation, women?s ?emotionality? is contrasted with men?s ability to reason. Such a contrast was also manifest in participants? responses in Anderson & Umberson?s study (2001, p. 374), where men highlighted their strength, power and rationality, whilst emphasising their women?s ?irrationality? and vulnerability. In the present research, failing the success of reasoning, violence was resorted to. P9 recalled a childhood encounter with his mother, one which evoked in him feelings of anger and frustration as a result of being forbidden to go to the movies. In this excerpt, a response of violence (or anger) was connected to women?s ?irrationality? or failure to rationalise: I?m the kind of person that wants a logical answer. I?ve always been that way ever since, being a lightie. If there?s a reason for something and it?s logical?alright, I can understand (2) but she [mother] wouldn?t argue rationally ? uh she was an intelligent woman, make no mistake she was ? but when it came to protecting us, particularly me, to..to preventing us from (.) bad behaviour [laughs] she became unreasonable and irrational herself and that?s?I responded in kind. I responded with anger and frustration. I remember grabbing her purse and (.) dashing out the front door and leaping over the front gate?with her purse in my hand, with the money for the [movies] The intolerance for women?s irrationality or emotionality in this instance, however, also begs a psychoanalytic interpretation. The ?voice of unreason?, as P9 noted, was a trigger for his violent outbursts with his wife, a prompt that would spark off childhood memories of being restrained by an ?overcontrolling? mother who would get ?panicky and upset? in response to his independence. This was made apparent in the following exchange: P9: ?although, she [wife] didn?t remind me of my mother. She didn?t speak like my mother. There?s was one, however?there were times when she did speak like her. It?s sort of a whiney voice?and that was identified as one of the triggers (.), this kind of worried voice, ?oh., why you doing this, why you doing that, why can?t you do this, why can?t you do that???whew! Used to (.), used to trigger off, um, an explosion Ursula: What would happen when you heard that, that voice? P9: I?d get very angry (.). Here we go again, type of thing. Gosh, why can?t we speak normally, why can?t we speak quietly, why do we always have to (.), why does that voice have to come in? That?s the voice of unreason?That?s the voice of (.) a person whose going to restrain, or restrict or forbid (.) you know (.) 81 Underlying the temper outbursts were feelings of ?frustration?, ?anger? and ?fear of belittlement? and ?fear of not being listened to?: ?but that?s how she [mother] expressed it on occasions and it led to huge frustrations with her (.) outbursts of temper with her?. The image of the ?overpowering? mother is contrasted with the ?little boy? who seeks to free himself from her grips. According to Chasseguet-Smirgel (as cited in Muller, 1999), femininity is perceived as an overpoweringly ambivalent experience ? one which ensures nurturance, at the same time, one which engenders feelings of dependency and powerlessness, the basis of narcissistic injury. The need for both nurturance and independence evolves into a masculine adult pattern of relating to an intimate other. As Gadd (2000) notes, these emotionally vulnerable experiences become rooted in the formation of adult masculinity. Internalised during the early years, P9 learned to ?split and project, onto his closest intimate, the uncomfortable emotions he [could not] quite repress? (p. 444). In the words of P9, violence was the natural outcome: And it worked. In the case of my parents, it worked. They were so terrified of this, it actually worked. They were afraid to realise it would happen again?Life got better (.) so the lesson is obvious. Shout and bawl a bit, throw things around? Violence was the ?defended? reaction against the potential resurfacing of vulnerable feelings, but it also developed into a habitual response to frustrations stemming from perceptions of being attacked (Gadd, 2000). In this latter sense, violence served to bolster his self-esteem and sense of reality, in Winnicottian terms, it was about ?feeling oneself? (Hagemann-White & Micus, 1999, p. 9). Men?s violence as an expression of intolerance for women?s emotionality is reminiscent of Chodorow?s (1978; 1989; 1999) notion that for boys, masculinity is learned via the repudiation of femininity. In other words, learning to be ?not-feminine? means to reject, perhaps by violent means, that which reminds him of his own dependency and vulnerability. The projection of empathy and emotions onto the feminine and the denial of dependence in himself is thus connected with the devaluation of womanly qualities (Hagemann-White & Micus, 1999). Somewhat aligned with this view, P2 highlighted the contrast between women?s expressive power and men?s propensity for physical expression: You know a man doesn?t have much of vocab ?cause she?s got vocab ? she?s a woman and then me, I?m like more in a physical way?you know so that?s why I became a man and hit her, but you know I wouldn?t say I became a man?I became you know a weak man. Drawing parallels to the irrationality of emotions, men?s violence is thus constructed as irrational and unpredictable, and is thus ascribed an uncontrollable quality. However, 82 such a view contradicts the widely subscribed patriarchal discourse on male rationality, which holds that men are agents of reason and are guided by a logical thinking style, whereas women are ruled by their emotions (Connell, 2005; Seidler, 1994; Whitehead & Barrett, 2001). Aggression and violence then becomes the medium by which emotions are to be expressed by men. In this sense, men?s violence itself takes the form of a masculine ?emotion?. Violence as an ?emotion? therefore has the element of both controllability and uncontrollability, rationality and irrationality. As a strategy, it serves to deny, denounce or repress vulnerable or ?feminine? feelings that threaten the masculine ideal of toughness and rationality (Chasseuguet-Smirgel, as cited in Muller, 1999). 4.5.2 Violence as a defence against own emotionality Intricately linked to the sub-theme, violence as a response to another?s emotionality, violence as a defence against own emotionality highlights another paradox of violence; that is, ?being a man? is expressed in violence in as much as is ?being a weak man?. P10, for instance, noted that violence may be the outcome in situations where he is made to feel less manly: ?I?ll say, in general, where a man feel less uh..or feel like he, he?s violated, I?m not sure, I?m not saying when they are violated they should act in this way, but, violence is bound to happen?? Beyond the level of discourse, what is revealed is a defensive ?split? between the ?real man? and the ?weak man?. Paradoxically, there is a disjuncture between the ?real man?, who asserts his independence needs and denies emotional vulnerability, and the ?weak man?, who (unconsciously) experiences powerless at his own vulnerability. The crux, therefore, ?is not [emphasis added] that the violent man ?is? powerless but that he cannot tolerate this normal and necessary part of being human? (Hagemann-White & Micus, 1999, p5). As reiterated by Goldner et al. (1990), the conflictual struggle between feelings of manliness and feeling macho often precipitate a violent escalation. Paradoxically, the same participant iterated that displaying violence toward a woman as an unmanly act: It?s not because eh that..that..that..that?s a choice, no, no one likes to be fighting ladies, no, no, I mean, it also puts you down as a man. You know she?s a lady, she won?t turn against you, but you are fighting her. Beating a woman is construed as a manly act in the same way that refraining from beating a woman is. P6, for instance, who witnessed his father?s beating of his mother as a child, explained the cultural reinforcement of the notion that ?being a man? necessitated physically dominating a woman: ?I was enjoying like beating up a woman. It?s my culture, my father told me that if you don?t beat a woman up, ah, you?re not a man?. Following his attendance of workshops on violence, P6 relayed his unsuccessful attempt to change his 83 father?s erroneous ways. Reconciling with him after a period of separation, he recounted his father?s attempts at making reparation for past mistakes by redefining manliness: ??Son you are a man??. So after my workshop, I started to workshop my father?didn?t understand me, but decided no, okay ?let me leave this, okay I?m going to get my place now, no longer staying with you ?cause you know too much? ? ?Okay, dad go?. But now he?s coming and tell me that um, ?Son, (.) a man, you good. As Chodorow (as cited in Goldner et al., 1990) notes, men?s need for intimate connection with the father is repressed due to cultural prohibitions. To attain worthiness for paternal love involves a concealment of or disowning of his vulnerabilities. Failing the possibility of an emotionally bonding relationship, the boy child, instead of ?being with? the father resorts to ?being like? him. As illustrated with reference to P6, by following in his father?s act of brutality towards his partner, he symbolically bonds with him: ?Ja, then my father used to beat up my mother, then I ?cause like aah my father is a man??. By being ?a man?, he repudiates his need for intimacy and closeness (Benjamin, 1990). Whilst ?being a man? reflected the patriarchal discourse of man ?having (rational) control? over his actions and those of others, ?being a weak man? reflected the emotional aspect of violence, the sense of irrationality and unreasonableness that is often ascribed to women (i.e. ?losing it?). The instrumental/expressive dichotomy (reflecting ?reasonable? man and ?emotional? woman) (Connell, 2005) is therefore evident in men?s experience of violence (see also Stamp & Sabourin, 1995). Being a man, moreover, was also about having charge over one?s own life, as articulated by P5: ?it?s just that I am in control of my life?. By extension, this also related to having control over his relationships and his vulnerable emotions. The ?ideal? of masculinity, as reflected by Connell (1995) is thus ?control? ? control over themselves, their partners, their domestic life and their broader environment. Taking charge of one?s life refers, not only to having financial stability, but also having the emotional strength to cope with life stress. In P5?s case, this was reflected in having to solve difficulties without feeling dependent on others: She [partner] asked me ?are you under a lot of stress??. I said ?no, I don?t have stress? (or else I know I was having stress). It?s like somehow neh, become, I want to, the thing that you cannot do on your own, but sometimes I push myself that this is my problem. I want to solve it on my own? In another instance, he remarked: I don?t like, ah seeking advice and help to friends is good, somehow, I avoid that, because I said ?ay, I won?t be having friends. How am I going to leave my situation?. I do have to come to a situation where I want to, can I do this on my own. So, you know 84 when you involved, the person comes neh, I don?t want to give the other person stress. So that?s why I distance myself from relationships, because it?s not fair. ?Control? becomes the means to disguise vulnerability, and distancing of self from others is a defence against the fear of rejection. As Seidler (1994, p. 138) substantiates: ?As boys we often learn to choose isolation out of a fear that we will be rejected if we reach out to others. We often learn to prefer to be on our own than to take the risk of making ourselves vulnerable to others?. The ability to manage aspects of one?s life was a reflection of manliness, such that failing to feel ?in control? produced feelings of frustration. Frustration for P9 was the disguise for fear: I like control. I like to have everything under my control. I like to have...I like to know what?s going on. When I don?t know what?s going on, I begin to feel a little restive, a little unhappy (.) um whether I?m running a company, or running my home ? whatever it is ? travelling, I like to know (.) what?s what (.) and feel that I?m in charge (.). If that situation changes, for whatever reason, I begin to feel (.) fearsome. It is fear. Similar to cultural injunctions of masculinity transmitted from father to son as with P6, being seen by peers as being ?in control? of his relationships was also a reflection of being a ?real man?. P3 remarked: ?My peers, it was important that they must see me (.) as a person who has control over (.) my..my..my love affair?. Not being able to ascertain the whereabouts of his girlfriend or having chosen the wrong ?uregte? (the ?main? girlfriend out of a series of girlfriends) would make him ?the joke? amongst his friends: ?Um (. ) so when they make you a joke and (.) come with the suggestions of how to (.) run your relationships (.) eh it will hurt you (.) because you know whom you love??. P6 reiterated this notion: ?My friend taught me that. If you beat your girlfriend, you show kind of love?. As reflected in the masculinities literature, domestic violence is a display of a masculine identity (Anderson & Umberson, 2001; Umberson et al., 2003). P1 distinguished between a ?real man? and the ?sissy?. Whereas the ?real man? uses brute force to control a woman: Some of my friends say, ?okay, this woman is treating you like a stupid. She?ll eat your money?and if you didn?t have sex with a woman, she is not yours, so you must forced her to make sex with you. the ?sissy? was potentially victimised by his peers: Those guys were romantic, they become even victims themselves?and say, ?you are sissy, I mean you are trying to be nice to the ladies, you are spoiling the ladies?, that?s what we?that?s the language was commonly used? On the psychic dimension, this is revealed as a need to safeguard one?s masculinity by denouncing any emotional identification with women (i.e. being a ?sissy? or the ?romantic?). In feminist psychoanalytic terms, any remote identification with the 85 feminine represents a return to the primitive state of empathic attunement once shared with the maternal figure, a state which threatens their independent selfness (Chodorow, 1978). Affirmations of a masculine selfhood are sought for through peer approval of behaviours that are distinctly ?not-feminine? (Hagemann-White & Micus, 1999). Some men also reflected on another code of masculinity: that emotions are evidence of masculine weakness. In P8?s case, resilience in suffering (i.e. having ?control? over emotional pain) was seen as the mark of a man: ?He [a friend] taught me ?no man this..this kind of things sometimes happen (.), you know, so be strong, be a man. Then he just give me courage eh most of the time, you know?. In another instance, P5 explained that survival during apartheid was predicated on disguising one?s vulnerable emotions: ?If you like show your enemy that you can cry, they will think you are weak as a man?. So even though it hurts neh, [inaudible] but like you don?t show your tears?only cry inside, it?s like a sheep, you see?. P8 also made reference to the agony of hidden pain and the necessity of maintaining ?emotional stoicism? (Umberson et al., 2003, p. 234): ?I don?t know why they [tears] didn?t come out, you know ?cause of eh most of the time I..I don?t usually cry you know (.) I..I..I don?t usually cry but I cry inside?. Showing emotional pain was seen as an indication of a ?weak man?. Revealing it invited humiliation, while at the same time, hiding it was a form of silent suffering that one bears privately, in the company of oneself: You are hurting inside, and nobody can see you. I?m not saying, I won?t say physical is better - I won?t compare - violence is violence, but physically, as soon as I, maybe hit you, maybe your skin is sensitive, the person will say, ?what happened???even if you hide, but people want to know, but eh emotionally, it hurts. You are suffering alone?it cannot disclose to anyone. So (.) emotionally is the one that I?m scared of? However, efforts to avoid or repress vulnerable feelings, whether related to relationship or job stress, nevertheless continue to evoke anxiety despite emotional suppression (Folkman, as cited in Umberson et al., 2003). As highlighted by Umberson et al. (2003), other than on the emotional level, anxiety may manifest in somatic or behavioural forms. P8, revealing the physical discomfort of disguising his emotions, attested to this notion: It?s hurting, sure, it?s hurting (.), find that eh (.) sometimes find that maybe one of that you know just comes out, just do it fast. You find that you just wipe it fast, you..ou know when you crying inside, eish it?s hurting ?cause find that eh..eh..eh..eh you can even lose weight. 86 As the excerpts reveal, masculinity, defined as ?not feminine? in a psychoanalytic sense, fears the emotional. All aspects of the self experienced as remotely feminine (weakness, fear, dependency, need for fusion, powerlessness and passivity) are projected onto women, the receptacle for his intolerable feelings (conscious or unconscious) that is kept at bay. Experiences of emotional vulnerability, perceived as a threat to the masculine identity are either devalued or repudiated (Muller, 1999). To preserve both the dignity of man whilst releasing emotional suffering, violence becomes the compromise that is made. The emotional pain that invites fear, as P5 articulated, is thus disguised in violent action. As an ?emotion?, it is expressed and translated into a form that is perceived as more acceptable than showing one?s tears. P5 recounted how violence was a means to relinquish his pain and reinstate control: It?s like (.) you know when you?re in pain, you don?t have like (.) to talk, what you have to do is to cry only, that?s what I had in mind, meaning she?s still talking, meaning (.), it means she can fight me back, so let me show her that (.) I?m in control of this relationship. P9 identified fear as the root of his anger/violence: ?The anger (.) that I uh expressed was rooted in fear /?/ fear of being belittled, fear of not being listened to (.), fear of (.) being interrupted in a conversation?. Violence, he asserted, was a state of ?being out of control? in order to regain control: It?s trying to regain control and using the wrong way of doing it, and realising that you?ve taken the wrong path (.) and feeling desperate, and more angry still, with yourself (.) as well as depressing (.) and the struggle to get back under control ? to calm down, to shut up. Men?s use of violence as a defence against vulnerabilities suggests an emotional experience that cannot merely be reduced to some ?monolithic masculine or patriarchal quality? but, as men?s narratives have revealed, suggest underlying conflictual and anxious emotions that are not easily expressed (Pattman, Rob, Frosh & Phoenix, as cited in Gadd, 2003). To preserve the masculine ideal is to assume a defensive position, one which overemphasises rationality, boundaries and a selfhood that is distinct from the feminine (Chodorow, 1978). Weakness is split off and projected violently onto another (Gadd, 2003). Whilst splitting was a defensive reaction characterised by disowning vulnerability, a ?split? or ?dual? self also represented a dissociation between the present ?talking self? and the ?violent self? of the past (Hearn, 1998). 4.6 THEME 5: THE VIOLENT SELF AS ?NOT ME? 4.6.1 ?I?m a changed man? Seven men asserted making the transition from ?violent man? to ?changed man?. Somewhat different to the ?rational self? versus ?irrational self? (as detailed in the previous 87 two themes), the split between the former self and the ?new self? in this instance reflected a disjuncture between ?the man I used to be? and ?the man I am now?. Gadd (2006) points out, however, that understanding the shifts in subjectivity from past to present cannot be reduced to ?a series of socially determined transitions? (p. 191). Consequently, these shifts in positioning are also considered in relation to the psychic dimension. The split between ?present self? and ?past self? was explicit in some instances, whilst in others change was alluded to in more subtle ways. P2 and P7 made specific reference to no longer being ?that person?. For P2, the ?past self? was a ?young boy? who reasoned with violence when talking lost its effectiveness: I told her ok, let?s move away here. I?m going home, leaving you here (.) so give me a company [accompany me] (.) ?cause I?m gonna stop hitting you and stuff now. Let?s stop fights, and then let?s talk?so we started talking. She was sorry, telling me how much she likes me ? she loves me - very much. She didn?t wanna do this and stuff, you know but already I was pissed off you know. I was still a young boy, you know. The ?changed man?, in contrast, was being ?a responsible man for my future? one with greater maturity and insight into relationships, as well as the ability to find alternative and healthier means of self-expression, such as through ?hip hop? and ?poetry?. Reflecting on the past self, he asserted: ?That?s why I?m saying like now, it?s a changed thing. It?s not gonna happen again, yeah, ?cause now there?s a better understanding on things and stuff?. Concurring with P2, P5 alluded to having matured from a ?small boy? to having ?grown up?. The present self is portrayed as a positive role model for others, one who has ?searched [his] inner self? and has ?survived? hardship: I got a baby now okay, and maybe along the way (.) her mom hears that [I am] still doing this and that. I?m not doing any favour, I don?t owe anybody?but the thing is that neh, if I was continue, leading a violent life, what are they going to tell my boy when he start to grow up? What kind of a father is he going to have? Analogously, P1 remarked: ?I must take decision that will not put my life in danger, that will not harm even young upcoming stars who are looking up to me as a role model?. Drawing on dual metaphors that contrasted the former from the present self, P9 also made reference to being taunted by his wife for being the ?little boy? who acted out when confronted with frustration: ?You got your sulky face on?, she?d [his wife] say [laugh]. I used to be furious with that /?/ I?m aware of what I used to do, and it?s quite funny, when you think about it, like a little boy? looking back at myself, I could see there was a little boy there ? a little boy who was?hew!...that?s the other thing, she used to [laughs] she said: ?you?re like a little boy in a short pants?. Oh, god! That wa..I would go ballistic ? (.) ?little boy in short pants?, you see there it was, it was, it was harking back to the childhood (.) 88 On the psychic level, being reminded of one?s ?smallness? (the ?little boy?) presented a threat to his masculine identity which, although he fended off by resorting to violence, became integrated into his sense of self. No longer engaging in this defensive splitting, the healed self achieves the ?true differentiation? ? not the striving for absolute independence and separation ? but the integration of ambivalent and contradictory feelings. The healed self is thus one which experiences ?the most intense sense of selfhood? as it negotiates the ?tension between negation and recognition, affirming singularity and connectedness, continuity and discontinuity at once? (Benjamin, 1980, p. 161). The contrast between present and past constructions of self is captured in P9?s labelling of his current identity as ?the ex-ogre?. In this image, he juxtaposed his children?s perceptions of him, both past and present (i.e. ogre/ex-ogre). The present self, conceived in their eyes, is that of a ?mentor?, conveying a sense of intimacy and trust which contrasts starkly with distance and fear. On a subjective level, the boundaries between the ?recognised? (the ?ogre? who forces his will upon others) and the ?nurturer? (the ?mentor? who recognises the feelings of others) are dissolved in the present self. The self of the past, moreover, is also portrayed as being the ?sick? or ?aberrant? self. This is reflected in two separate instances: ?Sick (.). Sick (.). Disgusted with myself, you know? and ?I thought no, this is an aberration?. Juxtaposing this against the current ?healed self?, P9 emphasised that ?the beginning of recovery? takes place only once ?I had admitted that it was my problem?. Parallel to this notion of the ?healed self?, P1 describes healing as synonymous with the ?purging? of oneself through talking or crying: Talking on itself is a healing process, even if you are in trouble, in order for you to feel very healed or relieved, it?s either you talk or you cry?so talking, it?s a healing on its own (.), it?s a healing on its own. So I?ve been doing this talking ever since?so right now I?m someone whose purified, whose healed (.) so?I don?t have any problem. I?m eh just a sweet and romantic guy (.). That?s how I can describe myself now? The uninhibited expression of emotional pain becomes the balm that heals, as the feminine aspects of himself are no longer split off and kept at a distance but are integrated into the present self. The ?sweet and romantic guy?, evoking traditionally feminine qualities of softness and sensitivity, is distinguished sharply from the former self, the ?beast? whose appearance instilled fear in women whom he had sexually violated: ?she must just be scared of me, she must fear me (.), she must fear me??. P6 similarly referred to his former self as one whose violence commanded others? respect through fear: 89 People must be afraid of me, this guy (.) is Mike Tyson /?/ Only sixteen years old, Mike Tyson, the young Mike Tyson /?/ It was good for me. Didn?t understand that ? another thing, like I can be behind bars, people will hate me. I didn?t worry by the way ? I was Mike Tyson. His present self, however, wavers between ?change in process? and complete change. The talk of ?doing good? represents a movement ?from the behaviour of non-offending to the assumption of a role or identity of non-offender? (Maurna & Farrall, as cited in Gadd, 2006, p. 180): ?Oh (.) actually when I started to meet that girl [?] I was abusive to her. Then after eh two years time, so I kind of changed now?I?m no longer beating her up, I?m a changed man now?. In more subtle ways, other men made reference to ?being a changed man?. P8 implied that his current self is one who is not only intolerant of woman abuse, but in witnessing such acts of violence, vicariously experiences women?s pain: Eish, you know I feel terrible actually (.). I feel (.) eish?you know, even right now, you know (2) I..I won?t..I won?t stand it, if I find that even, you know a guy outside is hitting a girl, you know (.) ?cause I know from my side that that is not good /?/ I also feel the pain, but sometimes I take a chance and I go there and I make a peace. The acknowledgement of pain, one?s own and another?s, perhaps represents an incisive step towards mutual recognition, recognising another as well as being recognised by another, both existing as ?two living subjects? (Benjamin, 1990, p. 16). In the two following sub-themes, men expressed remorse for the pain they inflicted onto their partners. 4.6.2 Expressing remorse: ?I feel bad for what I did? Six men expressed remorse for their violence. Remorse is defined as the acknowledgement of both the violence and its wrongfulness (Wood, 2004). Frequent references to feeling ?bad? was expressed in various forms. P6 expressed regret for the past and attempted to make a changeover in the present upon acquiring the skills to deal with his violence: Now I regret a lot (.) because I was doing things (2), what can I say, doing things in my way, not respecting her way?But now things are better. I?m workshopped now, Life is good for me, for her now. Nobody?s crying? For P1, his realisation of the pain and ?horror? he instilled in several women brought about feelings of shame and self-disgust: ?I felt sorry for myself and shame?. His narrative revealed a capacity to role-take or ?mentalise? (Fonagy, 1999). Role taking is defined as the ability to perceive another?s emotions (i.e. another?s physical or emotional pain), in short, the capacity for empathy (Goodrum et al., 2001). In the following excerpt, he admitted to being haunted by flashbacks of his past: 90 I feel terr..terrible you know?I feel very, very terrible. Actually there was a time whereby I will sleep and recall all these incidents ? ooh! this is what I was doing?especially when I was see victims of rape?people who have been raped, the way they, they respond, I say, ?mm!?, some of them they are so horrified? In the violent self, the fear of being perceived as weak, unmanly or ?sissy? supersedes any empathic identification with another (Gadd, 2003). For P5, in order to exercise violence, he needed to overcome any potential feelings of empathy he had for his partner. In his words: ?I didn?t think about her woman?s feelings?. The possibility of recognition of another, according to Gadd (2006), is achieved in the ability to hold in tension the different forms of identification (i.e. masculine and feminine, perpetrator and victim). His present self is portrayed as one who is able to recognise his own vulnerability in another and reconnect with those parts of himself that he previously disowned by projecting it onto his partner (Gadd, 2006): Eh, it?s like, there?s this thing that I have now. I think now, before I didn?t have it. It?s like o..like now I what avoids me to do these things, neh, when I think of hurting somebody, I think of me if somebody can do this to me, how am I going to think? By then, I didn?t have that. P8?s severe beating of his pregnant girlfriend (unbeknownst to him at the time) resulted in her miscarriage. Social withdrawal, oversleeping and drinking were ways in which he dealt with his emotional distress: ?No, you know I was not, I was not nice, to myself, you know (.) eh..eh..eh..eh ?cause you know when I?m alone I thought of death, you know of what..what..what happened then??. He recounted being laden with guilt and self-blame over the incident. As the following excerpt reveals, much of his remorse, however, was directed onto his unborn child who, to some extent, is seen as an extension of himself: No, I feel like crying ?cause of what I did was not good. It was a bad thing (.), you know. But I was angry when I was, when I did that (.), so well I came to my senses, you know?so whereby I find that I see but no (.) eh..eh..eh, if ever I beat a pregnant woman, I?m also, I?m also hitting a baby inside?you know, so I thought no, the..he..he baby so doesn?t know anything, did..did not do anything [?] So that?s where I..I..I felt the pain, you know and then I also felt that you know it?s also my baby she..he, he or she is our baby?you know, I?m the father of that baby? P5 offered two contrasting accounts of remorse. The first, located in the distant past, revealed an embittered and resentful self that allowed pride to become an obstacle to feeling remorseful: And the other thing is that this pride neh, I?m not going to say that I?m sorry to her, you know?I won?t feel pity for her /?/It?s eh (.) pride, pride of not saying ?I?m sorry??. In contrast, a positive and affirming ?present self? was also revealed ? the ?healed? self whose ?bad? feelings were outweighed by what he perceived as his partner?s mature 91 willingness to forgive him, and in response his offering of forgiveness to her. In the excerpt below, he recounted an exchange he had with her following their separation: Em (.) I can say, I feel bad, but the good part of it neh, it?s like after some long time neh, I was out of the relationship we did met with this girl [his former partner], it?s like we were reminiscing with the friends. You know, some?all the girls say eh, her boyfriend is abusing him. So she started to look at me, ?eh once I was involved with [P5] in a relationship, I didn?t do such things because he used to beat me up?, but I?m happy ?cause we are good friends. He has grown up?so that kept me (.) she showed me that she had forgiven me (.) and I try to explain that, I told her that I was upset, what, what, what, what, what, but now I?ve overcome it?and I wish her well, so (.) I do feel guilty somehow, but since, I did get a chance, to be forgiven? His reparative attempt is signalled by his former partner?s acceptance of both his good and bad qualities. Rather than positioning herself as the object or ?victim?, she becomes the ?autonomous and knowledgeable subject? who has the capacity to grant him recognition (Gadd, 2006, p. 193). In this instance, P5 re-established contact with the ?lost love object?. Her ability to acknowledge him for his present identity, namely the ?grown man? who is esteemed as a ?good friend? rather than the ?small boy? he once was, paves the way for his transition towards change and allows him to move beyond the ?upset? and the ?guilt? (Gadd, 2006). These illustrations reflect varying levels of ?feeling bad? for their violence. With two men specifically, identification with their partners was the vehicle for ?bridging difference without denying or abrogating it? (Benjamin, as cited in Gadd, 2006, p. 182). However, ?immediate remorse? alone does not necessarily constitute a true confession. The confession, according to Hearn (1998), is a statement reflecting accuracy, revelation or completeness, which may or may not contain elements of remorse. The cycle of violence theory suggests that once ?remorse? sets in, the honeymoon phase begins, and a perpetuating process of violence is reignited (Hearn, 1995, Walker, 1979). Five men reflected on the recurring pattern of violence, repeated remorse and subsequent violence. P7?s apologies for his violence ?the morning after? illustrated a return to rationality and the gradual insight and realisation of wrongdoing: We argue, argue, I end up beating her and really beat her with the bottle, so ei, I stop there and then, I was, we slept together, have sex and then in the morning I start up apologising, and I saw that what I have done, it?s a rape? The features of violence followed by remorse finds parallel in P5?s assertion: ?The only problem, it happens maybe (.) three times in those, three years. So it was I did beat her badly and then (.) the following day (.) I become like (.) feel remorse, you know?. P6 in a 92 similar manner confessed to a typical pattern of events surrounding the violent incident: violence, followed by regret and apologies. Like P7 and P5, insight into situation is gained in the quiet moment of reflection: When I?m alone, I regret it, you know. No, this is wrong, but I..I used to do it everyday (.) I regret it today, but tomorrow I make it until I was workshopped okay [joined a men?s programme], and then I understand it better now. Okay, so it would happen where in the situation, it felt okay, but afterwards it was wrong. Shocked at the reoccurrence of his violence toward his new wife just few weeks into their marriage, P9 expressed bewilderment as he consciously attempted to change his behaviour through counselling: Why does this happen? (.). Why does it happen with someone I respect and love? I didn?t respect the previous one (.). Why does it happen? What on..what on earth is going on? And we looked at the..the immediate, the usual suspects as (.) the movie said uh?my parents were not violent, not violent at all? Following a period of critical self-reflection through counselling, P9 was able to admit both responsibility and blame for his violence. He explained the characteristic response of remorse following each outburst of violence. ?Bitter disappoint[ment]? would follow each episode of violence and became part of the typical ?cycle of violence?, rather a genuine and sincere attempt at seeking reparation: ?After each one [violent incident], remorse would set in, either (.) ten minutes later, ten hours later, ten days later, remorse would always set in, usually very quickly and the uh determination not to do it again??. Whilst blame was initially directed at his wife for ?triggering off? his ?explosion[s] of temper?, healing was precipitated by when he identified himself as the source of the problem: Initially I would blame her ? initially. It must have taken me I guess about a year before I accepted that it was my problem, my fault?because without that, there?s no hope (.). And eventually after, I guess about a year, after the first incident. I had admitted that it was my problem and no matter what she said or did, it didn?t make excuses for me?and that was the beginning of the (.) recovery. ?Cause once you, once you reach that point, you can do something about it. If you don?t reach that point, you can forget it (.). You?ll go on? In comparison, P5 was critical of his failed attempts at changing and highlighted the conflict between his behaviour and the ideals of the men?s programme in which he was actively involved. Change was conceptualised as a rugged process rather than a smooth transition: I was worried because ei (.), I?ve joined an organ..an organisation /?/ It?s talking about men against women violence. I said ?what am I learning there because I?m still practising 93 that?so I told myself change doesn?t happen overnight. You come across lot of obstacles, so (.) a month ended. I went to her. She welcome me. I went and apologise, I say it?s what I wanted, I don?t want a relationship back, I just want to apologise. She said ?it?s fine with me, we can just be good friends ? even now we are still good friends. P5?s articulation of change and reconciliation of broken relationships reflected a movement of past to present as he gradually renounced his identity as ?abuser?: ?I was worried??; ?I?m still practising??; ??even now we are still good friends?. The content themes addressed throughout this section similarly reflected this movement of articulating ?what happened then? to ?who I am now? as a rehabilitated person. The next section will explore a similar movement, but will focus on the form, to use Hearn?s (1998) phrase, men?s talk of violence. By moving beyond the emotional content of experiences, men?s motives for violence are illuminated, in Goldner?s (1999, p. 330) words, ?the element of choice that hides behind the experience of being overtaken by an impulse?. Focus is then directed to the function of narrative as it applies not only to the research space, but also the broader societal arena in which both researcher and participants engage in. Part II: Establishing Credibility in Accounts of Violence In addition to content, the interview transcripts were analysed for their narrative structure. In the analysis of process, as opposed to content, Scott and Lyman (as cited in Jukes, 1999) proposed the notion of ?accounts?. Mills (as cited in Cavanagh et al., 2001) initially pointed out that individuals who commit acts that are socially deviant or morally offensive attempt to present them in ways that are ?both culturally appropriate and acceptable? (p. 700). To quote Cavanagh et al. (2001), ?accounting for an act involves seeking to absolve oneself from responsibility for the act, attempting to mitigate culpability by, for example, apportioning blame elsewhere, minimising the harmful effects of the act, and/or denying the act outright? (p. 700). Thus, the account becomes a way in which talking about unacceptable behaviour is managed and conveyed in a manner that is less implicating of its actor. This notion is aptly articulated by Bruner?s (as cited in Stamp & Sabourin, 1995, p. 287) statement: When you encounter an exception to the ordinary, and ask somebody what is happening, the person you ask will virtually always tell a story that contains reasons? [emphasis in original]. The story, moreover, will almost invariably be an account of a possible world in which the encountered exception is somehow made to make sense or to have ?meaning? (p. 49). 94 In this sense, the account is not only a means to manage impressions. As a narrative, it is also a means by which the storyteller constructs meaning in his/her story (Freedman & Combs, 1996). Significantly, meaning that is derived during storytelling (or the interview) is a process that takes place between the speaker and the listener. Meaning is thus co- constructed in the ?joint enterprise between interviewer and interviewee? (Elliot, 2005; Hollway & Jefferson, 2000; Hyden, 1994, p. 99; Riessman, 1993). In this light, issues relating to the interviewer?s self-presentation are thus pertinent and intrinsic to an analysis of accounts (Fontana & Frey, as cited in Boonzaier, 2001). The idea that meaning is derived in process reflects the social constructionist stance that reality is derived and constructed in talk/interaction (Burr, 1995). Adopting Riessman?s (as cited in Boonzaier, 2001) notion of the ?meta-story?, the analysis transitions from considering the content of men?s personal narrative to the process (how they tell their stories) and the broader narratives that exist on a societal level (Elliot, 2005). Taking these considerations into account, this section will engage in an analysis of men?s accounts of violence towards their intimate partners. The aim is not to present a ?narrative analysis? of accounts per se. Rather, emphasis is drawn to the form of accounts, specifically, to the intentions of the speaker in terms of how he constructs and negotiates his identity, or how he manages the listener?s impressions. Scott and Lyman?s (as cited in Jukes, 199) notion of accounts has value in its applicability to the present research. This follows the trend of past qualitative research on intimate partner abuse (Anderson & Umberson, 2003; Cavanagh et al., 2001; Boonzaier & de la Rey, 2004; Hearn, 1998; Ptacek, 1988; Stamp & Sabourin, 1995; Wood, 2004). Typically, these have been organised into various account types, namely, excuses, justifications, minimisation, denial and repudiations (Dobash et al., 2001; Goodrum et al., 2001; Hearn, 1998; Ptacek, 1988; Stamp & Sabourin, 1995; Wood, 2004). In the present research, accounts were predominantly organised into justifications, dissociations, and a third less common category called confessions, each reflecting different ways of negotiating an identity that was removed from the label ?abuser? (Jukes, 1999). As Hearn (1998; 1999) emphasises, it is necessary to examine how men locate their talk about their violence within particular contexts. In the present research, the exchange of words in the interview context becomes a pertinent one as it connects more broadly to the societal context. 4.7 TALKING ?VIOLENCE? IN ITS CONTEXT Whilst lending cognisance to the forms of accounts, Hearn (1998) cautions that an awareness of both their informativeness and limitations needs to be considered. Specifically, the notion of ?talk in social positionings? (p. 69) should be acknowledged, as these are made explicit (i.e. as articulated by the men themselves) and those that are 95 merely implied (i.e. observed through the analysis). Following Hearn (1998), this section brings to the fore the various subject positions that the men in the present research occupied in their talk of their violence, and will reveal how these shift within various discourses, thereby reflecting the notions of ?multiplicity? (Goldner, 1999). Present in these social positionings is also the researcher?s influence, particularly in terms of how and where she is socially located in relation to the participants. As the ?instrument?, the researcher is also the ?theorist, provocateur, observer, recorder, and interpreter? (Ptacek, 1988, p.136) and therefore has an impact on the subject of study. For this purpose, issues of reflexivity will be more closely examined. Existing research and literature have analysed men?s talk of violence, typically referring to terms such as, ?employment of strategies?, ?(rhetorical) devices?, ?techniques of neutralisation? and ?vocabularies of motive? to describe how men engage in impression management when narrating stories (Adams, Towns & Gavey, 1995; Cavanagh et al., Hearn, 1998; Ptacek, 1988; 1997; Scully & Marolla, 1984; Stamp & Sabourin, 1995). By implication, these are seen as intentional and deliberate means of impression management. As Cavanagh et al. (2001) point out, such devices are ?purposeful, reactive and proactive tactics designed to mitigate their responsibility for violent behaviour? (p. 700). Whilst this may certainly be the case, it is also emphasised that these ?strategies? may be employed outside of conscious awareness. Schlenker (1980, p. 6), for instance, defined impression management as ?the conscious or unconscious attempt to control images that are projected in real or imaginary social interactions?. According to Hearn (1998), context comprises three distinct material and discursive aspects: the ?strategic?, the ?societal? and ?agency? (p. 69-83). As evidenced in the present research, in talking about their violence in the research context, the men made reference to the societal context of gender relations. Noticeably, participants? meaning frames were also largely influenced by the language or talk of the organisation (i.e. Hearn?s ?agency context?). The researcher, who occupied the interpersonal space between the private setting, where men related their experiences, and the societal context, also influenced their manner of self-presentation. Rather than represented as discrete domains, these contexts dissolved into one another in the research setting. Direct reference in this regard is made to Hearn?s (1998) notion of ?violence as a strategic context?, ?agreeing to talk? and ?establishing credibility? (p.69-70). 4.2.1 Violence as a social positioning The strategic context denotes the ?strategic orientation of the man or strategic response of the man speaking about his violence in general [emphasis in original] as a 96 problematic activity? (Hearn, 1998, p. 69). Violence itself is a form of social positioning wherein most men are located. Within this positioning are differential positionings within which men situate themselves, for instance, ?never been violent?, ?previously violent? and ?presently violent?, or ?violent sometimes? or ?always violent? (Hearn, 1998). In the present research, most men positioned themselves as the ?changed man?, that is, the man who is no longer violent. By situating themselves as ?no longer violent?, they were able to persuade themselves, the researcher and their potential audience regarding their changed status as men. Noticeably, the interview exchange was mediated by a ?third presence? ? that of the audio recorder ? to which the men were conscious of. Two men in particular, both identifying themselves as being in the public eye, conveyed their message to the people ?out there?. In both instances, anti-violence was preached in the stance of being the different man. Towards the end of the interview, P2 took on the persona of the activist and reflected on creative ways of preaching non-violence to the youth. Using the personal pronoun ?we?, he allied himself with the interviewer in what he perceived to be a common cause. In so doing, he was able to persuasively relinquish the identity of abuser: If we can try and teach people in a nice way where they gonna accept it, like people love to be entertained, so if we can try and do this thing in an entertainment kind of style, I think they gonna do right ?cause the youth is the ones that is growing now. It is the one who is always talking things out, go and tell their parents what is happening in the school, or what is happening in the streets. So if we can try and feed the youth with information like this, they will grow up and make better communities, you know. Adopting a public persona, aware of the recording and its potential audience, he issued ?caution? to those listening: ?So if you can avoid yourself, stay away from trouble like jealousy, or [laughs] you know stuff like that. You can live a good relationship?. P1, as with P2, took on a public persona and offered ?inspiring words of wisdom? to people. In this manner, he positioned himself as the ?misogynist?-turned-activist who reaches out to men ill-informed about social gender stereotypes. Through his moral injunctions to men, he was able avert being labelled as ?abuser? or ?rapist?: No, the only thing I want to add, it?s my final word, I mean and it is the challenge to all men, I mean if you are a man, it doesn?t...your manhood won?t be determined by (.) your muscle or your physical strength, but determined by your wisdom, your wisdom. If you want to win a woman, you must use those nice words to win women, don?t use your muscles?/?/And men, when they see women, they must see human beings?not sex objects, not sex object. /?/ There are so many things that you can do and women are not trapped in the kitchen, they are so powerful, they can even rule the world. /?/Sometimes you must be a man in love (.) and help your woman when she does cooking, help your woman when she does 97 cooking. It?s not humiliating, it?s being man enough?So that?s the only thing I can say and people must definitely love their partners and they must be faithful?. 4.2.2 Men agreeing to talk However, the outright assertion that they are no longer violent or abusive is one form of strategic response by men speaking about their violence. As Hearn (1998, p. 20) notes, however, merely agreeing to talk (or being interviewed) itself is in some cases a way of demonstrating change. Through the act of talking, another self is created, not just a self who is able to confront his violence or ease his guilt, but also one who shows that he has talked about it. Among the participants interviewed, some men reflected on this process of talking about their violence within the research context. Both P3 and P1, for instance, made reference to healing themselves through the act of talking. P12 stated: ?I feel relieved, because I had no one to tell, but (.) ja, I had no one to tell?. P5, in one instance, also made reference to talking as a tool for coping: ?So it?s like, I can see that okay, I will overcome this by talking ? communication, it?s final survival. So I think communication for me is the main key??. In a similar vein, P3 voiced his intentions for participating in the present research. In the following excerpt, he made explicit the link between healing of himself through assisting others to heal: I feel that I had to do this and em (4) maybe to try and assist and help other people (.), especially myself in particular?/?/ (.). Um, because it?s important that these things are talked about (.) ah in order to be resolved (.) and not talked about generally, but talked about with people who got affected in real situations, um because what happened with me (.) those years ago doesn?t mean it?s not happening now, with someone else somewhere. More significantly, however, is his assertion that talk functions not only as a mechanism for resolving issues (both interpersonally and within oneself), but also as a tool to clarify and separate the ?assumptions? from the ?facts? to guard against misunderstandings: I think it?s important that these things are talked about so that when we help people, or whatever purpose that are developed to assist these people must be based on facts, eh, than assumptions. On a more latent level, what is suggested is that the act of talking allows the space for the ?other side of the story? is heard (i.e. the perpetrator?s as opposed to the victim?s). In a sense, the notion of ?consent? is reciprocated. For the participant, giving consent to being interviewed is exchanged for the researcher?s consent in allowing his marginal voice to be heard. Whilst ?marginal voices? reflect the so-called ?deviant?? or ?different?, those whom have been silenced in terms of political expression or historically denied a public voice (e.g. women, minority cultures/races, homosexuals, etc) (Boonzaier & Shefer, 2006; Eagle, Hayes & Sibanda, 1999; Shefer & Potgieter, 2006; Shefer & Ratele, 98 2006), the perpetrator of violence, by virtue of his display of ?anti-sociality?, is perhaps doubly stigmatised. Talking itself then becomes a means to (re)negotiate a non-deviant identity through unmasking the image of monstrosity to reveal in P5?s words what is ?only human?. P10 drives this point further in his repeated emphatic assertion that one needs to ?sit down with them [the men who use violence]? and to ?try to find out what is making them do this?. Rather than addressing the problem of men?s violence through the institution of laws and ?women?s rights?, he advocated an attitude that fosters assistance through understanding as opposed one which encourages punishment and alienation: When I say assist them, I don?t mean financially or power-wise but try to get means, of assist them, in a way that?s going to suit them, because the problem is with the people. You cannot impose what you saying to them. If they don?t respond, what they going to do? Now it seems like they are imposing this thing. No, we shouldn?t impose. I think what they should have done, they..they they should have involve?you know, just go to a certain street?you see we have dozen and dozen of streets, yes I agree, but twelve houses per street. No, just..ju..talk to the people, get involved, you know, find out exactly what it is that, that is bothering them about this, this, women abuse, women?s right, you know? /?/ Because men are there, they won?t go away. You try and help them, yes, they do have problems ? unemployment, drug abuse, this and that, alcohol, whatever. Then, find a way of helping them! P10?s plea to assist men?s ?deviant? behaviour by way of ?becoming involved?, by implication, is an invitation to ?come closer? to understand, rather than create distance and judge. Aligning men?s violence with other problems that are seen to be commonplace and intrinsically human (i.e. unemployment, drug and alcohol abuse), the abusive man is cast in another light, one who is just, in his words, an ?ordinary person?. As Hearn (1999) points out, the talk itself, particularly in the interview context, becomes the ?good investment? (p. 70). Made explicit by P3 and P12 in their accounts, and alluded to by P1, P5 and P8, the talk assists in the ?accumulation of resources? which, to some extent, alleviates them of their ?debt of violence?. As eloquently stated by Hearn (1998), the talk represents one of the ?positive gifts to the self that are of value to oneself and perhaps others? (p. 70; emphasis in original). How men talked about their violence, however, was also influenced by what Hearn (1998, p. 75) terms the ?agency context?. The agency context is another social positioning in which men talk about their violence. Hearn (1998) proposes that depending on the agency location (i.e. whether in prison, on probation, in men?s programmes or mental health agencies), men?s talk about their violence will differ in manner and emphasis. This was certainly the case for the men in the study who were 99 drawn from three distinct ?agencies?: men who were involved in an ongoing men?s programme, men who had attended a three-day workshop and one man who had undertaken in personal counselling/therapy. Consequently, their meanings of violence, and particularly their conceptualisations of ?change? was rooted in a particular societal and theoretical context. Three men, for instance, who were part of a men?s programme were also actively involved in anti-violence campaigns in their community. P6 explicitly drew upon an ?educational? discourse to describe his transition from a state of ignorance (?not knowing?) to one in which he realised his wrongdoing: But I you know our culture is like beating a woman is shows that they are man aah, I think?I didn?t realise. I started to join [the organisation], neh (.) They workshop me then, I started going workshop, (.) I realised no man, I?m wrong. Being ?workshopped? in P6?s words, suggests reformation via an external medium that provides instruction, teaching and correction. On another level, he implied that people are naturally ignorant (or by implication, ?innocent?) and therefore cannot be held accountable for their wrongs if they have not been informed or ?educated?, so to speak. By extension, he attempted to redeem himself from being blamed for his actions by pleading ignorance, in other words, that the principle of accountability only applies once you have been ?workshopped". This is reflected in the following statements: ?Anger is for everybody ? can?t control anger (.) but if you are workshopped, you can control it? and ?? you?ll do it again if you are not workshopped. If you are workshopped, then okay, start, you start to avoid things?. Pleading ignorance (?not knowing?) may be termed a form of ?quasi-repudiation? by which ?violence is acknowledged and not specifically [emphasis added] denied? (Hearn, 1998, p. 113). However, distinct from Hearn?s (1998) definition, some of the men in the present study rather than providing ?absent? descriptions of violence (e.g. in the form of forgetting: ?I can?t remember?), recounted their violent experiences, albeit removed the element of self-blame. This form of ignorance was also present in P7?s appeal: ??so [I] told them [the workshop facilitators] that I raped but I never knew I was raping?. As with P6 his insight into his actions were discovered following workshop attendance: ?It helped me very much because (2) this thing I never knew that (2) maybe when you (.) beat your girlfriend, after that you want to make love to her it?s a rape. If I never knew that, many things, many things, it helps me, it helps me?. However insistent he was in asserting this line of reasoning, when subsequently asked to confirm his lack of insight into his actions, he asserted: ?I knew at that time, just because of anger ? I forgot?. Like P6, P1?s account contained contradictions and tensions between ?knowing? and ?not knowing?. In one instance, he talked about misinterpreting his partner?s ?screams of pain? 100 for ?screams of pleasure? during sexual intercourse: ??and that lady, I mean, she was screaming unaware that that lady she scream, I thought maybe she was enjoying (.) and she wasn?t enjoying, she wasn?t enjoying that sex??. In another instance, he confessed to his own self-gratification at the expense of his partner?s apparent physical pain. Beyond the screams, he pointed out that she too expressed her pain in words. Thus, ?not knowing? became an unconvincing explanation for his violence: I think I was violent because I mean (.) she was scared of me and even when we were having sex, she was always screaming (.) screaming, telling me that I?m hurting her but I would continue?I would continue, because the only thing that I was more about concentrating was my feelings, not about her feelings. As with P6, he however resolved this incongruency by juxtaposing his ?not knowing? self during the act of violence with his post-violent self who is equipped with ?some skills? from ?attend[ing] trainings?, as well as legal knowledge: ?People have started to open their eyes right now with this new government that ei, so according to Constitution, which means what we are doing is wrong?. The law, like the men?s programmes, served to regulate behaviour, albeit more punitively. As illustrated, the men specifically involved in men?s programmes positioned themselves differently pre and post ?agency? intervention. Hearn (1998) points out that their stories are ones that have been told before and are simply rehashed in another context (i.e. the research interview). P7, substantiated on this point: ??Cause we used to talk when in the meetings such things like that?so maybe they ask you to tell the story?. However, by virtue of being involved in the men?s programme on an ongoing basis, they naturally adopted the ?talk? of the agency. Compared to participants who were not specifically associated with an organisation (i.e. those recruited from a three-day workshop), these men defined themselves directly in relation to their violence. When asked about their definitions of violence in the context of a relationship, one participant reiterated: ?If she says, or he says: ?No? ? it means no. If you beat her up, doesn?t mean that you love her?. In this statement, he drew upon the well known feminist slogan in anti-rape campaigns. By taking on the language adopted by men?s programmes, he repudiated the cultural injunction (i.e. ?to beat a woman is to shape her up?). Apart from the three participants who did not have prolonged associations with organisations, all participants defined violence in objective terms (i.e. having ?physical?, ?emotional? and ?financial? elements). This was evident despite the interviewer?s framing of the question to elicit a subjective and personalised meaning (?How do you understand violence in a relationship?? or ?What does violence mean for you??). 101 The agency context, however, not only influenced the way in which organisational definitions of violence were narrated as personal meanings. The political discourses adopted by particular ?agencies? dictated how men conceptualised violence. The participants involved in men?s programmes particularly took on socio-political discourses of violence, expressed in terms such as ?gender inequality? and ?women?s rights?. Locating violence against women in a broader societal context, particularly a ?heteropatriarchal? one, in some way ?normalises? it. By relating it to the issue of heteronormative relations or hegemony (Kandiyoti, 1994), the issue of intimate partner violence translates to a societal problem where responsibility is diffused and somehow shared, whilst personal ownership and responsibility is abdicated and dissolved. This collective approach to dealing with abuse and violence perhaps reduces shame and stigma. Notably, it was men who positioned themselves in this context that were more willing to take part in the research. Of all the participants in the study, however, one man framed his understanding of violence from a psychological perspective. His ?healing? moreover was conceptualised in relation to a different ?agency context?, namely that of counselling, therapy and hypnosis. Healing, rather than being a short-term educational learning process (i.e. being ?workshopped?), was constructed as a long-term therapeutic one whereby latent unresolved issues come to the fore: I went through hypnosis at one stage?and hypnosis showed very clearly, an example of what went wrong, what..what..what was going on in my head ? that one, under hypnosis, the uh doctor (.) recalled, or enabled me to recall? Framed from a psychological perspective, the onus rests on the individual to find the ?root? of his ?aberration? by seeking professional counselling. However, framing violence from such a stance arguably also deflects responsibility and ownership, particularly if constructed as a form of pathology (Jukes, 1999). This agency context then, rather than stressing change through education and shifting attitudes (e.g. ?No means no?), brings to the fore, the transforming of ?abnormality? to ?normality? through a personal journey/private context. Gaining access to participants in these contexts, however, proved a challenge. As it was noted by the co-ordinator of such a group, these men were too afraid to come forward and talk given the fear of stigma and the burden of shame. As revealed through those who offered to talk, their talk of violence was located in ?overlapping contexts? of the interview itself and the organisational setting. Just as agreeing to participate became a ?good investment?, a personal investment to make right what was wrong, being involved in men?s programmes itself was a means to make reparation on a societal level. As P2 commented, ?they [the organisation] told us we have to make amends for our wrongs by ourselves because nobody knew, I guess?. 102 4.2.3 Establishing credibility in accounts Although giving consent to talk is in some ways a reparative attempt, the talk itself also grants its persuasive power. Stated differently, the act of talking about their violence, men were able to ?establish credibility? in their accounts (Hearn, 1998). Echoing Scott and Lyman, Ptacek (1988, p. 141) notes that accounts of violence comprise ?a complex of anticipated judgement, face-saving, and status negotiation?. The talk or interview, therefore, is a persuasive medium by which the men attempt to convince the interviewer regarding their status as changed men. Anticipating judgement by the researcher, some of the men attempted to manage impressions by resorting to what Hewitt and Stokes (as cited in Willig, 2001) terms the use of a ?disclaimer?, ?a verbal device that anticipates, and rejects, potentially negative attributions? (p. 98). For instance, P10?s frequent use of disclaimers was manifest in statements: Aagh, I don?t know, I?m not sure about that (.), not sure (.). But I?ll say, in general, where a man feel less uh..or feel like he, he?s violated, I?m not sure, I?m not saying when they are violated they should act in this way, but, violence is bound to happen? You know she?s a lady, she won?t turn against you, but you are fighting her. I?m not saying it?s like a fight, but I mean those things which they do happen, get pushed, get slapped, I mean in a relation, sometimes the others use phones, to fight you? In both assertions, his initial phrase of ?I?m not saying?? disclaims potential judgements that the interviewer may make of him, based on his subsequent statements: ??but violence is bound to happen?? and ?but I mean those things which they do happen??. In safeguarding against possible judgements regarding his sanctioning of violence, he attempted to create and convey a non-violent self-image. Attaining credibility in accounts, however, can also be achieved through the elicitation of sympathy from the interviewer (Hearn, 1998). By conveying a vulnerable, impoverished or victimised self, the abuser re-positions himself ?not [emphasis in original] as a violent man needing social condemnation but as someone needing support? (p. 71). (This was illustrated in previous excerpts of P10, who attempted to cast the men who use violence as ?helpless men? rather than as ?perpetrators? requiring punishment). Such a strategy, however, was employed variously in other cases. P2?s account, for instance, was a personal narrative about losing out on love and close relationships as a child: You know growing up as a kid, thinking that no you never beat a woman and in my life, even she can beat me, ?cause I never find a thing like love or relationship. I was still a kid, so eish you know there was some other relationship, you know. Being deprived of love and a solid attachment with his parents, he attributed his lack of patience as an adult to being abandoned by his parents during childhood. The constant 103 waiting, the anticipation that they would one day arrive, hopes built up by his grandmother, at the same time dashed and disappointed by reality served to engender feelings of distrust, intolerance for waiting, and an underlying fear of abandonment. These sentiments were expressed by P2 as he reflected on the reasons for his impatience. Notably, it was the instance of not finding his partner at home that precipitated a violent response: Well I grew up with my grandmother you know, so I was always waiting for my parents, so I grew up pissed off and stuff, always waiting for empty promises and stuff?Yeah it?s one of the things that groomed me to be impatient in life, you know. So every time, you know I?m a little bit spoilt, but I don?t want to show that to the community ?cause she?s the one woman whose spoiling me, my grandwoman [grandmother] every time, yeah (.) So she?s trying to fill a space, like my mom and dad is not around most of the time. She has to tell me how they are coming. Every time she buys me nice things for me: ?Yeah your mom now she?s coming. She said I must give you this okay??. You know, poor woman, bought that from her money. There?s was nobody gave her that thing to promise me. So it came, I think it came from that, I don?t want to say I think ? I know? /?/ Yeah, ?cause like I?m used. I know that I?m gonna talk to someone today like ?I?m gonna come?do this and that na na??Ah okay, like okay I grew up being that kind of kid like okay?ah, everything?s everything for me but I?m still a kid, I?m not like I cry all days and ?like I know, even other parents, I used to think they are like that, besides my grandmother. So I lost trust. So that?s why I?m impatient. Cast in the light of an abandoned and disappointed ?kid?, he temporarily loses his identity as an abuser. The aggressiveness and violence typically associated with the image of perpetrator is shed away and replaced by the sobbing child who gradually learned to lose trust, or the heartbroken man who still yearns for ?true love?. Construed in this manner, his deprived circumstances therefore ?allowed? him to become as he is/was. As summarised by P2: I?m just dealing with other things like that you know in life?? In a similar, yet distinct way, P5 provided a persuasive account of his struggle for ?survival? and overcoming peer pressure, shortage of skills, unemployment and difficult relationships: There?s something like peer pressure, neh, but I?ve survived it. If I..I stand my ground if somebody tells me ?let?s go and do this?. If I don?t feel like it, I don?t because, I had a brother who passed away, used to do that, but (.) when he passed away he was sick, so I didn?t want to repeat the things that he did, because there is a younger brother and a younger sister. If I do that, maybe even in future, my younger brother did that, they?ll say he learned that from me, so I don?t want a bad experience to be a bad example?so, along the years I maybe I got temporary job by work. In fact, I was a person like, since I didn?t have any qualifications at school neh, so far, I don?t know my skill, I have been search my inner self. Each and every job that came, neh, I took it? 104 In contrast to P2, the emergent picture is one of victory and strength and the concomitant responsibility he takes upon himself to be the ?good example? to both his younger siblings and his son. Rather than the defenceless child in P2?s representation, P5 presented himself as ?no longer the small boy? who loses his temper but the responsible ?father? who has ?got to be there? for his ?baby boy?. By taking on the survivor position, he attempted to convince the interviewer that he has now attained maturity: ?I?ve overcome them. I?m growing, I?m 29 and I?m going to turn 30. I?m no longer a small boy, so I worked on my temper. I know how to handle myself in situations??. Metaphorically, he adopted the persona of a penitent sinner who now makes it part of his duty to ?preach the word? of non-violence. As reflected in these excerpts, some of the men utilised various means to render their account as a trustworthy or credible one. In P5?s case, he contextualised his ?survivor? position in the context of unemployment, difficulties with money, as well as personal struggles with built-up anger derived from the apartheid struggle. Using the social as a contextual background, the men?s accounts incorporated what has traditionally been termed ?excuses? and ?justifications? (Scott & Lyman, as cited in Stamp & Sabourin, 1995). The men?s accounts of violence were located variously in different societal contexts, namely the family, life stress, as well as what Hearn (1998) terms the ?heteropatriarchal context?. As reflected in the content themes, these centred on both ?losing control? and ?being in control?. As embedded in the narrative structure, they provide insight into how participants structured their stories. These narrative accounts predominantly reflected the use of excuses and justifications. In the present research, through dissociations, justifications and confessions, the men were able to ?establish credibility? in their narratives of violence. (i) Dissociations As Hearn (1998) argues, with excuses, the man is constructed as passive, an object controlled by external forces beyond his control, be it social, psychological, or chemical. A victimised and potentially violent self is thus often portrayed in a narrative of ?losing control?. Responsibility is variously assigned to the (distant) past (e.g. mothers, sexual abusers, school), alcohol or drugs, or located within himself (e.g. mental illness). In short, the potential for violence is construed as existing inside the man and triggered by circumstances in the present (Winstok et al., 2002). Constructing accounts of violence in 105 this manner suggests that, while the offence is acknowledged, ?its preventability is denied? (Stamp & Sabourin, 1995, p. 286). The majority of accounts in the present research typically classified as ?excuse-based?, contained the element of dissociation, whereby the participants experienced their violence as ?separate? to themselves. Following Wood?s (2004) line of reasoning, the label ?excuses? rather than denoting an ?insider? understanding, represents an ?outsider perspective? used to understand the motives in men?s accounts of violence. Rather than impose judgement on the accuracy of the accounts, the current research served to understand how participants give meaning to their experience (Stamp & Sabourin, 1995). Following this logic, ?excuses? were framed as ?dissociations?. Dissociation, according to Wood (2004), denotes the disconnecting of the self from the violence, such that the label of ?abuser? is to some extent dissolved. Hearn?s (1995) concept of distancing implies a similar understanding, whereby the participant in his account presents ?at least two selves ? the violent self who did the violence; and the talking self being interviewed?. Both concepts of dissociation and distancing apply in this study, whereby the present self was constructed as different or changed in relation to the past violent self. Wood (2004) notes that unlike dissociations, ?excuses? ?obscures many participants? active repudiation of identification with ?abusers?? (p. 571). Denoting both the experience of violence and the strategy for narrating a non-deviant identity, ?dissociations? were reflected in men?s narratives of ?losing control? which, according to Hearn (1998), are employed to ?excuse? violence, or reduce moral accountability (Hearn, 1998). As revealed in the content themes, the majority of men described feeling disoriented preceding and during the act of violence. Three men made explicit the link between disoriented thought and irrational behaviour by highlighting the role of alcohol in their state of ?reduced competence?. Although blame was acknowledged, the responsibility for violence was diminished. To paraphrase Cavanagh et al. (2001), guilt from being incompetent (i.e. being under the influence) is acknowledged, rather than guilt from the violent act. The contrast between the irrational and rational self is brought to the fore once again in P9?s reference to the dissociative experience of violence brought upon by alcohol: You come down, you?ve had some drinks with the boys. It?s a Friday night, you?re feeling full of yourself and top of the walk, you know, and suddenly you?re (.) attacked (.) by the person whose not supposed to attack you. And the reason, the rational self is not there, it?s not operating, because if I had been, if I had been sober, I probably would have said ?I?m terribly sorry and really awful, I apologise??But the alcohol allows the other side of the person to come into play. 106 As noted, the ?losing control? metaphor was also employed by men to ?excuse? their violence. Some located the context or origins of having ?learned? to lose control in a distant past. Referring to anger as residing inside oneself, two men attribute their anger as being ?inherited? during the apartheid past. P10, for instance noted: I mean, violence is triggered by something. I?m not saying you should, if ever there?s something, you should resort to violence, but eish (2) there are quite a number of things, I mean (.) that makes, especially us young guys to be violent, you know what I mean. For?our backgrounds, they do counts, where you come from. Yes, we do have our parents, mothers, for those like me who do have mother and father, but eh, I don?t, I don?t know this thing. We are filled with anger, there?s much, too much anger. Similarly, P5 poignantly related the source of anger to events during the South Africa?s political struggle. He recounted how suppressed pain brought on by physical humiliation by others served to create a destructive rageful self. In this account, he alluded to the social forces that have in a sense destroyed the ?real self?, replaced by a monstrous self: So that thing I think is what, eh caused a lot of people to riots because they..we are bottling that thing of which if we did cry, the people will see that this person is hurting. If like you are not crying, a person is busy beating you, looking at you, it cause eh, something really bad?they causing, they building somebody else inside and the time you let the person out, you see that you are piling, you are creating a monster inside. P6 recalled his reactions to witnessing his mother being beaten by his father: ?I was young, I was like crying, ?no, don?t beat up my mom, ei?? crying?. His difficulties with controlling his anger are rooted in violent family of origin experiences. He rationalised his violence by taking on a victim persona, as a child ?brainwashed? by his father into believing that beating a woman is acceptable: ?I was mentally abused by then?. (ii) Justifications Justifications contain a more conscious element, and draw on the interpersonal domain. Labelled as ?blame?, Goffman (as cited in Cavanagh et al., 2001, p. 703) explains that ?there are acts which the individual admits to doing but claims that he is not really at fault?. In this manner, responsibility is to some extent absolved by blaming other individuals or circumstances. As illustrated in the content themes section, men?s reasoning for their violence were premised upon gendered assumptions about women and their behaviour (Goldner, 1999). Some men, for instance, made explicit reference to the cultural sanctioning of violence against women. This was manifest on various levels of society. P6, for instance, drew upon the ?cycle of violence? discourse to justify his violence: ?My father told me that if you don?t beat a woman up, ah, you?re not a man /?/ 107 Ja, then my father used to beat up my mother, then I ?cause like aah my father is a man??. Campbell (as cited in Hearn & Whithead, 2006, p. 42) notes that the boy ?learns that aggression pays?, particularly where there is exposure to violence in the home. Family dynamics, in the case of husband-on-wife violence, therefore are seen as lending support to patriarchal understandings of gender roles and expectations (Jukes, 1999). On another social level, men justified their violence with reference to peer encouragement, acknowledgement and approval of beating a woman. The men looked up to their peers as holding the standards for how relationships with women should be run. Following ?role-model? behaviour for violence was a means to guard against potential humiliation, in P3?s words, ?trying to maintain that status that I must not be a joke?. Similarly for P5, enforcing respect in his partner was based on the need to secure his manly status among peers: And maybe at some stage neh, you are with friends, you know we do talk, ?how?s your relationship??, ?Ai, you know I?m warning her?. And maybe somebody say, ?Ay, you know that you talk too much, that she?s taking you for granted? ? you know some peer pressure some how does apply there. Maybe one of your friend, maybe her girlfriend is no longer cheating because your friend, your friend beat her up badly, so maybe, you say maybe ?if I can beat my girl, I show her a thing or two, maybe she?ll start to respect me? Apart from learning violence from family and peers, the men also attributed their past violence to the broader location of community. For P1, violence was a matter of ??survival of the fittest?? in which gangsterism offered some protection: Township by that time, township was very violent (.). We use to steal, I mean, hurting boys, hurting each other. So (.) in order by that time, there was, there was..there were gangsters, so I was, I was also one of them, those gangsters who were doing those funny things? Construed in this manner, violence is seen as something ?natural? and commonplace during apartheid, that being ?influenced? was almost inevitable: You know that I mean, when you are a teenager during the ?80s?because of the influence of the riots, the violent and lack of role models, so, you will obvious, idealise people who are doing wrong things (.), people are doing wrong things. And you know that especially with us, especially with us Blacks, especially boys (.) even in a community, when you become a teenager, you will tell yourself that now I am a man enough, no one will ever tell me?and that will influence by your surroundings (.), your surroundings. So when I was a teenager, I was very, very aggressive ? very naughty, very naughty when I was a teenager? 108 Whilst justification of violence relied on the discourse reflecting a ?culture of violence?, as seen from the excerpts, they also reveal men?s reliance on discourses reflecting a hegemonic masculinity. In P1?s words, ?manhood, it was, by that time, it was not determined by (.) no intelligent capacity ? it was determined by your muscle??. Participants? statements connect the social context of violence with social prescriptions of gender norms, and more particularly the relations between men and women. These actions reflect a set of beliefs, termed by Hochschild (as cited in Ptacek, 1997) as ?gender ideology?. Coming to the fore during adolescence, these beliefs are emotionally rooted and inform the cultural ideals of manhood and womanhood that individuals adopt. Masculinities literature has attempted to show that the image of masculinity is socially and culturally shaped. Moreover, aggression is considered the acceptable means by which emotions are to be expressed by men (Umberson et al., 2003). Previous research has identified that women were consistent targets of blame for the violence that their men directed toward them (Anderson & Umberson, 2001; Cavanagh et al., 2001; Dobash & Dobash, 1979; Fuller, 2001; Goodrum et al., 2004; Ptacek, 1988; Wood, 2004). The use of justifications in this sense is therefore typically gendered in nature. They are constructed as ?a response to something else in the present or recent past, particularly something not [emphasis in original] done? by the woman, thereby ?bringing forth an internalised response in the form of his violence to fill the gap/lack? (Hearn, 1995, p. 126), for example, failure to be the ?good wife?, neglecting household responsibilities and childcare duties, not being sexually responsive (Dobash & Dobash, 1979). As articulated by Dobash and Dobash (1979), the man who beats his partner for these reasons makes a powerful statement about his ?belief in her inability to be a good wife? (p. 125). As indicated in Part I of the analysis, men resorted to violence as a means of exercising and maintaining control over their partners, which were expressed either in the form of demonstrating ownership over partner, as a means to demand respect or correcting behaviour. As, such the man is typically constructed as the man having rights to possess and control women (Cavanagh et al., 2001). Their talk of their violence is therefore located within the broader social context of gender relations, what Hearn (1998) terms the ?heteropatriarchal context?. IPV is therefore contextualised within a broader societal arena that prescribes expectations for men and women and reinforces patterns of subordination and domination (Dobash & Dobash, 1979, Hearn, 1999). As Ptacek (1988) points out, the assumption of male entitlement and privilege extends beyond the battering context and is deeply entrenched within a society and culture at large. 109 Violence toward women was often carried out in contexts where women were perceived as failing to uphold societal norms that regulate women?s dress, appearance and behaviour. Thus whilst responsibility was acknowledged, women were consistently blamed for the violence, either due to failing to listen to their men, challenging their men (e.g. raised their voice at them), or acting in ways that were considered gender- inappropriate (e.g. being drunk or unfaithful). Munia (2000, p. 10 cited in Greig, 2001), in this regard notes: ?Any kind of rejection of conventional gender roles, whatever that might mean in a particular context, is going to be punishable vis-?-vis violent behaviour?. Toch (as cited in Cavanagh et al., 2001, p. 711) terms some abusers as ?norm enforcers?, whereby they take on the role of issuing ?just? punishments for those who have contravened the norm. In this manner, women?s behaviour is interpreted as flawed, thereby legitimising their abuse. The role of women is positioned in relation to culture: conforming to gender norms is equated with respect for patriarchy, whilst transgression is threatened with disownment or violence: I don?t go for girls wearing trousers (laughs), I don?t go for that ? with me. I know I stick to my culture. I know what I want, so (.) /?/ Of her, I just, I just told myself we don?t have a future, because my dad is my dad. I mean in the family as a whole, including that, the family as a whole, at some time, she?s going to be married to..to me, eh and pop out, wearing a miniskirt like that, I mean, it won?t make sense I mean ? respect? (P10) Just like in the past, we are telling ourselves, the girl can visit me wearing a miniskirt (.) it?s a challenge (.), ja it?s a challenge. So we tell ourselves, okay that there is a motive behind that manner of dress, so that?s what we are doing in the past. (P1) Different to excuse-based accounts, violence is thus construed as coming into play as a result of a social absence, rather than due to triggering an internal presence. Violence is thus seen as an instrumental response and agentic in correcting a particular behaviour in a relationship context, for instance, in the wife?s actual or assumed infidelity or failure to restrict her own movements (Hearn, 1998). By allocating blame to the woman, violence is construed as a woman?s problem rather than a problem with self (Cavanagh et al., 2001). As P9 noted: ?Initially I would blame her ? initially. It must have taken me I guess about a year before I accepted that it was my problem, my fault??. (iii) Confessions This attempt at seeking absolution was prevalent in men?s accounts. Previous research has variably termed this device ?confessions? (Hearn, 1998), ?remorse? (Presser, 2003; Wood, 2004) or ?apologies? (Cavanagh et al., 2001; Goffman, as cited in Cavanagh et al., 2001; Scully & Marolla, 1984), the latter defined as: Expressions of embarrassment and chagrin; clarification that one knows what conduct had been expected and sympathises with the application of negative 110 sanctions; verbal reflection, repudiation and disavowal of the wrong way of behaving along with vilification of the self; espousal of the right way and an avowal to pursue that course; performance of penance and the volunteering of restitution (Goffman, as cited in Cavanagh et al., 2001, p. 707) Different to Cavanagh et al.?s (2001) analysis which highlighted both exculpatory and expiatory discourses of men, men?s accounts in the present research reflected attempts at seeking absolution for their wrongs. As outlined earlier, men?s agreement to talk reflected both an attempt at reparation, as well as an act of self-redemption. As revealed, the men differentiated between the talking self/present self and the violent self/past self in the assertion ?I am a changed man?. The means by which this was achieved was through self-rebuke. As Cavanagh et al. (2001), however points out, castigation of self is itself a management element. By chastising themselves, they close off avenues for the listener (interviewer) to do so. In addition, they also attempt to ?prove? that they had changed. In recounting ?evidence? of their partners? forgiveness of them, they were able to give strength to the veracity of their statements. The men?s accounts could be typically classified as ?confessions?. (See ?The violent self as not me?). According to Hearn (1998), confessions are characterised by ?a relatively high degree of consciousness, completeness, accuracy or revelation? (p. 134). As such, these involve some level of acceptance of both responsibility and blame by the participant. In the majority of cases in the study, feelings of remorse were conveyed. However, as suggested earlier, confessions may or may not contain elements of remorse. The underlying feature of confessions, according to Hearn (1998), is an attempt by the individual to be transparent in his account, to accurately reveal or disclose things as they really were. The defining characteristic is thus the presence of ?honesty? whether remorse is manifest or not (Hearn, 1995). According to Hearn (1998), confessions incorporate two central themes based on the ?real self? of the man and the ?real power? of the man. Whilst the former represents a ?development of the inner-located account of violence found within excuses?, the latter centres on a ?development of the interpersonal, or relationship- based, account of violence found with justifications? (p. 134). Confession involves acknowledgement, disclosure, admission of fault or weakness (Besley, 2005). According to Foucault (as cited in Besley, 2005, p. 84), it is a form of knowledge-power, a discourse of truth: ?the penitent superimposes truth about self by violent rupture and dissociation?. The act of verbalisation, therefore, serves a performative function, that speaking truth about oneself enables the construction of a new self or selves. By narrating themselves, 111 they ?revise? the past and engage in a process of ?self-scrutiny, self-evaluation, and self- regulation? (p. 86). Moreover, as Besley (2005) notes, ?one is duty-bound to perform this confession as a repentance in the hope of absolution? (p. 85). Relating back to participants? ?agreeing to talk? (Hearn, 1998, p. 70), the research interview becomes the medium by which their ?truth? is facilitated. This is articulately captured in Besley?s (2005) statement: Accessing this inner self or ?truth? is facilitated by professionals in the psy sciences or helping professions (e.g. priests, doctors, psychiatrists, psychologists, psychoanalysts, counsellors) who may administer certain ?technologies? for speaking, listening, recording, transcribing and redistributing what is said, such as examining the conscious and unconscious, and confessing one?s innermost thoughts, feelings, attitudes, desires and motives about the self and one?s relationships with others. (p. 84) In constructing their accounts of violence by resorting to excuses/dissociations and justifications and confessions, the men were able, to some extent, able to exercise power in terms of their meaning of violence (Cavanagh et al., 2001). However, men?s accounts, rather than reflecting discrete categories, contained strands of inconsistencies and contradictions. Whilst excusing themselves from responsibility, they also attempt to find justifications for their behaviour (Ptacek, 1988). In one instance P2 recounted: ?You know growing up as a kid, thinking that no you never beat a woman??. In the following exchange, however, he reflected on a contradictory moral teaching (i.e. that beating a woman is justified). Moreover, he attempted to reconcile these inconsistent rationalities by juxtaposing them with the irrationality of violence. He countered his earlier view and resorts to a narrative of ?losing control? (?just happens?): P2: but like as you an African growing up knowing that a woman when is wrong, you have to beat her up and stuff, you know. Similar to that but not really? Ursula: Mmm mm. What do you mean by that? P2: Like it?s not something I can promote and say you know to beat a woman is to shape her up and stuff, no ?yeah?it something that just happens you know?? As evident, these tensions between ?losing control? versus ?being in control? explanations typically reflected dissociations in the former instance (or ?excuses?) and justifications in the latter. Moreover, whilst participants? form of talk reflected predominant themes of dissociations, justifications and confessions, other verbal ?strategies? were reflected. For instance, one 112 man initially denied being violent: ?Oh, I never hit a girl?, but subsequently admitted to violence which took the form of minimisation: ?We are used to fight with other girls in the past because (.) maybe they uh (.) they go out with so many boys, and most are included in that and, it was bad. It was bad. I didn?t like it so, I used to hit (.)?. Violence is reduced to a fight and the use of ?we? not only suggests joint accountability for violence, but a peer sanctioning of violence as the norm. P10 synonymously referred to violence as ?fighting?. In the following excerpt, he defines what constitutes violence and distinguishes between ?beating? and holding someone ?tight?: P10: Yes, mainly it was with words (.). And just hold her tight, in a manner that she doesn?t like, like shaking her, ?why did you do this, this and that??? Ursula: mm? P10: Yeah, but beating her, nooo Ursula: Okay, so it was restricted to holding her tight P10: Yes, sometimes push her Ursula: All right, and apart from holding her tight and pushing her, were there any other, ways which you used violence P10: I?ll not say violence, but I..I..I..I?ll get ways to get back at her, y?know, just go with other ladies y?know, just ladies, not something (.). Not relationship, just to get back at her, to make her feel that pain that I felt, you know? However, the interactive nature of talk is by no means a one-sided encounter. Much of these ?intertextual strategies? as Hearn (1998, p. 71) terms them, are in varying degrees achieved or unsettled in the negotiated space between participant and researcher. It is within this exchange that issues of reflexivity need to be critically examined. 4.8 REFLEXIVITY The interview context provided the structured setting in which men talked about their violence. Different to the ordinary conversation, the in-depth interview, by virtue of the specific roles assigned and adopted by researcher and participant (Kvale, as cited in Legard et al., 2003), power differentials are steeped in this ?dialogical? exchange (Kvale, in press). These power relations, however, are manifest in the interviewer not only having monopoly over the interview itself, but over the entire research process, from devising the research question, setting up the interviews, to the analysis and interpretation of the participants? speech (Banister et al., 1994; Smythe & Murray, 2000). Given this dynamic, a consideration of issues of reflexivity needs to be considered in its entirety with a view to making visible the research process (Banister et al., 1994). Nightingale and Cromby (1999) offer the following definition of reflexivity: Reflexivity requires an awareness of the researcher's contribution to the construction of meanings throughout the research process, and an acknowledgment of the 113 impossibility of remaining ?outside of' one's subject matter while conducting research. Reflexivity then, urges us to explore the ways in which a researcher's involvement with a particular study influences, acts upon and informs such research. (p. 228) 4.8.1 My Motivations In the early phases of constructing my research proposal, I reflected on my reasons for wanting to pursue research into men?s experiences of violence toward their intimate partners. Having had some exposure to the work in the gender-based violence sector, I felt the need to continue in contributing towards the societal change in the area of domestic violence. Having started out as a counselling volunteer with little knowledge about gender issues, I was naturally eager to embrace any new learning that came my way. My na?vet?/ignorance faded as I took on more learning in the field. However, I became gradually disillusioned as I began to realise that I had immersed myself in only a one-sided perspective. Given that the stress of the work was on victims of abuse (namely women), the predominant approach reflected a rather feminist one-sided intervention which, at times, I felt placed women in greater jeopardy. My curiosity as well as my need to understand the male abuser?s perspective of violence gradually deepened. Whilst in no means condoning men?s abuse, my overall dissatisfaction stemmed from I what I perceived to be a rather punitive and judgemental stance adopted short of understanding. By engaging with this research, I hoped to achieve an insider perspective of men?s violence toward their intimate partners with the anticipation that something may be derived to assist these men with change. 4.8.2 Positioning myself My wish to understand the subjective experiences of these men in some ways biased my theoretical positioning. I aligned myself strongly with an interpretive framework, which emphasised the importance of empathy in an interview encounter, underscoring the notion that people are regarded as the source of their thoughts, feelings and experiences and are considered real (Kelly, 1999; Terre Blanche & Durrheim, 1999). I attempted to adopt a non-judgemental listening stance. I felt this was necessary in order to elicit responses in men that would not be defensive and superficial (Ptacek, 1988). In some respects, I wanted to believe that they could convey the ?truth? about their experience of violence. My positioning, however, became gradually diluted as I learned to read into the meanings of their stories and adopt a more ?distanced, sceptical understanding? known as social constructionism (Kelly, 1999, p. 399). My initial discomfort with this more sceptical stance of interpretivism lay in what I considered to be a removal of the men?s subjective voice and an imposition of an outsider interpretation of their voices. Nevertheless, being prompted to be more critical and reflective of the process of ?talk? led 114 me to uncover the contexts in which men spoke about their violence, more specifically how they positioned themselves in relation to their violence. This process, moreover, allowed me to be critical of my own stance and subjective processes. Despite my attempts to satisfy both extremes on the ?interpretive continuum?, as Kelly (1999, p. 399) terms it, the tension between these stances was very conspicuous for me and I felt this throughout the research process. 4.8.3 Self presentation and power My initial contact with the participants was organised in a formal and structured manner, based on the organisational protocol. Given the sensitive nature of the topic as well as vulnerable nature of the sample, I was introduced to the group following a briefing of the research by the men?s programme coordinator. After the initial meeting, those who were interested in taking part in the study were asked to provide me with their contact details. Although the channels of access to the participants were initially guarded in a sense, I was subsequently able to make telephonic contact to set up interview times with each of the participants after they had agreed to participate. As the interviewer, the decision- making power rested with me in terms of when I contacted the participants, how I chose to set up time slots for the interviews (from which the participants merely had to select the most suitable), how I framed the questions, how I closed the interview, and most importantly, how I chose to be seen. Having adopted these structures within a framework of ?research?, the participants perhaps saw me as ?expert? having knowledge about the area of study. However, upon introducing myself to the men, I emphasised that I was a ?student? who was interested in learning about their experiences of perpetrating violence. Positioning myself in this way, I felt that I in some way diluted the power differentials between myself and the participants. However, it was also my own framing of the research (reflected in the ?participant information sheet? [see Appendix B] and in my interactions with the participants) which I believe had an impact on drawing men?s interest in the study. For one, I emphasised that much research has been conducted on women?s experiences of abuse and that I was interested in understanding the men?s perspective of violence. I further stressed that the purpose for the research was to assist men in similar situations and that their contribution would be of great value given that little has been done in this area. Although some of the participants expressed keenness to participate on our first encounter, as well as in my telephonic contacts with them, I nevertheless encountered what can be called a ?subtext of resistance? (Ptacek, 1988, p. 140) from a few of the men. Ptacek (1988) refers to this as a form of power play. One participant gave consent to participate but would not show up at the appointed time. Several attempts to 115 reschedule appointments followed by his failure to make them eventually led me to give up and exclude him from the study. Another participant also did not appear on the stipulated date. However, he did extend the invitation to reschedule for another day which was subsequently carried out. During my interviews, particularly with some of the men whom I identified as being of a similar age to myself, I noticed a greater ease and spontaneity in the way they related their stories to me ? or perhaps it was my own feelings that were being translated across to them, particularly a sense of comfort in knowing that in at least one respect (age and generational era) that we were similar, that I could identify. Apart from that, I was conscious of the social ?lines of difference? (Shefer, 1999, p. 158, cited in Boonzaier, 2001) between ?me? and ?them?. Perhaps difference could be read into all levels of identity. I am a young Asian female, a researcher, educated from a middle-class background. Apart from one white educated male in the study, the participants were black men, some unemployed, the majority holding a matric-level qualification or less, the majority being of a low socio-economic status. Some of the men seemed very aware of these differences too, particularly on the level of race/culture. This was manifest in their use of phrases, such as ?especially with us, especially with us Blacks? (P1), ?our culture? (P6), ?as a Zulu?, ?I stick to my culture? (P10). Age as an element in identity was also made conspicuous in one elderly participant?s statement: ?You were probably too young to remember that??. A disjuncture in understanding is assumed by the participant who locates himself in a distant time and space to me. Despite these disparities and dissimilarities between ?me? and ?them?, it is also relevant to emphasise that the idea of ?shared identity? is not a statement of fact, but rather contingent and based on individual perceptions and meanings (Lewin, as cited in Boonzaier, 2001). More significantly, however, was the more conspicuous difference between me and them, namely with respect to gender identity. In addition, I did not position myself as a perpetrator or as a victim of violence. Gender was a mode of distinction to which I was conscious, and perhaps even wary of in the conceptualisation phase of the research. According to Gordan (1984, p. 141, cited in Ptacek, 1988), ?no social relations are free of gender?. My own preconceived notions about ?abusive men? in some respects forced me to question whether I, as a female student and relatively inexperienced researcher, would have the necessary interview skills required to tap into the emotionally vulnerable areas of these men?s lives. Naturally, I contemplated what I thought to be a range of possible reactions and responses that my intrusions into their thoughts and feelings would elicit in them. My cautious attitude was somewhat affirmed as my research proposal was submitted for ethics clearance. Concerns about my ?safety? as a female 116 researcher who would be interviewing male abusers was voiced and that I was to take special precautions in ensuring that I was not alone with these men and that none of my personal contact information is divulged in any way. Whilst such measures were followed to guard against possible complications that might arise that would jeopardise my safety, I felt that it also introduced another dynamic. Through this dynamic of ?imbalanced access?, I perhaps created the impression that I while I was ?intruding? into their personal space, they were ?not allowed? access to mine. Banister et al (1994) in this regard points out the importance of considering the implications of the route by which the participants are contacted as this in turn ?structures the ways they see you? (p. 54). In this sense, the men may have seen me as somewhat aloof and inaccessible. Whilst ?permission? to access to their world was granted through signing consent forms, access to mine was perhaps seen as restricted and prohibited. P6, for instance, tried to alter this power dynamic by reversing roles with me during our interaction. In my attempt to seek closure in my interview with him, he retorted by asking me a question, particularly one which would reveal something about me and my relationship(s): Ursula: Okay (2). Is there anything else you want to tell me that we haven?t talked about? P6: haven?t talk about (5). Mmm, let?s see?are you in a relationship? Ursula: Mmm P6: How?s your relationship? Ursula: Mmm [laughs]. P6: No, I must understand yours too As illustrated in the excerpt, I was taken aback by the role reversal. My apparent discomfort prompted him to assert that he too needed to understand my world in the manner that I tried to understand his. This could perhaps be interpreted in many ways, whether it reflects an attempt to bring me down to his level, an offering to listen to my side, or more simply, a wish to connect with another through personally ?giving?, at the same time, hoping to ?receive? in return. However, the ?re-reversal? of power could also mean an attempt to subvert the power dynamic created during the entire research process. More specifically, this brief exchange signalled the end of the interview following which our respective roles as ?participant? and ?researcher? would be less apparent, perhaps dissolved. This transition back to ?reality? then becomes a reinstatement of my position as a woman and his as a man, as it is reflected in gender relations that are ordered and maintained by a broader patriarchal society. 117 4.9 CONCLUSION A data-driven approach to analysis revealed how the experience of violence is ?lived? from participants? point of view. More than mere understanding, however, this chapter attempts to offer an interpretation with respect to how men talked about their violence. By adopting both empathic and distanciated stances, the analysis considered parts of the data in relation to the whole, thus working through aspects of the hermeneutic circle (Kelly, 1999). The contextual nature of qualitative research implies that there are multiple frames of reference from which meaning is derived (Kelly, 1999; Rose et al., 1995). In order to account for ?transferability?, ?confirmablity? and accuracy of these findings (Lewis & Ritchie, 2003), it is necessary to validate them against existing theory. In combining these criteria for assessing rigour in the present research, in addition to examining the limitations of the present research, the next chapter attempts to situate these findings within a body of literature relating to men?s violence. In so doing, the significance of the present research is highlighted along with directions for future research. 118 CHAPTER 5: CONCLUSION 5.1 INTRODUCTION The present research aimed to understand how physically abusive men experienced their own violence in relation to their intimate partners. Specifically, an attempt was made to uncover participants? subjective experiences, the emotional and psychic dimensions in the context of their violent act(s). This allowed insight into personal meaning(s) they ascribed to their violence and how these were related to the larger social context. Based on the analysis of data, the findings reflected the men?s descriptions of their own violence as well as the means by which they talked about their experiences. A number of tensions and contradictions were evident in the men?s accounts. The present chapter sets out to explore these with reference to the existing literature. 5.2 THE INTERIORITY OF THE VIOLENT EXPERIENCE 5.2.1 ?Losing control? and ?having control? Previous qualitative research on intimate partner violence which has placed men at the centre of enquiry has revealed inconsistencies in understanding the nature of men?s violence. Cavanagh et al. (2001) contest the view that violence ?just happens?. Rather than construing it as an ?expressive release? (Boonzaier & de la Rey, 2004), according to the authors, violence constitutes a strategic and purposeful response ?calculated on the basis of a consideration of many factors including the man?s knowledge of the relationship and of his partner? (p. 699). In other words, violence is regarded as an instrumental response aimed at exercising power and control over one?s partner. Whilst men?s descriptions of their violence reflecting states of ?losing control? have also been reflected in numerous studies (Anderson & Umberson, 2001; Eisikovits & Buchbinder, 1999; Hearn, 1998; Jacobson & Gottman, 1998; James et al., 2002; Jukes, 1999; Umberson et al., 2003; Winstok et al., 2002), these experiences are often negated (Goldner, 1999) and dismissed as strategies by which men excuse, justify or deny responsibility (Adams et al., 1995; Boonzaier & de la Rey, 2006; Ptacek, 1988). Other researchers have distinguished clearly between men who use instrumental violence versus those who use expressive violence (James et al., 2002; Winstok et al., 2002). Winstok et al. (2002), for instance, imply that while instrumental abusers would be more receptive to change when the ?perceived benefits?(p. 141) of using violence is limited by increased punitive measures, expressive abusers, having a lower threshold of self-control and higher emotional rigidity, would respond more favourably to less severe interventions, such as therapy. In summary, expressive abusers are considered ?more 119 violent? than instrumental ones. James et al. (2002) similarly distinguish between tyrannical and exploder violence. ?Tyrants? use violence in order to assert their dominance (i.e. to frighten, intimidate or punish their partners for overstepping patriarchal norms) and typically describe being in control of their violence. ?Exploder? violence, by contrast, may not be pre-meditated and is typically experienced as a state of being out of control. Findings from the present research, however, did not reveal such neat and distinct categorisations. Rather, the men experienced their violence as being both ?out of control? and ?in control?. The ?losing control? narrative, taking the form of being controlled by their emotions and not being able to think straight stood alongside narratives of ?having control?. In the latter sense, the men expressed patriarchal ideas pertaining to male social ownership of their partners (I thought I owned her), male entitlements and privileges in terms of having dominion over women (She needed to respect me as a man) and men?s rightful obligation to control and enforce discipline (She needed to be punished). Different to research cited which presents these two types of experience as reflecting different abuser types, the present research revealed inconsistencies in these experiences. Isolating states of being or experience and attributing them to personality type or profile is in danger of harking back to earlier attempts to classify men who batter, based on their physiological experiences, to quote Connell?s (2001, p. 38) phrase, into ?fixed character type[s]? (see Gottman et al., 1995; Jacobson & Gottman, 1998). Such an understanding assumes that behaviour is unchanging, that experience is static and that all characteristic features define the essence of human behaviour. Contrary to the extreme view that male violence is invariably a deliberate strategy of control, rather than a loss of ability to exercise self-control (Cavanagh et al., 2001; Ptacek, 1988; 1997), Jukes (1999) notes that it is possible that men who report feeling ?out of control? are actually ??having the experience? of being out of control? (p. 52). According to Jukes (1999), this in itself reveals the confusion and bewilderment that these men experience in the moment of violence. Alternatively, these contradictions also reflect an endeavour to ?find an explanation for behaviour which he himself does not understand [emphasis in original] but is desperate to explain? (p. 51). In feminist-psychoanalytic terms, Benjamin?s (1990) notion of ?rational violence? explains the instrumental aspect of violent control as it was revealed in the men?s claims to ownership of their partners. Central to having control was denying the will of the woman (relegated to object status). Forcing recognition, however, was motivated by the need for emotional closeness but denying these feelings of dependency and vulnerability. 120 Sexually coercive behaviour, for instance, was an attempt at seeking an intimate connection with their partners, whilst denying feelings of insecurity and a fear of rejection. The ?struggle for recognition? (Benjamin, 1980, p. 162), moreover, played out in a dialectic of control in which the men felt compelled to violence by virtue of their partners? ?disobedience? or ?lack of respect?. Punishing their partners, in some ways, reflected, to quote Benjamin (1980, p. 156) a ?proof of servitude? and a means of diminishing their subjectivity. In the following statement Goldner (1999) captures the double meaning of the violent experience, showing the applicability of a ?double discourse?: ?men?s vernacular often emphasises the power of disembodied, overwhelming affects, as in the phrase, ?I don?t know what happened, I just ?lost it??. From a both/and position, the double-sidedness of a morally informed, psychological perspective captures something ?true? about the violent act and experience: that is both [emphasis added] volitional and impulse ridden and, thus it is both instrumental and dissociative. (p. 330) Umberson et al. (2003) offer a perspective which incorporates Goldner?s (1999) notion of ?both/and? positions, perhaps allowing some reconciliation between these seemingly contradictory explanations. Both are seen as reflective of wider discourses on masculinity around which men organised their lives (Connell, 2005). As illustrated in the content theme, violence as a masculine ?emotion?, being ?in control? is, to some extent, synonymous with exercising rationality, maintaining a stoic response in the face of stress, feeling financially capable, and having a say over one?s domestic affairs, such as one?s woman and children. Being ?out of control?, on the other hand, also supports a cultural image of masculinity in so far as it is expressed in ?masculinised? forms, whether it be drinking alcohol or showing violent behaviour (Umberson et al., 2003). Reaching a state of ?out of control?, however, is the end-state of tension build-up derived from repression of emotions and withdrawal, typically masculine stress responses, whether these originate from relationship strains or job difficulties. As proposed by Umberson et al. (2003), violence is thus the ?expression of extreme and cumulative emotional upset? (p. 244). Conceptualised in this manner, the seeming tensions between the two narratives find commonality in a gendered discourse of masculinity. 5.2.2 Love, violence and men?s emotions This gendered discourse also weaves explanatory power into the content theme, the continuum of love and violence. Mostly indicative of this was the cultural and social inscription echoed repeatedly by the men in the study: that ?by beating her, I show her 121 that I love her? simultaneously infused with ?if you beat a woman, you are a man?. Supporting previous studies, these findings reveal a hegemonic form of masculinity, as articulated by Connell (2001; 2005), which finds its place in gender relations between men and women in a patriarchal society. The men?s responses typically reflected dominant discourses pertaining to the male social control of women, male expectations for women to be the perfect ideal and masculine self-worth as tied to sexuality and sexual conquests. By engaging in language and culture, and drawing upon various cultural and social injunctions, whether through social learning, peer or cultural reinforcement, these dominant notions of masculinity were taken up, forming part of the men?s individual identities (Dimen, 1995). In so doing, these men ?presented themselves as masculine actors?, aligning themselves with ?masculine subjectivities? (Anderson & Umberson, 2001, p. 374). Through these instances of ?gendering of violence?, as Anderson and Umberson (1995, p. 375) propose, the men also reproduced ?gender as domination?. However, the themes reflecting the continuum of love and violence and violence versus emotionality was manifest in another form, not clearly accounted for in resorting to the notion of discourse alone. Significantly, some of the men in the present research alluded to feeling violence on a physiological level, in other words, through the arousal of bodily states. Moreover, feeling, as an emotional state, was also prevalent in men?s accounts, more specifically with reference to love: ?loving?, ?loved?, loving but with ?boundaries?, being loved, not being ?loved back?, not knowing ?thing[s] like love?. In these instances, love, as an intense, irrational, inexplicable feeling, was seen as a motive and source of their violence. In the words of the participants, love was an investment, described variously as ?obsession?, ?one bonding thing?, ?true love?, ?gonna be there forever?, ?the one you trust? and ?the best one?. The illusion of ?true love? or the ideal of the perfect woman was, in the majority of cases, shattered by their partner?s infidelity whether actual or suspected, signalling a (potential) threat to losing of one?s partner and more significantly oneself. Love and violence revealed itself in a continuum of experience in the context of intimate relationships. Beyond the discursive level, at the core of men?s oscillations between idealisation and devaluation of their partners was a vulnerable feeling state reflecting an (unconscious) emotional and sexual dependency. ?Love? encapsulated the true ideal wherein their partners? ?angelic status? (and associated qualities of submission, obedience and honesty) validated the masculine experience (Pleck, as cited in Wexler, 1999; Whitehead & Barrett, 2001), allowing him to be recognised (Benjamin, 1980). ?Violence? represented a negation or devaluation of the feminine where male vulnerability 122 and powerlessness at their dependence, once exposed, became intolerable to bear. In a psychoanalytic sense, wholeness is conditional upon maintaining the tension between recognition and negation. The breaking down of wholeness (i.e. splitting into contradictory halves), however, comes to define the violent relationship (Benjamin, 1980). The intolerance of women?s emotionality was further emphasised in men?s need to seek distance from or respond to with physical violence (Chodorow, 1978). To reiterate Hagemann-White and Micus? (1999) point, the violent man is not necessarily powerless, but rather is unable to accept feelings of powerlessness that is central to the human experience. These feeling states revealed as themes at various instances throughout the data are thus captured in a contemporary psychoanalytic framework. With its focus on ?relationality?, various theories within psychoanalytic thought are amalgamated: object relations theory, interpersonal psychology and self psychology (Hall, 2007, p. 29). The emotions underlying violence as a subjective experience is illuminated in such a perspective which lends voice to the ?power of feelings? (Chodorow, 1999). Subjectivity in a psychoanalytic sense then is more than being discursively and linguistically created. To quote Chodorow (1995), it involves: ?creation and re-creation, a merging and separation of fantasy and reality, ?inner and outer, unconscious and conscious, felt past and felt present, each element in the pair helping to constitute and to give meaning and resonance to the other. (p. 518) Through the uncovering of the, at times, unconscious ideas, beliefs and embedded internal representations of self and other, the psychoanalytic framework provides insight into the basis of men?s anxious attachments with their partners (Goldner et al., 1990). James et al. (2002) noted in their study that ?anxiety about attachment was the main predictor of violence? (p. 19). Other studies similarly have found fear of abandonment, intense jealousy and a need to control their partners was related to disrupted attachment with early caregivers (Dutton, 1995; Miehls, as cited in Zosky, 1999; Worely et al., 2004). Whilst such feelings of fear of losing their partner (or, abandonment) were implied by some of the men in the present research, one participant explicitly recounted experiencing disrupted bonding experiences with his parental figures which he felt created his intolerance for waiting for people in general. His relationship with his partner whom he had abused, according to him, was a ?bonding? experience which perhaps created a substitute for ?never [having found] a thing like love or relationship? in the first instance. Another also made a connection between losing his partner and in the process, his loss of meaning and happiness which he felt she brought to the relationship. Confessional accounts offered by some men also revealed emerging remorse and 123 feelings of gratitude. For others, confessions took the form of penalising guilt and punitive judgements toward the self. In this connection, Gadd (2000) articulates that explaining men?s violence necessitates accounting for the, at times, conflicting emotions (albeit without denying socio-structural factors): Subjectivity is not reducible to some monolithic masculine or patriarchal quality: it consists of an evolving set of difficult to express, and hence seemingly irreconcilable, anxious emotions. (p. 439) Unveiling the psychic dimension of men?s subjective experiences reveals what Kirkham and Thurmim (as cited in Yates, 2000) refers to as the inner struggle which lies at the core of masculinity. Symbolically, the conflict lies between the ?phantasies of power and phallic plenitude? versus the ??actual experience? of limited power and the problematic reality of dependence and relations of difference? (p. 77). As suggested earlier, violence as the substitute for repressed emotion can perhaps be construed as a form of ?externalising behaviour?, negative actions that are outwardly expressed that impact directly on others (such as threatening, attacking, fighting) (Miller & Eisenberg, 1988), which masks underlying vulnerable emotions (Scheff, 2006). Vulnerable emotions, such as feelings of rejection by one?s partner (e.g. refusing sex, refusing to go out), feelings of insecurity, fear of losing her (e.g. not finding her at home, partner?s attractiveness to other men), fear of not having a hold over her (e.g. coming home late), jealousy (e.g. seeing her with another man), the pain and sadness of actual loss, particularly to another man were part of the repertoire of experience which precipitated the violent episode(s). As Goldner et al. (1990) note, masculinity is premised on the culturally embedded assumption that men are not to feel fear, sadness, dependency and or the need for protection. Such cultural mores, however, are in direct conflict with the true psychological experience of what it means to be human. By giving significance to emotions, fantasy and bodily sensations, a contemporary psychoanalytic approach, therefore, offers a far more comprehensive account of gender and violence that goes beyond linguistic or cultural constructions. What is proposed here is a mutual influence of meanings of both personal and cultural gender (Bell, 2004; Chodorow, 1995; 1997; 1999; Dimen, 1995). As explanatory models, feminist psychoanalysis together with masculinity theory form the basis of understanding men?s subjective experiences of violence within a critical social psychological framework. 5.2.3 Changing Identities The men in the present research presented their talking self as being different from the violent self. Positioned as the ?changed man?, participants made reference to 124 descriptions, such as being responsible, having greater maturity and insight into relationships, being a positive role model for others, as well as being healed. Contrasts between the former violent self and the present self were reflected in juxtapositions such as ?beast? versus ?sweet romantic?, ?Mike Tyson? versus ?changed man?; ?ogre? versus ?mentor? and ?small boy? versus ?mature man?. These shifted or shifting positions from past to present reveal what Butler (as cited in Anderson & Umberson, 2001, p. 374) refers to as the ?instability of masculine subjectivities?. Although not emergent as a theme, this dynamic, moreover, was also reflected in two men positioning themselves as victims or powerless (e.g. feeling ?violated?) in the age during which society is progressively being influenced by women?s rights and empowerment. What both illustrations reveal is the notion that identity, rather than being a fixed essence, is malleable and subject to change as individuals draw upon new or alternative discourses that have become more readily available to them (Weedon, 1987). Being re-educated on men?s workshops on issues relating to gender-based violence, the men formulated new versions of themselves. On a subjective level, this reveals itself in an integration of contradictory feelings and aspects of oneself. For some of these men, crying and tolerating one?s own tears was a decisive step towards healing. Rather than repressing, denying or disowning these vulnerable feelings, their expression constituted a release from emotional pain. As opposed to projecting empathy and emotions onto the feminine (Hagemann-White & Micus, 1999), some of the men took ownership of these feelings and were able to ?experience? the pain of their partners. Most significant in the case of one participant was the critical role that recognition played in providing the momentum for change. Through the psychic survival of the Other who claims status as a subject, she was able to grant recognition to both his ?good (consciously owned) and bad (often unconsciously disowned) qualities?. On a strategic level, men constructing new identities enabled them to distance themselves from their past behaviours, thus averting the label of ?abuser? (Goodrum et al., 2001; Hearn, 1998; Wood, 2004). As revealed in the present research, the ways men accounted for their violence also revealed what has been termed dissociations, justifications and confessions. 5.3 DISSOCIATIONS, JUSTIFICATIONS AND CONFESSIONS Framed as ?exculpatory discourses? (Cavanagh et al., 2001), excuses and justifications have been identified as strategies by which men attempt to alleviate or absolve responsibility and/or blame for their violence (Anderson & Umberson, 2003; Cavanagh et 125 al., 2001; Boonzaier & de la Rey, 2004; Goodrum et al., 2001; Hearn, 1998; Ptacek, 1988; Stamp & Sabourin, 1995; Wood, 2004). Following Wood?s (2004) rationale, the present study refrained from using the category ?excuses?. As she notes, from an ?outsider perspective? the label could be applied to both justifications and dissociations. Imposing such a term, however, not only fails to represent the ?insider perspective? of the participants to capture their meanings of violence, it also obscures their ?active repudiation of identification with ?abusers?? (p. 571). Although findings of the present research did not deviate from previous research, an attempt was made to convey meanings that were more closely centred on participants? experiences. The term ?dissociations? was employed in the present research as an alternative to ?excuses? to reflect the participants? distancing of their talking self from their violent self (Hearn, 1998). More than a ?distancing strategy? however, it also captured men?s actual experiences of ?losing control?. Justifications, on the other hand, reflected men?s appropriation of discourses on a social and cultural level to explain their violence. Whilst justifications and dissociations were prevalent among the ?strategies? men in the study used to account for their violence, there were also sporadic references to strategies of minimisation and naturalisation (termed ?quasi-repudiations?) (Hearn, 1998). The use of denial, typically cited in few other studies (Cavanagh et al., 2001; Goodrum et al., 2001; Hearn, 1998; Ptacek, 1988) was, however, not strongly evident from the present analysis. Importantly, moreover, rather than reflecting these undiluted categorical distinctions of ?excuses? (dissociations) and justifications, contained in the majority of men?s talk of violence in the present research is the use of both verbal strategies. Narratives of ?losing control? (or ?excuses?/dissociations) were juxtaposed alongside those reflecting the instrumentality of violence (justifications). As with contradictions and inconsistencies reflecting in ?losing control? and ?having control?, the mix of both dissociations and justifications in single accounts of men are also open to speculation. On the one hand, they could reveal a desire to formulate meaning which, to some extent, absolves them of guilt and responsibility (Ptacek, as cited in Jukes, 1999). Hearn (1998), for instance, acknowledges that composite or contradictory accounts (i.e. containing a mix of justifications, excuses, confessions, minimisations, repudiations) are usually recounted by men who have had a long history of abusive behaviour. In other words, as they ?accumulate methods of violence?, they also ?increase their repertoire of ways of accounting, even if inconsistent? (p. 142). On the other hand, ?rather than simply lying or negotiating a non-deviant identity?, as Jukes (1999, p. 52) cautions, it is possible, as indicated earlier, that some of these men do experience a split between their emotions 126 and their actions (Goldner, 1999). Stated differently, they may be at fault in terms of their motivations for violence, but the experience of rage itself is not an inaccuracy (Jukes, 1999). Apart from dissociations and justifications, the men?s accounts of violence were also categorised into what Hearn (1998) terms ?confessions?. Acknowledged as ?expiatory discourses? (Cavanagh et al., 2001), previous studies have framed these in terms of ?apologies? that reflect men?s efforts in ?seeking absolution? from their partners (p. 711) in order to alleviate their guilt (Goffman, as cited in Cavanagh et al., 2001). In the present research, accounts identified as confessions were those containing a ?high degree of consciousness, completeness, accuracy or revelation? (Hearn, 1998, p. 134). Approximately half of the men in the present research expressed remorse for their wrongdoings. However, as has been highlighted, while remorse may provide the basis for change (Wood, 2004), elements of remorse are not necessary indicative of a confession. Based on this understanding, it becomes necessary to interrogate the surface meanings of remorse. 5.4 RECOMMENDATIONS 5.4.1 Implications for interventions The present research has revealed that both discourses of ?losing control? and ?having control? have a place in the experiences of the men in this study who enacted violence toward their intimate partners. Whilst adopting this ?and/both? position (Goldner, 1999; Goldner et al., 1990), at the same time, it is necessary to address potential implications that each discursive position has, not only for men in rehabilitation, but also for those engaged in assisting such men in the change process. Clinicians, scholars and batterers alike, according to Ptacek (1988), may inadvertently collude in stalling remediation efforts by ascribing to a common language that characterises violence as a state of loss of control. Terms such as ?impulse to batter?, ?uncontrollable rage?, ?psychiatric abnormality?, ?borderline? and ?sadistic? constructs the abuser as one who is controlled by forces outside his will, that he is ?not necessarily sick, but who is rather just temporarily insane? (Ptacek, 1988, p. 152; emphases in original). Construed in this manner, such men occupy the ?grey area? between being labelled ?psychopathic? or ?criminal? (i.e. not sufficiently abnormal in the former instance, and not sufficiently responsible in the latter). In the opposite extreme, other scholars reject the ?losing control? narrative (ascribing it to a strategy for excusing violence), in favour of one which places men directly in the seat of responsibility (i.e. violence is always a deliberate strategy of control). In dismissing ?losing control? narratives on the basis of their potential to ?explain away? men?s violence, 127 however, fails to acknowledge the subjective view and subsequently diminishes the subjective experience of violence (Goldner, 1999). As Wood (2004) proposes, the use of the term ?dissociations? provides new insights into the means by which violent men perceive their actions. In the present research dissociations reflect two levels of meaning. On the subjective level, men?s experiences in the moment of violence revealed a detachment or disconnection of emotions from rational thought. On a more strategic level, men while talking about their violence, maintained distance from their violent selves, at times with reference to the third person (e.g. ?the abuser? or ?that guy?). Feminist perspectives of intimate partner violence acknowledge abuse as a criminal offence and advocate a prosecution-punishment stance as the chief goal of perpetrator reform (Egger, 1993). Feminist-educational interventions focus on educating men on the social and cultural norms that underpin male dominance and violence towards women (Healey, Smith & O?Sullivan, 1998). As a dominant mode of intervention for batterers (Healey et al., 1988), however, this has been criticised as being too confrontational, often resulting in men becoming resistant, disengaged and hostile. Apart from counselling services offered by a few welfare and non-governmental organisations in South Africa, abuser intervention programmes at community level are scarce (Abrahams & Jewkes, 1999). Interventions assisting men to engage with alternative views of masculinity, as well as fostering personal accountability could prove beneficial (Wood, 2004). Whitehead (2001) proposes the development of self-reflexivity in men. Transformation, according to him, would need to be initiated in ?men?s subjective perception of their own gendered identity: they must come to some appreciation of how ?being a man? might affect and influence their expectations and experiences? (p. 351-2). Reflecting on the process of men?s ?talk? as the ?object to be revolved and examined in the process of therapy? (Frosh, 1997, p. 182), interventions could directly explore men?s strategies of impression management, as well as deconstruct notions of masculinity to address underlying contradictions (e.g. rationality versus emotionality, being a ?man? versus being a ?weak man?). Examined as such, men become the ?objects [emphasis in original] of inspection? (Frosh, 1997, p. 182) to both the therapist and themselves. In revealing both the rigidity and fragility of the masculine identity, to paraphrase Frosh (1997), rationality as the masculine defensive shield is uprooted and allows vulnerability to break through. In other words, men?s recognition that the rational organising principle fails in the expression of emotional states, is both crisis- laden and enabling as the ?real? emerges. In Frosh?s (1997) words, ?if masculinity deconstructs, it should implode, revealing an underside which is not merely ?feminine?, but non-rational and possibly gender-transgressive? (p. 183). 128 However, as Frosh (1997) points out, rather than being an exercise in cognitive restructuring, it is a deeply charged emotional process given the intense nature of investments in particular discourses. Given this view, the emotional reality of men?s experiences on a subjective level also need to be addressed. Interventions should also allow men the empathic space (Wexler, 1999) to explore personal meanings of violence. This becomes an avenue by which new meaning is created (Stamp & Sabourin, 1995), whereby their ?violent selves? are integrated into their global perceptions of self (Athens, as cited in Goodrum et al., 2001). In feminist-psychoanalytic terms, the disowned and vulnerable aspects of self (typically experienced as smallness, weakness, ?unmasculine?) need to be integrated with the more recognised and ?masculine? elements of self, ultimately achieving a ?selfhood? that is able to tolerate and maintain the tension between these contradictory states (Benjamin, 1980). By dissolving rigid boundaries of gender difference, healing is facilitated as men are able to develop greater tolerance for repudiated feminine aspects of themselves (Goldner et al., 1990). Ultimately, this would involve acceptance of his own dependency needs, acknowledging their partners? subjectivity (i.e. to see her as a separate person with own needs and feelings), whilst learning to empathise with their experiences (Benjamin, 1990; Goldner et al., 1990). As the present research has revealed, the capacity for mutual recognition plays a significant role in fostering change. Emphasising the intersubjective realm as the site for change, Gadd (2006) points out successful rehabilitation may hinge on work that empowers victims. Drawing upon an object relations framework, Gadd (2006) proposes a ?desistance? as opposed to an ?offence-focused? intervention that ?empowers the individual offender to rethink of the symbolism of the offence from the perspective of the other? (p. 197). In the same light, meanings and power relationships are renegotiated as men take ownership of the split-off aspects of themselves to achieve a true differentiation (Benjamin, 1980). To quote Gadd (2006): The experience of recognition ? especially when it runs against the tide of expected power relations ? can make the other seems less strange, as (psychic) sameness renders (social) difference less significant. In such circumstances, new subject positions can be tried on for size, the past can be viewed from different perspectives, lost love objects can be mourned and the fear evoking can suddenly appear reassuringly familiar. (p. 197) Such a view is encapsulated in P9 who reflected on his new perspective toward his partner: ?The little things that used to irritate me in the past don?t irritate me at all. In fact, some of them I find them quite, attractive?funny... I can see them with a balanced eye?. 129 As has been consistently argued, rather than adopt an extreme ?or? position in intervention efforts, it is possible and necessary as (Goldner, 1999) asserts to employ multiple discourses that acknowledge both the ?psychological interior? (Goldner et al., 1990, p. 578) of men?s experiences, at the same time, challenge moral accountability. In this manner, men are not alienated through punishment and confrontational strategies which potentially incite further aggressive responses (Wexler, 1999), but through the ?layering of meanings? are given the space to be understood whilst facilitating ownership and responsibility for their behaviour (Goldner, 1999, p. 330). Such a strategy could be offered in the following terms as proposed by Goldner (1999, p. 330): ?What happened inside of you at that moment that you let yourself feel justified to go against your promise [of non-violence]??. Framed in this manner, the psychological narrative of ?losing it? is explored; at the same time he is questioned on his perceptions of entitlement and directed towards personal choice and accountability. Stated somewhat differently, Zimmerman and Dickerson (as cited in Stamp & Sabourin, 1995, p. 304) suggest: ?By inviting clients to make meaning out of other aspects of their experience, the therapist creates a context of change or re-storying in a manner whereby the clients themselves become the ones to intervene in the problem?. 5.5 LIMITATIONS 5.5.1 ?Epistemological integrity? Epistemological integrity is defined as ?a defensible line of reasoning from the assumptions made about the nature of knowledge through to the methodological rules by which decisions about the research process are explained? (Thorne, 1997). In an attempt to uphold this criterion, decision-making at various levels of the research reflected the central aims of the research: how men experience and account for their violence toward intimate partners. As such, epistemological standpoints premised on an interpretive approach (embracing elements of both phenomenology and discursive psychology), were revealed in the research questions and carried through in strategies of data interpretation (Thorne, 1997). Thematic analysis of data based on narrative content and narrative form represented a ?contextualist? method of interpretation, bridging the gap between an essentialist approach that acknowledges subjective meaning of experiences and a constructionist method which is cognisant of how the societal context is imposed on subjective reality (Braun & Clarke, 2006). 5.5.2 ?Representative credibility? Various organisations in and around Johannesburg were contacted in order to recruit participants. By targeting these specific organisations, the researcher was facilitated in gaining access to what appeared to be an ?invisible? population. This convenience 130 approach to sampling, however, precluded other men from being included in the study, particularly those who do not have access to these groups or those who were unwilling to seek assistance from organisations. As such, the ?diversity? of experiences may not have been fully captured as these are mediated by socioeconomic status, culture and ?race?. One of the organisations, based on the reasoning that its members were fearful of coming to the fore to share their experiences, moreover, was only able to offer one participant who willingly agreed to participate in the research. In addition, one of the men who initially consented to participate in the present research consistently failed to make his appearance for his appointed interview. Whilst referred to as ?non-response? bias in quantitative studies (Fife-Schaw, 1995), the criterion of ?representative credibility? (Thorne, 1997, p. 120) may be more apt within a qualitative paradigm. Simply put, this pertains to the consistency between theoretical claims made by the research and its method of sampling. Reliance on a convenience sampling approach, as Ritchie et al. (2003) point out, potentially introduces another form of bias. The organisation may encourage participation from certain individuals perceived as more articulate or having more ?colourful stories? to tell, or alternatively from those from whom positive accounts about the organisation can be derived. The ?silent? population of men (i.e. those who were inaccessible, those who subsequently decided not to divulge information, and those who remained on the organisations? database but were not regular programme attendees), could have contributed significantly to an understanding of the wide-ranging experiences of men?s violence. Importantly, the research also focused on men outside of a prison or parole context. Whilst the diversity of sampling would potentially have yielded richer understanding of men?s violence, generalisablity, or transferability per se, the extent to which the findings are applicable beyond the men that comprised the present study (Kelly, 1999, p. 431; Lewis & Ritchie, 2003), was however not the aim of the present research. More pertinent to the research was the identification of themes that govern social transactions with a view to enriching understanding of men?s violence toward their intimate partners. 5.5.3 ?Interpretive authority? ?Confirmability? of findings (Lewis & Ritchie, 2003), or alternatively, the ?trustworthiness? or ?dependability? of the evidence (Lincoln & Guba, as cited in Lewis & Ritchie, 2003) is often synonymous with the notion of reliability or replication of findings in quantitative research. In proposing that qualitative research is evaluated in its own right, rather than based on ?borrowed terminology? from a quantitative science, Thorne?s (1997) term ?interpretive authority? (p. 121) acknowledges that ?all knowledge is perspectival?. Consequently, the researcher?s interpretations require grounding in an external truth beyond his or her own pre-existing assumptions and experience. The contextual nature 131 of qualitative research implies that meaning is linked to and derived from a specific setting (Banister et al., 1994). Despite the diverse contexts in which research on men?s violence has been conducted, the present research is consistent with ?the collective nature of the phenomen[on]? as reflected in previous studies, and is expected in findings of future research (Lewis & Ritchie, 2003, p. 271). It is acknowledged, however, that consistency checks were not carried out in the present research to ascertain the extent of consistency between independent assessments of the data (also known as ?inter-rater reliability?) (Armstrong, Gosling, Weinman & Martaeu, 1997; Lewis & Ritchie, 2003). The notion of ?multiplicity? implies that the researcher?s interpretations do not mirror underlying ?reality?. Given that each individual brings to the data his/her own world views, knowledge, experiences, theoretical frames and subjective feelings, it is expected that different researchers will provide an alternative conception of reality (Armstrong et al., 1997). Alternatively, replicability was achieved by means of researcher reflexivity (Lewis & Ritchie, 2003). Both personal and epistemological reflexivity was attended to, namely the researcher?s positioning with respect to studying men?s violence, as well as in relation to the participants and their stories (Banister et al., 1994). The attempt was to ?make transparent? the researcher?s motives, intentions and feelings during the research process and how these have potentially influenced the interpretations yielded. Apart from reflexivity issues, effort was made to ground particular interpretive claims in the evidence of the data themselves, as well as to find substantiation in the existing literature. 5.5.4 ?Correctness? of evidence In qualitative research, a research reading is said to reflect ?correctness? if it ?represents accurately those features of the phenomena that it is intended to describe, explain or theorise? (Hammersley, as cited in Lewis & Ritchie, 2003, p. 273). This relates to the precision of interpretation and its correct labelling (Miller, as cited in Lewis & Ritchie, 2003). One way to assess for ?correctness? of evidence is to verify the interpretations with participants for correspondences in meaning (Kelly, 1999). In terms of the present research, such a dialogue has not been formally engaged in at the time of write-up. However, the results will be made available to the participants and the wider scholarly audience at a future date. Apart from assessing final interpretations, the validity of conclusions may also rest with how fully the said phenomenon is captured (Kelly, 1999). Pointing towards the interview process itself, although the researcher made every effort to adopt phenomenological principles, such as ?bracketing? and ?intuiting? (Rose et al., 1995), thereby allowing participants to fully express subjective meaning, in hindsight there were few traces of inadvertent leading questions (perhaps based on pre-existing assumptions) that became more obvious to the researcher during the transcription process. For instance, assumptions about violence were reflected in the following 132 question with some of the participants interviewed earlier: ?what feelings did you feel were expressed in violence, in that act of violence what feelings came out for you?? or ?what did you achieve by being violent??. Legard et al. (2003) similarly caution against the researcher using extraneous remarks, such as, ?great?, ?I see? and ?that?s interesting? which may either discourage further elaboration by participants or perhaps reinforce participants? viewpoints. Apart from these, the inadvertent use of double barrel questions may have potentially confounded participants? understanding and thus influenced their responses. Neverthless, on a theoretical level, a theory triangulation approach (Kelly, 1999) provided the avenue to explore alternative levels of meaning. By giving voice to different theoretical perspectives at different points in the analysis of data, an attempt was made to accommodate diverse and contradictory perspectives in understanding men?s violence. 5.6 DIRECTIONS FOR FUTURE RESEARCH The present research comprised a sample of men who had enacted one or more acts of violence toward their intimate partner in the past. Men who were on parole, those in incarceration and those who at the time of study had legal proceedings against them were not selected for participation. Different to a prison population or men on parole, these men had voluntarily joined various organisations that provided workshops or counselling services for men. It is expected that a somewhat different picture may have emerged compared to previous qualitative studies with incarcerated men (e.g. Wood, 2004), men court-mandated to enter batterer counselling programmes (Anderson & Umberson, 2001; Goodrum et al., 2001), or men who had been convicted of a violent offence toward their partners (Cavanagh et al., 2001). In other studies, other than stating that the men were involved in a batterer treatment programme, information relating to voluntary or mandated entry into treatment, previous history of incarceration or current parole was not provided (Fuller, 2001; Ptacek, 1988; Stamp & Sabourin, 2005; Worley et al., 2004). Previous quantitative studies have shown that men classified as more severely ?pathological? or displaying ?antisocial? tendencies are more likely to enact violence in contexts outside the domestic relationship compared to the more ?normal? (Wood, 2004) population of men (Gondolf, 1988; Hamberger et al., 1996; Hanneke, Shields & McCall, as cited in Dutton, 1988; Hare, 1993). In a South African context, qualitative studies on diverse groups of men, not only with respect to previous arrest or conviction history, but also with regard to personal background ?variables? that mediate subjective experience, such as socioeconomic status, culture, ?race?, age, sexual orientation, religion, etc. may provide alternative insight into the complex phenomenon of men?s violence. Some participants in the present research also made reference to the cultural sanctioning of violence against women within a domestic context. Further 133 research could contribute to a better understanding of masculinity, particularly the interaction of a culture of patriarchy and the witnessing of violence between parents. This is pertinent within a South African context where childhood exposure to violence between parents is a widespread phenomenon (Bernhardt, 2004). Further research that explores masculinity as both a personal and social/cultural construction (Chodorow, 1995) in the context of violence would also shed light on the findings of the present research. In the present research, the men attempted to construct identities that deviated from their violent selves of the past. How they chose to present themselves, their agreement to talk, in addition to their means of establishing credibility in their narratives revealed instances of impression management, whether operating consciously or unconsciously. Giving acknowledgment to the notion of ?multiple realities?, in other words, the ?shifts in provinces of meaning in everyday life? (Burrell & Morgan, as cited in Cohen & Manion, 1989, p. 32), additional research could contribute further, as Goodrum et al. (2001), note by exploring how men construct disparate violent/non-violent identities in other interpersonal contexts, namely with friends and colleagues. The data in the present research further hinted at men?s capacity for empathy or ability to perceive their partners? pain. The avoidance or denial of another?s physical or emotional pain, according to Goodrum et al. (2001), further reinforces a non-violent view of the self. Inversely, a ?mentalising? stance (Fonagy, 1999) or capacity to ?role-take? (Goodrum et al., 2001) is fundamental to initiating change. According to Tangney (1991) empathy is a precondition for guilt, which in turn becomes the basis for reparative action. Further research exploring these possibilities would be fruitful in informing interventions. 5.7 SIGNIFICANCE OF THE RESEARCH The present research has attempted to move beyond a ?discourses of violence? approach (Gadd, 2000) to shed light on men?s subjective experiences of their violence toward their intimate partners. Whilst lending support to a critical social psychological framework adopted by emergent South African qualitative research (e.g. Boonzaier, 2001; Boonzaier & de la Rey, 2004; Boonzaier & Shefer, 2006; Wood & Jewkes, 2001), the present research has also made central to its focus the ?insider perspective? (Wood, 2004) of men?s subjective experience of their violence. In line with the emergent interest in the psychological dimensions of masculinities (Gadd, 2003), it represents an attempt to bridge the gap in understanding men?s violence by merging personal meaning with social constructions. In so doing, it has attempted to reveal how the experiences on the subjective level of emotions are processed in social discourses and narratives (Gadd, 2004). This has implications for rehabilitation interventions which, in addition to challenging men on issues of moral accountability for their violence, should also afford men the emotional space to articulate their feelings. 134 REFERENCES Abrahams, N. & Jewkes, R. (2005). Effects of South African men?s having witnessed abuse of their mothers during childhood on their levels of violence in adulthood. American Journal of Public Health, 95(10), 1811-1816. Abrahams, N., Jewkes, R. & Laubscher, R. (1999). ?I do not believe in democracy in the home?: Men?s relationships with and abuse of women. Cape Town, S.A.: Medical Research Council of South Africa. Retrieved May 26, 2007, from the Medical Research Council of South Africa website: www.mrc.ac.za/gender/nodemocracy.pdf Abrahams, N., Jewkes, R., Hoffman, M. & Laubsher, R. (2004). Sexual violence against intimate partners in Cape Town: prevalence and risk factors reported by men. Bulletin of the World Health Organisation, 82(5), 330-337. Abrahams, N., Jewkes, R., Laubscher, R. & Hoffman, M. (2006). Intimate partner violence: Prevalence and risk factors for men in Cape Town, South Africa. Violence and Victims, 21 (2), 249-265. Adams, P.J., Towns, A. & Gavey, N. (1995). Dominance and entitlement: The rhetoric men use to discuss their violence towards women. Discourse & Society, 6(3), 387-406. Anderson, S.A. & Schlossberg, M.C. (1999). Systems perspectives on battering. In M. Harway & J.M. O?Neil (Eds.), What causes men?s violence against women? (pp. 136- 152). London: Sage. Anderson, K.L. & Umberson, D. (2001). Gendering violence: Masculinity and power in men?s accounts of domestic violence. Gender & Society, 15(3), 358-380. Armstrong, D., Gosling, A., Weinman, J. & Martaeu, T. (1997). The place of inter-rater reliability in qualitative research: An empirical study. Sociology, 31(3), 597-606. Arthur, S. & Nazroo, J. (2003). Designing fieldwork strategies and materials. In J. Ritchie & J. Lewis (Eds.), Qualitative research practice: A guide for social science students and researchers (pp. 109-137). London: SAGE. Babcock, J.C., Green, C.E., Webb, S.A. & Graham, K.H. (2004). A second failure to replicate the Gottman et al. (1995) typology of men who abuse intimate partners?and possible reasons why. Journal of Family Psychology, 18(2), 396-400. Banister, P., Burman, E., Parker, I., Taylor, M. & Tindall, C. (1994). Qualitative methods in psychology. Bristol: Open University Press. Beall, A.E., Eagly, A.H. & Sternberg, R.J. (2004). Introduction. In A.H. Eagly, A.E. Beall & R.J. Sternberg (Eds.), The psychology of gender (2nd ed.). New York: Guilford Press. Beasley, R. & Stoltenberg, C.D. (1992). Personality characteristics of male spouse abusers. Professional Psychology: Research and practice, 23(4), 310-317. Bell, L.C. (2004). Psychoanalytic theories of gender. In A.H. Eagly, A.E. Beall & R.J. Sternberg (Eds.), The psychology of gender (2nd ed.) (pp. 145-168). New York: Guilford Press. 135 Bernhardt, S. (2004). A critical literature review of marital violence and the women who witnessed it during their childhood years. Unpublished masters dissertation, Rand Afrikaans University, South Africa. Benjamin, J. (1990). The bonds of love: Psychoanalysis, feminism, and the problem of domination. London: Virago Press. Benjamin, J. (1980). The bonds of love: Rational violence and erotic domination. Feminist Studies, 6(1), 44-174. Bersani, C.A., Chen, H.T., Pendleton, B.F, & Denton, R. (1992). Personality traits of convicted male batterers. Journal of Family Violence, 7(2), 123-134. Besley, T. (2005). Foucault, truth telling and technologies of the self in schools. Journal of Educational Enquiry, 6(1), 76-89. Bograd, M. (1991). Feminist approaches for men in family therapy. New York: Harrington Park Press. Bograd, M. (1988a). Feminist perspectives on wife abuse: An introduction. In M. Bograd & K. Yllo (Eds.), Feminist perspectives on wife abuse (pp. 11-27). Beverly Hills: Sage. Bograd, M. (1988b). How battered women and abusive men account for domestic violence: Excuses, justifications, or explanations? In G.T. Hotaling, D. Finkelhor, J.T. Kirkpatrick & M.A. Strauss (Eds.), Coping with family violence: Research and policy perspectives (pp. 60-75). California, U.S.A.: Sage. Bohan, J.S. (1997). Regarding Gender: Essentialism, Constructionism, and Feminist Psychology. In M.M. Gergen & S.N. Davis (Eds.), Toward a new psychology of gender: A reader (pp. 31-47). London: Routledge. Bohan, J.S. (2002). Sex differences and/in the self: Classic themes, feminist variations, postmodern challenges. Psychology of Women Quarterly, 26, 74-88. Boonzaier, F. (2001). Woman abuse: exploring women?s narratives of violence and resistance in Mitchell?s Plain. Unpublished Masters Dissertation, University of Cape Town, South Africa. Boonzaier, F. (2005). Woman abuse in South Africa: A brief contextual analysis. Feminism & Psychology, 15(1), 99-103. Boonzaier, F. & de la Rey, C. (2004). Woman abuse : The construction of gender in women and men?s narratives of violence. South African Journal of Psychology, 34(3), 443-463. Boonzaier, F. & Shefer, T. (2006). Gendered Research. In T. Shefer, F. Boonzaier & P. Kiguwa (Eds.),The gender of psychology (pp.3-11). Cape Town: UCT Press Borochowitz, D. Y. & Eisikovits, Z. (2002). To love violently: Strategies for reconciling love and violence. Violence against Women, 8(4), 476-494. Bowlby, J. (1984). Violence in the family as a disorder of the attachment and caregiving systems. American Journal of Psychoanalysis, 44(1), 9-27. Boyatzis, R.E. (1998). Transforming qualitative data. London: SAGE. 136 Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77-101. Brown, K. (1998). An introduction to sociology (2nd ed.). MA: Blackwell Publishers. Buchbinder, E. & Eisikovits, Z. (2004). Between normality and deviance: The breakdown of batterers? identity following police intervention. Journal of Interpersonal Violence, 19(4), 443-467. Burr, V. (1995). An Introduction to social constructionism. London: Routledge. Carden, A.D. (1994). Wife abuse and the wife abuser: Review and recommendations. The Counseling Psychologist, 22(4), 539-582. Carlson, B. (2005). The most important things learned about violence and trauma in the past 20 years. Journal of Interpersonal Violence, 20(1), 119-126. Carrigan, T., Connell, R.W. & Lee, J. (1985). ?Towards a new sociology of masculinity?, Theory and Society, 14(5), 551-604. Cavanagh, K., Dobash, R.E., Dobash, R.P. & Lewis, R. (2001). ?Remedial work?: Men?s strategic responses to their violence against intimate female partners. Sociology, 35(3), 695-714. Coleman, D.H. & Staus, M.A. (1986). Marital power, conflict, and violence in a nationally representative sample of American couples. Violence and Victims, 1(2), 141- 157. Chodorow, N.J. (1989). Feminism and psychoanalytic theory. Oxford, U.K. Polity Press. Chodorow, N.J. (1995). Gender as a personal and cultural construction. Signs, 20(3), 516- 544. Chodorow, N.J. (1999). The power of feelings: Personal meaning in psychoanalysis, gender and culture. New Haven: Yale University Press. Chodorow, N.J. (1978). The reproduction of mothering: Psychoanalysis and the sociology of gender. California: The University of California Press. Coffey, A. & Atkinson, P. (1996). Making sense of qualitative data: Complementary research strategies. California: SAGE. Cohen, L. & Manion, L. (1989). Research methods in education (3rd ed.). London, England: Routledge. Cohn, E.S. & White, S.O. (1986). Cognitive development versus social learning approaches to studying legal socialisation. Basic and Applied Social Psychology, 7(3), 195-209. Connell, R.W. (2005). Masculinities (2nd ed.). Berkeley: University of California Press. Connell, R.W. (2000). The Men and the Boys. Cambridge: Polity Press. Connell, R.W. (2001). The social organisation of masculinity. In S.M. Whitehead & F.J. Barrett (Eds.), The Masculinities Reader (pp. 30-50). Cambridge: Polity Press 137 Copenhaver, M.M., Lash, S.J. & Eisler, R.M. (2000). Masculine gender-role stress, anger and intimate abusiveness: Implications for men?s relationships. Sex Roles, 42(5- 6), 405-414. Cornwall, A. & Lindisfarne, N. (1994). Introduction. In A. Cornwall & N. Lindisfarne (Eds.), Dislocating masculinity: Comparative ethnographies. London: Routledge. Corvo, K. (2006). Violence, separation, and loss in the families of origin of domestically violent men. Journal of Family Violence, 21(2), 117-125. Craib, I. (1987). Masculinity and Male Dominance. The Sociological Review, 35(4), 721-743. Craib, I. (2001). Psychoanalysis: A critical introduction. MA: Polity Press. Craib, I. (1990). Psychoanalysis and social theory: The limits of sociology. MA: The University of Massachusetts Press. Davidson, R.J., Putnam, K.M. & Larson, C.L. (2000). Dysfunction in the neural circuitry of emotion regulation ? a possible prelude to violence. Science, 289 (5479), 591-594. Davis, S.N. & Gergen, M. (1997). Toward a new psychology of gender: Opening conversations. In M.M. Gergen & S.N. Davis (Eds.), Toward a new psychology of gender: A reader (pp. 1-27). New York: Routledge. Delsol, C., Margolin, G., & John, R. S. (2003). A typology of maritally violent men and correlates of violence in a community sample. Journal of Marriage and the Family, 65, 635-651. Denzin, N.K. (1984a). On understanding emotion. California: Jossey-Bass Inc. Denzin, N.K. (1984b). Toward a phenomenology of domestic, family violence. American Journal of Sociology, 90(3), 483-513. Dimen, M. (1995). The third step: Freud, the feminists, and postmodernism. American Journal of psychoanalysis, 55(4), 303-319. Dinn, W.M. & Harris, C.L. (2000). Neurocognitive function in antisocial personality disorder, Psychiatry Research, 97, 173-190. Dobash, R.E. & Dobash, R. (1979). Violence against wives: A case against patriarchy. New York: The Free Press. Dutton, D.G. (2003). MCMI results for batterers: A response to Gondolf. Journal of Family Violence, 18, 253-255. Dutton, D.G. (1988). Profiling of wife assaulters: Preliminary evidence for a trimodal analysis. Violence and Victims, 3(1), 5-29. Dutton, D.G. (1994). The ecological fallacy. Violence & Victims, 9(2), 125-140. Dutton, D.G. (1995). Trauma symptoms and PTSD-like profiles in perpetrators of intimate abuse. Journal of Traumatic Stress, 8(2), 299-316. Dutton, D.G. & Bodnarchuk, M. (2005). Through a psychological lens: Personality disorder and spouse assault. In D.R. Loeske, R.J. Gelles & M.M. Cavanaugh (Eds.), Current controversies on family violence (pp. 5-18). California: Sage. 138 Dutton, D.G. & Golant, S. (1995). The batterer: A psychological profile. New York: Basic Books. Dutton, D.G., Saunders, K., Starzomski, A. & Bartholomew, K. (1994). Intimacy-anger and insecure attachment as precursors of abuse in intimate relationships. Journal of Applied Social Psychology, 24(15), 1367-1386. Dutton, D.G. & Starmzomski, A.J. (1993). Borderline personality in perpetrators pf psychological and physical abuse. Violence and Victims, 8(4), 327-337. Dutton, D.G. & Strachan, C.E. (1987). Motivational needs for power and spouse- specific assertiveness in assaultive and nonassaultive men. Violence and Victims, 2(3), 145-156. Eagle, G., Haynes, G. & Sibanda, T. (1999). Standpoint methodologies: Marxist, feminist and black scholarship perspectives. In M. Terre Blanche & K. Durrheim (Eds.), Research in practice: Applied methods for the social sciences (pp. 438-451). Cape Town: UCT Press. Egger, S. (1993, June 15-18). Violence and masculinity: A commentary. Paper presented at Second National Conference on Violence convened by the Australian Institute of Criminology, Canberra, Australia. Retrieved August 17, 2007, from http://www.aic.gov.au/conferences/ncv2/ Elliot, J. (2005). Narrative and new developments in the social sciences. London: SAGE. Eisikovits, Z., Goldblatt, H. & Winstok, Z. (1999). Partner accounts of intimate violence: Towards a theoretical model. Families in Society, 80(6), 606-619. Eisikovits, Z. & Buchbinder, E. (1999). Talking control. Violence Against Women, 5(8), 845-868. Abstract retrieved March 31, 2006, from SAGE Criminology database. Eisikovits, Z.& Buchbinder, E. (1997). Talking violent. Violence Against Women, 3(5), 482-498. Eisikovits, Z. & Winstok, Z. (2002). Reconstructing intimate partner violence: The structure and content of recollections of violent events. Qualitative Health Research, 12(5), 685-699. Evans, S. (2005). Beyond gender: Class, poverty and domestic violence. Australian Social Work, 58(1), 36-43. Ezzy, D. (1998). Theorizing narrative identity: Symbolic interactionism and hermeneutics. The Sociological Quarterly, 39(2), 239-252. Ezzy, D. (2002). Qualitative Analysis. London: Routledge. Fife-Schaw, C. (1995). Surveys and sampling issues. In G.M. Breakwell, S. Hammond & C. Fife-Schaw (Eds.), Research Methods in Psychology (99-115). London: Sage. Flax, J. (1993). Disputed essays on psychoanalysis, subjects politics and philosophy. New York: Routledge. Flick, U. (1998). An introduction to qualitative research. London: Sage. 139 Fonagy, P. (1999). Male perpetrators of violence against women: An attachment theory perspective. Journal of Applied Psychoanalytic Studies, 1, 7-27. Fonagy, P. (2001, March). The psychoanalysis of violence. Paper presented at the Dallas Society for Psychoanalytic Psychology (DSPP) professional seminar ?Preventing Mass Murder in Schools: Understanding Violent Children from ?Peaceful? Families?, Dallas, U.S.A. Retrieved May, 4 2007, from DSPP website: www.dspp.com/papers/fonagy4.htm Fonagy, P., Moran, G.S. & Target, M. (1993). Aggression and the psychological self. International Journal of Psychoanalysis, 74, 471-485. Flournoy, P.S. & Wilson, G.L. (1991). Brief research report: Assessment of MMPI profiles of male batterers. Violence and Victims, 6(4), 309-320. Fuller, N. (2001). She made me go out of my mind: Marital violence from the male point of view. Development, 44(3), 25-29. Freedman, J. & Combs, G. (1996). The narrative metaphor and social constructionism: A postmodern worldview. New York: W.W. Norton & Company. Frosh, S. (1997). Masculinity, rationality, and family therapy. Systems Practice, 10(2), 179-190. Fuss, D. (1989). Essentially speaking. London: Routledge. Gadd, D. (2004). Making sense of interviewee ? interviewer dynamics in narratives about violence in intimate relationships. International Journal of Social Research Methodology, 7(5), 383-401. Gadd, D. (2000). Masculinities, violence and defended psychosocial subjects. Theoretical Criminology, 4(4), 429-449. Gadd, D. (2003). Reading between the lines: Subjectivity and men?s violence. Men and Masculinities, 5(4), 333-354. Gadd, D. (2006). The role of recognition in the desistance process: A case analysis of a former far-right activist. Theoretical Criminology, 10(2), 179-202. Gelles, R.J. (1985). Family violence. Annual Review of Sociology, 11, 347-367. Gelles, R.J. (1999). Male Offenders: Our Understanding from the Data. In M. Harway, & J.M. O?Neil (Eds.), What causes men?s violence against women (pp. 36-48). London: Sage. Goldner, V. (1999). Morality and multiplicity: Perspectives on the treatment of violence in intimate life. Journal of Marital and Family Therapy, 25(3), 325-336. Goldner, V., Penn, P., Sheinberg, M., & Walker, G. (1990). Love and violence: Gender paradoxes in volatile attachments. In M. Gergen & S. Davies (Eds.), Toward a new psychology of gender (pp.576-604). London: Routledge. Gondolf, E.W. (1987). Changing men who batter: A developmental model for integrated interventions. Journal of Family Violence, 2(4), 335-349. Gondolf, E.W. (1996). Characteristics of batterers in a multi-site evaluation of batterer intervention systems. Retrieved February 7, 2006, from Minnesota State University, 140 Minnesota Center Against Violence and Abuse website: http://www.mincava.umn.edu/documents/gondolf/batchar.html Gondolf, E.W. (1999). MCMI results for batterer program participants in four cities: Less ?pathological? than expected. Journal of Family Violence, 14, 1-17. Gondolf, E.W. (2001). MCMI results for batterers: Gondolf replies to Dutton?s response. Journal of Family Violence, 18(6), 387-389. Gondolf, E.W. (1988). Who are those guys?: Toward a behavioural typology of batterers. Violence and Victims, 3(3), 187-203. Goodrum, S., Umberson, D., & Anderson, K.L. (2001). The batterer?s view of the self and others in domestic violence. Sociological Inquiry, 71(2), 221-240. Gottman, J.M., Jacobson, N.S., Rushe, R.H., Shortt, J.W., Babcock, J., La Taillade, J.J. & Walts, J. (1995). The relationship between heart rate reactivity, emotionally aggressive behavior, and general violence in batterers. Journal of Family Psychology, 9, 227-248. Abstract retrieved June 4, 2007, from ScienceDirect database. Grafman, J., Schwab, K., Warden, D., Pridgen, A., Brown, H.R. & Salazar, A.M. (1996). Frontal lobe injuries, violence, and aggression. Neurology, 46, 1231. Abstract retrieved September 29, 2007, from SpringerLink database. Greene, A. F. (1999). Biological perspectives on violence against women. In M. Harway, & J.M. O?Neil (Eds.), What Causes Men?s Violence Against Women (pp. 51- 60). London, U.K.: Sage. Greig, A. (2001). Political connections: Men, Gender and Violence. Retrieved May 19, 2007, from United Nations International Research and Training Institute for the Advancement of Women website: http://toolkit.endabuse.org/Resources/PoliticalConnections.html Guba, E.G. & Lincoln, Y. (1983). Epistemological and methodological bases of naturalistic inquiry. In G.F. Madaus, M. Scriven & D.L. Stuffelbearn (Eds.), Evaluation Models: Viewpoints on Educational and Human Services Evaluation (pp. 311-333). Boston/Dordrecht/London: Kluwer Academic Publishers. Hagemann-White, C. & Micus, C. (1999, October). Explaining the inclination to use violence against women. Paper presented to the Seminar on ?Violence Against Women?, Strasbourg, Germany. Retrieved June 7, 2006, http://www.eurowrc.org/13.institutions/3.coe/en-violence-coe/06.en-coe-oct99.htm Hall, T.W. (2007). Psychoanalysis, attachment, and spirituality part II: The spiritual stories we live by. Journal of Psychology and Theology, 35(1), 29-42. Hamberger, L.K., Lohr, J.M., Bonge, D. & Tolin, D.F. (1996). A large sample empirical typology of male spouse abusers and its relationship to dimensions of abuse. Violence and Victims, 11(4), 277-292. Handrahan, L.M. (1999). Gender theory. Feminista: Journal of Feminist Construction, 3(3). Retrieved May 19, 2007, from http://www.feminista.com/archives/v3n3/handrahan.html Hansen, M., & Harway, M. (Eds.). (1993). Battering and family therapy: a feminist perspective. California: Sage Publications Inc. 141 Harcourt, W. (1994). Feminism, body, self: Third-generation feminism. In J.H. Smith & A.M. Mahfouz (Eds.), Psychoanalysis, feminism and the future of gender (pp. 70-90). Baltimore: The John Hopkins University Press. Hare, R.D. (1993). Without conscience : The disturbing world of the psychopaths among us. New York: Guilford Press. Harway, M. & O?Neil, J.M. (1999). What causes men to be violent against women? In M. Harway, & J.M. O?Neil (Eds.), What causes men?s violence against women (pp. 5- 11). London: Sage. Hastings, J.E. & Hamberger, L.K. (1988). Personality characteristics of spouse abusers: A controlled comparison. Violence and Victims, 3(1), 31-48. Healey, K., Smith, C. & O?Sullivan, C. (1998). Batterer intervention: Program approaches and criminal justice strategies. Retrieved May 28, 2007 from the United States National Criminal Justice Reference Service website: http://www.ncjrs.gov/App/Publications/abstract.aspx?ID=168638 Hearn, J. (2004). From hegemonic masculinity to the hegemony of men. Feminist Theory, 5 (49), 49-72. Hearn, J. (1998). The violences of men. London: Sage. Hearn, J. (1999). The violences of men: Men doing, talking and responding to violence against known women. Paper presented to the 7th International Interdisciplinary Congress on Women, Tromso, Norway. Retrieved January 19, 2006, from www.europrofem.org/contri/2_04_en/en-viol/Hearn_Jeff2.pdf Hearn, J. & Whitehead, A. (2006). Collateral damage: Men?s ?domestic? violence to women seen through men?s relations with men. The Journal of Community and Criminal Justice, 53(1), 38-56. Hollway, W. (1984). Gender difference and the production of subjectivity. In J. Henriques, W. Hollway, C. Urwin, C. Venn & V. Walkerdine (Eds.), Changing the subject: Psychology, social regulation and subjectivity (pp. 227-263). London: Methuen & Co. Hollway, W. & Jefferson, T. (2000). Doing qualitative research differently: Free association, narrative and the interview method. London: SAGE. Holtworth-Munroe, A. (1992). Attributions and martially violent men: The role of cognitions in marital violence. In J.H. Harvey, T.L. Orbuch & A.L. Weber (Eds.), Attributions, Accounts and Close Relationships (pp.165-175). New York: Springer- Verlag. Holtworth-Munroe, A. & Hutchinson, G. (1993). Attributing negative intent to wife behaviour: The attributions of martially violent versus non-violent men. Journal of Abnormal Psychology, 102(2), 206-211. Holtworth-Munroe, A. & Stuart, G.L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, 116(3), 476-497. Hook, D. (2006). Lacan, the meaning of the phallus and the ?sexed? subject. In T. Shefer, F. Boonzaier & P. Kiguwa (Eds.), The Gender of Psychology (pp.60-84). Cape Town: UCT Press. 142 Hyden, M. (1994). Woman battering as a marital act: Interviewing and analysis in context. In C.K. Riessman (Ed.), Qualitative Studies in Social Work Research (pp. 95- 112). California, U.S.A.: SAGE. Hyden, M. & McCarthy, I.M. (1994). Woman battering and father-daughter incest disclosure discourses of denial and acknowledgement. Discourse & Society, 5(4), 543- 565. Jacobson, N. & Gottman, J. (1998). Breaking the cycle: New insights into violent relationships. London: Bloomsbury Publishing. Jakupcak, M., Tull, M.T. & Romer, L. (2005). Masculinity, shame, and fear of emotions as predictors of men?s expressions of anger and hostility. Psychology of Men & Masculinity, 6 (4), 275?284. James, K., Sneddon, B. & Brown, J. (2002). ?Using it? or ?losing it?: Men?s constructions of their violence towards female partners. Retrieved February 7, 2006, from Minnesota State University, Minnesota Center Against Violence and Abuse website: http://www.austdvclearinghouse.unsw.edu.au/Occasional/James_et_al_research_pape r_final.pdf Jewkes, R. (2002). Intimate partner violence: Causes and prevention. Lancet, 359 (9315), 1423-1429. Jewkes, (2001). Reflections on gender violence in the South African public health agenda. Development, 44(3), 64-68. Jewkes, R., Penn-Kekana, L., Levin, J., Ratsaka, M. & Schreiber, M. (2001). Prevalence of emotional, physical, and sexual abuse of women in three South African provinces. South African Medical Journal, 91, 421-428. Jukes, A.E. (1999). Men Who Batter Women. London: Routledge. Jacobson, N. & Gottman, I. (1998). Breaking the cycle: New insights into violent relationships. London: Bloomsbury Publishing. Kalmuss, D. (1984). The intergenerational transmission of marital aggression. Journal of Marriage and the Family, 46, 11-19. Kanazawa, S. (2006). Violent men have more sons: Further evidence for the generalized Trivers-Willard hypothesis (gTWH). Journal of Theoretical Biology, 239, 450-459. Kandiyoti, D. (1994). The paradoxes of masculinity: Some thoughts on segregated societies. In A. Cornwall & N. Lindisfarne (Eds.), Dislocating Masculinity: Comparative ethnographies (pp. 197-213). London: Routledge. Kaufman, K. (1994). Cracking the armor: Power, pain and the lives of men. Toronto: Viking Canada. Kelly, K. (1999). Hermeneutics in action: empathy and interpretation in qualitative research. In M.Terre Blanche & K. Durrheim (Eds.), Research in practice : applied methods for the social sciences (pp. 398-419). Cape Town: University of Cape Town Press. Kelly, K. (2006). Lived experience and interpretation: The balancing act in qualitative analysis. In M. Terre Blanche, K. Durrheim & D. Painter (Eds.), Research in practice: 143 applied methods for the social sciences (2nd ed.), (pp. 345-369). UCT: University of Cape Town Press. Kohlberg, L. (1981). The philosophy of moral development. San Francisco: Harper and Row. Kohut, H. (1972). Thoughts on narcissism and narcissistic rage. Psychoanalytic Study of the Child, 27, 360-400. Kohut, H. & Wolf, E.S. (1978). The Disorders of the Self and Their Treatment: An Outline (pp.359-385). In P.H. Ornstein (Ed.), The search for the self: Selected writings of Heinz Kohut: 1978-1981 (Vol.3). Connecticut: International Universities Press. Ko Ling, C. (2001). An Evaluative Study of Group Therapy for Male Batterers Cum Intervention Strategies. Unpublished doctoral thesis, The University of Hong Kong, PROC. Kvale, S. (in press). The Dominance of Dialogical Interview Research. Qualitative Inquiry. Retrieved May, 1, 2007, from http://www.c4qi.org/qi2006web/archivepaper/Paper%203489.pdf Lansky, M.R. (1987). Shame and domestic violence. In D.L. Nathanson (Ed.), The many faces of shame. New York: Guilford Press. Lawson, D.M. (2001). The development of abusive personality: A trauma response. Journal of Counseling and Development, 79(4), 505-509. Leach, M. (1994). The politics of masculinity: An overview of contemporary theory. Social Alternatives, 12(4), pp. 36-37. Legard, R., Keegan, J. & Ward, K. (2003). In-depth interviews. In J. Ritchie & J. Lewis (Eds.). Qualitative research practice: A guide for social science students and researchers (pp. 139-169). London: SAGE. Lewis, H.B. (1971). Shame and guilt in neurosis. New York: International Universities Press. Lewis, J. (2003). Design Issues. In J. Ritchie & J. Lewis (Eds.), Qualitative research practice: A guide for social science students and researchers (pp. 47-75). London: SAGE. Loeske, D.R. (2005). Through a sociological lens: The complexities of family violence. In D.R. Loeske, R.J. Gelles & M.M. Cavanaugh (Eds.), Current controversies on family violence (pp. 35-47). California: Sage. Loeske, D.R., Gelles, R.J. & Cavanaugh, M.M. (2005). Current controversies on family violence. California: Sage. MacEwen, K.E. & Barling, J. (1988). Multiple stressors, violence in the family of origin, and marital aggression: A longitudinal investigation. Journal of Family Violence, 3(1), 73-87. Abstract retrieved May 30, 2007, from SpringerLink database. Maracek, J., Crawford, M. & Popp, D. (2004). On the construction of gender, sex, and sexualities. In A.H. Eagly, A.E. Beall & R.J. Sternberg (Eds.), The psychology of gender (2nd ed.) (pp. 192-216). New York: Guilford Press. 144 Marin, A.J. & Russo, N.F. (1999). Feminist perspectives on male violence against women. In M. Harway, & J.M. O?Neil (Eds.), What causes men?s violence against women (pp. 19-35). London: Sage. Marshall, C. & Rossman, G.B. (1995). Designing qualitative research (2nd ed.). California: SAGE. Mazur, A. & Booth, A. (1997, in press). Testosterone and dominance in men. Behavioural and Brain Sciences. Cambridge University Press. Retrieved May 18, 2007, from http://cogprints.org/663/00/bbs_mazur.html Smith M.D. (1990). Patriarchal ideology and wife beating: A test of a feminist hypothesis. Violence and Victims, 5(4), 257-273. Mathews, S., Abrahams, N., Martin, L.J., Vetten, L., van der Merwe, L. & Jewkes, R. (2004). ?Every six hours a woman is killed by her intimate partner?: A national study of female homicide in South Africa. MRC Policy Brief, 5. Cape Town: Medical Research Council. Mazur, A. & Booth, A. (1998). Testosterone and dominance in men. Behavioral and Brain Sciences, 21, 353?397. Mentis, H. M. (n.d.). Insight into Strong Emotional Experiences through Memory. Retrieved May 30, 2007, from Human-Computer Interfaces Group: http://www.sics.se/~kia/evaluating_affective_interfaces/Mentis.pdf Melzer, S.A. (2002). Gender, work, and intimate violence: Men?s occupational violence spillover and compensatory violence. Journal of Marriage and Family, 64 (4), 820-832. Miller, P.A. & Eisenberg, N. (1988). The relation of empathy to aggressive and externalizing/antisocial behaviour. Psychological Bulletin, 103(3), 324-344. Moffett, H. (2007). ?These women, they force us to rape them?: Rape as narrative of social control in post-apartheid South Africa. Journal of Southern African Studies, 32(1), 129-144. Moore, T.M. & Stuart, G.L. (2005). A Review of the Literature on Masculinity and Partner Violence, Psychology of Men & Masculinity, 6(1), 46?61. Muller, U. (1999, October). Explaining the inclination to use violence against women. Paper presented to the Seminar on ?Violence Against Women?, Strasbourg, Germany. Retrieved June 7, 2006, from http://www.eurowrc.org/13.institutions/3.coe/en-violence- coe/06.en-coe-oct99.htm Neuman, W.L. (1997). Social research methods: Qualitative and quantitative approaches (3rd ed.). Boston: Allyn and Bacon. Nightingale, D.J., Cromby, J. (1999). Social constructionist psychology: A critical analysis of theory and practice. Buckingham: Open University Press. O?Hearn, H.G. & Margolin, G. (2000). Men?s attitudes condoning marital aggression: A moderator between family of origin abuse and aggression against female partners. Cognitive Therapy and Research, 24(2), 159-174. Owens, D.J. & Strauss, M.A. (1975). The social structure of violence in childhood and approval of violence as an adult. Aggressive Behavior, 1, 193-211. 145 Padayachee, V. (2005). Offender rehabilitation and reintegration ? NICRO?s programme for perpetrators of intimate partner violence. Paper presented at the Criminal Justice Conference, Cape Town, South Africa, Retrieved March 5, 2006 from http://www.csvr.org.za/confpaps/padayachee.html Raine, A., and Scerbo, A. (1991). Biological theories of violence. In J.S. Milner (Ed.), Neuropsychology of Aggression (pp. 1-26). Boston: Kluwer Academic. Petrik, N.D., Olson, R.E.P. & Subotnik, L.S. (1994). Powerlessness and the need to control: The male abuser?s dilemma. Journal of Interpersonal Violence, 9(2), 278-285. Pratto, F. & Walker, A. (2004). The bases of gendered power. In A.H. Eagly, A.E. Beall & R.J. Sternberg (Eds.), The psychology of gender (2nd ed.) (pp. 242-268). New York: Guilford Press. Presser, L. (2003). Remorse and neutralization among violent male offenders. Criminal Justice Quarterly, 20(4), 801-825. Abstract retrieved from IngentaConnect database. Ptacek, J. (1997). The tactics and strategies of men who batter. In A.P. Cardarelli (Ed.), Violence between intimate partners: Patterns, causes and effects. Boston: Allyn and Bacon. Ptacek, J. (1988). Why do men batter their wives? In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse (pp. 133-157). California: Sage. Putallaz, M., Costanzo, P.R., Grimes, C.L. & Sherman, D.M. (1998). Intergenerational continuities and their influences on children?s social development. Social Development, 7(3), 389-427. Reitz, R. R. (1999). Batterers? experiences of being violent: A phenomenological study: Psychology of Women Quarterly, 23, 143-165. Retzinger, S.M. (1991). Violent emotions: Shame and rage in marital quarrels. Newbury Park, California: Sage. Riessman, R.K. (1993). Narrative Analysis. California: Sage. Ritchie, J., Lewis, J. & Elam, G. (2003). Designing and selecting samples. In J. Ritchie & J. Lewis (Eds.), Qualitative research practice: A guide for social science students and Researchers (pp. 77-107). London: SAGE. Rose, P., Beeby, J. & Parker, D. (1995). Academic rigour in the lived experience of researchers using phenomenological methods in nursing. Journal of Advanced Nursing, 21, 1123-1129. Rosen, K.H., Stith, S.M., Few, A.L. Daly, K.L. & Tritt, D.R. (2005). A qualitative investigation of Johnson?s typology. Violence and victims, 20(3), 319-334. Rosenbaum, A. & Leisring, P.A. (2003). Beyond power and control: Towards an understanding of partner abusive men. Journal of Comparative Family Studies, 34(1), 7-21. Rosenbaum, A. & O?Leary, K.D. (1981). Marital violence: Characteristics of abusive couples. Journal of Family Violence, 3, 91-104. Rossman, G.B. & Rallis, S.F. (2003). Learning in the field: An introduction to qualitative research. California: Sage. 146 Rupesinghe, K. (1994). The Culture of Violence. Tokyo: The United Nations University. Russell, M.N. (1995). Confronting abusive beliefs: Group treatment for abusive men. London: Sage. Ryan, G.W. & Bernard, H.R. (2003). Techniques to identify themes. Field Methods, 15(1), 85-109. Saunders, D.G. & Hamill, R.M. (2003). Violence Against Women: Synthesis of Research on Offender Interventions. Retrieved March 14, 2006, from University of Michigan School of Social Work website: http://www.ssw.umich.edu/research/saunddan/NijSynthNCJ201222.pdf Seale, C., Gobo, G., Gubrium, J.F. & Silverman, D. (2004). Introduction: Inside qualitative research. In C. Seale, G. Gobo, J.F. Gubrium & D. Silverman (Eds.), Qualitative research practice. London: Sage. Scheff, T.J. (1997). Emotions, the social bond and human reality: Part/whole analysis. Cambridge: Cambridge University Press. Scheff, T.J. (2006). Hypermasculinity and violence as a social system. Retrieved July, 13, 2006, from Human Dignity and Humiliation Studies website: http://www.humiliationstudies.org/documents/ScheffHypermasculinityandViolence.pdf. Scheff, T.J. & Retzinger, S.M. (1991). Emotions and violence: Shame and rage in destructive conflicts. MA: Free Press. Schlenker, B. R. (1980). Impression management: The self-concept, social identity, and interpersonal relations. Monterey: Brooks/Cole. Schmidt, M., Kolodinsky, J., Carsten, G., Schmidt, F., Larson, M. & MacLachlan, C. (2007). Short term change in attitude and motivating factors to change abusive behaviour of male batterers after participating in group intervention program based on feminist and cognitive-behavioural approach. Journal of Family Violence, 22(2), 91- 100. Abstract retrieved from IngentConnect database. Schwandt, T.A. (1994). Constructivist, interpretivist approaches to human inquiry. In N.K. Denzin & Y.S. Lincoln (Eds.), Handbook of Qualitative Research (pp. 118-137) California: Sage. Shefer, T. & Potgieter, C. (2006). Sexualities. In T. Shefer, F. Boonzaier & P. Kiguwa (Eds.), The gender of psychology (pp. 103-120). Cape Town: UCT Press. Shefer, T. & Ratele, K. (2006). New questions old answers? Gender and race in post- apartheid South Africa. In G. Stevens, V. Franchi, and T. Swart (Eds.). A Race Against Time ? Psychology and Challenges to Deracialisation in South Africa (pp. 229-244). Pretoria: University of South Africa. Scully, D. & Marolla, J. (1984). Convicted rapists? vocabularies of motive: Excuses and justifications. Social Problems, 31(5), 530-544. Seidler, V.J. (1994). Unreasonable Men: Masculinity and social theory. London: Routledge. Smith, M.D. (1990). Patriarchal ideology and wife beating: A test of a feminist hypothesis. Violence and Victims, 5(4), 257-273. 147 Smythe, W.E. & Murray, M.J. (2000). Owning the story: Ethical considerations in narrative research. Ethics & Behavior, 10(4), 311-336. Snape, D. & Spencer, L. (2003). The Foundations of Qualitative Research. In J. Ritchie & J. Lewis (Eds.), Qualitative research practice: A guide for social science students and researchers (pp. 1-23). London: Sage. Stahl, B.C. (n.d.). A critical review of the ethical nature of interpretive research: Paul Ricoeur and the other. Retrieved March, 11, 2005, from http://is2.lse.ac.uk/asp/aspecis/20050002.pdf Stamp, G.H. & Sabourin, T.C. (1995). Accounting for violence: An analysis of male spousal abuse narratives. Journal of Applied Communication Research, 23, 284-307. Sugarman, D.B. & Frankel, S. (1996). Patriarchal ideology and wife-assault: A meta- analytic review. Journal of Family Violence, 11(1), 13-40. Swart, L., Seedat, M., Stevens, G. & Ricardo, I. (2002). Violence in adolescents? romantic relationships: Findings from a survey amongst school-going youth in a South African community. Journal of Adolescence, 25, 385-395. Tangney, J.P. (1991). Moral affect: The good, the bad, and the ugly. Journal of Personality and Social Psychology, 61(4), 598-607. Terre Blanche, M. & Durrheim, K. (1999). Social Constructionist Methods. In M.Terre Blanche & K. Durrheim (Eds.), Research in practice : Applied methods for the social sciences (pp. 147-172). Cape Town: University of Cape Town Press. Terre Blanche, M. & Kelly, K. (1999). Interpretive Methods. In M.Terre Blanche & K. Durrheim (Eds.), Research in practice : Applied methods for the social sciences (pp. 123-146). Cape Town: University of Cape Town Press. Terre Blanche, M., Kelly, K. & Durrheim, K. (2006). Why qualitative research? In M. Terre Blanche, K. Durrheim & D. Painter (Eds.), Research in practice : Applied methods for the social sciences (2nd ed.) (pp. 271-284). Cape Town: University of Cape Town Press. Tesser, A. (1991). Social versus clinical approaches to self psychology: The self- evaluation maintenance model and Kohutian object relations theory. In R.C. Curtis (Ed.), The relational self: Theoretical convergences in psychoanalysis and social psychology (pp. 257-281). New York: Guilford Press. Thomas, S.P. (2003). Men?s anger: A phenomenological exploration of its meaning in a middle-class sample of American men. Psychology of Men & Masculinity, 4(2), 163- 175. Thorne, B. (1996). Person-centered therapy. In W. Dryden (Ed.), Handbook of Individual Therapy (pp. 121-146). London: Sage. Thorne, S. (1997). Completing a qualitative project: Details and dialogue. Thousand Oaks: Sage. Totten, M. (2003). Girlfriend abuse as a form of masculinity construction among violent, marginal male youth. Men and Masculinities, 6(1), 70-92. 148 Umberson, D., Anderson, K.L., Williams, K. & Chen, M.D. (2003). Relationship dynamics, emotion state, and domestic violence: A stress and masculinities perspective. Journal of Marriage and Family, 65 (1), 233- 247. Umberson, D., Williams, K. & Anderson, K. (2002). Violent behaviour: A measure of emotional upset? Journal of Health and Social Behaviour, 43, 189-206. Vetten, L. (1996). ?Man shoots wife?: Intimate femicide in Gauteng, South Africa. Crime and Conflict, 6, 1-4, Winter. Vetten, L. (2005, 17-20 May). Addressing domestic violence in South Africa: Reflections on strategy and practice. Paper presented at the Expert Group Meeting on ?Violence against women: Good practices in combating and eliminating violence against women?, Vienna, Austria. Retrieved March, 12, from http://www.un.org/womenwatch/daw/egm/vaw-gp-2005/docs/experts/vetten.vaw.pdf Walker, L.E. (1984). The battered woman syndrome. New York: Springer. Walton, C., Coyle, A. & Lyons, E. (2004). Death and football: an analysis of men's talk about emotions. British Journal of Social Psychology, 43(Pt 3), 401-16. Abstract retrieved March 7, 2007 from IngentaConnect database. Waltz, J., Babcock, J.C., Jacobson, N.S. & Gottman, J.M. (2000). Testing a typology of batterers. Journal of Consulting and Clinical Psychology, 68(4), 658-669. Weedon, C. (1987). Feminist practice and poststructuralist theory. Oxford: Basil Blackwell. Wetherell, M. & Edley, N. (1999). Negotiating hegemonic masculinity: Imaginary positions and psycho-discursive practices. Feminism and Psychology, 9(3), 335-56. Wexler, D.B. (1999). The broken mirror: A self psychological treatment perspective for relationship violence. Journal of Psychotherapeutic Practice & Research, 8(2), 129- 141. Whitehead, S.M. (2001). Man: The invisible gendered subject? In S.M. Whitehead & F.J. Barrett (Eds.), The masculinities reader (pp. 351-368). Cambridge: Polity Press. Whitehead, S.M. & Barrett, F.J. (2001). The sociology of masculinity. In S.M. Whitehead & F.J. Barrett (Eds.), The masculinities reader (pp. 1-26). Cambridge: Polity Press. Williamson, J.B., Karp, D.A. & Dalphin, J.R. (1977). The research craft: An introduction to social science methods. Toronto: Little, Brown and Company. Willig, C. (2001). Introducing qualitative research in psychology: Adventures in theory and method. Buckingham: Open University Press. Winstok, Z., Eisikovits, Z. & Gelles, R. (2002). Structure and dynamics of escalation from the batterer?s perspective. Families in Society: The Journal of Contemporary Human Services, 83(2), 129-141. Witt, D.D. (1987). A conflict theory of family violence. Journal of Family Violence, 2(4), 291-301. Wood, J.T. (2004). Monsters and victims: Male felons? accounts of intimate partner violence. Journal of Social and Personal Relationships, 21(5), 555-576. 149 Wood, K. & Jewkes, R. (2001). ?Dangerous? love: Reflections on violence among Xhosa township youth. In R. Morrell (Ed.), Changing men in Southern Africa (pp. 217- 336). Pietermaritzburg: University of Natal Press. Wood, K. & Jewkes, R. (1997). Violence, rape and sexual coercion. Gender & Development, 5(2), 41-46. Wood, K. & Jewkes, R. (2001). Violence, rape, and sexual coercion: Everyday love in a South African township. In S.M. Whitehead & F.J. Barrett (Eds.), The masculinities reader (pp. 133-140). Cambridge: Polity Press. Worley, K.O., Walsh, S. & Lewis, K. (2004). An examination of parenting experiences in male perpetrators of domestic violence: A qualitative study. Psychology and Psychotherapy: Theory, Research and Practice, 77(1), 35-54. Umberson, D., Anderson, K.L., Williams, K. & Chen, M.D. (2003). Relationship dynamics, emotion state and domestic violence: A stress and masculinities perspective. Journal of Marriage and the Family, 65(1), 233-247. Umberson, D., Williams, K. & Anderson, K. (2002). Violent behaviour: A measure of emotional upset? Journal of Health and Social Behaviour, 43(2), 189-207. Yates, C. (2000). Masculinity and good enough jealousy. Psychoanalytic Studies, 2(1), 77-88. Yllo, K. (1984). The status of women, marital equality, and violence against wives. Journal of Family Issues, 5, 307-320. Yllo, K.A. (2005). Through a feminist lens: Gender, Diversity, and Violence: Extending the feminist framework. D.R. Loeske, R.J. Gelles & M.M. Cavanaugh (Eds.), Current controversies on family violence (pp. 19-34). California: Sage. Zosky, D.L. (1999). The application of object relations theory to domestic violence. Clinical Social Work Journal, 27(1), 55-69. 150 APPENDIX A: ORGANISATION INFORMATION SHEET School of Human and Community Development Psychology Department Private Bag 3 Wits 2050 Johannesburg Tel: (011) 717-4541 Email: lauu2@students.wits.ac.za My name is Ursula Lau, and I am conducting research for the purposes of obtaining Masters Degree at the University of the Witwatersrand. The aim of this research is to understand the experiences of men who have acted violently against their intimate partners or spouses. In South Africa, violence against women is recognised as a pervasive problem, affecting women at all levels of society. While intervention and prevention efforts have targeted women victims of intimate partner violence, little has been done by way of understanding the subjective viewpoints of men in an effort to facilitate change. Such insight would allow for more effective ways of assisting men during the rehabilitative process. With your permission, I would like to invite the members of the men?s support group at [_______] to participate in this study. This research project would involve face-to-face in-depth interviews with 7 men lasting approximately 60 to 90 minutes. All participants in this study will be required to provide informed consent. (These are detailed in the attached ?Participant Information Sheet?). All interviews will be audio taped for the purposes of transcription. The information provided by individuals who choose participate in the study will be kept confidential. While texts from the interview material may be quoted verbatim in the research report, the identity of the participant will not be revealed. As a safety measure, information divulged in the interview context which indicates the intention to harm self or another person would need to be reported to a social worker on site. The process of interviewing would take approximately one to two weeks. Immediately following each interview, a follow-up debriefing session by a counselor or social worker at the organisation would be necessary. I am specifically looking for men who are able to converse fluently in English and have voluntarily presented themselves for treatment to participate in the study. Participants must have been involved in a relationship with their female partner for at least 6 months and must have engaged in at least one incident of physical abuse against their partner or spouse in the past year. In order to ensure uniformity and prevent any complications from arising, men who are on parole, as well as those who have legal proceedings against them will not be included in the study. If you choose to grant permission, your contribution would be valuable in providing greater understanding of how men?s subjective views can be used to inform treatment programmes which aim change men?s violent behaviour. Results of the research will be written up and made available to you. A time can be negotiated whereby these findings can be discussed with you in person. Findings of the research will also be made available to the participants of the study. The results of this research may also be published at a future date. Your participation in this study would be very much appreciated. If you choose to grant permission, I will be in contact with you to set up meeting time during which practical considerations for carrying out the research can be negotiated. Enclosed you will find a copy of the research proposal. Your participation in this study would be very much appreciated. Sincerely, Ursula Lau 151 APPENDIX B: PARTICIPANT INFORMATION SHEET School of Human and Community Development Psychology Department Private Bag 3 Wits 2050 Johannesburg Tel: (011) 717-4541 Email: lauu2@students.wits.ac.za My name is Ursula Lau, and I am conducting research for the purposes of obtaining a Degree at the University of the Witwatersrand. The aim of this research is to understand the experiences of men who have acted violently against their intimate partners or spouses. In South Africa, violence against women is recognised as a very serious problem, affecting all kinds of women. While efforts at helping have focused on women victims of violence in their relationships with their partners, little has been done to understand how men experience their violence. This is important in order to help them change their behaviours. This understanding would allow for better ways of helping men during their treatment. I would like to invite you to participate in this study. Participation in this research will involve taking part in a face-to-face interview at a time which is convenient for you. The interviews will take place at the premises of [ ] and will last between 60 to 90 minutes. The interview will involve questions about an experience you had when you acted aggressively towards your partner. Participation is voluntary, and you will not be advantaged or disadvantaged in any way if you choose or choose not take part in this study. If you choose to take part in the interview, you have a choice to stop at any time if you wish. You are also not expected to answer questions you feel are too sensitive. Your responses will be kept anonymous outside of the interview context. While questions of a personal nature will be asked, your responses will remain confidential. With your permission, the interview will be taped to ensure accuracy. The interview material (recordings and transcripts) will be viewed, heard and processed by myself only. Once they have been transcribed, the tapes will be destroyed. The interview transcripts will be stored at the Psychology Department at the university and will be destroyed one year after the research has been written up. Although parts of the interview material may be quoted directly in the write up of the research, you are assured that it will not be matched to your identity in any way. The results of this research may also be published at a future date. As a measure of safety, if you chose to provide information in which you reveal your intention to harm yourself or another person, the researcher has an ethical responsibility to report this to a social worker so that they can help you explore better options. If you choose to take part in this study, a debriefing (follow-up) session with a social worker at [ ] will be set up for you immediately after the interview. This is to give you the opportunity to express any upsetting feelings that may have come up during the interview, which could possibly lead to your acting out in a negative way. If you feel you would like further counselling at a later stage at no cost to you, you can contact ADAPT (Tel: 786-6608) or NICRO (Tel: 403-2953). The researcher is unfortunately unable to provide any counselling. If you choose to participate, your contribution would be valuable in providing a better understanding of how men?s personal views can be used in treatment programmes which aim to change men?s violent behaviour. The results of the research will be written up and delivered by myself to [name of organisation] where it will be made available to you. Your participation in this study would be very much appreciated. If you choose to participate in the study, please fill in your contact details on the form below. I will contact you within two weeks in order to discuss your participation. If you decide to participate at a later stage, you can leave your contact details with the [name of co- ordinator]. Otherwise, I can be contacted via the e-mail address provided. Kind Regards, Ursula Lau 152 APPENDIX C: PARTICIPANT CONTACT AND SCREENING FORM Name: _____________________________ Contact no: ___________________________ Please answer the following questions. These will be kept confidential 1. Are you currently involved in an intimate relationship? (dating, living with your partner or married?) Yes _______ No ________ 2. How long have you been attending the support group at [name of organisation]? ________________ 3. When was the last time a violent incident took place between you and your partner? _________________ 4. How were you referred to [name of organisation]? (Tick one) Self-referred __________ Referred by family/friend __________ Court-referred ___________ Other _________ Please specify 5. Home language ___________________ 6. Are you fluent in English? _______________ 153 APPENDIX D: CONSENT FORM (INTERVIEW) I _____________________________________ consent to being interviewed by Ursula Lau for her study on men?s subjective experiences of their violence toward their intimate female partners. I understand that: - Participation in this interview is voluntary. - That I may refuse to answer any questions I would prefer not to. - I may withdraw from the study at any time. - No information that may identify me will be included in the research report, and my responses will remain confidential. - Information that I provide in the interview may be quoted directly in the research report, but this will not be matched to my identity in any way. - A debriefing (follow-up) counselling session will be set up for me after the interview. This will allow me to express any upset feelings that may have come up during interview, so that there will be no negative consequences for myself or others close to me. Signed __________________________________________ 154 APPENDIX E: CONSENT FORM (RECORDING) I _____________________________________ consent to my interview with Ursula Lau being tape- recorded, which will be used for her study on men?s subjective experiences of their violence toward their intimate female partners I understand that: - The tapes and transcripts will not be seen or heard by any person in this organisation at any time, and will only be processed by the researcher. - All tape recordings will be destroyed once they have been transcribed. - The transcripts will be stored for safekeeping at the Psychology Department at the University of the Witwatersrand following submission of the research for examination purposes. They will then be destroyed one year after this submission date. - No identifying information will be used in the transcripts or the research report. Signed __________________________________________ 155 APPENDIX F: INTERVIEW GUIDE An Exploration of Men?s Subjective Experiences of their Violence toward Intimate Partners OBJECTIVES to explore men?s experience of their most memorable episode of enactment of violence toward their spouse/partner to explore men?s understanding of violence in the context of their relationship with their partner or spouse determine/identify the preceding events which led to the act of violence to explore thoughts and feelings that preceded the act of violence to explore men?s understanding of how emotions are expressed in the act of violence to gather men?s reflections on a particular act of violence INTRODUCTION introduce the study; ethical issues (confidentiality and anonymity, recording or data, how material is used); timing I. PRESENT CIRCUMSTANCES Age Level of education Marital status Home language Area of residence Summary of current activity (work/education/other) Date of joining support group Nature of relationship with current partner/spouse Duration of relationship with current partner/spouse II. MEANING OF VIOLENCE AS DEFINED BY PARTICIPANT Participant?s understanding of violence in an intimate partner relationship How participant describes current relationship with partner/spouse Any acts of violence towards partner at present Nature of violent acts (past or present) Previous relationship(s) where abuse took place III. MOST MEMORABLE EXPERIENCE OF VIOLENT ENACTMENT Description of incident - context of actual incident of violence (time and place) - nature of violence (specific acts/behaviours) - presence of others Events surrounding the context of the incident 156 Mood prior to violent episode - any perceived stressors outside of relationship - use of stimulants/substances IV. SUBJECTIVE EXPERIENCE Thoughts - preceding the incident - during enactment - about partner/spouse at the time - about self at the time Feelings - experienced preceding the encounter - expressed during the act of violence - how these feelings were aroused - feelings about self at the time - feelings about partner/spouse at the time V. V. REFLECTIONS ON THE VIOLENT EPISODE Aspects about the relationship perceived as provoking the violent act Issues relating to self perceived as provoking the violent act How emotions are expressed through violence Why violence was the outcome VI. VI. CONCLUSION: REFLECTIONS ON THEIR CURRENT SITUATION Thoughts and feelings about relationship with partner/spouse Reasons for joining a support group 157 APPENDIX G: TRANSCRIPTION NOTATION P1: participant (number used to distinguish between participants) wanna: want to ?cause; because y?know you know gonna going to .. stutter (.) pause (2) two second pause (number indicates duration) xxx untranscribable (xxx) indistinct/doubtful transcription Word in italics emphasis ? interrupted or new thought pattern /?/ omitted Modified from: Banister, P., Burman, E., Parker, I., Taylor, M. & Tindall, C. (1994). Qualitative Methods in Psychology. Bristol, U.S.A.: Open University Press. 158 APPENDIX H: INTERVIEW TRANSCRIPTS NOTE: For ethical reasons, the interview transcripts have not been included but are available upon request.