The normalization of violence in heterosexual romantic relationships: Women's narratives of love and violence

Julia T. Wood

University of North Carolina at Chapel Hill

Journal of Social and Personal Relationships, 18 (2), 239-261, 2001

ABSTRACT

Inductive analysis of in-depth interviews with 20 heterosexual women who had been in violent romantic relationships illuminated women's use of gender and romance narratives to make sense of violent relationships. All participants placed themselves within western culture's primary gender narrative, which prescribes and normalizes dominance and superiority for men and deference and dependence for women. Participants also relied on romance narratives -which entailed both fairy tale and dark versions -to make sense of violence in their relationships. Interrelated beliefs that emerged in the women's talk functioned to legitimise both fairy tale and dark romance narratives. This study highlights the urgency of weaving alternative gender and romance narratives into the structures and practices of the culture.

KEY WORDS: abuse. gender. narratives. violent relationships

Violence in personal relationships is widespread. In the mid-1980s, the FBI reported that in the USA a man beat a wife or girlfriend every 18 seconds; by 1992, the FBI reported that this happened every 12 seconds (Jones, 1994a). The National Coalition Against Domestic Violence (Hasenauer, 1997) states that at least four million incidents of violence against women by their intimate partners are reported each year and, on average, 10 women a day die at the hands of intimates. These figures are conservative because they summarize only reported violence between intimates; much violence is not reported.

In the vast majority of reported acts of violence between intimates, men are the perpetrators and women the victims. The National Coalition Against Domestic Violence (Hasenauer, 1997) estimates that 95% of reported victims who are physically abused by intimate partners are women. The National Violence Against Women Survey, jointly sponsored by the Centre for Disease Control and Prevention and the Centre for Injury Prevention and Control, found that women experience significantly more partner violence than men: 25% of women and 8% of men in the nationwide survey reported sexual or physical assault from a partner in the previous 12 months (Tjaden & Thoennes, 1998). That survey also found that 76% of women who were raped or otherwise physically assaulted after the age of 18 were attacked by current or former intimate partners.

This article furthers understanding of violent heterosexual romantic relationships by focusing on how women construct the meaning of violence inflicted by their partners. This study highlights the role of culturally endorsed gender and romance narratives in normalizing violence and identifies specific beliefs authorized by these narratives that women use to construct violence as tolerable in romantic relationships.

Review of the literature

Historically, violence against women by intimates has not been viewed as a serious issue (Del Mar, 1996; French, 1992). In the 19708, this began to change, largely as a result of the second wave of feminism in the USA (Atwood & Olsen, 1996; Tjaden & Thoennes, 1998). Heightened awareness has generated greater resources for victims, as well as legislation in many states to criminalize stalking and marital rape and to define violence against partners as a crime against the state that does not require a victim to press charges. A landmark in the struggle to raise awareness of this problem came in 1995 when President Clinton created the Violence Against Women Office.

Misconceptions about violent relationships

Many laypersons mistakenly think that violence against women is both unusual and a product of pathological personalities or relationships. The number one myth about violence between intimates is that it is rare (National Coalition Against Domestic Violence, 1999). Almost equally prevalent is the myth that anyone who commits violence against an intimate is psychologically unbalanced. Contrary to lay conceptions, most researchers and clinicians assert that violent relationships are not atypical (Bachman & Saltzman, 1995; Barnett & LaViolette, 1993; Caputi, 1993; Goldner, Penn, Sheinberg, & Walker, 1990). Although we recoil from labelling violence between intimates as normal, its frequency renders the adjective disturbingly appropriate. Years of clinical work led Jacobson and Gurman (1986) to conclude that a certain amount of violence in marriage is so common that it is 'normal' in the sense of adhering to norms.

Researchers and clinicians point out that many men who are violent toward romantic partners are normal on measures of mental stability, social adjustment, and other standard clinical criteria (Goldner et al., 1990; Jones, 1994a,b; Stoltenberg, 1989; Tjaden & Thoennes, 1998). Women who remain in violent relationships also are not demonstrably atypical of women in general (Atwood & Olsen, 1996; Goldner et al., 1990; Jacobson & Gottman, 1998).

Whether in the streets between strangers or in the home between intimate partners, the meaning and acceptability of violence are sculpted by cultural contexts. Although particular acts of violence arguably may result from psychological variables and/or unique situational factors, widespread violence defies individualistic explanations (French, 1992; West, 1995). It exists and continues to exist only if in some sense it is legitimated by social discourses that shape individuals' behaviours and modes of interpreting their own and others' actions (Cuklanz, 1996; Meyers, 1997). Writing in 1996, Del Mar stated that 'violence against wives will remain commonplace until we muster the will to examine how closely it is bound up with some of our most cherished values and most powerful cultural traits' (p. 174). Discourses that are available at any moment in cultural life constrain the options open to individuals for making sense of threats and acts of violence, as well as to the relationships in which they occur. Culture provides vocabularies of understanding, motive, and significance on which individuals rely, however unconsciously, to construct the meaning of 'personal' experience.

Narrative approaches to understanding violent relationships

Narrative theories offer insight into the cultural authorization of violence and women's toleration of it in romantic relationships. Although the literature on narrative approaches is vast and increasingly specialized, only three premises from narrative work are relevant to the current study.

Humans rely on narratives to make sense of their lives. The bedrock assumption of narrative approaches is that humans make sense of themselves through stories, or narratives (Fisher, 1987; Gergen, 1997; Shotter, 1993). In advancing the narrative paradigm, Fisher (1987) claimed that humans 'experience and comprehend life as a series of ongoing narratives' (p. 24). To make our lives coherent to others and ourselves, we rely on narratives, which Fisher defines as 'symbolic actions. . . that have sequence and meaning for those who live, create, or interpret them' (1987, p. 58). When we place ourselves within some narratives (and not others), we confer structure, sequence, and coherence on experiences that would otherwise be fragmentary and inchoate. In so doing, we create and recreate our identities and our lives (Gergen, 1997; Shotter, 1993).

Narratives are social in nature. Narratives are not strictly personal accounts, or stories; instead, they are decisively social, which is to say culturally constructed, sustained, reproduced, and sometimes altered (Berger, 1996; Shotter, 1993). MacIntyre (1984) notes that 'The story of my life is always embedded in the stories of those communities from which I derive my identity' (p. 221). Concurring, Frank (1995) observes that each person's story 'is embodied in a specific person, but it is equally social' (p. 18). In Realities and Relationships, Gergen (1997) emphasizes the communal origins of narrative meanings, insisting that social legitimation and reproduction of narratives imbues them with power and endurance.

Within western cultures, romance and gender narratives are imbricated, and they are pervasive. Children's fairy tales provide early tutelage in the central romance narrative in which Prince Charming rescues a damsel in distress (poisoned princess, unloved stepdaughter) and the two live happily ever after. The romance narrative is further bolstered through the media, including popular literature in which beautiful, but poor, women capture wealthy, worldly men and initially assertive heroines swoon demurely into the strong arms of handsome he-men (Holland & Eisenhart, 1990; Radway, 1991). The romance narrative portrays women as needing to be rescued by men and, further, as needing men to be complete and fulfilled. This narrative is reiterated in soap operas, novels, television programs including news shows (Meyers, 1997), films, magazines (especially those aimed at women), and popular advice columns (e.g., Ann Landers' perennial 'how we met/happily ever after' columns). As the media and other cultural institutions reproduce the gender and romance narratives, women and men learn the roles culture prescribes, or allows, for them. Women are taught to be accommodating and to seek and please men; men are taught to be dominating and to regard women as inferior.

Narratives are most urgently sought when experience does not make sense.

A third tenet of narrative theory pertinent to this study is that narratives are sought with particular fervour when experience feels chaotic and seems not to make sense. When our lives run smoothly and are untarnished by inconsistencies or chaos, we feel no urgency to locate a narrative that imposes coherence. However, when our experiences do not readily make sense, when chaos intrudes in our lives, we are compelled to find some way to generate coherence -or the illusion of it. For example, when people develop life-threatening medical conditions, they search for narratives to make sense of what is simply not sensible -bodies on which we rely for existence are threatening that very existence; the often impersonal, disembodying treatment of medical practitioners is at odds with the highly personal, bodily character of the experience (Frank, 1995).

In a similar manner, experience becomes incoherent when romantic relationships do not adhere to the central romance narrative -when Prince Charming is not charming, when he does not rescue the damsel in distress, when ever after is not happy. Confronted with a partner who hurts them, women cannot easily fit their experience within the central romance narrative offered by the culture. Thus, they are motivated to discover some way to make sense of what is not sensible: the simultaneity of professed love and enacted violence, the romantic times and the brutal ones, the tender embraces and the black eyes, the unpredictable transformations of Prince Charming to frog and back again.

Responding to incoherent experiences

Immersed in this incoherence, women search out a way to narrate themselves and their experiences. One option is to interpret a relationship so that it is consistent with the culturally favoured romance narrative and the gender narrative that infuses it. Entirely compatible with the fairy tale view of romance, the primary gender narrative casts men as domineering, superior, and aggressive and casts women as subordinate, forgiving, loyal, and accommodating (Gilligan, 1982; Goldner et al., 1990; Wood, 1994,2001). The established gender narrative defines women as needing men. Women are waiting to be rescued, waiting to have their lives and themselves completed by a man. Notwithstanding the notable accomplishments of women's movements, many women are still, to borrow the title of a recent book, 'educated in romance' (Holland & Eisenhart, 1990). They are socialized to believe that their value and happiness depend on 'catching a man.' Women who accept the gender narrative that links their self-worth inextricably to having a male partner are at risk for believing that they must sustain romantic relationships, even destructive ones. In fact, Sugarman and Frankel (1996) reported that wives who tolerated assault from husbands held more traditionally feminine gender identities than did wives who were not assaulted.

The established gender narrative in western culture is sustained through pervasive portrayals of women who subordinate themselves to men and songs that admonish women to 'stand by your man.' Women who accept this gender narrative may narrate themselves good women when they stand by and stay with violent partners. Women may also preserve belief in a romance narrative by focusing on the times when their partners are loving and minimizing the times when they are not. One of the most widely documented dynamics of violent relationships is the 'honeymoon period' (Jacobson & Gottman, 1998; Riessman, 1992; Walker, 1984; Yllo, 1988). Following an episode of violence, a batterer often expresses remorse and engages in behaviours typical of courtship to regain the commitment of a woman whom he has assaulted. The return of Prince Charming simultaneously resurrects the central cultural romance narrative and women's ability to view themselves as loved and their partners as loving. Even women who perceive their partners' violence as wrong often maintain that their partners love them (Atwood & Olsen, 1996; Jones, 1994a; Lempert, 1996).

A second option for women in violent romantic relationships is to give up the central cultural romance narrative and to adopt an alternate narrative. The alternate narrative is not as enchanting as the fairy tale one, but it nonetheless offers a coherent framework for understanding violence in romantic relationships. An alternate narrative that is well established in western culture portrays violence as a routine part of loving relationships (Blackman, 1989; Denzin, 1984; Jacobson & Gurman, 1986). Jones (1994a,b) asserts that there is a strong link between popular representations of love and romance and acceptance of violence in intimate relationships. She notes that the controlling and violent behaviours of men who batter are synonymous with portrayals of love in popular romance novels and other media. Prince Charming is strong, powerful, sure of himself, and commanding. These characteristics of the ideal man in the fairy tale script are not unlike the characteristics of men who are violent toward women (Adams, Towns, & Gavey, 1995; Coan, Gottman, Babcock, & Jacobson, 1997; Cook, 1995). Control, domination, and even violence fit equally well with Prince Charming and the Prince of Darkness. Women who seek to sustain a relationship that is fraught with chaos have available to them culturally legitimated narratives that reconcile what is irreconcilable, make sense of what is not sensible. These narratives, in allowing women to make sense of what is happening, simultaneously license women's oppression. They are a resource to be sure, but when used to justify violence in relationships, they are a resource with the potential for very troublesome consequences.

Within the dark romance narrative and also consistent with the primary gender narrative, women may blame themselves for their partners' violence and/or accept their partners' blaming of them. Research shows that both women and men often blame women for men's violence against women. Men who have been convicted of rape and other forms of violence against women assert that the women provoked them, had it coming, wanted" it, enjoyed it, and did not merit more respectful treatment (Adams et al., 1995; Breines & Gordon, 1983; Coates, Bavelas, & Gibson, 1994; Jacobson & Gottman, 1998; Wood & Rennie, 1994). Studies of women whose romantic partners committed violence against them reveal that they often feel that they provoked the violence, had it coming, or deserved it (Baker, 1997; Chang, 1989; Lempert, 1996).

A third option for women in violent relationships is to invent a new narrative that defines violence as unacceptable in romantic relationships, justifies leaving a violent partner, dissociates women's goodness from standing by their men in any and all circumstances, and maintains a woman's worth is not dependent on her ability to 'catch and hold' a man. In elaborating narrative theory, Fisher (1987) insists that humans are capable of creating or accepting new narratives when those they have grown up believing no longer suffice to explain or guide their lives. This option, however, is more difficult than accepting the ready-made narratives offered by the culture.

Research questions

Existing research has established that women often construct their romantic partners' violence as understandable, but it has not adequately explained how this occurs. In part, this may be because most research has accorded less attention to the voices of women than to the analysis of 'experts.' This study focused specifically on how women narrated violence in their romantic relationships. Two research questions were posed. The first asks whether previous findings that women excuse partners' violence apply to participants in this study. The second asks how they do this.

RQ1: Do participants describe violence against them as normal or understandable?

RQ2: If participants describe intimate partners' violence against them as normal or understandable, how do they narrate themselves and their partners' violence?

Procedures and methods

Participants

Participants were 20 heterosexual women who had been in romantic relationships that included emotional and physical violence. Because the researcher was not a trained clinician, the Board for Research on Human Subjects at the researcher's institution required that participants not currently be in violent relationships. Participants were recruited by announcing the study in classes at a south-eastern university and a north-western university. Because lesbian and gay relationships may be distinct from heterosexual relationships in some ways (Huston & Schwartz, 1996), it seemed inappropriate to merge lesbian and heterosexual women in this study. Participants were not in support groups and had not been counselled by staff at women's centres or similar agencies. This restriction was designed to exclude women whose understanding of their experiences was likely to have been reconstructed by feminist interventions, including naming of the cycle of abuse, building self-esteem, and defining batterers.

Participants were diverse in race, age, and economic class. Of the 20 participants, 12 were white, five were black, one was Native American, and two defined themselves as of mixed ethnicity. Three participants identified their relationships as mixed-race. The ages of participants ranged from 20 to 53 years; their ages at the time they were involved in the violent relationships ranged from 14 to 32. At the time of the relationships, eight of the participants defined themselves as middle class, one as upper class, seven as working class or lower middle class, and one as poverty class. The length of time that the relationships endured ranged from 8 months to 13 years with 3.4 years being the mean length; only four of the relationships lasted less than 1 year.

Procedures

All interviews were conducted by the author. Prior to conducting interviews, approval was secured from the author's university for research on human subjects. Each participant read a one-page statement that explained that the study was investigating violent romantic relationships, described provisions to ensure confidentiality, emphasized her right not to complete the interview, and identified the institutional office to contact if she had questions or concerns about the study.

The interview was minimally structured so that participants could present their experiences in their own words and follow the sequences that made sense to them. After reiterating the participant's right to end the interview at any point or to decline to discuss certain topics, the tape recorder was activated and each interviewee was asked:

“To start us off, could you help me understand who you were when this relationship began? Can you tell me something about what you were like then - how you felt about yourself, your self-esteem, your friends, and activities, and how much dating experience you had had?”

Each interviewee was then encouraged to tell the story of her relationship. Stories offer not only information about experiences, but also insight into attitudinal, cognitive, and affective process that shape the meaning given to experience (Bochner, Ellis, & Tillmann-Healy, 1997; Vangelisti, Crumley, & Baker, 1999). By not imposing on how interviewees narrate their experiences, researchers increase the likelihood of appreciating lived experience from respondents' perspectives. Prompting questions were used to encourage interviewees to elaborate and clarify their accounts, but were not used to direct or structure those accounts.

Interviews ranged in length from 48 minutes to 2 hours and 10 minutes with 1 hour and 16 minutes being the mean length. All interviews were audio recorded and transcribed for analysis. A total of 340 single-spaced pages of transcripts was generated. Analysis focused on parts of the narratives that participants linked directly to their feelings, perceptions, and thoughts about violence in the relationship (e.g., comparison to friends' relationships; parental comments about how relationships work; talking with friends and family about the relationship; witnessing violence in others' relationships).

Analysis

This study relied on inductive analysis as informed by grounded theory (Bulmer, 1979; Glaser, 1978; 1992; Glaser & Strauss, 1967; Strauss, 1987), which places priority on identifying respondents' versions of reality -the meanings they construct for their experiences. Analysis is guided by what emerges from the data, rather than by a priori theoretic formulations (Charmaz, 1983). The concepts that emerge become coding categories that provide conceptual frameworks for analysis. Coding categories are reshaped and refined throughout analysis and are then woven together to permit a holistic, processual rendering of lived experience. Categories are not viewed as distinct and unrelated, but instead are regarded as interrelated and interactive. This method has been useful in efforts to understand personal relationships, in general (Bochner et al., 1997; Harvey, Weber, & Orbuch, 1990; Orbuch, 1997; Vangelisti et al., 1999; Veroff, Sutherland, Chadiha, & Ortega, 1993) and violent relationships, in particular (Merritt-Gray & Wuest, 1995; Wuest, 1995).

Concurrent interviewing and analysis is a preferred methodological practice in analytic induction. In the process of analysing data, understandings derived from interviews conducted later in the research process are tested against and used to illuminate interviews conducted early in the project through a 'constant comparative process' (Strauss & Corbin, 1990, p. 65), which refines both broad coding categories and analyses of specific interviews. Analysis thus requires researchers to revisit both coding categories and interpretations of earlier interviews through the lens of understandings that emerge later in the process (Charmaz, 1983; Lempert, 1996).

In most research on violence against women, the voices of experts overshadow the voices of women and the stories they tell. In her work on narratives, Pollock (1990) cautions researchers not to 'subordinate first person narratives to their own' (p. 3). Following Pollock, this study foregrounds women's experiences as represented in their words. In working with participants I aimed to 'think with' the stories they told. Frank (1995) makes a useful distinction between thinking about and thinking with stories: 'To think about a story is to reduce it to content and then analyse that content. Thinking with stories takes the story as already complete. . . . Thinking with stories ultimately requires a highly personal sedimentation of experience: living with the stories and having them shape perceptions of various experiences over time' (pp. 23-24).

Results and analysis

Data provided a clear, positive answer to the first research question. The women in this study described their partners' violence against them as understandable at the time it occurred. They justified it using a variety of reasons: 'I deserved it,' 'He didn't really mean it,' 'He was drinking,' 'It could have been worse,' and so forth.

In answer to the second research question, the women's accounts were framed by gender and romance narratives that were used to explain and justify violence. Within any narrative, there are moments, or obstacles, that challenge the coherence of the story line. At these moments, beliefs authorized by particular narratives rescue the plot and sustain the narrative's coherence. Specific beliefs on which women relied to define and understand violence made sense within the context of gender ideology and its derivatives, either the fairy tale romance narrative or the dark romance narrative.

Participants discussed four beliefs that allowed them to define their relationships as consistent with the fairy tale romance narrative when that narrative's viability was jeopardized by violence. Participants also alluded to two beliefs on which they relied to reconcile themselves to an alternate dark romance narrative when the fairy tale version could not be sustained. In the following discussion, I begin by describing the dominant gender narrative in western culture. Next, I describe the romantic narrative in both its fairy tale and dark versions, and I highlight traditional gender ideologies that are woven into both narratives. Within this discussion, I define and illustrate beliefs that women referenced in describing how they sustained allegiance to the romance narratives. I present extensive excerpts from interviews because they offer the richest insight into how participants made sense of their experiences. All names used are pseudonyms.

Table 1 shows the number of participants who referenced each narrative and belief, lists the narratives and beliefs associated with them, and provides excerpts from interviews to illustrate narratives and beliefs. This table serves only to summarize data as information for interested readers. It should not be interpreted as reductionistic analytic categories, which are not capable of representing the full complexities of lived experience as embodied in narratives.

Gender narrative

Framing both versions of the romance narrative were knowledge and acceptance of the primary gender narrative authorized in western culture. This gender narrative stipulates that it is normal/appropriate for men to be controlling and dominating, and it is normal/appropriate for women to defer and subordinate themselves and their interests. Interviewees spoke of knowing that they needed to please their men. Explaining how she felt about her boyfriend's continuous criticisms of her body, Janelle described 'feeling like I had to look a certain way for him and definitely in terms of "well, I'm supposed to please him".' Denise recalled that 'I wanted to be like he wanted me to be, so I kept trying to figure out what he liked.' Mary gave up visiting her mother each afternoon in order to have her boyfriend's dinner ready when he came home and quit resisting when he 'made me get up every morning at like seven o'clock to start cooking his breakfast and start cleaning the house.'

The central gender narrative in western culture also explains participants' efforts to defend their partners from others' knowledge and criticism in order to shore up their own view of the relationship as a fairy tale romance. One woman narrated herself as shielding her boyfriend and herself by not letting others know about the violence in her relationship: 'I didn't want the girls on my hall to think he was crazy. . . didn't want any of them to know he hit me. . . . So, like, I'd tell them about the nice things he did.' Unwilling to let her parents know that her boyfriend had slammed a car door in her face, Nadine 'made up some story . . . some stupid cat story. Like some cat scratched me.' When the emergency room doctor asked Mehnessah about her injuries, she did not tell him her boyfriend had kicked her because 'I was kind of ashamed for Greg. It gave the wrong impression of him.' When the doctor asked if she wanted him to call the police, 'I said no. I didn't want to call in the police. I didn't want it to be public or anything, didn't want my folks to know.' Brandy was especially insightful in describing how protecting her boyfriend against her friends' criticism also helped her sustain belief in the fairy tale romance narrative. She said she told her girlfriends, '"He's sending me these little gifts and he does this" and trying to say that it was wonderful. . . trying to in my mind make it something that was romantic, was beautiful. So whenever I talked to them about it, I would embellish, and I would try to make myself believe it.'

A particularly prominent indicator of participants' acceptance of the established gender narrative was their internalisation of the expectation that women should care for their partners. Women narrated themselves as being responsible for comforting men who had beaten them and understanding the frustrations that 'drove' men to violence. The day after her boyfriend threw her out following forced intercourse, Mary went back to their apartment to get her clothes. Before she left, '1 washed his laundry and put it in a separate basket.' Beverly 'didn't want things to be bad between us. It was my responsibility to keep things okay.' Four women reported returning to abusive relationships when their partners threatened or attempted suicide. Bailey recalled that Jon 'talked about how he would kill himself and he couldn't make it without me.' Melinda's boyfriend said he would kill himself if she ever left him, a tactic that convinced her to stay, despite his holding a knife to her throat on repeated occasions. After attempting suicide and leaving her husband, Vera heard from a friend that he was in serious trouble. In response, she found her ex-husband, took him into her home, which was two thousand miles away, and told him, '1 will get you back on your feet.' She supported him financially and nursed him for months until he recovered his health.

Romance narratives

Infusing all women's accounts was a romance narrative, which had both idealistic and malignant forms. The romance narratives, in conjunction with traditional gender narratives, established a coherent frame which women used to narrate themselves and their relationships. The romance narratives also authorized specific beliefs that women used to sustain a coherent narrative of their relationships as viable and acceptable.

Fairy tale narrative. All participants recalled initially perceiving their relationships as fairy tale romances, complete with an adoring Prince Charming. Women fondly remembered how their partners had courted them with gifts and made them feel special. 'I was, like, happy at first, very happy. I had that glow, you know,' said Mary. Ellen cherished her boyfriend because 'he made me feel like I was the centre of his universe.' Echoing that feeling, Bailey said that Jon 'made me feel like I was his world.' Mehnessah 'was just in the clouds like you read about in fairy tales.' Brandy remembered that 'I was swept up in the romance of it. He was a Prince Charming. He would shower me with little creative gifts. Like on Valentine's Day he came with one rose and a card that said, "this is a rose for our first Valentine's Day together. It hopes to return next year with a twin." He pretty much swept me off my feet.' After Denise's first date with Gerald, 'he sent flowers. The note with them was so sweet too. It said, "More to come - Gerald." I was just totally excited that he would do that after we'd just met. I'd never known a man who treated me like that.' Charles delighted Beverly by calling after they had met at a party 'just to see if I got back safely. I was really impressed by that.' As Charles continued courting her, Beverly recalled that 'it was like a fairy tale, and he was Prince Charming. He was tall, dark, and handsome, and he was smart.'

Fairy tale views of romance were also revealed in women's desire to make a 'good catch.' Prince Charming is supposed to be handsome, professionally successful (or potentially so), and someone parents approve and friends admire or envy. Describing her boyfriend, Brandy remembered that he was 'very, very, very smart, very smart. He's an engineer, and he graduated magna cum laude . . . and he was very good looking. I thought my parents would be excited that I was dating an engineer that was very good looking, he will give her support and a good looking family one day.' Beverly expressed similar pleasure in Charles' prospects: 'He was very good looking. He was big and muscular, really pretty macho, if you know what I mean. And he was going to be a lawyer. I liked that and I knew my folks would be happy that I was seeing someone who was going to be a real professional.' Denise remembered that she 'was like totally blown away. . . I mean here is this really successful, brilliant, older guy who says he might marry me.' Later in the relationship after Gerald started hitting, then beating her, Denise recalled that she didn't want to give up her plan 'to marry him and have a successful life like he could make for us with his job and where he was going.'

The fairy tale narrative does not preclude problems, but it does maintain that love can conquer any hardship. Mehnessah reasoned that 'there were problems, but we loved each other enough to solve them.' After a lull in the violence, Mehnessah 'thought that love had finally conquered all and that we would have a good life together.' Recounting her 8-year relationship with Jon, Bailey said, '1 knew he loved me for sure. There were things I didn't question even when he hit me. He was my soul mate. I felt so complete with him.' Even when Jon impregnated another woman while in the relationship with her, Bailey said ‘I was willing to risk everything for him. I was determined to make it work because he was my world.'

Beliefs that bolster the fairy tale romance narrative. Participants described four specific beliefs that were both authorized by and sustaining of the fairy tale romance narrative. These functioned to convince the women and others that their violent relationships really did embody the fairy tale.

Not as bad as . . .The most prominent belief women relied on to bolster the fairy tale narrative was that the violence was not as bad as it could have been, had been before, or others experienced. All 20 participants mentioned this belief. Some regarded the violence as acceptable because it could have been worse. Bailey, for example, said, 'he would hit me on my face but it would be open-hand so I never had a mark or anything. Sometimes I'd have a handprint, but it would go away.'

Some women tolerated violence because it was less severe than that inflicted on women they knew or than previous violence they had experienced. Mary stated that her friend Amy 'had been beat up so bad that she's been in the hospital. And I mean, maybe I looked at it as I looked at her and I saw how Luke beat her and she still stayed with him, that I considered myself lucky. That, you know, I just had, like, bruises and I didn't have a broken jaw or a broken nose or anything like that.' Similarly, when her boyfriend assaulted her in front of a friend, Letitia thought, 'It's not like I was turning blue. It's not like, it's not like the other times.' After her boyfriend slapped her the first time, Mehnessah thought, 'He said he hadn't hurt me, and really he hadn't, not really.' Later, when the same man kicked her and cracked a rib, Mehnessah reasoned that 'a cracked rib wasn't like a broken leg or anything really serious.' Nadine recalled thinking, 'okay I may have had a bruise here and a bruise there, but the bruises went away.'

Underlining women's belief that they should tolerate violence that wasn't 'too bad' were four participants' explicit statements that the violence they had experienced had not been 'bad enough' to justify ending the relationship. They had an idea of what 'bad enough' was and, in fact, they recalled wishing that their partners had done something bad enough to legitimise their exits. Maggie said, 'I used to say I wish he'd actually hit me or something so I'd have a reason to leave him.' Brandy recalled that '1 wanted him, like, to really bruise my face, so I could go home, I could say, "look what he did. Is it okay for me to break up with him?" . . . I actually wished that he would hit me, that I'd have a bruise or a black eye. . . like almost that was a symbol that he really was doing this.' Against a constructed standard of what would be 'bad enough' to end a relation- ship, these women judged the violence inflicted on them as 'not so bad.'

Other women's narratives portrayed violence that would have been 'bad enough' early in a relationship as 'not so bad' by the time it did occur. Women explained that they became desensitised as they incorporated progressive violence into their narrative for romantic relationships. Reflecting on how she came to tolerate her boyfriend's pushing her around, Lynette said, 'At first it was just like, "What are you doing?" After a while, it was commonplace.' Denise observed that 'if he had kicked me on the floor and made me eat off it that first time when he just slapped me for looking wrong at a waiter, I would have been out of there. But it was like what he did to me when his bad side came out just got worse over time and so I did take it. . . . It was just so gradual like that I kind of got used to his bad spells.' Said Melinda, 'At this point he's violent so much and it's like I don't even, I'm not even trying to get away. At this point I've been strangled four times and I've got a knife to my throat and I'm just like, "what the hell?" Using a metaphor to explain the desensitising power of progressive violence Vera observed that 'it was like getting into a pot and then it slowly being warmed up.'

The good outweighs the bad Nineteen participants bolstered the fairy tale narrative by believing that violent incidents were aberrant and not part of the ordinary rhythms of their relationships. After her partner gave her a black eye, Denise thought, 'Gerald was more good than bad and I should be able to accept the bad times.' According to another woman, 'Most of the time he was real nice and thoughtful and everything, so I just tried to put it [being hit in the face] into perspective.' To get beyond her anger at abuse, Janelle reminded herself that 'Oh, but a week ago he was so nice.'

Women also recounted believing that the good outweighed the bad because increased intimacy followed a violent episode. This belief embodies the 'honeymoon period' and, with it, the return of Prince Charming. After leaving the hospital for treatment for an injury inflicted by her boyfriend, Mehnessah recalled, 'He was really good to me for a while - several months and it was just like the real Greg was there all the time.' Bailey said 'that after those instances sometimes we were closer. I was closer to him, we were both closer to each other after going through stuff like that.' After her boyfriend forced her to have sex and 'popped me in the face' on Valentine's night, Beverly recalled 'for a while after Valentine's, he was really nice. It was like that night had been a bad dream. He even brought me a dozen roses.' Roses were also the magic charm for Denise. The morning after her boyfriend ripped up her dress and bloodied her face, 'he brought me flowers - he got my favourite, which is roses. . . I knew he

loved me.' Following months of sexual abuse, including permanent, deliberate damage to her genitals, Vivian left her husband, but he won her back because he became 'just as sweet as he could be. . . . I mean, he was so nice it was basically like when I first met him.'

The honeymoon period fuelled women's belief that the violence was history. Bailey recalled that '1 thought it would get better. . . whenever there was an episode, say he went three months instead of one month, I thought "well, he's just regressed just a little bit. . . next time it will be longer."‘ Likewise, after being raped and beaten by her boyfriend, Mary remembered thinking 'he realises this time that it was bad, he won't do it again.' Ellen recalled that '1 always thought maybe he could change.' Despite years of abuse, including being drugged by her partner, Vivian 'thought it would get better, I really did, that it was just a phase.' Following being 'slammed against the wall' and forced to have sex during which her boyfriend called her a 'dumb bitch,' Dense remembered that the next day 'he said how sorry he was. . . how much he loved me and how he would never hurt me again and how sorry he was and all. And I could tell he was really upset and. . . I thought maybe we could go on and things would be okay after that.'

I can control it/stop it Sixteen participants sustained the fairy tale romance narrative by believing that they could avoid 'provoking' further violence, a strategy implicitly linked to believing they were to blame for the violence and to the gender narrative's expectation that women should please their men. Dense recalled that '1 learned how to read him and figure out what would make him mad and not do those things, or say those things. ...I kept trying to figure out what he liked and didn't like and to behave right around him. . .and things got better.' When her boyfriend Charles criticized her chewed fingernails and weight, Beverly reported' I quit biting my nails, and I did get more careful about what I ate. I even bought a scale, which I'd never owned or had one at home. And the first thing I did each morning was weigh myself and if I had gained a pound, I'd skip meals that day.' Knowing that her boyfriend despised mustard and would launch into violence if given food with mustard, Maggie placed an order for hamburgers 'and I told the people what I wanted, they got the order right on the screen and everything, and so, okay, and so I checked. I didn't see any mustard. I was looking for mustard. I was freaking out about the mustard.' Mary's boyfriend expected her to be home and to have dinner ready whenever he got off work. She recalled that 'he worked in construction and some days he got off at three, some he got off at seven, so I basically just sat in the house, all alone and just was there for him, like whatever he wanted.'

Not the real him A fourth means of bolstering belief in the fairy tale narrative was to believe violence could be dissociated from the men who enacted it (Wood, in press). Eighteen participants believed that the men they loved weren't 'really like that.' Some women separated 'the real him' from times when 'he was not himself' because he was in a bad mood, unhappy, drunk, or on drugs. And still others drew a distinction between who he (really) was and what he did. The women did not hold a partner responsible for violence if they attributed it to factors they constructed as beyond his control. By dissociating violence from the men they loved, the women could continue to love the 'real him.' Illustrating this belief in its variant forms are the following comments from, respectively, Beverly, Janelle, Denise, Mehnessah, and Bailey:

‘That wasn't the real Charles. . . . This wasn't the Charles I first met, the guy I fell in love with. It was just that he wasn't himself, you know. But nobody is nice all the time. So I figured he'd just been not himself or something, that he had been in a bad mood, but it was over and we were okay. I didn't want to make too big a deal over it or anything.’

‘I know he's a better person than that. . . maybe that wasn't him that night. . . . Well, he was a good person. He had a bad day.’

‘He just got in these mean moods, but most of the time he was his regular self. And so I tried to remind myself that this wasn't how he really was. I mean, everybody's got a bad side or has bad moods, so I understood that. He was only ugly or mean if he'd been drinking or doing drugs. The rest of the time he was his true self. . . real Greg -the nice one who cared about me and was warm and thoughtful and nice to me -that was the real Greg. . . . Sometimes the real Greg was hidden beneath all of the bad stuff, but that's who he really was. . . . His anger would just get away from him sometimes, that's all. It wasn't like that was who he was or anything.’

‘I hated those things he did, but I loved him. . . . He just couldn't control himself. . . I knew he would never purposely hurt me, but in that instant he was somewhere else.’

The dark romance narrative

Within the data, a second romance narrative was evident -one that constructs violence as typical in romantic relationships. The dark script claims that it is normal for men to have 'bad spells,' as Denise's mother referred to men's violent episodes, and it is normal for romantic relationships to be hurtful to women. The dark romance narrative also insists that abuse and unhappiness are not reasons to abandon the relationships because women are supposed to be for- giving and because they need men to be complete.

Referred to by 17 participants, the dark narrative encouraged women to believe 'You have to make it work, no matter what' and 'you deserve whatever a man does to you' because any relationship is better than none for a woman. Exemplifying the internalisation of this belief, Nadine explained that she stayed with her violent boyfriend because 'If something were to happen and I was not with him anymore, I wouldn't know what I would do in my life.' Explaining why she stayed, Janelle said, 'Everyone else always has someone and I was like, "this is just my someone and I'm going to have to put up with what little I have." I just felt like I needed to have someone.' Echoing Janelle, Maggie recalled '1 was afraid to be alone.' Reiterating this, another respondent said, '1 would just deal with things that I shouldn't have dealt with just 'cause I didn't want to be by myself.' Reflecting her need to stay in a relationship she perceived as unhealthy, Amelia said, '1 thought he was the best I could do. He felt like the best I could do.' The gender narrative's insistence that a woman needs a man encourages women to accept the dark narrative if the fairy tale one is unattainable.

The dark romance narrative portrays violence as normal and not a reason to end a relationship. Mehnessah explained that after talking with her friends and learning that their boyfriends hit them, '1 could see that things getting out of hand occasionally doesn't mean a relationship is bad or anything. These were nice girls. They wouldn't have put up with a guy who was bad, so I came to understand that Greg wasn't abnormal or anything.' Echoing this, Ellen said talking with her friends showed her that 'I'm not the only one having problems, and they're all dealing with it, so I can too.’ For Denise, the dark romance narrative was explained by her mother: 'Once when I'd told my mama that Gerald was sometimes mean, she said all men are and that's just how they are. So what she said kind of let me know that Gerald was not any worse than other men -that all of them have bad spells -that's what mama called them -and sometimes you just have to overlook those.' When Vivian told her father that her partner was abusing her, he ratified the dark romance narrative: 'my dad would say, "quit complaining and make the best of it."‘

Participants also reported coming to accept the dark romance narrative by observing friends and families. Mary stated that, 'I saw my dad beat his girl- friend, I've seen my mom get beat, I've seen my friends get beat, and everybody just lived with it. And I just thought it was something I was gonna have to deal with.' Melinda recalled that 'My family is violent. And that was a really import- ant part of what made me stay with Thad, 'cause these were people I loved, and I knew they loved me and, but they still hit me, so. . . .' Vivian recalled that 'My mom was shot by my father three times.' Reflecting on her experiences, Letitia commented 'In high school we didn't talk about it [abuse], because it was just part and parcel of relationships. It wasn't, like, a problem.'

Beliefs that bolster the dark romance narrative. When 17 women could not find ways to convince themselves that their relationships conformed to the fairy tale romance narrative, they turned to an alternate narrative of romantic relation- ships that is also well-established in western culture. According to the dark romance narrative, men are sometimes violent, relationships can be bad, and still women need to hang onto relationships and their men. Women described two beliefs that allowed them to perceive violence as tolerable within the dark romance narrative. Echoing previous research, women in this study blamed themselves for violence they experienced and they stayed because they felt they were better off with any relationship than without one.

I deserved it Consistent with many previous reports (Atwood & Olsen, 1996; Baker, 1997; Chang, 1989; Jacobson & Gottman, 1998), 17 of the 20 women in this study attributed violence inflicted on them to themselves, their actions, or their inactions. Many women stated that what they said or did caused their partner's violence toward them. Letitia believed that verbal and physical violence 'wouldn't have happened if I would just shut up,' while Dense 'thought he was right that if I had listened when he first said not to go, none of it [battering of her chest and face] would have happened.' After her boyfriend raped her, Letitia recalled 'thinking it was my fault because I put myself in the situation.' When a counsellor suggested Letitia press charges against her boyfriend, she said, 'Don't you get it? It's not his fault.' Bailey reasoned that 'Every time he hit me I had gotten him started over something and I always felt like I started it.' Ellen said, '1 never, ever thought to blame him. I was, like, maybe I did something wrong, maybe I shouldn't have said that, maybe I just need to be quiet. . . I had to be doing something wrong, because nobody would just react like this for no reason.' More emphatically Vera explained that she felt 'I'm the one that's in the wrong, so I just lived with it.' After her boyfriend dragged her around their house and cursed her, Mary said, 'I just assumed that I had been bad and I deserved everything I got.'

A second form of self-blame was to believe violence was motivated by partners' desire to help them overcome failings or punish them for bad behaviour. Brandy, who had won multiple beauty contests before and during her relation- ship with Martin, accepted his criticism that she was overweight and out of shape. When he hit her if she wasn't running 'fast enough: Brandy said, '1 thought he was right. I thought he was right. I took it as being right.' When Brandy discovered that Martin's father had brutally killed his mother and con- fronted Martin with her knowledge, he became enraged and lashed out at her for not understanding his situation. In the face of his attack, Brandy decided 'he was right. I wasn't understanding his point. I was being an emotional woman. I just didn't think in as clear terms.' After her boyfriend slapped her and called her a whore for dancing with another man, Denise reasoned, 'he was looking out for me and trying to protect my reputation, which I thought was good.' Later, at a restaurant, Denise's boyfriend threw her meal on the floor, telling her that fried food would make her fat, and she reasoned that 'he was trying to look out for me and even if it might have seemed rude, it was really because he wanted to help me do what was best for me.'

Being stuck The second belief that justified accepting the dark romance narrative was that there was no acceptable option. Sixteen participants said that they felt stuck, unable to end the relationship or live without a male, and, therefore, they had to accept whatever their partners did to them.

The most commonly expressed reason for feeling trapped in a relationship was belief that investments already made precluded exiting. Investments are things that people put into relationships that cannot be recovered if the relation- ship ends (Kelley, 1983). For Brandy, Beverly, and Bailey, the investment that they could not afford to lose was having lost their virginity to their partners before marriage, an action that, in their minds, required them to marry their partners: Brandy said, 'I lost my virginity to him. I equated sexual relationships with marriage, so now I was going to have to marry this guy. . . I felt trapped there. I felt like I was a bad person and that God and my parents would be mad that I slept with this guy out of marriage.' Beverly recalled that, 'After I lost my virginity to him, I felt like I was totally tied to him, like it had to be forever - sort of like it would still be okay, I'd still be moral if I wound up marrying the person I had sex with. Maybe if I hadn't done that [had sex], that I wouldn't have tolerated how he treated me, but once I did that, it was just like I had to take whatever he did. I didn't really feel I had any choice after that.' Bailey stated that, 'I thought of myself as a Christian and everything, wasn't gonna have sex before marriage and, like, I lost my virginity to him that spring. . .the fact that I did. . .would make me stay in the relationship.'

In a different way, Vera and Vivian cited their Christianity as what trapped them in abusive marriages. Vivian believed that 'I could never remarry if I did divorce him because he was still alive. I studied the Bible and he [partner] would use that against me.' Vera, married to a 'self-made minister,' said, 'We were very active in the church and nobody would support me. . . . He's a man of God - who would argue with that?'

The expectation of a desirable future - one that is the gold ring in the romance narrative - was another investment that some women did not want to lose. After her boyfriend kicked her to the floor, threw burning hot food on her, and forced her to eat off the floor, Denise still thought it would be 'hard to give up the idea of the future I thought he and I could have.' Rather than do that, Denise, like other women in this study, gave up the fairy tale narrative and substituted the dark narrative, which maintains that violence toward women is normal and women are supposed to tolerate it. As Mary explained, 'I just figured that every woman dealt with this. This is what happens in a relationship.'

The gender narrative was particularly pronounced in women's belief that they were stuck in violent relationships because they would be incomplete without men. Most women in this study narrated themselves as lucky to have any man, even an abusive one. Often the women's low estimation of self-worth echoed their partners' devaluations of them. Vivian's partner 'used to tell me all the time that no one else would ever have me. I felt like I wasn't worthy of any- thing. I mean, he led me to believe that I would never be nothing.' Similarly, Denise believed that 'Other men wouldn't be interested in me. I should count myself lucky that he was. ...He let me know that I wasn't perfect.' According to Mary, 'No one else would want me anyway, and I have it really good [with him].' Explaining why she tolerated regular beatings and forced sex, another woman said, '1 just figured it was his prerogative because I was less, I was a woman and I was younger than he was. He had the right to do that.'

Other women attributed their sense of entrapment to an inexplicable force that led them to stay with violent men. Exemplifying this, Amelia said, 'He just had a power over me. It was just something about him that I couldn't help myself. I felt I had no choices. I just felt trapped. . . I felt so trapped and so stuck. There was nothing I could do.' At the beginning of her four-year relation- ship with a man who eventually beat her, held a knife to her throat, and raped her, Melinda said, '1 felt this energy, just sexual energy the first time I saw him. I've never felt anything like that before or since. I was just totally drawn to him, unable to resist him.' Ellen's sense of helplessness to leave the relationship came through clearly in her statement that, 'It was like no matter what I did, I couldn't tear myself away.' Blaming herself for her entrapment, Nadine believed that 'It's just something in me. . . there's something inside of me. . . that is allowing this stuff to happen to me, that just found this relationship with a guy that treated me like I was nothing.'

No matter how violent a relationship was, some women felt that they couldn't leave it. Explaining this, Vivian said, 'If I walked away, I would not be able to make it, you know? I was just stuck.' Mary offered a similar account: 'It was like I was stuck. I just stayed there because I felt like I had to. . . this is what my life is gonna be.' Vera reported feeling so trapped that it seemed the only way out was to die: '1 just decided that this is useless. I had given up,' so she attempted suicide.

Several participants attributed their feelings of entrapment to losing perspective on themselves and life. Their comments dramatically illustrate their sense of being enmeshed in chaos and needing some way to make their relation- ships coherent. Said Amelia, 'When you're in it, you don't, you can't, you can't see above the swirling of what's happening around you.' In similar fashion, Letitia reflected that, 'It's not something that you have any control over. . . . There are warning signs and I know that now, but when you're in it, you don't see those warning signs.' Perspective also narrowed due to the well-documented tactic of abusers to isolate their partners from friends and family (Jacobson & Gottman, 1998; Lempert, 1996). Ellen's boyfriend edged her friends out of her life by always being with her. Thinking back on this, Ellen concluded '1 never really noticed it, but he like isolated me from all of my friends. He always wanted me to be with him and so it made me just depend on him more and more and more. . . . He was the only stability I had.'

Summary and suggestion for further research

Most research on violent relationships has emphasized the voices of 'experts,' especially clinicians and researchers. In contrast, this study accords centre stage to the voices of women themselves -their perceptions as embodied in their words. This focus generated findings that support and extend previous scholarship on violent relationships.

Summary of findings

The findings of this study are best understood through a narrative approach that sheds light on how existing, discrete findings cohere within culturally endorsed and supported romance and gender narratives. The stories told by the women in this study are both personal and social. They are personal in that they are accounts constructed by particular women to describe their individual experiences. They are also resolutely social because they reflect and embody culturally produced, sustained, and approved narratives of gender and romance. The narratives in circulation in western culture are resources available to individuals who are wrestling with experiences that do not make sense -with experiences that unite love and violence, Prince Charming and a brutal assailant.

Western culture offers two distinct romance narratives. One is the fairy tale romance in which Prince Charming courts a princess and makes her the centre of his world. Being in a relationship with him will lead to approval of family and friends and a good, complete life for a woman. Participants in this study worked hard to maintain the belief that their relationships embodied the fairy tale romance narrative. To do so, they relied on beliefs authorized by the narrative: the violence they experienced was not so bad, the good out- weighed the bad in their relationships, they could control or stop the violence, and their partners were not their 'real selves' when they were violent.

When women could not sustain belief in the fairy tale amidst continuing violence, another culturally sanctioned romance narrative was available to them. The dark romance narrative casts men as controlling and violent at times ('bad spells') and casts relationships as typically hurtful, yet necessary, for women. Women whose relationships fail to embody the fairy tale narrative are vulnerable to the dark romance narrative, which provides a coherent model for understanding and settling for the relationships they have. Women described two beliefs that supported and reflected the dark romance narrative. Consistent with previous research, women narrated themselves as to blame for the violence inflicted on them. They also portrayed themselves as trapped because they had made heavy investments or because they saw no acceptable alternative. Woven into both romance narratives was western culture's traditional gender narrative, which prescribes male dominance and aggression and female accommodation, caring, and dependence on a man for self-worth. In tandem, these sex roles can license men's violence and women's toleration of it.

The beliefs that romance narratives authorize are not a collection of discrete convictions. Rather, they function interactively and coherently to define violent romantic relationships as normal, tolerable, or at least prefer- able to no relationship. The systemic character of the beliefs that women narrated in this study invites investigation of how the beliefs interact and reinforce one another. Better understanding of this might inform efforts to craft interventions to disrupt the coherence and continuity of violent relationships.

Heuristic directions

This study invites research along several lines, two of which merit special attention.

Other populations. Future research might fruitfully investigate whether narratives and the beliefs they authorize apply to other kinds of relation- ships. For instance, researchers might study how gender and romance narratives operate in violent gay and lesbian romantic relationships. It would also be valuable to investigate narratives in violent parent-child and child-elder parent relationships. Presumably the romance narrative that frames heterosexual women's justifications of violence would not operate in parent-child and child-elder parent relationships. Other narratives, such as a happy family narrative or a stick together narrative, might substitute.

Remaking romance narratives. Culturally legitimated romance narratives, both dark and fairy tale forms, constrain both women and men. The fairy tale romance narrative, if realized and sustained, may be relatively benign, although it does encourage women to rely on men to rescue and complete themselves. The dark romance narrative has the potential to compel women to stay in harmful relationships.

Western culture's traditional gender ideology is woven seamlessly into both fairy tale and dark romance narratives. Central to traditional gender ideology is linking a woman's worth to having a man. This has the potential to authorize men's violence and women's toleration and forgiveness. As currently crafted, these narratives invite women to accommodate, defer, and tolerate unspeakable abuses in order to keep their men. The gendered romance narratives currently legitimated in western culture offer insufficient stories of individuals and relationships. They beg for revision.

And revision - even of powerful and long-held narratives -is not simply an abstract possibility. It has occurred again and again in human history. In his award-winning book, The Prophetic Tradition and Radical Rhetoric in America, Darsey (1997) notes that Americans periodically '(re )invent those principles that define us as a people' (p. x). One of the primary ways we reinvent who we are is by crafting new narratives to direct and make sense of our lives. Martin Luther King, Jr. created a new narrative of America with his '1 Have a Dream' speech; in collaboration with others, Elizabeth Cady Stanton created a new narrative of women in America and that story was eventually woven into laws and the Constitution; feminists created a new narrative of sexual harassment when they coined the term in the early 19708 and that story has gained both social and legal standing.

In his discussion of narrative theory, Fisher (1987) insists that humans are not constrained to accept the stories already established in a culture. People have the ability 'to formulate and adopt new stories that better account for their lives' (p. 67). Yet composing more healthy romance narratives requires more than personal creativity. Narratives are rooted in, and thus supported or undermined by, the larger culture. Thus, cultural structures and practices must participate vigorously in authorizing new romance narratives -ones that narrate violence as unacceptable, ones that narrate women as complete with or without male partners, ones that narrate individuals as responsible for their actions. To establish new romance narratives and to discredit ones that condone abuse, institutions must operate systemically. Families, schools, and the workplace must confer persuasive power on new narratives and diminish the acceptability of toxic ones.

The media is another cultural institution that can assume a powerful role in rewriting gender and romance narratives. Too often the media reproduce toxic romance narratives. Cuklanz (1996), for instance, points out that made-for-television movies tend to represent 'romantic endings as the resolution to ongoing marital violence' (p. 113). Cuklanz also notes that mediated representations of romantic relationships, including violent ones, typically address female audiences. For alternative, non-toxic romance and gender narratives to gain cultural legitimacy, they must be widely disseminated to multiple audiences, including, especially, men.

The narratives authorized by a culture are neither finite nor fixed. They are remade continuously as individuals and institutions decide that existing ones are inadequate to define and direct our lives. The human capacity for remaking the social world means that it is possible to imagine and bring into existence new gender and romance narratives that refashion understandings of what is acceptable, normal, and good in romantic relationships.

REFERENCES

Adams, P., Towns, A., & Gavey, N. (1995). Dominance and entitlement: The rhetoric men use to discuss their violence toward women. Discourse and Society, 6, 387-406.

Atwood, J., & 0lsen, M. (1996). Unscrupulous scripts: Scripts for family violence. In J. Atwood (Ed.), Family scripts (pp. 239-270). Washington, DC: Accelerated Development Press.

Bachman, R., & Saltzman, L. (1995). Violence against women: Estimates from the redesigned survey. NCJ-I54348. Washington, DC: Bureau of Justice Statistics, US Department of Justice.

Baker, P. (1997). And I went back: Battered women's negotiation of choice. Journal of contemporary Ethnography, 26, 55-74.

Barnett, O., & LaViolette, A. (1993). It could happen to anyone: Why battered women stay. Newbury Park, CA: Sage.

Berger, A. (1996). Narratives in popular culture, media, and everyday life. Newbury Park, CA: Sage.

Blackman, J. (1989). Intimate violence: A study of injustice. New York: Columbia University Press.

Bochner, A., Ellis, C., & Tillmann-Healy, L. (1997). Relationships as stories. In S. Duck (Ed.), Handbook of personal relationships: Theory, research, and interventions (2nd ed., pp. 107-124). Chichester, UK: Wiley.

Breines, W., & Gordon, L. (1983). The new scholarship on family violence. Signs, 8, 490-531.

Bulmer, M. (1979). Concepts in the analysis of qualitative data. Sociological Review, 27, 651-677.

Caputi, J. (1993). The sexual politics of murder. In P. Bart & E. Moran (Eds.), Violence against women: The bloody footprints (pp. 5-25). Newbury Park, CA: Sage.

Chang, D. (1989). An abused spouse's self-saving process: A theory of identity transformation. Sociological Perspective, 32, 535-550.

Charmaz, K. (1983). The grounded theory method: An explication and interpretation. In R. Emerson (Ed.), Contemporary field research (pp. 109-126). Boston: Little, Brown.

Coan, J., Gottman, J., Babcock, J., & Jacobson, N. (1997). Battering and the male rejection of influence from women. Aggressive Behaviour; 23, 375-388.

Coates, L., Bavelas, J., & Gibson, J. (1994). Anomalous language in sexual assault trial judgments. Discourse and Society, 5, 189-222.

Cook, S. (1995). Acceptance and expectation of sexual aggression in college students. Psychology of Women Quarterly, 19, 181-194.

Cuklanz, L. (1996). Rape on trial: How the mass media construct legal reform and social change. Philadelphia: University of Pennsylvania Press.

Darsey, J. (1997). The prophetic tradition and radical rhetoric in America. New York: New York University Press.

Del Mar, P. (1996). What trouble 1 have seen: A history of violence against wives. Cambridge, MA: Harvard University Press.

Denzin, N. (1984). Toward a phenomenology of domestic family violence. American Journal of Sociology, 90, 483-513.

Fisher, W. (1987). Human communication as narration: Toward a philosophy of reason, value, and action. Columbia: University of South Carolina Press...

Frank, A. (1995). The wounded storyteller: Body, illness, and ethics. Chicago: University of Chicago Press.

French, M. (1992). The war against women. New York: Summit Books.

Gergen, K. (1997). Realities and relationships. Cambridge, MA: Harvard University Press. Gilligan, C. (1982). ln a different voice: Psychological theory and women's development. Cambridge, MA: Harvard University Press.

Glaser, B. (1978). Theoretical sensitivity. Mill Valley, CA: Sociology Press.

Glaser, B. (1992). Basics of grounded theory analysis. Mill Valley, CA: Sociology Press. Glaser, B., & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine deGruyter.

Goldner, V., Penn, P., Sheinberg, M., & Walker, G. (1990). Love and violence: Gender paradoxes in volatile attachments. Family Process, 19, 343-364.

Harvey, J., Weber, A., & Orbuch, T. (1990). Interpersonal accounts: A social psychological perspective. Oxford, UK: Basil Blackwell.

Hasenauer, H. (1997). Taking on domestic violence. Soldiers, 52, 34-36.

Holland, D., & Eisenhart, M. (1990). Educated in romance: Women, achievement and college culture. Chicago: University of Chicago Press.

Huston, M., & Schwartz, P. (1996). Gendered dynamics in the romantic relationships of lesbians and gay men. In J. Wood (Ed.), Gendered relationships (pp. 163-176). Mountain View, CA: Mayfield.

Jacobson, N., & Gottman, J. (1998). When men batter women. New York: Simon & Schuster. Jacobson, N., & Gurman, A. (Eds.). (1986). Clinical handbook of marital therapy. New York: Guilford Press.

Jones, A. (1994a). Next time she'll be dead: Battering and how to stop it. Boston: Beacon Press.

Jones, A. (1994b, September/October). Where do we go from here? Unpublished manuscript, pp. 56-63.

Kelley, H. (1983). Love and commitment. In H. Kelley, E. Berscheid, A. Christensen, J. Harvey, T. Huston, G. Levinger, E. McClintock, I. Peplau, & D. Peterson (Eds.), Close relationships (pp. 265-314). San Francisco: Freeman.

Lempert, L. (1996). Women's strategies for survival: Developing agency in abusive relationships. Journal of Family Violence, 11, 269-289.

MacIntyre, A. (1984). After virtue: A study in moral theory (2nd ed.). Notre Dame, IN: University of Notre Dame Press.

Merritt-Gray, M., & Wuest, J. (1995). Counteracting abuse and breaking free: The process of leaving revealed through women's voices. Care for Women International, 16, 399-412.

Meyers, M. (1997). News coverage of violence against women: Engendering blame. Thousand Oaks, CA: Sage.

National Coalition Against Domestic Violence. http://www.ncadv.org (accessed 10 July 1999).

National Research Council. (1996). Understanding violence against women. Washington, DC: National Academy Press.

Orbuch, T. (1997). People's accounts count: The sociology of accounts. Annual Review of Sociology, 23, 455-478.

Pollock, D. (1990). Telling the told: Performing like a family. Oral History Review, 18, 1-36.

Radway, J. (1991). Reading the romance: Women, patriarchy, and popular literature. Chapel Hill: University of North Carolina Press.

Riessman, C. (1992). Making sense of marital violence: One woman's narrative. In G. Rosenwald & R. Ochberg (Eds.), Storied lives: The cultural politics of self-understanding (pp. 231-249). New Haven, CT: Yale University Press.

Shotter, J. (1993). Conversational realities Constructing life through language. Newbury Park, CA: Sage.

Strauss, A. (1987). Qualitative analysis for social scientists. Cambridge: Cambridge University Press.

Strauss, A., & Corbin, J. (1990). Basics of qualitative analysis. Newbury Park, CA: Sage. Stoltenberg, J. (1989). Refusing to be a man: Essays on sex and justice. Portland, OR: Breitenbush.

Sugarman, D., & Frankel, S. (1996). Patriarchal ideology and wife-assault: A meta-analytic review. Journal of Family Violence, 11, 11-40.

Tjaden, P., & Thoennes, N. (1998). Prevalence, incidence, and consequences of violence against women: Findings from the National Violence Against Women Survey. Atlanta, GA: Centre for Disease Control and Prevention, Centre for Injury Prevention and Control.

Vangelisti, A., Crumley, L., & Baker, J. (1999). Family portraits: Stories as standards for family relationship. Journal of Social and Personal Relationships, 16, 335-368.

Veroff, J., Sutherland, L., Chadiha, L., & Ortega, R. (1993). Newlyweds tell their stories: A narrative method for assessing marital experience. Journal of Social and Personal Relation- ships, 10, 437-457.

Walker, L. (1984). The battered woman syndrome. New York: Springer.

West, J. (1995). Understanding how the dynamics of ideology influence violence between intimates. In S. Duck & J. Wood (Eds.), Understanding relationship processes, 5: Confronting relationship challenges (pp. 129-149). Thousand Oaks, CA: Sage.

Wood, J. T. (1994). Who cares?: Women, care, and culture. Carbondale: University of Southern Illinois Press.

Wood, J. T. (in press). 'That's wasn't the real him': Women's dissociation of violence from the men who enact it. Qualitative Research Reports.

Wood, J. T. (2001). Gendered lives: Communication, gender, and culture (4th ed.). Belmont, CA: Wadsworth.

Wood, L., & Rennie, H. (1994). Formulating rape: The discursive construction of victims and villains. Discourse and Society, 5, 125-148.

Wuest, J. (1995). Feminist grounded theory: An exploration of congruency and tensions between two methods of knowledge discovery in nursing. Qualitative Health Research, 5, 125-137.

Yllo, K. (Ed.). (1988). Perspectives on wife abuse. Newbury Park, CA: Sage.

Correspondence concerning this article should be addressed to Julia T. Wood, Department of Communication Studies, University of North Carolina at Chapel Hill, 110 Bingham Hall, CB #3285, Chapel Hill, NC 27599-3285, USA. [E-mail: jwood1@email.unc.edu]. Terri Orbuch was the Action Editor on this article.