This article examines the theoretical perspective on gender that underlies the research on sex-symmetry in intimate partner violence. A neglected but central issue in this debate is the conceptualization and measurement of gender. Connections between gender and intimate partner violence have become complicated in recent years.
This article examines the theoretical perspective on gender that underlies the research on sex-symmetry in intimate partner violence. A neglected but central issue in this debate is the conceptualization and measurement of gender. Connections between gender and intimate partner violence have become complicated in recent years.
This article examines the theoretical perspective on gender that underlies the research on sex-symmetry in intimate partner violence. A neglected but central issue in this debate is the conceptualization and measurement of gender. Connections between gender and intimate partner violence have become complicated in recent years.
C 2005) DOI: 10.1007/s11199-005-4204-x Theorizing Gender in Intimate Partner Violence Research Kristin L. Anderson 1,2 Research ndings of sex-symmetry in the perpetration of intimate partner assaults have sparked vigorous debate about the appropriate denition and measurement of intimate vio- lence. A neglected but central issue in this debate is the conceptualization and measurement of gender. This article rst examines the often unstated theoretical perspective on gender that underlies the research on sex-symmetry in intimate partner violence. This perspective treats gender as an individual characteristic of persons. Next, I describe challenges to the in- dividualist model of gender from two emerging theoretical perspectivesinteractionist and structuralist gender theories. The article concludes with suggestions for research on intimate partner violence that are informed by these new gender theories. KEYWORDS: gender theory; sex differences; intimate partner violence; typologies; physical aggression. Connections between gender and intimate partner violence have become complicated in recent years. The classic feminist argument that battery represents a manifestation of patriarchal power is under attack in the political sphere (Currie, 1998; Worcester, 2002). Rancorous debates be- tween scholars who identify gender and power as the key processes in partner violence and those who view gender as just one component of the problem have dominated the literature during the past decade (Dobash, Dobash, Wilson, & Daly, 1992; Johnson, 1995; Kurz, 1993; Straus, 1993; Yllo, 1993). The controversy was initiated by a 197778 research study entitled The Battered Husband Syndrome that reported survey research ndings of sex-symmetry in husbands and wives self-reported partner assaults (Steinmetz, 197778). Although the methodology of this study was immediately criticized, ndings from more than 30 additional studies have corroborated Steinmetzs early report (Currie, 1998; Straus, 1993). These sex-symmetry ndings have been regularly cited to challenge fem- 1 Western Washington University, Bellingham, Washington. 2 To whom correspondence should be addressed at Depart- ment of Sociology, Western Washington University, Bellingham, Washington 98225-9081; e-mail: [email protected]. inist constructions of partner violence as a problem of gender and power (Currie, 1998). Mens rights groups have used these ndings to support their claim that feminist activists and scholars have hidden the problem of husband-abuse from public view (Messner, 1998). Researchers in the family violence tradition have proposed that domestic violence perpetrated by women is a social problem worthy of greater attention and investigation (Stets & Straus, 1990; Straus 1993). The sex-symmetry ndings have also been used to divert funding from battered womens shelters (Currie, 1998). Within the academy, debate about the sex- symmetry ndings has centered on the denition and measurement of intimate partner violence. Discus- sions about the validity and reliability of the Con- ict Tactic Scales (CTS), the dominant measure of partner violence among survey researchers, have ap- peared as regular features in social science publica- tions during the last 15 years (Straus, 1992). Critics of the CTS argue that this measure of partner violence is awed because it measures the acts of violence rather than consequences, fails to examine motives for violence, and ignores the larger context of gender inequality (Currie, 1998; Dobash et al., 1992; Kurz, 1993). Moreover, they note that women are vastly overrepresented among victims of partner assaults in studies of battered womens shelters, emergency 853 0360-0025/05/0600-0853/0 C 2005 Springer Science+Business Media, Inc. 854 Anderson rooms, and clinical populations (Dobash et al., 1992; Kurz, 1993; Stark & Flitcraft, 1996). More recently, Johnson (1995) proposed an in- uential methodological explanation for the sex- symmetry nding. Johnson argued that national sur- veys are not representative with regard to partner assault because most violent couples will refuse to participate in such surveys. National surveys thus measure common couple violencea form of part- ner assault that is gender-balanced and less severe than the type of partner assault encountered by re- searchers who study clinical populations, which John- son termed patriarchal terrorism (Johnson, 1995, p. 286). Johnson and Ferraro (2000) proposed that researchers should focus their attention on distin- guishing the distinct types of partner assault uncov- ered by different sampling techniques. According to this argument, the sex-symmetry debate reects the failure to separate these distinct types of violence rather than the inadequate measurement of violence; family violence and feminist researchers are simply studying different populations that engage in distinct types of intimate violence. In this article, I propose that a neglected but central issue in the sex-symmetry debates is theoret- ical rather than methodological. I argue that, amid the controversy over the denition and measurement of violence, intimate violence researchers have ne- glected to recognize that there is equivalent con- troversy over the conceptualization and measure- ment of gender. Confusion about how partner as- saults are gendered is a reection of a larger theoret- ical confusion about what it is that we mean by gen- der (West & Zimmerman, 1987). Theories of gender have changed dramatically in the past three decades, but recognition of these changes has not entered other substantive areas within the social sciences (Alway, 1995; Ferree, 1990; Stacey & Thorne, 1985; Williams, 1991). For example, although Johnsons (1995) proposed resolution to the sex-symmetry de- bate has been benecial in that it focuses attention on the issue of sampling limitations, his categoriza- tion of violence into typessome male-dominated and others gender-balancedrests on a specic conceptualization of gender as sex difference (e.g., this approach assumes that if women and men do the same thing then gender is balanced). Johnson (1995) writes that common couple violence is less a product of patriarchy, and more a product of the less gendered causal processes discussed at length by [scholars] working in the family violence tradition (p. 285). Critics have argued that this sex difference approach to gender vastly oversimplies the complex ways in which gender matters within social relation- ships. Recently, gender scholars have classied three divergent theoretical approaches to the study of genderindividualist, interactionist, and structural- ist (Risman, 1998). Answers to questions about whether and how partner violence is gendered change strikingly depending on the conceptualization of gender utilized by the researcher. For example, if we theorize gender from the interactionist or struc- turalist theoretical perspectives, we may nd that common couple violence is deeply connected to gen- dered social processes. This article rst examines the theoretical approach to gender that underlies most previous studies of sex-symmetry in intimate partner violencethe individualist approach. Next, I apply two competing theories of gender that have emerged in recent decadesinteractionist and structuralist to develop alternative explanations of the relation- ship between gender and intimate violence. I argue that interactionist and structuralist gender theories provide more fruitful approaches to understanding the complex ways in which intimate violence is gen- dered than individualist approaches. INDIVIDUALIST APPROACHES TO THE STUDY OF GENDER Individualist approaches to the study of gender propose that individual persons are gendered beings (Risman, 1998). In this view, masculinity and fem- ininity are traits that men and women incorporate into their selves or identities, either through a bi- ologically determined predisposition to these traits or through socialization processes. Individualist the- ories presuppose that a propensity to use aggression and violence is an innate or learned characteristic of masculine persons. This theory of gender has dominated the re- search on intimate partner violence conducted by scholars in the family violence tradition. However, family violence researchers have rarely described the theory of gender that informs their research. The assumptions about gender that underlie this research have remained implicit. For example, survey research that conceptualizes sex as an in- dependent variable to predict behavior such as violence rests on the implicit assumption that any statistically signicant associations with the depen- dent variable are caused by individual sex (Williams, 1991). Researchers who claim that domestic violence Theorizing Gender in Intimate Partner Violence Research 855 researchers should adopt a gender-neutral frame- work base their argument on ndings that the independent variable sex does not predict vio- lence perpetration in some national survey studies (Dutton, 1994; Felson, 2002; Straus, 1993). Moreover, Johnsons (1995) claim that patriarchal terrorism is male-dominated and common-couple violence is gender-balanced is based on the research nding that rates of common couple violence perpetration do not vary by individual sex. In addition, partner violence studies that uti- lize gender schema scales such as the Bem Sex Role Inventory (BSRI) take an individualist approach to gender. The BSRI and similar measures record in- dividuals assessments of the extent to which they possess traits characterized as masculine or feminine and use these self-reports to categorize people as masculine, feminine, or androgynous (Bem, 1974). In a third individualist approach, researchers classify individuals as traditional or egalitarian on the ba- sis of their self-reported attitudes about mens and womens roles within society. Studies of partner vi- olence that utilize individualist approaches hypoth- esize that biological males and more traditional and more masculine individuals (e.g., men who score high on masculinity or who report traditional gender atti- tudes) will be more likely to engage in partner vio- lence than women and more feminine or egalitarian men (Sugerman & Frankel, 1996). If gender is dened as a property of individu- als, substantial research evidence suggests that there is no relationship between gender and partner vi- olence. Studies that conceptualize gender as a per- sons self-reported maleness or femaleness often fail to nd a relationship between gender and partner as- sault perpetration and victimization (Archer, 2000; Felson, 2002). 3 If gender is dened as a masculine, feminine, or androgynous identity as measured by gender schema scales, the relationship between gen- der and partner assault contradicts the individual- ist hypothesis that more masculine persons are more violent. Previous researchers were surprised to nd 3 Recent victimization surveys are an exception to the pat- tern of sex-symmetry found in most survey studies. The Na- tional Crime Victimization Survey (Bachman & Saltzman, 1995) and the National Violence Against Women survey (Tjaden & Thoennes, 2000) nd substantially higher rates of domes- tic violence victimization among women than men. These sur- veys nd a lower overall incidence of domestic violence vic- timization than studies employing the CTS and may uncover more serious types of violence than surveys employing the CTS. either no relationship between a mans masculinity and violence, or that more feminine men and women reported higher rates of partner assault (Bernard, Bernard, & Bernard, 1985; Burke, Stets, & Pirog- Good, 1988). A meta-analysis of the voluminous re- search on gender role attitudes and partner violence has offered weak and inconsistent support for the hy- pothesis that traditional men engage in higher rates of partner violence than egalitarian men (Sugerman & Frankel, 1996). Findings from intimate violence research that uses individualist approaches to gender suggest that gender is not a particularly important predictor of in- timate partner violence. Some researchers have used these results to suggest that intimate violence is not a gendered phenomenon (Dutton, 1994; Felson, 2002; Straus, 1993). However, in recent decades, gender scholars have identied several limitations of the in- dividualist approach to gender (see Connell, 1995; Deaux & Major, 1990; Risman, 1998). First, and most importantly, the individualist approach reduces gen- der to the behavior of individual women and men. That is, this approach assumes that, if women and men are equally violent within intimate relationships, then partner violence is not gendered. Second, the sex as an independent variable approach cannot tell us why sex differences, if they are statistically signicant, exist. For example, the nding that men engage in higher levels of violence toward intimate partners than women can be inter- preted through theoretical perspectives as diverse and contradictory as feminism and evolutionary psy- chology (e.g., Archer, 2000). Third, this approach cannot easily explain differ- ences within groups of women or men. The empirical research clearly demonstrates that only a small per- centage of women and men perpetrate intimate part- ner violence (Archer, 2000). However, women are biologically similar to other women and men are bio- logically similar to other men. Moreover, within cul- tures, same-sex groups experience similar processes of gender socialization. Thus, it is difcult to explain why only some men and women perpetrate inti- mate partner violence from an individualist gender perspective (Dutton, 1994). Research that employs gender schema scales or measures of egalitarianism overcomes this limitation by recognizing that individ- uals can vary in the extent to which they incorporate gender ideologies into their sense of identity or into their attitudes (Bem, 1974). However, the research largely suggests that gender identities and attitudes are not associated with the perpetration of partner 856 Anderson violence (Bernard et al., 1985; Sugerman & Frankel, 1996). Finally, researchers who conceptualize gender from an individualist framework have used studies of partner assault among homosexual couples to but- tress their claims that partner violence is not a prob- lem of gender and power. Findings of equivalent or higher rates of partner assault among gay and lesbian couples as compared to heterosexual couples have led some scholars to claim that gender is unrelated to domestic violence (Dutton & Golant, 1995, p. 70; Felson, 2002, p. 42). Such claims rest on the implicit assumption that gender is limited to what women and men do in interactions with the opposite sex. This sex differences approach ignores the com- plex ways in which gender operates in social interac- tions between same-sex people. For example, stud- ies of masculinities and violence suggest that much of male violence perpetrated by youth is intended to impress an audience of male peers (Connell, 1995; Messerschmidt, 1993). Moreover, researchers who use the data on same-sex partner violence to claim that gender does not matter ignore the intricate ways in which gender and heterosexuality are constructed in mutually-reinforcing ways in contemporary West- ern societies (Butler, 1990; Connell, 1995). Gay men and lesbian women must negotiate a world in which their masculinity and femininity are called into ques- tion due to their sexual identity. Findings that part- ner assaults occur within gay male and lesbian re- lationships at the same, lower, or higher rates as within heterosexual relationships do not challenge the argument that gender is an important component of intimate partner violence. 4 Rather, these ndings lead to additional questions about how gender in- teracts with heterosexism to inuence the dynam- ics of violence in both heterosexual and homosexual relationships. 4 Several researchers claim that rates of partner violence are simi- lar in heterosexual and homosexual relationships (Dutton, 1994; Felson, 2002; Merrill, 1998). However, most studies of violence in gay male and lesbian relationships do not use probability sampling techniques (Renzetti, 1998). One recent study of part- ner violence in gay male relationships that uses a probabil- ity sample nds that rates of victimization are slightly higher than rates of male-perpetrated violence in studies of heterosex- ual relationships that use similar measures of violence (Green- wood et al., 2002). However, because rates of violence are in- uenced by both sampling techniques and the measurement of violence (Johnson, 1995; Straus, 1992), rates of reported vi- olence from different studies cannot be considered as valid comparisons. Researchers who believe that domestic violence is interconnected with sexism within society need a stronger theoretical language than that offered by the individualist perspective with which to counter the movement away from an analysis of gender within the eld. In recent years, gender theorists have pro- vided such a language. Interactionist and structural- ist theories understand gender as more than the be- havior of individual women and men. Gender exists in the expectations and demands that we place on people and in their desire to meet (or subvert) these expectations (Butler, 1990; Goffman, 1977; West & Zimmerman, 1987). It exists in the way in which sex is used as a basis for the division of labor, in the sex segregation of activities and leisure pursuits, and in other forms of social categorization (Connell, 1995; Risman, 1998). In sum, these theories propose that gender is a characteristic of interactive situations and of social structures in addition to being a character- istic of individuals (Risman, 1998). Interactionist and structuralist gender theories, applied to the data on partner assault, provide new interpretations and new questions about the relationship between gender and intimate partner violence. Moreover, these theories move us away from understanding gender as solely what individual women and men do in cross-sex in- teractions. In this regard, these theories can expose the overly-simplied gender framework that forms the basis for claims of gender-neutrality in intimate partner violence. INTERACTIONIST APPROACHES TO THE STUDY OF GENDER The interactionist approach to the study of gen- der emerged in the 1980s as, in part, a critique of indi- vidualist approaches (Deaux & Major, 1990; Risman, 1998; West & Zimmerman, 1987). This perspective treats gender as a characteristic of social interaction rather than of individual persons. In this view, indi- viduals do gender in daily interactions with others, performing masculinity or femininity in order to live up to social expectations (Butler, 1990; West & Zimmerman, 1987). From this perspective, gender is viewed as an outcome of social practices rather than as an individual characteristic that predicts behavior. Thus, the interactionist approach shifts our thinking from the question of how masculinity causes violence to the question of how violence causes masculinity. Interactionist gender theory suggests new ways of thinking about gender and intimate partner violence. Theorizing Gender in Intimate Partner Violence Research 857 Violence Is Not Gender-Neutral Behavior. Gender Can Be Constructed Through the Practice of Violence The practice of violence, in Western cultures, is perceived as masculine behavior. As gender schema scales such as the BSRI demonstrate, aggression is typed masculine in U.S. culture (Bem, 1974). Because aggression is a component of an idealized cultural image of masculinity, violence can be used to show others that one is a real man. Studies of part- ner violence that utilize interactionist approaches to gender theorize that violence is one means by which men can perform masculinity (Anderson & Umberson, 2001; Hearn, 1998). This approach illuminates qualitative research ndings that suggest that men use violence against their wives or partners in response to a perceived challenge to their position or authority (Dobash & Dobash, 1998; Ptacek, 1988; Totten, 2003). However, the use of violence against female partners to perform masculinity is not uncontested; because women are constructed as the weaker sex within popular culture, mens violence against women can be construed as cowardly and unmanly. Thus, many men minimize and deny the violence that they perpetrate against female partners (Anderson & Umberson, 2001; Dobash & Dobash, 1998; Ptacek, 1988). Mens Violence Is Evaluated and Interpreted Differently From Womens Violence Interactionists posit that gender is a social ac- complishment (West & Zimmerman, 1987) that is based on the denitions and interpretations of oth- ers. Scholars in the interactive tradition argue that audiences expect different performances from men and women and that the same behavior will be dif- ferentially evaluated depending on the perceived sex of the performer and the interactive context (West & Zimmerman, 1987). Thus, the very same violent act that conveys masculinity to others when the ac- tor is perceived to be a man (e.g., punching a man who has insulted his honor) will be interpreted in a different fashion when the actor is perceived to be a woman. This framework helps us to understand re- search ndings that mens violence is viewed more seriously than womens violence. Because violence is dened as masculine within popular culture, au- diences expect, acknowledge, and encourage mens violence as normal behavior. In contrast, audiences may discourage, trivialize, or mock women who en- gage in violence (McCaughey, 1997). There is some evidence that these gendered interpretations inuence the dynamics of violence within intimate relationships. A qualitative analysis of the accounts of violence offered by men in a court- mandated treatment program found that batterers depicted their violence as rational and effectual, but similar violent acts perpetrated by their female part- ners were perceived as irrational and ineffectual (Anderson & Umberson, 2001). College students of both sexes trivialize womens violence as compared to mens violence (Miller & Simpson, 1991). The em- pirical nding that men engage in stalking behav- iors that are physically threatening more often than women (Davis & Frieze, 2000) can be explained, in part, by the idea that women may not be perceived as dangerous in situations where the same acts done by men would be seen as threatening (Frieze, 2000, p. 682). Moreover, these gendered interpretations inu- ence how people respond to partner violence that they experience. Tjaden and Thoennes (2000) found that male victims of partner assaults were signi- cantly more likely than female victims to choose not to report the violence to the police because they de- ned the assault as a minor or one-time. In her study of violence in lesbian relationships, Renzetti (1998) found that women who were victimized by female partners who they dened as physically weaker and more feminine than themselves were unwilling to de- fend themselves against their partners. Gay men who are victims of partner assaults may also be reluctant to report the violence due to a cultural perception that they should be able to stand up to their partners assaults like a man (Merrill, 1998). This perspective also helps us to understand why male victims and criminal justice agents are reluctant to hold female perpetrators accountable for their vi- olence. To be labeled as a victim of an assault by a woman may threaten a mans sense of masculinity. Thus, even men who are injured by a female part- ner are reluctant to have their partners arrested or prosecuted (Anderson & Umberson, 2001). Several studies nd that male victims are less likely than fe- male victims to report the violence to the police and that police are more likely to make an arrest if the victim is female (Buzawa & Hotaling, 2000; Tjaden & Thoennes, 2000). Renzettis (1998) study of bat- tering within lesbian relationships found that many victims reported that their friends, families, and even womens shelters refused to believe that they were 858 Anderson battered due to the feminine appearance of the per- petrator. Violence Will Be Used to Construct Masculinity In Social Situations in Which Masculine Identity Is Threatened The interactionist perspective on gender sug- gests that the interactive context inuences violent behavior. If violence is a means by which men con- struct masculinity, violence should occur in situa- tions in which men desire to demonstrate or per- form masculinity for others (Anderson & Umberson, 2001; Messerschmidt, 1993). One such situation is when a mans masculinity is challenged or called into question. There is a small body of research that pro- vides support for the interactionist claim that part- ner violence is more likely to occur under conditions in which mens sense of masculinity is threatened. Qualitative studies nd that men report perpetrating partner assaults when they perceive a threat to their authority or status (Anderson & Umberson, 2001; Dobash & Dobash, 1998; Ptacek, 1988). Other stud- ies suggest that violence is used in a compensatory fashion when men lack other resources to display masculinity; for example, rates of physical and emo- tional abuse are higher among men whose female partners out-earn them (Anderson, 1997; Kalmuss & Straus, 1990; Kaukinen, 2004). In a recent study, Salari and Baldwin (2002) found that injurious vio- lence was more likely in couples with a greater in- come contribution by the woman. The interactionist approach might also help to explain why rates of intimate partner violence are higher among those who live in poverty. Financial success as a family breadwinner is another ideal- ized characteristic of masculinity in U.S. culture, but not all men are able to live up to this ideal (Fine, Weis, Addelston, & Marusza, 1997; Totten, 2003). Men who are unemployed or involved in part-time or low-wage work may feel that their masculine iden- tities are threatened because they are not breadwin- ners (Anderson, 1997). Research ndings that part- ner violence is correlated with unemployment and low levels of income and education can be under- stood through this interactionist framework (Totten, 2003). Moreover, men who are breadwinners may be able to exert control over their intimate part- ners through their control of economic resources rather than through violence (Stark & Flitcraft, 1996). In sum, interactionist gender theory proposes that violence may be a compensatory method of ex- erting control and constructing masculinity among men who feel that their authority and masculinity have been called into question. This argument ap- plies to mens violence perpetrated against other men as well as their violence perpetrated against female partners. In the language of causal order, the individ- ualist perspective posits that violence is an outcome of gender, whereas the interactionist perspective pro- poses the oppositethat gender is an outcome of violence. Applied to the study of partner violence, the interactionist perspective can help us understand why the practice of violence is evaluated differently depending on the gender attributions applied to the perpetrator and victim. For example, a common- couple violence scenario in which a woman slaps or shoves her male partner and he responds in kind is not gender-neutral in this framework; these equiv- alent acts will be treated differently based on the per- ceived sex of the actor and the audiences interpreta- tion (Was she out of line in slapping him? Is he a brute who deserved it? Is he a coward who should pick on someone his own size?). STRUCTURALIST APPROACHES TO THE STUDY OF GENDER A third and newly emerging approach to gen- der studies contends that gender is a form of social structure. This perspective emphasizes that gender organizes social institutions as well as identities, at- titudes, and interactions. Structuralists contend that gender is a system of stratication that places women and men into unequal categories, roles, and occupa- tions (Risman, 1998). The structural perspective em- phasizes that gender exists as a social force that op- erates independently of individual wishes or desires: Even when men and women do not desire to live gendered lives or to support male dominance, they often nd themselves compelled to do so by the logic of gendered choices (Risman, 1998, p. 29). The structuralist approach is methodologically distinct from the individualist and interactionist approaches in that it conceptualizes gender as a pattern of resource-distribution and social orga- nization rather than as a predictor of individual behavior (Risman, 1998). From a structuralist gen- der perspective, individual sex does not predict violent behavior. Rather, the gender structure inuences womens and mens opportunities and Theorizing Gender in Intimate Partner Violence Research 859 rewards for the use of violent behavior. A number of social structural processes inuence womens and mens experiences of intimate partner vio- lence. For example, marriage, as an institution, is organized by gender. We assign different roles and responsibilities to husbands and wives within marriage. This means that marriage is experienced differently by women and men (Hochschild & Machung, 1989). Similarly, the practice of violence is organized by gender (McCaughey, 1997). Men receive encouragement, support, and training in the use of violence whereas women are barred from its use in some situations (e.g., U.S. military service). Thus, violence is experienced differently by women and men. The structuralist approach to gender has long been utilized by feminist theorists of domestic violence although it has not been named as such. Feminist scholars have repeatedly emphasized that domestic violence must be understood in the context of the larger system of gender inequality (Brush, 1993; Currie, 1998; Dasgupta, 2002; Dobash et al., 1992; Yllo, 1993). However, feminist efforts to rede- ne domestic violence to include this larger context of gender inequality have centered on changing the denition and measurement of intimate violence rather than the denition and measurement of gender. For example, in a review of the gender sym- metry debate, Dasgupta (2002) claims that a broad denition of battering as a pattern of intimidation, coercive control, and oppression is needed so that researchers have to acknowledge the context of social norms and social power differentials between men and women (pp. 13671368). However, no def- inition of violence can fully encapsulate the myriad ways in which gender organizes social relationships. As the quotation from Dasguptas work suggests, what feminist theorists really want domestic violence researchers to consider is how the context of gender inequality inuences the experience of violence. Although efforts to redene domestic violence as a pattern of coercive control have been successful in shaping feminist research, they have not helped to stem the political and academic debate about sex- symmetry. It is not that the measures of violence are inaccurate, but that these measures fail to illuminate how gender inuences and structures violence. In ad- dition to criticizing the measures of violence utilized by sex-symmetry researchers, feminist researchers should note that the sex-symmetry research rests on individualist gender theory and is thus limited in important respects. Structuralist gender theory is needed to under- stand the ways in which women and men encounter different constraints to the perpetration of violence and different barriers to stopping the violence or leaving the relationship. The structuralist theory of gender suggests that an adequate understanding of gender and intimate partner violence must consider the ways in which gender is used to organize social life. Men Receive More Instruction in the Use of Violence Than Women Access to violence is unequally distributed by the gender structure. Boys and men are granted more opportunities to learn to use violence effectively than girls and women (Fagot, Hagan, Leinbach, & Kronsberg, 1985; McCaughey, 1997). Girls and women are discouraged from the use of violence as young children and throughout their formative expe- riences, whereas boys and men are more likely to be encouraged to play sports or apply for jobs in which violence and self-defense skills are taught (Fagot et al., 1985; Messner, 1998). Historically, women have been formally barred from some occupations that in- volve training in the use of violence and they con- tinue to be barred from military combat positions in the U.S. The ways in which gender organizes training in violence has implications for the sever- ity and success of violence perpetrated by men and women. The Consequences of Intimate Violence Will Differ for Women and Men Due to the Larger System of Gender Inequality Consequences for Injury Cultural expectations about male dominance in heterosexual pairings lead to situations in which women are disadvantaged in violent conicts with male partners. In U.S. culture, it is normative for men to have slightly more status than the women with whom they pair; male partners are typically older, physically larger, and have greater education and income than female partners (Goffman, 1977). These advantages mean that men and women will have different resources (physical, economic, psycho- logical) for gaining control over a partner. For ex- ample, numerous studies demonstrate that although 860 Anderson heterosexual men and women perpetrate assaults at similar rates, women are more likely than men to be injured (Archer, 2000; Brush, 1993; Stets & Straus, 1990). Data from the National Violence Against Women Survey (NVAWS), the largest national study of partner violence in the U.S., show that women victimized by male partners are twice as likely to be injured as men victimized by female partners (Tjaden & Thoennes, 2000). This nding has typi- cally been interpreted as resulting from the fact that men are physically larger and stronger than women (Archer, 2000; Felson, 2002). This explanation, how- ever, treats gender (in this case, physical size and strength) as an individual characteristic. Structural- ist gender theory would further explain that mens size advantage is not simply an individual traitit is built into the organization of heterosexual pair- ings. Although men, on average, are slightly heavier and taller than women, on average, there are many women who are larger and taller than many men. Gender norms organize heterosexual pairings such that tall women often seek male partners who are taller than themselves, and short men seek shorter women (Goffman, 1977). Mens size advantage is thus a structural component of heterosexual relation- ships; it is not simply a reection of individual gender characteristics. Consequences for Fear, Depression, and Self-Esteem Victimization studies have documented the neg- ative emotional, social, and psychological conse- quences of partner violence for female victims (Frieze, Hymer, &Greenberg, 1987; Kirkwood, 1993; Stark & Flitcraft, 1996). However, few studies com- pare consequences for male and female victims (Anderson, 2002). The available comparative stud- ies suggest that women experience more negative psychological consequences than men as a result of victimization. Jacobsen et al. (1994), in a laboratory study of arguments between heterosexual couples, found that female partners in violent relationships experienced higher levels of physiological arousal than their boyfriends or husbands, suggesting that women experience more symptoms of distress. Re- search on why victims are reluctant to report abuse to the police nds that women are signicantly more likely than men to choose not to report out of fear of retaliation by their attacker (Tjaden & Thoennes, 2000). In a study of married military couples that engage in mutual violence, Langhinrichsen-Rohling, Neidig, and Thorn (1995) demonstrated that wives were more likely than husbands to experience fear during a violent incident. Intimate violence victimization also has gen- dered consequences for depression. Stets and Straus (1990) found that, among heterosexual couples, fe- male victims were signicantly more likely than male victims to experience stress and depression. Anderson (2002) found that, although domestic as- sault victimization was associated with increased depression and lowered self-esteem among both women and men, the negative effects of violence on well-being were signicantly greater among women. These ndings suggest that the consequences of partner violence are more negative for heterosexual women than for heterosexual men. Although the ex- planation for why this should be the case has not been empirically demonstrated, the structuralist per- spective on gender suggests a number of hypothe- ses. Because women are assigned responsibility for the care and nurturance of others in the relationship, they may be more likely than men to feel guilt and re- sponsibility for their partners violent behavior and this may contribute to depression and lowered self- esteem (Kirkwood, 1993). Langhinrichsen-Rohling et al. (1995) found that wives were more likely than husbands to feel blame and guilt about the violence in their marriage. Women may also suffer more neg- ative psychological consequences from violence vic- timization because they are more likely than men to be injured by violence. Male batterers also isolate their female victims from support networks of fam- ily members and friends (Dobash & Dobash, 1998; Kirkwood, 1993; Walker, 1984). This pattern of iso- lation has also been found in studies of gay male and lesbian relationships (Kirkwood, 1993; Merrill, 1998; Renzetti, 1998). It is currently unknown whether het- erosexual men who are victims of intimate violence also experience this isolation. However, structural- ist gender theory suggests that heterosexual and gay male victims might experience less social isolation due the gendered organization of employment. Men are more likely than women to experience contact with others in the workforce and thus male victims of partner violence might be less likely than female victims to experience such isolation. Consequences for Economic Well-Being Heterosexual women are more likely than het- erosexual men to be economically dependent on their intimate partners. Moreover, women are more Theorizing Gender in Intimate Partner Violence Research 861 likely than men to experience poverty upon disso- lution of a marital or cohabiting relationship (Kurz, 1995). These patterns are consequences of the way in which gender organizes labor and compensation patterns and they have implications for the way in which women and men experience intimate assault. Qualitative studies of heterosexual womens experi- ences of partner violence victimization consistently nd that women feel trapped into violent relation- ships due to their inability to earn enough alone to care for themselves and their children (Kirkwood, 1993; Kurz, 1995). Moreover, the economic conse- quences of leaving a violent relationship are of- ten severe for gay and lesbian people. Studies of the dynamics of violence within gay and lesbian re- lationships nd that violent gay and lesbian part- ners rely on societal heterosexism and the threat of outing their partners to employers as a means of additional dominance and control (Merrill, 1998; Renzetti, 1998). The structuralist gender perspective thus sug- gests that sex and sexual orientation will inuence the barriers to leaving a violent relationship. Because the gender structure assigns women responsibility for the care and well-being of children, threatened cus- tody battles may inhibit womens plans to leave a vi- olent relationship; men have less reason to fear that they will be socially sanctioned if they do not obtain custody following a divorce. In a study of mothers experiences of divorce, Kurz (1995) found a signi- cant relationship between womens partner violence victimization prior to the divorce and the experience of fear during child custody negotiations. Moreover, lesbian mothers might remain in a violent relation- ship due to fears that they will lose custody of their children if they are outed by their partners. These ndings suggest that the gendered organization of the distribution of income, paid work, and responsibility for childrearing creates differential consequences for womens and mens experiences of intimate partner assaults. Rates of Intimate Partner Violence Perpetration Among Women and Men Will Vary by Structural Context Structuralist gender theory is also helpful for analyses of the ways in which rates of partner violence differ across social contexts. One well- documented research nding is that women perpe- trate partner violence more often than men within dating relationships (Frieze, 2000). In marriages, rates of male- and female-perpetration seem to be more equal (Archer, 2000). One explanation for this difference might lie in the way in which gender orga- nizes dating and marital relationships. In dating rela- tionships, women are granted a degree of power in that men are assigned the task of requesting dates and women have the right of refusal (Goffman, 1977). Moreover, women are able to freely leave a dating relationship (Frieze, 2000). Upon marriage, the balance of power shifts in that men are norma- tively considered to be heads of the household and they usually control the economic resources. The risks for womens violence perpetration increase upon marriage because women are often economi- cally dependent on the relationship and the relation- ship is less easily dissolved (Archer, 2000). The structuralist perspective on gender proposes that women and men will experience partner assault differently because they are differentially-situated within societies organized by gender inequality. This perspective suggests that although rates of perpetra- tion and victimization may not vary by sex, the con- sequences of victimization will differ for women and men because these groups face different structural opportunities and constraints. From this perspective, a common-couple violence scenario in which a woman slaps or shoves her male partner and he re- sponds in kind is not gender-neutral; this woman and man will in all likelihood be differentially posi- tioned in terms of size, strength, and economic de- pendency. RESEARCH AGENDAS Debates about howgender inuences partner vi- olence have ourished in recent decades. However, debate has centered on the denition of measure- ment of violence rather than the denition and mea- surement of gender. Bringing gender theory into the debate complicates the discussion of whether women and men are equally violent by asking us to recon- sider what we mean by gender. Do we mean, by gen- der, what women and men do? If so, the debate about sexual symmetry in intimate partner violence is over. We now have sufcient evidence that both women and men engage in partner violence to put the issue to rest (Archer, 2000). If, however, we apply inter- actionist and structuralist gender perspectives to un- derstanding gender and partner assault, there remain numerous unanswered questions. Such theories push 862 Anderson us toward questions of how gender matters rather than whether rates of intimate violence differ by sex. Interactionist theories of gender, as applied to the issue of intimate partner violence, suggest that violence is a means by which those who wish to be seen as manly can demonstrate masculinity for oth- ers. This approach suggests that interactive contexts are part of gender and that we can identify contexts in which male-perpetrated violence should be more likely to occur. The following hypotheses, suggested by interactionist gender theory, should inform future research. H1: Gender Attributions Inuence the Ways in Which Violence Is Dened and Evaluated by Observers Mens violence will be treated more seriously than womens violence due to our cultural stereo- types about masculinity and violence. Women will be viewed as more credible victims than men. Both women and men will trivialize and discount female- perpetrated violence. H2: Mens Perpetration of Violence Will Be More Likely to Occur in Contexts in Which Their Masculine Identities Are Threatened These contexts could include structural disad- vantage (e.g., unemployment) and interactive situa- tions in which men feel challenged or threatened by the actions of their intimate partners. H3: Mens Perpetration of Partner Violence Will, Temporarily, Reinforce Their Sense of Masculine Identity Interactionists propose that violence can be used to perform masculinity. Although a handful of qual- itative studies suggest that violence is successful in bolstering a sense of masculine identity (Anderson & Umberson, 2001; Totten, 2003), the outcome of vio- lence for gender identity remains an empirical ques- tion. Moreover, interactionists propose that gender identity is not a stable quality of individuals, but uid and shifting across time and place (Deaux & Major, 1990; West & Zimmerman, 1987). Thus, it will be methodologically difcult to assess the ways in which violence inuences mens sense of masculinity with- out access to data on the process and context of spe- cic violent incidents. H4: Womens Perpetration of Partner Violence Will Challenge Their Sense of Feminine Identity Women who perpetrate partner assaults might experience guilt and a sense of shame for behavior that does not meet cultural expectations for feminine comportment. Alternatively, as noted above, women might dismiss their own violence as trivial or unim- portant because they view themselves as too weak to do any harm. Structuralist gender theory suggests that the experience of partner violence will differ by sex due to the larger structure of gender inequality within society. The larger social and cultural cir- cumstances that shape gendered experiences of vio- lence have been undertheorized in intimate partner violence research. Specic hypotheses about part- ner violence suggested by structuralist gender theory include the following. H5: Differential Injury Rates by Sex Are Linked to Differential Training and Experience With the Use of Violence and to Advantages in Size and Strength Mens more successful use of violence is not sim- ply natural; it is organized by gender. Of particu- lar interest would be studies that assess whether in- jury differences by sex occur among the minority of heterosexual couples in which women are physically larger and stronger than their male partners or in which women have training in the use of violence. H6: Women Will Be More Likely Than Men to Experience Economic, Social, and Psychological Barriers to Leaving Violent Relationships Compared to male victims, female victims of partner violence will be more likely to be trapped in violent relationships due to economical depen- dency, social isolation, and damaging psychological consequences. H7: Rates and Experiences of Partner Violence Will Vary Across Structural Contexts A small body of existing research documents variation in rates of partner assault across structural contexts. Past studies nd that women perpetrate violence at higher rates in dating than in marital contexts and that rates of perpetration are higher Theorizing Gender in Intimate Partner Violence Research 863 among both women and men in cohabiting relation- ships as compared to marriages (Anderson, 1997; Frieze, 2000). Additionally, survey ndings suggest that U.S. women perpetrate intimate violence at higher rates than women in other nations, a nding that Archer (2000) suggests may be related to the success of feminist movements in the U.S. Although men perpetrate more intimate homicides than women in most societies, U.S. women are more likely than women in other countries to kill their intimate partners (Felson, 2002). Additional studies of contextual variation in the incidence of mens and womens partner violence will help us to understand the interactions between gender inequalities in the larger society and intimate violence. Comparative studies of womens and mens experiences of vio- lence in dating, marital, and cohabiting contexts are needed to explain why rates of intimate violence perpetration by sex vary across these contexts. CONCLUSION In their recent review of domestic violence scholarship in the 1990s, Johnson and Ferraro (2000) criticized domestic violence researchers for treating violence a unitary phenomenon. We need to further recognize that gender is not a unitary phenomenon that can be reduced to self-reported sex (Connell, 1995; Deaux & Major, 1990). Gender is multidimen- sional; it has consequences for individual identities and attitudes, social interaction, and structural position. Feminist critiques of the use of gender as a vari- able (e.g., Williams, 1991) have not been recognized within the academic debate about sex symmetry in intimate violence. The use of sex, uncritically, as a predictor of violent behavior suggests that individ- ual sex causes violent behavior, reifying essential- ist notions of gender difference (or a lack thereof). This individualist gender framework underlies recent claims of gender-neutrality in intimate partner vio- lence. Scholars who believe that a gender-lens is cru- cial for an understanding of partner violence need a stronger theoretical language with which to counter the claims drawn from sex-symmetry research. We must clarify that the individualist approach to gen- der reduces our chances of understanding the com- plex ways in which violence is gendered within societies. Interactionalist and structuralist approaches to gender may provide more fruitful contributions to our understanding of the relationship between gender and partner violence than individualist ap- proaches. The hypotheses informed by interaction- ist and structuralist gender theories that are pro- posed here will require comparative data on womens and mens experiences of intimate violence victim- ization and perpetration across social contexts. Pre- vious intimate partner violence research has largely studied women as victims and men as perpetrators (Anderson, 2002). Although there are a number of recent studies of women as perpetrators of domes- tic violence, these studies do not compare the corre- lates and characteristics of female versus male per- petrators (Babcock, Miller, & Siard, 2003; Stuart, Moore, Ramsey, & Kahler, 2004). Research on vio- lence within gay and lesbian relationships, to date, has largely focused on exploring violence within one group (e.g., gay men or lesbians) and thus does help us to understand how rates, correlates, or con- sequences of violence may differ by sex and sex- ual orientation (Greenwood et al., 2002; Merrill, 1998; Renzetti, 1998). Tjaden and Thoennes (2000) comparative analysis of intimate violence among same-sex and opposite-sex cohabiting relationships nds that rates of intimate violence victimization are higher among both men and women who have lived with male partners than with female partners. How- ever, their study does not assess whether correlates or outcomes of victimization vary by sexual orienta- tion. Additional comparative research that examines the correlates and consequences of intimate violence by data by sex and sexual orientation is needed to understand the ways in which gender and sexuality matter to the experience of intimate violence. The vast majority of previous studies on gender and partner violence have used an individualist ap- proach to gender. Interactionist and structuralist gen- der theories can help us to understand that, although men and women may report perpetrating violence at similar rates, this does not mean that intimate vi- olence is not a gendered problem. Gender is more than a characteristic of individuals. It also exists in the larger societal expectations for behavior and in the ways in which men and women are differentially situated within social institutions. REFERENCES Alway, J. (1995). The trouble with gender: Tales of a still missing feminist revolution in sociological theory. Sociological The- ory, 13(3), 209228. Anderson, K. L. (1997). Gender, status, and domestic violence: An integration of feminist and family violence approaches. Jour- nal of Marriage and the Family, 59, 655669. 864 Anderson Anderson, K. L. (2002). Perpetrator or victim?: Relationships be- tween intimate partner violence and well-being. Journal of Marriage and Family, 64, 851863. Anderson, K. L., & Umberson, D. (2001). Gendering violence: Masculinity and power in mens accounts of domestic vio- lence. Gender and Society, 15, 358380. Archer, J. (2000). Sex differences in aggression between hetero- sexual partners: A meta-analytic review. Psychological Bul- letin, 126, 651680. Babcock, J. C., Miller, S. A., & Siard, C. (2003). To- ward a typology of abusive women: Differences between partner-only and generally violent women in the use of violence. Psychology of Women Quarterly, 27, 153 162. Bachman, R., & Saltzman, L. E. (1995). Violence against women: Estimates from the redesigned survey August 1995 (NCJ- 154348 Special Report). Washington, DC: Bureau of Justice Statistics. Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155162. Bernard, J. L., Bernard, S. L., & Bernard, M. L. (1985). Courtship violence and sex-typing. Family Relations, 34, 573576. Brush, L. D. (1993). Violent acts and injurious outcomes in mar- ried couples: Methodological issues in the National Survey of Families and Households. In P. B. Bart & E. G. Moran (Eds.), Violence against women: The bloody footprints (pp. 240251). Newbury Park, CA: Sage. Burke, P. J., Stets, J. E., &Pirog-Good, M. A. (1988). Gender iden- tity, self-esteem, and physical and sexual abuse in dating rela- tionships. Social Psychology Quarterly, 51(3), 272285. Butler, J. (1990). Gender trouble: Feminism and the subversion of identity. New York: Routledge. Buzawa, E., & Hotaling, G. (2000). The police response to domes- tic violence calls for assistance in three Massachusetts towns: Final report. Washington, DC: National Institute of Justice. Connell, R. W. (1995). Masculinities. Berkeley, CA: University of California Press. Currie, D. H. (1998). Violent men or violent women: Whose deni- tion counts? In R. K. Bergen (Ed.), Issues in intimate violence (pp. 97111). Thousand Oaks, CA: Sage. Dasgupta, S. D. (2002). A framework for understanding womens use of nonlethal violence in intimate heterosexual relation- ships. Violence Against Women, 8, 13641389. Davis, K. E., & Frieze, I. H. (2000). Research on stalking: What do we know and where do we go? Violence and Victims, 15, 473487. Deaux, K., & Major, B. (1990). A social-psychological model of gender. In D. L. Rhode (Ed.), Theoretical Perspectives on Sex- ual Difference (pp. 8999). London: Yale University Press. Dobash, R. E., & Dobash, R. P. (1998). Violent men and violent contexts. In R. E. Dobash & R. P. Dobash (Eds.), Rethinking violence against women (pp. 141168). Thousand Oaks, CA: Sage. Dobash, R. P., Dobash, R. E., Wilson, M., & Daly, M. (1992). The myth of sexual symmetry in marital violence. Social Problems, 39, 7191. Dutton, D. G. (1994). Patriarchy and wife assault: The ecological fallacy. Violence and Victims, 9, 167182. Dutton, D. G., & Golant, S. K. (1995). The batterer: A psychologi- cal prole. New York: Basic Books. Fagot, B., Hagan, R., Leinbach, M. B., & Kronsberg, S. (1985). Differential reactions to assertive and communicative acts of toddler boys and girls. Child Development, 56, 1499 1505. Felson, R. B. (2002). Violence and gender reexamined. Washing- ton, DC: American Psychological Association. Fine, M., Weis, L., Addelston, J., & Marusza, J. (1997). (In)secure times: Constructing White working-class masculinities in the late 20th century. Gender and Society, 11, 5268. Ferree, M. M. (1990). Beyond separate spheres: Feminism and family research. Journal of Marriage and the Family, 52, 866 884. Frieze, I. H. (2000). Violence in close relationshipsDevelopment of a Research area: Comment on Archer (2000). Psychologi- cal Bulletin, 126, 681684. Frieze, I. H., Hymer, S., & Greenberg, M. S. (1987). Describing the crime victim: Psychological reactions to victimization. Profes- sional Psychology: Research and Practice, 18, 299315. Goffman, E. (1977). The arrangement between the sexes. Theory and Society, 4, 301331. Greenwood, G. L., Relf, M. V., Huang, B., Pollack, L. M., Canchola, J. A., & Catania, J. A. (2002). Battering victimiza- tion among a probability-based sample of men who have sex with men. American Journal of Public Health, 92(12), 1964 1969. Hearn, J. (1998). The violences of men: How men talk about and how agencies respond to mens violence against women. Thou- sand Oaks, CA: Sage. Hochschild, A., & Machung, A. (1989). The second shift. New York: Avon. Jacobsen, N. S., Gottman, J. M., Waltz, J., Rushe, R., Babcock, J., & Holzworth-Munroe, A. (1994). Affect, verbal content, and psychophysiology in the arguments of couples with a violent husband. Journal of Consulting and Clinical Psychology, 62, 982988. Johnson, M. P. (1995). Patriarchal terrorism and common couple violence: Two forms of violence against women. Journal of Marriage and the Family, 57, 283295. Johnson, M. P., & Ferraro, K. J. (2000). Research on domestic vi- olence in the 1990s: Making distinctions. Journal of Marriage and the Family, 62, 948963. Kalmuss, D. S., & Straus, M. A. (1990). Wifes marital dependency and wife abuse. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp. 369382). New Brunswick, NJ: Transaction. Kaukinen, C. (2004). Status compatibility, physical violence, and emotional abuse in intimate relationships. Journal of Mar- riage and Family, 66, 452471. Kirkwood, C. (1993). Leaving abusive partners. London: Sage. Kurz, D. (1993). Physical assaults by husbands: A major social problem. In R. J. Gelles & D. R. Loseke (Eds.), Current con- troversies on family violence (pp. 88103). Newbury Park, CA: Sage. Kurz, D. (1995). For richer, for poorer: Mothers confront divorce. New York: Routledge. Langhinrichsen-Rohling, J., Neidig, P., & Thorn, G. (1995). Vio- lent marriages: Gender differences in levels of current vio- lence and past abuse. Journal of Family Violence, 10, 159176. McCaughey, M. (1997). Real knockouts: The physical feminism of womens self-defense. New York: New York University Press. Merrill, G. S. (1998). Understanding domestic violence among gay and bisexual men. In R. Bergen (Ed.), Issues in intimate vio- lence (pp. 129141). Thousand Oaks, CA: Sage. Messerschmidt, J. (1993). Masculinities and crime: A critique and reconceptualization of theory. Lanham, MD: Rowan & Little- eld. Messner, M. A. (1998). The limits of the male sex role: An analy- sis of mens liberation and mens rights movements discourse. Gender and Society, 12, 255276. Miller, S. L., & Simpson, S. S. (1991). Courtship violence and so- cial control: Does gender matter? Law and Society Review, 25, 335365. Ptacek, J. (1988). Why do men batter their wives? In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse (pp. 133157). Newbury Park, CA: Sage. Renzetti, C. M. (1998). Violence and abuse in lesbian relation- ships: Theoretical and empirical issues. In R. Bergen (Ed.), Theorizing Gender in Intimate Partner Violence Research 865 Issues in intimate violence (pp. 117127). Thousand Oaks, CA: Sage. Risman, B. (1998). Gender vertigo: American families in transition. London: Yale University Press. Salari, S. M., & Baldwin, B. (2002). Verbal, physical, and injuri- ous aggression among intimate couples over time. Journal of Family Issues, 23, 523550. Stacey, J., & Thorne, B. (1985). The missing feminist revolution in sociology. Social Problems, 32(4), 301316. Stark, E., & Flitcraft, A. (1996). Women at risk: Domestic violence and womens health. Thousand Oaks, CA: Sage. Steinmetz, S. K. (197778). The battered husband syndrome. Vic- timology, 2, 499509. Stets, J. E., & Straus. M. A. (1990). Gender differences in re- porting marital violence and its medical and psychologi- cal consequences. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp. 151165). NJ: Transaction. Straus, M. A. (1992). The conict tactics scales and its crit- ics: An evaluation and new data on validity and reli- ability. In M. A. Straus & R. J. Gelles (Eds.), Phys- ical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp. 4973). NJ: Transaction. Straus, M. A. (1993). Physical assaults by wives: A major social problem. In R. J. Gelles & D. R. Loseke (Eds.), Current con- troversies on family violence(pp. 6787). Newbury Park, CA: Sage. Stuart, G. L., Moore, T. M., Ramsey, S. E., & Kahler, C. W. (2004). Hazardous drinking and relationship perpetration and victim- ization in women arrested for domestic violence. Journal of Studies on Alcohol, 65, 4657. Sugerman, D. B., & Frankel, S. L. (1996). Patriarchal ideology and wife-assault: A meta-analytic review. Journal of Family Vio- lence, 11, 1340. Tjaden, P., & Thoennes, N. (2000). Extent, nature, and conse- quences of intimate partner violence: Findings from the Na- tional Violence Against Women Survey. Washington, DC: Na- tional Institute of Justice/Centers for Disease Control and Prevention. Totten, M. (2003). Girlfriend abuse as a form of masculinity con- struction among violent, marginal male youth. Men and Mas- culinities, 6, 7092. Walker, L. (1984). The battered woman syndrome. New York: Springer. West, C., & Zimmerman, D. H. (1987). Doing gender. Gender and Society, 1, 125151. Williams, C. L. (1991). Case studies and the sociology of gender. In J. R. Feagin, A. M. Orum, & G. Sjoberg (Eds.), The case for the case study (pp. 224243). Chapel Hill: University of North Carolina Press. Worcester, N. (2002). Womens use of force: Complexities and challenges of taking the issue seriously. Violence Against Women, 8, 13901415. Yllo, K. (1993). Through a feminist lens: Gender, power, and vio- lence. In R. Gelles & D. L. Loseke (Eds), Current controver- sies on family violence (pp. 4762). Newbury Park, CA: Sage.