Introduction

For too long now, it has been considered taboo to study the behaviors of women involved in violent relationships. Well-known feminist researchers have expressed the view that “spousal violence is to all extents and purposes wife beating” (Dobash and Dobash 1977–78, p. 439) and thus “wives [are] the ‘appropriate’ victims of marital violence” (Dobash and Dobash 1977–78). Unfortunately, these sorts of attitudes seem to have inadvertently stifled progress in the research and treatment of partner violence. In a subset of violent relationships, women are the exclusive victims of abuse. However, in many violent couples, both partners engage in physical aggression (Fergusson et al. 2005; Straus 2008). The restriction to limit consideration of IPV to only that which is perpetrated by men is itself sexist. Thankfully, Langhinrichsen-Rohling’s (2010) article serves as an important indicator that some in the field of partner violence research are making progress and moving past this short-sighted and sexist view, to allow equal consideration of the behaviors and needs of men and women involved in violent relationships.

Langhinrichsen-Rohling (2010) outlines five controversies or challenges that exist within the field of intimate partner violence (IPV) in the U.S. today. Prominent among these are the conflicting theories and findings of “feminist” (see Yllö 1993) versus “family violence” (see Straus 1999) groups (or “camps,” as they are often called), which center on the question of whether or not relationship violence is symmetrical. Langhinrichsen-Rohling (2010) provides a fair assessment of the feminist and family violence camps, including the different methodologies and major strengths and weaknesses of each view. Unfortunately, the methodologies of both groups appear to contribute to a problem of restricted range among their data, with feminist researchers typically studying more severe forms of violence among clinical populations and family violence researchers examining community samples of (less) violent couples. Considering the problems that sampling issues cause for both of these groups, it seems inappropriate for either group to claim to be entirely correct, or that the other is entirely wrong. Both camps have only part of the entire picture.

Highlighting the discrepancies between feminist and family violence research is an important first step in resolving this controversy, which has to be made a priority, as outsiders surely are made more skeptical of the validity of discrepant research findings within a single, although sometimes divided, field of study. Agreeing to disagree is not an option. Resolution, via explanation of discrepant findings, is a must. Without this, movement from theoretical research to clinical application will continue to be hindered. This underscores the need for unifying theories in the field of partner violence which can help to explain empirically valid but conflicting data. Conveniently, the second controversy discussed by Langhinrichsen-Rohling (2010) has helped move the field forward, toward this goal.

The second issue addressed by Langhinrichsen-Rohling (2010) relates to the state and usefulness of research into typologies within IPV. One major advantage of the typology literature has been the ability of these researchers to begin to bridge the gap between the feminist and family violence fields. Most notably, in his typology of violent couples, Johnson (1995; Johnson and Ferraro 2000) was able to validate the findings from both camps by illustrating the different types of couples being studied by each group. Langhinrichsen-Rohling integrates the “meta-typology” proposed by Holtzworth-Munroe and Stuart (1994) with the Johnson and Ferraro (2000) typology, to create a new typology, incorporating concepts of intimate partner control and emotion dysregulation. While the idea of typologies is intuitively appealing and a convenient heuristic, clustering batterers into groups is empirically flawed. An important next step may involve deconstructing such typologies in order to examine dimensionally the major factors that comprise them (see Ross and Babcock (in press)). A dimensional application of the principles underlying typologies may more adequately represent reality, as not all couples fit neatly into theoretically delineated categories, a finding that Johnson and colleagues recognized (Johnson and Ferraro 2000) and empirically validated (Leone et al. 2004).

The third controversy discussed in the paper involves the importance of understanding bi-directionally violent relationships. This is perhaps the most controversial and most important point made by Langhinrichsen-Rohling (2010). Evidence that women and men engage in similar rates of IPV has existed for at least the last 30 years (see Straus and Gelles 1986), yet these data were explained away or ignored, ironically, in an effort to avoid being labeled as sexist. For example, Langhinrichsen-Rohling writes about how researchers who uncovered IPV symmetry via empirical research have “struggled to interpret these... findings in… non-victim blaming ways” (2010, this issue). But who is “the” victim in bi-directionally violent relationships? It seems that, in reality, the “struggle” has been trying to interpret findings of IPV symmetry in non-woman blaming ways, whether appropriate or not. As a result, we know relatively little about the behavior of partner violent women compared to partner violent men.

Feminism, by definition, is the belief in equality between men and women. Women, then, should be held equally accountable for their behavior. Some women use drugs, commit crimes, and do morally reprehensible things and it is not anti-feminist to study them and their behavior. In the case of IPV, equal consideration of perpetration may ultimately portray some women in a less than flattering light but it is not anti-feminist. Unless her aggression is enacted in self-defense, a woman who engages in partner violence should bear responsibility for her actions. More importantly, information about the causes, consequences, and context of women’s IPV is sorely needed. Lack of research on potential treatment targets has left us in the dark as to how to best to intervene clinically with partner assaultive women.

In the past, bi-directional partner violence was thought to be the fault of the man. Only he was arrested and sent to treatment. While women benefitted by this preferential treatment, they were also at a disadvantage because they were clinically ignored. Their needs were unexamined and unfulfilled. Just as research on partner violent men has shown a number of correlates that, in and of themselves, could benefit from intervention (e.g., histories of violence in their families of origin, personality dysfunction, substance use, and emotion dysregulation; Feingold et al. 2008), aggressive women are likely to have similar psychosocial problems. Indeed, family of origin violence, substance misuse, and trauma histories have been found to be problems at least among certain subtypes of aggressive women (Babcock et al. 2003; Luthra and Gidycz 2006). If women are violent toward an intimate, and research indicates that partner violence is frequently accompanied by other characteristic dysfunction, then not allowing investigation into this population has seriously hindered the ability of researchers and practitioners to gain insight into and help modify the maladaptive, aggressive behaviors of these women.

In her discussion of bi-directionally violent couples, Langhinrichsen-Rohling (2010) goes one step further by boldly pointing out the need for models that also help to explain uni-directionally violent relationships characterized by a female perpetrator and male victim. Even in recruiting samples for male-perpetrated IPV, we have found relationships marked by one-sided, female-perpetrated violence (Ross and Babcock (2009)). Still, examination of women’s behaviors should not be reserved for those cases in which she is the sole perpetrator of aggression. Even in the case of male-perpetrated violence, examination of women’s behaviors may help to highlight antecedents to men’s violence and/or avenues for treatment for female victims (e.g., safety planning or reasons for not leaving an abuser). The implicit warning against examining the behavior of women involved in violent relationships has hindered progress in the field of IPV for too long, and largely to the detriment of those who this mindset was intended to protect.

While multiple typologies exist within the field of IPV research, they tend to focus on male perpetrators (e.g., Gottman et al. 1995; Hamberger et al. 1996; Holtzworth-Munroe and Stuart 1994). Relatively little attention seems to have been paid to mutually-violent couples, which is unfortunate considering that bi-directional IPV seems to be the most prevalent form in the US. With most partner violence being bi-directional, it is also likely that subtypes of mutually violent couples exist, each with different motives and consequences.

In response to the immediate need for a greater understanding of bi-directionally violent relationships, Langhinrichsen-Rohling (2010) outlines a new typology of violent couples, comprised of three subtypes. Broadly speaking, two subtypes would be considered to be “characterological” in nature, suffering from personality disorder features or high need for control (Babcock et al. 2007). The third type would be a “situational” type, characterized by communication deficits but low levels of psychopathology.

In Langhinrichsen-Rohling’s (2010) model, subtypes are referred to as “dyadic,” in part, because each member of the couple is expected to show the characteristic dysfunction. For the “Dyadic Dysregulation” bi-directionally violent couples, both partners are thought to have preoccupation with their partner, abandonment fears, and emotional dysregulation, much like Holtzworth-Munroe and Stuart’s (1994) Borderline-Dysphoric batterer. Both individuals are also assumed to encounter similar problems related to emotionality in their other, very close relationships, such as those with very close friends or family. In the “Dyadic Domination” subtype, which is based on the Mutually-Violent Control couple subtype (Johnson and Ferraro 2000), both partners are thought to have preoccupation with intimate partner control. These individuals are expected to struggle for power and control with others, across their intimate relationships, not just with their current partner. The third subtype, “Dyadic/Reciprocal Couple Violence,” is a blend of Holtzworth-Munroe and Stuart’s (1994) Family-Only batterer and Johnson’s (1995) Situational Couple Violence. Here, violence occurs as a type of cycle within a particular relationship that results from initiation of aggression in a form that is relatively, socially supported (e.g., women’s use of minor violence in response to male impropriety), followed by a type of retaliatory violence by the partner.

While there is considerable overlap with existing typologies, Langhinrichsen-Rohling’s (2010) dyadic typology differs in some important ways. First, it focuses specifically on bi-directionally violent couples and does not attempt to explain the uni-directionally violent extremes (e.g. Johnson and Ferraro’s (2000) “Intimate Terrorists”). Unlike the Holtzworth-Munroe and Stuart (1994) typology of batterers, Langhinrichsen-Rohling’s (2010) model focuses on the dyad, the couple, not the individual, to understand the violence in a relational context. Langhinrichsen-Rohling’s dyadic typology, although narrower than other typologies, has a number of strengths. In addition to building on existing typology research, Langhinrichsen-Rohling (2010) presents a model, the “Dyadic Culture-Family-Attachment-Skill Deficit Model,” which highlights key variables associated with bi-directionally violent couples. This model suggests that the development of bi-directional relationship violence is a function of each partner’s attachment style and personality, which is influenced by broader, cultural norms (e.g., norms about violence and gender). However, it is somewhat unclear how the variables in the “Dyadic Culture-Family-Attachment-Skill Deficit Model” contribute to the development of partner violence over time. While the author provides some detail on the developmental process for the Dyadic/Reciprocal couple, there is no such a discussion on how the remaining two subtypes develop over time. A more complete theory would outline a differential developmental pathway for each subtype.

Langhinrichsen-Rohling (2010) also outlines a number of predictions based on her typology. She describes certain behaviors that one should expect from members of each subtype, such as emotion dysregulation across close relationships for partners engaged in Dyadic Dysregulation and control struggles with others, in Dyadic Domination. Still, further development of this model may help to prioritize research into the typology. In particular, it would be helpful if a broader set of predictions based on each subtype were outlined. For example, aside from attempting to control others, what other behaviors might one expect to see from an individual engaged in Dyadic Domination?

One of the most important contributions and greatest strengths of the Langhinrichsen-Rohling (2010) dyadic typology is the explicit focus on the couple as the unit of investigation. Indeed, the symmetry or asymmetry of relationship violence can only be assessed when data is collected on both partners. From a systems perspective, the couple represents a dynamic unit where the affects, behaviors, and cognitions of each individual continually influence the affects, behaviors, and cognitions of the other. Outside of this context, an individual’s behavior becomes difficult to fully understand. Understanding a phenomenon or behavior is one of the most basic goals of research, as well as an important prerequisite in predicting, and ultimately influencing or controlling, that behavior. If the field of psychology hopes to have a positive influence on IPV, a greater understanding of this behavior, in the context of the relationship where it occurs, is required.

While an examination of bi-directional IPV is important, such an endeavor is inherently limited in the scope of information it can provide about partner violence overall. By excluding one-sided violence from her model, Langhinrichsen-Rohling’s (2010) dyadic typology cannot satisfy the need for well-developed, unifying theories of violence that will ultimately help to bridge the gap between feminist and family violence researchers. In limiting subtypes to bi-directionally violent couples, this typology does not bring us closer to the goal of explaining the conflicting data that has emerged from the two divided sides of partner violence research. Ideally, a typology of IPV would encompass all subtypes of partner violence, including relationships marked by one-sided violence. Arguments can be made for the importance of specialized research aimed at providing detailed information on a specific subtype of IPV, and this is one step toward improving our understanding of partner violence in general. However, less progress has been made in “bridging the gap” when each camp goes about studying only one particular type of violence to the exclusion of the others. At some point, typologies of partner violence must take into consideration the broad variability in types of violent couples that exist. Yet research may demonstrate that dimensional approaches to the study of IPV are preferable, versus attempting to generalize or synthesize across different typologies.

A dimensional approach to studying relationship violence may also help to resolve one of the potential limitations of Langhinrichsen-Rohling’s (2010) dyadic typology. This typology involves three subtypes in which partners are matched on a defining characteristic. For example, in Dyadic Dominance, both partners are expected to be controlling toward one another. In Dyadic Dysregulation, both partners should have deficits in emotional and behavioral self-control. While there is evidence that both of these characteristics (dominance/control and dysregulation) are helpful in making predictions about the behavior of violent individuals (Graham-Kevan, and Archer 2008; Holtzworth-Munroe et al. 2003), it is not necessarily the case that both partners will always be matched on these traits. For example, Johnson’s (1995) subtype called “Violent Resistance” describes an individual who engages in low-level violence and whose partner is thought to be substantially more controlling than them. Additionally, it should not be assumed that these traits will never occur within the same individual. On the contrary, with the co-occurrence of “Cluster B” personality disorders, and Antisocial and Borderline Personality Disorders in particular, there are surely some individuals who engage in intimate partner control while also having difficulty regulating their own emotions. In fact, an individual with Borderline traits (e.g., emotional instability, dependence, fear of abandonment) may attempt to control his or her partner’s proximity via antisocial means, such as violence, as a result of their emotional upset at the thought of being abandoned (Gottman et al. 1995).

Moreover, the field of psychology is moving away from typologies. There have been increasing questions about adequacy of categorical approaches and dimensional alternatives may be methodologically superior (Jablensky 2005; Widiger 2005). In fact, there is evidence that some phenomena central to the study of IPV, which have been treated as categorical in nature, are better represented dimensionally (Rothschild et al. 2003). Even the next version of the DSM is expected to replace personality disorder diagnostic categories with a dimensional approach (see Widiger and Trull 2007).

The goal of typology research in IPV is to help differentiate “types” of batterers in order to improve our ability to make predictions about their behavior and, ultimately, to influence this behavior (e.g., in treatment). However, not all individuals will fit neatly within a particular subtype. There may be important differences between members of the same group and/or important similarities across groups (see Holtzworth-Munroe et al. 2003). Instead of synthesizing across typologies, an alternative approach would involve selecting out the major factors which have been used to differentiate subtypes and examine these as dimensional predictors of IPV. In some cases, important information may be gleaned from a dimensional approach that might otherwise be obscured with the use of categorical approaches (e.g., Ross and Babcock 2009). Perhaps instead of attempting to predict the behavior of a batterer based on group membership, or trying to tailor treatments to a particular subtype of batterer, IPV researchers could begin to extract those variables that have consistently been shown to contribute to IPV and identify specific intervention techniques for a particular behavior. In this way, a type of a la carte treatment menu could be tailored to fit the needs of each individual.

Research on IPV has already provided the background to allow selection of a number of important treatment targets (proximal risk factors), such as emotion dysregulation, acceptance of violence, insecure attachment, patriarchal or sexist views, impulsivity, substance abuse, personality pathology, and mood disorders (Holtzworth-Munroe and Meehan 2004; Holtzworth-Munroe and Stuart 1994; Johnson 1995; Schumacher et al. 2001). Notably, a majority of this research has been aimed at uncovering antecedents to male-perpetrated IPV. More research is needed to determine whether these risk factors apply equally to female-perpetrated IPV.

The need for more research on the topic of female-perpetrated IPV is supported in Langhinrichsen-Rohling’s (2010) discussion of the fourth challenge involving gender and IPV: gender differences in motives for partner violence. Relatively little is known about motives for women’s violence, as compared to men’s. Although a pattern of physical self-defense has been described among women in certain types of violent relationships (e.g., Johnson’s “Violent Resistance”), some women do initiate relationship violence (Archer 2000; Capaldi et al. 2007). In some relationships, women admit to being the sole perpetrator (Ross and Babcock 2009). Self-defense is clearly not the only motive for all women’s violence. The degree to which findings on men’s IPV motives generalize to women can only be established by empirical investigation of women’s motives.

Despite the reticence to study women’s violence and the fear that one might be accused of “victim-blaming” or being part of a feminist “backlash” movement, there are a number of important reasons for this line of research. First, as mentioned above, perpetration of partner violence is often associated with related problems that could benefit from intervention. If women’s violence, like men’s, is related to emotion dysregulation, mood disorders, substance abuse, or personality dysfunction (see Dutton et al. 2005), then researchers and practitioners should be invested in identifying these women and addressing their treatment needs. Second, some men are victims of IPV and they deserve acknowledgment and assistance, just as female victims of IPV do. Furthermore, children exposed to interparental violence deserve assistance, regardless of the gender of the offending parent. Even if severe, one-sided IPV is perpetrated relatively less often by women, does this mean we should ignore the issue? In homes where women are the sole aggressor, primary aggressor, or even the equal aggressor, their partners and/or their children may be suffering because of this behavior. It is socially irresponsible to ignore this. Finally, a woman’s violence may increase her risk of IPV victimization. In a study of antecedents to men’s IPV, women’s violence was one of the most common predictors of men’s violence (Ross and Babcock 2007). As Langhinrichsen-Rohling (2010) suggests, an examination of women’s motives for IPV is an important step in understanding IPV, and possibly in interrupting the cycle of family violence.

As intimate partner control has been one major factor used to subtype batterers (e.g., Johnson 1995), and a potential moderator of violence severity or predictor of partner injury (Graham-Kevan and Archer 2008), control as a motive for women’s violence is one important avenue of research. However, this may require examining the construct of control first, in order to determine whether the same behaviors that are controlling for men are also controlling when enacted by women. It is doubtful that women’s control tactics are exactly the same as those of men.

The fifth and final controversy addressed by Langhinrichsen-Rohling (2010) involves the need to improve upon current IPV treatments. Considering the prevalence of IPV as a major public health concern in the U.S., the effectiveness of current treatments is unacceptable. Even more unacceptable is the fact that most treatment programs do not systematically collect data to verify the effectiveness of their intervention methods. Here, Langhinrichsen-Rohling highlights one of the biggest problems in the field of IPV today-- the disconnect between empirical findings and public policy.

If IPV is a male-perpetrated crime against women resulting from patriarchal beliefs, and if clinical/treatment samples contain only these types of men (intimate terrorists), then treatments rooted in feminist theory should be effective. Unfortunately, research suggests otherwise. Data on the effectiveness of current IPV interventions, most of which are based on these types of assumptions, is discouraging (Babcock et al. 2004). The ineffectiveness of current IPV treatments may stem, at least in part, from the fact that IPV is not always a male-perpetrated reflection of patriarchal sexism and not all members of clinical samples are engaged in intimate terrorism. This may be increasingly the case with the implementation of mandatory arrest policies. For example, in their sample of women arrested and convicted on a domestic violence charge, Henning et al. (2006) found that the largest group of women (33%) could be classified as partners in mutually-violent relationships.

For some perpetrators of IPV, their violence is thought to be a reflection of the underlying pathology that they carry with them across all of their relationships. For these types of aggressors, individual treatment to address their pathology may be most suitable (e.g., cognitive behavioral therapy, substance abuse treatment, Duluth-type interventions targeting power and control issues). In contrast, some IPV is low-level, bi-directional, and the result of the particular relationship dynamic that has developed between two relatively normal individuals. As Langhinrichsen-Rohling (2010) reports, data suggest that conjoint treatment may be helpful for some of these couples. In determining the appropriate type of treatment, it might be necessary to first assess whether IPV in a particular relationship is a reflection of one partner’s pathology, a dysfunctional relationship dynamic between two functional individuals, or both. Langhinrichsen-Rohling’s (2010) dyadic typology addresses both of these points. Conjoint treatment may be best suited for the dyadic/reciprocal couple, while individuals involved in with emotional dysregulation or dysfunctional needs for control may be benefit more from individual treatment to address their personal pathology.

Just as it is difficult to understand a particular behavior out of context, it is also difficult to change an individual’s behavior without also working to change the family system to which they belong. Family systems theory may help to highlight the importance of conjoint treatment for couples who intend to remain intact (see Murray 2006). For example, if therapeutic changes are made with the perpetrator and this individual returns to the original system (i.e., the original family dynamic), his/her behavior may revert back to that which fits well in the niche they traditionally occupied within that system, particularly problematic if the system itself was dysfunctional. If the relationship is going to continue, if the personality issues are resolved in individual therapy, and if safety is established, perhaps even couples with a characterological batterer may ultimately benefit from some type of conjoint therapy. Certainly, public policy regarding the types of interventions offered or required of couples experiencing IPV should be informed by empirical evidence.

Summary

Langhinrichsen-Rohling (2010) presents a dyadic typology of bi-directionally violent couples that incorporates existing knowledge and empirical findings about subtypes of violent relationships. She appropriately focuses on the couple as the unit of investigation, and she calls attention to the serious need for improvement in IPV intervention programs. Although the model is somewhat limited in scope in that it does not include uni-directional IPV and in its assumption that partners in these subtypes will be matched on pathology, it does attempt to clarify subtypes of the most prevalent form of couple violence, in which both partners engage in IPV. While Langhinrichsen-Rohling (2010) makes only limited predictions about how the “Dyadic Culture-Family-Attachment-Skill Deficit Model” relates differentially to the development of each subtype, the model provides a framework from which to make these predictions and test these hypotheses. Most importantly, it provides a reasoned argument as to why to study both men’s and women’s violence in the context of the relationship. In the future, researchers may wish to examine dimensionally the major factors of this typology--control and emotion dysregulation--since not all bi-directionally violent couples are expected to fit neatly into one of the subtypes. For example, some bi-directionally violent couples may include one partner who is preoccupied with power and control and one partner with substantial difficulties in emotion regulation. Researchers should continue to focus on IPV within the context of the relationship and assess both partners’ behaviors (including violence). An explicit examination of women who engage in IPV should be an important goal in the field of IPV research and should not be considered anti-feminist. In fact, such research may help to identify additional treatment needs of these women. Finally, efforts are needed to inform policy makers of the current research on IPV treatment and to ensure that we are putting reality above political correctness.

In her review of major controversies facing IPV researchers and practitioners, Langhinrichsen-Rohling (2010) thoroughly describes each controversy and tells what is currently being done to resolve the issue, highlighting the need for future research. She presents a comprehensive discussion of the history, current state, and future directions of the field in a “must read” paper for those entering the field of IPV research. It is the hope of these authors that Langhinrichsen-Rohling’s (2010) paper will spark some much-needed research and additional debate on the phenomenon of mutual violence within romantic relationships, the needs of both male and female perpetrators and victims of IPV, and an increased awareness of the importance of studying partner violence within the context of the couple.