INTIMATE ABUSIVENESS

DONALD G. DUTTON

DEPARTMENT OF PSYCHOLOGY

UNIVERSITY OF BRITISH COLUMBIA

Running Head: Abusiveness

Preparation of this manuscript and research cited in it was made possible through grants from the Solicitor General of Canada, Health and Welfare Canada and the Social Sciences and Humanities Council of Canada. Requests for reprints should be sent to Dr. Don Dutton, Department of Psychology, University of British Columbia, Vancouver, B.C., V6T 1Z4, Canada.

ABSTRACT

The practice of psychology includes treatment groups for perpetrators of wife assault, treatment for victims and policy work on court-treatment models. Such practice must be rooted in solid empirical data on the problem of wife assault. Existing theories of wife assault perpetration (psychiatric, sociological, sociobiological, and social learning) fail in a variety of ways to explain the empirical data generated by research in the last two decades. A new theory is proposed to account for individual differences in the tendency to be assaultive in intimate relationships. The focus of this theory is on men whose abuse is specific to intimate relationships and is manifested through cyclical mood swings. This group appears to comprise about 40% of all men who are present for treatment for wife assault. For these men, abusiveness is triggered by their internal mood states rather than by external events. Several studies are cited which indicate that intimate attachment generates rage in wife assaulters. The origins of this attachment-rage are traced to early development. This template generates a complex of perceptions (attributions and projections) and behaviors (abusiveness) specific to intimate relationships. Such males are shaped differentially from other males by a patriarchal culture that provides justifications for their abuse of women in intimate relationships. A personality profile of intimately abusive males has been strongly related to intimate partners' reports of psychological abusiveness in several samples of males, including physically assaultive males, non-violent controls, clinical outpatients, college students and gay males.

Abuse within families is arguably the "discovery" of the latter part of the twentieth century that has the greatest implications for psychological practitioners. It is rare to provide service in a contemporary private practice and not encounter clients who have been either victims, perpetrators (or both) of some form of intimate abuse. Psychologists regularly appear in court in both custody disputes and self-defense cases where abuse is an issue. Despite its' extensive impact on practice, intimate abuse has not received the attention from academic research that one might expect. Most research the area of wife assault, for example, is interdisciplinary and has not been viewed as "mainstream" by clinical, social or personality researchers. Because of the nature of academic disciplines being split into departments and "areas", those attempts that have been made to understand wife abuse have strongly reflected the paradigms of the discipline. In so doing the strengths of each perspective have been revealed.

INCIDENCE AND PREVALENCE RATES

The private nature of family violence makes it difficult to develop accurate estimates of its incidence and prevalence. Victim surveys have assessed victimization experiences of representative sample from the general population. Three U.S. national surveys have been implemented in 1975 ( Straus, Gelles and Steinmetz, l980), 1985 (Straus and Gelles, l985) and 1992 (Straus and Kantor 1994). All used the Straus Conflict Tactics Scale (see Straus, l979a) as a measure of the type of actions used to resolve family conflicts. This self-report measure asks respondents to estimate the number of times during their marriage they and their partner dealt with conflict in various non-violent and violent ways. "Serious" violence was defined as kicking, biting, hitting with a fist or object, beating up, or using a weapon against the victim.

When considering "serious" violence only, the victimization rates for husband-to-wife violence on the three surveys were 3.8% (Straus et al., 1980), and 3.0% (Straus & Gelles, 1985) and 2.0% (Straus & Kantor, 1994) for violence occurring during the year preceding the survey. This decline occurs regardless of whether males or females report male violence. Whether this represents an actual drop or a reluctance to report is unknown, although drops in reports by victims suggests some veridical decrease in actual incidence. Dutton & Hemphill (1982), for example, found that females' reporting of victimization by male violence was unrelated to response styles such as social desirability, increasing our confidence in victim reports of incidence.

For the entire marriage, corresponding rates are l2.6% (Straus et al., 1980), and 11.3% (Straus & Gelles, 1985). Some comparable rates are reported from other countries that have employed the CTS and done stratified sampling similar to the U.S. surveys. Kennedy & Dutton (1987) reported rates for Canada that were about .66 of the U.S. rates. Kim & Cho (1992) reported rates for South Korea that were 3.5 times the U.S. rate. comparison of the Straus et al. (1980) data from the U.S. National survey in 1975 with the Straus and Gelles (1985) data from the U.S. National Survey in 1985 leads to the tentative conclusion that reports of severe husband-wife violence (both by perpetrators and victims) had dropped by about 21% during the decade. Straus & Gelles (1992) report assault rates by women against men that are generally higher than comparable rates for men assaulting women. Furthermore, in contradistinction to the decline in male rates, female rates for spousal violence seem to be increasing (Straus & Kantor 1994). Gender comparisons are controversial though, as the acts are not considered comparable by some (e.g. Browne 1992).

The Straus et al. (1980) survey reported single versus repeat assault in their data, specifically asking respondents the number of times they had been assaulted in the past twelve months. These data indicate that assaults are likely to be repeated within the year in about 66% of cases when no outside intervention occurs. Hence, the question of "spontaneous desistance" from violence continues to intrigue some researchers (see Dutton, Hart, Kennedy & Williams, 1990).

EXPLANATIONS

Wife assault has had a variety of "single factor" explanations ranging from the intrapsychic focus of psychiatry to the breadth of feminist sociology. Dutton (1988, 1995) critically reviewed these explanations and argued that each was incomplete. Originally, psychiatric explanations were based on small incarcerated samples, biased by the tendency of the times not to incarcerate except for the most heinous cases. These explanations helped reinforce the view of wife assault as rare and the men who committed it as unusual, atypical, and pathological (Faulk, 1974). Furthermore, early studies were reductionistic; "explaining" wife assault as caused by a neural mechanism (Elliott, 1977) such as temporal lobe epilepsy leading to the diagnosis of Intermittent Explosive Disorder (312.34 in the DSM-IV). Neurological explanations should lead to the prediction of random explosive outbursts yet wife assault typically occurs only in the context of an intimate relationship and usually in private. More recent psychiatric studies are sounder methodologically, but still suffer from interpretative problems, primarily due to the psychiatric tendency to insert diagnostic labels in lieu of etiology . Such studies never developed etiological models that explained the abuse, focusing instead on associations of abuse with established diagnostic categories (e.g., Bland & Orn (1986)).

Sociological/Feminist Explanations of Wife Assault

Feminist/sociologists provided an antidote for the reductionism of the early psychiatric view by arguing that wife assault was a common event, generated by social rules that supported male dominance of women (Dobash & Dobash, 1978; Goode, l97l) and tacit approval by society (Straus, l976, l977a, b, c; Bograd, 1988). As Dobash and Dobash (l979) put it "men who assault their wives are actually living up to cultural prescriptions that are cherished in Western society--aggressiveness, male dominance, and female subordination--and they are using physical force as a means to enforce that dominance" (Dobash and Dobash, l979, p. 24) and Bograd (1988, p.14) claimed that "the reality of domination at the social level is the most crucial factor contributing to and maintaining wife abuse at the personal level" and that "feminists seek to understand why men in general use physical force against their partners and what function this serves for a society in a given historical context " (Bograd, 1988, p. 13). Hence, domination of women is viewed, from the feminist perspective, as a cultural prescription, and violence against women as a means to that end. This emphasis on the sociocultural is reflected in the feminist distrust of psychological causes of male violence (e.g. Goldner, Penn, Sheinberg, & Walker, 1990) as potentially "exonerative" of male violence and by the lack of empirical studies of putative interactive causes conducted within a feminist perspective. In general, feminist analysis has not included psychological explanations for wife assault, focusing instead on the question of "why men in general use physical force against their partners" (Bograd, 1988, p. 13).

I have critiqued the feminist position elsewhere (Dutton, 1994c) and will simply summarize the essence of that criticism here. Feminist theory has difficulty accounting for individual differences in male assaultiveness or explaining several key empirical findings: that all abuse rates (physical, verbal, sexual) are higher in lesbian relationships than in heterosexual relationships (Lie, Schilit, Bush, Montague, & Reyes 1991); that abuse rates are lower in more patriarchal cultures (Sorenson & Telles, 1991; Campbell, 1992); that there is no linear correlation between dyadic power and wife assault (Coleman & Straus, 1990); that there is a "near zero" correlation between structural patriarchy (defined as male-female inequality) and individual patriarchal beliefs (Yllo & Straus 1990, p.395) and that direct tests of structural patriarchy and patriarchal beliefs yield only weak relationships to assaultiveness (Smith, 1990). Indeed, what the latter study found, rather than invariant male patriarchal ideology, was evidence for the heterogenity of male beliefs about women. This is matched, I might add, by a heterogenity of male actions: only about 12% of men would have lifetime incidence of "serious" violent acts (punching, kicking or worse on the Conflict Tactics Scale (CTS: Straus, 1979)) at any time towards their wives (Straus, Gelles, & Steinmetz, 1980), according to surveys done by women interviewers with women respondents. About 8% are repeatedly assaultive during the duration of the marriage with the percentage dropping further as incidence increases. When we add to this the finding that only about 9% of men are dominant within their family according to a "final say" measure of power (Coleman & Straus, 1990), we get a picture of male dominance and violence very different from that painted by sociological feminism. Furthermore, in contradistinction to the feminist view, male violence seems to be more strongly associated with powerlessness in males, not to power maintenance. One of the problems, I suspect, is that sociological feminism has attempted to graft a sociopolitical model onto male-female relationships. Attempting to reduce the complexity of intimate power dynamics in this way has led to an inevitable simplifying. Apart from the inexplicable data sources reported above, the sociological/feminist model cannot account for male feelings of powerlessness in intimate relationships (Dutton & Strachan, 1987), various subtypes of wife assaulter (Dutton, 1988, 1995; Saunders, 1992, Holtzworth-Munroe & Stuart, 1994), or cyclical or phasic battering (Dutton, 1994 a & b; Dutton & Starzomski, 1993; Walker, 1979).

Sociobiology

Sociobiologists have extended Charles Darwin's (l87l, l872) notion that physical characteristics and behaviours of species develop over time through the process of natural selection. Characteristics and behaviours that enable a species to function in a specific environment are maintained and gradually evolve to enable the species to survive. In more recent years, the evolutionary point of view has been extended to account for the social behaviour of animals (Wynne-Edwards, l962) and humans (Wilson, l975, l978).

In extending evolutionary ideas to human social behaviour, sociobiologists attempt to account for cooperation, competition, and aggression from the standpoint that each behaviour has an evolutionary function; that is, it maximizes the likelihood that individuals who demonstrate the behaviour will survive and that their offspring will survive (Bigelow, l972) or, that their contribution to the gene pool is maximized.

In so doing, sociobiologists tend to focus on social behaviour that is common to humankind across all cultures (rather than on cultural variations as anthropologists do). They regard these common types of social behaviour as part of an evolutionary heritage and an attempt to maximize one's contribution to the gene pool. Sociobiological theory comes in for strong attack by most social scientists who claim that it underestimates the impact of socio-cultural factors on contemporary social behaviour and overestimates the biological or inherited aspect of such behaviour (Gould, 1983). It is not necessary to get into the current controversy, however, in order to raise a question that is difficult for sociobiologists to answer: How does one explain domestic violence and, in particular, the killing of one's spouse? It seems a less-than-optimal way to "increase one's contribution to the gene pool;" in fact, it seems evolutionarily unsound, and yet it has occurred with alarming frequency since the beginnings of recorded history (Davidson, l977a; deReincourt, l974). Buss (1994) argues that abuse serves the sociobiological function of coercive control and is motivated by the male need to guarantee his paternity by ensuring compliance through coercion. Physical abuse, however, also risks the physical welfare of both the mate and the offspring if she is pregnant. Neither of these outcomes seem evolutionarily sound. It seems also that sociobiologists should predict an inherited predisposition of aggression toward invading males but little description of this occurs in the sociobiological literature.

Simeons (l962) argues that men have a genetic predisposition to react with rage to sexual threat. However, the stimuli that constitute "sexual threat" are often socially determined, and the response to such stimuli, while most certainly a form of physiological arousal, is itself often labeled (as rage, hurt, anxiety, etc.) in ways that are also culturally shaped. Finally, the behaviour that follows from the emotion is again directed by what the culture deems more or less acceptable. Indeed, most recently sociobiological writers (e.g., Symons, l980) on the evolution of human sexuality argue that molar behaviour (with the exception of simple motor patterns) "is too variable and too far from the genes" (Symons, personal communication, 1981). Hence, it appears that the strongest statement that can be made on sociobiological grounds is that men may have an inherited tendency to secrete adrenalin when they believe themselves to be sexually threatened and that they will experience this state as arousing. The label applied to this arousal, however, will be socially determined, and the tendency to label this arousal as anger, the behavioral response of acting out anger aggressively, and the choice of a target for aggression, are all shaped by societal values and learned dispositions. It is for this reason that so much individual variability, inexplicable from a sociobiological position, occurs in male responses to common "threatening" stimuli.

The above arguments notwithstanding, one major empirical work on intimate violence has been produced from a sociobiological perspective: Homicide by Daly and Wilson (1988). In this book, the authors argue that intimate homicides follow patterns that are uninterpretable without the aid of a sociobiological perspective. Daly and Wilson examine 212 homicides in Detroit categorized by motive (in 164 homicides a male killed a male, in 19 a male killed a female, in 18 a female killed a male, and in 11 a female killed a female). A sociobiological perspective argues jealousy as a motive for homicide should be more frequent with male perpetrators. The Detroit data found that 13.6% (25/183) of male homicides were jealousy driven. By comparison, they found that 31% (9/29) female homicides were jealousy driven. A comparable study in Canada found that 24% (195/812) of male and 7.6% (19/248) of female homicides were jealousy driven. The Detroit homicides were largely black, the Canadian mainly white. The jealousy-driven rate for the female homicides in Detroit is 4.5 times that of the Canadian data, while the male rate for Detroit drops by almost half (the reason being a much larger number of homicides in Detroit falling into other categories). What this exemplifies to me is the inability of sociobiological explanations to account for cultural variability.

From a sociobiological perspective, jealousy motives should influence the homicide rates of various age categories as well. For example, so called "May-December" (younger woman-older man) relationships should be at risk because of greater jealousy. Daly and Wilson cite statistics which show that marriages with high age disparities have four times the risk of marriages with a small (2 year gap). The highest risk for homicide victimization is for wives +10 years younger than their husbands. There are alternative psychological interpretations for these data, of course. including the notion that identity issues are bound up with intimacy in relationships in a way that is naturally confounded with what sociobiologists call kinship; the closer the person, the more our personal identity is founded upon our relationship with them. Hence, if our identity is unstable and the relationship is troubled, the greater may be the risk of anger and violence. It may not be kinship or zealous protection of the gene pool that drives rage, it may be an aspect of threatened loss of identity that is heightened when intimacy is imperiled. The two are inextricable. Among the cases of jealousy-precipitated homicides in Detroit cited by Daly and Wilson as evidence for their sociobiological position are two cases of homosexuals killed by lovers for alleged infidelity. I would see this as supporting an intimacy interpretation of intimate violence rather than a sociobiological interpretation. Buss (1994) concedes that it is difficult for sociobiology to explain homosexuality.

Social learning explanations for wife assault emphasize violence as a learned response to stress supported by the immediate rewards gained through its use (Dutton, 1988; Ganley, 1989). Dutton (1988, 1995) described some of these rewards as feelings of agency and control for the male, cessation of aversive stimulation provided by losing a verbal conflict, and cathartic expression of anger. Some evidence exists for the observational learning aspect of social learning theory. Straus, Gelles, and Steinmetz (l980), for example, found that males who had observed parents attack each other were three times more likely to have assaulted their wife (35% of men who had seen this had hit their own wife in the year of the study compared to l0.7% of men who had not witnessed this event). However, something more than simple modeling appears to occur; Kalmuss (1984) found that modeling was not sex-specific. Exposure to fathers hitting mothers increases the likelihood of both husband-wife and wife-husband aggression in the next generation and both sons and daughters who witness paternal hitting are more likely to be both victims as well as perpetrators of violence against their spouse. Hence, something more than sex-role specific behavioral modeling seems to occur.

Dutton (1988, 1995) expanded social learning theory's notion of instigators of aggression to include perceived abandonment and control of intimacy. The control of intimacy seemed to define a special category of aversive stimuli frequently mentioned by assaultive males in treatment groups. Dutton (1988, 1995) proposed that changes in socioemotional distance between the man and his wife could serve as instigators of wife assault. In an empirical test of these notions, Dutton and Browning (1988) found that wife assaulters reacted with more anger than did control males to videotape scenarios depicting uncontrollable abandonment of a male by his female partner. The anger was significantly greater than that reported for equally conflictual arguments without abandonment themes.

A limitation of social learning theory is its' emphasis on immediate external stimuli producing "aversive arousal". This somewhat passive view of perpetrators sees them as responding to proximal stimuli with rage and abuse. However, based on certain descriptive features of batterers (Walker, 1979) provided by their victims, these men appear to respond to internal stimuli which seem to increase over time in the absence of any increase in real "aversive stimuli" and to disappear temporarily upon catharsis. There is no systematic explanation for the development of these apparently cyclical internal states within a social learning framework.

THE CYCLE OF VIOLENCE

In her seminal study of 120 battered women, Walker (1979) described what she termed the "cycle of violence" whereby relationships go through a three stage cycle with stages described as tension building, acute battering, and contrition. The first phase is characterized by escalating anger and outbursts on the part of the man accompanied by a recognition that his behavior is wrong (but feeling that it is out of his control), fear of the woman leaving and greater oppression, and jealousy and possessiveness to keep the woman captive. Phase two "is characterized by the uncontrollable discharge of tensions that have built up during phase one" (Walker, 1979, p. 59) The trigger for moving into phase two, as Walker's respondents reported it, was either "an external event or the internal state of the man" (Walker, 1979, p. 60). This uncontrollable rage generates acute battering until the batterer becomes "exhausted and emotionally depleted" (Walker, 1979, p.61), typically from two to twenty four hours. Victims report the battering occurring (occurred or occurs?) in response to anything that they do (e.g., both staying quiet and answering back escalate anger). Upon exhaustion, phase three begins characterized by contrition, confession, promises of reform and by attempts to convince the victim and others that the abuse will not recur. Walker's respondents depicted these phases as predictable. Until recently, little attention has been paid to the dynamics generating cyclical abuse by males. Our recent work has searched for a personality basis for cyclical male-initiated abuse. We believe such a basis exists in borderline personality organization (BPO).

Borderline Personality Organization and Assaultiveness: The Theoretical Connection

BPO is a clinical category characterized by intense, unstable interpersonal relationships, an unstable sense of self, intense anger, and impulsivity (Gunderson, 1984), characteristics which, prima facie, should increase the likelihood of violence in intimate relationships. BPO is a less severe form of the more rare Borderline Personality Disorder (BPD: Kernberg 1977).

As Gunderson (1984) describes the Borderline Personality (BP), the essential characteristics (in order of importance) are: a proclivity for intense, unstable interpersonal relationships characterized by intermittent undermining of the significant other, manipulation, and masked dependency; an unstable sense of self with intolerance of being alone and abandonment anxiety; and intense anger, demandingness, and impulsivity, usually tied to substance abuse or promiscuity.

The theoretical initiative for examining BPO came from the work of Gunderson (1984) who described a three level defense structure of Borderline Personality that produces sudden shifts in "phenomenology", affect, and behaviour. This defense structure could theoretically produce the kinds of behavior depicted by Walker's (1979) "abuse cycle" description of some wife assaulters. Gunderson described the BP as existing in a "dysphoric stalemate" in relationships, where intimacy needs are unmet but where the requisite motivation and skills to assert the needs were non-existent. This first stage resembled the "tension building" phase of the abuse cycle, during which frustrations increase. Stage Two, according to Gunderson, occurred when the BP perceived an intimate relationship as possibly lost. The defense structure at this stage expressed itself as anger, devaluation of the significant other, or open rage. This appeared to correspond to Walker's "battering" stage of the abuse cycle. Stage Three occurred when the significant other was lost. At this point, the BP engaged in behaviours designed to ward off the subjective experience of aloneness. Impulsive substance use and promiscuity were the examples offered by Gunderson. We speculated that the exaggerated "appeasement" behaviours that assaultive husbands engage in after their wife has temporarily left the relationship might constitute another example. These behaviours persist until the woman has emotionally returned, then the cycle repeats itself. In a similar description of homicide perpetrators, Revitch and Schlesinger (1981) described a "catathymic personality" which underwent three stages: incubation, violent act, and relief. The primary affect during incubation is depression, accompanied by a constant sense of tension buildup. When the violent act was homicide, it appeared to be preceded by a period of brooding. Retrospective accounts reveal that, during this incubation period, the perpetrator viewed the tension buildup as being "caused" by the victim who acted in a perceived malevolent and persecutory manner toward the perpetrator. The relief experienced after catathymic violence came from a release of stored up affect: specifically, depressive and anxious symptoms and muscular tension. Although Revitch and Schlesinger's analysis was meant to be applied to intimate homicides, it appears, prima facie, to also describe the abuse cycle. Both wife assault and borderline personality have been empirically linked to being a victim of childhood abuse (Kalmuss 1984, Herman & van der Kolk 1987). Hence, there is a prima facie case for systematically examining the relationship between wife assault and the BP.

Borderline Personality Organization and Assaultivness: The Empirical Studies

In a series of studies, Dutton and his colleagues (Dutton, 1994 a, b; Dutton, in press a, b; Dutton & Landolt, 1995; Dutton, Saunders, Starzomski, and Bartholomew, 1994; Dutton & Starzomski, 1993, 1994; Hart, Dutton, & Newlove, 1993; Starzomski & Dutton, 1994) have examined personality profiles of assaultive males. The overall strategy of this work has been based on self-report scales filled out by abusive men as part of an assessment procedure for treatment and corroborated through the female partners' reports of the man's abusiveness. Both self-referred and court-referred men have been compared to demographically matched controls (Dutton & Starzomski, 1994). Extensive analyses of the men's reporting tendencies have been made through the use of both the Marlowe-Crowne scale (Crowne & Marlowe 1960), the Balanced Inventory of Social Responding (Paulhus 1984, 1986) and through the use of the Disclosure, Debasement and Desirability Scales of the Millon Clinical Multiaxial Inventory-II (MCMI-II: Millon 1987). An extensive report on the relationship of social desirability to all self-report scales described below is available in Dutton and Hemphill (1992) and Dutton & Starzomski (1994). Self-reports of the man's anger, jealousy, experience of trauma symptoms and abusiveness, and of the man's abusiveness (both physical and psychological) made by his female partner have constituted the dependent variables in these studies. Dutton & Starzomski (1994) argued that self-referred assaulters constituted a more "pure" group of abusive personality (typically self-referring during the contrition phase of an abuse cycle) while court-referred samples were more heterogeneous. Consistent with this view was the finding that 45% of self-referred but only 27.5% of court-referred wife assaulters reached the 85th percentile on the borderline scale of the MCMI- II (Hart, Dutton, & Newlove, 1993). Self-reports were made on a scale measuring borderline personality organization (BPO: Oldham, Clarkin, Appelbaum, Carr, Kernberg, Lotterman, & Haas, 1985). This scale has three subscales; Identity Diffusion (a poorly integrated sense of self), Primitive Defenses (projection and splitting)and Reality Testing (transient psychotic states). The scale has good psychometric properties and Cronbach's alphas for its' subscales of .84 - .92.

In an initial sample of 80 wife assaulters and 40 demographically matched controls, Dutton (1994a) found BPO scores to be similar to those for diagnosed borderlines. The mean BPO score for the sample of wife assaulters was 71.3 (S.D. = 17.1) whereas the score for diagnosed borderlines was 74.8 (Oldham et al., 1985). By comparison, Oldham et al reported a mean score of 61.3 for a non-borderline sample and our controls scored 60.0 (S.D.= 17.0) on the BPO scale. Furthermore, BPO scores were significantly related to chronic anger, jealousy, use of violence, and experience of adult trauma symptoms in the wife assault group. High BPO scorers reported significantly more anger; of greater frequency, magnitude, and duration. They also reported greater jealousy and more trauma symptoms; dissociation, anxiety, sleep disturbance, depression and post-sexual abuse trauma. Finally, they reported significantly more abuse towards their wives: both verbal-symbolic and physical as measured on the Conflict Tactics Scale (CTS: Straus 1979). Correlations of BPO to other measures are reported in Table 1. Analysis of response styles indicated that these associations were not mere disclosure or social desirability effects.

______________

insert table 1 here

______________

Dutton & Starzomski (1993) corroborated these findings by focusing on wives' reports of abusive treatment by their husbands through assessment of both physical abuse using the CTS and emotional abuse using the Psychological Maltreatment of Women Inventory (PMWI: Tolman, 1989). The latter 58 item scale yields two factors called Dominance/Isolation and Emotional Abuse. Correlations of men's self reported BPO scores with women's reports of the men's abusiveness are displayed in Table 2. Strong associations of men's BPO scores with women's reports of male abusiveness were found.( see Table 1) A multiple regression indicated that BPO scale scores combined with scores from a self report for anger (the MAI: Siegel 1986) accounted for 50% of women's reports of Dominance/Isolation (Factor 1: PMWI) and 35% of Emotional Abuse scores (Factor 2).

Saunders (1992), developed a three part typology of wife assaulters. Essentially the three groups, based on behavioral descriptors were overcontrolled, antisocial and borderline. The latter group constituted about 30% of all assaultive males. We are not attempting to account for all assaultive behavior on a personality basis. Rather, we argue that men who are exclusively violent in intimate relationships and repeatedly initiate the violence appear to have borderline features. Both the borderline personality and the abusive behavior may have a common origin. Interestingly, the empirical relationship between BPO and women partners' reports of abusiveness was maintained for psychological abuse in non-physically abusive control groups, implying that BPO and anger scores may define a general profile for intimate abusiveness that is differentially manifested in different groups of men, possibly as a function of abuse styles learned in the family of origin.

Profiling Abusiveness

We established an "abuse profile" (Dutton, in press b) by combining those self report scores that most differentiated high from low abusive men as reported on their wives' PMWI scores. These items originated from the Multidimensional Anger Inventory, the BPO, the MAI, Trauma Symptom Checklist (TSC: Briere & Runtz, 1989), and the EMBU (Perris, Jacobsson, Lindstrom, von Knorring, & Perris, 1980), a self-report instrument for recollections of early parental treatment. Items with the highest loadings on this discriminant function were then selected and combined into a scale.

The resulting 29 item scale (called the Propensity for Abuse Scale (PAS: Dutton, in press, b) correlated +.51 (p <.001) with wives reports of Dominance/Isolation on the PMWI, and +.37 (p <.001) with the Emotional Abuse scale from that same inventory. Predictive validity was again assessed by using the scale to discriminate high and low abuse groups. Using the criterion of one standard deviation from the mean, the PAS correctly identified 82.2% of men on Dominance/Isolation and 81.3% of men on Emotional Abuse. The scale has been cross validated successfully and has also been found to identify abusiveness in populations of clinical outpatients (Dutton 1994d), blue collar workers, gay males (Dutton & Landolt 1995) and college students (Dutton, Landolt, Starzomski ,& van Ginkel 1995). In effect, the same constellation of personality features (BPO, high anger, fearful attachment, chronic trauma symptoms and recollections of paternal rejection) accounts for reports of abusiveness by one's intimate partner in all of the above groups.

The Origins of Abusiveness

Having developed this assessment of the abusive personality, we attempted to ascertain the etiology of this personality. First (Dutton, 1994 b), we performed a discriminant function analysis on abusive personality scores themselves. Family of origin factors (CTS + EMBU) correlated +.41 (p <.001) with scores on the PAS. The largest contributor to a discriminant function on abusiveness was paternal rejection, correlating +.87 with the discriminant function.

Next, composite scores were developed for abusive personality, abusive behavior, and abuse experiences in the family of origin (Dutton, Starzomski, & Ryan, 1994). The relationship among these composites is shown in Figure 1.

_________________

insert figure 1 here

_________________

As can be seen from Figure 1, the association of abuse in the family of origin with abusive personality (+.41)is stronger than its' association with abusive behavior (+.22). This result suggests a more complex model than direct modeling of abusive behaviors as suggested by social learning theory. Both borderline personality (Herman, Perry & van der Kolk, 1989; Zanarini, Gunderson, Marino, Schwartz, & Frankenburg, 1989) and wife assault (Kalmuss, 1984) have been empirically shown to have childhood origins. The possibility exists that some common set of early experiences serves as a common cause for both.

Attachment and Abusiveness

If early experiences influenced adult abusiveness, attachment theory might provide a valuable perspective its' etiology. Bowlby asserted that "attachment behavior is held to characterize human beings from the cradle to the grave" (1977, p. 203) and that adult "introjects"- confidence in the availability of attachment figures is built during the years of immaturity --infancy, childhood, and adolescence-- and that "whatever expectations are developed during those years tend to persist relatively unchanged throughout the rest of life" (1973, p. 235). Bowlby viewed interpersonal anger as arising from frustrated attachment needs and functioning as a form of "protest behavior" directed at regaining contact with an attachment figure.

In turn, chronic childhood frustration of attachment needs may lead to adult proneness to react with extreme anger (which I refer to as "intimacy-anger") when relevant attachment cues are present. Thus, attachment theory suggests that an assaultive male's violent outbursts may be a form of protest behavior directed at his attachment figure (in this case, a sexual partner) and precipitated by perceived threats of separation or abandonment. (Similar to the argument made by Dutton & Browning, 1988). A "Fearful" attachment pattern may be most strongly associated with intimacy-anger. Fearful individuals "desire social contact and intimacy but experience pervasive interpersonal distrust and fear of rejection" (Bartholomew, 1990).

This style manifests itself in hypersensitivity to rejection (rejection-sensitivity), and active avoidance of close relationships where vulnerability to rejection exists. While the Fearful share anxiety over abandonment with another insecurely attached group (called Preoccupied), their avoidance orientation may lead to more chronic frustration of attachment needs. Subjects were assessed using the RSQ (Griffin & Bartholomew, 1994), a 30-item self-report measure. Table 2 shows the associations of attachment to other aspects of abusiveness.

________________

insert table 2 here

________________

Fearfully attached men experience high degrees of both chronic anxiety (as measured by the TSC-33) and anger (as measured by the MAI). In addition, both trauma scores in general, and dissociation scores in particular, were highest for the fearful group. Fearful attachment alone accounted for significant proportions of variance in both emotional abuse criterion factors. Fearful attachment was also strongly correlated with borderline personality organization. Since anxiety (+.42) and anger (+.48) are both strongly associated with fearful attachment, one could argue that an emotional template of intimacy-anxiety/anger is the central affective feature of the fearful attachment pattern. These correlations maintain in the control sample (fearful-anxiety +.53, fearful-anger +.52), suggesting that this emotional template does not only reside within physically abusive men.

A prominent feature of BPO is intimacy-anger (Gunderson, 1984; Dutton,1994). The correlation of fearful attachment to BPO is so strong (+.62) that one could argue BPO is a personality representation of this particular attachment style (see also Mahler, 1971).

Attachment Disruption and Trauma

In abused boys, another prominent sequela of abuse victimization is hyperaggression. Carmen, Rieker, and Mills (1984) suggested that abused boys are more likely than girls to identify with the original aggressor and to eventually perpetuate the abuse on their spouse and children. In their view, an effect of physical maltreatment by a parent is to exaggerate sex role characteristics, possibly as a means attempting to strengthen the damaged self. van der Kolk (1987) noted that traumatized (including physical abuse) children had trouble modulating aggression and included being physically abused as a child as a trauma source.

Some evidence is also beginning to accumulate that shows abuse in the family of origin to be a risk factor for the development of post-traumatic stress disorder (PTSD). Herman & van der Kolk (1987) noted how PTSD included poor affect tolerance, heightened aggression, irritability, chronic dysphoric mood, emptiness, and recurrent depression and was "described in patients who have been subjected to repeated trauma over a considerable period of time" (Herman & van der Kolk, 1987, p. 114). Hence, the possibility is presented, that PTSD may be another link or mediating variable between childhood abuse victimization and adult perpetration of intimate abuse.

________________

insert figure 2 here

________________

In order to test this notion, wife assaulters were compared to two groups of diagnosed PTSD men from independent studies (Dutton, in press b). The results can be seen in Figure 2. In the wife assault sample, 45% of all men met criteria for PTSD. (75% on the MCMI-II on the "82C profile" (avoidant-passive-aggressive-borderline)). On the Millon Clinical Multiaxial Inventory - 11 (Millon, 1987), wife assaulters and PTSD men were similar on all profile peaks with the following exceptions: wife assaulters had significantly higher scores on antisocial personality and PTSD men had higher scores on anxiety and dysthymia. This result implies that the acting out of aggression by wife assaulters may dispel pent-up dysthymia (as Revitch and Schlesinger suggested happens with catathymics).

What is the source of trauma for these men? Is it the childhood experience of abuse, rejection and disturbed attachment seen in their self reports? At this point, the data suggest the answer in the affirmative.

Projection and Attribution

How are chronic abuse habits maintained? One mechanism appears to be cognitive. Starzomski & Dutton (1994) replicated the BPO results reported above with a non-clinical population of college students. Fearful attachment in this group was again strongly associated with anger (+.58, p <.001) as with the wife assault group (Dutton , Saunders, Starzomski & Bartholomew, 1994) (see Table 3).

________________

insert table 3 here

________________

Furthermore, one mechanism was unearthed in this study that contributes to the maintenance of chronic intimate anger. Fearfully attached college males make attributions for intimate events differently than other attachment groups. Fearful attachment is correlated +.33 (p.<.01) with a tendency to make blaming attributions to the partner on the Relationship Attribution Measure (Fincham & Bradbury, 1992). This finding is again consistent with a feature of BPO described in the clinical literature. Gunderson and others had identified the tendency to project blame as a defining feature of BPO and Oldham et al. included measures of projective identification as part of their measure of Primitive Defenses, (a subscale of the BPO self report instrument). Although couched in attributional terms, the empirical results of Starzomski & Dutton (1994) support this clinical claim. Scores on fearful attachment also correlated significantly with the BPO subscale for primitive defenses, which assesses the tendency to split women into ideal and devalued objects and to project angry impulses onto the devalued woman-object. This projection or blaming attributional tendency serves to maintain high degrees of chronic intimate anger and may have its origin in attempts to ward off shame. By projecting unacceptable impulses outward, their potential to threaten a weak self- concept is neutralized. A summary of the relationships observed between BPO and other measures for the assaultive sample is outlined in Figure 3.

___________________

insert figure 3 here

___________________

Shame

Some recent work by Tangney (1991) has presented a more focused analysis of the potential role of shame as a mediator between the early experiences of assaultive men and their adult experience of anger and abusiveness. Tangney (1991) differentiates shame-proneness and guilt-proneness as two moral affective styles where the former has to do with "global, painful, and devastating experience in which the self, not just behavior, is painfully scrutinized and negatively evaluated" (op. cit., p. 599). In this sense, shame-inducing experiences which generate a shame-prone style, may be viewed as attacks on the global self and should produce disturbances in self-identity. Shame-prone individuals have been found to demonstrate a limited empathetic ability (Tangney, 1991), a high propensity for anger and self-reports of aggression (Tangney, Wagner, Fletcher, & Gramzow, 1992). Dutton, van Ginkel, & Starzomski (1994) found recollections of shame-inducing experiences by parents of assaultive men to be significantly related to the men's self reports of both anger and physical abuse and to their wives reports of the man's use of Dominance/Isolation. Miller (1985) and others (e.g., Wurmser, 1981; Tompkins 1987, Lewis 1987, Tangney et al. 1992) have commented on the relationship of shame to anger, deriving from Helen Block Lewis' description of "humiliated fury" (Lewis 1971). Tangney et al. (1992) found significant correlations between TOSCA shame-proneness and hostility, anger arousal and tendencies to blame others for negative events. Novaco (1976) described anger as serving the function of overriding less acceptable emotions. Shame and guilt would appear to be two such emotions. Shame-proneness as Tangney described it was a tendency to experience global attacks on the self. Since such affective experiences are aversive, shame-prone individuals attempt to ward them off by externalizing cause for negative events, thereby avoiding personal responsibility which they experience as shameful. Dutton, Starzomski & van Ginkel (1994) found three recalled sources of shame in assaultive males. These were public scolding, random punishment and generic criticism. All three were recalled as generating experiences of shame. These, in turn, were correlated with adult anger and tendencies to project blame. Not surprisingly, given these tendencies, abusive actions also correlated with recalled shame experiences. Partial correlations revealed that parental shaming still correlated significantly with measures of abusive personality after physical abuse by the parents had been partialled out. The converse however, was not true. With parental shaming partialled out, physical abuse by parents did not correlate significantly with abusive personality measures. Dutton, van Ginkel, & Starzomski, (1994) suggested that a two step learning process may be involved in the acquisition of abusiveness. Shaming experiences appeared to contribute to the formation of the abusive personality, physical abuse to the modeling of its' behavioral manifestations. In non-assaultive control groups, abusiveness personality is still significantly related to verbal and emotional abusiveness. These men report less modeling experiences in their family of origin (see Figure 4 ). Hence, abuse or rejection by parents, insecure attachment, and parental shaming all appear to contribute to the formation of a personality type in males associated with abusiveness.

_________________

insert figure 4 here

_________________

In both clinical and nonclinical samples, BPO is highly correlated with both anger and Fearful attachment and in the nonclinical sample, anger was also significantly correlated with, and sustained by, the tendency to make blaming attributions to the woman (Starzomski & Dutton, 1994). Figures 2 and 3 demonstrate these relationships.

Why do these men get so angry in intimate relationships? The answer may lie in part in the meaning they ascribe to intimacy. For men high in BPO, intimate relationships serve the unenviable task of maintaining their ego integrity. With an unstable sense of self and an inability to tolerate aloneness , these men depend on their relationship with their female partner to prevent their fragile selfhood from disintegrating. Yet that very relationship is fraught with "dysphoric stalemates" (being unable to communicate intimacy needs while being extremely demanding), chronic intimacy-anger, and a tendency to project unacceptable impulses onto the significant other, while denying and masking his unconscious dependency. The intimate relationship of the high BPO scorer is asked to do the impossible, and when it fails, or appears in his eyes to fail, anger escalates because his very sense of self is threatened and because he projects blame to his partner for the dysphoria that he experiences inevitably during intimacy. He views her, at this phase of the relationship, as "all bad." If that impasse is resolved, he then tends to view her as "all good," and himself as bad, and enters the contrition phase of the abuse cycle (cf. Walker, 1979).

The unavoidable aspect of attachment-anger, the tendency to blaming and projection anger onto the partner and the inability to verbalize dysphoric states defines a personality profile that generates both relationship conflict and abuse. I have come to call this personality profile the "abusive personality" and to assess it through use of the Propensity for Abusiveness Scale (PAS). The behavioral manifestation of abusiveness (e.g., whether it takes a physical or verbal form) may still depend on modeling and autodidactic ontogenetic experiences (Dutton, 1988). BPO correlates with abusiveness in our control groups, but only with emotional abuse because men in these groups are, according to their wives, non-violent. The mood cycles which characterize Borderlines are essential features of the disorder, but the behavioral forms of abuse may be learned reactions, not just to external stress, but to the internal cues of dysphoria. It is for this reason that victims of the abuse describe it as self-generated by the perpetrator (e.g., Walker, 1979).

Relationship to Socializing Culture

If early experience contributes toward a propensity for assaultiveness, it may be that elements of Borderline Personality Organization seek out aspects of the culture to direct and justify abuse. The primitive defenses of the Borderline Personality, for example, which involve "splitting" good objects from bad objects are reinforced by cultural judgments about female sexuality. Cultures which divide women into "Madonnas and whores" provide a sanctioned reinforcement of the object split in the assaultive Borderline male. Cultures which socialize men to expect the woman to be responsible for relationship outcome provide a rationale for the Borderline Personality's expectation that his intimate partner should mitigate his dysphoria. From this perspective, the personality pattern contains emotional demands which it directs and justifies through drawing on the ambient culture. We are currently testing the hypothesis that high BPO men will rationalize their maltreatment of intimate partners by adopting more negative sexual stereotypes of women. In a non-clinical sample (n = 72), Starzomski (1993) found that BPO scores correlated +.57 (p <.01) with scores on a measure of hostility toward women. This was replicated in our control group of blue collar workers. At present, we are attempting to replicate this relationship within several distinct cultural groups. If early experience contributes toward a propensity for abusiveness, it may that high BPO men seek out aspects of the culture to rationalize their anger and behavior toward women. In my clinical experience, abusive men frequently blame their female partner for their own behavior. After a series of relationships in which their abusiveness continues to re-manifest itself, they "progress" from blaming a woman to blaming women in general.

Summary

At this juncture, our data reveal a model for the etiology of the abuse cycle that is more comprehensive than previously available. Broader aspects of early experience such as recollections of paternal rejection are the strongest contributors to an Abusive Personality. Shaming experiences and insecure attachment both appear to contribute to this link between parental treatment and subsequent personality. The Abusive Personality itself is highly correlated with both the Fearful attachment style and with borderline personality organization, although both of these characterizations tend to underexpress the anger component of the Abusive Personality. The cyclical aspects of the borderline shifting phenomenology may originate in early alternate reward-punishment experiences such as ambivalent attachment. The cognitive aspects of the Abusive Personality, especially its' tendency to project/attribute blame to the intimate partner for felt dysphoria and to ruminate on these blaming attributions, serves as a strong maintenance mechanism for anger. Experienced anger in intimate relationships is probably inevitable for the Abusive Personality, and coupled with the repetitive blaming of the partner sets the stage for abusive outbursts, either verbal or physical.

Arrest and Desistance

Dutton (1995) has reviewed criminal justice system involvement in wife assault and suggested a model for criminal justice-psychological treatment. Arrest policies have significantly effects on reducing recidivism (Sherman & Berk, 1984), although not for "marginal groups" (black, unmarried, unemployed) who show time-lagged increases in recidivist wife assault (Sherman et al., 1992). This latter study however, did not report post-arrest events such as whether conviction and court-mandated treatment occurred. The study has been used as an argument against mandatory arrest but other studies reveal that arrest-treatment combinations are quite effective at reducing recidivism (Dutton 1986, Dutton 1995). An alternative interpretation of the Sherman et al. results is to maintain mandatory arrest policies and work at a broader societal level toward reducing marginality.

Court Mandated Treatment

Group therapy for physically abusive males was developed by Anne Ganley (Ganley,1981) and Daniel Sonkin (Sonkin, Martin, & Walker, 1985). The treatment draws on cognitive- behavioral theories for aggression (Bandura 1979, Novaco 1975) that treats aggression as a learned response to stressors or "aversive stimuli". For a number of reasons, such groups have been the therapy of choice for court mandated programs. The emphasis of the group is on teaching the man to take personal responsibility for controlling his violence. That philosophy of is consistent with the criminal justice notion, although the groups frequently extend it, so that alcohol use, for example, is not accepted as an "excuse" for violence. Also, the group nature of treatment maximizes learning opportunities, so that the groups are more cost-effective than individual treatment.

Rosenfeld (1992) reports the results of twenty-five outcome studies on group treatment for assaultiveness. These generated self reports of recidivism (any kind of post-treatment violence on the Conflict Tactics Scale) of 21%, wives' reports of 33% and police reports (arrests) of 8.4%. By comparison untreated men generated recidivism reports from police of 23.4% (about three times that of the treated men). Dutton (1986) reported a matched group study of 50 treated and 50 untreated controls. Eight of the untreated controls had been rejected for treatment for motivational reasons. The subsequent recidivism rate for the controls was 40% within two years. The comparable rate for treated men was 16%. Based on wives' reports, 84% of the men were not violent post-treatment. Rates of both physical and emotional abuse dropped by over sixty- six percent. The problem with all treatment evaluations is that none have been based on randomized allocation of subjects to treatment conditions. Courts may not order to treatment men who are unmotivated to change, hence, comparing treated with untreated groups may inflate treatment effect size. Randomized designs are extremely difficult to implement simply because, court officials and therapists are reluctant to deny men treatment for research purposes.

Despite these problems, the evaluation of treatment groups seems to have been harsh. Rosenfeld concludes that arrested-untreated men have a recidivism rate that is no different from arrested-treated men. In fact, the data he reviewed showed post treatment recidivism rates of 30% (for court-ordered men who refused treatment based on police reports) compared to 8.4% for treated men. Similarly, wives reports of post treated violence for treatment dropouts was 47% compared to 33% for treated. Given that most men who present for treatment in court-ordered programs have no prior treatment experience and ambivalence about attending, these results seem acceptable. Prochaska, DiClemente & Norcross (1992) reviewed treatment effects on a variety of addictive behaviors and developed a six stage spiral model that described an addicts progress from "precontemplation" (no acknowledgment of problem) to termination. Few progressed linearly to termination, most experienced relapses. Use of violence in intimate relationships can be conceived of as a form of addiction. Violence or expressed abuse is often the only way an abusive man has of releasing tension and generating short term interpersonal control. From this perspective, it seems overly optimistic to expect a sixteen week treatment group to stop all abuse immediately after treatment. We have recently completed a ten year follow-up of a sample of 586 treated and untreated men, examining their post arrest recidivism rates (Dutton, Bodnarchuk, Kropp, Hart ,& Ogloff, 1995). Compared to untreated men, treated men had 350 fewer arrests per thousand men during the decade. Since follow-up interviews with a subset of female partners revealed that for every arrest, there are about thirty unreported assaultive actions (on the CTS), this translates to 10,500 fewer assaults against the female partners of men who completed treatment. It is important to note that treatment did not appear to stop recidivist arrest immediately but appeared rather to shorten what criminologists call the "criminal career" of wife assaulters. The reader is cautioned however, that these were not randomly selected groups. Hence, these differences in outcome cannot be directly attributed to treatment.

The question still to be addressed is not "do treatment groups work?" but "for whom do they work best/least?". We have begun to answer that question by assessing, amongst other measures, the borderline personality organization dimension described above. In a study of sixty men completing treatment (Dutton & Ogloff, 1994), the largest predictor of the man's post treatment abusiveness was his score on the BPO self report instrument taken post treatment which accounted for 50% of the variance in his wife's reports of abusiveness (measured by the Psychological Maltreatment of Women Inventory (Tolman 1988)). Both BPO and MCMI scores of Antisocial Personality were significant predictors of post treatment abusiveness.

A new wave of treatment modalities have potential applicability to wife assault treatment. The connection of abusiveness with fearful/angry attachment described above (Dutton et al, 1994) suggests that therapy focusing on attachment issues may be beneficial. A recent family- based approach to attachment problems by Doane & Diamond (1994) is an example. Linehan (1987) developed what she called a "dialectical behavior therapy" for working with borderlines. Clearly, given the nature of the data presented above, her techniques, although intended for more extreme forms of Borderline Personality Disorder, may have applicability to work with assaultive men.

Surprisingly, little has been written about the treatment ramifications of shame-proneness in assaultive men. Wallace & Nosko (1993) argue that shame is associated with anxious attachment in batterers and that shame drives the contrition phase of the abuse cycle (Walker 1979). Our current data are consistent with this notion. Dutton (1995) showed how men high in BPO most resemble cyclical or phasic personalities who might generate abuse cycles. Dutton, Starzomski & van Ginkel (1994) found that BPO was strongly related to shame experiences and shame proneness. Wallace & Nosko (1993) advocate group therapy as an effective technique for working with shame-prone individuals because of the safety and co-confessional aspects of groupwork. Before this advantage can be realized however, hurdles must be crossed. Having to attend a group through formal or informal coercion, and the requirement of the group to reveal past transgressions can themselves be shame inducing. The exercise of confession, typical in most treatment groups, can allow men to overcome their shame through what the authors call "vicarious detoxification"; listeners encounter their own shame as each member of the group confesses. Projective identification, permits men to encounter parts of themselves otherwise split off through hearing other men's confessions and sets the stage for working through shame issues. Furthermore the requirement that each man confess, builds group solidarity. Eventually, anger, rage and violence, the original defenses against shame, no longer become necessary.

What remains to be done is to incorporate these techniques into a group treatment format. The treatment program of the future may begin at a court mandated level with a sixteen week cognitive-behavioral treatment followed by more prolonged follow up of men with relevant personality disturbances. The innovative techniques described above could constitute part of this follow up. The evolution of diagnosis of spousal assault into more differentiated categories with differential etiologies has yet to be matched by parallel developments in treatment. This may be about to change.

Court mandated treatment for wife assault can only reach about .5% of all assaultive males (Dutton 1986, 1995). Clearly, effective secondary and primary intervention techniques are also required. Some promising high school intervention programs have been reported by Jaffe et al. (1992) designed at educating all students about abusiveness. However, more in-depth follow-ups may be required with boys showing high risk profiles on the PAS scale. Jaffe et al (1992) found a small group of boys who exhibited a boomerang effect to high school abuse workshops. These boys might be candidates for more extensive preventive work. A central implication of our work on the abusive personality is that intimate abusiveness has psychological origins. A society that wishes to seriously reduce abusiveness must address both the family factors that create this personality form and the social values that justify and sustain it.

REFERENCES

Bandura, A. (1979). The social learning perspective: Mechanisms of Aggression. In H. Toch (ed.) The Psychology of Crime and Criminal Justice. Holt, Rinehart & Winston:New York.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of Social and Personal Relationships, 7, 147-178.

Bigelow, R. (1972). The evolution of co-operation, aggression and self-control. In J.K. Cole & D.D. Jensen (eds.), Nebraska symposium on motivation (pp. 1 - 58). Lincoln: University of Nebraska Press.

Bland, R., & Orn, H. (1986, March). Family violence and psychiatric disorder. Canadian Journal of Psychiatry, 31, 129-137.

Bograd, M. (1988). Feminist perspectives on wife assault: An introduction. In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse (pp. 11-26). Newbury Park, CA.: Sage.

Bowlby, J. (1973). Attachment and loss. Volume 2. Separation. New York: Basic Books.

Bowlby, J. (1977). The making and breaking of affectional bonds. British Journal of Psychiatry, 130, 201-210.

Briere, J., & Runtz, M. (1989). The trauma symptom checklist (TSC-33): Early data on a new scale. Journal of Interpersonal Violence, 4(2), 151-162.

Browne, A, (1992). Are women as violent as men? Unpublished manuscript. University of Massachusetts.

Buss, D. (1994). The evolution of desire: Strategies of human mating. Basic Books: New York.

Campbell, J. (1992). Prevention of wife battering: Insights from cultural analysis. Response, Issue 80, 14(3), 18-24.

Carmen, E.H., Reiker, P.P., & Mills, T . (1984). Victims of violence and psychiatric illness. American Journal of Psychiatry, 141: 378-379.

Coleman, D. H., & Straus, M. A. (1985). Marital power, conflict, and violence. Paper presented at the meeting of the American Society of Criminology, San Diego, CA.

Crowne, D. P., & Marlowe, D. A. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24, 349-354.

Daly, M., & Wilson, M. (1988). Homicide. New York: Aldine.

Darwin, C. (1872). The expression of emotion in man and animals. London: Murray.

Darwin, C. (1871). The descent of man and selection in relation to sex. Vol. 1. London: Murray.

Davidson, T. (1977). Wife beating: A recurring phenomenon throughout history. In M. Roy (ed).), Battered women: A psychosociological study of domestic violence (pp. 1 - 23). New York: Van Nostrand.

deReincourt, A. (1974). Sex and power in history. New York: Delta.

Doane, J.A., & Diamond, D. (1994). Affect and Attachment in the Family. Basic Books: New York.

Dobash, R. E., & Dobash, R. P. (1978). Wives: The appropriate victims of marital assault. Victimology: An International Journal, 2, 426-442.

Dobash, R. E., & Dobash, R. P. (1979). Violence against wives: A case against the patriarchy. New York: Free Press.

Dutton, D.G. (1986). The outcome of court-mandated treatment for wife assault: A quasi-experimental evaluation. Violence and Victims, 1, 163 - 175.

Dutton, D. G. (1988). The domestic assault of women: Psychological and criminal justice perspectives. Allyn & Bacon: Boston.

Dutton, D.G. (1994a). Behavioral and affective correlates of Borderline Personality Organization in wife assaulters. International Journal of Law and Psychiatry, 17(3),

265 - 277

Dutton, D.G. (1994b). The origin and structure of the abusive personality. Journal of Personality Disorders, 8(3), 181 - 191.

Dutton, D.G. (1994c). Comparing assaultive males with clinical outpatients: A cross validation of the Propensity for Abusiveness Scale. Unpublished manuscript., Department of Psychology, University of British Columbia.

Dutton, D.G. (1994d). Accuracy of the Propensity for Abusiveness Scale (PAS) in assessing abuse in a clinical outpatient sample. Unpublished ms., Department of Psychlogy, University of British Columbia.

Dutton, D. G. (1995). The domestic assault of women: Psychological and criminal justice perspectives (Second Edition). Vancouver: University of British Columbia Press. Dutton, D.G. (in press, a). Patriarchy and wife assault: The ecological fallacy. Violence and Victims, 9(2), 125 - 140.

Dutton, D.G. (in press, b). Trauma symptoms and PTSD profiles in perpetrators of abuse. Journal of Traumatic Stress.

Dutton, D.G. (in press, c). The propensity for abusiveness scale. Journal of Family Violence.

Dutton, D.G., Bodnarchuk, M., Kropp, R., Hart, S. & Ogloff, J. (1995). A ten year follow-up of treated and untreated wife assaulters. Unpublished ms., BC Institute on Family Violence, Vancouver, British Columbia.

Dutton, D.G., & Browning, J.J. (1988). Power struggles and intimacy anxieties as causative factors of violence in intimate relationships. In G. Russell (Ed). Violence in intimate relationships. Great Neck, New York: PMA Publishing.

Dutton, D.G., Hart, S.D., Kennedy, L.W., & Williams K.R. (1990). Arrest and the reduction of repeat wife assault. In E.S. Buzawa and C.G. Buzawa (eds.), Domestic violence: The changing criminal justice response (pp. 111 - 127). Westport, CT: Greenwood. Dutton, D.G., & Hemphill, K.J. (1992). Patterns of socially desirable responding among perpetrators and victims of wife assault. Violence and Victims, 7(1), 29-39.

Dutton, D.G., & Landolt, M. (1995). Personality correlates of abusiveness in Gay and Heterosexual males. Unpublish ms. (under review), Department of Psychology, University of British Columbia.

Dutton, D.G., Landolt, M., Starzomski, A., & van Ginkel, C. (1995). The centrality of abusive personality in several populations of males. Unpublished ms., Department of Psychology, University of British Columbia.

Dutton, D.G., & Ogloff, J. (1994). Preliminary report on predictors of post treatment success/failure in court mandated treatment of wife assault. Unpublished manuscript. Department of Psychology, University of British Columbia.

Dutton, D.G., Saunders, K., Starzomski, A. & Bartholomew, K. (1994). Intimacy anger and insecure attachment as precursors of abuse in intimate relationships. Journal of Applied Social Psychology, 24(15), 1367 - 1386.

Dutton, D.G., & Starzomski, A. (1994). Psychological differences between court-referred and self-referred wife assaulters. Criminal Justice and Behavior, 21(2), 203 - 222.

Dutton, D.G., & Starzomski, A. (1993). Perpetrator characteristics associated with women's reports of psychological and physical abuse. Violence and Victims, 8(4), 326 - 335.

Dutton, D.G., Starzomski, A.J. & Ryan, L. (1994). Antecedents of abusive personality and abusive behavior in wife assaulters. Unpublished manuscript.(under review). Department of Psychology, University of British Columbia.

Dutton, D. G., & Strachan, C. E. (1987). Motivational needs for power and dominance as differentiating variables of assaultive and non-assaultive male populations. Violence and Victims, 2(3), 145-156.

Dutton, D.G., van Ginkel, C. & Starzomski, A.J. (1994). The role of shame and guilt in the intergenerational transmission of abusiveness. Unpublished manuscript.(under review) Department of Psychology, UBC.

Elliott, F. (1977). The neurology of explosive rage: The episodic dyscontrol syndrome. In M. Roy (Ed.). Battered women: A psychosociological study of domestic violence. New York: Van Nostrand.

Faulk, M. (1974). Men who assault their wives. Medicine, Science and the Law, 14, 180-183.

Fincham, F.D., & Bradbury, T.N. (1992). Assessing attributions in marriage: The relationship attribution measure. Journal of Personality and Social Psychology, 62(3), 457-468.

Ganley, A. (1981). Court-mandated counseling for men who batter. Center for Women Policy Studies: Washington, D.C.

Ganley, A. (1989). Integrating Feminist and Social Learning Analyses of Aggression: Creating multiple models for intervention with men who batter. In P.L. Caesar, L.K. Homberger Treating Men who Batter: Practice and Programs. Springer, N.Y.

Goldner, V., Penn, P., Sheinberg, M. & Walker, G. (1990). Love and violence: Gender paradoxes in volatile attachments. Family Process, 29(4), 343 - 364.

Goode, W. G. J. (1971). Why men resist. Dissent, 181-193.

Gould, S.J. (1983, June). Genes on the brain. New York Review of Books, p. 5.

Griffin, D.W., & Bartholomew, K. (1994). The metaphysics of measurement: The case of adult attachment. In K. Bartholomew & D. Perlman (Eds.), Advances in personal relationships vol. 5: Attachment processes in adulthood (pp. 17-52). London: Jessica Kingsley Publishers.

Gunderson, J.G. (1984). Borderline personality disorder. Washington, DC: American Psychiatric Press, Inc.

Hart, S. D., Dutton, D. G., & Newlove, T. (1993). The prevalence of personality disorder amongst wife assaulters. Journal of Personality Disorders, 7(4), 328-340.

Herman, J.L., & van der Kolk, B. (1987). Traumatic antecedents of Borderline Personality Disorder. In B. van der Kolk (Ed), Psychological Trauma, American Psychiatric Press, Washington, DC, pp. 111 - 126.

Herman, J.L., Perry, J.C., & van der Kolk, B. (1989). Childhood trauma in Borderline Personality Disorder. American Journal of Psychiatry, 146:, 490-495.

Holtzworth-Monroe, A., & Stuart, G.L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, 116(3), 476 - 497.

Jaffe, P., Sudermann, M., Reitzel, D., & Killip, S.M. (1992). An evaluation of a secondary school primary prevention program on violence in intimate relationships. Violence and Victims, 7(2) 129 - 154.

Kalmuss, D. S. (1984, February). The intergenerational transmission of marital aggression. Journal of Marriage and the Family, 46, 11-19.

Kennedy, L.W. & Dutton, D.G. (1989). The incidence of wife assault in Alberta. Canadian Journal of Behavioural Science, 21(1), 40 - 54.

Kernberg, O. (1977). The structural diagnosis of borderline personality organization. In P. Hartocollis (Ed.). Borderline Personality Disorder: The concept, the syndrome, the patient. New York: International Universities Press.

Kim, K.-I., & Cho, Y.-G. (1992). Epidemiological survey of spousal abuse in Korea. In E. Viano (ed.), Intimate Violence: Interdisciplinary Perspectives. Hemisphere: Washington.

Lewis, H.B. (1971). Shame and guilt in neurosis. NewYork: International Universities Press.

Lewis, H.B. (1987). Shame and the narcissistic personality .In D. Nathanson (ed). The many faces of shame New York: Guilford.

Lie, G., Schilit, R., Bush, J., Montague, M., & Reyes, L. (1991). Lesbians in currently aggressive relationships: How frequently do they report aggressive past relationships? Violence and Victims, 6(2), 121-135.

Linehan, M. (1987). Dialectical behavior therapy for Borderline Personality Disorder. Bulletin of the Menninger Clinic, 51, 261 - 276.

Mahler, M. (1971). A study of the separation-individuation process and its possible application to borderline phenomena in the psychoanalytic situation. Psychoanalytic Study of the Child, 26, 403-424.

Miller, S. (1985). The Shame Experience. Erlbaum: Hillsdale NJ.

Millon, T. (1987). Manual for the Millon Clinical Multiaxial Inventory-II. National Computer Systems, Minneapolis, Minnesota.

Novaco, R. (1975). Anger Control. Lexington Books: Lexington. Oldham, J., Clarkin, J., Appelbaum, A., Carr, A., Kernberg, P., Lotterman, A. & Haas, G. (1985). A self-report instrument for Borderline Personality Organization. In: T.H. McGlashan (Ed.) The Borderline: Current empirical research. The Progress in Psychiatry Series (pp. 1-18). Washington, D.C.: American Psychiatric Press.

Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46, 598-609.

Paulhus, D. L. (1986). Self-deception and impression management in test responses. In A. Angleitner & J. S. Wiggins (Eds.), Personality assessment via questionnaire. (pp. 142-165) New York: Springer.

Perris, C., Jacobsson, L., Lindstrom, H., von Knorring, L., & Perris, H. (1980). Development of a new inventory for assessing memories of parental rearing behaviour. Acta psychiat. Scand., 61, 265-274.

Prochaska, J.O., DiClemente, C.C., & Norcross, J.C. (1992). In search of how people change: applications to addictive behaviors. American Psychologist, 47(9), 1102 - 1114.

Revitch, E., & Schlesinger, L.B. (1981). Psychopathology of Homicide. Charles Thomas: Springfield, Ill.

Robert, J.A., Ryan, J.J., McEntyre, W.L., McFarland, R.S., & Lips, O.J. (1985). MCMI characteristics of DSM-III PTSD in Vietnam veterans. Journal of Personality Assessment, 49, 226 - 230.

Rosenfeld, G. (1992). Court ordered treatment of spouse abuse. Clinical Psychology Review, 12, 205 - 226.

Saunders, D.G. (1992). A Typology of men who batter: Three types derived from cluster analysis. American Orthopsychiatry, 62(2), 264-275.

Sherman, L. W., & Berk, R. A. ( 1984). The specific deterrent effects of arrest for domestic assault. American Sociological Review, 49, 261 - 272.

Sherman , L. W., Schmidt, J.D., Rogan, D.P., Smith, D.A., Gartin, P.R., Cohn, E.G., Collins, DJ., & Bacich, A.R. (1992). The variable effects of arrest on criminal careers: The Milwaukee domestic violence experiment. Journal of Criminal Law and Criminology, 83(1), 137 - 161.

Siegel, J.M. (1986). The multidimensional anger inventory. Journal of Personality and Social Psychology, 5(1), 191 - 200.

Simeons, W. (1962). Man’s presumptuous brain. New York: Dutton.

Symons, D. (1980). The evolution of human sexuality. Cambridge, MA: Cambridge University Press.

Smith, M. (1990). Patriarchal ideology and wife beating: A test of feminist hypothesis. Violence and Victims, 5(4), 257-273.

Sonkin, D.J., Martin, D., & Walker, L. (1985). The male batterer: A treatment approach. Springer: New York.

Sonkin, D.J. (1987). The assessment of court-mandated male batterers. In D.J. Sonkin (Ed.), Domestic violence on trial: Psychological and legal dimensions of family violence (pp. 174-196). New York: Spring Publishing Company.

Sorenson, S.R., & Telles, C.A. (1991). Self-reports of spousal violence in a Mexican-American and a non-Hispanic white population. Violence & Victims, 6(1), 3 - 15.

Starzomski, A. (1993). Attachment style, affect and the construal of interpersonal conflict. Unpublished master’s thesis, Department of Psychology, University of British Columbia.

Starzomski, A.J., & Dutton, D.G. (1994). Attachment, anger and attribution. Unpublished manuscript. (under review). Department of Psychology, University of British Columbia.

Straus, M. A. (1976). Sexual inequality, cultural norm and wife beating. Victimology, l, 54 - 76.

Straus, M. A. (1977a). Societal morphogenesis and intrafamily violence in cross cultural perspective. Annals of the New York Academy of Sciences, 285, 718-730.

Straus, M. A. (1977b). Wife beating: How common and why? Victimology, 2(3-4), 443-459.

Straus, M. A. (1977c, March). Violence in the family: How widespread, why it occurs and some thoughts on prevention. In Family Violence: Proceedings from Symposium, United Way of Greater Vancouver.

Straus, M. A. (1979). Measuring family conflict and violence: The conflict tactics scale. Journal of Marriage and the Family, 41, 75-88.

Straus, M. A., Gelles, R. J., & Steinmetz, S. (1980). Behind closed doors: Violence in the American family. Garden City, NY: Anchor Press/Doubleday.

Straus, M.A. & Gelles, R.J.(1992). How violent are American families? Estimates from the National family violence resurvey and other surveys. In M.A. Straus & R.J. Gelles (eds.) Physical Violence in American Families. Transaction: New Brunswick, N.J.

Straus, M.A. & Kantor, G.K. (1994). Change in spouse assault rates from 1975 to 1992: A comparison of three national surveys in the U.S. Paper presented at the 13th World Congress of Sociology, Biedefeld, Germany.

Tangney, J.P. (1991). Moral affect: The good, the bad and the ugly. Journal of Personality and Social Psychology, 61(4), 598 - 607.

Tangney, J.P., Wagner, P., Fletcher, C., & Gramzow, R. (1992). Shamed into anger? The relation of shame and guilt to anger and self-reported aggression. Journal of Personality and Social Psychology, 62(4), 669 - 675.

Tolman, R.M. (1989). The development of a measure of psychological maltreatment of women by their male partners. Violence and Victims, 4(3), 159-177.

Tompkins, S.S. (1987). Shame. In D. Nathanson (ed). The many faces of shame. New York: Guilford.

van der Kolk, B. (1987). Psychological trauma. Washington, DC: American Psychiatric Press.

Wallace, R. & Nosko, A. (1993). Working with shame in group treatment of male batterers. International Group Psychotherapy, 43(1), 45 - 61.

Walker, L. E. (1979). The battered woman. New York: Harper & Row.

Wilson, E.O. (1978). On human nature. Cambridge: Harvard University Press.

Wilson, E.O. (1975). Sociobiology. Cambridge: Harvard University Press.

Wurmser, L. (1981). The mask of shame. Baltimore: Johns Hopkins University Press.

Wynne-Edwards, V.C. (1962). Animal dispersion in relation to social behavior. Edinburgh & London: Oliver & Boyd.

Yllo, K. & Straus, M. (1990). Patriarchy and violence against wives: The impact of structural and normative factors. In M. Straus & R. Gelles (eds.), Physical violence in American families (pp. 383-399). New Brunswick, NJ: Transaction Publishers.

Zanarini, M.C., Gunderson, J.G., Marino. M.F., Schwartz, E.O., & Frankenburg, F.R. (1989). Childhood experiences of borderline patients. Comprehensive Psychiatry, 30(1), 18-25.

Table 1

Correlations of BPO with Self Reported Abusiveness and Abuse Related Scales

n = 160 (males), n = 80 (female partners)

BPO TOTAL

Verbal/Symbolic Aggression (CTS)	.32**
Physical Aggression (CTS)	.24*
Trauma Symptom Checklist	.73***
Jealousy	.41***
Anger (MAI)	.62***
Partner’s Report of Domination/Isolation	.58***
Partner’s Report of Emotional Abuse	.55***
Partner’s Report of Physical Abuse	.29**

MAI - Multidimensional Anger Inventory

CTS - Conflict Tactics Scale

* p< .05, ** p< .01, *** p < .001

Table 2

Correlations of RSQ with Total Scores on Other Measures for Entire Sample

n = 160

	Secure	Fearful	Preoccupied	Dismissing
BPO	-.35***	.58***	.42***	-.04
Trauma Symptons	-.28***	.50***	.34***	-.03
Anger	-.36***	.49***	.20	.02
Jealousy	-.16*	.34***	.18*	-.015
Verbal Abuse (PMWI)
Domination/Isolation	-.30*	.46**	.27*	.06
Emotional Abuse	-.09	.52***	.26*	-.20

*p<.05 **p<.01 ***p<.001

Table 3

Fearful attachment, BPO in Assaultive, Controls and College Males

	Assaultive^1	Control^1	College Students^2
BPO	+.58***
MAI	+.49***	+.52***	+.58***
PMWI
Domination	+.46**	+.50**	+.45***
Emotional Abuse	+.52***	+.40**	+.50***
Attribution of Blame			+.33

n = 160 n = 90 n = 72

**p<.01 ***p<.001

¹ Dutton & Starzomski, 1993

² Starazomski & Dutton, 1994; Dutton, Landolt, Starzomski, & van Ginkel, 1995

Figure 1

Model of Family of Origin Effects on Abusive Personality and Behavior

Figure 2

MCMI Profiles of PTSD, Wife Assault, and Control Groups

Figure 3

The Centrality of BPO in an Assaultive Group of Males

Figure 4

The Centrality of BPO in a Non-Assaultive Group of Males

Invited Paper: Clinical Psychology: Science and Practice

First Draft:Revised: Sept. 20, 1994

Second Draft Revised: Oct 26, 1994

Third Draft Revised: Nov. 24, 1994

Reference: Clinpsyc2.doc

Nov. 5, 1994 AK

Nov. 25, 1994 AK

November 29, 1994 AK