Dissociative Pathways to Homicide: Clinical and Forensic Implications Andrew K. Moskowitz, PhD ABSTRACT. In a earlier review, dissociation was found to be linked to violence in a wide range of populations, including college students, young mothers, psychiatric patients, and criminal offenders, and was of- ten expressed in the violent act itself, in the form of depersonalization or subsequent amnesia (Moskowitz, 2004). While that review focused on “state” dissociation–i.e., dissociation during an act of violence, this pa- per looks at links between “trait” dissociation and violence–evidence that long-term dissociative processes may predispose vulnerable indi- viduals to violent behavior, and even homicide. Specifically, it is argued that dissociation is related to the development of four “types” (probably not mutually exclusive) of homicide offenders: (1) violent individuals suffering from dissociative identity disorder, whose violence is ex- pressed through certain dissociated personality states, (2) fantasy- driven violence, often accompanied by some form of identity alteration, (3) paroxysms of “dissociative rage” in individuals normally presenting as polite or meek, characterized as “overcontrolled hostile” (Megargee, 1966), and (4) persons designated as psychopathic, whose emotional numbing may be trauma-based and similar to that seen in depersonal- ization disorder. While the last three areas are somewhat speculative, research findings supporting each proposed link are presented. In con- junction with Moskowitz (2004), this review strongly suggests a signifi- cant role for dissociation not only in the commission of certain violent acts, but also in the development of some violent individuals; the clinical Please note that this electronic prepublication galley may contain typographical errors and may be missing artwork, such as charts, photographs, etc. Pagination in this version will differ from the published version. Andrew K. Moskowitz is affiliated with the Department of Psychology, University of Auckland, Auckland, New Zealand. Address correspondence to: Andrew K. Moskowitz, PhD, Department of Psychol- ogy, University of Auckland, Private Bag 92019, Auckland, New Zealand (E-mail: a.moskowitz@auckland.ac.nz). Journal of Trauma & Dissociation, Vol. 5(3) 2004 http://www.haworthpress.com/web/JTD 2004 by The Haworth Press, Inc. All rights reserved. Digital Object Identifier: 10.1300/J229v05n03_02 5 and forensic implications of this are discussed. Further research system- atically exploring the role of dissociative processes in violent and homi- cidal behavior is called for. [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: <docdelivery@haworthpress.com> Website: <http://www.HaworthPress.com> 2004 by The Haworth Press, Inc. All rights reserved.] KEYWORDS. Dissociation, psychopathy, homicide, fantasy, deper- sonalization Having previously argued that dissociative episodes are commonly experi- enced by individuals engaging in violent behavior, i.e., “state” dissociation (Moskowitz, 2004), I suggest in this paper several ways by which dissociative processes, i.e., “trait” dissociation, may lead to the development of homicidal behavior in vulnerable individuals. Broadly speaking, dissociation may relate to homicidal behavior in three ways. The first involves individuals who report having experienced diss- ociative symptoms, such as depersonalization or amnesia, while committing a homicide, but show little evidence of ongoing dissociative pathology either before or after the violent episode (Tanay, 1969). For example, studies previ- ously reviewed (Moskowitz, 2004) suggest that up to a third of individuals convicted of homicide claim total or partial amnesia for the violent attack, many of whom nonetheless accept full responsibility for their actions. These cases tend to involve extreme emotional states (Kopelman, 1987; Schacter, 1986), with the amnesia most often associated with the most emotionally aroused segments of the violent act, a phenomenon which has been termed a “red-out” (Swihart, Yuille, & Porter, 1999). As some of these individuals did not appear dissociative either before or after the incident, the dissociative ex- periences may result from the traumatic nature of the violent behavior on the perpetrators themselves (Spitzer et al., 2001; Steiner, Garcia, & Matthews, 1997), or may only be expressed under the extreme, highly emotional, circum- stances that led to the attack (Swihart et al., 1999). This point is made explic- itly in a study of police officers involved in shooting incidents, in which over 90% reported experiencing dissociative symptoms at the time (Rivard, Dietz, Martell, & Widawski, 2002): (T)he occurrence of dissociative symptoms among such a high propor- tion of shooting-involved officers casts serious doubt on the credibility of those who argue that dissociation at the time of a crime is a mental dis- ease or defect. . . . It would be more reasonable to believe that, in general, dissociation is a normal response of some criminals to the traumatic events they create. (p. 6) 6 JOURNAL OF TRAUMA & DISSOCIATION Such cases, in which there is evidence of dissociation at the time of the vio- lent incident (“state”), but no ongoing dissociative pathology (“trait”), are ex- plored in detail in Moskowitz (2004), and will not be discussed further here. It contrast to Rivard et al., however, it is contended here that dissociation at the time of a crime may in some cases indicate the presence of a dissociative disor- der prior to the offense, and thus possibly a mental disease or defect for the purposes of an insanity defense. 1 Thus, the second possible relationship between dissociation and homicide involves individuals engaging in violent behavior who clearly met criteria for a dissociative disorder–generally dissociative identity disorder–prior to the incident (though possibly only in retrospect). These cases, in which there is evi- dence of both state (i.e., at the time of the violent incident) and trait dissociation, while also discussed in Moskowitz (2004), will be presented here as well. Finally, in between these two extremes are cases, it is argued, in which dis- sociative processes play a major role in the development of homicidal behav- ior, in individuals who may or may not currently meet criteria for a dis- sociative disorder, and may or may not have experienced dissociation at the time of the crime. In this section, making up the bulk of the paper, three types of violent offenders (possibly in the order of decreasing dissociative pathol- ogy) are presented: (1) fantasy-driven homicide offenders, whose attacks are preceded by some form of identity confusion or alteration, (2) meek or mild-mannered individuals who erupt (usually on only one occasion) in a par- oxysm of violence, characterized as “overcontrolled hostile” (Megargee, 1966), and (3) violent persons characterized as psychopathic, whose emo- tional numbing may be similar to that seen in depersonalization disorder. There is no assumption that these groups are necessarily mutually exclusive. VIOLENCE AND HOMICIDE IN PERSONS WITH DISSOCIATIVE IDENTITY DISORDER (DID) Dissociation researchers have long wondered about a relationship between dissociation and violence. Some have argued that the overwhelming prepon- derance of women among those diagnosed with dissociative disorders could partly be due to males with dissociative disorders being funneled into the criminal justice system and their diagnoses missed, and have called for a system- atic exploration of this important area (Carlson & Putnam, 1993; Steinberg, 1995). Indeed, while there is an extensive literature on the relationship between mental disorders and violence (see Monahan, 1992, and Eronen, Angermeyer, & Schulze, 1998 for useful reviews; and Woodward, Nursten, Williams, & Badger, 2000, specifically on homicide), dissociative disorders are not among the mental disorders assessed in such studies (diagnoses gener- Andrew K. Moskowitz 7 ally include psychoses such as schizophrenia and affective disorders such as bipolar disorder and major depression, along with substance abuse and per- sonality disorders). Of interest, posttraumatic stress disorder (PTSD), also rarely assessed in this literature, was found in almost 10% of a large sample of New Zealand male and female prisoners; schizophrenia was found in just over 4% (Brinded, Simpson, Laidlaw, Fairley, & Malcolm, 2001). As PTSD is closely related to dissociative disorders (Bremner et al., 1992; Carlier, Lam- berts, Fouwels, & Gersons, 1996), such findings support calls for research ex- ploring dissociation in prison samples. Such calls have recently been answered with six studies published since 2000 assessing the prevalence of dissociation or dissociative disorders in prison, jail or offender populations (Campbell, 2000; Carrion & Steiner, 2000; Friedrich et al., 2001; Simoneti, Scott, & Murphy, 2000; Stein, 2000; Walker, 2002). This spate of recent studies has allowed a review of the area, which re- vealed that approximately 25% of prison or jail inmates demonstrate “patho- logical” levels of dissociation (primarily determined by DES scores of 30 or above), and that a somewhat smaller percentage meet diagnostic criteria for a dissociative disorder (see Moskowitz, 2004, for more details). A few studies specifically addressed dissociative identity disorder (DID), with the most con- servative finding a prevalence of 6.3% in a hospitalized (psychiatric and medi- cal) male prison sample, most of whom had been convicted of violent offences (Stein, 2000). In addition, scores above 50 on the DES have been found in be- tween 7.0 and 9.5% of prison or jail populations (Graham, 1993; Moskowitz, 2001; Snow, Beckman, & Brack, 1996), a substantial portion of whom would have likely met criteria for DID (Carlson et al., 1993). These results, in con- junction with several mental health professionals reporting finding more than a dozen inmates with DID in their clinical samples (Carlisle, 1991; Lewis, Yeager, Swica, Pincus, & Lewis, 1997; Snow et al., 1996), suggest that DID may not be an uncommon phenomenon in prisons. Unfortunately, no study to date has determined whether such diagnoses are more common among homi- cide and violent than non-violent offenders. The parallel literature, exploring violence and homicide in persons diag- nosed with DID, has no such shortcomings, but does suffer from a lack of vali- dation of reported violent behavior. Several studies have examined violent and homicidal behavior in individuals diagnosed with DID, as well as the preva- lence of violent or homicidal “alters.” It has been suggested that such alters, or personality states, may develop through a process of identification with an abuser, or as a “protector” personality which has become aggressive over time (Putnam, 1989). Putnam et al. (1986) found violent alters in 70 out of 100 DID patients, only eight of whom were male. Twenty percent claimed to have com- mitted a sexual assault, and six percent claimed to have committed homicide. In a study of adolescents with DID, 82% had aggressive “persecutor” alters, 8 JOURNAL OF TRAUMA & DISSOCIATION and over a third reported having been violent or threatened violence (Dell & Eisenhower, 1990). In the largest study of men with DID to date, Loewenstein and Putnam (1990) compared 21 men with DID to a sample of female DID pa- tients from the Putnam et al. (1986) study. Violent alters were more common in the male (90%) than female (74%) participants, but there was no difference in the rate of homicidal alters (present in about a third of each gender). In addi- tion, a history of criminal conviction was more common in the male than the female DID participants (47% compared to 35%). All of the convicted men had been incarcerated; for more than 40% of them (20% of the entire male sample), and for 20% of the incarcerated women, the convictions were report- edly for committing homicide. Unfortunately, these were all clinically gener- ated samples, with no reported external validation for the claims of homicidal or violent behavior. 2 Nonetheless, it is worth noting that the homicide rates re- ported in these two DID samples, 6% for the women and 20% for the men (Loewenstein & Putnam, 1990; Putnam et al., 1986), is considerably higher than those reported for individuals with schizophrenia, less than 0.5% over a 12-year period in one study (Tiihonen, Hakola, Eronen, Vartiainen, & Ryynanen, 1996). Thus, while there are clear gaps in both sets of literature, taken together they suggest that violent alters may express violent or even homicidal behav- ior in a significant minority of individuals with DID. As suggested below, and as seen in cases of PTSD, such violence may occur in situations (or towards persons) that are reminiscent of an earlier traumatic event (Silva, Derecho, Leong, Weinstock, & Ferrari, 2001). Nonetheless, it is important to note that there is no evidence to suggest that most persons with DID are violent to oth- ers; many are violent to themselves, and others manage to avoid engaging in any form of violent behavior whatsoever. Dorothy Otnow Lewis, in her book Guilty by Reason of Insanity (1998), discusses one such case (see also Keyes, 1981, for a famous treatment of an in- dividual with DID who committed a violent crime and was acquitted by reason of insanity) in which an elderly nun was raped and murdered by a young man initially considered by her and almost all parties involved to be suffering from schizophrenia. Only after several years of contact (during which time he was found guilty and sentenced to death), numerous interviews in which personal- ity shifts were observed, and the confirmation of childhood sexual abuse by other family members, did Lewis conclude that he was actually suffering from dissociative identity disorder. 3 Her final thoughts on this case provide a fitting close to this discussion of DID and violence: We shall never know for sure who among Johnny’s alters actually mur- dered the aged nun. Was it Aaron? Aunt Barbara? My own guess is that it was the powerful, controlling figure with the fiery eyes–the one who Andrew K. Moskowitz 9 most intimidated Aaron, the one who refused to give me his name. We can be fairly certain, however, that whoever committed the murder, whoever slashed the throat of the innocent nun, did not see Sister Catherine’s face when he did it. He saw instead the face of Granny. Al- ters, stuck in time, are always mistaking one situation for another, for- ever confusing someone with someone else. (Lewis, 1998, pp. 235-236) FANTASY-DRIVEN VIOLENCE IN SERIAL AND MASS MURDERERS Several authors have identified fantasy as an important factor in the de- velopment of homicidal behavior, particularly with regard to serial or mass murderers (MacCulloch, Snowden, Wood, & Mills, 1983; Meloy, Hempel, Mohandie, Shiva, & Gray, 2001; Prentky et al., 1989; Ressler, Burgess, & Douglas, 1988). This has been stated perhaps most clearly by Ressler, Douglas, and others from the United States FBI (Prentky et al., 1989; Ressler et al., 1988; Ressler & Shachtman, 1992). Following a detailed study of 36 murder- ers whose crimes all had sexual components, many of whom were serial kill- ers, Ressler et al. (1988) concluded that the murders were motivated by detailed fantasies, which often had “strong visual components.” Importantly, violent fantasies were much more commonly reported in serial (86%) than in single (23%) sexual murderers (Prentky et al., 1989). In exploring the histories of serial and mass murderers, high levels of child- hood abuse and neglect have been found (Levin & Fox, 1985; Ressler et al., 1988). Several authors have speculated that the intense fantasy lives reported by these individual may emerge out of their traumatic childhood experiences (Dietz, Hazelwood, & Warren, 1990; MacCulloch, Gray, & Watt, 2000; Prentky et al., 1989). Ressler and colleagues (1988) provide an example of how this might happen: A child abused by an adult caretaker begins to think about being hit ev- ery time an adult comes near him, dwelling on the hitting. He may imag- ine (fantasize) about someone coming to help him by beating up the adult. This thinking pattern may bring relief, because someone has pro- tected him in his fantasy. (pp. 34-35) 4 When the serial murderers in Ressler’s study were asked what led to their first murder, they commented on their “long-standing preoccupation and pref- erence for a very active fantasy life,” which was primarily violent and sexual in nature (Ressler et al., 1988; p. 33). Of interest, the murderers who had been sexually abused began fantasizing significantly earlier, on average three and a 10 JOURNAL OF TRAUMA & DISSOCIATION half years, than the men who did not report sexual abuse. Crimes were often rehearsed in these fantasies; as Ressler (1992) states, “they murdered to make happen in the real world what they had seen over and over again in their minds since childhood and adolescence” (p. 84). A similar pattern was found in stud- ies of adolescent mass murderers, in which 44% reported daily indulgence in violent fantasies (Meloy et al., 2001). Fantasy is a common childhood experience that several authors have con- sidered to be a developmental substrate for dissociation (Kluft, 1994; Putnam, 1989). While most children with active fantasy lives do not go on to develop dissociative disorders, several recent studies have strongly linked “fantasy proneness” with dissociative pathology, in both substance abuse and uni- versity undergraduate populations (Merckelbach, Horselenberg, & Schmidt, 2002; Merckelbach, Rassin, & Muris, 2000; Pekala, Angelini, & Kumar, 2001; Pekala et al., 1999; Rauschenberger & Lynn, 1995; Waldo & Merritt, 2000). All studies, with a combined total of over 4000 participants, found DES scores to be strongly related to “fantasy proneness,” as measured on the Inven- tory of Childhood Memories and Imaginings (Wilson & Barber, 1983, cited in Rauschenberger & Lynn, 1995) or the Creative Experiences Questionnaire (Merckelbach, Horselenberg, & Muris, 2001). Further, a recent study found fantasy proneness highly predictive of dissociation, considerably more so than a diagnosis of PTSD (Muris, Merckelbach, & Peeters, 2003). These results support an early study’s claim that a “growing body of evidence” pointed to “parallels between imagination and fantasy-based activities and dissociative experiences and symptoms” (Rauschenberger & Lynn, 1995, p. 378). With regard to possible etiological mechanisms, Rhue and Lynn found that participants scoring high on fantasy proneness reported having experienced a greater frequency and intensity of physical punishment, and more revenge fantasies, than participants scoring in the medium or low ranges (Rhue & Lynn, 1987). They speculated that fantasy proneness in the individuals they studied developed either through encouragement from an important adult to fantasize, or as a means of escaping from child abuse, harsh punishment, or other aversive environments. It can thus be proposed that the active fantasy lives seen in many serial and mass murderers, developing in response to childhood abuse and neglect (Dietz et al., 1990; MacCulloch et al., 2000; Prentky et al., 1989), is explicitly dissociative in nature, or progresses to dissociative pathology at some point. What remains to be determined, however, is how and under what circum- stances violent fantasies lead to violent behavior. One possibility has been proposed by Carlisle (1991, 1993), who describes a progression from fantasy to violence as resulting from a dissociation of iden- tity, produced through intense immersion in violent fantasies. He describes the Andrew K. Moskowitz 11 way in which fantasy can drive the development of an “evil” alternative iden- tity or personality, often referred to as the “shadow,” “beast,” or “dark side.” Based on interviews with a number of serial murderers, Carlisle (1993) de- scribes a dialectical process, a tension between fantasy and reality, ultimately leading to the development of a more violent personality state and the expres- sion of violent behavior: As the person shifts back and forth between the two identities in his at- tempt to meet his various needs, they both become an equal part of him, the opposing force being suppressed when he is attempting to have his needs met through the one. Over time, the dark side (representing the identity or entity the person has created to satisfy his deepest hunger) be- comes stronger than the “good” side, and the person begins to experi- ence being possessed, or controlled by this dark side of him. This is partly because the dark side is the part anticipated to meet the person’s strongest needs, and partly because the good side is the part that experi- ences the guilt over the “evil” thoughts, and therefore out of necessity is routinely suppressed. Thus, the monster is created. (p. 27) This fantasy world becomes of great importance to these individuals, taking over more and more of their lives (Carlisle, 1993). As it does, the person be- comes more secretive and concerned about protecting the fantasy world from discovery from others (through diaries, books, magazines, etc.), further driving the compartmentalization process. Such a process can be seen in a number of adolescent mass murderers, including those at Columbine (Meloy et al., 2001). Thus, the existence of an intense violent fantasy world, developing in re- sponse to childhood abuse or neglect, has been found to be common in adoles- cent and adult sexual, serial and mass murderers. Persons who intensively fantasize may be highly dissociative, particularly when those fantasies have strong visual as well as verbal components (Rauschenberger & Lynn, 1995). Finally, there is evidence that ongoing indulgence in violent fantasies can lead to the expression of murderous impulses in vulnerable individuals, or as Carlisle has described, the emergence of a “dark side.” In contrast to violent individuals with DID, the fantasy-driven violent indi- vidual described above would likely not have met criteria for a dissociative disorder prior to the violence (though their fantasies may well have been dissociative in nature), but likely exhibited dissociative symptoms at the time of the crime (such as derealization and depersonalization), and may (due to the identity alteration process described above) meet criteria for a dissociative disorder, such as dissociative disorder, not otherwise specified (DDNOS), subsequently. 12 JOURNAL OF TRAUMA & DISSOCIATION “OVERCONTROLLED” HOSTILITY AND DISSOCIATED RAGE The previous two sections discussed violent individuals meeting diagnostic criteria for DID, and those individuals driven to murder through long-term vi- olent fantasies, often highly visual in nature, ultimately leading to the develop- ment of a violent personality state. Dissociation can also be linked to a type of homicide offender who in most circumstances appears anything but violent, what Megargee (1966) has termed the “overcontrolled hostile” offender. Megargee (1966) argued that there are primarily two types of violent of- fenders, those whom he characterized as being “undercontrolled” with regard to hostility and aggression, and those whom he described as being “over- controlled.” The anger and hostility of undercontrolled individuals is obvious to those around them, and they often engage in violent incidents (usually non-lethal) when provoked (Megargee, 1966). In contrast, those characterized as overcontrolled are uncomfortable with openly expressing, or even ac- knowledging, anger or frustration and are often seen as quiet or even meek. Paradoxically, if they do ultimately explode, the crime is often more violent than those committed by the undercontrolled group (Megargee, Cook, & Mendelsohn, 1967). These are the individuals whose violence appears to come “out of the blue,” and about whom media, acquaintances, and possibly even friends and family express “amazement” over their actions. Megargee and colleagues (1967) describe these violent individuals as follows: [They are] characterized by excessive inhibitions against the expression of aggression in any form. Since even the normal socially approved out- lets for anger are unavailable, people of the chronically overcontrolled type are often subjected to extreme frustrations as they are exploited by spouses, co-workers, and peers. In the absence of outlets, instigation to aggression may build up over time to the point where, in some cases, even the most massive defenses are overwhelmed and an aggressive act takes place. Because of the extreme amount of instigation which this en- tails, the aggressive act is likely to be of extreme or homicidal intensity. Hence, a person who has never been known to speak a harsh word may suddenly become a murderer. (p. 520) Megargee’s theory led to the development of an MMPI scale designed to identify such overcontrolled offenders, known as the Overcontrolled Hostility or O-H Scale (Megargee et al., 1967), which has been largely validated by sub- sequent research (Greene, Coles, & Johnson, 1994; Lane & Kling, 1979; Quinsey, Maguire, & Varney, 1983; White & Heilburn, 1995). One recent study (Verona & Carbonell, 2000) found a majority of violent female inmates to fit this profile. The group scoring high on the O-H MMPI-2 Scale had sig- nificantly shorter criminal records, but the crimes they had committed were Andrew K. Moskowitz 13 more violent, and included a larger percentage of the homicides, than the group scoring low on the O-H scale. Many of the homicides committed by the high O-H group were of abusive partners. Verona and Carbonell (2000) con- cluded that this group was a majority in the violent female offenders, as op- posed to a minority in the violent men (Megargee et al., 1967) because O-H behaviors fit gender stereotypes of appropriate female behavior (i.e., not ex- pressing anger or frustration openly). The process Megargee describes can be conceptualized as dissociative in nature. In this formulation, the anger, frustration, and rage of an overcon- trolled individual, unable to be tolerated consciously, would be split-off or dis- sociated from the individual’s normal state of awareness or sense of identity, where it would remain (possibly increasing) until released cataclysmically under extreme provocation. Alternatively, this rage could be thought of as a dissociated “part” of the individual, emerging only under the extreme circumstances previ- ously noted. Support for this thesis comes from Tanay (1969), whose “dissociative ho- micides” (over two-thirds of his sample of 53 homicidal men) are described as having an “overdeveloped” superego (due to “violent child-rearing practices”) that is “intolerant of any overt expressions of aggression” (p. 151). Tanay ex- plicitly links this group of men to Megargee’s overcontrolled aggressive group. Cartwright’s (2001) review of “rage-type homicides” is also consistent with this conception. He concludes that a “pattern of overcontrol” best ex- plains the “apparent normality” of the rage-type offender, whose violence is attributed to “dissociated split-off aspects of the personality” (Cartwright, 2001, p. 15). Over the years, a number of other categories of violent or homicide offend- ers have been proposed which appear similar to Megargee’s concept of overcontrolled aggression, and could be linked to dissociation, such as the “sudden” murderer (Weiss, Lambert, & Blackman, 1960), catathymic vio- lence (Kirschner & Nagel, 1996; Schlesinger, 1996), and the DSM-IV catego- ries of Intermittent Explosive Disorder and the culture bound syndrome Amok (American Psychiatric Association, 2000). While the emphasis differs some- what between these categories, the largely unplanned and spontaneous violent behavior subsumed under each is broadly consistent with the expression of dissociated rage in individuals who do not easily express aggression or hostil- ity (Felthous, Bryant, Wingerter, & Barratt, 1991; Weiss et al., 1960). Of the above categories, only Amok will be discussed, as it is often results in homi- cide, and is characterized in the DSM-IV as a dissociative disorder. Amok, which is most common in individuals from Southeast Asian coun- tries (the term comes from Malaysia), is important for this discussion as it is the only disorder in the DSM-IV in which the concepts of dissociation and vio- lence are linked. It is defined as: “a dissociative episode characterized by a pe- 14 JOURNAL OF TRAUMA & DISSOCIATION riod of brooding followed by an outburst of violent, aggressive, or homicidal behavior directed at people and objects” (APA, 2000; p. 899). Amnesia for the episode is also common. Cases of Amok often involve multiple, frenzied, apparently indiscriminate, homicidal attacks, with no immediate warning–though there is often a perceived insult or slight experienced over the preceding days (Hatta, 1996). Similar to the overcontrolled individuals described above, Ma- lay society has traditionally been characterized as tremendously polite, with essentially no normal channels for expressing frustration or aggressive im- pulses (Gaw & Bernstein, 1992; Raffles, 1965, cited in Hatta, 1996). It has been suggested that Amok may be a “culturally prescribed” form of violent behavior, allowing an outlet for the expression of normally suppressed aggres- sive impulses (Carr & Tan, 1976). Certainly, in individuals who rarely allow themselves to express negative affect, the emergence of intense feelings of anger or rage are likely to feel for- eign–producing sensations of depersonalization. In addition, if negative affect is not dealt with in the normal course of daily life, an individual’s capacity to modulate that affect will be limited, leading to the increased likelihood of their becoming overwhelmed and driven by the affect when it surfaces. In overcontrolled violent individuals, it is primarily the expression of anger or frustration that is not normally tolerated; the individual may have some fleeting awareness of the negative feelings, and may have little difficulty ex- periencing or expressing other emotions. In contrast, individuals classified as psychopathic have “deadened” emotions, and are sometimes characterized as being unable to feel almost anything. PSYCHOPATHY AND DEPERSONALIZATION James Gilligan, in his book Violence: Our Deadly Epidemic and Its Causes (1996), offers an apt description of men he calls “the living dead,” who would likely be classified as “psychopathic.” Many murderers, both sane and insane, have told me that “they” have died, that their personality has died, usually at some identifiable time in the past, so that they feel dead, even though their bodies live on. When they say they feel dead they mean they cannot feel anything–neither emotions nor even physical sensations. I have seen many who admit to killing others without so much as a flicker of remorse or any other emo- tion. (p. 33) Gilligan describes the extraordinarily violent childhoods experienced by these men, which he believes led to their state of “deadness.” As can be seen in Andrew K. Moskowitz 15 this section, psychopathy and depersonalization share this emotional “numb- ness,” and may share etiology as well. In addition, their responses to emotional stimuli appear to be quite similar. It will be argued that these two concepts, while clearly not identical, are close enough in several important ways to war- rant further investigation. Psychopathy is an old concept first re-popularized by Cleckley in his book The Mask of Sanity (1941), and modernized by Hare with the development of the Psychopathy Checklist (PCL; Hare, 1980). As originally envisioned by Cleckley, the psychopath is an individual who suffers from a deficit in emo- tional experience, which expresses itself in symptoms such as general poverty of affect, lack of insight, absence of nervousness, superficial charm, patholog- ical lying, egocentricity, inability to love, and an inability to establish close or intimate relationships. He felt that the behavioral features of psychopathy (e.g., impulsivity, antisocial acts, etc.) were secondary to this affective deficit. Cleckley believed that psychopaths could be found in many areas of life, and did not see violence as one of their core characteristics. Of interest, Cleckley, who was also an author of the first popular book on multiple personality, The Three Faces of Eve (Thigpen & Cleckley, 1957), ac- tively considered, in the early editions of The Mask of Sanity , dissociation as a potential explanation for psychopathy. In a several page discussion of dissoci- ation, Cleckley (1950) comments on Janet’s use of the concept, and then states, “Let us for the moment refer to that which is missing in the psycho- path’s response to life (i.e., a deficit in emotional experience) as dissociated instead of using terms that would imply with insistence that it has been dy- namically repressed or, on the other hand, absent through congenital defect” (p. 402). While he ultimately does not, in these early editions, clearly embrace (or clearly reject) the concept of dissociation as an explanatory principle for psychopathy, the fact that it is considered as potentially relevant is signifi- cant. 5 With the development of the Psychopathy Checklist (PCL; Hare, 1980), which was originally designed to identify Cleckley-type psychopaths within a prison population, the concept of psychopathy was broadened, with antisocial behaviors being given equal weight with the affective deficiency. Factor anal- yses of the PCL have most often revealed two factors, which have been labeled “emotional detachment” and “antisocial behavior” (Harpur, Hakstein, & Hare, 1988; Harpur, Hare, & Hakstein, 1989; Patrick, Bradley, & Lang, 1993). 6 In- dividuals called “psychopaths” are currently identified on the basis of high scores on the PCL or PCL-Revised (PCL-R; Hare, 1991). These individuals have consistently been found to be more violent that persons scoring low on the PCL or PCL-R both inside institutions and in the community (Hart, Hare, & Forth, 1994; Patrick & Zempolich, 1998). They also commit a significant portion of homicides 27% in one study covering two large Canadian prisons 16 JOURNAL OF TRAUMA & DISSOCIATION (Woodworth & Porter, 2002). Of interest, almost all the homicides committed by the psychopathic participants in this study were planned, i.e., not reactive or spontaneous in nature, and not associated with emotional arousal, which led the authors to refer to the crimes as “cold-blooded” (Woodworth & Porter, 2002). There is some evidence that the “emotional detachment” factor (which cor- responds largely to Cleckley’s concept of psychopathy) predicts violence equally as well as, if not better than, the “antisocial behavior” factor (Serin, 1996). This is consistent with the thesis presented here, if the “emotional de- tachment” or “deficient affective experience” (Cooke & Michie, 2001) factor of psychopathy is equated with the dissociative symptom of depersonaliza- tion. The basis for such a position is provided below. The author who has most prominently linked the concepts of dissociation and psychopathy is J. Reid Meloy, in his book, The Psychopathic Mind (1988). Meloy, coming from a strongly psychoanalytic perspective, dedicates over thirty pages of the book to a discussion of “splitting and dissociation,” where he addresses topics such as “depersonalization and derealization,” “psychopathy and hysteria,” and “psychopathy and multiple personality disor- der.” Meloy reviews a number of cases of apparently dissociative experiences in murderers such as Ted Bundy and others, and concludes that dissociation is “ubiquitous” in the psychopath. He argues that dissociation relates to the psy- chopathic process in four ways: (1) it is common in the highly aroused emotional states that psychopaths seek out to alleviate their base state of under- arousal; (2) the psychopath’s impaired capacity to form attachments make dissociative experiences unusual but tolerable; (3) during acts of violence dis- sociation allows the victim to be seen as an object; and (4) memories of dissociative experiences can be used as rationalizations for avoiding responsi- bility, because they seem dreamlike or unreal (Meloy, 1988, p. 164). Thus, he seems here to be primarily describing experiences of depersonalization and derealization. 7 Meloy (1988) concludes, “[T]here is a growing empirical basis for the hypothesis that some varieties of dissociative states are significantly re- lated to psychopathy . . .” (p. 164). While Meloy argues for links between psychopathy and dissociation on largely phenomenological grounds, there is some basis to conclude that etiol- ogy may be shared as well. Porter (1996) makes an explicit argument to this ef- fect, noting that childhood abuse can produce difficulties in experiencing affect that could lead to a form of psychopathy, which he calls “secondary” psychopathy, “phenotypically (and diagnostically) indistinguishable” from “classic” or “fundamental” psychopathy (p. 183). Porter argues that, while the “secondary” type is due to abusive childhood experiences, the “fundamental” psychopath is inborn. This argument is premised on the position that “most psychopaths do not come from obviously dysfunctional backgrounds” (p.182). Andrew K. Moskowitz 17 However, this position has been disputed by several recent studies. Koivisto and Haapasalo (1996) and Weiler and Widom (1996) both found a correlation between a history of documented childhood abuse and neglect and high scores on the PCL-R. Moeller and Hell (2003), while not focusing on childhood trauma per se, also found a strong correlation ( r = .62) between the number of reported traumatic experiences and PCL-R scores in a young male prison pop- ulation, and Forth (1994, cited in Forth, 1995) found a variety of childhood variables, including physical punishment, to predict scores on a version of the PCL modified for use with adolescents (Hare, Forth, & Kosson, 1994). Fur- ther, Weiler and Widom (1996) found psychopathy to significantly mediate the relationship between childhood abuse and later violent behavior. Non- etheless, it is possible that these positive results could be due to a subgroup of per- sons meeting criteria for psychopathy, perhaps Porter’s “secondary” psycho- paths, with strong trauma histories, and that others diagnosed as psychopathic do not have such histories. This remains to be explored, as none of these stud- ies was designed to address that possibility. The form of dissociation most relevant to an understanding of psychopathy is depersonalization. Depersonalization Disorder (DD), an uncommon and relatively poorly researched dissociative disorder, is defined in the DSM-IV as “persistent or recurrent experiences of feeling detached from . . . one’s mental processes or body” (APA, 2000; p. 490). According to the DSM-IV-TR, per- sons with DD often experience “sensory anesthesia, lack of affective response, and a sensation of lacking control of one’s actions” (APA, 2000; p. 488). While transient episodes of depersonalization are frequent after exposure to severe stress (APA, 2000), individuals suffering from DD are reported to typi- cally have such experiences on a chronic basis (Phillips et al., 2001). An in- creased prevalence of childhood emotional abuse (but not physical abuse or neglect) was found in one study comparing persons with DD to “healthy con- trols” ( p < .001; Simeon et al., 2001). In addition, a link between trauma and depersonalization experiences explicitly exists in the DSM-IV criteria of Acute Stress Disorder and PTSD (APA, 2000). A PTSD diagnosis may in- clude aspects of depersona