It. \{0"-/~D~ V-::f. iI. W.:£e1 / /7

CHAPTER 25 Pathologies of Attachment, Violence, and Criminality

1. REID MELOY

THE ORIGINS OF PATHOLOGIES OF ATTACHMENT 513 AND RESEARCH 509 ATTACHMENT AND VIOLENCE 514 THE PSYCHOBIOLOGY OF ATTACHMENT 510 Intimate Partner Violence 514 Attachment as a Behavioral System 510 Violence and Criminality 518 Biology and Attachment 511 NEW AVENUES OF FORENSIC RESEARCH Emotion and Cognition 511 AND APPLICATION 519 Attachment and Exploration 512 : The Preoccupied Crime 519 Attachment and 512 : The Dismissive Criminal 521 Attachment and Socialization 512 CONCLUSION 521 Attachment and Caregiving 512 REFERENCES 522 Attachment Behavior and the Attachment Bond 512

.c of the great paradoxes of human existence is that most clinical and empirical evidence in two emerging areas of interpersonal violence occurs between people who are at­ criminality, develop theoretical links to other areas of foren­ tached or bonded to each other. Proximity seeking toward an­ sic knowledge, and suggest directions for both future forensic other and acute distress when unpredictably or permanently research and practical applications. separated, the empirical components of attachment, appear to be the most fertile territory for physical combat. This is an as­ sociation filled with irony, reminding one that the tendency to THE ORIGINS OF ATTACHMENT THEORY "debasement in the sphere of love" (Freud, 1912, p. 177) is a AND RESEARCH widely observed phenomenon. Violent attachments (Meloy, 1992) are not lost in the Attachment is a biologically rooted, species-specific behav­ commonsense behavior of those professionals charged with ioral system that, when activated, maintains close proximity risk managing violent individuals: Judges are most likely to between a child and his or her caretaker. It was first proposed issue protection or restraining orders to prevent domestic vi­ and investigated by John Bowlby, James Robertson, and olence; homicide detectives first suspect sexually or affec­ Mary Ainsworth at the Tavistock Clinic in London following tionately intimate members of the victim's kinship network World War II (Ainsworth & Bowlby, 1954; Bowlby, 1953; when investigating a murder; and child as a form of in­ Robertson & Bowlby, 1952). Attachment behaviors are evi­ terpersonal violence has received an enormous amount of dent in both birds and mammals, but are generally absent in publicly funded legal, clinical, and research attention during reptiles. Individuals with reptiles as pets often misinterpret the past quarter-century. their thermotropic (heat-seeking) behavior as an emotion re­ The clinical and forensic investigation of the relationship lated to attachment or bonding and project on the animal their among attachment, violence, and criminality is quite recent own affectionate feelings. and very promising. In this chapter, 1 summarize and high­ John Bowlby was the fourth child born to a London sur­ light this situation, argue for its relevance on the basis of geon and a country parson's daughter. He was trained as a child psychiatrist and joined the British Psychoanalytic Society at a time when there was great turmoil between the This chapter was supported by a grant from Forensis, Inc. followers of Melanie Klein and those of Anna Freud. (www.forensis.org). Bowlby's personal analyst was Joan Riviere, and one of his

509 510 Pathologies ofAttachment, Violence, and Criminality supervisors was Klein. Troubled by the dogmatism of psy­ predicted the nature of the mother-infant relationship in Ih,' choanalysis at the time, its extrapolations from the couch to last quarter. The Baltimore work also led to the formulalillll the crib, and its dismissive attitude toward empirical investi­ of the "Strange Situation," a 20-minute contrived naturalislil gation of normal development, Bowlby's long-standing inter­ experiment that examined attachment and exploration undl" ests in Darwinism led him to the new science of ethology. minimal and maximal stress, Mother and baby would play. ;, This provided him with a truly scientific framework within stranger would enter the room, mother would leave bricll, which to reformulate his psychoanalytic knowledge. Attach­ and then return. The various stages of this experimenl ;11 ment theory began (Holmes, 1995). lowed Ainsworth to discern differences in the infants' \l' The origins of attachment theory are found in three psycho­ union behavior with their mother. Most of the infants WCll' analytic papers (Bowlby, 1958,1960,1961) which were later immediately soothed by their mother's return and quickly 1\' expanded into Bowlby's trilogy of books: Attachment (1969), sponded to her nurturing. A few, however, were very an~n. Separation (1973), and Loss (1980). The early papers empha­ cried and wanted contact, but would not cuddle and aCCCl'1 sized three findings: (a) There is a primary attachment the nurturing. They were markedly ambivalent. Others wOllld between mother and child that is "hard-wired" and whose evo­ dismiss and ignore the mother even if they searched for hCI lutionary purpose is to protect the infant from predators; until she returned. They were avoidant. Robertson (1953) had (b) anxiety is an affective response to either separation from a documented similar behaviors in his film, A Two Year Old loved one or external threat; and (c) infants and children expe­ Goes to Hospital, and Harlow (1961) had noticed similal rience grief when they experience loss. Although these postu­ patterns in some monkeys. The Baltimore studies are remel11 lates are accepted by most contemporary psychologists, they bered for the development of the Strange Situation cia' were revolutionary during their time because they challenged sification system, which identified three attachment styb the primacy of sexuality in development and emphasized the secure. avoidant, and ambivalent/resistant (Ainsworth, Blehal, impact of evolution and biology on personality. Attachment Waters, & Wall, 1978). theory was an interpersonal theory of mind that stressed an essential harmony between mother and child unless it was disturbed. Bowlby unified the psychoanalytic world against THE PSYCHOBIOLOGY OF ATTACHMENT him for nearly 20 years; he began to achieve a rapprochement only after his appointment as Freud Memorial Professor of Secure or normal attachment assures proximity of the childl' I Psychoanalysis at University College in London, an appoint­ the attachment figure, usually the mother. Smiling, vocalil ment now held by his heir-apparent, Peter Fonagy. ing, and approaching are signaling behaviors that communi Mary Salter Ainsworth, a Canadian psychologist who cate a desire on the part of the child for interaction; otlll'! studied at the University of Toronto, accompanied her hus­ behaviors, most notably , are aversive events for th,' band to London in 1950 and answered a job advertisement in mother and bring her close to the child to terminate them. ;1 the London Times for a research position investigating the negative for both caregiver and child throu~h impact of maternal separation on personality development. the alleviation of their mutual distress. This serendipitous event changed her life, and she collabo­ From an evolutionary perspective, attachment behavill/ rated with Bowlby for many years to come. She left for ensures the survival of the child by protecting him or hl'l Africa with her husband in 1953 and conducted the first em­ froIU predators. Although Bowlby (1969) originally emph;1 pirical study of normal attachment among 26 families with sized survival of the species as the goal of attachment, COli unweaned babies in Uganda. It was here that she began to temporary evolutionary thinking has refocused on thl' validate Bowlby's ethological theory of attachment and also reproductive fitness of the child if he or she grows up, thus in the importance of maternal sensitivity in attachment quality. creasing the probability that the genes of the individual wi II The genesis of secure and insecure attachment can be found survive into the next generation. Ifthe child is eaten, a seClI1 in the "Ganda data" (Ainsworth, 1967). ingly universal and unconscious fear that has sparked both ill While mulling over the findings from Africa, Ainsworth tellectual curiosity (Freud, 1919) and enormous cinemalil began a second observational study with 26 families after she success (Jaws), there will be no future children. relocated to Baltimore in 1963. She collected 72 hours of data during home visits that spanned the first year of the new­ Attachment as a Behavioral System borns' lives. These meticulous narratives documented the difficulties some mothers had responding to their baby's Attachment is a species-specific system of behaviors th;11 cues, and the interactions in the first quarter of observation leads to certain predictable outcomes through organizatiol1, II The Psychobiology of Attachment 511

"goal-correct," depending on the behavior of the care­ arouses anxiety and actual loss gives rise to sorrow; whilst giver. My dog Rubin shows this very clearly. If I call him each of these situations is likely to arouse anger. The unchal­ from a distance, he begins running toward me; if! move from lenged maintenance of a bond is experienced as a source of my original location, he will adjust the vector of his approach joy" (p. 130). The evolutionary purpose of emotion in rela­ . to most efficiently arrive next to me. His goal-proximity to tion to attachment is that humans actively work to maintain a his caretaker-

able to empirically measure the parent's capacity to mentally however, is not attachment, although it does appear to hl' ;1 represent the child as a whole, real, and meaningful human behavioral system that is activated under certain circuIlJ being, and has shown its causative impact on the child's se­ stances. Children, for example, engage in more playful aCli\ cure or insecure attachment behavior. ity with their peers when their attachment to a priman caretaker is secure (Bowlby, 1969). Harlow (1969) showcd, Attachment and Exploration moreover, that monkeys reared with their mother but withlllli peers were subsequently impaired in their adult social, mal There is an exploratory behavioral system that is biologically ing, and parenting behavior. Bowlby understood affiliation a, based and complements attachment. When a child feels s~­ a broader concept than attachment, the former, covering aII cure, what Ainsworth (1963) called a secure base from which "friendliness and goodwill, of the desire to do things in thl' to explore, the attachment system is not activated and the company of others," but without the object specificity (.1 child can go forth and gather new information about how the attachment (p. 229). world works. This dynamic equilibrium is mutually inhibit­ ing; when there is a threat or a potential hazard, exploratory Attachment and Caregiving behavior will diminish or cease altogether as the attachment system activates. Empirical research has demonstrated that There also appears to be a biologically based caregiving sy' the infant's belief that the mother will be available when tem that protects the child and works in concert with attach needed enhances exploration (Sorce & Emde, 1981). In sev­ ment. When caregiving is activated by the parent, the child', eral studies (Ainsworth & Wittig, 1969; Carr, Dabbs, & Carr, attachment seeking is unnecessary and deactivated. When 111l' 1975), the mother's physical or psychological presence was child is an infant, the chief caregiving behavior is retrieval experimentally manipulated, producing data that strongly Exploratory behavior is also enhanced if caregiving is aell supported the theoretical association between maternal avail­ vated. Cassidy (1999) noted that a child exploring a park will ability and infant exploration, what Ainsworth referred to cover much more territory if the mother actively follow, as an "attachment-exploration balance" (Ainsworth, Bell, & Caregiving is activated by a variety of internal (hormonn, Stayton, 1971). beliefs, fatigue states, emotions, and attachment style of I Ill" mother) and external clues (familiarity of the environmenl, Attachment and Fear presence of danger, and behavior of the infant). Cassidy pI'! I posed that soothing also facilitates caregiving by ensuring till' Although fear is evoked when there is a real threat, there monitoring of potential or real dangers to the child; for ill appears to be a fear behavioral system that initiates attach­ stance, continued holding of the child after his or her distil'" ment seeking when danger is likely. Bowlby (1973) called the subsides may reveal a splinter in the child's finger. stimuli that trigger this system "natural clues to danger." He included such things as high places, darkness, loud noises, Attachment Behavior and the Attachment Bond aloneness, and sudden looming movements. These clues are not inherently dangerous, but provoke attachment behaviors Attachment behavior is not the same as a bond to another pl'l that, in tum, diminish fear if the caretaker is accessible. These son. Empirical research has substantiated that attachnil' II I clues are distinguishable from other objects that provoke fear behavior exists throughout the human life cycle, and early al that are inherently dangerous to infants, such as poisonous tachment experiences predict to a certain degree later attach 1 and predatory creatures (some spiders, snakes, and large ment expectancies and behaviors (Cassidy & Shaver, 199 ) I mammals). The infant's capacity to experience fear in all of Crowell, Fraley, and Shaver (1999) caution, however, thai these situations is an evolutionarily adaptive trait that con­ measurement of attachment across methods (interview veNI, tributes to its survival and eventual reproductive success. self-report) and domains (parent versus romantic partnl" I produces different correlations, averaging r = .39 and r = .3 I , Attachment and Socialization respectively. Ainsworth (1989) described six criteria for all attachment bond: (a) it is persistent; (b) it involves a speeill' Individuals in the company of others are much less likely to person; (c) it is emotionally significant; (d) proximity wilh be killed by predators (Eisenberg, 1966). In addition, there the person is wished for and sought; (e) distress is felt wlll'lI are other important survival advantages to spending time there is involuntary separation; and (f) the relationship brill!" with people, including food gathering, building shelters, security and comfort. Although activation of the attachmcllt learning, and finding a mate. Affiliative or social behavior, behavioral system is situational, and often initiated by all Pathologies of Attachment 513

Jtemal or external threat, an attachment bond exists over Disorganized attachment in infants has been associated time and can be inferred, but not observed. with severe maternal psychosocial problems, including de­ The importance of this distinction is clear if we assume, pression, history of violence or abuse, inpatient psychiatric for a moment, that there is no distinction. Then we would history, and the mother's own abuse of the infant (Lyons­ wrongfully conclude that a child who fearfully clings to his Ruth, Repacholi, McLeod, & Silva, 1991). By 6 years of age, or her mother is securely bonded to her; a child who has lost disorganized attachment often becomes controlling behavior his or her mother and temporarily seeks comfort from a toward the mother, either caregiving or coercive, and this role stranger is bonded to the stranger; and a child who confi­ reversal is often accompanied by childhood and a dently plays with another child in the presence of his or her disparity between verbal and performance IQ (Lyons-Ruth mother, but does not seek her comfort, is not bonded to her. et al., 1991). It appears strongly related to diagnoses of oppo­ The difference is even more apparent when we turn to abnor­ sitional defiant disorder, conduct disorder, and other exter­ mal or pathological attachments. Paradoxically, the absence nalizing problems in childhood (Lyons-Ruth, 1996). Most of a secure bond activates the attachment behavioral system interestingly, disorganized attachment in infants is reliably in unusual, strange, and sometimes dangerous ways. predicted by the mother's lack of resolution of a previous loss or trauma, measured before the birth of the child (van > Ijzendoorn, 1995). For example, a mother who suffered from PATHOLOGIES OF ATTACHMENT posttraumatic stress disorder due to chronic by her ex-boyfriend is at great risk to raise a child who evi­ Two pathological forms of attachment were first discovered dences disorganized attachment within the first few years of by Ainsworth et al. (1978) in the Strange Situation. The life. Main and Hesse (1990) theorized that this intergenera­ avoidant infants (Type A) were exploratory without paying tional transmission of attachment is related to frightening or attention to mother's location, were minimally distressed frightened parental behavior and may be a product of dissoci­ when she left, and largely ignored her when she returned. ation in the parent. Other psychiatric disorders in parents may 'le secure infants (Type B) competently expressed their also be strongly related to risk of disorganized attachment in deeds and accepted maternal care. The ambivalent/resistant infants (Lyons-Ruth, 1996). infants (Type C) had difficulty separating from their mother Models of adult attachment have been developed by Hazan and exploring or playing in the environment, were also very and Shaver (1987), Main, Kaplan, and Cassidy (1985), and distressed when mother left, but could not "settle in" with Bartholomew (1990, 1994, 1997). The latter's work is most her when she returned. Separation distress did not distin­ promising because it is consistent with earlier infant and child­ guish secure from insecure (avoidant or ambivalent) infants; hood theories of attachment, further delineates avoidant strate­ all three groups evidenced such distress to one degree or an­ gies, and incorporates an object- and self-representational other. The reunion behaviors most clearly demarcated the perspective. It contains three pathological types: groups. These three forms of attachment behavior worked well in 1. Preoccupied individuals have a negative perception of research for many years and were used to successfully test self and a positive perception of others. Attachment be­ the hypothesis that attachment types were generally stable haviors and internal regulation of arousal have been con­ from childhood to adulthood (Goldberg, Muir, & Kerr, 1995). ditioned by inconsistent parenting in childhood. They There were always some subjects, however, who could not be themselves for a lack of love and appear to be very classified, especially in research with clinical samples. Main dependent in their attempts to gain others' approval and and Solomon (1986, 1990) subsequently developed criteria for acceptance. a fourth type: disorganized/disoriented attachment (Type D). 2. Fearful individuals have anegative perception ofboth self These infants had no organized strategy for managing arousal and others and avoid close contact, usually due to a history during the activation of their attachment behavioral system of rejecting or unresponsive parents. Others are viewed as while seeking comfort and security. Behaviors included uncaring due to the fearful individual's unlovable nature. apprehensiveness, helplessness, , unexpected al­ Although they desire acceptance, they fear rejection. terations in approach or avoidance toward mother, prolonged 3. Dismissing individuals have a positive perception of self ""reezing," and psychomotor slowing. Cortisol levels re­ and a negative perception of others. They have managed .tined significantly elevated and higher than other, more rejecting or unresponsive parents by distancing and be­ organized attachment types, whether secure or insecure coming self-reliant, inoculating themselves against the (Spangler & Grossman, 1993). devaluation they have learned to expect. 514 Pathologies of Attachment, Violence, and Criminality

Bartholomew has used a circumplex model of interper­ and men (N = 5,056). The strongest predictor of victimit;1 sonal behavior to validate her attachment prototypes along tion by an intimate partner for both men and women was phy~ dimensions of control and affiliation (Bartholomew & ical assault as a child by a caretaker. Other predictors inclmk,l Horowitz, 1991). There has also been substantial research unmarried (but cohabitating) status, African American ra,','. on both concurrent and predictive validity of her model by the partner, jealousy or possessiveness, and (Bartholomew, 1997). The fourth type of adult attachment educational or racial disparities between the partners. The all pathology, which is not included in her model, is the disorga­ thors wrote, "Violence perpetrated against women by mail" nized individual. Although research with such adults is lim­ partners is part of a systematic pattern of dominance and COil ited, it appears closely associated with severely disturbed trol, or what some researchers have called 'patriarchal terror clinical and forensic samples of individuals (Fonagy, 1999b). ism'" (p. 34). Despite the merit of these empirical finding~, These four adult attachment pathologies-preoccupied, attachment theory, even the word attachment, was never used fearful, dismissing, and disorganized-are becoming keys to throughout this study. Instead, the authors chose to interprel unlocking the raison d'etre for violent attachments. their findings in a narrower feminist sense, which begs 'Ill" question: IfAmerican society is suffused with "patriarchalll'l' rorism," why is it that most men do not assault their intimal,' ATTACHMENT AND VIOLENCE partner? I think the psychosocially deeper and more comprehcn· Although there have been many models proposed for classi­ sive answer to this question is that most men form secure al· fying violence, converging lines of theory and empirical tachments. The ones who are not capable of forming sud] research have divided violence into two modes: predatory attachments are at greatest risk for intimate partner violenc,'. (instrumental, premeditated, attack) and affective (impulsive, Research continues to accumulate that empirically support~ reactive, defensive). Labels have varied, but the underlying the general hypothesis that insecure attachments are sig· characteristics have been similarly described and, in some nificantly associated with, and in some cases intergenera· cases, measured by different research groups (Barratt, tionally predict (Adamson, 1998), intimate partner violenCl' Stanford, Felthous, & Kent, 1997; Cornell et aI., 1996; (Holtzworth-Munroe, Stuart, & Hutchison, 1997). Most stud· Meloy, 1988, 1997; Raine et aI., 1998). Predatory violence is ies have focused on the male's attachment pathology (Dutton. planned, purposeful, and emotionless, with absent autonomic 1995a), but some recent studies indicate that securely ai­ arousal. Affective violence is a reaction to a threat, accompa­ tached individuals are more likely to form sexual pair bonds. nied by anger and fear, and involves high levels of autonomic devoid of violence, with each other. The sexual intimates 01 (sympathetic) arousal. The evolutionary basis of predatory insecurely attached individuals, on the other hand, are also violence is hunting; affective violence is rooted in a protec­ likely to have a history of insecure attachment, thus embark· tive and defensive response to an imminent threat. Both serve ing on a pathogenic dance that is at greater risk of violenCl' reproductive success and genetic viability. In other words, (Babcock, Jacobson, Gottman, & Yerington, 2000; Irwin. our ancestors thousands of years ago were adept at both 1999). Birds of a feather appear to flock together. Tjaden and predatory and affective violence (more so than their neigh­ Thoennes (2000) also found that intimate partner violenc,' bors who did not survive to reproduce and raise their young). was highest among homosexual males and lowest among Research on attachment and violence during the past hom?sexual females, an important point of reference thai decade has largely focused on intimate partner, or domestic, underscores the biological propensity of men to be more vio· violence. There has been limited research on attachment lent than women regardless of their target. Such finding_~ and violent criminality. The discoveries are new and very also contradict the argument that any psychopathology, in promising. cluding attachment, in the (usually) female victim of domcs· tic violence is irrelevant to understanding the violence, amI i~ nothing more than "blaming the victim," Attachment pathol· Intimate Partner Violence ogy in the male perpetrator of also appear~ In July 2000, the U.S. Department of Justice published find­ to be a more stable correlate than anyone specific person· ings from the National Survey con­ ality disorder (Tweed & Dutton, 1998; Waltz, Babcock. cerning the extent, nature, and consequences of intimate Jacobson, & Gottman, 2000). partner violence (Tjaden & Thoennes, 2000). Risk factors as­ Several researchers and their colleagues dominate the work sociated with intimate partner violence were discerned using in this area, and despite different foci, each has shaped scien· logistic regression on separate samples ofwomen (N = 4,896) tific thinking about intimate partner violence and attachment Attachment and Violence SIS

,(esner conducted two studies (Kesner, Julian, & McKenry, Dutton's work is unique among these researchers because 1997; Kesner & McKenry, 1998) drawing on Bowlby's (1984) he has used both attachment and object-relations theories to notion that intimate violence may be a product of maladaptive propose hypotheses concerning batterers, developed instru­ anger to keep the partner from separating. In his first study of ments when needed to measure his hypotheses, and then violent male spouses (Kesner et aI., 1997), he found that at­ tested them on various samples of batterers in treatment pro­ tachment variables served as unique predictors of male inti­ grams and in prisons around Vancouver, Canada. He has mate violence: (a) the male's recollection of his relationship shown that the etiology of battering is not simply with his mother (a perceived deficiency in love and caring), of the batterer when he was a young boy. Instead, shaming of and (b) his perceived relationship support from his spouse. In a child by a caretaker, witnessing violence directed toward Ris second study (Kesner & McKenry, 1998) of heterosexual the self or mother, and insecure attachment (fearful or preoc­ couples, he found that attachment factors of both the male and cupied) form a triad that predicts battering as a adult (Dutton, female partners were unique predictors of male violence; 1999). All contribute to the formation of a borderline person­ specifically, the males were more fearfully attached and less ality organization (Kernberg, 1984) that stimulates an "inti­ secure, and their female partner had more of a dismissing at­ macy anger" when in a relationship. This largely impulsive tachment and were less secure. He wrote, "The anger that acts group of batterers are prone to experience rejection anxiety, to communicate fear of separation in the secure relationship which is quickly converted into rage when loss intensifies into violent behavior by the fearful individual in a is imminent and is then violently expressed to diminish ten­ gross escalation of this anger" (p. 429). Concurrent life stres­ sion (Dutton, 1994, 1995a, 1995b, 1995c). His empirical SOl'S failed to predict violence. Kesner's work empirically sup­ findings using various predictive statistical models have val­ ports the importance of a negative maternal transference in idated his etiological theories (Dutton, 1998). male batterers and the contribution to the violence of the vic­ The continuation of violence by the batterer, largely tim's insecure attachment, and also emphasizes the affective, through the inability of the victim to leave the relationship, rather than predatory, mode of violence in the couples whom led Dutton and his colleagues to apply the theory of traumatic 'le has studied. bonding to such phenomena. Drawing on the social psychol­ Downey and her colleagues have made similarly impor­ ogy hypothesis of a traumatic bond that forms between hos­ tant contributions. In two recent papers (Downey & Feldman, tage and hostage taker, the so-called , 1996; Downey, Feldman, & Ayduk, 2000), she has studied he and his student (Painter & Dutton, 1985) posited that "rejection sensitivity" as predictive of male violence toward reinforcement mechanisms interact with extreme power dif­ romantic partners. Defining her construct as "the disposition .ferentials to constitute traumatic bonding. For example, both to anxiously expect, readily perceive, and intensely react to intermittent (the onset of violence) and nega­ rejection by significant others" (p. 45), Downey has shown tive reinforcement (the termination of violence) can further that it is a vulnerability factor for two maladaptive styles of cement the relationship. In a subsequent study (Dutton & with intimacy: fearful avoidance of such an intimate, Painter, 1993), their hypothesis was empirically tested and and a preoccupied search for an unconditionally supportive demonstrated that 55% of the variance in their attachment intimate. The latter style predicted relationship violence, usu­ measure of female victims six months after separation was ally affective, in a large nonclinical sample of college males accounted for by the traumatic bonding variables. They em­ (Downey et al., 2000). Rejection sensitivity (what Gabbard, phasizect the prolonged effects of abuse and dismissed 1989, termed "hypervigilant narcissism") may be an impor­ other, more static, theories of female victimization, such as tant personality trait that is a product of insecure attachment. masochism. In a related study, Oderberg (1995) found in a sample of Another area in which Dutton has made important contri­ college undergraduates that witnessing parental violence as a butions is the development of a typology for batterers. child was positively associated with insecure attachment as Drawing on the earlier work of Hamberger and Hastings a young adult. (1986) and Holtzworth-Munroe and Stuart (1994), which Dutton (1995a, 1998) and his colleagues have made enor­ identified three subgroups of batterers-the generally violentJ mous contributions to our understanding of domestic vio­ antisocial, dysphoric/borderline, and family only/overcon­ lence. Their discoveries have emerged along three lines of trolled-his work has refined our understanding of the first research: the etiology of intimate violence; the perpetuation two groups. Dutton (1998) has referred to the dysphoric/ Jf intimate violence (in particular, the reasons why a victim borderline group of batterers as the "abusive personality." stays in the abusive relationship); and typologies of inti­ Characteristics of this subgroup include a fearful attachment mately violent men. style, borderline personality organization, chronic anger, and 516 Pathologies ofAttachment, Violence, and Criminality

impulsivity. They are withdrawn, asocial, moody, hyper­ made enormous contributions. Most recently, they have vali­ sensitive to slights, volatile, reactive, and oscillate rapidly dated the three types of batterers, which they call generally between indifference and rage. The modal Diagnostic and violent, pathological, and family only (Waltz et aI., 2000), the Statistical Manual ofMental Disorders (DSM-IV; American typology originally proposed by Holtzworth-Munroe and Psychiatric Association, 1994) diagnosis appears to be bor­ Stuart (1994). Although the first two groups did not differ on derline personality disorder. Saunders (1992) referred to personality disorder-both elevated on the borderline and them as the Type 3, Emotionally Volatile group. I earlier de­ antisocial scales of the MCMI-II-the types differed as pre­ scribed the etiology of this group, which appear to make up dicted on the frequency of their emotionally abusive behav­ 25% of batterers in treatment. Their violence is affective ior, their history of witnessing parental violence, attachmenl rather than predatory. pathologies, jealousy, and presence of chemical abuse. (The The generally violent/antisocial group is the most psycho­ authors also noted the high overlap between these scales on pathic of the batterers. Although psychopathy has yet to be the MCMI-II and their correlation of 0.64. The MCMI-lli. directly measured in a study of spousal batterers, this however, shares only 18% of items between these two scales. group tends to elevate on the Antisocial, Narcissistic, and suggesting further research may benefit by using the lattcr Aggressive-Sadistic scales of the Millon Clinical Multiaxial measure.) The generally violent men were dismissing and Inventory II (MCMI-II; Millon, 1981). They are also more se­ avoidant, whereas the pathological men were preoccupicd verely physically violent, are narcissistically entitled, and ma­ and ambivalent. The "family-only" batterers showed a "com­ nipulative. In contrast to the dysphoricfborderline group, they pulsive care-seeking" attachment style. exhibit low levels of depression and anger. Their abuse of Their most compelling work, moreover, has been in the drugs and alcohol is frequent, and they are more violent out­ area of physiology, emotional regulation, and marital vio­ side the home than other groups. They also represent approxi­ lence. In an earlier study which received considerable atten­ mately 25% ofbatterers (Holtzworth-Munroe & Stuart, 1994). tion, Gottman et al. (1995) recruited a sample of couples witll They have moderate marital satisfaction, but fail to achieve maritally violent histories and measured their physiology in any sense of relationship reciprocity or whole-object related­ the laboratory while the pair engaged in conversations about ness to their partner (in analytic terms, she remains a selfobject highly conflicted issues in their relationship. They identificd or part-object). Most interestingly, their violence is predatory two groups: Type I batterers «20% of their sample) demon­ (instrumental): planned, purposeful, and emotionless. Their strated heart rate decreases during intimate conflict; Type II attachment pathology, however, appears to be preoccupied, batterers demonstrated heart rate increases. They referred \0 but not fearful. Although at first blush, this appears contradic­ the former as "vagal reactors," in reference to the vagus nerve. tory, and one would expect a dismissive attachment style, which, when activated, reduces autonomic arousal. This group Tweed and Dutton (1998) note that such an attachment pattern was also more likely to be generally violent and antisocial. would not motivate a high investment in a troubled relation­ and had scale elevations on the MCMI-II for Antisocial and ship nor an ongoing effort to use violence to control a partner. Aggressive-Sadistic behavior. Although they did not measure What appears to be present is, instead, a preoccupation with psychopathy, nor discuss it in this study, their findings were attaining a relationship in which the psychopathic batterer , highly consistent with the autonomic hyporeactivity that hw.. dominates and controls his partner. Measurement of psy­ been documented in psychopaths for decades, particularly in chopathy in a future study of batterers will clarify this issue. aver~ve circumstances (Meloy, 1988). Low resting heart rate The severe psychopath-individuals that score ~30 on the is also one of the most replicated physiological finding ... Psychopathy Checklist-Revised (PCL-R; Hare, 1991 )-may among adolescent delinquents, and is a measurable aspect 01 be relatively infrequent among spousal batterers because all of the chronic cortical underarousal seen in habitual criminab his relationships are generally fleeting. Without any capacity (Raine, 1993). Most interestingly, during long-term follow­ to attach or bond, he moves on to another sexual object, en­ up, none of the women married to these men had left them. gaging in a pattern of search polygyny (Meloy, 1992, 20oob) Babcock et al. (2000) further advanced their work in a that precludes any sustained effort to control a noncompliant study that integrated attachment pathology as an "index 01 mate. The batterers in Dutton's instrumental group are likely emotional regulation" (p. 392) and the function of violence ill to be significantly more psychopathic than his impulsive marital relationships. They viewed insecure attachment group, but may not be severe psychopaths. along a dimension of deactivation (dismissing) versus hyper

Gottman and Jacobson (Gottman et aI., 1995) have di­ activation (preoccupied) attentional systems that serve III rected their research efforts toward understanding and treat­ regulate affect during stress. They showed, first of all, thai ing marital violence for a number of years, and have likewise violence was most likely to occur in an insecure attachment Attachment and Violence 517

according to the Adult Attachment Interview (AAI; Main & violence. and (cl the batterer's psychopathology or person­ Goldwyn, 1984). ality disorder. This dimensional approach yielded three The AAI is a semistructured interview about childhood at­ types previously mentioned: the family-only batterers, the tachment experiences that has been refined and expanded dysphoric-borderline batterers, and the generally violent­ over the past 15 years (Main & Goldwyn, 1994) but has yet antisocial batterers. They further proposed a developmental to be published. Extensive training is required to use the in­ course for the three types, which included both historical (e.g., strument. The nalTative of the interview is transcribed and genetic and prenatal factors, violence in the family of origin) scored according to three criteria: (a) the coder's assessment and proximal correlates (e.g., attachment style. dependency, of the subject's childhood experiences; (b) the language used hostility toward women, social skills). They built predictions by the subject during the interview; and (c) the individual's for their model based on their proposed types and risk factors. ability to give an integrated and credible account of his or her One hundred and two men were recruited from the com­ experiences as a child. Two sets of scales. Parental Behavior munity to test their model (Holtzworth-Mumoe, Meehan, and State of Mind, result in the assignment of the subject to Hen-on, Rehman. & Stuart. in press), selected on the basis of one of three major classifications;'secure, dismissing, or pre­ a wide range of violence toward their spouse. Two nonviolent occupied. Individuals may also be classified as "unresolved" comparison samples were also recruited (distressed and not (what I have termed "disorganized" in this chapter) and distressed). When they completed their analyses of both their "cannot classify." The AAI is the gold standard for assess­ dependent and independent variables, their three predicted ment of attachment (Crowell et aI., 1999). subgroups emerged, along with a fourth group. The sub­ Babcock et al. (2000) further demonstrated two functional groups generally differed along their three descriptive dimen­ patterns of intimate partner violence that were tied to both at­ sions and their proposed developmental risk factors. One tachment pathology in the male and triggering behavior in the independent research group has also found three subgroups female. The dismissing batterers were most likely to use vio­ that closely fit the Holtzworth-Munroe and Stuart typology lence instrumentally (a predatory mode) to control the behav­ (Hamberger, Lohr, Bonge, & Tolin, 1996). )r of their spouse. This subgroup also had the most extensive The fourth unpredicted cluster was labeled the "low-level antisocial traits and were likely the most psychopathic of antisocial" group. These men appear to fall within an inter­ their types; although, once again, psychopathy was not di­ mediate range on a number of variables between the family­ rectly measured. They were also most likely to be violent only and the generally violent-antisocial groups. Holtzworth­ .when the spouse became defensive during an argument. The Munroe (2000) argued that the family-only group in their preoccupied batters were most likely 10 use violence expres­ community sample probably represents the young, newlywed sively (an affective mode) to regulate affect in their interac­ couples where low levels of aggression are almost normative tion with their spouse. They were most likely to be violent (O'Leary et aI., 1989). These men, however, did not differ on when she attempted to withdraw during an argument. Both measures of attachment or psychopathology from the non­ attachment pathologies tended to be more domineering than violent but distressed comparison group. It may be that their the secure husbands. The researchers hypothesized that the violence is socioculturally based, rather than rooted in any dismissing batterers used a controlling and distancing style of psychological abnormalities. interaction to get what they wanted, whereas the preoccupied Holtzworth-Munroe (2000) also proposed a condensing of batterers were remarkable for their inability to use distancing her three descriptive dimensions into two: an antisocial con­ and disengage from contlict: When their spouse withdrew, tinuum (measurement of psychopathy would work best here) they perceived imminent abandonment, and their anger esca­ and a borderline continuum (perhaps a measure of borderline lated into dysregulated fury and violence. personality organization) to account for the severity of vio­ Holtzworth-Munroe and her colleagues (Holtzworth­ lence and the degree of attachment pathology, respectively. Munroe & Stuart, 1994) have been the undisputed leaders in She also emphasized the dynamic, rather than static, nature of the formulation of a reliable and valid overall batterer typol­ spousal violence, and endorsed, at least in theory. the appli­ ogy. Their theory is an intrapersonal one, focusing on the pre­ cation of predatory versus affective modes of violence in de­ disposing and precipitating factors within the batterer that marcating the behavior of the generally violent-antisocial contribute to his violence. from the borderline-dysphoric batterer. Following a review of the existing research, Holtzworth­ Fonagy (1999c) and his colleagues have charted exciting .unroe and Stuart (1994) theorized that batterers could be new territory in our understanding of violent attachments and categorized along three descriptive dimensions: (a) the sever­ the psychology of the self. Approaching attachment theory ity and frequency of marital violence, (b) the generality of from the perspective of psychoanalysis, their theoretical and 518 Pathologies of Attachment, Violence, and Criminality empirical work has focused on the "mentalizing" function persecutory mental representations of the parent. These hate­ and the reflective self, the capacity of an individual to recog­ ful introjects then become a source of emotional volatility nize subjective states and the subjectivity, or inner states, of and turmoil in subsequent attachments throughout their life, others. This is the experience of oneself and others as having as they continuously project them onto their intimates as a wishes, feelings, thoughts, desires, beliefs, and expecta­ means of evacuating and controlling them. These individuals tions-in short, an "intentionality" that is motivated by an are clinically observed as impulsive, emotionally unstable, internal psychology. Fonagy's work is a deepening and and prone to violence toward self and others; the diagnosis broadening of Bowlby's (1961) early theory on internal is often borderline personality disorder. Fonagy (1999a) has working models that, in essence, posits that the absence of a emphasized the importance of trauma and disorganized at­ theory of mind (a theory of self) is a fundamental cause of in­ tachment in the genesis of such a personality disorder. security of attachment and, in certain cases, intimate vio­ Although Fonagy's (1999b) theory of male violence to­ lence. As Fonagy (1999a) wrote: ward female intimates has yet to be empirically tested, it is an elegant conceptual extension of his other work. The frequenl The child finds himself in the caregiver's mind as an intentional childhood abuse and shaming of the male (Dutton, 1998) being motivated by mental states, beliefs, and desires. This rep­ when he is little is managed by refusing to acknowledge his resentation is internalized as the core of the psychological self. caretaker's thoughts about him and his wish to harm him. The Thus, the realization of subjectivity might be more accurately lack of safety with his caretaker continuously triggers his at­ stated: "My caregiver thinks of me as thinking, therefore I exist tachment behavioral system, which is responded to with ne­ as a thinker." (pp. 12-13) glect or abuse. The nascent mentalizing stance in the child is disavowed, and under the combined pressure of needing Fonagy has empirically tested his theory in a number of comfort and escaping abuse from the same person, he dis­ ways. Fonagy, Steele, Moran, Steele, and Higgitt (1991) rupts his capacity to represent the mental states of himsel r found that the capacity for caregivers to reflect on mental and others. People become objects or bodies, rather than states in themselves and others when describing their own whole, real, and meaningful individuals. A failure of mental­ childhood predicted their children's security of attachment. izing also causes a moral disengagement for four reasons: Reflective self ratings were reliable (r > .80) and provided a (a) Individuals without a well-established sense of them­ good prenatal prediction of their child's behavior in the selves have no sense of personal agency; (b) they cannot an­ Strange Situation experiment. Highly reflective parents were ticipate the psychological consequences of their actions on three or four times more likely to have secure children than others; (c) others are treated as objects; and (d) rationaliza­ low-reflective parents. In another study, Fonagy, Steele, tion and minimization (plausible but false fluidities of think­ Steele, Higgitt, and Target (1994) factored in social depriva­ ing) are more prominent (Fonagy, 1999a). Violence toward tion of the mother (single parent, living in overcrowded con­ the intimate results from a maladaptive escalation of anger 10 ditions, unemployed father, low income, etc.) to see if it keep the partner from neglecting or abandoning, as well as an would affect the impact of the reflective self on secure at­ overwhelming need to control the other so that intolerable tachment. It did not. The deprived mothers with a capacity to self states can be projected (or projectively identified) inlo fully represent and reflect on themselves and others all had her. One 26-year-old male who killed his estranged wife told securely attached infants, and virtually all of the deprived me,· "I didn't know what to feel. I was in a rage and also mothers who could not reflect had insecure infants. In a third numb. I needed to shoot the pain ... I killed the woman I study, children securely attached in infancy were more likely loved." A 38-year-old male who sexually assaulted and killed to cognitively understand the affective states in others at a 12-year-old girl told me that his father would always say 10 5 years of age when compared to insecurely attached children him, "The best part of you got spilled on your mother's bed­ (Fonagy, Redfern, & Charman, 1997). sheets." These devaluing and hateful self and other represen­ The psychoanalytic basis of Fonagy's work is that chil­ tations constantly oscillate between two insecurely attached dren find themselves in the mind of their caretaker, and the partners who attempt to manage, often unsuccessfully, a psychobiological vehicle for this discovery is a loving and volatile interpersonal space. secure attachment. When this is not available, when, for in­ stance, the parent is constantly angry at or even hates the Violence and Criminality children, the children's contemplation ofthe parent's feelings toward them is intolerable. Therefore, they do not think of The research on attachment and other forms of criminal vio­ themselves; rather, they internalize the hateful, perhaps lence is much more limited than the intimate partnl'/ New Avenues of Forensic Research and Application 519 research. Antisocial personality disorder (DSM-IV), or con­ for the prominence of dismissing attachment pathology duct disorder in adolescence, appears to be associated with among criminals, likely related to the construct of psychopa­ dismissing or disorganized attachment pathology. Allen, thy, that may instead have its roots in a temperament­ Hauser, and Borman-Spurrell (1996) found that both patholo­ environmental misfit that leads to avoidant strategies by both gies predicted criminality in a sample of adolescents 10 years mother and child (Shaw & Bell, 1993). It also does not leave after their attachment was measured. This prospective study room for the possibility that a constitutional defect in the capac­ compared adolescents who were psychiatric inpatients with a ity to bond may exist in the child, and despite heroic efforts by group of high school students. Derogation of attachment (dis­ the securely attached parents to stimulate a bond, nothing works. missing) and lack of resolution of trauma (disorganized) were In the domain of attachment and violent criminality we are the best predictors, and did so when psychiatric hospitaliza­ left with intriguing theory, very little research, and some ten­ tion was controlled as a confounding variable. Likewise, tative findings: (a) Insecure attachment is a risk factor for vi­ Rosenstein and Horowitz (1996) found in a small sample of olent criminality; (b) secure attachment may be a protective conduct disordered adolescents (N = 7) that most were clas­ factor against violent criminality, particularly when the child sified as dismissing and none were classified as unresolved is raised in a deprived economic or social environment (disorganized). Fonagy et al. (1996) found that most paranoid (Klevens & Roca, 1999; Marcus & Gray, 1998); (c) the re­ and antisocial personality disordered adults in a nonrandom flective function may be an important mediating variable for sample were classified as disorganized, with clearly unre­ understanding affective violence in particular; and (d) dis­ solved trauma, when a four-category system of attachment missing and disorganized pathologies of attachment may classification was used. correlate with constitutional and traumagenic pathways to The most compelling theory and supportive empirical violent criminality, respectively. findings concerning pathological attachment as a risk factor for violence and criminality have been advanced by Fonagy NEW AVENUES OF FORENSIC RESEARCH (l999a; Fonagy et aI., 1997). In a small study comparing AND APPLICATION rison inmates, psychiatric patients, and controls (Levinson & Fonagy, cited in Fonagy, 1999a) using the AAI, the vast ma­ If we conceptualize attachment pathologies as lying on a con­ jority of the prisoners were classified as either dismissing tinuum between hyperarousal (the preoccupied type) and (36%) or preoccupied (45%). Although 82% of psychiatric hypoarousal (the dismissive type), and see this autonomic ac­ patients were disorganized, only a minority of prisoners were tivation or deactivation (whether acquired or inherited) as disorganized (36%). However, most of the prisoners had being related to both attention and emotion (Babcock et aI., been physically or sexually abused, and was also 2000), two intriguing new areas of forensic research and ap­ prevalent. Anger was highest among the prisoners, and their plication become apparent: understanding the nature and dy­ reflective function was lowest among the three groups. Re­ namics of stalking and psychopathy. flective function among the violent prisoners, as measured by index offense, was significantly lower than among the non­ Stalking: The Preoccupied Crime violent prisoners. Fonagy (1999a) argued that these findings, although only Stalking i.s an old behavior but a new crime (Meloy, 1999). a pilot study, support the theory that weak: bonding and the First codified in California in 1990, stalking laws now exist dismissal of objects is a risk factor for violent criminality, a throughout the United States, Canada, Great Britain, relatively consistent finding over the past 50 years (Bowlby, Australia, and New Zealand. Typically defined as "the 1958; Meloy, 1992); more important, "criminal behavior may willful, malicious, and repeated following and harassing of be seen as a socially maladaptive form of resolving trauma another that threatens his or her safety" (Meloy & Gothard, and abuse. Violent acts are committed in place ofexperienced 1995, p. 259), stalking victimization affects a large propor­ anger concerning neglect, rejection, and maltreatment. Com­ tion of both the adult and adolescent populations. mitting antisocial acts is facilitated by a nonreflective stance Stalking laws typically have three elements: a pattern of toward the victim" (Fonagy, 1999a, p. 64). unwanted pursuit, a credible threat, and the induction of rea­ This thinking is in accord with other work concerning dis­ sonable fear in the victim. In California, the current stalking -ganized, traumagenic attachment in infants and the emer­ law reads as follows (Penal Code Section 646.9): 6ence of coercive and aggressive behavior in later childhood (Lyons-Ruth, 1996). It usefully extends it into the object rep­ Any person who willfully, maliciously, and repeatedly follows or resentations of the violent criminal. But it does not account- harasses another person and who makes a credible threat with the 520 Pathologies of Attachment, Violence, and Criminality

,tent to place that person in reasonable fear for his or her safety, observe and document two empirical findings that strongly or the safety ofhis or her immediate fami!y, is guilty of the crime suggested attachment pathology in stalking cases. In a small of stalking. sample of incarcerated stalkers in a Missouri prison, the ma­ jority had lost a primary caretaker in childhood and had had a Although the law is new, Mullen, PatM, and Purcell major loss, usually a personal relationship, in the six months (2000) note that the first attempt to prosecute stalking behav­ preceding the onset of stalking. The researchers proposed ior was brought before the English court in Dennis v. Lane in that these two findings respectively predisposed and precipi­ 1704. Dr. Lane, a physician, engaged in an unwanted pursuit tated the criminal behavior. Although Meloy (1996, 1999) of Miss Dennis. During the course of his stalking, he as­ focused on a preoccupied attachment style among stalkers in saulted two parties, a man accompanying Miss Dennis on a subsequent writings, Kienlen (1998) reported case examples trip and a barrister who had escorted her to London. He was and theory consistent with a variety of attachment patholo­ eventually ordered to pay 400 pounds as security to ensure gies among stalkers. the peace. The eventual outcome of the case is unknown. The preoccupied, hyperaroused nature of stalkers has been At the end of the twentieth century in the United States, it supported by several negative findings. Most individuals who appears that 8% of adult women and 2% of adult men will be stalk are not antisocial personality disordered (Meloy et aI., stalked some time in their life (Tjaden & Thoennes, 1997); 2000), and the psychopathic stalker is a rare event (Meloy, approximately 25% of college-age students will be victim­ 1999). These empirical findings are consistent with the hy­ ized by stalking behaviors, although most incidents do not pothesis that chronically emotionally detached individuals arise to the level of criminal activity (McCann, 2001). who evidence a "dismissing" attachment would not waste Stalking and violence are closely allied. Rates of violence their time stalking someone; they do not fonn an enduring, are disturbingly high, usually directed toward the target of the meaningful emotional bond with another. Instead, they ma­ stalking. They range from 25% to 40%, but they typically nipulate, exploit, and then dispose of their objects. It is also exceed 50% when there has been a prior sexual intimacy consistent with findings I described concerning the surpris­ leen the stalker and his or her victim (Meloy, in press). ingly preoccupied attachment pathology among some anti­ '1 ne nature of the stalking violence is also being studied. In social batterers (Tweed & Dutton, 1998); they are probably most cases of "private" stalking in which there has been a pre­ not psychopaths. vious known relationship, the violence is affective: Victims More recent studies continue to verify the hyperaroused, are pushed, shoved, grabbed, choked, slapped, punched, fon­ preoccupied pathology of individuals who stalk prior sexual dled, or their hair is pulled. There is typically no weapon used. intimates. Mechanic, Weaver, and Resick (2000) found in a In cases of "public" stalking, in which the target is a public large sample of battered women that emotional and psycho­ figure such as a celebrity or politician, the violence is preda­ logical abuse in the relationship were strong predictors of tory: Victims are attacked with a weapon, usually a firearm, postrelationship stalking, even when the effects of physical after a lengthy period of obsessive thought, dysphoric rumi­ violence were controlled. They wrote, "It appears that one nation, planning, and approach (Fein & Vossekuil, 1999; function of pursuit-oriented behaviors, of which stalking is a Meloy, 1999; 2001). Mark David Chapman, the assassin of particularly virulent form, is to regulate attachment and prox­ John Lennon, traveled from Hawaii, where he was living, to imity seeking via coercive control strategies" (p. 70). Others New York City and back, only to return again in December have found -that attachment disturbances (preoccupied and 1980 to carry out his killing. He made himself known to the fearful) are related to jealousy, following, surveillance, and doonnen at the Dakota Building as a fan of Lennon over separation behaviors (Dutton, Saunders, Starzomski, & the course of a number of days and actually got Lennon's au­ Bartholomew, 1994; Guerrero, 1998; Holtsworth-Munroe tograph on a compact disk before he murdered him later that et al., 1997). Research among college students is promising. evening by shooting him in the back using a .38 caliber Langhinrichsen-Rohling, Palarea, Cohen, and Rohling (2000) revolver (Jones, 1992). found in a large sample of undergraduates that unwanted Meloy (1989, 1992) first proposed that stalking may be a pursuit behaviors were significantly predicted by an ex- . pathology of attachment in relation to unrequited love and the partner who was anxiously and insecurely attached and evi­ wish to kill. His clinical and theoretical assertion was largely denced higher levels of "possessive" and "dependent" love. . 'ed on the obsessive nature of the cognitions and the labil­ These latter tenns concerning "love styles" have recently

_ J and intensity of the affect apparent in the rejected (either played a role in the research of Cupach and Spitzberg in fantasy or reality) individual. Kienlen, Binningham, (1998), who have made important contributions to our un­ Solberg, O'Regan, and Meloy (1997) were the first to derstanding of "obsessive relational intrusion," a typically Conclusion 521

;1onviolent and less severe form of stalking, among college number of antisocial samples-including children, adoles­ students. cents, and adults-that a Rorschach measure of attachment, Love styles were first proposed by Lee (1976) and mea­ the texture response, was significantly less frequent than in sured by Hendrick and Hendrick (1986). A secondary style, normal samples. As degree of psychopathy increased across calleo "," blends eros (passion and romance) and ludus these subjects, the frequency of the texture response de­ (game playing and exploitation); it is possessive, dependent, creased. Meloy (1988) described this measure, which in­ and addictive. In a large study of undergraduates, Spitzberg volves the perception of a tactile quality to the inkblot, as a (200I) found that both a preoccupied attachment pattern and somatosensory analog for early skin contact with the mother, manic love had small but significant associations with some the primary vehicle of affectional relatedness for the infant of the obsessively intrusive tactics of relational pursuit, and perhaps the corporal genesis of secure attachment. specifically, physical threats and hyperintimacy (unwanted Attachment and psychopathy have been measured among messages, intruding on interactions with others, monitoring, female inmates. Both Strachan (1993) and Taylor (1997) exaggerated affection). found that a dismissive attachment pathology, inferred by the This new area of forensic research-stalking as a pre­ voluntary relinquishment of their children, significantly cor­ occupied crime-is important because of the high rates of related with psychopathy in samples of incarcerated women, violence associated with it, its prevalence in society, its rela­ even when other confounding variables, such as drug abuse tionship to domestic violence, and accumulating evidence and prostitution, were controlled. On the other side of this that it is a chronic behavior for which a hyperaroused, preoc­ coin is the finding by Raine, Brennan, and Mednick (1997) cupied attachment pathology may be central. Empirical that birth complications and maternal abandonment during studies, however, that directly measure the attachment the first year of life were significant predictors of early-onset pathologies of samples of convicted stalkers, both men and violent criminality in their adult male offspring. women, have yet to be done. This new area of forensic research, the psychopath as a dismissive criminal, is important because of his high rates of violence and the chronic, nonviolent destruction he causes 'sychopathy: The Dismissive Criminal through dominance, manipulation, and exploitation of At the other end of a hypothetical attachment continuum is others---despite his apparent conscious disavowal of any the underaroused, affectively avoidant, chronically emotion­ need for affectional relatedness, a striking paradox. Attach­ ally detached individual. This dismissing attachment pathol­ ment theory also can bring to the psychopathy research an ogy, in its most extreme and virulent form, is likely found empirically based, psychobiologically informed construct in the psychopath. A plethora of research during the past that may help complete the unfinished patchwork quilt that 20 years has shown the construct of psychopathy-a constel­ best describes the current findings within the neurobiology of lation of behaviors and traits (Hare, 1991 )-to be both the psychopath (Millon, 1998). For example, I would hy­ reliable and valid, particularly as a predictor of violent crim­ pothesize that a dismissing attachment pathology may be in­ inality (Millon, 1998). Psychopaths, when compared to other herited in some cases, rather than acquired through parental nonpsychopathic criminals, are more frequently and severely abuse, neglect, or an unreflective parent, a possibility hereto­ violent, are more likely to target strangers, engage in both af­ fore unacknowledged among attachment researchers. Testing fective and predatory violence, perpetuate violent criminal of this hypothesis may contribute to our fuller understanding acts for a longer period of time across their life span, and are of the exact nature of heritability of psychopathy. Another often found among the most feared and unpredictable offend­ intriguing area of investigation is the role that deficiencies in ers: those who commit sexually sadistic acts and serial sexual vasopressin and oxytocin, two hormones apparently related homicides (Meloy, 2000a, 2000b). to attachment (Fisher, 1998), may play in the biology of psy­ Curiously, there are no published studies that have directly chopathy, two biochemicals unexplored by psychopathy re­ measured psychopathy (Hare, J991) and attachment (using searchers. This might help us understand the psychopath's the AAI or other direct self-report measures) in samples of lack of empathy and enormous capacity for cruel aggression. male inmates, despite the work cited earlier concerning the externalizing, disruptive, and controlling behavior found in "hildren and adolescents with various attachment patholo­ CONCLUSION ,Jes, and the chronic cortical underarousal found in habitual criminals (Raine, 1993). There has, however, been work in It may become an empirically grounded truism, years from two related areas. Gacono and Meloy (1994) found in a now, that attachment pathology is a centrally necessary, but 522 Pathologies of Attachment, Violence, and Criminality alone insufficient, component to explain violence: whether it Barratt, E., Stanford, M., Felthous, A, & Kent, T. (1997). The ef­ is the hyperaroused, preoccupied attachment pathology of fects of phenytoin on impulsive and premeditated aggression: A stalking behavior that often results in affective violence, or controlled study. Journal ofClinical Pharmacology, 17, 341-49. the hypoaroused, dismissive attachment pathology of the Bartholomew, K. (1990). Avoidance of intimacy: An attachment psychopath that often results in predatory violence. In a more perspective. Journal of Social and Personal Relationships, 7, applied context, violent attachments and their measurement 147-178. through the use of reliable, valid, and normed forensic instru­ Bartholomew, K. (1994). The assessment of individual differences ments, none of which currently exist, may become de in adult attachment. Psychological Inquiry,S, 23-27. rigueur, a standard of practice requirement, for the forensic Bartholomew, K. (1997). Adult attachment processes: Individual psychologist of the future. Current research is certainly light­ and couple perspectives. British Journal ofMedical Psychology. 70,249-263. ing the way. Bartholomew, K., & Horowitz, L. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Per­ REFERENCES sonality and Social Psychology, 61, 226-244. Bowlby, J. (1953). Some pathological processes set in train by early mother-child separation. Journal of Mental Science, 99. Adamson, J. L. (1998). Predicting attachment and violence in rela­ 265-272. tionships: An investigation across three generations (Order No. AAM9805493). Unpublished doctoral dissertation, Univer­ Bowlby, J. (1958). The nature of the child's tie to his mother.lnter­ sity of Nebraska-Lincoln. national Journal ofPsychoanalysis, 39, 350-373. Ainsworth, M. (1963). The development of infant-mother interac­ Bowlby, J. (1960). Separation anxiety. International Journal of tion among the Ganda. In B. Foss (Ed.), Determinants of infant Psychoanalysis, 41, 89-113. behavior (Vol. 2, pp. 67-112). New York: Wiley. Bowlby, J. (1961). Processes of mourning. International Journal of Ainsworth, M. (1967). Infancy in Uganda. Baltimore: Johns Psychoanalysis, 42, 317-340. Hopkins University Press. Bowlby, J. (1969). Attachment and loss. Volume I: Attachment. New Ainsworth, M. (1989). Attachments beyond infancy. American York: Basic Books. Psychologist, 44, 709-716. Bowlby, J. (1973). Attachment and loss. Volume II: Separation: Ainsworth, M., Bell, S., & Stayton, D. (1971). Individual differences Anxiety and anger. New York: Basic Books. in strange situation behavior of one-year-olds. In H. R. Schaffer Bowlby, J. (1979). The making and breaking of affectional bond\. (Ed.), The origins of human social relations (pp. 17-52). New London: Tavistock. York: Academic Press. Bowlby, J. (1980). Attachment and loss. Volume Ill: Loss: Sadne.l.\ Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978). Patterns and depression. New York: Basic Books. ofattachment. Hillsdale, NJ: Erlbaum. Bowlby, J. (1984). Violence in the family as a disorder of the at· Ainsworth, M., & Bowlby, J. (1954). Research strategy in the study tachment and caregiving systems. American Journal of Psycho­ of mother-child separation. Courr. Cent. Int. En/, 4, 105. analysis, 44, 9-27. Ainsworth, M., & Wittig, B. (1969). Attachment and exploratory be­ Carr, S., Dabbs, J., & Carr, T. (1975). Mother-infant attachment: The havior of one-year-olds in a strange situation. In B. M. Foss importance of the mother's visual field. Child Development, 46. (Ed.), Determinants of infant behavior (Vol. 4, pp. 113-136). 331-338. London: Methuen. Cassidy, J. (1999). The nature of the child's ties. In J. Cassidy & P. Allen, J., Hauser, S., & Borman-Spurrell, E. (1996). Attachment Shaver (Eds.), Handbook of attachment (pp. 3-20). New York: theory as a framework for understanding sequelae of severe ado­ Guilford Press. lescent psychopathology: An I I-year followup study. Journal of Cassidy, J., & Shaver, P. (Eds.). (1999). Handbook of attachment. Consulting and Clinical Psychology, 64, 254-263. New York: Guilford Press. American Psychiatric Assoc. (1994). DSM-N. Washington: author. Cornell, D., Warren, J., Hawk, G., Stafford, E., Dram, G., & Pine, D. Ashby, W. R. (1956). An introduction to cybernetics. New York: (1996). Psychopathy in instrumental and reactive violent of· Wiley. fenders. Journal of Consulting and Clinical Psych~logy, 64. Babcock, J., Jacobson, N., Gottman, J., & Yerington, T. (2000). At­ 783-790. tachment, emotional regulation, and the function of marital vio­ Crowell, 1., Fraley, R. C., & Shaver, P. (1999). Measurement of lence: Differences between secure, preoccupied, and dismissing individual differences in adolescent and adult attachment. In violent anQ nonviolent husbands. Journal of Family Violence, J. Cassidy & P. Shaver (Eds.), Handbook of attachmefJt 15, 391-409. (pp. 434-465). New York: Guilford Press. References 523

Alpach, W., & Spitzberg, B. (1998). The dark side ofclose relation­ classification, and response to psychotherapy. Journal of Con­ ships. Hillsdale, NJ: Erlbaum. sulting and Clinical Psychology, 64, 22-31. Dennis v. Lane (1704). 87 English Reports (Queens Bench), Fonagy, P., Redfern, S., & Charman, T. (1997). The relationship 887-888. between belief-desire reasoning and a projective measure of at­ Downey, G., & Feldman, S. (1996). Implications of rejection sensi­ tachment security (SAT). British Journal of Developmental Psy­ tivity for intimate relationships. Journal ofSocial and Personal­ chology, IS, 51--61. ity Psychology, 70, 1327-1343. Fonagy, P., Steele, M., Moran, G., Steele, H., & Higgitt, A. (1991). Downey, G., Feldman, S., & Ayduk, O. (2000). Rejection sensitivity The capacity for understanding mental states: The reflective self and male violence in romantic relationships. Personal Relation­ in parent and child and its significance for security of attach­ ships, 7, 45--61. ment.lnfant Mental Health Journal, 13,200-216. Dutton, D. (1994). Behavioral and affective correlates of borderline Fonagy, P., Steele, M., Steele, H., Higgitt, A., & Target, M. (1994). personality organization in wife assaulters. International Jour­ Theory and practice of resilience. Journal of Child Psychology nal ofLaw and Psychiatry, 17, 26-38. and Psychiatry, 35, 231-257. Dutton, D. (1995a). The domestic assault of women. Vancouver, Fox, N., & Card, 1. (1999). Psychophysiological measures in the Canada: University of British Columbia Press. study of attachment. In J. Cassidy & P. Shaver (Eds.). Handbook Dutton, D. (1995b). Intimate abusiveness. Clinical Psychology: ofattachment (pp. 226-248). New York: Guilford Press. Science and Practice, 2, 207-224. Freud, S. (1912). On the universal tendency of debasement in Dutton, D. (1995c). Male abusiveness in intimate relationships. the sphere of love (contributions to the psychology of love II). Clinical Psychology Review, 15, 567-581. The standard edition of the complete psychological works of Sigmund Freud (Vol. pp. 177-190). London: Hogarth Dutton, D. (1998). The abusive personality. New York: Guilford 11, Press. Press. Dutton, D. (1999). Traumatic origins of intimate rage. Aggression Freud, S: (1919). A child is being beaten. The standard edition of and Violent Behavior, 4, 431-447. the complete psychological works of Sigmund Freud (Vol. 17, pp. 179-204). London: Hogarth Press. 'tton, D., & Painter, S. (1993). Emotional attachments in abusive relationships: A test of traumatic bonding theory. Violence and Gabbard, G. (1989). Two substypes of narcissistic personality dis­ Victims, 8, 105-120. order. Bulletin ofthe Menninger Clinic, 53, 527-532. Dutton, D., Saunders, K., Starzomski, A., & Bartholomew, K. Gacono, C., & Meloy, J. R. (1994). The Rorschach assessment (1994). Intimacy-anger and insecure attachment as precursors of of aggressive and psychopathic personalities. Hillsdale, NJ: abuse in intimate relationships. Journal of Applied Social Psy­ Erlbaum. chology, 24, 1367-1386. Goldberg, S., Muir, R., & Kerr, J. (1995). Attachment theory: Eisenberg, J. (1966). The social organization of mammals. Hand­ Social, developmental, and clinical perspectives. Hillsdale, NJ: buch Zoologie, 8, 1-92. Analytic Press. Fein, R., & Vossekuil, B. (1999). Assassination in the United States: Gottman, J., Jacobson, N., Rushe, R., Shortt, J., Babcock, J., An operational study of recent assassins, attackers, and near­ LaTaillade, J., et al. (1995). The relationship between heart rate lethal approachers. Journal of Forensic Sciences, 44, 321-333. reactivity, emotionally aggressive behavior, and general violence in batterers. Journal ofFamily Psychology, 9,227-248. Fisher, H. (1998). Lust, attraction, and attachment in mammalian re­ production. Human Nature, 9, 23-52. Guerrero, 1:. (1998). Attachment-style differences in the experience and expression of romantic jealousy. Fonagy, P. (1999a). Attachment, the development of the self, and its Personal Relationships,S, 273-291. pathology in personality disorders. In Derksen et al. (Eds.), Treatment of personality disorders (pp. 53--68). New York: Hamberger, L., & Hastings, J. (1986). Personality correlates of men Kluwer AcademiclPlenum. who abuse their partners: A cross-validation study. Journal of Fonagy, P. (1999b). Male perpetrators of violence against women: Family Violence, I, 323-341. An attachment theory perspective. Journal ofApplied Psychoan­ Hamberger, L., Lohr, J., Bonge, D., & Tolin, D. (1996). A large sam­ alytic Studies, 1,7-27. .ple empirical typology of male spouse abusers and its relation­ Fonagy, P. (1999c). Psychoanalytic theory from the viewpoint of at­ ship to dimensions of abuse. Violence and Victims, II, 277-292. tachment theory and research. In J. Cassidy & P. Shaver (Eds.), Hare, R. D. (1991). Manualfor the Psychopathy Checklist-Revised. Handbook of attachment (pp. 595--624). New York: Guilford Toronto, Ontario, Canada: Multi-Health Systems. Press. Harlow, H. (1961). The development of affectional patterns in infant .-onagy, P., Leigh, T., Steele, M., Steele, H., Kennedy, R., Mattoon, monkeys. In B. M. Foss (Ed.), Determinants of infant behavior G., et al. (1996). The relation of attachment status, psychiatric (pp. 75-97). London: Methuen. S24 Pathologies of Attachment, Violence, and Criminality

Harlow, H. (1969). Age-mate or affectional system. In D. Lehrman, Kienlcn. K .. Birmingham, D., Solberg, K., O'Regan, J., & Meloy. R. Hinde, & E. Shaw (Eds.), Advances in the study of behavior J. R. (IYY7). A comparative study of psychotic and non­ (Vol. 2, pp. 334-383). New York: Academic Press. psychotic stalking. Journal ofthe American Academy ofPsychi­ Hazan, c., & Shaver, P. (1987). Romantic love conceptualized as an atry LUll'. 25. 317-334. attachment process. Journal of Personality and Social Psychol­ Klevens. J., & Roca, J. (1999). Nonviolent youth in a violent og); 52. 511-524. society: Resilience and vulnerability in the country of Colombia. Heider, F. (1958). The psychology of interpersonal relations. New Violence ([nd Victims. 14, 311-322. York: Wiley. Langhinrichsen-Rohling, J., Palarea, R., Cohen, 1., & Rohling, M. Hendrick, c.. & Hendrick, S. (1986). A theory and method of love. (2000). Breaking up is hard to do: Unwanted pursuit behaviors Journal ofPersonality and Social Psychology. 50, 392-402. following the dissolution of a romantic relationship. Violew(' and Victims, 15, 73-90. Hofer, M. (1995). Hidden regulators: Implications for a new under­ standing of attachment, separation, and loss. In S. Goldberg, R. Lee, J. (1976). The colors of love. Englewood Cliffs, NJ: Prentice­ Muir, & J. Kerr (Eds.), Attachment theory: Social, developmen­ Hall. tal. and clinicClI perspectives (pp. 203-230). Hillsdale, NJ: Lewin, K. (1933). Environmental forces. In C. Murchison (Ed.), i\ Analytic Press. handbook ofchild psychology (2nd ed., pp. 59Q-625). Worcester. Holmes, J. (1995). "Something there is that doesn't love a wall": MA: Clark University Press. John Bowlby, attachment theory, and psychoanalysis. In S. Lyons-Ruth, K. (1996). Attachment relationships among children Goldberg, R. Muir, & J. Kerr (Eds.), Attachmellt theory: Social, with aggressive behavior problems: The role of disorganized developmental, and clinical perspectives (pp. 19-44). Hillsdale, early attachment patterns. Journal of Consulting and Clinical NJ: Analytic Press. Psychology, 64. 64-73. Holtzworth-Munroe, A. (2000). A typology of men who are violent Lyons-Ruth, K., Repacholi, B., McLeod, S., & Silva, E. (1991). Dis­ toward their female partners: Making sense of the heterogeneity organized attachment behavior in infancy: Short-term stability. in husband violence. Current Directions in Psychological Sci­ maternal and infant correlates, and risk-related subtypes. Devel­ ence, 9, 140-143. opmellt and Psychopathology, 3, 377-396. Holtzworth-Munroe, A, Meehan, J., Herron, K, Rehman, U., & Main, M. & Goldwyn, J. (1984). Adult attachment scoring and clas­ Stuart, G. (in press). Testing the Holtzworth-Munroe and sification system. Unpublished manuscript. University of Cali­ Stuart batterer typology. JOl/rnal of Consulting and Clinical fornia at Berkeley. Psychology. Main, M., & Hesse, E. (1990). Parents' unresolved traumatic expe­ Holtzworth-Munroe, A., & Stuart, G. (1994). Typologies of male riences are related to infant disorganized attachment status: Is batterers: Three subtypes and the differences among them. Psy­ frightened andlor frightening parental behavior the linking chological Blllletin, i 16. 476-497. mechanism? In M. Greenberg, D. Cicchetti & E. Cummings Holtzworth-Munroe, A., Stuart, G., & Hutchinson, G. (1997). Vio­ (Eds.), Attachment in the preschool years (pp. 161-184). lent versus nonviolent husbands: Differences in attachment pat­ Chicago: University of Chicago Press. terns, dependency, and Jealousy. .loumal of Family Psychology. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy. 11,314-331. childhood, and adulthood: A move to the level of representation. Irwin. H. (1999). Violent and nonviolent revictimization of women MO/lographs of the Society for Research in Child Developmelll. abused in childhood. Journal of lllterpersonal Violence, 14, 50,66-104. 1095-1110. Main,· M., & Solomon, J. (1986). Discovery of an insecure­ Jones, 1. (1992). Let me take you down. New York: St. Martin's disorganized/disoriented attachment pattern. In T. B. Brazelton Press. & M. W. Yogman (Eds.), Affective development in infann' Kernberg, O. (1984). Severe personality disorders. New Haven, CT: (pp. 95-124). Norwood, NJ: Ablex. Yale University Press. Main, M.. & Solomon, J. (1990). Procedures for identifying infants as Kesner, 1.. Julian, T., & McKenry, P. (1997). Application of attach­ disorganized/disoriented during the Ainsworth Strange Situation. ment theory to male violence toward female intimates. Journal In M. Greenberg. D. Cicchetti & E. Cummings (Eds.), Attachment (!f Family Violence. i2, 211-228. in the preschool years (pp. 121-160). Chicago: University or Kesner, J., & McKenry, P. (1998). The role of childhood attachment Chicago Press. factors in predicting male violence toward female intimates. Marcus, R., & Gray, L. (1998). Close relationships of violent and Journal ofFamily Violence, 13,417-432. nonviolent African American delinquents. Violence and Victim,l. Kienlen, K. (1998). The social antecedents of stalking. In J. R. i3,31-46. Meloy (Ed.), The psychology of stalking: Clinical and forensic McCann, J. (2001). Stalking of children and adolescents: The primi­ perspectives (pp. 52-67). San Diego, CA: Academic Press. tive bond. Washington, DC: American Psychological Association References 525

,G. H. (1934). Mind, self, and society. Chicago: University of between spouses. Journal ofConsulting and Clinical Psychology, t:hicago Press. 37, 263-268. Mechanic, M., Weaver, T., & Resick, P. (2000). Intimate partner Painter, S., & Dutton, D. (1985). Patterns of emotional bonding in violence and stalking behavior: Exploration of patterns and cor­ battered women: Traumatic bonding. International Journal of relates in a sample of acutely battered women. Violence and Vic­ Women sStudies, 8, 363-375. tims; 15, 55-72. Polan, H. 1., & Hofer, M. (1999). Psychobiological origins of infant Meloy, 1. R. (1988). The psychopathic mind: Origins, dynamics, attachment and separation responses. In 1. Cassidy & P. Shaver and treatment. Northvale, N1: Aronson. (Eds.), Handbook of attachment (pp. 162-180). New York: Meloy, 1. R. (1989). Unrequited love and the wish to kill: The diag­ Guilford Press. nosis and treatment of borderline erotomania. Bulletin of the Raine, A (1993). The psychopathology of crime. San Diego, CA: Menninger Clinic, 53, 477-492. Academic Press. Meloy, 1. R. (1992). Violent attachments. Northvale, N1: Aronson. Raine, A, Brennan, P., & Mednick, S. (1997). Interaction between Meloy, 1. R. (1996). Stalking (obsessional following): A review of birth complications and early maternal rejection in predisposing some preliminary studies. Aggression and Violent Behavior, 1, individuals to adult violence: Specificity to serious, early-onset 147-162. violence. American Journal ofPsychiatry, 154, 1265- 1271. Meloy, 1. R. (1997). Predatory violence during mass murder. Jour­ Raine, A, Meloy, 1. R., Bihrie, S., Stoddard, 1., LaCasse, L., & nal ofForensic Sciences, 42, 326-329. Buchsbaum, M. (1998). Reduced prefrontal and increased Meloy, 1. R. (1999). Stalking: An old behavior, a new crime. Psy­ subcortical brain functioning assessed during positron emission chiatric Clinics ofNorth America, 22, 85-99. tomography in predatory and affective murderers. Behavioral Meloy, 1. R. (2oooa). The nature and dynamics of sexual homicide: Sciences and Law, 16, 319-332. An integrative review. Aggression and Violent Behavior, 5, 1-22. Robertson, 1. (1953). A two year old goes to hospital [Film). Meloy, 1. R. (2ooob). Violence risk and threat assessment. San London: Tavistock Child Development Research Unit. (Avail­ Diego, CA: Specialized Training Services. able from Penn State Audiovisual Services, University Park, PA). Meloy, 1. R. (2001). Communicated threats toward public and pri­ Robertson, 1., & Bowlby, 1. (952). Responses of young children to ~ targets: Discerning differences among those who stalk and separation from their mothers. CourT: Cent. Int. Enj, 2, 131-142. _•.lick. Journal ofForensic Sciences, 46. Rosenstein, D., & Horowitz, H. (1996). Adolescent attachment Meloy, 1. R. (in press). Stalking and violence. In 1. Boon & L. Sheri­ and psychopathology. Journal of Consulting Clinical Psychol­ dan (Eds.), Stalking and psychosexual obsession. London: ogy, 64, 244--253. Wiley. Saunders, D. (1992). A typology of men who batter women: Meloy, 1. R., & Gothard, S. (1995). Demographic and clinical com· Three types. American Journal ofOrthopsychiatry, 62, 264--275. parison of obsessional followers and offenders with mental dis­ Shaw, D., & Bell, R. (1993). Chronic family adversity and infant at­ orders. American Journal ofPsychiatry, 152, 258-263. tachment security. Journal of Child Psychology and Psychiatry, Meloy, J. R., Rivers, L., Siegel, L., Gothard, S., Naimark, D., & 34, 1205-1215. Nicolini, R. (2000). A replication study of obsessional followers Sorce, 1., & Emde, R. (1981). Mother's presence is not enough: and offenders with mental disorders. Journal of Forensic Sci­ Effect of emotional availability on infant explorations. Develop­ ences, 45, 189-194. mental Psychology, 17, 737-745. Millon, T. (1981). Disorders of Personality DSM-I11: Axis 11. New Spangler, G., & Grossman, K. (993). Biobehavioral organization in York: Wiley. securely and.insecurely attached infants. Child Development, 64, Millon, T. (1998). Psychopathy: Criminal, violent, and antisocial 1439-1450. behaviors. New York: Guilford Press. Spitzberg, B. (2001). Forlorn love: Attachment styles, love styles, Mullen, P., Pathe, M., & Purcell, R. (2000). Stalkers and their vic­ loneliness, and obsessional thinking as predictors of obsessive tims. New York: Cambridge University Press. relational intrusion. Unpublished manuscript. San Diego State Nachmias, M., Gunnar, M., Mangelsdorf, S., Parritz, R., & Buss, K. University. (1996). Behavioral inhibition and stress reactivation: The mod­ Strachan, C. (1993). The assessment of psychopathy in female of­ erating role of attachment security. Child Development, 67, fenders. Unpublished doctoral dissertation, University of British 508-522. Columbia, Vancouver, Canada. Oderberg, N. (1995). History offamily violence, adult attachment, Taylor, C. (1997). Psychopathy and attachment in a group of and child abuse potential: Interrelationships in a sample ofcol­ incarcerated females. Unpublished doctoral dissertation, , 'e students (Order No. AAM9506364). Unpublished doctoral California School of Professional Psychology, San Diego. .ertation, University of Colorado, Boulder. Tjaden, P., & Thoennes, N. (1997). Stalking in America: Findings O'leary, K., Barling, 1., Arias, 1., Rosenbaum, A, Malone, J., & from the National Violence Against Women survey. Denver, CO: Tyree, A. (1989). Prevalence and stability of physical aggression Center for Policy Research. 526 Pathologies of Attachment, Violence, and Criminality 1 j Tjaden, P., & Thoennes, N. (2000). Extent, nature, and conse­ on the predictive validity of the Adult Attachment Interview. quences ofintimate partner violence (NCJ 181867). Washington, Psychological Bulletin, lJ7, 387-403. DC: U.S. Department of Justice. Waltz, J., Babcock, J., Jacobson, N., & Gottman, J. (2000). Testing Tweed, R., & Dutton, D. (1998). A comparison of impulsive and in­ a typology of batterers. Journal ofConsulting and Clinical Psy· strumental subgroups of batterers. Violence and Victims, 13, chology,68,658-669. 217-230. van Ijzendoom, M. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis

, 1