Daniel Sonkin, PHD. Licensed Marriage & Family Therapist HOME | CONTACT | ABOUT
header articles

Treating Assaultive Men from an Attachment Perspective

Sonkin, D. and Dutton, D (2003). Treatment assaultive men from an attachment perspective. In Dutton, Don and Sonkin, Daniel (eds). Intimate Violence: Contemporary Treatment Innovations. New York: Haworth Publishing. Copyright 2002.


Daniel Jay Sonkin, Ph.D. is a Licensed Marriage and Family Therapist in an independent practice in Sausalito, California. His work focuses on the treatment of individuals and couples facing a variety psychological problems including domestic violence and child abuse.  In addition to his clinical experience, he has testified as an expert witness in criminal cases where domestic violence is an issue.  He has also evaluated defendants facing the death penalty conducting social histories with a focus on their childhood abuse and it’s impact on adult criminal behavior.  He has also testified as an expert wintess in malpractice cases and licensing actions.  As one of the early investigators and specialists in the field of family violence he has developed a widely used protocol for treating male batterers.   He is the author of numerous articles and  books on domestic violence and child abuse including: Learning to Live Without Violence: A Handbook for Men, The Male Batterer: A Treatment Approach, Domestic Violence on Trial: Psychological and Legal Dimensions of Family Violence, and Wounded Boys/Heroic Men: A Man’s Guide to Recovering from Childhood Abuse, The JurisMonitor Stabilization Program for Stalkers, A Counselors Guide to Learning to Live Without Violence and Domestic Violence: The Court-Mandated Perpetrator Assessment and Treatment Handbook.  He has written software for assessing perpetrators of violence for both the Macintosh and IBM compatible computers. He has conducted trainings nationally and internationally for mental health professionals on the treatment of male batterers.   He is a former chair of the state ethics committee of the California Association of Marriage and Family Therapists and a former member of the Board of Directors for that organization.  In addition to his clinical practice, he is an adjunct faculty in the Department of Counseling at Sonoma State University.  Dr. Sonkin provides consultation, training and supervision in his unique model outlined in his books to individuals and agencies providing services to male batterers and their families.

Don Dutton received his Ph.D. in Psychology from the University of Toronto in 1970.  In 1974, while on faculty at the University of British Columbia, he began to investigate the criminal justice response to wife assault, preparing a government report that outlined the need for a more aggressive response, and subsequently training police in “domestic disturbance” intervention techniques. After receiving training as a group therapist at Cold Mountain Institute, he co-founded the Assaultive Husbands Project in 1979, a court mandated treatment program for men convicted of wife assault. During the fifteen years he spent providing therapy for these men, he drew on his background in both social and clinical psychology to develop a psychological model for perpetrators of intimate abuse. This model views intimate abusiveness as emanating from a trauma triad and comprised of witnessing abuse, being shamed and experiencing insecure attachment. He has published over 100 papers and three books, including the Domestic Assault of Women (1995), The Batterer:A Psychological Profile(1995) and  The Abusive Personality (1998). The Batterer has been translated into French, Spanish, Dutch and Polish and Dutton has provided numerous workshops to professionals based on this work, including talks at the Sorbonne in Paris, Washington, D.C.and New York City.  Dutton frequently serves as an expert witness in civil trials involving domestic abuse and in criminal trials involving family violence, including his work for the prosecution in the O.J. Simpson trial (1995). The latter led to an interest in spousal homicide and to “abandonment killing”. He is currently Professor of Psychology at the University of British Columbia, Vancouver, BC, Canada.


This chapter explores the relationship between attachment theory and treatment of perpetrators of domestic violence.  First the authors present a brief overview of attachment theory.  This is followed by a discussion of how domestic violence research findings suggests that attachment theory is a good paradigm to understanding the phenomenon of intimate violence.  Lastly, the authors describe the elements of attachment oriented psychotherapy as they might apply to working with perpetrators of domestic violence.

Keywords:  attachment theory, attachment, psychotherapy, domestic violence, perpetrator treatment.

In a landmark series of studies entitled Attachment and Loss, Bowlby (1969, 1973, 1980) outlined a remarkable theory that posited that early attachment had sociobiological significance and constituted a powerful human survival motive.  The theory has implications for anger in interpersonal relationships and for the seemingly irrational outburst that accompany real or imagined separation.  Primary attachment (usually to the mother[i]) is governed by three important principles: first, alarm of any kind, stemming from any source, activates an attachment survival system in an infant that directs and motivates it to seek out soothing physical contact with the attachment figure.  Second, when activated, only physical attachment with the attachment figure will terminate it.  Third, when the system has been activated for a long time without soothing and termination, angry behavior appears; if soothing and protection is not eventually found, the system can then become suppressed.  Bowlby  reported observations he made of young children (15 –30 months) separated for the first time from their mothers. He witnessed a three phase behavioral display: anger, despair, and detachment.  He concluded from these observations that the primary function of anger was to generate displays that would lead to the return of the absent mother.  Anger is thus an attempt to recapture the object that can soothe tension and anxiety at a developmental stage where the child cannot yet self soothe through signaling the mother that she is wanted and/or needed.  Anger is an emotion “born of fear” of loss.  Dysfunctional anger, occurring later in adult affectional bonds, was defined as anger that distanced the attachment object instead of bringing (her) closer.

Subsequent empirical studies by Ainsworth and her colleagues (1978) showed that different “attachment styles” existed for infants.  Initially these were classified as “secure,” anxious-avoidant, and anxious-ambivalent.  A fourth category emerged in their research that was eventually called disorganized.  Subsequent terms for the three insecure patterns were dismissing, preoccupied, and fearful (see Figure 1).  The pre-occupied and fearful types sought attachment but experienced anxiety as a consequence of attachment.  Also, both experienced anxiety at the disappearance of the mother and were difficult to soothe upon reunion.  The fearful children were particularly ambivalent upon reunion with their attachment figure, both approaching and avoiding contact.  Bowlby (1969) described these children as “arching away angrily while simultaneously seeking proximity” when re-introduced to their mothers.  Interestingly, although the avoidant or dismissing children seemed content in the absence of their attachment figure and not particularly interested in reconnecting upon reunion, when physiological measures were taken, these children were quite anxious during separation, but somehow learned to repress their feelings.

In 1987, Hazan and Shaver published a landmark study that showed that adult “attachment styles” resembled infant attachment styles.  (The spate of research that emerged on adult attachment styles is too voluminous to review here; however, the interested reader is referred to Karen (1977) or Sperling and Berman (1994) for a discussion on the history of attachment theory).  Assessment of adult attachment can be done through interviews (Main & Goldwyn, 1998), projective tests (West & George, 1999) or self reports (see http://psyweb2.ucdavis.edu/labs/Shaver/).  Sperling and Berman (1994) define adult attachment as “the stable tendency of an individual to make substantial efforts to seek and maintain proximity to and contact with one or a few specific individuals who provide the subjective potential for physical and/or psychological safety and security” (p. 8).

One of ways attachment styles have been deconstructed involves what are called representational models of self and other (Bartholomew & Horowitz, 1991).  Each of these representations is a network of beliefs and expectancies about how the relationship will function.  It is a cannon of attachment theory that these representational models are internalized through the attachment process.  They include positive and negative views of self, expectancies about what will be received from another and generalized projections about relationship outcome.  Bartholomew (1990) systematized these into a 2x2 arrangement for each of four attachment styles, each having a positive or negative self-concept and expectation of another (via relationships).  These beliefs present another aspect of attachment that is open to therapeutic intervention.  In Bartholmew’s schema, Preoccupied attachment styles have negative self-images, while Dismissing attachment styles have negative other images.  Fearful attachment styles have both negative self and other images.  Fearful attachment styles also expect the worst from an intimate relationship but need such a relationship to heal their damaged self-image.  Hence, they are thrown into an ambivalent double avoidance (aloneness versus engulfment) that may serve as the basis for borderline alternation (see Dutton, 1998).

Dutton, Saunders, Starzomski, and Bartholomew (1994) attempted to relate attachment style in adults to abusive behaviors.  In a sample of 120 men in treatment for wife assault and 40 demographically matched controls, they assessed men’s attachment style using a self report measure called the Relationship Style Questionnaire (Griffin and Bartholomew, 1994) and abusiveness through wives’ reports using the Psychological Maltreatment of Women Inventory (PMWI; Tolman, 1989).  A Fearful attachment style in the male perpetrator was highly related to abusiveness, correlating +.46 with the Domination/Isolation subscale of the PMWI, and +.52 with the Emotional Abuse subscale.  These relationships were highly significant.  To a lesser extent, an attachment style called Preoccupied also correlated significantly with abuse.  Fearfully attached men also reported high levels of chronic anxiety and anger.  Bartholomew, Henderson, and Dutton (2001) also found that women in shelters escaping abusive relationships could be classified with Preoccupied (53%) or Fearful (35%) attachment styles.  These styles were related to the woman’s difficulty in leaving the relationship.

Dutton (1998) described what he called the “abusive personality,” a constellation of psychological traits that, when assessed in males, are highly related to partners’ reports of abusiveness.  Fearful attachment was an important component of this personality constellation and, according to Dutton, directed the anger to an intimate target.  Men whose violence was predominantly or exclusively in intimate relationships probably have an attachment disorder.  This disorder may be related to personality disorder diagnoses such as Borderline or Dependent personality; however, it has an attachment aspect to its’ origin and plays itself out in intimate relationships.  Dutton has suggested that such men have both a public and private (intimate) personality that may be quite dissimilar.

Typically, batterer treatment has not included specific work on attachment, yet therapists regularly hear of delusional construals of or pre-occupation with the spouses behavior (deemed “conjugal paranoia”).  For example, one of the authors (Don Dutton) had a client who was convinced that his wife was having an affair when he found “a key with a man’s name on it” (the key manufacturer).  Many batterers present as cold, unemotional, and non-empathetic, similar to persons with avoidant attachment and those suffering from psychopathy.  Likewise, it is common for batterers to show patterns of approach/avoidance as seen with disorganized attachment and borderline personality disorder.  Spousal homicide committed by males is frequently in response to real or perceived abandonment (Dutton & Kerry, 1999).  Browning and Dutton (1986) obtained pronounced anger/arousal responses in batterers who witnessed a videotape depiction of an “abandonment” (a woman unilaterally deciding to visit another city with female friends and join a woman’s consciousness- raising group).  Their anger/arousal scores were significantly higher than control groups of men, and were especially pronounced on this “abandonment” scenario.  The relationship between fear of abandonment and rage thus appears strong in this group of partner abusive men.  The conversion of fear to rage could occur because the latter is more consistent with male sex role conditioning.  Regardless, the confrontation of this emotional contribution to abuse deserves therapeutic attention.

These data suggests that incorporating attachment theory into batterer treatment is well founded.  First, it can enable batterers to perceive a broader pattern in their reactions to loss and separation in their intimate relationships.  Second, this theory supports the prevailing notion that clients need to learn emotion self-regulation during periods of attachment-anxiety.  Third, attachment theory suggests that through altering the internal working models of self and other the client can break a perceptual mold in which attachment-anxiety is reduced to either distancing, clinging, or approach/avoidance.

Theory to Practice: Attachment Theory Informed Psychotherapy

Although psychotherapy with adults from an attachment perspective is still in early development, some significant clinical ideas and applications exist.  Some clinical scholars have incorporated attachment theory into other theories (Masterson & Klein, 1995; Schore, 1994), which has served to enhance general psychoanalytic theory and practice.  However, other psychoanalytically oriented theorists have criticized this theory based on it’s interpersonal versus intrapsychic focus and the categorical, mutually exclusive attachment categories (Fonagy, 1999).  To date, there exists little (Slade, 1999) or no specific models of attachment theory informed psychotherapy with adults.  It is beyond the scope of this chapter to debate the strengths and weaknesses of attachment theory as it applies to psychotherapeutic intervention, but regardless of the final outcome of such a debate, Arietta Slade sums up the controversy by stating, “In essence, attachment categories do tell a story.  They tell a story about how emotion has been regulated, what experiences have been allowed into consciousness, and to what degree an individual has been able to make meaning of his or her primary relationships” (p. 585).  Given this perspective, let’s first look at the road map Bowlby has laid out when applying his theory to clinical practice. 

Bowlby explicitly saw the therapist as a surrogate mother who encouraged the client to explore the world from a secure base he or she creates.  In the context of therapeutic work with individuals, Bowlby (1988) defined five tasks:

  1. Create a safe place, or Secure Base, for client to explore thoughts, feelings and experiences regarding self and attachment figures;
  2. Explore current relationships with attachment figures;
  3. Explore relationship with psychotherapist as an attachment figure;
  4. Explore the relationship between early childhood attachment experiences and current relationships; and
  5. Find new ways of regulating attachment anxiety (i.e., emotional regulation) when the attachment behavioral system is activated.

Each of these five tasks is described in detail below.

Creating a Secure Base

The primary task that Bowlby states as necessary to addressing attachment in psychotherapy is the development of the secure base.  In this section, we will define the secure base, and discuss its development and function in the therapeutic relationship.

What is a Secure Base?

In order to understand Bowlby’s concept of the secure base in psychotherapy, one must look at how this is developed between the mother and child.  The infant’s inability to communicate in adult terms makes parenting a challenging task.  Parents (and mothers in particular) must develop skills in empathy and attunement in order to understand the needs of the developing child.  An attuned mother (or father) can tell the difference between a full diaper cry, a hungry cry, and a tired cry.  Even if they can’t tell the exact difference, they are quick to assume that the baby is distressed and in need of some form of caretaking, and if in their response one strategy does not work, they quickly employ another.  Compare this to an insensitive or misattuned parent, who either ignores that child’s needs altogether, considers the crying a problem and loses sight of the underlying needs, or is overwhelmed by the baby's needs. 

The attachment behavioral system, according to Bowlby however, does not just activate when the child is hungry or needs a diaper change.  The attachment system activates when there is fear or vulnerability for some reason.  Perhaps the baby heard a loud noise or woke up in the dark.  These experiences activate the attachment system, which serves to motivate the infant to seek protection from threatened danger.  The infant is like, as Cassidy (1999) describes, a heat-seeking missile, looking for an attachment object (typically the parent) that is sufficiently near, available, and responsive.  When this attempt for protection is met with success, the attachment system de-activates, the anxiety is reduced, the infant is soothed, and play and exploration can resume.  When these needs are not met the infant experiences primal anger accompanied by extreme arousal and terror.  These reactions, according to Bowlby, set a template for later adult reactions to abandonment.

The parental caretaking system compliments the infant’s attachment behavioral system.  It is the caretaking system that responds with the goal being to protect and reassure in order to reduce the child’s anxiety.  Behaviors that can accomplish this goal can range from the practical (e.g., putting the child down for a nap, or feeding or removing a child from a dangerous or frightening situation) to the more complicated process of mirroring the child’s inner life in words that help the child to learn self-reflection and understanding.  For example, for the two-year-old who is involved in a full-fledged tantrum, the parent may reflect the child’s feelings (e.g., “you must be so tired,” or “I know it hurts when I say no sometimes”).  As the child gets older, this mirroring process becomes more complex reflecting the child’s more sophisticated understanding of their feelings, needs, and relationships. 

No parent always knows how to respond, or even how to respond constructively.  Mis-attunements are an inevitable part of the parenting process.  This is beneficial for the child, because if a child grew up with a perfectly attuned parent, they would not be prepared for the vicissitudes of life.  They would be sadly disappointed to discover that other people in the world did not provide the same sensitivity as their mother or father.  Mis-attunements are opportunities for the child to develop realistic expectations about the world in response to their needs.  These mis-attunements and attunements are also an opportunity for parents to help children learn about the give and take of relationships.  Through the rupture and repair process, children learn about how people become intimate in spite of differences and conflict.  They develop a sense of poignancy and tolerance for the ambiguities of intimate relationships.

This process of the activated attachment behavioral system and the complimentary caretaking system helps to create the secure base necessary for healthy development, exploration, and play.  According to Bowlby’s theory, this healthy developmental process gets derailed when the parental caretaking system is not adequately or appropriately near, attentive, or responsive to the child’s attachment behavioral system.

The Secure Base in Psychotherapy

The parent who provides a secure base for their child through attunement, sensitivity, caring, setting limits, and teaching helps the child to learn to soothe the anxiety generated by the activated attachment system, and hence return to exploration and play.  It is through this exploration and play process that the child is developing a sense of self.  In the case of psychotherapy, the clinician is the caretaking figure who likewise provides a secure base so that the client’s attachment system is sufficiently deactivated and the client is free to explore and play.  In therapy, however, the exploration is the inner world of feelings, thoughts, and experiences, and the play is, for example, trying on new identities and responses to stress and conflict.

Developing a secure base in psychotherapy would be very easy if it were as simple as therapists being available, attentive, responsive, and attuned.  Unfortunately, it is not so elementary. What is interesting about psychotherapy is that, like the strange situation (Ainsworth et. al., 1978), it too creates a degree of emotional stress, can be threatening emotionally to clients, and can be detected by observing the coherence of the client’s stories about their attachment experiences (Main & Weston, 1981).  Sitting in the room with a stranger and talking about emotionally laden material can be quite anxiety provoking and likely to activate the attachment behavioral system right from the start of therapy.  Unlike the infant whose attachment experiences are not yet solidified into firmly established working models of self and others, the adult client has already developed a response set to stress and vulnerability within the interpersonal context.  That set, depending on the attachment style, will be similar to responses to other interpersonal relationships in their life, behaviors that contribute to problems that they are seeking help for in the first place.   Those attachment behaviors may be obvious, but can also be so subtle that the therapist will not recognize that they are present and interfering with the change process.  So on one hand the client is seeking help, yet on the other hand the client’s attachment behavioral system may be the very thing that presents obstacles to actually receiving assistance from the therapist. 

Research in domestic violence suggests that male batterers represent all three insecure attachment classes: avoidant, pre-occupied, and disorganized or fearful (Holtzworth-Monroe, et. al., 2000).  Each form of insecure attachment has particular defense mechanisms as a method of coping with attachment anxiety.  Batterers with an avoidant style present as disconnected emotionally, lacking empathy, cold, and uninterested in intimate relationships.  They can vacillate between being distant and cut-off emotionally to critical and controlling.  These clients need to incorporate an emotional soundtrack, as one client put it, into their life.  Batterers with a pre-occupied style try to please others in order to receive approval.  They can present as extremely self-controlled except when experiencing loss anxiety, when they can become extremely clingy and angry.   When experiencing emotion, these clients are overwhelmed by their attachment needs and are often unable to contain themselves.  Unlike their avoidant counterparts, these clients need cognitive structures necessary to contain their intense emotional reactions.  The fearful, or disorganized, batterer can manifest elements of both the avoidant and pre-occupied batterers.  They experience attachment anxiety, and fear of rejection or being hurt if they are too close and anxiety if they are too distant.  Like the disorganized children in the strange situation, these clients do not have an organized strategy for dealing with attachment anxiety.  Dutton (1998) has written extensively about the fearful/disorganized or borderline batterer.

How the therapist proceeds in the early stages of therapy with domestic violence clients is critical to the creation of the secure base.  If the therapist fails to notice the client’s strategies and their psychological function, the therapist’s responses will most probably confirm the client’s inner working models of self and others and reinforce the attachment behavioral system as it currently manifests.  However, if the therapist responds with empathy and attunement, two things can happen.  First, the client gets a different experience of him/herself.  The attuned therapist, like the attuned parent, will look beyond the client’s response set and help them recognize their unconscious motivations, needs, and emotions.  The therapist also helps the client view their response set (attachment behaviors) from a different perspective - how they undermine their getting their needs met in interpersonal relationships.  This rudimentary process is the beginning of the client altering their inner working models of self. 

The second possible outcome of therapist empathy and attunement is that the client experiences the therapist in a positive way, in that they feel understood, seen, and cared for by the therapist.  When the client feels understood and not judged, that experience in and of itself can be relieving and soothing, thereby beginning to alter their inner working models of others.   These processes, feeling understood and recognizing underlying needs and feelings, is the rudimentary beginning of the creation of a secure base in psychotherapy, a necessary first step in the process of altering the attachment behavioral system so that it is not likely to wreak havoc in interpersonal relationships.

However like parenting in the real world, even the most sensitive and talented therapists are not always going to be perfectly attuned; therefore, clients are likely to experience ruptures in this state of understanding and perfect attunement by the therapist.  Like with the developing child, these ruptures are not only inevitable but necessary to the process of therapy and the development of a more adaptive attachment behavioral system.  We will discuss these opportunities later in the section on utilizing the therapeutic relationship to effect change.

However, most batterers present in therapy with severe acting-out problems.  These can range from physical or non-physical abuse towards their family members or others to substance abuse, missing sessions, hostility toward the therapist, or other oppositional behavior.  The therapist is confronted with the following dilemma.  On one hand, the client requires understanding and support for the pain they are experiencing that leads to these behaviors.  On the other hand, continued acting-out will interfere with the client benefiting from the therapeutic experience.  Therefore, a combination of interpretation, which is necessary to facilitate the development of a secure base with the therapist, and confrontation, which is also necessary in setting limits on the self or other destructive acting-out behaviors, is needed. 

An Empirical Descriptions of Secure Base Priming

The idea of creating a secure base in psychotherapy sounds good, but is this a real concept or just another variation of the therapeutic alliance?  Researchers in adult attachment have been able to empirically test the notion that creating a secure base experience for individuals may temporarily alter an individual’s inner working models of others and therefore change behaviors or emotional states.  The idea of “secure base priming” has been gaining attention in the adult attachment literature.  Mario Mikulincer and Phil Shaver (2001) examined the effects of secure base priming on intergroup bias.  They hypothesized that having a secure base could change how a person appraises threatening situations into more manageable events without activating insecure attachment-like behaviors such as avoidance, fear, or preoccupation. They utilized a series of well-validated secure base priming techniques (described below) that have appeared to create in subjects a sense of security one would find in individuals who would might otherwise be assessed as having a secure attachment style.  These techniques were quite creative and had powerful effects on subjects.

One group was primed using subliminal presentation of words that exemplify a secure schema (e.g., love, support) within a word relation task (Arndt, Greenberg, Pyszczynski, & Solomon, 1997).  This is not unlike the therapist who gives verbal as well as non-verbal messages to clients communicating support, caring, and empathy.  In another study, participants performed a guided imagination task in which they visualized an interpersonal episode containing the prototypical if-then sequence of the secure base schema (Mikulincer & Arad, 1999).  This method seemed close to the process of helping clients imagine a situation with positive outcomes, such as one used by cognitive-behaviorists called rehearsals with a positive outcome.  What would it be like if they got the love and support that they deserve?  The third priming technique was Baldwin, Keelan, Fehr, Enns, and Koh Rangarajoo's (1996) visualization task, in which participants visualized a real person who served as a secure base for them.  Here again, it is not unusual to ask clients to talk about positive experiences in their life, or for the client to report thinking about the therapist (or another positive attachment figure such as a peer in the batterer's group) outside of the session as a means to self-soothe, feel reassured, or bolster confidence.

In all five of these studies, those subjects exposed to secure base priming acted in the experimental condition similar to securely attached individuals who did not receive priming but were nevertheless exposed to similar conditions assessing intergroup bias.  The authors suggest that secure base priming enhances motivation to explore by opening cognitive structures and reducing negative reactions to out-group members or to persons who hold a different world view.  The observed effects of secure base priming may reflect cognitive openness and a reduction in dogmatism and authoritarianism (Mikulincer & Shaver, 2001).  Other similar studies have found that secure base priming will have a positive effect on cognitive and affective states (Mikulincer, 1998).  Although these studies are not meant to be applied to clinical situations, they have powerful implications for the clinical setting.  Aspects of the psychotherapy process are similar to these descriptions of secure base priming and through that process clients may begin to change their internal representations of self and others or attachment styles. 

Exploring Current Relationships with Attachment Figures

As the therapy proceeds and the therapist works to create the secure base environment, Bowlby’s second task eventually begins to become a focus of the psychotherapy: exploring current relationships with attachment figures.  These attachment figures include family members, friends, relatives, partners, and spouses.  Here the client is exploring patterns in their close relationships, while the therapist is listening for patterns of relating that suggest secure or insecure attachment patterns, and if the latter, which particular insecure attachment style.  The exploration of these relationships helps the therapist understand the client’s attachment style as it manifests in the significant relationships of his or her life.  Research suggests that people may demonstrate different attachment styles in different relationships (Feeney, 1999).  This makes a certain amount of sense.  Since the attachment system is closely tied to the attachment figure’s caretaking system, then how the attachment figure responds to the client will in part determine the client’s response to attachment system activation.  In addition, in adult relationships (unlike a child-mother relationship) both adults are acting in the capacity of caretaker and seeking attachment for their own needs.  This fact is likely to complicate the issue of stability of attachment style within differing contexts. 

It is not completely clear how attachment style correlates with the issue of personality disorders (Dozier, Stovall, & Albus, 1999).  It is generally thought that people suffer from one personality disorder rather than multiple personality disorders.  Neither attachment theory or the empirical literature on personality disorders can say they have spoken the final word on this issue.  What is seen clinically, however, is that people do seem to have consistent core issues, but these issues may manifest differently in different contexts.  Like attachment relationship dynamics, personality disorders are likely to manifest differently depending on the context or relationship.  It is believed that attachment styles are not so much categorical as much as degrees; hence, different client-attachment figure relationships are likely to evoke different degrees of insecurity.  For example, one relationship may generate a mild avoidant response by the client, whereas another relationship may evoke an extreme avoidant reaction.  Even in the same relationship, different degrees of avoidance or anxiety may be evoked depending on the situation.  The same can be said about personality disorders.  Therefore, determining the attachment style of a particular client is only part of the goal of this process; more importantly, assessment is also done on how the attachment system is being activated with the client in a particular relationship or context.

Domestic violence perpetrating clients spend a great deal of time talking about their experiences with the partner they have abused.  The tendency to focus on the relationship or partner is great in this population.  These clients grew up in families where the attachment figure was not sufficiently present, attentive, or responsive; therefore, a great deal of personal energy was expended focusing on the attachment figure - are they present?  Are they going to respond positively?  Are they even going to know what I need?  These same questions are evoked in their adult relationships, either consciously or unconsciously.  Directing the clients to their inner experience is key to turning this pattern of externalizing behavior to one of personal awareness and responsibility.  Because so much focus in traditional domestic violence treatment is on anger management and power and control dynamics, therapists do not pay enough attention to the client’s inner psychological experience of relationships.  Here attachment theory can enhance the current domestic violence treatment paradigms.  By exploring the unconscious internal working models of self and other, clients can begin to understand why they may have the difficulties in regulating affect or why they experience a need to control others as a means to regulate attachment-related affect.

Exploring the Relationship between Early Childhood Attachment Experiences and Current Relationships

An important and necessary aspect of psychotherapy from an attachment perspective is the exploration of early childhood experiences and their effect on the inner experience of self and others.  Those experiences with caregivers formed the representational models of self and others from which the client views self and the significant attachment relationships in their life.  Although Bowlby’s description of this process seems primarily cognitive in nature, there is a significant emotional component to this task of psychotherapy.  In many cases, domestic violence perpetrators present with unresolved trauma, loss, and other emotionally laden relationship experiences that must be worked through cognitively, emotionally, and physically.  Victims of physical, sexual, and psychological maltreatment will experience a range of emotional reactions to this exploration process from depression to rage.  The therapist must be willing to work these painful minefields with the client.  Much has been written on addressing childhood abuse in psychotherapy (e.g., Herman, 1992; Van der Kolk, McFarlane & Weisaeth, 1996), a topic that is beyond the scope of this chapter.  But even with clients whose experiences would not be classified as “abuse,” painful recollections of subtle and no-so-subtle rejections and misattunements by parents evoke powerful feelings of sadness, loss, and anger.   Research in domestic violence treatment outcomes suggests that some perpetrators may need to address unresolved trauma before, or at least concurrently to, addressing violent acting-out behavioral patterns.

An important part of this process involves the exploration of the representational models of self and attachment figures that resulted from these experiences with the goal being to reappraise them and restructure them in light of the understanding and insight gleaned from this process.  Most often children’s strategy for dealing with unpleasant experiences is to put them out of mind.  In psychotherapy, the client can revisit these experiences but with the benefit of having an adult mind that can understand the reasons for their experiences and how they affected them psychologically.  Where the therapist has the most leverage in assisting the client in changing these representational models is through new relational experiences that the client has in therapy with the therapist him/herself.  The goal of this historical exploration is helping the client to be less “under the spell” of historical experiences with attachment figures.  In doing so, current relationships with attachment figures will be less charged.

Another important aspect of this process is to explore the more pathological aspects of insecure attachment.  Jealousy in batterers was first described by Walker (1979) and reiterated in Sonkin, Martin, and Walker (1985).  It was described as taking the form of frequent questioning of whom a spouse has been with or where she has been, accusations of her attraction to other men, and suspiciousness that she being flirtatious with other men.  In extreme cases, this serves as a motive for “pseudo-incarceration,” the literal isolation and confinement of the woman to the home and monitoring of her phone contacts.  It can also involve frequent phone calls to her place of work and insistence upon picking her up from work.  Duluth Model “explanations” for these behaviors has been to label them as “Power and Control.”  Dutton (1998) pointed out that the use of power and control was relationship-specific to batterers, and that people exercise control most when they are anxious and afraid.  The control of batterers is exercised because of a fear, the same anger “born of fear” that Bowlby described.   Because men often look to external causes of their discomfort, they assuage the fear and anxiety within themselves by controlling their partner, who is the perceived source of their anxiety. 

Although insight into attachment patterns is an important task in treating male batterers from an attachment perspective, the strong agent of change in this form of psychotherapy is the development of new strategies for coping with attachment related anxiety.  On a practical level, one immediate therapeutic objective is developing the ability to recognize an anxious reaction to loss and the ability to self-soothe.  However, because this ability should have developed through sensitive attunement by the attachment figure as a child, it now must also be learned through the attunement of an attachment figure such as a therapist.  The therapist must be that soothing voice until the client learns to find that voice within him or herself. 

In an group psychotherapy format, this could be established through the introduction of a topic such as “fear of losing her,” in which “abandonment” scenarios are described (e.g., you call and she’s not home, she’s late returning from work or shopping, or she pursues a job or hobby that takes more of her time).  It is possible to have men generate loss-fear diaries the same way they would generate anger diaries.  A discussion of the timing and frequency of daily contact might help establish a pattern: who initiates the contact?  Is it by phone?  How frequently does it occur?  What are the reactions to a failure to establish contact? 

In a more unstructured domestic violence therapy, the client will eventually bring in material where attachment or separation anxiety has been triggered and the therapist can be a soothing voice with a more objective perspective that helps the client learn to do similarly for him or herself.  It is also possible to structure systematic desensitization exercises to loss-fear in the same fashion as any other fear based cognitive-behavioral intervention (e.g., fear of flying), where an anxiety gradient is established with the most fear-inducing scenarios at the top, less serious at the bottom.  The client then visualizes the less serious scenarios and is taught relaxation techniques to extinguish the anxiety at the lower levels.  When these are mastered, the therapist proceeds to a more anxiety-producing level.

Clulow (2001) discussed working with insecure attachment in a couples therapy context.  In this context the focus is on establishing a secure base in the couples relationship.  Although couples therapy is not advisable in some domestic violence situations, attachment theory can provide a valuable perspective to understanding and treating domestic violence with couples as well as individual or groups.

The secure base relationship creates the safe container from which representational models of the client and his attachment figures can be explored. Bowlby (1988), in one of his last papers, outlined this surrogate task as follows:

A therapist applying attachment theory sees his role as being one of providing conditions in which his client can explore his representational models of himself and his attachment figure with a view to reappraising and restructuring them in the light of the new understanding he acquires and the new experiences he has in the therapeutic experience. (p. 138)

 The therapeutic alliance appears as a secure base, an internal object as a working, or representational, model of an attachment figure, reconstruction as exploring memories of the past, resistance as a deep reluctance to disobey the past orders of parents not to tell or not to remember.  (p. 151)

Exploring the Relationship with the Psychotherapist

In any ongoing psychotherapeutic process, the client may begin to consciously or unconsciously view the therapist as an attachment figure (Farber, Lippert, & Nevas, 1995).  If this indeed occurs, there is a great possibility that the attachment behavioral system will activate at various points in the therapy process.  Although talking about events and relationships outside of therapy is helpful, therapy from an attachment perspective must include, at some point, a discussion of the attachment dynamics between the therapist and the client - Bowlby’s third task for the attachment informed psychotherapist.  Psychotherapy may be viewed as common place for many people who have participated in the process, particularly for therapists who live and breathe the profession.  However, for most domestic violence clients, the act of entering a therapist’s office and disclosing private thoughts and feelings is likely to raise a degree of attachment-related anxiety.  Therefore, it is important that therapists pay close attention to their client’s verbal and non-verbal behaviors from the moment they make contact to begin to hypothesize how their particular attachment behavioral system is activated.

Most clients rarely readily admit to having feelings about their therapist, or at least being in therapy.  Their rational mind takes over and they tell themselves, “of course I feel comfortable with my therapist” or “why would I be here if I didn’t feel comfortable?”  In reality, however, it would be considered highly problematic if the client only had positive feelings while in therapy.  Not all clients will be able to directly confront their feelings about the therapy and therapist early in the therapeutic relationship.  Individuals with some attachment styles are not likely to admit that the relationship is significant, let alone admit that they have deep emotional reactions to the therapist.  Just as differential diagnosis guides the clinician about treatment planning and pacing, so does understanding a clients particular attachment style inform the attachment-oriented psychotherapist about how and when to address the therapeutic relationship with a particular client.

Addressing therapeutic relationship from an attachment perspective is important for a number of reasons.  First, it is through the intimate relating that occurs within the clinical hour that there is the opportunity to explore and hopefully change the representational models that determine a client’s attachment style.  Second, working with the client when feelings arise in therapy helps him/her find ways of regulating attachment anxiety and patterns of avoidance when attachment system is activated.  Viewing attachment from the perspective of anxiety and avoidance (Hazan and Shaver, 1987) suggests that changing attachment styles involves the client learning to regulate attachment anxiety and/or finding other means of expressing attachment needs other than through avoidance.  Lastly, there is some evidence that long term psychotherapy can affect the neuro-circuitry that gives rise to attachment related representations as well a emotion regulation (Perry, 1995; Vaughan, 1997). 

Regulating Attachment Anxiety when Attachment System is Activated

As mentioned above, the activation of the attachment behavioral system in the therapeutic hour can be the most effective way to address attachment anxiety with the client.  The distancing of the dismissing attachment style, the pleasing and idealizing behaviors of the preoccupied attachment style, and the erratic dependency and distancing of the fearful attachment style will eventually manifest in therapy in subtle and no-so-subtle ways.  When the therapist develops a secure base relationship with the client and the client has some of the above mentioned insight into his/her attachment relationships, the ground is set to address these behaviors as they manifest in the relationship with the therapist.  Through both the interpretation and confrontation of these behavioral manifestations of the activated attachment system, the client can learn to face the pain and vulnerability that underlie these defenses. This approach also allows the clients to understood and supported by the therapist, and eventually develop within themselves new skills in self-soothing, reassurance, and relaxation.  The net result is the client is able to reduce the reactivity and sensitivity to perceived cues of threatened safety or protection.

Much of what’s published in the domestic violence field speaks to this task of the attachment-oriented therapist.  Education, cognitive interventions, and behavioral therapy all focus their efforts at assisting the client (or student in the case of educational based programs) in learning new methods of coping with anger, conflict, or any emotionally difficult situations.  Although some programs address childhood abuse issues (Bowlby’s fourth task), it has been promulgated by leaders in the domestic violence field that it is more effective to focus on the here and now and less on childhood abuse experiences, which can be addressed later in the treatment process.  This mythology seems to contradict research that suggest otherwise (Saunders, 1996).  Saunders found that some batterers may actually improve faster by focusing on childhood abuse issues earlier in psychodynamically oriented treatment.  Additionally, Dutton’s (1998) research on male batterers suggests that for a significant percent of men, childhood trauma has led to borderline personality organization.  Thus, it appears that addressing childhood abuse issues is a necessary element of the treatment process.   Although cognitive and behavioral interventions are an important element in domestic violence treatment, they clearly are not sufficient given the fact that a significant percentage of persons who complete domestic violence treatment do seem to re-offend.  In addition, even though physical rates of violence do significantly decrease post treatment, psychological or non-physical violence do persist at relatively high rates (Rosenberg, 2001).  Dutton (1998) found a 21% arrest rate for an eleven-year follow-up, with partner interview violence rates at approximately 16%.  These data suggest that at least 20% of persons completing treatment will re-offend.  Higher rates of physical and non-physical violence have been found in other studies (Gondolf, 1997).

Taken together, this data suggests that the treatment programs developed to date may still be missing important elements necessary to long term cessation of physical and non-physical abuse.  Cognitive and behavioral interventions are necessary but not sufficient for long-lasting, successful treatment.  Treatment of domestic violence from an attachment-informed perspective may include the missing elements that can ultimately lead to lasting change with clients, manifested not only by the cessation of violence but also by a significant change in their experience of close relationships in general.

Dutton’s research suggests that batterers will present with all three types of insecure attachment styles in similar frequencies (Dutton, Saunders, Starzomski, & Bartholomew, 1994).  However, Fearful, and to a lesser extent Preoccupied, styles are correlated with partner’s reports of abuse.  These findings suggest that the subcategories/typologies of batterers are sufficiently different enough to justify therapists approaching treatment from an assessment based perspective, as opposed to using a cookie-cutter approach to treatment whereby all batterers are treated as if their violence has a single origin or etiology.  In addition, studies on drop-out rates of individuals in domestic violence treatment (Daly and Pelowski, 2000) suggest that one factor, psychopathology, may be related to this phenomenon.  Therapists who begin to recognize that they will need to vary their conceptualization and intervention with different clients may be able to reduce the drop-out rates in their treatment program.  Attachment informed psychotherapy recognizes that different attachment styles may need different therapeutic conceptualizations and interventions (Slade, 1999).

The Assessment of Adult Attachment Status

Numerous measures of adult attachment have been developed over the past ten years each with their own strengths and weaknesses (Crowell & Treboux, 1995).  Generally, these measures fall into two categories: self-completed questionnaires (questions or statements responded to with a Likert type scale) and those administered by a trained evaluator.  We would like to discuss three of these instruments because each one deconstructs attachment somewhat differently, and we believe that each method has clinical relevance to treating domestic violence clients. 

The Adult Attachment Interview.  Main and Goldwyn (1998) developed the Adult Attachment Interview (AAI), as system based on the structural qualities of narratives of early experiences.  The interview consists of eighteen questions about childhood experiences with attachment figures.  The trained evaluator is not so much interested in the content, as much as the coherence of the interviewee’s narrative.  Arieta Slade (1999) explains Main’s definition of coherence as the following:

For Main, the capacity to represent past experiences in a coherent and collaborative fashion is the most significant and compelling aspect of adult security, and is clearly the most predictive of infant security.  A coherent interview is both believable and true to the listener; in a coherent interview, the events and affects intrinsic to early relationships are conveyed without distortion, contradiction or derailment of discourse. The subject collaborates with the interviewer, clarifying his or her meaning, and working to make sure he or she is understood.  Such an subject is thinking as the interview proceeds, and is aware of thinking with and communicating to another; thus coherence and collaboration are inherently intertwined and interrelated. (p. 580)

While autonomous individuals value attachment relationships and are able to integrate memories into a coherent narrative, insecure individuals are poor at integrating memories of experience with the meaning of that experience.  Those persons classified as having a dismissing attachment style tended to deny negative memories, and idealize early relationships.   Their stories were very brief, general, and often full of contradictory data (e.g., describing negative experiences but talking about the parent in a positive light).  Preoccupied individuals tend to be preoccupied with childhood attachment experiences, often still complaining of childhood slights, echoing the protests of the resistant infant.  Their stories are often long and grammatically entangled with vague usages (“dadadadada,” or “and that”).  Unresolved individuals give indications of significant disorganization in their attachment relationship representation via either semantic or syntactic confusions in their narratives concerning childhood trauma or a recent loss (Fonagy, 1999).  These individuals show striking lapses in monitoring of reason or discourse (George, Kaplan and Main, 1996).

The relevance of the AAI to clinical work with batterers is that clinicians can listen to their client’s narratives from the beginning of treatment so as to begin to form hypotheses about attachment status.  Additionally, as the narratives begin to evidence certain forms of incoherence, the clinician can also strategize treatment interventions that specifically address the client’s defensive patterns that have led to the particular form of incoherence.  For example, for the avoidant or dismissing client who presents little data, idealizes their attachment experiences, and is unable to express affect, the therapist can begin to formulate strategies that help draw out the client’s story, listen for inconsistencies in their recollections of childhood experiences and begin to point them out, and slowly help the client connect with the emotional track of their narratives.  The pre-occupied client, whose narratives are convoluted and saturated with uncontained affect about attachment experiences, will need to learn how to better self-soothe so that their narratives will have a certain degree of objective distance or cognitive structures that contain the appropriate degree of affect.  With the fearful or disorganized batterer, the therapist will need to address the early childhood trauma experiences, whose resultant repressed affect leads to dissociation and other forms of maladaptive emotion regulation.  When treating domestic violence perpetrators, it would be our hope that as the client learns more about himself and his attachment relationships and becomes more effective at modulating attachment anxiety, his/her narratives will become more coherent.

Experiences in Close Relationships Questionnaire.  Brennan, Clark and Shaver (1998) developed the Experiences in Close Relationships (ECR) questionnaire, a self-report measure that assesses adolescent and adult romantic-attachment orientations (secure, anxious, and avoidant--the three patterns identified by Ainsworth, Blehar, Waters, & Wall, 1978 in their studies of infant-caregiver attachment).  They deconstruct attachment on two continuums: anxiety (need for approval, preoccupation with relationships, fear of being abandoned) and avoidance (discomfort with intimacy and closeness).  Persons with low anxiety and low avoidance are within the secure range.  Those with high anxiety and low avoidance are within the preoccupied range, while those with low anxiety and high avoidance are within the dismissing range.  Finally, persons with high anxiety and high avoidance are within the fearful range.  Clients can fill out the 36 questions fairly quickly.  The client is asked to read each statement and answer to what degree it reflects how they see themselves.  They can even take the test online and receive the results immediately (http://www.geocities.com/research93/).  Unlike the AAI, the ECR scores the person in degrees of avoidance and anxiety, and therefore is somewhat less categorical in nature.

The Relationship Questionnaire.  Another self-report adult attachment measure is the Relationship Questionnaire, developed by Bartholomew and Horowitz (1991).  This measure, although similar in form to Brennan, Clark and Shaver’s (1998), conceptualizes attachment in terms of internal working models of self and others.  This deconstruction of attachment is based on Bowlby's (1973, 1979) original conceptualization of attachment.  Bartholomew provides two theoretically unrelated dimensions giving four quadrants or categories.  Positive working models of the self and positive working models of others give rise to the secure attachment status.  Negative working models of the self and positive working models of others give rise to the preoccupied attachment status, while positive working models of the self and negative working models of others give rise to the dismissing attachment status.  Finally, negative working models of both the self and others give rise to the fearful attachment status. 

Understanding attachment from the internal working model perspective helps to explain many of the behaviors evident in perpetrators of domestic violence.  The pre-occupied client who is trying to please or receive validation from the therapist, or his partner, is avoiding experiencing the sense of defective self or self-hatred that would result from focusing on himself.  Addressing issues of self-esteem is critical with this client, whereas the avoidant client has learned to protect himself from others by distancing and may experience his partner or the therapist as intrusive and/or controlling and may act out violently or aggressively in retaliation.

An important issue being discussed among researchers developing these methods of measuring or identifying attachment styles is the notion of categorical typologies versus dimensions of security or insecurity.  In the real world, clients present with varying degrees of mental illness.   Therefore, it would be expected that attachment status would be no different.  The strength of the ECR and the Relationships Questionnaire is their use of the Likert-type scales that allow respondents to rate themselves in degrees of similarity or dissimilarity to each attachment related statement, rather than the categorical nature of the AAI.

Discussion of Adult Attachment Measures.  Each of these models of adult attachment (coherence, anxiety/avoidance, and internal working models of self/others) can be useful in understanding psychotherapy with perpetrators of domestic violence.  Although there is considerable overlap in how each of these attachment categories manifest interpersonally, they each suggest unique treatment goals.  Based on the AAI, the goal of therapy is helping the client reduce their anxiety sufficiently to reconstruct a coherent narrative of their attachment-related experiences, both in the past as well as currently.   As Jeremy Holmes suggests (2001), attachment based psychotherapy is a process of story-making and story-breaking.  One needs to break the rigid, unemotional, and unrelated story of the avoidant individual and create a story with greater emotional content, better balance of positive and negative experiences, and a more descriptive and realistic narrative description of relationships.  With the pre-occupied individual, one must break the emotional dysphoria by creating one that is also infused by logic and perspective and balance of affect and reflective understanding.

Similarly, the Brennan, Clark, and Shaver (1998) model suggests that by learning to self-soothe attachment anxiety and find other mechanisms beside avoidance to deal with the fear and vulnerability that can be activated within close relationships, clients can begin to develop more secure relationship experiences.  The Bartholomew and Horowitz (1991) model suggests that working more on improving self esteem and reassessing feelings of distrust and fear of others will ultimately allow the client to experience relationships from a secure perspective.

There is some question as to whether or not Bowlby's concept of "internal working models" is the same as attachment styles described in the current literature.  At a recent meeting of the American Psychological Association, Adult Attachment Discussion, the issue of working models attachment styles or attachment representations were explored among researchers in the field and the following was noted (Adult Attachment Lab web site, 1998):

There was some initial disagreement over the use of the terms "working models," "attachment styles," and "attachment representations." It was generally agreed that the term "attachment style" is best reserved for describing observable or manifest patterns of behavior, and the term "working models" is best reserved for describing the latent mental structures giving rise to variability in attachment styles.

It was suggested that the concept of working models is of relatively little use in describing the psychological dynamics of attachment because the concept brings to mind conscious-evaluative belief systems (positive/negative models of self/others) operating with little input from motivational and defensive goals or over-learned strategies of behavioral and emotional regulation. In contrast, but also speaking to the limitation of the concept of working models, it was suggested that the concept was broad enough to refer to both declarative and procedural aspects of cognition and behavioral/emotional regulation.

It was generally agreed that the concept of working models is most useful when referring to organized strategies for regulating emotion, attention, and behavior with respect to attachment concerns. It was also suggested that a number of social-cognitive techniques exist that can be exploited to investigate the procedural and unconscious aspects of working models."

Main (1999) notes that there is research that suggest there are in fact neurological correlates to internal working models as either neurological circuits or patterns that are ingrained from experience or a function of working memory.  In either case, understanding the neurological basis of internal representations of self and others may be an important element to understanding attachment patterns in children and adults.


Psychotherapy with perpetrators of domestic violence from an attachment perspective involves creating a secure base environment so that clients can explore their current and past attachment relationships within the safety of the therapeutic relationship.  Safety is critical, because many insecure batterers have experienced tremendous loss, hurt, and disappointments within their close relationships; they therefore enter therapy with fears and anxiety about opening up to someone who is perceived as having power over them.  This is particularly true for the court-mandated client, where the therapist may indeed have a great deal of influence over their criminal justice experience.  For these reasons, creating a secure base environment is a critical first step to achieving therapeutic goals, such as learning emotional self regulation or resolving childhood trauma.  Another critical element to attachment oriented psychotherapy with perpetrators of domestic violence is the "not-one-size-fits-all" maxim.   Different attachment styles need different interventions and approaches.  The batterer with the overly structured dismissing attachment style needs to connect to their emotional life and acknowledge the importance of attachment in their lives.  They need to learn that attachment relationships do not need to be exploitative, hurtful, controlling, or rejecting.  The batterer with the preoccupied attachment style needs structures necessary to contain their emotional reactivity in attachment interactions, while learning greater self-sufficiency and less dependency on attachment figures for self-definition and security.   The batterer with the fearful attachment style likewise needs to heal the split that exists within them from childhood trauma and losses so that they can both learn to self-soothe their attachment anxiety through means other than avoidance or pushing others away through anger and violence.

Using the therapeutic relationship (or peer relations when utilizing a group intervention modality) is the most powerful means to highlighting attachment behavioral system patterns in psychotherapy.  Through these in-the-moment experiences, therapists can help raise the client's awareness of these patterns, but most importantly strategize more adaptive responses to attachment anxiety.   This process takes therapists out of their heads and challenges them to work within the here and now with clients.  Quick thinking, self-awareness, and sensitive attunement to the client are critical to making use of these "now-moments" (Stern, 1998).  Making use of them on a continual basis gives the client the message that he/she and the therapist can go to that frightening place of emotions and the meaning of intimacy. 

Because domestic violence clients are a heterogeneous population, clinicians are likely to encounter all three insecure attachment styles.  An assessment of the client's attachment status is necessary to understand how the client's attachment behavioral system activates and the mechanisms they use to cope with the anxiety associated with attachment.   Understanding the client’s attachment status helps us to form some hypotheses about the etiology of the client’s violence.  Psychotherapy offers the domestic violence client the opportunity to learn more adaptive methods of regulating attachment anxiety, reevaluate internal models of self and others, and experience intimate relationships (with the therapist or fellow group members) in new and positive ways.  Through long term exposure to these therapeutic experiences, changes in the internal working models and attachment style is not only possible but inevitable.

Although John Bowlby began his work on attachment theory over fifty years ago, there are still varying ideas about how one approaches psychotherapy from an attachment perspective.  Unlike most clinical theories, attachment theory has had the benefit of more than forty years of empirical research before discussions even began on the clinical applications to adult psychotherapy.  So as the clinical application of this theory evolves, clinicians will have at their disposal a continually growing body of empirical data that will hopefully meld with clinical experience.  Through a positive attachment between clinicians and academics, the application of this theory will unfold in the years to come.


Adult Attachment Lab. (1998).   Notes from the Adult Attachment Discussion Session at the 1998 meeting of the American Psychological Association. Retrieved March 17, 2002, from http://psychology.ucdavis.edu/Shaver/apasum.html.

Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, N.J.: Erlbaum.

Arndt, J., Greenberg, J., Pyszczynski, T., & Solomon, S.  (1997).  Subliminal exposure to death- related stimuli increases defenses of the cultural worldview. Psychological Science, 8, 379-385.

Baldwin, M. W., Keelan, J. P. R., Fehr, B., Enns, V., & Koh Rangarajoo, E. (1996). Social-cognitive conceptualization of attachment working models: Availability and accessibility effects. Journal of Personality and Social Psychology, 71, 94-109.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment  perspective,  Journal of Social and Personal Relationships, 7, 147 –158.

Bartholomew, K., Henderson, A., & Dutton. D. G. (2001). Insecure attachment and abusive intimate relationships. In C. Clulow (Ed.), Adult attachment and couple psychotherapy (pp. 43-61) Taylor & Francis: Philadelphia.

Bartholomew, K., & Horowitz, L. M. (1991) Attachment styles among young adults: A test of a four category model. Journal of Personality and Social Psychology, 61(2), 226 – 244.

Bowlby, J. (1969).  Attachment and loss: Vol. 1. Attachment (2nd Ed). London: Hogarth Press.

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

Bowlby, J. (1980).  Attachment and loss: Vol. 3. Loss, sadness, and depression. New York:  Basic Books.

Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.

Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46-76). New York: Guilford Press.

Browning, J. J., & Dutton, D. G.  (1986).  Assessment of wife assault with the conflict tactics scale: Using couple data to quantify the differential reporting effect.  Journal of Marriage and the Family, 48, 375-379.

Cassidy, J. (1999). The nature of the child’s ties. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355-377). New York: Guilford Press.

Clulow, C. (2001). Adult attachment and couple psychotherapy.  Brunner- Routledge: London.

Crowell, J. A., & Treboux, D. (1995) A review of adult attachment measures: Implications for theory and research. Social Development, 4. 294-327.

Daly, J. E., & Pelowski, S. (2000).  Predictors of dropout among men who batter: A review of studies with implications for research and practice.  Violence and Victims, 15(2), 137- 160.

Dozier, M., Stovall, K. C., & Albus, K. (1999). Attachment and psychopathology in adulthood. In J. Cassady & Phillip Shaver (Eds.), Handbook of attachment theory: Research and clinical implications (pp. 497-519).  Guilford: New York.

Dutton, D. G. (1998). The abusive personality.  Guilford: New York.

Dutton, D. G., & Kerry, G. (1999). Modus operandi and personality disorder in incarcerated spousal killers.  International Journal of Law and Psychiatry, 22(3-4), 287 –300.

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

Farber, B., Lippert, R., & Nevas, D. (1995).  The therapists as attachment figure.  Psychotherapy, 32(2), 204-212.

Feeney, J. (1999).  Adult romantic attachment and couple relationships. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355-377). New York: Guilford Press.

Fonagy, P. (1999). Psychoanalytic theory from the viewpoint of attachment theory and research.  In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355-377). New York: Guilford Press.

George, C., Kaplan, N., & Main, M.  (1996).  Adult Attachment Interview Protocol (3rd Ed.) Unpublished manuscript, University of California at Berkeley.

Gondolf, E. (1997). Patterns of reassault in batterer programs.   Violence and Victims, 12(4),373-387.

Hazan, C., & Shaver, P. (1987). Conceptualizing romantic love as an attachment process. Journal of Personality and Social Psychology, 52, 511 –524.

Herman, J. L. (1992). Trauma and recovery. New York: Basic Books.

Holmes, J. (2001).  In search of the secure base. London: Routledge.

Holtzworth-Monroe, Amy; Meehan, Jeffrey; Herron, Katherine; Rehman, Uzma; and Stuart, Gregory (2000).  Testing the Holtzworth-Monroe and Stuart (1994) batterer typology.  Journal of Consulting and Clinical Psychology. Vol 68,  No. 6,  pp 1000-1019.

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

Karen, R.  (1977). Becoming attached: Unfoding the mystery of the mother-infant bond and its’ impact on later life.  Warner Books: New York.

Main, M. (1999).  Attachment theory: Eighteen points with suggestions for future studies. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355-377). New York: Guilford Press.

Main , M., & Goldwyn, R. (1998). Adult attachment classification system. Unpublished manuscript. University of California: Berkeley, CA.

Main, M., & Weston, D. (1981).  Quality of attachment to mother or father: Related to conflict behavior and the readiness for establishing new relationships. Child Development, 52, 932 – 940.

Masterson, J. F., & Klein, R. (Eds.). (1995). Disorders of the self: New therapeutic horizons. New York: Bruner Mazel.

Mikulincer, M. (1998). Adult attachment style and affect regulation: Strategic variations in self-appraisals. Journal of Personality and Social Psychology, 75, 420-435.

Mikulincer, M., & Arad, D. (1999). Attachment, working models, and cognitive openness in close relationships: A test of chronic and temporary accessibility effects. Journal of Personality and Social Psychology, 77, 710-725.

Milkulincer, M., & Shaver, P. (2001).  Attachment theory and intergroup bias: Evidence that priming the secure base schema attenuates negative reactions to out-groups.  Journal of Personality and Social Psychology, 81, 97-115.

Perry, B. (1995). Incubated in terror: Neurodevelopmental factors in the “Cycle of Violence”. In J.D. Osofsky (Ed.), Children, youth and violence: Searching for solutions (pp 124-148). Guilford: New York.

Rosenberg, M. (2001). Domestic violence in Sonoma County: An outcome study of court-mandated offenders. Unpublished manuscript.

Schore, A. N. (1994).  Affect regulation and the origin of the self : The neurobiology of emotional development. Erlbaum: Hillsdale, N.J.

Slade, A. (1999).  Attachment theory and research: Implications for the theory and practice of individual psychotherapy with adults.  In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355-377). New York: Guilford Press.

Saunders, D. (1996).  Feminist-cognitive-behavioral and process-psychodynamic treatments for men who batter: Interaction of abuser traits and treatment models.  Violence and Victims, 11(4), 393-414.

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

Sperling, M. B., & Berman, W. H. (1994). Attachment in adults: Clinical and developmental perspectives. Guilford: New York.

Stern, D. (1998).  The process of therapeutic change involving implicit knowledge: Some implications of developmental observations for adult psychotherapy.   Infant Mental Health Journal, Volume (19), 300-308.

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

Van der Kolk, B., A., McFarlane, A. C., & Weisaeth, L.  (Eds.) (1996). Traumatic stress: The effects of overwhelming experience on mind, body, and society. New York: The Guilford Press.

Vaughan, S. (1997).  The talking cure: Why traditional talking therapy offers a better chance for long-term relief than any drug.  New York: Owl Books.

Walker, L. E. A.  (1979). The Battered Woman.  Harper & Row: New York.

West, M., & George, C. (1999). Violence in intimate adult relationships: An attachment theory perspective. Attachment and Human Development, 1, 137-156.

[i] [i] From his early wiritings, Bowlby used the terms “principal attachment figure” or
“mother figure” rather than the word “mother.”  This usage underscored his belief that although the principal attachment figure is often the mother, it may also be another person (such as father or grandparent).  For the purposes of this chapter, we shall use the term “principal attachment figure” keeping in mind that although for most of our patients this is the biological mother, it is not necessarily so.