The link between childhood trauma and later violent offending

Research findings relating to young offenders show a history of maltreatment and loss in up to 90% of the sample population (Boswell, 1996; Fonagy, Target, Steele et al., 1997). These findings accorded with my clinical experience when I worked as a probation officer with adult offenders in the community.

In particular, I found that those who had committed violent offences had themselves been victims of childhood abuse and/or suffered neglect or loss experienced as catastrophic. Indeed, the acting out of unresolved childhood trauma in a criminal way was a consistent feature in the behaviour of those with whom I worked and, moreover, was strongly associated with substance misuse.

The study that follows illustrates the clinical application of attachment theory in a probation setting. It is presented as an example of the work I undertook with violent offenders and with the intention of emphasising the connection between childhood trauma and subsequent violent offending. It elucidates the way in which attachment theory may be used to explicate offending behaviour and to assess risk in a forensic setting.

In line with Boswell’s (1998) advocacy of research-minded practice, the study seeks to demonstrate the importance of asking offenders about their traumatic backgrounds at the point of assessment. The therapeutic model shows that the application of attachment theory in brief, time-limited work may enhance the offender’s capacity for narrative intelligibility, leading to an integration of dissociated thoughts and emotional affect and to a concomitant cessation of violent behaviour.

Theoretical paradigm

Attachment theory played a central part in my assessment of the offender and I anticipated that this approach would underpin my intervention with him. As indicated above, I found attachment theory a powerful tool in explicating offending behaviour and assessing risk and, furthermore, eminently adaptable to working effectively with offenders in a probation setting. From this developmental perspective the person’s inner world of subjective experience is structured, shaped and organized by patterns of attachment and interpersonal interactions into representational models (Bowlby, 1969, 1980).

With regard to traumatic childhood experiences involving separation and loss, Bowlby (1969) found that when a young child is unwillingly separated from the attachment figure, he or she shows distress. In the event of the separation being prolonged, necessitating the child being placed in unfamiliar surroundings, such distress is likely to become intense.

Typically, the child’s distress follows a sequence of protest, despair and emotional detachment. Bowlby (1969) suggests that these phases may be linked to three types of responses, viz., separation anxiety, grief and mourning, and defence. Further, he argues that these responses are phases of a single process - that of mourning separation and loss. The traumatic quality of the child’s grief reaction is encapsulated in Bowlby’s poignant observation that “Loss of a loved person is one of the most intensely painful experiences any human being can suffer” (Bowlby, 1980, p. 7).

Bowlby (1979, 1980) emphasises that the crucial process of mourning generally takes place in the context of the family’s characteristic attachment behaviour towards the child. He contends that the family may either facilitate the expression of grief by responding sympathetically to the child’s distress or adopt an inhibiting attitude that causes the child to suppress or avoid typical feelings of fear of abandonment, yearning and anger. Bowlby (1979, 1980) stresses that a supportive and sympathetic attitude within the family may lead to a process of healthy mourning in children as young as two years.

The process consists of normal behavioural responses of anxiety and protest, despair and disorganization and detachment and reorganization. By means of this process, the loss is gradually accepted by the child whose capacity to form new attachment bonds is restored following a period of disorganization.

By contrast, in pathological mourning the child’s unexpressed ambivalent feelings of yearning for and anger with the attachment figure are split off into segregated or dissociated systems of the personality, and the loss may be disavowed. As a consequence, and in the absence of a trusted substitute attachment figure, the child has little alternative but to move precipitously to a defensive condition of emotional detachment, thereby internalizing a mental model of attachment that is dismissing or avoidant of affective states associated with separation and loss. In such instances, the child’s attachment behavioural system remains deactivated because attachment-related information is being defensively and selectively excluded from consciousness (Bowlby, 1980, 1988).

In describing childhood pathological mourning, Bowlby (1979) makes the important point that his hypothesis is not confined to the actual death of or separation from the attachment figure. Indeed, he stresses that the child may experience separation and loss in numerous, less overt ways, for example, in the form of threats of abandonment, parental rejection, depression, neglect and/or abuse, as well as loss of love (Bowlby, 1979, 1988). Bowlby (1979) emphasises that the common factor in these various situations is loss by the child of a parent figure to love and to attach to.

In developing Bowlby’s theoretical concepts, Main, Kaplan and Cassidy (1985) suggest that patterns of secure and insecure attachment organization, internalized in the form of working models, are representational of states of mind in relation to patterns of attachment. Further, Main et al.’s (1985) research indicates that, once established, patterns of attachment tend to persist over time and become actively self-perpetuating because information experienced as potentially disruptive is countered by perceptual and behavioural control mechanisms. Internal working models are, therefore, thought to mediate experience of actual relationships and events, and to guide and direct feelings, behaviour, attention, memory and cognition.

The authors’ findings support Bowlby’s (1979) contention that mental models shaped by childhood experiences of pathological mourning may be activated under conditions of separation and loss in adulthood, together with the expression of dysfunctional anger, hatred and aggression.

In the course of their research, Main et al. (1985) employed the Adult Attachment Interview (George, Kaplan and Main, 1984) in order to classify parental states of mind with respect to attachment. Using this research tool and Ainsworth, Blehar, Waters and Wall’s (1978) Strange Situation procedure, which observes and classifies the attachment status of children, Main et al. (1985) established a link between four distinct discourse styles and four corresponding patterns of attachment behaviour.

These findings were confirmed by Main in a follow up study in 1991, which, together, demonstrate that each discrete pattern of attachment organization has, as its precursor, a specific pattern of caregiver-infant interaction and its own behavioural sequelae. Again, this research suggests that malignant childhood events relevant to attachment, such as separation and loss, may cause difficulty in integrating and organizing information, and that such difficulty may play a determining role in the creation of security in adulthood.

As noted above, Main et al. (1985) conclude that internal working models derived from insecure patterns of attachment organization are resistant to change because error-correcting information is being defensively and selectively excluded from consciousness, resulting in perceptual distortion of relationships and events. The authors’ findings of research undertaken in Baltimore have been replicated in studies carried out in Germany by Grossman and Grossman (1991).

Along similar lines, Peterfreund (1983) suggests that different internal working models are in operation during different activities and in different situations making predictive calculation and adaptive behaviour possible. In advocating a “heuristic”, as opposed to a “stereotypical”, approach to the process of psychoanalytic therapy, he, too, stresses the significance of information processing and error-correcting feedback in this process, arguing that these are the means by which perceptually distorted internal working models are modified, updated and fine-tuned.

Peterfreund’s synthesizing approach reflects Bowlby’s emphasis both on empirical observation of human relationships and the fact that many of the concepts underpinning attachment theory are derived from cognitive psychology and developmental psychology. Attachment theory, therefore, may be seen as acting as a bridge between cognitive science and psychoanalysis (Holmes, 1993).

In line with this thinking, my work with offenders was informed by findings from developmental studies, adult attachment research, trauma research and neurobiology. In combination with interactional and developmental perspectives in psychoanalysis, such findings provide both a particular way of listening to the offender’s narrative and of understanding the clinical process (Slade, 1999).

In accordance with this view, Stern (1998) argues that “search strategies” which explore the client’s past are an integral aspect of the therapeutic process, contending that “In good part, the treatment is the search” (p. 203). As with attachment theory, Stern’s (1985, 1998) perspective views psychopathology as arising out of an accumulation of maladaptive interactive patterns that result in character and personality types and disorders in adulthood.

The offender - John: personal history 

John, the subject of this case study, has given his permission for this paper to be published. Names have been changed, however, and personal circumstances disguised in order to protect identities. John is 48 years old and grew up in a large family, being one of eight children. He is the youngest of four brothers, one of whom died several years ago, and has two older and two younger sisters. He spoke of his father as being “distant and always at work”, and his mother as “over-protective”, recalling that she had played out an elaborate pretence in respect of his father’s occupation by telling neighbours he worked in a bank, whereas, in fact, he was a barman.

John completed his secondary education at the age of 15, leaving school with no exam qualifications. By this time, he was misusing illicit drugs and alcohol. He went on to develop a dependency on the latter. As a consequence of this problem, John’s employment record is inconsistent and, in the main, comprised of manual and semi-skilled work.

John has had a series of unstable relationships with women characterized by violent, controlling behaviour, possessiveness and sexual jealousy on his part. Because of his problematic attitude to women, John consulted his GP when aged 18. He was referred for psychiatric assessment but not offered ongoing treatment. John married when aged 28, but insisted “there never was a true love”, adding “I haven’t wanted to commit myself”. He avoided doing so in part by “always having relationships with two women at the same time”.

This situation obtained during the course of his 13-year-long marriage, which John described as an “on-off affair”. He related how he would often pick fights with his wife so as to give himself an excuse to leave home and go on a drinking binge. The marriage was childless, but John has three children from a subsequent relationship, which, typically, was brief, intermittent and volatile. He has had no contact with his children for several years and was unaware of their current whereabouts. At the time we met, John was largely estranged from his own family and not in an intimate relationship.

Forensic history 

John has been involved with the criminal justice system for over 30 years, appearing before the courts for the first time as a juvenile. Though he has convictions for motoring offences and, when younger, burglary, the most prominent and consistent feature of his offending behaviour is drink-related violence. The latter commenced in adolescence and, as mentioned above, was the reason why John was eventually referred for psychiatric assessment.

He has convictions for grievous bodily harm, assault with intent, assault on the police, possession of a firearm and criminal damage, on one occasion going to his ex-partner’s home armed with an axe, which he used to break in. John has been subject to a range of sentences including discharges, fines, probation, community service, and imprisonment. He has had numerous sojourns in rehabilitation units for his alcohol problem but always returned to misusing drink. One of his brothers also has an alcohol problem but John is the only member of the family to become embroiled with the law.

The index offence 

The index offence consisted of a serious assault on John’s partner, Sylvia. The couple had been in a relationship for 2 years but lived separately. John came to suspect Sylvia of being sexually involved with someone else. He went to her home in a drunken state one evening and accused her of having sex with another man, calling her a “slag and a whore”.

When Sylvia denied John’s accusation he attacked her with his fists and feet in a blind, uncontrollable rage causing serious injury to her head and body, desisting only when finally she told him “what I wanted to hear”. At court, the photographic evidence of Sylvia’s injuries was said to be “horrific”. John denied the offence when arrested, maintaining that Sylvia’s injuries were self-inflicted. He was convicted following a jury trial and sentenced to 2½ years imprisonment.

First contact

John’s case was allocated to me when I transferred to the probation office in his home area. His reputation at the office was that of a perpetual offender with whom everything had been tried. I wrote to John in prison to introduce myself as his new throughcare officer. In his reply, he alluded to the attack on Sylvia, saying “It wasn’t anger, it was alcohol talking, I’m not angry by nature”. Though clearly John was distancing himself from his anger and violence in this statement, there was at least an implicit admission of his assault on Sylvia.

The initial assessment 

I met John for the first time during his temporary release from prison on home leave. He had managed to retain his local authority tenancy by sub-letting to a male alcoholic friend, but previously had lived alone. John was due to be released on parole licence a month later and his period of supervision would run for 8 months. At this first meeting, I asked John specific questions about childhood experiences in respect of separation, loss and abuse. He was clearly surprised and puzzled by the tenor of my questions as he had not been asked about such issues before.

After some initial hesitation, John spoke of having had frequent separations from his family from about the age of 5 years. These were the result of a series of operations for ENT problems necessitating his hospitalization. He recalled struggling with the nursing staff on one occasion as he fought to retain consciousness whilst being held down and given “gas”.

Despite these traumatic experiences, John’s discourse style when discussing them was dismissive, in that he did not believe they had had an adverse effect on him. The dismissive quality of his narrative, together with his propensity for violence in intimate relationships, indicated that he might have developed an insecure-avoidant pattern of attachment organization (Main, 1991; Main et al., 1985; Main and Weston, 1982).

Given these clinical features, I held in mind the possibility that John may have responded to the enforced separation from his family by precipitously entering a state of emotional detachment (Bowlby, 1973, 1979). In reviewing studies linking insecurely attached children and subsequent criminal behaviour, Fonagy et al. (1997) suggest that insecure attachment constitutes a distinct risk factor. Further, Fonagy et al. (1997) argue that patterns of attachment operate as mechanisms of defence to help the child cope with idiosyncrasies of parental caregiving and that criminality involves disturbance of attachment processes. These findings accord with de Zulueta’s (1993) proposal that “violence is attachment gone wrong”.

It soon became clear that John’s ideal view of himself was that of a passive, non-violent man who, in his own words, “wouldn’t hurt a fly”. My tentative hypothesis at this point was that John was carrying powerful unprocessed emotional pain; that he was disowning feelings of anger and hatred and that, lacking the capacity to contain and transform such emotions, these built up in response to stressor events, generating intense internal conflict which eventually became overwhelming.

At such times, John resorted to binge drinking. Under the disinhibiting effect of alcohol his split-off, dissociated emotional turmoil was unleashed and acted out in the form of a violent rage. This clinical picture indicated that John might be prone to experiencing a traumatic stress reaction when embroiled in an intense, emotionally-charged situation (de Zulueta, 1993; Herman, 1992). From this perspective, traumatic affect is seen as having a disorganising effect on mental functioning, and to be a significant motivating factor in the manifestation of violent behaviour (Tyson and Tyson, 1990).

I harboured reservations about John’s ability to engage in a therapeutic process. These misgivings centred on the fact that he was denying the index offence and that his record of attending appointments when supervised in the past had not exactly been exemplary. Further, as noted above, John was resistant to the idea that past experience may have a maladaptive effect on behaviour in the present, specifically in relation to his alcohol misuse, as he had been told at a rehabilitation unit that “alcoholism is a disease”.

He therefore expressed a good deal of scepticism about the prospect of change, having passively accepted this fatalistic diagnosis. Nevertheless, I explained what our work together would involve, should he decide on this option, emphasising the collaborative nature of the process. John responded by saying that he would “give it a go” as nothing else he had tried had been successful. He signed a standard medical consent form giving me permission to contact his GP in order to discuss any relevant issues.

Time framework

Given the setting within which I then worked my intervention with John would be brief and time-limited. Hence speed in assessing the clinical issues was a major consideration. In fact, I had a total of 13 sessions with John, each session lasting an hour. A follow-up meeting was held 6 months after his supervisory period had ended. I incorporated such meetings into my practice, as I saw these as serving a dual purpose of evaluating the effectiveness of my work and providing the offender with a sense of continuity and connection to a secure base (Bowlby, 1988), or at least to one experienced as secure-enough.

In my experience, the availability of an ongoing link at this critical time helped to preclude the often noted (though anecdotal) phenomenon whereby the offender re-offended towards the end of the supervisory period, seemingly in reaction to the loss of a relationship that had become significant.

Therapeutic intervention

The first meeting following John’s release from prison focused on helping him to recognise and own disturbing thoughts and feelings. An example of this difficulty arose when John spoke in mild terms about the friend whom he had allowed to stay at his home whilst he was in prison. John returned to find the place a complete tip and rent arrears of over £1000 owing to the housing department.

At first John spoke of feeling “a bit let down” and, later, when I questioned his passive response, of being “angry and annoyed”, vacillating between these two attitudes. It seemed to me that John was quite confused as to how he actually felt about his friend and in two minds about how to respond, speaking in the same breath of going to reason with this person and of beating hell out of him!

This narrative appeared to provide a glimpse of the conflict and disorganisation characterizing John’s representational world of object relations or confused, unstable internal working models of attachment. On the one hand, he seemed to be identifying with the hurt, angry, disappointed child who had been let down and whose trust had been betrayed; on the other, to be identifying with a dismissing parent who deflected and perhaps even forbade the expression of difficult thoughts and painful feelings.

This situation seemed to be re-created in the session, in that John anxiously deflected any attempt on my part to connect with him on an emotional level. Indeed, I felt under immense pressure not to talk about meaningful issues and events and I experienced a sense of futility and despair.

It would have been all-too easy to have succumbed to the sense of hopelessness that I was experiencing and thus given up the attempt to engage John. Instead, I sought to understand his emotional state and subjective experience. In ways not too dissimilar from the stress-inducing aspects of the Strange Situation procedure (Ainsworth et al., 1978) and the Adult Attachment Interview (George et al., 1984), I viewed the emotionally heightened exchange with John as having elicited archaic interactive patterns of attachment.

These mental models took the form of non-verbal, non-reflective procedural memories, and were expressed in his behavioural performance, the observation of which provided a basis for me to experience, share and match his affective state (Beebe, Jaffe and Lachmann, 1992; Beebe and Lachmann, 1992; Stern, 1985, 1998).

Despite the intensity of the interaction, my reading of John’s overt behaviour influenced me to stay in the affective moment, which I viewed as an unconscious communication of unmanageable feelings. I decided to share aspects of my countertransferential experience of being with him, wondering whether the powerful thoughts and feelings stimulated in me mirrored something of his own experience (Casement, 1990; Maroda, 1991).

John confirmed that he had felt a mounting sense of anxiety, verging on panic, adding that he usually avoids talking about his feelings. Avoidance of this sort, particularly in men, may reflect the way in which gender, culture and inner prohibition coalesce, resulting in a defensive splitting of thought from feeling. From a developmental perspective, such behaviour may also indicate a failure of interactive affective development and a concomitant incapacity to self-regulate emotional states when under stress (Schore, 1994). Research suggests that in such instances, the lack of a contingent parental response to the child’s attachment signals may, if characteristic of the relationship, come to be associated with negative affect and escalating arousal, leading to prolonged and severe states of withdrawal. The internalization of such interactive patterns may interfere with the person’s optimal regulation of arousal and thus compromise his or her capacity to stay attentive and process information (Beebe, Jaffe and Lachmann, 1992).

Disclosure of childhood trauma

Somewhat paradoxically, John appeared relieved by the dawning realization that inner emotions may be recognized, shared and understood (Benjamin, 1992). Seemingly in consequence of this intersubjective experience, a more reflective mood and positive affective state prevailed. This exchange, in turn, appeared to evoke in John memories of a traumatic event that had taken place when he was aged 8. Tentatively, John related how he and his then best friend, Ricky, had been playing near a fast-flowing river. John’s memories of the event were somewhat vague and hazy, but he recalled that Ricky had slipped on the moss-covered embankment into the river and drowned. John came to believe that people suspected him of having pushed his friend into the river. Indeed, I found myself silently questioning whether John might have had a hand in Ricky’s death.

Again, I observed the nuances of his facial expressions, direction of gaze, vocal inflections, bodily orientation and gesture when discussing this traumatic event, as well as monitoring my own affective and bodily responses. I detected nothing in John’s overt behaviour at that time, or subsequently, to indicate that his narrative was anything other than authentic in regard to this matter. I therefore concluded that Ricky’s death had, indeed, been a tragic accident.

In addition to feeling blamed and accused, John came to view himself as a “bad” person because his attempts to save Ricky had failed. John went on to speak of having confused and intangible memories of being in court in the aftermath of Ricky’s death and of growing up feeling burdened by “guilt”.

The court in John’s memory was probably that of the coroner who carried out the inquiry into the circumstances of Ricky’s death. It seemed likely that, in a similar way to those who live through a major disaster, John experienced a deep sense of guilt at having survived when Ricky had died (de Zulueta, 1993; Herman, 1992) and that this whole situation was exacerbated by his having to appear at the coroner’s court.

I wondered whether the trauma of Ricky’s drowning had activated John’s earlier trauma, that of being separated from his family and held down and “gassed” in hospital - drowning in gas, as it were. This observation seemed to resonate with John’s subjective experience. He became deeply thoughtful and reflective, sitting in silence for a considerable time. He looked sad and forlorn and his eyes brimmed with tears.

When he surfaced from this pensive mood he appeared to recognise aspects of himself as if for the first time. He spoke about persistent feelings of sadness, anxiety and watchfulness, and he questioned whether these could be linked to his disturbing childhood experiences. It seemed that the recollection of these state-dependent memories had started the process of unlocking the affective components of John’s unresolved trauma (Stern, 1985). Despite the similarity in our ages, my primary countertransference at this point was that of a benign, concerned parent seeking to understand and ameliorate a child’s confused state and emotional distress.

Although the session had been challenging and intense, John seemed buoyed up and expressed the hope that ghosts could finally be laid to rest. His positive affective response and disclosure of unresolved childhood trauma suggested that a secure-enough therapeutic alliance had been speedily established.

Repetition of the trauma

During subsequent sessions, John and I tried to give meaning to what, in symbolic terms, he was acting out unconsciously by means of his offending behaviour. It seemed to me, at least in part, that he was re-enacting a destructive and self-destructive pattern of behaviour in identification with the “bad”, traumatized 8-year-old child who had been unable to mourn Ricky’s death and was left carrying a tremendous burden of shame and guilt.

My hypothesis was that an aspect of this re-enactment involved John being drawn compulsively and repetitively to stand accused in the dock of a court, thereby reliving the trauma and, at the same time, confirming his negative core assumption or fantasy of himself as a bad, guilty person. Moreover, I wondered whether the experience of being adjudged guilty and sentenced to a period of incarceration had the temporary effect of assuaging John’s deep and pervasive sense of shame.

This hypothesis rang true for John and became a key therapeutic metaphor in our work together (Stern, 1985). Further, the co-construction of significant events in John’s childhood seemed to go some way towards filling gaps in his personal history by beginning to provide his fragmented experience with coherent narrative meaning (Holmes, 1996; Main, 1991; Main et al., 1985). John elaborated on these thoughts, saying that he feels safe and secure in prison, whereas on the outside he is continually assailed by feelings of panic, anxiety and an impending sense of danger, as if something dreadful were about to happen. In phenomenological terms, it would seem that John’s unresolved trauma was experienced as “a fear of a breakdown that had already happened” but had not been “remembered”, and thus was prone to being repeated at an unconscious level of mental functioning (Winnicott, 1974, p. 104).

Continuous assessment and ongoing intervention

At this point in my work with John my assessment had crystallized. Keeping relevant research findings in mind, I based my assessment of John on the theoretical premise that cognitive-affective states associated with his traumatic experiences had been subject to perceptual distortion, defensive exclusion and selective inattention (Bowlby, 1980; Main et al., 1985). In line with Herman (1992), I surmised that the lack of an appropriate response to John’s trauma had left him with a pervasive sense of alienation and disconnection in his relationships.

I concluded, therefore, that the main therapeutic task was to facilitate a process of mourning by assisting him to make connections between dissociated thoughts and feelings associated with the traumatic events he had described (Bowlby, 1973, 1979, 1988; Spezzano, 1993). This work had, of course, already commenced to some extent during the assessment process, reflecting Stern’s (1998) contention, alluded to above, that the search process is, in itself, therapeutic.

By this stage, John seemed committed to working on these unresolved issues and he stuck doggedly to the task, appearing to have an active need to tell his story and create a narrative. He admitted to being desperate for a drink after the previous session, but told me that, instead, he had made a conscious effort to think about what we had discussed. In line with Main’s (1991) research into metacognitive monitoring, I had enjoined John in quite a directive way to develop a dialogue with himself.

This involved using thought and his mind in a new and novel way so as to contain and assimilate raw psychic pain; of stepping outside of himself in order to observe and monitor his thoughts, feelings and behaviour. Following Fonagy et al.’s (1997) development of Main’s (1991) research, these strategies were designed to enhance John’s reflective functioning, thereby increasing his capacity to contemplate and understand (mentalize) both his own and others’ psychological states in a coherent way. The overall therapeutic aim was to assist John better to regulate his emotional anguish and physical states without becoming overwhelmed to the extent that he acted out by misusing alcohol and behaving violently. As we have seen, the process involved the evocation of key traumatic experiences preserved in his childhood memories, making these available for dyadic regulation and ideational elaboration (Schore, 1994; Spezzano, 1993; Stern, 1985).

As the weeks went by, John reported that he was keeping his drinking within sensible limits. He looked healthier with clear eyes and a better colour to his complexion, and he seemed more at ease with himself. The impression of John being less anxious and conflicted was quite pronounced and he related how, prior to this improvement, the mere act of leaving home to catch a bus to the town centre would engender anxiety, panic and a sense of danger which he would quell with drink. John also reported noting changes in the way he was responding to others, and they to him, acknowledging that in the past he would often deal with his aggressive impulses by provoking aggression in others, thereby giving himself a ready excuse to be violent and resort to alcohol misuse.

He spoke of decorating his flat, and we came to see this as an external manifestation of the tidying up his inner life was undergoing. He went on to recall feeling acutely persecuted and paranoid as a child following Ricky’s death, saying that he lived in a state of fear and anxiety about the prospect of being attacked by Ricky’s family because “they thought I’d killed Ricky”. Such fears may well have been realistic but, in my opinion, were also likely to have been fuelled by fantasies of retaliation which flourished in the absence of an affectively attuned, containing parental response.

It seemed to me that John’s later violent behaviour reflected Greenberg, Speltz and DeKlyen’s (1993) proposal that children who are insecurely attached develop an internal working model of relationships characterized by anger, mistrust and hostility. As noted above, Main et al.’s (1985) longitudinal research supports the proposition that internal working models developed in childhood tend to persist over time.

Certainly, lack of trust became a major issue for John as he developed into adolescence and adulthood, together with clinical issues of affect regulation, control, autonomy, dependence, separation and loss. He spoke of his surprise at being able to talk to me about personal and painful matters, and he went on to risk rejection by asking if he could contact me after his parole licence had ended should a crisis arise. I agreed to this request, viewing it in terms of an adult relational need rather than an infantile desire, and, as mentioned previously, said that I would like to have a follow-up meeting in any event.

This exchange seemed to indicate that John was internalizing his relationship with me in the form of a secure-enough base from which to explore his traumatic experiences, but that he still needed to feel there would be the opportunity for direct proximity-seeking should something untoward occur (Bowlby, 1980, 1988). I was encouraged by the fact that John was beginning to make links between his traumatic childhood experiences and the anxiety, panic and aggression manifested in later years. He seemed increasingly able to appraise the significance and meaning of these distressing affects (Schore, 1994), and to use the working alliance, or developing attachment relationship, to negotiate and reorganize unresolved clinical issues (Stern, 1985).

Childhood amnesia/dissociation

Some confirmation of this progress occurred towards the end of the session. John related that when aged 28 and on the point of marrying, “my mother told me I’d changed when I was 8”. Apparently, she had offered no explanation as to why this should have happened. Significantly, John went on to say that, to this day, no one in his family has ever alluded to Ricky’s death, adding that he had suffered “amnesia” between the ages of 8 and 11.

It seemed to me that, lacking the emotional and cognitive capacities to assimilate the traumatic event unaided, John’s only option was to resort to a form of dissociation, that is an altered, detached state of consciousness (de Zulueta, 1993; Herman, 1992). For whatever reason, it would appear that John’s parents were insensitive to his needs and unable to help him deal with the aftermath of the tragedy, perhaps misguidedly believing that ignoring the event was for the best. Indeed, there was nothing to suggest they were intentionally cruel or malign but that for reasons stemming from their own attachment histories, they were defensively excluding from consciousness John’s attachment behaviour cues (Main et al., 1985).

The information provided by John, when listened to with relevant research findings in mind, led me to surmise that his parents had been unable to comfort and soothe his distress because of the fear this evoked in themselves (Lyons-Ruth and Jacobovitz, 1999; Schore, 1994). In this event, it is likely that John came to perceive his parents’ dismissing, non-reflective response to his fear as both frightening and frightened, and thus to experience his own state of arousal as a danger signal for abandonment (Main and Hesse, 1990).

In consequence of the family’s disorganized caregiving-attachment system, and the fear and insecurity to which this relational matrix gave rise, it would seem that John’s attentional strategies were compromised and that he developed an exquisite vulnerability to trauma (Lyons-Ruth and Jacobovitz, 1999). His subsequent behaviour suggests that he adapted to this unhappy situation by inhibiting his mentalizing capacity, becoming increasingly detached from his parents, as well as from aspects of his subjective experience, particularly affective states of anxiety, fear, shame and rage (Fonagy, 1999). Thus, although John’s mother was physically present, she appears to have been inaccessible psychologically and emotionally and, therefore, unavailable to help John develop the capacities to regulate, reflect upon and process negative affect and traumatic states of mind.

His subsequent misuse of addictive substances may be seen as having its aetiology in these very incapacities, with first drugs and then alcohol being used to suppress dreaded psychobiological states and hence restore a semblance of affect regulation (Schore, 1994). In this context, it would seem reasonable to hypothesize that John’s childhood attachment to his mother was characterized by what Settlage, Rosenthal, Spielman et al. (1990) and Schore (1994) term “proximal separations”.

Discussion 

John’s material brought to mind research that addresses the aetiology of cognitive-affective disturbance in children. For example, Liotti (1992), following Main (1991), posits a connection between disorganized/disoriented attachment and dissociative disorders. According to Liotti’s hypothesis, the child’s disorganized/disoriented attachment behaviour corresponds to the construction of an internal working model of self and attachment figure that is multiple and incoherent, as opposed to singular and coherent. Liotti (1992) suggests that a multiple internal working model of this kind may predispose the child to enter a state of dissociation in the face of further traumatic experiences. Similarly, Davies and Frawley (1994), in their work with adult survivors of childhood sexual abuse, view dissociation as existing on a continuum, with multiple personality disorder or dissociative identity disorder (MPD/DID) representing the most extreme form of mental defence against severe, protracted trauma.

This opinion is shared by Mollon (1996) who questions whether MPD/DID should be conceptualized as part of a broad grouping of trauma-based psychiatric disorders or as a unique form of personality organization deriving from dissociative and post-traumatic factors.

From a social constructivist perspective, Stern (1997) views cognition as an amalgam of thought and feeling and an integral aspect of a continuous phenomenological process operating within the interpersonal field. Under optimal conditions, this process functions to organize, structure and unify subjective experience, thereby providing the individual with a sense of coherence and meaning. However, Stern (1997) argues that experience may be split for defensive reasons in reaction to trauma and result in the isolation of emotion from mentation.

Van der Kolk and Fisler (1995) found that, in effect, the traumatized subject is left in a state of “speechless terror”. Lacking the words to describe the traumatic event or construct a coherent personal narrative, the individual experiences great difficulty in regulating internal states. Moreover, the authors’ findings show that subjects traumatized in childhood experience more pervasive biological dysregulation than those first traumatized in adulthood. In both instances, however, the traumatic incident is initially “remembered” in the form of fragmented somatosensory experiences (van der Kolk, 1994). Similarly, McDougall (1985; 1989) argues that cumulative trauma consequent on a mother’s insensitive way of handling and interacting with her infant may, during the course of development, lead to a split between word-presentations and affect-laden experiences.

McDougall (1985, 1989) adopts Nemiah’s (1978) and Sifneos’ (1973) concept of alexithymia, that is the inability to recognise and describe discrete emotional states. She postulates that affective reactions associated with the traumatising caregiving process are either avoided or rapidly ejected from consciousness. As a result of this developmental failure, the individual may be susceptible to psychosomatic illness in later life.

Fonagy et al. (1997), in building on Main’s (1991) research into metacognition, posit that the child’s capacity to explore the mind of the other and develop as a thinking and feeling being arises within the matrix of a secure attachment relationship. Insecurity of attachment, on the other hand, undermines the child’s capacity to reflect on and integrate mental experience. Such individuals, it is argued, lack insight into the representational basis of human interaction and intentionality. This being so, they resort to concrete solutions to intrapsychic and interpersonal problems, attempting to control their subjective states and self-cohesion through physical experience such as substance misuse, physical violence and crime.

Van der Kolk (1989), in reviewing studies pointing to the underlying physiology of attachment, posits that endorphin releasers are laid down in the early months of life within the context of attachment to caregivers with different styles of care-giving. He concludes that affectively intense experiences are accompanied by the release of these neurochemicals, and that this psychobiological process comes to be associated both to states of security and trauma. With these findings in mind, Mitchell (2000) comments on the seemingly addictive propensity repeatedly to forge intimate adult relationships redolent of ties to early objects, even when these are traumatic. He suggests that such behaviour may reflect neurochemical, as well as psychological and emotional derivatives.

Much of the aforementioned theory and research is derived from Bowlby (1988) who presented a paper in 1979 entitled “On knowing what you are not supposed to know, and feeling what you are not supposed to feel”. Here, Bowlby cites findings by Cain and Fast (1972) to show how distorted communications between parent and child, which disconfirms the child’s thoughts and feelings of real events, may engender intense guilt. Cognitive dissonance of this kind may lead the child to develop a chronic distrust of other people and of his or her own senses, together with a tendency to find everything unreal.

The response of John’s parents (though he spoke only of his mother in this context) would seem to suggest that emotional states were characteristically dismissed and deflected. Moreover, as we have learned from John, his mother appears to have entered prolonged periods of denial during his childhood, seemingly prompted by feelings of shame and social embarrassment, as evinced by her refusal to acknowledge the reality of her husband’s actual employment status. As noted above, research has demonstrated a significant link between such parental characteristics in terms of a dismissing narrative style on the one hand and insecure-avoidant attachment behaviour in children on the other (Main, 1991; Main et al. 1985; Main and Weston, 1982).

Further, as previously pointed out, children with an insecure-avoidant pattern of attachment have been found to show a marked lack of empathy towards peers in distress. Indeed, Main and Weston (1982) observed a distinct tendency in such insecurely attached children to behave in an aggressive and hostile way, as did Grossman and Grossman (1991). In John’s case, as with so many men who suffer unresolved childhood trauma, substance misuse and violent behaviour followed. The links between these factors were, again, highly reminiscent of the work on trauma by de Zulueta (1993) and Herman (1992). They also accord with findings cited by West and George (1999). These show that male perpetrators of adult relational violence report a high incidence of childhood histories of severe abuse and trauma (Downey, Khoun and Feldman, 1997; Herman and van der Kolk, 1987; Kalmuss, 1984).

Session eight: gender identity

During the eighth meeting with John issues surrounding sexuality and gender were discussed. Given the avoidant pattern of attachment behaviour characterizing John’s adult intimate relationships, I silently questioned the security of his masculine identity. Despite being a stocky, powerfully built and somewhat gruff and macho man, John was sporting a long ponytail hairstyle. Moreover, as already noted, he generally adopted a passive, non-aggressive stance, seemingly disowning authentic thoughts and feelings in a way reminiscent of Winnicott’s (1988) concept of the false self. John’s style of relating at this point elicited feelings within me of inauthenticity and emotional disconnection. Thus prompted, I asked myself whether he might be employing a feminine identification in his interpersonal relationships as a defence against being overwhelmed by anger and rage deriving from archaic ambivalent feelings of separation from and engulfment by the symbiotic mother (Stoller, 1988).

On discussing the way in which men and women incorporate both masculine and feminine attributes, John’s behavioural responses were initially averse. His reaction put me in mind of the fact that heterosexual men not infrequently form temporary homosexual liaisons when imprisoned for any length of time. On discussing this delicate subject, I keenly observed John’s behaviour for any signs of intrusiveness or persecution as I sought to establish a sense of emotional connection or intersubjective relatedness with him (Stern 1985; 1998).

Again, I felt rather parental in the countertransference, as though fulfilling functions that were containing emotionally as well as informative on a cognitive level. The interactional aspect of the therapeutic process would seem to confirm the importance of applying a developmental model in clinical work with offenders, given the high incidence of unresolved childhood trauma in the offender population.

Following these exchanges John was able to elaborate on the experiences he had had whilst in prison. More generally, we explored how heterosexual men with a confused sense of gender identity may manifest homosexual panic and deal with feelings of shame and anger by denying aspects of their sexuality that create anxiety, and instead project these onto others.

A dialogue developed exploring the way in which defensive behaviour of this kind, allied to a morbid fear of difference, may act as a touchstone for violence that targets minority groups, for example, “gay bashing” and racist attacks. Although the latter appeared not to be features of John’s pattern of offending behaviour, this discussion seemed further to enhance his reflective functioning or mentalizing capacity. He appeared more able gradually to perceive others as separate from himself and as having distinct feelings, intentions and desires (Fonagy et al., 1997).

Session nine: relapse and transference issues 

Before the ninth session, John attended the funeral of a family friend. During this session, John volunteered the information that he had consumed about six pints of beer at the wake. We discussed this in the context of what the death of his friend had evoked in John. He said that his predominant feelings were of anxiety and guilt. John linked these feelings directly to a fearful anticipation that I would “misjudge” him for drinking. I wondered whether this dynamic again constituted a transference re-enactment connected with John’s childhood trauma, particularly the unmourned loss of Ricky. At that time, whether in reality or in fantasy, John did indeed feel misjudged and blamed for Ricky’s death.

Further, as we have seen, it would appear that his parents’ response lacked empathy and was dismissive of his emotional pain and distress. On an unconscious level, therefore, John may well have been expecting a similar response from me, as it would seem that I was being attributed an archaic parental role in the transference (Sandler and Sandler, 1998). His relapse provided the opportunity to resolve some of these issues, in that it enabled him to re-experience his traumatic attachment to a dismissing, emotionally unavailable parent in a way that was bearable (Holmes, 1996). My task at such times was to survive John’s omnipotent destructive fantasies without collapsing or retaliating (Winnicott, 1988).

This “holding” response appeared to help John recognise my existence as a separate person available to be used and related to intersubjectively (Benjamin, 1992). Moreover, by relating to John in this unfamiliar way, I became a new developmental object, different from the original pathogenic object being sought in transferential re-enactments (Fonagy, 1998; Hurry, 1998; Schore, 1994).

Session 10: the aetiology of John's violent behaviour 

John opened the tenth session by saying he felt on an “even keel”, adding that he was continuing to spend a good deal of time thinking about past experiences, as well as monitoring his thoughts and feelings in the here-and-now, particularly when in an emotionally disturbed mood. At such times, in line with my suggestion, he would try to trace the immediate trigger of the affective distress and then make links between the past and the present.

The therapeutic purpose of setting John this “homework” was to encourage him to stay with the dreaded lived experience for long enough to reflect on and attempt to self-regulate primitive, unintegrated affective states. These problematic experiences would then be brought to sessions for collaborative exploration of their aetiology and meaning. My expectation was that this therapeutic intervention would gradually assist John to develop the capacity to elaborate and transform disturbing bodily experiences into a coherent narrative (Schore, 1994; Spezzano, 1993; Stern, 1998).

With regard to this process, Schore (1994) emphasises that the therapist’s own tolerance of affects will critically determine the range and types of emotion that may be explored or disavowed in the transference-countertransference relationship. It may be accepted that this consideration is of particular relevance in a forensic setting because the practitioner is often starkly confronted with the bleaker aspects of human experience and the darker side of human behaviour.

The current session focused in a direct way on John’s violence to women. This issue had been a delicate subject up until this point because his violent behaviour, especially in regard to women, jarred with his ideal self, leaving him feeling deeply shamed. At our first meeting John had displayed a pronounced tendency to minimize his culpability and blame the victim. Indeed, we will recall that he had completely denied the offence initially and was convicted following a jury trial. John’s capacity for denial brought to mind his mother who, as we have seen, deployed the self-same defence mechanism with equal conviction. In this context, I have learned from hard experience that working precipitously with denial is clinically sterile and counter-productive, generating intense feelings of frustration and rage in the participants as early parent-child roles and patterns of interaction get re-enacted in the transference-countertransference matrix.

Fortunately, by this stage in my relationship with John a secure-enough working alliance had been forged, and thus the time seemed ripe for us to explore this form of defensive behaviour. As we did so, John’s dissociated affect of shame and anger became increasingly available for interactive regulation. This process facilitated his gradual acceptance and active responsibility for his violent behaviour, and an enhancement in his ability to empathise with Sylvia.

My thinking in respect of John’s violence was that he had dissociated states of anger, rage and yearning as a child, primarily in relation to his mother. This adaptive defence was needed because separations from her and the family had been managed insensitively, as had the later trauma in respect of Ricky, reflecting the family’s disorganized caregiving-attachment system (Lyons-Ruth and Jacobovitz, 1999).

As noted above, the effects of these events and relational patterns had become frozen in time, being preserved and represented internally as non-reflective, non-verbal procedural memories in the form of pre-symbolic interaction structures and internal working models. These unmodified mental models were, in turn, expressed in violent behavioural enactments in John’s relationships, particularly at times of intense interpersonal stress (Beebe, Jaffe and Lachmann, 1992; Fonagy et al., 1997; Main et al., 1985; Schore, 1994; Stern, 1985, 1998; West and George, 1999). This hypothesis received some confirmation when John went on to speak of becoming angry with a man who had recently beaten his wife.

On discussing the incident, it became clear that much of the anger generated in John was not only because of the man’s physical abuse of the woman. His feelings were also inflamed because this person had subsequently flatly denied that the assault had taken place, even though all his acquaintances knew full well that it had.

This scenario appeared to have powerful associations and resonances with John’s childhood, in that in a similar way everyone had known that Ricky had drowned. As we have seen, despite the reality of this traumatic event, John’s emotional and cognitive experience had been denied, or, at best, unacknowledged, with Ricky’s death becoming, in effect, a well kept family secret (Pincus and Dare, 1990). Thus John’s capacity for metacognitive monitoring or reflective functioning was fatally compromised since the information he was receiving about the traumatic event was contradictory and distorted.

This, in turn, seems to have led to the development of a multiple, incoherent internal working model in respect of his attachment to his mother, and a concomitant state of disorganisation and dissociation (Liotti, 1992; Main, 1991). As Solomon and George (1996) found, disorganized attachment is characterized by controlling behaviour towards the attachment figure in the context of the child feeling abandoned, helpless and vulnerable. This fraught situation is likely to have been exacerbated in John’s case by the separations he had experienced at a younger age when hospitalized and subjected to surgical intervention. The clinical evidence attested to these factors having contributed to the development of a predominantly insecure-avoidant pattern of attachment organization which, in line with Fonagy et al. (1997), I viewed as an adaptive defence mustered in the face of unattuned care-giving. Moreover, the overall clinical picture suggested that insecurity stemming from separation anxiety had interfered with John’s capacity to differentiate himself psychologically from his attachment figure and, thereby, attain a state of “mature dependence” (Fairbairn, 1996).

In this context, it is of interest to note that West and George (1999) suggest that psychological merging may explain the conflict between engulfment and abandonment that appears to be so characteristic of the physically abusive male.

As we have seen, Bowlby (1973) emphasises that anger at an attachment figure who fails to provide the expected comfort at times of stress is a normal and integral aspect of the attachment system. It would appear that this safety valve was not available to John, as he lacked both external and internal permission to experience disturbing negative affect (Spezzano, 1993). He had little option, therefore, but to develop a defensive organization against anger and rage, in part, identifying with the dismissing, non-reflective qualities and characteristics of his relationship with his mother as a way of defending against feelings of guilt and anxiety (Fonagy et al., 1997; A. Freud, 1993). When these mental defences were overwhelmed, John’s split off infantile murderous rage was acted out with the violent, destructive force of an adult.

In the latest incident of this repetitive pattern, this internal dynamic was externalized, being displaced or redirected from John’s original primary attachment figure (his mother) and projected onto Sylvia who, at the point of the breakup of their relationship, he perceived as untrustworthy, rejecting and abandoning. Indeed, John’s “theory of mind” appears to have led him to expect betrayal and misattuned regulation at the hands of Sylvia (Fonagy et al., 1997; Schore, 1994; West and George, 1999). Thus, she became a vehicle for his intolerable and persecutory self states, that is, for the internalized aspects of his relationship with his mother that he experienced as frightening and unmanageable (Fonagy, 1999). Further, aggression, rather than love, seems to have become an emotionally “rewarding” way for John to express his ambivalent world of object relations (Dicks, 1993).

Fear of abandonment, then, seems to have been the primary affect that led to the sudden activation of John’s attachment behavioural system and identification with the disorganized internal working model of the relationship with his mother. As already noted, a salient feature of this mental model was John’s expectation that his primary attachment figure would not be available or accessible at times of affective stress to provide comfort and protection (Schore, 1994; West and George, 1999). The clinical evidence, combined with John’s forensic history, indicated that his mentalizing capacity was prone to becoming compromised and disorganised by intense separation anxiety and dysregulated fear, shame, hatred and rage when he felt threatened by the loss of a female partner with whom he had formed an intimate attachment.

The theoretical model of male violence proposed here converges with that delineated by West and George (1999). These authors contend that intimate adult relational violence is rooted in attachment disorganization, viewing this pattern as inextricably linked to unresolved trauma and to a segregated representational system characterized by dysregulated affect and pathological mourning. West and George (1999) suggest that the perpetrator’s defensive and tightly controlled regulation of his attachment system breaks down at the moment of the assault as a consequence of his becoming flooded with negative affect and distorted perceptions deriving from his personal trauma.

With regard to the index offence, the significance for John of Sylvia’s perceived sexual infidelity lay, in part, in the fact that it represented her independence of mind and psychological separateness. In my opinion, any move by her towards a separate, independent existence would have conflicted with the explicit and implicit role expectations that John had brought to the relationship, being construed as a threat to his sense of security (Dicks, 1993). Thus, John’s violence was, in part, a frantic attempt to control Sylvia so as to ensure her continued availability, both to protect him against infantile loneliness and immature dependence (Dicks, 1993; Fairbairn, 1996) and to carry the alien, persecutory parts of himself (Fonagy, 1999). The thought of being abandonment by Sylvia instilled terror in John because loss was experienced as a re-traumatization and, therefore, as a threat to the coherence and stability of his sense of self (de Zulueta, 1993; Fonagy, 1999; Herman, 1992).

I was struck, moreover, by the fact that John’s vicious assault on Sylvia had been triggered at the very point that she denied sexual infidelity, and that John continued to beat her mercilessly until she told him the “truth”. Again, I silently wondered to what extent this ghastly episode was a re-creation in the present of unresolved aspects of John’s childhood relationship with his mother who, throughout his life, adamantly denied and invalidated the reality of his traumatic experience (van der Kolk, 1989; 1994). Thus, it would appear that John’s violence not only had the effect of making him feel coherent and real, but also of eliciting the response from Sylvia he so desperately needed to hear - a voice that validated the “truth” of his subjective experience (Fonagy, 1999).

Clearly, John’s violent behaviour could easily have escalated out of control and led to a charge of murder or manslaughter, as in the case of other insecurely attached men who had responded with extreme violence when abandoned by their female partners. Sharing these thoughts with John seemed to have a sobering effect on him and again engendered a state of deep, prolonged and silent reflection.

Session 12: indications of change 

At the twelfth session, John announced with great confidence that he no longer thought of himself as an alcoholic. I was surprised to see that he had had his hair cut short. I silently wondered whether this dramatic change in his personal appearance was emblematic of a firmer sense of masculine identity, together with a concomi

THE THERAPEUTIC PROCESS The progress made by John was dependent on his ability gradually to organize and integrate error-correcting information received as an ongoing aspect of the therapeutic process (Bowlby, 1980, 1988; Main et al., 1985; Peterfreund, 1983). Following Tronick, Als, Adams, Wise and Brazelton (1978), I viewed this process as consisting, in significant degree, of the moment-to-moment micro-repair of attunement or misaligned interaction - an intersubjective process operating at the level of procedural or implicit relational knowing (Stern, Sander, Nahum et al., 1998). This process was informed by the tracking and matching of subtle and dramatic shifts in John’s mood-state as he narrated his story (Schore, 1994). This interactive process led, in turn, to the recognition of a shared subjective reality (Fonagy, 1998). By these means, my facilitating behaviours combined with John’s capacity for attachment. Though operating largely out of conscious awareness, this mutually reciprocal process permitted the development of a working alliance or attachment relationship (Schore, 1994) that was secure-enough to facilitate a collaborative exploration of painful, unresolved clinical issues linked to his misuse of alcohol and violent offending behaviour. As we have seen, key aspects of this intersubjective and reparative process were the dyadic regulation of dreaded states charged with intense negative affect (Schore, 1994) and the co-construction of a coherent narrative (Holmes, 1996). Thus, I became a new developmental object, the relationship with whom disconfirmed John’s pathogenic transference expectations and enhanced his capacity for reflective thought (Fonagy, 1998; Hurry, 1998; Schore, 1994). From a neurobiological perspective, Schore (1994) postulates that the process of affect regulation links non verbal and verbal representational domains of the brain, thereby facilitating the transfer of implicit information in the right hemisphere to explicit or declarative systems in the left.

FOLLOW-UP CONTACT As agreed, I contacted John for a follow-up discussion 6 months later. His progress had been sustained, in that he was still in work, keeping his consumption of alcohol within sensible limits, and had not re-offended. Prior to the ending of John’s supervisory period, I had liaised with his GP who, in consultation with John, agreed to refer him to the local mental health resource centre. John had attended an assessment session there with a clinical psychologist. It was mutually agreed that no further work was needed at that stage.

CONCLUSION Given the wide incidence of intimate adult relational violence in Western society, understanding the clinical issues underlying such behaviour, and developing an effective therapeutic model to address the problem, is a pressing social concern. An important consideration in this context is the traumatic effect on children who repeatedly witness scenes of abusive male violence in the home (Cawson, Watton, Brooker and Kelly, 2000). I would argue that an attachment theory conceptualisation, as outlined in this study and proposed by West and George (1999), has a significant contribution to make in this area of work.

With regard to my brief intervention with John, it remains to be seen whether this will prove to be effective in the long term. I was keenly aware that far more could have been achieved therapeutically, not least on consolidating the progress he had made in regulating his affective states without resorting to alcohol misuse. This, however, was not a viable option, given the constraints of time and resources obtaining within the probation service and the limited period of his parole licence. Nevertheless, I consider that the work undertaken with John helped to resolve the childhood trauma underlying his adult violent offending behaviour. This, in turn, enhanced his sense of security and capacity for narrative competence and reflective functioning, thereby strengthening his ability to activate alternative models of interaction and take the other’s perspective. These interlinking positive therapeutic changes should potentiate further personal growth and so provide John with a greater ability to empathize with others and make more moral, reasoned choices in the future (Fonagy and Target, 1998; Holmes, 1996).

More generally, I consider that this case study illustrates that it is not the traumatic childhood event in itself that is salient in personality development and adult psychopathology, but rather the characteristic caregiving-attachment system within which the child experiences the trauma. As noted above, research has demonstrated that the securely attached child develops the capacity to stay attentive and responsive to the environment, and to use error-correcting information to construct a coherent narrative when presented with scenarios involving separation and loss (Main, 1991; Main and Hesse, 1990; Main et al., 1985). It would seem reasonable to hypothesize, therefore, that the secure/autonomous adult has developed the mental capacity to process information more readily in the aftermath of a traumatic event than the insecure, disorganized subject whose ability to regulate states of arousal at moments of stress was compromised during early development. As Bowlby (1973, 1979) has observed, the quality of the emotional bond between the child and the caregiver will vitally influence whether mourning proceeds along a healthy path or takes a pathological course. From an attachment theory perspective, therefore, the overarching therapeutic task in my work with John was to help him to recover and express ambivalent thoughts and feelings linked to the trauma of unmourned childhood loss (Bowlby, 1979).

OUTCOME In a meta-analysis of outcome studies, Roth and Fonagy (1996) found that the extent to which ruptures to the working alliance were adequately addressed during the course of the therapy was predictive of the efficacy of the intervention. The authors conclude that the relationship component is the common effective ingredient in positive outcomes. Their findings accord with my clinical experience in applying attachment theory in a probation setting, and would seem to confirm the respective findings of Schore (1994) and Stern et al. (1998) viz., that the interactive emotion-transacting aspect of the therapeutic encounter is the main mechanism of intrapsychic change.

Specific to a probation setting, it has been suggested that reconviction rates should be used to evaluate effectiveness and measure outcome (Chapman and Hough, 1998). At that I wrote this paper, John’s name had not appeared on the local court list, nor had any requests for pre-sentence reports been received from non-local courts. It seemed reasonable to assume, therefore, that 5 years had elapsed since John last offended. Further, it may be accepted that identifying the links between John’s insecure attachment, unresolved childhood trauma, emotional detachment/dissociation, substance misuse and violent offending behaviour was vital in order to work with him both effectively and with expedition.

POSTSCRIPT I met with John some six months after our follow-up contact to obtain his consent to publish an account of our work. He had remained largely free of symptoms of panic, persecution and depression, telling me “I feel a lot more confident than I used to”, adding “I now think before I act”. John was still in work, even though this entails travelling some distance from his home. As we parted, he told me “I don’t need the drink and violence any more because I’ve accepted myself for who I am”.

REFERENCES Ainsworth, M., Blehar, M., Waters, E. and Walls, S. (1978). Patterns of Attachment: Assessed in the Strange Situation and at Home. Hillsdale, NJ: Lawrence Erlbaum.

Beebe, B. and Lachmann, F.M. (1992). The Contribution of Mother-Infant Mutual Influence to the Origins of Self- and Object Representations. In Relational Perspectives in Psychoanalysis, eds. N.J. Skolnick and S.C. Warshaw, 83-117. Hillsdale, NJ: The Analytic Press.

Beebe, B., Jaffe, J. and Lachmann, F.M. (1992). A Dyadic Systems View of Communication. In Relational Perspectives in Psychoanalysis, eds. N.J. Skolnick and S.C. Warshaw, 61-81. Hillsdale, NJ: The Analytic Press.

Benjamin, J. (1992). Recognition and Destruction: An Outline of Intersubjectivity. In Like Subjects, Love Objects: essays on recognition and sexual difference, Ch. 1, 27-48. New Haven: Yale University Press.

Boswell, G. (1996). Young and Dangerous: the Backgrounds and Careers of Section 53 Offenders. Aldershot: Avebury.

Boswell, G. (1998). Research-Minded Practice With Young Offenders Who Commit Grave Crimes. In Probation Journal. 45: 202-207.

Bowlby, J. (1969). Attachment and Loss, Vol. 1: Attachment. London: Pimlico.

Bowlby, J. (1973). Attachment and Loss, Vol. 2: Separation: Anger and Anxiety. London: Pimlico.

Bowlby, J. (1979). The Making and Breaking of Affectional Bonds. London: Routledge.

Bowlby, J. (1980). Attachment and Loss, Vol. 3: Loss: Sadness and Depression. London: Pimlico.

Bowlby, J. (1988). A Secure Base: Clinical Applications of Attachment Theory. Bristol: Arrowsmith.

Cain, A. C. and Fast, I. (1972). Children’s disturbed reactions to parent suicide. In Survivors of suicide, ed. A. C. Cain, 93-111. Springfield, Illinois: C . C. Thomas.

Casement, P. (1990). On Learning from the Patient. London: Routledge.

Cawson, P., Watton, C., Brooker, S. and Kelly, G. (2000). Child Maltreatment in the United Kingdom. London: NSPCC Publications.

Chapman, T. and Hough, M. (1998). Evidence Based Practice: A Guide to Effective Practice. HM Inspectorate of Probation, ed. M. J. Furniss. London: Home Office Publications Unit.

Davies, J. M. and Frawley, M. G. (1994). Treating the Adult Survivor of Childhood Sexual Abuse. New York: Basic Books.

De Zulueta, F. (1993). From Pain to Violence: The traumatic roots of destructiveness. London: Whurr Publishers.

Dicks, H. V. (1993). Marital Tensions: clinical studies towards a psychological theory of interaction. London: Karnac Books.

Downey, G., Khoun, H. and Feldman, S. (1997). Early interpersonal trauma and later adjustment: The mediational role of rejection sensitivity. In Development perspectives on trauma: Theory, research and intervention, eds. D. Cicchetti and S. L. Toth, 85-114. Rochester, New York: University of Rochester Press.

Fairbairn, W.R.D. (1996). Psychological Studies of the Personality. London: Routledge.

Fonagy, P., Target, M., Steele, M., Steele, H., Leigh, T., Levinson, A. and Kennedy, R. (1997). Morality, Disruptive Behaviour, Borderline Personality Disorder, Crime, and Their Relationships to Security of Attachment. In Attachment and Psychopathology, eds. L. Atkinson and K. J. Zucker, 223-274. New York: Guilford Press.

Fonagy, P. (1998). Moments of Change in Psychoanalytic Theory: Discussion of a New Theory of Psychic Change. Infant Mental Health Journal, 19 (3): 346-353.

Fonagy, P. and Target, M. (1998). An interpersonal view of the infant. In Psychoanalysis and Developmental Therapy, ed. A. Hurry, 3-31. London: Karnac Books.

Fonagy, P. The Male Perpetrator: The Role Of Trauma And Failures Of Mentalization in Aggression Against Women - An Attachment Theory Perspective. Unpublished paper given at the 6th John Bowlby Memorial Lecture, London, 20th February, 1999.

Freud, A. (1993). The Ego And The Mechanisms Of Defence. London: Karnac Books.

George, C., Kaplan, N. and Main, M. (1984). Adult Attachment Interview, (1st edition). Unpublished manuscript, Department of Psychology, University of California at Berkeley.

Greenberg, M. T., Speltz, M. L., and DeKlyen, M. (1993). The role of attachment in the early development of disruptive behaviour problems. In Development and Psychopathology. 5: 191-213.

Grossman, K. E. and Grossman, K. (1991). Attachment quality as an organiser of emotional and behavioural responses in a longitudinal perspective. In Attachment across the Life Cycle, eds. C. M. Parkes, J. Stevenson-Hinde and P. Marris, 93-114. London: Routledge.

Herman, J. L. and van der Kolk, B. A. (1987). Traumatic antecedents of borderline personality disorder. In Psychological trauma, ed. B. A. van der Kolk, 111-126. Washington, DC: American Psychiatric Press.

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

Holmes, J. (1993). John Bowlby and Attachment Theory. London: Routledge.

Holmes, J. (1996). Attachment, Intimacy, Autonomy: Using attachment theory in adult psychotherapy. New York: Jason Aronson.

Hurry, A. (1998). Psychoanalysis and Developmental Therapy. In Psychoanalysis and Developmental Therapy, ed. A. Hurry, 32-73. London: Karnac Books.

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

Liotti, G. (1992). Disorganized/Disoriented Attachment in the Etiology of the Dissociative Disorders. In Dissociation. 4:196-204.

Lyons-Ruth, K. and Jacobovitz, D. (1999). Attachment disorganization: unresolved loss, relational violence and lapses in behavioural and attentional strategies. In Handbook of attachment: Theory, research and clinical applications, eds. J. Cassidy and P.R. Shaver, 520-554. New York: Guilford Press.

Mahler, M. S., Pine, F. and Bergman, A. (1985). The Psychological Birth of the Human Infant: Symbiosis and Individuation. London: Karnac Books.

Main, M. and Weston, D. (1982). Avoidance of the attachment figure in infancy. In The Place of Attachment in Human Behaviour, eds. C. M. Parkes and J. Stevenson-Hinde, 31-59. London: Tavistock.

Main, M., Kaplan, N. and Cassidy, J. (1985). Security in Infancy, Childhood, and Adulthood: a move to the level of representation. In Growing points in attachment: theory and research, eds. I. Bretherton and E. Waters, 66-104. Monographs of the Society for Research in Child Development. Chicago: University of Chicago Press.

Main, M. and Hesse, E. (1990). Parents’ unresolved traumatic experiences are related to infant disorganized attachment status: Is frightened and/or frightening parental behaviour the linking mechanism? In Attachment in the preschool years: Theory, research and intervention, eds. M. Greenberg, D. Cicchetti, and E. M. Cummings, 161–182. Chicago: University of Chicago Press.

Main, M. (1991). Metacognitive knowledge, metacognitive monitoring, and singular (coherent) vs multiple (incoherent) models of attachment: findings and directions for future research. In Attachment across the Life Cycle, eds. C. M. Parkes, J. Stevenson-Hinde and P. Marris, 127-159. London: Routledge.

Maroda, K.J. (1991). The Power of Countertransference: Innovations in Analytic Technique. New York: John Wiley & Sons.

McDougall, J. (1985). Theatres of the Mind: Illusion and Truth on the Psychoanalytic Stage. London: Free Association Books.

McDougall, J. (1989). Theatres of the Body: A psychoanalytical approach to psychosomatic illness. London: Free Association Books.

Mitchell, S.A. (2000). Relationality: From Attachment to Intersubjectivity. Hillsdale, NJ: The Analytic Press.

Mollon, P. (1996). Multiple Selves, Multiple Voices: Working with Trauma, Violation and Dissociation. Chichester: John Wiley & Sons.

Nemiah, J.C. (1978). Alexithymia and psychosomatic illness. Journal of Continuing Education in Psychiatry (1978): 25-37.

Peterfreund, E. (1983). The Process of Psychoanalytic Therapy. New York: The Analytic Press.

Pincus, L. and Dare, C. (1990). Secrets in the Family. London: Faber and Faber.

Roth, A. and Fonagy, P. (1996). What works for whom: Limitations and implications of the research literature. New York: Guilford Press.

Sandler, J. and Sandler, A-M. (1998). Internal Objects Revisited. London: Karnac Books.

Schore, A.N. (1994). Affect Regulation and the Origin of the Self: The Neurobiology of Emotional Development. Hillsdale, NJ: Lawrence Earlbaum.

Settlage, C. F., Rosenthal, J., Spielman, P.M., Gassner, S., Afterman, J., Bemesderfer, S. and Kolodny, S. (1990). An exploratory study of mother-child interaction during the second year of life. Journal of the American Psychoanalytic Association, 38, 705-731.

Sifneos, P.E. (1973). The prevalence of “alexithymic” characteristics in psychosomatic patients. Psychotherapy and Psychosomatics, 22: 255-262.

Slade, A. (1999). Attachment Theory and Research: Implications for the Theory and Practice of Individual Psychotherapy with Adults. In Handbook of attachment: Theory, research and clinical applications, eds. J. Cassidy and P.R. Shaver, 575-594. New York: Guilford.

Spezzano, C. (1993). Affect in Psychoanalysis: a clinical synthesis. Hillsdale, NJ: The Analytic Press.

Soloman, J. and George, C. (1996). Defining the caregiving system: Toward a theory of caregiving. Infant Mental Health Journal, 17, 183-197.

Stern, D.N. (1985). The Interpersonal World of the Infant: A View from Psychoanalysis and Developmental Psychology. New York: Basic Books.

Stern, D.N., Sander, L.W., Nahum, J.P., Harrison, A.M., Lyons-Ruth, K., Morgan, A.C., Bruschweiler-Stern, N., and Tronick, E.Z. (1998). The Process of Therapeutic Change Involving Implicit Knowledge: Some Implications of Developmental Observations for Adult Psychotherapy. Infant Mental Health Journal. Vol.19 (3), 300-308.

Stern, D.N. (1998). The Motherhood Constellation: A Unified View of Parent-Infant Psychotherapy. London: Karnac Books.

Stern, D.B. (1997). Unformulated Experience: From Dissociation to Imagination in Psychoanalysis. Hillsdale, NJ: The Analytic Press.

Stoller, R. J. (1986). Perversion: The Erotic Form of Hatred. London: Karnac Books.

Tronick, E., Als, H., Adams, L., Wise, S. and Brazelton, T.B. (1978). The infant’s response to entrapment between contradictory messages in face-to-face interaction. Journal of American Child Psychiatry. 17: 1-13.

Tyson, P. and Tyson, R.L. (1990). Psychoanalytic Theories of Development: An Integration. New York: Yale University.

van der Kolk, B.A. (1989). The Compulsion to Repeat the Trauma. Psychiatric Clinics of North America, 12. 2: 389-411.

van der Kolk, B.A. (1994). The body keeps the score: Memory and the evolving psychobiology of post-traumatic stress. Harvard Review of Psychiatry, 1, 253-265.

van der Kolk, B.A., and Fisler, R. (1995). Dissociation and the Fragmentary Nature of Traumatic Memories: Overview and Exploratory Study. Journal of Traumatic Stress, Vol. 8, No 4, 505-521.

West, M. and George, C. (1999). Abuse and violence in intimate adult relationships: New perspectives from attachment theory. In Journal of Attachment and Human Development, Vol. 1, No 2.

Winnicott, D.W. (1974). Fear of Breakdown. International Journal of Psycho-Analysis. 1: 103-107.

Winnicott, D. W. (1988). Playing and Reality. London: Penguin.