Articles‎ > ‎

The Death of a Leader Homicide as a Means of Group Disengagement

The Death of a Leader: Homicide as a Means of Group Disengagement

Janet Haines, Ph.D.

School of Psychology

University of Tasmania


Christopher L. Williams, Ph.D.

School of Psychology

University of Tasmania



Discipline of Psychiatry

University of Tasmania


Jodi Glading, B.S., John Davidson, Ph.D.

School of Psychology

University of Tasmania



This paper presents a single case study of a cult group member who murdered the group leader as a means of escaping the group. The authors undertook an assessment to examine the pre-homicide and peri-homicide factors that influenced the individual’s homicidal behavior, in an endeavor to establish this man’s motive for ending the group leader’s life. The authors administered psychological tests to assess the individual’s intellectual functioning, psychological adjustment, and group psychological abuse. In addition, they employed a personalized guided-imagery methodology to recreate the homicidal act so that they could examine the individual’s stage-by-stage psychophysiological and psychological responses to the behavior. They then made a comparison of the individual’s responses to the homicidal act with his responses to other control scenes to establish the differential influences operating at the time of the homicide. The results of the research highlight the influences of psychological manipulation and the constriction of problem-solving options that can operate as a function of cult-group membership.

Membership in cultic groups is associated with a range of specific influences, including thought reform (Lalich, 1992; Lifton, 1991) and mind control (Langone & Chambers, 1991), which function to maintain obedience and reduce individuality. In a situation in which mind control is in operation, making everyday decisions can be problematic for group members. It is hardly surprising, then, that a major decision on the part of a member to disengage from a controlling group presents an enormous challenge.

Some researchers have suggested that there is little robust evidence of individuals’ pre-cult psychopathology (Martin, Langone, Dole, & Wiltrout, 1992) or maladjustment during the period of membership (Aronoff, Lynn, & Malinoski, 2000). It is after individuals’ separation from the group that elevated levels of psychopathology become apparent (Aronoff et al., 2000; Swartling & Swartling, 1992; Walsh, Russell & Wells, 1995). There is evidence to suggest that the process of disengagement from the group is what is detrimental to the individual’s well-being (Walsh et al., 1995; Wright, 1991). Merely being away from the group does not produce psychopathology. The struggle to reject previously held beliefs, and the need to engage in high-level decision making prior to disengagement, might contribute to the development of symptomatology.

Although well-publicized deaths have been associated with cults, such as homicidal and suicidal behaviors that occur in relation to group beliefs or because of external threat (e.g., Lasaga, 1980; Nishida, 2001; Robinson, 1997), it seems that homicide as a means of disengaging from a cultic group has received less attention.

In this paper we will describe a case of a young man who was charged with and found guilty of the murder of the leader of a group of which the young man was a member. We will show that the characteristics of the group suggested a cultic influence.

Others (e.g., Tunnell and Cox, 1995) have endorsed a single case-study approach to the examination of motives for homicide. We obtained information about this young man as part of a larger research project examining motives for homicide that two of the authors undertook with the approval of the University of Tasmania Research Ethics Committee and the relevant government agencies. They obtained written informed consent from the participant for publication of this material.

The Case

The perpetrator of the homicide, named Matthew, was the middle child of working-class parents. Matthew was of less than average height and of average build. He had a fair complexion, short-cropped and balding hair, and wore glasses. His general demeanor was unassuming, and during interview and assessment he was obliging and eager to please.

Matthew’s parents divorced when he was 10 years old, and he continued to reside with his mother. The family frequently moved, living in city, suburban, and rural areas in Tasmania. Matthew was a quiet and unassuming child, and his schooling was without incident. The frequent moves during his childhood and adolescence would not have allowed for stability of school peer relationships. He completed his education at the end of high school and obtained employment at the age of 16 years. Matthew was continuously employed in manual labor and sales positions until his arrest. He had no history of serious medical problems and no evidence of psychological disturbance prior to his involvement in the group.

During his early twenties, Matthew was introduced to the victim, Kane, who encouraged him to join a group. Kane, a middle-aged man who was known to police and who was a somewhat colorful character, led the group. This man called the group members “ninja warriors,” and the main purpose of the group was to become proficient in martial arts so that when it became necessary to rise against the government, the group would be prepared to do so.

Kane’s parents had started their own religious group, and they ran a boarding house that was decorated as a medieval residence. Kane had married and was the father of three children, although he was living separately from his wife at the time of his death. He was highly proficient at martial arts, and Matthew had witnessed the ferocity with which Kane applied his skills against people who displeased him. An interesting note is that post-mortem examination identified a structural abnormality of the left temporal pole of Kane’s brain. Anger and anxiety have been demonstrated to elicit left temporal lobe activity (Kimbrell et al., 1999). Temporal lobe abnormalities have been associated with aggressiveness and violence (Amen, Stubblefield, Carmichael, & Thisted, 1996; Wong, Lumsden, Fenton, & Fenwick, 1994), and delusions (Liddle, 1997; Trimble, 1992), particularly hypochondriacal delusions (Wada, Kawakatsu, Komatani, Okuyama, & Otani, 1999). Kane had told Matthew and others that he believed he had terminal cancer.

Matthew was told that the group had many members. It was explained to him that the group was divided into smaller warrior groups who trained together. The army as a whole would not meet until necessary. In this way, Matthew did not know of other members beyond his own sub-group, although he was convinced that they existed. He was encouraged to live at the boarding house along with other group members, and Kane closely monitored his activities. Matthew was led to believe that he was part of an inner circle of members who would reap financial and other benefits when the group acted against the government. Matthew was encouraged to believe that the police and government personnel were enemies.

Matthew became disenchanted with Kane, the group, and the way of life. He left the group and avoided his associates from the group. However, when Matthew was 26 years old, Kane again made contact with him and encouraged him to rejoin the group. Matthew agreed, although he did not return to live at the boarding house despite considerable pressure to do so. By this time, he had begun a relationship and was residing with his girlfriend.

Matthew again commenced martial arts training. At this time, however, Kane demonstrated an escalation of eccentric and bizarre behavior. He was reported to have become increasingly volatile, and his behavior was reported to be erratic. He told Matthew of a plan to gain control of the state. It was at this time that Matthew learned that his girlfriend was pregnant with their first child.

According to Kane’s plan, activities of the group were to be financed by extorting money from a local, wealthy businessman. The plan entailed the group breaking into a federal armory to amass an arsenal of weapons and explosives. Group members were then to cause an explosion at a fuel-storage facility on the outskirts of the city, which would result in massive local devastation. Kane’s intention was for group members to storm police headquarters and gain control of the state from there while police and emergency services were dealing with the aftermath of the explosion.

To this end, Kane instructed Matthew and other group members to engage in a range of illegal activities including vandalism, assault, and extortion. Group members deemed many of these actions to have symbolic importance. The criminal activities were interesting for two reasons. First, the majority of group members, including Matthew, had no history of criminal behavior before they engaged in such activities at the direction of the group leader. Despite this, they did not question Kane’s demands of them. Second, at no time did Kane participate in these activities; group members acted as the perpetrators of the crimes while the leader isolated himself from the activity.

The plan clearly was a grand one. However, certain factors would have influenced the extent to which Matthew and fellow group members would have believed that Kane was capable of coordinating the activities necessary to achieve the goal. First, Tasmania is an island state with a population of only approximately 500,000 people. The capital city, Hobart, is located at the south of the state and is the only city with a population greater than 100,000. In theory, with enough personnel and firepower, isolating the state from mainland Australia would be possible. Worth remembering is that Matthew was unaware of the number of group members, but he had been led to believe that there were sufficient numbers to form an army. Second, it was clear that Kane already had a considerable number of weapons and sufficient explosives to cause devastation at the fuel depot. Kane had shown Matthew the explosives and weapons. Third, the fuel-storage facility was located adjacent to a suburban area. An explosion at this site potentially would have caused loss of life, destruction of property, and disruption to infrastructure. The group members believed that these events would occur because of the group leader’s degree of influence.

During a period of escalation of criminal activity by group members, and an intensification of the bizarre and erratic behavior of the leader, Matthew reported becoming frightened by the uncontrollability of the events. He stated that he wanted to leave the group at this time. However, he was faced with a dilemma. Kane had made it clear to all members that there was no possibility of leaving the group alive: “the only way out is death.” At the time, Matthew did not doubt that he would be killed if he left the group. He reported intense fear for himself, his pregnant girlfriend, and his family. The other factor that reduced his options in this situation was the fact that he wholeheartedly believed the police were the enemy, a belief that led to his inability to seek assistance from this source.

Matters quickly escalated. Several group members were arrested for a range of criminal activities they had perpetrated in the process of achieving group goals. Kane visited Matthew at his place of employment and provided him with weapons that he told him to use against the police if they visited his residence. Although desperate to disengage from the group, Matthew could see no way that he could stop the escalation of events and protect himself and his loved ones. He reported intense fear and marked depersonalization.

After a day of intense contemplation of his problem situation, Matthew met with Kane on June 1, 1997 at 1:00 a.m. in a semi-isolated area in a suburb of Hobart. Kane was reported to be agitated and threatening. Despite Matthew’s suspicion that Kane was being protected by his “ninja army,” and that Matthew would not escape their wrath, he repeatedly shot, and killed Kane as he stood on the street. Matthew stated that he believed at the time that he had no choice but to do what he did because his life and the lives of his family were at risk.

All the victim’s known associates were interviewed at the police station in Hobart. During routine questioning, the police asked Matthew whether he had killed Kane. Matthew acquiesced and confessed to the crime, despite the fact that the police had no evidence or particular suspicion that he was the perpetrator.

Although Matthew pleaded self-defense, he was found guilty of murder in the Tasmanian Supreme Court in Hobart and sentenced to 18 years’ imprisonment. However, with recent sentences ranging from 21 years to 25 years (information available at, this sentence was less severe than what most others received. Matthew is currently a maximum-security prisoner. He has accessed psychological support within the prison.

In summary, Matthew was an unassuming young man with a good work ethic whose history of geographic mobility made it difficult for him to develop good and lasting peer relationships. He had no history of criminal behavior, and he appeared to have been relatively well adjusted psychologically, albeit unassertive and somewhat isolated. His interest in martial arts brought him into contact with the leader of the group of which he was to become a member. Following his membership, evidence suggests that his behavior radically changed: He demonstrated aggressive behaviors and engaged in criminal activities that were outside his realm of experience. In retrospect, Matthew finds it inconceivable that he acted in such a way. His belief in the power of the leader of the group was absolute. As a consequence, Matthew felt that he had to take the radical action of killing the leader to extricate himself from the group. On the basis of the above information, we consider the influence of the group to be cultic in nature.



Intellectual functioning. We administered the Wechsler Adult Intelligence Scale–III (WAIS-III) to Matthew to assess intellectual functioning and to calculate Total, Verbal, and Performance IQs. We used the American standardization norms.

Psychological functioning. The Millon Clinical Multiaxial Inventory-III (MCMI-III; Millon, Davis, & Millon, 1997) is a self-report measure of Axis I and Axis II psychopathology, compatible with the DSM-IV diagnostic system. We used this test to assess Matthew’s psychological functioning. This instrument has subscales that fall into four categories: clinical personality patterns (schizoid, avoidant, depressive, dependent, histrionic, narcissistic, antisocial, sadistic, compulsive, negativistic, masochistic); severe personality disorder (schizotypal, borderline, paranoid); clinical syndromes (anxiety, somatoform, bipolar, dysthymia, alcohol dependence, drug dependence, posttraumatic stress disorder); and severe clinical syndromes (thought disorder, major depression, delusional disorder).

The MCMI-III allows for the calculation of Base Rate scores (BR scores), standard scores that take into account the prevalence of particular attributes in the psychiatric population and that are an appropriate alternative to norm based T scores when measuring personality traits (Millon et al., 1997). We administered a range of other psychological tests as part of a larger research project, but we will not report those results here.

Group psychological abuse. The Group Psychological Abuse Scale (Chambers, Langone, Dole, & Grice, 1994) is a self-report measure of group processes for former members of various groups that provides subscales relating to compliance, exploitation, mind control, and anxious dependency. The scale measures how characteristic each of 28 functions is on a scale from 1 (not at all) to 5 (very characteristic). We obtained normative data from a sample of former American cult members.

Assessment of Peri-Homicidal Reactions

The examination of peri-homicidal reactions represents an assessment challenge. Although recall of events might be enhanced through hypnotism (e.g., Scheflin, Spiegel, & Spiegel, 1999) or abreaction interviewing techniques that rely on the administration of disinhibiting drugs (Kopelman, Christensen, Puffett, & Stanhope, 1994), the methods used mask the psychophysiological arousal that would normally occur in response to recollection. To overcome these problems, guided imagery has been used to recreate the memory of a past event so that psychophysiological reactions and psychological responses to such an event can be examined. Success with this type of procedure has been achieved, for example, in the examination of traumatic events (e.g., Blanchard, Hickling, & Taylor, 1991; Holmes, Williams, & Haines, 1998; Orr et al., 1998; Shalev, Orr, & Pitman, 1993) and acts of self-injury (Brain, Haines, & Williams, 1998; Haines, Williams, Brain, & Wilson, 1995; Wells, Haines, Williams, & Brain, 1999).

The use of guided imagery is based on the propositions of Lang (1979), who suggested that cognitive events (thoughts about a behavior) result in measurable efferent outflow. It has since been established that personalized imagery depicting actual events the individual has experienced result in a more appropriate response to imagery content than does standard imagery wherein individuals are asked to image an event they have not experienced (e.g., Haines et al., 1995; Pitman et al., 2001; Shin et al., 2000). The function of personalized guided imagery is to recreate the memory of an event, present the imagery to the individual in a structured way, and measure the efferent outflow and the psychological response. In addition, by structuring the imagery into identifiable stages, it is possible to examine the development of the reaction over the course of an event.

Although the reliability of memories elicited by guided imagery has been criticized by some (e.g., Schacter, 2001), others have reported that the content of traumatic memories remains unchanged over time (Koss, Tromp, & Tharan, 1995). This methodology has been successfully applied to the examination of homicidal behavior (e.g., Glading, Williams, & Haines, 2001; Glading, Williams, Haines, & Sale, 2001; Haines, Williams, Sale, & Glading, 2001; Williams & Haines, 2001; Williams, Haines, & Casey, 2000; Williams, Haines, Sale, & Glading, 2001).

Imagery Scripts

To elicit Matthew’s reactions to the homicide, we used a personalized, staged, imagery script.  In addition, we used two personalized imagery scripts for comparison purposes; an aggressive interpersonal interaction script that did not result in violence, and an emotionally neutral script.

We used personalized imagery scripts to guide the perpetrator (Matthew) through the images and scenes and to assist him in re-experiencing the psychological and psychophysiological reactions that he had at the time of an act, be it neutral or violent. We drafted scripts using information Matthew had detailed. The script interviews consisted of in-depth descriptions of the three scenes in terms of environment, cognitions, behaviors, emotions, and psychophysiology. Personal reactions were included in the scripts to help trigger personal images. The emphasis was placed on using Matthew’s own view of what happened. We audio-taped the interviews to aid us in preparation of the scripts. In developing the scripts, we used only information that Matthew provided.

We divided each script into five stages (setting the scene, approach, incident, consequence, and resolution), which represented a sequence of moments from immediately before the event of interest to the time immediately after the event of interest. Each stage took approximately 60 seconds to verbally administer. The content of each script is outlined in Table 1.

Table 1

The Content of Each Stage of Each Script






Matthew climbed a hill to meet with Kane in the early hours of the morning.


Kane was erratic and threatening. A discussion took place.


Matthew repeatedly shot, and killed the victim.


Matthew stood over the victim to ensure that he was dead.


Matthew fled the scene.



Matthew at work. Problem worker late back from lunch.


Matthew talked to customer about problem worker as truck was being loaded.


Matthew went into office, threw keys on bench, and said he was going to quit job.


Matthew picked up gear from behind the bench with intention of leaving the workplace.


Matthew walked across car park toward motorbike and away from workplace.



Matthew stood in prison yard waiting for lunch.


Custodial officers prepared to call prisoners to parade.


Signal given and men lined up for parade.


Names called, then people file off to go to lunch.


Matthew walked past custodial officer and out of yard toward lunch room.

Visual Analogue Scales

We used visual analogue scales (VASs) (McCormack, Horne, & Sheather, 1988) using bipolar dimensions of not anxious-anxious, not angry-angry, unreal-real, not fearful-fearful, not agitated-agitated, and not guilty-guilty to record Matthew’s subjective reactions to the imagery. The scales were scored from 0-100 (with a higher score reflecting a more negative experience) on a 100mm line. Matthew’s psychological response was considered negative if the rating he gave was higher than 50, moderate for a rating from 65 to 75, marked for a rating from 75 to 85, and extreme for a rating from 85 to 100.


We made psychophysiological recordings using an Acer TravelMate 514T computer linked to a PowerLab 4/20 portable data-acquisition system using Chart 4.0.1. Although we made multimodal recordings, we will report heart rate here. We recorded heart rate using Unilect high resolution, Ag/AgCl adhesive ECG electrodes at the second rib on either side of the torso, with an earth reference on the mastoid process.


Two of us collected the data as part of a larger study that had the approval of the University Human Research Ethics Committee and the Department of Justice and Industrial Relations. We obtained written informed consent from the participant, Matthew.

We interviewed Matthew at the prison hospital at Her Majesty’s Prison Risdon to obtain the information required for script construction. We audio-taped this interview. We administered the WAIS-III to Matthew and gave him the MCMI-III and the Group Psychological Abuse Scale to complete.

We obtained information about the event from Matthew before trial, although the actual assessment did not occur until after trial. In general, deterioration of memory has been cited as a problem with retrospective evaluation of responses to events. However, research evaluating the efficacy of traumatic memories indicates that little change occurs in either the factual or the emotive nature of these types of memories (e.g., Koss et al., 1995).

One of us prepared the scripts from the audio-taped interview. At the second session, we applied electrodes using standard electrode placements. We took a 60-second baseline recording as Matthew sat with his eyes closed. One of us then administered each stage of the first script, while the other monitored the recording of the psychophysiological data. There was a 10-second pause between each stage, during which Matthew was allowed to open his eyes. At the end of the script, we asked Matthew to rate his psychological response to each stage of the preceding script on the VASs. To facilitate these ratings, we reiterated key elements of each stage. We administered subsequent scripts in the same way.

We debriefed Matthew at the end of the session. Further, we visited him on another occasion, when we described the results of the assessment and did additional debriefing.

Data Transformation, Scoring, and Analysis

We took a 30-second scoring period from the baseline and during each stage of each script. We took this scoring period from approximately 15 seconds into each stage. This scoring method has been successfully used elsewhere (e.g., Haines et al., 1995; Haines, Josephs, Williams & Wells, 1998). We obtained a mean heart rate for the scoring period.

We handled the data in a number of ways. We have presented descriptive data from psychological testing with either an indication of clinical significance or a comparison with normative data. We have described across-stage responses to imagery relative to script content. For the VASs, we determined between-script differences at each stage to be significant if they were more than 3.29 standard deviations above the mean of the stages of the neutral script. (See Tabachnick & Fidell, 1989, for a discussion of the identification of outliers.)

To provide a more comprehensive assessment of the heart-rate data, we used time-series analysis to evaluate changes from baseline and in response to the various components of the imagery scripts. Because recording was discontinued between scripts, we performed separate analyses for the three scripts.


Intellectual Functioning

Matthew obtained a Full Scale IQ of 95, which placed him in the average range together with 50% of the general population. He obtained a Verbal IQ of 91, which placed him at the lower end of the average range. Results included some variability of scores related to verbal skills. He obtained better scores in relation to general information and arithmetic abilities than those he obtained for social comprehension, vocabulary, and verbal abstract thinking. Matthew obtained a Performance IQ of 114, which placed him in the high average range together with 16.1% of the general population. Again, there was some variability in these scores, with the highest scores relating to constructional abilities and the lowest to processing visual information. His low average verbal score is consistent with what might be expected for conditions of environmental/educational impoverishment.

Psychological Adjustment

Table 2 presents Matthew’s final BR scores for each of the subscales of the MCMI-III. Clinically significant scores are indicated. The validity index indicated that this protocol was valid. The disclosure index, desirability index, and debasement index all were within normal limits. The results indicate that Matthew was an isolated, dependent, depressive, and self-defeating young man who would have experienced difficulties with relationships. At the time of assessment, Matthew was anxious, was experiencing a low-grade depression, and was reporting symptoms of posttraumatic stress that were consistent with his experiences at the time of the homicide.

Table 2

The Final BR Scores for the MCMI-III




Clinical personality patterns















Sadistic (aggressive)




Negativistic (passive-aggressive)


Masochistic (self-defeating)


Severe personality disorder







Clinical syndromes





Bipolar: Manic




Alcohol dependence


Drug dependence


Posttraumatic stress disorder


Severe clinical syndromes

Thought disorder


Major depression


Delusional disorder


*clinically significant personality traits

**clinically pervasive personality pathology

#clinically prominent syndrome

Group Psychological Abuse

We compared the scores Matthew obtained on this test with those scores three other group members obtained, and with the normative sample of former cult members’ characterizations of their groups. Because of ethical constraints, we are unprepared to publish the scores obtained from the other three group members; suffice it to say that the results were comparable with Matthew’s results. Table 3 presents the scores Matthew obtained on each of the subscales.

Table 3

Scores Obtained on the Group Psychological Abuse Scale, Together with the Mean Scores and Standard Deviations for the Normative Sample (Chambers et al., 1994), and the Degree to Which the Case Scores are Characteristic of Cultic Groups


Participant Score

Normative Sample













Very characteristic

Mind control





Anxious dependency





Total scale




Psychophysiological Response to Imagery

Figure 1 presents Matthew’s mean heart-rate scores for each stage of each script. From baseline, he demonstrated a substantial increase in heart rate to the scene stage, in which he recalled an image of the climb up a hill to meet Kane for a discussion. Matthew believed at the time that it might be necessary to defend himself and kill Kane. In the approach stage, we provided a description of the meeting with Kane, and this coincided with a reduction in Matthew’s psychophysiological arousal. During the incident stage, he recalled firing his shotgun at Kane, nearly losing control of the gun, and then repeatedly firing at the victim. The consequence stage coincided with Matthew standing over the victim, feeling relief that it was over and that Kane was no longer a threat. In the resolution stage, Matthew left the scene. The final two stages of imagery were associated with the further reduction of his heart rate from it rate during the incident stage.

Matthew’s pattern of response to the homicide script is markedly different from his response to the aggression script that described a stressful interpersonal interaction at work. In that case, his arousal increased from baseline to the setting-the-scene stage and remained high thereafter.

Figure 1. Matthew’s mean heart rate for each stage of each of the three scripts.

The neutral script elicited from Matthew the lowest level of arousal, as we had anticipated. Interesting to note is that his level of arousal in the final two stages of the homicide script, when Matthew described being relieved, is not markedly different from his level of arousal to the neutral script, which involved his being on parade in the prison yard.

To provide a more comprehensive analysis of the data, we divided the recording periods into 2-second intervals, and we estimated the heart rate for each interval. Because we discontinued recording between scripts, we fitted separate time series to the data for each script. We used interrupted time-series analysis, modeling the underlying intervention effect of the imagery at each individual stage by a step function (whose size is designated by omega). Thus, beginning with the baseline, there are five steps between adjacent stages, each of which can be tested for statistical significance. The positive and decreasing lagged autocorrelations, together with a single, large, partial autocorrelation for each series, suggested an autoregressive model with a single parameter, which indicates dependence of each observation on the one preceding, but without the necessity of considering further lagged dependencies. This identification was supported by non-significant autocorrelations in the residual series, the residuals being approximately normally distributed. The various parameters in the first-order regression equation are the constant term, which reflects the average level of the series; the first-order regression coefficient p (1), which indicates the degree of dependency of each heart rate observation on the preceding one; and the step size omega (1) to omega (5). Table 4 shows the estimated values of these parameters and the associated t tests for statistical significance for the three scripts.

Table 4

Parameters in the First-Order Autoregression Model for the Time Series Representation of Heart-Rate Responses to the Neutral, Aggression, and Homicide Scripts

Homicide script

Aggression script

Neutral script

















p (1)







omega (1)







omega (2)







omega (3)







omega (4)







omega (5)







Note. Omega (1) to omega (5) constitute steps in heart rate between adjacent stages: baseline, setting the scene, approach, incident, consequence, resolution.

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

For all scripts, the imagery produced a significant overall increase in Matthew’s heart rate. In the homicide script, this increase was followed by a significant reduction in the rate from the scene to the approach, with a further reduction in rate that approached significance from the consequence to the resolution stages. For the aggression script and neutral script, there were no significant steps after the first. A trend analysis demonstrated a substantial and significant decreasing linear trend across the stages of the homicide script from the scene stage to the resolution stage, t(263) = 6.19, p<.001. There was a smaller but significant decreasing trend across the same stages of the aggression script, t(283) = 2.94, p<.01. The peaking of arousal at the beginning of the homicide script is an interesting manifestation of Matthew’s reaction to the imagery.

Psychological Response to Imagery

Figure 2 presents the ratings for the VAS not fearful-fearful for each stage of each script. Matthew reported intense fear in the time leading up to the homicide and during the homicidal act itself. The relief that Matthew reported during the consequence stage when he realized that his problem was solved gave way to increased fear when he realized that supporters of the victim might be waiting to attack him and that he needed to leave the scene.

Matthew’s pattern of response to the homicide script was noticeably different from that associated with the aggression script. Although the distressing nature of the aggressive interaction elevated his fear ratings, these ratings were only moderate at the highest level, at the incident stage when he imaged expressing his anger about a work colleague.

The neutral script elicited little from Matthew in relation to fear ratings, although a very small increase was noted at the resolution stage when he was leaving the prison yard to go to lunch after parade. However, this elevation in response still increased fear ratings only to a very low level.

When considering the differences between scripts, we noted that all stages of the homicide script elicited higher ratings of fear in Matthew than did the neutral script.  A comparison of his responses to the aggression script and the neutral script demonstrated elevated ratings at the incident, consequence, and resolution stages.

Figure 3 presents Matthew’s ratings for the VAS not agitated-agitated for each stage of each script. The moments leading up to the homicide and the actual homicide itself elicited the greatest feelings of agitation in him. Again, the ratings at the consequence stage were low and coincided with imagery of the immediate aftermath of the homicide when he realized that he was safe from the victim. Matthew reported increased agitation to a moderate level during the resolution stage, when he reported that it became imperative that he leave the scene in case supporters of the victim were present.

Matthew’s levels of agitation to the stages of the aggression script were extreme, with the exception of the approach stage, when he reported trying to come to terms with the situation by discussing it with another person, who agreed that the target of Matthew’s animosity was worthy of that animosity. The neutral script did not elicit feelings of agitation from Matthew, and there was no variation across the stages of this script.

Figure 2. Matthew’s ratings for the VAS not fearful-fearful for each of the three scripts.

Figure 3. Matthew’s ratings for the VAS not agitated-agitated for each stage of the three scripts.

For the between-script differences, Matthew demonstrated elevated ratings of agitation for all stages of both the homicide and aggression scripts relative to the neutral script.

Figure 4 presents Matthew’s ratings for each stage of each script for the VAS not anxious-anxious. In relation to the homicide script, he demonstrated an increase from a moderate level of anxiety at the scene stage to a peak of extreme anxiety at the incident stage. During the consequence stage, he reported his anxiety to be low, and this coincided with imagery relating to his standing over the body, feeling relief that the threat had passed. His return to an extreme level of anxiety at the resolution stage was associated with an awareness that danger might still be present, and of a need to leave the scene.

In contrast, the aggression script elicited a different pattern of response in Matthew, with an initial high level of anxiety in the scene and approach stages, escalating to an extreme level of anxiety from the incident stage onward, with no reduction over the course of the final three stages. The neutral imagery script did not elicit feelings of anxiety in him.

When making comparisons between scripts, we noted that both the homicide script and the aggression script elicited higher ratings of anxiety at all stages compared with the neutral script.

Figure 4. Matthew’s ratings for the VAS not anxious-anxious for each stage of the three scripts

Figure 5 presents Matthew’s ratings for each stage of each script for the VAS not angry-angry. In relation to homicide imagery, Matthew’s anger increased from a moderate level at the scene stage to a peak of anger at a high level in the approach stage, in which imagery depicted the verbal interaction between Matthew and Kane. At the incident stage, Matthew’s anger again was at a moderate level; it decreased further during the consequence and resolution stages so that no anger was evident by the end of imagery presentation.

In comparison with the homicide script results, the aggression script elicited high levels of anger in Matthew that increased to an extreme level by the incident stage and remained at this elevated level for the remainder of the script. The pattern of his anger in reaction to the aggression script mirrored that of the pattern of his anxiety to the same script. Neutral imagery did not produce feelings of anger in him.

Between-script comparisons indicated elevated ratings of anger to aggression imagery at all stages relative to the neutral script. Elevated ratings to the homicide script also were evident at the scene, approach, incident, and consequence stages, but not at the resolution stage, where ratings of anger were comparable with the neutral script.

Figure 5. Matthew’s ratings for the VAS not angry-angry for each stage of the three scripts.

Figure 6 presents the ratings for Matthew for each stage of each script for the VAS real-unreal. Homicide imagery was associated with Matthew’s feelings of unreality that increased from a high level at the scene stage to an extreme level at the incident stage, with these feelings not abating for the remainder of the homicide imagery script stages.

Ratings of feelings of unreality were at a moderate level throughout the aggression script, with the exception of a peak at the incident stage when ratings increased to a moderately high level that was still lower than the lowest rating associated with the homicide script. Neutral imagery did not elicit feelings of unreality in Matthew.

When we make comparisons between scripts, we note that the homicide and the aggression scripts were associated with higher ratings of unreality than was the neutral script at all stages.

Figure 6. Matthew’s ratings for the VAS real-unreal for each stage of the three scripts.

Figure 7 presents Matthew’s ratings for each stage of each script for the VAS not guilty-guilty. From a moderate level of guilt at the scene stage of the homicide script, Matthew’s guilt was rated as extreme thereafter.

In reaction to the aggression script, Matthew’s ratings of guilt at the incident and resolution stages were higher, reaching a moderate level, although overall the ratings of his guilt to the aggression script were much lower than to the homicide script. Neutral imagery did not elicit feelings of guilt from Matthew.

Both the homicide and the aggression scripts were associated with elevated ratings of guilt at all stages in comparison with the neutral script.

Figure 7.  Matthew’s ratings for the VAS not guilty-guilty for each stage of the three scripts.


The case presented here was of a young man with average intelligence who found himself in extraordinary circumstances and who was ill-equipped to deal with the problem situation. A number of factors contributed to his inability to cope.

First, interviews about Matthew’s childhood history indicated that he lacked stability, both geographic and within the family, and this instability would have influenced his ability to develop and maintain long-term relationships with peers. This factor, in itself, might have created a need in him to belong to a stable and supportive group. We could speculate that the disruption to his childhood experiences was sufficient to have affected his intellectual development. In the absence of neurological injury, the discrepancy between performance and verbal IQ in favor of performance abilities has been identified as potentially attributable to environmental and educational impoverishment (Walsh, 1990), and linked with antisocial personality traits (Snow & Thurber, 1997) and delinquent or criminal  behavior (Cornell & Wilson, 1992; Walsh, 1992; Wong & Cornell, 1999).

Second, perhaps because of his early experiences, it was evident from psychological testing that Matthew demonstrated a range of personality traits that would have made him vulnerable to the influences of a group such as the one with which he became involved. He demonstrated tendencies to be depressive, schizoid, avoidant, dependent, negativistic, and self-defeating. These types of characteristics generally are pervasive and of long duration, so they probably pre-existed the homicide. These traits would have translated into a moody, isolated young man who craved acceptance. These personality characteristics might have seriously influenced the quality of his life choices and resulted in his not always acting in his own best interests. All this would have been compounded by the borderline personality pathology that was evident on testing. Borderline personality disorder is characterized by increased impulsivity and instability of mood, self-image, and relationships (American Psychiatric Association, 1994).

The third factor that would have influenced Matthew’s difficulty in adequately resolving the problem situation was the nature of the group of which he became a member. As the measurements of the Group Psychological Abuse Scale indicated, it was evident that the group had all of the characteristics of an abusive cult (see Chambers et al., 1994).

The first subscale of this test measures compliance, or the degree to which the group leader insists that the individual behave in a way the leader demands. Matthew obtained a compliance score that was comparable to those obtained by former cult members. The second subscale measures exploitation, or the unethical use of power within the group by the leader. Matthew obtained scores that were substantially higher than those obtained for the former cult members, indicating that he perceived (the GPA measures perceptions, not objective reality of the group) a very high degree of exploitation operating in this group.

The third subscale measures mind control, which assesses the nature of the relationship between the leader and the group members in terms of power and manipulation, and the extent to which the leader reduces members’ capacity to engage in critical thinking. The scores Matthew obtained were comparable to the normative scores for the group of former cult members.

The final subscale measures anxious dependency, which describes the nature of the physical, psychological, and spiritual needs of the group as defined by the leader. The dependency resembles an unconditional bond, and removal of favor by the leader results in anxiety in members of the group. In a cult/group situation, dependency can be absolute, and the fear associated with this dependency can color all member experiences. Matthew obtained a score that was comparable with the normative cult members’ scores.

Within the context of the manipulation of the leader and the negative influences of the group, Matthew engaged in a range of criminal behavior despite having no history of such activity prior to joining the group. It is interesting that, similarly, the perpetration of serious criminal activity by Aum Shinrikyo (Supreme Truth) members was reported to have occurred, in part, as a result of their obedience based on a fear that they themselves would be killed if they did not perform the criminal actions (Nishida, 2001). Matthew, too, feared for his life if he acted against group beliefs or thwarted the leader.

A confluence of particular circumstances would have created an intense conflict for Matthew: He was facing a need to disengage from the group; he had a constriction of problem-solving options; and he had a decreased ability to make decisions, particularly decisions that would move him away from the group’s belief system. These factors might have led him to believe that he had no choice but to take the life of the leader.

Clearly, the disrupted childhood, the personality pathology, and the influences of the group of which he was a member all contributed to influence this act of homicide. However, the study would be incomplete if we were to examine these factors without considering the peri-homicide reactions of the perpetrator to the act of homicide. We took these reactions into consideration by using a personalized, staged, guided-imagery methodology.

In Matthew’s response to homicide imagery, after a peak of arousal in the approach stage, he demonstrated a reduction in arousal. From the incident stage, this reduction in heart rate coincided with his self-report of intense depersonalization and extreme ratings of unreality as measured by the VASs. This pattern of heart-rate reduction associated with reports of dissociative experiences has been identified elsewhere by means of this guided-imagery methodology (Williams, Haines, Sale, & Doherty, 1999). Intense anxiety leads to depersonalization experiences as a mechanism for coping with the distress (Nuller, 1982; Williams et al., 1999). However, whereas criminal law covers complete dissociation under the insanity plea pertaining to automatism (http://www., depersonalization is not accommodated as easily; and there is debate about the extent to which depersonalization actually lessens the normal constraints on behavior and increases the likelihood of impulsive action (Frasquilho & Oakley, 1997; Kuley & Jacobs, 1988).

Nevertheless, the results in Matthew’s case have explanatory power in relation to his motive for homicide. The results are consistent with an explanation that Matthew was trying to solve a problem situation. The reduction in tension and negative affect following the act can be interpreted as an indication that Matthew felt safe from the power of the group leader after he was dead, and that Matthew did not have to fear for his life any longer.

To support this notion, he demonstrated consistent reductions in negative affect at the consequence stage, in the moments after the death of the victim. Feelings of fear, agitation, anxiety, and anger all reduced to markedly low levels at this stage. However, the impact of his actions, and the potential of ongoing threat from other group members who were believed to have been protecting the leader, resulted in an increase in ratings on these dimensions during the resolution stage, which was coupled with a desire to flee the scene.

In general, motive for homicide is understood in terms of an expressive-instrumental dichotomy (Decker, 1996). Expressive homicide is the result of an unplanned escalation of interpersonal conflict that results in the death of another person. Instrumental homicide has been described as a premeditated action that is carried out to achieve some goal such as robbery or rape.

It is not a straightforward matter to assign this particular homicide into an expressive or an instrumental category. Although the homicide was preceded by an escalation in distress and the worsening of a conflict situation that are characteristic of an expressive homicide, the degree of premeditation seems to preclude labeling this homicide expressive. Alternately, instrumental homicide is commonly associated with a more dispassionate attitude toward the homicidal action than was evident in relation to this homicide.

We might be able to classify this homicide as instrumental if we consider the nature of the factors driving the homicidal action. The dispassionate, instrumental homicide seems to be related to an internal motive. That is, personal gratification of impulses or desires leads to single-minded, unemotional, homicidal behavior. What might be in operation here is an instrumental homicide driven by an external motive — in this case, the need to solve a seemingly unsolvable problem. It might be that internally driven, instrumental homicide is associated with an unemotional response to the homicidal behavior, whereas externally driven instrumental homicide is associated with a hyper-reactive emotional and psychophysiological response.

Interesting to note is that instrumental homicide generally is associated with less psychological effect for the perpetrator (e.g., Pollock, 1999). That is, the perpetration of instrumental homicide is considered to be less traumatic for the perpetrator than is the commission of an expressive homicidal act. However, that may be the case only with internally driven instrumental homicide. Externally driven instrumental homicide might be associated with circumstances that would give rise to psychological maladjustment in the perpetrator. Certainly, in this case, Matthew currently is reporting anxiety, depressive symptoms, and reports of manifestations of posttraumatic stress that appear to be related to the group influences and the commission of the homicidal act. Presently, he responds with anxiety and agitation to perceived undue pressure to engage in activities that he would prefer not to do. This response is likely a residual effect of group membership but, unfortunately, the requirement to do non-preferred activities is a common scenario within a prison system.

Matthew reported intense feelings of guilt in relation to the perpetration of the homicide, both at the incident stage and at subsequent stages of homicide imagery. The issue of remorse, however, is more complicated. Matthew now finds it difficult to conceive that he ever was so strongly influenced by Kane that he felt pressured to commit a homicidal act or engage in other criminal behaviors. In contrast, he is relieved that Kane is dead and is no longer a threat to himself or others.

In general, the research results support Matthew’s descriptions of his fear of the victim. The intensity of the circumstances surrounding the homicidal act, including the leader’s manipulation and the escalation of criminal and police-related activity, created in Matthew a desire to be away from the group. At the time he felt he had no choice but to do what he did.


Amen, D.G., Stubblefield, M., Carmichael, B., & Thisted, R. (1996). Brain SPECT findings and aggressiveness. Annals of Clinical Psychiatry, 8, 129-137.

American Psychiatric Association (1994). The diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Aronoff, J., Lynn, S.J., & Malinoski, P. (2000). Are cultic environments psychologically harmful? Clinical Psychology Review, 20, 91-111.

Blanchard, E.B., Hickling, E.J., & Taylor, A.E. (1991). The psychophysiology of motor vehicle accident related posttraumatic stress disorder. Biofeedback and Self-Regulation, 16, 449-458.

Brain, K.L., Haines, J., & Williams, C.L. (1998). The psychophysiology of self-mutilative  behavior: Evidence of tension reduction. Archives of Suicide Research, 4, 227-242.

Chambers, W.V., Langone, M.D., Dole, A.A., & Grice, J.W. (1994). The Group Psychological Abuse Scale: A measure of the varieties of cultic abuse. Cultic Studies Journal, 11, 88-117.

Cornell, D.W., & Wilson, L.A. (1992). The PIQ>VIQ discrepancy in violent and nonviolent delinquents. Journal of Clinical Psychology, 48, 256-261.

Decker, S.H. (1996). Deviant homicide: A new look at the role of motives and victim-offender relationships. Journal of Research in Crime and Delinquency, 33, 427-449.

Frasquilho, F., & Oakley, D. (1997). Hypnotizability, dissociation and three factors of eating behavior. Contemporary Hypnosis, 14, 105-111.

Glading, J., Williams, C.L., & Haines, J. The psychophysiology of acts of murder. Paper presented at the 1st Forensic Psychology Conference, Sydney, Australia, February, 2001.

Glading, J., Williams, C.L., Haines, J., & Sale, I. Motives for female homicide: A comparison of situational conflict and chronic abuse. Paper presented at the 21st Annual Congress of the Australian and New Zealand Association of Psychiatry, Psychology and Law, Melbourne, Australia, November, 2001.

Haines, J., Josephs, S., Williams, C.L., & Wells, J.H. (1998). The psychophysiology of obsessive-compulsive disorder.  Behavior Change, 15, 244-254.

Haines, J., Williams, C.L., Brain, K.L., & Wilson, G.V. (1995). The psychophysiology of self-mutilation. Journal of Abnormal Psychology, 104, 471-489.

Haines, J., Williams, C.L., Sale, I., & Glading, J. Fearful problem solving: The influence of cult activity on subsequent homicidal behavior. Paper presented at the 21st Annual Congress of the Australian and New Zealand Association of Psychiatry, Psychology and Law, Melbourne, Australia, November, 2001.

Holmes, G.E., Williams, C.L., & Haines, J. Psychophysiological responses to posttraumatic imagery following road trauma: A comparison of Posttraumatic Stress Disorder and Acute Stress Disorder. Paper presented at the 9th World Congress of the International Organization of Psychophysiology, Taormina, Sicily, Italy, 1998.

Kimbrell, T.A., George, M.S., Parekh, P.I., Ketter, T.A., Podell, D.M., Danielson, A.L., Repella, J.D., Benson, B.E., Willis, M.W., Herscovitch, P., & Post, R.M. (1999). Regional brain activity during transient self-induced anxiety and anger in healthy adults. Biological Psychiatry, 46, 454-465.

Kopelman, M.D., Christensen, H., Puffett, A., & Stanhope, N. (1994). The great escape: A neuropsychological study of psychogenic amnesia. Neuropsychologia, 32, 675-691.

Koss, M.P., Tromp, S., & Tharan, M. (1995). Traumatic memories: Empirical foundations, forensic and clinical implications. Clinical Psychology: Science and Practice, 2, 111-132.

Kuley, N.B., & Jacobs, D.F. (1988). The relationship between dissociative-like experiences and sensation seeking among social and problem gamblers. Journal of Gambling Behavior, 4, 197-207.

Lalich, J. (1992). The cadre ideal: Origins and development of a political cult. Cultic Studies Journal, 9,1-77.

Lang, P.J. (1979). A bio-informational theory of emotional imagery. Psychophysiology, 16, 495-511.

Langone, M.D., & Chambers, W.V. (1991). Outreach to ex-cult members: The question of terminology. Cultic Studies Journal, 8, 134-150.

Lasaga, J.I. (1980). Death in Jonestown: Techniques of political control by a paranoid leader. Suicide and Life-Threatening Behavior, 10, 210-213.

Liddle, P.F. (1997). Dynamic neuroimaging with PET, SPET or fMRI. International Review of Psychiatry, 9, 331-337.

Lifton, R.J. (1991). Cult formation. Cultic Studies Journal, 8, 1-6.

Martin, P.R., Langone, M.D., Dole, A.A., & Wiltrout, J. (1992). Post-cult symptoms as measured by the MCMI before and after residential treatment. Cultic Studies Journal, 9, 219-250.

McCormack, H.M., Horne, D.J., & Sheather, S. (1988). Clinical applications of visual analogue scales: A critical review. Psychological Medicine, 18, 1007-1019.

Millon, T., Davis, R., & Millon, C. (1997). MCMI-III Manual (2nd ed,) Minneapolis: National Computer Systems.

Nishida, K. (2001). A social psychological analysis of Aum Shinrikyo’s criminal behavior. Japanese Journal of Social Psychology, 16, 170-183.

Nuller, Y.L. (1982). Depersonalisation: Symptoms, meaning, therapy. Acta Psychiatrica Scandinavica, 66, 451-458.

Orr, S.P., Lasko, N.B., Metzger, L.J., Berry, N.J., Ahern, C.E., & Pitman, R.K. (1998). Psychophysiologic assessment of women with posttraumatic stress disorder resulting from childhood sexual abuse. Journal of Consulting and Clinical Psychology, 66, 906-913.

Pitman, R.K., Lanes, D.M., Williston, S.K., Guillaume, J.L., Metzger, L.J., Gehr, G.M., & Orr, S.P. (2001). Psychophysiologic assessment of posttraumatic stress disorder in breast cancer patients. Psychosomatics, 42, 133-140.

Pollock, P.H. (1999). When the killer suffers: Post-traumatic stress reactions following homicide. Legal and Criminological Psychology, 4, 185-202.

Robinson, W.G. (1997). Heaven’s Gate: The end? Journal of Computer Mediated Communication, 3, NP.

Schacter, D.L. (2001). The seven sins of memory: How the mind forgets and remembers Boston: Houghton Mifflin.

Scheflin, A.W., Spiegel, H., & Spiegel, D. (1999). Forensic uses of hypnosis. In A.K. Hess & I.B. Weiner, (Eds.), The handbook for forensic psychology (2nd ed.), 474-498. New York: John Wiley.

Shalev, A.Y., Orr, S.P., & Pitman, R.K. (1993). Psychophysiologic assessment of traumatic imagery in Israeli civilian patients with posttraumatic stress disorder. American Journal of Psychiatry, 150, 620-624.

Shin, L.M., Dougherty, D.D., Orr, S.P., Pitman, R.K., Lasko, M., Macklin, M.L., Alpert, N.M., Fischman, A.J., & Rauch, S.L. (2000). Activation of anterior paralimbic structures during guilt-related script-driven imagery. Biological Psychiatry, 48, 43-50.

Snow, M., & Thurber, S. (1997). Cognitive imbalance and antisocial personality characteristics. Journal of Clinical Psychology, 53, 351-354.

Swartling, G., & Swartling, P.G. (1992). Psychiatric problems in ex-members of World of Life. Cultic Studies Journal, 9, 78-88.

Tabachnick, B.G., & Fidell, L.S. (1989). Using multivariate statistics. New York: Harper Collins.

Trimble, M.R. (1992). The schizophrenia-like psychosis of epilepsy. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 5, 103-107.

Tunnell, K.D., & Cox, T.C. (1995). Applying a subculture of violence thesis to an ongoing criminal lifestyle. Deviant Behavior, 16, 373-389.

Wada, T., Kawakatsu, S., Komatani, A., Okuyama, N., & Otani, K. (1999). Possible association between delusional disorder, somatic type and reduced regional cerebral blood flow. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 23, 353-357.

Walsh, A. (1990). Illegitimacy, child abuse and neglect, and cognitive development. Journal of Genetic Psychology, 151, 279-285.

Walsh, A. (1992). The P>V sign in corrections: Is it a useful diagnostic tool? Criminal Justice and Behavior, 19, 372-383.

Walsh, Y., Russell, R.J.H., & Wells, P.A. (1995). The personality of ex-cult members. Personality and Individual Differences, 19, 339-344.

Wells, J.H., Haines, J., Williams, C.L., & Brain, K.L. (1999). The self-mutilative nature of severe onychophagia: A comparison with self-cutting. Canadian Journal of Psychiatry, 44, 40-47.

Williams, C.L., & Haines, J. Psychophysiological characteristics of punitive interactions in a case of filicide. Paper presented at the 1st Forensic Psychology Conference, Sydney, Australia, February, 2001.

Williams, C.L., Haines, J., & Casey, S.L. (2000). Filicide and the insanity plea: The use of guided imagery. Paper presented at the 10th European Conference of Psychology and the Law, Limassol, Cyprus.

Williams, C.L., Haines, J., Sale, I., & Doherty, M.R. (1999). The psychophysiology of dissociative identity disorder. Paper presented at the 34th Australian Psychological Society Annual Conference, Hobart, Australia.

Williams, C.L., Haines, J., Sale, I., & Glading, J. Personality disorder and homicide: A comparison of psychopathic homicide and sexually-motivated homicide. Paper presented at the 21st Annual Congress of the Australian and New Zealand Association of Psychiatry, Psychology and Law, Melbourne, Australia, November, 2001.

Wong, M.T.H., Lumsden, J., Fenton, G.W., & Fenwick, P.B.C. (1994). Electroencephalography, computer tomography and violence ratings of male patients in a maximum-security mental hospital. Acta Psychiatrica Scandinavica, 90, 97-101.

Wong, W.K., & Cornell, D.G. (1999). PIQ>VIQ discrepancy as a correlate of social problem solving and aggression in delinquent adolescent males. Journal of Psychoeducational Assessment, 17, 104-112.

Wright, S.A. (1991). Reconceptualizing cult coercion and withdrawal: A comparative analysis of divorce and apostasy. Social Forces, 70, 125-145.