Document text
rs. nnwmmwrnunzerax M-s<...,ss.a,at<..~......._W_.m...»~ s.-W ...-JOURNAL OF INTERPERSONAL VIOLENCE / September I986 Z
identity. Although this kind of crime has existed throughout history Lunde,
1977!, the number of such murders has never been as high Ressler et al., I985!.
According to the 1984 FBI Uniform Crime Report, 22.1% of murders com-
mitted in the reporting year had an unknown motive as analyzed by law
enforcement. This figure takes on added meaning when it is compared to
earlier reporting figures. In 1976, murders with an unknown motive repre-
sented 8.5% of all murders, l7.8% in 198], and 22.1% in I984 or an increase of
160% in an 8-year period.
Such seemingly motiveless murders were first covered extensively by the
news media when New York City's Son of Sam" killer David Berkowitz
stalked victims, apparently chosen at random, and killed them with a .44
caliber pistol. Since then there has been considerable attention to these types
of murders. People fear becoming the next random victim of these violent,
often grisly crimes.
Sexual homicide results from one person killing another in the context of
power, control, sexuality, and aggressive brutality. The psychiatric diagnosis
of sexual sadism, sometimes applied to the victimizer, states that the essential
feature of this deviant behavior i.e., paraphilia! is the iniction of physical
or psychological suffering on another person in order to achieve sexual
excitement.
It has been difficult to gather dependable statistics on the number of sexual
homicide victims for several reasons: ! the victim is officially reported as a
homicide statistic and not as a rape assault Brownmiller, 1975; MacDonald,
1971!, ! there is a failure to recognize any underlying sexual dynamics in a
seemingly "ordinary" murder Cormier 8: Simons, I969; Revitch, 1965!,
! those agencies that investigate, apprehend, and assess the murderer often
fail to share their findings, curtailing the collective pool of knowledge on the
subject Ressler, Douglas, Groth, 8: Burgess, I980!, and ! conventional
evidence of the crimes sexual nature may be absent. ,
When law enforcement officials cannot readily determine a motive for
murder, they examine its behavioral aspects. In developing techniques for
profiling murderers, FBI agents have found that they need to understand the
thought patterns of murderers in order to make sense of crime scene evidence
and victim information. Characteristics of evidence and victims can reveal
much about the murderer's intensity of planning, preparation, and follow-
through. From these observations, the agents begin to uncover the murderers
motivation, recognizing how dependent motivation is to the killer's domi-
nant thinking pattems. In many instances, a hidden, sexual motive emerges,
a motive that has its origins in fantasy.
The role of fantasy in the motive and behavior of suspects is an important
factor in violent crimes, especially sexual murders Ressler et al., in press!. In
the last 20 years, the role of sadistic fantasy has been explored in several
studies Brittain, I970; Reinhardt, 1957; Revitch, 1965, 1980; West, Roy, 8c
4O:-t an
6'=: eg-=-
= . r
. E2»¬ E?
Z3=2:
ii;
%?>%
1%:
g?»
% ::
2:
§<
E;
FZCi /"
I5' ?cl:'2i?Zi
I <a2z'a:;,li:';'::,/:1"/
sf
2%:
45
§;
aé
I3
4.
?;'.'
zBurgess et al. / SEXUAL HOMICIDE
Nicholas, l978!, with MacCulloch and colleagues I983! suggesting thatsadistic acts and fantasy are linked and that fantasy drives the sadistic be-havior. Current realization of cognitive structures, which help maintainbehavior patterns Beck, 1976!, combine with investigation of sadistic fan-tasies Brittain, l970; MacCulloch, Snowden, Wood, 8: Mills, 1983; Ressleret al., l985!, criminal reasoning Yochelson 8: Saminow, 1977; Saminow,1984!, and Uiminal fantasy Schlesinger 8: Kutash, l98l!, and serve as pri-mary foundations for our conceptualization of a motivational model of
sexual murder.
THE STUDY
Many people have speculated on variousaspects of murder: epidemiolog-ical studies report on demographic data concerning victims and perpetrators Constantino, Kuller, Perper, 8: Cypress, 1977! and patterns of homicide Rushforth, Ford, Hirsch, Rushforth, 8: Adelson, l977; Wolfgang, 1958!;murderers have been categorized in terms of motive Revitch, 1965!, intent Kahn, 1971 !, number of victims Frazier, l974! and type of victim Oormier 8:Simons, 1969!. Our study of S6 sexual killers was not designed to examinemotivation, yet our research yielded rich descriptive data about what moved
these men to kill.The basis for the Patterns of Homicide Crime Scene Project, from whichthis article is derived, has been reported elsewhere in this joumal Ressler,Burgess, Douglas, Hartman, 8: D'Agostino, this issue!. The project can betraced to the early 1970s, when agents of the FBIs Behavioral Science Unit BSU! began, on an informal basis, to deduce certain offender characteristicsby examining crime scene information. As a result, a preliminary frameworkfor crime scene analysis and criminal profiling was formulated. Concurrentwith the development of the criminal profiling project, a study was proposedto analyze crime scene patterns. Using case record review, direct observation,and first-hand investigative interviews, the study would examine convicted,
incarcerated offenders.This law enforcement study focused on analyzing crime scene evidence inorder to identify the murderer. Data collection, which took place in variousU.S. prisons between 1979 and 1983, was performed by special agents of theFBI. The data set for each murderer consisted of the best available data fromtwo types of sources: official records I psychiatric and criminal records, pre-trial records, court transcripts, and/ or prison records! and interviews with theoffenders. The majority of offenders provided written consent to be inter-viewed. Interviews were all conducted in prisons with the cooperation of
officials at the various correctional institutions.
41
]OURNAL OF INTERPERSONAL VIOLENCE / September I986
Standard data collection forms were used. The forms not only provided
guidelines for interviewing subjects but also established a system of recording
and coding relevant data to permit computer analysis and retrieval. Informa-
tion was requested about the offender and his background, about the offense,
about the victim, and about the crime scene. Subjects were asked questions
about childhood, adolescent, and adult behaviors or experiences that might
be related to violence. In this article, we present a motivational model of
sexual homicide based on I! quantitative analysis of background data, and
! qualitative analysis of interview data from murderers.
This was an exploratory descriptive study of a smalLa_vailab_le sample of 36
sexual murderers. Because of the limitations of the study design, we present
critical variables not for generalization purposes, but as hypotheses for exam-
ination in subsequent research of sexual and motiveless murders. We have
no data on a comparable control group; thus these findings should not be
interpreted as showing a predictive role for certain childhood or adolescentexperiences. Instead, we use the data in developing a motivational matrix for
sexual murder.
FINDINGS
The Murderers
The 36 men in the study began their lives with certain advantages. Most of
them grew up in the 1940s and 1950s, a period when attitudes in the United
States favored oldest, white male children; all subjects were male, the major-
ity 3! were white, and many were eldest sons were only children, and 4
were adopted!. They were of good intelligence, with 29% classified in the
average range, 36% in the bright normal and superior range, and 15% in the
very superior range. These attributes fostered in the offenders a certain sense
of privilege and entitlement.
Initially, the majority of the men began life in two-parent homes. Half of
the mothers were homemakers; three-quarters of the fathers earned stable
salaries. Over 80% of the offenders described their family socioeconomic levels
as average self-sufficient! or better. Thus mothers were in the home raising
the children; fathers were earning stable incomes; poverty was not a factor in
the financial status of families.
Although the families initially appeared to be functional with both par-
ents present, problems were noted within the parents backgrounds. Families
had criminal 0.0%!, psychiatric 3.3%!, alcohol abuse 9.0%!, drug abuse
3.3%!, or sexual 6.2%! problems in their histories. It appears that parents
of these men were often absorbed in their own problems. Thus, while
being offered little guidance because of their parents preoccupation with
42II #1
5%,
ii;
.
;2.2".
Lt:
31
.
3;.
§¢
$5
5
2
it .
Ia
'2
Z?
"2
<
Burgess et al. / SEXUAL HOMICIDE
their troubles, the murderers as young boys were witness to these deviantrole patterns of criminal behavior, substance abuse, and poor interpersonal
relationships.ln 47% of cases, the father left the home before the subject was l2; in 43% ofthe cases at least one parent was absent at some time prior to the subject'sreaching age l8. This loss of the father required many of the offenders toadjust to a new male caretaker during childhood and adolescent years.Instability in the family residence was also noted in many cases 8%!. Inaddition, 40% of the subjects lived outside the family home before age l8.Locations included foster homes, state homes, detention centers, and mentalhospitals. The histories of frequent moving reduced the boys opportunitiesto develop positive outside relationships that might have compensated for
family instability.Examination of performance behaviors of the subjects revealed that despitetheir intelligence and potential in many areas, performance in academics,employment, and military was often poor. Only one-third did average orbetter in school, with 68% receiving a fair to poor academic rating. Themajority did not finish high school. Thus, although these men were intellec-tually bright, they did not perform to their abilities.There was confirming evidence of abuse in the childhood histories of the36 murderers. Physical abuse 3/31!, psychological abuse 3/51!, and sex-ual abuse l2/ 28! were noted. This reveals that many of the men experiencedsome type of childhood abuse. It is noteworthy that 25 of the 36 men had sometype of psychiatric assessment or confinement as a child or adolescent.
Behavioral IndicatorsOur analysis examined the results of a checklist of symptoms and behav-ioral experiences see Table l!. This checklist was derived from a standard listof self-report indicators, used in research on a wide variety of psychosocialstudies, and also included indicators of thinking patterns daydreams! andbehaviors derived from the FBI profilers' understanding of criminal behavior.However, readers should keep in mind that many of these behavioral symp-toms have no consistent definitions or ways of measurement. For example,there is no method of measuring a pattern of lying or masturbation.Childhood. An analysis of 24 checklist items indicates that over 50% of themurderers reported the following present in childhood: daydreaming 82%!,masturbation 82%!, isolation l%!, chronic lying 1%!, enuresis 8%!,rebelliousness 7%!, nightmares 7%!, destruction of property 8%!, fire-setting 6%!, cruelty to children 4%!, and poor body image 2%!.Adolescence. An analysis of 24 checklist items indicates that over 50% ofmurderers reported the following behaviors: assaultive to adults 84%!, rebel-
43
IOURNAL OF INTERPERSONAL VIOLENCE / September 1986
liousness 84%!, masturbation 82%!, stealing 81%!, daydreaming 81%!, iso-
lation 7%!, chronic lying 5%!, nightmares 8%!, poor body image 3%!,
cruelty to children 4%!, destroying of property 2%!, enuresis 0%!, and
firesetting 2%!.
Adulthood. An analysis of 24 checklist items indicates that over 50% of
murderers reported that during adulthood, the following behaviors were
present: assaultive to adults 86%!, daydreaming 81%!, masturbation 81%!,
isolation 3%!, rebelliousness 2%!, chronic lying 8%!, poor body image
2%!, stealing 6%!, and nightmares 2%!.
For descriptive purposes, we use the terms internal behaviors and external
behaviors. Internal behaviors include thinking pattems and experiences
within or unique to the individual; external behaviors are those overt actions
that can be observed by others. The internal behaviors most consistently
reported over the three developmental periods are daydreaming, compulsive
masturbation, and isolation. The external behaviors most consistently
reported include chronic lying, rebelliousness, stealing, cruelty to children,
and assault on adults.
ROLE OF FANTASY
The central role of daydreaming and fantasy in the lives of the 36 mur-
derers is critical to what motivated them to kill. Daydreaming has been
defined as any cognitive activity representing a shift of attention away from a
task Singer, I966!. A fantasy, as we define it, is an elaborate thought with
great preoccupation, anchored with emotion and having origins in day-
dreams. A fantasy is generally experienced as thoughts, although the indi-
vidual may be aware of images, feelings, and internal dialogue. Some people
may be conscious only of thoughts, whereas others are conscious only of
feelings. Fantasy is a normal way for adults as well as children to obtain and
maintain control of an imagined situation.
However, the level of fantasy development may differ among people and is
generally based on the individual's ability to identify certain thoughts as
daydreams, to articulate their content, and retrospectively to recall this con-
tent. Singer l966! observed that 96% of adults report that they daydream
several times a day, and Beres 961! noted that fantasy may either substitute
or prepare for action. For various groups of people, fantasies may be sadistic
MacCulloch et al., 1983!. It is not known how many people activate their
sadistic fantasies and in what context this may occur, but Schlesinger 8:
Revitch 980! caution that once the fantasy builds to a point where inner
stress is unbearable, the way for action is prepared.
Whereas psychological motives for violent behavior are usually concep-
tualized in the literature as having roots beginning with trauma, insult,
44.§M' 11155333? Q
2% 1?:X14
5
2
2322112: .:;;':'-/.- -Burgess et al. / SEXUAL HOMICIDE
TABLE 1Frequency of Reported Behavior Indicators in
Childhood, Adolescence, and Adulthood
for Sexual Murders
I Frequency
Childhood Adolescence AdulthoodBehavior n '76! n 96! n %!
Daydreaming 28 82 27 81 27 81Compulsive masturbation 28 82 28 82 27 81Isolation 28 71 26 77 26 73Chronic lying 28 71 28 75 28 68Enuresis 22 68 20 60 20 15Rebellious 27 57 25 84 25 72
24 67 22 68 21 52Destroying property 26 58 26 52 23 35Fire setting 25 55 25 52 25 28Stealing 27 56 27 81 25 55Cnlelty to children 28 54 28 64 27 44Poor body image 27 52 27 63 .26 62Temper tantrums 27 48 26 50 25 44Sleep problems 23 48 22 50 22 50Assaultive to adults 25 38 25 84- 28 86Phobias 24 88 23 4-3 24 50Running away 28 36 26 46 26 llCruelty to anirnals 28 36 26 4-6 25 36Accident prone 24 29 22 32 22 27Headaches 21 29 21 33 22 45Destroying possessions 25 28 23 35 28 35Eating problems 26 27 25 36 26 35Convulsions 26 19 24 21 23 13Self-mutilation 26 19 24 21 25 32Nightmares
NOTE: n = number of subjects with data.
and/ or overstimulation in early childhood, our thesis is different. We hypoth-esize that these men are motivated to murder by their way of thinking. Overtime, their thinking patterns emerged from or were inuenced by early lifeexperiences. For example, a child abused by an adult caretaker begins to thinkabout being hit every time an adult comes near him, dwelling on the hitting.He may imagine fantasize! about someone coming to help him by beating upthe adult. This thinking pattern may bring relief, because someone hasprotected him in his fantasy. In addition, while being abused the child maypsychologically remove himself from the pain. He may pride himself on hiscontrol over pain in the face of abuse; for example, while being beaten he doesnot inch or blink. This thinking pattern gives the child a sense of controland, as a result, tension is relieved. The child can increase or decrease terror
45
-nun ..-.-....._.,.»,.,,-..- W AMWW "~692-nan-JOURNAL OF INTERPERSONAL VIOLENCE / September 1986
with different levels of arousal through fantasy. Development of this type ofthinking pattern does not necessarily mean a child will grow up to be an
abuser; our example illustrates the role of fantasy and its development.In analyzing the data we obtained through interviews with the murderers,we attempted to link our quantifiable findings with indications from themurderers themselves of long-standing, aggressive thoughts and fantasiesdirected toward sexualized death. The findings suggest that these thoughtpatterns were established early and existed in a context of social isolation.Murderers were consciously aware of the central role of fantasy in their
lives and of their preference for fantasy over reality. Even those men unawareof this reported that their thoughts became retaliatory or vengeful when theyperceived themselves as being slighted, rejected, frustrated, or betrayed. Suchthinking becomes an important component in the maintenance of sexually
aggressive violent behavior.The central role that fantasy plays in the thinking patterns of these men isnoted in one of the subjects statements: "All my life I knew I was going to endup killing." It also was observed in the statement of a parent who, after herson was convicted of fetish robberies, feared the outcome of her sons moodi-
ness and isolation would be something really terrible and tragic."
Early Fantasy Development
It is important to keep in mind that not all children respond to theirenvironment with violent fantasies and not all children who fantasize vio-lence act out these fantasies. Nevertheless, from our interviews with the
murderers in our study, the high degree of egocentricity in the murderers
negative, aggressive, sexualized fantasy and play is revealed.As children, the murderers often thought of other children and family
members as extensions of their inner worlds. The revealed intermittent
awareness of the impact of their early childhood behavior on others. Theywere not influenced by the response of others to their behavior. It continuedand repeated itself. They recounted tying up a smaller child and scaring himor destroying another child's toy. A man who eventually beheaded his victimsdid not associate that action with his early childhood activities that involvedthe systematic decapitation of his sister's dolls. He saw his actions as aresponse to his annoyance with his sister, not to his desire to dominate, bully,
and hurt. Ritualistic play of tying up and scaring a younger child was not
associated to one murderers abuse as a child. His play was a reenactment.
Murderers recounting violent and sadistic behavior as adolescents were more
aware of the intentionality of their acts.The interviews with the offenders are remarkable in the absence of anyaccounts of positive childhood fantasies. However, it is unclear whether such
fantasies were actually nonexistent or whether early positive fantasies were
46
z
at.
31.
gt.
¢
¢
4»Burgess et al. / SEXUAL HOMICIDE
lost in later negative perspectives and behaviors. The following exampleillustrates the pervasive nature of the child's negative daydreaming:
l felt guilty for having those thoughts [toward family] and submerged them andbuilt up lots of hostility and then it got off into fantasy. . . . They should havenoticed it at school, so excessive was my daydreaming that it was always in myreport cards. . . . I was dreaming about wiping out the whole school.
Early Sexualization of FantasyThe childhood onset of sexual fetish interests was noted in the subjects.Several subjects described strong interest and attraction at approximately thea e of five to high heeled shoes, female underwear, and rope. The men weregaware of the carryover of sexualized fantasies about these items into adoles-cence and adulthood. When the subjects began to murder, these items took on
importance in ritualized aspects of their murders.E 'd nce of childhood sexualized play in the lives of the murderers wasvi ealso revealed by parents and caretakers. Some parents provided informationabout preschool sexual fantasies. In one case, a mother recalled finding hei. . . . . d f3- ear-old son with one end of a string tied to his penis and the other en oY !the string shut in a bureau drawer, leaning backward to exert a pullingtion on his penis The boy's behavior suggested he had engaged in suchsensa .activity earlier. It is most likely that this behavior was introduced initially byan older person; evidence suggests the boy's babysitter was responsible.
Aggressive Components of Early Fantasy
Wh n murderers were asked to describe their early favorite play activities,esome revealed a repetitive acting out of a core aggressive fantasy. Thesechildhood fantasies were so dominant that they became persistent themes inlay with other children or alone. In some cases, an original violation orP
assault was expressed. For example, one subject at age l5 took youngerboysinto the bathroom of his residential facility and forced oral and anal sex onthem, reenacting his own victimization at age l0 but reversing his role fromictim to victimizer However, he did not consciously connect this behaviorV .with his own earlier assaults. The assaultive rituals were his attempts at
mastery and control over people and situations.Another offender as an adolescent openly masturbated in his home, espe-cially in front of his sisters, using their underwear in his masturbation rituals.He appeared oblivious to the inappropriate nature of his acts, and wasoffended by his family's response, feeling that family members were rejecting,intolerant, and unfair. It is speculated that his behavior represented a hyper-
arousal state derived from a repressed memory from childhood.Aggressive content in the form of death themes was also noted for those1 .murderers who described their early fantasies. Death is an example of u ti-
4'7
JOURNAL OF INTERPERSONAL VIOLENCE / September I986
mate control. When directed toward oneself in childhood, it can be a counter-
reaction to overwhelming fear, and its dominance in thought and play reveals
the child's troubled state of mind.
In one case, the 12-year-old subject repeatedly played "gas chamber" with
his sister. This game required his sister to tie him up in a chair, throw an
imaginary switch, and when "gas" was introduced, the subject would grasp
at his throat, drop to the floor, writhe convulsively, and "die." This game
combined a sexual theme compression of the carotid arteries for a sexual
sensation! and death theme fantasy.
Early Expressions of Sexual
and Aggressive Fantasy
Early fantasies often give rise to behavior tryouts that are precursors to
criminal behavior MacCulloch, 1983!. These precursor behaviors have the
capacity to move the child into pain-inflicting acts and to break through in
subtle, as well as overt, ways. They may emerge as play-engagement be-
haviors with others i.e., bullying younger children or putting pins in the rug
for a sister to walk on! or actions involving only themselves i.e., tying ropes
around their necks or cutting parts of their bodies!. Such behavior was noted
in the reported cruelty to other children by the offenders as youths as well as
in the offender's earliest encounters with law enforcement. Although such
encounters are often dismissed as adolescent adjustment problems, they may
be evidence of an escalation in aggressive acts toward others.
Especially illustrative of such escalation of fantasy expression is the pre-
viously cited example of the S-year-old boy who was observed by his mother
with his penis tied to a bureau drawer. As a young adolescent, he was found by
his parents in the bathtub practicing autoerotic asphyxia with his penis and
neck tied to the cross-bar of the faucets. At age 14 his parents took him to a
psychiatrist after noticing rope burns on his neck. At age l7 this same subject
abducted a girl at knife-point, took her to a deserted area where he kept her all
night, and released her in the morning. The adolescent was apprehended and
then released; the charge on his record was "girl trouble." Of importance is
the offender's shift in the object of aggression from himself to a woman. Not
until late adolescence, when the offender began following women, confront-
ing them with a knife, binding them, and fondling them was the offender sent
to prison. After release from prison, his crimes escalated to the murder of three
young women by asphyxia.
The early expressions of aggressive fantasies were often painful memories
for the offenders to reveal and the ones, for various reasons, that were never
discussed. They may have realized that they could have controlled their
actions and that they were aware that they had crossed the line between
fantasy and reality. In cases where the men were not apprehended for their
early crimes, they learned they were not controlled by authority and that they
48:§
55%
ii?
31>?
;=2%;
%
z
Ex.
gs»
E
is
§§/
4
?;
4
e
X
:
2
2
Ax
El
1
v§, >1 , $5
Ja
§~ :~» ,am,
K rvlBurgess et al. / SEXUAL HOMICIDE
could act violently and kill with impunity. Of the 36, 10 murdered as juve-
niles, thus realizing that they had the power over life.
MOTIVATIONAL MODEL OF SEXUAL HOMICIDE
To illustrate our hypothesis of the various factors that inuenced the 36
sexual murderers to kill, we present a motivational model for understandingsexually oriented murder and sadistic violence. In addition to the data we
' f h' model. collected, the interviews with the murderers serve as a basis or t is
The murderers early development of an active, aggressive fantasy life day-dreams! combined with later sexual reinforcement compulsive masturba-tion! and increasing detachment from social rules of conduct social isolation!provide a framework that reinforces his subsequent violent behavior.The model has five interacting components emphasizing interre anon-h' n l! the murderer's ineffective social environment, ! child ands ips amo gadolescent formative events, ! patterned responses to these events, ! resul-tant actions toward others, and ! the killer's reactions, via a mental feed-
back filter," to his murderous acts see Figure l!.
l! Ineffective Social Environment
It is often suggested by child and family theorists that the structure andf f il and social interaction, especially in the way the childquality o am yperceives family members and their interaction with him and with each
th re im rtant factors in a childs development. For children growingo er, a poup, the quality of their attachments to parents and to other members of thefamily is most important in how these children later as adults relate to andvalue other members of society. Essentially, these early life attachments
h h'ld 'll sometimes called "bond.ing"! translate into a blueprint of how t e c i wi
' ituations outside of the family. Thus one of the primary functionsperceive s _of family life is to develop a child who has a positive bonding with his social
environment.In our population of murderers, this social bonding fails or becomes
1.narrow and selective. Caretakers either ignore, rationalize, or norma izevarious behaviors in the developing boy or, through their own problems. . . . , d I . e.g., criminal behavior or substance abuse!, support the child s eve opingdi ti ns and rojections "I was framed"!. People significant to the boy dostor o pnot provide nurture and protection; rather, they impose adult expectationson the boy Boys should be strong and take care of themselves"!. Adults arenonprotective and rioninten/ening on behalf of the boy. The boy may bepunished for a specific antisocial act but the social restriction does not regis-ter in an experiential and cognitive way; that is, the boy is reprimanded or
"All b into brought to court but he normalizes the behavior as, oys get
49
_| !URNAL OF INTERPERSONAL VIOLENCE / September I986
' at EIVIRUIIEII E l' PATTERIED RESPONSES 1lllEFFEETlllE SUCK
Ion-lntervenlnq
Support Distortions
hon rotectivr
__-_ l#_.
Child/Adolescent
lFOIIAIIVE EVENTS
l 69222
1 Physical
1 5e|ua|
1 P£!!l9!!£!!!1-!!!l!!!Negative sacral attachlent
Diminished emotional response
FEEDBACK FILTER
lustllres acts
Sorts out errors
Discovers increased
arousal states
Discovers rncreesed areas
ol oonrnancer power,
control
Knows hon to continue
ants without detection
or punishlent
QbilineelCruelty to llllli
Cruelty to children
J0yle55| hostile, aggressive,
repetitive play patterns
Disreqaro lor others
Firesettinq
Stealing
Destroying prooerty
llqnore Behavior '%-CRITIEAL PERSONAL TRAITS
§0 ll| isolation
Prelerenoe lor iUt0erOt|i attrvrtles
Fetishes
Rebellious
Aggressive
LyrnqW<z
Entitlenent I
l l
lgtgrgersonal failure
y inconsistent care and contact
Deviant garental lodelsC05hl1lVE HAPPINS I PIUCESSIIE
Persistent and repetitive
itzeetuis
Daydreans
Fantasies
Thoughts with strong visual
conoonent
lrqhtlares
lQ!!!!!l.!t!l9l!!
Strong, lrertrnq presuppositions
re: cause. ellect. probability
Deals lh ahsolutes
Generalizations
Iheess
Dominance Poverlcontrol
Revenge Death
Violence lortore
Rape/nolestation lutilation
lnlllctinq pain on others/sell
liueethetLs-ereusel_1e:e1§
Prruarrly aroused by hiqh levels ol
aggressive experience5@>la
Ye
if
?
Q
as
%T
Q
¥
equires high levels of stimulation
hCil0lS TOIARDS UTHERS/SELF
tdelesseossltdult
Assaultive behaviors
lurqllryArson
Abduction
Rape
Hurder lnon-seruall
Se:-oriented ourder
rape
torture/lutllatron
|_ necrophilia
Figure 1: Sexual Homicide: Motivational Model§
*1 .
5Q2
2.
5g
7
21?:
ao
5
F?
wen
..
1?.
é
a
¢
z
52.»
.2..1; Max:211:
3 / -11""
isBurgess et a1. / SEXUAL HOMICIDE
trouble." The ineffective social environment expands from caretakers toindividuals in a community whose work brings them into contact with theyoung person e.g., teachers, counselors, ministers, police!.
! Formative EventsThere are three factors that contribute to the formative events componentof our model. The first of these is trauma, in the form of physical or sexualabuse. The developing child encounters a variety of life events, some norma-tive e.g., illness, death! and others nonnormative. Those nonnormativeevents in the murderer sample include direct trauma physical and/ or sexualabuse! and indirect trauma witnessed family violence!. Within the context ofthe child's ineffective social environment, the child's distress caused by thetrauma is neglected. The child is neither protected nor assisted in recoveryfrom the abusive and overwhelming events; the external environment does
not address the negative consequences of the events.One assumption regarding early traumatic events is that the child'smemories of frightening and upsetting life experiences shape the child'sdevelop ng thought patterns. The type of thinking that emerges developsstructured, p .-erned behaviors that in turn help generate daydreams andfantasies. The literature on children traumatized by sexual and physicalabuse and by witnessing violence reports the occurrence of dreams, night-mares, and disturbing memories of the trauma Burgess 8c Holmstrom, 1974,1979; Conte, 1984; Pynoos 8: Eth, 1985!. Other studies have documented thesechildren engaging in painful, repetitive acting-out of the traumas Axline,1969; Gardner, 1971; Terr, 1979, 1981a, 1981b, 1983!. Play of emotionallydisturbed and troubled children often contains conflicted and. obsessivethemes, contrasting with the creative and flexible themes noted in nondis-turbed children. We believe the traumatized child's play remains fixed onthoughts associated with the traumatic event and is held separate or encap-sulated Hartman 8c Burgess, in press! rather than integrated in play activitiesor in art expression through drawings Wood, Burgess, Sc McCormack, inpress!. Successful resolution of traumatic events results in the child being ableto talk about the event in the past tense and with equanimity. Unsuccessfulresolution of the trauma underscores the victim's helplessness often with theemergence of aggressive fantasies aimed at achieving the dominance andcontrol absent from reality Burgess, Hartman, McCausland, 8: Powers, 1984;
MacCu1loch et a1., 1983; Pynoos 8c Eth, 1985!.A second assumption regarding early traumatic events is that manifesta-tions of the impact of distressing events, such as direct sexual and physicalabuse, are influential in the child's $042131 development BUTQCSS Ct 31., 1984;Come, 1984; Pynoos Sc Eth, 1985!. Concurrent with the abusive event, thechild may experience a sustained emotional/physiological arousal level.When this sustained arousal level interacts with repetitive thoughts about the
51
_lOURNAL OF INTERPERSONAL VIOLENCE / September I986
trauma, the child's perceptions and patterns of interpersonal relationships
may be altered. For example, the child may show hyperaggressive behavior by
striking out at parents or repeatedly assaulting a favorite pet.
The second factor contributing to the formative events component of our
model is developmental failure. For some reason the child does not readilyattach to his adult caretaker. This is the child who does not listen or respond
to any limit setting and who often is described as aloof, cold, and uncaring. As
a result of this negative social attachment bonding!, the taretalter has no
influence initially over the child and later over the adolescent. In cases where
the child has been psychologically deprived or neglected, he may feel a
diminished emotional response.
Interpersonal failure, the third factor in this model component, is thefailure of the caretaking adult to serve as a role model for the developingchild. There are various reasons for this failure including the caretaker being
absent or serving as an inadequate role model e. g., a parent with problems of
substance abuse or an abusive parent!. The child may experience a violent
home environment where he sees aggression drunken fights! associated with
sexual behavior of adults.
! Pattemed Responses
The patterned responses component of the motivational model includes
two subcategories: I! critical personal traits, and ! cognitive mapping and
processing. These subcategories interact with each other to generate fantasies.
Critical Personal Traits
In the normal growth and development of a child, positive personalityuaits of warmth, trust, and security help establish the child's relationships
with others. These critical traits, in combination with an effective socialenvironment, allow the child to develop competence and autonomy.
In the murderer group, there was a propensity for the 36 men to develop
negative rather than positive personal traits. These negative personal traitsinterfere with the formation of social relationships and the development of an
emotional capacity within the context of human encounters. Increased social
isolation encourages a reliance on fantasy as a substitute for human en-
counter. In turn, individual personality development becomes dependent onthe fantasy life and its dominant themes, rather than on social interaction.Without human encounters and negotiations, there is failure to develop the
corresponding social values, such as respect for others lives and property.
The personal traits critical to the development of the murderers in ourstudy include a sense of social isolation, preferences for autoerotic activities
and fetishes, rebelliousness, aggression, chronic lying, and a sense of entitle-ment. The offenders chronic lying underscores their lack of trust and com-mitment to a world of rules and negotiation. Rather, distrust and a sense of
52- 1%,
We "as
-, §.. If ,. .,§'2 E93Z92»t
e. :~
st:
is ..
5' i
g;§.§§37
3
inW
4..
ac/v
3.
é.
36.5:.
z
E4;
Pf;
I /i.;.
2
..
/is£5».
$2:
aw-»1.5Z~:
>1
Z
23
Q :
3:
$
s v .
Q7.
%2:.
3'
1;
/
ii
7
2
4
/
,1
v
,<
?
J
1
aMBurgess et al. / SEXUAL HOMICIDE
entitlement to whatever they can get dominate their perceptions. Their social
isolation and aggression interact, restricting sexual development based on
caring, pleasure, and companionship. Because they are so isolated, the men
have little opportunity for interpersonal experiences that might modify their
misconceptions about themselves and others. Their personal affective lives
become dependent on fantasy for development. ln tum, fantasy becomes the
primary source of emotional arousal and that emotion is a confused mixture
of sex and aggression.
Cognitive Mapping and Processing
Cognitive mapping refers to the structure and development of thinking
patterns that give both control and development to one's intemal life e.g.,
one's sense of self and beliefs about the world! and link the individual to the
social environment e.g., one's interpretation of others!. The process of
cognitive mapping generates the meaning of events for an individual and
mediates sensory arousal pattems. Additionally, it is a filtering system that
allov s for interpretation of new information e.g., l'm always being singled
out", It " -ny life and l can live it my own way !. Cognitive mapping and
processing are lmed at self-preservation and equilibrium through the reduc-
tion of the negative affects of helplessness, terror, and pervasive anxiety.
In the murderers, the mapping is repetitive and lacking socially enhancingcognitions, moving the individual to an antisocial position and view of the
world. What emerges is a primary sense of entitlement to express oneself
regardless of its impact on others. The thought and action are justified
through the cognitive mapping of the murderer. The individual does not
experience a positive impact with the social environment. This occurs
because his fantasies and thinking patterns are a substitute for social relation-
ships. They are designed to stimulate and reduce tension. A sense of self is
developed and bolstered by the fantasies. The self-image is terrifying to
imagined others and contributes to further social isolation. The process
continues and becomes the primary source of energy for the psychological life
of the individual. Imagined outcomes of control and dominance over others
become a substitute for a sense of mastery of internal and external experience.
Parallel with the repetition of ideation of cognitive mapping is the neuro-
hormonal influence on sensory arousal levels. The neurohorrnonal basis of
the pleasure associated with aggressive fantasy activity is unknown. There is,
however, substantial evidence that stressors elicit a central nervous system
CNS! opiod responses in both animals stress-induced analgesia! and human
beings van der Kolk, Greenberg, Boyd, 8c Krystal, 1984!. Elevated plasma
levels of endogenous opiates have been documented in marathon runners
Colt, Wardlaw, 8: Frantz, 1981!, individuals who have undergone surgery
Cohen, l982!, and patients who engage in self-mutilation Coid, l983!. Thissuggests that the source of the stressor may be external or internal. It also has
53
JOURNAL OF INTERPERSONAL VIOLENCE / September I986
been reported that removal of the stressor may be associated with opiate
withdrawal-like symptoms: anxiety and irritability Backland, 1970; van der
Kolk et al., I984!. In a recent analysis of the psychobiology of posttraumatic
stress PTS!, van der Kolk et al. 1984! have suggested that the stress-approach
behavior consistently displayed by individuals who suffer from PTS disorders
may involve a conditioned CNS opiod response followed by withdrawal
hyperreactivity. By analogy, compulsive aggressive fantasy activity may also
involve such a psychobiological mechanism. In this case individuals re-
expose themselves to traumatic situations through fantasy activity. The
internally induced stressor elicits the opiod response, which brings relief
and/or pleasure as well as avoidance of the noxious symptoms of opiate
withdrawal.
Structures of cognitive mapping and processing include daydreams,
nightmares, fantasies, and thoughts with strong visual components. There is
internal dialogue of limiting beliefs regarding cause, effect, and probability.
The subjects deal in absolutes and generalizations. The themes of their
fantasies include dominance, revenge, violence, rape, molestation, power,
control, torture, mutilation, inflicting pain on self/others, and death. High
sensory arousal levels become the preferred state. The preoccupation with the
aggressive themes, the detailed cognitive activity, and elevated kinesthetic
arousal state eventually move the person into actions.
! Actions Toward Others
Childhood actions are based on the child's regard and caring for others as
well as on self-respect and flexibility. In other words, behavior patterns reflect
the private, internal world of the child.
Interviews with the murderers in our study revealed that their internal
world is often preoccupied with troublesome, joyless thoughts of dominance
over others. These thoughts are expressed through a wide range of actions
toward others. In childhood, these include cruelty toward animals, abuse of
other children, negative play patterns, disregard for others, firesetting, steal-
ing, and destroying property. In adolescence and adulthood, the murderer's
actions become more violent: assaultive behaviors, burglary, arson, abduc-
tion, rape, nonsexual murder, and finally sexual murder involving rape,
torture, mutilation, and necrophilia.
The early expression of cruelty toward both animals and humans when
not intervened and stopped, we believe, sets the stage for the future abusing
behavior in two ways. First, the early violent acts are reinforced, as the
murderers either are able to express rage without experiencing negative con~
sequences or are impassive to any prohibitions against these actions. Second,
impulsive and erratic behavior discourages friendships. The failure to make
friends leads to isolation and interferes with the ability to resolve conflicts, to
54a 2
_ = ;:z:;:5z=:5-5:5;5;=
=:=:, :2? E
1
sQ
~ 3 zsxxr, W .,W*1 :2:iii :3:
>= ta 4
lg..w<. ~1*? '1'9% i
.
:1:
9
.1Z
2
'3
4
2
s
/
ex.
2::mi.
4»,
Q.
;.
ea
$2.
3%
£2,
£2:21:
~
,r:?%§:::h .r=¬t>:.:: ?? .¬>¢*~: .¢<,.,,
:=:i7§i5:.
1,. .I 11; ?
:Z
43%.. .
,. <<,...:~z,.,;'~¢,§4:
We/Z.:
::+»g»:;::
gétaisx ,Burgess et al / SEXUAL HOMICIDE
develop positive empathy, and to control impulses Furthermore, there is nochallenge to their beliefs that they were entitled to act the way they do TheAlth he ther as children or adolescents feel estranged from people ougmen ithat does not mean that superficially they cannot relate to people, it doesindicate that in terms of socially effective learning, they have mayor deficitsThey are loners, they are self-preoccupied Either by daydreaming or fan-tasies, they become absorbed in their own thoughts.
! Feedback FilterGiven the detailed and repetitive thinking patterns of these murderers, it isnot surprising to learn that the murderer reacts to and evaluates his actionsrd others and toward himself. These reactions and evaluations influencetowa
his future actions. We term this reacting the feedback filter, because it bothfeeds back into the killer's patterned responses and filters his earlier actions
into a continued way of thinking.Through the feedback filter, the murderers earlier actions are justified,errors re sorted out, and corrections are made to preserve and protect the' t na fai y world and to avoid restrictions from the extemal environ-ln Cl A .ment. The mui .-rer experiences increased arousal states via fantasy varia-tions on the violent actions. Feelings of dominance, power, and control are' c eased The murderer develops increased knowledge of how to avoidll! I .unishment and detection. All this feeds back into the patterned responsesP
and enhances the details of the fantasy life. For example, one oi the murderersreported how he sat in prison ruminating on his fantasies regarding killingwomen and dismembering their bodies. As time went on, he became muchmore excited by his thoughts of disposing of the victims bodies and trickinglaw enforcement agents. In this peculiar evolution of events, he now experi-
d.enced himself as more involved in the social worl
Model Summaryd' d ' terms of the contri-Wh d lescent and adult criminals are stu ie inen a 0bution of past events to their criminality, emphasis previously has been onthe event itself rather than on the sub3ect's response and reaction to the event.. . edIn part, psychological models of motivation for sexual murder have focuson models of displacement of rage and frustration from primary caretakers inthe lives of sexual murderers. Although these symbolic artifacts may operate, . . . . . ba more direct understanding of the potential for violence and criminal e-havior resides in the fantasy life and basic cognitive operations of murderers.A context of justifying socially abhorrent acts provides support for themurderers aggressive, violent fantasies. This structure, limited to its sensoryts.arousalcapacities, maintains and perpetuates the destructive ac
5 5
JOURNAL OF INTERPERSONAL VIOLENCE / September I986
Our motivational model suggests that traumatic and early damaging
experiences to the murderers as children set into motion pattems of cogni-
tion. Although there may be initial attempts to work through the trouble-
some effects of the experience, attempts to do so become pattems for limiting
choices such as aggression being the only method for dealing with conict. In
addition, a structure of thinking that motivates and sustains deviant behavior
through developmental and interpersonal failure and through the alliance of
distorted perceptions and affect begins to emerge. Of particular importance is
the activation of aggression and its link with sexual expression. The laclt of
attachment to others gives a randomness to the sexual crimes; however,
scrutiny of the thinking patterns of the offenders indicates that there is
planning of the crimes whether they rely on chance encounters with any
victim or whether they are planned to snare victims.
IMPLICATIONS
There are clinical implications from this study. Understanding the rein-
forcing quality of actions, be they in fantasy, play, or acting out behaviors,
may lead to different notions regarding not only motivation but also behavior
change. Exploratory efforts by clinicians are needed for methods to alter the
structure of these fantasies. For example, the offender might be forced to
relate to the victim position in the fantasy as a way to stimulate compassion
for rather than violence to the victim.
This study raises concern about how to deal therapeutically with the
notion of fantasy in the criminal population. We note that some levels of
dwelling on fantasies has the capacity to escalate rather than diminish the
power of the fantasy. Unless one alters the structure of the fantasy that moves
toward the aggressive acts, the power can be increased.
This exploratory study suggests avenues for further research. Basic re-
search in biological and psychosocial factors is necessary to explore the
biochemical hormonal sensory levels associated with deviant fantasies of
both youth and adults. We know that pessimistic cognitions are associated
with lowered epinephrine levels in endogenous depression; what might
research suggest regarding violent sexualized fantasies?
Basic research on the sensory arousal levels of people during fantasy might
answer the question: Is there a basis of hormonal release addicting the person
to violent fantasy and violent acts? And does the structure of fantasy differ
between various groups of deviant offenders?
Research on a longitudinal basis of children's response to and recovery
from sexual and physical/psychological abuse and research on the social
context in which the child survives and recovers from abuse are important to
any understanding of motivational factors. In this context, a control group
56s
1
2%
§ 121%"<:
" ~%z=;~" 2211
§§¬?' xzzzcz
¢=f2:-21??
f3%» Aklf '
5,
E,
1'8?..'Burgess et al. / SEXUAL HOMICIDE
of abused males who do not commit criminal acts is essential to identify thefactors that help the victim recover and survive the abuse.Our work and the research of others Prentky et al., 1985! suggest that atypology of murderers is essential to investigate for a variety of reasons. First,any understanding of the typology may enhance law enforcement efforts bothat persuading certain offenders to turn themselves in and at more narrowlyfocusing investigative efforts. Second, a classification system will give profes-sionals working to curtail violent behavior a focus for intervention effortsthat address the need to monitor, evaluate, and change salient personalitycharacteristics. Measurements of these characteristics and methods of evalu-ating positive change are essential to prevent the tragic reality of releasedviolent criminals repeating their crimes. Third, a classification system wouldfacilitate dialogue between the various disciplines working with offendersand would encourage research into profiling of suspects from crime sceneevidence, a technique currently in progress at the National Center for theAnalysis of Violent Crime Douglas, in preparation!. Further, behavioralesearch efforts by law enforcement agencies are important to their develop-ment " additional skill in reading the seemingly inert characteristics of crimescene eviuence. Understanding the motivational and behavioral matrix of theoffender increases law enforcements utilization of the connection between
patterns of thinking and behavior.
REFERENCES
Axline, V. M. 969!. Play therapy. New York: Ballantine.Baekeland, F. 970!. Exercise deprivation: Sleep and psychological reactions. Archives
of General Psychiatry, 22, 365-369. Beck, A. T. 976!. Cognitive therapy and emotional disorders. New York: Simon 8:
Schuster.Beres, D. 961!. Perception, imagination and reality. International journal of Psycho-
analysis, 4I, 327-334. The sadistic murderer Medical Science and the Law, 10, 198-207.Brittain, R. P. l970!. . Brownmiller, S. 975!. Against our will: Men, women and rape. New York: Simon 8:
Schuster.Burgess, A. W., Hartman, C. R., McCausland, M. P., 8c Powers, P. 984!. Response' ' ' ed h h sex rings and pornography.patterns in children and adolescents exploit t roug' I of P chiatry, 141 !, 656-662.American journa sy R trauma syndrome. American journalBurgess, A. W., Sc Holmstrom, L. L. 974!. ape
P chiatry 131 981-986MD:of rv . » - Burgess, A. W., 8: Holmstrom, L. L. 979!. Rape: Crisis and recovery. Bowie,
Brady. ' ' S induced plasma betaR Pichas, D Dubois, M et al 982!. tress Cohen, M. ., ., ., .' ' ' ' edict postoperative morphine usage. Psychi-endorphin immunoreactivity may pr
atric Research, 6, 7-l2.
57
I
2
l
i
I
.s
,.
3
.v
x
ll
/XJOURNAL OF INTERPERSONAL VIOLENCE / September I986
Coid, _]., Allohio, B., 8: Rees, L. H. I983, September 3!. Raised plasma metenkephalin
in patients who habitually mutilate themselves. The Lancet, pp. 545-546.
Colt, I-I.W. D., Wardlaw, S. L., 8: Frantz, A. G. I981!. The effects of running on plasma
B-endorphin. Life Sciences, 28, I637-I640.
Constantino, _]. P., Kuller, L. H., Perper, _]. A., 8: Cypress, R. H. I977!. An epidemio-
logic study of homicides in Allegheny County, Pennsylvania. American journal of
Epidemiology, 106 !, 3:4-324.
Conte, _]. R. I984!. Progress in treating the sexual abuse in children. Social Work,
pp. 258-263.
Cormier, B. S., 8: Simons, S. P. I969!. The problem of the dangerous sexual offender.
Canadian Psychiatric Association journal, I4, 829-334.
Douglas, J. E. in press!. Profiling as an investigative technique.
Eth, S., 8: Pynoos, R. I984!. Developmental perspectives on psychic trauma in child-
hood. In C. R. Figlcy Ed. !, Trauma and its wake. New York: Brunner/Mazel.
Frazier, S. H. I974!. Murder-single and multiple. Aggression Research Publication
ARNMD, 52, 304-312.
Gardner, R. A. I971!. Therapeutic communication with children. New York: Science
House Books.
Hartman, C., 8: Burgess, A. W. in press!. Causes and consequences of sexual assault. In
V. Carlson 8: D. Cicchetti Eds.!, Child maltreatment.
Kahn, M. W. I971!. Murderers who plead insanity: A descriptive factor-analytic study
of personality, social and history variables. Genetic Psychology Monographs, 84
!, 275-360.
Lunde, D. I977!. Murder and madness. Los Angelesz University of California Press.
MacCulloch, M. C., Snowden, P. R., Wood, P.].W., 8: Mills, H. E. I983!. Sadistic
fantasy, sadistic behaviour and offending. British journal of Psychiatry, I43, 20-29.
MacDonald, J. M. I971!. Rape offenders and their victims. Springfield, IL: Charles C
Thomas.
Prentky, R., Cohen, M., 8: Seghorn, T. I985!. Developmentofa rational taxonomy for
the classification of sexual offenders: Rapists. Bulletin of the American Academy of
Psychiatry and the Law, 13, 39-70.
Pynoos, R. S., 8: Eth, S. I985!. Children traumatized by witnessing acts of personal
violence: Homicide, rape, or suicidal behavior. In S. Eth 8: R. S. Pynoos Eds!,
Post-traumatic stress disorders in children pp. I7-44!. Washington, DC: American
Psychiatric Press.
Reinhardt, _]. M. I957!. Sex perversions and sex crimes: A psychocultural examination
of the causes, nature and criminal manifestations of sex perversions. Police science
series. Springfield, IL: Charles C Thomas.
Ressler, R. K., Douglas, ]. E., C-roth, A. N., 8: Burgess, A. W. I980!. Offender profiling:
A multidisciplinary approach. FBI Law Enforcement Bulletin, 49 I0!, I6-20.
Ressler, R. I ., et al. I985!. Violent crimes. FBI Law Enforcement Bulletin, 54 8!, I-33.
Revitch, E. I965!. Sex murder and the potential sex murderer. Diseases of the Nervous
System, 26, 640!-648.
Revitch, E. I980!. Genocide and unprovoked attacks on women. Correctional and
-Social Psychiatry, 26, G-l l.
Rushforth, N. B., Ford, A. B., Hirsch, C. S., Rushforth, N. M., 8: Adelson, L. I977!.
Violent death in a metropolitan county: Changing pattems in homicide 958-74!.
New England foumal of Medicine, 274!, 53-58.
582; 1%
1
.5
,.
. aw.
gm.?
,7...
ghcxié / < ii
52:1;3 6%
,.. ..
z
7
»s
%
?'
%
2
.
1,,
1%:it.
I
2&2:
Z.
E92
1 ?<
W
ii
§¢.-3;: :
ixi./
W"E?Burgess et al. / SEXUAL HOMICIDE
Samenow, S. E. 980!. Inside the criminal mind. New York: Time Books.Satten, _]., Menninger, K., Rosen, 1., 8: Mayman, M. 960!. Murder without apparentmotive: A study in personality disorganiiation. American journal of Psychiatry,
II 7, 48-53. 1981! The criminal fantasy technique: A compar-Schlesinger, L. B., 8: Kutash, 1. L. .ison of sex offenders and substance abusers. joumal of Clinical Psychology, 3 7 !,
210-218. ' . S violence and crime. In 1. L. KutashSchlesinger, L. B., 8: Revitch, E. 981! tress, g' d . Handbook on stress and anxiety. San Francisco:8c L. B. Schlesinger E s!,
jossey-Bass.Seghorn, T. 1 ., Boucher, R. ]., 8: Prentky, R. A. in press!. Childhood sexual abuse in
the lives of sexually aggressive offenders.Singer, ]. L. 966!. Daydreaming. New York: Random House.Terr, L. 979!. Children oi Chowchilla: A study oi psychic trauma. Psychoanalytic
Study of the Child, 34, 547-623. T L. 981a!. Forbidden games: Post-traumatic child's play. journal of the Ameri-
C11,can Academy of Child Psychiatry, 20, 741-760.' h'ldren. American journal of Psychiatry, I38,Terr, L. l98lb!. Psychic trauma in c 1
14-19.Terr, L. 983!. Lile attitudes, dreams and psychic trauma in a group oi "normal"childre ,'0vrnal of the American Academy of Child Psychiatry, 22, 221-250.van der Kolk, t.., Greenberg, M., Boyd, H., 8: Krystal, ]. I984!. inescapable shock,neurotransmitters, and addition to trauma: Toward a psychobiology of p0st-trau-
matic stress. Biological Psychiatry, 20, 314-S25.West, D. ]., Roy, C., 8: Nicholas, F. L. 978!. Understanding sexual attacks. London:
Heinemann.Wollgang, M. E. 958!. Patterns of homicide. Philadelphia: University oi Pennsyl-
vania Press.Wood, ]., Burgess, A. W., 8: Mctlormack, A. in press!. Runaways and their drawings:
Response patterns associated with abuse.h 'minal personality. New York: ]. Aronson.Yochelson, S., 8c Samenow, S. 977!. T e cri