REFERENCE: Meloy JR, Rivers L, Siegel L, Gothard S,
Naimark D, Nicolini R. A replication study of obsessional follow-
ers and offenders with mental disorders. J Forensic Sci 2000;45
ABSTRACT: The purpose of this study was to compare certain
demographic, clinical, and criminal variables within subgroups of
obsessional followers, and compare them to a group of offenders
with mental disorders to attempt to replicate earlier findings. A
static group archival design utilized a non-random group of con-
venience and a randomly selected comparison group. Sixty-five
obsessional followers and 65 offenders with mental disorders
were evaluated by psychiatrists and psychologists for court or-
dered reasons during their criminal proceedings. Both groups were
evaluated during the same period, in the same court diagnostic
clinic, and generally for sentencing determinations. The obses-
sional followers were measured on demographic, diagnostic, pur-
suit, victim, threat, violence, emotional, motivational, and defense
variables. Inferential comparisons that used parametric and non-
parametric statistics were done within and between groups on se-
lect variables. The obsessional followers had significantly greater
estimated IQ than the offenders with mental disorders, but were
neither older nor better educated. There were no significant dif-
ferences in the high prevalence of both DSM-IV Axis I and II di-
agnoses. Obsessional followers who stalked prior sexual intimates
were significantly more likely to have a substance abuse or de-
pendence diagnosis. Obsessional followers who stalked strangers
or acquaintances were more likely to be delusional. The majority
of the obsessional followers, primarily motivated by anger, both
threatened and were violent toward person or property. The modal
obsessional follower is an average or above IQ, unemployed, un-
married male in his fourth decade of life, chronically pursuing a
prior sexually intimate female. He is diagnosed with substance
abuse or dependence and a personality disorder NOS, and has a
prior psychiatric, criminal, and substance abuse history. He is an-
gry, likely to threaten her, and assault her person or property with-
out causing serious injury.
KEYWORDS: forensic science, forensic psychiatry, stalking, ob-
session, violence risk, mental disorder
Stalking laws now exist throughout the United States, Canada,
and Great Britain. A recent study jointly funded by the National In-
stitute of Justice and Centers for Disease Control and Prevention
found that the lifetime risk for victimization is 8% for women and
2% for men (1), eclipsing earlier, more speculative estimates (2).
One percent of women and 0.4% of men report being stalked dur-
ing the past year, causing a variety of psychiatric symptoms and
major social disruption in their lives. This criminal behavior ap-
pears to be a significant public health and social problem that af-
fects both adults (1) and adolescents (3).
Stalking is often legally defined as, “the willful, malicious, and
repeated following and harassing of another that threatens his or
her safety,” (4, p. 258) without any inference concerning the psy-
chiatric status of the offender. Meloy and Gothard (4) coined the
clinical and behavioral term “obsessional following” for three rea-
sons: to further clinical research through the use of a clearly de-
fined label that was untouched by sensationalism; to identify the
most common prohibited act of stalkers, that is, following (1); and
to highlight what appears to be an important cognitive and motiva-
tional component of stalking, that is, obsessions (5). An “obses-
sional follower” is a person who engages in an abnormal or long
term pattern of threat or harassment directed toward a specific in-
dividual. Threat or harassment is more than one overt act of un-
wanted pursuit that is perceived by the victim as harassing (4,6).
Meloy (7) reviewed the existing research on obsessional follow-
ers who had law enforcement contact, utilizing the above defini-
tion, and identified 10 studies which included 180 individuals.
Most of the subjects were gathered from three large urban areas in
the United States (4,6,8), with a few subjects from Australia (9) and
Great Britain (10). Preliminary findings indicated that obsessional
followers were men in their fourth decade of life who pursued prior
female sexual intimates or acquaintances. They were suggestively
more intelligent and better educated than other mentally disordered
criminals, and had prior psychiatric, criminal, and drug abuse his-
tories. The majority of the subjects had both Axis I and Axis II dis-
orders, and had chronic histories of failed heterosexual relation-
The obsessional followers, moreover, had substantially lower
rates of antisocial personality disorder (ASPD) than other incarcer-
ated males, which Meloy attributed to the intense, preoccupied na-
ture of their attachment—a finding which would be inconsistent
with ASPD (7). A psychodynamic hypothesis was also formulated
which asserted that such men develop “narcissistic linking fan-
J. Reid Meloy,1Ph.D.; Lynette Rivers,2Ph.D.; Liza Siegel,3Ph.D.; Shayna Gothard,4Ph.D.;
David Naimark,1M.D.; and J. Reese Nicolini,5Ph.D.
A Replication Study of Obsessional Followers and
Offenders with Mental Disorders
1Associate and assistant clinical professors of psychiatry, respectively, Uni-
versity of California, San Diego, CA.
2Centinela State Prison, El Centro, San Diego, CA.
3California School of Professional Psychology, 27 Cornerstone Court, San
4Forensic Evaluation Unit, Superior Court, San Diego, CA.
5Psychiatric Health Systems, San Diego CA.
Received 9 March 1999; and in revised form 6 April 1999; accepted 12 April
Copyright © 2000 by ASTM International
JOURNAL OF FORENSIC SCIENCES
istics of the obsessional followers was gathered from the forensic
evaluations written by the clinicians and submitted to the superior
court. The clinicians gathered their data from a clinical interview,
review of criminal and medical records, and telephone contact with
various collaterals. A data collection protocol was developed by
the researchers based upon extant clinical and empirical research
concerning stalkers (1–12). Comparisons within and between the
obsessional followers and offenders with mental disorders were
made on a priori selected demographic, criminal, and clinical vari-
ables; particular attention was paid to the variables which yielded
significant differences in our previous study (4). Psychodiagnosis
had been determined by each evaluating clinician following a re-
view of all data. No structured diagnostic interviews were used,
and all diagnoses were determined by the evaluating psychiatrist or
psychologist. These shortcomings represent limitations of the
study. IQ had been determined through clinical measurement using
the Shipley Institute of Living Scale (N ? 67) or clinical estimate
(N ? 28). Each subject for whom IQ data were available from both
groups (N ? 95) was then assigned to one of four categories: supe-
rior (WAIS-R equivalent ?120), above average (110–119), aver-
age (90–109), or below average IQ (? 89). Continuous variables
were compared using two-tailed independent samples t-tests. Cat-
egorical variables were compared using chi-square analysis and
followup z-tests for proportion when indicated. Significance was
set at p ? .05.
Table 1 presents the demographic and clinical characteristics of
the sample of obsessional followers.
Although there were missing data for the IQ measures, only one
obsessional follower had an estimated below average IQ. Sixty-
eight percent of these subjects (N ? 44) tested or were estimated to
have an average IQ or better.
Most of the subjects (72%) were single, divorced, or widowed at
the time of the evaluation. Only 23% reported a current sexually in-
timate partner, but there were substantial missing data for this vari-
able (N ? 29). The subjects had prior psychiatric histories (N ? 40,
62%), were currently unemployed (N ? 40, 62%), and had incon-
sistent employment histories (N ? 29, 45%). Although 72% (N ?
47) had prior criminal histories, most of the subjects, when data
were available (N ? 43), did not have a childhood history of anti-
social behavior (N ? 26).
Eighty-six percent (N ? 56) of the subjects had an Axis I (DSM-
IV) diagnosis at the time of evaluation. The most common diagno-
sis was substance abuse or dependence (N ? 31, 48%), usually al-
cohol, amphetamine, or both. Other Axis I disorders included a
mood disorder (N ? 15, 23%), usually dysthymia or bipolar, or a
schizophrenia (N ? 7, 11%). Delusional disorder was uncommon,
occurring in only four subjects (6%).
Fourteen subjects were determined to be psychotic at the time of
their stalking (22%), and one half of these individuals visited the
home of the victim. Nine subjects (14%) showed evidence of ero-
tomanic delusions, 11 subjects (17%) evidenced persecutory delu-
sions, and three subjects (5%) evidenced grandiose delusions at the
time of their stalking.
Axis II disorders were also present in the majority of subjects (N
? 40, 62%). Nine percent (N ? 6) met criteria for Antisocial Per-
sonality Disorder. Two subjects met criteria for Borderline Person-
ality Disorder, and one subject each met criteria for Narcissistic,
Schizoid, Paranoid, Histrionic, and Dependent Personality Disor-
der. The most likely Axis II diagnosis was Personality Disorder
tasies” to their objects which are challenged by real world rejection
and lead to shame and humiliation. These emotions are defended
against with rage, which fuels the pursuit to control and devalue the
object, and thus restore the narcissistic fantasy (7).
Among the larger studies were two in which clinical-forensic
evaluations of the subjects were conducted (4,8). Since the publi-
cation of this review (7), three other studies deserve comment.
Menzies et al. (11) completed the first predictive study of danger-
ousness among a small sample of erotomanics (N ? 29) and found
that two variables—multiple victims and unrelated antisocial be-
havior—correctly classified 88% of the sample as dangerous or
not. Kienlen et al. (12) conducted the first comparative study of a
small sample (N ? 25) of psychotic and non-psychotic stalkers,
and identified major childhood attachment disruptions and recent
significant losses in a majority of their sample. Schwartz-Watts et
al. (13) confirmed many of the findings from previous studies, in-
cluding better education and significant mental illness and sub-
stance abuse among stalkers. They also found high rates (50%) of
organicity in their stalking sample (N ? 18) based upon MRI scans
or head trauma with documented personality change.
We decided to conduct this comparative clinical study to attempt
to replicate the findings of our first effort (4). We also hoped to ex-
tend and refine the suggestive findings on threats and violence,
both quantitatively and qualitatively, among obsessional followers
and harassers in general (7,8). We tested the null hypothesis that a
forensic cohort of obsessional followers would not significantly
differ on certain demographic and clinical variables from a ran-
domly selected group of offenders with mental disorders. The lat-
ter group was selected for comparison because of the finding in the
stalking research that most subjects have prior criminal and psy-
chiatric histories (4–8).
The study was a static group archival design comprising a non-
random group of convenience and a randomly selected comparison
group. Both the obsessional followers (N ? 65) and the offenders
with mental disorders (N ? 65) were selected from the case files of
approximately 2300 adults whom the Superior Court of San Diego
County referred between January, 1994, and June, 1996, for a clin-
ical evaluation by the Forensic Evaluation Unit, a publicly funded
court psychodiagnostic clinic composed of two board-certified
psychiatrists and three licensed clinical psychologists with exten-
sive experience in forensic evaluations. The forensic evaluators
were blind to the methods and hypotheses of this study, and had
been randomly assigned subjects in both groups by the supervising
psychiatrist in the normal course of daily evaluations before the on-
set of this study. The study was approved by the Human Subjects
Research Committee of San Diego County Mental Health Services.
Patient consent was unnecessary due to the archival nature of the
All case files were screened for subjects who would meet the
previously defined criteria for obsessional followers. Sixty-five
cases were identified. The offender group was filled by selecting at
random from the case files of the clinic subjects who were evalu-
ated by the same clinicians during the same period of time, but did
not meet the definition of obsessional follower. A majority of the
subjects (N?108) were referred for presentencing evaluation. Vir-
tually all stalking and harassment cases in San Diego County are
referred for an evaluation to this court clinic.
Two data analyses were completed. Descriptive information
concerning various demographic, clinical, and criminal character-
MELOY ET AL. • A REPLICATION STUDY 149
NOS (N ? 22, 34%), and the most common descriptors were “nar-
cissistic” (N?19, 30%), “paranoid” (N?14, 24%), “antisocial” (N
?8, 12%), “dependent” (N?7, 11%), “compulsive” (N?7, 11%),
and “histrionic” (N ? 5, 8%). The modal Personality Disorder NOS
descriptor was “narcissistic, paranoid, and compulsive.” One sub-
ject was diagnosed with borderline intellectual functioning.
The majority of the victims of the obsessional followers were
prior sexual intimates (N ? 37, 57%) and female (N ? 53, 82%).
Eleven victims (17%) were prior acquaintances, and nine victims
(14%) were strangers. The average victim age was 33 (SD ? 15,
Range 16–64). All the subjects, except for one, pursued opposite
The obsessional followers had multiple and varied means of con-
tact with their victims. The most common method of pursuit was
telephone calling (N ? 46, 71%), visiting the home (N ? 43, 66%),
behavioral following (N ? 31, 48%), visiting the workplace (N ?
15, 23%), and letter writing and slandering (N ? 9, 14%). Three
percent engaged in sexual assaulting, exposing self, and sending
gifts (N?2). One subject used the internet and one subject tortured
his victim. Two subjects (3%) sought restraining orders against
their victims. Average duration of contact prior to arrest was 11.46
months (SD ? 20, Range 0–120). Average number of incidents
was 6.24 (SD ? 1.89, Range 2–7). Twelve of the subjects (18%)
reported 17 prior stalking victims (M ? 1.42, Range 1–5). Half of
these subjects reported prior stalking of strangers, but only two had
ever been criminally charged with stalking.
All of the subjects were criminally charged at the time of the pre-
sent evaluations. The most common criminal charge was stalking
(N ? 17, 27%), followed by violation of a restraining order (N ?
6, 9%), terrorist threat (N ? 4, 6%), and stalking after a restraining
order was issued, assault with great bodily injury, or kidnap (N ?
The majority of the evaluations were done post-conviction (N ?
46, 71%) but prior to sentencing. Eighteen (28%) subjects were
evaluated for competency to stand trial, and one subject was eval-
uated for insanity at the time of the crime.
The vast majority of the obsessional followers threatened their
victims (N ? 49, 75%). Most of the subjects directed their threats
toward the victim (N ? 42), but seven subjects also included prop-
erty threats. No one only threatened property. Twenty subjects
threatened third parties (31%).
Thirty-four of the subjects (52%) were physically violent: 15
toward the victim, four toward the victim’s property only, and fif-
teen toward the victim and her property. Only two subjects (3%)
assaulted a third party. A weapon was used by 10 subjects (15%)
who were violent, and usually consisted of a car, knife, or
firearm. In no cases where these weapons were used was the vic-
tim hit by the car, cut by the knife, or shot by the gun. If violence
was committed toward property only, the likely target was the
victim’s automobile. Serious physical injury to the victim re-
sulted in only two cases when there was physical violence, and
there were no homicides in our study. Victims were typically
grabbed, choked, hair pulled, thrown, shaken, hit, slapped,
kicked, or punched.
When threats and subsequent violence toward person and prop-
erty were considered (N ? 34, 52%), the false positive rate was
60% and the false negative rate was 11%. When threats and subse-
quent violence only toward persons were considered the false pos-
itive rate increased to 72% and the false negative rate also in-
creased to 15%. In both cases where a third party was assaulted (N
? 2) there had been an articulated threat, but 90% of the threats to-
ward third parties were false positives.
TABLE 1—Demographic and clinical characteristics of obsessional
followers (n ? 65) and offenders with mental disorders (n ? 65).
Not Otherwise Specified
Not Otherwise Specified
other than ASPD
Substance Abuse or
Various drugs (most
alcohol, or both)
4 6.25 7.7
6 9.27 10.8
3147.7 40 61.5
* t(128) ? 0.43, p ? .667.
† t(109) ? 0.81, p ? .419.
‡ X2? 10.24, df ? 3, p ?.025.
§ X2? 0.29, df ? 2, p ?.50.
? X2? 1.14, df ? 1, p ?.20.
JOURNAL OF FORENSIC SCIENCES
We also gathered data on defenses, emotions, and other psycho-
logical factors that influence and motivate obsessional followers,
as illustrated in Table 2. Our list of emotions and other psycholog-
ical factors was constructed from previous research (4,12).
We then divided the obsessional followers into two subgroups:
prior sexual intimates (N ? 37) and strangers/acquaintances (N ?
20) to further test for differences on four a priori select variables.
The prior sexual intimates were significantly more likely to have a
substance abuse or dependence diagnosis (X2 ? 4.66, df ? 1, p ?
.03), and significantly less likely to be diagnosed with schizophre-
nia (X2 ? 4.85, df ? 1, p ? .027) or have any psychotic diagnosis,
excluding drug-induced psychosis (X2 ? 5.36, df ? 1, p ? .02)
when compared to the stranger/acquaintances. The prior sexual in-
timates were significantly more violent than the stranger/acquain-
tances (X2? 36.24, df ? 1, p ? .001).
Demographic and psychodiagnostic data for the comparison
group of mentally disordered offenders are also listed in Table 1.
We conducted only five a priori inferential statistical tests between
groups to reduce Type I error. There was no significant difference
between age and education of the obsessional followers and the of-
fenders with mental disorders. The obsessional followers had a sig-
nificantly higher IQ distribution than the comparison group (X2 ?
10.24, df ? 3, p ? .025). Followup z-tests for proportion indicated
the obsessional followers were significantly less likely to have an
estimated below average IQ than the offenders with mental disor-
ders (z ? ?2.13, p ? .0332). There was no significant difference
between the frequency of antisocial personality disorder (ASPD),
other personality disorder, or substance abuse/dependence diagno-
sis between the two groups.
TABLE 2—Defenses, emotions, and other psychological factors
motivating obsessional followers (N ? 65)*.
* Percentages exceed 100% due to more than one category being scored
in many subjects.
Projection of blame
(Attribution to and control of victim)
Obsession (preoccupation involving
Devaluation of victim
Idealization of victim
Emotions and other psychological factors
Anger and hostility
Need for power/control
Perceived mistreatment by public
Distress over custody fight
Seeking benefits from public figure
The demographic characteristics of obsessional followers appear
to be surprisingly consistent across studies. They are generally
males of various racial backgrounds in their fourth decade of life.
In fact, this is the fourth study that has found the average age to be
34–35 (4,6,13). We did not, however, replicate our earlier finding
that obsessional followers are significantly older than other offend-
ers with mental disorders. The subject sample does appear, how-
ever, to be at least a decade older than other arrested offenders in
We also did not replicate our earlier finding that obsessional fol-
lowers are better educated than other mentally disordered offenders
(4); but we did find, once again, greater estimated IQ. We consider
this replication tentative, however, since there were missing data on
subjects, and only 50% were tested with the Shipley. The work of
Harmon and her colleagues (8) does lend credence to the higher IQ
hypothesis for obsessional followers and harassers, and the recent
noted high prevalence of stalking on college campuses (3,15),
where better than average intelligence is expected, also support our
findings. Clinical cases also continue to document the resourceful-
ness and manipulativeness of stalkers (16), a likely correlate of in-
Our findings of prior psychiatric, criminal, and drug abuse histo-
ries in our sample of obsessional followers were replicated and pre-
dicted from the extant research (7). These individuals are not
paragons of the community who have suddenly and unexpectedly
engaged in this criminal behavior. The more precise characteriza-
tion would be a multi-problemed individual who engages in a num-
ber of antisocial activities, obsessional following being one of
them. The nature of the referral process to this courthouse clinic,
moreover, indicates that our sample is likely representative of the
population of obsessional followers in San Diego County, the sixth
largest metropolitan area in the United States.
Our research also continues to support the hypothesis that indi-
viduals who obsessively follow or harass have a history of failed
heterosexual relationships (7), and extends our belief that the aver-
age subject has had at least a decade of chronic heterosexual failure
before he begins this aberrant behavior. The crime of stalking may
be a “courtship” disorder, similar to the paraphilic disorders identi-
fied by Freund (17), although we do not think that stalking is usu-
ally an explicitly sexualized behavior. It may be that one who stalks
has decided that romantic pursuit, reciprocated and invited by the
object, does not work, and unwanted, aggressive pursuit is neces-
sary to form an attachment.
It is also clear that most victims were prior sexual intimates, a
finding echoed by recent studies (1,16) but contrary to Meloy’s ear-
lier assertions (7). This study also lends empirical support to the be-
lief that stalking and domestic violence are closely linked (24) and
there is a likely overlap in perpetrator psychopathology (27).
The plethora of Axis I disorders in this sample is also consistent
with extant research and our previous work (4,7). Once again, drug
abuse and dependence is most common, followed by a mood disor-
der or schizophrenia. There are likely differences in Axis I diag-
noses, however, when the subject pursues a prior sexual intimate or
a stranger/acquaintance. As we theorized in our first study (4), and
others have found (12), delusional beliefs are significantly more
likely in stranger/acquaintance cases, and will result in a major
mental disorder diagnosis at the time of evaluation. On the other
hand, subjects who pursue prior sexual intimates are significantly
more likely to be chemically fueled (alcohol or stimulants) during
their crime. Delusional disorder, erotomanic subtype, is also rare,
MELOY ET AL. • A REPLICATION STUDY 151
contrary to earlier research and speculation (7). Our frequency of
the diagnosis at the time of evaluation (N ? 2, 3%) and the pres-
ence of erotomanic symptoms during the stalking (N ? 9, 14%)
suggest that: a) erotomanic symptoms, if present, are likely to be
secondary to another psychiatric diagnosis; b) most individuals
who obsessively follow do not show any erotomanic beliefs; and c)
erotomania is still much more frequent in samples of obsessional
followers than among normals or other clinical samples (18).
We also replicated our finding that most obsessional followers
have a diagnosable personality disorder, typically Cluster B (4,7).
Although we did not find significantly less ASPD among the ob-
sessional followers when compared to the other mentally ill of-
fenders as we did before, the frequency of ASPD in this study, 9%,
was virtually identical to our earlier effort (4) and the work of oth-
ers (12). Harmon et al. (8) found higher rates in their New York
sample. A 9% rate of ASPD among obsessional followers is
markedly less than the 60–75% rate expected in incarcerated pop-
ulations in general (19).
We continue to theorize that the low rate of ASPD among ob-
sessional followers is expected given their attachment pathology:
the intense and pre-occupied seeking of a bond to a rejecting ob-
ject. Antisocial individuals, particularly psychopaths, are chroni-
cally emotionally detached, and would not be predicted to exhibit
such behavior. They are much more likely to use others and then
dispose of them (20,21).
The “pre-occupied” attachment pattern (22) among some ob-
sessional followers has been explored by Kienlen (12,16). She
also found that the majority of subjects in her sample had experi-
enced the loss of a primary object in childhood and a major loss
in adulthood, usually a job or divorce, in the half year preceding
the onset of stalking. This pattern of abnormal attachment may
also explain the recent disparity in findings concerning the effec-
tiveness of protection orders in domestic violence situations; a
majority of studies find them effective (23,24), while the largest
survey study, to date, of victims of stalking per se found them vi-
olated in the majority of the cases in which they were utilized (1).
Protection orders are likely to be more effective in the absence of
obsession and preoccupation with the victim. This is a psycho-
logical factor that should be considered, alongside aggressive en-
forcement of the protection order by police and a careful assess-
ment of the subject’s history of violence toward the protectee,
when an order is sought.
Although most cases of obsessional following appear to in-
volve both sexes, and in most cases the victim is a female, the be-
havior does not appear to be paraphilic or explicitly sexual. We
found very low rates of paraphilia (5%), sexual preoccupation
(14%), and sexualized pursuit (sexual assault or exposure, 6%)
among our subjects, similar to other research (7). Although a
component of obsessional following could be a disorder of sexual
desire, as Kaplan (25) theorized in erotomania, we think that it is
primarily fueled by anger, hostility, and abandonment rage—the
defensive byproducts of the shame of rejection. Our findings and
others (12) support this hypothesis as noted in Table 2. Meloy has
further theorized (4,5,7,21,26) that the narcissistic psychopathol-
ogy of the obsessional follower, supported by the Axis II findings
in this study, prevent him from grieving a rejection as a loss, and
instead fuel a desire to restore an idealized “linking” fantasy with
the object. This psychodynamic may paradoxically necessitate
the destruction of the object in real life (usually assault and bat-
tery) to restore this narcissistic fantasy. O.J. Simpson, found
civilly responsible for the death of his ex-wife, captured this psy-
chodynamic in a reported dream: she “comes to me from time to
time in my dreams and it’s always a positive dream. Occasionally
I dream that I single-handedly solve the case” (Newsweek, June
23, 1997, p. 43).
The pursuit patterns in this study were also similar to other re-
search (7), with the exception that letter writing is much less fre-
quent (14%). The most common pursuits continue to be telephon-
ing and seeking physical proximity for the majority of subjects. We
did not validate the finding of Kienlen (12) that psychotic subjects
were more likely to visit the home of the victim than non-psychotic
subjects. Cyberstalking is still unusual, but may increase in fre-
quency in future research. The most common form of cyberstalk-
ing appears to be e-mail harassment (16).
Our study found very high rates of both threats and violence to-
ward victims. Threats occurred at a frequency of 75%, substantially
higher than the 50% average reported in other studies (7). Most
threats, however, were not acted upon, a finding concordant with
previous research which indicates that three-fourths of threats are
false positives (7). Most threats among stalkers, moreover, are di-
rected at the object of desire, and are not directed at property or
Violent behavior, like in other studies, occurs frequently but is
generally not physically injurious to the victim. The 46% rate of vi-
olence toward persons in this study, however, was higher than the
range in most other studies (7,3–36%). When violence occurred, it
was done with a weapon in less than one out of three cases, and
when a weapon was used, most curiously, the predictable injury,
based upon the nature of the weapon, never occurred. We think this
empirical finding supports the assumption that weapons are used
by obsessional followers to intimidate, frighten, or control the vic-
tim, and not to injure her. There were also no homicides in this
study, consistent with extant research which indicates that homi-
cide occurs less than 2% of the time (7).
One of the most troubling findings of our study is the false neg-
ative rate for personal violence of 15%—this is the frequency with
which violence occurs in the absence of a previous threat. One in-
terpretation of this finding is that the violence was impulsive, af-
fective, and unplanned. Another interpretation is that the violence
was planned, purposeful, and carried out in a clandestine manner:
the subject as “hunter” rather than “howler” (28). Although hunt-
ing, or what has also been called “predatory violence” (20), cer-
tainly occurs among some stalkers, particularly those targeting
public officials or public figures (29), it is apparent to us by the na-
ture of the attacks that most violence, particularly toward prior sex-
ual intimates, is “affective violence:” emotional, unplanned, a re-
sponse to a perceived threat (rejection), and the result of high levels
of autonomic arousal (30).
Summary and Limitations
Obsessional followers are very similar to other offenders with
mental disorders. The one finding that appears to discriminate the
two groups is IQ. Our study, however, is limited by its archival na-
ture, the absence of some data, the absence of structured diagnos-
tic interviews, and the relatively small sample sizes. Generalizabil-
ity to other samples of obsessional followers seen in large urban
forensic settings appears warranted.
Obsessional following, and its related legal corollary, stalking,
are public health and social problems that are scientifically com-
ing of age. The previous decade has produced a growing body of
research that has recorded: a) the widespread nature of these
problems; b) the substantial impact on victims; c) similar demo-
graphic and clinical characteristics of the offenders; d) similar
152 Download full-text
JOURNAL OF FORENSIC SCIENCES
pursuit patterns; and e) substantially high rates of both threats and
personal violence, usually without serious physical injury, to the
victims. Further research needs to explore the prediction of vio-
lence among obsessional followers, risk management and treat-
ment outcomes, and variables among prior sexual intimates, par-
ticularly domestically violent men, that may predict such
behavior in the first place.
We would like to thank Ansar Haroun, M.D., and the entire staff
of the Forensic Evaluation Unit, San Diego County Courthouse, for
their support in the completion of this study. Funding for this study
was provided by Forensis, Inc., with a grant from the Susan Stein
1. Tjaden P, Thoennes N. Stalking in America: findings from the National
Violence Against Women Survey. Denver, CO: Center for Policy Re-
2. Senate Judiciary Committee Hearing, S.B. 2922, Sept. 29, 1992.
3. Fremouw W, Westrup D, Pennypacker J. Stalking on campus: the preva-
lence and strategies for coping with stalking. J Forensic Sci 1997;42:
4. Meloy JR, Gothard S. Demographic and clinical comparison of obses-
sional followers and offenders with mental disorders. Am J Psychiatry
5. Meloy JR. The psychology of stalking. In: Meloy JR, editor. The psy-
chology of stalking: clinical and forensic perspectives. San Diego: Aca-
demic Press, 1998;1–23.
6. Zona M, Sharma K, Lane J. A comparative study of erotomanic and ob-
sessional subjects in a forensic sample. J Forensic Sci 1993;38:894–903.
7. Meloy JR. Stalking (obsessional following): a review of some prelimi-
nary studies. Aggress Viol Behavior 1996;1:147–62.
8. Harmon R, Rosner R, Owens H. Obsessional harassment and erotoma-
nia in a criminal court population. J Forensic Sci 1995;40:188–96.
9. Pathé M, Mullen P. The impact of stalkers on their victims. Br J Psychi-
10. Taylor P, Mahendra B, Gunn J. Erotomania in males. Br J Psychiatry
11. Menzies R, Fedoroff JP, Green C, Isaacson K. Prediction of dangerous
behavior in male erotomania. Br J Psychiatry 1995;166:529–36.
12. Kienlen K, Birmingham D, Solberg K, O’Regan J, Meloy JR. A com-
parative study of psychotic and nonpsychotic stalking. J Am Acad Psy-
chiatry Law 1997;25:317–34.
13. Schwartz-Watts D, Morgan D, Barnes C. Stalkers: the South Carolina
experience. J Am Acad Psychiatry Law 1997;25:541–50.
14. Federal Bureau of Investigation: Uniform Crime Reports 1997. Wash-
ington DC, Department of Justice.
15. Romans J, Hays J, White TK. Stalking and related behaviors experienced
by counseling center staff members from current or former clients. Prof
Psychol: Res Pract 1996;27:595–9.
16. Meloy JR, editor: The psychology of stalking: Clinical and forensic per-
spectives. San Diego, CA, Academic Press, 1998.
17. Freund K, Scher H, Hucker S: The courtship disorders. Arch Sex Be-
18. Sandberg DA, McNiel DE, Binder RL: Characteristics of psychiatric in-
patients who stalk, threaten, or harass hospital staff after discharge. Am
J Psychiatry 1998;155:1102–5.
19. Hare RD: Manual for the psychopathy checklist-revised. Multihealth
Systems, Toronto, 1991.
20. Meloy JR: The psychopathic mind: Origins, dynamics and treatment. Ja-
son Aronson, Northvale, NJ, 1988.
21. Meloy JR: Violent attachments. Jason Aronson, Northvale, NJ, 1992.
22. Bartholomew K: Avoidance of intimacy: an attachment perspective. J
Soc Pers Relat 1990;7:147–78.
23. Meloy JR, Cowett PY, Parker S, Hofland B, Friedland A. Domestic pro-
tection orders and the prediction of criminality and violence toward pro-
tectees. Psychotherapy 1997;34:447–58.
24. Violence Against Women Grants Office: Domestic Violence and Stalk-
ing: The Second Annual Report to Congress. Washington, DC, U.S.
Dept. of Justice, 1997.
25. Kaplan HS. Erotic obsession: relationship to hypoactive sexual desire
disorder and paraphilia. Am J Psychiatry Festchrift Supple. 1996;153:
26. Meloy JR. Unrequited love and the wish to kill: diagnosis and treatment
of borderline erotomania. Bull Menn Clinic 1989;53:477–92.
27. Dutton D. The domestic assault of women. Vancouver, Univ. of British
Columbia Press, 1995.
28. Calhoun F. Hunters and howlers: threats and violence against federal ju-
dicial officials, 1789–1993. Arlington, VA, U.S. Marshals Service,
29. Fein R, Vossekuil B. Preventing attacks on public figures and public of-
ficials: a Secret Service perspective. In: Meloy JR, editor. The Psychol-
ogy of stalking: clinical and forensic perspectives. San Diego, Academic
30. Eichelman B. Aggressive behavior: from laboratory to clinic. Arch Gen
Additional information and reprint requests:
J. Reid Meloy, Ph.D
Clinical and Forensic Psychology
964 Fifth Avenue, Suite 409
San Diego, California 92101