Sunday, April 26, 2009

Craigslist Killer? What do we know about Physician Serial Murderers?



The so-called Craigslist Killer - hasn't even been convicted, at least not in a criminal court

The Craigslist Killer has been the recent subject of media attention. Since I seem to recall my law professors teaching me that an individual is presumed innocent until proven guilty at trial, I shall refrain from offering worthless speculation and adding to the hype. Instead, I'll try to do what I always do - provide you with some forensic knowledge about the issue in general, like I try to do with juries, and you can take this knowledge and form your own conclusions.

So - what do we really know about physicians who kill?

A physician swears by the hippocratic oath, and also swears to "first, do no harm," while putting his patients' best interests above all else. What could be more unnerving than a person who has taken this oath, looks the part, talks the talk of the helping profession, and appears to walk the walk as well? The truth is that killer doctors are nothing new, yet perhaps the aura of the profession has buffered these individuals from being examined too closely by the research - until relatively recently.

The study of medical serial killers may have gone overlooked due to an unwillingness to perceive sworn “healers” as potential murderers. However, research has revealed that medical killers may actually be the most prolific of all serial killers. Doctors who serially murder their patients are considered to belong to a larger group of “career-assisted killers.” The term “clinicide” has been used to describe “the unnatural death of multiple patients in the course of treatment by a doctor.”[i] Such murders may be difficult to detect, since they often occur in settings where death is expected to happen. Doctors accused of clinicide will be likely to put forth the defense that they were relieving suffering or providing euthanasia. Clinicidal doctors may have extreme narcissistic personalities, and may obtain pleasure by “determining” when a person will die.

One of the most deadly doctor serial killers may also hold the dubious distinction of being one of the most prolific serial murderers to date. Dr. Harold Shipman, a UK physician, was convicted of killing 15 patients with lethal injections of narcotics. In a post-trial investigation, it was concluded that Shipman was responsible for 218 known victims.[ii] Other estimates have suggested the number is closer to 450.[iii] Most of Shipman’s victims were not terminally ill, nor did they have an immediate life threatening illness. Shipman refused to speak to anyone, and no complete psychological assessment was ever performed on him.29 He committed suicide in prison in 2004.

Other healthcare professionals have been implicated in serial murder. In a study of 90 healthcare killers, 86% were nurses and 12% were doctors.[iv] Injection was the most common method used, followed by suffocation, poisoning and tampering with equipment Fifty-four of the 90 cases were ultimately convicted. A total of 2, 113 deaths could be attributed to these 54 convicted healthcare killers.

The motives? Most have speculated that it has much to do with the feelings engendered in the killer by the power of life over death - perhaps the very thing that drew them to the profession in the first place. Does the Craigslist killer appear to fit this profile? Unknown - because a legitimate forensic scientist will not draw such conclusions based only on what the media presents us with. We will just have to wait for the case and evidence to unfold.


[i] Kaplan R. The clinicide phenomenon: an exploration of medical murder. Australasian Psychiatry 2007 15(4): 299-304.
[ii] Esmail A. Physician as Serial Killer – The Shipman Case. N Eng Jour Med 2005 352(18): 1843-1844.
[iii] The Shipman Inquiry. First Report, Volume One Death Disguised. COI Communications, Manchester, 2002.
[iv] Yorker B., Kizer K., Lampe P., Forrest A., Lannan J., Russell D. Serial murder by healthcare professionals. J Forensic Sci 2006 51(6): 1362-71.

Friday, April 24, 2009

Analyzing Threats: Forensic Psycholinguistics


Above: Threatening letter to a friend intercepted by police
Middle: Rationale of a "Rejected" type stalker
Bottom: Same stalker - ultimately killed his victim


A “threat” may be defined as a declaration of intent to harm. It may be the basis for criminal or civil liability. Threats are common, and most are not carried out. There is only a weak association between threats and violence, but there is an association. Consider the following statistics[1]:

§ 75% of threateners are not violent
§ When a threat is made, there is a 52 – 83% false positive prediction of violence
§ Attacks on public figures are rarely preceded by threats


In contrast to a clinical risk assessment done by a treating mental health clinician, a threat assessment is typically done by an expert with training and experience in the field of threat assessment. Competence in threat assessment comes from: 1) specialized training, 2) familiarity with current literature & research, and 3) experience in the field. The average mental health clinician would not reasonably be expected to perform a formal threat assessment of the type described in this course. The following table lists some differences between threat assessment and clinical risk assessment:


Differences Between Threat & Risk Assessment

Threat Assessment
§ Case-specific
§ Target specific
§ Usually not clinical
§ Goal: protect target, apprehend perpetrator
§ Procedure: threat mgt. plan


Risk Assessment
§ Population-specific
§ Actuarial or Structured Clinical approach
§ Clinical scenario
§ Goal: “predict” likelihood, reduce risk
§ Procedure: risk reduction/treatment plan



There are three possibilities in relation to a threat. The individual:

1. Made a threat, but does not pose a threat
2. Made a threat, and does pose a threat
3. Made no threat, and does pose a threat


Although there are many different types of threats, clinicians may be likely to encounter “instrumental” and “expressive” threats[2]:

Instrumental – made to control or influence the target’s behavior. They can be recognized by their conditional nature: “If you ____, then I’ll ____!”

Expressive – made to control or influence the target’s emotions. They can be recognized by their affective nature (ie., “blowing off steam”): “I could kill you!”

Expressive threats may be easier to spot and abort with management interventions. For example, some therapists handle them by allowing the patient to “vent” in a therapeutic manner, and then invite the patient to come up with nonviolent solutions to the problem. Instrumental threats may be more likely to be used by manipulative and/or antisocial individuals who may be less likely to respond to clinical interventions alone.

In a recent large DOJ study[3], 56.8% of stalkers made no threats, whereas 43.2% did make threats. The most common threats were:

1. Hit/slap/harm – 13.6%
2. Kill victim – 12.1%
3. Harm or kill self – 9.2%

This DOJ study considered stalking only from the standpoint of stalker and victim, and did not involve any stalker typologies. Among all stalkers, the following rates of violence were found:

1. Property damage – 24.4%
2. Attacked victim (hit, choked, raped, used weapon) – 21%
3. Attacked 3rd party or pet – 15%



The 3 Principles of Threat Assessment:

1. Targeted violence is neither impulsive nor spontaneous. Targeted violence results from an understandable process of thinking and behavior.

2. Violence is situational and contextual. Violence stems from an interaction among the potential attacker, past stressful events, a current situation, and the target.

3. “Attack-related” behaviors must be identified. Investigation and resolution depend on identifying the discrete behaviors preceding and liked to the attack. Attack-related behaviors move along a continuum from idea – to behaviors/communications – to preparations.


Key Questions for Threat Assessments:

1. What motivated the subject to make the statements, or take the action?

2. What has the subject communicated to anyone concerning his intentions?

3. Has the subject shown an interest in targeted violence, perpetrators of targeted violence, weapons, extremist groups, or murder?

4. Has the subject engaged in attack-related behavior, including any menacing, harassing, and/or stalking type behavior?

5. Does the subject have a history of mental illness involving command hallucinations, delusional ideas, feelings of persecution, etc. with indications that the subject has acted on those beliefs?

6. How organized is the subject? Is he/she capable of developing and carrying out a plan?

7. Has the subject experienced a recent loss and/or loss of status, and has this led to feelings of desperation and despair?

8. Corroboration – What is the subject saying and is it consistent with his/her actions?

9. Is there concern among those that know the subject that he/she might take action based on inappropriate ideas?

10. What factors in the subjects life and/or environment might increase/decrease the likelihood of the subject attempting to attack a target?

Examining the Data: Communications & Behaviors

The forensic psycholinguistic analysis is best done as a part of a team approach, where the forensic psychiatrist can meet with and obtain feedback from law enforcement, the victim, and the DA assigned to the case. From the outset, the forensic psychiatrist should pay careful attention to the quality of the evidence examined. It is recommended that the forensic psychiatrist obtain the best possible sample of the writing, communication, etc. A trip to the detective’s office may be necessary to view the original evidence.[1] This may also be useful where the detective involved can provide other helpful background information to the investigator in person.
If copies are used, make sure they are high quality and complete. It may be helpful to make multiple copies of single documents so that highlighting, notes, etc. can take place, yet you will still have a clean copy.

Review document(s) carefully, slowly and multiple times. Each review may be done with a different primary purpose. For example, the first review may simply be to obtain a general first impression. Subsequent reviews may be done for:

Identifying themes, motives
Identifying evidence of mental illness
Identifying threatening language, types of threats
Identifying idiosyncratic language or symbol use
Comparisons with other related documents
Identifying basic features, such as type of medium, style of handwriting, dates, drawings, postage markings, etc.
Method of delivery of the threat


After careful, multiple examinations, it may be possible for the investigator to determine important basic information such as the unknown stalker’s: age, sex, ethnicity, geography, educational level, religious orientation, and other valuable data.[2]

The use of language may also suggest different types of mental illness, such as schizophrenia, or depression.[3], [4] In particular, the excessive use of pronouns has been associated with high levels of psychological distress.[5] The use of metaphor or metonymy may also lend clues about an individual’s past experience, ethnic background, primary motivations and level of distress.[6] One psycholinguistic study of threateners from the FBI’s NCAVC database found that higher conceptual complexity and lower ambivalent hostility/paranoia were more strongly associated with predatory violence.[7]

A working knowledge of recent internet communication trends is important, as 25% of victims reported some form of cyber or electronic stalking.[8] For example, even a piece of data as seemingly unimportant as an E-mail address may suggest clues about the stalkers personality structure.[9] Do not fail to listen to any CDs, audiotapes or other recorded media, as stalkers may communicate what they believe are their most “important messages” in seemingly unimportant “gifts.”

Keeping the above principles in mind during the analysis, one should consider paying close attention to the following data[10]:

§ Changes in tone, affect, organization, etc., over time in serial communications.
§ Statements suggesting the stalker has knowledge of victim’s location
§ How the stalker perceives consequences of re-contact
§ Stalker’s responses to any past victim actions
§ Presence/absence of violent intent
§ Duration and intensity of infatuation
§ References to third parties, suggesting some degree of morbid jealousy. Thus, third parties may also be at risk if he perceives them as thwarting his goal.
§ Violation of personal boundaries (visits to the victim’s home or office to deliver communications)
§ Evidence of persecutory delusions coexisting with linear thought processes
§ Evidence of erotomanic delusions, and intensity of intimacy fantasies and unrealistic desires
§ Severity of mental disorder as suggested by communications
§ References to access to weapons or weapon usage
§ Language suggesting that the stalker views himself as an “aggrieved victim.” In some cases, this may suggest a lowered threshold for acting on threats, as the stalker may feel justified in seeking retribution against his perceived persecutors.
§ Evidence of or references to possible substance use
§ Antisocial and Borderline personality styles – (past failure to conform to the law, lack of remorse, fear of abandonment, difficulty controlling anger, impulsivity)
§ Overall personality structure suggestive of an “externalizing” style of coping
§ Statements demonstrating a forceful sense of entitlement (eg., “I’m not going to ask for you, I’m going to take you”)
§ Frequency and intensity of relevant cognitive distortions – (eg., minimization, denial and externalization of blame)

Finally, consider consulting with a qualified forensic document and/or handwriting examiner in difficult cases.


[1] Morris R: Forensic Handwriting Identification: Fundamental Concepts and Principles. San Diego, Calif: Academic Press, 2000.
[2] Smith S, Shuy R: Forensic Psycholinguistics: Using Language Analysis for Identifying and Assessing Offenders. FBI Law Enforcement Bulletin. April 2002, pps. 16-21.
[3] Stephane M, et al.: Empirical evaluation of language disorder in schizophrenia. J Psychiatry Neuroscience, 2007;32(4):250-8.
[4] Pennebaker J, Stone L: Katie’s Diary: Unlocking the Mystery of a Suicide. (D. Lester, Ed.). Ch. 5, pps. 55 – 79; Routledge Press, 2003.
[5] Henken V: Banality reinvestigated: A computer-based content analysis of suicidal and forced death documents. Suicide and Life-threatening Behavior, 1976; 6: 36-43.
[6] Eynon T: Cognitive Linguistics. Advances in Psychiatric Treatment, 2002; 8: 399-407.
[7] Smith S: From Violent Words to Violent Deeds: Assessing Risk From FBI Threatening Communication Cases. In: Stalking, Threatening, and Attacking Public Figures: A Psychological and Behavioral Analysis. (J. Meloy, L. Sheridan, J. Hoffman, Eds.) New York, NY: Oxford Press, 2008; Ch. 20, pps. 435-455
[8] Baum K, Catalano S, Rand M: Stalking Victimization in the United States. Bureau of Justice Statistics Special Report, U.S. Dept. of Justice. January, 2009: pps. 1-15
[9] Back M, Schmukle S, Egloff B: How extraverted is honey.bunny77@hotmail.de?: Inferring personality from e-mail addresses. Journal of Research in Personality, 2008; 42: 1116–1122
[10] This list of evidence consists of both risk factors from the literature, as well as the author’s experience.
[1] Calhoun F, Weston S: Threat Assessment and Management Strategies: Identifying the Howlers and Hunters. Boca Raton, FL: CRC Press, 2009.
[2] Meloy J: Violence Risk and Threat Assessment: A Practical Guide for Mental Health and Criminal Justice. San Diego, Calif.: Specialized Training Services, 2000.
[3] Baum K, Catalano S, Rand M: Stalking Victimization in the United States. Bureau of Justice Statistics Special Report, U.S. Dept. of Justice. January, 2009: pps. 1-15


Sunday, April 5, 2009

Mass Murder




Above: Mass Murder is not new
Middle: Charles Whtiman - died 8/1/66

Below: Life cover of U.T. Austin Tower shooting by Whitman



With the recent tragic mass murder in Binghamton, NY, I felt compelled to post some of basic research knowledge about Homicide-Suicides, of which Mass Murder is a category. My deepest sympathies go out to the survivors and community. I hope that the Binghamton community can come together to grieve, support one another, and not let this isolated act of selfish anger and resentment define them.


There is often the (false) impression in the media that mass murder more or less began with the Charles Whitman incident, but this is simply not the case. They have been occuring probably for longer than we have recorded history available to us. It is simply that the media has become more sophisticated and adept at reporting it (See above table). Here, I will briefly outline the phenomenon of Homicide-Suicide, and then discuss Mass Murder.

Homicide – Suicide is the phenomenon in which an individual commits a homicide, and subsequently (usually within 24 hours) commits suicide (Felthous, 1995; Marzuk, et al, 1992). The dramatic nature of a completed homicide-suicide frequently captures media attention, while efforts at recognition and prevention have received much less consideration. Because the event leaves no living victim or perpetrator, input from a mental health professional is typically not sought.

Most information about homicide-suicide has been gathered from data present in police and coroner’s reports. (Malphurs, 2002; Felthous 2001; Morton, 1998). Few studies have utilized interviews of family members, in addition to record reviews, to enhance the psychological autopsy approach.

The rate of homicide – suicide has been found to vary only slightly throughout the world. In the United States, rates have been reported as more or less consistently between 0.21 to 0.55 per 100,000 (Coid, 1983; Milray, 1995).

Although it is a relatively rare event, it is likely responsible for 1,000 to 1,500 deaths per year in the United States (Marzuk, et al, 1992). Coid (1983) found that countries with a high homicide rate had the lowest rate of homicide – suicide.

In their review of the literature Marzuk, et al (1992) were the first to propose a clinical typology for the classification of homicide – suicide. Their system categorizes perpetrators based on victim – perpetrator relationship, and by class of precipitants or motives. For example, it has been reported that the most common type of homicide – suicide is spousal killing, usually with the male killing his female “consort” due to a breakdown of the relationship (Milray, 1995). Marzuk, et al (1992) have classified this type as a Spousal homicide – suicide of the “Amorous Jealousy” class.

Depression was found to be the most common diagnosis in perpetrators of spousal homicide-suicide. These dyads were commonly characterized as chaotic, abusive relationships. In addition, histories of alcohol abuse and violent behavior are frequently found among this type of perpetrator (Rosenbaum, 1990). The common thread running through this type of homicide – suicide appears to be the precipitating factor of a loss of a previously intimate consort. Indeed, recent estrangement of a partner increases the risk of both homicide and homicide – suicide (Darpet, 1966; Currens, 1991).

Another common type is the Spousal homicide – suicide of the “Declining Health” class. In this group a male, usually elderly, kills his spouse and then himself because of declining health and it’s associated hardships. In actuality, both may be suffering declining health, or conversely, only one has health issues while the other suffers from depression. There may have been some form of threat (eg, financial) to a spouse’s ability to continue functioning in the caretaker role. Beginning in the 1990’s, younger couples suffering from AIDS have been classified in this group. Cohen argues that homicide – suicides of this class are not acts of love or altruism, but of depression and desperation (Cohen, 1998).

Other typologies of homicide – suicide seen with less frequency are the filial, familial, and extra-familial types. Filicide – suicide usually involves the classic scenario of a depressed and psychotic mother who kills her infant in an “extended suicide” (Resnick, 1970; Marzuk, et al 1992). A familicide – suicide is usually committed by a depressed man who kills his entire family. He is likely to view his act as a delivery of the family from continued hardships or stressors (Selkin, 1976).

The Extra-Familial type is sometimes referred to as the “Adversarial” type because the most common event involves an offense against a perceived “enemy” who is unrelated to the perpetrator. Adversarial homicide-suicides typically consist of disgruntled employees, or antagonistic, hate filled individuals. They are likely to be suspicious loners who have had a recent social stressor. They are prone to perceiving themselves as persecuted, and seek revenge in the workplace, or indiscriminately in public (Dietz, 1986; Felthous, 1995; Marzuk, 1992).

Mass murderers who commit suicide would fit into this category, as their relationships to their victims are often extra-familial and adversarial in nature. The U.S. Bureau of Justice has defined “mass murder” as the killing of four or more victims at 1 location, within 1 event. Thus, the following categories of homicide-suicide could also potentially be considered mass murders: 1) the “disgruntled” (ex) employee, 2) the “class room avenger,” and 3) the “pseudo-commando.”


Adversarial Homicide-Suicide (Extra-familial)

This type involves a disgruntled employee who has recently been dismissed or is experiencing work stress. He externalizes blame onto his supervisors or co-workers, and feels wronged in some way. He is very likely to have depression, as well as paranoid narcissistic traits. Actual persecutory delusions may be seen. Variants of this type include disgruntled students, patients, and litigants.

The phenomenon of mass murder described by Dietz (1986) has some over-lap with homicide-suicide. A mass murder occurs when multiple victims are intentionally killed by a single offender in a single incident. The "pseudo-commando" subtype of mass murder can be considered a homicide-suicide in certain cases. The pseudo-commando is usually a man who is feeling strong anger and resentment, in addition to a paranoid character(Dietz). He kills indiscriminately in public during the day time. He uses a powerful arsenal of weapons, and has no escape planned. This may sometimes involve a "passive suicide" in that he forces police to kill him in a last stand "blaze of glory."

Pseudo-commando mass murders have been described as often possessing the following characteristics (Mullen, 2004):

§ Bullied or isolated as a child
§ Loners who are socially excluded and despair over feeling excluded
§ Suspicious, resentful, grudge holders
§ Obsessional or rigid traits
§ Narcissistic, grandiose traits
§ Externalizers – unable to take responsibility for their distress and place responsibility on others
§ Weapons collector, preoccupied with weapons
§ Strong feelings of persecution or mistreatment
§ World seen as rejecting, uncaring
§ Resentful with rumination on past humiliations – “collectors of injustice” (Dietz)
§ Fantasize about violent revenge
§ No significant criminal or violence history
§ No significant mental health history or serious mental illness
§ No significant substance abuse


The massacres carried out by pseudo-commandos are often characterized by the following:

§ Well planned out – not impulsive, did not “snap” (JK: If you investigate closely enough, you will find that no one "just snaps." This is a lay-myth. The act is the culmination of a long period of harboring/collecting resentment and fantasizing about violent retribution)
§ Set out to kill as many people as they can
§ Come well prepared and well armed, often in camo or “warrior” gear
§ Pursue a highly personal agenda of “pay back” to an uncaring, rejecting world
§ Sometimes also against those he has a grievance with


Mullen (2004) raises the possibility that these “autogenic massacres” (mass murders) are unique to western society. Further, subsequent pseudo-commandos appear to have been inspired or influenced by previous ones via the media. The perpetrators “welcome death,” and perceive it as bringing them fame with an aura of power.

Here are more results from some interesting studies:


Mass Homicide & Suicide: Deadliness & Outcome (Lester, ’05)

This study examined a nonrandom sample of 98 Lone “rampage” killers. Lester defined "deadliness" by number of victims.

In terms of deadliness, here's how the perpetrators fell out:

  • Most deadly: Killed by police
  • 2nd most deadly: Committed suicide
  • 3rd most deadly: Captured by police

Lester also found that perpetrators often showed an interest in guns, had past violence, and demonstrated paranoia or paranoid traits. Interestingly, disgruntled employees were more likely to commit suicide.


Mass Murderer Characteristics (Hempel, ’99)

This study looked at a sample of 30 nonrandom mass murderers and found:


•Paranoid and/or depressive traits
•Personality disorders
•Major loss precipitating the event
•A “Warrior mentality” among perpetrators - coming to the event "decked out" in military-like garb, heavily armed and even sometimes shouting particular "war cries."

Another common theme running through these types of events is the toxic effects of social isolation or rejection. This phenomenon has been well studied by Baumeister, who has shown that social rejection engenders feelings of nihilism, hopelessness, anger, as well as impaired cognition and ability to make cautious decisions.