Skip to main content

Verified by Psychology Today


Are All Mass Killers "Sick"?

Personality disorders are “mental disorders,” but are they “illnesses”?

Key points

  • A diagnosis of a personality disorder is often interpreted as connoting someone who is afflicted with a mental illness.
  • We are often unrealistically optimistic about the potential for psychiatric treatment to change who a person is.
  • In the extreme case of mass murderers, we assume that what they need is psychiatric treatment. Let’s be more realistic.
  • Our intuitive dualism leads us to think of mental disorders as separate from who the person is—as things that “happen” to the “self.”
soupstock | AdobeStock
Source: soupstock | AdobeStock

In the 26 years that I have been a psychiatrist, I have seen many diagnostic fads come and go—some driven by my profession and some by the general public. One of the more surprising ones has been a fairly recent trend of patients self-diagnosing with borderline personality disorder. (I'm mentioning BPD here not to imply that people with that disorder are likely to become mass killers, but to make a more general point about personality disorders. We'll get to the topic of mass killers in a minute.)

Personality Disorder as an Affliction1

The desirability of a BPD diagnosis is particularly strange to any clinician who has been in practice for a while, as BPD used to be a highly undesirable and stigmatizing label. It typically describes a very dysfunctional, emotionally unstable, self-destructive person, someone who might also be very manipulative and who is likely to drive damaging wedges into their relationships with other people. It's worth noting as well that the criteria for BPD overlap with other conditions and personality traits, so it’s prone to overdiagnosis—especially among those who are self-diagnosing on the internet.2

Perhaps part of BPD’s sudden unexpected appeal stems from the oversimplified notion that all mental disorders are illnesses, separable from the self—that they are things that “happen” to the brains of people, perhaps early in development, rather than (as is the case for personality disorders) descriptions of who the person is.

A diagnosis of BPD has somehow come to connote a person who is afflicted with a mental illness, and therefore a victim—rather than someone who is a very difficult person (albeit due to factors partly beyond their control). Other “cluster B” personality disorders3—such as narcissistic or antisocial personality disorders—have not yet acquired the same kind of sympathetic connotations or self-diagnosis trends that BPD has begun to acquire, but they also tend to be regarded as mental illnesses.

“Sick” Antisocial and Homicidal Behaviors

A common reaction by the public and the media to horrendous crimes, such as school shootings and other mass killings, is to label the perpetrator as mentally ill and to lament the shortage of psychiatric care, presuming that if only the perpetrator had been referred to and followed up by a psychiatrist then the tragedy could have been averted. Fingers are often pointed at overloaded hospitals that had sent a future mass killer home after he—it’s usually ”he”—had been taken there because of having made suicidal or homicidal threats. This is a particularly common reaction by people who would prefer to divert attention from more obvious solutions like gun control.

(Another well-known yet still insufficiently-adhered-to solution is for the media to completely stop publishing the killers’ names and manifestos, and to stop publishing a running tally of who has achieved the highest “kill count” so far. We know that mass killings are often a copycat phenomenon and a perverse competition for infamy).

Only a minority of homicides and mass homicides are committed by people with potentially treatable mental illnesses. Research suggests that individuals with schizophrenia and bipolar disorder, which are the two major mental illnesses associated with psychosis—a loss of touch with reality—are responsible for approximately 10 percent of all homicides in the United States and 33 percent of mass killings (at most).4

This is an important and significant minority of cases, to be sure, as psychosis can potentially be controlled with medications. But those who go on to kill others are a small percentage of all people suffering from psychosis, and it can be extremely hard to prospectively predict who will do so (it’s notoriously hard enough even to predict who will die by suicide). Hardly any psychiatrists will ever come across a patient who goes on to commit mass murder.

People with antisocial, psychopathic, narcissistic, and paranoid personality disorder traits likely account for a major proportion of mass murderers, and these are mostly untreatable.5, 6 Most psychiatrists have no more desire than anyone else to deal with these unpleasant and often dangerous individuals. And let’s be realistic here, if you’ll forgive an extreme example to make the point: Would World War II have been averted if only young Adolf had seen a psychiatrist as a teen, or even as a child?

Personal Responsibility and Free Will

How should we strike the balance between mental disorder diagnoses explaining away and excusing or absolving people’s personal failures and transgressions, while still preserving a societal value of personal responsibility? How much of human behavior can and should be reasonably attributed to involuntary impingement of self-control by a mental disorder (i.e. ”My brain made me do it”)?

On the one hand, if we really understand free will from a neuroscientific point of view, we ought to believe that there is no such thing. On the other hand, for practical purposes, it seems reasonable to expect people to exert something resembling or approximating free will under ordinary circumstances. We need to preserve some basic societal expectation of personal responsibility. Major mental illnesses do reduce cognitive and behavioral control and flexibility and thus the “degrees of freedom” of volition. But not all mental disorders do so to the same degree.

Some Misconceptions About Mental Disorders Are Rooted in Dualism

Our intuitions can easily mislead us into holding dualistic notions about the self as an independent entity that “has” or is afflicted with or is suffering from a disorder or illness. Those misconceptions have been unintentionally reinforced by the language of mental health education and destigmatization campaigns, suggesting explicitly or implicitly that all mental disorders are separate from who the person is—things that happen to the self. (This seems to be part of what people are thinking when they eagerly embrace a self-diagnosis of borderline personality disorder.)

Another myth is that the self is a permanent unchanging essence. Whereas, in reality, the self changes all the time, including when afflicted by a mental disorder and—in the case of disorders that are reversible or episodic—when recovering from it.

Some mental disorders can probably be understood as mostly separable from the self—illnesses that have happened to, "landed on," or “infected” the self. Perhaps a stronger argument can be made in this regard for episodic, fully reversible disorders, as these represent more of a state than a trait—a clearer out-of-character change from the person’s usual personality, particularly if followed later by a return to that person’s normal baseline.

These are the more textbook-typical mental illnesses. Even relentlessly progressive, deteriorating, or degenerating illnesses such as dementia and some of the more devastating forms of schizophrenia can be understood this way—as separable from the self—although, sadly, in these diseases, we can only talk of the person’s former self. It makes no sense to imagine that these individuals continue to have a healthy self that is somehow being suppressed by the disease and is still underneath it, struggling to break free. That former self no longer exists.

Not Separable from the Self

But personality disorders, by definition, are not separable from the self. They are the self. This is not a matter of blaming the individual—they didn’t choose to have such maladaptive traits, and in many cases desperately wish not to. They do deserve a great deal of understanding from the rest of us who, by dumb luck, were dealt a better hand in terms of temperament.

Still, it can be very hard to be sympathetic toward people with really malignant personalities whose maladaptive traits lead them to behave in ways that harm the rest of us. And when some of those individuals commit heinous crimes, society must be protected from them through the criminal justice system. We simply don’t have any reliable treatments or therapies to rehabilitate or change the personalities of those at the far end of the spectrum of destructive personality traits.

The solutions, or at least partial solutions, are primarily societal, and the principal ones, such as gun control, are maddeningly obvious to many.

Realistic Expectations

We psychiatrists don’t have the answers that we and society wish for. Human nature is complex. Expectations need to be more realistic regarding the potential for psychiatric treatments and therapies to change who people are.

With appreciation to Dr. Graham Glancy, Director, Division of Forensic Psychiatry, Department of Psychiatry, University of Toronto, for reviewing this article prior to its publication.


1. Parts of this post are adapted from my Psychology Today print edition article “The Curious Cachet of a Psychiatric Diagnosis.” 2022 Jan/ Feb; 55(1)

2. In my clinical experience, I actually find that what a lot of people diagnosed or self-diagnosed with BPD really have is a combination of high levels of anxiety (accounting for their emotional sensitivity and volatility), and strong ADHD characteristics (accounting for their tendency toward chaotic lifestyles and difficulties with self-regulation). Thus, they are like a pot on a stove that is quickly and easily brought to the boil, with a lid that is insufficiently tight.

3. “Cluster B” personality disorders are marked by dramatic, overly emotional, or unpredictable thinking or behaviors.

[Click 'More' to view footnotes 4-6].

4. See and These estimates are probably at the higher end. See also:

– Peterson, J. K., Densley, J. A., Knapp, K., Higgins, S., & Jensen, A. (2022). Psychosis and mass shootings: A systematic examination using publicly available data. Psychology, Public Policy, and Law, 28(2), 280–291. In this dataset of 172 mass shooters from 1966 to 2020, psychosis played no role in 69% of cases, but may have played a minor role in 11% of cases, a moderate role in 9% of cases, and a major role in 11% of cases.

– Brucato, G., Appelbaum, P. S., Hesson, H., Shea, E. A., Dishy, G., Lee, K., Pia, T., et al. (2021). Psychotic symptoms in mass shootings v. mass murders not involving firearms: findings from the Columbia mass murder database. Psychological Medicine, 1–9. Cambridge University Press. In this study of 1315 mass murders reported globally in English from 1900 to 2019, 65% of which involved firearms, lifetime psychotic symptoms were noted among 11% of perpetrators, including 18% of mass murderers who did not use firearms and 8% of those who did.

5. Psychopathy is not a DSM-5 category. Psychopaths can be thought of as being at the more severe end of the spectrum of Antisocial Personality Disorder, but psychopathy additionally implies a lack of empathy—a more cold-blooded, heartless type of person, whereas the DSM-5 criteria for Antisocial Personality Disorder focus mainly on antisocial / criminal behavior ("a pervasive pattern of disregard for, and violation of, the rights of others")—these behaviors tend to be driven mainly by impulsivity, recklessness, deceitfulness and irresponsibility, usually involves a lack of remorse, but may or may not involve a severe lack of empathy. A particularly malevolent combination of overlapping personality traits that has gained recognition in the psychological literature in the last couple of decades is the so-called dark triad, marked by narcissism (characterized by egocentrism), Machiavellianism (exploitativeness), and psychopathy (lack of empathy). Milder versions of these traits are not uncommon and usually do not lead people to serious criminal behavior. The really malevolent and potentially dangerous people are the small minority of the population with more marked versions of these traits.

6. A variety of other psychiatric disorders are also sometimes implicated in these kinds of crimes. One in particular that has attracted attention is Autism Spectrum Disorder. It is unclear to what degree that disorder could adequately account for such murderous behavior, and courts have generally not accepted a diagnosis of ASD as a defense. After all, the vast majority of people with ASD would never commit such a crime.

A familiar, common profile of a mass killer these days is that of an angry, resentful, socially isolated, insecure young man with rigid personality traits who feels aggrieved and vengeful, has been radicalized by the internet and social media, and is seeking fame through notoriety.

More from Ralph Lewis M.D.
More from Psychology Today
More from Ralph Lewis M.D.
More from Psychology Today