THE PSYCHOLOGY OF SERIAL KILLERS: WHAT SCIENCE KNOWS

An investigator studies a board covered with photographs. The temptation is to imagine that the connecting thread will eventually reveal a single explanation: a damaged brain, an abusive childhood, a personality without conscience. The board in our illustration is fictional. The desire for an explanation is real. Repeated murder makes us search for a boundary between ordinary people and those capable of extraordinary harm.

Science offers no clean boundary of that kind. It offers definitions, imperfect observations and competing explanations. A case file can establish acts; an interview can suggest motives; a brain scan can reveal a statistical difference. None provides an unobstructed view into a person's inner life. This investigation draws on official records and peer-reviewed research checked in October 2026, while keeping victims and the limits of knowledge in view.

A category of crime, not a diagnosis

The FBI's 2008 multidisciplinary report defines serial murder as the unlawful killing of at least two victims by the same offender or offenders in separate events. This is an investigative definition, not a mental disorder. It separates a repeated pattern from killings within one event without telling us why either occurred. The report rejects a universal offender profile and cautions against familiar myths about intelligence, motives and social isolation. FBI multidisciplinary report, 2008

That distinction changes the question. Instead of asking what every serial killer must be like, researchers ask which processes may contribute to particular offenders' behaviour. Sexual gratification, power, financial gain and other motives can overlap. Repeated violence may involve planning, opportunism or both. A label that groups offences together should not erase differences between people, victims, circumstances and evidence.

The evidence is uneven. Researchers cannot randomly assign people to childhood adversity or criminal careers. Much forensic knowledge comes from convicted offenders, archived cases and interviews conducted after the outcome is known. Uncaught offenders are missing from that picture. Successful investigations may be unusually well documented, while failures leave gaps. Such selection makes confident claims about all serial murderers especially difficult to defend.

Psychopathy, narcissism and the problem of empathy

Psychopathy describes a researched constellation of traits involving interpersonal manipulation, callousness, reduced remorse and aspects of impulsive or antisocial behaviour. It is not another word for psychosis, which can involve impaired contact with reality. Nor is it interchangeable with antisocial personality disorder, a clinical diagnosis with substantial overlap but different criteria. Assessments such as the Psychopathy Checklist-Revised require trained evaluation and records, not a viewer's reaction to a documentary. De Brito et al.: psychopathy, 2021

Empathy also contains different abilities. Understanding another person's perspective is not identical to sharing their distress or caring about their welfare. A person might read vulnerability while showing little concern for it. Conversely, difficulty interpreting emotions does not imply cruelty. Researchers debate how emotion, attention and learning interact in psychopathy. Describing someone as having no empathy compresses those distinctions into a misleadingly simple switch.

Narcissistic traits, including entitlement and an inflated self-image, can be relevant to aggression without explaining serial homicide. A 2021 meta-analysis by Sophie Kjærvik and Brad Bushman combined 437 independent studies involving 123,043 participants and found an association between narcissism and aggression. Those outcomes were broader than serial murder. The result neither diagnoses a named offender nor makes narcissistic traits a reliable forecast of killing. Kjærvik and Bushman, 2021

Ordinary vanity, interpersonal unpleasantness and psychiatric labels should never become shortcuts to suspicion. Traits lie on dimensions, measurement varies and the same trait can coexist with very different conduct. Explaining an association requires asking who was studied, how the trait was measured and which behaviour counted as aggression. Without those questions, a research finding easily becomes a stereotype disguised as a clinical fact.

What a brain scan can actually show

In a 2013 JAMA Psychiatry study, Jean Decety, Laurie Skelly and Kent Kiehl examined 80 incarcerated men across different psychopathy scores as they viewed scenes of pain. Brain responses differed between groups in networks involved in affective processing. The pattern was complex, with stronger responses in some regions and weaker responses in others. It did not reveal a uniformly inactive empathy system or identify a brain signature of serial murder. Decety, Skelly and Kiehl, 2013

Functional MRI measures changes associated with blood oxygenation, not thoughts or moral character directly. A cross-sectional prison study cannot establish whether differences preceded offending, followed adversity or reflected other influences. Group comparisons are not individual diagnoses. An apparently vivid colour map does not tell an investigator that a particular person will commit a crime, and these participants were not a representative sample of serial killers.

Broader neuroscience reviews examine fear learning, reward, decision-making and connections between brain systems. They also identify inconsistent findings and developmental questions. Substance exposure, injuries, social experience and differences in tasks can complicate interpretation. Biological research matters because behaviour has biological mechanisms. It becomes misleading when a mechanism is treated as a complete cause, or a statistical association is offered as a verdict about responsibility. Frazier et al.: biology and environment review

Born or made is the wrong fork in the road

Genes and environments interact throughout development. In a widely discussed 2002 Science study, Avshalom Caspi and colleagues investigated whether variation in the MAOA gene moderated the relationship between childhood maltreatment and later antisocial behaviour. The reported interaction concerned a male cohort and broader antisocial outcomes. It did not discover a serial-killer gene, establish that a variant inevitably causes violence or provide a screening test for dangerous children. Caspi et al., Science, 2002

Candidate-gene findings require particular caution: effects can depend on samples, exposure definitions and analytical choices. A gene's biological function does not establish a direct route from DNA to homicide. The broader literature suggests multiple influences rather than one hereditary switch. A genetic contribution to variation across a population cannot tell us what percentage of one individual's crime was inherited, or whether that individual lacked meaningful alternatives. Frazier et al.: biology and environment review

Childhood maltreatment is associated with later violent outcomes in prospective research, according to a systematic review published in its 2020 journal issue. Association is not inevitability, and these studies address violence generally rather than serial killing. Most survivors do not become violent offenders. Trauma deserves care because it harms children, not because every harmed child should be treated as a possible future criminal. Fitton et al.: prospective-study review, 2020 issue

Retrospective stories require another layer of scrutiny. An offender may remember selectively, minimise agency or present a childhood narrative that makes later acts seem inevitable. Records can support or contradict that account. Development involves protective relationships and changing circumstances as well as vulnerability. An explanation should illuminate opportunities for prevention without converting suffering into destiny or confusing understanding with an excuse.

Motives, fantasies and the illusion of an unstoppable compulsion

The FBI report describes varied motivations, including sexual, financial and power-related ones, and notes that offenders can stop killing for reasons unrelated to capture. Its practical implication is that repeated homicide is not necessarily a continuous, irresistible urge. Investigators must distinguish an offender's explanation from independently established conduct. A claim of compulsion may describe subjective experience, serve self-justification or do both; it does not settle legal responsibility. FBI multidisciplinary report, 2008

Fantasy alone offers no dependable prediction. Many people experience disturbing thoughts they reject and never act upon. The relevant investigative concern is behaviour: targeting, coercion, threats and actual harm supported by evidence. Even when fantasy appears in an offender's history, a retrospective account cannot tell us how many people with similar thoughts never offended. Studying only those who killed makes that missing comparison easy to overlook.

Four cases, four limits on easy explanations

Ted Bundy's case illustrates how apparent social competence can coexist with serious violence. The FBI's account documents interstate investigations, escapes from custody and his February 1978 arrest after an officer stopped a stolen car. Identification material helped establish who was in custody. This was an investigation involving witnesses, coordination and physical evidence, not the discovery of a unique psychological type. His uncertain total victim count should not become a macabre measure of achievement. FBI: Bundy investigation

After Jeffrey Dahmer's arrest in Milwaukee in 1991, the FBI assisted with laboratory examination and victim identification. Its account describes DNA profiling and other forensic analyses, followed by Wisconsin proceedings in 1992 and a subsequent Ohio conviction. Such evidence establishes acts more securely than speculation about an offender's emotional life. Neither his sexual orientation nor that of his victims is a scientific explanation for homicide. FBI: Dahmer investigation

John Wayne Gacy's arrest in December 1978 followed a missing-person investigation by Des Plaines police, according to the Cook County Sheriff's Office. Searches and recovery work were central to the case. The lesson is not that a familiar occupation or respectable public presentation conceals a particular diagnosis. It is that reports of disappearance require attention, evidence must be pursued and victims' identities matter beyond the offender's notoriety. Cook County Sheriff: investigative history

Richard Ramirez's offences are documented in the California Supreme Court's 2006 appellate opinion. The record describes murders alongside sexual assaults, burglaries and other offences, and examines the evidence and trial issues. It is a legal record, not a neuroscience experiment. Turning his public persona into a universal explanation would confuse presentation with mechanism. None of these four cases supplies a controlled comparison proving what produces serial murder. California Supreme Court: Ramirez, 2006

Profiling is a hypothesis, not a fingerprint

Behavioural analysis can help investigators organise questions about offence patterns, access to victims and links between cases. Its evidential status depends on the particular method. A prediction about an unknown offender needs independent testing; a persuasive story after arrest is easier to construct. Profiles should guide inquiry without excluding contradictory evidence. The distinction resembles the problem examined in our investigation of lie detectors: inference must not masquerade as direct observation.

In 2004, David Canter and colleagues examined 39 crime-scene features across 100 murders committed by 100 US serial offenders. They did not find support for a clear division into the familiar organised and disorganised types. The study challenged a popular dichotomy, not every possible form of investigative analysis. Its own archival selection and coding limitations also matter. Scientific criticism should improve methods rather than replace one sweeping certainty with another. Canter et al.: typology test, 2004

Investigators face confirmation bias: once an appealing explanation forms, ambiguous details can begin to look like corroboration. A profile is not a fingerprint, and resemblance to a description is not guilt. Linking crimes requires converging evidence. Comparison groups, transparent methods and prospective evaluation help distinguish useful predictions from hindsight. Intellectual humility is practical here because a mistaken focus can consume resources and leave other victims unprotected.

Myths, warning signs and the cost of stigma

The supposed childhood triad of bedwetting, fire-setting and animal cruelty is often presented as an early serial-killer alarm. Parfitt and Alleyne's critical review, first published online in 2018 and in a 2020 journal issue, found inadequate support for that combined premise. These behaviours may indicate distress or dysfunctional environments. Cruelty and dangerous fire-setting warrant intervention for their immediate harms, while bedwetting is not evidence of a future murderer. Parfitt and Alleyne: critical review, 2018/2020

Another myth treats mental illness as a synonym for violence. Most people with mental disorders are not violent. The APA's review explains that prior violence, substance use and social circumstances can be important alongside specific symptoms. A diagnosis alone is insufficient for assessing danger. This does not mean symptoms can never contribute to risk; it means responsible assessment considers context rather than casting an entire patient population as a threat. APA: mental illness and violence, updated 2022

There is no validated everyday checklist that identifies a future serial killer. Concrete threats, stalking and escalating coercion deserve a proportionate response because they can cause harm now. Support and professional assessment should follow observed needs, not lurid labels. If danger is immediate, seek local emergency assistance. Prevention also means taking missing-person reports seriously and avoiding assumptions that some lives are less worthy of investigative effort.

Understanding without surrendering judgment

The most useful scientific question may be where harm could have been interrupted. Better coordination, careful evidence handling, victim support and access to appropriate care offer practical directions without promising that serial murder can be eliminated. Research should test interventions and investigative methods, while respecting the people whose histories are being studied. Explanation has value when it increases accountability and protection, not simply when it produces a memorable character portrait.

Serial murder exposes a limit in our appetite for certainty. We want one damaged part to explain an entire human life, but findings about traits, development and neural systems remain partial. As our feature on human nature and morality also explores, describing the conditions of behaviour is not the same as resolving moral responsibility. Science can sharpen questions, correct myths and challenge comforting stories. It cannot make victims' losses intelligible by giving the perpetrator a single definitive label.