The Dark Triad: narcissism, Machiavellianism and psychopathy as traits, not labels

In 2002 Delroy Paulhus and Kevin Williams published a short paper asking whether three personality constructs that had been studied separately — narcissism, Machiavellianism and psychopathy — were really distinct when measured in ordinary university students. They found that the three overlapped considerably but were not identical, and gave the cluster a memorable name.

Roughly:

• Narcissism — grandiosity, entitlement, a need for admiration and a sense of superiority.
• Machiavellianism — strategic manipulation, cynicism about human nature, and a focus on self-interest over principle.
• Psychopathy (the subclinical trait, not the forensic diagnosis) — impulsivity, thrill-seeking, callousness and low remorse.

What they share, in the meta-analytic evidence, is a core of low agreeableness and low honesty-humility: a willingness to put oneself first at others' expense.

Two decades of work, summarised in Furnham, Richards and Paulhus's reviews and in Muris and colleagues' 2017 meta-analysis, support a few conclusions and undermine a few others.

Supported

• The three traits are substantially intercorrelated. Muris and colleagues found the overlap large enough to question whether they are three things or one thing with three accents.
• All three are more common in men than women on average, with modest effect sizes.
• All three relate to lower agreeableness (Big Five) and lower honesty-humility (HEXACO), and to a range of negative psychosocial outcomes: aggression, exploitative behaviour, poorer relationship quality.
• Jones and Paulhus's Short Dark Triad (SD3), a 27-item questionnaire, gives a reasonable picture of what each trait covers — and shows how ordinary the items are ("People see me as a natural leader"; "It's not wise to tell your secrets").

Shaky

• The field relies heavily on self-report in student samples. People high in these traits are not always reliable narrators of themselves.
• The distinctness of the three is contested. Machiavellianism in particular is hard to separate from psychopathy in most measures.
• Effect sizes for most outcomes are small to moderate. The traits predict tendencies, not acts.

This distinction does most of the work in using the Dark Triad responsibly.

A high score on a trait scale is not a disorder. A disorder is not something a partner, friend or internet article can identify. And the recent research on narcissism in particular — Pincus's work on the Pathological Narcissism Inventory, Miller and colleagues' 2011 analysis, Krizan and Herlache's Narcissism Spectrum Model — shows that even the clinical construct has two very different faces, grandiose and vulnerable, which look nothing alike from the outside. That is covered in the profile on narcissistic traits.

The research on relationship behaviour is consistent in direction, if modest in size. On average, people higher in Dark Triad traits are more likely to:

• Pursue short-term, low-commitment relationships and to report more infidelity.
• Use manipulation tactics — charm, coercion, guilt — to get what they want from partners.
• Show lower empathy and respond to a partner's distress with irritation or indifference.
• Engage in game-playing and keep partners uncertain, which (see why it's so hard to leave) is precisely the pattern that makes attachment stickiest.
• In the case of grandiose narcissism, to begin relationships with intense idealisation — the pattern described as love bombing — and to devalue when admiration is no longer forthcoming.

Two caveats matter. First, these are average tendencies across many people; any particular person high in these traits may do none of this, and plenty of people who do these things are not high in these traits. Second, the trait is not what harms you; the behaviour is. Which brings us to the practical question.

When someone has hurt you, the pull to classify them is strong. It makes the pain make sense and promises that you will recognise the type next time. The research suggests a different question works better: what is the pattern of behaviour, and what does it cost me?

• Pattern, not incident. One lie, one cold response, one grand gesture tells you little. The same behaviour across months, especially when it is inconvenient to the other person to maintain, tells you a lot.
• Response to boundaries. How does the person react to a clear, polite "no"? Respect, negotiation and disappointment are ordinary. Punishment, escalation and reframing you as the problem are not.
• Accountability. Can they own a specific mistake without turning it into your fault, a competition of grievances or a performance of remorse that changes nothing?
• Cost. Are you smaller, more anxious, more isolated, less sure of your own perception than before? Those outcomes matter regardless of what causes them.

This approach has two advantages. It keeps you observing behaviour you can actually see rather than inferring a trait you cannot. And it leads to the same practical decision — protect yourself, set limits, consider leaving — whether or not the other person would score high on a questionnaire they will never take.

Some behaviour associated with dark traits — intimidation, isolation, monitoring, financial control, threats — is not a personality question at all. It is the pattern Evan Stark named coercive control, and in England and Wales it is a criminal offence. If a relationship involves fear, the question of what trait explains it can wait.

• Narcissistic traits: grandiose and vulnerable — the two faces of narcissism, and the gap between trait and diagnosis.
• The Controlling Partner — coercive control as a pattern of behaviour.
• Why it's so hard to leave — the reinforcement and investment mechanisms that keep people in exploitative relationships.
• Love bombing vs genuine affection — idealisation and devaluation at the start of a relationship.