Language and Labels
What to Say Instead of “Narcissist”
The laundromat on a Sunday morning is the warmest room on the block, it smells like hot cotton, and every third dryer has a coin-sized rattle in it.
Josiah is thirty-four and has been talking about his manager for nine minutes over the noise, which everybody at the folding table has agreed to without agreeing to it.
The complaint is one he has made before, in the same order. He was promised the early shift when it came free, it came free in June, and he has been re-promised it twice since. He reads the message out. Absolutely, next cycle, I’ve got you. Dated the ninth of July.
Then his friend sets down a folded towel.
“Josiah. Come on. He’s a narcissist.”
What happens next takes about two seconds. The shoulders come down, all of them, including his. Somebody laughs — not cruelly, a finally laugh, the sound a room makes when a long problem turns out to have a short answer. Within four minutes the folding has resumed and the conversation is about somebody’s car.
Nobody asked how many times a promise had been made and dropped, three or eleven, because nobody counted and neither had he.
The relief was real, and I am not sneering at it. Four people wanted their friend to stop hurting and gave him the fastest thing they had.
On Monday, Josiah did exactly what he had done the Monday before.
Three objects, and the door one of them opened
Most arguments about this word are two people holding different objects. There are three.
A clinical diagnosis is a trained person’s judgment about a long pattern that is impairing somebody’s life, assessed against written criteria, requiring persistence across a whole life rather than one bad department — and it is uncommon. A trait score is a position on a continuum in a room of ordinary people; when Paulhus and Williams introduced the Dark Triad, their opening line called these “offensive yet non-pathological” personalities. An epithet is what somebody calls their ex, and Josiah’s friend, who has never met the manager, is not making a mistake. She is naming an experience.
Only one of the three can be checked. Writing in the British Journal of Psychiatry, Freestone, Osman and Ibrahim put it about as bluntly as a journal permits: in public usage the term has lost accurate meaning and become a label used routinely to describe and dismiss people who appear self-serving. Describe and dismiss, both in one move, in under a second. I have used it too, at a volume suggesting I had been saving it up.
The word was a door, and that has to be said first
Before any of this asks anybody to set the word down, it has to say what the word has done.
Marta was married eleven years to a man everyone found generous and funny and a little unlucky. Money left the joint account and appeared in another one, and the explanations were always plausible and always after the fact. She was confused not because she was foolish but because the experience had no shape: every piece of it was defensible, the whole of it was crushing her, and there is no ordinary English sentence for that.
Then, past two in the morning, under a video she had not gone looking for, she read a description of a pattern, and the pattern had a name. The relief was physical — the relief of the thing being a thing, with edges, that other people had lived inside and come out of. Altmann and colleagues found that applying a diagnostic label to a marginal case increased empathy toward the person and increased judgments that help was warranted — alongside costs the same work measured, since labeling also reduced how much personal control people saw over the problem and raised how persistent they expected it to be. The benefit came first in their findings, so it comes first here.
Nothing here tells Marta she was wrong. The fair question to ask of a tool that has already done its job is: what is the word for now?
How a word gets bigger, and who pays
Concept creep. Plain sentence: words for harm get bigger over time, coming to cover milder things and new things. Picture a smoke alarm installed to tell the house about a fire, which now goes off reliably when somebody makes toast.
Haslam named the process. It runs downward, to milder cases, and outward, to things that were never in the category — narcissist downward to a man who talks too much at dinner, outward to the posted photograph and the curated feed. Haslam and colleagues call the consequences genuinely mixed rather than bad, and they are right: the alarm that goes off for toast is the alarm that catches the fire at three in the morning. So it is a trade, and the useful question about a trade is who pays.
Somebody measured that. Dakin, McGrath, Rhee and Haslam published a paper under a title that gives away the ending: “Broadened concepts of harm appear less serious.” Across five experiments, people shown marginal examples of a harm concept went on to judge severe instances of the same harm as significantly less serious.
That is who pays. The woman whose husband has moved the money needs this word to still be heavy on the day she says it to her brother, and he has spent a decade hearing it applied to a manager who takes credit and a cousin who talks over people at Christmas. He is not callous. He has been calibrated.
One honest limit: those experiments were about workplace bullying, trauma, and sexual harassment, not about narcissist, and the inference is mine. Almost nothing has been studied about this word as a lay diagnosis. Its neighbors are no better established — when Darke, Paterson and van Golde took gaslighting into a laboratory, participants called nearly everything gaslighting, including scenarios missing its defining features, and trauma bonding rests on no established measurement literature at all. That does not make the underlying experiences imaginary. These terms are simply running far ahead of anybody’s ability to check them.
What the label does to the person holding it
A label closes inquiry. Once you have an answer you stop asking questions, the way somebody says it’s an old car about a noise on the highway — which is true, and which is why the noise is still there in November.
Josiah is the demonstration. Eight months later he is on the same shift, and he no longer describes the corridor or the message, because he has something more compact. He says his manager is a narcissist, people say oh, God, I’ve had one of those, and the exchange is warm and quick and closed. He has explained the man to eleven people and decided nothing.
Inquiry is uncomfortable — holding a situation open, keeping a count of something you would rather not count. It is also the only process that produces a decision, and the label stops the accumulation while supplying the exact sensation of having concluded something.
He is a narcissist is not actionable. He has agreed to four things since February and done none of them, and each time I raise it we end up discussing my tone is actionable — not because it is nicer, but because it has moving parts. A number that can go up. A pattern with a next occurrence. A specific thing that happens when it is raised.
There is also a mirror. Do we love people for who they are, or for the functions they perform for us? The label runs that compression the other way, and a man with a history becomes, in one word, a function.
If any of this arrived as an accusation rather than a description, that reaction is shame, and shame converts a sentence about conduct into a sentence about worth. Noticing yourself in it is not a verdict.
Three sentences, and the test that matters
Replace the label with a description, and a description has three parts. What happened — concrete, the kind of thing a camera would have caught, not a motive. How often — since three times in five years is an entirely different sentence from three times this month. What happened when it was raised — the part people skip, which carries the most information, because it tells you whether the situation has any repair mechanism in it.
The label: “My brother-in-law is a narcissist, he makes everything about himself.”
The description: “He has offered four times since February to help clear out my mother’s house, and he has come once, for about ninety minutes, and spent part of that on the phone. When I mentioned it in June, he said he was hurt that nobody in this family notices what he does do. We have not discussed the house since.”
The second version is longer and also worse for him. The label asks nothing of anybody; the description gives you a count, a pattern, and the fact that raising it moved the subject from the house to his feelings.
Now the test, and it is done entirely in your own head, on a page nobody else sees. Imagine reading it aloud to a lawyer, a pastor, your sister — and to the person it is about. A description survives all four imaginings. A lawyer could work with it, because it has dates. A pastor could respond, because it does not require him to condemn anybody’s soul first. Your sister could hear it without picking a side. And in the fourth version the sentences still hold their shape, because the most anybody could do with them is dispute a particular fact. Whether that fourth conversation should ever actually happen is a separate question, and a real one, and this is not the place it gets answered. The label survives none of the four.
None of this suggests your situation is milder than you think. It might be worse. The label feels like knowledge and works like a door closing; the description feels like homework and works like a window.
One thing you can do this week
The three sentences. Half an hour, once, on paper. Done alone; nobody is told, nothing is sent.
Pick one person you have a word for — an ordinary one rather than the worst one, the first time. Write the word at the top of the page, draw a line under it, and do not use it again there. Then write what happened, with a date if you have one. Then how often. Then what happened when it was raised, and if it never has been, write that instead, in those words.
Most people find one of the three will not fill in, and which one tells you something. An empty first sentence usually means you are carrying a conclusion rather than a situation. An empty second means you have never counted. An empty third is the most common of all, and it is not a criticism: there are excellent reasons not to raise things, including safety, and this never asks you to.
The short version
- A diagnosis is a clinician’s judgment against written criteria. A trait score is a position on a continuum. An epithet describes how somebody made you feel. Only one can be checked.
- Words for harm expand over time. Haslam and colleagues call this concept creep, and judge its consequences genuinely mixed rather than bad.
- Dakin, McGrath, Rhee and Haslam found across five experiments that people shown marginal examples of a harm judged severe instances as less serious. Broadening a word is paid for by the people at its severe end.
- A label closes inquiry, and inquiry is the only process that produces a decision.
- A description has three parts: what happened, how often, and what happened when it was raised. Imagined aloud to a lawyer, a pastor, a sister, or the person it is about, it survives all four. A label survives none of them.
- If you have used the word, and most of us have, noticing that is not a verdict on you.
Josiah and Marta are composites. The details are borrowed from a great many people and belong to none of them.
One note about Marta’s situation in particular. Where money, movement, or permission have come under somebody else’s control, that is not a language problem. There are people who do this work for a living — an advocate who works with domestic abuse, an attorney, a therapist who has seen the shape before — and asking is neither a confession nor an accusation.