When Therapy Speak Becomes a Weapon in Your Relationship

The fight is about the dishes, or the schedule, or a comment made three days ago. Then it happens. One of you says, “I’m not going to discuss this anymore. That’s my boundary.”

The conversation does not pause. It ends. The person who said the sentence has walked off the field and taken the ball with them, and they have done it in a way that makes following them look like the violation.

The person on the other end feels a sudden, sharp silence. They are not only unheard. They have been told that the act of trying to speak is now a clinical infraction.

Why does that one sentence work, when “I don’t like this” never does?

We are living through something the psychologist Nick Haslam named concept creep. His observation is that the concepts psychology uses for the harmful side of human experience have expanded their meanings, so that they now cover a much broader range of phenomena than they used to. Some of that expansion is horizontal, meaning the words reach new kinds of behavior. Some is vertical, meaning the same words now reach far milder versions of the old thing.

The result is that clinical vocabulary has outgrown the clinic. Words that once described severe pathology now describe a bad Tuesday.

“Boundary” is the clearest case. It was a word about professional conduct before it was a word about your boyfriend. In its clinical life it described the line a therapist does not cross with a patient, and later the line a family system draws around itself. Both of those are about a professional’s behavior toward someone in their care. Neither is a tool for telling a partner what he is allowed to do on a Thursday.

Haslam is careful in a way that gets lost when this work is quoted. He does not argue that the expansion was a mistake. He notes that conceptual change is inevitable and often well motivated, and he also warns that it risks pathologizing everyday experience. Both halves are his. Hold onto the second half and the first one keeps this honest.

When we reach for this language we are usually trying to be healthy. The paper that this article leans on is explicit that normalizing the vocabulary of mental health has real benefits, including resisting the tendency to dismiss people who are struggling and raising general awareness of mental health problems. That upside is genuine, and it is why the words spread.

In a 2025 paper called “Unmasking therapy-speak,” Cristina Isern-Mas and Manuel Almagro argue that the same language is often turned into a weapon, and that the turning relies on three main ingredients. Their words, not mine: the appeal to a certain epistemic authority, the confusion of descriptive and evaluative uses of language, and the linguistic strategy of deniability.

The appeal to authority. Clinical words arrive carrying the prestige of medicine. Say a thing in the register of a diagnosis and it stops sounding like your opinion and starts sounding like a finding. The paper points out that this is not just tone. It is a claim on credibility that the speaker has not earned and the partner cannot easily contest, because contesting it makes you look like you are against mental health.

The confusion of description and evaluation. You describe a feeling, and then use the description to issue a rule. “I have trauma around yelling” is a description of you. “Therefore you may never raise your voice, in any tone, ever” is a rule about him. The clinical vocabulary lets the second sentence travel inside the first one, unexamined.

Deniability. Because the words sound healthy, the speaker can always retreat. “I was only expressing a need.” “I was protecting my peace.” The move is unanswerable while it is happening, because objecting to it requires arguing with someone who has already framed themselves as the one doing the emotionally mature thing.

Notice that none of this requires a bad person. It requires fluency. That is what makes it so common in couples who have read the same books.

The vocabulary is too imprecise to settle this by labels, so settle it by grammar. What follows is a definitional argument, not a scientific finding. It holds up because of how the sentences are built, and you can check it in about four seconds on any sentence you like.

Three different speech acts get collapsed into one word. Here is what they actually are.

So the test is this: who is the subject of the sentence, and what happens if the other person just says no?

A boundary survives a no. It was never waiting on his answer. A demand dressed as a boundary collapses into an accusation the moment it meets resistance.

One thing this test does not license. A real limit your partner has stated is not up for debate, and you do not get to run this test on his sentence in order to keep doing what he has asked you to stop. The test is for auditing your own language, not for cross-examining his.

I am not a therapist. I read the research because I want to understand my own marriage, and I am writing this as a man who has been married a long time and has read more than his share of the literature, not as anyone who can diagnose you or your partner.

Which is why I have to put myself first here. I have used the word “boundary” to end a conversation I was losing. At the time it felt like self-respect. It was not. It was a way to stop being asked a question I did not want to answer, with a word that made stopping sound like growth.

The test is much easier to run on your own sentence than on your partner’s, which is exactly why almost nobody runs it that way. Run it on yours first. If the subject of your sentence turns out to be him, you have your answer, and it does not require him to be in the room.

When these words get used in a fight, what usually drops out is the requirement that made them mean anything. Each of these has a real meaning, and each one is wider in the wild than it is in the source.

Here is the part I would rather not have to write, because it would be a stronger article with a number in it.

Nobody has measured how often this happens. There is no couple-level survey, no diary study, no observational coding of how often clinical vocabulary gets used to end a negotiation, and no experiment on what it does to the outcome when it does. No base rate. No effect size. Isern-Mas and Almagro state plainly that despite the growing public interest in therapy-speak, its epistemic and ethical risks had not been examined in the scholarly literature at all.

A lack of measurement is not a lack of reality. Anyone who has been on the receiving end of that one sentence knows exactly what it does. But I would rather tell you the evidence is thin than hand you a statistic that does not exist, and you should be suspicious of any article about this that has one.

Since the research on this subject is thin, the honest move is to fall back on what is known about couple conflict generally, and to say what grade of evidence each thing carries.

Watch the first three minutes. This is the best-supported item on the list. In a study of 124 newlywed couples, researchers coded one marital conflict discussion, and were able to predict marital outcome over the following six years from the first 3 minutes of that one argument. It is prediction, not causation, from a single sample, but it points somewhere useful: how a conversation starts is not a preview of it. It is most of it.

Describe, do not adjudicate. This one rests on the logic of the paper above rather than on a trial. Instead of handing your partner a verdict on his character, describe the event and say what you will do. “You checked out at dinner and I felt alone at my own table” is a description. “You’re being toxic” is a ruling, and rulings get appealed, not answered.

Use plain-language phrasing, and know its ceiling. Nonviolent Communication is the version most people have heard of, and its evidence is more modest than its reputation. A 2026 systematic review of 31 studies found some support in some contexts, concluded that the picture points to a more modest pattern of potentially effective uses than is usually assumed, and noted that the research to date is almost entirely based on self-report. Many of those studies were run in prisons and hospitals. It is a reasonable tool. It is not a proven cure for how you talk to your husband.

Agree on a stop signal in advance. This one is a suggestion, not a finding. If you both decide, on a good day, on a word or a gesture that means “we are now fighting about the words instead of the problem,” you have somewhere to go that is not a clinical verdict. It works because it is mutual. Nobody gets to deploy it as a rule against the other.

What is not on this list is a better vocabulary. If the underlying move is a demand that sounds like self-care, learning more precise clinical terms will only make the demand harder to answer.

There is a line, and I want to name it clearly rather than let this article get used on the wrong side of it.

Using “boundary” to dodge an argument about the dishes is a vocabulary problem. Repeated, patterned conduct aimed at erasing your grip on reality, or at controlling where you go and who you see, is not a vocabulary problem. That is abuse, and Isern-Mas and Almagro say directly that therapy-speak can be used by people in positions of power, naming abusive partners among them. This article is about imprecise words. It is not a guide to surviving a controlling relationship.

If the words are being used to isolate you, to make you doubt your own memory, or to convince you that your perception of events is the problem, the answer is not a better turn of phrase. It is help from outside the relationship. In the US, the National Domestic Violence Hotline is free, confidential, and available around the clock, and you do not have to be sure you qualify to call.

The goal is not to win the fight. The goal is to stay in the conversation, which is a different and harder thing, because staying in it costs you the move that would have ended it.

So the next time you feel the pull of a clinical term to close a dispute, ask yourself one question. Is this a statement about my own behavior, or is it a rule for my partner?

If it is a rule, it is not a boundary. It is a demand. You are allowed to make demands. You are allowed to want things from the person you married, and saying so plainly is not a failure of emotional sophistication. Call it a demand, and notice what happens next. Sometimes the argument starts moving again the moment you stop pretending your demand is a medical necessity, because your partner can finally respond to the thing you actually want instead of defending himself against a diagnosis.