For clinicians

Coercive Control Is Not an Attachment Problem

By Dr. Donetta D. Quinones, PhD, LMHC, LPC  ·  6 min read

Every framework has a boundary, and the honest ones say where it is.

Everything in attachment-informed couples work describes two people in distress, each running a protective strategy learned somewhere else, each contributing to a loop neither of them built alone. It is a good description of a great many relationships. It is a false description of a relationship in which one person is running a regime and the other is managing it — and the falseness is not academic. Delivered to a controlled partner in the presence of the person controlling them, it does harm.

The distinction

In a distressed relationship, two protectors are firing and both people are losing. In a controlled one, one person’s behaviour is organised around constraining the other, and the other’s behaviour is organised around managing the risk of that.

The behaviours can look similar across a session or two. Both couples argue. Both may report escalation.

What separates them is fear, and the direction it runs.

The concept that carries the weight here is not incident frequency but entrapment: the cumulative restriction of a person’s liberty through monitoring, isolation, rules, and threat. Counting incidents misses it entirely, which is why a couple can present with no reported violence and a situation that is nonetheless dangerous.

Why the framing itself is the hazard

Consider what a standard, well-intentioned cycle formulation says to a controlled partner:

"You each contribute to this. When you pull back, he pursues; when he pursues, you pull back further.

Neither of you started it."

To a distressed couple this is liberating and true. To a controlled woman it is a clinician, in a position of authority, telling her partner in front of her that her attempts to protect herself are half the problem. She will not correct you. She will use what you said to recalibrate what is safe to say next week, and she will be right to.

This is the specific reason attachment framing must be gated behind screening rather than applied first and revised later. The revision comes too late.

Screening that actually works

Four properties, and all four are load-bearing.

• Individual. Not a joint conversation with a question about safety in it. Separate rooms.

• Routine. Part of your intake for everyone, stated as such before you separate them: I meet with each partner on their own at the start. I do it with everybody — it is not about anything either of you has said.

• Universal. Asked of every couple, in the same words, regardless of presentation. Screening only when something worries you tells the person in front of you that something worried you.

• Repeated. A first screen is a snapshot. Disclosure follows trust, and trust takes weeks. Re-screen around week six, individually, framed as routine.

State the confidentiality limit before you ask, not after: what you tell me in here, I will not bring into the joint room without asking you first. The exceptions are the usual ones — I would rather say that now than after you have told me something.

Then ask behaviourally, in order, without commentary between items, and do not react. Fear. Monitoring of phone, location, money, movements. Restriction of contact with family and friends. Threats — leaving, money, the children, immigration status, disclosure. Sexual coercion. Physical contact in anger, including blocking a door, taking keys, driving dangerously, punching a wall. Weapons in the home. Strangulation, ever, asked directly. Then the open question: is there anything I have not asked about that I should have?

When something is disclosed

The first ninety seconds determine whether the rest comes out.

• Receive it flatly. Thank you for telling me. I’m glad you did.

• Do not name it, categorise it, or tell her what it is. She knows more about her situation than you do.

• Ask the question that matters most: what would make it worse for you, if I got this wrong?

• Then: what do you want to happen next? I’ll tell you what I can and can’t do, but I want to know what you want first.

The sentence never to say. “You need to leave.” It is the most dangerous sentence in this work.

Separation is the period of highest risk, and it is a plan she has not made. Your alarm is not a plan.

Follow her answer, not your alarm.

Disposition

If control is indicated: stop conjoint work. Do not name it to the couple jointly — that discloses her disclosure. Do not prescribe contact and do not prescribe forgiveness; neither is yours to schedule, and both are handed out routinely as though they were interventions. Document what was asked and what was decided, including your reasoning, not only your conclusion.

If it is unclear: treat it as control until it is not, and re-screen before the next joint session. The asymmetry of the errors is not close.

On the evidence

Typologies distinguishing coercive controlling violence from situational couple violence are well developed and clinically load-bearing. They are also contested at the margins and hard to measure reliably, and any clinician who tells you the classification is clean is overselling it.

That uncertainty does not soften the clinical rule. It sharpens it: because you cannot classify reliably from the joint room, the individual screen is not a formality you complete before the real work. It is the thing that makes the real work permissible.

What to do with the joint session you have already booked

A practical question clinicians actually face: you have screened, something is indicated, and the couple is coming in on Thursday expecting conjoint work.

• Do not cancel abruptly. An unexplained ending is itself a piece of information, and the controlling partner will interpret it.

• Have a reason that is true and non-specific. Sequencing is usually available: before we do joint work I want a couple of individual sessions with each of you. It is honest, it applies to everyone, and it is unremarkable.

• Do not create a secret you cannot hold. Never imply to the couple jointly that you know something.

Never let a joint session become a place where she has to perform normality while you both know.

• Get consultation the same week. Not because you cannot manage it, but because this is the category of case where a second clinician’s read is worth more than your own confidence.

On your own reaction

Two things happen to clinicians in these cases and both are worth naming in advance.

The first is urgency. You want her out. The urgency is a decent impulse and a poor plan — the risk arithmetic of separation is not intuitive, and she is the one who has been assessing it, continuously, for years. Your job is to be a durable resource rather than a decisive one.

The second is doubt. Controlled situations frequently present with a partner who minimises, retracts, and returns. That is a feature of entrapment, not evidence that you were wrong. Retraction is one of the most common sequelae of disclosure, and a clinician who treats it as a correction has closed the only door that was open.

From the book: “Who Are You Fighting With?” treats this across three chapters — the only ones in the book described as not optional — with the full protocol, a disposition table, and the wording above set out verbatim. The companion workbook contains it as a reproducible screening record.

Approx. 1182 words in the body above.