Disorganized Attachment in Relationships: The Push, the Pull and the Repair

Readers meet two names for this and reasonably assume they are reading about two things. Disorganized attachment is the older term, from child development research, where Mary Main and Judith Solomon used it in 1990 for infants whose behavior toward a caregiver looked contradictory in an observed separation and reunion procedure. Fearful-avoidant is the adult term, from Kim Bartholomew's work, describing someone who wants closeness and expects that depending on another person will cost them.

Everyday writing treats the two as interchangeable. The research does not, and Duschinsky's 2018 clarification paper says plainly that it remains unclear whether or how the adult fearful attachment style and infant disorganization relate to each other, conceptually or empirically. Different methods, different ages, different questions. An adult questionnaire result cannot tell you how you behaved at eighteen months.

For the rest of this page, the pattern being described is the adult one: high on both attachment dimensions at once. It is the only one of the four adult styles where both the pull toward someone and the pull away from them run at full strength, which is why it behaves so differently in a relationship from the steady, low-turbulence distance of the dismissive-avoidant pattern.

Where the dismissive-avoidant pattern starts slow, this one often starts very fast. Long first conversations, early disclosure of things that usually take months, a sense within weeks that this person understands you in a way nobody has. The intensity is genuine and it is also doing a job: speed shortens the uncertain stretch where you do not yet know whether you matter to someone.

The cost arrives on the other side of it. Having told someone a great deal very early, there is now a person who holds something, and the part of the system that treats closeness as a risk wakes up to that fact. People describe the days after an unusually open evening as a kind of exposure hangover: a wish to take it back, an urge to find something wrong with the other person, a sudden interest in whether this was a mistake.

A partner has no way to read that. From their side, the person who wanted to talk until four in the morning on Tuesday is distant and formal on Saturday, with no event in between. They search for what they did. Usually they did nothing. The intensity itself was the trigger.

If this is your pattern, deliberately slowing the opening is one of the few changes that alters the whole arc. Not withholding, which is a different move and tends to read as mystery, but spacing. One real thing at a time, with a week in between, lets safety build at a rate the rest of the system can keep up with.

Needing something is the dangerous position in this pattern, and not for the reason the dismissive-avoidant pattern finds it dangerous. There, dependence is unnecessary and slightly undignified. Here, dependence is wanted badly and expected to end in being let down, so asking carries a real risk of confirming the worst thing you believe about relationships.

The behavior that comes out of that is often an ask that gets retracted. Someone says they are having a terrible week, a partner clears the evening, and by six o'clock the message arrives saying it is fine, do not change anything, they are better now. The retraction is not politeness. It is the cheaper option: a withdrawn request cannot be refused.

Partners frequently learn to stop offering, which is the worst available outcome for everyone. If you recognize the retraction, the practical move is to hold the ask for one hour longer than feels comfortable before you take it back, and to tell your partner this is a thing you do. I sometimes take back a request about an hour after I make it, and when I do, it usually means I still want it.

There is an honest caution here too. A pattern explains a behavior; it does not excuse its effects. Repeated withdrawn requests, repeated silence after an opening, and repeated changes of direction are genuinely hard to live with, and the person doing them is the one who has to work on them.

Conflict in this pattern tends to escalate on the inside faster than it does on the outside. Heart rate rises, the room starts to feel smaller, and the capacity to hold both sides of the argument disappears early. That is the flooding part, and it is the same thing a dismissive-avoidant partner experiences before leaving the room.

What follows is different. The dismissive-avoidant exit is usually followed by relief and a lower temperature. Here the exit is followed by the second fear: that leaving the argument has ended the relationship. So the evening after a conflict can contain both an absolute refusal to talk and a string of messages at midnight, and both are genuine.

This is the part partners find hardest to describe to anyone else. They are not dealing with someone who went cold. They are dealing with two opposite signals arriving within hours, and the natural response, asking which one is real, has no good answer because both are.

The practical structure that helps is a pause with a fixed, short return. Twenty minutes, not two days, because a long pause in this pattern gives the fear room to grow into a conclusion. Agreeing the length in advance, when nobody is upset, works much better than negotiating it in the middle of an argument.

A week where nothing goes wrong is not neutral in this pattern. Warmth that has nowhere obvious to come from often arrives as a question rather than a comfort: what is this for, what does it cost, what is about to happen. The type page describes this as difficulty trusting positive signals, and inside a relationship it turns into specific behavior.

The most common version is a small test. Raising something with more edge than it needs, going quiet to see whether anyone comes, mentioning an old relationship to watch for a reaction. The test is rarely planned. It is closer to pressing a bruise to find out whether it still hurts, and the unbearable part is that passing it does not settle anything for long.

The second version is pre-emptive withdrawal: deciding privately that it will end anyway, and starting the ending early in order to be the one who does it. People describe this with the phrase leaving before I get left, and it can look from outside like a clean, considered decision.

The counter to both is not reassurance, which this pattern absorbs and discounts. It is naming the test out loud while it is happening, or afterward if that is all you can manage. I think I just picked a fight to find out whether you would stay. That sentence gives a partner a way to respond to the real thing instead of the fight, and it takes the move out of the dark.

Start with the body rather than the insight. Most people with this pattern already understand it intellectually and still cannot think their way out of a flooded moment. Slowing the breath, putting both feet on the floor, naming three things in the room: these are not a cure, they buy the ninety seconds in which a different choice becomes possible.

Put a delay on relationship-ending decisions. The urge to end it arrives at full confidence and tends to arrive after closeness rather than after a problem. A standing rule that no decision about the relationship gets acted on for twenty-four hours removes very little that was worth keeping and prevents a lot of avoidable damage.

Space the disclosures on purpose. Intensity feels like proof of connection in this pattern and is not a reliable indicator of it. Building slowly is the uncomfortable version and the one that tends to hold.

Say the pattern in advance, in plain words, to someone you trust. I want to be close to you and part of me will want to run, and when that happens I will tell you rather than disappear. A partner can work with that. They cannot work with the behavior alone.

Keep one thing that is yours outside the relationship. Where the anxious pattern tends to collapse everything into the bond and the dismissive pattern tends to keep too much outside it, this pattern does both in turn. A friendship, a piece of work or a regular commitment that does not move when the relationship moves is a practical stabilizer.

Consistency is worth more than intensity here, and it is worth more than patience as an attitude. Showing up in the same way on the good week and the difficult one, keeping small commitments exactly, and being reachable in a predictable rhythm all do more than any single reassuring conversation.

Do not chase harder when pushed away, and do not disappear when pulled in. Both responses confirm the expectation. The steady middle, I am here, I am not going to pressure you, let me know when you want to talk, and then actually being there when they do, is the thing that slowly contradicts the old model.

Ask for the content rather than the motive. I noticed we stopped talking after Tuesday, are you willing to come back to it, is answerable. Why do you always do this is not, and in this pattern it tends to land as proof of the thing already feared.

Keep your own limits in plain view. Being on the receiving end of push and pull is genuinely hard, and partners often last longer than they should by redefining what is tolerable. Decide what you need in ordinary terms, say it out loud, and keep track of whether anything moves. The one exception to all of this: if there is intimidation, coercion or control, that is not a pattern to be patient with, and safety comes before understanding.

Repair here has an extra step at the front, and skipping it is why so many attempts fail. Before any content, the relationship itself has to be put back on the table: we are okay, I am not leaving, this is not the end of us. Until that is said, the other half of the conversation cannot be heard, because the fear in the room is not about the subject of the argument.

Then name the move without defending it. I went silent for two days. Then one true sentence about what was happening: it got close and I panicked, or I was sure you were about to end it so I started ending it first. Then the original issue, which is still unresolved and still matters.

Then the next time, agreed in advance and in specifics. If I go quiet, I will text you the word pause and I will come back the next evening. Having a word turns the worst recurring event in the relationship into a known procedure, which is a far more realistic goal than never going quiet again.

In her 1990 paper, Bartholomew drew the distinction this whole page rests on: fearful avoidance, where contact is wanted but inhibited by what is feared, against dismissing avoidance, where the need for contact is played down. That is a theoretical distinction with a long life, and it is more useful than treating every withdrawal as a loss of interest.

On outcomes, Li and Chan's 2012 meta-analytic review of 73 studies found both attachment anxiety and avoidance associated with poorer relationship quality, with avoidance tied more closely to lower satisfaction and connectedness and anxiety tied more closely to conflict. A pattern high on both sits in the uncomfortable position of carrying both associations. These are population averages and correlations, not forecasts for anyone's relationship.

Change is real and unglamorous. Arriaga and colleagues' 2018 review describes security growing from repeated ordinary situations that contradict the old expectation, rather than from insight or from one good conversation. For this pattern that means many small instances of closeness that is not followed by harm, and many small instances of conflict that is not followed by an ending.

Psychology.com's Attachment Style Test is a self-report tool for reflection. A fearful-avoidant result is not a diagnosis, it cannot establish that anything happened to you, and it does not predict whether a relationship will last.

Treat this as a decision rather than a test anyone can score. The signs that usually mean outside help is the efficient next step: the push and pull cycle is now running weekly rather than occasionally, conflicts are ending in a decision to leave that gets reversed the next day, you are flooded so fast that nothing you have read is reachable in the moment, or the relationship has become the only place your sense of safety lives.

Individual work suits the regulation problem, because learning to notice flooding at the two-minute mark instead of the twenty is a skill rather than an insight. Work with both partners present suits the cycle, because by the time push and pull is established, the two of you are running it together. Some people with this pattern have a history that a trauma-informed therapist is better placed to work with, and that is worth asking about directly when choosing someone. None of this requires a label, and a questionnaire result is not a diagnosis of anything.

If you are the partner rather than the person with the pattern, your own support counts too. Living inside a cycle you did not create is tiring, and having one place to think about it that is not the relationship itself tends to make you steadier inside it.

The free Attachment Style Test asks how you usually think, feel and act in close relationships, using items in the spirit of the widely used Experiences in Close Relationships framework, and names the pattern your answers lean toward with a plain-language explanation. An optional in-depth report is available afterward for a closer look at where you fall on both dimensions, which is the useful number here, since this pattern is defined by being high on both rather than by a single score.

If you pursue closeness without the pull to retreat from it, the anxious-preoccupied style is the closer description, and the reassurance that helps there does not work the same way here. If the distance is steady and arrives without much fear of losing the person, the dismissive-avoidant pattern fits better. Reading the secure pattern is also worth the time, not as a standard to measure yourself against, but because it describes what asking, pausing and returning look like when none of those moves carries a threat.