What is avoidant attachment?

Why Does Just Open Up Fail So Reliably?

Because it treats a trained suppression as though it were a missing skill.

The standard advice is familiar to anybody who has looked this up before: open up more, communicate your needs, let people in, practice vulnerability. All of it assumes the person does not know how, and would do it if they had a better script or a bit more courage.

Deactivation sits below where that advice can reach. The numbing runs first and the conscious choice comes afterward, if it comes at all. By the time somebody is aware enough to apply the advice, the need has usually already been filtered out of the picture, so there is nothing available to communicate. Ask a person in this pattern what they need in the middle of a difficult evening and the honest answer is often a genuine blank, which then gets interpreted by everybody present as stonewalling.

There is a second problem, and it is worse. Telling somebody to open up asks them to override the exact protective response that a real, earlier bid for closeness taught them to build. From inside the system, the advice does not read as growth. It reads as an instruction to make the mistake that hurt last time, again, deliberately, at a moment when nothing about the current situation has been established as safer.

And there is a third. Because deactivation includes minimizing the emotional stakes of relationships in general, a large number of people in this pattern do not experience themselves as being in any pain. They feel fine. Mildly disengaged, perhaps, a bit flat, a bit tired of being told they are distant. So advice about communicating more arrives as generic relationship chatter addressed to somebody else entirely, and gets filed accordingly.

Consider what this looks like in practice. Someone praised their whole life for being so independent, never any trouble, ends up seeking help not because a relationship is failing but because they cannot understand why they felt almost nothing at their own wedding. The numbness there is the same system that made them no trouble at seven, still competently doing its job. Nothing shifted quickly. What helped was lowering the target from feeling present at a wedding to noticing one moment in a day where they were not performing being fine.

Is This Avoidant Attachment or Genuine Independence?

Worth checking carefully, because several very different situations produce identical-looking weeks, and the right response to each one differs completely.

How much solitude somebody wants tells you very little on its own, since plenty of securely attached people want a great deal of time alone and always have. What tells you something is what happens to that preference under pressure. Does the pulling away scale specifically with how close somebody is getting? Does it show up right when the relationship starts requiring something real, reliance, disclosure, a shared plan with a date on it? And does it reverse once things cool back to a safer distance? Withdrawal that tracks intimacy is a defense doing its job. Withdrawal that stays flat across every season of a relationship is a temperament.

Introversion is the version of that question people ask most, and it sits on a different axis rather than being a milder degree of the same one. Somebody low on social energy wants fewer hours of company and recovers alone, and inside those fewer hours can be entirely willing to depend on a person, say what is wrong, and let a friend see the unflattering part. Avoidance is about the depending. So the two can live in one person without either one accounting for the other, and separating them means you stop counting how much solitude you want and start marking each pullback against what happened in the two days before it. One quiet month will not settle it. A dozen entries with dates will.

The first row is the one that hides longest, because both versions produce an identical record of competence and get praised identically by everybody watching. Competence itself tells you nothing here. Only the internal cost of the exception does.

Sit with this table and pick the row matching the last three months, rather than the row matching the worst evening. Most people reading a page like this have already run the argument in their head forty times without ever pinning it to a timeline.

Where Does It Hide When You Would Not Call It Avoidance?

In seven ordinary arenas, and most people recognize themselves in only two or three of them at first.

The second row deserves its own paragraph, because it is the one people find hardest to see in themselves. Being needed is safe. Needing is not. That asymmetry gets experienced as generosity, and it functions as a one-way valve installed a long time ago. Somebody can be calm, capable, and genuinely excellent in an emergency for twenty years while never once making the reverse call, and never find it strange until a friend asks the blunt version of the question: when was the last time you asked somebody for something and let them actually give it to you? The absence of an answer is the thing worth working on, rather than the crisis management skill.

The third row is worth watching too, in its physical forms. Frequent redecorating, moving flats, changing cities, restarting from a clean slate. A fresh environment offers the feeling of a new beginning at no emotional cost, which makes it an extremely efficient substitute for the beginning that would actually change something.

What Is the Difference Between Anxious and Avoidant Attachment?

Direction. Both patterns start from the same failed bid for closeness and then take opposite exits, one turning the signal up until somebody responds and one turning it down until it stops going out at all. The question underneath is identical in both cases, whether a need said out loud will actually be met, and the two answers run in opposite directions from there.

What separates them day to day is what sets each one off. An anxious system keeps its alarm on distance, so the expensive hours are the ones after a message goes unanswered or a partner goes quiet, and the move is to reach, ask, check, and close the gap. An avoidant system keeps its alarm on closeness, so the expensive hours are the ones after a good weekend, a first trip, or a sentence somebody meant, and the move is to get busy, cool the contact down, and produce a reason. If you want to place yourself, skip the question of how much you love anybody and ask which week costs you more, the one where somebody pulled back or the one where somebody moved in.

Neither end is the healthier one, and a good number of people run both, in different rooms and in different decades. The anxious side of this comparison is set out in full on its own page, written from the chair of the person doing the reaching.

What Is Dismissive Avoidant Attachment, and How Is It Different From the Fearful Version?

Dismissive avoidant attachment, or the dismissive avoidant attachment style, is the phrase most people type when they come looking for this, and it names the pattern this page has already spent every section describing: deactivation, the distance that arrives after things have gone well, the reliable turning down of a signal that asks for closeness. The term describes a strategy somebody uses with closeness rather than a kind of person, and what marks this version off from the fearful one is how steadily that strategy runs.

Consistency. Dismissive avoidance stays regulated by deactivating reliably, every time. The fearful version oscillates between the anxious and avoidant strategies inside the same relationship, sometimes inside the same week.

Somebody in the fearful pattern wants closeness genuinely, moves toward it, gets it, and then flees from the thing they just asked for. Both get called avoidant in ordinary conversation, and the fearful version sits closer to a blend of the two strategies than to pure deactivation.

The reason this distinction earns its place here is what people conclude about themselves without it. The oscillation gets read as a personality defect. I'm just inconsistent. I don't know what I want. I ruin things. Seeing it as a second, distinct pattern with two engines running at once, rather than as evidence of a broken character, changes what a person does next. It also explains the experience of wanting connection and panicking the moment it arrives, which is two accurate systems reporting at the same time rather than one confused one.

The four-category split has an author and a date. Bartholomew and Horowitz proposed it in 1991, cutting the avoidant end into a dismissing version that holds closeness away and reports being comfortable with the arrangement, and a fearful version that wants closeness and expects to be hurt by it, with the limit that the four boxes arrived as a proposed scheme checked afterward against interviews and questionnaires in samples of young adults, which leaves a great many people sitting between two boxes rather than inside one.

Two more words attach to that same oscillating pattern, and one of them is on far heavier rotation than the other. Fearful avoidant attachment is the term from the four-category model already described above, and it is used accurately most of the time. Disorganized attachment is the one people actually search for, along with the same word spelled with an s in place of the z, the British form. So is anxious avoidant, a fourth name doing double duty, since the infant research used it for the plain avoidant type this page has been describing all along while popular writing now attaches it to the oscillating one. In everyday use the four get treated as synonyms. They did not arrive from the same place, and where the second one came from decides how much weight it can carry.

Disorganized began as a description of infants. The separation-and-reunion procedure that produced the original three classifications had trained observers watching through one-way glass while a caregiver left a room and later came back. Most of what they saw sorted into three organized strategies, the avoidant one among them, which is where the pattern on this page starts. A fourth group did not sort. Going back through the recordings years afterward, coders found children who froze partway across the floor, or moved toward a parent with the head turned away, or produced an approach and a retreat inside the same second. The word picked for that group named the absence of any single organized strategy across a few observed minutes. It described what a small child did in a room on one afternoon. Nobody at the time was handing it to adults.

The popular adult usage takes that word and puts it on something built quite differently. One instrument is a trained coder rating observed behavior from a recording. The other is a person answering questions about themselves on a form. The two were developed for separate purposes and checked in separate ways, and where researchers have given the same people both a coder-rated measure from the developmental tradition and a self-report questionnaire, the agreement between them has come out close to nothing, which is what Roisman and colleagues found in 2007, running both kinds of measure on the same participants. That result is missing from nearly every popular retelling of this vocabulary. Both measures still describe something real inside their own territory. The narrower and more useful conclusion is that a word carried across from one of them to the other arrives carrying a claim the evidence underneath it does not cover.

There is a cleaner way to hold all four names at once, and it removes most of the damage the boxes do. Adult attachment questionnaires are generally built on two continuous dimensions rather than on four types: roughly, how much anxiety somebody carries about where they stand with people, and how much they avoid depending on anyone. Secure sits low on both. The pattern this page describes sits low on the first and high on the second. The anxious or preoccupied pattern sits high on the first and low on the second. Fearful sits high on both, which is the two engines described a moment ago, drawn as a location instead of as a contradiction. That is a described model, assembled by several research groups across decades rather than produced by one researcher, which is why no name is attached to it here and why there is no single study to send you to. What it gives you is a position on two sliders instead of membership in a box, and sliders move.

Now the limits, stated plainly, because almost nothing written about these four words states them. Attachment style is measured as a tendency, never as a fixed property, and it drifts in three directions at once. It drifts between relationships, so the same person can score one way about a partner and another way about a parent, which is the effect described further up where somebody runs anxious with a mother and avoidant with a partner. It drifts across years. Where the same people have been followed from infancy into adulthood, an early classification carries real predictive weight and later experience keeps overwriting it, which is a considerably softer claim than the one in circulation, that a childhood category accounts for an adult relationship. And it drifts with the instrument, since two questionnaires asking slightly different questions will place the same person in slightly different spots. Reanalyses of the original infant data have gone further and suggested the underlying variation may be continuous rather than made of discrete types at all, which would leave all four of these names as convenient regions on a map rather than things a person has. None of that makes the description useless, and this page would not exist if it did. It does mean that a category collected once, from one form, on one afternoon, is a snapshot rather than a verdict, and that what is circulating under these words is a research vocabulary rather than a clinical category, however confident the account you met it in sounded.

Which leaves the division this page has been making since the top. The four names describe what somebody does with closeness and how much of each strategy is currently showing. None of them says why yours in particular is set where it is. Two people can land in the identical corner out of an early absence and a later ending with nothing in common between them, and the description will read as accurate to both while pointing at completely different work. The end of the same two dimensions where both of them run low has its own page. The word names the position. The reason sits back in the section on where this comes from, and the reason is the part that has to move before the position does.

Held side by side, the three names come apart like this.

Read the middle column as what somebody is doing with closeness at present, which is the limit the rest of this section has already described.

What Does the Secure End Actually Look Like?

Concrete enough that you could film it, and considerably less impressive than most people expect.

Every one of these patterns is a polarity with two ends, and describing only the wounded end leaves a reader with an accurate account of their problem and nowhere to walk. So here is the other end, translated out of abstraction and into an ordinary Tuesday.

Read the right-hand column carefully, because it is deliberately unremarkable. Nobody in it is confessing anything. There are no overdue conversations and no speeches. The available end of this pattern is mostly a person who can let one fact about their life become evidence that somebody matters to them.

Don't judge yourself if that column reads as slightly out of reach today. The distance between the two columns took a childhood and then two decades of daily rehearsal to build, and nobody crosses it in a good two weeks.

What Happens to Your Relationships While You Practice This?

They usually get briefly stranger before they get more honest, which is worth knowing in advance.

Some of the distance was load-bearing for both people. A partner who spent three years learning not to ask you anything directly may not immediately notice that you have started answering, and may test it a few times in ways that feel clumsy. A friend used to a certain register may find a true sentence from you unsettling. Saying out loud what you are practicing handles most of that.

If you are with somebody who moves toward you when you move away, expect something counterintuitive. Staying present for one extra hour frequently produces less pursuit from them, sometimes within days, because the thing they were reaching for is no longer receding. That is disorienting, and it is also useful information about what the loop was actually made of.

The honest limitation belongs here too. A small number of arrangements were genuinely built on your unavailability, and those get renegotiated or they end. Friendships usually improve first, since the stakes are lower and the repetitions are easier to run. The parent relationship moves slowest, which is normal, because it has been running the longest. Start smaller there than anywhere else.

How Long Does This Take, Honestly?

Longer than a week, and the honest answer runs at two different speeds.

Recognition arrives quickly. Most people can catch a deactivating move mid-flight within a few weeks of deliberate attention: seeing the flaw assemble itself, watching the busy two weeks get scheduled, feeling the chest tighten a second before the not yet leaves your mouth. Catching it counts even on the days you obey it anyway.

The reflex takes far longer, months to years rather than weeks, because it was rehearsed daily for decades and it fades at the pace of accumulated experience rather than understanding. Reading this page changes remarkably little on its own. Fifty small needs voiced and met, in rooms where almost nothing was at stake, change a great deal. Expect a step pattern with long flat stretches and a bad month after any real ending.

I spent five years and more than five thousand hours doing this kind of work on myself before I built anything around it, and more than fifty thousand dollars on seminars, therapy, books, and teachers, some found in the jungles of Vanuatu, some in the jungles of Peru. The thing I underestimated longest was how much of this is capacity rather than character. People who look self-sufficient are usually people who never got fifty low-stakes repetitions at needing somebody and having it go fine. Where it sits for me today is worth saying plainly: the reflex to handle it alone has quietened a lot over the years, and it has not gone to zero, and I would rather tell you that than sell a finished version of something nobody finishes.

Where Does This Thinking Come From?

From attachment research that has been public for decades. Five strands carry most of the weight on this page, and every one of them can be checked without going through me.

John Bowlby, a British psychiatrist, set out the idea that closeness runs as a behavioral system with its own signals, across three volumes published between 1969 and 1980. Mary Ainsworth made it observable, running the Strange Situation procedure in Baltimore through the 1960s and 70s and reporting it in Patterns of Attachment in 1978, and the avoidant classification described on this page starts there. Cindy Hazan and Phillip Shaver carried the vocabulary into adult romantic relationships in a 1987 paper, which is the reason anybody discusses attachment styles in couples at all, and Kim Bartholomew and Leonard Horowitz split the avoidant end into its dismissive and fearful versions in a four-category model published in 1991. The words deactivating and hyperactivating, used throughout this page, are laid out in Mario Mikulincer and Phillip Shaver's Attachment in Adulthood. The finding that somebody can report feeling nothing during a separation while their body says otherwise comes from Dozier and Kobak, whose 1992 study measured skin conductance climbing during attachment interviews in exactly the people who described the least distress.

The rest of it is mine and carries no research weight at all: the five years and five thousand hours described above, with no clinical qualification behind them. None of the researchers named has any connection to this page or to the seminar. Where their published work and my own experience disagree, theirs is the one to go and check.