In most couples who argue, there is one who gets visibly upset and one who goes quiet. Ask either of them who the dysregulated one is and you will usually get the same answer. Ask the friends. Ask the family. Everybody agrees.
The measurements do not.
What happens in the body of the person who stays calm
Mary Dozier and Roger Kobak put adults through a long interview about their childhoods while recording their skin conductance, which rises with sympathetic arousal. The people who talked about separation, rejection and being frightened by a parent in the most composed, unbothered way, who played the whole thing down, were the ones whose readings climbed most sharply on exactly those questions.
This is the adult version of something I wrote about last week. Infants who have learned to go quiet when a parent leaves look calmer than the ones who cry, and their stress hormones afterwards are higher.
Same shape, forty years apart. The composure is real as a behaviour. It is not a report on the inside.
The claim worth making, and the one that is not
It is tempting to go on from there to “actually the quiet one is the more dysregulated of the two”. I would rather not, because it is not quite honest. On visible disruption and on how bad they feel, the upset partner scores higher. Which of them is more aroused depends on what you choose to measure.
The claim that survives is narrower and more useful. Both of them are aroused. One of them can feel it.
Arousal and access to your own arousal are two different things, and they come apart. That is the whole of it.
Why this makes avoidance the harder pattern to shift
Because you cannot help somebody regulate a feeling they cannot detect.
Attachment anxiety announces itself. It arrives in the room already named, already uncomfortable, already motivating. There is something to work with from the first minute. Deactivation produces no such material. The person is not withholding it. They go looking and find nothing there, and then, often, they apologise for having nothing to say.
The research points the same way. Reviews of what changes during therapy find attachment security going up and attachment anxiety coming down fairly consistently, while the picture for avoidance stays unclear. A meta-analysis of thirty-six studies found that people who start therapy more securely attached do better, and that avoidance specifically predicts poorer outcomes in younger clients.
Why you may never have met one
Here is the part that took me a while to see properly. Avoidance is not the absence of attachment insecurity. It is attachment insecurity with a defence over it that is working.
And because the defence is working, nobody can see the insecurity. Not the partner. Not the friends. Not the person themselves. They are not concealing anything. They look inward and find, accurately, nothing much.
You can watch this break in a laboratory. Mikulincer, Dolev and Shaver asked people to push away thoughts of a relationship ending, and then gave them a second task to do at the same time. With a light load, the avoidant participants pushed the thoughts away completely and thought better of themselves than before. Add enough load, and the suppression failed, and what came up was a set of harsh beliefs about themselves that had not been available a moment earlier. The researchers call these the hidden vulnerabilities of avoidant people.
Real life supplies heavier loads than a laboratory can. Berant, Mikulincer and Shaver followed mothers of babies born with serious heart conditions for seven years, and it was the avoidant mothers whose mental health and marriages deteriorated most over that time, more than the anxious ones. Avoidance is not a strength that holds. It is a strategy that holds until it does not.
Which is when they arrive
So this is a pattern that mostly does not come to therapy on a Tuesday because it has been thinking things over. It comes after something has broken the defence. A bereavement. A marriage ending. Redundancy. Something that happened at work, or in a war, or in a hospital corridor.
And what the person describes is not really the event. It is the state they find themselves in, which is completely unfamiliar and frankly frightening, and they will almost always say some version of “I am not normally like this.”
That sentence gets heard as minimising. It is not. It is an exactly accurate report on how well something has been working for about thirty years.
And then the crisis recedes, and this is the part worth naming: the defence goes back up. The unfamiliar state subsides. The reason for coming disappears. And so, often, do they.
I want to be careful here, because that last bit is my own clinical impression and not something I can show you a study for. But if it is right, it changes what “avoidance is harder to treat” actually means. Not that the pattern resists change. That the window in which the person is reachable is narrow, and closes by itself, and closing is what it feels like to get better.
It would also mean that some of the unclear picture in the research is people leaving before anything could be measured, which is a different problem with a different answer.
What this does not excuse
Everything above explains why a person goes quiet. It says nothing at all about what the silence does to the person on the other end of it, and it must not be allowed to stand in for that.
In England and Wales, controlling or coercive behaviour in an intimate or family relationship is a criminal offence, and sustained silent treatment can form part of such a pattern. The law does not require an intention to harm. It asks whether the person knows, or ought to know, the effect their behaviour has. “I was trying to calm myself down” describes a mechanism. It is not a defence.
What separates the two is not how aroused somebody was. The physiology cannot tell you, because a person can be flooded and punishing at the same time. What separates them is what happens once the effect is known. Somebody whose withdrawal is genuinely a flooding response is usually appalled to learn what it does, and wants to do something about it. A pattern that carries on after its effect has been made plain is a different thing with a different name.
If you are reading this and recognising a pattern that frightens you rather than one that frustrates you, then that is the part to attend to, and it is worth talking to somebody outside the relationship about it.
What actually helps
Not insight, and not being told what you are. A person who cannot feel their own arousal is not helped by being handed the word for it.
The work is building the signal, and it starts lower down than people expect. Noticing that your jaw has set. That your breathing has gone shallow and high. That you have stopped hearing the other person properly. None of that requires you to know what you feel yet. It only requires you to notice that something in your body has changed, which is a smaller and much more learnable thing.
There is a distinction worth knowing about here, because it changes what the work is. For some people the signal exists and drops out under pressure, and the job is to make the pressure survivable so it stays online. For others it was never built in the first place. There is no faint version to recover, and it has to be constructed from the ground up, slowly. In my own caseload, among people carrying difficult childhoods, the second is the more common of the two. Being asked to “just notice what you are feeling” when nothing was ever built to notice with produces blankness, then compliance, then shame. It is worth knowing which problem you have before you start.
And do not take the label home with you
One more thing, because it matters more than it sounds. “Avoidant” is not a kind of person. The same human being withdraws in one relationship and pursues in another, and most of us do both depending on who we are with. If you plotted yourself on the grid in last week’s worksheet and came out in different places for different people, that was not an error. That is how it works.
A caveat that matters, in case you are reading this and it does not fit. What I have described, the insecurity that its owner cannot see, is the dismissing version. There is another kind of avoidance where a person very much can feel the insecurity and pulls away anyway, because closeness has been dangerous rather than merely disappointing. If that is you, nothing here is saying you ought to be feeling less. It is a different arrangement, and it needs writing about separately.
The pattern is something you learned, in particular conditions, with particular people. Which is also why it is something that can be learned differently.
Where to start
The grid worksheet is a reasonable first move if you want to see where you actually sit, relationship by relationship. The previous piece, on why insight alone does not shift any of this, sets out the mechanism. More on the patterns themselves is on the attachment styles page.
References
Dozier, M and Kobak, R R (1992). Psychophysiology in attachment interviews: converging evidence for deactivating strategies. Child Development, 63(6), 1473–1480.
Mikulincer, M, Dolev, T and Shaver, P R (2004). Attachment-related strategies during thought suppression: ironic rebounds and vulnerable self-representations. Journal of Personality and Social Psychology, 87(6), 940–956.
Berant, E, Mikulincer, M and Shaver, P R (2008). Mothers’ attachment style, their mental health, and their children’s emotional vulnerabilities: a 7-year study of children with congenital heart disease. Journal of Personality, 76(1), 31–65.
Spangler, G and Grossmann, K E (1993). Biobehavioral organization in securely and insecurely attached infants. Child Development, 64(5), 1439–1450.
Taylor, P, Rietzschel, J, Danquah, A and Berry, K (2015). Changes in attachment representations during psychological therapy. Psychotherapy Research, 25, 222–238.
Levy, K N, Kivity, Y, Johnson, B N and Gooch, C V (2018). Adult attachment as a predictor and moderator of psychotherapy outcome: a meta-analysis. Journal of Clinical Psychology, 74(11), 1996–2013.
Serious Crime Act 2015, section 76, as amended by the Domestic Abuse Act 2021.
