Attachment and parenting: how the pattern passes to a child

“I heard myself talking to him. That was my mother’s voice. The same words, the same tone. I swore I would never say that to my own child.”

That is what a mother of two preschoolers told me after one evening, tying shoelaces, she heard herself. She was not a bad mother. She was an exhausted mother who had, in a single sentence, met something she had carried for thirty years without knowing she was carrying it.

I wrote about how attachment patterns repeat in my piece on attachment patterns. This one is about something more specific and, for many people, more painful: how what you experienced with your own parents passes on — and, more importantly, why that transmission is not fate, and why the bar a child actually needs is considerably lower than you think.

What actually gets passed on

Let us clear up a misconception first. A child does not inherit “anxious attachment” from a parent as a trait. What is inherited is something smaller and far more concrete: what happens between the two of you when the child is in distress.

Every time a child feels distress and sends a signal — crying, arms outstretched, whingeing, or in an older child a slammed door — something happens in the seconds that follow. Someone comes or does not come. They arrive settled, or they arrive with their own panic. They recognise what is going on, or they read the signal as an attack. One such episode means nothing. A few thousand of them mean an expectation.

And it is precisely that expectation which is passed on: the child’s working assumption about whether signalling is worth it, and what happens if you do. Not the words, not the parenting principles, not whether the parents are still together. Responsiveness in moments of distress.

The loop that repeats a thousand times

This loop is simple, and its simplicity is what misleads. In daily life it never looks important — the child falls, comes to you, you pick them up, a minute later they are running again. Yet it is in that minute that the model is built.

When the response does not come — not once, but systematically — the child does not give up. The child adapts. One solution is to turn the signal up: cry louder, cry longer, cling, do not let go. The other is to turn it down: settle yourself, do not ask, do not complain, become “good”. Both solutions are intelligent in that environment. Both become a burden once the child grows up and the environment is no longer the same.

The prediction holds, but not completely

Now to the numbers that surprised me most as a student. Back in the nineties, researchers found that from an interview with an expectant mother — from how she narrates her own childhood — you could predict reasonably well what her child’s attachment would look like at one year old. The prediction beat chance and it replicated.

But it was never complete, and the difference even got its own name: the transmission gap. Part of the association remains unexplained. A large synthesis of three decades of research showed that the link between a parent’s narrative and a child’s attachment is weaker than first believed, and that the path in between — sensitive responsiveness — is only one part of the explanation.

That gap is not a statistical awkwardness. It means that between a parent’s own story and a child’s attachment stand things researchers have not fully captured: the child’s temperament, what was happening in the family at the time, how much support was around, and not least the other adults available to the child. Which is why the same mother is often not the same mother with her first and her third child — and that is explained by context, not by her childhood.

To me this gap is the most hopeful finding in the whole of attachment theory. It means there is no straight line between your childhood and your child’s. There is space in between. And in that space, work is possible.

“Good enough” is not a modest standard — it is a measured one

The most relieving finding from observational studies of mothers and infants is this: the match between the parent’s state and the child’s was, on the recordings, far from perfect. They were mismatched considerably more often than matched. And those children still developed secure attachment.

Security, then, does not come from flawless reading of a child. It comes from mismatches being repaired often enough. A child who never experienced a mismatch would never get the chance to learn that things can be put right — and it is precisely that experience they carry into every relationship afterwards.

If you have ever heard the phrase “good enough parent” and assumed it was consolation for those who fall short, it is not. It is a description of what actually worked on the recordings.

Repair matters more than perfection

In practice this means the single most useful thing you can do is this: when you hear your mother’s voice coming out of your mouth, come back half an hour later and say it. “Earlier I was angry and I shouted. That was not all right. It is not your fault that I was tired.”

What the child takes from that is not that you are perfect. They take three far more useful things: that adults get it wrong, that things get repaired, and that what they experienced was not imagined. The last is the most important — the greatest harm is done not by the outburst, but by an outburst that is afterwards denied.

There is a limit here too: repair is not pleading with your child, and it is not a request to be comforted. A child must never become the person who regulates how you feel.

Where your pattern shows up day to day

In parenting, an attachment pattern almost never shows in the calm moments. It shows when the child is in distress — because that is when your own system switches on.

If you sit more on the avoidant side, you will probably notice that your child’s crying repels you physically. Not because you do not care, but because you once learned that distress is something to be got rid of. A pull towards fixing appears: “it is nothing, look what a lovely toy you have”. The child gets comfort, but does not get company inside the feeling.

If you sit more on the anxious side, the reverse can happen: the child’s distress floods you more than it floods them. Goodbyes, separations, the first day at nursery become hardest for you, and the child reads that. Sometimes a quiet expectation appears that the child will confirm you are a good parent — and that is a weight which does not belong to them.

Neither description is an accusation. Both are simply information about where you will need more awareness, because it will not happen by itself.

Why this is not an accusation against parents

Now the part that matters most to me here. When we talk about intergenerational transmission, it quickly starts to sound as though we are hunting for a culprit one generation back. We are not.

A parent’s responsiveness does not depend only on their inner story. It depends on how much they sleep, whether there is money at the end of the month, whether anyone helps them, whether they are ill, whether they are in a relationship where they feel safe. A parent living in poverty or in a violent relationship is not less loving — they have less attention available, because it is occupied elsewhere. Your mother in the year everything fell apart was not the same mother as five years later.

So in my room I never make a list of what a person’s parents did wrong. We do something else: what you took from it, and what you do with that now. This is often where guilt shows up with nowhere to go — towards your parents, for “blaming them”, and towards your children, for “passing it on”. Neither guilt improves responsiveness. As a rule it makes it worse.

What actually helps

Research on parenting interventions has one counter-intuitive finding: short, focused programmes that work on concrete responsiveness achieve more than long, broad ones. Working on a parent noticing the child’s signal and answering it is more effective than explaining theory.

That shapes what I suggest to people. First, sleep and relief from load wherever it is at all possible — depletion does more damage to responsiveness than any unprocessed story. Then a single question, asked before you respond: “what is going on with them right now?”, not “how do I stop this?”. And third, the one that sounds least clinical and changes most: self-compassion towards yourself as a parent. A parent who tears themselves apart after a mistake has less capacity for the next round than one who admits it and moves on.

If you do choose therapy, you do not have to walk through the whole of your childhood. It is often enough to turn your story into a coherent narrative — not a pretty one, a coherent one. I described what such a conversation is like in my piece on preparing for a first session.

Where this explanation does not hold

Now let me argue against myself, which on this topic is essential. Attachment has become a fashionable explanation for almost everything that happens with a child, and that is dangerous.

Children are born with very different temperaments. A child who is more easily aroused from birth, harder to soothe and a poor sleeper is not evidence that you answered their signals badly. Nor does attachment explain autism, ADHD, learning difficulties or speech delay — and explaining those conditions by a mother’s responsiveness is an old, harmful and discredited idea.

The predictive power of attachment is also moderate. A securely attached child is not insured against difficulties, and an insecurely attached child is not condemned to them; most children with insecure attachment grow up without mental health problems. If you are worried about a specific behaviour — if your child is not sleeping, not developing as expected, or hurting themselves — the first port of call is the paediatrician, not a reflection on your own childhood.

And one more thing I hear often: if you already have a teenager, it is not too late. The working model consolidates with age, it does not close. Repair at fifteen works on the same principle as at three, only it takes longer and the first move has to come from you.

What to remember

A child does not inherit your attachment pattern; they inherit what happens between the two of you in their moments of distress. That loop repeats a thousand times, and the expectation grows out of it — not out of your parenting principles.

Transmission is real, but it is not a straight line: between your childhood and your child’s there is a gap, and in that gap there is room to change. Secure attachment does not require perfect attunement; it requires frequent enough repair after a mismatch, and enough resource to respond at all. And when you next hear your mother’s voice coming out of your mouth, that is not proof that you have passed everything on — it is a moment in which you can come back half an hour later and say that it was not all right.