Anxious and Avoidant Attachment, Explained

You keep doing the same thing in relationships. You know you are doing it. You watch yourself do it — clinging too hard, or going cold, or cycling between both — and still you cannot stop. That is not a character flaw. It is a pattern your nervous system built before you were old enough to choose differently. Attachment theory explains where that pattern came from, how it wired itself into your body, and what it takes to actually change it.

Where Attachment Theory Came From — and Why It Matters Now

John Bowlby was a British psychiatrist. In the 1950s and 60s, he did something unusual: he looked at what happened to children who were separated from their caregivers — in hospitals, orphanages, wartime evacuations. What he found was not what most psychologists expected.

Children did not simply grieve and adapt. They went through a predictable sequence: protest, despair, then detachment. And the quality of care they had received before the separation shaped how badly the separation hurt them, and how they related to people afterward (Bowlby, 1969).

Bowlby called the bond between child and caregiver the attachment system. He argued it is biological — a survival mechanism. Infants who stay close to a protective adult live. Infants who wander do not. The attachment system is evolution's answer to that problem. It motivates the child to seek proximity — closeness — whenever they feel threatened. And it learns, very quickly, whether the caregiver is reliable or not.

In the 1970s, Mary Ainsworth expanded on this work. She designed a study called the Strange Situation — a controlled sequence of brief separations and reunions between mothers and one-year-old infants. What she found changed psychology permanently. Infants did not all react the same way. Their reactions fell into distinct, recognisable patterns (Ainsworth, Blehar, Waters, & Wall, 1978).

Those patterns are what we now call attachment styles.

What Anxious Attachment Actually Is

Imagine a mother who is loving, but inconsistently so. Some days she is warm, attentive, present. Other days — because she is depressed, or overwhelmed, or simply distracted — she is not really there. The baby cannot predict which version of the mother is coming. So the baby learns to stay on high alert, to protest loudly when the mother leaves, to cling when she returns.

This is how anxious attachment starts. Not with cruelty. Often with love that is real but unreliable.

In Ainsworth's Strange Situation, these infants are easy to spot. When the mother leaves the room, they cry hard. When she returns, they do not settle quickly. They push her away and cling at the same time. They cannot use her as a safe base to go explore, because they never fully trust she will still be there when they turn back (Ainsworth et al., 1978).

In adulthood, this pattern looks different on the surface, but the logic underneath is the same. The person with anxious attachment craves closeness and worries constantly that it will disappear. They check their phone. They read into small signals. They need frequent reassurance — not because they are weak, but because their nervous system learned, very early, that connection is fragile and must be monitored.

This matters because anxious attachment is often mistaken for neediness, jealousy, or emotional immaturity. Those labels assign blame. The research tells a different story. The alarm is real. The body is doing exactly what it was trained to do.

There is a shadow side to this pattern that does not get talked about as much. Because anxiously attached people so often suppress their own needs to protect the relationship, they can develop what psychologists call a false self — a version of themselves built around what they think the other person needs them to be. This connects directly to the way Jung understood the persona, the mask we wear in relationship. You can read more about that dynamic in the article on authenticity versus performing the persona.

What Avoidant Attachment Actually Is

Now imagine a different kind of mother. She is present, but she is uncomfortable with strong emotion. When the baby cries, she soothes the crying — but in a slightly stiff, managed way. When the baby reaches for closeness, she responds, but keeps just a little distance. The message, delivered without words and without any conscious intention, is: manage yourself, you are okay, you do not need that much.

The baby learns. It stops sending the signal. Not because the need for closeness goes away — it does not — but because the need no longer seems to work. The child becomes self-sufficient earlier than is natural. They get good at regulating their own feelings alone.

In the Strange Situation, these infants look, on the surface, remarkably calm when the mother leaves. They explore the room. They do not seem distressed. Ainsworth initially thought this might mean they were well-adjusted. Then researchers measured their heart rate and cortisol (the stress hormone). It was elevated the whole time. The calm was not peace — it was suppression (Spangler & Grossmann, 1993).

This is the central fact about avoidant attachment: the distress is still there. It is just not expressed outward. Instead, it gets turned off — what researchers call deactivation of the attachment system. The nervous system learns to dial down the signal of need before it can be sent.

In adulthood, dismissive-avoidant people often appear very competent and self-contained. They are not cold — they are defended. They fall in love, but closeness creates a quiet discomfort they cannot always name. They want the relationship and simultaneously feel a pull to have more space, more distance, more time to themselves.

Under real stress, avoidant individuals often withdraw. Not because they do not care. Because their entire regulatory system is built around the idea that needing people leads to disappointment. The body learned, very early, to need less.

There is a fearful-avoidant pattern too — sometimes called disorganised — where the person both wants closeness and fears it at the same time. Often this comes from caregivers who were themselves a source of fear. The person who was supposed to provide safety was also frightening. This leaves the attachment system with nowhere to go (Main & Hesse, 1990). That pattern is worth its own full treatment; this article focuses primarily on the anxious and dismissive poles.

Why These Two Styles Find Each Other

Here is the thing that confuses most people: anxious and avoidant individuals pair up constantly. Not occasionally. Consistently. The research confirms what anyone who has lived it knows.

The reason is almost mechanical. The anxiously attached person interprets the avoidant's pull-back as the exact signal their nervous system has been trained to watch for — withdrawal, potential loss. Their alarm activates. They pursue. The avoidant person interprets the anxious person's pursuit as exactly the kind of engulfment their nervous system has been trained to protect against. They pull back further. Which escalates the anxious person's alarm. Which makes the avoidant retreat more.

This cycle can run for years. Both people are in genuine pain. Neither is behaving badly. Both are responding, in real time, to what their own nervous system learned was the most dangerous thing in the world.

Fraley and Brumbaugh (2004) studied how people with different attachment styles responded to romantic loss. Avoidant individuals showed less conscious distress but showed more intrusive thoughts when distracted — meaning the grief was still active, just being actively suppressed. Anxious individuals showed intense conscious distress but recovered their baseline faster once reassurance was available. These are not personality types in the way astrology signs are. They are regulatory strategies — different tools for managing the same underlying biological need.

Understanding this reframes the question. The question is not "why are they so clingy?" or "why are they so cold?" The question is: what did this person's nervous system learn was necessary for survival, and how is that learning playing out right now?

This is also, interestingly, where Jung's framework becomes useful. Jung understood the psyche as compensatory — the parts of ourselves that were not safe to express get pushed into what he called the shadow (see our article on the anima and animus in Jungian psychology for related dynamics). The avoidant person's need for closeness becomes shadow material — present but unacknowledged. The anxious person's capacity for self-containment gets buried under the urgency of maintaining connection. The work of individuation, for both, involves reclaiming what has been disowned.

What the Brain Is Actually Doing

Attachment is not just a psychological concept. It is a biological one. The patterns laid down in infancy show up in measurable ways in the adult brain.

The amygdala — a small, almond-shaped structure deep in the brain, responsible for detecting threats — responds differently in people with different attachment histories. Securely attached adults show less amygdala activation when they see images of interpersonal threat. Anxiously attached adults show more. The threat-detection system is calibrated differently (Lemche et al., 2006).

The prefrontal cortex — the part of the brain that handles reasoning, long-range planning, and the ability to step back from immediate emotion — is central here too. Under stress, it needs to stay online to regulate the amygdala's alarm. Secure attachment in infancy is associated with better development of this regulatory capacity. When early care is inconsistent or dismissing, that regulatory pathway is less well developed — not permanently, but as a default (Schore, 2001).

This is also why talking about your attachment style is not enough to change it. Insight sits in the prefrontal cortex — the thinking brain. The attachment pattern lives deeper, in implicit memory — the kind of memory that does not require conscious recall, the kind that runs your body's reactions before your thoughts can catch up. You can understand the pattern completely and still feel the alarm spike when your partner does not text back for four hours.

What changes the pattern is repeated new experience — new relationships where the prediction fails, where the person stays, where reaching out leads to comfort rather than disappointment. The brain updates its predictions when the evidence changes consistently over time. This is neuroplasticity (the brain's physical ability to change with experience) applied to relationship: not a metaphor, but an actual structural process (Siegel, 2012).

It is slow. That is the honest answer. Faster with a skilled therapist, particularly one using approaches that work with the body and with implicit memory — like somatic therapy or EMDR (eye movement desensitisation and reprocessing). But slow even then.

How Attachment Patterns Travel Through Generations

One of the most striking findings in attachment research is this: you can predict, with reasonable accuracy, how a baby will be classified in the Strange Situation — before the baby is born — by asking the mother about her own childhood attachment experiences.

The measure used to do this is called the Adult Attachment Interview (AAI), developed by Mary Main and colleagues (George, Kaplan, & Main, 1985). It does not ask whether a person had a good or bad childhood. It looks at how they talk about it — whether the narrative is coherent, whether feelings and facts are integrated, whether there are blank spaces where memory should be.

A parent who can tell a coherent, honest story about their own early experiences — including difficult ones — is very likely to raise a securely attached child. Not because they had an easy childhood. Because they have processed it. The technical term for this is earned security: security arrived at through reflection and relationship, not through luck of birth.

This is where the generational dimension becomes clear. Unprocessed attachment pain travels. Not genetically, but through the micro-moments of caregiving: the tone of voice, the timing of response, the quality of attention. A parent who learned not to need too much will, without meaning to, teach their child the same. A parent whose own alarm system is always slightly elevated will, without meaning to, keep their child's slightly elevated too.

Main and Hesse (1990) showed that parents classified as "unresolved" with respect to loss or trauma were significantly more likely to have children classified as disorganised. The mechanism is not cruelty. It is dysregulation — a parent who is themselves frightened cannot be a steady enough presence to regulate a frightened child.

This matters because it removes blame from the equation, for both the parent and the child. The patterns that hurt you were almost certainly patterns that hurt your parents first. Understanding that changes something — not everything, but something. It is the beginning of recognising the beliefs and patterns that came down through generations, not from your own choices.

Can This Actually Change?

The research is careful here, and so should we be.

Fraley (2002) conducted a meta-analysis — a study of studies — examining whether attachment patterns are stable or changeable over time. The finding was nuanced: patterns are moderately stable across the lifespan, meaning they tend to persist, but they are not fixed. Significant life events — loss, new relationships, therapy — can shift them.

The most robust evidence for change comes from therapy, particularly from approaches that target the implicit, body-level patterns rather than just conscious understanding. Johnson's Emotionally Focused Therapy (EFT), for example, has shown in multiple randomised controlled trials that it can move couples from insecure to more secure interaction patterns — and that these gains hold at follow-up (Johnson et al., 1999). EFT works directly with the attachment system: it helps partners recognise the cycle they are in, identify the underlying need beneath each defensive move, and risk asking for that need directly.

Individual therapy, particularly with a therapist who maintains what Bowlby called a secure base — reliable, non-reactive, consistently present — provides a corrective experience. The client's nervous system encounters, over and over, an attachment figure who does not withdraw, does not punish emotional expression, does not disappear. The prediction begins to update.

Outside of therapy, close friendships and romantic relationships with securely attached people can do similar work, though more slowly and less predictably. The key variable is consistency. One exception does not update a deep prediction. A hundred consistent exceptions, over months and years, can.

There is also something that does not require another person at all: the capacity to become, in some sense, a secure base for yourself. This is not a self-help platitude. It is a finding. Mikulincer and Shaver (2007) showed that activating mental representations of safe figures — even imagined ones — reduced anxiety and improved coping in experimental conditions. The brain does not fully distinguish between a remembered safe person and an imagined one. You can practise security internally, building a kind of inner steady presence, before you have found it reliably in others.

Jung called something similar the Self — the deeper centre of the personality that can hold complexity without fragmenting. The alchemists, working before the language of psychology existed, described it as the fixed point around which transformation could occur: the lapis, the stable stone. The language differs. The function is the same. A steadier centre allows you to stay present in closeness without being overwhelmed, and to be alone without the silence becoming unbearable.

Frequently Asked Questions

Can you change your attachment style?

Yes, but slowly. Attachment patterns are stored as implicit memory — the kind that runs on autopilot. Consistent new relationships, and sometimes therapy, give the brain enough repeated evidence to update that autopilot. Research by Fraley (2002) shows attachment patterns are moderately stable but not fixed. Change is real; it just takes more than one good conversation.

What does anxious attachment actually feel like?

It feels like a low hum of alarm that only gets quiet when the other person confirms they are still there. You check your phone. You replay the last conversation looking for signs of withdrawal. When they pull back even slightly, the alarm spikes. It is not drama — it is a nervous system that learned that connection is fragile and must be monitored constantly.

What does avoidant attachment actually feel like?

On the outside, avoidants look calm and self-sufficient. On the inside, closeness triggers a quiet discomfort — a pull to get some distance, to think rather than feel, to do something productive rather than sit with someone. It is not coldness. It is a nervous system that learned that needing people leads to disappointment, so it stopped asking.

Why do anxious and avoidant people keep ending up together?

Each triggers the other's deepest pattern. The anxious person's pursuit activates the avoidant's need for space. The avoidant's withdrawal activates the anxious person's alarm. This creates a cycle that feels like passion but is actually two nervous systems locked in a familiar loop. Both partners are, in different ways, replaying the attachment pattern they know best.

Is secure attachment something you can learn as an adult?

Yes. Researchers call this 'earned security' — security that comes not from a perfect childhood but from relationships that repeatedly prove safe. A good therapist, a consistently reliable friend, or a patient partner can all serve as what Bowlby called a 'secure base.' The brain updates its predictions when the evidence changes. Security is not a personality trait you are born with or without.