Attachment—δεσμός—التعلق
A Proper Introduction to Attachment Theory (Because 'TikTok Psychology' has Flattened it)
Love has a way of humbling our belief that we are self-determining creatures. Again and again, in friendships, families, novels, and therapy rooms, I find myself returning to the familiar question: Why have we ended up here?
Falling in love with the same flavour of person again and again. Staying too long, or leaving before we can be left. Finding someone steady and feeling, inexplicably, bored. Finding someone unavailable and feeling, inexplicably, alive. Pushing people away at the exact moment they get close enough to matter. Giving too much, too fast, and calling it love when it's closer to panic. Giving nothing and calling it self-sufficiency when it's closer to grief. Choosing the chaotic relationship over the peaceful one, because peace feels like waiting for something to go wrong. Being chosen by someone wonderful and spending the entirety of the relationship waiting to be abandoned by them.
Surely none of it is random. My average reader will have spent long enough on the internet by now to have come across ‘attachment.’ You'll have seen the quizzes, the infographics, the TikToks telling you you're anxiously attached in twelve seconds flat. Some of it is useful. Most of it is the map without the territory. What I want to offer here is something closer to the territory itself.
Origins
Let’s start at the beginning, which is always, in this field, John Bowlby.
In 1969, Bowlby published Attachment, the first of three landmark volumes that would permanently alter how we understand human development1. He was asking what is, in retrospect, a deceptively simple question: why do early relationships with caregivers have such a profound and lasting effect on everything that follows?
What made Bowlby radical for his time, and a bit lonely in his own field for a while, was his insistence on looking at this question from multiple directions at once (psychological, biological, evolutionary). He drew on ethology (the study of animal behaviour), psychoanalysis, and systems theory, arguing that none of these alone was sufficient. Bowlby was attempting something that both Darwin and Freud had gestured towards but never fully achieved: a genuine dialogue between the biological and the psychological. He wanted to show how the inner life of a child, their feelings, their sense of self, their expectations of others, is shaped by the actual, physical, relational world they are born into.
Bowlby’s core insight was that the infant is not a passive recipient of care. The baby is an active participant in a system, a reciprocal, dynamic, biological relationship with a caregiver. And that system, in its earliest form, literally shapes the brain.
What Is Attachment, Actually?
It's worth pausing on the word itself, because "attachment" gets used to mean almost anything these days. People talk about being attached to their phone, to a TV series, to a city they used to live in, to almost anything emotionally involving. That's not what Bowlby meant.
In the technical sense, an attachment relationship is one that serves specific functions. Attachment figures are the people we turn to when we feel threatened or distressed, they are our safe haven. They are also our secure base: the platform from which we feel confident enough to go and explore the world. Not every close relationship is an attachment relationship. Your favourite colleague might matter enormously to you, but you probably don’t find yourself reaching for them when the ground falls away beneath your feet.
What Bowlby identified, and what subsequent researchers have built upon, is that attachment is not merely a psychological preference. It is, as Allan Schore puts it, “the interactive regulation of biological synchronicity between organisms.”2 Dense wording, but it points to something marvellous: when a caregiver and infant are in attunement with each other, when they’re in that face-to-face dance of eye contact, cooing, mirroring, soothing, they are quite literally synchronising their nervous systems. The caregiver is helping to regulate the baby’s internal states: calming arousal, amplifying joy, soothing distress (read Maternal Psychology is Nothing Short of Miraculous for more on this process).
What’s Happening in the Brain
Schore’s contribution to attachment theory is to map what Bowlby described in psychological terms onto the developing brain, and the picture that emerges is pretty extraordinary.
The right hemisphere of the brain is dominant in human infants for the first three years of life. This is the hemisphere that processes emotion, holds the body's internal regulatory states, and carries the implicit, pre-verbal, pre-conscious, sense of who we are in relation to other people. It is through right-brain-to-right-brain communication (the caregiver's face, voice, timing, touch, their gaze) that the earliest attachment experiences are transmitted, received, and encoded.3
At the centre of this is a region called the orbitofrontal cortex, which Schore describes as the ‘senior executive of the emotional brain.’ This is the system that learns to regulate affect, to appraise situations for their emotional significance, to shift flexibly between states. And crucially, it is sculpted by experience during a critical developmental window: roughly ten months to two and a half years. It is built in the fire of early attachment relationships.
So what does all of this actually look like? Think about a mother who is warm, emotionally open, comfortable with her own inner life. When she sits with her baby, she creates a rhythm. The infant cues, she responds. Equilibrium is disturbed, equilibrium is restored. Over and over, a thousand times. She is, in a very real sense, willing to go where the infant leads, emotionally, moment to moment. It is precisely these exchanges, these unremarkable right-brain-to-right-brain exchanges, these daily moments of being met, repeated thousands of times across the first years of life, that wire the infant’s emotional brain for optimal development.
But here is the part I find genuinely astonishing, the part that stops a person in their tracks every time I explain it. The mechanism of soothing is not what most people imagine. We tend to picture a mother swooping in with cheerfulness, distracting the baby, lifting the mood, offering something brighter. That is not what happens in attunement. What actually happens is almost the opposite.
When the infant begins to cry in distress, the attuned mother does not immediately try to pull the baby upward into calm. Instead, she herself descends. Unconsciously, her own nervous system moves to meet the infant’s emotional state, her face softens into something that mirrors the distress, her voice carries the colour of it, her body comes close. She sinks, emotionally, to where the baby is. The infant is met there. Joined. The terrible aloneness of the distress is dissolved, because now there are two people in it. And then, and only then, does she begin to lift. Gradually, her own emotional state begins to move toward regulation, toward ease, toward warmth. And the infant, already attuned to her, follows. Unconsciously, automatically, the baby’s nervous system tracks hers upward, out of the distress and into something more settled.
That is the sequence: descent into meeting, then a shared ascent. It is, I don’t use this word lightly, miraculous. The child is not talked out of their distress. They are accompanied through it. And it is the experience of being accompanied, repeated again and again, that teaches the developing brain that emotional states are not permanent, that distress has an ending, that another person can be a source of regulation rather than threat. The implications are significant. When a caregiver is consistently responsive in this way, not perfectly, but reliably, the orbitofrontal system develops in ways that support emotional flexibility, stress regulation, and what researchers call secure attachment. The child internalises a fundamental assumption: that when things go wrong, they can be set right. That help is available. That connection and repair, after rupture, is possible.
However, when caregiving is inconsistent, frightening, or chronically misattuned, the system develops differently, not defectively, but adaptively, in ways shaped by a more difficult relational environment. This is where the different attachment styles emerge.
The Attachment Styles
Mary Ainsworth, Bowlby’s great collaborator, gave us the empirical frame that transformed his theory into something measurable. Her famous “Strange Situation” (a structured experiment in which a toddler experiences brief separations from and reunions with their caregiver) revealed something important: children's responses are not random expressions of temperament. They are coherent, organised adaptations to the specific relational environment each child has been living in. In other words: attachment styles.4
Secure Attachment
The securely attached child has a caregiver who is generally reliably present and responsive, not perfect (there is no such thing), but available enough that the child has learned that the world is basically safe and relationships are basically trustworthy. When distressed, these children seek comfort, accept it, and return to play. They have what we might call a confident expectation of repair.
Schore frames this elegantly: in the securely attached individual, internal working models “encode an expectation that homeostatic disruptions will be set right.” This is the foundational emotional grammar: things go wrong, and then they get better. I am not alone in my distress.
Anxious (or Ambivalent) Attachment
This style develops when caregiving is inconsistent, sometimes warm and responsive, sometimes distracted, overwhelmed, or emotionally preoccupied. The child cannot predict when the caregiver will be available. So what does an intelligent, adaptive nervous system do with that sort of unpredictability? Turn up the volume. Stay close. Don’t fully self-soothe, because the caregiver’s availability is uncertain and you need to keep the signal running.
Anxiously attached children appear inconsolable during separation and are difficult to soothe on reunion, not because they’re wilful, but because their system has learned that the attachment figure’s reliability is not to be trusted. 'There will not always be repair for my rupture. The strategy is hyperactivation of attachment behaviour. More distress, more proximity-seeking, less independence.
Avoidant Attachment
When caregivers are consistently emotionally unavailable, dismissive of emotional expression, or actively uncomfortable with closeness and need, the child faces a very particular bind. Expressing attachment needs leads to the withdrawal of the very connection being sought. The adaptive strategy here is deactivation: minimise emotional expression, suppress attachment needs, develop an exaggerated self-reliance.
These children look remarkably calm during separation and on reunion. Physiological measures, however, tell a different story. Their stress hormones are just as elevated as other children’s. The distress is there; it has simply been driven underground. The strategy works in one sense (keeps the caregiver from withdrawing further) and fails in another (leaves the child’s internal distress unregulated, unshared).
Disorganised Attachment
This is the most complex and, in therapeutic terms, the most significant pattern to understand. It emerges when the caregiver is simultaneously the source of fear and the source of comfort, as in cases of abuse, severe neglect, or a parent whose own unresolved trauma causes them to behave in frightening or frighteningly helpless ways.
Here the child faces what researchers call ‘fright without solution.’ 5Every instinct says: go to the caregiver. But the caregiver is the threat. The attachment system is activated and simultaneously blocked. There is no coherent strategy available. What you see is behavioural collapse: freezing, contradictory movements, dissociation. The system short-circuits because no response is safe.
Disorganised attachment is consistently associated with the most significant difficulties in adult relationships and mental health and it is more common than the available literature would have you believe.
Internal Working Models
One of Bowlby's most enduring ideas is the internal working model and I think this is worth spending some real time on, because it's where the child development story connects directly to the adult experience of why we do what we do. As the child moves through these early attachment experiences, they are not merely reacting to what happens in the moment, they are building a representational map of how relationships work. How available are others likely to be? Am I the kind of person who deserves care, time, interest, desire, love? Is it safe to show need?
These models, encoded in implicit memory, the kind of memory that lives in the body and the nervous system, not in conscious narrative, become the templates through which we perceive, interpret and navigate all subsequent relationships. They operate beneath awareness. They are not decisions. They are assumptions so deep they feel like reality itself.
Schore makes the point that these models encode not just cognitive expectations but strategies of affect regulation (habitual ways of managing emotional states in relational contexts). The avoidant child who learned to suppress need has an internal working model that says, roughly: “I am on my own with this. Showing it will make things worse.” The anxiously attached child’s model says something like: “I must stay vigilant. I cannot trust that a connection will last.”
And here is what I find myself saying to clients, again and again, because I genuinely believe it: these are not pathologies. They are wisdom. They were brilliantly creative adaptations to, once, intolerable problems. Ingenious ways to navigate specific relational environments those children found themsleves in. The tragedy isn't that the strategies formed. The tragedy is that they don't tend to update themselves automatically once the environment changes.
This is the question. Why now? Why still? Why do I keep doing this? Why, knowing what I know, do I still react this way?
The answer lies in the nature of the unconsicous or, if you prefer, implicit memory. Internal working models are not stored as conscious beliefs they are procedural, embodied, automatic. They are activated not by thinking about the past but by anything in the present that resembles the relational conditions under which they were formed. The smell of a particular anxiety. The sound of silence in a room. A boyfriend’s slight withdrawal. A girlfriend’s raised eyebrow that carries, for a moment, the ghost of a parent’s disapproval.
These models function as guides for future action. The child who learned that need leads to rejection will, as an adult, not consciously decide to be self-sufficient to the point of isolation, they will simply find themselves unable to ask for help, and they may not even know they need it. The person who learned that love is unpredictable will find themselves preoccupied in relationships, reading every small signal for signs of withdrawal, amplifying distress because the system never learned that calm waiting is a viable option.
And here is the rude rub: we are often drawn to relational dynamics that replicate what we know. That is the paradox. In love, we seek to re-find all or some parts of the people to whom we were attached as children. And, simultaneously, we ask our beloved to correct all of the wrongs that these early parents or siblings inflicted. Not because we want to suffer, but because familiarity is attractive, it feels like safety, even when it isn’t. The person who grew up walking on eggshells may feel oddly comfortable with a partner who runs hot and cold. It’s familiar. And we will choose a familar hell over an unfamiliar heaven almost every time. The nervous system recognises it. There’s a terrible kind of homecoming in it.
It’s also worth pausing here to bring in a voice that complicates all of this in the most useful ways. Jacques Lacan, the French psychoanalyst who spent his career making straightforward things productively difficult, would take issue with the very idea that we are seeking a person when we fall in love. For Lacan, what we are seeking is something more fundamental and more elusive: the objet petit a (the object-cause of desire) and his central, rather devastating, insight was that we don’t actually want to find it. Because if we found it, desire itself would end. And so we are drawn, structurally, to people and dynamics that keep desire alive. Meaning, in practice, people who are just out of reach, relationships that are never quite resolved, love that always has a gap at its centre. From this angle, the patterns attachment theory describes are not only the residue of early experience, they are also the expression of something more fundamental in the human condition: the fact that we are beings who are constituted by a lack we cannot name, and who spend our lives reaching for something we cannot quite articulate and would not know what to do with if we found it.
This makes my mind, a religious mind, think immediately of God. Are depictions of God not almost unanimously depictions of the ultimate answer to precisely this human longing? All-loving, all-forgiving, all-knowing, omnipresent: the perfect attachment figure, the perfect secure base? A presence imagined as incapable of abandonment, incapable of misunderstanding, incapable of withdrawal. And yet, at the same time, never fully graspable. God remains pursued, prayed toward, longed for, experienced in flashes, but always somehow beyond complete possession. From this perspective, religion can begin to look less like an escape from desire, than its highest articulation. But I digress. Back to human-human attachment.
Adult Attachment
Mary Main, one of the most important figures in the second generation of attachment research, extended all of this into adulthood through the development of the Adult Attachment Interview (a measure of how adults narrate their own childhood experiences).6 Fascinatingly, it’s not what happened to people that predicts their adult attachment style; it’s the coherence of their account of what happened. People who can speak about their childhood with perspective, nuance, and emotional integration, whether or not that childhood was easy, tend to be securely attached as adults. Those whose narratives are dismissive, confused, or flooded with unprocessed emotion tend to show the adult equivalents of avoidant, anxious, and disorganised attachment. Hence, the utility of undergoing a relational mdoel of psychotherapy.
The adult versions look something like this:
The secure adult can tolerate intimacy and separateness. They can ask for help without it feeling catastrophic. They can be disappointed by a partner without it meaning the relationship is over. They have, as a baseline, the felt sense that they are loveable, that others are basically available, and that repair is possible.
The anxiously attached adult tends to preoccupy with relationships. They may worry excessively about abandonment, seek constant reassurance (‘clinginess’), interpret ambiguity as rejection. They can be exhausting to themselves, and sometimes to their partners. They feel love intensely, often because love has always felt precarious.
The avoidantly attached adult has learned to be independent to a fault. They may intellectualise their inner life, keep emotional distance, feel uncomfortable when partners want more closeness than they can give. They often don’t know they’re lonely. They tend to perform health.
The disorganised adult (often described as unresolved in the literature) may oscillate between desperate closeness and sudden shutdown. They may have experienced relational trauma that makes intimacy simultaneously the thing they most want and the thing they most fear.
And here is the thing no one tells you in the neat typologies: these are not fixed identities. They are patterns, and patterns can shift.
Can They Shift?
The research is now clear, from multiple directions, that the brain retains far more plasticity than we once believed. The orbitofrontal system, Schore is explicit about this, remains capable of experience dependent change well beyond childhood and early development. Attachment patterns can shift, not easily, not quickly, but they can, through new relational experiences that provide something different from what what the internal working model has learned to expect.
This is, at its core, what therapy offers. Not insight alone (though insight matters), but a sustained relational experience in which the implicit system gets to encounter something new: a person who stays present when things get difficult. Who can hold and tolerate your distress, perhaps even share it, without becoming overwhelmed themselves, or withdrawing, or getting defensive. Who notices your pulling away but doesn't punish it. Who repairs ruptures, consistently, with care. Who is, boringly and dependably, there.
The nervous system notices. Slowly, incrementally, and software begins to update.
The language of neuroscience for this is that the right prefrontal cortex learns to modulate limbic activation, that higher-order neural networks develop the capacity to regulate the deeper, faster, older threat responses. My preferred language is simpler: something shifts in the body's felt sense of what is possible in a relationship.
Earned security, the term researchers use for people who develop security in later life despite difficult early attachment histories, is real.7 It is documented and it is not a therapeutic fiction. It requires, almost always, sustained experiences of being genuinely known and held over time. It can happen in a long, committed partnership. In a deep and durable friendship. In a community. And it can happen in therapy.
Bowlby, J. (1969). Attachment and Loss, Vol. 1: Attachment. New York: Basic Books.
Schore, A. N. (1994). Affect Regulation and the Origin of the Self: The Neurobiology of Emotional Development. Hillsdale, NJ: Lawrence Erlbaum Associates.
Schore, A. N. (2001). Effects of a secure attachment relationship on right brain development, affect regulation, and infant mental health. Infant Mental Health Journal, 22(1–2), 7–66.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of Attachment: A Psychological Study of the Strange Situation. Hillsdale, NJ: Lawrence Erlbaum Associates.
Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the Preschool Years (pp. 121–160). Chicago: University of Chicago Press.
George, C., Kaplan, N., & Main, M. (1985). Adult Attachment Interview (unpublished manuscript). University of California, Berkeley.
Pearson, J. L., Cohn, D. A., Cowan, P. A., & Cowan, C. P. (1994). Earned- and continuous-security in adult attachment: Relation to depressive symptomatology and parenting style. Development and Psychopathology, 6(2), 359–373.