The Internal Working Model: The Relationship Map You Don’t Know You’re Using

A woman I’ll call Nina came to see me after her second divorce. She was forty-four, careful and deliberate in the way that people become after they have been hurt enough times to stop trusting their first instincts. She told me she had a type , not in the affectionate, self-deprecating way people sometimes say this, but with genuine bewilderment, as though she were reporting a recurring mechanical failure she could not diagnose. Both husbands had been emotionally unavailable. Both had been brilliant, professionally driven, and fundamentally not quite present in the way she needed. She had known it about both of them, she said, long before the marriages ended. She had believed, each time, that love would be enough to change what she now recognized as a fixed pattern in the men she chose.

I want to say something about Nina that I want to be careful with, because it would be easy to miss. The problem was not that she had bad taste in partners. The problem was that she had been navigating her relationships using a map that was drawn a very long time ago, in conditions very different from the ones she was in now, and that the map described those original conditions accurately. It just did not describe reality. She kept getting lost because she kept trusting the map.

What Bowlby meant by a working model

John Bowlby introduced the concept of the internal working model in the 1960s as part of his effort to explain why early attachment experiences have the lasting effect that they clearly do. The idea was this: as the infant moves through its early relational experiences, it builds a mental representation of how relationships work. Not a conscious representation, not a set of beliefs the child could articulate, but something more like a template , an implicit prediction system that encodes what to expect from close relationships, from the self in those relationships, and from the world those relationships inhabit.

The model is built from experience, the same way all neural architecture is built from experience: through repetition, through the accumulated weight of thousands of ordinary interactions that teach the nervous system what to anticipate. A child who reaches for comfort and finds it, repeatedly, builds an internal model that says: when I am in distress, reaching out tends to work. People can be relied upon. I am worthy of care. A child who reaches and finds absence or hostility or inconsistency builds a different model: people are unreliable. Reaching out may make things worse. I am not reliably worth attending to. Both models are rational conclusions drawn from the available evidence. Both are, in the context where they were formed, accurate.

The problem is what happens next.

The model that doesn’t know it’s a model

The internal working model does not present itself as a model. It presents itself as reality. This is the most important thing I can say about it, and also the hardest thing to convey to people who have not yet been able to see their own.

Nina did not think: I have an internal working model that predicts emotional unavailability from partners. She thought: men are emotionally unavailable. She did not think: I am attracted to people who match my early relational template. She thought: there is something wrong with me that I keep choosing the wrong person. The model was running her interpretations, her predictions, her attraction responses, her alarm system , all of it , without announcing itself. It felt like simply seeing the world clearly. And in a way it was, because the model is built from genuine experience and produces genuine expectations. The trouble is that it was built in one world and is operating in another.

What makes this particularly difficult to see is that the model is self-reinforcing. People tend to interpret ambiguous information through the lens of their existing expectations, and they tend to behave in ways that elicit responses consistent with what they expect. A person who expects to be abandoned often behaves in ways that create distance, which creates the conditions for abandonment, which confirms the expectation. A person who expects that they are not worthy of real attention often diminishes their own needs in relationships, which results in their needs being unmet, which confirms that they were right not to expect them to be met. The model produces the evidence that sustains it. This is not a moral failing. It is how expectation systems work, in the same way that a thermostat set to a particular temperature will keep cycling the heat on and off to maintain that temperature even if the weather outside changes.

What the model contains

Bowlby described the internal working model as actually two interrelated models: one of the self and one of other people. They develop together and reinforce each other.

The model of other people encodes expectations about how close relationships function. Whether the people you depend on will be available when you need them. Whether closeness brings comfort or danger or disappointment. Whether the people you love can be trusted to remain. These expectations are not held as conscious beliefs , they operate below the level of deliberate thought, shaping perception and response at a speed that precedes reflection.

The model of the self encodes a sense of whether you are the kind of person who is worthy of care. Whether you deserve to have your needs attended to. Whether you are, at some fundamental level, lovable. A child who is responded to warmly and consistently develops a sense of self that includes the assumption that they matter, that their inner experience is worth attending to, that their reaching will be received. A child who is chronically not responded to, or who is responded to only when performing in certain ways, develops a different sense of self, one organized around the idea that their worth is conditional or partial or perpetually in question.

The two models work together in the predictable way: a person who expects other people to be unavailable and who also believes themselves to be less than fully worthy of care will approach relationships with a particular set of strategies, calibrated to that combination of expectations, and will interpret relational events through a lens that confirms both.

Why it feels like the truth

I have sat across from enough people to know that what I am about to describe is almost universal among people who carry insecure attachment models: the model feels like clear-eyed perception. It does not feel like a filter. It feels like vision itself.

When Nina told me that men are emotionally unavailable, she was not stating a prejudice she was aware of holding. She was reporting what she had observed, across two marriages and several relationships before them, to be reliably true. And she had observed it. The data was real. What she could not yet see was how much her model was selecting for the data, directing her attention toward confirming evidence and away from disconfirming evidence, making her feel at home in dynamics that matched the model because the match felt like recognition.

Familiarity is one of the more powerful and least examined forces in how people choose relationships. The nervous system does not always register familiar as the same as good. It registers it as known. And known, in the calculus of the attachment system, carries a weight that new does not. A relationship that matches the model feels legible, navigable, even when what it is navigating is pain. A relationship that does not match the model can feel destabilizing even when it is safe, because the nervous system does not have a template for it, and what it cannot predict, it often cannot fully trust.

This is why many people who have worked hard to understand their patterns still find themselves drawn to familiar dynamics. Understanding the model does not automatically update it. The model was built experientially, at a level beneath conscious access, and it updates the same way , through new experience, accumulated over time, in relationships that provide different information about what to expect from closeness.

What happens when it updates

The research on earned secure attachment describes people who were insecurely attached in childhood and who developed genuinely secure patterns in adulthood. These people did not change by gaining insight, though insight was often part of it. They changed through relational experience , through having, in a partnership or in therapy or in a close friendship, the repeated experience of reaching and being met, of being seen without it being weaponized, of closeness that did not produce the punishing consequence the model had predicted.

The model updates slowly, the way it was built slowly, through accumulated evidence that runs counter to its predictions. And the process of updating it is not comfortable, because what it requires is tolerating the dissonance between what the model predicts and what is actually happening, holding the two things at the same time long enough for the new experience to register as data.

Nina eventually said something that I think about still. She said that she had started seeing someone who was different from anyone she had chosen before, who was present and warm in a way that should have felt good but that she kept waiting to be disproved. She said it was like learning to read in a different language , she had to go slowly, had to resist the impulse to translate everything into the familiar, had to stay with the discomfort of not yet being fluent in something good. She said: I think I’m learning that the map might just be wrong.

That is the beginning of it. Not the end, but the beginning. The map was not wrong about the original territory. It was simply made for a place she no longer lives, and the work, now, is drawing a new one.

Daniel

Daniel Wethington is a Licensed Professional Counselor and the founder of Arise Counseling Services in York, Pennsylvania, where he specializes in attachment trauma, anxiety, and gaming and technology addiction. He is the author of four books, including Shaped in Silence and Breaking Free, which explore how early attachment experiences shape adult relationships and compulsive behaviors, which you can find here. Dan works with individuals and couples across Pennsylvania, both in person and through telehealth. If something in this article resonated with you, you can schedule an appointment or learn more at http://arise-pa.com.