Most people assume the guarded adult was always that way, and that a wall put up in someone’s forties points back to something that went wrong in childhood. Often the pattern runs the other way. Plenty of people who grew up secure, affectionate, and open still end up building a wall later in life, and when they do, it usually follows a specific run of adult betrayals rather than anything that happened when they were small.
The psychologist Jennifer Freyd, who developed the concept of betrayal trauma, made one observation that helps explain why: the injuries that reshape us most are the ones delivered by the people we had already decided to depend on. By her definition, this kind of harm occurs when the people or institutions a person relies on for survival violate that trust. The key word is depend. A stranger cannot betray you in this sense. Only the people you had already let in can.
Seen this way, the wall is less a personality flaw than a filing decision. Somewhere in the second or third round of adult evidence, a person can quietly move intimacy from the necessary column to the optional one, and walk out of that reclassification changed.
The old assumption, and why it wobbled
For most of the twentieth century, attachment style was treated as something set in early childhood and largely fixed by adolescence. John Bowlby’s original framework described the pattern a child learned with a primary caregiver as a working model that the adult would carry into every future relationship. Secure kids became secure adults. Avoidant kids became avoidant adults. The map, once drawn, stayed drawn.
That model held for a long time. Then the longitudinal data got more complicated.
A large, decades-long study covered by Scientific American found that childhood relationships — especially with mothers, and with close childhood friends — go on to shape adult attachment styles across the board, though in more nuanced ways than researchers once assumed. Crucially, the researchers also stressed that a person’s style is not locked: there is always room to move. Childhood sets a starting point; later experience can still shift it. And when people who once felt securely attached describe feeling more guarded decades on, the change tends to track to identifiable adult events — a divorce, a workplace collapse, a friend group that turned, someone close who lied about something structural.
What the sequence tends to look like
It rarely happens after a single incident. One affair, one lie, one falling-out tends to produce grief, not a wall. The wall more often goes up after the third or fourth data point, when the pattern becomes hard for the person collecting the evidence to ignore.
A partner of eleven years turns out to have been living a parallel life. Six months later, a close friend from college quietly disappears without explanation. A year after that, a sibling repeats a story that was told in confidence. The person absorbing all of this is not a child. They are 38, or 45, or 52. They have adult tools. They can name what happened. And that is part of why the wall goes up: the evidence is legible in a way it never could have been at seven years old.
People who describe this arc often say the decision felt almost administrative. They were not raging. They were doing math. Being fully known had produced repeated, verifiable hurt. Being less known would have cost less. The equation resolved.
Why a naturally open temperament doesn’t protect you
You might expect a warm, trusting baseline to insulate someone from this. In practice it often does the opposite. People who are naturally open extend more credit, sooner, and closer to the chest. When that credit is broken, the correction tends to be proportionally larger. The person still wants connection. They have simply concluded, on the strength of adult experience, that connection is also where the injuries came from.
None of this means the working model Bowlby described is carved in stone at age three. Adult experience is data too, and a mind that updates its expectations will revise old estimates when new ones keep arriving.
Why divorce hits so hard
Divorce is a common trigger, and the reason is structural rather than emotional. Marriage is the adult relationship most explicitly built on public, legal, and financial interdependence. Someone entering a marriage has, by definition, decided to depend on another person in the most concrete ways — housing, income, health insurance, parenting, future planning, the phone call from the hospital.
When that arrangement collapses through betrayal rather than mutual drift, the injury lands squarely on the person you had chosen to rely on. And the lesson a person takes from it is often not a story about that one partner. It becomes a revised estimate of what depending on anyone might cost.
The friend group that turns
Another common trigger is quieter and, in some ways, more corrosive: the friend group that reorganizes without the person. Someone gets sick, or loses a job, or goes through a public embarrassment, and the group that seemed permanent thins to two people, then one, then none. No fight. No confrontation. Just a slow social evaporation.
People who have lived through this tend to use the same phrase — they say they learned who their friends actually were. The wall that goes up afterward is calibrated specifically to that lesson. It is not a wall against strangers. It is a wall against people who look like friends but have not yet been stress-tested.
The familiar attachment categories — secure, anxious, avoidant, disorganized — were first mapped in children, but they are better understood as movable positions rather than fixed labels, with adult experience acting as the wind that nudges people between them.
What the wall actually does
The wall is not silence. Most people who put one up after adult betrayal stay sociable, funny, competent at work, warm in surface interactions. What changes is the depth setting. They stop volunteering the second-layer information — the fear, the health worry, the money problem, the crush, the doubt. They keep the first layer wide open and the second layer sealed.
A Forbes piece by psychologist Mark Travers describes this pattern from the inside. The person has excellent communication skills, strong self-awareness, and clear boundaries. They can talk about their feelings fluently. What they will not do is put themselves in a position where being known could cost them again — an approach that, as Travers puts it, can create the appearance of openness while keeping tight control underneath. Insight becomes a tool for staying safe rather than getting closer.
It is one of the hardest presentations to notice, because nothing looks wrong from the outside. The person is functioning. They have friends. They date. They just do not let anyone through the second door.
Whether the wall comes down
The honest answer is: sometimes, partly, and only with new evidence. The same mechanism that built the wall can lower it. If experience can revise an estimate downward, it can also revise it upward — but only if the person lets themselves into situations where new, contradictory data can accumulate. That requires the very openness the wall was built to prevent.
People describe a slow-motion version of this. They let one new person past the first layer, then wait. Nothing bad happens. They let a second in. Nothing bad happens. Over three or four years, sometimes longer, the estimate shifts. The wall does not vanish. It gets shorter, and the door in it gets easier to open.
What Bowlby’s later thinking and the more recent attachment research converge on is a quiet correction to a very old assumption. The idea that character is fixed by childhood is comforting for people who had good childhoods and unkind to people who did not, and it turns out to be shaky in both directions. The working model updates.
Which means the person who arrives at 45 with a wall they did not have at 25 has not become someone else. They have become the same person carrying newer information. The openness they were born with is still there, filed behind adult evidence, waiting to see whether the next set of data is worth revising for.
