Earned Secure Attachment

What it is

The Adult Attachment Interview scores coherence rather than content. An adult is asked to describe their childhood relationships and then to produce specific memories supporting each description. What is coded is whether the account holds together: whether someone who reports real harm can describe it steadily, without either dismissing it or being pulled back into it.

That creates a category. Some adults describe a hard childhood and describe it coherently. They are classified secure. Because their narrated history is difficult and their current state is secure, the obvious inference is that something happened in between, and the label "earned secure" encodes that inference.

The inference is the part that was never tested, until it was.

In effect

In 2002 Glenn Roisman, Elena Padrón, Alan Sroufe and Byron Egeland ran the test using the Minnesota cohort, a group followed since infancy with observational records of their actual childhoods.

Adults classified as retrospectively earned-secure were not more likely to have been anxiously attached as infants. Observed at the time, they had received some of the most supportive and structured maternal parenting in the sample.

So the category was not picking up people who had overcome adversity. It was picking up people who narrate their childhood negatively now, which tracks a depressive recall style more closely than it tracks a documented history. The authors' recommendation was to reserve the term for adversity that can be established prospectively, preferably from observation showing actual change.

Whether attachment security can be improved in adulthood is a separate question and the honest answer is that the evidence is thin. A review of fourteen studies found security rising after therapy, but most of them had no control group, and people improve over time in uncontrolled studies of almost anything.

What it does not say

It does not mean nobody recovers from a difficult childhood. It means this particular label is not evidence that they did.

It does not make the interview useless. Narrative coherence is a real and measurable thing, and it predicts things. It is just not a window onto what happened.

And it is not a reason to distrust your own account of your childhood. It is a reason not to treat a coherence score as a historical record.

---