Two questions, asked before language and answered for life unless revised: am I lovable? Will they come?
Overview
Internal working models are the cognitive heart of Attachment Theory — John Bowlby's answer to how a pattern laid down in infancy could still be steering a marriage forty years later. Borrowing from Kenneth Craik's idea that minds navigate by building small internal models of the world, Bowlby proposed that the infant abstracts, from thousands of interactions with caregivers, a working model of the attachment world. The model has two faces that develop together. A model of the other: are attachment figures available, responsive, likely to come when called? And a model of the self, built as the mirror image: am I the kind of being whose calls get answered — am I lovable? The child whose cries are met composes one answer; the child whose cries are rebuffed or gambled on composes others. Each pattern in the Strange Situation is a working model performed in miniature: secure (self worthy, others reliable), anxious (others unreliable, self insufficient at natural volume), avoidant (others unavailable, self alone but coping), disorganized (multiple models that cannot be held in one frame).
How Models Perpetuate Themselves
The models' power comes from their position: they do not sit among the mind's contents but operate on its inputs. A working model filters perception — the ambiguous text message, the partner's tired sigh, the friend's delayed reply all arrive pre-interpreted by a template built before memory. It shapes attention (the anxious model scans for signs of leaving; the avoidant model scans for signs of encroachment), shapes interpretation (the same neutral face reads as withdrawal to one model and as demand to another), and — most insidiously — shapes behavior that elicits confirming responses. The person braced for abandonment clings until the partner pulls away; the person certain that needs drive people off never voices one, and is never answered. The model manufactures its own evidence. This is why Bowlby observed that the models tend to become automatic and to persist: they stop being beliefs the person holds and become the environment the person lives in.
Models Are Revisable
But Bowlby chose the word working deliberately, on the engineering sense: a working model is a functional draft, built to be updated when reality outruns it. This is the theory's bridge to therapy and its refusal of fatalism. Models installed by relationship are revised by relationship — a therapist, a partner, a friendship that responds off-script long enough and consistently enough that the disconfirming evidence finally gets counted. Reflection is the other lever: a model examined stops being the weather and becomes a fact about one's history, which is a thing that can be worked with. That is much of what parts-work approaches like Internal Family Systems are doing in another vocabulary — finding the young part still running the old model and updating its information — and the documented endpoint, security built despite the original blueprint, is Earned Security.
Related
- Attachment Theory — The framework this construct anchors
- John Bowlby — Its author
- Strange Situation — The models performed in miniature
- Secure Attachment · Anxious Attachment · Avoidant Attachment · Disorganized Attachment — The four templates
- Earned Security — Revision, documented
- Internal Family Systems — Parts-work as model revision in another vocabulary
- Healing — What updating the draft belongs to
- Psychology — Domain index
A working model is a draft. The tragedy is forgetting that; the therapy is remembering it.