Where attachment styles come from
Seventy years of research, in the time it takes to finish a coffee.
Attachment describes what your nervous system does when closeness is at stake. Not how much you love people. What happens automatically when someone you care about goes quiet, or needs more than you have.
The research behind it is real. It is also narrower than the way the words get used online.
It began with one year olds
John Bowlby argued in the 1950s that the drive to stay near a protective adult is a system in its own right, as basic as hunger. Mary Ainsworth made it measurable: the Strange Situation, twenty minutes in a lab watching a one year old separate from a parent and reunite.
Three patterns came out of it, and Mary Main and Judith Solomon added a fourth in 1990 for children whose behaviour followed no strategy at all. They called that one *disorganized*.
Then it became a question about adults
Cindy Hazan and Phillip Shaver asked in 1987 whether adult bonds ran on the same system. Kim Bartholomew and Leonard Horowitz proposed four adult types in 1991.
Then the field dropped the boxes. In 1998 Kelly Brennan, Catherine Clark and Phillip Shaver factor-analysed every attachment measure in English and found they all reduced to two dials. That produced the ECR, revised in 2000 into the ECR-R, still the standard questionnaire today.
- Instead of: Attachment anxiety Try: How much you fear being left, and how much reassurance you need that the bond is safe.
- Instead of: Attachment avoidance Try: How uncomfortable closeness feels, and how strongly you pull toward handling things alone.
- Instead of: Low on both Try: Secure.
- Instead of: High anxiety, low avoidance Try: Anxious, sometimes called preoccupied.
- Instead of: Low anxiety, high avoidance Try: Avoidant, sometimes called dismissing.
- Instead of: High on both Try: Fearful-avoidant.
Nobody sits exactly on a corner. Everybody sits somewhere on the grid, and the four names are a convenience for describing which part.
Why we do not say disorganized
Our fourth result used to carry that name. We changed it.
*Disorganized* belongs to the infant procedure. Its adult counterpart is scored from the Adult Attachment Interview, an hour of recorded conversation about your childhood, coded by a trained rater listening for breaks in how you talk about loss. Questionnaires cannot reach it, so they call this corner fearful-avoidant.
The gap is measurable. A 2007 meta-analysis pooling 961 people found the two kinds of measure agree at r = .09, which is close to not at all. Using the interview's word for a questionnaire's result would borrow authority we have not earned, about trauma we have no way of assessing.
What holds up, and what does not
- Holds up The two dials are reliable across samples and languages, and they predict real things: how couples behave in conflict, how people seek support under stress.
- It is relationship-specific Your pattern with a partner is not automatically your pattern with your mother or your child. People score differently depending on who they are asked about.
- It moves Moderately stable, not fixed. It shifts with circumstances, relationships, and therapy.
- The measures disagree The r = .09 above. Anyone describing one unified science of attachment styles is skipping this.
Why our eight questions look like they do
Each one puts you somewhere specific and asks what you actually do. Asking how anxious you are gets you your self-image; asking what you do when a partner goes quiet gets closer to the behaviour.
One choice per question rather than a seven point scale is the trade we make for two minutes. A scale would tell you how far along each dial you sit. A forced choice can only tell you which way you lean.
Eight questions about what you do when a relationship gets strained, and a reading of which way you lean on each dial.
Free to start, no card needed. Your result is saved to your profile so Avani can use it when you talk something through.
Create a free account to take it
- Sources
- Ainsworth, Blehar, Waters & Wall (1978), Patterns of Attachment. Main & Solomon (1990), on identifying disorganized infants in the Strange Situation. Hazan & Shaver (1987), Journal of Personality and Social Psychology 52, 511-524. Bartholomew & Horowitz (1991), on the four-category model. Brennan, Clark & Shaver (1998), on self-report measurement of adult attachment. Fraley, Waller & Brennan (2000), the ECR-R. Wei, Russell, Mallinckrodt & Vogel (2007), the ECR Short Form, Journal of Personality Assessment 88(2). Roisman, Holland, Fortuna, Fraley, Clausell & Clarke (2007), the AAI and self-reports of attachment style, Journal of Personality and Social Psychology 92(4), 678-697.