A room nine feet square, a scattering of toys, two chairs and a one-way mirror. That was the entire apparatus, and it has been shaping how people talk about love for half a century.
The psychologist was Mary Ainsworth, who built the procedure with Barbara Wittig in the late 1960s and reported it in detail with Silvia Bell in a 1970 paper in Child Development. Fifty-six infants, every one of them between 49 and 51 weeks old, went through eight short episodes of roughly three minutes each: mother and baby in the room, then a stranger joins, then the mother leaves. She returns, leaves once more, then returns for good.
Ainsworth was not really watching the crying.
What the reunion showed
Her interest was the moment of return. Ainsworth and Bell found that the mother’s presence encouraged babies to explore the room, while her absence flattened that exploration and pushed clinging and searching upward. On reunion, most infants went to her and held on. A sizeable minority did something stranger: they wanted contact and resisted it at the same time, arching away while gripping. Others ignored her completely.
Those three patterns became the taxonomy everybody now half-remembers from a podcast. Secure, avoidant, resistant. A fourth, disorganised, was added years later by Mary Main and Judith Solomon to cover babies whose strategy seemed to collapse mid-episode.
How twelve months became a dating theory
The jump from nursery to bedroom took two decades. In 1987, Cindy Hazan and Phillip Shaver proposed in the Journal of Personality and Social Psychology that romantic love is itself an attachment process, and translated Ainsworth’s three infant categories directly into adult ones. Roughly the same proportions showed up in their adult samples, and the three groups described falling in love in predictably different ways.
That paper was two questionnaire studies. Adults recalling their own childhoods, adults describing their own love lives. Nobody had gone back to the infants themselves.
Then somebody followed the babies
Jeffry Simpson and colleagues at the University of Minnesota tracked 78 people from infancy into their mid-twenties. Writing in the same journal in 2007, they reported a chain rather than a straight wire. Babies rated secure at twelve months were judged more socially competent by their primary school teachers. That competence predicted the quality of their friendships at sixteen. Those friendships predicted the emotional weather of their romantic relationships at twenty to twenty-three.
Every link in that sequence is doing work. Pull one out and the current stops.
The numbers are smaller than the story
This is where the tidy version starts to fray.
A team led by Marissa Nivison and Glenn Roisman went back to two of the largest long-running studies in the field, wrote down exactly how they would analyse the data before they looked at it, then published whatever came out. Once they accounted for other obvious factors, a baby’s reaction during that test explained only a sliver of how the child turned out later, roughly one to two percent of the variation in their social and emotional adjustment as adults. Small, but genuine, and it wasn’t a one-off. It lines up closely with an earlier pooling of dozens of studies by Ashley Groh and colleagues, which found early attachment explained a similarly small share, under four percent, of how well a child got on with peers.
Two other findings in the Nivison paper are more interesting than the deflation.
Infant attachment predicted objective academic test scores more strongly than it predicted emotional and social adjustment, which is the reverse of what attachment theory had long expected. Other work by the same research group, tracking maternal sensitivity directly across children’s first three years, found that those repeated observations forecast later outcomes considerably better than the twenty-minute test did.
What the test was actually sampling
That second result reframes the whole business. The Strange Situation is a photograph. What travels forward into adulthood is not the photograph but the thousands of hours of ordinary responsiveness it caught a single frame of, and that kind of slow accumulation is a much harder sell as a headline.
So nobody in that room was being sentenced. They were being sampled, briefly, on the strength of what one adult had already taught them to expect.
Which means the interesting question was never what a twelve-month-old did when the door opened. It is what the person walking back through it had been doing for the previous three hundred and sixty days.
