Your Closest Friendships Were Decided by a Seating Chart
Beneath the Surface · By Ali Daudji ·
There is someone you met at another person's birthday dinner two years ago. You ended up beside each other because of where the coats went, and you talked past the jobs and the commute into something you had not intended to say to a stranger. At the door you put your number in their phone. You both said we should do this again, and you both meant it.
Neither of you did.
I have a collection of these now. People I liked immediately and without reservation, then never saw again. The story I have always told myself about it is a story about effort. Adult life is full and somebody has to text first, and neither of us was ever going to be that person. On my more honest days I have called it a defect in me, a small failure of follow-through. I no longer think that is what happened.
In the fall of 2017, three researchers went into forty primary schools in rural Hungary. They took 182 classrooms, close to three thousand children between the ages of eight and seventeen, and randomized the seating charts. The classes and the schools stayed exactly as they were. What changed was who sat beside whom at the two-person desks, and only in three school subjects, which came to somewhere between a quarter and a half of the week. Then they left it alone for a semester. In the spring, the children were asked to name up to five best friends in their class.
Julia Rohrer, Tamás Keller and Felix Elwert published the result in PLOS ONE in 2021. Sitting next to someone raised the chance of a mutual friendship, both children naming each other, from 15.3 percent to 22.3 percent.
Seven pairs in every hundred. I have not been able to put that number down.
These children already knew each other. Most of them had been in that room together since first grade, had looked around long ago and sorted everyone into the ones who mattered and the ones who did not, and the sorting was finished. Then an adult with a spreadsheet moved a desk, and seven pairs in every hundred ended the year naming each other as best friends who otherwise would not have.
The friendship had been sitting there the whole time, a row away, for years. What stood between those two children was furniture.
They were children, and I am not. Childhood is soft, and eight-year-olds have time and no mortgage. It would be convenient to leave the finding there.
Except that those rooms were full of seventeen-year-olds as well as eight-year-olds, in classes that had been together since the first grade, with a social order nobody had questioned in years. The authors put it plainly in their conclusion. Friendship networks stay malleable long after the friendships inside them have been established. The children were not the pliable part. The seating was.
I cannot prove that carries into adult life, and the study does not claim it does. What I would say is that the difference worth arguing about is not how open we are. It is how much of this we get. Those children spent months in a room they did not pick, on a schedule nobody let them cancel. The proximity was the standing condition of their lives, and all the researchers did was move it a few feet. We have stopped being assigned anywhere at all.
Every close friendship I have came out of a situation somebody else arranged, and I have never heard anyone describe theirs differently. Everybody's list is a list of rooms. A residence floor, a first job with assigned desks, a course someone signed up for half by accident. Afterward we all credit recognition, how fast we understood each other. I don't think that part is false. But the recognition had months of enforced nearness underneath it and would not have arrived without them, and there were other people in those same buildings who would have been understood just as fast, if the seats had been handed out differently.
The friendships I have do not feel smaller for knowing this. The ones I never had have started to feel specific, which is worse.
It is also where I stop believing the advice. The advice is addressed to the will. Be intentional. Prioritize your friendships. Be the one who reaches out. I have taken it seriously for years, and it produces roughly one pleasant dinner every four months with a person I would like to know far better than I do.
It fails for a reason that has nothing to do with how much anyone meant it. Contact you have to choose and contact you cannot avoid are not the same material. Anything you have to decide to do, you have to decide again every time, and a decision that repeats will eventually meet a bad week, or a stretch where you are not at your best and would rather not be seen that way. Deliberate friendship has to survive your worst month by being actively chosen during it. Most of it doesn't.
The Hungarian classroom asked nothing of those children, and prioritizing each other never came up. There was a desk, and it was Tuesday.
The study is honest about its limits and I should be too. When boys were seated next to girls the effect shrank hard, to about two percentage points. Within a gender it ran to nine for girls and thirteen for boys. A desk was not enough to cross that line.
That is the finding, not the exception to it. What beat proximity was another structure, older and better installed than a seating chart could reach. Willingness had nothing to do with it. Nearness opens a door. It does not walk anyone through it, and it cannot undo something already standing in the way. It is a precondition, and preconditions are dull. So we hurry past them to talk about chemistry and whether we clicked.
I have stopped asking who my people are, as if they exist somewhere already formed and the work is to go locate them. The question I ask now is duller and more useful. Where in my week am I near the same people repeatedly, without either of us having arranged it?
For most adults the honest answer is almost nowhere, followed by a short list nobody has ever examined. The neighbor two doors down. The same six faces at the same Wednesday class. We read them as scenery, because they arrived with no introduction and no moment where anyone decided they mattered. They are also the last people most of us see on a schedule we did not set.
I do not find that consoling. Somewhere in the last ten years there were people I stood near every week who would have been, if the rooms had fallen differently, among the three or four I call when something goes wrong. I never learned their names.
What I notice is that I still do not want to. The people already nearest me do not register as candidates. They arrived without anyone choosing them, and something in me treats that as a disqualification rather than as the mechanism itself. I have understood this for exactly as long as it took to write it down, and it has not yet changed anything about my Wednesday.