Picture the moment a friendship actually began.
Most people, if they think hard enough, land on a neighbor, a coworker two desks over, a regular at the same Tuesday spin class. Not a soulmate discovered through long conversation. Not someone with a perfectly matching personality. Just someone who kept showing up in the same physical space. The question researchers have spent decades circling is deceptively simple: why does that keep working?
The brain bias that builds a bond before words are exchanged
The mechanism is called the mere exposure effect, and it operates almost entirely below conscious awareness. The mere exposure effect refers to the tendency for people to develop a preference for things simply because they are familiar with them, even if they have no specific information about the merits of those things. In the social world, that stimulus is another human face.
Repeated, unreinforced exposure to a stimulus is sufficient to enhance its affective standing: repeated exposure leads to familiarity, and familiarity leads to liking. So every time a neighbor passes on the stoop or a coworker fills a coffee mug at the same moment, the brain is quietly upgrading its rating of that person, even if no word is spoken.
This works primarily through repeated unplanned encounters. A scheduled coffee date demands social energy and decision-making. Passing someone in a hallway costs nothing and still counts.
What the numbers look like when researchers mapped real buildings
The principle gained rigorous empirical grounding in 1950 when social psychologists Leon Festinger, Stanley Schachter, and Kurt Back examined friendship formation among 240 married students living in MIT’s Westgate housing complex, a community where residents were largely similar in background and life goals, and apartment assignments were essentially random. Their findings were striking: 65% of reported friendships formed among residents living within five doors of one another, with friendship rates dropping sharply as physical distance increased. That is not five blocks. Five doors.
The study revealed a second force beyond raw physical distance: functional distance. Some apartments were more likely to be passed by, if a unit sat at the bottom of a staircase leading to the second floor, all upstairs residents had to pass that door, both physical and functional proximity shaping how often any given resident would be seen by the others. Residents were much more likely to become friends with people on the same floor than with even very similar people on a different floor, and the only residents who had friends across multiple floors were those who lived next to the stairwells. The stairwell was a social infrastructure feature as much as a fire-safety one.
The same Westgate data sharpened the picture at finer grain: 41 percent of next-door neighbors reported being close friends, compared to 22 percent of those two doors apart, and only 10 percent of those at opposite ends of the hall. Each additional door was a measurable tax on the odds of friendship forming.
How classroom seating experiments confirmed the same rule
Because housing assignments are hard to randomize at scale, researchers turned to classrooms, where seating charts give them a cleaner lever. A large pre-registered field experiment randomized seating in 182 third- through eighth-grade classrooms (2,966 students) for the duration of one semester, finding that being seated next to each other increased the probability of a mutual friendship from 15 to 22 percent on average.
A study of 235 primary school students in grades 3 through 5 found that students sitting next to or nearby one another were more likely to be involved in reciprocated friendships, and that classroom seating proximity was associated with the formation of new friendships over time. That result held even when researchers accounted for shared interests and personality overlap.
Friendship networks are strongly shaped by homophily, the tendency to form relationships with those who resemble us in gender, age, and background. The large randomized experiment found that induced proximity increased the propensity toward friendship equally for all students regardless of educational achievement, gender, and ethnicity, though the probability of an actual manifest friendship still increased more among similar students, a pattern mainly driven by gender. Proximity raised the odds of friendship across social lines that gravity usually keeps in place, without erasing the pull of similarity entirely.
Where the mechanism breaks down and what it costs adult social lives
The propinquity engine is powerful, but it needs fuel: regular, repeated contact in a shared physical space. School supplies it automatically. Office work supplied it through shared kitchens and hallway collisions. Remote and hybrid arrangements have removed both.
Research suggests that bonds typically need a consistent run of encounters before a connection registers as real to both parties. Without a structure that enforces that repetition, most adult acquaintances never clear the bar.
As one researcher in this field has noted, committing to the same recurring event, Tuesday night volleyball or a Wednesday morning class, gives people the consistency that allows recognition and memory to build.
That gap shows up in the data on cross-generational connection. Only 22 percent of US adults have a close friend from a different generation, a number that reflects, in part, how rarely different age groups now share the same recurring physical spaces. And a panel study on dating apps found that people who treat repeated digital contact as a substitute for real connection do not experience the same warmth that genuine propinquity delivers. Scrolling a profile photograph does not fire the same familiarity loop as seeing someone in person: researchers are still asking whether repeated digital contact can create the same sense of familiarity as physical presence, or whether the body in space still holds a unique place in how relationships develop.
What people can deliberately do with a mechanism they already carry
The good news is that the engine does not require charm, social skill, or even a gregarious personality. The most reliably replicated finding in friendship research is that consistency of exposure matters more than the specific context. Joining a recurring activity creates the conditions that make friendship formation more likely.
Researchers point to what sociologists call third places, spaces that are neither home nor work, as the adult world’s best remaining propinquity engines. A gym, a library reading room, a community garden, a neighborhood bar: each is a stairwell in disguise, supplying the repeated unplanned contact that closeness requires.
None of this makes personality or shared values irrelevant. Research identifies key maintenance behaviors that sustain close friendships: being positively engaging, being supportive, being open, and maintaining regular interaction. But those deeper forces only get a chance to operate once proximity has done its opening work. The brain decides it likes a face long before the heart decides it trusts a person.
