Two people sit across a kitchen table with a phone propped between them, reading questions off a list. Would you like to be famous? What would constitute a perfect day for you? When did you last cry in front of another person? They have been told — by a friend, by a podcast, by half the internet — that if they make it through all thirty-six, they may well fall in love. Somewhere around question twelve, one of them wonders whether the study behind this actually promised anything of the kind.
It didn’t. We went back and read the original paper in full, and the distance between what it says and what it has come to mean is one of the cleaner examples we know of how a careful piece of research becomes a myth — not through fraud or error, but through a parenthesis, an essay, and twenty years of retelling.
A methods paper, not a love story
The list comes from a 1997 study by Arthur Aron, Edward Melinat, Elaine Aron, Robert Vallone and Renee Bator in Personality and Social Psychology Bulletin. Its title gives the game away: “The Experimental Generation of Interpersonal Closeness: A Procedure and Some Preliminary Findings.” This was a methods paper. The team wanted a way to create a feeling of closeness between strangers in a laboratory, on demand, so that relationship questions that had only ever been studied by observation could finally be tested as experiments.
Their tool was structured mutual self-disclosure: pairs of strangers worked through three sets of slips — each carrying a question or a small task — over 45 minutes, each set more personal than the last, from dinner-party hypotheticals up to “share a personal problem and ask your partner’s advice.” In the first study, pairs who got the escalating questions reported feeling markedly closer afterward than pairs who spent the same 45 minutes on small-talk tasks — describing their last trip to the zoo, naming a favorite holiday. The difference was large by the conventions of the field.
Notice what is missing. Nobody measured romantic love or romantic interest — the closest the paper comes is a two-item liking scale in one study, and even liking turned out to be unmoved by the study’s matching and expectation manipulations. Many of the pairs were two women assigned to each other in a psychology class. The authors state plainly that the goal was “a temporary feeling of closeness, not an actual ongoing relationship,” and they compare their procedure to a mood induction — a way of switching on a state so it can be studied, not a way of changing anyone’s life.
Where the love story came from
The romance entered through a single parenthetical line. Describing an earlier, longer pilot version of the task, the authors mention “anecdotal reports of the impact of the experience over the next few months (including one pair who married!).” One pair, in a preliminary study, noted in passing.
That parenthesis slept quietly for eighteen years until January 2015, when the writer Mandy Len Catron published a Modern Love essay in The New York Times called “To Fall in Love With Anyone, Do This,” recounting how she and an acquaintance tried the questions — and did fall in love. The essay was read millions of times, becoming one of the paper of record’s most-read pieces of that year, and the “36 questions to fall in love” were born. Even the four minutes of sustained eye contact that most versions now attach to the exercise comes from the essay’s retelling, not from the 1997 procedure — there is no eye-gazing task anywhere in the paper. The only four minutes in the original is slip eleven: “Take 4 minutes and tell your partner your life story.”
Catron never claimed to be reporting science; her essay is honest about being one story. The claim hardened downstream, in the thousand summaries that dropped her hedges and kept her wedding-adjacent ending.
The finding, and the quieter one behind it
Read in full, the paper’s real findings are stranger and more interesting than the legend. The closeness people reported after 45 minutes was substantial: on the study’s main scale, the average stranger-pair rated their hour-old relationship as closer than the closest relationship in the lives of about 30 percent of comparable students — a comparison the authors themselves flag as imperfect, since people may use rating scales differently for an hour-old acquaintance than for a sibling. Seven weeks later, 57 percent of pairs in one study had had at least one further conversation, and around a third had done something together.
The quieter surprise sits in the second and third studies. The researchers tested whether the machinery around the questions mattered — matching pairs so they agreed on important attitudes, telling people their partner was expected to like them, making “getting close” an explicit goal. None of it made a measurable overall difference, despite the studies being designed with good power to detect it — the one wrinkle was that making closeness an explicit goal did seem to help introverts specifically. The only ingredient that clearly mattered overall was the task itself: escalating, reciprocal, personal disclosure. Not compatibility. Not expectation. Just two people taking turns being progressively more honest.
What forty-five minutes cannot buy
It would be easy to swap one overclaim for another here, so let us be precise about the limits. This was a set of preliminary studies, mostly on university students, with self-reported closeness measured minutes after the interaction; the authors called their theoretical findings “highly preliminary” themselves. The follow-up numbers had no control group — some classmates would have talked again anyway. The closeness-generating procedure has since been used widely in relationships research as a standard tool, which speaks to its reliability for its intended purpose, but the popular claim — that the questions produce romantic love — has never been a research finding at all. It is one anecdote from a pilot study plus one beautifully written essay. And the study says nothing about whether lab-built closeness lasts: the authors are explicit that loyalty, dependence and commitment “might take longer to develop,” which is a researcher’s way of saying that 45 minutes is 45 minutes.
We should also say what we are: writers who read studies closely, not psychologists or therapists. What follows is our reading of one research program, not counsel about anyone’s relationship.
What actually travels beyond the lab
The questions cannot make anyone fall in love, and treating them as a spell tends to end in a slightly awkward evening. What the research does support is smaller and more portable: closeness between two people is not only a matter of luck or chemistry but partly a matter of structure — of whether the conversation is built so that both people reveal themselves, gradually, and in turn. That holds for strangers in a classroom, and it is at least suggestive for the marriages and friendships where conversation has flattened into logistics. If distance in a relationship has become painful rather than puzzling, a couples or family therapist is the right room for that conversation — a list of questions is not an intervention, and it was never designed to be one.
The 36 questions were never a recipe for love. They were a demonstration that intimacy has a mechanism — and the mechanism is still available to anyone willing to answer honestly and then ask.
