The 36 questions, and what they actually do
By God's Closet
They are known as “the 36 questions that lead to love”. The title is catchy and slightly wrong, which probably explains its longevity. The experiment does exist, it is published, and its mechanism is more interesting than the legend.
The 1997 experiment
Arthur Aron and colleagues published in 1997, in the Personality and Social Psychology Bulletin, a paper titled “The Experimental Generation of Interpersonal Closeness”. The starting question is methodological: can you produce in a laboratory, in under an hour, the closeness two people normally take months to build.
The setup is simple. Two strangers sit facing each other and receive a list of questions to ask in turn. The list holds three sets of twelve questions, and each set is more personal than the last. A control group receives a list of ordinary conversation topics instead.
At the end, participants in the first group report feeling markedly closer to their partner than those in the second. That is the finding, and it is more modest than its reputation: the study measures a feeling of closeness, not love.
One detail the legend lost along the way: the paper reports three experiments, not one. Only the first establishes the famous result. The other two asked whether the effect depended on being matched on opinions, or on knowing in advance that the other person would like you. Neither found anything significant. In other words, what brings people closer is neither similarity nor intent: it is the exchange itself.
Why the order matters more than the questions
The heart of the setup is not the list, it is the escalation. The first set asks whom you would invite to dinner if you could choose anyone. The last asks which object you would save from a burning house, and which death in your family would affect you most.
Between the two, each question licenses the next. This is the principle of reciprocal, gradual self-disclosure: nobody answers an intimate question asked cold, but almost everyone answers it after twenty minutes of an exchange in which the other has taken risks too.
- Reciprocal: both answer, in turn. A one-way interrogation produces nothing.
- Gradual: difficulty rises in steps. A question that is too personal too early makes people retreat.
- Framed: the list exists so neither person has to decide what to ask. That is what makes the exercise bearable.
The protocol has in fact served well beyond the romantic question. Researchers have reused it as a laboratory tool to create, in a single session, closeness between people who had no particular reason for it — notably in work on relations between groups, where the aim was to manufacture a budding friendship between participants from different backgrounds. That is a sign the mechanism interests research on its own terms, independently of what popular culture took from it.
What people wrongly make it say
In 2015, a newspaper piece made the experiment famous by recounting that its author had tried it with an acquaintance and that they had fallen in love. The text travelled everywhere, and two things attached themselves to it along the way.
The first is the headline: the study never showed that a list of questions makes people fall in love. It showed that it brings them closer, which is already remarkable for forty-five minutes in a laboratory.
The second is the four-minute mutual gaze, often presented as the protocol’s thirty-seventh step. It is not part of the protocol published in 1997: it comes from the 2015 piece, whose author adds it to her own essay. It is an interesting addition, but it is not the result of an experiment.
Ten questions rather than thirty-six
The Arsentis protocol takes the principle, not the list. It holds ten questions, designed for two people who already know each other, which changes everything, since half of Aron’s questions only make sense between strangers.
The duration changes too. Forty-five laboratory minutes become about thirty at home, with no stopwatch. That is where couples naturally stop, and it is plenty for the material an artwork needs.
There is nothing to get right. A body produces variations constantly, including when you think you feel nothing.
The protocol is a document to print, not an app to consult. A screen between two people talking is exactly what should be avoided, and it is also why no data is displayed during the capture: there is nothing to watch climb.
References
Aron, A., Melinat, E., Aron, E. N., Vallone, R. D., & Bator, R. J. (1997). The experimental generation of interpersonal closeness: A procedure and some preliminary findings. Personality and Social Psychology Bulletin, 23(4), 363–377. https://doi.org/10.1177/0146167297234003