Dr. Paul Eastwick's work on attraction and compatibility does something quietly radical: it demonstrates, across dozens of studies and thousands of participants, that stated preferences — the lists people carry into dating of what they think they want — have almost no predictive power once a real person is standing in the room. The traits you swore mattered dissolve on contact with an actual face, an actual laugh, an actual nervous habit. What predicts attraction instead is stranger and less controllable: proximity, repeated exposure, the slow accumulation of what he calls 'relationship experience' that cannot be simulated in advance. I want to take this seriously — not as a cute finding about dating but as something that has consequences for how we understand human knowledge itself. Because what Eastwick has shown, rigorously and repeatedly, is that there is a category of self-knowledge that does not exist until the thing it refers to is already happening.
You cannot know what will move you before you are moved. This is not ignorance. It is the structure of the phenomenon.
The current discourse frames distrust between men and women as a wrecking ball aimed at the cathedral of society. But societies are not cathedrals. They are weather systems. You don't destroy a weather system — you perturb it into a different regime. The question that matters is never who is swinging; it's what made the atmosphere unstable enough that any nucleus could trigger condensation.
And Eastwick's research, without intending to, identifies one source of that instability: the structural mismatch between what people say they want and what proximity and time actually produce in them. When an entire culture builds its romantic infrastructure — its apps, its advice columns, its billion-dollar matching industry — on stated preferences that the data show are predictively worthless, you have created a system that systematically disappoints. The gap between declared desire and lived experience is the humidity. It was always there. The apps just made it measurable.
Trillions of neutrinos pass through your body every second without interacting with a single atom. Most of what moves between people — profiles swiped past, glances on the streetcar, conversations that open and close like automatic doors — passes clean through without collision. Eastwick's finding that repeated exposure and physical co-presence matter more than preference profiles is, beneath the methodology, a finding about rarity. About how few of our encounters are encounters at all. The moments of genuine contact are so improbable they cannot be optimized into existence.
They can only be made more likely by a condition the research keeps pointing toward: unstructured time in shared space.
The algorithm assumes contact is the problem to be solved. But contact was never the problem. The problem is that we've built systems to manage collision while living in a famine of it.
I walked east along Queen Street past midnight — the Portuguese bakeries dark, the laundromats dark — and watched a man mop a convenience store floor in long unhurried arcs, the water darkening and wrung clear and darkening again, and I thought: this is what the research is actually about, underneath the regressions. Not optimization. Presence. The mop moves because the man is there and the floor is there and the hour asks nothing of either of them. Eastwick's data keep arriving at a version of this: that what moves us in another person cannot be known in advance because it is not a property of the other person.
It is a property of the interaction — of two bodies in a room on a specific Tuesday, generating something that did not exist before they were both there. And this means the not-knowing that precedes love is not a bug in human cognition. It is the physics of the thing. You cannot predict an emergent property from the components alone. You have to run the experiment.
You have to be in the room. The atmosphere has to be there — humid, unstable, charged — and then the particular rain happens or it doesn't, and no one scheduled it, and it is not repeatable, and this is not a failure of science. It is what science found.
