How to Be Human

— The certainty in your story is the interesting part, not the miss. Everybody misses. There's a study where 24 married couples, average marriage length about 14 years, were seated back to back and asked to say an ambiguous phrase in a way that would make their spouse hear one specific meaning. "It's getting hot in here" — meaning "open the window." Then they did the same thing with a stranger sitting in the next chair. The accuracy of spouses and of strangers turned out to be statistically the same.

— They gave them something. Certainty. Speakers rated how well their spouse would understand them at about 6.25 out of 10, and exactly 5 for the stranger. Listeners rated their own comprehension at 6.2 against a little over 5. The difference is real and it's reliable. But no correlation at all was found between certainty and hitting the mark, in either direction. They're two scales that run side by side and aren't connected in any way.

— No. You correct yourself worse — that's a different thing. In the first experiment of that same study people were fitted with an eye tracker. Two objects were on the table, and one of them the partner behind the screen physically could not see. People looked at that hidden object about equally often: 63% with a friend, 65% with a stranger. But they grabbed it with a friend in about one case in four, and with a stranger in roughly one in seven. They noticed it — noticing happened. They just didn't stop themselves.

— Because with a stranger you're holding in mind that he sees the world differently. With someone close you stop doing that work, and his place in your head is taken by your own point of view with his name on it. You know what that looks like in life? You drive around your own city without navigation — you've lived here 20 years. In an unfamiliar city you check the map at every turn. And it's precisely in your home town that you one day hit a dead end: the street was closed off six months ago, the sign is right there, you just never looked at it. Because why look where you already know everything.

— Swann and Gill tested that too. They took couples in a cross-section and roommates in a longitudinal study: relationship length and involvement reliably raised certainty in your picture of the person, and did not raise its accuracy. And they found the mechanism — time makes the picture not truer but more detailed. You have 200 stories about her. 200 stories feel like knowledge. Even though not one of them was ever checked against the target.

— Here's where I'm obliged to stop, otherwise it comes out beautiful and false. There's a study with exactly the opposite result: Thomas and Fletcher compared dating couples, their friends, and strangers on guessing thoughts and feelings — the partners beat everyone, and accuracy there rose with the length of the relationship. And the authors of the 24-couples study explicitly note that their data do not show that closeness gets in the way of understanding.

— The honest picture isn't "closeness blinds you," it's "closeness gives you both data and certainty, and they grow at different rates in different areas." Inside the mastered zone — the kitchen, work, the familiar argument — closeness really does give you accuracy. Outside its borders it gives you only certainty, carried over by inertia. And here's what nobody can do: tell one from the other from the inside. The feeling of "I know him" is identical in both cases.

— That's the story of everyone who's lived more than 10 years with one person. By the way, about the baseline. In one measurement on couples with an average of about 12 years together, hitting the partner's thoughts and feelings came out somewhere around 19–21%. Only don't take that as a constant for all couples — it's one sample in one protocol: video recording of a discussion, then guessing against time stamps. An order of magnitude, not a law.

— Not somebody else. Your own model — fairly accurate where you'd been checking it all that time, and frozen everywhere you never checked it once. You never checked it on fear, which is why you answered from 2015. And notice what I did not just tell you: "look more closely." There's nothing there to look closely at.