Are You Seeing Them, or Your Guess About Them?

Seema is certain that Lalit is angry.

He put his keys down harder than usual. At least she thinks he did. When she asked about dinner he said “hm”. Now she is watching his jaw.

By nine she has gone quiet and clipped. He looks up and asks what is wrong with her, and there is an edge in his voice. So he is angry. She knew it.

A brain that guesses first

The philosopher Andy Clark’s Surfing Uncertainty is a book about how brains work, and it contains almost nothing about marriage. It does contain an account of Seema’s evening.

The older picture of perception has the world pouring in through the senses and the brain assembling it. Clark sets out a newer theory in which the traffic runs mostly the other way. The brain is constantly predicting what it is about to see and hear, and it uses what arrives chiefly to check the guess. We tend, he says, to see what we expect.

He gives a neat demonstration. Two researchers told students that a recording of noise had a faint snatch of the song “White Christmas” buried in it. Almost a third reported hearing the song begin. There was no song.

How much to trust the guess

A guessing brain needs a second skill. It has to judge how far to trust its expectation and how far to trust the incoming evidence. Clark’s example is a driver. On a familiar road in thick fog you lean on what you know. On a strange mountain road you lean on your eyes.

On this theory, attention is that adjustment. To attend to something is to turn up the weight given to one stream of evidence. Because the budget is limited, other streams are turned down.

Jack and Jill

Clark borrows a scenario described by Susanna Siegel. Jill believes, for no good reason, that Jack is angry with her. When she looks at him, the belief does its work. Faint signs that could be anger are given extra weight. Ordinary signs that he is fine are treated as noise. Clark is precise about the result. Jack looks angry to her. This is in the seeing itself and is not a judgement added afterwards.

Then her eyes go hunting, over his face, his shoulders, his choice of words. A guess directs the search for its own proof.

And Jack is the same kind of creature. He notices that she is tense. He wonders whether she is annoyed with him. Now she looks a little angry to him, and his own signs of strain may by this point be real. Clark’s phrase for where this leads is “self-fulfilling psycho-social knots and tangles”.

How solid is this?

I should be straight about the evidence. Jack and Jill are an illustration and not an experiment. Clark cites one study, by Teufel and colleagues, showing that what we believe about another person’s state of mind alters how we perceive their gaze and their movements. The wider theory is one he presents as promising and unfinished.

Still, most couples will recognise the knot.

What follows is mine, not his

Clark offers no advice to couples. He does make one remark that I find encouraging. Because the brain can also turn up its trust in the evidence, “we are not slaves to our expectations”.

So here is what I would try. Treat the certainty as a guess. Describe the evidence to yourself in bare physical terms: keys on a table, one syllable. And then get evidence of a kind that is harder to bend, which means asking him.

Had Seema asked at seven, she might have heard about a bad meeting. By nine there were two angry people in the room, and one of them had been made.