Is Batman Working Hard?

You are four years old. Somebody has brought you to a quiet room on a university campus, and everyone has been very nice about it, and now they have gone.

In front of you is a computer. The job is easy and it is extremely boring. When a piece of cheese appears, press the space bar. When a cat appears, press nothing. Cheese, cheese, cat, cheese. Then a wait. The wait is a second and a half long, which at four years old is roughly a geological era. Cheese again.

Across the room, on a low stool, there is an iPad. It is switched on. It is running a game about a small alligator who wants a bath, and the water has to be got to him through pipes, and it is one of the best games in the world. Nobody has told you not to play it. In fact they said, quite clearly, that you may go and play it whenever you like, for as long as you like, and then come back, or not.

You have ten minutes. And once every minute, from a small speaker beside the computer, a recorded voice asks you one short question.

Here is the entire experiment. That question — the same question, asked once a minute, for ten minutes — comes in three versions, and which one you get was decided before you walked in.

The children in version three got to choose a character first — Batman, Bob the Builder, Rapunzel, or Dora the Explorer — and then they got the costume. A cape, if you chose Batman. And for ten minutes a voice asked, once a minute, whether Batman was working hard.

The three versions are not three different questions. They are the same question at three different distances from you.

What happened, across a hundred and eighty children, was this. The further the question got from me, the longer children stuck with the boring job. Ask a child “am I working hard?” and they worked least. Use their own name and they worked more. Put them in a cape and ask about Batman and they worked most — a straight line running away from the self.

Not a big line. Across every condition, children spent about sixty-three per cent of the ten minutes on the iPad, because they were children and it was an iPad. Six-year-olds worked far more than four-year-olds. But the trend was real, and it survived when the researchers controlled for the children's vocabulary, their executive function and their theory of mind. It did not matter how clever you were. It mattered where you were standing when you asked.

The obvious objection is that four-year-olds will do almost anything if you give them a cape. So here is the version for people old enough to be embarrassed.

In a psychology building at the University of Michigan, a person is told they have five minutes to prepare a speech about why they are qualified for their dream job. They may not write anything down. They will deliver it to two evaluators who have been instructed not to smile. It is a standard laboratory device for making an ordinary adult miserable, and it works.

Half of them are told to spend the preparation time working through their nerves using the word I. The other half get the same instruction with one small change:

That is the whole intervention. One pronoun. Judges who did not know which group anyone was in rated the name-users' speeches as better. The name-users felt less shame afterwards. And — this is the part that stays with me — they spent less of the following days re-running the speech in their heads. The technical phrase is post-event processing. The ordinary phrase is lying awake at 2 a.m. thinking about the thing you said.

Later work put people in a brain scanner and found that the effect arrives fast — within the first second of seeing something upsetting — and, oddly, without the fingerprints of effortful self-control. It does not look like gritting your teeth. It looks more like the picture simply arriving at a different size.

There is an old and slightly unkind observation about King Solomon: a man so wise that rulers travelled for months to put their disputes in front of him, and whose own household was a disaster. The book of Proverbs, which tradition assigns to him, contains the line “The way of a fool is right in his own eyes.”

In 2014 two researchers gave this a name and then tested it. They handed nearly seven hundred people a scenario — a partner has been unfaithful with a close friend — and asked them to reason it through. Half were told it had happened to them. Half were told it had happened to someone else.

Then the answers were scored for four things that show up whenever people reason well about a mess: do they admit the limits of what they know? do they look for a compromise? do they consider how it looks from other seats? do they allow that things might change or go several ways?

People reasoned much more wisely about someone else's disaster than their own. Then the researchers did one more thing: they asked a third group to think about their own disaster in the third person — using their name, asking “why is she feeling this way?” — and the gap closed. Not narrowed. Closed.

Which is a strange and rather cheering thought. The wisdom was never missing. It was in the room the whole time. It just could not see anything from where it was sitting.

All of this assumes there is a voice to change the pronoun of. Which raises a question that sounds childish and turns out not to be: when you think, does it come in words, and can you hear them?

In 1990 a researcher tried to measure the speed of inner speech. People solved problems silently, reported the clipped little fragment they'd actually thought, and then unpacked it into the full sentence it stood for. Unpacked, the rate came out at over four thousand words a minute — roughly twenty times faster than anybody can talk. It is a wonderful number and it is worth knowing exactly what it is: not a recording of your head, but a reconstruction of it, and the fragments were much shorter than the sentences they were expanded into.

In 2024 two researchers reported something stranger. They split ninety-three people by how much inner speech they said they experienced, and gave both groups four tasks. On two of them — remembering lists of words that rhyme, deciding whether two pictures rhyme — the low-inner-speech group did worse. On the other two, no difference at all. They proposed a name for having little or no inner voice: anendophasia.

And then, in 2025, the same journal published a short reply arguing that they had proved no such thing — that scoring low on a questionnaire about your inner voice is not the same as not having one, and the word had run ahead of the evidence.

Both of those are in the record. That is what a live question looks like while people are still arguing about it, and it is a much more honest picture of your inner life than any quiz that tells you what kind of thinker you are.

Here is the question to walk out with. If the kindest, most sensible, most patient version of you is the one that only ever comes out for other people — where has it been all this time, and what would it take to get it to answer to your own name?