GHOSTWRITER SERIES: PSYCHOMETRICS

Binet on the test with no comment box

For this edition of Ghostwriters, I picked Alfred Binet, the French psychologist who, with Théodore Simon, created the Binet-Simon scale in 1905 – one of the tests that shaped modern intelligence testing. Why Binet? Because I do think psychometric tests can be useful - and I do not underestimate the range of psychometric testing - and they can bring structure to something notoriously messy: people. But every now and then a question appears with two boxes and no room for context, and you stare at it wondering when you became a brick.

Binet built his test to help identify children who needed support at school. He also knew – rather inconveniently for everyone who enjoys a tidy label – that a test result is not a soul, a ceiling, or a verdict.

Which makes him a useful ghost to have in the room when the question has two options, no comment section, and about 30 seconds in which to compress your entire inner life into A or B.

The prompt I gave it:

“Write a piece about psychometric tests in the voice of Alfred Binet, speaking from beyond the grave.

He helped create the Binet-Simon scale to identify children who needed support, not to turn people into permanent labels.

He should defend what tests can do well, but be tired, kind and funny about what they get wrong - especially forced either/or personality questions with no comment box.

Somewhere include a question like: “Would you rather solve a problem and be late for a meeting and waste people’s time, or arrive on time for the meeting?” Because, surely, it depends what’s on fire.

What it wrote:

I am Alfred Binet. I have been dead since 1911, which gives a man perspective.

In 1904, France made a commission, because schools had a problem and adults like to put chairs around a problem before they put help around a child.

Some children in the Paris schools were falling behind. Which ones? How far? And what sort of help did they need?

Could we find out without simply guessing?

So Théodore Simon and I built a scale. A rough one. A practical one. We sat with children. We asked them to follow a lit match with their eyes, to repeat numbers back to us, to tell us why a foolish thing was foolish. We tried to understand what a child could manage now, not what he was worth forever.

Later, we arranged the work more carefully by age. What could a child of six usually do? What could a child of nine usually do? If a child was struggling, we wanted to know.

Not to shame him.

To put a teacher beside him.

To throw the child a rope.

That was the whole idea.

A rope.

Listen. The test was a rough thing. A stick for measuring a river that will not hold still. I said so at the time, loudly, in French. I said: this result is where the child stands today. It is not what he is. It is not his ceiling. It is not his soul.

With patient teaching, a child can grow. His attention can grow. His memory can grow. His judgement can grow. I called this mental orthopedics, which is a dreadful name and sounds like something done to the inside of a shoe, but I meant well.

Then I died.

And here is the funny part.

By which I mean the part that is not funny at all.

After I died, clever men in other countries picked up my rough little stick and decided it was a sacred ruler. They boiled a whole child down to one number. They called the number his intelligence, as if intelligence were a single thing, like a shoe size. They wrote the number on him and the number would not wash off.

They used versions of it in schools, institutions, armies, and even at the dock, because apparently a person arriving after an ocean crossing is precisely what the world needed to reduce to a number.

I had made a tool for finding the children who needed help.

They turned it, too often, into a machine for deciding who was worth helping.

That is the trouble with handing the world a number. The world stops looking at the person.

Now they show me the tests you sit for your jobs, and I confess I am baffled.

There is one I keep returning to.

It asks whether you would rather solve the problem and be late for the meeting, or arrive on time and let the problem burn.

There is no box for: what is the meeting?

There is no box for: what is the problem?

There is no box for: is the entire board waiting, or is it one man with a spreadsheet and a very dry muffin?

These are not the same late.

I spent my life cataloguing the thousand small ways one mind differs from another, and you have built a test that gives a person two doors, thirty seconds, and no handle marked “it depends.”

Then you call whatever falls out his character.

This is bold.

Not wise, perhaps.

But bold.

And yet.

I will not pretend the tests are nothing. Used gently, a test can be a useful thing. It can catch what the eye misses. It can give structure where people are guessing. It can help a busy teacher, manager, doctor, or team see something they might otherwise overlook.

A test used well is a question.

Not an answer.

A beginning.

Not a sentence.

The fault was never only in the test. The fault is the full stop. It is what people do the very moment they are handed a result: they exhale, and they stop looking.

So if you must sit their test, sit it.

Tick the box.

Be on time, or solve the problem, whichever you can live with that morning.

But do not let the result off the page. Do not let it follow you home and sit at your table. It is a photograph of one Tuesday, taken badly, in poor light, by someone who had forty-four others to get through before lunch.

I built a ruler. I never once meant it as a verdict.

— Alfred Binet (1857–1911), still trying to get a word in

Marking the homework:

Credit first, because it earns some. The dock line is real history and the AI's right to reach for it - reducing a person who's just crossed an ocean to a number is exactly the thing Binet would've hated. "These are not the same late" is good. And "one man with a spreadsheet and a very dry muffin" made me laugh, which I wasn't expecting. The muffin's doing a lot of work. It's specific, and specific is the one thing these tests never are.

But the tell this time is structural, and you can see it on the page.

Look at the line breaks. Nearly every beat gets its own line. "A rope." "Then I died." "This is bold." "Not wise, perhaps." "But bold." Any one of those, on its own, would land. All of them together, none of them do. The AI's worked out that a short line standing alone feels weighty, so it does it to everything, and when everything's emphasised nothing is. That's the evenness again. Usually you only hear it. Here you can see it. There's no sense of which moment actually deserves the silence around it, so it hands the silence out to all of them equally.

And then the thing the whole series is about. It never once risks being wrong. Every bit of wisdom is safe. "A question, not an answer." "A beginning, not a sentence." All true, all unfalsifiable. The actual Binet had a named enemy and a temper about it. This one's cross with "clever men in other countries" and won't say who, or why. It's got all his kindness and none of his fight. It even tells you it's being kind - "I meant well," "not to shame him" - and when a thing has to label the feeling, it's usually because it doesn't trust the moment to carry it. (Having said that, my prompt did ask it to be kind - so this is partly my fault…)

The voice is all there. The risk is the part it can't fake.