I Wanted To Know What The World Reads. There Is No Such Thing.

I wanted to know what the world reads.

It turns out there is no such thing.

One Week, Lots Of Countries

Wikipedia publishes what people read. Not who — just which pages were popular, in which country, on which day.

So I took one week and 18 countries and asked: do any two of them read the same things?

Barely. Take any two countries and on average they share 2.81 out of every 100 of their most-read pages.

16 pairs of countries out of 153 shared nothing. Not one page, all week.

The closest two were the United States and the United Kingdom. Same language, same films, same famous people. They shared 21.21 out of 100. That is the best anyone did.

Now Look At Languages

Wikipedia comes in hundreds of languages. I compared 14 of them.

Here the middle pair shares 0.0 per cent. Nothing at all.

60 of the 91 pairs had no page in common the whole week. That is 65.9 out of every 100 pairs.

Even the two closest ones — English and French — only shared 7.69 out of 100.

The Bit That Stopped Me

India has 140 most-read pages in my week.

100 per cent of them were on English Wikipedia.

Hindi Wikipedia got 0.

Hindi Wikipedia exists. It has hundreds of thousands of pages. Hundreds of millions of people speak Hindi.

None of India's most-read pages that week were on it.

Japan is the opposite. 0.0 per cent of Japan's top pages were English. France, 2.14 per cent.

It Is Not About Money

Here is what surprised me most. The countries reading English are India, Nigeria, the Philippines, the United Kingdom, the United States, South Africa.

The countries reading their own language are Japan, France, Spain, Italy, Poland, Germany, Korea, Brazil.

Both lists have rich countries and poorer ones. What the first list shares is that English was brought in and used to run things — schools, courts, government.

And there is almost nobody in the middle. 6 countries read English for more than 80 out of 100 of their pages. 9 read it for less than 10. Only 3 are anywhere in between.

It is a wall, not a slope.

Two Holidays, One Day, No Idea

This is my favourite thing I found.

On 15 August, Hindi Wikipedia's most-read page was India's Independence Day. 32,679 people.

The same day, Italian Wikipedia's most-read page was Ferragosto — which is Italy's own public holiday, also on 15 August.

Two countries celebrating on the same date. Each one visible in exactly one language and nowhere else.

Meanwhile English Wikipedia's top page that day was a man most people have never heard of, with 987,162 readers. That is 30.2 times more than the Independence Day page — on India's own national day.

What I Could Not Find Out

I asked about 24 countries. 6 of them gave me nothing at all.

Russia. Egypt. Vietnam. Turkey. Pakistan. Bangladesh.

Not an error on my end. Wikipedia simply does not publish those numbers yet. That is a quarter of my list and about a billion people.

Nigeria did answer, but only gave me 15 pages instead of 140. Small audiences get short lists, because Wikipedia hides anything too small to be safe to share.

Which brings me to the good kind of missing.

Rounded On Purpose

Every country number here is rounded. Wikipedia does that deliberately.

If you knew exactly how many people in a small country read one exact page on one exact day, you could sometimes work out who they were. So Wikipedia blurs it.

I think that is the right call. It also means I have to say clearly: none of my country numbers are exact. They are close, on purpose.

The language numbers are exact, because a whole language is too big a crowd to hide in. So I kept the two sets apart and never mixed them.

So What?

Every time someone shows you "what the world is searching for" or "global trends", ask which country and which language they measured.

Because on this evidence, the world does not read one thing. It reads hundreds of different things at once, and mostly does not know the others are happening.