“Half of France supports” is not a statistic until you tell me whose poll it is

France 24 can do numbers like a grown-up. It proves it in the same protest coverage where it asks you to swallow a different number whole.

On the French high-school protests, it reports that “141 high schools remained closed” — and it gives you the basic handles: France’s Ministry of Education is the issuer, and the count is as of Friday. That’s ledger prose. You can audit it, dispute it, ask what “closed” means in practice.

Then, in another story on the same protests the same day, France 24 tells you that “half of France supports” the movement.

Half is a number that wants to settle an argument — not about logistics, but about legitimacy. And it arrives with none of the things that make polling a measurement rather than a mood:

• which pollster
• field dates
• sample/method
• question wording
• what “supports” means (approve in principle? back the blockades? sympathise but oppose disruption?)

That isn’t “missing context.” It’s a pseudo-operand: precision theatre. A percentage costume on an unaudited claim.

I’m not calling this a pattern. I’m describing what’s on the page today: France 24 alternates between audit-ready numerals and vibe numerals on the same story — and the one that would most need its keys is the one that doesn’t have them.