Policy corner: Do we know a thing happened, or only that someone said it happened?

Sometimes we have to be careful how we, in a headline, declare that something happened.

For the front page of our Aug. 24 edition, we ran the news that Yevgeny Prigozhin, the leader of the Wagner mercenary group in Russia who led a brief armed rebellion against the Russian military earlier this year, was killed - killed in a plane crash, is how we said it in a first draft of the big headline.

But there was a catch. The news came from Russia's civil aviation agency, which cited the airline. And the news was greeted with skepticism: Was it really him? And was it really a simple crash?

In the end we changed the main headline to be "Mercenary boss said to be dead." We would learn over subsequent days that some Western officials believe that basically a bomb was aboard the plane, speculating a targeted hit typical of Russian leader Vladimir Putin; Russian officials deny that.

We learned the lesson of the importance of attribution in headlines in 2006, when 12 men died in a mine collapse in West Virginia. One night, then-West Virginia Gov. Joe Manchin announced that 11 of those 12 men had survived. We declared that 11 men survived in a big headline on the front page. Three hours later that night, the CEO of the mine's owner declared that only one man survived.

We scrambled back to the newsroom to update our front page, but the damage had been done: The incorrect news was on the front page for thousands of our readers for whom papers already had been printed.

If we had said in the headline "11 men survived, governor says," that would have mitigated the situation at least somewhat, but without the attribution, we declared the news as fact as if we knew it to be true, and we were very wrong.

So when we have to rely on reports that we're not sure we can trust as certain just yet, we have to couch even our headlines: We didn't know for sure Prigozhin was killed in a simple plane crash; we knew only that an agency said so.