Remember this? There is a new chapter in the mother of all narrative fails, and it’s hilarious.

That was the cover graphic on the New York Times Magazine on the day the US Navy SEALs killed Osama bin Laden. If you’re ever looking for a definition case of awful timing, this was it.

Over at Blackfive, McQ took that story and gave it a proper fisking. It was, said McQ, the Times’ attempt to resurrect the hoary old “crazy vet” meme from Vietnam. Because every war America gets involved in from now until the end of time will be Vietnam. Not the big wars of arms and ideas America won. Just the little one we lost.

But now, there’s more to the story. Over at Contentions, Jonathan Tobin squares off the New York Times against…itself. The Times of 20 years ago, it turns out, was smarter than the Times of today.

The Times cites some war crimes statistics in an attempt to demonstrate that American troops who commit such crimes are actually not aberrations. And those stats are the problem. To be more precise, they’re fake.

The source of these statistics was General S. L. A. Marshall, a military historian who included it in his 1947 book Men Against Fire. Mogelson pulled them from a more recent book by retired military psychiatrist Lt. Col. Dave Grossman, whom Mogelson quotes as accusing the military of “programming” soldiers to kill indiscriminately.

But what Mogelson fails to disclose in his article is that, more than 20 years ago, the New York Times itself published an article debunking the numbers upon which his entire argument rests.

On February 19, 1989, the Times published a front-page story by Richard Halloran detailing the findings of historians who had probed Marshall’s research and discovered it was completely fabricated.

It’s one thing to have new information come to light to debunk a story — we’ve all been there. It’s another thing to have your story debunked by old information that has been laying around a while. But it’s downright priceless when the old information that debunks the story comes from your own publication.

(thanks to Mark)