Deed

Deed had grown up in the reading room of an old records-house, among moles who kept every original a town ever produced and distrusted, on principle, anything that had been copied.

Her grandfather ran the deepest stacks, and he had a saying he repeated until it settled into her bones: A retelling is a rumor with good manners. Go find the deed. Whenever anyone repeated a claim in his hearing, he would send young Deed to fetch the original and lay the two side by side.

"What did the story say," he would ask, "and what does the record say?"

The gap between the two answers, he taught her, had a shape you could learn to predict. Retellings did not drift at random. They drifted toward the interesting — a modest number swelling into a fortune, a "considered" hardening into a "decided," a "some residents" ballooning into "the whole town." Each teller, meaning no harm, nudged the claim a little nearer the version that made a better story, and a dozen small honest nudges added up to a fiction nobody had ever actually told.

Once, young Deed brought her grandfather a claim she was sure of — she had heard it from someone reliable, who had heard it from someone reliable. He did not argue. He simply walked her back down the chain, hand-off by hand-off, until they reached the beginning: a single offhand remark, misheard, that had been growing ever since like a snowball rolling downhill. "Reliable people," he said, holding up the tiny true thing at the bottom of the pile, "passed along a false thing reliably. That is the trap. The chain can be honest at every link and still deliver you a lie — because no one in it ever went back to the deed."

She arrived at the NewsForge newsroom with her grandfather's lantern and a habit of counting hand-offs.

Scoop, the editor who met every newcomer, looked down at her over his galleys. "What is a primary source?"

Deed didn't recite. She set two objects on his desk: a folded newspaper clipping, and beneath it a thick, plain envelope. "This," she said, tapping the clipping, "is an article about a scientist's study. It's a retelling — a reporter's summary of a press office's summary of the study." She slid the envelope forward. "This is the study. The thing itself. The place the claim actually lives." She opened it a crack. "The article says the study 'proves' a food is dangerous. The study says it found a small effect, in mice, at doses no person would ever eat, and calls for more research. Same study. The article isn't lying — it's just three retellings downstream, and 'more research needed' has drifted all the way to 'proven dangerous.'"

Scoop read the two side by side. Then he set his pencil down. "Most of this newsroom quotes the article and calls it reporting," he said. "You're the one who'll open the envelope. You belong here."

Deed's corner filled, over time, with reporters who'd been burned by the drift.

A young writer slumped onto the stool across from her, holding his phone like it had betrayed him. "I reported that a company admitted its product was faulty," he said. "I got it from a really widely-shared summary. Turns out the company said the opposite — that an independent test cleared it. I trusted the summary because it was everywhere. How was I supposed to check the original for every single thing? I don't have all day."

"You don't need all day," Deed said, not unkindly. "You need a rule about which things to walk back to the source — and it's a simple one." She held up a claim. "Walk it back whenever the claim is surprising, important, or convenient — surprising, because surprise is exactly what a drift produces; important, because the cost of being wrong is high; convenient, because a claim that's suspiciously perfect for the story is the one most likely to have been nudged into shape." She set it down. "'Everywhere' is not a source. A false thing shared a million times is still a false thing — it just has company. The number of retellings tells you how popular a claim is. It tells you nothing about whether it's true."

The writer stared at his phone. "So 'it's everywhere' was the reason I should have checked — not the reason I didn't have to."

"That's the whole reversal," Deed said. "Popularity is a reason for suspicion, not for trust. The more a claim has traveled, the more hands have had a chance to nudge it — and the more it's worth going down the stairs to find the deed."