The trail of manuscripts posted to the preprint server arXiv shows how cosmologists rapidly embraced, then doubted and gradually lost interest in one of last year’s most sensational announcements: the discovery of gravitational waves generated at the birth of the Universe.

Soon after the announcement last March, papers questioning the result started to emerge. The final nail in the coffin came last week, when researchers officially conceded that dust in the Milky Way was responsible for the signal seen by the South Pole-based telescope BICEP2.

Credit: Source: Paul Ginsparg/arXiv

Such was the intial excitement over the result that between March 2014 and last week, physicists and astronomers posted more than 1,000 research manuscripts about the findings to arXiv — more than 3 per day, says Paul Ginsparg, a physicist at Cornell University in Ithaca, New York, who founded arXiv. “This really was an extraordinary (albeit hallucinatory) phenomenon to enter into so many articles,” he says. As Ginsparg’s data show, a flurry of manuscripts appeared about BICEP2 in March, and older papers were updated to include mention of the findings. But interest quickly waned — and mention of dust began to take up a larger proportion of the postings.

“I was motivated to track this in real time because it did appear so problematic right from the outset to so many of us — and with so many important lessons, like don’t get videoed accepting a bottle of champagne,” says Ginsparg. (For Nature’s view on the episode's lessons about communicating science findings, see ‘Dust to Dust').