In 2014 BICEP2 made a very controversial press release. It was rushed, given before any papers where published, because they feared they would be scooped by another similar experiment. In the end it turned out that their analysis of their data was flawed. It was particularly dramatic because if they had been correct it would have been a significant development.

Are there any similar cases in other sciences? Not necessary an identical situation, but a high profile result that was released in a hurry that later turned out to be incorrect.


A textbook case of this is the famous cold fusion debacle. It explains the temptation, if cold fusion worked out, producing a cheap and abundant energy source, there would be a Nobel Prize and world fame. Would all the cut corners still have mattered then? Maybe not that much.

Back in the 1960s, Fleischmann noticed that palladium can absorb copious amounts of hydrogen, which meant that the packing of hydrogen atoms in it had to be very dense. This gave him an idea that if deuterium, whose atoms are larger, is used in place of hydrogen the packing will be dense enough to fuse two atoms together and produce helium-4, an unstable element that quickly decays into helium-3 releasing large amount of heat. If it worked, cold fusion would have solved the energy crisis. In 1983, in collaboration with Pons, Fleischmann performed an experiment with palladium and platinum rods submerged in heavy water and deuterium was produced by electrolysis. They did detect a large amount of excess heat, not accounted for by chemistry, and applied for an NSF grant. One of the reviewers, Jones, was working on a similar setup, but tried to detect neutrons produced by the helium-4 decay rather than excess heat. The amounts he detected, however, weren't nearly enough to account for the heat Pons and Fleischmann reported. Still, it looked promising and Jones proposed a collaboration.

Pons and Fleischmann did not take to Jones kindly, thought he stole their idea, and the race was on. They decided to measure the neutrons themselves, which was not in their area of expertise. The results did not match either Jones's or their own excess heat measurements. Nevermind. In 1989, Jones proposed submitting publications simultaneously to the same journal to share the credit, the Journal of Electroanalytical Chemistry, in 18 days. Pons and Fleischmann agreed despite the fact that by their own estimation they needed 18 more months to gather data. In only 5 days they submitted their paper. Given the historical importance of the paper and about-to-be-solved energy crisis, the editor put it through an expedited review process, giving reviewers only one week instead of several. Even that was not expedited enough apparently. With a nudge from the University of Utah, Pons and Fleischmann held a press conference where they announced their results without providing much detail on the experiments.

Then the troubles began. The unpublished paper was circulated and several groups tried to reproduce the cold fusion heat generation. Most did not succeed and the few that did later retracted due to an equipment error, or because they could not reproduce their own results. The paper also did not have enough details to reproduce the original experiment exactly, and Pons refused to provide them, the University of Utah did not want disclosure until a patent could be filed. Even when palladium rods from his own laboratory were inspected the results were negative. Pons "explained" it by heat output being smaller than he originally reported. In the meantime, errors in Pons and Fleischmann's experimental design came to light, and the fact that fusion in it was at variance with nuclear physics. Finally, Salamon was allowed to perform independent measurements on the original apparatus, and detected no required neutrons even when Pons claimed the fusion was on. Pons threatened him with legal action if he did not retract. After a year the headlines read: Fusion or Illusion.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.