Example · part of What shortening does to truth

In December 2010 NASA held a press conference for a paper in Science. A bacterium from California’s Mono Lake, the authors said, could build its DNA with arsenic in place of phosphorus. Stories about “alien life on Earth” began within days.

Within days, the microbiologist Rosie Redfield had posted a detailed critique of the methods. Within weeks, labs were trying to repeat the result. Within eighteen months Science published two independent refutations: the bacterium’s DNA contained no arsenic, and the original measurement had been contaminated. Coverage spiked for about two days and went quiet.

The book includes this case to show the system working. A finding passes several checkpoints on its way to the public, each one a round of selection. Here the scientific community’s check caught the error before the journalists and popularizers of the bridge zone could build a lasting story on it. Without this case the book’s diagnosis would read as blanket pessimism. Set it beside power posing, where the correction came after the story had spread. The book suggests the check works better for extraordinary claims in the hard sciences than for crowd-pleasing claims in social science, and ties the difference to the economics of attention.

Read the full argument

How it connects


← Power posing: the meme that survived its own corrective · What shortening does to truth (5 of 6) · Stanford Prison Experiment correctives: died at curation →

Explore all the ideas · Read the book in order · Open the interactive map