It seems dangerously easy to get scientific nonsense published
IN 1996 Alan Sokal, a physicist at New York University, submitted a paper to Social Text, a leading scholarly journal of postmodernist cultural studies. The journal’s peer reviewers, whose job it is to ensure that published research is up to snuff, gave it a resounding thumbs-up. But when the editors duly published the paper, Dr Sokal revealed that it had been liberally, and deliberately, “salted with nonsense”. The Sokal hoax, as it came to be known, demonstrated how easy it was for any old drivel to pass academic quality control in highbrow humanities journals, so long as it contained lots of fancy words and pandered to referees’ and editors’ ideological preconceptions. Hard scientists gloated. That could never happen in proper science, they sniffed. Or could it?
Alas, as a report in this week’s Science shows, the answer is yes, it could. John Bohannon, a biologist at Harvard with a side gig as a science journalist, wrote his own Sokalesque paper describing how a chemical extracted from lichen apparently slowed the growth of cancer cells. He then submitted the study, under a made-up name from a fictitious academic institution, to 304 peer-reviewed journals around the world.
The publications Dr Bohannon selected for his sting operation were all open-access journals. These make papers available free, and cover their costs by charging authors a fee (typically $1,000-2,000). Policymakers have been keen on such periodicals of late. Since taxpayers already sponsor most academic research, the thinking goes, providing free access to its fruits does not seem unreasonable. But critics of the open-access model have long warned that making authors rather than readers their client risks skewing publishers’ incentives towards tolerating shoddy science.
Dr Bohannon has shown that the risk is real.