RADIO, TV, the Internet, even mobile phones constantly bombard us with snippets of news from around the globe. Trouble is who and what to believe? Science has always been lucky in this regard. Peer-reviewed journals ensure that protocols are followed, discussions rational and conclusions justified. Which makes it all the more disastrous when peer review breaks down.
On its website this week and in next week鈥檚 journal, Nature disowns a paper it published last year. In it, David Quist and Ignacio Chapela of the University of California, Berkeley, reported finding DNA from genetically modified plants in wild maize in Mexico. They also showed that the foreign DNA had somehow become scattered throughout the maize鈥檚 genome. Now it seems they got it wrong. Next week, Nature will carry two articles showing that the scattering is a mirage鈥攁n error created by the method chosen by the researchers. Yet, the journal also carries more work from Quist and Chapela, which they claim bolsters their original finding that maize has been contaminated by artificial genes (see A natural mistake?). What does it all mean?
An editorial note helpfully guides the reader. First, the original paper should not have been published. Second, the journal鈥檚 reviewers disagree over whether Quist and Chapela鈥檚 new work should be published. And, thirdly, the journal is publishing everything so that readers can make their own minds up. Confused? You鈥檙e not alone. The point is not so much to criticise Nature鈥攖hough by leaving readers to make up their own minds about research a reviewer rejected, they seem to be failing in their duty鈥攂ut to show the confusion caused when peer review goes wrong. Nobody knows what to believe.
Advertisement
There is one other recent, important example of peer review going AWOL鈥攚hen The Lancet published research by Arpad Pusztai of the Rowett Research Institute in Aberdeen (16 October 1999, p 1353). That research, which described abnormalities in the guts of rats fed on GM potato, was widely regarded as flawed. Some of The Lancet鈥檚 reviewers urged the journal not to publish, yet the editor argued that it would be better if the results were in the public domain. That episode too caused great confusion. Publication did not clarify the issues at stake. Indeed, campaigners against GM foods still argue that the article added credibility to Pusztai鈥檚 claims that GM foods could damage people鈥檚 health.
Why do these events happen? Much of the blame must lie with the pressure on researchers to publish as often as possible in high-status journals. In response, researchers are submitting for publication ever more preliminary results, which by their nature are more difficult to put into perspective. In the rush to be first, scientists also tend to be more secretive, so they may not circulate papers for review among colleagues who would spot flaws early on.
Journals are in a difficult position, too. Rejecting controversial research can lead to accusations of bias or censorship. But they aren鈥檛 absolved of blame. Science and Nature, for example, battle like fury for supremacy in the US and Europe and each one knows that if it says 鈥渘o鈥�, its rival may say 鈥測es鈥�. So the pressure is on to find reasons to publish rather than turn a paper down.
What鈥檚 clear is that on some important, contentious issues, science relies too heavily on peer-reviewed journals. To make sense of these cases we need a different system altogether, and you don鈥檛 have to look far for a model. Bodies such as Britain鈥檚 Royal Society and the US National Academy of Sciences already initiate investigations into broad areas of science. It wouldn鈥檛 be much of a stretch for them to form scientific 鈥渉it squads鈥� to report quickly on narrow but important issues.
The benefits of this could be enormous. First, an expert group could give a complete assessment, tying up loose ends so that nothing is left in limbo. If more research is needed for the sake of clarity then it can point research teams in the right direction. Funding bodies would be daft to ignore that advice. And last but not least journalists and the public would have an authoritative source.
It is a curious coincidence that Quist and Chapela鈥檚, and Pusztai鈥檚 work deal with people鈥檚 twin fears about GM crops鈥攖heir danger to the environment and to our own health. Even if their research is flawed, someone still has to answer the questions they posed. At a time when public confidence in science is low, it would do no harm to see science fighting to regain people鈥檚 trust by injecting some sense into areas of confusion.
