On 24 July, “” was finally unannounced.
That quotation was part of a media release from the United States National Aeronautics and Space Administration (Nasa) about a to announce the “discovery” of a microbe that could survive even in the absence of phosphate – an element previously assumed to be essential to life. The bacterium in question could supposedly grow on arsenate in the absence of phosphate and incorporate arsenic instead of phosphorus into its DNA and RNA.
The conference was filled with grand statements of the significance of the research, described in an at the same time. If life were possible in such unlikely conditions, who knew in what unlikely places – terrestrial or extraterrestrial – it might be found.
Yet even during this, frankly, embarrassing event, alarm bells should have been ringing. One of the panellists took gentle issue with the claims. Describing himself as a curmudgeon, the chemist Steven Benner calmly and kindly described the unscientific rationale of the study, thereby demolishing its plausibility.
Indeed, that the article was nonsensical was immediately manifest to the overwhelming majority of the scientific community. The proposal violated well-established chemistry (arsenate cannot stably form the necessary chemical bonds), and the evidence in the article provided no support for the claims of the 12 authors. Yes, the bacterium was highly resistant to arsenate, but?it was not unique in this trait: even nematodes and brine shrimp? from where the bacterium was isolated. Besides, resistance does not imply support for the proposition. For example, a bacterium can be resistant to an antibiotic without making use of it.
But most of the media trumpeted the paper uncritically as heralding a scientific revolution. Indeed, even some scientists got in on the act; that the paper “means that every biology textbook now has to be revised. Even the very definition of life may have to be changed.”
But others pushed back hard. Some, including microbiologist?, focused on the lack of real evidence for the incorporation of arsenic into the bacterial DNA. Redfield disseminated her opinions – a milestone in unfiltered post-publication review. Still others, Istvan Csabai and Eors Szathmary and I, found that the calculations and statistics that the authors were using were suspect. And it was noted that the authors had presented a distorted interpretation of the previous literature.
Moreover, a number of scientists, including Benner, and myself, recognised that the most likely explanation for the apparent ability of the bacterium to grow in the presence of arsenate was that it was, in fact, growing on trace phosphate that was present in the growth medium, probably as a contaminant of the arsenate itself.
The sad part is that the authors were aware of this problem, as illustrated by data in a supplementary table in their paper. But they were apparently so enamoured of their hypotheses that they wished it away and doubled down on the rigour of both their results and their interpretation of them. And even now that the paper has , as it was last month, the authors are still unwilling to concede that the paper was fundamentally flawed.
Yet that fact was readily apparent even by the time the paper was published in the print version of Science. During a highly unusual six-month delay, the criticisms had piled up to such an extent that the journal felt obliged to publish an unprecedented eight “technical comments” alongside it, which thoroughly eviscerated nearly every aspect of it. Nor was the authors’ case helped by the extremely weak, disingenuous and mostly irrelevant response from them that was also published.
You might argue that the inclusion of the technical comments validated the print publication of the article, allowing readers to draw their own conclusions about how flawed it really was. But the comments were subjected to review, whereas the authors’ response was allowed to include assertions inconsistent with the data and science. Moreover, a journal cannot justify publishing rubbish by saying that it also allowed others to explain why it was wrong. To do so would be to make it look as if there was a legitimate dispute about “interpretation” (the desperate argument of the authors), whereas, in fact, the article represented bad science. I’m sure that Science has received articles that state that the Earth doesn’t spin on its axis, but it doesn’t publish them to “stimulate debate” and allow readers to come to their own conclusions.
The fact is that the article should never have been published online, never mind in print. Publication in Science is based on two criteria. First, the studies must be “important”. Second, they must pass prepublication peer review. Clearly the studies fulfilled the first criterion. An article that both overturned the laws of chemistry and changed how we would engage in the search for life on other planets would be irresistible. Tossed into the mix was the stubborn advocacy by one of the authors, Paul Davies, of the possible existence of a “shadow biosphere” on Earth, consisting of microbes with highly divergent biochemical processes from those of known life.
That hypothesis, too, enjoys neither evidence nor rationale. Yet the paper passed peer review anyway. How?
The , Dan Vergano, published in USA Today in 2013, demonstrated that the peer reviewers were substantially unqualified and incompetent. It was also difficult for the editors to reject the assurances of the senior authors, Ron Oremland, Ariel Anbar and Davies, about the reliability of the data. The authority of the last author, however, was, by his own admission, of limited value. A well-known physicist at Arizona State University, Davies after the press conference about his pride in his lack of expertise in chemistry (“I had the advantage of being unencumbered by knowledge. I dropped chemistry at the age of 16, and all I knew about?arsenic?came from Agatha Christie novels”) and how one prominent scientist had asserted that “You’d be off your trolley to go searching for?arsenic-based life”. The implication was that the glare of world attention had refuted the naysaying scientist – and, astonishingly, that ignorance was a positive value in scientific research.
The rejection of informed criticism was to be a recurrent motif in the saga. The authors appeared to have the erroneous belief that the article’s publication was a demonstration of its validity.
Even now, after a number of articles have since disproved the “arsenic life” paper’s conclusions, the authors continue to stand by their findings and refused to sign the . In their attached letter criticising it, they say that “while our work could have been written and discussed more carefully, we stand by the data as reported. These data were peer-reviewed, openly debated in the literature, and stimulated productive research.”
They also note that “no misconduct or error is alleged” in the retraction. But Science’s rationale is that the “paper’s reported experiments do not support its key conclusions”. In the authors’ view, this criterion represents “a major shift from the standards?Science?adhered to in the past, which aligned with those of the Committee on Publication Ethics (COPE).
“”, they note, “state that ‘Retraction might be warranted if there is clear evidence of major errors, data fabrication, or falsification that compromise the reliability of the research findings’. In going beyond COPE, the editors of?Science?explain that ‘standards for retracting papers have expanded’. We disagree with this standard, which extends beyond matters of research integrity. Disputes about the conclusions of papers, including how well they are supported by the available evidence, are a normal part of the process of science.”
But COPE also makes clear that “retraction is a mechanism for correcting the literature and alerting readers to articles that contain such seriously flawed or erroneous content or data that their findings and conclusions cannot be relied upon. Unreliable content or data may result from honest error, na?ve mistakes, or research misconduct.” There is no doubt that the content of the arsenic life paper was erroneous and the findings and conclusions untrustworthy.
Some have argued in the wake of the retraction that the floodgates are now open. Any papers whose findings or interpretations have been superseded by subsequent work – an inherent feature of scientific advancement –?are?now in danger of being retracted, with their authors’ reputations being trashed in the process.
But that peril is vastly overstated. This is a particularly high-profile and egregious case of a paper that was only published in the first place because the editor was out of her depth and the reviewers were unable to recognise the sheer chemical impossibility, irrelevance of the experiments to the conclusions, and the statistical ineptitude. Any journal that commits such gross negligence has an obligation to correct the literature.
And the journal can hardly plead ignorance of its negligence. One lesson from this case is that post-publication peer review is frequently of higher quality than prepublication peer review, and it should be acknowledged as such. I do not advocate for eliminating peer review, even if it needs reform; I merely insist that prepublication review should not have a privileged position.
So why did the retraction take so long? A comparison with the infamous Wakefield Lancet article affair is illustrative. The 1998 publication of Andrew Wakefield’s paper falsely alleging a link between the MMR vaccine and autism was accompanied by a press conference in which Wakefield advocated changes in vaccination practice for which even the data in the Lancet paper provided no support. The article and Wakefield gained enormous attention, even though it was refuted comparatively rapidly.
In this case, 10 of the 12 authors retracted the major conclusion of the Lancet article in 2004, but it was only six years later that the journal finally retracted the article for ethics considerations. The major factor in the delay was the refusal of Wakefield to accept the scientific results, as a consequence of his apparent craving for the publicity that his initial unfounded assertions had generated.
Attention-seeking behaviour will achieve its goals. However, in the case of a scientific article, this means that the research will be scrutinised more closely, another case in point being Francesca Gino, the fired earlier this year for extensive data fraud. And while the psychopathology of exaggerated self-promotion may lead to denial of the validity of any criticism no matter how indisputable, the fact that authors have deluded themselves should not be a reason for a journal to delay retraction.
There are also reputational issues for journals in admitting error, of course. However, I argue that the reputational damage to Science is greater because of the long deferral of action – not to mention the reputational damage to science itself.
This brings me back, finally, to the media. Although there were a few journalists who expressed scepticism about the soundness of the research and the conclusions of the authors, journalists bear substantial responsibility in this affair. Major journalistic failings include the widespread uncritical and breathless repetition of the unfounded claims of the authors, the presentation of the matter as if there were some possible validity to the article, the framing of the authors as victims, and even the reporting of the authors’ objections to the retraction as something more than an act of desperation.
Yes, reputable journalists strive to be balanced in their reporting, allowing both sides of legitimate debates to air their views. But reporters question the factuality of things they are told all the time: why not in this case? And why did outlets that trumpeted the paper’s results in 2010 not proclaim with equal exuberance their falsity when that quickly became indisputable – or even after the retraction?
There is an important message to convey to the public here, particularly amid increasing political mistrust of science: that, albeit sometimes senselessly slowly, the trash does eventually get taken out of the literature. Unfortunately, the media’s own negligence means that that crucial truth will not be widely heard.
?is associate professor of biological sciences at?Purdue University.
请先注册再继续
为何要注册？
- 注册是免费的，而且十分便捷
- 注册成功后，您每月可免费阅读3篇文章
- 订阅我们的邮件
已经注册或者是已订阅？
