The worldwide toll of deaths from Covid-19 has just hit 6 million, nearly 1 million of which are in the United States. Few science stories are more important than understanding where the Covid virus came from. Yet the science writers’ section of the press corps has proved strangely incapable of telling the story straight.
Two hypotheses have long been on the table. One is that the virus jumped naturally from some animal host, as many epidemics have done in the past. The other is that it escaped from a lab in Wuhan, where researchers are known to have been genetically manipulating bat viruses in order to predict future epidemics. Both hypotheses are plausible but, so far, no direct evidence exists for either.
The rule for covering such a story is obvious: write about both possibilities as evenhandedly as possible until the truth emerges. But science writers have consistently trumpeted any developments favoring natural emergence while downplaying or ignoring those pointing to a lab leak.
Press reports have abounded in the last few days about new studies claiming to prove that all early cases of Covid were associated with a wet market in Wuhan and hence that the virus must have jumped to people from an animal there. “Two new papers make the case robustly,” ran the credulous headline in The Economist, typical of many articles implying that the case was closed. “The lab leak theory is dead,” declared The New Republic, with great exaggeration.
But the new papers are mere preprints, unpublished drafts that have not yet been exposed to the rigors of peer review. One, from an American group led by Kristian Andersen of the Scripps Research Institute, claimed the wet market was the source of the epidemic. The second paper’s group of authors—headed by George Gao, director of the Chinese CDC—had every incentive to back the Andersen claim but did not do so. They said only, as was already well known, that “the market might have acted as an amplifier due to the high number of visitors every day”—in other words, the crowded, closed atmosphere in the market helped the virus spread from person to person, but the epidemic didn’t necessarily begin there.
“Scientists who weren’t involved in the research papers are calling the new data ‘very convincing’ and a ‘blow’ to the lab-leak theory,” asserted NPR. In fact, the data don’t detract in any way from the substantial evidence favoring a lab leak. They also offer no evidence that SARS-CoV-2 ever existed in nature, the key to proving that it emerged naturally. Not a single animal tested in the market bore the virus. That being so, it’s surely more economical to assume the many positive environmental samples from the market were contaminated by infected people, not infected animals.
Here are three flaws in the new papers that most science writers ignored.
First, even if all the early cases originated from the wet market, as the Andersen paper contends, there is no way of telling whether the virus was carried into the market by an animal or by a person whose infection came from a laboratory. That leaves the debate at square one.
Second, the Andersen group is looking at the wrong time period for determining the virus’s origin. The epidemic probably started somewhere between September and early December 2019. But the cases analyzed in their paper date from mid- to late December 2019, when the epidemic was probably well-entrenched and its origin already clouded. “These authors are trying to close the case using only mid-to-late December cases and this is unlikely to lead to any scientifically robust conclusion of how or when the outbreak began,” says Alina Chan of the Broad Institute and coauthor of Viral.
A third flaw is a probable error in the authors’ statistical assumptions. It has long been known that many of the early cases of Covid occurred in people who had no known association with the wet market, which would seem to rule out the market as the source of the pandemic. The Andersen group has plotted the mid-to-late December 2019 cases on a map, showing that the spatial distribution of those with no known link to the market is very similar to that of the market-related cases. This shows, they say, that even the apparently unassociated cases probably had hidden chains of infection with the market, which is therefore the sole source of the epidemic.
The argument is ingenious. Its fatal flaw lies in assuming that the non-market cases chosen for study were selected at random by the Chinese authorities. In fact, as Chan has noted, one of the authorities’ criteria was closeness to the market. The spatial pattern of non-market cases reflects this selection bias, not a hidden chain of infection to the market. “Since their assumption of no ascertainment bias is most likely incorrect, their analysis is therefore also meaningless,” Chan says.
Unlike most journalists, science writers seldom consider the motives of their sources. Few or none remarked on Andersen’s deep personal interest in the result he was trying to prove. He and his colleagues concluded on January 31, 2020, that the Covid virus did not have a natural origin. But Francis Collins, then director of the National Institutes of Health, immediately decreed this view to be a conspiracy theory that will do “great potential harm to science and international harmony.” Not to mention to his own reputation and that of his lieutenant Anthony Fauci. Both have long advocated for gain-of-function research—enhancing the infectivity of natural viruses—and they funded such research involving bat viruses at the Wuhan Institute of Technology.
No scientist wishes to get on the wrong side of NIH administrators, the major funders of biomedical research. If Collins said the lab leak was a conspiracy theory, why then, so it must be. A mere four days later, Andersen changed his mind and derided lab leak as a conspiracy theory. No one in his group has provided a convincing explanation for this 180-degree reversal. Andersen’s new paper, if true, would go a long way to justifying his otherwise unsupported second take on the issue.
Why are science writers so little able to report objectively on the origin of the virus? Innocent of most journalists’ skepticism about human motives, science writers regard scientists, their authoritative sources, as too Olympian ever to be moved by trivial matters of self-interest. Their daily job is to relay claims of impressive new discoveries, such as advances toward curing cancer or making paralyzed rats walk. Most of these claims come to nothing—research is not an efficient process—but science writers and scientists alike benefit from creating a stream of pleasant illusions. The journalists get their stories, while media coverage helps researchers attract government grants.
Dulled by the advantages of this collusion, science writers pay little attention to in-house problems that seriously detract from the credibility of the scientific research enterprise, such as the astounding fact that less than half the high-profile findings in some fields can be replicated in other laboratories. Fraud and error in scientific papers are hard to detect, yet nonetheless some 32,000 papers have been retracted for various reasons. The reliability of scientific claims is a formidable problem but one of strangely little interest to many science writers.
If the Covid virus should be found to have indeed escaped from a lab in Wuhan, a tidal wave of public rage may shake the temple of science to its foundations. It’s in reflection of their sources’ interests—though political polarization is also involved—that science writers jump on any evidence favoring natural emergence and ignore everything that points toward a lab leak.
Science writers need to decide whether their duty lies to their readers or to their sources. One choice makes them real journalists, the other just unaccredited PR agents for the scientific community.
Photo by Chip Somodevilla/Getty Images