Correction/Clarification: The original headline of this piece said that the relevant PNAS paper did not pass peer review. As pointed out in the comments, it did not go through the same standard peer review required of most other papers.
It might seem like a classic tale of triumph for the underdog: A group touting a theory long considered fringe gets published in PNAS. And yet, it wasn’t what most reporters made it appear.
The claim was that scientists finally found “conclusive” evidence that a comet or asteroid wiped out the big mammals in North America some 12,900 years ago – and according to some sources, spurred the start of civilization.
Anyone who has followed this might wonder about earlier claims by the same group. A breathless NOVA documentary a few years earlier showed what appeared to be the final definitive evidence for this scenario. That time the group found “nanodiamonds” embedded in a particular layer of Greenlandic Ice, and the lead researcher broke down in tears of joy near the end. This time it was BB sized “spherules” found both in Canada and Pennsylvania.
This new “conclusive” evidence led to less-than-critical stories in places such as Livescience. There, skeptics were quoted saying they doubted the finding, but it wasn’t clear whether they represented token skeptics or the view of the community. The Huffington Post went so far as to credit the impact with starting civilization. Neither story offered a clue as to whether anyone outside the paper’s author list believes any of it. No one except those making the claim seems to find anything conclusive.
Here’s the Huffington Post:
"Experts agree on the climate change that marked the start of the Younger Dryas period, but are divided about its cause. The new research is said to provide "conclusive" evidence that the trigger was a meteor impact."
New research “is said” to provide conclusive evidence? Who says? The story doesn’t have a single voice in favor of this idea outside the group behind the paper.
"Before now the most popular explanation has been an ice dam rupture releasing huge amounts of freshwater into the Atlantic. "
Has anyone changed his or her mind because of this paper? The story, believe it or not, doesn’t say.
Several interesting facts came out in a more reasoned take by Nicola Jones in Nature.
First, there’s more to this impact debate than just piecing together the earth’s past:
"The leading theory is that shifting North American glaciers allowed freshwater melt to pour into the Atlantic or Arctic oceans, slowing ocean circulation and cooling the Northern Hemisphere. This idea has spawned worries that fresh water from today’s melting ice might also spur rapid climate change.
According to Nature, this is still the leading theory. But more importantly, Nature notes that this paper didn’t pass peer review. Nor did earlier papers by this same group:
"Steven Stanley, a paleobiologist from the University of Hawaii at Manoa in Honolulu, has acted as a ‘personal editor’ on several of the PNAS papers on this topic, including the new work by Sharma and the original 2007 paper proposing the idea. “It has been very controversial,” he admits. “It’s my view that I should help to get this stuff published. It needs to be aired; it’s not outlandish.” PNAS typically uses a 'personal editor' option for papers considered too controversial to receive a fair hearing from the standard review process.
Stanley says he is increasingly convinced by the impact theory as a mechanism for what prompted the freshwater floods. “I’m not sure how people can be so negative at this point. The case just builds.”
Standard peer review isn’t always required if a member of the National Academy of Sciences wants a paper to be published. Why would these impact papers fail peer review? Are scientists in this field unusually closed-minded or is there something wrong with this research? Is it possible the data don’t support the conclusion?
Another PNAS paper that didn’t go through peer review at all was published in 1987 by Peter Duesberg, a biomedical researcher who has insisted for more than 25 years that HIV does not cause AIDS. He can get such a paper published, apparently, because he is member of the academy.
It’s pretty clear now that he was wrong, but did Duesberg’s case have merit back then? You can learn a lot about the issue in this story by Jon Cohen in Science, who shows how real science reporting is done. There are no token skeptics or empty, ambiguous quotes. Instead, there’s a genuine assessment of the biomedical research community’s view of the role of HIV in AIDs, and of Duesberg’s contrarian stance. Scientists quoted there give specific, understandable reasons for doubting Duesberg and believing that HIV does, indeed cause AIDs.
As sexy as it is to report that the brave outliers are right after all, often the mainstream view prevails in the long run.