One paper from 2012 linked darker skin to aggression and sexuality in humans. Another from that year claimed to show that women with endometriosis are more attractive. A third, published last December, lamented physicians who posted casual pictures of themselves online—including some in which they’re wearing bikinis—as being unprofessional.
All three of these articles have recently been retracted after outraged readers took to social media. In the past three months, at least four other articles, too, have been called out for both their content and their lack of scientific rigor, and then either flagged or withdrawn by their science publishers.
It’s playing like a preview for “The Purge: Academia.” Just as politicians and entertainers are confronted with years-old tweets that aren’t quite in keeping with the image they want, journals have been faced with ugly papers from their archives—some old and long-ignored—that their readers find unsettling. These papers were deeply flawed, and removing them from the literature is a good thing. But the reactive nature of the moves raises questions. Publishers’ typical narratives would suggest that problems such as these would be caught by peer review, before a manuscript is accepted; rather than acknowledged only later, in the middle of a public backlash.
For some of these retracted papers, the question is not whether they’re offensive but rather how they managed to get published in the first place. Take, for example, the one that argued that Blacks and Hispanics lack the cultural fundamentals for success in the American economy; or the commentary in a leading chemistry journal that was hostile to efforts to increase diversity. Another argument against affirmative action, published in the Journal of the American Heart Association, was taken down on account of its “many misconceptions and misquotes,” as well as “inaccuracies, misstatements, and selective misreading of source materials.”
That hasn’t stopped conservatives from decrying the “censorship” of cancel-culture mobs on social media, and dismissing the recent moves as an exercise in virtue-signaling. Indeed, the fact that journals waited nearly a decade after publication to issue some of these retractions—and then moved very quickly—hints at a sliver of truth in the conservative critique. If a paper from 2012 didn’t meet a journal’s standards for scholarship to begin with, what’s so different now?
The critics are right: journals do have a double-standard, and it is political. They move briskly to pull unworthy papers tinged by politics while ignoring hundreds, or likely thousands, of credible allegations of fraud or major error. Just ask Elisabeth Bik, who some five years ago carefully documented and reported evidence of image manipulations in around 800 academic papers, often to no avail. In many cases, the publishers of those articles are the same as those that hop to it when social-media-powered petitions make their way to their inboxes.
Of course, no one (that we know of, at least) is arguing that #AllPapersMatter, and not all bad articles are created equal. For example, a paper that advocates racist pseudoscience almost certainly would cause more harm than one that over-hypes the benefits of doing a Superman flex in the bathroom mirror before a job interview, or an uncited and entirely forgettable article with one or two duplicated figures. Papers claiming benefits of snake oil should also be prioritized for retraction. Journals should act quickly to pull these more dangerous studies, while a little less alacrity on the others is understandable—to a degree, as long as they do something at some point.
But journals often act as if they’re monuments to their own rectitude rather than repositories of valid scientific information. The Lancet took a dozen years to retract the bogus study linking autism to early childhood vaccinations. Science has yet to pull a 2011 article, which was almost immediately debunked, claiming to have found a bacterium that lives on arsenic. And the infamous “Study 329,” in which SmithKline Beecham (now GlaxoSmithKline) downplayed the potential harms of its mood drug Paxil, remains in the pages of the Journal of the American Academy of Child and Adolescent Psychiatry, almost 20 years later.
Every time a reader sees a paper with fatal flaws that hasn’t been retracted, [their] trust erodes a little further.
To be fair, small publishers may not have the person-power to conduct a thorough review of their back catalog. Retractions aren’t always, or even generally, simple administrative matters. Clear-cut claims of plagiarism, for example, must still be vetted with software and the human eye to compare text and make sure that overlapping sections are indeed theft. Allegations of manipulated images require investigation of figures that even experts may find difficult to decipher. Questions about shoddy statistics and tortured methodology, which may blur the line between acceptable practice and bad science, often require adjudication by independent experts. Oh, and authors of papers on the chopping block don’t always agree that their work should be retracted. They can drag things out for months or years, or, in some cases, even sue journals in response.