Lies, damn lies and social media

Abstract

In his article on the rapid proliferation of misinformation on social media, David Robert Grimes (2020) quotes Jonathan Swift: “falsehood flies, and truth comes limping after it”. A more popular version, “A lie can travel halfway across the world while truth is putting on its shoes”, was ostensibly penned by Mark Twain in 1919. Twain—who inspired the title of this editorial—actually died in 1910 but he may have appreciated the irony of the misattribution.

Two articles in this issue, by Grimes and Emilia Niemiec (2020), discuss how the COVID‐19 pandemic highlights the dangers of an “infodemic” that spreads lies, bogus claims and misinformation with potentially serious consequences. While Niemiec looks at how social media companies have been dealing with misinformation on COVID‐19, Grimes’ article examines the enormous harm that false information on health can cause. This is not a new phenomenon: falsehoods about the safety of vaccines have been a chronic problem since Andrew Wakefield’s infamous paper on the purported link between the MMR vaccine and autism. And it is not limited to health and medical issues either: social media are systematically being used by political, commercial or ideological interests to manipulate public opinion.

The most obvious solution is censorship or filtering of information by the social media companies themselves. Two years ago, a Google search of “MMR vaccine” listed outrageous anti‐vaccine websites right on top. After US Congressman Adam Schiff (D‐CA) contacted Google and Facebook about their role in spreading anti‐vaccine propaganda, both companies pledged to address the problem. The same search term now yields public health authorities and academic sources on vaccine safety; anti‐vaccine propaganda comes much further down the screen. Facebook has also made anti‐vaccine information less visible on its site. Note that this is not censorship per se: Google does not withhold anti‐vaccine websites from its search results, and Facebook does not remove posts; they just make these harder to find. Nonetheless, the question remains why did it take so long and required the intervention of a congressman to do the right thing?

As Niemiec and Grimes explain, social media companies are not primarily interested in whether something posted is true or not; their business model is based on the number of clicks a post generates. Counting on them to fight the infodemic is therefore akin to ask the foxes to guard the chicken pen. But there are other reasons why censorship is not a desirable solution, Niemiec points out. First, who is to decide what is “true” and what is not? This is an easy decision in a clear‐cut case such as vaccine safety, but “truth” becomes harder to define in a murky situation with high uncertainty and experts disagreeing over the right course of action: welcome to COVID‐19. Moreover, censorship would withhold potentially important information from scientists and clinicians who increasingly use social media to widely share and comment on publications and other evidence.

In the end, it comes down to what Grimes calls individual “information hygiene”—prudence and caution when dealing with social media—and/or mandating companies to fact check content in the same way as the media are supposed to. This raises the question of where the information comes from in the first place. And fact‐checking implies a reputable source against which to check. In short, both social media companies and their users are confronted with the bigger question of what is a reliable source of information and what is not. The friend of my neighbour’s sister should be regarded with the same caution as an anonymous Twitter account, whereas a peer‐reviewed publication in a reputable journal or a scientist would be more trustworthy.

In general yes, but, as Niemiec argues, not necessarily. At the heart of the anti‐vaccine movement is Wakefield’s fraudulent paper in The Lancet, which is still cited as evidence by anti‐vaccine propagandists. The same journal retracted a paper on treating COVID‐19 patients with hydroxychloroquine after readers raised serious doubts about the underlying data. A similar paper was retracted by the New England Journal of Medicine for the same reason, which illustrates that journal peer review is not without flaws (Gannon, 2020).

Just because a scientist says so does not mean that it is true either; a hallmark of science is that its practitioners are wrong most of the time. It is only by patiently weeding out false ideas and hypotheses with new data and evidence that science exerts its power to get to the truth. Many epidemiologists and virologists have therefore cautioned against believing in any explanations or recommendations regarding COVID‐19, including their own, given that they could be wrong tomorrow in a situation with high uncertainties. The recent “Great Barrington Declaration” with its polarized view against lockdown measures is a case in point, again, of the danger of drawing conclusions from insufficient evidence by academic scientists (https://www.sciencemediacentre.org/expert-reaction-to-barrington-declaration-an-open-letter-arguing-against-lockdown-policies-and-for-focused-protection/).

A lack of reliable information is nothing special for science—it is, after all, where and how front‐line research operates. But to confirm or refute ideas and hypotheses with new evidence requires openness and transparency—not only for sharing information but also to evaluate the quality and veracity of the underlying data. Hence the concept of Open Science to encourage and enable sharing and transparency—which, in a perfect world, should include social media’s search algorithms.

It is here where journals could do more—mandating sharing of data, protocols and materials; transparent review; free access to the underlying source data in published articles—would help to advance Open Science. Such polices—as established by EMBO Press journals more than a decade ago—would not solve the problems of an infodemic. But they would help to maintain and increase the reliability and credibility of the scientific literature as a trustworthy source of information—even if takes its time to put its shoes on.

EMBO Reports (2020) 21: e51877.