Publicly questioning things is what research scientists always do, but that didn’t matter. The artful message from the tobacco industry to smokers was “This is complicated, and we’ll pay attention to it so that you don’t have to.” When we are confronted with unwelcome evidence, we don’t need much of an excuse to reject it.
Trump seemed to channel this body of thought when he seized upon a moral panic about a few transparently silly stories—“fake news”—and created a catchphrase to smear serious journalists. While we in the media wrung our hands at the idea that people might believe the Pope had endorsed Trump, Trump himself realized that the real danger—and for him, the real opportunity—was different. It was not that people would believe such nonsense, but that they could be persuaded to disbelieve authoritative, carefully sourced journalism.
“Deepfakes”—the technology that creates plausible footage of people saying and doing things that they did not—provide a similar lesson. (Deepfakes of Tom Cruise seem to be popular right now.) One researcher reassured Radiolab that “if people know that such technology exists, then they will be more skeptical.” She may be wrong about that, but I am more worried that she is right—that deepfakes create a world of unlimited deniability. Say anything, do anything, and even if the cameras are rolling, you can claim it never happened. We’re not yet at that point, but the trajectory is hardly reassuring.
From the May 2018 issue: The era of fake video begins
Journalists need to take the problem of weaponized doubt more seriously. Fact-checking outfits in particular, such as PolitiFact, FactCheck.org, and Snopes, must take care not to breed cynicism. The risk is of creating the sense that lies are ubiquitous—which is why the best fact-checkers spend as much effort explaining what is true as they do exposing what is false.
The ultimate cautionary tale here is Darrell Huff’s 1954 classic, How to Lie With Statistics. Huff’s book is clever, insightful, and impish, and it may be the best-selling book about statistics ever written. It is also, from cover to cover, a warning that statistics are all about misinformation, and that one should no more believe in them than in stage magic. Huff ended up testifying at a Senate hearing that the evidence linking smoking and cancer was as spurious as the evidence linking storks and babies. His unpublished sequel, How to Lie With Smoking Statistics, was paid for by a tobacco-lobby group.
Yes, it is easy to lie with statistics, but it is much easier to lie without them. It is dangerous to warn that the lies are universal. Skepticism is important—but we should recognize how easily it can curdle into cynicism, a reflexive dismissal of any data or testimonies that do not fit neatly into our preconceived ideas.