Get Access to Print and Digital for $23.99 per year.
Subscribe for Full Access

On the morning of May 26, Donald Trump posted a pair of tweets about voting by mail. “There is NO WAY (ZERO!) that Mail-In Ballots will be anything less than substantially fraudulent,” Trump wrote. “This will be a Rigged Election.” Soon after, a monitor flagged the tweets for violating Twitter’s “civic integrity policy,” which reads, in part, “You may not use Twitter’s services for the purpose of manipulating or interfering in elections or other civic processes.” In response, Twitter appended to the tweets an unobtrusive link that read “Get the facts about mail-in ballots” and brought users to a page noting: “Trump’s claims are unsubstantiated, according to CNN, Washington Post and other fact checkers.”

Outraged by this rather mild clarification, Trump issued an executive order threatening protections contained within Section 230 of the Communications Decency Act, the 1996 law that gives “interactive computer services” like Twitter immunity from legal responsibility for user-generated content. (There was a certain irony to this response, since it is precisely Twitter’s status as a “platform” rather than a “publisher” that has allowed it to host Trump’s defamatory statements with impunity.) The day after Trump issued this order, he posted a tweet calling black Americans protesting police brutality “THUGS,” adding that “when the looting starts, the shooting starts,” a threat that was judged—quite correctly—to have violated the site’s policy against glorifying or inciting violence. This time, Twitter went a step further. The company hid the tweet from view, forcing users to click through in order to read it and preventing them from replying to it. But they did not take it down, as they likely would have if any other user had posted it.

Trump has violated one or another of Twitter’s policies on a near-daily basis in the decade since he joined the site. But until now, the platform has taken a laissez-faire approach, allowing him to communicate without editorial oversight to his eighty million followers, not to mention the readers of the countless media outlets that treat his tweets as inherently newsworthy and breathlessly amplify them. So even these minor interventions represented a major policy shift, as Trump himself obviously recognized. While many users believed that the site had not gone far enough—Trump’s account should have been suspended, they said—there was a widespread belief that this was a step in the right direction. Finally, Trump’s tweets would be fact-checked.

Fact-checking is often identified as one of the features that distinguish so-called legacy publications such as Harper’s Magazine from the “publish first, ask questions later” world of new media. And it’s true that we consider getting things right to be an essential part of what we do. But I’m skeptical that Twitter’s apparent embrace of this ethos will amount to much. While Trump is not the first serial liar to occupy the White House, he is the most aggressively “fact-checked” president in history. The Duke Reporters’ Lab, which conducts an annual “fact-checking census,” has found that since Trump took office the number of outlets that “actively assess claims from politicians and social media” has more than doubled. Over the years, the New York Times has attempted to keep a “definitive list” of Trump’s lies. The Washington Post even encourages the public to get in on the fun, with its interactive fact or fiction game, which invites readers to guess whether various Trump statements “pass the Pinocchio test.” Needless to say, none of this work has been particularly effective in changing anyone’s mind about Trump, and it’s tough to see how the occasional Twitter alert about “unsubstantiated” claims will do any better.

One thing that years of work as a fact-checker teaches is the limits of what can be checked. The prevailing philosophy of Silicon Valley—not just of its social-media platforms, but of such data-driven, explainer-journalism sites as FiveThirtyEight and Vox, and of the rapidly proliferating online fact-checking projects—is a kind of positivism that treats arriving at the truth as a simple matter of data collection: the more facts we have, the closer we are to a complete picture of reality.

This philosophy has also made its way to mainstream outlets like the Times. When a recent op-ed by Senator Tom Cotton, headlined send in the troops, caused widespread outrage both inside and outside the Times, the paper responded with a statement acknowledging that the op-ed “did not meet [its] standards”—because it contained a handful of factual errors. As a corrective, the Times committed to expanding its checking operations.

Like so many effective ideologies, this elision of information and truth persuades precisely by presenting itself as the absence of ideology, the neutral view that is laid bare once the facts are allowed to speak for themselves. But facts cannot speak for themselves. Even if they could, they could not speak all at once—the result would just be noise. The truth can’t be arrived at by accumulating atomized data points, no matter how scrupulously they have been vetted. Truth requires a shared context within which the relative meaning and importance of various facts can be judged. It is this sort of context that magazines like Harper’s seek to provide, and getting our facts straight is a necessary but not sufficient part of that work.

Social media—by design—strips this context away. On Twitter, an anguished lament about police brutality follows an absurdist riff on the distracted-boyfriend meme follows an invitation to a friend’s book reading follows an engagement announcement. None of these tweets is “false,” but what is the truth to which they add up?

This lack of context is what makes Trump such a natural fit for the platform. Not because he is the master of the impulsive non sequitur, but because he brings his own context—his own values and worldview—with him wherever he goes. While that context has been painfully manifest in recent weeks, he has carried it with him for his entire public life. It is the context of a man whose first appearance in a major newspaper was as a defendant in a Justice Department suit for housing discrimination against black tenants, a man whose transition from tabloid clown to commander in chief began with his championing of the birther conspiracy. Trump’s supporters know that every tweet carries this context with it, but the structure of social media allows him to deny it when politically expedient. If this feels dispiriting, never fear: both the Times and the Post recently fact-checked Trump’s claim to have done more for black people than any president since Lincoln, with the Post awarding it a rare “Four-Pinocchio” rating.

This checking obsession recently reached a kind of apotheosis after the U.S. Park Police cleared peaceful protesters from Lafayette Square, near the White House, for a Trump photo op. Protesters said the police had used tear gas against them; the government insisted it had merely used pepper spray. Luckily, the Associated Press stepped in to fact-check the matter. Is pepper spray a tear gas? It turns out this depends on whether the term was used in a “common or formal” way. This was as clear a picture as one could have of where we’ve arrived: the president is gassing his own people, and the media is fact-checking the gas.

More from

“An unexpectedly excellent magazine that stands out amid a homogenized media landscape.” —the New York Times
Subscribe now