(Damian Dovarganes/AP)

If your social media filter bubble is more blue than red, you’ve probably seen the following, terrifying question recently: “Trial balloon for a coup?” The question, like most questions on the Internet these days, pertains to President Trump.

It is the headline of a viral Medium post making the case that the administration’s chaotic rollout of Trump’s executive order temporarily banning people from seven Muslim-majority countries was actually “the trial balloon for a coup d’état against the United States.” It is, for many reading and sharing it, a convincing argument, or at least one worth passing along to their friends in case it turns out to be correct.

The story was endorsed by woke celebrities:

As it went viral, the post more or less served to scare the crap out of a lot of people about the near future. At the center of the post, written by Google privacy engineer Yonatan Zunger, is the theory that the “inner circle” of the Trump administration is systematically, and deliberately, “probing the means by which they can seize unchallenged power” — to put the entire power of the federal government into the hands of a tiny group of Trump loyalists, backed by a “shadow government” of supporters in the police and intelligence community.

Like any successful speculative theory, the “trial balloon for a coup” is based on observation of real-world phenomena. The Washington Post’s own reporting on the Trump administration indicates that the White House’s vision for America appears to be driven by a small group of very powerful Trump loyalists and that it is aggressively aiming to fulfill Trump’s most controversial populist campaign promises. But its central claim of intent, that all those actions are leading up to a coup, remains unproven, even as it is responsible for the entirety of its momentum online.

A viral, speculative scenario can be either a fantasy or a nightmare, but both usually share one thing: a promise of certainty. There’s a weird sort of relief that comes with believing you understand, at least, why a bad thing is happening, even if you’re powerless to stop it. Increasingly, the paranoid Medium post has entered to provide that relief.

Medium, an online publishing platform, has allowed just about anyone to publish their theories about the Trump administration into a sleek-looking article. Sometimes, as was the case for an essay by a fired “Marketplace” reporter, Medium provides a direct line for a different perspective into a conversation that might not otherwise be available through a more traditional media organization. But it has also started to become a home for frightening, unproven speculations about our future in Trump’s America.

Another viral Medium post declares: “The Immigration Ban Is a Headfake, and We’re Falling for It.” That post received its own celebrity boost:

The Medium posts appear to be a cousin to the Twitter “thread” phenomenon, where speculative analyses of the current state of things are laid out in a long series of tweets instead of in an actual blog post. And they also share a tendency to overstate the certainty of what it is they’re writing about.

When Tom Pepinsky, an associate professor of government at Cornell University, saw the “trial balloon” post, his expertise led him to a very different possibility from the same evidence. In an article rebutting Zunger’s approach, Pepinsky argued that it was possible “everything that Zunger identifies is evidence not of a deliberate planning by an aspiring authoritarian, but of the exact opposite: the weakness and incoherence of administration by a narcissist.”

But more than that, the point of Pepinsky’s piece was to identify the flaw in trying to completely understand and explain the Trump administration’s intentions in real time at all: “observational equivalence.”

“We have two theories of why something is happening, and yet we cannot tell which is the ‘correct’ theory based on the data that we observe,” he wrote. “We have precious little evidence about what is happening within President Trump’s administration. What we observe is its output: executive orders, staffing decisions, and personnel management. What we don’t observe is everything that we need to know to interpret those outputs.”

More reading: