Here's the science behind Trump's ability to control the media through Twitter

“Bush Lied/People Died.” It was simple and to the point: There were no weapons of mass destruction in Iraq and no reason to invade. People paid for it with their lives. Iraq Body Count kept track of how many. Trump lies too (bigly!), but in a fundamentally different way that journalists can't seem to get a handle on — well into his third year in office — leaving themselves (and too many of us) floating around in an ever-fluctuating state of befuddlement.

No one can imagine what Trump’s equivalent of Iraq Body Count might be. As Trump ramps up his racist attacks with the 2020 election cycle starting in earnest — while obfuscating his own far-flung misconduct — it’s long past time for journalists to do something serious to curtail the power of his lies and deceptions. The answers are there, if they’d start asking experts, because there are social scientists who understand a great deal about what’s going on — not just with Trump himself, but with the larger environment in which he thrives. There’s still a great deal more to be learned, but they know enough to provide a vaccine against the needless spread of Trump’s venomous disinformation.

One such scientist is Stephan Lewandowsky, whose work I’ve written about before (here and here). He has a new paper in progress — with co-authors Michael Jetter and Ullrich Ecker — covering the first two years of Trump’s time in office, dealing directly with his strategy of distraction and diversion, a key weapon in his arsenal of disinformation, and one of four main types of tweets identified by cognitive linguist George Lakoff just before Trump took office. (More on that below.)

“We were interested in establishing whether there is any evidence in support of the view voiced by numerous observers that President Trump systematically uses Twitter to distract us from issues that he finds threatening,” Lewandowsky told Salon.

“Up till now, this was supported only anecdotally, for example by noting that Trump tweeted extensively about a Broadway play, haranguing the performers, at the same time that he settled the lawsuit against Trump University. Our study provided quantitative evidence not only for the distraction but, even more strikingly, that the media respond to the distraction by reducing their coverage of the threatening theme that triggered the distraction in the first place.”

That last bit speaks directly to journalists’ culpability: Not only are they distracted by Trump’s shiny-object trick—along with Trump’s millions of followers — they lose focus on what’s threatening Trump, thus amplifying Trump’s distraction with their own. Given how complicated and damaging Robert Mueller’s investigation was, this distraction helped prevent casual news consumers from developing a sense of urgency commensurate with what was being revealed. It’s as if the press was running interference for Trump and shielding him from public scrutiny, rather than exposing him, as it is supposed to do when politicians violate the public trust.

Making sense of the misinformation landscape

To really appreciate what the paper found, we need to place it in wider context, as Lewandowsky did in a recent keynote conference presentation entitled “'Post-Truth': What, Why, and How Do We Respond?" He began by addressing the differences between the Bush and the Trump style of lying, drawing on a 2017 paper, "Combatting Misinformation" for short, by Aaron McCright and Riley Dunlap. It lays out a typology of four distinct types of misinformation, but warns that “combinations of these types of misinformation synergize to have even more complex and recalcitrant impacts.”

Their typology employs two axes — one based on style (formal vs. informal) and target audience the other on ontology (using the terms "realism" vs. "constructivism") Lewandowsky focused on  “the distinction between curated lies and shock and chaos,” both of which are stylistically formal, targeting institutions and systems. These were also the focus of a paper he presented to the Stuengmann forum on deliberate ignorance in Frankfurt, Germany, last March, which will appear in a forthcoming volume from MIT Press.

The term "curated lies" describes the Iraq WMD narrative, carefully supported by U.S. and British intelligence agencies, Lewandowsky explains in “Two Ontologies,” also pointing to global warming denial and the decades-long denial of tobacco health risks as similar examples of the deliberate, carefully curated creation of ignorance in others. Although they are monstrous lies, causing widespread suffering and death, they are built on realist foundations — they acknowledge realty, even as they selectively seek try to obscure it.

“[P]urveyors of this type of misinformation target organizations, movements, and institutions they perceive as threatening their interests,” McCright and Dunlap wrote. “Systemic lies align closely with what we have termed ‘anti-reflexivity,’ or the defense of the industrial capitalist system from the claims of scientists and social movements.”

Perhaps most notably, “The success of the climate change denial countermovement owes much to the Right’s superior effectiveness in framing and re-directing public discourse toward advancing their ideological interests,” they write. “Indeed, the Right seems especially adept at using Orwellian language to promote their ideological and material interests via what we would argue are systemic lies,” going on to cite a list of examples, including:

• “Right-to-work” laws that further weaken labor unions and the very mechanisms (e.g., collective bargaining) that earned workers hard-fought rights in the first place;
• “Religious liberty” bills designed to legalize discrimination against the LGBTQ community, based on a narrow, fundamentalist interpretation of the Christian Bible;
• Focus-group generated terms that conservative activists have infiltrated into public discourse: e.g., “family values,” “junk science,” “partial birth abortion,” “death panels,” “death tax,” “job creator” and, most recently, “fake news.”

Curated systemic lies like these have been the bread and butter of movement conservatism since at least the 1980s, with roots going back even further. But Trump represents something completely different: a liar so prolific the media can’t possibly keep up. Even so, he racked up an eight-point lead over Hillary Clinton in perceived honesty at the end of the 2016 campaign. (More on this below.) Since taking office, his rate of recorded falsehoods has risen dramatically, with no discernible impact on his political support, which remains within a relatively narrow range.

“This type of misinformation is not carefully curated but is showered onto the public as a blizzard of confusing and often contradictory statements,” Lewandowsky’s “Two Ontologies” paper says. “Indeed, some of the claims, for example that people went out in their boats to watch Hurricane Harvey, have an almost operatic quality and are not readily explainable by political expediency.”

In fact, belief is not the point, nor is persuasion: the aim is to bludgeon, not persuade, and — as Hannah Arendt warned — to get you to believe in nothing, leaving you open to accepting almost anything.

McCright and Dunlap describe this as “misinformation intended to destabilize social relations and societal institutions so that its proponents may consolidate power and force unpopular decisions on a confused and/or distracted public. As such, it is a mix of the ‘shock doctrine’ strongly critiqued by Naomi Klein (2008) and postmodern authoritarianism championed by Vladimir Putin’s key advisors, Vladislav Surkov and Aleksandr Dugin.” They observe that this is most common in nations like Russia, North Korea and Iran, and “involves weaponizing misinformation to secure the allegiance of followers and to root out and suppress potential dissidents.”

Underlying all this is an extreme "constructivist" view: There is no truth, just competing stories supported by “alternative facts.” And in part, that’s why our media has such a hard time dealing with it: In a way, they believe the same thing. It’s central to their faith in both-sides-ism, and why they have been easy marks for climate denialism: Everyone’s views are equally valid, and let the market decide whose views win out. Reality? What’s that?

Lying as a feature, not a bug

As I noted above, Trump had an eight-point lead in perceived honesty at the end of the 2016 campaign. There’s an obvious quandary here: How does someone whose own supporters know that he’s lying manage to be trusted? Indeed, a post-election survey found that most Trump supporters recognized one of his most notorious lies as false, but nonetheless saw him as highly authentic, while Clinton supporters did not see her as authentic, but emphasized other positive attributes instead, such as competence.

The answer was provided in "The Authentic Appeal of the Lying Demagogue: Proclaiming the Deeper Truth about Political Illegitimacy," by Oliver Hahl and colleagues. As explained in the abstract: “for the lying demagogue to have authentic appeal, it is sufficient that one side of a social divide regards the political system as flawed or illegitimate.” A key part of the their explanation is that “public compliance with norms often masks the suppression of widespread private dissent,” and the gap between the two “creates an opening for a demagogue to claim she is conveying a deeper truth and is the authentic champion of those whose voices have been muzzled by the established leadership.”

Thus, they not only explain how a lying demagogue may be seen as authentic, but also illuminate why Trump repeatedly attacked “political correctness”: as his way of describing that muzzling. This was also an act of deflection. Trump’s political persona as a Republican was entirely founded on his birtherism, which he further embellished by questioning Obama’s educational record and demanding to see his papers.  All this was far more blatantly racist than national politics would allow, thus creating the need for deflection, for blame-shifting. Trump’s charge of “political correctness” was a racist cri de coeur: “You’re not the victim! I am! You’re oppressing me by telling me what I can and can’t say!”

The topic of deflection brings us back to Lakoff’s taxonomy of Trump’s tweets, as described on WNYC’s “On the Media" just before Trump’s inauguration. Tweets fall into one of four categories, Lakoff said — pre-emptive framing, diversion, deflection and trial-balloon — though he suggested another category too: the "salient exemplar," which means presenting a single, isolated event as typical in general, such as the Trump handles crimes committed by immigrants.

It has also been suggested that projection should be added to the typology — but projection is more  a pervasive feature of Trump’s thinking that shows up in multiple different contexts: pre-emptively blaming others for his own faults, trial balloons testing how well such framing might work, or accusing the mainstream media of spreading “fake news” (a prime example of deflection). The four-fold typology is neatly summarized in this graphic Lakoff tweeted out a month later:

Deflection and diversion sound similar, so it’s worth pausing for clarification. “Deflection means blaming others when Trump is being blamed,” Lewandowsky said. “The classic recent case was his accusation of racism against the congresswomen whom he initially insulted by telling them to ‘go back to where you come from,’” which in turn is grounded in the deflection mentioned above — the narrative of political correctness used to portray white men as the real victims of oppression in today’s America.

On the other hand, Lewandowsky said, Diversion is better described as "a shift of topic to a shiny new object. So Trump tweets about the 'Hamilton' play on the day that he settles a multimillion dollar lawsuit against Trump University. This drowned out the bad news for him and is classic diversion.”

Media helps spread Trump’s lies, even as he attacks them

While Lakoff’s taxonomy has circulated through social media, drawn critical attention and inspired research projects, it’s been largely ignored by the media practitioners it was primarily intended to inform, who still function primarily as amplifiers of Trump's messages. Lakoff has provided a vaccine, but too many journalists are, in effect, anti-vaxxers: They refuse to take it. So the media Trump regularly attacks must share some of the blame: They've been warned, and have chosen to ignore it.

This willful ignorance is particularly striking given how much Trump attacks the press on Twitter. In fact, the first study of Trump’s use of deflection focused precisely on that: “Discursive Deflection: Accusation of ‘Fake News’ and the Spread of Mis- and Disinformation in the Tweets of President Trump,” by Andrew Ross and Damian Rivers. The paper discusses Lakoff’s taxonomy, with examples and sample tweets for each category. They collected all of Trump’s tweets from Nov. 9, 2016 through Aug. 7, 2017 — 1,416 original tweets containing 30,928 words, and did a comparative keyword analysis, using a collection of tweets from all serving state governors and members of Congress. The words “fake,” “media” and “Russia” were the top three words in the resulting keyword list. They found “a high frequency of words used in relation to Lakoff’s strategy of deflection”:

[W]hen expanded into clusters and the entire tweet, words such as “fake,” “media,” “news,” “phony,” and “dishonest,” which all featured in the top 20 words of the keyword list, were almost completely used in reference to the media and his claim that the mainstream media were disseminators of fake news.

Strikingly, Trump used the word “fake” 103 times during this period, followed by “news” 86 times and by “media” 11 times.

They did find "some instances" of keywords linked to the other three of Lakoff’s strategies, though this occurred at a "much lower frequency," which makes sense from a strategic point of view. Recall that shock and chaos is “intended to destabilize social relations and societal institutions so that its proponents may consolidate power.” Delegitimizing all other sources of information is central to this mission.

Other strategies — and even other examples of deflection — are nowhere near as central to the long-term goal. They apply to situations as they arise, rather than to the authoritarian project as a whole. (There’s one notable exception we’ll take up in a moment, where the lines are blurred: the need to divert attention from the Mueller investigation into Russia’s role in bringing Trump to power in the first place.)

In fact, “fake news” deflection was used so frequently that they were able to discern three distinct sub-categories: “direct accusation, accusation as signal of allegiance, and intratweet accusation of fake news and dissemination of mis- and disinformation.” The first, direct accusations are straightforward but the second is a little less so: typically attacking a specific target (CNN, the New York Times) while praising an ally (Fox News). "Intratweet" accusations are more variable in form, but generally use the accusation as a sort of booster shot, to heighten the emotional intensity behind the mis- and disinformation being spread.

Trump/Russia: The grand distraction

As mentioned above, there’s one exceptional case where the distinction between Trump’s authoritarian project and specific situations breaks down: the Mueller investigation, which went on for two years and called into question Trump’s legitimacy, both directly and indirectly. It was both an ongoing threat to his authoritarian project and the source of specific news stories to which Trump needed to respond. It makes sense that this would be the most prominent example of diversionary tweeting on Trump’s part, which is where Lewandowsky’s forthcoming paper with Jetter and Ecker comes in.

As already noted, deflection and diversion and differ in a significant way: the first shifts blame, the second shifts attention to an entirely different topic. This difference explains why a different approach is needed than simply examining tweet texts alone. We need a way to see the whole picture: What’s happening that gives rise to the distraction effort, the effort itself, and what happens as a result.  This is what Lewandowsky’s new study does — looking at coverage in the New York Times and ABC World News Tonight in tandem with Trump’s tweeting to record the diversionary effects.

Of course a diversionary tweet need not have any impact on the world in order to qualify as such. Lakoff’s initial focus was on understanding the tweets themselves, and what Trump was trying to do. What this new study looks at, then, is not whether Trump engages in such tweeting, but rather how effective it is in terms of agenda setting, and causing the media to focus on anything other than the Mueller investigation and Trump's cooperation with Russia’s attack on our democracy.

“The literature on agenda-setting basically invokes multiple actors, with a primary focus on the media. They are seen to be the principal agenda-setters,” Lewandowsky said. He provided two papers on the subject. “It’s a nuanced literature and evolving rapidly now, but it’s pretty clear that in the past the media were agenda-setters whereas now it’s been handed over to social media and fake news.” In short, Trump’s use of social media to shape coverage of him needs to be seen as part of a larger long-term shift.

What Lewandowsky’s team found was a distinct, statistically significant impact. They looked for, and found, words that Trump used more frequently after news of Mueller’s investigation appeared, and which were followed by reduced Mueller coverage afterwards. They did this first with a targeted analysis using combinations of keywords for diversionary topics taken to play to Trump’s agenda. They found a diversionary effect for "jobs," "China" and "North Korea," but not for "wall" or "immigration."

Second, they conducted an expanded analysis using word pairs drawn from Trump's entire Twitter vocabulary of words used 150 times or more. This allowed them to see if other words (and associated topics) might also have a diversionary effect. They did find some, but relatively few.  Mostly, they found additional confirmation of the targeted analysis: The word “job” or “jobs” was present in 56 out of 73 diversionary word pairs found for the New York Times, and all 48 of them for ABC News.

"The diversion works," Lewandowsky said in his presentation. Both the Times and ABC reduced their inconvenient coverage, and thus the public was "less likely to be interested in an inconvenient issue." Thus, he went on to say, "the president sets the agenda — contrary to decades of conventional wisdom on agenda-setting."

What the press could do to fight back is hardly a mystery. “Cover the real issues no matter what the president says,” Lewandowsky told Salon. “If he talks about 'Hamilton' on Broadway, publish an article about the real story — the Trump University settlement — and point out how Trump tried to divert. Reveal his techniques.” The same applies to deflection as well. Show what he’s up to, and don’t hand him the microphone to self-describe, blame-shift and gaslight. When he lies, report it using Lakoff’s “truth sandwich” approach. It’s not that complicated.

Trump’s ability to distract from such a consequential investigation led me to ask about the media's obsessive focus on Clinton's pseudo-scandals in the 2016 election, and its relative neglect of Trump's real ones. “Well, that’s a bit of a mystery,” Lewandowsky conceded. “The media ended up normalizing Trump while pathologizing Clinton — the reverse of reality.

“And this happened despite media coverage of Trump actually being quite critical," he continued. "However, even the critical coverage of his outrageous behavior ultimately just turned into free publicity for him. The reason this works is because his base considered him authentic because of those transgressions,” as described in Hahl’s “Lying Demagogue” study above. “In those circumstances, critical coverage doesn’t necessarily harm a candidate."

The deeper question here is why Trump's base feels this way. Why does one large segment of a polarized public feel that the whole system is illegitimate, to the extent that they embrace a pathological liar because of his lies, not in spite of them? Part of the answer may come from a 2016 study by Manuel Funke and colleagues in the European Economic Review, to which Lewandowsky drew my attention.

After constructing a database of more than 800 elections covering 20 advanced economies over 140+ years, the authors' analysis found that, “After a crisis, voters seem to be particularly attracted to the political rhetoric of the extreme right, which often attributes blame to minorities or foreigners. On average, far-right parties increase their vote share by 30% after a financial crisis,” an effect not seen after normal recessions or non-financial shocks.

Perhaps the rise of the Tea Party, Trump’s 2011 embrace of birtherism and his 2016 election need to be considered as typical examples of this same broad phenomenon. People who feel the system has betrayed them are looking for someone to blame. They want someone profane to fight for them. They want it so badly they’ll embrace a charlatan who’ll happily pick their pockets all over again. There is likely not much we can do about them, in the short run, at least.

But we can stop making things even worse. We can stop contributing to the ongoing destruction of our democracy.  Trump’s base is not a majority — not even close. To the contrary, once safe red states like Texas, Arizona and Georgia are now becoming electoral battlegrounds. We can reclaim our democracy — but not if the media persists in letting Trump set their agenda for them. And make no mistake, Lewandowsky’s study shows that’s exactly what much the media has done. The vaccine is out there, but our supposedly free press refuses to take it.