On Metapragmatic Gaslighting: Truth and Trump’s Epistemic Tactics in a Plague Year

Abstract

This article focuses on the nexus between political discourse and contemporary concerns about the arbitration of truth to argue that Trump’s way of using speech about speech (i.e., his metapragmatic discourse) resembles the manipulation tactic commonly called gaslighting. Based on examples drawn from 2020 (i.e., White House press conferences and electoral presidential debates), I explore Trump’s metapragmatic gaslighting both as an epistemic tactic for the manipulation of information and as an effective style of political self-presentation. By analyzing specific instances of Trump’s metapragmatic commentaries that blatantly contradict shared pragmatic principles for the interpretation of utterances’ illocutionary force and denotational content, I show how Trump is able to present himself as a champion of semantic and moral candor and simultaneously promote a highly personalist view of the meaning-making process. In so doing, I also propose a metapragmatic approach to the epistemological and political predicaments posed by the “post-truth” epistemic regime.

On October 7, 2020, during what in Europe and North America became known as the second wave of the COVID-19 pandemic, US vice president Mike Pence and his Democratic Party rival (and US senator) Kamala Harris took part in a widely watched 90-minute-long faceoff leading up to the November presidential election.¹ Minutes into the debate, Pence began to speak while Harris was still holding the floor. Reacting to the disruption caused by her opponent’s overlapping voice, Harris raised her left hand in a halt gesture and uttered in a poised but markedly firm voice what became a memorable metapragmatic statement: “Mr. Vice President, I’m speaking.” Then, after a pause, Harris repeated: “I am speaking,” at once describing what she was doing (i.e., speaking) and requesting not to be interrupted during a tightly timed debate (fig. 1).² Harris’s remark quickly became buzzworthy and turned into one of the most popular lines of the entire electoral race. In a matter of seconds, the five-word sentence originated a flurry of citational posts on social media, whereby celebrities and common users—unwittingly following the recycling and recontextualization process illustrated by Asif Agha (2011, 166–67)—would reiterate slightly rephrased versions of the statement applauding its self-assertive subtext (Spencer Saraim 2020). The statement soon began to circulate within the North American semiosphere and political mediascape in a multifariousness of verbal, visual, as well as material instantiations—through a series of viral videos on YouTube, a countless number of memes, and as a political claim embellishing a variety of electoral merchandise, ranging from T-shirts to coffee mugs, from stickers to face masks, from pin badges to fridge magnets. Undoubtedly, there was something inherently political in Harris’s metapragmatic utterance: its capacity to combine a matter-of-fact assertion with a performative evocation of women empowerment discourse.³

Figure 1. Kamala Harris: “Mr. Vice President, I’m speaking” (screenshot by author)

My argument in this article is that to understand the impact and political potential of Harris’s “I’m speaking” move (and the strict interplay of evidence and agency embedded therein) we need to discuss the role of metapragmatics (i.e., speakers’ use of language to comment upon language-in-use) within contemporary debates on fake news and post-truth. From a more general perspective, my analysis aims at showing that while post-truth poses a number of dilemmas to the anthropological imagination, a metapragmatic lens may help us recast the debate through a different (and perhaps more anthropologically viable) framework. Rather than focusing exclusively on the referential function of language (and on the correspondence between words and the world), a metapragmatic perspective reveals how, during the course of situated interactions, speakers may use language not only to manufacture specific versions and interpretations of reality, but also to lay claims on the truth value of what they say they are doing with language. Put differently, I propose a linguistic anthropological antidote to the epistemological and political challenges posed by the current crisis of facticity.⁴ Instead of getting caught into a dead-end polarization between radical forms of positivist objectivism and social constructionism, a close analysis of metapragmatic practice may further our understanding of how by making statements on their language-in-use, political actors simultaneously construct their political self and their stance on moral norms and epistemic standards of interaction. Such an approach may open up a different way of addressing the conundrums of what is currently perceived as the blurring of the distinction between facts and fiction.

During the last five-plus years, the notion of “post-truth” has saturated our mediascapes becoming a buzzword both within lay public discourse and academic debates. Dating back to the early 1990s, when it was first used in an editorial that Steve Tesich (1992) wrote for the magazine The Nation, the term reached unprecedented popularity in the mid-2010s (Kozinets et al. 2020). According to the Oxford Dictionaries, which in 2016 named it “Word of the Year,” the term post-truth denotes an epistemic regime in “which objective facts are less influential in shaping public opinion than appeals to emotion and personal belief.” In Tesich’s (1992) terms, post-truth is evocative of a “Government of Lies,” that is, a political system based on the dissemination of fake information and the legitimation of blatantly untrue facts paired with the social compliance of the citizenry, which passively accepts the systematic manipulation of truth and prefers the comfort of falsehood to the critical scrutiny of information.

The recent skyrocketing popularity of the term post-truth is clearly connected to the growing anxiety regarding truth and facticity triggered by global events and fueled by the circulation of a number of semiotic artifacts: from deep fake videos to erudite digital narratives denying climate change,⁵ from news and intelligence reports on nowhere-to-be-found weapons of mass destruction to multimodal alt-right memes denying systemic racism, from antivaxxer banners coopting slogans of the reproductive rights movement to newspaper editorials normalizing sexual assaults.⁶ As we are still trying to understand whether our present moment is the aftermath of one of the greatest pandemic outbreaks of modern history or the beginning of a new viral endemic era, we grapple with concerns about the manipulation of truth and the arbitration of reality.

Indeed, the US mediascape has become increasingly saturated with epistemic angst and social debates aimed at discriminating objective truths from so called “alternative facts.” For example, a few months prior to the Far Right US Capitol insurrection of January 6, 2021, that led Twitter and Facebook to ban Donald Trump from using their platforms, the US digital public sphere was shaken by another (only apparently less dramatic) controversy: During the tumultuous months of electoral campaign and pandemic emergency that preceded the November 2020 vote, Twitter started to apply a fact-checking label to President Trump’s tweets (fig. 2). When, on May 26, 2020, the former president posted a series of tweets suggesting that postal vote ballots would result into a rigged election, Twitter took action and applied the tag “Get the facts about mail-in ballots” to Trump’s tweets. Invoking their antimisinformation policy, a Twitter spokesperson explained to the Washington Post (Dwoskin 2020) that the rationale for the unprecedented fact-checking label stemmed from the “potentially misleading information about voting processes” contained in Trump’s tweets, which had thus to be “labeled to provide additional context around mail-in ballots.”

Figure 2. Twitter’s fact-checking label (courtesy of Twitter)

Trump’s responses were predictably enraged. On the evening of May 26, he posted abrasive reactions to Twitter intervention: “Twitter is now interfering in the 2020 Presidential Election. They are saying my statement on Mail-In Ballots, which will lead to massive corruption and fraud, is incorrect, based on fact-checking by Fake News CNN and the Amazon Washington Post,” and “Twitter is completely stifling free speech and I, as President, will not allow it to happen!”⁷

A few months later, in early October 2020, a similar type of flagging was applied to a tweet (fig. 3) in which President Trump, while still recovering from COVID-19 one day after being discharged from the hospital, had downplayed the dangers of the coronavirus pandemic.⁸ Unsurprisingly, the measure triggered a flurry of enraged tweets against “Fake News CNN and the Amazon Washington Post” by Trump and his supporters. However, while the public commentary about this mediatic dispute focused almost exclusively on issues of free speech, other aspects of the controversy remained relatively overlooked. For example, Trump’s use of his own healing body as evidence of the truth value of his claim about the allegedly mild risks of the virus epitomizes evidentiary regimes pivoting on notions of transparency and accountability and common to both populist and neoliberal democratic ideologies (Strathern 2000; West and Sanders 2003). Further, by using his body to corroborate his own medical theory, Trump positions himself as the ultimate locus of truth and promotes a “populist epistemology” characterized by the rejection of scientific knowledge and the proclamation of a highly personalist theory of meaning, which views meaning “as owned by the individual speaker and exclusively defined by his intentions” (Duranti 1988, 13; see also Rosaldo 1981; Hill 2000).⁹

Figure 3. Twitter’s reaction to Trump’s tweet about COVID-19’s mild lethality (courtesy of Twitter).

Even more interesting, aside from the discussion about free speech—an argument that was repeatedly invoked by the president himself—Trump’s stated intention to repeal Section 230 of the Communication Decency Act is noteworthy.¹⁰ By signing (on May 28, 2020) an executive order that could potentially undermine the legal immunity of social media companies like Facebook and Twitter from being liable for third-party content posted on their sites, Trump gestured toward an absolute liability regime (whereby platforms are responsible for all third-party content even without any knowledge) as a reaction to the notion of epistemic objectivity (i.e., that facts are independent from opinions and interpretations) postulated by said platform’s fact-checking practices. Put differently, Trump’s legal intervention implied a significant interconnection between the epistemic and the moral: If Twitter and Facebook have the right to fact-check the truth value of their users’ content, then they should also be liable for what their users publish on their platforms.

For anthropologists who have long conceived their work as a critical exploration of the daily operations of ideological formations based on particular interests (and for all those scholars interested in the social construction of reality) the notion of “post-truth” poses a series of problems. As Ho and Cavanaugh (2019, 160) point out, post-truth hints at a binary between self-evident objective facts and unfounded opinions—a binary that may naturalize a positivist view of science and thus neglect how scientific knowledge is always constructed, situated, and potentially usable to legitimize power.¹¹

In this article, I draw on linguistic anthropological insights (Hill and Irvine 1993; Kuipers 2013; Dent 2019) to suggest that a fine-grained analysis of the specific discursive practices through which socially situated actors construe authoritative interpretations of facts and lay claims to truth may help resolve the moral, political, and epistemological predicament anthropologists face in a time in which post-truth (and the rigid binary between objective factual evidence and “alternative facts”) has become a seemingly unavoidable frame of thought. In this light, the concern about the arbitration of reality that characterizes our contemporary mediascapes is, at least in part, a function of a referentialist bias, which reductively conceptualizes language as a symbolic-denotational system made of words that stand for things in the world and designed to communicate information (Silverstein 1976).¹² Indeed, as Hill (2008, 39) highlighted, this “referentialist ideology makes the question of whether or not statements are ‘true’ into a very salient issue.” I use examples drawn from Trump’s discursive practices and epistemic tactics to argue that an important (but not adequately explored) side of contemporary concerns with facticity involves the metapragmatic function of language, that is, not so much how language is used to represent facts (i.e., its denotational function), but how language is used to reflexively represent how language is used.

Trump and the Epistemic Violence of Metapragmatic Gaslighting

The term metapragmatics was originally introduced by Michael Silverstein (1976, 1993) to capture a specific form of metasemiotic activity whereby speakers explicitly (e.g., through overt verba dicendi, as in Kamala Harris’s case) or implicitly (i.e., through a variety of nonreferential indexical cues, such as prosody, voice quality, and indirect appeals to socially identifiable genres and registers) interpret and provide signals on how to construe ongoing discursive activity and appropriately use language.¹³

Put differently, metapragmatics can be understood as a specific dimension of the reflexive function of natural languages—one that concerns language’s potential to be used to describe how language is or should be used to perform actions. In linking the performative and reflexive function, metapragmatics constitutes a key dimension of political talk. It may be argued that what makes language distinctively political is its capacity to be used not only as a performative dispositive to “do things with words” but also as reflexive instrument to express explicit commentaries on (and implicit allusions to) how language itself is or should be used (see, e.g., Lempert and Silverstein 2012).

Since his rise to the political stage in 2015, Donald Trump’s unconventional rhetoric and “unpresidential” style of self-presentation have attracted the attention of laypersons, political commentators, as well as scholars of language and communication.¹⁴ Several commentators have highlighted Trump’s linguistic incoherence (Slotta 2020), inarticulacy (Leith 2017), unintelligibility (Sclafani 2017), abrasiveness (McIntosh and Mendoza-Denton 2020), childishness (Donzelli and Budgen-Powell 2019), and defiance of Grice’s conversational implicatures (Hodges 2020). Drawing on these analyses, this article further explores Trump’s irreverent subversion of shared principles of communicative interaction by focusing on a type of epistemic violence that goes under the label of gaslighting.

The term comes from the title of a 1938 stage drama, Gas Light, which was later adapted for the screen and made famous by George Cukor’s Oscar-winning film starring Ingrid Bergman and Charles Boyer. In this black and white slow-burn mystery thriller, a husband (played by Boyer) develops an elaborate plan to rob his wife Paula (Bergman) of some valuable jewelry and have her committed to a mental institution. To accomplish his manipulative project, the husband tries to convince Paula that she is imagining things, in particular the occasional dimming of their home’s gas lamps. Over time, Paula starts to believe her husband’s lies and to question her own grasp on reality.

In more recent times, the term gaslighting has become current in popular psychology and self-help literature as a label for radically deceptive behavior deployed for bluntly denying infidelity. Even when confronted with factual evidence, gaslighters firmly deny that they are cheating and claim that their partners’ suspicions are not based on facts but derive instead from paranoia and unfounded jealousy, driving, in the long run, the gaslit counterpart to become unable to discriminate facts from fantasies. This lay notion has also inspired a number of scholarly analyses from various disciplinary perspectives, ranging from sociology (Sweet 2019) to psychology (Stern 2007; Stark 2019), from epistemology (Ivy 2017) to communication studies (Graves and Spencer 2022). In this scholarly literature, gaslighting refers to a form of psychological abuse and epistemic violence pivoting on a discursive strategy whereby an alternative and unrecognizable reality is presented to the interlocutor as unquestionably true. Gaslighting thus entails distorting an empirically verifiable reality and deliberately presenting false information to the gaslightee with the aim of making them question their own memory and perception of events and destabilize their sense of reality. As Graves and Spencer (2022, 49) point out, however, this perspective assumes, often implicitly, “that the competing knowledge claims of gaslighter and gaslightee should be adjudicated with reference to an objective truth criterion.” Such a conceptualization, modeled on the plot of Cukor’s movie and its emphasis on the manipulative distortion of objective reality, neglects the essential insight shared by scholars interested in the social construction of reality “that all knowledge claims are necessarily partial” (49).

In what follows, I offer a reflection on this model’s shortcomings (and the general predicament underlying mainstream notions of post-truth) by adopting a metapragmatic approach to gaslighting—one that instead of seeking to assess a simple correspondence between language and objective reality focuses on what speakers claim they are doing as they are using language to objectify what they are doing with language. More specifically, I draw on examples extracted from White House press conferences held between March and August 2020 and on the presidential debates held in the fall of the same year to argue that Trump’s reflexive use of language to referentially describe how language is used is driven by a gaslighting approach to the meaning-making process.¹⁵ My goal is to show that Trump’s metapragmatic discourse is both an epistemic tactic for the manipulation of information and an effective style of political self-presentation. By analyzing specific instances of Trump’s metapragmatic commentaries, which blatantly contradict shared pragmatic principles for the interpretation of utterances’ illocutionary force and denotational reference, my analysis shows how Trump challenges liberals’ allegedly circumlocutory speech and presents his manipulative discursive practice as a paradigm of semantic directness and moral candor, promoting at once an agentive political self and a highly personalist view of language and meaning. Let me start with an example taken from the first presidential debate held in the fall of 2020.

The Metapragmatics of Political Condemnations

On the last Tuesday of September 2020, the incumbent president and his challenger Joe Biden appeared in front of the American public for what was supposed to be the first of three 90-minute-long presidential debates.¹⁶ About one hour into what was described as one of the most chaotic presidential debates in history, the moderator, Chris Wallace of Fox News, asked President Trump if he was willing to condemn white supremacists and right-wing militia members who had contributed to create a climate of civil unrest in several US cities:¹⁷

In spite of its semantic content, Trump’s interjection “That’s right!” (line 2) is all but an indication of agreement. As signaled by its overlap with the qualifier “left-wing” uttered by Wallace, Trump’s exclamation “That’s right!” wants to underline his criticism of Antifa and of the Democratic nominee’s (to whom Wallace refers as vice president) lack of condemnation of Antifa militants.

Wallace thus explicitly tries (lines 3 and 7) to engage President Trump in an unequivocal expression of moral and political disapproval of the actions (and presumably the ideas) of the white supremacist groups who constitute part of his base. The prosody of Wallace’s speech is characterized by shifts in volume and intonation and meaningful pauses as a way to invite Trump to provide a definitive answer to the question (fig. 4). But once again, after expressing twice his alleged willingness to discredit white supremacist groups (lines 4 and 6), Trump backs off:

Indeed, in spite of his repeated declarations of willingness (“Sure!” at lines 4 and 6) to comply with Wallace’s (and then Biden’s at line 12) request to perform an expressive act of political condemnation, Trump first pivots back (line 8) to denounce the left and then pretends not to be able to identify who the target of the required condemnation should be: “What do you wanna call them? Give me a name.” “Who do you want me to condemn?” (lines 13 and 15):

According to Searle’s (1976) classification of illocutionary speech acts, condemnations are expressive speech acts in that they entail expressing a critical stance toward a violation of a rule and/or strong moral disapproval of prior transgressive acts. As Kampf and Katriel (2016, 4) point out, in order to be successful, condemnations require that (1) speakers identify a past action as transgressive of a norm (preparatory condition); (2) perceive such action as deplorable, that is, as (a) having some negative effect and/or (b) as something they would not do (sincerity condition); (3) represent the transgressive act in their speech (propositional condition); and (4) use a linguistic formula or a performative verb expressing a condemnation (essential condition).¹⁸

Figure 4. Chris Wallace: “Are you willing to condemn?” (screenshot by the author)

Interestingly, despite his apparent uptake and stated intention to comply with Wallace’s request, Trump fails to perform all the four required conditions. Not only he refuses to identify the facticity and (moral unacceptability) of prior deplorable acts and to identify the perpetrators of such acts (conditions 1, 2, and 3), but when it comes to actually expressing the required condemnation and to commit his white supremacist supporters to a future course of events, he changes the form and the meaning of the directive verb (“stand down”) Wallace had invited him to use and opts, instead, for a different directive. Through a masterful play with prepositions, Trump transforms the phrasal verb originally used by Wallace “stand down” into an appeal to white power groups to be ready for insurgence (i.e., “stand back” and “stand by” at line 18). Unable to contain his bewilderment, Biden intervenes, highlighting how even FBI director Christopher Wray expressed concern about domestic terrorism and white supremacist violence:

The metapragmatics of Trump’s condemnation of the Proud Boys’ actions blatantly contradicts shared pragmatic principles of interaction. Indeed, rather than condemning white power groups, Trump expressed “an unambiguous call to them to be ready” (Belew 2020). Trump’s metapragmatic gaslighting conduct was also recognized and criticized by Republican Tim Scott, the only Black Republican in the Senate, who, suggesting that the president “misspoke” in failing to explicitly condemn white supremacists and violent right-wing groups, urged him (unsuccessfully) “to correct” his comments. The next section examines a different case of “misspeak,” which, according to Trump, entailed a deliberate misquote by the media. Let us take a closer look at this alleged incident.

The Metapragmatics of Quotes and Misquotes

Another remarkable instance of Trump’s metapragmatic gaslighting practices occurred during one of the daily Coronavirus Task Force briefings held at the White House during the first pandemic spring. In this second example, Trump recruits another person, CDC director Robert Redfield, to rectify the fact that the media had “totally misquoted him.” In order to understand the excerpt some additional background context is needed. On April 21, 2020, a Washington Post headline announced that: “CDC director warns second wave of coronavirus is likely to be even more devastating.” The article (Sun 2020) stated that:

Even as states move ahead with plans to reopen their economies, the director of the Centers for Disease Control and Prevention warned Tuesday that a second wave of the novel coronavirus will be far more dire because it is likely to coincide with the start of flu season.

In her piece, Sun (2020) went on to quote what Redfield himself had said in his interview with the Washington Post:

“There’s a possibility that the assault of the virus on our nation next winter will actually be even more difficult than the one we just went through. … And when I’ve said this to others, they kind of put their head back, they don’t understand what I mean. We’re going to have the flu epidemic and the coronavirus epidemic at the same time.”

The article then continued to quote Redfield, albeit this time without enclosing his words within quotation marks:

Having two simultaneous respiratory outbreaks would put unimaginable strain on the health-care system, he said.

The following day (April 22, 2020), during the usual briefing, President Trump expressed great discontent toward the Press and accused the Washington Post of having “totally misquoted” Redfield’s original words. Here Trump’s (vicarious) metapragmatic gaslighting tactic aims to characterize an accurate quote as misquote. The speech below is a slight adaptation of the transcript available on C-Span.org (an online cable channel that offers complete coverage of the House and Senate congressional hearings and White House events). The most significant passages are highlighted in bold and the time codes (TC) refer to the video that can be viewed on said platform:¹⁹

Trump: I do wanna mention a man who’s done a very good job for us: Dr. Robert Redfield. He was totally misquoted in the media on a statement about the fall season and the virus. Totally misquoted. I spoke to him. He said it was ridiculous. He was talking about the flu and corona coming together at the same time. And corona could be just some little flare-ups that we’ll take care of. We’re going to knock it out. We’ll knock it out fast. But that’s what he was referring to “coming together at the same time.” (TC 01:08)

After having introduced the metapragmatic issue he wanted Redfield to rectify, Trump expressed his usual abrasive criticism toward the media and invited Dr. Redfield to take the floor to straighten things out:

And I think rather than waiting, I’d ask Dr. Redfield to come up, say a couple of words just to straighten that out, because he didn’t say it was a big—a big explosion. The headline in the Washington Post was totally inaccurate. The statement wasn’t bad in the Post, but the headline was ridiculous, which is—as I say, that’s fake news. And CNN is fake news like crazy, and they had just totally the wrong story, which they knew. They were asked to change it, and they wouldn’t do that. And it was false. … So I’ll ask Dr. Redfield, … to come up and explain. Please. Thank you, Doctor. (TC 01:40)

Redfield promptly complied with the president’s request and launched himself into a rather clumsy explanation of how he was allegedly misquoted:

Thank you, Mr. President. I really do think it’s important to clarify this as we build the confidence of the American people. … When I commented yesterday that there was a possibility of the fall, winter … it could be more difficult, more complicated when we had two respiratory illnesses circulating at the same time: influenza and the Coronavirus-19. (TC 02:23)

Redfield first tried to draw a distinction between “worse” and “more complicated”:

But I think it’s really important to emphasize what I didn’t say. I didn’t say that this was going to be worse. I said it was going to be more complic— or more difficult and potentially complicated because we’ll have flu and coronavirus circulating at the same time. (TC 02:49)

As it soon became clear, however, the Washington Post did not misquote Redfield. On the contrary, the paper accurately reported Redfield’s actual words. As his speech unfolded, Redfield appeared caught up in a nearly impossible attempt to disentangle himself from the metapragmatic predicament created for him by Trump.²⁰ Redfield, thus, sought to explain that although he was not actually misquoted, the Washington Post did not adequately understand the perlocutionary goal of his statement, which should have been interpreted as “an appeal” to the American citizens “to embrace” the flu shot “with confidence”:

The key eh eh to my comments and the reason that I really wanted to stress them was to appeal to the American public to embrace the flu vaccine with confidence. One of the greatest tools we have as we go through the fall-winter season that we’re into is to get the American public to embrace the influenza vaccine and thereby minimize the impact of flu to be the co-respiratory disease that we confront. Thank you very much. (TC 04:18)

Noticing Redfield’s embarrassed reaction, a journalist began to press him about the accuracy of the metapragmatic qualification of “having been misquoted”:

Press: Could I just ask a follow-up on that, Dr. Redfield? Sir? Just to clarify your comments? (TC 04:50)

In vain Trump sought to prevent the questioning by claiming that there was nothing to follow up to. He kept on insisting that the Washington Post completely misquoted Redfield and misunderstood the ultimate goal (i.e., perlocutionary intention) of the Doctor’s statement, that is, encouraging people to get their seasonal flu shot:

Trump: I don’t know what’s to follow up. He was misquoted. Totally misquoted. He said they could come together. … And his whole purpose in making the statement was to get a flu shot, so that next fall we don’t have such a big season of flu. (TC 04:54)

The journalist, in fact, persevered in his attempt to clarify the metapragmatic point raised by Trump’s claim that Redfield had been misquoted and misinterpreted. To do so, the journalist directly addressed Redfield, reading out to him the actual quote from the Washington Post. As Nakassis (2012, 624) points out, citations (whether direct or indirect) are always made visible through a series of semiotic metamarks, which may include punctuation (e.g., colons and quotation marks), sentence structure (e.g., deployment of verba dicendi and of the complementizer “that”), indexical elements (e.g., pronouns, demonstratives, temporal and locative adverbials, etc.) that may shift according to direct and indirect speech frames. In reanimating the quote from the Washington Post, the journalist, thus, emphasized the semiotic cues indicating that Redfield’s words were actually a citation (i.e., pronominal deixis I and they and the citational framing “they quoted you as saying”):

Press: Okay. So, Dr. Redfield, the Washington Post—which, you did the interview with them—they quoted you as saying: “There’s a possibility that the assault of the virus on our nation next winter will actually be much—even more difficult than the one we just went through. And when I’ve said this to others, they’ve kind of put their head back, they don’t understand what I mean. We’re going to have the flu epidemic and the coronavirus epidemic at the same time.” Is that what you said to the Washington Post? (TC 05:34)

Redfield had thus to admit that the quote read aloud by the journalist was accurate, but insisted in claiming that what he meant by it was not that the situation was going to be “worse,” but just “more difficult,” due to the likely circulation of two viruses at the same time:

Redfield: Yeah, that’s what I was trying to say to you just a minute ago—that the issue that I was talking about, about being more difficult, is that we’re going to have two viruses circulating the same time. … so the comment that I made “it’s more difficult.” doesn’t mean it’s going to be more impossible. It doesn’t mean it’s going to be more, as some people have said, “worse.” It just means it’s going to be difficult because we have to distinguish between the two. And what I was wanting to do and what I want to do again here is appeal to the American public to … help, … by getting the flu vaccine and taking flu out of the picture. (TC 05:58)

The journalist insisted:

No, but—but, I’m sorry, but that quote that I just read was accurate—right, sir? Because that’s the quote from the Washington Post. You were accurately quoted, correct? (TC 07:06)

Eventually, Redfield gave in and slightly bowing toward the journalist (fig. 5) had to admit that he had not been misquoted: “I’m accurately quoted in the Washington Post as ‘difficult.’” But he insisted in adding that: “the headline was inappropriate.”

Figure 5. Robert Redfield bowing toward the press as he acknowledges he had not been misquoted (screenshot by author).

Redfiled’s capitulation shows how Trump’s attempts at coopting others into his signature gaslighting are not always successful, which, in turn, suggests how the tactic is strongly associated with Trump’s political persona. Metapragmatic gaslighting is, indeed, inflected with a markedly personalist bend. In her discussion of the metapragmatic production of specific structures of agency and personhood, Jane Hill (2000) described the key role played, within US political discourse, by the implicit belief that the meaning of utterances is determined by the speaker’s intentions. As my analysis illustrates, Trump’s manipulation of facts is mediated by deceptive metapragmatic commentaries on what he actually does with language. By engaging in metapragmatic gaslighting, Trump projects a highly agentive political persona—one who is capable of subverting implicit norms of interaction (e.g., what counts as a misquote or how to perform a condemnation) and, at times, even reformulate the pragmatic rules that regulate the relation between words and their denotata. A case in point is illustrated by another dispute (discussed in the next section) on how to identify the anaphoric reference of a pronoun in a given sentence.

The Metapragmatics of Pronominal Reference

My final example of Trump’s metapragmatic gaslighting entails the manipulation of pronominal reference and illustrates how Trump seeks to subvert established pragmatic norms for inferring relations of co-reference between a pronoun and a nominal antecedent. In this excerpt drawn from the Corona Task Force briefing occurred on March 16, 2020, a journalist explicitly questioned Trump’s use of language, by calling him out on the expression “under tremendous control.” The question came after Trump had been offering a somewhat muddled report on the initiatives undertaken by the federal government regarding paid sick leaves for employees infected with COVID-19, the availability of ICU beds, and the provision of medical equipment: “we’re ordering tremendous numbers of ventilators, respirators, masks … and they’re coming, and we have quite a few at this point. I think, Mike, we have a lot.”²¹ In light of the somewhat confused account provided by the president, reporter Kaitlan Collins pointed out that the picture emerging from the report did not suggest that the pandemic was on the verge of disappearing anytime soon, which contradicted the bold statement Trump had made the day before:

Journalist: Okay. And yesterday, you said that this was “under tremendous control.” Do you want to revisit that statement if we are going to be experiencing this until July or August—five more months ahead of where we are now? (TC 16:49)

Indeed, during the briefing occurred the day before (March 15, 2020), Trump had claimed that: “It’s a very contagious virus. It’s incredible. But it’s something we have tremendous control of.”²² As the journalist pointed out in her question, the scenario provided by the experts clearly clashed with the idea of the virus was being under tremendous control. Indeed, in that same briefing, Anthony Fauci, the chief medical advisor to the president, had declared that “the worst is yet ahead of us.”²³

Trump’s answer to the journalist, however, did not address the factual domain of the pandemic situation, but it rather focused on the anaphoric reference of the pronouns he had used and revolved around an attempt at providing a different interpretation for the full noun phrase the pronouns it and something referred to. Contrary to the journalist’s understanding and to the general pragmatic rule according to which a pronoun anaphoric reference is to be retrieved in the most recently mentioned entity that syntactically matches the pronoun, Trump explained that his sentence should not be construed as meaning that he had said that COVID-19 was under control. What he meant was, instead, that the government’s efforts to control the situation displayed remarkable control and coordination. Trump’s metapragmatic gaslighing effort was thus aimed at suggesting that the correct interpretation for his sentence was that the “control was under tremendous control”:

Trump: Well, when I’m talking about control, I’m saying WE are doing a very good job within the confines of what we’re dealing with. We’re doing a very good job. There’s been a— there’s been a tremendous eh amount eh of the way they’re working together. They’re working hand-in-hand. I think they’re doing, really, a great job. And from that standpoint, that’s what I was referring to. Yeah, Steve, go ahead. ((Trumps dismisses the journalist to take another question)) (TC 17:30)

Trump’s twisted tautological self-exegesis left the questioner baffled. The journalist, thus, pressed Trump to clarify the actual referent of the pronoun it. Attempting to expose Trump’s contradiction with his own March 15 statements, the journalist (fig. 6) first tried to lead him to agree that the COVID-19 outbreak was all but under control:

Journalist: But you’re not saying “it’s [COVID-19 is] under control,” right? (TC 17:20)

Trump readily agreed that no, by “it,” he did not mean COVID-19, which could not possibly be under control:

Trump: I’m not referring to “it” meaning the= (TC 17:22)

Journalist: =coronavirus? ((offering the word)) (TC 17:24)

Trump: Yeah, if you’re talking about the virus— (TC 17:26)

Journalist: Yeah. (TC 17:27)

Trump emphasized that indeed nowhere in the world the virus was under control and implied that nobody of sound mind could make such claim:

Trump: No, that’s not under control for any place in the world. … (TC 17:28)

It is worth noticing that by stating that COVID-19 was rampant everywhere in the world, Trump could both downplay responsibility for the virus not being under control in the United States and provide evidentiary corroboration for his denial that he ever intended that “it” referred to COVID-19. The journalist thus remarked that by saying that “it” should not be construed as referring to “COVID-19” (being under control), Trump was contradicting himself given that the day before he had actually stated that “it” (meaning COVID-19) was indeed under control:

Journalist: Okay. Yesterday you had said it was, so I just … wanted to clarify. (TC 17:30)

Trump, however, firmly sticking to his metapragmatic gaslighting, insisted in denying that he ever intended to connect the pronoun it to the noun phrase “the virus” and eventually dismissed the journalist by taking another question:

Trump: No, I didn’t. I was talking about what we’re doing is under control. But I’m not talking about the virus. Yes, please ((turning to another journalist)) (TC 17:32)

As many have noticed, Trump’s irreverent use of profanities, his exclamatory and hyperbolic style, and his discursive incoherence often paired with straight ungrammatical constructs (see McIntosh [2020] for an excellent overview) are far from casual and have greatly contributed to Trump’s political success. Most observers and commentators, however, have focused on a number of overt semantic and grammatical features, paying less attention to more subtle and less segmentable and lexicalized features of language.²⁴ The exchanges I analyzed above indicate that an important (albeit not fully explored) aspect of Trump’s eccentric discursive style concerns his metapragmatic subversion of language use rules.

Figure 6. Reporter Kaitlan Collins: “Do you want to revisit that statement?” (screenshot by author).

In a seminal article, Silverstein (1981) pointed out the limits of ordinary metalinguistic awareness and noted that the referential function of language is generally available to speakers’ for conscious comment, while other pragmatic and indexical functions seem to escape speakers’ awareness (especially when these indexical and pragmatic elements are capable of creatively transforming the context of interaction). Drawing on the analysis of Trump’s metapragmatic subversions, I argue that departing from an exclusive focus on the referential function of language (and from the narrow correspondence between language and object) and attending to linguistic processes that fall beyond speakers’ metalinguistic awareness may provide an exit strategy from the intellectual bottlenecks of our present post-truth era.

Conclusion

Our present moment is fraught with concerns about the manipulation and the arbitration of truth and is characterized by a pervasive reformulation of long-standing structures of knowledge and agency. New notions of truth and responsibility are redefining contemporary forms of material and immaterial production disseminating novel standards of transparency, accountability, and new notions of citizenship and entrepreneurialism (see, among others, Strathern 2000; West and Sanders 2003; Gershon 2011). Often delegated to the media industry and to fact-checking firms, the efforts to ascertain real facts from fake news constitute a prominent feature of the public sphere in the United States and beyond (Ho and Cavanaugh 2019; Overell and Nicholls 2019). The ongoing viral contagion has enhanced the tensions between reality and its representation, saturating our mediascapes with a profusion of epidemiological and statistical data and exacerbating collective anxieties about the arbitration of truth. This article engaged the nexus between political discourse and contemporary concerns about truth and facticity. Based on examples extracted from public speeches performed during the first phase of the COVID-19 pandemic outbreak, I explored Trump’s metapragmatic discourse as a tactic of political self-presentation and epistemic manipulation.

Since his first appearance on the political stage, Trump’s talk has become the object of remarkable attention due to a series of distinctive features such as a profuse use of adjectival phrases, expletive utterances, hyperbolic statements, and performative epistemic modals (i.e., commissives like “I promise,” “I swear,” “I assure”). By deploying these discursive features Trump expresses “depth of conviction” (McIntosh 2020, 27) and epistemic commitment to the truth of his sentences’ propositional content. Although (or precisely because) he provides no evidential grounds (i.e., no indication of the source of information his claims are based on and proof that corroborates the truth value of his assertions) for his statements, Trump is able to present himself as referentially straightforward and capable of assertive stance-taking acts. Through these discursive moves, Trump conceals his metapragmatic manipulations and the ideological coefficient of his statements, appearing as a charismatic speaking subject who has full cognitive and pragmatic control of the world.

These discursive features have greatly contributed to Trump’s success as a populist president in that they seem to respond to the sweeping metalinguistic preoccupations (held by wide sectors of the conservatives and the progressives alike) that—due to the overly cautionary approach embraced by advocates of the so-called political correctness (PC)—words have departed from their referents and utterances have lost their illocutionary force. By defying the allegedly unassertive and hypocritical talk of PC liberals, Trump promotes a highly personalist view of meaning (Rosaldo 1981; Duranti 1988, 1993; Hill 2000, 2008) and a ludic form of authoritarianism (Dent 2019), which, in turn, are based on a bold and irreverent approach to expert opinion and scientific evidence but also on open disobedience to the conventional pragmatic principles of interaction. Aside from conferring him an aura of authenticity (Stolee and Caton 2018, 157), Trump’s norm-shattering approach to speech has been driven by the effort to manufacture a distinctive way of using language as a means to project the image of an unfiltered, transparent, emotionally earnest, referentially direct, and morally authentic political leader, who is a champion of both semantic candor and moral character. In this perspective, Trump’s techniques of self-presentation are meant to appeal to those “who are not afraid of speak their minds,” who refuse to self-censorship to protect the oversensitive and who, like him, “do not have time for total political correctness” (McIntosh 2020, 9–10).

My analysis aimed to illustrate the important role that metapragmatic mediation plays in Trump’s tactics of epistemic manipulation. Originating from the title of a 1940s film, gaslighting has been described as a form of psychological abuse whereby false information is deliberately presented as true. By representing as a misquote what was in fact a verbatim citation or by blurring the distinction between political endorsement and condemnation of alt-right white supremacists, Trump is not only able to practice “plausible deniability” (Hodges 2020) and assert his preferred version of truth, but he can also perform a charismatic political persona based on a proprietary ideology of meaning: “You are not entitled to interpret the meaning of my utterances because I am the only one who gets to decide over their interpretation.” These epistemic tactics allow Trump to promote a personalist view of meaning-making processes—one that denies the interlocutors’ collaboration and represents meaning as established solely by the speaker’s will (Duranti 1993; Hill 2000).

To go back to my opening vignette, within a public discourse pervaded by continuous gaslighting, Kamala Harris’s “I’m speaking” statement stood out as a memorable (and nearly revolutionary) example of metapragmatic candor. Indeed, one of the consequences of Trump’s flabbergasting metapragmatic conduct is that, much like gaslit people exposed to the epistemic violence of protracted manipulation, those who are critical of his politics often strive to establish some sort of universal immovable truth as a reaction to Trump’s blatant and reiterated discursive deception. The risk, however, is that, as we try to ascertain the real from the fake, we plummet into a fact-checking obsession and, clinging to an untenable binary between objective truths and “alternative facts,” we lose track of the importance of taking a stance—a political one—about what we deem right and what we consider discriminatory and unjust.