Skip to navigation – Site map

HomeAnglophonia31Varia‘I will never concede’: Donald Tr...


‘I will never concede’: Donald Trump’s discourse of denial on Twitter (Nov. 4th 2020 – Jan. 8th 2021)

Sandrine Sorlin


This article concentrates on US President Donald Trump’s last months in power and studies how he manages to deny defeat until the very last day of his presidency. Focusing on his Twitter communication extracted from the Trump Twitter Archive from November 4th 2020 to January 8th 2021 when his Twitter account was suspended, Sandrine Sorlin shows the continuity and changes in Trump’s style at the very end of 2020, highlighting how he goes about sustaining faith in his victory through the specific use of negation. This article also delves into Trump’s strategic use of personal pronouns and (in)direct address that seek to intimidate weak Republicans. Through what she calls a ‘rhetoric of the iceberg’, presenting only some evidence of the Republican victory while giving the illusion of ‘massive’ proof underneath, and through constant repetition in the Twitter echo chamber, she demonstrates how Trump succeeds in kindling his fans’ frustration and leading them to the Capitol on January 6th 2021.

Top of page

Full text

1. Introduction

1Trump’s distinguished ‘linguistic style’ has been much written about (Hodges 2019, McIntosh & Mendoza-Denton 2020, Scalfani 2018, Schneider and Eiterlmann 2020, Skinnell, 2018, Stolee and Caton 2018 among many others). This article focuses on the specific features in terms of linguistic and rhetorical strategies of Trump’s final two months on Twitter in the context of the disputed elections on the Republican side. It offers a qualitative study that aims to answer the following questions: how did Trump manage to ‘never concede’ and sustain belief in his victory until the end? I will show how the 45th president of the USA ‘weaponises’ Twitter (Bialy, 2017) transforming a social media into a battlefield where soldiers are being prepped up to fight and defend their candidate. The mirror effects produced by the medium, where only ‘digital doppelgangers’ (Kalpokas, 2019: 60) meet, block access to any form of dialogism or contradiction in Trump’s tweets and retweets, the point being not to convince anyone but to sustain people’s faith and turn it into an incentive for action. Twitter has definitely proved to be the most fitting medium for a President ready to push his fans into total denial through direct communication with them and through them.

2This article intends to show how Trump succeeds in denying defeat in spite of the stubborn resistance of the facts and how he manages to construct an alternative world in which he has always won. As the chance of contesting consent depleted and after Vice President Pence’s refusal to block the confirmation of Biden’s election at the joint session of Congress, Trump seems to have turned to his fans as a last resort. The article shows the evolution of his strategy of intimidation towards a variety of targets including the traitors in his own party.

2. Corpus and Methodology

  • 1 All the emphasis in bold in the following tweets is mine.

3The corpus under study extends from the day after the US presidential election (November 4th 2020) until the day Trump’s Twitter account (@realDonalTrump) was suspended on January 8th 2021, amounting to 1523 tweets in total.1 They were all retrieved from Trump Twitter Archive:​.

4Between February 2013 and 2019, the number of postings by Trump has been shown to average ‘15-18 tweets a day’ (Ouyang and Waterman, 2020: 55). On each of the 64 days between the election and the assault on the Capitol, Trump tweeted 23.76 messages on average, which is above what previous studies have noted, demonstrating the urgency of the moment for the president. It should be noted that more than half of Trump’s tweets (810) in this period are retweets of others’ messages of support. I have chosen to leave aside the retweets that did not emanate from Trump directly, which reduces the corpus to 713 tweets – a mean number of 10 a day and a total number of 17,628 words.

5My interest is not to give quantitative analyses of Trump’s tweets as compared to other politicians’ use of tweets or retweets but to determine the specific features of Trump’s rhetoric in a specific period of time. I started following his tweets carefully from November 4th on because my first intention was to study how he was going to eventually concede victory and/or recognize defeat. As the days went by, not only did Trump go further and further into denial but he adopted strategies that gave his fans the certitude of imminent victory. This article intends to bring to the fore the linguistic and rhetorical tools he used to sustain belief in victory until January 2021.

6Although I annotated the 713 tweets manually, I tested hypotheses about recurrent features using the software AntConc in order to establish frequencies.

7I will show in the following sections of my analysis that while Trump’s characteristic style remains fundamentally the same, there are crucial differences due to the unprecedented context at the end of 2020: yesterday’s alliances crack apart, bringing Trump to dis-align with former allies. The last section foregrounds Trump’s strategic use of personal pronouns, address, and negation at the end of 2020. Lastly it delves into a rhetorical strategy – that I choose to call ‘the rhetoric of the iceberg’ – consisting in making people believe there is more evidence than meets the eye, which is a way for Trump to constantly maintain a sense of stolen victory. His iceberg rhetoric gives the illusion throughout the period under consideration that victory is easy and within sight, arousing people’s frustration and (indirectly) asking them to act on his words.

3. Old style and new alignments

3.1. Vague quantifications and impersonal stance

8 Much of what is now recognized as Trump’s characteristic style is still vibrant at the end of 2020 (see also McIntosh, 2020 for a synthesis of newspaper articles commenting on his language). One of his trademarks is the ‘epistrophic punctuation’, that is ‘the repetition of short phrases, often ones that convey an affective or epistemic stance’ appearing at the end of rhetorical units in his tweets (Scalfani, 2018: 39). Although the remark ‘so sad’ that Trump uses frequently still occurs seven times in the corpus (‘so sad for our country!’), from the day after the election on it has been superseded in quantity by a new epistemic stance. Trump repeatedly punctuates his Twitter statements with ‘rigged election’ (29 occurrences, almost one every other day) – often capitalised for additional emphasis.

  • 2 Taking up Altenberg’s classification, Stange focuses on ‘amplifiers’ as a subcategory of intensi (...)

9Keeping up with his tendency to use imprecise quantifications (see Egbert and Biber, 2020: 28-29 for instance), Trump is able to convey the impression of rampant fraud. His favourite quantity (‘hundreds of thousands’) recurs 17 times, notably to refer to the votes that Republican observers were not allowed to see, or to illegal tampering, for instance. Here is one example: ‘They shut out our Poll Watchers and Observers. Those hundreds of thousands of votes should not be allowed. Corrupt politics’ (Nov. 10th 2020). At a time when precise quantifications mattered more than ever, Trump’s imprecise language did not undergo the slightest change. On the contrary, it is politically expedient to conceal his lack of tangible evidence (see next section). If Trump still ‘sprinkles’ semantically vacuous intensifiers (Stange, 2020: 88-89) such as the maximizers ‘totally’ (9 occurrences) and ‘really’ (10), or the boosters2 ‘so’ (24) and very (48), he also has heavy recourse to the adjective ‘massive’ that has a similar amplifying effect. Occurring 31 times in 64 days, ‘massive’ seems to have replaced Trump’s most frequently used adjective – ‘tremendous’ appears only 7 times in this corpus – hyperbolically expressing great quantity and sparing him from supplying accurate numbers. Interestingly, this adjective co-occurs with both negative notions such as ‘fraud’ or ‘corruption’ but also referents that are supposed to enhance Trump’s position, such as the massive evidence gathering to prove corruption or people’s massive support in his favour. Moreover, the specific details about where his evidence comes from and the origins of his supporters are intentionally left unclear. He creates the impression of a generalised phenomenon creeping up ‘all over the country’ (6 occurrences).

10Trump’s identified habit of sheltering behind what Egbert and Biber (2020: 32) call ‘impersonal stances’ continues in the post-election tweets. Trump has indeed been shown to favour persuasive statements ‘that are high in certainty but lacking in evidentiality’ (2020: 33): his assertive statements are never substantiated by precise proof or associated with the source from which he derives his statements. With the exception of the assertion ‘I believe we will prevail’ (Nov. 23rd) positioning Trump as the source of the belief, Trump usually leaves the source of his stances uncertain and never uses evidential markers that would convey a lower degree of certainty. The ad populum fallacious argument of the type ‘everybody knows…’ he recurrently uses also partakes of the ‘impersonal stance’, turning a controversial issue into common knowledge for all people:

(1) Everybody knows that we won the state (Nov. 13)

(2) Everybody knows it was Rigged. They know Biden didn’t get more votes from the Black community than Obama & certainly didn’t get 80,000,000 votes. (Nov. 29)

To the indefinite totalizing pronoun (‘everybody’) must be added the use on many occasions of the indefinite third person plural ‘they’ whose reference is not specified, as in (2).

  • 3Wow. This is exactly what happened to us. Great courage by judge!’ (Nov (...)
  • 4Michigan AG Dana Nessel Pursues Sanctions Against Lawyers Questioning Election (...)

11 Lastly, the defeated President maintains the superpower attitude of the former reality showman firing people on The Apprentice, putting people in their places through insults or praise (see Schubert, 2020: 199). In these two months during which Trump has been seeking support, those who are not worthy of his trust or have nothing to bring him anymore are denigrated or roughly fired. The President carries on his practice of distributing stars to the good pupils. The Secretary of State Mike Pompeo denying the election result is referred to in these laudatory terms: ‘That’s why Mike was number one in his class at West Point!’ (Nov. 10th 2020). Intelligence is assessed in accordance with the level of support shown for Trump. Senator John McCain (regarding his involvement in the passing of documents alleging Trump’s collusion with Russia) does not fare so well: ‘Check out last in his class John McCain, one of the most overrated people in DC’ (Dec. 17th 2020). Conversely, he heaps praise on the whistle-blowers, tipsters or other judge and lawyers of ‘great courage’3 that have come forward with evidence of voter fraud, granting them the status of ‘true patriots’4. By retweeting his supporters’ messages, Trump both gives them the attention they seek from the president and also prompts others to follow suit.

3.2. Disalignment with former allies

  • 5 ‘The Radical Left Democrats, working with their partner, the Fake News Media, are trying to STEA (...)

12What changed dramatically at the end of 2020 is the nature of the enemies. Apart from fake news media and the radical left – that Trump indifferently brings together5 – the new culprit of Trump’s woes is the Dominion machine accused of tampering with the election. He indeed likens the company owning the faulty machine to his political enemies:

(3) People are not going to stand for having this Election stolen from them by a privately owned Radical Left company, Dominion, and many other reasons! (Nov. 14th 2020)

  • 6 This is not the first time some Fox News presenters are being criticized for their unfaithfulnes (...)

The big difference in the post-election period, highlighted by Trump himself, is what he perceives as the betrayal from a press that has usually 6shown him in the most favourable light.

(4) @FoxNews daytime ratings have completely collapsed. Weekend daytime even WORSE. Very sad to watch this happen, but they forgot what made them successful, what got them there. They forgot the Golden Goose. The biggest difference between the 2016 Election and 2020 was @Foxnews! (Nov. 12th 2020)

  • 7 French and Raven identify five bases of power among which ‘reward power’ that is predicated on a (...)

By reminding them he has ‘reward power’7 (French and Raven, 1959: 263) over them, Trump warns Fox News against pursuing coverage that acknowledges Biden’s victory. In (4), he explicitly tells the channel who made them all rich – which incidentally emphasises his collusion with the media in a most open-ended way. The most faithful channel has parted ways with the President who sees this as a realignment with the mainstream ‘fake’ media. By encouraging people to turn to other, less ‘boring’ channels, Trump sentences Fox News to death on Twitter:

(5) @FoxNews daytime is not watchable. In a class with CNN & MSDNC. Check out @OANN, @newsmax and others that are picking up the slack. Even a boring football game, kneeling and all, is better!’ (Dec. 6th 2020)

  • 8 We recognise here Trump’s tendency to nickname, putting people in their places (see Scalfani 201 (...)

13Trump establishes hierarchies in the media world in terms of ratings, entertainment and the ‘true’ rendering of events. Increasingly, he is brought to adopt the same rating scale within his very own party, sorting the wheat from the chaff to establish who should be exposed as traitors to the cause, like Fox News. Trump gives weakness and cowardice marks to the nicknamed ‘Republicans In Name Only’ (RINO8) who either fail to have power over their people or who give in too easily. As with Fox News presenters, Trump does not hesitate to implicitly threaten RINO by reminding them of what they owe him (namely their political career, past and future) as a way to scare them into allegiance:

(6) I saved at least 8 Republican Senators, including Mitch, from losing in the last Rigged (for President) Election. Now they (almost all) sit back and watch me fight against a crooked and vicious foe, the Radical Left Democrats. I will NEVER FORGET. (Dec. 24th 2020)

A poll asking GOP voters about their preferred candidate for Governor in Georgia in a potential primary shows Governor Brian Kemp whom Trump keeps accusing of weakness only coming in second. In response to this poll, Trump recalls what daring to oppose him entails for a Republican’s future career:

(7) Wow, I guess it’s not good to go against a President who everyone in Georgia knows got you into office!’ (Jan. 2nd 2021)

As a final blow, Trump goes as far as calling on his followers to vote the traitors ‘out of office’ in the next elections:

(8) They allowed states that I won easily to be stolen. Never forget, vote them out of office!’ (Dec. 12th 2020)

The following tweet is a derisive response to a fan (@PhilipRucker) retweeting the Washington Post survey stating that out of all 249 Republicans in the House and Senate ‘only 25 would acknowledge that Joe Biden is president-elect’9:

(9) wow! I am surprised there are so many. We have just begun to fight. Please send me a list of the 25 RINO. I read Fake News Washington Post as little as possible! (Dec. 5th 2020)

Asking his followers to send him a list is nothing short of a call for denunciation. Treasons today will be paid highly tomorrow: the menacing ‘I will NEVER FORGET’ in (6) serves here again to warn other Republicans against deserting the President and to keep the troops on the alert.

4. Trump’s rhetorical strategies

4.1. Personal pronouns and (in)direct address

  • 10 Fuchs indicates that Trump hardly uses the pronoun ‘us’, preferring the showcasing of himself th (...)
  • 11 Trump can’t help reminding people of his uniqueness though: ‘Another Vaccine just announced. Thi (...)

14Although Trump’s rhetorical style of self-promotion or other-denigration still obtains in the corpus, it has changed slightly in pronominal appearance. The hyperpersonalisation that many scholars have noticed – the overwhelming presence of ‘I’ (see Coe 2016, Fuchs 2018: 166, 18110) – does not seem to pervade after the election.11 An analysis of the personal pronouns in the short period under study shows the prevalence of the first-person plural pronoun compared with the first-person singular (see Figure 1).









Fig. 1. The distribution of personal pronouns in the post-election corpus

  • 12 In their comparative study of all the transcripts of presidential debates since the first offici (...)

15While the referentially non-specific third-person plural pronoun ‘they’ occurs the most frequently, as often evinced in studies of Trump’s speeches12, ‘we’ appears to be the new favourite pronoun (with ‘us’ appearing 25 times and ‘our’ 115 times). What is interesting is that Trump scrupulously avoids a personalisation of the election, turning it into a fight that ‘we’ need to take on for the whole country. The depersonalisation is a way to further implicate his supporters in the fight they have now to lead together. This decentring from the self is manifest in the blending of plural and singular entities in one of Trump’s tweets: ‘We received more votes that any sitting President in US history – and we will win’ (Nov. 11th 2020). The plural ‘we’ is a mere extension of the presidential leader. This use of ‘we’ brings into relief Trump’s use of the pronoun in other speeches, as underlined by Sidnell (2020: 68): Trump’s rhetoric bringing into existence the group ‘to whom power [is] to be returned’ is at the heart of his ‘we the people rhetoric’ in which people are brought to act as ‘his agents’ (2020: 70).

16If during the campaign the recourse to ‘you’ as a direct address to potential voters is expected, in the post-election corpus, ‘you’ tends to directly address those among the RINO who fail to do what is required to win the day. Among them is Governor Brian Kemp of Georgia (already referred to in Tweet (7)) who gets the most tweets (35 in total) for having ‘stupidly’ signed the ‘consent decree’ in his state. Trump seems to be directly addressing him, as he uses the imperative form followed by the Governor’s Twitter account: ‘Open up unconstitutional Consent Decree Now! @BrianKempGA’ (Nov. 6th), ‘Get it done! @BrianKempGA’ (Nov. 19th), ‘Do something @BrianKempGA’ (Dec. 1st), ‘Move fast @BrianKempGA’ (Dec. 21st). But interestingly, the direct address gives way to forms of indirect address to the same protagonist. He is still very much the target of the address but not strictly speaking the addressee. Through the use of questions (see also Section 4.3 below), he targets the ‘hapless’ governor but includes his audience in the questioning, which is a strategic way to kindle his followers’ frustration:

(10) Why won’t Governor @BrianKempGA, the hapless Governor of Georgia, use his emergency powers, which can be easily done, to overrule his obstinate Secretary of State, and do a match of signatures on envelopes. (Nov. 30th 2020)

(11) What is Secretary of State and @BrianKempGA afraid of. They know what we’ll find!!! (Dec. 1st 2020)

The indirectness of the question is at its acme in the following tweet addressing an indefinite ‘anyone’. It has the effect of accentuating the target effect, rallying people on Trump’s side against the Governor:

(12) Has anyone informed the so-called (says he has no power to do anything!) Governor @BrianKempGA & his puppet Lt. Governor @GoffuncanGA, that they could easily solve this mess, & WIN. Signature verification & call a Special Session. So easy! (Dec. 6th 2020)

Trump is addressing Kemp through indirect means – the real addressee is ‘anyone’, but Kemp occupies the speech role of the target (see Truan, 2021 on the pragmatic continuity between second and third-person address). Through the third-person question targeting him, he further pits people against Kemp, thereby intensifying intimidation.

17By contrasting the governor’s inefficacy and ‘the great people of Georgia’, Trump turns the governor into an easy scapegoat, crystalizing the anger of the people:

(13) Republicans in Georgia must be careful of the political corruption in Fulton County, which is rampant. The Governor, @BrianKempGA, and his puppet Lt. Governor, @GeoffDuncanGA, have done less than nothing. They are a disgrace to the great people of Georgia! (Jan. 3rd 2021)

Earlier on in December, Trump had singled out the smartness of the people of Georgia the better to contrast it with the dumbness of their governors through a damning direct address:

(14) @brianKempGA and puppet @GeoffDuncan GA. Your poll numbers are tanking like seldom before. Always said Georgians are smart!’ (Dec. 8th 2020)

  • 13 I agree with Schneider and Eitelmann (2020: 249) that Trump is ‘best described as rhetorically p (...)

18 The reference to his followers in the period under study goes from mainly indirect to more intensely direct address towards the end of the year. Sharpening his populist13 rhetoric, he refers to them as ‘the people’ brandishing their anger with the primary aim of scaring the establishment. The third-person reference to his ‘massive’ angry followers is a way to put pressure on all parties concerned, including those in his own Party:

(15) People are not going to stand for having this Election stolen from them by a privately owned Radical Left Company, Dominion, and many other reasons! (Nov. 14th 2020)

(16) People are coming forward like never before. (Dec. 1st)

(17) The people of the United States were cheated, and our Country disgraced. Never even given our day in Court! (Dec. 12th)

(18) People are angry! (Dec. 16th)

(19) I hope the Democrats, and even more importantly, the weak and ineffective RINO section of the Republican Party, are looking at the thousands of people pouring into DC. They won’t stand for a landslide election victory to be stolen. (Jan. 5th 2021)

He also accuses Congress of depriving the American people of care in these pandemic times by refusing to grant a generous stimulus package: ‘Why isn’t Congress giving our people a Stimulus Bill. It wasn’t their fault, it was the fault of China’ (Dec. 20), ‘Give our people the money’ (Dec. 25th), ‘I simply want to get our great people $2000, rather than the measly $600’ (Dec. 26th). His third-person references to ‘our people’ further consecrate the dichotomy between people versus the establishment, indirectly kindling people’s outrage against the ones in power. Such mentions also serve to remind Republicans that he has a huge base behind him. This is the forceful effect of the indirect address. Trump’s use of the ‘people’ brand becomes more and more frequent as his support in court and among his own Party dwindles. Asking Republicans to get ‘tough’ (used 6 times, for instance ‘GET TOUGH REPUBLICANS!’ Dec. 4) is also indirectly asking his people to get tough. But Trump’s goading remains forcefully indirect, as in this instance where he imagines an alternative storyworld:

(20) If a Democrat Presidential Candidate had an Election Rigged & Stolen, with proof of such acts at a level never seen before, the Democrat Senators would consider it an act of war, and fight to the death. Mitch and the Rep do NOTHING, just want to let it pass. NO FIGHT! (Dec. 26th)

Hypothesizing that Democrats faced with the same situation would equate the stolen victory to an ‘act of war’ entailing a metaphorical ‘fight to the death’ is a means for Trump to kindle his followers’ reaction and bolster their right to take strong action.

19When all the rhetorical tools of intimidation become blunted, interestingly, direct address comes back in force in December. Trump’s ‘people’ are not referred in the third person any more but directly addressed:

(21) Big protest in D.C. on January 6th. Be there, will be wild! (Dec. 19th 2020)

(22) See everyone in D.C. on January 6th (Dec. 26th)

(23) See you in Washington, DC, on January 6th. Don’t miss it. Information to follow! (Dec. 27th)

(24) January 6th. See you in D.C. (Jan. 1st 2021)

(25) You will see the real numbers tonight during my speech, but especially on JANUARY 6th (Jan. 4th 2021)

The more and more isolated President may have seen in his followers his ultimate solution to carry on the fight.

4.2. Negating negation

20Never once in the corpus does Trump acknowledge that he lost the election. Only in the alternative world created by fake media can Biden be designated as president:

(26) He only won in the eyes of FAKE NEWS MEDIA. I concede NOTHING. We have a long way to go. (Nov. 15th 2020)

(27) Will never concede to fake ballots & ‘Dominion’. (Nov. 23rd)

  • 14 The third and only other instance is put under the form of a poll question: ‘Should President Tr (...)

(26) and (27) are the two instances where the long-awaited ‘concede’14 word appears only to be negated via absolute negators (‘concede nothing’, ‘never concede’). The absolute negators are telling of an attitude that consists in negating a whole reality outright. The negative adverb ‘never’ (occurring 38 times) or the pronoun ‘nothing’ (19 occurrences) express his all or nothing stance allowing for no doubt, nuance or middle ground. His use of ‘no way’ partakes of the same attempt at refusing to accept defeat by making it ‘unthinkable’, condemning all discussions that would entail explanatory arguments and supportive evidence. Tangible proof is replaced by a firm belief strongly asserted. The synthetic negation ‘no’ combines the function of marking negation and determining the noun ‘way’ in a forceful manner that sounds without appeal in Trump’s world: ‘no way’ (Nov. 16th), ‘no way the Dems could have won’ (Nov. 19th), ‘Just saw the vote tabulations. There is NO WAY Biden got 80,000,000 votes! (Nov. 25th), ‘NO WAY WE LOST THE ELECTION’ (Nov. 29th, Dec. 6th).

  • 15 Metalinguistic negation is defined as ‘a means for objecting to a previous utterance’ (Horn, 198 (...)

21To the absolute negators must be added the use of affixal negators that carry the same function. The victory of Joe Biden is an ‘impossibility’, not even conceivable, and the proof ‘pouring in’ is ‘undeniable’ (Nov. 21st 2020). Saying that something is ‘impossible’ with the affixal negator (im-) is not the same as negating the adjective ‘possible’: ‘Joe Biden’s victory is not possible’ is not rigorously equivalent to ‘Biden’s victory is impossible’. As Huddleston & Pullum (2002: 821) explain, using the examples ‘not common’ and ‘uncommon’, the first clausal negative is contradictory of the positive, the affixal negative is not. To claim a victory to be impossible is not to negate its possibility but to reject it altogether, as hammered home by Trump: ‘Impossible for Biden to have overcome these, and even greater, odds’ (Nov. 26th), ‘Impossible result!’ (Dec. 1st), ‘data taken after the vote says that it was impossible for me to lose, unless FIXED!’ (Dec. 9th), ‘Statistically impossible to have lost the 2020 Election’ (Dec. 19th). When Trump uses verbal negation, it is metalinguistic negation15 through which he shows he does not accept the formulation used in the media:

(28) Biden did not win, he lost by a lot! (Nov. 12th)

(29) He didn’t win the Election. He lost all 6 Swing States, by a lot. (Dec. 19th)

Trump is not disputing the statement ‘Biden didn’t win’ – nothing to be asserted or contested here – he is objecting to the size of the defeat. He did not just lose, he lost ‘by a lot’, which becomes one of Trump’s set-phrases in his end-of-year tweets (13 occurrences).

  • 16 An exception is when Trump speaks of the fake pollsters who ‘said I would lose’ (Dec. 30th).
  • 17 66 days until Trump’s Twitter suspension on January 8th 2021.

22 In the two months after the election until the 6th of January (and ever since), Trump has never admitted defeat. This denial is derived from a refusal to concede to what was never even envisaged. This de-negation tallies with the attitude of the competitor which Shields and Bredemeier (2011: 33-5) call ‘decompetition’ – a negative conceptualization of competitors knowing from the start they will win and therefore allowing themselves to bend all the rules that can get them there. As losing is only the prerogative of the weak (see Koth 2020), we understand why the verb ‘concede’ cannot enter Trump’s vocabulary. For that matter the verb ‘lose’ never appears in connection with the first-person pronoun16. Trump never says ‘I didn’t lose’ – a negative assertion which would consider the possibility of a loss as reported in the mainstream media. What he does, instead, is negating this negation by asserting that ‘he won’. The word ‘win’ appears 66 times in 66 days17 either as a noun or a verb, and the preterit version (‘we won’) considering victory as an acknowledged fact occurs 55 times. Trump’s discourse of denial goes against the traditional working of negation consisting in asserting the negation of a proposal. As Cambier (2019: 144) shows, denying a proposal is predicated on the concept of ‘disbelief’, which boils down to discrediting proven facts, refuting their obviousness and undermining any opposite objective belief.

  • 18 The word ‘truth’ appears 10 times and ‘lie’ only 5 as compared to the adjective ‘fake’ which rec (...)

23 In building an alternative world in which he has always won, Trump’s relationship to ‘negation’ is illustrative of the post-truth age which displays a growing indifference between the truth and lies18. What only matters is the annihilation of all the arguments that could contradict his newly-created world. In his defeat-denying, Trump eclipses an annoying reality and transforms facts as announced on mainstream media into subjective beliefs. Through his winning stance, conveyed via the modal ‘will’ (‘we will win’ appears 20 times), Trump imposes his will through the strength of his assertions, thereby bending reality to his own wishes. By using the modal ‘will’ rather than ‘be going to’ for instance, Trump makes the predicted victory depend on his capacity and desire or willingness to make it happen whilst ‘we are going to win’ would emphasize predictability based on the circumstances and the progression of an engaged process. Moreover, to give his people the constant impression of (imminent) victory, he makes sure that the word ‘win’ appears repeatedly, using it in connection to other current victories (‘a vaccine WIN’ for instance) which, through association, place him on the side of the winners. By repeatedly seeing this very word in Trump’s tweets and being carried away by the force of the denial through his performative language, his followers are incited to start believing that victory is indeed inevitable.

24 But how could Trump, in spite of the fact that he lost the 63 lawsuits contesting election processes he filed, manage to sustain faith in victory? It must be noted that, as part of the denial strategy, these failures get absolutely no mention in his tweets engaged in sustaining the belief in victory. As Kalpokas (2019: 12) argues, ‘political actors and their truth-claims become true through affective investment’, emotional opinions taking precedence over any documented truth. Trump definitely appeals to feelings over facts, drowning any rigorous demonstration of evidence in an ocean of wishful-thinking assertions. In his post-truth world, wilful assertions turn wishes into truths. In the following tweet, the juxtaposition of the words ‘feel’ and ‘know’ equates instinct with knowledge in a revealing way:

(30) 78% of the people feel (know!) the Election was RIGGED’ (Dec. 10th)

The transformation of gut beliefs into facts finds an echo in Trump’s question to the crowd at the January 6th rally: ‘And by the way does anybody believe that Joe had 80 million votes, does anybody believe that?’.

25Twitter has served its role of tightening the link between Trump and his supporters. As McIntyre (2018: 102) puts it, ‘if we are already motivated to want to believe certain things, it doesn’t take much to tip us over to believing them, especially if others we care about already do so’. Giving stars or massively retweeting favourable news items and videos has been a way for Trump to seek even more sympathisers. The numerous retweets – more than half of the tweets of the period as mentioned in the corpus and methodology section – gives the impression of constant overbid in the production of evidence while a thorough study shows that Trump keeps repeating the same things, betting on the repetition-induced truth effect whose influential efficiency has been noted by many a socio-psychologist (see for instance Unkelbach & Koch, 2019, Greifeneder et al, 2021). But he also sustains interest through suspenseful storytelling, using the attention-grabbing imperative form ‘stay tuned’ for instance, advertising what will come next as if Trump was himself promoting his own road to success. ING forms also contribute to an impression of action taking place on all fronts: ‘Blockbuster testimony taking place right now in Georgia’ (Dec. 3rd) ‘Big news coming out of Pennsylvania’ (Dec. 21st). As represented in Figure 2 extracted from AntConc, the impression of ongoingness is reinforced by the use of imperfective aspectuality, conveying the illusion of continuous (victorious) fighting.

Fig. 2 Lexical verb ‘fight’ with imperfective aspectuality (from AntConc)

Fig. 2 Lexical verb ‘fight’ with imperfective aspectuality (from AntConc)

26Trump keeps his followers on the alert, directing their attention to evidence that confirm what they already believe in. In this echo chamber, the same version of the story infinitely reverberates, sustaining what Gerbaudo (2018: 7) calls ‘online mass co-operation’ turning followers into a ‘militant support base’. Trump fully exploits this capacity, advising his fans on what they should watch in a most explicit imperative manner: ‘A MUST WATCH!’ (Dec. 6th), ‘Hope everybody is watching @OANN right now. Other media afraid to show’ (Dec. 1st), ‘Watch @SeanHannty tonight at 9:00 PM Eastern – Story on Voter fraud!’ (Dec. 4th). Should they make the trip to Washington on January 6th, they will be rewarded with new evidence: ‘Massive amounts of evidence will be presented on the 6th. We won, BIG!’ (Jan. 1st 2021).

4.3. Iceberg rhetoric

27In this January rally, Trump announces the revelation of ‘some of the evidence’, which implies a selection among much more that he will not bore his audience with. This technique of pretending there is more by saying so, will be referred to as the iceberg rhetoric. The iceberg image means that he is only showing a little evidence (because that is all he needs) but this little evidence conceals a much greater quantity of evidence (which he does not ‘even’ need). The adverb ‘even’ plays a primordial role here along with others such as ‘on that alone’, ‘only’, or ‘enough’ as well as other comparative phrases contributing to this iceberg effect:

(31) They wouldn’t let our Poll Watchers and Observers into the Counting Rooms. We win on that alone! (Nov. 14th 2020)

(32) Before even discussing the massive corruption which took place in the 2020 Election, which gives us far more votes than is necessary to win all of the Swing Sates (only need three), it must be noted… (Jan. 1st 2021)

(33) (…) will insure a Georgia Presidential win (very few votes are needed, many will be found) (Dec. 3rd 2020)

(34) These States ‘election laws’ were made up by local judges & politicians, not by their Legislatures, & are therefore, before even getting to irregularities & Fraud, UNCONSITUTIONAL! (Jan. 3rd 2021)

(35) The numbers are far greater than what is necessary to win the individual swing states, and cannot even be contested. (Dec. 26th 2020)

(36) We now have far more votes than needed to flip Georgia in the Presidential race (Dec 30th 2020)

(37) my investigators have found hundreds of thousands of fraudulent votes, enough to ‘flip’ at least four States, which in turn is more than enough to win the Election. (Nov 21st 2020)

This rhetoric of the iceberg is his way of overcoming the paradox of having to reconcile a proclaimed victory on the one hand and the need to find a few more votes on the other.

28Trump indeed keeps repeating that victory is just around the corner and few votes are needed to win the day. He puts his legal failures (due to lack of evidence confirmed by the state courts and the Supreme Court) on some Republicans’ unwillingness to do what he presents as ‘easy’. As intimated above (Tweets 10 to 13), he blames RINOs for not doing what is within their easy reach. Through ‘easy’ steps (such as signature verification on ballots and envelopes), victory will ‘easily’ ensue: ‘so simple, and so easy to do’ (Nov. 30th), ‘Signature verification & call a Special Session. So easy!’ (Dec. 6th), ‘so easy to do’ (Dec. 18th). By reiterating the easiness of what should be done, Trump exacerbates his fans’ frustration, turning them against those who incomprehensibly refuse to do the ‘easy’ thing. The adverb ‘easily’, repeated 17 times, epitomises the contrast between his legal difficulties in the real world and the easy route to victory he proclaims in his own.

29 The rhetoric of the iceberg also allows Trump to feed into people’s belief in a conspiracy against them, the smell metaphor conveying suspicion: ‘something smells fishy’ (Dec. 3rd). This plot against him and his supporters is led by an obscure league denying them access to victory. Instead of telling why courts rejected his team’s suits, Trump shouts that people are being unfairly treated and prevented access to evidence by the establishment. He construes their fight as a heroic journey against impediments to be overcome on their way to victory, stirring affects along the way: ‘What are they trying to hide. They know, and so does everyone else’ (Nov. 14th), ‘They know what they are trying so hard to hide. Terrible people!’ (Dec. 23rd). Trump adopts conspiracy theory rhetoric that consists in ‘just asking questions’ (Byford, 2011: 88-93), fighting against his opponents rather than bringing strong and rational evidence for his own case (Wood & Douglas, 2013, see also Oswald, 2016). Indeed, his grammar of conspiracy takes the form of constant (rhetorical) questions, which produces the forceful effect of letting his fans provide the answer. Asking the questions is also a way to confer the burden of proof to the other side. Trump’s favourite questions raising doubts and creating suspicion start with ‘how’ (16 occurrences) and ‘why’ (23). Here is a selection of both:

(38) How can you have a presidency when a vast majority think the election was RIGGED? (Dec. 9th)

(39) Why is Joe Biden so quickly forming a Cabinet […]? (Nov. 21st)

He maintains doubts in his people’s minds by asking questions that do not call for legitimate answers but serve to enhance their suspicion and fuel their rage against guilty Democrats dubiously hurrying to take office. Poking his fans’ fear of being cheated by the establishment, Trump’s rhetoric may have goaded his supporters into feeling it was their constitutional right to invade Congress on January 6th.


30After repeated attempts at intimidation towards ‘weak’ Republicans, Trump could only count on fans bred through self-reinforcing tweets and retweets to take the call to ‘war’ and the ‘fight to the death’ mentioned in tweet (20) literally. His rhetoric of ‘decompetition’ pushed them to go ‘wild’ (tweet (21)), which conveniently enabled him to ‘never concede’ (tweet (27)).

31In this Orwellian dystopia where numbers can be disputed, the aim is to deny a world that does not suit one’s wishes and contest it all the more strongly as it is showing stronger signs of resistance every day. The more it resists, the stronger the optimistic rhetoric of the winner must be proclaimed loud and clear. In this attempt at sustaining belief in victory against obvious proof against it, Trump’s stubborn rhetoric of optimism evokes Hitler’s rhetoric of certitude as recounted in Klemperer’s notebooks, published as The Language of the Third Reich: LTI, Lingua Tertii Imperii. The philologist details how the dictator’s discourse managed to minimize the German defeats on the Eastern borders through a language giving the illusion of dynamism over inaction and stalemate. As Klemperer (1996: 294) indicates, the words ‘defeat’, ‘retreat’ or ‘escape’ were never part of Hitler’s language of defeat denial. The enemies never made breakthrough (Durchbrüche) but only irruptions (Einbrüche) on the ‘elastic’ German front, for instance.

32This comparison adds grist to the mill of those who saw in Trump’s use of language and the media the premises of authoritarian politics. In his persuasive rhetoric, Trump did manage to turn obvious facts into mere opinions, to transform defeat into imminent victory, to have his fans perceive Democrats’ taking office as a suspicious rush, and to arouse their fear about a reinforced alliance between the Radical Left and the fake news media. Quoting Timothy Snyder’s book On Tyranny (2017) in which the author perceives in fake news and alternative facts the seeds of fascism, McIntyre (2018: 179) raises the question: ‘This may seem a heavy conclusion to draw from something as facile as fake news. But with today’s social media to facilitate the spread of misinformation faster than a propagandist’s dream, shouldn’t we at least be awake to this possibility?’. The assault on the Capitol on 6 January 2021 that seems to have taken everyone by surprise might be one answer to the question.

Top of page


Altenberg, Bengt. “Amplifier Collocations in Spoken English.” In Stig Johansson and Anna-Brita Stenström (eds.), English Computer Corpora: Selected Papers and Research Guide, 127-47, Berlin: Mouton de Gruyter, 1991.

Biały, Beata. “Social Media – from Social Exchange to Battlefield.” The Cyber Defense Review 2.2 (2017): 69-90.

Byford, Jovan. Conspiracy Theories. London: Palgrave Macmillan, 2011.

Cambier, Alain. Philosophie de la post-vérité. Paris: Hermann Editeurs, 2019.

Coe, Chelsea. “WIRED Opinion: The Way Trump Talks in Debates is Contagious. Wired October 3 2016. Available online: https//

Egbert, Jesse and Biber, Douglas. “It’s just words, Folks. It’s Just Words: Trump’s Distinctive Linguistic Style.” In: Schneider Ulrike and Eitelmann Matthias (eds) Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp. 17-40.

French, John R.P and Raven, Bertram. “The Bases of Social Power.” In: Cartwright Dorwin (ed.) Studies in Social Power. Ann Arbor: University of Michigan, 1959, pp. 150-167.

Fuchs, Christian. Digital Demagogue. Authoritarian Capitalism in the Age of Trump and Twitter. London: Pluto Press, 2018.

Gerbaudo, Paolo. “Social media and populism: an effective affinity?” Media, Culture & Society(2018)

Greifeneder, Rainer, Jaffe, Mariela, Newman, Eryn and Schwarz, Norbert. The Psychology of Fake News. Accepting, Sharing, and Correcting Misinformation. London & New York: Routledge, 2021.

Hawkings, Kirk. “Is Chávez populist? Measuring Populist Discourse in Comparative Perspective.” Comparative Political Studies 42.8 (2009): 1040-67.

Hodges, Adam. When Words Trump Politics: Resisting a Hostile Regime of Language. Stanford: Stanford UP, 2019.

Horn, Laurence R. “Metalinguistic Negation and Pragmatic Ambiguity.” Language 61.1 (1985): 121-174.

Huddleston, Rodney and Pullum, Geoffrey K. The Cambridge Grammar of the English Language. Cambridge: CUP, 2002.

Kalpokas, Ignas. A Political Theory of Post-Truth. Basingstoke: Palgrave Macmillan, 2019.

Klemperer, Victor. LTI, La Langue du IIIe Reich. Carnet d’un philologue. Guillot E Trans. Paris: Albin Michel, 1996.

Koth, Anthony. “Either We WIN this Election, or We are Going to LOSE this Country! Trump’s WARLIKE COMPETITION Metaphor.” In: Schneider Ulrike and Eitelmann Matthias (eds) Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp. 155-172.

McIntosh, Janet. “Introduction: The Trump Era as a Linguistic Emergency.” In: McIntosh Janet and Mendoza-Denton Norma (eds) Language in the Trump Era. Scandals and Emergencies. Cambridge: CUP, 2020, pp. 1-43.

McIntosh, Janet and Mendoza-Denton, Norma (eds) Language in the Trump Era. Scandal and Emergencies. Cambridge: CUP, 2020.

Moffitt, Benjamin. The Global Rise of Populism: Performance, Political Style, and Representation. Stanford: Stanford UP, 2016.

Oswald, Steve. “Conspiracy and Bias: Argumentative Features and Persuasiveness of Conspiracy Theories.” In: Bondy Patrick and Benacquista Laura (eds.) Argumentation, Objectivity, and Bias: Proceedings of the 11th International Conference of the Ontario Society for the Study of Argumentation (OSSA), 18-21 May 2016. Windsor, ON: OSSA, pp. 1-16.

Ouyang, Yu and Waterman, Richard W. Trump, Twitter, and the American Democracy. Political Communication in the Digital Age. Basingstoke: Palgrave Macmillan, 2020.

Pichard, Alexis. Trump et les médias, l’illusion d’une guerre ? VA Editions, 2020.

Scalfani, Jennifer. Talking Donald Trump: A Sociolinguistic Study of Style, Metadiscourse, and Political Identity. London & New York: Routledge, 2018.

Schneider, Ulrike and Eitelmann Matthias (eds). Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020.

Schneider, Ulrike and Eitelmann, Matthias. Great movement versus crooked opponents: Is Donald Trump’s language populist? In: Schneider Ulrike and Eitelmann Matthias (eds) Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp. 235-250.

Schubert, Christoph. “Donald Trump’s ‘Fake News Agenda. A pragmatic account of rhetorical delegitmation.” In: Schneider Ulrike and Eitelmann Matthias (eds), Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp. 194-214.

Schwarz, Norbert and Jalbert, Madeline. “When (Fake) News Feels True. Intuitions of Truth and the Acceptance and Correction of Information.” In: Greifeneder Reiner, Jaffé, Mariela, Newman, Eryn and Schwarz, Norbert, The Psychology of Fake News. Accepting, Sharing, and Correcting Misinformation. London & New York: Routledge, 2021, pp. 73-89.

Sidnell, Jack. “‘Get ‘Em Out!’: The Meaning of Ejecting Protesters.” In: McIntosh Janet and Mendoza-Denton Norma (eds) Language in the Trump Era. Scandals and Emergencies. Cambridge: CUP, 2020, pp. 63-73.

Shields, David and Bredemeier, Brenda. “Contest, Competition, and Metaphor.” Journal of the Philosophy of Sport 38.1 (2011): 27-38.

Skinnell, Ryan (ed). Faking the News: What Rhetoric Can Teach Us about Donald J. Trump. Imprint Academic, 2018.

Snyder, Timothy D. On Tyranny: Twenty Lessons from the 20th Century. New York: Tim Duggan Books, 2017.

Stange, Ulrike. “Very Emotional, Totally Conservative, and Somewhat all over the Place.” In: Schneider Ulrike and Eitelmann Matthias (eds), Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp. 87-108.

Stolee, Galen and Caton, Steve. “Twitter, Trump, and the Base: A Shift to a New Form of Presidential Talk?” Signs and Society 6.1 (2018): 147-65.

Truan, Naomi. The Politics of Person Reference: Third-person Forms in English, German, and French. Amsterdam: John Benjamins Publishing, 2021.

Tyrkkö, Jukka and Frisk, Irina. “Crooked Hillary, Lyin’ Ted, and Failing New York Times. Nicknames in Donald Trump’s Tweets.” In: Schneider Ulrike and Eitelmann Matthias (eds) Linguistic Inquiries into Donald Trump’s Language. From ‘Fake News’ to ‘Tremendous Success’. London: Bloomsbury, 2020, pp.109- 139.

Unkelbach, Christian and Koch, Alex. “Gullible but Functional? Information Repetition and the Formation of Beliefs.” In: Forgas Joseph and Baumeister Roy F (eds), The Social Psychology of Gullibility. Fake News, Conspiracy Theories and Irrational Beliefs. New York and London: Routledge, 2019, pp. 42-60.

Wood, Michael and Douglas, Karen. “What about Building 7? A Social Psychological Study of Online Discussion of 9/11 Conspiracy Theories.” Frontiers in Psychology 4 (2013).

Top of page


1 All the emphasis in bold in the following tweets is mine.

2 Taking up Altenberg’s classification, Stange focuses on ‘amplifiers’ as a subcategory of intensifiers. They are opposed to ‘downtoners’ which have a ‘lowering effect’ (Stange 88). The maximizer category is subdivided into the boosters (that is adverbs that indicate a ‘high point on a scale’ such as greatly, highly, so) and the maximizers that point to the ‘upper extreme of the scale’ with adverbs such as entirely, perfectly, extremely (89).

3Wow. This is exactly what happened to us. Great courage by judge!’ (Nov. 13th 2020); ‘Now that the Biden Administration will be a scandal plagued mess for years to come, it is much easier for the Supreme Court of the United States to follow the Constitution and do what everybody knows has to be done. They must show great Courage & Wisdom. Save the USA!!!’ (Dec. 11th 2020).

4Michigan AG Dana Nessel Pursues Sanctions Against Lawyers Questioning Election via @BreitbartNews These lawyers are true patriots who are fighting for the truth and, obviously, getting very close. AG should be sanctioned. Fight on!’ (Dec. 27th 2020)

5 ‘The Radical Left Democrats, working with their partner, the Fake News Media, are trying to STEAL this election. We won’t let them’ (Nov. 16), ‘The media is just as corrupt as the Election itself!’ (Nov. 21).’

6 This is not the first time some Fox News presenters are being criticized for their unfaithfulness by Trump (see Pichard 2020) but criticism and intimidation become more vicious at the end of 2020.

7 French and Raven identify five bases of power among which ‘reward power’ that is predicated on a person’s perception of another person’s ability ‘to mediate rewards for him’ (French & Raven, 1959: 263).

8 We recognise here Trump’s tendency to nickname, putting people in their places (see Scalfani 2018, Tyrkkö & Frisk 2020).

9 @Philip Rucker, Dec. 5th 2020:

10 Fuchs indicates that Trump hardly uses the pronoun ‘us’, preferring the showcasing of himself through the first-person pronoun: ‘Trump is a brand. Trump is a strategy. Trump is entertainment. Trump is a spectacle. Trump is politics. Trump is the instrumentalization of everything surrounding him. Trump is the absolute commodification of the self’ (Fuchs, 2018: 166).

11 Trump can’t help reminding people of his uniqueness though: ‘Another Vaccine just announced. This time by Moderna, 95% effective. For those great “historians”, please remember that these great discoveries, which will end the China Plague, all took place on my watch!’ (Nov. 16).

12 In their comparative study of all the transcripts of presidential debates since the first official one (between Kennedy and Nixon in 1960), Egbert & Biber (2020: 22) show that ‘the pronouns that were used more by Trump than other presidential candidates include third-person pronouns (she, it, they), the second-person pronoun you, and the impersonal pronoun it’.

13 I agree with Schneider and Eitelmann (2020: 249) that Trump is ‘best described as rhetorically populist rather than politically populist’. Following Moffitt (2016) and Hawkings (2009), they show that Trump’s populism can less be defined as ideology than as a political style. For Moffitt, rhetorical populism requires a leader as performer and is predicated on the opposition between the people and the elite (whom, with Trump, is a blend of Big Media, Big Tech along with the establishment), bad manners and the tendency to sustain a sense of tension and crisis among the people.

14 The third and only other instance is put under the form of a poll question: ‘Should President Trump concede to Biden? Poll results. No: 190,593 (98.9%) Yes: 2,181 (1.1%)’ (Nov. 24).

15 Metalinguistic negation is defined as ‘a means for objecting to a previous utterance’ (Horn, 1985: 38).

16 An exception is when Trump speaks of the fake pollsters who ‘said I would lose’ (Dec. 30th).

17 66 days until Trump’s Twitter suspension on January 8th 2021.

18 The word ‘truth’ appears 10 times and ‘lie’ only 5 as compared to the adjective ‘fake’ which recurs 56 times.

Top of page

List of illustrations

Title Fig. 2 Lexical verb ‘fight’ with imperfective aspectuality (from AntConc)
File image/png, 32k
Top of page


Electronic reference

Sandrine Sorlin, ‘I will never concede’: Donald Trump’s discourse of denial on Twitter (Nov. 4th 2020 – Jan. 8th 2021)Anglophonia [Online], 31 | 2021, Online since 13 October 2021, connection on 12 August 2022. URL:; DOI:

Top of page

About the author

Sandrine Sorlin

Univ Paul Valéry Montpellier 3, EMMA, F34000, Montpellier, France

Top of page



Creative Commons - Attribution-NonCommercial-NoDerivatives 4.0 International - CC BY-NC-ND 4.0

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search