Far-right infighting in digital spaces during the Capitol Hill riot

Abstract

Reports of the Capitol Hill Riot (CHR) on 6 January 2021 emphasized how far-right activists converged from all over the United States and coalesced in violent public action. And yet, the riot itself was highly contested within online far-right communities, revealing underlying infighting. This paper draws on data from three forums and image boards, 8kun, 4chan, and TheDonald.win, to explore how online users interacted on the current political situation in the run-up to and the aftermath of the CHR. We explain far-right infighting through competing conspiratorial narratives that entail different attributions of who is deemed a traitor to the movement and then result in diverging logics for and legitimacies of political action. Drawing on qualitative content analysis of approximately 3000 posts, we demonstrate how these narratives shape internal conflicts, infighting, and proposed action. One perspective supports the riot as a necessary response to save Donald Trump after an allegedly stolen election, rendering anyone not fully in support of the CHR as hybrid; the second perspective depicts Trump as a traitor and part of more fundamental, systemic problems, therefore advocating for violence that supports a fundamental renewal of society. Our analysis thus traces the roots of political infighting amongst far-right actors to two differing conspiratorial narratives and diverging logics of political action.

Introduction

On January 6, 2021, far-right protesters stormed the Capitol Hill in Washington D.C. Images from the Capitol Hill Riot (CHR) ¹ spread across the world, shaping the post/pre-Trump era like no other event. This provoked considerable political discussion on how to evaluate the CHR and much analysis in academia has sought to understand the CHR’s genesis and aftermath (Clarke et al., 2021; Kasimov et al., 2025; Rudden, 2021; Vidino et al., 2021). Online platforms have been shown to be key to the riot, particularly far-right platforms (Molas, 2023; Munn, 2021), and analysts have frequently pointed to the CHR as a moment of unity with a seemingly cohesive coalition of different groups and unaffiliated individuals (Hughes and Miller-Idriss, 2021). However, the American far right is not, and has not been traditionally, a coherent movement; it consists of various ideological groups with internal divisions (Blazak, 2022; Lyons, 2003; Sunderland, 2023). Much like after the “Unite the Right” rally in Charlottesville in 2017, the weeks after the CHR revealed a tendency of the far right to engage in internal conflicts following pivotal events (Kunzelman and Richer, 2021; Riley, 2022).

All groups negotiate fundamental understandings of the world and current politics. However, sometimes conflicts arise with competing discourses and ideological battles, including who is seen as loyal or treacherous to the ingroup (Berger, 2021), and such fissures within a community can ultimately impede a group’s development (Buyukozturk et al., 2018). The concept of “infighting” is useful because it describes “how group cultures can diverge, despite a seeming commonality of interest” (Ghaziani and Fine, 2008: 52). A closer inspection of infighting, especially during offline trigger events, can cast light on why ideological beliefs, organizational goals, and tactical decisions diverge and what consequences this has for future political action (Ellefsen, 2018). In this context, Riley’s analysis of TheDonald.win highlights how shifting ingroup and outgroup memberships shape these dynamics, noting that “the collapse of the In-Group and the transfer of members and components of the In-Group to the Out-Group appears to be a trigger point for increased violence and rhetoric” (2022: 11).

Social media and forums are particularly relevant to studying infighting, as they are discursive spaces for organizing action, building networks, and advancing personal claims (Bennett and Segerberg, 2012; Theocharis et al., 2015). Previous research, for example, has explored the spread of misinformation and conspiracy theories (Allcott et al., 2019; Enders et al., 2023; Vanderwee and Droogan, 2025), the mechanisms and effects of polarization (Dunaway, 2021; Gupta et al., 2023), and how digital environments shape the boundaries of collective identity (Gaudette et al., 2021; Priante et al., 2018). Conflict dynamics in digital spaces are thus highly relevant for understanding discourses of the far right.

Thus far, research on the far right has mainly focused on ideologies on online platforms (Holt et al., 2022; Hutchinson et al., 2023; Jasser et al., 2023; Scrivens et al., 2023), their conflicts with opposing groups (Gallacher et al., 2021; Klein, 2019; Lee and Knott, 2020), their transnational networks (Froio and Ganesh, 2019), or mechanisms of online radicalization and extremism (Mølmen and Ravndal, 2023; Youngblood, 2020). However, less attention has been paid to ideological conflict within far-right communities. And yet this is important, as Pond and Lewis point out, because “networks assemble and mobilize through the activation of discourse within a wider media sphere of competing discourses” (2019: 214). These insights go beyond the case of the far right and can be seen in splits within movements ranging from the split of IS from Al-Qaeda to divisions within the American Civil Rights Movement or the Apartheid movement, divisions between different factions in Nazi Germany, or infighting and purges within the Khmer Rouge. Each of these cases shows how ideological infighting connects existing narratives and tropes of archetypal actors assigned to the ingroup, outgroup, or hybrid group, with present events to unleash divisions within a movement. In this way, present events are tied to existing tropes that in turn rely on ideological interpretations of historical events. The focus of our research is on a recurring motif that has as of yet remained understudied, namely, to identify specifically how the identification of hybrid actors plays a crucial role in the infighting and how it is used to draw the ideological line between the sides engaged in infighting.

To better understand the mechanisms of infighting in far-right online communities, this paper addresses the following research question: Why does infighting emerge within far-right online communities following trigger events such as the CHR? To answer this question, we conduct a qualitative analysis of approximately 3000 posts from 8kun, 4chan, and TheDonald.win, three platforms commonly linked to far-right activity and extremism, focusing on the days before, during, and after the CHR. We conceptually draw on Baele’s work on violent language and conspiratorial narratives (2019). We find that rather than one unifying conspiratorial narrative, two competing conspiratorial narratives dominate the discourses and result in competing conceptions of who is seen to have betrayed the ingroup. These narratives stipulate different logics of political action and legitimations for violence that result in different interpretations of the riot. The concept of conspiratorial narratives is useful because it emphasizes the role of discourse in the construction of collective identities that can then rupture consensuses otherwise held within the far-right community.

We begin with a contextualization of the CHR as a trigger event and the use of digital spaces by the far right. Next, the concept of conspiratorial narratives is explored, which extends beyond rigid ingroup-outgroup distinctions to include hybrid ² actors, that is, groups or individuals perceived as having betrayed the ingroup (Baele, 2019; Berger, 2021). We explain how it is possible for more than one conspiratorial narrative to be relevant within an online community and theorize this to be relevant for explaining infighting. After outlining our methodological approach and its limitations, we discuss points of consensus within the far-right content but focus on divergent perspectives within the digital communities under examination. We find considerable evidence for the role of hybrid actors in understanding infighting and demonstrate how competing conspiratorial narratives manifest in competing logics of political action surrounding the CHR.

The Capitol Hill riot as a trigger event and the far right online

On 6 January 2021, thousands of far-right demonstrators stormed the U.S. Congress on Capitol Hill to overturn the Electoral College’s certification process that would confirm Joe Biden as the winner of the 2020 presidential election. The CHR was preceded by a large rally called for by former President Donald Trump and can be viewed as a pinnacle of a series of contentious events beginning after the election in November 2020. Trigger events such as this help capture groups’ attention, highlighting the interdependence between offline and online spaces. Previous research has demonstrated that offline events can significantly influence online interactions; for example, Bliuc et al. (2019) examine the impact of offline intergroup conflicts on the collective beliefs of Stormfront Downunder members, concluding that such conflicts fostered unity and a sense of purpose within the far-right community. The far right includes political actors who view the world through an authoritarian and nativist lens, marked by xenophobic and exclusionary nationalism (Mudde, 2007). While distinctions exist between the radical right (illiberal but democratic) and the extreme right (anti-democratic) (Bjørgo and Ravndal, 2019), these categories often blur, reflecting the “increasingly porous borders and growing links” that exist within the far right (Pirro, 2023: 107). We thus draw on a broad understanding of the term to include “those who fall in between the far right and the mainstream via casual engagement with extremist content online but may nonetheless be radicalized in the long run” (Zhang and Davis, 2024: 2957).

Far-right groups have taken advantage of the Internet since before the advent of social media (Ganesh, 2020). Internet services offer far-right sympathizers global reach in spreading myths, connectedness with like-minded people and expansion of their networks (Kasimov et al., 2025; Simi and Futrell, 2006). The Internet fosters political socialization (Caiani, 2019) and radicalization toward political violence (Conway, 2017). By maintaining a presence across various online platforms, the far right uses the discursive opportunities offered to its followers, who engage in meaningful conversations regarding real-world events and personal experiences, allowing activists to “negotiate the use of violence, what national identities are valued, and what racial identities are targeted” (Hutchinson et al., 2023: 216; Törnberg and Törnberg, 2025; Wahlström and Törnberg, 2021). However, online spaces with changing contributors are prone to internal conflicts and infighting, causing disputes over the group’s mission and political targets (Sunderland, 2023).

Infighting and competing conspiratorial narratives

Internal conflicts are frequent within the far right (Hughes and Miller-Idriss, 2021). Whilst offline behavior might have demonstrated unity, online behavior in the context of the CHR exhibited considerable infighting (Kunzelman and Richer, 2021). Scholars often emphasize that the beliefs and identity of the far right revolve around a dichotomous “us versus them” understanding of the world (Van der Vegt et al., 2021: 13). Comparatively little research examines conflicts within these groups, although previous research has shown that the online far-right ecosystem is decentralized and consists of multiple groups (O’Callaghan et al., 2013). By using the term ecosystem, Baele et al. (2023: 1600) push for a “new understanding of the far-right internet as a dynamic and multidimensional entity; that is, an entity made of an ever-changing number of different components whose natures and interconnections are in constant evolution.” Berger’s (2021) research on texts written by white nationalist James Mason shows, for example, that extremists can turn on their fellow members and criticize their actions and political stance. Riley (2022) also shows that certain offline events, such as speeches or protests, can change how ingroup members are perceived by their peers. The author notes that some individuals may be pushed into the outgroup when their statements or actions fail to meet the ingroup’s expectations. In extreme forums, such as Iron March and Fascist Forge, research has found that some posts may target members of the far right by users trying to distance themselves from other nationalists “seen as the mainstream and nonviolent White supremacist movement” (Scrivens et al., 2023: 2162). This is evident in how U.S. white nationalists have tried to rebrand their movement by aligning with traditional conservatism, leading to internal conflicts (Sunderland, 2023).

Infighting can be rooted in competition for leadership roles or disagreements about identity, morality, goals, and tactical decisions (Ghaziani and Kretschmer, 2018; Hurwitz and Crossley, 2018; Suttmoeller et al., 2015), ranging from simple ideological differences to moderate or severe polarization (Orian Harel et al., 2020). Relational aspects of a movement’s development are key to understanding radicalization as not only interactions with opponents but also competitive dynamics within the group can determine political violence (Della Porta, 2013). Further, infighting can arise when members of a group “criticize or even attack” others for being “misinformed, complacent, or have become corrupted by contact with an out-group” (Berger, 2021: 23). Thus, infighting constitutes a crucial aspect of group interactions and brings to the fore ideological divergences, “while simultaneously debating the identity contours of the group” (Ghaziani and Fine, 2008: 53; Tan and Snow, 2015).

Narratives are key to understanding infighting. Narratives are not simply stories but rather cultural products of constant social interactions that lead to re-evaluations of how the world is perceived (Archetti, 2017). Narratives connect people through shared symbols and stories that impact collective identities and act as interpretive mechanisms for people to understand socio-political phenomena and portray who they are to others and to themselves (Marcks and Pawelz, 2022), thereby “forming a whole that is greater than the sum of its parts” (Halverson et al., 2011: 1). Narratives link previous and current experiences with expectations about the future, contributing to a sense of community and demarcating the group as different (Polletta, 1998).

One powerful form of narrative are conspiracy myths. Based on distorted versions of reality (Kerodal et al., 2016), conspiracy myths are “suspicions that a number of actors join together in secret agreement, and try to achieve a hidden goal, which is perceived as unlawful or malevolent” (Van Prooijen et al., 2015: 571). Doing so, they describe grievances via a set of simplistic explanations, compensating for uncertainty and powerlessness (Swami and Furnham, 2014), narratively centering social actors. Similarly, Baele (2019: 711) defines a conspiratorial narrative as a “story which integrates, in a single teleological explanation for the alleged suffering of a given social group, a large range of events from past and present that are allegedly hidden and provoked by various nefarious archetypal actors.” These insights are not unique to the far right and are applicable to cases from jihadist terrorist to communist parties, from infighting within the Nazi party to civil rights movements. And yet, we show here how the infighting is rooted most specifically in mechanisms of group identification.

As Baele (2019) demonstrates, these narratives identify ingroups, outgroups, and hybrid groups as those members of an ingroup who have become contaminated or traitors. ³ The role assigned to the hybrid and outgroups shapes the level of imminent threat to the ingroup, therefore impacting and legitimizing who is targeted with violence and when. The outgroup or hybrid group considered most responsible in different conspiratorial narratives may shift, explaining why levels of violence toward certain groups can vary. To understand the relationship between these groups and their actors, Halverson et al.’s (2011) conceptualization of archetypes is particularly useful. They define archetypes as “standard characters that one might expect to find in a story” which can “unlock motives and operate as ‘shorthand terms for situations’ in which characters might find themselves” (Halverson et al., 2011: 21). Thus, in using an archetype, one can invoke pre-existing and overarching narratives, immediately connecting individual events to larger historical developments. This approach helps understand the power of hybrid actors, such as traitors or pawns, who can be conceptualized as a form of archetype. The attribution of actors to such an archetype is particularly powerful, as the hybrid actor is thereby tied and interpreted through a pre-existing and powerful narrative.

In this article, we explain infighting using the idea of competing conspiratorial narratives and argue that it is the competitive attribution of political actors as hybrid and resulting logics of political action that underlies much infighting surrounding the CHR and results in considerable denigrating language (see Figure 1). The CHR is a useful case for exploring this concept, as it resulted from the growing presence of “conspiracist rhetoric in various socio-cultural milieus in the United States” (Wiggins, 2023: 382).

Figure 1.

Process of infighting (authors).

Methods, limitations, and ethics

As the far right is active across multiple sites (Baele et al., 2021), we engage in a cross-platform analysis of data collected from 8kun, 4chan, and TheDonald.win to study infighting. Although several platforms played a role in the attack—such as Twitter, Gab, Parler, and Telegram among others—we focus specifically on 8kun, 4chan, and TheDonald.win. These platforms were central “free speech” forums in which far-right actors discussed politics and the CHR and had a key position in spreading conspiracy theories, emotional messages, and misinformation during the riot (Ghaffary and Heilweil, 2021; Paul et al., 2021; Rudden, 2021; Timberg, 2021). Except for that, these non-mainstream platforms have long been associated with far-right ideologies and the dissemination of extreme and violent political discourse. Their platform affordances and limited moderation created environments where more radical views could be expressed, facilitating as a result the spread of misinformation, identification of threats, and mobilization (Baele et al., 2021; Klein, 2025; Riley, 2022). Focusing on the events of January 6^th, Klein (2025) argues that anonymity enabled the circulation of extreme violent posts on 8kun and 4chan, whereas on TheDonald.win, the permissive approach of platform moderators allowed users to issue threats and coordinate activities by redirecting discussions to private channels such as Discord.

We sampled 333 posts from each platform from the 7 days before the CHR, on the day itself, as well as in the 7 days after the CHR, ⁴ resulting in 2997 posts (Kasimov et al., 2025). ⁵ This sample size is considerably larger than comparable, qualitative studies, allowing a broader analysis of the dynamics of infighting across these three platforms (Jasser et al., 2023; Scrivens et al., 2023). For 8kun and 4chan, we used the 4CAT Capture and Analysis Toolkit (Peeters and Hagen, 2021) to collect data, drawing on generic keywords derived from preliminary analysis of the case that capture far-right discourses around the CHR on /pnd and /pol boards, respectively ⁶ ; using the toolkit’s date-range options, we collected posts from periods before, during, and after the CHR, after which we conducted sampling for analysis. For TheDonald.win platform, we drew a random selection of posts, downloaded from Kaggle through the Social Media Analysis Toolkit (Distributed Denial of Secrets). Given the focus of the platform on Trump, no specific keywords were needed to filter out politically irrelevant topics for this study (Hendrix, 2022). Not gathering threads but individual posts, this data collection method allows us to parse the data into the three time periods of before, during, and after the attack.

We subsequently analyzed the data with qualitative content analysis, combining deductive coding drawn from previous research with inductive coding generated from within the empirical data itself (Miles et al., 2014). This process continued in iterative rounds of coding and analysis that sought to make sense both of the textual data and ideological and attitudinal information garnered from titles or usernames (Holliday, 2007; Holt et al., 2022). It is important to emphasize again that 8kun and 4chan are fully anonymous, while TheDonald.win allows pseudonyms, meaning that we cannot attribute for the first two platforms which users are posting which content. While this means it is not possible to trace which users are particularly influential on the discussions, we are primarily interested in the way infighting is constructed narratively so this is not detrimental to the research design.

To mitigate ethical and data protection issues, ⁷ all data were stored on secure university servers and access was limited to the researchers directly involved. Further, data were collected only from public platforms, “where individuals have no perception or expectation of privacy” (Mahoney et al., 2022: 233; The British Psychological Society, 2021). All quotes were anonymized and post and user identification numbers removed to prevent the identification of individuals.

Findings

Far-right consensus and infighting in far-right online platforms

Within the studied communities, there is consensus on understandings of the ingroup as white and male, with racism and misogyny going unconfronted across the platforms. Discourses on the platforms construct a common enemy and antisemitic, racist, anti-establishment, anti-leftist, and misogynistic sentiments disparage identified outgroups, mirroring broader anti-minority sentiments in far-right inspired narratives (Gaudette et al., 2021; Holt et al., 2022; Jasser et al., 2023). Across all platforms, we find a sense of inequality and victimization of the ingroup that ties into ideas of victimhood and imperilment (Marcks and Pawelz, 2022), juxtaposed by the moral integrity of the cause pursued by those connected via digital platforms. Based on their analysis of Telegram data, Boussalis et al. (2025: 1421) further argue that the victimhood narrative is not surprising, as “the offences committed by the ingroup will either been seen as justified or minimized by the ingroup.” Finally, posts reveal a latent assumption of an either upcoming or ongoing existential battle (Ahmed and Pisoiu, 2019), although how to deal with this differs strongly.

Beyond these consensuses, the data reveal significant infighting across more than one in six posts, containing hostility toward others on the platform or other far-right actors. This infighting is prominent on 4chan (16.52% of posts) and 8kun (30.5%), with markedly less infighting on TheDonald.win (4.1%); this discrepancy can be explained to some extent by the narrower ingroup boundaries on the platform TheDonald.win, which consists mostly of individuals who supported Donald Trump (Riley, 2022), as well as due to the stronger hierarchies within the space compared with the looser structures of anonymous boards such as 4chan and 8kun. The virulence of infighting is demonstrated by the frequent calls for users to hurt themselves, with multiple posts even telling others: “go kill yourself.” This infighting manifests in the use of insulting language, tailored to the discourses they are embedded in: insults commonly drew on Antisemitic (Ascone et al., 2022; Riedl et al., 2024), anti-Chinese (He et al., 2021; Kim and Kesari, 2021), or otherwise racist slurs or questioned the masculinity of other users (Saresma et al., 2021). Infighting draws on common understandings of the outgroup and by connecting this outgroup to the person being insulted, the posts imply a betrayal of the ingroup. Similarly, some insults connected the targeted individual with a hybrid actor or affiliated political positions, denigrating the individual but also marking a stake in who should be seen as hybrid.

In the following sections, we introduce the two core conspiratorial narratives and then show how these result in diverging constructions of hybrid actors, before delineating how this results in conflicting logics of action and legitimations of violence.

Competing conspiratorial narratives

The stolen election

We find two different conspiratorial narratives in the data. The first conspiratorial narrative asserts that the election was stolen by Biden and that Trump was the legitimate victor. Beyond discussions of how the election was stolen (manipulated counting machines, etc.), Trump is depicted as the savior of the American (white) nation whose contribution is key to winning the raging battle between good and evil. This view ties into broader insights on social media support for Trump, as well as those who follow Trump’s battle cry to “Make America Great Again” (Berger, 2018). Some posts include assurances of loyalty to Trump as leader, referring to how he had fought for the ingroup. This likely includes some from the group who amassed for the “Stop the Steal” rally in Washington D.C. on 6 January and were involved in the CHR (Donovan et al., 2022). In this way, we can see how Trump is assigned the archetype of hero or savior within this narrative.

The Zionist Occupied Government

The second popular conspiratorial narrative in the data draws on the idea of a Zionist Occupied Government (ZOG) (Holt et al., 2022), “which reflects the common white supremacist belief that the U.S. government is controlled by Jews” (ADL, 2023). Here, there are frequent references to the New World Order (NWO) conspiracy, which purports the existence of a Jewish world power that controls Western governments (ISD, 2022). Antisemitic stereotypes play into the construction of a “cabal” of leaders that restrict individual freedoms and control the world’s political, economic, and media systems. Next to the evil, all-powerful villain archetype assigned to Jews in this narrative, we can see that those who are “controlled” by these villains fall into the archetype of a puppet or pawn, whose weakness means they turn their back on their actual allies or ingroup.

Diverging attribution of hybrid actors

Hybridity due to inaction—Pence and Trump as saviors who disappointed

Each conspiracy theory constructs specific actors as hybrid, that is as somehow betraying the interests of the ingroup, and are assigned archetypal roles such as pawns or traitors. Within the first conspiracy theory of the stolen election, there are two camps. First, some continue to loyally see Trump as their savior, while outgoing Vice President Mike Pence is communicated as a hybrid actor because he did not undermine the certification of the election. This position is especially prominent on the pro-Trump platform TheDonald.win and one post reads: “Mike Pence is a fucking traitor. I didn’t want to believe it, but he’s finally ripped his mask off” (TheDonald.win, 06.01.2021). A similar hybridity is also suggested for Mitch McConnell and Ted Cruz, whom Trump loyalist contributors frame as traitors who have betrayed or backstabbed Trump and his supporters, some even claiming they work for Biden.

In the second camp, Trump’s hesitant reaction to the riot is understood as betrayal and Trump himself is perceived as a hybrid actor. One contributor asks: “Why did Trump start a fight then run and leave his supporters to die?” (4chan, 06.01.2021) while another states: “He literally called in the national guard on his own supporters after whipping them up” (8kun, 06.01.2021). Though our sampling method does not allow us to observe the individual shifts of specific contributors, at a discursive level we see positive references to Trump become less frequent during the CHR, with an increase of references to Trump as a hybrid actor. These portray Trump as weak, inactive in a time of need or having turned on his supporters. Some users even ask whether he may actually be part of the establishment, while even more posts suggest that Trump is actually Jewish. Thus, Trump’s hybridity stems from the potential his supporters believed he had to protect them from myriad threats. By not acting successfully to remain in power, he is seen as betraying the trust that was put in him.

Both camps subscribe to an idea that actors have become hybrid due to their inaction in times of need, particularly with Pence not intervening adequately in the electoral process and a lack of support for rioters and Trump telling them to go home, respectively. This relates to ascriptions of cowardice or weakness and the idea that they are pawns. Interestingly, this ties into the construction of archetypal actors, in particular in the interpretation of Trump. For some in this camp, Trump continues to be the archetypal patriotic hero risking everything for his nation. For others, Trump acts like the archetype of a pawn, backstabbing followers and their cause, or even fulfilling the archetype of fallen heroes or weak rulers, failing to live up to the hope imbued in them. In drawing on these archetypes, the commentators are lending credence to their own interpretation of present events, by tying them to existing narratives for understanding the past. The choice of archetype is further useful, for it positions the commentators themselves as victims of such archetypal actors, thereby freeing themselves of any wrongdoing and enabling a continued belief in their ideological interpretation of the world.

Hybridity due to implication: Trump as a Zionist puppet

The second conspiratorial narrative is opposed fundamentally to the current political system as it is controlled by the ZOG and thus sees Trump as a more hybrid actor and fundamental part of the larger problem. Here, the hybridity of Trump is not newly revealed but deeply rooted in larger discourses about how his position within the establishment makes him a member of the Jewish outgroup or hybrid in that his aims align with an outgroup. One post on 8kun on the day before the CHR reads:

I don’t support Trump, I support the chaos. I think a Siege-like situation can come from this, Trump was a kike ⁸ pawn and they pumped and dumped him and he’s not following the script I dunno what’s going to happen tomorrow but it’s definitely the closest to something legitimate in US history (8kun, 05.01.2021).

Trump is perceived to have made fundamental mistakes, that all patriots must pay for, by working within the established system. This is reflected in terms such as Trumpn*, ZogEmperor, or MAGA emperor. Furthermore, the Republican establishment and conservatives, as well as many state institutions such as the military, police, and national guard are seen as mere puppets of the ZOG. Trump and the broader Republican party are suggested to being pro-Jewish and acting as puppets for Israel or Jews, exemplified in a post on 8kun: “Fuck off, kike. If you’re white and republican, you are a traitor to all white people. Neck yourself” (8kun, 07.01.2021). As such, posts embed current political events in a distorted understanding of history in which nefarious Jews manipulate others to gain global power. In this way, powerfully and pre-existing antisemitic archetypes are drawn on to directly tie Trump to them and position himself in the archetype of a traitor. The power of archetypes is here revealed, as little more than key words, such as Zog or puppet is required to frame Trump as an archetypical hybrid actor within an antisemitic master narrative shared by the far right.

These posts label Trump’s followers as MAGAmorons, MAGAtard, or MAGA idiots. Besides portraying Trump supporters as stupid, they are frequently labelled either Jewish or tightly connected to Jews, provoking antisemitic slurs, particularly on 8kun. Notably, not only Trump but his followers and those who maintain their loyalty to him despite his actions proving his hybridity are constructed as hybrid actors and defamed. Whilst such references are particularly frequent on 8kun, less on 4chan, and almost never on TheDonald.win, the defamation occurs not through association to or making fun of different online platforms but rather by focusing specifically on their ideological allegiance to Trump (by referencing MAGA). Thus, online ecosystems show a preference for certain conspiratorial narratives over others; the battle lines are not drawn via belonging more to one social media platform—as Baele et al.’s (2023) paper on far-right online ecosystems might suggest—but view the conspiratorial narratives that gain traction within each platform.

In this context, the purportedly stolen election figures prominently. Some argue that Trump lost the election, but most dispute the election’s legitimacy, tying it into clearer attributions of fundamental problems: “We now live in an era where you can openly commit voter fraud in the US, and get away with it. Arguably our votes have not matter [sic!] for a long time, but now that illusion is shattered” (8kun, 06.01.2021). Others critique the attempts to “save” the election by rioters, rationalizing that there is no choice between the Left and Right, given that all are perceived to be controlled by Jews. Also, hostility toward the state manifests in reduced trust in authorities and critiques of reductions of freedom, such as a post reading: “Fuck America larping freedom state” (4chan, 03.01.2021). ⁹ As the next section shows, the conspiratorial narratives regarding Trump and the establishment being controlled by (Jewish) overlords lead to different logics of political action.

Conflicting logics of action and violence legitimations

Storming the Capitol as an attempt to maintain Trump’s power

From these competing conspiratorial narratives, there result conflicting logics on the right course of political action and how the violence of the CHR is evaluated. The first perspective building on a belief that the election was stolen sees the CHR as an act of resistance, heroism, and martyrdom. With the goal of maintaining Trump in power and attributions of those who attempted to prevent this as traitors or puppets emerges a logic of action that makes mobilization and a takeover of Capitol Hill laudable. While some people call on senators, particularly Republican senators, to do their part to “stop the steal,” there are many more calls to action on behalf of contributors in the run-up to and on 6 January.

Some calls to action remain vague, such as “Senate and pence has fled. If we can keep them constantly fleeing they can’t count their fraudulent votes” (4chan, 06.01.2021), while others are considerably more direct demands to participate, including mentions of the location or the planned timing of the event. Some calls to action make implicit positive references to violence, whilst others directly call for violent action.

While some posts discuss that Biden won and Trump lost and that the community needs to move forward, many posts include resounding repetitions of hope including a belief that there is a larger plan and that Trump is in control. Here, TheDonald.win was most prominent, featuring posts such as “Don’t worry, IT’S NOT OVER YET!” (TheDonald.win, 06.01.2021) and “Sounds like we all just need to trust the plan even harder. Extra, extra hard. Like so hard. Trust trust trust” (TheDonald.win, 06.01.2021). As the day of the riot progressed, the increasing nature of Trump’s hybridity becomes clearer as contributors assumed he was no longer supporting the riot and uncertainty on whether he had called the National Guard.

Despite many calls to violence, there is little discussion of the destruction caused by those entering the building; instead, their actions are portrayed as peaceful, legal, and legitimate. Contrasting this, there are several discussions of violence against protestors who are construed as unarmed and defenseless, thereby tying into the consensual perspective of the historic victimization of the ingroup. There are many mentions of protestors being killed or murdered, with far-right veteran Ashli Babbitt frequently mentioned and cast in the archetype of martyr. This is juxtaposed with Black Lives Matter action which is argued to have been disproportionately treated more leniently than the police response to the CHR. The sense of inequality resonates with the broader narratives of inequality and heightens the degree of martyrdom and heroism constructed around CHR participation.

Storming the Capitol as a contribution to broader chaos

The second perspective that sees Trump as a part of the fundamental problem offers little support for the CHR because its purpose is to maintain the power of Trump. Political competition between Republicans and Democrats is portrayed as farce given that both are perceived to be puppets of the ZOG.

Given the hybridity of the political elites, this perspective has no interest in taking power through the ballot box. There is a severe sense of paranoia regarding the role of mass media and technology in supporting the establishment and directly threatening the ingroup:

Do not trust any form of digital communication. Everything is copied, stored, and analyzed. Even what you type in the reply box and delete before posting. Google is here slurping up everything. Google = Alphabet = US Government = NOW […]

Return to the old ways. Don’t trust anyone but family. Don’t let your family watch TV, use cell phones, or any type of screen […] The revolution will not be televised (4chan, 30.12.2020).

From this dire reading of the world comes an acute sense of threat and the urgency to enter battle with those powers that can hinder the radical renewal of the political system. While there is often some degree of regret that this is necessary, contributors argue that it is imperative given the state in which the world now finds itself. These posts suggest a logic of political action that embraces palingenetic fantasies “of a nation in demise that could only be resurrected through purges and uncompromising measures” (Marcks and Pawelz, 2022: 1417). While some demand that neighbors should be educated into the thinking of this perspective (they need to be “redpilled”) to grow the movement, the more accepted route is to sow dissent and disorder, following an approach of “accelerationism”; the idea that by sowing chaos the inevitable demise of the system can be brought around more quickly (Johnson and Feldman, 2021). In this sense, re-installing Trump and the CHR are not seen as worthy goals, but they follow a logic of political action that supports the unrest they create as useful tools in an accelerationist agenda.

Posts suggest further forms of action beyond the CHR that support this goal, including revolution, the destruction of political groups, or waging physical war against outgroups. Violence against outgroups is particularly prominent: “US Civil war imminent. Joggers gets ropes. Kikes gets ropes. Crackers reign supreme. Chinks get ropes. 1000 years of peace and prosperity. Trannies sent to Venus permanently. Sensual nudity allowed on prime-time TV. Hulk Rules” (4chan, 03.01.2021). This reference to the U.S. Civil War shows how constructions of ingroup, outgroup, and hybrid group enable commentators to draw on pre-existing and shared interpretations of the past to position themselves as archetypal heroes battling against archetypal enemies and thereby legitimize the use of physical violence.

Even if there are mixed impressions of whether the larger fight is already lost or being won, many posts claim that Trump supporters will remain more broadly inactive or criticize the lack of action during the CHR, suggesting that the CHR will not unfold a lasting effect. In this way, they contrast their preferred logic of action (accelerationism, chaos, and physical violence) with the actions of hybrid actors they denigrate and evaluate as failed. The construction of hybrid actors becomes a differentiating line not only along ideological infighting but also central to understanding divergences based on different logics of action proffered within the far right. One contributor on TheDonald.win asks “But why aren’t people burning the shit?” (TheDonald.win, 06.01.2021) whilst an 8kun post reads:

Revolution lol. Its just a bunch of fucking mutts going into the capitol building, breaking some stuff, backing down and complying like bitches when encountering [sic!] real armed resistance inside.

If this was a revolution they would have gone in with arms and we would be watching a livestream about them hanging the senate (8kun, 06.01.2021).

Discussion and conclusion

We have demonstrated that to understand infighting in far-right online communities during offline trigger events we must look to competing conspiratorial narratives in which diverging hybrid actors are defined; from these, conflicting logics for political action result and violence becomes legitimated in different ways. By developing clear lines of difference between the ingroup and hybrid group, conspiratorial narratives sharpen distinctions within far-right online communities regarding interpretations of past and present political events, as well as a distinct logic for action proposed for the present and future; as there are starkly different conspiratorial narratives leading to different calls for action, these become the locus of far-right infighting on social media and forums.

As already discussed, Berger (2021) argues that when an extremist movement does not receive support from those it perceives as part of its ingroup, it may respond by criticizing these members. Thus, such dynamics may help explain why infighting functions as a process of boundary construction, specifying who belongs to the group (Ghaziani and Kretschmer, 2018; Holt et al., 2022). More broadly, this also shows how events such as the CHR can be interpreted through distinct group cultures, producing competing meanings and justifications for action.

The CHR was a watershed moment in current American politics that revealed fundamental opposition to the democratic system. While there are key points of consensus within the online communities we studied, our research shows that deep rifts exist that manifest in hostility and infighting. We argue that such infighting is grounded in fundamentally diverging perspectives on the world and current politics. These perspectives are embedded in conspiratorial narratives that outline and legitimize who should be targeted by violence (Baele, 2019). These competing conspiratorial narratives share definitions of who is part of the ingroup and the outgroups but are split by diverging conceptions about who is part of the hybrid group and why. It is this definition of who is hybrid and why they are defined as such that is key to defining a logic of action in reaction to the situation, with the CHR being only one among several different actions that were perceived as necessary.

Existing scholarship has shown that ingroup dynamics, such as competition and contested tactical decisions, play an important role in understanding infighting (Ellefsen, 2018). Moreover, infighting has been associated with radicalization processes online and the emergence and form of political violence (Mølmen and Ravndal, 2023; Törnberg and Törnberg, 2025; Wahlström and Törnberg, 2021; Youngblood, 2020). By bringing the narrative turn in social studies to our study of infighting and developing the idea of competing conspiratorial narratives, this paper has shown how discussions on logics of action take place through and because of diverging interpretations of the past and present. Narratives provide a temporal linking of past, present, and future understandings of the world that are in turn intimately connected to identity formation (Marcks and Pawelz, 2022; Polletta, 1998). As our analysis has demonstrated, competing narratives are based not solely on disagreements about the unfolding of certain events, but are rooted in conceptions of identity, with the delineation between ingroup and outgroup becoming actively contested through narratives of hybrid actors. In the case of the CHR, these competing conspiratorial narratives resulting from diverging understandings of the hybrid group created the foundation for contesting the next steps for the far right. Therefore, debates about narratives are a fruitful avenue for researching infighting online, as they enable an understanding of both the identity formation of online communities and the emergence of proposed logics of political action, including political violence or radicalization.

With the idea of competing conspiratorial narratives, we provide a heuristic to discuss far-right infighting by disentangling it, explaining how and why different actors are seen as hybrid and thus suggesting differing logics of action. We hope that this concept will support research on other cases of online interactions in relatively non-hierarchical settings to understand how hate against outgroups as well as other (previous) members of one’s ingroup develops.