Social media discourses amidst ethnopolitical extremism and conflict: The case of Ethiopia

Abstract

This article examines how influencers and political armies in Ethiopia use social media specifically Facebook to propagate sectarian rhetoric and ethnopolitical extremism within a society grappling with protracted internal conflicts, prolonged ethnopolitical extremism, and armed clashes. The study employs qualitative research methodology, using critical discourse analysis (CDA) alongside facets of rhetorical analysis, to examine Facebook posts and comments produced in English and Amharic languages during the post-Tigray War politically tense period (the final quadrimester of 2023) amidst armed conflicts in the Amhara region. By purposively sampling content from pages of political influencers and elites, the study investigates the rhetorical patterns and discursive constructions shaping ethnopolitical narratives. The analysis reveals that the rhetorical and discursive patterns and textual trajectories on Facebook in Ethiopia are characterized by hostilities and animus, which are prone to promoting ethnopolitical radicalization and deepening the schism among ethnic and political groups.

Introduction

During its 27-year rule, the Ethiopian People’s Revolutionary Democratic Front (EPRDF) – a coalition of the Tigray People’s Liberation Front (TPLF), Amhara Democratic Party (ADP), Oromo Democratic Party (ODP), and Southern Ethiopian People’s Democratic Movement (SEPDM) – was widely criticized for deepening sectarian divisions and undermining pan-Ethiopianism by privileging ethnicity as the primary mode of political mobilization (Abbink, 2011; Kelecha, 2025; Levine, 2011; Mekonnen, 2015). Mounting protests culminated in 2018 with the TPLF’s displacement from federal power and the rise of Abiy Ahmed as prime minister. In December 2019, Abiy dissolved the EPRDF to form the Prosperity Party (PP), a move rejected by the TPLF, which opposed the centralization of authority. Relations deteriorated further when the Tigray regional government defied federal ¹ directives by holding regional elections in September 2020. The federal government declared the elections illegal and suspended funding, triggering a devastating 2-year war (November 4, 2020–November 2, 2022) that claimed over a million lives.

While the Pretoria peace agreement in November 2022 between the government and TPLF sparked hope for stability, unresolved political tensions and armed conflicts have persisted in many parts of the country, specifically in Amhara and Oromia, two most populous regions. In Amhara, frequent attacks on civilians, widespread extrajudicial killings (Office of the United Nations High Commissioner for Human Rights [OHCHR], 2024), and the federal government’s decision to disarm the Fano forces, previously a key ally during the Northern Ethiopia War, has led to violent resistance. The situation escalated since early August 2023 despite the central government declaring a consecutive state of emergency. By late 2023, the region’s political and security structure teetered on the brink of collapse. The war between the two forces continued, and the government used heavy artillery, including military drones which resulted in huge casualties including the killing of innocent people (Human Rights Watch, 2024; OHCHR, 2024). The armed group, Fano, is organized into commands and fighting throughout the region. Similarly, the Oromo Liberation Army (OLF), (commonly referred to by the government as “OLF-Shene”), formerly designated a terrorist group, sustains insurgencies in Oromia. These conflicts underscore Ethiopia’s deteriorating stability, reflected in its 2024 ranking at the bottom of the Global Peace Index and its designation as a fragile state (IEP, 2024).

In such a situation, the media, specifically social media (SM) play key roles. For instance, Levine (2011) highlighted that the rise of SM has created new platforms for identity formation, conflict, and political mobilization, often fueling long-distance nationalism and virtual political activism. This was also observed in the case of Ethiopia. The tightly controlled media landscape in the country that were found weak in reporting ethnic conflicts (Moges, 2017) resulted in creating a vacuum which increasingly open spaces to SM platforms such as Facebook (Workneh, 2021). However, these platforms are frequently rife with hate speech and polarization, exacerbating divisions (Skjerdal and Moges, 2021). For instance, Facebook has disseminated extensive ethnic divisive content (Chekol et al., 2021) and widespread hate speech (Mulugeta, 2019). Orgeret et al. (2025), moreover, examined “hatetaging,” the use of hate-filled hashtags on X during Ethiopia’s Tigray conflict. Against this backdrop, this study explores how discursive acts on Facebook contribute to incitement to violence and political polarization in Ethiopia’s recent volatile context.

Like offline public discourse, Facebook discourses are constitutive of identity (Georgalou, 2017). Polarizing and divisive content tends to outperform other types of content on the platform in terms of engagement and virality (Stray et al., 2023). For instance, former Facebook employee-turned-whistle-blower Frances Haugen described Facebook’s role during Ethiopia’s Tigray War as alarming, with the platform fanning ethnic violence. Studies by Akinwotu (2021), Megersa and Minaye (2023), and Roychowdhury and Gessesse (2024) also confirmed that the Northern Ethiopia War was also fumed by the social media. Similarly, Skjerdal and Moges (2021) and Moges et al. (2022) noted that given the country’s history of political turbulence and a polarized media landscape SM, while not the root cause of conflict, intensifies violence in unprecedented ways. Particularly, Facebook, as a primary source of miss/disinformation in Ethiopia (European Institute of Peace, 2021), is weaponized by institutional and non-institutional political actors to incite sectarian conflicts (Proctor, 2021; Workneh, 2021). A notable example here is Prime Minister Abiy Ahmed’s Facebook post during the Northern Ethiopia War calling for Ethiopians to “bury the terrorist TPLF” using terms like “weeds” and “cancer,” which Meta (Facebook’s mother company) later removed from its page. ² While such incidents do not necessarily establish direct causation, studies reveal a strong correlation between incitement on SM and real-world violence against targeted groups (European Institute of Peace, 2021; Stray et al., 2023). Above all, this demonstrates a complex interplay between ethnic violence, political polarization, and SM in Ethiopia.

While SM is accused of spreading hate speech and dangerous content, how their content incites and exacerbates conflict and expands the ethnic polarization in the current unrest in Ethiopia remains unstudied. This article thus seeks to explore how ethnopolitical narratives on Facebook reinforce division and hate among communities, and promote violence, political extremism, and ethnic conflict in the country. It critically analyzes the discursive and rhetorical strategies deployed on Facebook during political crises, particularly after the Tigray conflict and the ongoing unrest in Amhara. Rather than evaluating Facebook as a corporation, the study focuses on how influencers ³ use the platform – emphasizing their objectives, techniques, and audience engagement – within the broader media ecology (Riedl et al., 2023). Exploring the nuanced human meanings and interpretations embedded in Facebook discourse, the study stresses the importance of a context-sensitive lens in understanding SM dynamics.

Research methods

The study examines Facebook posts and comments produced in Amharic and English languages during the post-Northern Ethiopia War politically tense period, specifically the final quadrimester of 2023 amid rising tensions in the Amhara region. ⁴ It employed Critical Discourse Analysis (CDA) along with rhetorical frameworks to investigate how narratives and discourses are strategically crafted to incite ethnopolitical animosities.

The study utilized several frameworks based on Wodak and Meyer’s (2009) classification of CDA approaches. The socio-cognitive perspective (Van Dijk, 2016) was deployed to analyze discourse structures that reflect socially shared perceptions and cognitive polarizations, particularly in portraying in-groups and out-groups. The social actors’ approach aided in exploring the formation of echo-chambers and epistemic conflicts driven by SM influencers positioning themselves as sociopolitical “elites.” This aligns with Bouvier and Machin’s (2018) assertion that SM discourse analysis must consider how audiences use texts, not solely the texts themselves. By analyzing influencers’ discourses, the study reveals how they frame interpretations of reality, echoing Van Leeuwen’s (2016) argument that discourse serves as a lens for understanding sociopolitical contexts. In relation to the dialectical-relational approach, Fairclough (2016) further examined the linguistic and semiotic dimensions of subtle polarization, illustrating how Facebook discourse reflects broader sociopolitical relations. This perspective enabled an understanding of Facebook as a site of human interaction and a reflection of sociopolitical dynamics, integrating semiosis with contextual analysis.

In addition to CDA, rhetorical theories, particularly Kenneth Burke’s and Jeanne Fahnestock’s frameworks, guided the analysis. Burke’s (1969) concepts of symbolic action and identification (through rhetoric devices) were pivotal in understanding how influencers foster alignment and engagement, forming echo-chambers. His focus on rhetorical motives (presenting where it is not usually recognized) and social context (i.e., the social use of language) underscores the complexity of discursive strategies beyond mere persuasion. Similarly, Fahnestock’s rhetorical style was instrumental in informing language analysis methods rooted in the rhetorical tradition, including figures of word selection, amplification, and sentence structure (Fahnestock, 2011), to discern the intentions of Facebook post authors. The study also incorporated her concepts of rhetorical figures in scientific narratives to examine how Facebook users construct narratives to frame issues, influence perceptions, and engage audiences. Given her focus on “the technique of rhetoric itself,” Fahnestock (1999: viii) and the universal applicability of rhetorical tactics, devices such as repetition, emotional appeals (e.g., anger, fear, and indignation), antithesis, and force or vividness proved essential for analyzing Facebook posts.

This critical SM approach departs from dominant methodologies, such as digital positivism (Fuchs, 2017), by emphasizing qualitative analyses of relatively “small data” rather than large-scale quantitative analysis. Specifically, the focus here is on explicating patterns of rhetorical discourses and Facebook narrations by non-regular political actors. Fuchs (2017) notes that critical SM analysis can focus on various units of study – whether individual user posts, specific periods, discourse topics, or particular social groups – enabling an exploration of intertextual and interdiscursive dynamics that connect texts, rhetoric, and contexts. By diverging from digital positivism, this study seeks to provoke discussions that enrich scholarly conversations and complement big data analytics, which often overlooks the agency of individual actors in shaping online discourses and rhetoric in the SM.

The study purposively selected Facebook discursive artifacts from the final quadrimester of 2023 (September to December) for analytical focus. In Ethiopia, Facebook functions as a key platform for (socio)political conversations (Gagliardone et al., 2015). This period also coincided with the active conflict between the Ethiopian central government and Fano forces in the Amhara region, a context that significantly shaped SM interactions. Drawing on Tagg et al.’s (2017: 5) observation that SM users “imagine and respond to a particularly complex set of contextual variables as they design their posts and interactions,” this timeframe offered a lens into the dynamics of Facebook discourse during conflicts. By focusing on this specific timeframe, the study seeks to elucidate the interplay between conflict-driven exigencies and discursive practices within digital spaces.

Data collection began with identifying prominent SM actors and activists through consultations with Ethiopian journalists, academics, and SM participants. Reflecting the ethnic fragmentation of Ethiopian politics, SM activism often mirrors these divides. The study prioritized Facebook pages of influential actors representing diverse political and ethnic affiliations, including Amhara, Oromo, pan-Ethiopianists, and pro-government influencers. Inclusion criteria emphasized four factors: (a) prominence in articulating concerns of the Amhara and Oromo groups, whose rivalry over Ethiopian identity and politics has intensified since the rise of an ethnic Oromo Prime Minister, (b) prominence in advancing pan-Ethiopian views, (c), views advocating government political interventions, and (d) active engagement with issues such as the Amhara conflict, ethnic strife, identity politics, and political rivalries between and among ethnic groups.

Prominence was also assessed based on consistent posting on these themes and follower counts, with selected influencers (as of September 2023) having at least 30k followers, one exceeding 600k. The analysis concentrated on posts from Amhara, Oromo, pro-government, and pan-Ethiopianist influencers to illuminate historical contestations and the escalation of discursive conflicts amid political and military developments in the regions. To enhance the depth of analysis, the study also examined audience engagement, including comments, offering insights into collective meaning-making. A preliminary dataset of 206 posts was gathered from the timelines of 10 selected (political) influencers. These posts were then filtered based on four inclusion criteria: (1) focus on ethnic/historical tensions or rivalries between the Amhara and Oromo groups, (2) emphasis on the ongoing conflict in the Amhara region, (3) engagement with narratives of central government militarism, and (4) rhetoric surrounding pan-Ethiopianism, nationalism, or pro-government interventions. This resulted in a final dataset of 96 Facebook posts. From this dataset, 10 key posts were selected as focal examples to anchor and contextualize the analysis.

Regarding translation, interpretation, and data presentation, the article followed a systemic approach. As most posts were in Amharic, they were translated by authors of this article into English using a combination of literal and contextual translation methods. Literal translation prioritized lexical fidelity, while contextual translation emphasized cultural and political resonance, aiming for equivalent impact between source and target texts (Georgalou, 2017). The translated multimodal resources were read line-by-line to familiarize the researchers with the data, followed by coding relevant text segments and paragraphs. Codes were subsequently collated and synthesized into initial themes, which were refined into the following eight: Contestation of national identity narratives; identity and historical memory; politicization of historical narratives; ethnic polarization through historical memory and perceived suppression; historical revisionism and epistemic manipulation, resentment and politics of suppression; influencer-driven grouping/ethnic mobilization; narratives of suppression and retributive justice, including like-mindedness, antagonism, erosion of political ideals. Through further refinement, these themes coalesced into four overarching themes/frames of SM discourse (during political upheaval): discourses of suppression, echo-chambers, grievance salience, and epistemic manipulation. These final themes were defined and reviewed iteratively, enabling progressive focusing, described by Simons (2014: 464) as a “gradual reframing” of the data. This process facilitated meaning-making, and the interpretation of evidence embedded within the themes, culminating in the findings presented below.

Data presentation and analysis

Discourses of systemic injustice

Numerous Facebook posts and their responses consistently depict various groups framing themselves as victims of historical and political injustices. Whether linked to ethnic, political, or religious identity, nearly all factions articulate grievances related to systemic oppression. Terms such as “genocide,” “savage regime(s),” “persecution,” “fascism,” “imperialism,” “colonizers,” and “ethnic cleansing” are commonly used, particularly by influencers identifying with the Amhara or Oromo communities, to describe the perceived oppression their ethnic groups have been facing under various regimes or dominant ethnopolitical forces. As illustrated by a post (Figure 1), attributed to an Amhara influencer (quoting a Fano regiment leader), for instance, reads that “the Oromo do not allow the Amhara people to live even if they convert into Oromo.”

Figure 1.

The post roughly translated as, “Even if the Amhara were to say, ‘we will identify as Oromos as you wish’, the Oromos would still not allow them to live.”

Amhara influencers often assert that extreme Oromo political actors (such as Oromo Liberation Army) are committing genocide against the Amhara people, that being an Amhara is criminalized under the current regime, and that the Oromo-led government is hostile to Amhara. Influencers repeatedly used this rhetoric. For instance, an Amhara influencer used the term “fascist” 8 times in 18 posts. Such messages, which focus on emotional appeals and ethnic tensions, often obscure the root causes of violence and political conflict.

Habtamu Ayalew, a former opposition leader in Ethiopia and current self-proclaimed journalist and political analyst based in the US, frequently asserts that the Ethiopian government indiscriminately targets the Amhara people based solely on their ethnicity. In one of his videos posted on Facebook and shared by Amhara influencers and others voicing the Amhara cause, he frames the issue not as a matter of governance but as a fight for the survival of the Amhara people, claiming that merely being Amhara now warrants execution by the Oromo-dominated government. He calls for unity among the Amhara and other Ethiopians to resist what he labels a “fascist” regime under Prime Minister Abiy Ahmed. In this video, showing a throat-slitting gesture, Habtamu expresses his willingness to sacrifice everything, even his life, to protect the Amhara from what he perceives as ongoing systemic persecution.

Similarly, Oromo activists/influencers/political elites and other ethno-nationalists, including TPLF supporters, engage in comparable rhetorical strategies, invoking historical grievances related to the cultural and linguistic dominance of the Amhara under Emperors such as Menelik II. These groups argue that the Amhara imposed their cultural values on other ethnic groups, with symbols of Amhara supremacy permeating what is considered Ethiopian culture. Such discourses, illustrated in posts by Oromo influencers (see Figures 2 and 3), tend to focus on historical fault lines and exacerbate destructive patterns of behavior, reinforcing a reductive argument that Ethiopia’s complex problems stem from the ethnic supremacy of the Amhara. Such historical animosity has contributed to tensions, particularly between the Amhara and Oromo elites.

Figure 2.

Post claiming that, as underlined in the book, Shoa and Wollo (in the Amhara region) are Oromo areas, with Amhara as minorities.

Figure 3.

The post author claiming that the animosity between Oromo and Amhara is deep-seated.

Figure 3 translation and interpretations

The post author addresses those attributing blame to the TPLF for perpetuating the narrative that Menelik mutilated breasts to instill enmity between the Amhara and Oromo. He contends that the Oromo people do not need anyone to tell them what has happened. According to the post, the Oromo people were prohibited from using their language by the invaders (Menelik’s army). Hence, during the Italian invasion, the Oromo chose the lesser of two evils, favoring (Italian invaders) over Menelik, as the following poem encapsulates the sentiment. It can be roughly interpreted as: although white, gabi* ⁵ is better than shash* ⁶ ; notwithstanding evil, (fascist) Italia is better than neftegna. The term “neftegna” (Amharic for musketeers) is used by Oromo political elites (and other ethno-nationalists) derogatorily towards Amhara (elites). The term is often associated with the consolidation of Ethiopian imperial authority in the Southern regions. Therefore, the post highlights the deep-seated animosity of the Oromo (elites) towards their Amhara counterparts.

Thus, SM influencers particularly those who supposedly represent Oromo assert that Amhara political elites foster intentions of exerting dominance and transforming the nation into an Amhara-led entity. Hence, they reignite their resentment and embody narratives of their perceived historical oppression under Amhara emperors, characterizing Fano, the Amhara militia (that portrays itself as a resistance movement against persecution by the central government and any aggressors), as “unitarist,” “Jawisa” (loosely translating to an idle wander who shirks working), “Janjaweed organization,” and an “invading force.” This discourse of historical and systemic injustice and domination, therefore, amounted to an unrelenting war of rhetoric among different ethnic and political groups/elites about the Ethiopian’s history and each group’s place in it. Whether the claims of systemic injustice put forth by different groups hold merit or otherwise, they create affective polarization and stoke animosity between groups.

Figure 4 translation and interpretations

In this picture, the initial post appears to be authored by Mesay Mekonnen, a self-claimed journalist based in the US. Mesay’s post reflects his bafflement regarding the widespread display of the Oromia regional state flag in various places in Addis Ababa – Ethiopia, including around Menelik Square, which features the statue of Emperor Menelik II. Mesay frames it as an act of “racism and teregninet ⁷ ” by the government. An Oromo influencer reposted Mesay’s remarks, suggesting instead that Menelik’s statue be removed to honor the flag – a reflection of contested historical narratives, as some Oromo view Menelik as a conqueror and perpetrator of ethnic violence, while Amhara elites and Ethiopianists revere him.

Figure 4.

An online dispute over the display of the Oromia regional state flag around the statue of Menelik II.

In sum, the discourses and rhetoric of historical and systemic (political) injustice often emphasize emotional appeals – pathos – and exploit sentiments such as fear, anger, or pride to garner support for specific viewpoints that can yield violent reactions or conflicts. To an extent, the rhetoric of oppression appears to be an expression of actual or perceived existential fear of another group posing a threat against one’s own. The challenge lies in the fact that whether the rhetoric of oppression is established on credible evidence of real danger or not, instilling fear among one’s group is a key mechanism that may lead to violence than even propagating hate speech that aims at engendering hatred against the other group (Buyse, 2014: 785).

Echo-chambers

In Ethiopia’s SM landscape and public discourse, divisions are stark among ethnic-based activist groups – such as Oromo activists, Amhara activists, Tigray activists, PP supporters, etc. This segmented communication strategy appears to create isolated enclaves of like-minded individuals, resulting in individual-level filter bubbles that expose audiences predominantly to in-group narratives. As demonstrated by the responses of group followers (see Figures 5 and 6), this dynamic has given rise to a social proof phenomenon, wherein a few prominent SM influencers amplify specific issues to achieve virality and bolster their popularity within their respective groups.

Figure 5.

Pro-Amhara influencer post urging the “Amhara army” controlling the country to prevent Oromo rule, with unanimous commenter support.

Figure 6.

A re-post showing overwhelming support for the view of an Oromo influencer suggesting the removal of the statue of Menelik II to honor the Oromia flag.

Figure 5 translation and interpretations

A pro-Amhara influencer calls for Fano (“the Amhara army”) to seize control of Addis Ababa and prevent Oromia (“Region 4”) from ever gaining power, portraying the region as a genocidal threat comparable to ethnic Hutu extremists. “We should not forget that Region 4 is a force prepared to carry out a genocide in which the public will participate en masse,” the post claims. The post’s comments unanimously endorse this perspective. The first comment vows – “we will correct emperor Menelik’s mistakes eternally.” The subsequent comments echo similar sentiments: “true; correct (victory for Amhara); you are the only one who dares to tell this fact; you are a true Amhara; bless you! Its name [Oromo] should be wiped off from the global map!; it’s true; 100 [% true]; exactly; exactly; true.”

In this discourse, the notion of “truth” does not function to validate the factual alignment of the opinions presented. It signifies an emotional alignment with the influencer’s perspective. Truth, here, reflects enthusiasm, approval, and preference for the post’s rhetoric, independent of its factual accuracy. Within this discursive framework, truth is constructed through consensus, creating a perception of reality grounded in collective agreement.

Figure 6 translation and interpretations

As shown earlier (see Figure 4), Mesay Mekonnen’s post expresses surprise at the widespread display of the Oromia flag in Addis Ababa, including around Emperor Menelik II’s statue. An Oromo SM influencer reposted it, suggesting the statue be removed in honor of the flag, a view overwhelmingly endorsed in the comments. The comments can be summarized as follows respectively: it will happen, this beast; it is not your concern, shut up; swallow it; we will hang it at your door; finally the country will be returned to its owner, Oromo; close your eye and swallow it; Mesay Mekonnen this comes by the blood of Qeerro [youth movement for freedom organized with Oromo nationalism] not by kererto [patriotic songs often used by the Amhara people]; [Mesay]to reconcile with us, it would be better if you wear a towel with Aba Gadaa flag; you can hang [the old Ethiopian flag] in Bahir Dar[capital of Amhara region]; it is your right to hang your flag in your region and capital city, those who don’t accept it can swallow a metal; who should we ask for permission in our own Oromia regional capital; insane Mesay, we hanged the flag in the capital city of Oromia, it is better to remove the statue in honour of the flag; envious, if this angered him, it will be an overdose for him when we do more; it’s just a beginning; besides Oromo, what do other regions do here?;[Menelik] must get off from our horse; the flag is hanged in its region; which region is prohibited to wave its flag? Finfinnee is Oromia, OK; after this, we will erect Hachalu’s [he was a popular ethnic Oromo singer] statue where the horse stands; you are trapped, monster corpse, your children will reap your crime when the time comes; it is enough for you [Amhara] now, take it somewhere [Menelik’s statue] and you can throw the stones into the water; it’s only a matter of time; instead of dismantling it, it will be good to write the historical violence and abuses he committed for people to visit; it makes me happy when they enraged; you will be incensed more.

The uniformity of comments on the respective posts (Figures 5 and 6) illustrates how Ethiopianist and Oromo influencers’ narratives are reinforced through confirmation bias and in-group expectations, reflecting identity negotiation processes in close-knit discursive contexts. These dynamics foster polarization and negative out-group attitudes by creating an echo-chamber effect, wherein repetitive, insular exchanges amplify affirming narratives. Importantly, this effect extends beyond the immediate commenters to wider networks who may overhear or indirectly engage with the discourse (Tagg et al., 2017). Recognizing these dynamics is essential for understanding their broader sociopolitical implications.

Moreover, political scientists assert that cooperative meaning construction is a fundamental aspect of violent conflicts (Jabri, 1996). For instance, in Figure 5, expressions such as “prevent Oromia from ever gaining power,” and “a region prepared to carry out a genocide,” used by an influencer, construct narratives that promote violence against the opposing group. Similarly, Figure 6 illustrates how comments by individuals engaging with an Oromo influencer’s post – such as “monster corpse,” “your children will face your crime,” “it’s just the beginning [of humiliation and ‘teregnet’ – our turn to seize power],” “swallow it [the Oromo dominance],” and “you will be incensed more” – perpetuate hostility toward the “other,” in this case, the Amhara people. These examples highlight how a small but active group of SM advocates strategically shape political and ethnic narratives to advance their agendas. By presenting themselves as representatives of ethnopolitical groups, these influencers cultivate an ethos of credibility (Burke, 1969), fostering dialogic exchanges and ambient affiliations (Jones, 2023; Zappavigna, 2011) with followers, often without direct interaction. Such mediated engagements amplify sociopolitical tensions and risk escalating conflicts.

In the context of the Amhara region’s ongoing war between the central government and Fano, rhetorical clashes between Oromo and Amhara Facebook activists are especially pronounced. Hostile discourse on both sides entrenches animosities and deepens polarization. Although one Oromo activist in the study seeks to analyze Ethiopia’s politics beyond ethnic binaries, most posts analyzed here reinforce antagonism by attributing national crises to rival groups’ political elites. In doing so, influencers consolidate their ethos as arbiters of perception and action in ethnopolitical struggles, illustrating Bitzer’s (1968: 8) concept of rhetorical constraints, wherein discourse shapes collective decisions and conflict trajectories.

As reiterated, posts and comments are polarized along Amhara-Oromo lines, reflecting positive self-presentation and negative other-presentation (Van Dijk, 1998), further fueling intergroup animosity. The narratives advanced by political influencers and their followers are embedded in sociopolitical and cultural histories of ethnic belonging that compartmentalized communities into exclusive territories. This polarization sustains ethnicized echo chambers online and offline, reinforcing divisive rhetoric, extremist identity-based narratives, and violent antagonism. Consequently, zero-sum ethnopolitical discourses privilege group-specific “facts” and agendas over broader national cohesion.

Importantly, although some voices attempt to occupy a middle ground, the dynamics of social media visibility often coerce them into alignment with antagonistic rhetoric. Efforts to foster dialogue and mutual understanding are thus subsumed within polarized contestations, where SM visibility and influence are contingent upon taking sides.

Figure 7 translation and interpretations

Lidetu, long identified with the “Third Way” – a centrist alternative to Ethiopia’s polarized ethnopolitical order – now contends that regime change is the sole viable path forward. If not, according to him, “we [Ethiopians] will fall apart, killing each other as we are doing it now.”

Figure 7.

Ethiopian politician Lidetu Ayalew, known for his centrist political ideology, advocating for changing Abiy Ahmed’s government.

Partly, due to the virality of divisive content, even moderate figures such as Lidetu inadvertently perpetuating cycles of outrage, sustaining a vicious loop of rage amplification. This dynamic often exacerbates conflict by fostering insulated online public spheres that fuel offline polarization. Ethiopia’s digital media ecosystem reflects this trend: online audiences primarily consume content that reinforces preexisting beliefs, shaped by entrenched ethnopolitical divisions (Internews, 2023). Such dynamics foster echo chambers in which influencers mobilize followers less through substantive analysis than through emotional resonance, reinforcing confirmation bias (Nickerson, 1998) and generating “informational cascades” within closed communities (Quattrociocchi et al., 2016: 1–2). This explains why contentious political and conflict-related topics drive engagement on platforms like Facebook (Internews, 2023).

In sum, far from serving as arena of open deliberation, the observed trends in Ethiopia suggest that Facebook amplifies exclusionary narratives and rhetoric, reinforcing niche politics and shaping the broader political landscape (Akinwotu, 2021; Amnesty International, 2023; Chala, 2019; Roychowdhury and Gessesse, 2024; Stremlau, 2022). As Vaidhyanathan (2018) notes, such dynamics undermine foundational principles of deliberation. Discursive practices of unscrupulous Ethiopian SM influencers have deepened ethnic and political divides, complicating efforts to build inclusive democratic culture. However, these outcomes cannot be solely attributed to SM; they are rooted in Ethiopia’s historical ethnopolitical conflicts, which digital platforms such as Facebook – through algorithmic design and affordances – serve to reproduce and intensify.

Grievance salience

Grievance salience here refers to the rhetorical frameworks deployed to amplify real and perceived historical and contemporary suffering of a group. Such frameworks construct narratives of struggle that affirm one’s own history and realities while negating those of others. These narratives employ various discursive strategies, as elaborated in the following paragraphs.

Figure 8 translation and interpretations

The post discusses Tesfaye Gebreab, ⁸ an Ethiopian-born Eritrean writer, and the controversies surrounding his literary works. It recounts a conversation among friends, including the author, regarding the hostility Tesfaye faced – particularly accusations, slander, and denunciations from those identifying as the “sole heirs of Ethiopia” (namely, the Amhara) – in response to works such as YeBurka Zimta and Chaltu Ende Helen. An elderly Oromo man, overhearing the discussion, inquired about Tesfaye and the reasons for his condemnation. Upon hearing the explanation, the old man replied that Tesfaye had done nothing wrong if he was being criticized merely for addressing Oromo oppression. When pressed further, the man explained metaphorically that Tesfaye had only “touched the key to the big box” without opening it, suggesting that the deeper truths of Ethiopian history and Oromo repression remained unexposed. He concluded that true reckoning would occur only when “the big box” was opened in a language comprehensible to the dominant Amhara.

Figure 8.

A pro-Oromo influencer reflecting on Tesfaye Gebreab’s works and engaging in discussions with friends and other people.

This post (Figure 8), for instance, leverages humanization by adopting a conversational tone and featuring an elderly man – a figure traditionally associated with wisdom and credibility in Ethiopian culture – to highlight Oromo grievances. The use of storytelling, reinforced by cultural respect for elders, amplifies the salience and emotional impact of the narrative. By foregrounding the Oromo-Amhara divide, the post not only articulates Oromo discontent but also risks inflaming interethnic tensions. Further, the kairos of the message is evident. Metaphors such as “opening the big box” and “opening the box in a language comprehensible to them” suggest imminent confrontation, given the current political dominance of the Oromo. These metaphors play vital roles in heightening the perceived urgency of the conflict, aligning with rhetorical principles that prioritize salience in discourse (Fahnestock, 2011).

Conversely, Amhara SM influencers, political actors, and elites employ an alternative rhetorical strategy to make grievances salient. They invoke parallels with infamous ethnopolitical conflicts and mass atrocities, such as the Rwandan genocide and the breakup of Yugoslavia, to aggrandize the Amhara plight. Terms like “genocide,” “Nazism,” and “Hutuawi” (denoting “Hutu-like ethnic power” in Amharic) frequently characterize Oromo politics and the actions of the ethnic-Oromo Prime Minister toward the Amhara people. For instance, two influencers referenced “genocide” 16 times across 27 posts. Such comparisons serve not only to highlight the instability in the region but also to indict the leadership. Facebook’s algorithm, which amplifies emotionally charged content (Vaidhyanathan, 2018), further intensifies these narratives, enabling exaggeration and hyperbole to mobilize audiences and incite action.

Amid widespread armed conflict over the past 4 years, drawing attention to the Amhara conflict requires distinguishing it from other crises, such as the war in Tigray. By framing the Amhara situation as analogous to globally recognized atrocities, influencers reinforce its prominence. This aligns with Buyse’s (2014) observation that violence deemed exceptionally threatening exacerbates group antagonisms, severing intergroup social ties. This phenomenon is evident in Ethiopia, where tensions between Amhara-Tigray and Amhara-Oromia persist. Rhetorically, the posts employ vivid and evocative language, such as “Hutuawi” and “Nazism,” to make the grievances of the Amhara people salient and outgroup persecutions ⁹ more conspicuous. As Fahnestock (2011) notes, such strategies endow specific elements with conceptual importance, overwhelming audiences with claims and evidence of persecution to achieve rhetorical potency.

Some posts (see Figure 9) also leverage analogies from Bible stories.

Figure 9.

Excerpt of a post using the story of Saul and Samuel as analogy to inspire the Amhara to fight for nationhood and identity.

The post employs a biblical analogy to frame resistance against the central government as a moral imperative for the Amhara. Drawing on the story of God urging Samuel to move beyond mourning Saul’s demise, the author urges the Amhara to reject passive lamentation, reclaim their land, and assert their rights. Invoking divine authority, the analogy positions active resistance as necessary for survival and presents Ethiopia’s fragmentation as irreversible. While advancing Amhara nationalism as the only viable path, the post simultaneously denounces Tigray and Oromo nationalisms as fascistic and as the causes of Ethiopia’s disintegration. This rhetorical strategy privileges Amhara grievances as uniquely legitimate, despite similar historical struggles experienced by others. The paradox underscores the contested nature of ethnopolitical narratives and the broader struggles for power shaping Ethiopia’s political landscape.

Biblical analogies are particularly potent in Amhara discourse due to the deeply entrenched Orthodox Christian heritage of the community. Such references resonate profoundly, reinforcing the notion that resistance against injustice aligns with divine principles. By intertwining religious rhetoric with calls for political action, the post not only amplifies the salience of Amhara’s struggles but also legitimizes resistance as a sacred duty within the ethnopolitical landscape.

As reiterated, while ethnopolitical polarization in Ethiopia long predates SM, platforms such as Facebook have intensified existing divides. The circulation of emotionally charged messages that highlight ethnic grievances does not necessarily create new conflicts but entrenches pre-existing schisms. Such rhetoric, designed to evoke frustration and resentment, reinforces perceptions of irreconcilable difference (Schirch, 2020). As Buyse (2014) observes, ethnic identity becomes particularly salient under real or perceived threats of violence, rendering it a powerful mobilizing force. Accordingly, SM discourse may not directly incite violence but can weaponize latent tensions, demonstrating how Ethiopia’s political landscape remains acutely vulnerable to digital manipulation.

Epistemic manipulation

For this research, epistemic violence/manipulation denotes the distortion or strategic misuse of information to shape perceptions of reality. Within Ethiopia’s online ethnopolitical activism, dominant actors – often self-identified political elites claiming to represent ethnic groups – employ falsehoods, disinformation, and contradictory narratives to steer online discourse and advance their agendas. Such practices constitute epistemic violence by producing misleading knowledge and undermining the epistemic agency of (ethnic) groups (Pérez, 2019). Consequently, facts are routinely misrepresented, obscuring the distinction between truth and propaganda. Ethiopian Facebook discourse, in particular, is saturated with ethnic slurs, fabrications, and hate speech designed to delegitimize rivals. For instance, terms like neftegna (“musketeer”) and Fano are frequently mobilized to portray the Amhara as coercive unitarists, thereby reinforcing divisive and exclusionary narratives. Yet, historically, neftegna was also associated with maintaining order during Menelik’s Southern consolidation (Bitew, 2024), while Fano was linked to anti-colonial resistance.

The following two posts – one by an Amhara influencer and the other by a pro-government activist – illustrate the contested struggle to frame narratives about the ongoing war in the Amhara region. These posts reflect the broader struggle to control interpretations of the conflict, offering contrasting accounts of the realities on the ground. Such instances underscore the challenges of discerning objective truth amid a proliferation of conflicting narratives. The first (Figure 10) asserts that “only 40% of the Prosperity Party ¹⁰ infantry remains” and predicts the government’s inability to sustain fighting beyond a month. In contrast, the second (Figure 11), by a pro-government activist, portrays former Amhara fighters, now in reformation training, as pledging to correct their “mistake” of siding with Fano. These divergent accounts exemplify how digital platforms function as arenas of narrative contestation, where conflicting representations of the same events complicate efforts to establish an objective understanding of the conflict.

Figure 10.

A post by an Amhara influencer alleging that only 40% of the Ethiopian infantry remains.

Figure 11.

A pro-government activist claims former Amhara forces who accepted the peace call are undergoing reformation training.

Posts like these (Figures 10 and 11) illustrate two interrelated dynamics of epistemic manipulation and violence. First, the dissemination of misleading information enables social media influencers to assert intellectual dominance, undermining citizens’ epistemic agency to interpret and share their lived realities. This reflects what Pérez (2019) describes as a division of intellectual labor, in which influencers act as gatekeepers of knowledge, relegating the public to passive recipients of curated narratives. Such epistemic domination marginalizes alternative voices, obscures facts, and positions influencers/activists as ideological architects who mobilize audiences toward particular ends.

Second, those posts highlight the entrenched refusal of political actors to accommodate competing perspectives, thereby deepening ideological polarization. Narratives are constructed and weaponized to perpetuate epistemic violence, especially during periods of conflict like the Northern Ethiopia War and the ongoing unrest in the Amhara region. Warring parties instrumentalize facts and knowledge to dominate the narrative and delegitimize the “other.” Detached from objective realities, these narratives seek to overwrite opposing histories with emotionally charged rhetoric and discourse designed to evoke fear and sadness. Influencers thus rely on pathos-driven strategies to inscribe their ideological positions into the affective lives of their followers.

These observations align with findings from studies on various SM platforms. For instance, Pohjonen (2022) highlights X’s role as a battleground for epistemic proxy wars during the Tigray War, where warring sides offered conflicting accounts of events to assert narrative dominance. Such practices constrain audiences’ ability to make informed choices, as influencers manipulate information to serve their interests. Misrepresentation of facts and opaque communication – such as claims based on unverifiable “internal sources” – are standard tactics, further muddying the waters of public discourse.

The posts by both Oromo and Amhara influencers not only engage in epistemic manipulation through the distortion of facts but also manipulate perceptions of identity. This manipulation refers to the deliberate alteration or control of understanding about the identities of others, particularly concerning the concepts of neftegna and Oromummaa. In this context, epistemic manipulation involves the intentional distortion of knowledge, narratives, and perceptions regarding these notions and their connections to Amhara and Oromo identities, respectively. For instance, despite Oromummaa refers to Oromo culture and nationalism (for Oromo political elites), the following posts show that it is often depicted distortedly as a political apparatus of the regime.

The fascist Oromummaa government soldiers

The Oromummaa barbaric regime

In recapping, it is often argued that SM platforms have empowered ordinary people to engage in political and cultural discourse, facilitating the expression, formation, and dissemination of ideas outside of traditional, often restrictive, official channels (Khosravinik, 2017). While this assertion holds some merit, particularly when considering the traditionally rigid control of discursive power by mainstream media, the situation in Ethiopia – at least within the scope of this study – reveals that the purported freedom of citizen participation is often constrained. Citizens’ access to information, forms of engagement, and ideological exposure are frequently manipulated by online influencers/activists aligned with specific ethnic or political groups. Therefore, the influence and manipulation by these influencers extend beyond the creation of false narratives. In the short term, epistemic violence can fuel misinformation, leading to intergroup conflicts. This occurs when the epistemic agency of ordinary people is marginalized within alternative public spheres, such as Facebook, enabling political actors to construct divisive groups based on ethnicity, religion, geography, and other sociopolitical factors. Over time, these dynamics risk producing significant ontological shifts in the country’s political and sociological landscape, contributing to the normalization of political unrest.

Conclusion

The analysis reveals that discursive practices on Facebook often exacerbate political polarization, ethnic divisions, and communal violence. Through varied rhetorical strategies – such as antagonistic relationship-building, reinforcing animosities, and fostering hostile sociopolitical climates – SM influencers manipulate the ethno-politics narratives to shape online and offline discourse. Narrative techniques and frames like affective polarization, epistemic warfare, and victimization enable these actors to assert control over sociopolitical discourses, intensifying tensions.

Misleading and exaggerated content on Facebook frequently constructs alternate sociopolitical realities. While elements of these narratives reflect Ethiopia’s volatile sociopolitical landscape, they are marred by false narratives/misinformation, distrust, and ethnic divisions, offering an incomplete representation of offline realities. Viewing Facebook as a discursive system (Albert and Salam, 2013), the platform is often used to emphasize destructive semiotic patterns – chauvinism, vehemence, and polarization – thereby reflecting and constructing realities that challenge the sociopolitical status quo. As sociopolitical and cultural identities are dynamic, Facebook’s discourses may not only influence offline debates but also destabilize conceptions of Ethiopian nationhood and identity (Brooks, 2022).

However, Ethiopia’s complex history – marked by authoritarianism, contentious multi-ethnic relations, and ethnified political structures – renders it simplistic to attribute online polarization solely to SM interactions. Discourses on Facebook must be contextualized within the country’s historical and sociopolitical dynamics. As Fuchs (2017) and Mare (2020) note, discourses emerge from broader societal, political, and cultural contexts. Local agents shape SM logic to align with indigenous political realities (Gagliardone and Stremlau, 2012; Tagg et al., 2017), making Facebook an extension of offline practices (Georgalou, 2017). The enduring legacies of ethnic federalism and tribalistic political structures in the country further amplify sectarian rhetoric online, where ethnicity becomes central to political mobilization (Kelecha, 2025). Thus, online discourses reflect, rather than exist apart from, Ethiopia’s offline sociopolitical complexities.

Supplemental Material

sj-pdf-1-dis-10.1177_14614456251374251 – Supplemental material for Social media discourses amidst ethnopolitical extremism and conflict: The case of Ethiopia

Supplemental material, sj-pdf-1-dis-10.1177_14614456251374251 for Social media discourses amidst ethnopolitical extremism and conflict: The case of Ethiopia by Kibrom Berhane Gessesse and Mulatu Alemayehu Moges in Discourse Studies

Supplemental Material

sj-pdf-2-dis-10.1177_14614456251374251 – Supplemental material for Social media discourses amidst ethnopolitical extremism and conflict: The case of Ethiopia

Supplemental material, sj-pdf-2-dis-10.1177_14614456251374251 for Social media discourses amidst ethnopolitical extremism and conflict: The case of Ethiopia by Kibrom Berhane Gessesse and Mulatu Alemayehu Moges in Discourse Studies