Interview: Affective-discursive ecologies of the far right (Ben Nangle)
This interview is based on Ben Nangle and Leen d’Haenens’s article ‘Affective-discursive ecologies of the far right’
Ben Nangle and Leen d’Haenens’s article ‘Affective-discursive ecologies of the far right: Multimodal softening, mobilisation and algorithmic proximity on Instagram’, New Media & Society, 2026: https://doi.org/10.1177/14614448261428632.
The interview was conducted by Georgios Samaras.
***
How do you distinguish between genuinely far-right content and content that is just ironic or edgy?
That’s a good question, and I think the answer requires both a theoretical and a methodological reflection
On the theoretical side, the distinction between “genuinely” far-right and “just edgy” or ironic content is often analytically undecidable at the level of individual items, but I’d argue that’s not a limitation so much as a finding in itself. The intended ideological function of any given post is largely impossible to establish: producers are frequently pseudonymous or anonymous, motivations are often mixed or opaque (e.g., financial incentives, engagement-seeking, ironic performance) and in many cases the content isn’t being produced by a sincere ideological actor at all, but by sockpuppets, bots, or coordinated inauthentic networks.
But more fundamentally, producer intent is arguably less analytically important than effect. What matters is that the content disseminates and normalises far-right messaging regardless of whether anyone behind it consciously intended that outcome. The ecology produces ideological effects that exceed and outlast the intentions of any individual actor within it, which is what makes it difficult to counter such messaging through approaches focused only on identifying and removing deliberate bad actors.
Building on work by Nagle, Petersen & Johansen, and many others, in the paper we contribute to existing arguments that irony and ambiguity are constitutive features of contemporary far-right communication. We try to develop this argument further, the importance of this ambiguity isn’t only visible at the level of individual content, but is structural. What we call the affective-discursive ecology describes how the context in which content circulates itself becomes ideologically productive, as semi-relevant, ostensibly or genuinely neutral, or even oppositional content gets drawn into the gravitational field of explicitly extremist material through algorithmic clustering, proximity, and affect, contributing to dissemination and normalisation without requiring any conscious ideological alignment from producers or audiences.
The theoretical move we tried to make is to synthesise several existing frameworks that each capture aspects of this dynamic without fully connecting them. Ahmed’s concept of affective economies gives us the foundational logic: emotions aren’t private states but circulating forces that acquire political value through repetition and attachment to signs and bodies. Collins’s affective communities and Papacharissi’s affective publics both describe how networked formations are bound together through shared feeling rather than shared belief, you don’t need to consciously endorse an ideology to be part of the affective community it generates.
Petersen and Johansen’s concept of hybridised prefatory extremism spaces captures something important too: how conspiracy cues, grievances, extremist narratives, and pop-cultural references intersect within loosely structured online arenas without fixed ideological anchoring, producing a kind of ambient or undefined extremism. We engage with that framework, but we also wanted to push against a tendency toward post-ideological framing that can, even unintentionally, euphemise what are in many cases clearly far-right, racist, and dangerous tendencies. The ambiguity and hybridity are real, but so is the reactionary turn visible across much of contemporary politics, and that turn has identifiable ideological content and real-world consequences. Treating online extremism as purely post-ideological or ambiguous risks obscuring the extent to which specific, elite-driven ideological projects are being normalised through these affective dynamics. We wanted the paper to be analytically precise about that without sacrificing the nuance that the ecological framing offers.
What we attempt to add, then, is an infrastructural and ecological dimension that the existing frameworks don’t fully develop: these affective circulations don’t happen in a neutral space but within algorithmically curated environments that actively cluster, amplify, and contextually politicise content. The ecology is the mechanism through which ambient familiarity with extremist ideas is produced and extended to audiences who may never consciously identify with them, and it is that mechanism, operating at the level of infrastructure rather than individual susceptibility or deliberate strategy, that we think needs to be at the centre of how we understand contemporary far-right normalisation, though this has aspect has certainly been described by others.
Again it’s near impossible to judge the deliberateness or conscious strategy behind much of these dynamics, but it’s immaterial, this dynamic operates largely independently of conscious strategy. Whether or not any individual actor intends these effects, the ecology produces them, and that’s what makes it so expansive in reach.
Methodologically, our response to this ambiguity was triangulation. In keeping with multimodal discourse analysis, no single indicator is sufficient: explicit markers like fascist symbols or recognisable narratives such as great replacement; intertextual context, what surrounds a post algorithmically, what community produces it, which accounts are likely to encounter it; and cross-modal cues, what a caption does to an image, what a soundtrack does to archival footage etc. For instance, a Hitler joke in isolation is ambiguous; the same joke appearing between Holocaust denial content and a remigration reel is considerably less so, regardless of the original intent of the joke. This is because meaning isn’t fixed in the content itself but imbued by the context in which it is encountered and interpreted, and that context is itself constructed through the interaction of platform logics, algorithmic incentive structures, and deliberate human input and guidance. Whether or not this construction is conscious is, again, somewhat beside the point: the effect is produced regardless. These dynamics reward strongly affective and controversial content while remaining beholden to financial and political pressures, and the result is an interpretive environment in which adjacency itself becomes a form of meaning-making. The ecology doesn’t just distribute content but actively produces the conditions under which that content acquires ideological significance.
Does Instagram mainly amplify, or does it actively shape the softening of the message?
I think it’s difficult to distinguish between the two aspects. The platform essentially does both, mediating and amplifying messaging, while also shaping the form that messaging takes and how it is interpreted. These mechanisms are intertwined rather than sequential.
Amplification is the more straightforward aspect: engagement-maximising recommendation rewards strongly affective content, which maps onto what far-right communicators already produce. But far-right actors also strategically adapt their content to fit platform logics, so the causal arrow runs in both directions. The platform shapes content more actively through its affordances: the Reels format, short, looping, audio-visual, template-based, structurally favours the aesthetic modes identified in the paper. Humour, wholesomeness, and nostalgia translate well into that format; explicit propaganda often less so. Creators adapt to the format, and the format selects for softened content. Moderation asymmetry compounds this: explicitly ideological content gets removed, softer affective content doesn’t, creating evolutionary pressure toward precisely the affective and stylistic modes we document. The opaque, black-box nature of social media algorithms makes definitive claims about prevalence or effectiveness difficult, but our observations suggest a dissemination advantage for far-right content that adopts a softer register, and, returning to the previous question, the combination of that soft content with more overt messaging likely plays an important role in the overall dynamic.
Framing this purely as amplification versus shaping, however, risks obscuring something more fundamental. Treating the platform as essentially neutral infrastructure that happens to have certain effects understates the role of ownership and governance. Decisions about what gets amplified or suppressed are not just algorithmic outputs but organisational outcomes reflecting executive authority, shareholder interests, strategic priorities, and the broader political environment in which platforms operate. When ownership or leadership changes, tolerance thresholds and enforcement regimes change with them. The far-right content we encounter on social media is not only a product of format affordances; it is also a product of business models that may make affectively engaging, borderline content economically and politically functional, and ownership structures will variably determine how much extremity a platform tolerates or promotes at any given moment.
The more precise, but still convoluted answer, then, is that Instagram shapes the form of harmful content through its affordances and moderation architecture, variably amplifies what results, and ownership and governance structures set the parameters within which both processes operate, parameters that are neither fixed nor politically neutral.
Which mode is most effective at pulling people in, and why?
I’d resist ranking the modes too definitively, because their real effectiveness likely comes from their combination and variety rather than any single register, and because different modes probably work differently depending on the audience. Less extreme or ambiguous content may function as a kind of warm-up, gradually desensitising audiences to more explicit material, material they will then encounter either through algorithmic recommendation driven by their own engagement, or through more direct funnelling: many humorous or ambiguous far-right posts include links to YouTube playlists, comment sections saturated with explicitly reactionary content, or Telegram groups, creating deliberate pathways from soft entry points into harder ideological spaces. The variety of affective registers (humour, wholesomeness, outrage, pride) also likely plays a role in preventing audience fatigue, enabling sustained consumption without the overwhelm that a diet of pure outrage or explicit propaganda would eventually produce. What the varied repertoire does collectively, then, is extend appeal across a wider public, and sustain engagement over time, in ways that any single mode could not achieve alone.
Another reason why it’s difficult to isolate any single mode as most effective is the inherently subjective nature of humour, and the fact that the emotional impact of any given content will vary considerably depending on the audience, their prior exposure, their social context, and their existing dispositions. What lands as transgressive and funny for one viewer reads as offensive and alienating to another. The same applies to nostalgia, pride, or outrage, these are not universally experienced affects, and their persuasive force is contingent rather than guaranteed. It’s this variability that makes the combination so important: a varied repertoire hedges against any single mode failing to land, ensuring that the ecology as a whole retains its pull across heterogeneous audiences even where individual elements don’t.
From a psychological perspective, part of the explanation for why this works likely lies in dual-process thinking: when people encounter humour or emotionally resonant content, they’re typically not engaging deliberative reasoning, they’re reacting, laughing, sharing, without pausing to critically evaluate what they’re actually amplifying. This is a cognitive dynamic these modes of content exploit, and it’s part of why we’re interested in developing interventions that interact with harmful content at that same affective level rather than relying on rational counter-argument.
With that said, I would think that humour probably has the lowest barrier to entry of the three, you don’t need to identify with any ideology to find something funny or share it, and it operates through plausible deniability in a way the other modes don’t. The meta-ironic content is striking, the self-referential “I’ve been radicalised” jokes, content that describes itself as propaganda. That reflexive layer is disarming precisely because it mimics critical distance while actually performing ideological participation, and it pre-empts the obvious critique before you can make it. The nostalgic-wholesome mode probably works better for retention than initial recruitment, requiring some prior exposure to land. The inspirational-glorifying mode is the most nakedly ideological and likely least effective as a first point of contact, but potentially the most radicalising for those already inside the ecology.
I should be clear that we haven’t done audience reception research, so these are theoretical mechanisms rather than empirically demonstrated effects. But taken together they describe an affective-discursive ecology that is structurally very conducive to getting people to entertain and potentially adopt far-right ideas, not through overt persuasion, but through ambient familiarity, emotional resonance, and cumulative exposure across a varied affective landscape.
Any concluding statement?
I’m aware I haven’t given particularly straightforward answers to any of these questions, and may have raised more questions than I’ve resolved, but I think that’s actually reflects the complexity of the affective-discursive ecology itself. It is produced through intertwined mechanisms and forces, some deliberate, some incidental, some structural, and it resists the kind of clean causal account that a more straightforward answer would imply. What the ecological framing tries to capture is that the outcome, the amplified dissemination of highly evolved and persuasive reactionary propaganda, exceeds any single explanation, any single actor, and any single platform dynamic. Understanding it requires holding multiple levels of analysis simultaneously.
***
Ben Nangle is a PhD researcher at KU Leuven, focusing on the far right and migration issues, with a particular interest in digital communication, propaganda, and social media discourses. He is currently involved in investigating far-right persuasion strategies and developing countermeasures and interventions to reduce youth susceptibility as part of the DEMINE project. He recently published a book with Palgrave Macmillan on the construction of hostility towards migrants during the Lithuanian–Belarusian border crisis. Outside of research, he has significant experience as a humanitarian worker in the refugee and asylum sector, as well as political organising experience.
Dr Georgios Samaras is a Lecturer (Assistant Professor) in Public Policy at the School for Government and Policy Institute, King’s College London. Previously, he held a Lectureship in Political Economy within the Department of Political Economy during the 2022/23 academic year. He completed both his MA and PhD in Political Science at King’s College London, where he examined the Greek far-right party Golden Dawn. His research encompasses far-right extremism, political communication, and policy responses to illiberal political practices, while also exploring the dynamics of European and US politics and the global rise of authoritarianism. His work has been published in several reputable academic journals, including Distinktion, Empirical Theology, Visual Studies, and Frontiers in Political Science. His forthcoming monograph, Who Rules the Land of Denial?, is due for publication in August 2025 by Palgrave Macmillan and Springer Nature, with a Greek translation forthcoming.
Ben and Georgios are members of the RPRN
Creative Commons Attribution-NonCommercial 4.0