Affect, Certainty, and Mis/Disinformation

Afterword

GANAELE LANGLOIS

RORY SHARP

York University, Canada

This afterword examines the centrality of affect as a key lever and critical site of intervention in mediated misinformation and disinformation. It starts by centering affect as a preconscious guide for navigating the world, explaining how consciousness-level interventions, such as fact-checking, ignore key psycho-physical-social dynamics through which affective reactions lead to personal and collective certainties. It then explores how mediation enables specific experiences of affect that serve political means and shows how this very question of affective experiences has to become central to countering misinformation and disinformation.

Keywords: misinformation, disinformation, authenticity, affect, emotions, populism, manipulation, social media

Ganaele Langlois: [email protected]

Rory Sharp: [email protected]

Date submitted: 2024-05-02

Let us start with a well-known analogy from Indian philosophy: the snake and the rope. A man enters a dark room, notices a coiled shape on the ground, and immediately flees in fear, screaming that there is a snake in the room. Upon further inspection, however, the snake is revealed to be a rope. Traditionally, this analogy illustrates the fallibility of human perception and the tendency to reify false perceptions of reality. In this sense, the parable remains relevant today for understanding mis- and disinformation . Countless people worldwide now mistake ropes for snakes by believing fabricated stories that deny scientific consensus (e.g., climate change, vaccine benefits), further ostracize already marginalized groups (e.g., immigrants, visible minorities, LGBTQ2S+), and normalize extreme antidemocratic political views (Benkler, Faris, & Roberts, 2018). While the idea of one absolute truth is a fallacy (Marres, 2018), mis/disinformation belongs to a longstanding tradition of manipulation and propaganda. These traditions are mobilized by specific political and economic interests whose strategies for gaining power are rooted in breaking down the status quo and essentializing gender, social, environmental, economic, and racial inequalities (Freelon & Wells, 2020). The stories carried through mis/disinformation are sometimes based on complete lies (e.g., a U.S. Democrat pedophile ring ran through a pizza shop), but more often than not, they require sustained and elaborate effort by specialized public relations (see, for instance, Lukito et al., 2020) to ensure the rope appears like an authentic snake. Such stories are not new, especially in politics, and have been used by nonelected and elected officials many times before (recall the weapons of mass destruction that were used to justify the invasion of Iraq) (Altheide & Grimes, 2005). However, the phenomenon has grown to a much larger scale through digital networked media (Bennet & Livingstone, 2018). What makes mis/disinformation distinctive today is that it relies on corporate social media logics that mine users’ attention and emotions and market them to specific interest groups, including political ones (Chadwick, 2019; Munger, 2020). The twin logics of social media’s user data capture and recommendation systems, combined with long-standing populist strategies to enflame segments of the population and deepen social division and inequalities (Berry & Sobieraj, 2016), are at the core of what we understand as contemporary mis/disinformation (Zelenlauskaite, 2022).

Departing from the traditional conclusion of the parable, which blames the man who enters the room for his faulty perception, we instead imagine the story taking place today. Which malicious deceiver (Jucan, 2023) left the rope there, and why? Was it planted by a TikTok prankster seeking followers or by somebody wanting to kickstart a larger vendetta against snakes? Was the rope put there to teach our man a lesson, to make a fool of him, to encourage him to spread false information about a snake invasion, or to use him to spread fear in the neighborhood? All these scenarios have one thing in common: provoke fear in our anonymous protagonist/victim. This raises a second question: Can the man’s reaction truly be blamed? His instinctual response—to jump back in fear—would have been most appropriate if it were indeed a venomous snake in the room. His quick reaction, in that case, would have been applauded. Rather, let us imagine the aftermath of the event. If we were to further embellish the story, we imagine the man refusing to acknowledge that it was indeed a rope: “I swear I saw it move! Somebody must have planted the rope after I ran out to make me look bad! There’s a bigger conspiracy at play!” It is incredibly hard to admit to oneself, let alone publicly to others, that one was indeed delusional, even if only for a split second. This—our profound reliance on our preconscious responses to survive and navigate immediate, urgent situations—is key to understanding the difficulty of addressing mis/disinformation. If we recognize that mis/disinformation fundamentally works by triggering affective responses to information, then combating it becomes much trickier than setting the record straight through fact-checking (Graves, 2016), calling in more witnesses, and similar measures.

As our contributing authors and many others have pointed out, fact-checking alone does not solve mis/disinformation (Andersen & Soe, 2020; and from this Special Section, Akhavi & Sharp; Alloing, Fournier, Reid, Germain, Turbide, & Yates; Sylvester, Hughes, Canute, Ruiz-Soler, & Chun; O’Meara, Morales, & Hodson). At times, it can worsen the phenomenon, as fact-checked people cling all the more ferociously to their perception of the world (Walter & Tukachinsky, 2020). As we explore in this afterword, the problem lies in the relationship between affect—our response to informational input about the world—and how it can morph into a certainty that authenticates what reality really is. Affective certainty is one of many modes of authentication that establishes beliefs beyond doubt. It may manifest as performative utterances or competing citational practices, blurring the boundaries between gut reactions and socially conditioned responses (in this Special Section Nguyễn, Le, & Moran; Sylvester, Hughes, Canute, Ruiz-Soler, & Chun). Certainty, especially in its moral and spiritual aspects (consider, for instance, martyrs), is a potent means of challenging established social truths. One might recall Jodi Dean’s (2009) exploration of certainty among conspiracy theorists as a response to the collective trauma and loss of 9-11, or the rhetorical “transgression of conventions” during and after the COVID-19 pandemic, as explored by Sylvester, Hughes, Canute, Ruiz-Soler, and Chun in this Special Section. Certainty is an affective enclosure resistant to corrective intervention, and while certainty is highly personal, it is also highly mediated and socially interconnected. In the context of mis/disinformation, how do the mechanisms of social media, with their personalization algorithms, data infrastructures, and platform logics, contribute to making people certain of the existence of snakes when they are factually surrounded by ropes? Across this special issue, our contributing authors have each addressed this question in their own way, exploring how these affordances manipulate space and time to amplify certain narratives (Weltevrede & Burton; Rogers & Koronska; Noordenbos & Tuters) and enable unique performances of authenticity (Akhavi & Sharp; Bimo; Nguyễn, Le, & Moran; O’Meara, Morales, & Hodson). Such processes not only involve the crafting of believable stories but also require a circuitry that triggers, spreads, cultivates, and modulates affects. This circuitry articulates the relationship between affect and experience, transitioning from a preconscious state to certainty about one’s experience of reality.

In this afterword, we focus on the circuitry of negative affects like fear, anger, sadness, and hopelessness that are in the shadows of this Special Section’s focus on mis/disinformation. We recognize the key importance of social media for affective exploitation, which in turn provokes and capitalizes on all kinds of affect (Sampson, 2020), and we follow many others in highlighting that affect is both a site of capture and of transformation and intervention (Ahmed, 2004; Blackman, 2019; Ngai, 2015). Because they are preindividual and entwined with preconscious vital instincts, affective responses are challenging to counteract. Looking at affect as mis/disinformation requires acknowledging that while affect is at the core of mis/disinformation, it is not the culprit or a problem that a serious dose of objective rationalization could neutralize. Rather, what is critical is to understand and unravel the industry that automates the viral affectivity on which mis/disinformation campaigns rely. And here, perhaps, lies a ray of hope: Dealing with affect is messy, and the multiple actors involved in online affect prodding and management create a chaotic situation, one where many of us find ourselves experiencing affective exhaustion. Consider, for instance, the assemblages of affect explored by Rogers and Koronska; Noordenbos and Tuters; and Alloing et al. in this issue, or the distended temporalties outlined by both Bimo and Akhavi and Sharp. Of importance, then, in considering how to address mis/disinformation—apart from much needed, and perhaps belated policy and legislative responses—is understanding the paucity of affective management processes provided to users. This afterword starts with considerations around the politics of affect to focus specifically on how it has been shaped as mis/disinformation via media infrastructures. Arguing for the messiness of how affect is treated within the technocapitalist context of social media, this article turns to potential sites of intervention where digital methods might play a role, particularly in the design of affect exploration tools, while reflecting on the key contributions of this special issue.

The Politics of Affect, or What Does Affect Do?

Contemporary mis/disinformation models, such as those emerging from Cambridge Analytica, propose a quasi-hypodermic needle model of manipulation: provoke the right affect, cultivate it over a period of time, and people will act in a certain way, for instance, in their voting. One of Cambridge Analytica’s key claims to fame was crafting messages that targeted affective reactions specific to different user groups. For example, in promoting gun rights, “neurotic” users were targeted with fear-inducing messages, while those more prone to nationalistic pride received messages linking guns to family values and longstanding American ideals of freedom and tradition. While the geopolitical operation of mis/disinformation highlights industrial-scale manipulation, it is worth considering the complexity of affect and how it does not quite fit into this simple causality model. Approaching affect can be daunting: There are many theoretical camps and debates (Wetherell, 2015) under the umbrella of affect theory, and the concept itself is notoriously slippery (Gregg & Seigworth, 2010). The theoretical framework presented here to situate affect within and beyond mis/disinformation processes is agnostic, borrowing from different affect theories to understand how affect is mobilized, shaped, managed, and experienced through both contemporary politics and corporate media and social media. In other words, we are more concerned with conceptual tool-building (Deleuze & Guattari, 1996) to critically explore affect in relation to mis/disinformation than with adhering to specific theoretical dogmas or dividing our contributors into mutually exclusive tendencies. Though sensitive to the limitation, repeated early and often in most primers on affect theory, that “[n]o one emotional state can encompass all the depth and breadth of our experiencing of experiencing—all the ways our experience redoubles itself” (Massumi, 2002, p. 5), we view that limitation as a productive constraint. Thresholds between affects—between affect, emotion, and sentiment more broadly, where one enhances or negates the other—become sites for more fulsome interventions.

The question is, how does it affect the move from physiological to political? Here, we look at affect as the result of relational interactions. We borrow from the Spinozist strand of affect theory, particularly Gilles Deleuze (2022) and Frédéric Lordon (2014), which highlights how affect (to affect and to be affected) is an expression of power relations. Power here should be understood as both power over—as control—and as power as potential, as a creative force. How, in a specific situation, I am affected by others (humans or otherwise) and how I affect others are the key questions here. The Spinozist understanding of affect focuses on action and capacity: I am affected by something that acts on me and either fosters or inhibits my capacity for action. Affect then becomes a question of the ethics of power relations. What is important here is examining which sets of actions an affective interaction provokes. The question is how affect shapes one’s drive to act—what Spinoza names conatus and what Tomkins, perhaps following him, calls “motivation”: the drive to act in the world in ways that enhance and transform one’s body and mind, unfolding one’s being to the world and allowing new capacities for action to emerge. Notably, from this perspective, the social context defines the power of affect: fear, for example, may be mobilized to preserve (e.g., “beware of snakes”) or diminish life (e.g., “I will kill you”).

Two other aspects of affect from the Spinozist tradition are important to highlight as well: First, affect is not simply about impact on bodies but also on individual and collective ideas, and second, ideas themselves have affective impacts, which are modulated by the social-biographical history of the person being affected. Frédéric Lordon (2016) gives the example of the idea of a dog, which can create opposite affective responses in people depending on whether they were viciously bitten by a dog in childhood or, conversely, had a dog as a trusted companion (p. 21). One can thus be affected by what is immaterial and transmitted through language and media. Affect, therefore, works across distance, transversing time and space. We also know that affect can be a shared experience, as sociobiographical histories often follow similar, shared trajectories. Revolutions and, on the opposite spectrum, war and terror create shared affective reactions and landscapes for everyone involved. Similarly, individual experiences can foster how one refuses, neutralizes, or nullifies an affect: The capacity to resist is found through layers and folds of both mundane and key life experiences, enabling a degree of refusal (Lordon, 2016).

Concerning interpersonal and political relations, the ethics of affects involves choosing what kind of power to exert on an Other: Power to let the other grow into existing and new capacities and to expand their agencies and power to affect in turn, or, conversely, power to diminish and destroy their capacities for action. Thus, we might think of affective relations in polarized terms. To a large extent, we could say that politics is often about exerting control over others, and that relentless, demeaning attacks against a political opponent exemplify affective power that stunts and denies the capacities of an Other—such as the capacity to be recognized as worthy of being heard. However, it is important to attend to another murkier type of affective relation: the ability to activate and cultivate specific capacities in an Other. This is akin to growing a bonsai tree, which involves careful pruning and controlled restriction to maintain the tree’s growth, but in a stunted and controlled way. Similarly, mis/disinformation works by provoking and directing specific affective responses—anger and fear being common examples—such as when a marginalized group is defined as an attacker posing real danger. These emotions are politically useful because they not only undermine the capacity of the marginalized group to be seen as equal but also intensify the manipulated group’s need to act in a certain way, for example, to defend themselves. Anger and fear can also be incredibly contagious; one of our authors recalls the mass panic in a popular area of Paris, France, provoked by the sound of a blowing fuse in a bar a few days after the 2015 terrorist attacks. What happened was a conjunction of an original affect—terror from the attacks—that then primed an intense affective response to loud sound, which quickly propagated itself. Negative affects such as terror are, in this situation, intensively productive and politically impactful: They create large-scale actions and enshrine terror as a dominant, shared affect. It is important to note that the panicked people mentioned here are acting out of self-preservation. They were not denied the possibility of action, but rather the triggering of a specific affect narrowed down the range of their possible actions. In Spinozist parlance, their conatus—that is, their drive to self-fulfillment—was deeply shaped in a specific direction.

Mediating Affect

Encountering an affecting event or object is not enough to consider the relationships between affect and mis/disinformation, which oftentimes bypass direct encounter altogether and rely on the media to provide a certainty of something that might neither be true nor exist. Thus, while the previous section explained how ropes can be confused with snakes, this section is concerned with how our man could further swear, with the utmost certainty, that he saw a snake, not a rope. As Bruno Latour (2004) showed, for a fact to be accepted or authenticated as fact, it must be experienced. There is, of course, an affective component to this experience: I know that water boils at 100°C at sea level, not necessarily because I have actually experimented this with a thermometer, but because I have either burned myself with water before (a direct experience), have seen others hurt (a witnessing experience), or have been warned countless times about boiling water by trusted people in positions of authority (a mediated experience), so I exercise some level of caution when I strain my pasta. As Sylvester et al.’s article from this Special Section shows, certitude requires “competing modes of authentication” with diverse citational strategies to affirm information as authentic. The problem is that some facts are too complex, too large in scope, or too insidious to be fully experienced directly—this is the case with climate change or long histories of racial exclusion—so they tend to be particularly instrumentalized within mis/disinformation networks. Furthermore, such complex facts, because they pose existential threats at a scope and complexity unheard of before, foster uncertainty. Typically, we tend to rely on mediated experience for complex facts. It is not surprising, then, that mistrust in governing institutions and political actors usually creates a breakdown of mediated experience that can lead many to deny specific facts (Bennet & Livingstone, 2018; in this Special Section Alloing et al.; Nguyễn, Le, & Moran; O’Meara, Moreales, & Hodson; Rogers & Koronska). Similarly, media industries, including groups that purport to be in the business of news like Fox News, privilege political bias and disinformation to the detriment of fact-checking (Gil de Zúñiga, Scheffauer, & Zhang, 2023). Doubt adds itself to existential uncertainty and unease within this specific milieu that mis/disinformation operates. How, then, do mediated experiences shape and direct affective reactions, and how are these affective reactions stabilized into durable certainties, such as a political position, an identity, and so on, within this maelstrom of doubt and uncertainty?

Here, it is key to rethink the relationship between affect, feelings, and emotions. Typically, we posit an arc from affect (preindividual) to feelings (individual) to emotions (social). In the rope example, the affect was fear, which then became a feeling (“I’m scared”) and was subsequently expressed as a recognizable emotion (expressing one’s feeling in ways that are socially recognized by others, who can then share the experience). Because they are described as primordial reactions, affects often feel like the origin of emotions, but this is actually not the case: Affect can be triggered by propagating and integrating emotions in the form of stories, discourses, representations, and so on. This is important for our argument: The relationship between affect, feelings, and emotions is circuitous, enabling multiple points of strategic manipulation where certainties arise. A famous instance of this is the Monosodium glutamate (MSG) controversy, which has been ongoing since the late 1960s in North America, with the publication of an article alleging that MSG had detrimental health effects (Sand, 2005). People started complaining of various ailments, including headaches, body pain, and overall feelings of sluggishness, after eating at Chinese restaurants. This has been debunked over the years: Large doses of MSG are present in many popular savoury packaged snacks and other common foodstuffs, and studies have proved that most people who complain about MSG did not react to it when they were unaware that it was present in their food. This intense ongoing shared affective reaction is not a physical reaction to an allergen, but the bodily preconscious integration of anti-Asian racism as latent fear, particularly as it propagates itself through countless fabricated stories about Chinese food. That is to say, these people complaining about MSG were not lying; they felt something, but what they felt was an affective reaction to an object (Chinese food) long associated with anti-Asian racism. Here, several racist discourses, fantasies, and conspiracies recede under the level of consciousness into the physiological and the affective.^[1] The potency of this move—to translate from representational and linguistic to embodied sensation—shows how affect can make racism so visceral that it becomes a felt certainty. Here, a latent affect is undeniably provoked to the surface by internalizing broader socialized emotions via mediated discourses.

Acknowledging that the relationship between affect and emotions is multidirectional opens the door to imagining diagrams and maps to trace their recursive and circular movements. The structuring of collective emotions can shape a primary affect and direct it toward shared certainties that can then be expressed as collective actions and reactions. A key example here is from Eva Illouz’s (2023) Emotional Life of Populism (published before the October Hamas attacks) concerning soldiers and their capacity to kill without feeling guilt or remorse about it. Illouz explores the politics of emotions in Israel and highlights how the articulation between affect and the manufacturing and structuring of collective emotions serves right-wing ideology. She highlights two sides of emotion politics: First, they are based on shared social, cultural, and economic experiences, which are then, in a second stage, given form by diverse groups through discursive and institutional practices. Here, such “structures of emotions” define an object to channel affect and activate specific capacities of action. Illouz (2023) cites compulsory military service in Israel as an example, where the natural affect of terror in the face of potential death fuels shared emotions of fear and rage for all enlisted. Such fear and rage are fed and shaped through populist discourses about an “Arab” enemy that seeks complete annihilation, so much so that when faced with a Palestinian individual, military trainees do not see a human being, but an instance of a larger, omnipresent, dangerous enemy that has to be terminated. In other words, we see a movement from original affects (terror, rage) that is canalized and fed through emotional discourses on the Arab enemy, which then enables and sanctions an ideo-affective reaction, in this case, the shared collective certainty of killing as morally necessary, if not good.

While Illouz points out the mediated discursive and institutional structures of emotions that channel affect and shape affective reactions and the resulting certainties that emerge from them, we need to further expand toward the nonlinguistic, nonrepresentational aspects of mediation. Crucially, there are two key aspects of media as affective structures here: their ritualistic capacities, which create shared, repeated affective and emotional experiences, and their formalization of affect through sensory forms, both of which serve to give a sense of certainty about one’s place in society and in the world. Benedict Anderson’s (2006) Imagined Communities—a classic example of how print media, especially newspapers, helped build shared nationalism through rituals—explains both aspects. National belonging requires that individuals identify with countless others they do not know and who are not part of their immediate community. The ritual of reading the daily newspaper played a key role in the emergence of the nation as a shared emotional and affective object. Not only does my awareness of reading content about the same important events of the day connect me to an imagined national community, but the knowledge that many others share the same ritual of physically reading the paper in the morning also creates a sense of belonging among us, despite our anonymity to each other. The industrially mass-produced newspaper is about how mental actions (reading, thinking) and physical gestures (picking up the paper, sitting down, turning the page) come together to form atomized yet shared habits with shared affects and emotions. We see the same trends in the national sentiments evoked by diasporic media (Nguyễn, Le, & Moran) and chauvinistic rhetorics of Asian inclusion/exclusion (Sylvester et al.). In other words, media as rituals articulate together individual and collective affective and emotional experiences: They make abstract ideas, like those of the nation, tangible and physical, giving them certainty. For contemporary media consumption, considering the role played by the loss of specific media rituals—through the switch from analog to digital—in fostering uncertainty, fear, and the collapse of shared affective and emotional ideas helps illuminate how mis/disinformation overtakes rituals and content. As our contributors show, whether through the rise of “appligandists” greasing the chutes of mis/disinformation networks (Rogers & Koronska), “multidirectional” propaganda (Noordenbos & Tuters), or the ease with which astroturfing is accomplished online (Alloing et al.), the “grammars of action” that have accompanied the “appification” of digital media are uniquely susceptible to manipulation (Weltevrede & Burton). Today, this convergence of the mental and physical is evident in online communities that promote specific types of subjectivity, reaching a toxic level in the “manosphere,” where misogyny and mis/disinformation go hand in hand. Jordan Peterson’s paternalistic exhortation to “clean your room” combined with discourses of male heteronormative superiority exemplifies this incentive to link innocuous habits with specific ideas to create shared rituals and collective imaginaries. At the other extreme, groups like the Proud Boys or QAnon adopt extreme rituals (from foregoing masturbation and submitting to group beatings as initiation rituals to obsessively interpreting textually dense clues) that aim toward bonding and camaraderie (McCann Ramirez, 2023), functioning as “affective glue rather than a cause of disillusionment” (Bimo in this Special Section). Common to these is a call to adopt mental and physical practices: patterns of thinking akin to a self-hypnosis and delusion aimed at embodying emotional structures at the level of affect and eventually becoming one’s authentic self.

Social Media, Affect, and Certainty

Social media, of course, inscribe themselves directly in the history of mediated affect: They advertise on behalf of specific interests that snakes are everywhere, articulate individual and collective affective experiences through viral repetitions of this claim, and provide addictive, ritualistic routines of constant scrolling and doom-scrolling (Lovink, 2019; Sampson, 2020). What distinguishes social media is their capacity to create affective refrains. As multiple authors in this issue illustrate (Alloing et al.; Rogers & Koronska), corporate social media platforms were never just spaces for self-expression: They function like fishnets for social interactions, capturing and enclosing them as marketable commodities. The primary commodity is personal and relational data, with platforms—and their operators—promising sociobiographical data that reveal and authenticate the affect to be triggered. While this operationalization of affect within a smooth machinery of recommendation has to be taken with some skepticism, it nevertheless indicates a pattern of how affect is conceptualized and mobilized within the corporate social media model.

Second, social media platforms’ algorithmic logics, combined with content creators’ understanding of affect as key to the attention economy, provide a continuity of affective prodding for users. Affect is one lever of the online attention economy. Niche content producers focus on creating endlessly similar types of content that repeat the same affect, from rants to provoke anger and despair (i.e., “doomscrolling”) to Autonomous Sensory Meridian Response (ASMR; Smith & Snider, 2019). To increase followers and views, professional content creation on platforms relies less on radically new content than on creating a recognizable, affective experience through formal repetition, similar to the participatory structure Bimo observes in “Q drops”—whether high-impact (e.g., an unsavoury Pepe meme) or low-stakes (e.g., cooking or kitten videos). Recommendation algorithms will then pick up users’ attentional patterns to this specific type of affect-laden content and repeat the exact affective experience: Endless scrolling is the repetition of the same affective prodding, constituting a continuum of experience. With affect related to content with shock value, algorithms’ demands on content producers will typically encourage the creation of content that intensifies such affective reactions. Hence, for instance, the spate of increasingly violent and cruel prank videos on YouTube that eventually led to child abuse charges (Hern, 2018).

Digital networks make the spread of mis/disinformation viral and unwieldy, but the role social media plays in mis/disinformation goes beyond the transmission and recommendation of content. Provoking affective experiences is part of the political economy of social media platforms. Byung-Chul Han (2017) relates the temporality of digital media to “the pre-reflexive, half-conscious, physio-instinctual level of action that escapes full awareness” to show how “neoliberal psychopolitics seize on emotion in order to influence actions on this pre-reflexive level” (p. 48; emphasis in original). The drive to maximize extractable data, both through binge-like scrolling sessions and a constant push to respond to notifications, places incredible value on platforms that can condition their users to treat such use as a persistent, unquestioned habit. As Karppi’s (2018) study of Facebook affirms, cultivating affective bonds is key to consolidating both a constant user base and steady corporate profits. Social media platforms’ logic is to privilege affective continuity over the accuracy of content. The corporate social media ecosystem is only very reluctantly addressing mis/disinformation through content moderation while continuing to privilege the right of any content that does not infringe existing laws to be transmitted, even if factually inaccurate or inflammatory. Some fringe and mainstream platforms (e.g., Elon Musk’s X) welcome mis/disinformation content for neoliberal and libertarian ideological and political reasons. This reluctance enables a continuity between information and mis/disinformation, as users continue to consume the same affective forms, repeatedly articulating their gut reactions as proof of the information’s authenticity (see Bimo in this Special Section). Here, mis/disinformation infiltrates recommendation flows by mimicking existing genres and mining the logics of recommendation that seek to provoke a continuous affective experience to maintain a user’s attention.

Overall, mis/disinformation on corporate social media platforms relies on orchestrating affective resonances—provoking and repeating over and over again that gut feeling in users that is so important to navigating one’s environment by preconsciously developing a sense of certainty. In so doing, affect is “mobilized, fomented, unleashed” as much as it is measured, tracked, and rationalized (Bösel & Wiemer, 2020, p. 8). It is cultivated by insincere organizers and turned back to modulate their supporters’ sincerity (Alloing et al., in this Special Section). It is here that the “certainty” of mis/disinformation takes over truth: the gut feeling, the affective eruption, tells us that whatever provoked this reaction is worth pursuing and that it reflects a reality somewhere, despite whatever anyone else says. What we see in these dynamics is a stickiness of affect (Smith & Snider, 2019), which makes not only fact-checking but also content moderation limited. As de Keulenaar, Burton, and Kisjes (2021) show in their study of YouTube’s content moderation techniques, the direct intervention of the platform into the content it transmits is perceived by users as “operational biases” (p. 137) and therefore as a betrayal that further feeds conspiracy theories. The stickiness of affect—that one just wants to keep experiencing an affect that becomes a reassuring factor in one’s life—goes beyond individual platforms as well. The search for affective reactions jumps from one mainstream platform to a fringe one, from watching TV to talking to neighbors to posting online. It shapes forms of collectivity, just as it amplifies patterns of isolation and exclusion (in this Special Section, Akhavi & Sharp; Alloing et al.; Nguyễn, Le, & Moran; Sylvester et al.; Rogers & Koronska; Noordenbos & Tuters). Mis/disinformation actors then use this to articulate specific capacities for action from users: on the platform itself, by continuing to scroll and post content; for the mis/disinformation actors, by articulating and shaping these affective responses within broader social emotions (e.g., nationalistic pride) and potential actions based on reinforced certainties (e.g., voting for a specific candidate, donating money, participating in a protest, and so on).

Decoupling Affect from Authenticity

The machinery of affect as mis/disinformation as it spreads via social media platforms, both mainstream and fringe as well as through the media, might seem unassailable. From “post-truth spaces” that interpolate their views in participatory propaganda (Rogers & Koronska) and recommendation algorithms that automatically repeat affective triggers, through the ease of astroturfing online to regurgitate the same stories (Alloing et al.), not to mention the growing ubiquity of AI-generated text and images, to PR companies that specialize in crafting messages based on sociobiographic profiling and well-funded populist groups at the local and global levels (Noordenbos & Tuters), mis/disinformation emerges as vast networks that escape borders and national jurisdictions. While taking on mis/disinformation requires a multisited approach, from engaging with major epistemological pivots (Sylvester et al.) and ontological shifts to foster a new planetary thinking (Hui, 2024) to focused media policy interventions, we suggest that further inquiry into the user experience is warranted. As many have pointed out, the user in the corporate social media model is a site of extraction of psycho-social materials (Sampson, 2020). Users both willingly and unwittingly give up personal data, in our case, socibiographical information, as well as how such information articulates itself more readily toward some affective reactions and states than others. This is where we see a potential site of intervention to develop new digital and affective literacies.

Specifically, meaningful intervention can occur by remediating how users experience and express their affective and emotional experiences against extractive models of corporate social media sentiment analysis. Where the social media user can be visibly active and thus tracked and read by social media platforms is by posting or reacting to content, content that is continually assessed and aggregated to assemble robust demographic models of affective orientation. Such reactions, whether through text, emoticons, or swiping left or right, constitute the basis of online sentiment analysis, alongside comparisons of time spent watching different types of content to gauge interest and attention. Typically, social media platforms tend to identify a dominant sentiment, although hashtagging offers more variety. Historically, sentiment was first understood as either negative or positive, though sentiment analysis has transitioned from these polarities toward a more gradated understanding to detect anger, sadness, fear, surprise, and disgust (Munezero, Kakkonen, & Montero, 2011).

Yet it is clear that the theories of affect undergirding these instruments of affective exploitation are seriously deficient: corporate social media platforms see affect not as something to be understood and cared for, but as something to be captured and used: affects are datafied and metrified. Such an extractive posture resembles Tomkins’ (1995) assessment of a strong affect theory, one that can handle expansive data sets at the expense of more nuanced outputs. As Crawford and Paglen (2021) observe, the enormous training sets that feed corporate social platforms reflect their creators’ personal valences and biases—the “supervision” in supervised machine learning systems—and, within machine learning and natural language processes, these sentiments are “flattened,” erasing the gradients of human understanding and experience. Furthermore, the strategies social media giants offer to grant their users both small- and large-scale knowledge about how they feel posit a zero point between affect and sentiment, ossifying affective experience through false hierarchies and mutual exclusions (for instance, that anger and joy typically cannot coexist). Sentiment analysis codes words, facial expressions, and algorithmically recognized patterns into an inner core of authentic feeling that is both universally expressed and semantically ambiguous. Emoji-based reaction options, for instance, are singular and delimited, and though they may be deployed ironically, the scrolling subject is only ever thought of as feeling a base emotion. Sentiment analysis thus represents another leg in the passage from affect to feeling to emotion, one that arranges affective data demographically to infer a population’s motivations or collective conatus.

What happens when affect is translated into sentiment within the corporate social media environment? One key thing is that sentiment involves the disembodiment and objectivization of affect, as affect ceases to be something that is felt and becomes an extracted data point that is then given back to the user as further recommendations, potentially establishing an unstable experience where users become aware of the extractive processes they are subjected to while being encouraged to act with spontaneity and authenticity. In so doing, affect becomes its own object, the consequences of which are poorly understood. Han (2017) links the proliferation of affect theory itself to the accession on the part of researchers to the extractive demands of neoliberal capital, but these inquiries also outline the paucity of the theorization of affect that guides corporate social media platforms. We might position this as an effect of a strong theory of affect—one animated by private interest to monopolize all incoming information as a source of value extraction—or as the ethical bankruptcy of platforms that disregard their deleterious social influence. As users, our affective experiences become layered because we are affected by how our affects are rendered to us. This kind of branching off of affect has two consequences: On the one hand, we follow the recommendations and can play with our feedback to maximize a specific kind of affective experience, acting in that sense like we are expected to by the platform. On the other hand, we also experience a feeling of detachment, which can build up to disaffection, exhaustion, digital lethargy (Hu, 2024, p. xvii), and fatigue from the constant barrage of an “experience capitalism” that breaches the boundaries of wakefulness and sleep (Sampson, 2020), as our affects become more and more detached from our embodied experiences. We join Ngai (2015) and Paasonen (2021) in recognizing that in this very context of rationalizing, prodding, and mining affect, there is an ambiguity that allows for intervention. In other words, when affect ceases to lead to certainty, but rather to the nascent realization of an emptied-out, burnt-out self (Han, 2015), there lie new possibilities.

Digital research methods and tools can provoke fulsome contemplation of how and why affect is experienced in ways that do not necessarily lead to increasingly fascistic certainties. While Alloing et al. rightly note that online platform affordances are easily exploited to influence affect, digital research methods, such as the research persona method employed by Weltevrede and Burton and Akhavi and Sharp in this issue, show how these affordances can be manipulated to expand the spaces where affect is experienced. If affect is innate, preconscious, and durationally specific, it is not possible to preemptively intervene in affective responses triggered by mis/disinformation. Rather, it is in the space between affective experience and its ossification as certainty that we can intervene to redefine affect’s potentialities. The development of interface tools that allow users to experience and explore affect differently could open up new possibilities for experiencing affect as something creative and full of potential.

One such instance is the “Image-Affect” browser plug-in we are helping design, which allows users to record all the affects they experience while viewing an image and revisit them over time (forthcoming). In layering affect as a coassembly of neurological, physiological, and aesthetic feedback, we draw from Tomkins’ (1995) work on affect theory to merge multiple levels of experience and experimentation without subordinating one to the other. Tomkins’ theory of affect is notoriously ambiguous because of its indistinction between what we might prefer to delineate as instinctive reactions and encultured responses, especially when attempting a definitive taxonomy of primary affects. Yet, as his more recent interlocutors (Frank & Wilson, 2020; Sedgwick, 2003) propose, this indistinction is precisely the strength of Tomkins’ theorization. Though some critics read his theorization as biologically deterministic (Leys, 2017), we interpret it as a theory of thresholds, tracing the immediacy of affective experience as it animates and is absorbed into frameworks for collective understanding. By installing an axis of sliding scale affect/intensity categories alongside a streamlined interface that makes it easy to tag content encountered online, this tool expands the spaces available for registering affect. Instead of an aggregable system of likes and dislikes, our tool provides a persistently fuzzy catalogue of affective responses for the individual user and a shifting tableau collaging shared affects for groups. This fairly simple intervention frees affect, and in our group discussion with users of the tool, it became clear that what was mundane submission to the affective demands of a social media platform became a way for users to understand affect as layered, complex, and playful. Overall, work on creating new user interfaces can be incredibly useful for a digital literacy strategy, as it allows users to understand the demands imposed on them by social media platforms and how they are affectively manipulated.

We theorized through this afterword that mediation shifts the transmission of affect from presubjective experiences into formal certainty, with or without translation into feeling or emotion. Certainty requires a certain level of meaningfulness (“I know that it was a snake!”), which might become shaped into a representation, but is also defined by repeated, ritualistic, habitual, and asignifying formalizations that, while figural, remain meaningful. Like the institutional continuities between affect and emotion outlined above, the interaction between unmediated affectivity and formalized meaning is a complex issue in affect theory. Some polemical approaches suggest that affect theory (Bollmer, 2023; Leys, 2017), because of its roots in psychological experimentation, cedes the capacity for meaning-making to the instruments and interfaces that index the autonomic “experience of experiencing,” thereby isolating affect from emotion (Massumi, 2002, p. 5). Yet, like the distinction from emotion, we view this as a productive constraint, one that affirms affect theory’s capacity to operate as a “variant of materialist media theory” (Bollmer, 2023, p. 206), where affect performs intra-actively with its medium of measurement. Seizing on the indistinctions enabled by a Tomkins-inspired approach to affect theory, our work on research methods and tools relies on experimentally layering analog and digital representations of affect to shift the forms available to measure its transmission (Sedgwick & Frank, 1995). Following the insights of film and media theorist Eugenie Brinkema (2014), our analysis works from the assertion that we cannot “have already determined that forms and bodies have nothing to say to each other” (p. 40). The autonomy of affect as a natural, preindividual intensity does not contradict collective meaning. Though Brinkema herself is skeptical that the neuropsychological taxonomy of affects associated with Tomkins can adequately operate within a Spinozist or Deleuzian interpretation of affect, there is a shared preoccupation with the speculative in affect theory.

In conclusion, we perceive affect as a potent, inherent element of meaning that can enhance and alter conscious understanding of and establish what is felt as certainty. This is not to flatten affect toward a teleology of stabilized, subjective cognition, where every form has a finite feeling to be uncovered through interpretation. We approach affect intra-actively, as a quality that is not reducible to either preexisting personal interiority or abstract structures of feeling. At the level of affect, form familiarizes us with the branching paths that affective feedback might take, entrenching emotional investments in specific objects and encounters. The patterning of affect in repeated, formal experiences generates an incipient tendency toward “self-fulfilling prophecies,” allowing specific affective forms to propagate and monopolize all incoming experiences (Tomkins, 1995, p. 141). Whether in the serpentine form of a coiled rope in shadow or the everyday ritual of looking both ways before crossing the street, affect emerges through specific, materially embodied configurations irreducible to the experience they engender. Fear-terror does not transcend the situation it indexes, but informs it, installing the difference that will make the difference when a similar stimulus is encountered in the future. In this way, algorithmically optimized media forms that direct our engagement online are not merely bearers of pregiven affect, but complex bundles of collective interpretive clues that shape both how we feel and what we feel.

Focusing on the nonlinguistic, nonrepresentational mediating processes that produce affective responses explains why content-focused interventions like fact-checking and moderation, while needed, cannot be the only solution to mis/disinformation. What they may tackle in content, they neglect in form. This is especially clear with corporate social media platforms. Thus, focusing on affect as a central site of mis/disinformation involves developing tools that recognize the inherent ambivalence of affect, which is typically excluded from corporate social media approaches to the mining of affect. Our endeavors in designing the image and affect tool, rather than attempting to freeze affective experience in a singular expression, prioritize the virtualities of affect through a modular design that emphasizes intensity and ambivalence. Such experiments at the site of users’ affective experience offer promises for developing digital literacy and combating mis/disinformation. Affective responses are both the means and goals of mis/disinformation, enabling an expanded critique of information that moves beyond facticity toward culturally troubling experiences of authenticity, gut reaction, and preconscious instincts.

References

Ahmed, S. (2004). Affective economies. Social Text, 22(2), 117–139.

Altheide, D. L., & Grimes, J. N. (2005). War programming: The propaganda project and the Iraq War. The Sociological Quarterly, 46(4), 617–643. https://doi.org/10.1111/j.1533-8525.2005.00029.x

Andersen, J., & Søe, S. O. (2020). Communicative actions we live by: The problem with fact-checking, tagging or flagging fake news—the case of Facebook. European Journal of Communication, 35(2), 126–139. https://doi.org/10.1177/0267323119894489

Anderson, B. (2006). Imagined communities: Reflections on the origin and spread of nationalism. New York, NY: Verso Books.

Benkler, Y., Faris, R., & Roberts, H. (2018). Network propaganda: Manipulation, disinformation, and radicalization in American politics. New York, NY: Oxford University Press.

Bennett, W. L., & Livingston, S. (2018). The disinformation order: Disruptive communication and the decline of democratic institutions. European Journal of Communication, 33(2), 122–139. https://doi.org/10.1177/0267323118760317

Berry, J. M., & Sobieraj, S. (2016). The outrage industry: Political opinion media and the new incivility. New York, NY: Oxford University Press.

Blackman, L. (2019). Haunted data: Affect, transmedia, weird science. London, UK: Bloomsbury Publishing.

Bollmer, G. (2023). The affect lab: The history and limits of measuring emotion. Minneapolis: University of Minnesota Press.

Bösel, B., & Wiemer, S. (Eds.). (2020). Affective transformations: Politics - algorithms - media. Luneburg, Germany: Meson Press.

Brinkema, E. (2014). The forms of the affects. Durham, NC: Duke University Press.

Chadwick, A. (2019). The new crisis of public communication: Challenges and opportunities for future research on digital media and politics. Loughborough, UK: Loughborough University.

Chen, M. Y. (2012). Animacies: Biopolitics, racial mattering, and queer affect. Durham, NC: Duke University Press.

Crawford, K., & Paglen, T. (2021). Excavating AI: The politics of images in machine learning training sets. AI & SOCIETY, 36, 1105–1116. https://doi.org/10.1007/s00146-021-01162-8

de Keulenaar, E., Burton, A. G., & Kisjes, I. (2021). Deplatforming, demotion and folk theories of Big Tech persecution. Revista Fronteiras, 23(2). https://doi.org/10.4013/fem.2021.232.09

Dean, J. (2009). Democracy and other neoliberal fantasies: Communicative capitalism and left politics. Durham, NC: Duke University Press.

Deleuze, G. (2022). Spinoza: Practical philosophy (R. Hurley, Trans.). San Francisco, CA: City Lights Publishers.

Deleuze, G., & Guattari, F. (1996). What is philosophy? New York, NY: Columbia University Press.

Frank, A. J., & Wilson, E. A. (2020). A Silvan Tomkins handbook: Foundations for affect theory. Minneapolis: University of Minnesota Press.

Freelon, D., & Wells, C. (2020). Disinformation as political communication. Political Communication, 37(2), 145–156. https://doi.org/10.1080/10584609.2020.1723755

Gil de Zúñiga, H., Scheffauer, R., & Zhang, B. (2023). Cable news use and conspiracy theories: Exploring Fox News, CNN, and MSNBC effects on people’s conspiracy mentality. Journalism & Mass Communication Quarterly, 101(4), 889–910. https://doi.org/10.1177/10776990231171929

Graves, L. (2016). Deciding what’s true: The rise of political fact-checking in American journalism. New York, NY: Columbia University Press.

Gregg, M., & Siegworth, G. (2010). The affect theory reader. Durham, NC: Duke University Press.

Han, B.-C. (2015). The burnout society (E. Butler, Trans.). Redwood City, CA: Stanford University Press.

Hern, A. (2018, July 19). FamilyOFive: YouTube bans “pranksters” after child abuse conviction. The Guardian. Retrieved from https://www.theguardian.com/technology/2018/jul/19/youtube-bans-familyofive-pranksters-michael-heather-martin-child-abuse-conviction

Hu, T.-H. (2024). Digital lethargy: Dispatches from an age of disconnection. Cambridge, MA: The MIT Press.

Hui, Y. (2024). Machine and sovereignty: For a planetary thinking. Minneapolis: University of Minnesota Press.

Illouz, E. (2023). The emotional life of populism: How fear, disgust, resentment, and love undermine democracy (A. Sicron, Trans.). Medford, MA: Polity.

Jucan, I. B. (2023). Malicious deceivers: Thinking machines and performative objects (1st ed.). Redwood City, CA: Stanford University Press.

Karppi, T. (2018). Disconnect: Facebook’s affective bonds. Minneapolis: University of Minnesota Press.

Latour, B. (2004). Why has critique run out of steam? From matters of fact to matters of concern. Critical Inquiry, 30(2), 225–248.

Leys, R. (2017). The ascent of affect: Genealogy and critique. Chicago, IL: University of Chicago Press.

Lordon, F. (2014). Willing slaves of capital: Spinoza and Marx on desire (G. Ash, Trans.). New York, NY: Verso.

Lordon, F. (2016). Les affects de la politique: Pour un structuralisme despassions [The politics of affect: Toward a structuralism of feelings]. Paris, France: Éditions du Seuil.

Lovink, G. (2019). Sad by design: On platform nihilism. London, UK: Pluto Press.

Lukito, J., Suk, J., Zhang, Y., Doroshenko, L., Kim, S. J., Su, M.-H., … Wells, C. (2020). The wolves in sheep’s clothing: How Russia’s internet research agency tweets appeared in U.S. news as vox populi. The International Journal of Press/Politics, 25(2), 196–216. https://doi.org/10.1177/1940161219895215

Marres, N. (2018). Why we can’t have our facts back. Engaging Science, Technology, and Society, 4, 423–443. https://doi.org/10.17351/ests2018.188

Massumi, B. (2002). Parables for the virtual: Movement, affect, sensation. Durham, NC: Duke University Press.

McCann Ramirez, N. (2023, January 23). The Proud Boys have really strict rules about jerking off. Rolling Stone. Retrieved from https://www.rollingstone.com/politics/politics-news/proud-boys-trial-rule-book-masturbation-ban-1234666317/

Munezero, M., Kakkonen, T., & Montero, C. (2011). Towards automatic detection of antisocial behavior from texts. In Proceedings of the Workshop on Sentiment Analysis where AI Meets Psychology (SAAIP 2011), 20–27. Chiang Mai, Thailand: Asian Federation of Natural Language Processing.

Munger, K. (2020). All the news that’s fit to click: The economics of clickbait media. Political Communication, 37(3), 376–397.

Ngai, S. (2015). Our aesthetic categories: Zany, cute, interesting. Cambridge, MA: Harvard University Press.

Paasonen, S. (2021). Dependent, distracted, bored: Affective formations in networked media. Cambridge, MA: The MIT Press.

Sampson, T. D. (2020). A sleepwalker’s guide to social media. Medford, MA: Polity.

Sand, J. (2005). A short history of MSG: Good science, bad science, and taste cultures. Gastronomica, 5(4), 38–49.

Sedgwick, E. K. (2003). Touching feeling: Affect, pedagogy, performativity. Durham, NC: Duke University Press.

Sedgwick, E. K., & Frank, A. (1995). Shame and its sisters: A Silvan Tomkins reader. Durham, NC: Duke University Press.

Smith, N., & Snider, A.-M. (2019). ASMR, affect and digitally-mediated intimacy. Emotion, Space and Society, 30, 41–48. https://doi.org/10.1016/j.emospa.2018.11.002

Tomkins, S. S. (1995). Exploring affect: The selected writings of Silvan S. Tomkins. Cambridge, UK: Cambridge University Press.

Walter, N., & Tukachinsky, R. (2020). A meta-analytic examination of the continued influence of misinformation in the face of correction: How powerful is it, why does it happen, and how to stop it? Communication Research, 47(2), 155–177. https://doi.org/10.1177/0093650219854600

Wetherell, M. (2015). Trends in the turn to affect: A social psychological critique. Body & Society, 21(2), 139–166. https://doi.org/10.1177/1357034X1453902

Zelenkauskaite, A. (2022). Creating chaos online: Disinformation and subverted post-publics. Ann Arbor: University of Michigan Press.

https://doi.org/10.65476/9xn78026

^[1] See a similar point made by Chen (2012) in Animacies: Biopolitics, Racial Mattering, and Queer Affect