Introduction

Affect, emotions and moods all play an important role in social and political life. They motivate, excite, colour experience, are core to communication, help us perceive value and inform our judgements (including those of a moral sort). This chapter accounts for the energising role of feelings in relation to false information throughout the civic body. Using feelings as a catch-all term to describe affects, emotions and moods, as well as reactions to stimuli we may not be aware of, we start by charting the trajectory of the role of feelings in understanding citizen-political communications. Their persuasive importance was recognised millennia ago and this has been recognised anew in recent decades with the advent of neuroscience and the understanding that emotions are important for decisions and judgements. Many studies address how governments can try to best manage public feeling, and hence behaviour, and we highlight three main mechanisms: discursive, decision-making based and datafied.

Claims that we live in a post-truth condition are prevalent, with appeals to emotion and personal belief argued to be more influential in shaping public opinion than objective facts. While the relative importance of emotion and facts in everyday life is difficult to ascertain, we demonstrate that the media from which people would normally derive their facts (namely, news media and social media) have become more emotionalised and affective. We suggest that we live in an informational environment that is sub-optimal for a healthy civic body. We exemplify this by examining challenges faced by governments in managing their population’s feelings during the COVID-19 pandemic where uncertainty, anxiety and false information proliferate.

Feelings in Citizen-Political Communications

The role of affect, emotions and moods in understanding citizen-political communications has been uneven across the centuries. Their importance for persuasion was recognised several millennia ago, as well as by those engaged in twentieth-century social engineering and propaganda. Although challenged by Enlightenment-oriented discourse and much communications research, their importance has been refreshed in recent decades with the advent of identity-based conceptions of the political, as well as neuroscientific understanding that emotions underpin our decisions. In this section, we chart this trajectory, but first, we define our terms.

The distinction between affect and emotion has long been debated (Döveling et al., 2011). Writing on the ‘passions’ goes back at least as far as Aristotle in the Nicomachean Ethics. The word ‘emotion’ came into use in the English language in the seventeenth and eighteenth centuries, but the idea of emotions as mental states that can be systematically studied only emerged in the nineteenth century (Dixon, 2012). Although emotions have been studied across diverse disciplines, we are apparently ‘not much closer on reaching consensus on what emotions are than we were in Ancient Greece’ (Scarantino, 2018, p. 37). We recognise that emotion has physiological qualities, but also that emotion is not synonymous with these. Emotion has episodic, experiential, intentional, communicative, historical, cultural and social qualities too. A reasonable working definition is that emotions are ‘internal states that arise following appraisals (evaluations) of interpersonal or intrapersonal events that are relevant to an individual’s concerns’ which in turn ‘promote certain patterns of response’ (Cowen et al., 2019). Affect, again studied by diverse disciplines, is broadly understood as feelings that are less fully formed and a ‘general property of experience that has at least two features: pleasantness or unpleasantness (valence) and degree of arousal’ (Barrett et al., 2019, p. 51). Important, too, is the more literal definition of affect, to be moved by something, potentially without a person or group being aware. Affect is part of every waking moment of life and not specific to instances of emotion, although all emotional experiences have affect at their core. A mood is a longer-term condition, and drawing on Heidegger’s (2011 [1962]) phenomenology of moods and experience, McStay (2018, p. 164) defines moods as ‘a way of being-in-the-world. They represent an attunement, that characterises “being-there” and the disclosure of how and what things are’.

The powerful role of feelings in persuasive communications was recognised in the era of classical Greek Democracy (5 BC) by Aristotle. He advocated that rhetors use ‘pathos’ (appeals to emotion), as well as ‘ethos’ (appeals to the speaker’s character and personality) and ‘logos’ (appeals to rationality) (Aristotle, 1991). The contagious nature of emotion has also long been observed. David Hume, for example, in A Treatise of Human Nature (1978 [1739]) spoke of affections passing from one to another. Likewise, Gustave Le Bon’s The Crowd (2016 [1896]) spoke of emotional contagion as undermining individual rational thought, exaggeration of sentiment, impulsiveness, force, destruction and absence of critical spirit.

As the early twentieth century ushered in the rise of mass communication technologies alongside expanded electorates, American scholars reflected on how these could be combined with psychological research to manage populations and engineer consent (Bernays, 1928a, b; Lasswell, 1936). For instance, Bernays applied the concept of the subconscious mind (pioneered by his psychoanalyst uncle, Sigmund Freud) to management of mass communications, blending the idea of subconscious messaging with theories of crowd psychology and herd instinct (Bernays, 1928a). Describing mass psychology and the ‘group mind’, Bernays posits that: ‘In place of thoughts it has impulses, habits, and emotions’ (Bernays, 1928a, p. 73). Also influenced by psychoanalytical concepts, Lasswell (1936, p. 317) describes how elite propaganda, which he regarded as necessary to manage the masses, is complicated by the ‘changing emotional requirements of the community, moods of submissiveness, moods of self-assertion’. Certainly, studies of propaganda frequently analyse how it is designed to bypass rational thought (Quaranto & Stanley, 2021).

Despite these forays into mass media’s influence on public feeling, early communication research largely emphasised cognitive aspects such as recall, learning, thoughts and beliefs, with emotions regarded as mere ‘noise’ (Konijn & ten Holt, 2011). Indeed, until the 1990s, information processing models assumed that affect and cognition were two antagonist forces, with cognition the pre-eminent force and affect something requiring control (Spezio & Adolphs, 2013 [2007]). However, neuroscience has since challenged presumptive claims that devalue emotional processes in decision-making (Barfar, 2019), with the influential works of António Damásio (1994, 2010) suggesting that emotions can enhance information processing. Damásio (1994) showed in Descartes’ Error that people with brain damage that makes them incapable of experiencing emotion or detecting it in others cannot function rationally: they cannot feel what decisions will make them (or others) happy or unhappy.

Within contemporary communications and cultural studies scholarship, the ‘affective turn’ has also become more pronounced in recent decades (Clough & Halley, 2007; Döveling et al., 2011; Bösel & Wiemer, 2020). As McStay (2013) notes, it was Deleuze’s (1988 [1970]) monograph on Spinoza, and his later work with Guattari (Deleuze & Guattari, 2011 [1994]), that helped to return affect to contemporary critical attention within this field. Spinoza (1996 [1677]) opposed mind–body dualism, preferring instead a monism that more directly links mental goings-on with the body. For Spinoza, the mind and body work in parallel (as with neuroscientific accounts) and are indivisible, somehow made of the same substance. Such affective accounts explore drives, motivations, will, emotion, feelings and sensations. For Spinoza, these were central to being human. In Deleuzo-Guattarian terms, affect is an aesthetic activity in the way that artists are interested in generating intense experiences that take the body (including the brain) from one condition to another. McStay (2013, p. 4) argues that such conceptions of affect allow us to analyse media in terms of attention, attraction, stimulation, sensation, context and corporeal events.

Despite these interventions, our understanding of the significance of affect, emotion and mood in citizen-political relations has been hampered by the suffusion with Enlightenment principles of liberal democratic theory (the dominant mode of political organisation in Western democracies) (Wahl-Jorgensen, 2019). The Enlightenment path to knowledge idealises rational, dispassionate, autonomous and informed citizens; and it regards passions as something to be controlled and channelled, without recognising the orienting role of affect and emotion (Kant, 1998 [1781]). Following Enlightenment principles, Jurgen Habermas argues for a privileging of the rational in his (now archetypal) democratic ideal of the public sphere. This ultimately seeks a consensus among citizens by enabling all to speak rationally, through listening to each other’s viewpoints and agreeing the best way forward (Habermas, 1984).

By contrast, questioning the very desirability of Habermas’ rational, consensus-oriented model of deliberative democracy, Chantal Mouffe (2005) regards ‘the political’ as a space of power and unavoidable conflict between adversaries. In this affective, identity-based conception of the political, people are deeply embedded within their own communities and passionately attached to their own conceptions of the common good. For Mouffe, a heartfelt clash of perspectives (rather than a universal, rational consensus) is central to democracy (Mouffe, 2005, p. 11) as it produces an agonistic debate where citizens can be heard and choose between real alternatives (Mouffe, 2013) rather than adopting a technocratic consensus. From Mouffe’s perspective, communication that is persuasive, passionate and conflictual is desirable, as long as it does not seek to exclude others either discursively or literally. Discursive exclusion could constitute rendering adversaries as enemies through hate speech (that creates an othered object of disgust). Literal exclusion might constitute advocating physical elimination of the other side (as in genocide). Indeed, there is a long tradition of work within feminist, Black and queer scholarship that values the power of public feelings as important sources of knowledge about power, oppression and governance (Blackman, 2022).

As such, the ‘emotional public sphere’ (Lunt & Pantti, 2007; Richards, 2007) where emotions are expressed, shared and managed draws attention to civic gains from affective, mediated engagement. Often, it is the advent of new media forms that prompts scholarship on their affective affordances. While today it is the affective nature of social media that is attracting widespread attention, and tomorrow it may be the affective nature of biometric forms of emotional AI, in the early 1990s it was the rise of satellite TV and its live imagery of far-flung conflicts that absorbed many scholars of media, politics and tele-diplomacy. Arising, the term ‘the CNN effect’ was coined to describe how media influence foreign policy by evoking audience responses through concentrated, emotionally based coverage, which pressurises governments to respond (Livingston, 1997, June). Almost two decades later, Papacharissi’s (2015) analysis of events on Twitter (such as the 2011 Arab Spring and Occupy) conceptualises an ‘affective public’ where people use social media platforms to facilitate engagement, shape solidarity and make their voices matter in everyday politics.

These affective affordances play out differently across diverse geopolitical contexts. In a context of authoritarian silence in Tunisia and Egypt, Sumiala and Korpiola (2017) explain the construction of digital solidarities in the circulation and remediation of martyr narratives of the suicide protest of Tunisian fruit seller, Mohammed Bouazizi and the death of a young Egyptian man, Khaled Saeed, after being beaten by police. In a geopolitical context of strong, complex institutions for Internet censorship, Song et al.’s (2016) study of emotional expression on political aspects of food safety issues on China’s microblogging site, Weibo, notes the benefit of forming like-minded clusters around emotions expressed. It concludes that such activity can convey political opinions that resonate, helping to hold the state accountable, which is beneficial in an authoritarian society that values rational social engineering to efficiently achieve order (also see Tong, 2015). Given the importance of feelings in citizen-political communications worldwide, it is unsurprising that scholarship has turned to addressing how governments can best try to influence and manage public feeling within the civic body.

Managing Public Feeling: Discourses, Decision-Making and Datafication

We highlight three mechanisms that power-holders use in efforts to manage public feeling and hence behaviour: one is discursive, one is decision-making based and one is datafied.

Managing Discourses

The management of public feeling can be attempted through carefully constructed discourses in public communications. Wahl-Jorgensen (2019) sees the discursive construction of emotion through media texts as carefully staged strategic performances for specific purposes and audiences, driving social and political action. She argues that societies have always been preoccupied with managing emotions, with eras characterised by distinctive ‘emotional regimes’, namely, normative emotions and ways of expressing them in public. As already noted, early twentieth-century American mass communication scholars explored how mass media could be combined with psychological research to manage the population (Bernays, 1928a; Lasswell, 1936). A century later, political rulers continue to attempt emotional influence of their populations through media. For instance, in China, the 14th Five-Year Plan for National Informatisation issued by the Central Commission for Cybersecurity and Informatization (2021, December 28) states its aim of expanding ‘diversified online propaganda platforms and channels’ and strengthening ‘the propagation of positive energy information’ in cyberspace.

Likewise, in political campaigning, the discursive manipulation of the electorate’s feelings has long been attempted. During India’s 2019 General Election, the ruling Hindu-nationalist party, the BJP, used humour, wit and sarcasm in its digital campaigns, helping entrench conversations about Hindu nationalism, stretching the boundaries of what could be said in public and creating familiarity with nationalist vocabulary while stirring Hindu majority fears against the 14% Muslim minority (Naumann et al., 2019). Such discursive manipulation during elections has been extensively studied in the USA. A longitudinal study of television advertising in US presidential campaigns across the second half of the twentieth century (1952–1996) finds that emotional appeals are more often dominant than logical or ethical ones (Kaid & Johnston, 2001). Negative campaigning (namely, attacking an opponent) has also long featured in American political campaigning (evident since at least 1800) as well as in other democracies (Fowler et al., 2016) and continues on social media (Haselmayer, 2019). A meta-analysis of 111 studies (mostly from a pre-social media ecology) finds that negative campaigning is more memorable and stimulates knowledge about the campaign yet, less positively, also slightly lowers feelings of political efficacy, trust in government and possibly overall public mood (Lau et al., 2007). Worryingly, experimental work (on American college students) in cognitive psychology shows that negatively valenced false political information tends to be more durable than positive or neutral false information, even after the false information is corrected (Guillory & Geraci, 2016).

Managing Decision-Making

A second mechanism for trying to manage public feeling focuses on people’s decision-making processes. The twenty-first century has seen a surge in research suggesting that emotions guide formation of opinions and decisions to take political action (Brader & Wayne, 2016). For instance, analysis of survey data from the 1996 US presidential election shows that voters’ opinions of candidates eventually converge with their initial emotional responses (Just et al., 2013 [2007]). American research on campaign ads evidences the importance of enthusiasm and fear in increasing desire to volunteer and vote (Brader, 2006). When people search for information online, campaign-related and experimentally induced fear consistently makes voters more inclined to pay attention to candidates and debates (Valentino et al., 2008). Incivility and negative political speech can lead to higher participation and stimulate voter turnout (Lu & Myrick, 2016).

The importance of emotions and gut feeling in decision-making more generally has been studied both by behavioural economics and cognitive psychology. It has led governments and corporations to design interventions or ‘nudges’ to help us make better (or different) decisions (The Behavioural Insights Team, 2020). Nudges are ‘any aspect of the choice architecture that alters people’s behaviour in a predictable way, without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid’ (Thaler & Sunstein, 2008, p. 6). Nudges inform people of factual information (such as via warnings, reminders, personalisation, framing, timing and increases in salience); make certain choices easier (via simplification, ease, convenience and active choosing); use the power of default and procrastination (such as default rules on opting in or out); or exploit social influences (for instance, being told what other people do and leveraging social norms) (Sunstein, 2016).

Operationalising nudges, the first Behavioural Insight Team institutionalised in government was in the UK, created in 2010 (Sunstein, 2016). By 2020 its work spanned 31 countries, applying behavioural insights to inform policy and improve public services (The Behavioural Insights Team, 2020). Global digital platforms also embrace nudging to engage and track their users. For instance, Facebook’s get-out-the-vote button nudges users to vote. This message displayed in users’ Facebook News Feed on election day encourages voting, provides a link to local polling places, shows a clickable ‘I Voted’ button, and a counter indicating how many other Facebook users reported voting. The button was first used in the USA in 2008 and has since been used in elections and referenda in multiple countries. Facebook’s own studies show that this slightly increases voter turnout (Bond et al., 2012; Jones et al., 2017). Yet, even this ostensibly positive nudge has raised concerns about uneven exposure across different parts of the population (as not everyone is on Facebook and not everyone is shown the button). It has also raised concerns about interfering in foreign elections (as Facebook, in all countries except the USA, is a foreign power and legally should not be interfering in their elections) (Grassegger, 2018, April 15). Also problematic are ‘dark patterns’, namely, design choices that alter users’ decision-making for the designer’s benefit. For instance, across 2022 Google was sued in the USA over ‘deceptive’ location tracking policies that make it hard for people to understand (and hence control) when or why Google collects and retains their location. The legal action refers to ‘dark patterns’ that include complicated navigation menus, visual misdirection, confusing wording and repeated nudging towards a particular outcome (Wakefield, 2022, January 26).

Managing Datafication (Optimisation)

A third mechanism for attempting to manage public feeling involves gauging citizens’ emotions through their datafied behaviour, gleaned via big data and profiling technologies. Across the twenty-first century, big data has been combined with psychological science to optimise and target individual desires and vulnerabilities, whether for political campaigning or governing.

These profiling and targeting technologies are utilised by political campaigners, as will be elaborated in Chap. 6. Several studies evidence the nefarious practices of digital targeting of misleading messages designed to bypass thoughtful deliberation in favour of emotionalised engagement and culture wars. For instance, Kim et al.’s (2018) study of social media ads run by anonymous groups in the 2016 US presidential election demonstrates that they largely focused on divisive issues and that lower-income, White voters in swing states were most likely to be targeted, especially by ads on immigration and race. They found ads run by these groups to be largely misleading, emphasising negative emotions and political attacks. Attempted emotional manipulation of targeted electorates via social media has also been observed during the Catalan referendum for independence from Spain on 1 October 2017. Analysis of nearly four million Twitter posts collected during the referendum finds two polarised groups of Independentists and Constitutionalists. Bots targeted the most influential humans of both groups. They also bombarded Independentists with violent contents, increasing their exposure to negative, inflammatory narratives (for instance, that inspire fight, violence and shame against government and police), so exacerbating social conflict online (Stella et al., 2018).

Beyond political campaigning, some advocate for real-time social media surveillance to enable timely assessment of the public’s emotional and behavioural responses to governments’ actions. This includes measures to engage the public during crises such as riots and natural disasters (for instance, in the UK, USA and Indonesia), through to full incorporation into epidemic preparedness and response systems for public health communication and control in more centralised countries like China (Chen et al., 2020; Ni et al., 2020). Clearly, feelings deserve serious attention when considering citizen-political communications. This is especially so when dissecting contemporary false information and wider questions of post-truth.

Post-truth? Assessing Emotionalised Media

According to the dictionary definition of post-truth environments, appeals to emotion and personal belief are argued to be more influential in shaping public opinion than objective facts. Scholars point to multiple practices indicative of a post-truth era (Balaskas & Rito, 2021; Blackman, 2022; Capilla, 2021; Farkas & Schou, 2020), many of them discussed in our book. Empirically we cannot speak to whether emotion is more, or less, important than facts currently. However, we can point to the increasing emotionality of media environments from which citizens draw their facts. Below, we examine the emotionality of two such media forms: news media and social media.

The Emotionality of News

Despite journalism’s long-standing ideals of objectivity and privileging of facts over values, emotionality has always been part of the profession (Beckett & Deuze 2016; Peters, 2011), although its extent varies across different types of news genre and outlet, geography, platform and time (Pantti & Wahl-Jorgensen, 2021; Wahl-Jorgensen, 2013). For instance, a century ago, Lippman (1922) (himself a former US journalist) regarded news stories as lacking ‘truth’ because they are dominated by the emotions and hopes of those working in the news organisation. A hundred years later, Glück’s (2021) interviews with Indian and British broadcast journalists within commercial networks and public service broadcasters find that they consider emotionalising elements as indispensable to engaging audiences: Indian producers appear particularly open to interventionist (rather than detached) roles, combining ideas of national development with motivating citizens and government-critical journalistic elements.

Other scholars observe that emotionality is an increasing feature of journalism. Pantti and Wahl-Jorgensen’s (2011) study of British press coverage (1952–1999) of human-made disasters finds that from the 1980s onwards, there is a shift to a more open emotional regime, which values journalists’ individualised emotional expressions, allowing them to raise structural questions of collective significance. Coward (2013) observes that ‘objective’ reporters often want to be known for a distinctive personal voice in the ‘confessional society’. Indeed, digital native news outlets BuzzFeed News (UK) and Vice News (UK) used subjective, confessional and personalised forms of expression to engage young audiences when reporting the 2017 UK General Election (Dennis & Sampaio-Dias, 2021). In the USA, Benkler (2020) observes the changing political-economic and commercial imperatives fuelling ‘outrage’ discourse in right-wing media, going back to the repeal by the Federal Communications Commission in 1987 of the fairness doctrine (which had required broadcasters to offer public affairs programming and a balance of viewpoints).

Others blame the digital media ecology itself for the rise in emotionality in news (Al-Rawi, 2020; Beckett & Deuze, 2016; Peters, 2011). For instance, Al-Rawi’s (2020) concept of ‘networked emotional news’ comprises news stories posted on social media that generate quantifiable collective emotional responses due to audiences’ strong involvement facilitated by Facebook’s Reactions features. Beckett and Deuze (2016) observe that mobile digital media are increasingly personalised and intimate (as in always-on smart devices where personal and public networks interconnect) and that journalism turns to emotion to virally engage news consumers in an increasingly economically competitive news ecology. Not all types of emotion are viral, and there are national differences. An analysis of 9.6 million comments on the New York Times website between 2007 and 2013 finds that comments featuring partisan incivility receive the most engagement, but comments with swearing do not drive engagement (Muddiman & Stroud, 2017). However, a study of user comments from 26 news websites in South Korea in 2012 finds that swearing increases interaction with comments, especially for political discussions (Kwon & Cho, 2017).

Expectations of emotionality in news also play out differently in different countries and different demographics. For instance, whereas in the USA only 29% of people surveyed in 2022 think that journalists should be able to express personal opinions as well as reporting news, this figure is far higher in Japan (44%) and Brazil (60%); and across all countries, younger adults are more prone to this view (Newman et al., 2022). The affective nature of news consumption is further evident in a study of 56 Dutch users’ news browsing which finds that affective considerations influence their clicking patterns (Kormelink & Meijer, 2017, p. 678). That audiences want a news experience that accords with their worldview is found in a qualitative study of far-right citizens in Norway. These citizens perceive that mainstream press do not cover the perceived threats of immigration and Islam objectively and are angry that far-right political actors are silenced and ridiculed in the news. They seek alternative news sources that support their worldview (Ihlebæk & Holter, 2021). There is also a growth worldwide in selective news avoidance. Although across 46 countries surveyed in 2022, the most commonly cited reason for news avoidance is the repetitiveness of the news agenda (43%), and other reasons include emotionality, namely, that the news brings down their mood (36%), feeling worn out by the news (29%) and that the news leads to arguments they would rather avoid (17%) (Newman et al., 2022). Conversely, the role of news on social media platforms in eliciting positive emotions is shown by Al-Rawi’s (2020) study in 2016 of over 12,000 news items on Facebook pages of mainstream American and British news outlets. This finds that social media readers are emotionally engaged with news that involves positive feelings (especially love). Another study of American mainstream news on Facebook shows how different ways of framing protests influences emoji engagement: for instance, if protests are framed as legitimate, this decreases emotional reactions from audiences (Kilgo & Harlow, 2021).

In terms of studying audiences’ emotional relationships with fake news, there are few studies, and these are focused on the USA. Martel et al.’s (2020) experiments into the relationship between experiencing 20 specific emotions and believing fake news find that heightened emotionality at the study’s outset predicts greater belief in fake (but not real) news posts. In other words, there are notable increases in belief in fake news as emotionality increases. Their study also finds correlational and causal evidence that audience’s reliance on emotion increases their belief in fake news. Another US-based study finds that participants are more likely to believe fake news political headlines that align with their existing beliefs (for instance, liberals are more likely to believe negative news about conservatives); react with more negative emotions to such headlines that attack their party; and are more likely to report intentions to suppress fake news that attacks their own party. Furthermore, participants who reported high levels of emotions are more likely to take actions that would spread or suppress the fake news; and participants who reported low levels of emotions are more likely to ignore or disengage from the spread of false news (Horner et al., 2021).

The Emotionality of Social Media

While social media can support rational, deliberative discourse (Jakob, 2020), more studies point to highly emotional content circulating on social media. This section examines areas of emotionality on social media closely associated with false information: incivility, hate speech and conspiracy theories. Notwithstanding everyday realities of how people receive, understand, negotiate and subsequently circulate content, we observe that many studies highlight the various affordances of social media platforms that are then exploited by architects of disinformation.

Internet scholarship has long noted uncivil behaviour online. This includes ‘trolling’, an antagonistic rhetorical practice that aims to elicit emotional responses from unwitting or unwilling targets (Phillips, 2015); online shaming through ‘viral outrage’ (Sawaoka & Monin, 2018); ‘oppressive outrage’ where marginalised voices are silenced with coordinated harassment (Brady & Crockett, 2018); and ‘hate speech’, commonly understood to be bias-motivated, hostile, malicious language targeted at people because of their actual or perceived innate characteristics (Sellars, 2016; Siegel, 2020, p. 57). The overall incidence of hate speech in social media appears to be rare, but there are few systematic studies (Siegel, 2020, p. 66). However, hate speech promotes reactions and travels further. This virality is evident in a study on Facebook surrounding Ethiopia’s 2015 General Election (Gagliardone et al., 2016). Similarly, a big data study across 2016–2018 on Gab (a site created in 2016 as a free speech alternative to Twitter that mainly attracts alt-right users) finds that content generated by hateful users tends to spread faster, farther and reach a wider audience compared to content from non-hateful users (Mathew et al., 2019). Unsurprisingly given its virality, a cross-national survey of youths and young adults from Finland, Germany, the UK and the USA suggests that many have been incidentally exposed to online hate speech (53% of Americans, 48% of Finns, 39% of Britons and 31% of Germans), especially those who use online social networks often and visit ‘dangerous’ sites (Hawdon et al., 2017).

Multiple reasons are posited for the rise of online incivility and hate speech, many pointing to the affordances of social media platforms. While their design would not have intended such anti-social behaviour, it is a regular outcome. A notorious example is 4Chan’s affordances of anonymity and ephemerality that enable what Tuters and Hagen (2020) call ‘memetic antagonism’, namely, the use of memes as vehicles for antagonistically articulating an out-group, unbound by civility. On more mainstream platforms such as Twitter, Ott (2017) suggests that incivility is due to its informality and depersonalisation of interactions with others. Others discuss the emotional disinhibition and lack of social control prevalent online as, for various reasons, including anonymity, participants feel free from social convention (Suler, 2016). Crockett (2017) suggests that digital media may exacerbate expression of moral outrage and viral online shaming in three ways. Firstly, digital media inflate its triggering stimuli (as people are more likely to learn about immoral acts online than in person, as online algorithms promote content most likely to be shared and as people are more likely to share content that elicits moral emotions). Secondly, digital media reduce the costs of online shaming (the tools for quickly expressing outrage online are at our fingertips while hiding the target’s suffering). Thirdly, digital media amplify personal benefits from online shaming (such as virtue signalling moral authority to large audiences).

Conspiracy theories also proliferate on social media (Bessi et al., 2015; Zollo et al., 2017). Conspiracy beliefs are attempts to explain the ultimate causes of significant social and political events and circumstances with claims of secret plots by two or more powerful, malevolent actors. Such beliefs are widespread and long-standing in modern Western societies (Allcott & Gentzkow, 2017; Sutton & Douglas, 2020). The affordances of social media are not a dominant explanation for conspiracy theory proliferation. Rather, explanations point to complex psychological, political and social factors, this demonstrated by a review of studies from psychology, political science, sociology, history, information sciences and the humanities (Douglas et al., 2019). For instance, US experiments and surveys show an association of anxiety and personal uncertainty with conspiracy perceptions: as such, increasing anxiety or personal uncertainty levels (potentially induced by disinformation architects) may lead ordinary people (not just the paranoid) to become conspiracy theorists (Radnitz & Underwood, 2017; Miller, 2020). Conspiracy theories appear to provide broad, internally consistent explanations that help people to preserve beliefs in the face of uncertainty and contradiction, helping them see the world as orderly, understandable and predictable following threatening societal events (van Prooijen & Jostmann, 2013). There is also a relationship between conspiracy belief and distrust in governments, authorities and scientists (Jensen et al., 2021; Lindholt et al., 2021; Sutton & Douglas, 2020). Across these studies, the direction of causality remains unclear. More clear-cut, however, is a demographically representative survey of Americans in 2020 that finds that women are significantly less likely than men to endorse COVID-19 conspiracy theories and that this cuts across political party lines (Cassese et al. 2020).

That conspiracy theories proliferate on social media, then, is not reducible to social media affordances. Yet, these affordances certainly have some bearing. Given that social media are designed to maximise user attention and affect (see Chap. 2), it is unsurprising that conspiracy theories proliferate there. Indeed, a US national online survey (760 adults) into conspiracy beliefs finds that those with heavy reliance on, and trust in, social media news have the highest level of general and COVID-19-related conspiracy beliefs. Furthermore, those who blindly trust social media news are more likely to fall prey to conspiracy theories even if they can identify the false information (Xiao et al., 2021). A UK-wide national survey (May 2020) of over 16-year-olds finds that those who believe in COVID-19 conspiracy theories are far more likely than non-believers to get their information about the virus from social media (Duffy & Allington, 2020). Demography, nation and its media ecology clearly make a difference as an online survey of adults in China, where the information environment is strictly controlled and rumours are banned on social media, shows that social media use was not associated with conspiracy theory endorsement (Su et al., 2021; also see Jensen et al. 2021). An experimental study into how 50 German university students emotionally cope when confronted with an opinion-challenging YouTube clip propagating conspiracy theory disinformation about causes of climate change finds highly varied coping strategies. Of concern is that, for many participants, their climate change problem awareness decreased following exposure to the conspiracy clip (Taddicken & Wolff, 2020).

Indeed, the importance of social media’s affordances in proliferating emotional content more generally is indicated in studies that find that expression of emotion is socially contagious on social media (meaning that a perceiver’s emotions become more similar to others’ emotions as a result of exposure to these emotions), with caveats that such causality is difficult to prove (Goldenberg & Gross, 2020; McStay, 2018). Facebook’s infamous mood study conducted in 2012 secretly optimised 689,003 people’s News Feeds to understand ‘emotional contagion’ on its platform. (This is the only published study that has manipulated users’ emotions without their knowledge on a digital media platform.) When users logged into their Facebook pages, some were shown News Feed content with a greater number of positive words, while others were shown sadder than average content. After the week of exposure to either more positive or negative content, manipulated users were more likely to post either especially positive or negative status messages. When the experimenters reduced the positive and negative content (making News Feeds lacklustre), people reduced the overall amount they posted (Kramer et al., 2014). Certainly, Del Vicario et al.’s (2016) computational, comparative study of Italian Facebook pages’ reporting on two polarised communities (scientific and conspiracy) across 2010–2012 shows that in both communities, emotional behaviour (ascertained by sentiment analysis of users’ posts) is affected by how often users post comments. More posting of comments resolves in a more negative emotional state; and on average, more active users show a faster shift towards negativity than less active ones. This emotional contagion is also found in studies on Twitter (Brady et al., 2017; Ferrara & Yang, 2015; Goldenberg & Gross, 2020; Stieglitz & Dang-Xuan, 2013). For instance, Brady et al. (2017) find that presence of moral-emotional words in 563,312 tweets on three polarising issues increased their transmission by approximately 20% per word.

Emotional contagion is also found on non-US-based social media platforms. A big data analysis of the discussion network on Chinese microblogging site, Weibo, regarding political aspects of food safety (43,575 posts, June–August 2014) finds that compared with non-emotional posts, emotional posts are more likely to be spread through reposting and that political discussions expressing anger are most likely to generate responses (Song et al., 2016). Another big data study of Weibo in 2017 unpicks the massive-scale network of emotion contagion underpinning the anger of online activism. It finds that this is driven by broadcasters (presenting emotionally neutral posts, but signalling that the Chinese authorities are open to public discussion of the topic); celebrities (whose emotional venting acts as ‘emotion initiators’, provoking emotion contagion); and micro-celebrities (who act as ‘emotion brokers’ by connecting diverse subgroups) (Liu & Liu, 2021). Such emotional contagion is not an accident but the result of social media algorithms that are constantly tweaked to optimise engagement.

In short, intense emotions (including incivility, outrage and hate speech) and conspiracies proliferate online at least partly because of the affordances of social media platforms, which as Chap. 2 explained are geared towards eliciting high arousal and viral emotions to further their attention economy. This emotional virality is not just evident on Facebook (the mechanics of which have been revealed by whistleblowers, as discussed in Chap. 2) but on other US and non-US-based social media platforms. Such affordances are exploitable by the varied architects of disinformation: partisanship can be stoked, and money can be made, from civic bodies undergoing strong conflicted emotions.

To summarise, the media from which people would normally derive their facts (namely, news media and social media) have become more emotionalised. Alongside the prominence of false information (see Chap. 4), is it any wonder that claims for a post-truth condition are prevalent? We cannot assess the general accuracy of whether emotion and personal belief play a greater role than facts in shaping public opinion, but we do observe that some (US-based) studies show that people do prioritise emotion over fact in political arenas. We recommend more studies in different affective contexts across the world and demographically to empirically scrutinise the claim that we live in a post-truth condition. From what we have evidenced in this chapter, we can assert that we live in an informational environment that is sub-optimal for a healthy civic body. We exemplify this below by examining challenges faced by governments in managing their population’s feelings during the COVID-19 pandemic where uncertainty, anxiety and disinformation prevail.

Affective Challenges in Managing COVID-19: Uncertainty, Anxiety and False Information

During the COVID-19 pandemic, inherently uncertain facts raised anxiety levels and provided fertile ground for false information worldwide. This made it harder for governments to manage their population’s feelings to secure behaviour changes deemed necessary to combat this highly infectious respiratory disease.

COVID-19 (‘coronavirus disease 2019’) was first reported in Wuhan, China, on 31 December 2019. By March 2020, the virus had spread to over 120 countries, leading the World Health Organization to declare it a pandemic. Governments, to various degrees across the world, and with vastly different resources and states of preparedness, simultaneously mobilised their healthcare systems to cope with an influx of patients requiring prolonged intensive care; attempted to track and curtail the exponential spread of the disease; and instructed citizens to engage in profound and rapid behaviour change including wearing masks, washing hands and engaging in prolonged and repeated lockdowns. Messaging was often mixed, and some governments played down health risks to keep public confidence in the economy. A year following the outbreak, there had been 2.75 million deaths globally, but several vaccines had been developed. Two years following the initial outbreak, the global death toll was over five million, and vaccine roll-out remained highly uneven worldwide, partly because of lack of supply but also because many refused to take the vaccine (Mallapaty et al., 2021). COVID-19 proved to be an inherently affective issue: alongside high death tolls, absence of cure; onsets of new, more transmissible and potentially vaccine-resistant variants; and extreme behaviour change required to quell the death spikes, increased anxiety and depression were reported across multiple countries (Sigurvinsdottir et al., 2020).

While false information in individual countries takes shape under specific affective contexts (as explored in Chap. 3), COVID-19 adds to this the sociological characteristics of being a ‘risk issue’ (Beck, 1992). Like other risk issues (such as climate change), it induces systematic, often irreversible harm (such as death and the debilitating condition of ‘long COVID’). It also makes it hard for people to find trustworthy, reliable information because of three other core features that breed uncertainty. The first of these features is immateriality: risk issues generally remain invisible, giving them an air of unreality. Indeed, across 2020, the visibility of COVID-19 would only become apparent some 2–14 days from infection on manifestation of symptoms (such as loss of smell) or, as many infected people were asymptomatic, on reliable testing for the infection itself or for antibodies (World Health Organisation, 2020a, b, April 17). A second feature breeding uncertainty is reliance on causal interpretations: we only know how COVID-19 is likely to spread because experts have modelled this. For instance, the UK government consulted scientists, as part of its Scientific Advisory Group for Emergencies (SAGE) team, to model different interventions. Experts realised that COVID-19’s reproduction rate, if left unchecked, is exponential, meaning that its effects on populations would largely be invisible in initial weeks, but would rapidly spike thereafter, overwhelming health services. A third feature that breeds uncertainty is that people must respond to the risk without an adequate foundation of knowledge. COVID-19 has no cure and, until almost a year after it was first identified, had no vaccine. Also, the virus mutates, producing new variants with resistance to some of the vaccines. As such, there was incomplete understanding of who is most at risk and what will best prevent it. As the pandemic progressed, some of these knowledge gaps closed, but many remained. These features produce Beck’s (1992) ‘risk society’—a society that is uninsured and incapable of providing for the uncertainties it faces.

How do people react when living with such uncertainty? Uncertainty is strongly related to information seeking, especially with health information online (Lin et al., 2016). Certainly, COVID-19 saw a substantial increase in consumption for mainstream news media and online sources, evident in all six countries surveyed in 2020 before and after the pandemic took effect (Argentina, Germany, Spain, South Korea, the UK and the USA). Four months after the emergence of the disease, people considered the news media to have done a good job in helping them understand the crisis (60%) and in making clear what they can do to mitigate the impact (65%) (Newman et al., 2020). Of course, negotiating such crisis communication is not straightforward for journalists. In countries with more authoritarian tendencies such as Slovenia, where the governing Slovenian Democratic Party seeks to politically instrumentalise and economically devastate the media, scholars find journalists juggling their facilitative role (in helping the public to understand the health crisis, promoting the official discourse and pointing out false information) with a watchdog role critical of those in power (Pajnik & Hrženjak, 2022).

While in five countries surveyed in 2020 (the UK, Ireland, the USA, Spain and Mexico) public belief in false information about COVID-19 is rare, a substantial proportion views such false information as highly reliable. Furthermore, a small group finds common factual information about the virus highly unreliable (Roozenbeek et al., 2020). More commonly, public health scholarship demonstrates that when the public is exposed to novel or contradictory health information, people experience more uncertainty and disorientation, and decreasingly trust scientists issuing these competing recommendations (Chang, 2015; Clark et al., 2019). More broadly, trust in experts has long been in global decline: as far back as 2005, trust shifted from authorities to peers (Edelman, 2021). Analysis of vaccine misinformation across the twentieth and early twenty-first centuries in the USA, UK and Nigeria helps illuminate public reactions in the face of uncertainty. It finds that lack of trust in the science, government and money-hungry pharmaceuticals, alongside rare but heavily reported vaccine accidents, side effects or dangerous experimental tests, exacerbates the public’s vulnerability towards conspiracy theories as they seek explanations (Cabrera-Lalinde, 2022). Even if there is quality, trusted information available, multiple psychological biases that help people reduce uncertainty may prevent them from acting on public health messages. For instance, we may limit our exposure to conflicting information by defaulting to information channels we deem credible (the ‘channel heuristic’), which may generate large variations in beliefs about what is true, especially when issues become politicised, as COVID-19 became in many countries including the USA and Brazil (Dunwoody, 2020; Gramacho et al., 2021, Hamilton & Safford, 2021).

Under such conditions of uncertainty, governments have had to formulate responses that maximise public safety, effectively use finite health services and minimise adverse economic impacts and disruption to people’s lives, but also avoid undesirable population responses such as panic or indifference. In liberal democracies, scholars proffered advice on managing populations based on likely affective reactions to COVID-19. Petersen (2020, March 9) advocated that ‘optimistic anxiety’ (but not insecurity), and telling people the truth about the pandemic, would affect citizens’ political behaviour and information seeking in positive ways (regarding compliance with government measures) while averting panic. Fear is a central emotional response during a pandemic. A meta-analysis reports that appealing to fear leads people to change their behaviour if they feel capable of dealing with the threat, but produces defensive reactions when feeling helpless. Furthermore, people often exhibit ‘optimism bias’: the belief that bad things are less likely to befall oneself than others. Behavioural and social scientists therefore recommended that COVID-19 communication strategies should strike a balance between breaking through optimism bias without inducing excessive anxiety and dread. The study also notes that an emerging sense of shared identity and concern for others arises from the shared experience of being in a disaster and that this feeling can be harnessed by urging ‘us’ to act for the common good (Bavel et al., 2020).

In the UK, Independent Scientific Pandemic Insights Group on Behaviours (SPI-B) provides independent, expert behavioural science advice to SAGE, which in turn advises government ministers and officials. By February 2021, they found low vaccine take-up in certain groups (such as those shielding, and from deprived socio-economic circumstances, and also from non-White groups). Accordingly, a core SPI-B recommendation in March 2021 was to develop communications from ‘a more data-driven approach that moves beyond aggregated headline percentages and flags important disaggregated, nuanced sub-groups, confounders and intersectionality to more efficiently target low uptake and hesitancy in a more tailored manner’ (SPI-B, 2021, March 9, p. 14). SPI-B also recommends that the government should: ‘continue messaging about positive effects of behavioural interventions such as face coverings, high vaccine uptake, low vaccine hesitancy, hope and return to longer goals and avoid blame or enforcement’ (SPI-B, 2021, March 9, p. 2). Positivity, giving hope and encouragement, and avoiding blame, then, was the UK government’s desired emotional regime for managing COVID-19 during 2021.

However, of particular concern to governments trying to change population behaviour is that when health messages are unclear, people are less likely to change behaviour (Chang, 2015; Taber et al., 2015). It was therefore of grave concern to governments and health organisations worldwide that as COVID-19 spread globally, so did emotive, false information. In China, conspiracies circulated that the virus was part of the American trade war, or a biological weapon, or brought into China by American military members. Conversely, in the USA, conspiracies abounded that the virus may have originated in a Chinese lab and was a Chinese bio-weapon (Su et al., 2021). Indeed, a demographically representative online survey of US adults in 2020 found that 52% believe the virus was accidentally released by China and 49% believe it is a Chinese biological weapon (Miller, 2020).

Harmful disinformation about COVID-19 went particularly viral in smaller media markets, where technology companies face lower incentives to take adequate countermeasures, according to a report from East Stratcom Task Force (an organisation set up in 2015 to increase public awareness, understanding of, and resistance to, Russia’s disinformation) (EUvsDISINFO, 2020). Given Facebook’s research showing a small number of posters and commenters were responsible for much anti-vaccine content, an internal memo from 2 April 2021 saw Facebook reducing the number of comments a person could make on posts from authoritative health sources from 300 to 13 per hour. However, a leaked Facebook memo shows that in the first few months of 2021, about 41% of comments on English-language vaccine-related posts risked discouraging vaccinations; and even authoritative sources of vaccine information were becoming ‘cesspools of anti-vaccine comments’ (Schechner et al., 2021, September 17).

There are varied actors and motivations behind COVID-19 disinformation. As well as anti-vaccine activists worldwide whose existence predates COVID-19 (Cabrera-Lalinde, 2022), such disinformation is spread by Russian and Chinese state media, aiming to undermine the European Union and its crisis response and to sow confusion about COVID-19’s origins and health implications (according to a report from East Stratcom Task Force (EUvsDISINFO, 2020) and from an American think tank, the Council on Foreign Relations (Kurlantzick, 2020)). By April 2021, a rumour-tracking program from US-based analytics company, Novetta, found that Russia targets African countries to discredit Western vaccines in favour of its own Sputnik V (Hotez, 2021). In an African context where COVID-19 has exposed poor health systems, governments’ default response, in line with years of official practices, has been denial, secrecy and false information spread through state-controlled media (Ogola, 2020). Religious actors also spread emotive false COVID-19 information. In a Middle Eastern and North African context, Alimardani and Elswah (2020) find that Islamic misinformation and clickbait on social media became more acute during the pandemic. This took shape in false Hadiths (fabrications of retellings of the Prophet’s words and deeds). For instance, a flood of Arabic-speaking YouTube videos prophesised that a divine sound that would take 70,000 souls and leave 70,000 deaf would be heard on the night of the 15th of Ramadan 2020, based on a false Hadith. After being viewed millions of times, their virality and fear led the official Egyptian religious entity, Al-Azhar, to pronounce the videos false. Religious misinformation draws on fear, emotional appeals, or the credibility of religious authority to persuade. It is harder to fact-check and requires a deeper knowledge of religion and its socio-political context to discern.

Such false information harms the civic body mentally and physically. A study from a technology company that monitors and disrupts violent extremism online finds that in terms of mental harms, COVID-19 created a spike in online hate speech against China and Jews, with racially linked incitements of violence, hate speech and a rebirth of old conspiracy theories on Twitter (Moonshot, 2020, April). The physical health of the civic body was harmed as anti-vaccination attitudes hardened: refusal to get vaccinated effects the individual (a greater risk of getting the disease severely and of resulting hospitalisation) and the community (greater virus transmission and strain on health resources). While vaccines were developed by December 2020, vaccine hesitancy remained a major hurdle across 27 countries surveyed in 2020. On average, only one in three would take the vaccine as soon as possible. Those with poor information hygiene were 11% less likely than those with good information hygiene to say that they would take the vaccine within a year (Edelman 2021; also see Roozenbeek et al., 2020). With a large cross-country survey (conducted across September 2020 to February 2021) finding large variations in acceptance of an approved COVID-19 vaccine (ranging from 83% in Denmark to just 47% in France and Hungary), the study finds that lack of vaccine acceptance is associated with conspiratorial thinking (namely, that the government is hiding information about the virus and its cures), as well as with lack of trust in authorities and scientists and a lack of concern about COVID-19 (Lindholt et al., 2021).

Conclusion

Feelings have enduring importance in citizen-political communications, fuelling collective identities and solidarities, and helping form opinions and decisions to act. This chapter highlighted three mechanisms that are used in efforts to manage public feeling and hence behaviour: one is discursive, one is decision-making based (often involving ‘nudges’) and one is datafied (often involving social media platforms and optimisation). We see these mechanisms at play in attempts to gain power (via political campaigning) and in attempts to govern once in office (for instance, in pandemic mitigation).

Our examination of the emotionality of two key media forms from which citizens garner their facts (news and social media) finds these to be highly emotionalised environments and hence fertile grounds for post-truth. Emotionality has always been a part of journalism despite long-standing ideals of objectivity; and it appears to be an increasing feature in the digital ecology given the rise of confessional journalism; changing political-economic imperatives; the personalised, always-on nature of digital media; and use of emotion to virally engage news consumers. Audiences’ expectations of emotions in news vary across countries and demographics. Audiences share news for emotional (as well as other) reasons, and studies show that some audiences want a news experience that accords with their worldview, avoiding news exposure that elicits negative feelings while emotionally engaging with news that involves positive feelings. Indeed, studies suggest that as far as fake news and American audiences are concerned, there are notable increases in belief in fake news as audience emotionality increases and that people are more likely to believe fake news political headlines that align with their existing beliefs. While social media can support rational discourse, more studies point to the often highly negative and positive emotions they circulate worldwide. This is enabled by the affordances of social media platforms and their optimisation of emotions, these then exploited by the architects of disinformation to spread incivility, outrage, hate speech and conspiracy theories. In this media ecology, it is easy to see how and why objective facts may have become less influential in shaping public opinion than appeals to emotion and personal belief, but studies on this causal link are mostly lacking.

Our examination of the resulting harms to the civic body highlights the challenges it poses to governmental efforts to manage their population’s feelings and behaviour during the COVID-19 pandemic where uncertainty, anxiety, false information and conspiracy theories proliferated, where the issues became politicised and where people lack trust in authorities and scientists. While the news in multiple countries is generally regarded as having helped people understand the crisis, the facilitative versus watchdog role of journalists in negotiating such crisis communication is not straightforward, especially in countries with authoritarian tendencies. Both mental harms (online hate speech) and physical harms (reduced vaccine uptake) were evident. We conclude that we live in an informational environment that is sub-optimal for a healthy civic body.

Of note for horizon scanners is that although social media platforms have developed and honed the practice of profiling and targeting individual desires and vulnerabilities, they are now being joined by more emergent forms of emotional AI that claim to read and react to emotions through text, voice, computer vision and biometric sensing (McStay, 2018). While the biometric part of this is not yet widespread enough to have had significant empirical impacts on false information, in our final chapter (Chap. 9), we reflect upon near-horizon futures and how emotional AI may further incubate false information. Before such future-gazing, however, we turn next to examine the role of profiling and targeting in incubating false information online.