Understanding Fake News: Technology, Affects, and the Politics of the Untruth

. This paper provides epistemic and conceptual tools for a better understanding of fake news. It begins by looking for a definition of fake news that distinguishes between fabricated news stories and biased claims, showing that fake news stricto sensu is less frequent and influential than expected. Then a multicausal account is provided that identifies six main drivers for the multiplication and spread of fabricated news: erosion of truth, digitization, changes in journalism, digital illiteracy, emotional biases, rise of populism. Finally, the implications of fake news for democracy are explored. It is suggested that fakes, their statistical irrelevance notwithstanding, contribute to the weakening of trust and the epistemic disorientation of citizens in a digitized public sphere.


Introduction
Tanner 88 is an innovative TV series created by cinematographer Robert Altman and writer Garry Trudeau. It narrates the misadventures of a fringe contender in the 1988 primaries of the Democratic Party by way of mixing him with real events and candidates. In the desperate attempt of Jack Tanner -a representative from Michigan-to gain attention from voters, an element is highlighted: the construction of his persona by political advisers and communication experts. Ironically, the best footage Tanner has to offer is inadvertently captured by his camera operator while in the middle of a private, emotional outburst. But the question Altman is posing concerns the theatricality of politics -the blurring of lines between reality and fiction, truth and untruthin a heavily mediated world. Thirty years later, however, the quality of mediation has changed and television is not the only game in town. On the contrary, the Internet has transformed the mediascape in significant ways, adding new layers of complexity to the question of political truthfulness. Hence the importance of fake news.
The main difference with Jack Tanner's world is obvious: social networks have widened the public sphere, making possible for nearly everyone to take part in the public conversation. In this context, fake news is often perceived as a threat to the ideal of a well-ordered, rational democracy. Deliberately fabricated news might lead people to take wrong political decisions, thus precipitating the deterioration of liberal societies. Or perhaps not? The magnitude and relevance of fake news are in dispute. Media effects, i.e. the specific impact that a discrete piece of information produces on a particular individual, are hard to measure. There is thus a visible contrast between fake news as a good story about democratic regression and the systematic analysis of the phenomenon. Blaming fake news for all that has gone wrong in representative democracies is tempting. Yet social sciences should be cautious and empirically informed before formulating judgements about it. At the same time, its ultimate impact may not be so easily discerned -it can be indirect, further eroding public trust in media. This paper will proceed as follows. Firstly, it will define fake news and reflect upon its real impact. Secondly, a multicausal explanation will be offered that identifies up to six drivers behind its spread: epistemological, economic, affective, demographic, technological, and political. Finally, it will consider the implications that this phenomenon has, or can have, for liberal democracy and the good functioning of its public sphere. A brief conclusion closes the paper.

Defining and measuring fake news
What, exactly, is fake news? How can it be identified? And how can its effect on the public be measured? After all, false news have always existed in various forms -ranging from fabrications to rumours and conspiracy theories. A conceptual stretching would deprive the notion of «fake news» of its usefulness -thus the semantic field must be carefully demarcated.
Although the label is new, scholars have been discussing satirical pastiches and parody journalism since the 1990s (see Jones, 2007). Television formats such as those from The Daily Show or The Onion in the US elicited reflection upon the blurring of the dividing line between fact and fiction, mostly in connection to the pervasiveness of irony as a mode of communication (see Wallace, 1998). On the philosophical front, Harry Frankfurt (2005) explored the concept of «bullshit» and called attention to a type of discourse disengaged from the pursuit of truth. Yet in the early 60s political historian Daniel Boorstin (1997) had already claimed that Americans mistake fabricated «pseudo-events» for «real» news, thus adopting a typically faked view of social reality. He was pointing to an erosion of truthfulness that seems connected to contemporary fake news. The latter can be seen as expressing a crisis of trust in elites as much as a technologically mediated phenomenon facilitated, in turn, by psychological traits common to all individuals.
For news to be fake, it must be both news and fake. This means that a deliberated attempt to distort information is presented with the appearance of news information. For Lazer et al. (2018), the focus is on publishers rather than stories, lest the accuracy of every single news story needs be evaluated. Thus they define fake news as fabricated information that mimics news media content in form but not in organizational process or intent. Fake-news outlets, in turn, lack the news media's editorial norms and processes for ensuring the accuracy and credibility of information (Lazer et al., 2018(Lazer et al., : 1094 However, this may be too rigid. According to these scholars, fake news overlap with false or misleading information purposely spread to deceive people without the appearance of «real» news. If a particular individual disseminates a false story through her Twitter account, therefore, this should not count as fake news. This certainly makes sense for observational purposes, but leaves out a part of what makes fake news a potential danger for democratic processes. An example will help: Internet users spreading images of police violence supposedly taking place in Catalonia during the illegal referendum for independence on October 2017 were fabricating news -they circulated as such and were often echoed by media outlets which took them as real. Only later were most of them revealed as coming from different contexts. Such «misappropriations» -occuring when factual photos are used to supported a concocted narrative related to different facts-must be considered fake news (Tandoc et al., 2018: 143). After all, the news consumer faces the same difficulty in verification as in more orthodox fabrications.
An altogether different matter is the need to distinguish fakes from ideologically biased opinions. Mcnair has insisted on this point, emphasizing that demonstrably false or misleading news should not be conflated with that which is «merely contested by various political and social actors, usually because it is at odds with a particular ideology or worldview» (McNair, 2018: 17). To put it differently, an ideological bias that produces a given interpretation of facts or even a slightly distorted account of them is not the same as a fabricated story that seeks to mislead readers. This is an important distinction that points to the key question of truth and the role it plays in the democratic discussion. Journalistic bias should not be conflated with fabrication of facts, lest a wrong impression is given about the attainability of «truth». Fake news should then be understood as false pieces of factual information that are presented as accurate by media outlets -either established or themselves faked-or by individuals who introduce them in the public sphere as if directly collected from a real event.
How important is fake news? As Lazer et al. (2018) point out, there are few scientific answers to the question of how common is fake news and how does it impact on individuals. The main reason is the difficulty for measuring any media effect: We do know that, as with legitimate news, fake news stories have gone viral on social media. However, knowing how many individuals encountered or shared a piece of fake news is not the same as knowing how many people read or were affected by it (Lazer et al., 2018(Lazer et al., : 1095 Attempts, however, have been made. Grinberg et al. (2019) examined exposure to and sharing of fake news by registered voters on Twitter during the 2016 US presidential campaign and found a high concentration of engagement: only 1% of individuals accounted for 80% of fake news source exposures, and 0.1% accounted for nearly 80% of fake news sources shared. News exposure for most people came from mainstream media, thus questioning the premise that the Internet reinforces selective exposure and the formation of echo chambers. Interestingly, individuals most likely to engage with fake news were older, conservative, and already active news consumers. This makes sense: intense news consumers that also show partisan behavior tend to enter into frequent contact with fake news.
Other studies reach a similar conclusion. Watts and Rotschild (2017) warn that there is a great difference between the attempted influence pursued by fake news producers and their real. It is partly a matter of how numbers are presented: whereas the New York Times identified more than 3000 Facebook ads payed for by false accounts of Russian origin during the 2016 campaign, the total sum of money thus earned by Facebook represents just 0.1% of its daily revenue. In a similar manner, BuzzFeed may have reckoned that the 20 most circulated fake news in 2016 generated 8.711.000 user actions in Facebook between August 1st and election day in November, yet some perspective is needed: as Facebook had 1500 million active users at the time, if we assign just one daily action related to fake news to every user, the result is that such 20 top stories produced just 0.006 of total user actions. This is hardly a massive contagion, even though the transmission of infection diseases has been proposed as a model for understanding the propagation of fake news among digitally connected humans (see Kucharski, 2016). Moreover, it is consistent with statistical findings that confirm how the media ecosystem is still dominated by traditional sources (Zuckerman et al., 2017). Most information still comes from legacy media, digital outlets that have consolidated an audience in the last decade included.
By no means does this involve that fake news is a marginal or irrelevant phenomenon. Apart from its indirect effect on the formation of public opinion, to be discussed later, there is the risk that the phenomenon still grows in the future. Only 14 percent of Americans called social media their most important source of information in the 2016 election, but that proportion is surely to increase and this makes for a worrying prospect if we take into account that half of those who recalled seeing fake news believed it (see Hunt and Gentzkow, 2017). All the more reason, then, to identify the drivers that are behind the phenomenon.

Fake news: a multicausal explanation
Fake news cannot be explained as the result of a single factor, but as the combined outcome of a number of them. Some operate in the long run, while some others appear more recently. One of them is indispensable for the others to produce any effect, namely the digital technology that allows for an easier production and faster dissemination of fake news. But digitization would not cause such effects on its own in the absence of epistemological or political drivers. What follows, then, is a brief account of factors that should always be seen as operating together and never in isolation from each other.

Epistemological factors: the erosion of truth
Fake news points to a wider phenomenon: the erosion of truthfulness in liberal societies. The rationale goes like this: as the collective faith in a single truth has waned, people are more ready to believe that the latter is context-dependent, so that different worldviews will produce different «truths» whose relative merits cannot be decided upon. This cultural development is captured by the term «post-truth», which in turn is related to the postmodern turn that discredits grand narratives and fixed meanings. As D'Ancona has argued, the emphasis on the social construction of meaning, irony, and fragmentation has lead to the belief that reality is now so elusive and our perspectives as individuals and groups so divergent that it is no longer meaningful to speak of, or seek, the truth (D'Ancona, 2017: 98). Thus the emergence of «tribal epistemologies» that correspond to how discrete groups in conflict with each other see the world -there is no such thing as an «overlapping truth» that covers them all (see Roberts, 2017). Moreover, liberal critics of mainstream media have spent decades denouncing the manipulations of capitalist media. Yet if mainstream media coverage is presented as a fabrication by media scholars, what can be said of Trump when he makes the same point? McNair summarizes this reappropriation: the old left critique of «dominant ideology» (and its variants) in the practice of mainstream liberal journalism has been co-opted by the populist movement and its alt-right media cheerleaders as they seek to destabilise long-established norms of truth, objectivity and accuracy in news and in the process replace what they see as exhausted elites of both right and left» (McNair ,2018: 29).
The practical applications of this manoeuvre have in turn been explored by McIntyre (2018), showing how the campaigns against evolutionary theory and climate change have taken advantage of postmodernist arguments about the social contamination of science. So-called «right-wing postmodernism» use doubts about objectivity and truth to assert that all truth claims are politicized and thus can be reduced to ideological statements. Now, a fabrication can be recognized as such and thus rejected, but the dividing line between them becomes more tenuous in this new cultural context. In other words, nobody would dispute that there are indeed true facts that can be truthfully reported, and yet something of the openness associated to the interpretation of meanings has pervaded factual discourses -hence the abused idea of the «narrative», i.e. a way of inserting facts into a story that gives them a particular meaning. This is hardly an inhospitable environment for fake news.

Technological factors: the impact of digitization.
Technology is key to understand the production and propagation of fake news. It has become cheaper and easier to fabricate them, as well as incorporating them into the flow of information -taking advantage in the process of the very speed and seeming disorder of the digitized public sphere. Once the fake is inside the public sphere, it can circulate quickly. On the one hand, there is people: sharers who in good faith disseminate information that they take as truthful. But on the other hand there are bots that automatically propagate fake news, as well as people who know it to be fake but take an interest in its circulation. Together with innocent citizens, then, it makes sense to talk of «underground markets» and «public opinion manipulation campaigns» (Gu et al., 2017: 66). As far as bots are concerned, Lazer et al. (2018Lazer et al. ( : 1095 reckon that between 9 and 15% of active Twitter accounts are bots, whereas Facebook calculates that as many as 60 million bots infest its platform. On their part, Shao et al. (2018) suggest -after analyzing 14 million messages on Twitter during the 2016 US election-that bots play indeed a key role in spreading fake news, especially in the early phases of viral claims, as they usually target influential users that are vulnerable and can lend further credibility to particular pieces of misinformation.
Interestingly, digitization involve a partial regression to communitarian ways of news consumption. As Pettegree (2014) has shown, our medieval ancestors suffered even greater problems of corroboration in the face of dubios information whose truthfulness could not be esily «fact-checked». Thus the credibility that enjoyed information conveyed by people who was personally trustworhty -be it a friend or a reliable messenger. This tradition could not possibly be maintained in mass markets of news information, but has resurfaced with digital peer-to-peer communication. News are not produced within the community, but they are received and discussed within them. People who share an ideological viewpoint will be more ready to believe a piece of news that is endorsed by other members of the group. This has made itself clear with the case of private messaging systems such as Whatsapp, where news circulate outside the public gaze: 55.000 billion daily messages among 3300 million users worldwide. It has been reported that Catalan separatists, as well as Brazillian president Jair Bolzonaro relied heavily on them during their respective political campaigns (see Terrassa, 2017;The Economist, 2018). Technology, therefore, creates the echo chamber with the help of algorithms that select the information we receive on the basis of our previously expressed preferences (see Pariser, 2012). This idea is supported by the so-called «silo effect» that describes how Internet users tend to inhabit networked communities where all members belong to the same moral tribe (see Sunstein, 2008;Reese et al., 2007).
However, the qualities of such chamber must be rightly understood. Despite the insistence on selective exposure and the massive fragmentation of the public, there is no empirical support for the idea that people are now less exposed to different views or news sources. On the contrary, the audience for news websites is highly concentrated and politically extreme sites gets little traffic (Gentzkow and Shapiro, 2011). Most news consumers come into regular contact with stories, images, and comentaries that differ from their own (see Webster, 2014: 111). Granted, people who are deeply engaged with politics will naturally form communities of mutual support and interpretation, a connection that is facilitated by technology and explains the rise of the «political fan» (see Sandvoss, 2013). Coming across contrarian information, however, is less relevant than what is actually done with it. In this regard, deliberative thinkers have reasons to be dismayed, since it looks as if encountering other opinions and views pushes us to reject them more explicitly, reinforcing our own ideological identification (see Bail et al., 2018). Rather than the echo chamber, what technology creates are the conditions for disagreement and polarization. Unsurprisingly, online conversations are less polite and asymmetrical in sites where anonymity is allowed (see Halpern and Gibbs, 2013), as opposed to what happens in online groups where deliberative rules are enforced (Strandberg, Himmelroos and Grönlund, 2017).
That polarization could be avoided in groups where deliberative norms are enforced does not bear much importance in the digital public sphere, as the latter is precisely defined by the absence of any norm. As McNair points out, the «discursive anarchy» that reigns online increases the urgency of «reaching some shared understanding of what we mean when we talk of news being faked, as opposed to merely objectionable» (McNair ,2018: 13). In such context, fake news may not be the main factor driving polarization -despite contributing to it. It is just an additional expression of moral tribalism and deliberative neutralization (see Haidt, 2012). But it is not less than that either.

Demographic factors: the role of digital illiteracy
A technology-related factor that begins to appear as relevant in the analysis of fake news consumption has to do with age. Guess et al. (2019) have convincingly shown that there is a strong age effect in sharing activity: on average users over 65 shared nearly seven times as many articles from fake news domains as the youngest age group. This is a significant gap, especially if we consider that sharing fakes is a rare activity -much rarer than sharing as such. They conclude that «we are challenged to view demographic traits not as controls to be ignored but as central explanatory factors above and beyond the constructs standard in the literature (Guess et al., 2019: 12). Again, it is hard to know how much of this behavior is motivated by the lack of digital literacy as opposed to debilitated cognitive faculties -answering empirically to this question will require new, sophisticated tools. In the meantime, it should be incorporated as an intriguing hypothesis that bears importance in a time when a large cohort enters into retirement across Western societies while technological and cultural change elicit new fears.

Economic factors: changes in journalism
Another technology-related factor concerns economic changes in the media sector. Not all fake news is politically motivated. On the contrary, it can be comercially propelled: news outlets have proliferated and they -as much as traditional media that has gone online-need consumer attention. Hence the «clickbait» strategy: exciting such attention by offering consumers a sensational or shocking content. This comercial behavior is attuned to the logic of media itself, which centers on the new and thus the sensational (see Luhmann, 1996). Interestingly, fringe political messages share this logic too: they are usually presented in the form of provocations against political correctness and the liberal consensus -they are, if you may, political baits. But extreme political views can produce biased journalism without it leading necessarily to fakes -as demonstrated by American alt-right outlet Infowars. McNair presses this point: «Opinion journalism cannot be «fake news» unless there is deliberate use of information which the author knows to be fake» (McNair, 2018: 63).
However, there is no such thing as a clear-cut distinction between objective and biased journalism. The ensuing epistemological confusion is reinforced by another economic factor connected to the digital revolution: barriers to entry in the media industry have dropped dramatically. It is now easier to set up a website or monetize web content through advertising. As a consequence, there is greater competition for readers and this also explains the different forms of clickbaiting and the formal similarities between fakes and baits. Yet the effect of this crowded field for fake news is ambiguous: while it is easier to imitate respectable news outlets in order to lend crebibility to a fabricated story, reputational concerns will discourage real ones from knowingly reporting false stories (see Hunt & Gentzkow, 2014). Honest truth-seeking is still prized by consumers.

Affective factors: the emotional response to information
Fake news that feels right can be validated as truthful by those who are willing to reinforce their beliefs. This hypothesis summarizes the affective driver behind fake news. In itself, it does not refer to a new phenomenon: psychological biases and emotional distorsions have always existed. However, the affective turn and the rise of neurosciences have combined to produce a novel conceptualization of the limits to political rationality. Hence the description of the political brain as an «emotional brain» (Westen, 2009) or the claim that cognition is always «hot», i.e. operates under the influence of emotional pressures (see Marcus and Neumann, 2007). According to psychologist Daniel Kahneman (2011), people decide in two separate ways -one is quick and emotional, the other is slow and rational. In order to make use of the second system, individuals have to make a conscious effort to counteract habits and biases. The so-called «confirmation bias» is key to understanding how humans process information: we seek to confirm our beliefs while rejecting opposite views. Such bias seems to be physiologically pleasurable, as the body releases dopamine when we come across information or opinions in agreement with our views (see Kolbert, 2017). To this we should add the well-known effects of groupthink, be them contemplated from the viewpoint of social psychology (see Sunstein and Hastie, 2008) or even psychoanalisis (see Bion, 1989). On this lecture, sharing views with others serves the purpose of maintaining a collective identity that in turn sustains our personal identity.
Fake news does benefit from this push, as we will feel driven to accept as real news fabricated stories that are aligned with our worldview or political ideology. As Sunstein (2010Sunstein ( , 2018 has remarked, rumour and falsities gain traction in a digiticized public sphere because they fit the previous beliefs of those who receive and spread them. This also says something about democracy and truth: it is not that truth does not matter anymore, but rather that each social group defends its own truth as the only valid one. In other words: «post-truth is not so much a claim that truth does not exist as that facts are subordinate to our political point of view» (McIntyre, 2018: 11). This obviously applies to fake news, which benefit from the same affective response.

Political factors: the rise of populism
At least since the economic crash that started back in 2008, political populism has been on the rise across Western democracies. Populism contributes to the spread of fake news in several ways. An obvious one is related to the anti-establishment quality of the populist discourse (see Müller, 2016). As it creates a divide between a virtuous people and a morally corrupt establishment, populism encourages mistrust of inherited narratives and expert views -journalists included. Truth should not be found in the mainstream media, but in those movements and outlets that dare to challenge the liberal consensus and «speak truth to power». Populism fosters «oppositional forms of political rhetoric that can be directed against or in tactical collaboration with media» (Higgins, 2008: 2). Again, there is the need to distinguish between invented news stories that aims to be taken as genuine thanks to the deprofessionalisation of news gathering (see Wardle, 2017) and the distribution of contestable but legitimate material that supports political contestation (Corner, 2017). Yet it makes sense to expect that the populist rejection of mainstream media will lend credibility to alternative sources of information no matter how obscure they are, thus increasing the odds that at least some fabricated news gets a welcoming response. Now, it is useful to conceive populism not as an ideology but as a political style (see Moffit and Torney, 2014). As a political style, populism employs the antagonism people/elite, exhibits "bad manners", and stages a crisis or a threat to acquire public support (Moffit, 2016: 45). This approach takes into account the medialization of contemporary life and highlights the performative aspects of populism, including its aesthetic and dramatic qualities. On this account, populists leaders either create or disturb the public's subjectivity through a number of discursive and non-verbal strategies, ranging from provocation to impersonation and using as many communicative channels as are avalaible. What interests me here is how this political style is related to the realm of affects. Populist leaders make use of simplistic and politically incorrect language, a «tabloid style» (Canovan, 1981) that privileges «bad manners» as an expression of defiance against the system. They make a conscious effort to isolate its supporters, creating echo chambers from which established parties and media outlets are seen with suspicion. The populist public is thus an affective public that filters information through a feeling of mistrust (see Papacharissi, 2015). Social networks are key for such distribution of affects.
Finally, there is what semiotician Massimo Leone (2019) has called «cognitive populism». It refers to how examples acquire a rhetorical force in the populist narrative that ends up substituting conceptual thinking -putting the «token» in the place of the «type». As Leone argues, meaning cannot be explained without accounting for the human capacity to activate the switch between general representation and particular occurrences. Yet we are witnessing an strenghtening of the example that is related to broader cultural trends. This includes the perversion of exemplification in the political discourse. The force of singularity is thus reaffirmed against generalization, leading to a «violent desintermediation» that continuously strengthen the «thirst for singularity» of individuals and groups. Fake news are relevant in this respect. As Müller (2019) has shown, they support the narrative of populism with the rhetorical force of the example (or token). A particular story, such as those which are routinely «shared» in digital platforms, is meant to demonstrate by itself the truthfulness of a whole narrative. The closing of a local factory can thus be presented as evidence of America's decline, whereas stories about subsidized migrant families would show that migrants predate national social security systems. These stories can be true, but they can also be fabricated -and believed.

Democracy, fakes and the politics of untruth
Shocking political events and the rise of fake news have produced a narrative of crisis about liberal democracy in which hysteria has often prevailed over analysis. Causal links have been established between the digitization of the public sphere and the rise of populism -a plausible hypothesis that overlooks the fact that populism pre-exists the Internet. However, this concern has led to an important conversation about the nature of truth and its relation to liberal democracy. Social psychology and affect theory have made valuable contributions to it, the result being a more sophisticated account of the way in which «truth» is produced at both the individual and the collective level. At the same time, a useful separation between facts and their meanings has come to the fore.
It is under this perspective that the significance of fake news should be considered. On the one hand, it matters because it puts into circulation invented or distorted facts. Following Arendt, this undermines the basis upon which the discussion about meanings is to take place (see Arendt, 2006). Fake news is certainly a small portion of the total amount of news and its effects on individuals cannot be easily discerned (see Neuman, 2016: 4). Yet its effect may be indirect, i.e. its combination with other contemporary phenomena can contribute to the erosion of public trust. Digitization is one of them, as social networks significantly increases the reach and range of fake news. But this effect also applies to biased information and fringe views -they can all be produced and spread more easily than ever. The result is that fakes and fringe views share space with the mainstream, so that the boundaries that used to separate them are blurring. Moreover, fakes are parasitic on standard news outlets, as they benefit from them while simultaneously undermining their credibility (Lazer et al., 2018(Lazer et al., : 1094. The integrity of the public sphere is thus affected, as the citizen can encounter more difficulties in sorting out honest information. McNair puts it this way: Alas, the collateral damage inflicted by the deployment of fake news is borne by the average citizen, the non-aligned and the neutral, by those who genuinely wish to know the truth and understand the world around them. It is they who are left confused and frustrated by news media they seem no longer to be able to trust, but upon which they continue to rely as democratic resources and information sources» (McNair, 2018: 13). This problem is reinforced by the populist contamination of the political discourse, as well as by the strategical use that parties and politicians make of fakes as an accusation against uncomfortable information. Donald Trump regularly tweets that mainstream media such as The New York Times produce fake news, but he is not the only one nor will he be the last to question the credibility of those news stories that harm them. Thus the credibility of the media is further questioned -not a single piece of news looks reliable. Moreover, political parties have partly adopted the language of fake news, as many of their claims consist of false claims. This may be done carefully, for instance making use of the conditional or appealing to the future: during the last electoral campaign, the Spanish vice-president said that the Spanish constitution does not guarantee the equality between men and women, which is not true. Yet the constitution literally speaks of «equality between the sexes», so she might say that her claim was not false.
Yet it is not only politicians and political parties who take advantage of digital technologies in order to generate messages that are ideologically biased or contest the accuracy of factual reports. Journalists and ordinary citizens do the same on social networks -a telling reflection of hyperpolitization. This is especially worrisome as far as journalists are concerned, since they embody an ethos of truthfulness that is eroded wherever they place their neutrality aside and pursue an agenda on behalf of their perceived audience (Higgins, 2008). As much as McNair is right when claiming that there is indeed a difference «between the routine informational game playing of political actors and the systematic deployment of untrue "facts" at any and every opportunity» (McNair, 2018: 78), the difference can cease to matter if citizens become disoriented in a cacophonous public sphere.
Nevertheless, this effect should also be seen as an unavoidable consequence of pluralism. As liberal societies become more individualistic and diverse, the conflicts about meanings and values cannot help but be more acute and salient. After all, disagreement is to be expected from human communication. As Russell Neuman (2016) has argued, the latter tends to be «valenced», as it is imbued with the identities and interests of those who speak. This includes the interpretation of messages in a way that reinforces our identities -the ambivalence that is inherent to language allowing for such recodification. This is hardly new: Hobbes himself was worried about the openness of language and its effects on political order (see Hobbes, 1998). Still, there are reasons to think that pluralism has increased in the last decades. On the one hand, there is the general development of modernity towards a more differentiated and expressive social body -a movement towards «hyperdemocracy» (Welch, 2013). On the other, this trend has received an additional impulse with digitization. With remarkable foresight, Bruce Bimber (1998) advanced twenty years ago that the Internet would intensify group-centered, pluralistic politics. Digital communications have thus been described as resulting in a «chaotic pluralism» where a highly reactive public operates under the influence of social networks (see Margetts et al., 2016: 206).
But which is exactly the danger that this phenomenon, fake news included, entails for liberal democracy? The answer can only be speculative. There is no way to ascertain whether the rise of populism and nationalism is causated by or merely correlated with digitization and the erosion of trust. Only the future will clarify whether liberal societies can adapt to the chaotic pluralism just mentioned or, contrariwise, the latter becomes an unmanageable source of instability. A golden age where truth reigned supreme never existed, though, nor can it ever exist. Tellingly, Hobbes devised a solution that consisted in giving the sovereign commonwealth a last word on the linguistic conflicts arising from society (Hobbes, 1998: 215). But in a liberal society there is not, nor can ever be, such thing. On the other hand, as much as contaminations and distortions cannot be prevented, it can at least be hoped that the public anxiety that has followed the digitization of the public sphere still produces some positive effects. Thus it should be expected that a heightened awareness on the part of the public may eventually act as an antidote to the politics of untruth that fake news have come to embody.

Conclusion.
Jack Tanner, the fictional character created by Robert Altman and Garry Trudeau for their TV series back in 1988, has in the meantime become an apt metaphor for our times. As he moves around real candidates during the Democratic Party primaries, an outer observer should be excused for taking him as a real contender -he looks as credible as the others. And yet his appearance is deceptive: he is a fake. Likewise, news consumers in the digital public sphere often come across fabricated news and, more generally, experience a growing disorientation as to what is a reliable news source. As I have shown, therein lies the significance of fake news, a phenomenon that in itself lacks the direct influence usually attributed to it. This paper has also identified a number of drivers whose combination explains the current spread of fabricated news: the erosion of truth, the digitization of communications, changes in journalism that makes it easier to produce fake news, digital illiteracy on the part of older Internet users, emotional biases in how human beings process information, the rise of populism. It has been suggested that while there is no obvious solution to the problems posed by digitization and post-truth, a heightened awareness of the ambiguities of communication might serve as an antidote to distorted information in democratic societies.