Truth,Deliberative Democracy,and the Virtues of Accuracy: Is Fake News Destroying the Public Sphere?

Introduction

A February 2017 Dan Piraro cartoon captures the anxiety, the danger, and, in some cases, the absurdity of the epistemic uncertainty that has accompanied the rise in fake news (Piraro, 2017). A couple is watching a weather report on television. The weather man says, “A category 4 hurricane is on track to make landfall by morning and residents are advised to evacuate. Those who do not trust the media, of course, can just stay put.” As the cartoon pointedly shows, there are brute facts out there and sometimes having trusted access to those facts can be the difference between life and death. Less dramatically, the cartoon also suggests that citizens generally need facts and accurate knowledge in order to make practical judgments. And, finally, the cartoon illustrates that citizens’ access to facts is often highly mediated. This mediation is unavoidable and not a predicament brought on by new communication technologies, but new communication technologies have facilitated the spread of fake news. The question I ask in this article is, do the epistemic uncertainties connected to fake news represent just another layer to our mediatized access to factual truth, or is fake news and what some have labeled our post-truth predicament a new and more deadly challenge to the epistemic presuppositions of political communication and the public sphere? The answer I develop in this article is that the digitalization of the public sphere is indeed a new challenge but that it does not have to be deadly. Deliberative democracy offers some insight into how to meet that challenge.

I begin by outlining contemporary threats to information acquisition in the public sphere. These are by now familiar and involve the rise of fake news and epistemic uncertainty connected to the digital expansion of the public sphere, especially the growth of social media as a major source of information. Layered on top of that are public and private actors exploiting the uncertainties and technology of the new digital landscape to sow doubt about many of the traditional sources of mediated information. Post-truth is a term that often suggests that the threat to truth is not simply bad actors weaponizing the Internet but also the declining salience of truth or objective facts for citizens, and elite opinion and will-formation. Here, the worry is that people do not really care that much about the truth or they care about other things (partisan causes) more than the truth.

While many commentators have invoked Hannah Arendt to help get a firm conceptual grasp of our epistemic situation, I argue that Arendt is the wrong place to look. Instead, I suggest that deliberative democracy and Jürgen Habermas’ idea of democracy as truth-tracking offer a more helpful framework for assessing and combating the threat of fake news and post-truth attitudes. With the democracy-as-truth-tracking frame in place, I return to the threat assessment and suggest that there is empirical evidence that the democratic public sphere is recalibrating after the epistemic shocks of 2016. This recalibration involves rethinking, reforming, and regulating social media from the point of view of the role they play in political communication, truth-tracking, and public opinion formation. But slowing down, debunking, and exposing fake news will only have an impact if citizens care about, and make efforts to acquire, the truth.

Turning from the structural to the citizen side of the equation, I look at data suggesting that populist/right-wing social media users are more likely to trade in fake news and perhaps more likely to believe fake news, but in any case are more likely to hang on to fake news in the face of debunking and correction than other partisan groups. What can account for this asymmetry in the relationship of citizens to truth in the public sphere? Empirical social sciences favor explanations drawn from social psychology that tie resistance to truth, to cognitive weakness and biases built into our psychological makeup. I suggest that this strategy is problematic for a number of reasons and turn instead to Bernard Williams to shed light on this predicament and more importantly to suggest some strategies to combat creeping post-truth. I argue that Williams’ virtues of accuracy are citizen virtues and a necessary condition for a working public sphere. These virtues are more important than more commonly identified virtues of deliberation, for example, respect and civility. The virtues of accuracy can be facilitated and encouraged through structural and regulatory features in the public sphere. Indeed, the recalibration that we are seeing is due, I argue, to both structural changes and ordinary citizens exercising the virtues of accuracy by becoming more knowledgeable about and aware of the ways that digital communication can mislead. Along with demands on citizens to exercise the virtues of accuracy more vigilantly in the age of fake news, owners, users, designers, and regulators of social media must embrace and recognize the fact that even if this is not how social media were originally envisioned, they now have a political democratic function in the public sphere and we must all act accordingly.

Fake News and the Rise of Epistemic Uncertainty

The circulation of hoaxes, fabrications, and lies in the public sphere is not new. But there appears to be both a spike in the sheer number of such stories circulating and an increase in the politically relevant partisan leaning of the false and misleading information being fed into the public sphere by questionable actors (Allcott and Gentzkow, 2017; Lazer et al., 2018). This phenomenon has come to be referred to very generically as fake news. The two salient features in my use of the term fake news are that, first, the information is false or fabricated, not just biased and partisan, and, second, the information is intentionally introduced to sway people’s political opinions, judgments, and behavior. We are not talking about UFO sightings in Central Park; we are talking about Hillary Clinton selling arms to ISIS. The two cases that have received the most attention have been the Brexit campaign and the US 2016 election (Graves and Cherubini, 2016). But no public sphere is immune to fake news. The structural features that have facilitated the rise and salience of fake news—the growing reliance on social media for news and information—are a global phenomenon, even if they have not been politically weaponized in every context.

The rise of fake news as well as the growing public recognition of the phenomenon leads to epistemic uncertainty; citizens are having a harder time telling fact from fiction and are alarmed and worried about the amount of misleading information circulating online (Matsa and Shearer, 2018; Mitchell et al., 2016, 2019). This growing uncertainty about access to facts is being exploited by strategic actors in a number of ways, for example, by what I call fake fake news. This is the rhetorical and partisan attack on “real” facts and/or fact-based journalism by the accusation of fake news. This and other strategic moves add to the confusion and doubt regarding the trustworthiness of facts and information in the public sphere.

Post-truth is a term difficult to pin down, as it is used to mean many things (Hyvönen, 2018; McIntyre, 2018; Rosenfeld, 2019; Sleat, 2018). This article is not about post-truth and I do not think we are living in a post-truth era or predicament. But there is one sense of post-truth that is useful for my analysis. It points to the subjective or attitudinal side of our epistemic crisis. Fake news is about access to truth in our new media landscape. Post-truth is about citizens’ attitudes toward the truth. How much do they care about the truth? Why do they pass on fake news if they might have reason to think that it is not true? Why aren’t they more up in arms about the vast number of (easily verified) lies that seem to be coming out of the White House? Post-truth in this sense postulates that, with an increase in the number and political effectiveness of falsehoods entering the public sphere, one sees a decrease in the salience of truth or objective facts in opinion and will-formation.

Both the circulation of fake news and the effectiveness of fake fake news claims (perhaps also the rise of post-truth attitudes) are facilitated and amplified by the growing role of social media as purveyors of news and information. Sixty-eight percent of US adults get their news on social media (Matsa and Shearer, 2018). I want to briefly mention three inter-related dimensions of social media that significantly affect the circulation and transmission of information: technology, financial structure, and functional position in civil society.

A recent study completed at MIT looked at the circulation of information, about 126,000 contested news stories, on Twitter over 10 years (Vosoughi et al., 2018). The conclusion was startling and depressing. “Fake news and false rumors reach more people, penetrate deeper into the social network and spread faster than accurate stories” (Meyer, 2018: 2). In short, social media seem to systematically amplify falsehood at the expense of the truth. The authors of the study speculate as to the reasons for this and suggest that human psychology is at the root of this phenomenon: “content that arouses strong emotions spreads further, faster, more deeply and more broadly on Twitter” (quoted in Meyer, 2018: 10). It is, however, the technological features of social media that harness this emotional response and translate it into a viral tweet that can be consumed and passed on by literally hundreds of thousands of users. Legacy media, which can also arouse emotions and can of course also be false, just do not have the power of virality. Social media

have dramatically different structure than previous media technologies. Content can be relayed among users with no significant third party filtering, fact-checking, or editorial judgement. An individual user with no track record or reputation can in some cases reach as many readers as Fox News, CNN, or the New York Times (Allcott and Gentzkow, 2017: 211).

The ease with which fake news enters communication networks and the speed with which it can circulate and be passed on is made possible by social media. The massive amount of information both fictional and real that is available to citizens and the very short time in which they either click or not click, like or not like, repost or not repost are technological features that appear to lean away from considered evaluation of the information.

In addition to the interface between the technology and the human frailty of users is the interface between technology, the actors inserting fake news into the information flow, and the financial model of social media. Here, big data, micro-targeting, bots, and proprietary algorithms dominate the input side of social media. The MIT study speculated about why people pass on false information. But they did not speculate about why people introduce the false information to begin with. In the case of the 2016 election, there appear to have been three main actors: random entrepreneurs who wanted to make money; partisan players in the election drama interested in swaying votes; and malicious third parties, a.k.a., Russia, whose motives are murky but whom one might want to suppose has an interest in a dysfunctional liberal democratic public sphere (Jamieson, 2018). These actors then exploited a system designed to maximize advertising dollars through targeted messaging. The messages now were not about what music to listen to or vacation spot to visit; the messages were about politics, parties, and candidates.

Finally, I want to turn to the functional place of social media in civil society and the public sphere. Lies, rumors, and misinformation have always been a staple of any public sphere where strategic actors vie for power and influence. And legacy media and news outlets have acted both as watchdogs and exposers, as well as complicit perpetrators of these actions. But legacy news and media outlets, even when they fail to do their job, are understood to have a job in a democracy. They are in the business of communicating news and information and it is acknowledged that that job is centrally important to democracy. I am not claiming that legacy media and news outlets have always tried to be “neutral” or “impartial,” nor am I denying that financial interests dominate the decisions of most legacy media owners. The claim is that even when acting as ideological outlets, party organs, or rapacious capitalists, the political function of mediating between citizen and the state or between citizen and citizen was understood by most of the players. ¹

Social media came into the world not as an organ of the free press, not as news and information services, not as the mediator between citizen and state, and not even as a mediator between citizens and citizen, but as a fun way for people to socialize (and, it turns out, a good way to make money). Their epistemic role and function as transmitters of important facts about what is going on the world, including the world of politics, was an unintended development. Perhaps this role could have been predicted, but it is not what they were designed, or apparently prepared, for. As Nathaniel Persily notes, “Reluctantly or not, these platforms are the new intermediary institutions for our present politics. The traditional organizations of political parties and the legacy media will not reemerge for the Internet age in anything like their prior incarnations” (Persily, 2017: 74). Legacy media, for all their faults and pathologies, were primarily designed to communicate the news and information essential for a well-functioning democracy.

Politics and Truth

So the picture I have painted so far is of an old and persistent threat to democracy, namely, misinformation, lies, and strategic manipulation, becoming surreptitiously injected and exponentially amplified via social media. Before assessing the level of the threat, I turn to the question of why we should care about a threat to truth.

It might seem obvious that we are in big trouble if citizens cannot tell fact from fiction. But for a long time, the mainstream study of democracy has thought that citizens were, on one hand, shockingly ill-informed about the issues and, on the other hand, very poor truth seekers anyway (Achen and Bartels, 2016). So, for many, the problem is not so much that fake news is undermining the rational capacities of citizens, it is rather that fake news is a form of elite manipulation exploiting the fundamental epistemic weakness of citizens. Thus, it is the manipulative (and propagandistic) nature of the epistemic threat that is the real danger. And here, Hannah Arendt is often the go-to political theorist to articulate this threat. ² I want to suggest that Arendt, while insightful, is not very helpful.

Arendt wrote a number of powerful essays about truth and politics that took on the propaganda machines of both the East and the West. She was concerned with the way totalitarian governments (but also to some lesser extent liberal democratic governments) sought to rewrite history, manipulate the facts, and undermine truth. Politics cannot go on under these conditions. In a much-quoted passage from “Truth and Politics,” she says that “Freedom of opinion is a farce unless factual information is guaranteed and the facts themselves are not in dispute. In other words, factual truth informs political thought just as much as rational truth informs philosophical thought” (Arendt, 1954: 238). There is no doubt that her essay “Truth and Politics” speaks to our present situation in many ways. This essay gives us insight into why factual truth is both hard and wobbly: “facts have no conclusive reason whatever for being what they are; they could always have been otherwise” (Arendt, 1954: 242). The contingent nature of facts makes it possible to lie. This leads to two types of strategies for those in power who do not like the facts. They can lie and on a grand scale wipe out whole swaths of history with the airbrush of power. But also effective, especially in non-totalitarian situations, is for those in power to turn fact into opinion. There are two sides to every question, after all. We see this happening all the time, not just with questions like climate change but also with fake news itself. In shifting the meaning from false news to partisan news, whether or not something is fake news becomes a matter of opinion. So Arendt appears immediately relevant.

I want to suggest, however, that there is a limit to how helpful Arendt is in navigating our present epistemic situation. It seems to me that she contributes more to the panic about post-truth than to thinking through solutions or ways to address the situation. I hold this opinion for two reasons. First, the essay “Truth and Politics” is not about democracy. It focuses on power and the way power holders exploit the frailty of truth. Democracy and citizens are not mentioned once in this essay. It fails to assess any of the resources and possibilities at the disposal of citizens to push back against the manipulation and distortion of truth. In another much-quoted passage, Arendt sees little hope on that front:

the result of a consistent and total substitution of lies for factual truth is not that the lies will now be accepted as truth, and the truth be defamed as lies, but that the sense by which we take our bearings in the real world—and the category truth vs. falsehood is among the mental means to this end—is being destroyed. And for this trouble there is no remedy (Arendt, 1954: 257).

There is, perhaps, no remedy for the wobbly and frail nature of facts; that is to say, there will always be lies. But movements like Solidarity and individuals like Václav Havel are testament to a stubborn resistance of citizens to “the total substitution of lies for factual truth.” Deliberative democracy, as we will see, takes different perspectives on lies and truth, a perspective from which to develop strategies and practices to guard against attacks on the truth.

There is a second reason why Arendt in the end is not helpful. What she gives with one hand she takes away with the other. While she says factual truth is essential to politics, she also says that all truth—including factual truth—is the enemy of politics. This essay is also famous for arguing that truth itself is despotic. “It may be the nature of the political realm to be at war with truth in all its forms, and hence to the question of why a commitment even to factual truth is felt to be anti-political attitude” (Arendt, 1954: 239). She appears to think that truth claims are conversation stoppers that silence opinion—and politics is all about the clash of opinion (Hyvönen, 2018). But I resist the idea that truth and facts are in some essential sense conversation stoppers. It is not truth that is antithetical to politics (and democracy); it is Arendt’s embrace of a quasi-Platonic idea of truth as singular and final that clashes with politics (Beiner, 2008; Sleat, 2018). If we insist instead on a corrigible and fallible view of all truth claims, redeemed through discourse, then the conversation is never over. This is the view of truth that stands at the center of the deliberative perspective to which I now turn.

Deliberative Democracy and Truth-Tracking

One strand of deliberative democracy has become associated with the mini-public wherein citizens come together in face-to-face designed settings with good information, trained moderators, and procedural norms that promote participant equality in the deliberative and decision-making process. In deliberative mini-publics, there are no social media, no fake news, no fake fake news, and many of the pathologies associated with post-truth are absent or mitigated. Furthermore, there is ample empirical evidence that citizens in these contexts are effective rational problem solvers pursuing evidence-driven solutions to political problems (Dryzek et al., 2019; Gastil et al., 2008). Although there are all sorts of ways in which mini-publics can contribute to and enhance the broad public sphere in mass democracies (Fishkin, 2018; Gastil et al., 2018), they cannot replace that public sphere. Citizens out in the open of mass democracies cannot be shielded from the epistemic onslaughts of fake news and fake fake news the way participants in mini-publics can.

Political communication in the broad public sphere is not anything like a face-to-face mini-public. Nevertheless, there are two ways that the dynamics of well-designed mini-publics can contribute to an understanding of the epistemic potential of the public sphere. ³ The first is that it is important to remember that when citizens step out from a mini-public into open spaces, their cognitive capacities and competencies do not change—the circumstances and forces change. We should be able to leverage some of what we learn about individual-level cognitive capacities in mini-publics to the outside context. Second, it is possible to use the normative structure of face-to-face discourse (equality, openness, accountability) as a systemic framework for conceptualizing democracy more broadly. There are a number of versions of the system view of deliberative democracy (Bächtiger and Parkinson, 2018; Dryzek, 2017; Mansbridge et al., 2012) but I want to take up Habermas’ and in particular his claim that from a discursive point of view, democracy can be seen as having a truth-tracking potential (Curato et al., 2019; Habermas, 2009). What does he mean by this?

Habermas (2005) argues that all truth claims are fallible, corrigible, and redeemed in discourses. Habermas is more interested in the ways, means, and procedures through which we establish and justify truth claims than in defending any particular epistemic theory of truth. ⁴ So he is not interested in justifying, say, a correspondence theory of truth. He is interested in how we establish or know that, say, “p” corresponds to a state of affairs in the real world. For statements like “it is raining,” we can sometimes just point to the corresponding state of affairs. But for anything complicated, we need discourses where we bring forward arguments, evidence, and justifications as well as defend against criticisms and challenges. Those discourses have certain procedural requirements: no coercion, freedom to raise any objection, all inquirers have equal standing, and so on. We can see these procedural requirements clearly in scientific discourse. If pharmaceutical companies introduce threats or bribes into scientific discourse, for example, by paying for results or interfering with who gets tenure and so who is part of the discourse, we would worry that the scientific discourse was no longer truth-tracking. Discourses, then, are how we establish, test, justify, and redeem truth claims, and discourses have some invariable, if very general and abstract, procedural norms that ensure the integrity of the process. Sometimes, we have procedurally independent ways of testing whether the outcome of discourses is tracking truth: for example, some equivalent of pointing to the rain or more often pointing out some other established truths, meaning outcomes of other trusted discourses. But often we also rely on the procedural integrity of the process itself to give us confidence that outcomes have at least a presumption of truth.

Habermas thinks that, looked at structurally, many procedural elements of democracy can be seen as performing a similar function as procedural elements of discourse. ⁵ This is what Habermas means by saying democracy is potentially truth-tracking. Habermas is saying that democracy as a structural and procedural matter has a logic that pushes toward the truth. This seems counter-intuitive in a world where we have the rise of populism, the election of Donald Trump, and the Brexit vote. But he is not saying that majorities or elections are truth-tracking. Nor is he saying that any particular democracy actually tracks the truth as an empirical fact. He is saying that the core features of democracy from a deliberative or discursive perspective—such as open and free debate, equal status of citizens, a porous and critical public sphere, an independent, active, and accessible free press, the circulation of information, and pluralism—can be understood in terms of the conditions needed to test truth claims. These things are the structural and procedural conditions that potentially push toward better answers to our questions and better solutions to our problems. The truth of truth-tracking is not an idea of truth as singular and final. Truth here refers to epistemic robustness. It begins from the idea that there are better and worse answers to many political questions, and democracy is the system that generally and over time is best designed to get better answers (Landemore, 2017). ⁶ Real existing democracies are hugely complex and deeply inadequate instantiations of this idea. Nevertheless, this pragmatic view of the truth-tracking potential of democracy offers a more useful way to analyze our present predicament than Arendt’s essential (and essentialist) clash between truth and politics.

The Public Sphere, Social Media, and Fake News

The claim that democracy has a truth-tracking potential is pitched at a high level of abstraction. To specify what this might mean for fake news, we have to take a closer look at the function of the public sphere in a truth-tracking democracy and then the function of social media within that public sphere.

I use the term public sphere to refer to a sphere of political communication that stands between civil society and the state. Following Habermas, I understand this sphere to be “wild,” unstructured, and contain multiple forms of political communication (Habermas, 2009: 155). Social media are now a vital and permanent part of that public sphere, if for no other reason than that so many people get their basic information about what is going on in the world from social media (including whether it is raining or not). Although individuals, groups, and institutional contributions to the wild and open communication of the public sphere may be very deliberative or flow from a deliberative venue (e.g. mini-publics), taken as whole what goes on in the public sphere cannot be characterized as deliberation and, whatever function social media do have in the system, it is not primarily as a venue for deliberation strictly speaking. ⁷ One reason why the communication in the public sphere is not a deliberation is that it is not attached to decisions in any direct sense. Another reason is that communication is highly mediated and disaggregated. Deliberation does not always have to be literally face-to-face, but the level of mediation and disaggregation (not to mention unruliness and chaos) of communication in the broad public sphere stretches even analogous or metaphorical suggestions that a conversation or deliberation is going on here. Finally, communication is asymmetrical in the sense that the majority are consumers of messages rather than producers of messages. The participatory potential of social media, where anybody can have their tweet go viral, has not changed this fundamental asymmetry in communication. It is no accident that @therealDonaldTrump has almost 59 million followers.

Although the public sphere cannot be understood as a deliberation writ large, it can be understood as performing important functions in the deliberative system. While the public sphere does not perform a decision-making function, in our truth-tracking ideal, it performs a function of articulating and raising the problems, claims, and interests (not to mention hopes and dreams) upon and about which decisions are taken. The public sphere produces public opinion. Public opinion articulates the problems for which a truth-tracking democratic system seeks solutions. Public opinion sets the broad agenda in the deliberative system. Public opinion, although ideally reflecting the real-life concerns, interests, and problems of the public, is constituted through communication. Democracies function properly when they respond to and act on problems, concerns, and issues that confront real people in civil society. But not all ideas, claims, positions, and demands raised in the periphery can be justifiably translated into a legislative agenda, and in any case, all of them need clarification, articulation, and translation in order to make it onto an agenda. So the system must “filter,” to use Habermas’ word, the claims and demands by putting them through a feedback loop of public scrutiny and then ever more rigorous processes of justification (Habermas, 2009: 159). As the political talk moves closer to the center, traditional rules of deliberation and arguing become more rigorous and the conversation looks more and more (or ideally) like a deliberation. The system as a whole is characterized by a significant epistemic division of labor (Chambers, 2017). The more epistemically rigorous function of arguing and deliberating over clear policy options takes place higher up the system ladder. But the content of deliberation, that is, what gets deliberated, comes out of a communicative process that filters, clarifies, and prioritizes claims and narratives in the public sphere. When this is working well, the result is considered public opinion. But it rarely works well and so a critical theory of deliberative democracy is constantly studying and exposing the ways that real existing democratic public spheres derail and distort this process; there are many such ways and fake news is one of them.

There are three ways fake news undermines the process of opinion formation. The first is that as a form of intentional manipulation, it pushes to reverse the communicative flow from periphery to the center. Rather than lived experiences and problems bubbling up into the public sphere to be framed, articulated, and responded to in the public sphere, there is a concerted effort to insert elite agendas into the space of public opinion. This is a problem of all manipulation of which fake news is a subset. The second way fake news disrupts the process of public opinion formation is that, to the extent that citizens believe it and incorporate falsehood into their opinions, it undermines the epistemic credentials of the opinion and so undermines the epistemic process of truth-tracking. Imagine if totally fabricated scientific data make it past the gatekeepers and become part of the scientific discourse. Finally, to the extent that people are aware and worried about the entrance of falsehood into the communicative situation, they distrust public opinion and lose confidence in the democratic process. How damaging has fake news been in these three areas?

Pushing Back on Fake News

The 2016 election in the US as well as Brexit saw unprecedented levels of partisan misinformation disseminated through social media (Allcott and Gentzkow, 2017). This, I have argued, undermines the process of public opinion formation, which in turn weakens the truth-tracking potential of the democratic system. In the intervening years, we have seen an unprecedented level of alarm about this situation as well as an explosion of research and study about what exactly happened, why it happened, and how we can stop it happening. The level of alarm and the frenzy of analysis in the press, in popular and scholarly books, academic journals, and on social media about fake news and post-truth is quite astounding and appears to cross all ideological, methodological, and philosophical lines. This also appears to be the case for ordinary citizens where survey data indicate that both sides of the partisan divide are worried about accuracy and truth (Mitchell et al., 2016, 2019). The level of concern about fake news and post-truth is indirect evidence, it seems to me, that many people still think there is a truth to be had and we need to think carefully about how to access it and make it relevant to politics. Perhaps we are already making headway on that front.

There is evidence that, in the 2018 midterm elections, there was less fake news in information feeds and the fake news that was there reached fewer people (Guess et al., 2018a; Resnick, 2018). It is too early to say for certain what can account for this or how significant are these findings. But they do make sense if we look at the efforts made to address fake news in social media. Social media are here to stay. But there is a new awareness of the dangers and limits of social media. Data that point to older users as being more likely to be susceptible to fake news suggest that digital illiteracy is an important factor in the spread of fake news, and there are growing number of digital literacy programs (Graves and Cherubini, 2016; Guess et al., 2019). New data also suggest that most citizens are not caught in partisan media bubbles or echo chambers (Guess et al., 2018b). There has been an explosion of fact-checking initiatives attached to news organization, sponsored by non-governmental organizations (NGOs), and instituted by social media platforms themselves (Graves and Cherubini, 2016), and there is encouraging survey data that suggest American citizens are more vigilant than they once were about the accuracy and reliability of news and news sources (Mitchell et al., 2019). ⁸ In addition to armies of fact checkers, Facebook and other platforms are experimenting with different ways to combat fake news, including algorithms that are powered by crowdsourced judgments of media trustworthiness (Pennycook and Rand, 2018b). The crowdsourcing experiment is both innovative in the way it uses algorithms to fight fake news and is testament to the fact that people generally care about reliability. That general care about reliability can be harnessed to correct limitations in social media. Facebook, Google, and Twitter have all revamped their financial structure to limit fake news. A Council of Europe report on information disorder and fake news ended with 36 recommendations for policy initiatives to address the new communication reality confronting citizens (Warble and Derakhshen, 2017).

The money and research being poured into looking at technological solutions as well as the tightening up of oversight regarding the authenticity of users have been the result of the acknowledgment on the part of social media platforms of the role these platforms played in the election dramas of 2016. But the acknowledgment of the role is not the same as embracing the responsibility that comes with that role. Social media are the new digital public sphere, and as a public sphere, they perform political functions. Even if they were not designed as news, fact, and information outlets, or as central hubs of political communication, they perform these functions now. There is evidence, however, that social media platforms, while willing to clamp down on questionable content, are not willing to embrace their new political role. This is how I interpreted a 2019 vision statement coming out of Facebook (Yurieff, 2019). Privacy concerns will dominate the next phase of Facebook’s development, not its relationship to democratic citizens. Creating more insulated and protected conversations between friends rather than communication between citizens is the heart of this latest move. Nevertheless, the universal condemnation of fake news, the heightened scrutiny placed on social media’s political function, and growing regulatory pressure (especially in Europe) mean that they are being forced to address many of the political pathologies that have come to light, even if they resist embracing a democratic responsibility.

The problem of fake news points to new challenges brought about by the rise of social media as central actors in the public sphere. We do have a new media landscape and that new media landscape will demand new strategies and initiatives to combat new pathologies. There is evidence, however, that we are coming up with new strategies and initiatives. And there is a growing set of resources for citizens to rely on to help navigate this new media landscape. But in addition to structural resources to restore some basic elements that contribute to democracy’s truth-tracking potential, we also need citizens who care about the truth.

The Asymmetries of Post-Truth

A potential epistemic crisis of the public sphere is made possible by structural features of the Internet, especially the economic and social model of social media, and malicious, partisan, and strategic actors exploiting those structures. While these structural features are present for all users no matter their political or partisan leanings, data suggest that populist/right-wing causes (in the US and Europe anyway) consume, proliferate, and sometimes create more fake news than other groups, and public figures associated with these groups are more likely to challenge the authenticity of more “fact-based stories” (engage in fake fake news) than other groups as well (Khazan, 2017). Scholars trying to unpack the role and impact of fake news in the 2016 US election all agree that there were significantly more pro-Trump fake news stories circulating than pro-Clinton stories (Allcott and Gentzkow, 2017; Guess et al., 2018b). Study after study has confirmed that core Trump voters as well as right-wing enclaves are more likely to visit fake news websites and more likely to pass on fake news and more likely to be resistant to debunking efforts and fact checkers. The Brexit debate also saw a lopsided circulation of fake news, with the more fake news being pro-leave and more leave-side voters circulating and sharing fake news than remain voters (Persily, 2017).

Despite this growing body of research, many still look at the dangers of fake news and post-truth as reflecting a global problem with democracy and democratic citizens. For example, it is not uncommon in commentary about post-truth to cite the Gallup poll that shows a drastic drop in Americans’ level of trust in the media (Hyvönen, 2018: 45; McIntyre, 2018: 86). For Americans, trust that the mass media report the news “fully, accurately, and fairly” went from a high of 72% in 1976 to an all-time low in 2016, coming in at 32% (Swift, 2016). But these numbers are misleading. Broken down by party affiliation, the data indicate that sinking trust in mass media appears to be largely driven by Republicans and to some extent Independents, with Democrats’ trust dropping only very slightly (Swift, 2016). Couching the problem as a global problem affecting all citizens equally also dominates what has become one of the most common frameworks for understanding the fake news phenomenon—psychology (Lazer et al., 2018; Resnick, 2018; The Economist, 2018).

Two overlapping areas of research inform the psychological account. First, there is cognitive-psychology research that suggests that at a deep level we are “motivated reasoners,” and that our motivations are not dominated by an interest in truth but instead by an interest in winning or not wanting to admit that we are wrong (Kahneman, 2013; Lodge and Taber, 2013). Motivated reasoning leads to biases (e.g. confirmation bias, desirability bias, in-group bias, and observational-selection bias) that skew our cognitive process toward what we want to be the case rather than what is actually the case. This research then dovetails with more specifically political studies that look at the causes and sources of partisanship. Here, evidence is mounting that political preferences are driven by identity rather than ideology (Iyengar et al., 2012; Mason, 2016). Identity here means affective group attachment. Citizens join a team, they stick with their team, root for their team, and vote for their team. In-group/out-group identification affects political preferences and voting more than do the policies, positions, or platforms put forward by candidates and parties. Identity, on this view, is an affective attachment impervious to evidence, facts, and reasoning, or at least is resistant to them. On this view, then, people are susceptible to fake news, even easily checked and corrected fake news, because rather than being a rational truth seeker, the human mind is wired to protect and defend its beliefs from reason and truth if those do not support partisan affective identification.

I have three reservations about using these mostly experimental findings in analyzing democratic citizenship in a digital age. First, there is a growing body of research, some of which is focused on fake news, that questions the “motivated reasoner” paradigm by suggesting that rather than essentially poor reasoners, we are lazy reasoners (Pennycook and Rand, 2018a). Citizens are generally capable of employing evidence-driven reasoning when reflexively encouraged to do so and that reasoning often is not dominated by partisan motivation. These experimental data are also confirmed in deliberative mini-publics. This implies that properly constructed cues and messaging in social media news feeds that I described above can trigger the sorts of cognitive effort needed to identify fake news and ignore it even in the presence of partisan bias. This is one of the ways to think about leveraging what we learn about cognitive capacities from mini-publics in the digital public sphere.

My second reservation about this research is that the reliance on psychology often downplays or fails to thematize variation across time and space. I do not want to deny that humans have cognitive weaknesses or that group attachments can be a driving force in the development of political preferences. But these general psychological facts about us only become pathological under certain conditions or in particular times and places. That humans are psychologically capable or even psychologically prone to ignore facts and resist truth does not tell us much about when and where humans can resist these tendencies, when and where these tendencies become toxic or pathological, or why some groups appear more susceptible to these weaknesses. Thus, there is often a failure to address the asymmetry of the circulation of and susceptibility to fake news. This in turn is tied to my third concern. The discussion of the pathologies of fake news is often subsumed under a larger discourse about the competence or lack thereof of democratic citizens (Achen and Bartels, 2016; Brennan, 2017). Fake news and the growing influence of social media is just another chapter in the long story about citizen incompetence. Thus, democracy is to blame for Brexit, Trump, and other populist movements gaining ground in democracies.

The Virtues of Accuracy

The relationship of individuals to truth is certainly structured by the mind, and so psychology can perhaps help to set the limits of the possible (although the science is not there yet), but citizenship—individuals interacting in a political way—is more than a psychological capacity. It also involves ethical and structural dimensions often ignored or downplayed by social and political psychology. Truth-seeking is an ethical political commitment or goal as well as a psychological capacity. Perhaps we are not perfect truth seekers (I am not denying that we are prey to many cognitive biases), but we can do better or worse—if we care. Whether or not we make the effort to correct our beliefs, check our facts, or put truth before interest depends on internal ethical dispositions and external structural conditions. Thus, to the structural conditions outlined above that focus on making facts available (digital literacy, reducing fake news, finding ways to authenticate facts, etc), we add measures to encourage and foster truth seeking. Bernard Williams has interesting things to say about truth seeking that supplement and in some ways challenge the narrative of citizen epistemic incompetence.

Williams argues that truthfulness involves two separate sets of virtues (Williams, 2002). The first and the one that most people focus on is sincerity. Telling the truth involves saying or reporting what one knows or believes to be true. Lying, then, is saying or reporting what one knows or believes to be false. But truthfulness involves more than this. It also involves the virtues of accuracy. If one cares about being truthful then not only does that come with the obligation to only say and report things that one believes to be true but it also involves making a goodwill effort to find out what is true or to acquire true belief. If you care about the truth then “you do the best you can to acquire true beliefs, and what you say reveals what you believe” (Williams, 2002: 11). The acquisition of true belief involves virtues because it often takes effort, as there are many obstacles in the way. Some obstacles are external: the plant you wish to study only grows in a remote part of Uganda or the entry of fake news into one’s news feed requires extra care and effort to check. But there are also internal obstacles to accuracy and one of those is “wishful thinking.”

Like Arendt, Williams sees that the contingent nature of factual truth makes it possible to lie but it also opens the door to wishful thinking. We can wish facts to be otherwise because they could be otherwise; we can have no such wish for axiomatic or mathematical truths. If the money in my pocket does not add up to the price of what I want to purchase, I do not wish that mathematics worked differently, I wish that I had more money. The contingent nature of facts makes it possible to wish that things were otherwise. We are susceptible to wishful thinking and so the virtues of accuracy include “dispositions and strategies for sustaining the defenses of belief against wish, and against one of the products of wish, self-deception” (Williams, 2002: 125).

Thus, to psychology, Williams adds ethics. To be truthful, we need to cultivate both sets of virtues, sincerity and accuracy. We often forget this and focus on the insincerity of telling lies, but it is “at least as important to improve the caution of the potentially deceived as it is to censure the motives of the deceiver” (Williams, 2005: 156). It seems to me that improving the caution of the deceived is a much sounder strategy and one that holds out more hope than transforming the motives of the deceivers. This is why Habermas is more useful than Arendt in approaching fake news. A Habermasian theory of the public sphere and the conditions of truth-tracking focuses on improving the defenses of the potentially deceived.

Accuracy, I would argue, is a citizen virtue. This suggests that, as structural features of the methods of truth transmission change (social media) and as the tactics, strategies, means, and methods of deceivers change (fake news-fake fake news), so must the tactics, strategies, means, and methods of information consumers become updated to the threat. There is ample evidence that many, perhaps most, democratic citizens still care about factual truth and understand the catastrophic consequences of having no trustworthy sources of factual information. They do not always know what to do about it. But there are ways to educate for the new digital age and monitor, regulate, fact check, and debunk fake news as well as call out and expose fake fake news. There is also evidence that the present epistemic crisis has triggered a more robust exercise of the virtues of accuracy as citizens are more aware of the ways they may be duped and misled (Mitchell et al., 2019). Thus, the virtues of accuracy are a potent force for resisting the lies of the powerful.

I want to also suggest that, from the point of view of deliberative democracy and at the level of the individual, the virtues of accuracy are more important than the virtues of respect or civility for a well-functioning democratic public sphere. Democracies need respect in the system, and certain types of incivility in certain places can threaten the very fabric of social order or minimum levels of trust. Toxic trolling and white supremacist propaganda are scourges of the Internet and surely symptoms of deep pathologies in our democratic societies. One does not need to be a deliberative democratic to recognize the harm and danger of this type of behavior. But most citizens do not engage in this behavior. Here, I am interested in the individual-level, everyday use of social media and the connection of that to the democratic public sphere. If ordinary citizens turn away from social media because of toxic content and they turn to other, less toxic sources of information, this is not necessarily a problem for our deliberative system. I am not saying that disrespect is fine. I am saying that because ordinary citizens are not usually engaged in deliberation when they use social media but they are consuming information, the virtues of accuracy are more relevant than those of respect and civility.

The issue that concerns me is that many citizens get the basic information that informs public opinion formation from social media. Public opinion formation is not a face-to-face deliberation where disrespect would surely derail the process. The irony here is that although citizens receiving messages and communicating in the public sphere do not engage in epistemically rigorous deliberation, their epistemic disposition (i.e. virtue of accuracy) is more important to the process than the ethical disposition of respect. A lack of respect, for example, aggressive trolling, can derail any online conversation. But as social media do not primarily perform a directly deliberative function but rather an information function in the process of opinion formation, trolling is less damaging than fake news. Finally, the virtue of accuracy can be a stand-in or proxy for respect. Accuracy implies an epistemic openness even to opinions, statements, and claims made by people you disagree with and might even want to shout at. Thus, while respect is necessary for face-to-face deliberation, and certain levels of public and visible disrespect and incivility can be deeply threatening for a democratic order, caring about truth and taking extra care in verifying the information one is receiving is a more relevant individual disposition in public opinion formation than respect.

Conclusion

Social media have taken on important functions in the public sphere and it is not likely that we will see a reversal of this trend. Furthermore, social media have drastically amplified the potential reach of fake news and misinformation, and this represents new challenges. But there are structural, regulatory, and ethical responses to this new challenge and we are seeing some of those taking shape. I do not have an explanation for the asymmetry we are seeing in the susceptibility to fake news. But this asymmetry is evidence that human psychological and cognitive weakness cannot be the sole or even primary determining factor. Psychology is at most a condition of possibility. This suggests then that the interplay between structural reforms, on one hand, and the virtues of accuracy, on the other hand, can have an impact on the spread of fake news, misinformation, and the susceptibility to the rhetoric of fake fake news. I have argued that deliberative democracy, Jürgen Habermas, and Bernard Williams’ virtues of accuracy are helpful in thinking through the structural reforms as well as citizen dispositions called for in this situation. But a full and adequate response would require that owners, users, designers, and regulators of social media embrace and recognize the fact that, even if this is not how social media were originally envisioned, they now have a political democratic function in the public sphere. This does not mean that Facebook and YouTube need to see themselves as primarily news and information outlets. But it does mean thinking of users as democratic citizens and not simply as private consumers.