The Concept of Epistemic Power

We will now explain the form of power that is our focus of concern, epistemic power.^{Footnote 2} Epistemic power is the power one possesses in one’s capacity as an epistemic agent. Epistemic power is best understood in terms of two related abilities.^{Footnote 3} First, is the ability to influence what others believe, think, or know.^{Footnote 4} A teacher, for example, possesses this ability in relation to her students. The second is the ability to enable and disable others from exerting epistemic influence. This is done by way of believing others (“giving them credence”) or by discrediting them. By trusting and hence validating a person or institution as a proper source, we enable that person or institution to have influence. For example, a teacher might praise a student’s mathematical ability, leading others to believe that student when they provide answers. Here, the teacher has increased that student’s epistemic power. In contrast, by disabling a person or institution we preclude that person or institution from having influence. For example, we might publicly denounce a news outlet as “fake news.” When this public denunciation is successful, the effect will be a decrease in the epistemic power of the news outlet.

To summarize, our account of epistemic power is the following:

Almost everyone possesses some degree of epistemic power, though some have more than others. There are different sources of epistemic power, some of which are entirely legitimate: an articulate person who possesses detailed knowledge of a particular subject may be able to influence what people believe simply by providing clear explanations to others. Institutional positions may also bring epistemic power. Professors, for example, possess epistemic power by virtue of their position as a professor, which grants them a special status as experts on their topic.

Ideally, this institutional source of epistemic power will only be granted to those who deserve it. However, this will not always be the case. Someone could be granted an institutional position that brings epistemic power without possessing the relevant expertise. For example, a corrupt university could appoint an underqualified candidate to a professorship. Similarly, we can imagine cases where someone’s expertise in one area could allow them to bypass the normal procedures by which someone is granted the position of an expert. For example, a leading chemistry professor may use his prestige and profile to make public statements about how to treat cancer that are unrelated to his field of expertise (Ballantyne Reference Ballantyne2019). This professor uses his title and the platform it provides to exert epistemic power, whilst venturing into a scientific field that is not his own area of competence.^{Footnote 5} Of course, people are allowed to give their opinion, but they should ensure that it remains clear when they speak from personal conviction, rather than a position of expertise. The worry with this kind of procedure-skipping is that there is a danger that people will be placed into institutional positions that bring epistemic power without possessing the relevant expertise.^{Footnote 6}

Credibility

One source of epistemic power is perceived credibility. The extent to which someone is perceived to be credible has a major impact on how much epistemic power that person possesses. Our focus will be on credibility excesses, cases where people are perceived to be more credible than they really are. All else being equal, the more credible someone is perceived to be, the more power they have to influence what people think, believe, and know, and to influence who else can exercise epistemic influence. So when someone possesses a credibility excess, they will possess a greater degree of epistemic power than is warranted by their epistemic abilities and expertise.^{Footnote 7}

Do celebrities possess higher levels of perceived credibility than other people? The idea that celebrities are taken as reliable sources on topics for which they lack expertise is certainly familiar and seems to be accepted by a number of commentators. For example, Traub described singer Bono’s position in his political campaigning in the following way: “Bono offered decision makers an implicit bargain: do the right thing, and I’ll say so in public. His currency was not just his fame but his credibility” (2008).

The long-running use of celebrities in advertising and political campaigning might seem to provide further evidence for this link. However, despite the popularity of this technique, the connection between celebrity and credibility is a complex one. In a comprehensive meta-analysis, Knoll and Matthes (Reference Knoll and Matthes2017) found that celebrity endorsements had a positive effect on consumers’ attitudes, but no average effect on their behavioral intentions. The effectiveness of the celebrity endorser appears to depend on a variety of other factors, such as their gender (male celebrities performed better than females), their profession (actors outperformed models, musicians, and TV hosts) and the fit between celebrity and product (2017, 67).

People’s willingness to accept celebrity endorsements is not restricted to trivial matters. In 2013 Angelina Jolie wrote a widely read editorial in the New York Times about her decision to undergo a mastectomy after discovering she faced a high chance of developing breast cancer. Jolie urged women to investigate their own chances of developing breast cancer. An observational study found a 64% rise in American women undergoing the same genetic test in the fifteen days following the article’s publication (Desai and Jena Reference Desai and Jena2016). Further, Freed et al.’s (Reference Freed, Clark, Butchart, Singer and Davis2011) study into who parents view as credible sources of information on the safety of vaccinations found that, while parents were most likely to trust their doctor, 24% of parents gave some credibility to what celebrities had to say on the topic, with 2% giving celebrities a lot of credibility (Freed et al. Reference Freed, Clark, Butchart, Singer and Davis2011, 109). While these numbers may seem low, they do show that there are people who are willing to trust the word of celebrities on important matters like the health of their children.

Relatedly, celebrities have been claimed to have a major influence on political opinion. Garthwaite and Moore’s (Reference Garthwaite and Moore2012, 358) study of the effects of Oprah Winfrey’s endorsement of Obama in the 2008 presidential primary estimated that Winfrey’s intervention was responsible for over one million votes for Obama. Similarly, a study by Jackson and Darrow (Reference Jackson and Darrow2005) into the impact of celebrity political endorsements among Anglophone Canadian youths found that celebrity endorsement influenced the extent to which respondents were willing to endorse certain political opinions. However, as the authors note, the link between celebrity endorsement and political opinion is a complex one, dependent on an interaction between the celebrity, audience, and endorsed opinion (Jackson and Darrow Reference Jackson and Darrow2005, 94). Veer, Becirovic, and Martin’s Reference Veer, Becirovic and Martin2010 study of the effects of celebrity endorsement among UK voters also found these endorsements to be influential but far more so amongst those who spent little time thinking about politics. Harvey’s (2017. chap. 4) research on people’s credibility ratings of celebrities and politicians found that in most cases the ratings between the two groups were indistinguishable. Finally, Becker’s (Reference Becker2010, 112-116) research into the effects of political endorsement found that they were effective with young people (especially women) and Democrats, especially on social rather than economic issues. These results suggest clear evidence for a link, albeit a complex one, between celebrity and perceived credibility in the political domain.

One important factor in the effectiveness of a celebrity endorsement is whether the respondents are fans: as Jackson and Darrow (Reference Jackson and Darrow2005) noted, while some celebrities were capable of influencing their entire sample of Canadian political science students, others were only effective over fans of their music style. This suggests that some celebrities possess a credibility excess only amongst their fans. One way to explain this is that admiring someone in some way can often lead us to admire him or her in other ways as well.^{Footnote 8} Another, complementary explanation is that the fan/celebrity relationship is one of “intimacy at a distance” (Horton and Wohl Reference Horton and Richard Wohl1956, 215; Van Krieken Reference Van Krieken2012, 83), where interaction with celebrities involves coming to know them “in somewhat the same way they know their chosen friends” (Horton and Wohl Reference Horton and Richard Wohl1956, 215). This may lead people to attribute greater credibility to celebrities than they would to unknown strangers.^{Footnote 9Footnote 10}

In summary, there is good reason to think that some celebrities will be subject to credibility excesses in the political arena among certain groups of people, especially those who admire and are familiar with them, and those with little interest in politics. The processes by which these credibility excesses are constructed will take place without the conscious awareness of the person making the credibility judgement. However, the celebrity/credibility connection is complex, and highly context dependent. In the next section, we will explain a more general connection between celebrity and epistemic power.

Attention

Another source of epistemic power is attention. Celebrities are often the focus of people’s attention. This forms a core part of the economic value of celebrity, especially in an information-rich society. Of course, celebrities often also have economic value stemming from their talents. However, the economic value that celebrities possess by virtue of being a celebrity comes from their ability to capture attention. Importantly, the attention people pay to celebrities distinguishes them from ordinary people: celebrities are paid attention to in a way that ordinary people are not (Van Krieken Reference Van Krieken2012, 60). They are, in Marshall’s words, “a voice above others” (1997, xlviii). Moreover, the attention paid to celebrities often outweighs that paid to politicians or to those with relevant expertise. As the former United States senator Arlen Specter said, “When Hollywood speaks, the world listens. Sometimes when Washington speaks the world snoozes” (cited in Ross Reference Ross2011, 5).

Politicians also pay attention to celebrities in a way they do not to other people. Activist Han Shan explained his movement’s decision to utilize celebrity campaigners in the following way: “if a celebrity backs your cause, then immediately the public and the media and even policy-makers will sit up and pay attention” (cited in Duncombe Reference Duncombe2007). A number of high-profile celebrities have leveraged this fact. For example, when Natalie Portman became concerned about the Israel-Palestine conflict, she contacted Queen Rania of Jordan to see if she could help (Traub Reference Traub2008). Non-celebrities are unlikely to be able to command the attention of those in power in the same way.

Being the focus of attention is a source of significant epistemic power for celebrities. This point may not be immediately obvious, as being the focus of attention does not necessarily enhance credibility. For example, American terrorist Ted Kaczynski (aka the Unabomber) captured attention for his anti-technology manifesto Industrial Society and Its Future by stating that he would continue his bombing campaign until it was published in a major newspaper. However, the attention this brought to Kaczynski’s views is unlikely to have increased many people’s perceptions of his credibility.

Nevertheless, this case shows why attention is a source of epistemic power distinct from credibility. Attention by itself is unlikely to change people’s credibility perceptions about ideas they are already aware of. But it can alert people to ideas they had not heard of before. For those disposed to find Kaczynski’s ideas appealing, exposure to these views may well lead them to start accepting these ideas. While attention may not boost the speaker’s credibility, it does provide a platform for one’s testimony to be heard. This is crucial, as it does not matter how credible someone would be perceived if he or she does not have a means of being heard in the first place. Having a platform provides new opportunities to influence what people think, believe, and know. In other words, attention can increase one’s epistemic power without one’s perceived credibility increasing.

Being the focus of attention can also work to redirect attention towards others, with celebrities using interviews or social media to encourage people to pay attention to other people’s ideas. For example, during a 2016 appearance on the Today Show, Robert De Niro encouraged people to watch the controversial anti-vaccination film Vaxxed. In this case, De Niro did not use his platform to get his own testimony accepted. Rather, he directed attention towards the testimony of people who would not have received anywhere near the same levels of attention without his assistance. In doing so he enabled others to exert epistemic influence, and hence increased their epistemic power.

Another way in which being the focus of attention can be a source of epistemic power is through providing people with the tools to set political agendas. Agenda setting is the process by which the media shape which issues are deemed to be of political importance. McCombs and Shaw (Reference McCombs and Shaw1972) developed their theory of agenda setting based on their studies of US presidential elections, arguing that:

According to McCombs and Shaw, the media do not simply reflect political reality, they also structure and shape it. The more attention the media gives to a certain issue, the more likely people will consider it important. As Cohen says, the media “may not be successful much of the time in telling people what to think, but it is stunningly successful in telling its readers what to think about” (Reference Cohen1963, 13).^{Footnote 11} The ability to shape the political agenda is therefore also a source of epistemic power for those to whom attention is paid. It enables them to influence the topics people think about and, as a result, what people come to believe and know.

One way celebrities set the political agenda is through traditional forms of media. As Harvey’s (2017, 65) research shows, a celebrity’s intervention in a political cause can raise the media visibility of that cause. Additionally, the rise of social media has allowed celebrities to reach large numbers without having to navigate the gatekeepers of the traditional media. Social media also allows celebrities and politicians to influence the shape of the coverage of the news media (Parmelee and Bichard Reference Parmelee and Bichard.2011; Parmelee Reference Parmelee2014). A simple indicator of social media’s power is that in 2017 the estimated combined print and online daily newspaper circulation in the United States was 31 million on weekdays and 34 million on Sundays (Pew Research Center 2018). In comparison, Donald Trump has 57.2 million Twitter followers, at the time of writing (January 13, 2019). Trump utilized this platform to great effect in his 2016 presidential election campaign, gaining far more attention than his opponent Hillary Clinton. Trump’s posts were retweeted 6,000 times on average, compared to Clinton’s average of 1,500 (Pew Research Center 2016). As several commentators observed, this provided Trump with extraordinary power to set the news agenda. In the words of Trump’s former press secretary Sean Spicer, “Whatever he tweets is going to drive the news” (cited in Borchers Reference Borchers2017). A large-scale analysis of mainstream and social media coverage of the 2016 election campaign highlighted the success of Trump’s agenda setting. Although the study found that coverage of both candidates was mostly negative, it “largely followed Donald Trump’s agenda” (Faris et al. Reference Faris, Roberts, Etling, Bourassa, Zuckerman and Benkler2017, 5).

In summary, attention is a source of epistemic power for celebrities. First, attracting large amounts of attention provides a platform for their testimony to be heard. Second, it allows them to redirect people’s attention to other sources of testimony. Third, it allows them to set the political agenda. Of course, celebrities are not the only people to garner attention in the political arena. Leading politicians receive attention even if they were not previously celebrities, and economists, sociologists, psephologists, geographers, and other experts all receive attention on issues relevant to their expertise. This attention is a source of epistemic power as well. And these experts may even achieve a level of celebrity themselves. The difference is that, in these cases, epistemic power is granted to people with some legitimate claim to it. In the case of politicians, it is focused on those making the decisions and asking them to account for their actions; in the case of experts, attention is given in light of their expertise. This is not the case with celebrities, who receive attention even when they may lack any relevant expertise.

Democratic Theory

We begin by considering how celebrity epistemic power may interact with two standard approaches to democratic theory: deliberative and epistemic democracy. These terms denote a vast array of distinct views, and we will not attempt here to examine epistemic power alongside each. Nor will we survey other influential theories such as pluralism, agonism, or elitism.^{Footnote 12} Rather, we will identify some core commitments of selected standard approaches, broadly conceived, and analyze how they survey the importance of celebrity epistemic power. Moreover, we will argue that celebrity epistemic power creates challenges for grounding the legitimacy of democracy by appealing to these democratic ideals. We end with a brief discussion of an emerging democratic theory, namely Green’s (Reference Green2010) plebiscitary democracy. While plebiscitary democracy may avoid some problems facing the traditional accounts, it is perhaps even more vulnerable to some of the concerns associated with the workings of celebrity epistemic power. We provide here a brief introduction to deliberative and epistemic theories of democracy.

Deliberative models of democracy ground the legitimacy of democratic rule in public debate (i.e., the public giving and receiving of reasons) informed by citizens’ mutual respect for others as moral equals. In contrast to aggregative models, which interpret moral equality as the requirement of giving one equally weighted vote to each citizen, deliberative democracy pairs equal formal influence with deliberative ideals that are intended to protect equal opportunity to participate in and influence political outcomes. Protecting this equal opportunity to influence is taken to have both procedural and substantive requirements. Procedurally, participants are required to advance claims solely on the force of argument and reason, and not trade on power inequalities (Knight and Johnson Reference Knight, Johnson, Bohman and Rehg1997, 288). In Habermas’s words, “no force but that of the better argument is exercised” (Reference Habermas and McCarthy1975, 108). Substantively, asymmetries of power or resources should not unfairly (dis)advantage participants in deliberation; participants should not be prevented from deliberating as a result of a lack of resources or power and should not be subject to coercion.

In response to accusations that these ideals are unrealistic and over-demanding, some have developed a related, systemic model of deliberative democracy that aims to retain the original deliberative values while allowing for more expansive notions of deliberative processes. Systemic models acknowledge a variety of interactions and associations as forms of deliberation which can be treated as component parts of a larger deliberative system.^{Footnote 13} Such systems are characterized by the distribution of deliberation across components, and a division of deliberative labor. As a result, no individual component or interaction must itself meet the full and rigorous criteria for deliberative democracy.

Pure deliberative models ground the value and legitimacy of democracy in its relation to these deliberative ideals, independently of the epistemic value of the outcome. Conversely, epistemic models of democracy suggest that in addition to protecting equal opportunity to influence political outcomes (i.e., its procedural fairness), democracy’s legitimacy lies in part in its epistemic value. Estlund (2003, 69), for instance, argues that the typical dichotomy between procedural fairness and (epistemic) quality of outcomes is a false dilemma (2003, 69).^{Footnote 14} Rather, appropriately rigorous democratic procedures can be described as having (modest) epistemic value, tending to “track truth” to a sufficient extent to ground democratic legitimacy.

Recent work in this area has revisited the core epistemological assumptions, developing further criteria for legitimate democracy.^{Footnote 15} Peter (Reference Peter2007) for instance, offers an alternative (pure) account of Estlund’s Reference Estlund and Christiano2003 rationalist epistemic proceduralism. Rather than evaluating democratic outcomes against a procedure-independent standard of correctness, we can think of knowledge as the product of a deliberative process and focus our efforts on defining “the conditions that knowledge producing practices should satisfy” (2007, 341).^{Footnote 16} These conditions support epistemic fairness, and can include “criteria that regulate public deliberation at the fact-gathering and analysis stages of the processes of policy formation” (2007, 344).^{Footnote 17}

Similarly, Young (Reference Young2000) characterizes democracy as a form of practical reasoning aimed at finding correct and just solutions. While different social groups have access to different forms of situated knowledge, it is a mistake to think that enjoying “epistemic privilege”—in the form of having first-hand knowledge of the experience of subordination—yields epistemic power. Marginalized groups may lack perceived credibility or suffer from deficits of attention. Young therefore advocates inclusion as a normative ideal of democratic communication. Inclusion describes not only formal opportunity to participate in deliberation, but also effective opportunity to influence others and to contest dominant notions of the common good that underpin deliberations. This effective opportunity could be read as enabling appropriate degrees of perceived credibility and paying attention to all participants.

In addition to deliberative or epistemic accounts of democratic legitimacy, there are mixed theories that draw on the insights of both in various ways. Some theorists propose drawing on a deliberative stage to produce a considered set of options that are then voted on. This model is defended on epistemic grounds, wherein incorporating a deliberative stage in the generation and vetting of possibilities produces a list of options that is better able to capitalize on the epistemic power of aggregative methods.^{Footnote 18} Landemore (Reference Landemore2013) for instance, advocates an inclusive form of deliberation characterized by cognitive diversity.^{Footnote 19} When combined with a “stopping rule,” such deliberation is best able to produce intelligent answers to collective problems.

Analysis of Democratic Theory and Celebrity Epistemic Power

We turn next to the question of how celebrity epistemic power creates challenges for those wishing to ground the legitimacy of democracy by appealing to these democratic ideals. We introduce three points of critique, namely the illegitimate ability to set political agendas, the relative immunity from familiar checks on power, and the undermining of democracy’s epistemic qualities as a result of the disconnect between expertise and epistemic power. Importantly, we do not argue that celebrity epistemic power always works to undermine democratic goals or epistemic value. Particular uses of this power could compensate for systemic failings elsewhere. Moreover, we do not suggest that celebrity epistemic power poses the biggest problem for democracy as compared to more familiar threats tied to inequalities of wealth, education, or leisure. We aim primarily to highlight existing democracies’ vulnerability to unchecked epistemic power, which we have argued that celebrities tend to enjoy, and therewith to show important lacunae in how deliberative and epistemic theories of democracy understand attention and influence.

Agenda Setting

Our first point concerns the potential for celebrity epistemic power to influence the subject of deliberation: the greater attention that celebrities attract could translate into greater attention for the causes or issues they find important, thus influencing the public agenda.

Agenda setting can refer to the setting of the objects of public interest (i.e., what people think about), but also the attributes or framing of a debate (i.e., the terms in which people think about the subject). Much research has demonstrated the influence of the news media on both levels of the public agenda and thus as constituting an important subject of democratic concern (McCombs Reference McCombs2002, 3). As McCombs writes:

While celebrity power to attract and direct attention arguably does not (yet) rival that of the news media, it can function in similar ways to influence the objects of public debate and their attributes. That is, it can transmit the salience of a subject, or of particular attributes of the subject, through volume of attention.

For example, it is likely that the celebrity profile of individuals subjected to sexual harassment in Hollywood played a large role in the recent rise in awareness of the issue of sexual harassment reflected in the #metoo and #timesup movements. A pervasive but previously seldom discussed issue has in a short time come to occupy a more substantive position on the political agenda. As Brooke notes, “[#metoo] brought the idea of sexual harassment and assault into the public consciousness” (in Seales Reference Seales2018). It has also been described as challenging the dominant “crisis” framing of sexual assault and related policies (Abrams Reference Abrams2018). Placing sexual harassment on the political agenda is a positive achievement and could be described as an instance of celebrity epistemic power working to supplement the credibility and attention deficits that have wrongly kept this issue off the agenda for so long. But while the phrase “me too” was first used to highlight the prevalence of sexual assault by activist Tarana Burke in 2007, it only became the focus of widespread attention when high-profile celebrities such as Alyssa Milano began to use it.^{Footnote 20} The way this issue has (finally) been brought to the fore, i.e., via the attention-directing dimension of celebrity epistemic power, seems contrary to central deliberative ideals of fairness and equal opportunity to influence. As such, this example illustrates how celebrity epistemic power has influenced the public agenda in a way that, at least on one level, weakens the legitimacy of existing democracy and shows just how far away existing democracies are from meeting deliberative ideals.

Granted, neither the news media nor celebrities fully determine the public agenda. The influence of news media partially depends on an individual’s level of media reliance or need for orientation (McCombs Reference McCombs2002). However, as Yang and Stone (Reference Yang and Stone2003) show, the media agenda is also able to indirectly influence others via their interpersonal connections with those who do follow the news media (described as “opinion leaders”). This raises two concerns: first, to the extent a celebrity echoes the news media’s agenda, they may function as an opinion leader in ways that further the news media’s agenda-setting power, and second, the special attention-directing power of celebrities likely amplifies this indirect means of agenda-setting.

Such methods of agenda-setting are problematic for grounding a democracy’s legitimacy in the deliberative ideals of procedural fairness and the equal standing of citizens. Procedurally, viewpoints may be excluded not as a matter of having been formally rejected, but because our limited capacity for attention has been devoted to some issues over others, resulting in a de facto setting of the agenda. Regarding standing, it seems celebrities likely enjoy a higher effective status in deliberative contexts (as opinion leaders) by virtue of their epistemic power. Consequently, any democracy seeking to appeal to ideals of procedural fairness and equal standing to ground their legitimacy ought to find a way of addressing these power imbalances.

Substantive Equality and Checks on Power

Plausibly, we ought to add epistemic power to the list of inequalities deliberative democrats ought to be concerned with. In particular, inequalities in epistemic power seem likely to threaten the (substantive) requirement that participants have equal opportunities to influence others. While the simple existence of an inequality in power is not itself a threat to a deliberative democracy’s legitimacy, substantive controls are required to prevent this inequality leading to unequal opportunities to influence deliberation. And such controls do seem to be lacking with celebrity epistemic power currently lying outside the familiar mechanisms for checking power. First, unlike with economic power, there are no formal mechanisms aimed at regulating or monitoring celebrity contributions to political campaigns whether in the form of explicit endorsements or indirect political commentary via the media (social or traditional). Second, there are problems with relying on informal mechanisms such as public debate as a way of identifying and contesting power. Such measures may prove impotent in the case of celebrity epistemic power. To the extent that an issue becomes affiliated with a celebrity, any debate on that issue is likely to be linked to its celebrity advocate. Thus, attempts to challenge celebrity statements may work to draw further attention to the celebrity. In other words, the attention tied to publicity can bolster, rather than check, a source of celebrities’ epistemic power. Recognition of a similar point has led some journalists to advocate ignoring rather than contesting provocative comments (Borchers Reference Borchers2017).

Additionally, the deliberative process should be driven by engagement with, and reflection on, reasons rather than the imposition of power. Thus, the political process ought to be protected from coercive or manipulative influences, but also from unequal opportunities to influence that derive from wealth, educational, or other social inequalities.^{Footnote 21} Deliberative theorists are keen then to distinguish between these improper sources of unequal influence and proper sources tied to the persuasive power of good argument and evidence.^{Footnote 22} However, as our analysis of the sources of epistemic power shows, these forces cannot be neatly separated—the very force of reasons can be tied to workings of epistemic power. In such cases, non-epistemic reasons (such as being the subject of greater attention) influence the extent to which some reasons are deemed legitimate or some reasoners are thought credible. To return to the previous #metoo example, it is largely because of the collective action of a handful of celebrities that widespread sexual harassment and assault in the workplace—and particularly the entertainment industry—has come to be taken seriously. While not speaking on issues outside their expertise, the actors who advanced the #metoo movement plausibly enjoyed greater perceived credibility than women generally receive when reporting sexual harassment. Moreover, their capacity for persuasion was not tied to greater abilities to reason or present evidence (proper sources of influence), nor to economic or educational resources. And, such influence seems of a different type than the forms of manipulation enabled by economic or educational inequalities. Instead, recognition of celebrity epistemic power reveals the ready notion of autonomous influence, understood as deliberation resulting from “free reflection on what one takes to be relevant reasons” (Kolodny Reference Kolodny2014, 310), may be too simplistic. Again, in this case, epistemic power may have helped advance democratic goals (i.e., social equality), though this will not always be the case.^{Footnote 23} Thus, to the extent that deliberative democratic theory is ignorant of this form of informal influence, it will be ill-equipped to determine when no force but that of the better argument is in play, and hence when deliberative outcomes are legitimate. There are two points to draw from this. First, the prospects for deliberative democratic theory are tied in part to whether its proponents can offer a satisfying account of how to distinguish and control for this influence. Second, the phenomenon of celebrity epistemic power presents a real challenge to anyone seeking to ground the legitimacy of any actual democracy by appealing to deliberative democratic theory.

Defenders of systemic accounts of deliberative democracy might reply that there are ways in which democracies could respond to the challenges posed by celebrity epistemic power. First, even if celebrity epistemic power risks violating deliberative ideals in traditional deliberative interactions (as argued earlier), the systemic approach suggests that other institutions or associations could counter this influence and thus neutralize its anti-deliberative effects. Moreover, this countering could work as an opportunity to assert valuable epistemic norms and practices and thus drive the revaluing of an important deliberative virtue.^{Footnote 24} Finally, by not relying on specific static institutions and defined checks and balances to define deliberative democracy, the systemic approach might be less vulnerable to the criticism that celebrity epistemic power largely bypasses, or falls outside, the traditional means of protecting against undue influence. While not accounted for explicitly, a systemic approach retains the possibility of identifying countervailing practices or norms that serve to check celebrity epistemic power where it threatens to undermine the deliberative system.

It may be true that any anti-deliberative effects of celebrity epistemic power may be neutralized by other systemic components, or even work to prompt the re-assertion of valuable deliberative norms. There is evidence, for example, for thinking celebrity influence in the domains of health and science has prompted experts to publicly respond to pseudo-science and to emphasize the value of scientific literacy.^{Footnote 25} However, while scientists have increased engagement and publicly supported valued epistemic norms, it is unclear whether the scientific experts are winning. This is despite the well-established and widely endorsed role of science in society. This gives reason for worry that the deliberative systems in the political domain, which are less well established or endorsed, may be comparatively vulnerable.

It is also important to consider what kinds of practices or institutions could counter any anti-deliberative effects of celebrity epistemic power, as it is not obvious how this could be done. For example, even if relevant experts respond to celebrity comments, their response could work to establish an issue as a subject of public debate. Responding to issues selected by celebrities enables celebrities to set the agenda. These worries relate to the ability to identify and manage (anti)deliberative components within a democratic system, though again, in theory, it may be possible to achieve this. The task then for proponents of systemic accounts of deliberative democracy is to acknowledge the ways in which celebrity epistemic power could interact with other components of a deliberative system and to conceptualize its potential anti-deliberative effects.

Celebrity epistemic power thus poses a potential practical challenge for democracies striving to achieve the ideals of deliberative democratic theory by insulating the deliberative process from power inequalities, particularly as active efforts to contest celebrity statements could increase their power. In addition, deliberative theories of democracy need more robust notions of autonomous influence that can account for the ways in which attention and perceived credibility affect the force of reasons.

Decoupling

One important feature of celebrity epistemic power is that it frequently extends beyond the celebrity’s legitimate area of expertise. That is, celebrities are routinely perceived as credible on subjects outside their relevant expertise. The problem is not simply that some individuals have greater epistemic power than others, as this may be entirely legitimate. Scientists, for instance, may be due more epistemic power than anti-vaxxers on the subject of vaccine safety, due to their relevant expertise and credentials. The problem is that epistemic power should be tied to relevant expertise if democracy is to be defended on its ability to generate better choices. The decoupling of epistemic power and expertise then creates additional problems for those seeking to ground the legitimacy of a democracy by appealing to epistemic theories of democracy.

First, the gap between expertise and epistemic power can threaten a democracy’s potential for truth-tracking. For rational epistemic proceduralists such as Estlund (Reference Estlund2007), to the extent that celebrity epistemic power undermines the tendency of democratic procedures to track the truth, the epistemic value of democracy could be undermined. And there seems particular cause for worry on this issue given the nature of celebrity epistemic power. Granted, this is not to say that celebrities will necessarily lead us further from truth, but to the extent that celebrity epistemic power is not tied to systems of credentials, training, and expertise which are meant to aid truth-tracking, it is more difficult for someone like Estlund to claim that the democratic procedure is sufficiently epistemically rigorous to support claims of legitimacy.^{Footnote 26}

This point is also worrisome for pure epistemic proceduralists like Peter (Reference Peter2007) and Young (Reference Young2000). On Peter’s account, democracy contributes to the production of knowledge. Again, strict equality of epistemic power is not necessarily desirable, as we can also recognize the distinct expertise that social groups have in certain areas. Catala (Reference Catala2015) makes a similar point when she argues that dominant majorities ought to epistemically trust a minority group on a relevant issue by virtue of their expertise. That is, on some issues, minority groups are epistemically privileged such that special, rather than equal, status is due (2015, 431-434).^{Footnote 27} Thus, pure epistemic proceduralists also have reason to be concerned that epistemic power aligns with relevant expertise—namely epistemic privilege tied to one’s social group. As noted in the earlier discussion concerning rationalist accounts, celebrity epistemic power is not the product of, nor is it reliably connected to, relevant forms of expertise, and thus is worrisome. To the extent that pure epistemic proceduralists characterize democracy as a site for the production of knowledge, unchecked celebrity epistemic power that contradicts or crowds-out relevant expertise threatens democracy at a more fundamental level.^{Footnote 28} Moreover, Turner warns that despite the growing appearance of celebrity diversity, “celebrity still remains a systemically hierarchical and exclusive category” largely under the control of media industries driven “to operate this economy in the service of their own interests” (Reference Turner2006, 157). Alongside the obvious issue here regarding substantive equality, there is also the concern that such dominant interests translate to a homogenization of perspectives and interpretations that undermine the epistemic value of Landemore’s deliberative component.

Finally, this credibility-creep could be problematic for those seeking to appeal to systemic accounts of democratic legitimacy. Mansbridge et al. (Reference Mansbridge, Bohman, Chambers, Christiano, Fung, Parkinson, Thompson, Warren, Parkinson and Mansbridge2012) describe the related defect of decoupling as a possible threat to a deliberative system. Decoupling occurs when “parts of the system become decoupled from one another in the sense that good reasons arising from one part fail to penetrate the others” (2012, 23), such as when politicians fail to acknowledge scientific data on climate change. The spreading features of celebrity epistemic power represent the complementary defect (encoupling), where reasons arising from one part of the system fail to remain isolated. Celebrities are taken to have expertise even when they lack credentials and stand outside the peer communities that help support legitimate expert authority. In many ways, celebrity epistemic power is distinctly contrary to the systemic components that support healthy expert contribution. Thus, given the potential for celebrity epistemic power to lessen the tendency of democratic procedures to track truth, or to undermine democracy’s contribution to the production of knowledge, it should be of concern to those aiming to ground democracy’s legitimacy in its epistemic value.

Plebiscitary Democracy

We end this analysis with a discussion of an emerging model of democracy that reduces the emphasis on the traditional deliberative and epistemic values noted earlier, namely Jeffrey Green’s (Reference Green2010) theory of plebiscitary democracy. Green argues that theorists must acknowledge that very few citizens in today’s existing democracies are empowered to influence politics via expression of preferences—that is, via their voice—but are better described as spectators in the democratic system. He rejects the view, however, that this condition is a lamentable one of powerlessness. Rather, via recognition of three central shifts in focus he develops an alternative, ocular paradigm that identifies a new source and object of democratic power. First, rather than aiming to shape legislation, plebiscitary democracy is concerned with leaders; that is, the character and conduct of political leaders are of primary rather than secondary instrumental importance. Second, rather than exercising power via speech, citizens are empowered via the gaze: a “hierarchical form of visualization that inspects, observes, and achieves surveillance” (2010, 9). Finally, the third shift replaces the critical ideal of autonomy with that of candor. This is the institutional (rather than personal) requirement that leaders should not have full control over their appearances and presentation. The people ought to secure “control of the means of publicity” (2010, 14) such that public appearances are spontaneous rather than scripted, rehearsed, or managed and are thus “worthy of being watched” (2010,15).

Green’s emphasis on the gaze as the new site of people’s democratic power partly aligns with Keane’s (Reference Keane2009) account of monitorial democracy. For Keane, contemporary democracy is being redefined as a system of complementary monitory measures that work to identify and challenge power, rather than a system of popular self-rule (2009, 706). Enabled by a media revolution that has produced a communicative abundance, “every nook and cranny of power becomes the potential target of ‘publicity’ and ‘public exposure’” (2009, 740). These monitoring bodies can be thought of as providing material for Green’s plebiscitary gaze and thus as a source of empowerment.

As plebiscitary democracy does not aim to produce decisions, it avoids many of the concerns that applied to epistemic and deliberative models of democracy. However, celebrity epistemic power raises problems for this view, too. First, while Green identifies the spectator gaze as a site of empowerment, our analysis of celebrity epistemic power and attention suggests this relationship is more fraught. Increasing scrutiny and surveillance of celebrities can be cast as increasing the attention directed towards them and thus increasing their epistemic power.^{Footnote 29} But, Green might object, this result is less damaging for his model, as the related agenda-setting effects of attention that worried us on other accounts of democracy lose their bite on the plebiscitary model. In the world of issueless politics that characterizes plebiscitary democracy, the object is not to decide on options (i.e., engage with an agenda), but to observe the leader’s character and personality (2010, 45).

However, this only deepens the concern. Today’s celebrities are well equipped to manage their image and thus control the public’s knowledge, not (only) of issues and facts, but of the celebrity’s personality and individual qualities. Many have ready access to public relations consultants and media experts, and moreover, as private individuals, celebrities can retreat behind calls for privacy to more fully control information about themselves. The private world not only lacks some of the powerful monitoring capacities of the political realm—for example, being subject to Freedom of Information requests, or being required to disclose sponsorship sources—it also enjoys certain protections explicitly against such monitoring.^{Footnote 30} Furthermore, celebrity publicity does not seem subject to Green’s ideal of candor. While the public may value authenticity in celebrities, this does not translate to an institutional call for the public to control the means of celebrity publicity. Without this institutional dimension, celebrities retain control and are able to create the appearance of authenticity (Click, Lee, and Holladay Reference Click, Lee and Willson Holladay2013). Thus, on plebiscitary democracy, celebrity epistemic power to control the public’s knowledge of them as individuals is increased (via attention) while avoiding the institutional requirement for candor and the checking potential of the gaze.

In sum, celebrity epistemic power poses a variety of challenges to a democracy’s ability to realize its normative ideals. This power threatens provision of equal opportunities to influence political outcomes in both procedural and substantive respects. The source of celebrity epistemic power is of particular concern given its potential to come apart from relevant expertise and hence derail productive and epistemically fruitful deliberative practice, including the knowledge-production capacities of democracy. Moreover, the deliberative and epistemic theories of democracy examined here were shown to be lacking in their respective abilities to address the workings of celebrity epistemic power.

The analysis given here represents only an initial examination of the relationship between celebrity epistemic power and selected models of democracy. There are additional models that merit analysis, and further interactions within our selected theories to consider. One such issue concerns how celebrity epistemic power relates to the problem of “constituting the demos” (Goodin Reference Goodin2007, 40). In particular, celebrities can, and do, engage in international deliberation: they are free to endorse, comment, and exercise influence at a global level. Examples include American actors Susan Sarandon, Lena Dunham, and Mark Ruffalo expressing support for Jeremy Corbyn in the UK’s 2017 snap election and British singer Adele endorsing Hillary Clinton. Celebrity epistemic power, then, could result in illegitimate forms of deliberative influence that undermine democratic legitimacy. Finally, more work is needed to determine the relative magnitude of the threats posed by celebrity epistemic power to existing democracies as compared to other phenomena such as economic inequality, apathy, corporate interest lobbying, or educational deficits. This work will require substantive empirical research to determine the extent of celebrity epistemic influence, particularly when interacting with these other phenomena.

While we have identified a number of serious concerns, we do not take any of these points as proving fatal to the legitimacy of existing democracies or to the related democratic theory. Rather, we take this discussion to have revealed gaps that could help inform future developments in democratic theory and practice. In particular, if Green and Keane are correct that contemporary democracy is increasingly characterized by spectatorship and a politics of publicity, democratic theorists ought to pay special attention to celebrity epistemic power.

Responsibility

One avenue of inquiry is to revisit the subject of responsibility and the importance of robust moral-epistemic norms for checking celebrity epistemic power. We have identified three immediate domains for which such norms seem essential. The first includes the role of media companies in facilitating and shaping what gets attention and thus (among other things) in setting the political agenda. There is a case to be made that those companies have a moral and social responsibility to oversee who gets to use their services—in particular, they have responsibility to prohibit the spreading of falsehoods or “fake news.” Some companies seem to have begun to recognize this responsibility. For example, conspiracy theorist Alex Jones was banned from YouTube, Apple, and Facebook in August 2018 (Newton Reference Newton2018).

Our point is not that Jones is a celebrity (in fact he may not be by our definition) but rather to point to this as an example of a case where media companies have recognized and responded to a misuse of epistemic power. Given this, it seems reasonable for these companies to also consider the ways in which celebrities might misuse their epistemic power. Of course, this leaves the question of what counts as a misuse. While some cases are clear, we need a more substantive standard to determine what counts as a misuse of epistemic power. A future project for political scientists, alongside social epistemologists, is to investigate whether such a standard can be determined.

The second domain of responsibility concerns celebrities’ responsibility for the use of their epistemic power. Their responsibilities would involve, most minimally, doing adequate research before they speak out about a topic. This might involve contacting relevant experts and checking the credentials of those experts with other experts, such as when Bono consulted Professor Jeffrey Sachs at Harvard to learn more about developing world economics (Harvey Reference Harvey2017, 46).

The final domain of responsibility concerns ordinary citizens. Citizens effectively give celebrities their epistemic power by paying attention to them and, as Van Krieken notes (2012, 73), citizens can stop paying attention at any point. This relation resembles the role of citizen voters in empowering government that Beerbohm (Reference Beerbohm2012) discusses. For Beerbohm, citizens share responsibility for their country’s actions by virtue of their relation as voters. He builds on this point to develop an ethics of justified political belief and to map the contours of this responsibility. Beerbohm’s views could theoretically be expanded to acknowledge that ordinary citizens endow celebrities with their epistemic power and are thus partly responsible for its existence and exercise. His ethical program could similarly be extended to citizen responsibility for directing attention.

In order to combat the unwarranted epistemic power of celebrities, we must therefore hope that citizens, celebrities, or media companies, upon seeing their responsibility in creating this power, will help to control its use. One avenue for investigation, then, is for political scientists to garner insights from contemporary epistemology about which intellectual and moral virtues should be developed in citizens, celebrities, and media companies.^{Footnote 31}

Whether this can be done by citizens, celebrities, or media companies individually or must be done collectively is a question that requires further investigation. However, even if an adequate solution is found, it seems somewhat naïve to think that each or any of these groups will in fact do this. This presents a serious challenge for political scientists and highlights the limits of this line of research.

Celebrity Creation

A different direction to pursue concerns the creation of celebrity. It is important to recognize that celebrities are not inevitable; as products of profit-driven media industries, political ideologies, and technological developments, celebrity is a contingent category that could theoretically be dissolved. The history of the American film industry provides a nice example of the contingent nature of celebrity culture. In the beginning of the twentieth century The Edison Trust monopolized the industry and prevented actors from gaining celebrity status by banning artistic credits and denying licenses for films featuring star-making performances (Wu Reference Wu2010, 61-73). Thus, one approach is to challenge the forces that produce celebrities and eliminate (or reduce) the very existence of celebrity epistemic power. Conversely, one could advocate exploiting this category rather than eliminating it. Perhaps the better strategy is to promote the creation of celebrity experts and diverse celebrity voices in order to counter the effects of an exclusive celebrity population and to help foster informed deliberation. Physics professor Lawrence Krauss advocates a related strategy, claiming that scientists have a responsibility to inform the public and are at least as worthy of helping to steer public debate as other celebrity figures (2015, 32). Others, including Keane, worry that media saturation can negatively affect democracy by overloading citizens with demands for their attention. “Profusion breeds confusion,” thinks Keane, and can lead to escapism, disaffection, or cynicism (2009, 746-7).

Ethics and Politics of Attention

Implicit in our study is a focus on attention and publicity. In this way, our analysis shares similarities with Green’s insights concerning the role of publicity (or candor) in democratic theorizing. Green gave candor—that is, the people’s institutional control over the means of publicity—a central place in his democratic theory. He also emphasized the people’s gaze as an alternative source of power in place of their voice. Future research in democratic theory could build on these insights in light of the connections we have drawn between epistemic power, attention, and credibility. Green’s concern could be framed as one regarding the allocation of epistemic power—namely, for the reallocation of the power to influence belief, to citizens via control over the means of publicity. But we ought also be concerned with how epistemic power is used—that is, how citizens use their epistemic power to allocate credence. Moreover, these distributive and ethical questions might be fruitfully framed in terms of attention as well as in terms of publicity. For instance, concerns about opportunities to influence deliberation via speech could be supplemented with opportunities to be the focus of attention. One possible outcome of this study is to consider publicity and attention as a subject of justice.

A related line of research concerns the ideal of equal opportunity for influencing political outcomes. This ideal presumed a clean distinction between the legitimate influence of reason and evidence and the illegitimate influence of economic or educational advantage. But as discussed in the section on substantive equality and checks on power, celebrity epistemic power represents a form of influence that resists description in these terms. Thus, theorists could revisit the ideal and explore ways to address forms of influence that trade on attention and perceived credibility.

Expertise and Power

A final direction to pursue would be to investigate exactly what kind of link should exist between expertise and epistemic power. It is important for political scientists and political theorists to acknowledge both the epistemic power celebrities have and that this power is often unaccompanied by relevant forms of expertise. Thus, the ideal distribution of epistemic power would be one in which someone’s epistemic power regarding a particular subject was proportional to their expertise. However, such an approach runs the risk of leading to a society governed by technocratic experts. An ongoing project in political science and political theory is to investigate how the seemingly anti-democratic phenomenon of expertise can be legitimately incorporated into democratic decision-making. For example, Alfred Moore (Reference Moore2017) recently defended a form of “critical elitism” in which expert authority can be made democratically legitimate by being made subject to public contestation. Our discussion can contribute to future theorizing on this topic, first by emphasizing the importance of epistemic power, and second by raising the question of whether there is a legitimate role for those whose epistemic power is disconnected from expertise, such as with celebrities, to play in this process.