Democratic Equality Beyond Deliberation

Abstract

Many theorists believe ideals of deliberation realize democratic equality by ensuring each speaker’s influence is commensurate with the strength of the reasons they give. This article argues that view is incomplete. When the frames, interpretations, and concepts speakers bring into debate disproportionately reflect the perspectives of socially advantaged groups, they contribute to hermeneutic injustices that impede marginalized citizens’ voice even under the most favorable deliberative rules. In response, I describe the role institutions of mass communication play in shaping the epistemic resources available to deliberators and show how structural biases toward advantaged groups within those institutions generate shared understandings that undermine deliberation’s egalitarian potential. Pursuing democratic equality, then, requires not only the fair exchange of reasons but also proactive efforts to identify and correct institutional mechanisms of hermeneutic injustice.

INTRODUCTION

Many political theorists argue that democracy’s attractiveness (Anderson 1999; Beitz 1989; Kolodny 2014; Wilson 2019) and authority (Christiano 2008; Viehoff 2014) stem from citizens’ equal say in both formal decision-making and the informal exchange of ideas surrounding it. “The liberties protected by the principle of participation,” John Rawls (1999, 197–8) writes, “lose much of their value whenever those who have greater private means are permitted to use their advantages to control the course of public debate.” Unequal voice in informal discourses provides “good reason to think” silenced groups’ interests are less likely “to be advanced by the collective decision‐making of the society,” even if they have equal voting power (Christiano 2008, 201–2).

But what, then, does democratic equality require of public debate? Among the most influential answers is the one put forward by deliberative democrats: informal equality is best captured by ideals of deliberative reason-giving in which each speaker’s influence is commensurate with the strength of the reasons they give (Gutmann and Thompson 1996; 2004; Habermas 1996; Johnson and Knight 1997). Many egalitarian theorists have followed deliberativists’ lead, describing democratic equality’s communicative dimension in terms of the relations of mutual respect and equal consideration secured by idealized deliberative exchange (Anderson 1999, 313; Christiano 2008; Scheffler 2015; Wilson 2019). In turn, the qualities of ideal deliberation offer evaluative standards for assessing how well communication systems promote egalitarian democracy generally (e.g., Bächtiger and Parkinson 2019; Cohen and Fung 2021; Mansbridge et al. 2012).

This article argues that focusing on deliberation, whether as a practice or an organizing principle, tells an incomplete story about what equality requires in an institutionally dense public sphere. Such a view rests on an inflated sense of deliberation’s independence from the background conditions of speech. When citizens meet as deliberators, they arrive already equipped with common frames, concepts, and interpretations inherited from their shared communicative context that structure how they think and talk about politics. By prioritizing how persons with formed conceptual vocabularies communicate, theorists threaten to lose sight of how those vocabularies come about and their consequences for citizens’ capacity to contribute to public discourse as equals.

Sustaining communicative relations among equals, I argue, depends not only on the fair exchange of reasons but also on common epistemic resources that reflect persons’ equal participation and consideration as democratic citizens. Building on ideas from feminist epistemology (Collins 2000; Davis 2018; Dotson 2014; Fricker 2007; 2013; 2015; Harding 1991; Medina 2013; Nelson 1993; Pohlhaus 2012), I show that when marginalized groups are systematically denied equal voice in the processes by which shared understandings are created and circulated, the understandings those processes yield come to disproportionately reflect the perspectives, interests, and prejudices of dominantly situated groups. Such a tilted hermeneutic playing field impedes marginalized citizens’ equal influence and consideration within both interpersonal deliberation and wider systems of communication.

Theorizing democratic equality’s communicative dimensions requires going beyond deliberation to capture persons’ participation as equals in the processes of collective knowledge-making that precede and structure our communicative acts. I argue that in a polity of hundreds of millions of strangers, the epistemic resources strangers share are primarily products not of interpersonal discussion but of the impersonal political–institutional context to which they belong. Digitally enabled circuits of mass communication like political representation, news journalism, social media, and popular culture bind diverse and far-flung publics together through common flows of frames and interpretations. ¹ At best, these circuits function as intermediaries that empower publics to introduce grassroots understandings into wider public consciousness (e.g., Barvosa 2018; Woodly 2015; 2022) and enable individuals to incorporate a maximally wide range of relevant perspectives into commonsense schemas. ² But when such institutions systematically prioritize advantaged groups, they not only “shift the balance of reasons” (Mansbridge et al. 2012, 24) but also corrupt how persons make and consider reasons in the first place: crime reporting targeting white suburbanites perpetuates stereotypes of Black criminality (Dixon 2008), political campaigns adopt color-blind frames to avoid alienating white moderates (Gillion 2016; Stephens-Dougan 2020), and “rags-to-riches” entertainment refracts views of poverty as a personal failure (Kim 2023).

Without a well-ordered institutional foundation capable of securing background conditions of hermeneutic justice, deliberation cannot sustain the egalitarian virtues theorists so often ascribe to. In response, I recast democracy’s communications infrastructure as a critical site of democratic equality and defend equal epistemic participation as a regulative ideal for guiding its organization and operation. While the scale and hierarchical nature of most mass media make direct public participation difficult, equal epistemic participation works primarily in the negative. It pushes us to correct mechanisms responsible for institutions’ unequal attention to and consideration of socially disadvantaged groups in the production of shared understandings. As an example, I diagnose how market pressures incentivize mainstream news media to pander to white suburban audiences with equality-undermining frames and show how an ideal of equal epistemic participation supports reforms to shift news to a not-for-profit basis.

Theorists have long noted how, in Iris Marion Young’s (2000, 71; see also Bohman 2000, 114–23) words, “the assumptions, experiences, and values of some members of the polity dominate the discourse and that of others is misunderstood, devalued, or reconstructed to fit the dominant paradigms.” But their responses have primarily focused on recalibrating deliberation to counter the downstream effects of hermeneutic inequality, rather than challenging the structural barriers at its source. In centering those structures, my argument advances recent efforts to theorize the limits of deliberation’s “foundational claims to moral and political equality” in unjust social-institutional contexts (Drake 2023, 94; see also Liveriero 2020; Scudder 2023). And like that work, it deepens egalitarian theories of deliberation by better identifying their epistemic preconditions, as well as the practices and institutions that support them. ³

I begin by canvassing egalitarian defenses of deliberation and then show how reliance on flawed epistemic resources undermines groups’ treatment as equals within discussion. The next section outlines how flawed resources emerge through the interaction of top–down institutional and bottom–up interpersonal communication and defends an ideal of equal epistemic participation mediated through these institutional processes. I conclude by showing how deliberation itself cannot correct flawed resources without addressing their pre-deliberative sources.

DELIBERATIVE EQUALITY AND ITS LIMITS

The most popular approach to theorizing equality in the public sphere has been to define it in terms of deliberative practices that reflect participants’ status as equals or as an ideal of public discourse that approximates those practices in aggregate. How should we talk to one another to ensure each is given due respect? Deliberators aspire to persuade through the exchange of reasons while holding fast to norms of justification, fallibility, and rational updating. Convincing our peers by the light of their own reason shows respect for their autonomy as rational agents, such that deliberation manifests “mutual respect among free and equal citizens” separate from its outcomes (Gutmann and Thompson 2004, 23; see also Cohen 2003, 23; Mansbridge et al. 2012). In this sense, deliberation is motivated by our interest in being respected as equals and constituted as a practice by speaking in ways that reflect what it means to do so.

Equality in deliberation is often theorized as the “equal opportunity of access to political influence” (Johnson and Knight 1997, 280). Deliberative norms ensure this equality of opportunity by minimizing “judgment-independent” inequalities between speakers stemming from unequal material resources or arbitrary prejudices while still permitting unequal influence based on the quality of arguments (Kolodny 2014, 332–6). An alternate view emphasizes deliberation’s relational dimension. Deliberators express equal respect for one another’s capacity “to render authoritative judgments as to how to organize and regulate all citizens’ common life” (Wilson 2019, 49, 170). These conceptions of deliberative equality help theorists navigate the details of deliberative practices, such as whether deliberators must appeal to shared values (Christiano 2008, 190; Cohen 2003, 23; Gutmann and Thompson 1996; 2004, 3; Habermas 1996, 166; Rawls 2001, 27, 34–5; Young 2000, 51) or whether speakers may be treated differently based on their relative social power (Beauvais and Bächtiger 2016; Wilson 2019, 160–1). ⁴

Nonetheless, these conceptions realize equality in roughly similar ways. Deliberation rests on the idea that as equal members of a decision-making community, citizens are entitled to contribute to discussion about issues that affect them jointly. ⁵ This requires that deliberation not only be formally open to all but also that deliberators’ contributions receive fair consideration and uptake based on their merits, with norms of deliberation calibrated to that end (Young 2000, 54–5). Socially dominant groups’ biases toward dispassionate reason-giving, for example, silence speakers who express claims through emotional appeals, nonlinear argument, or personal anecdote. Thus, deliberation must be sufficiently inclusive of diverse modes of expression, especially those associated with disadvantaged groups (37–40). Similarly, prejudices against an identity-group’s trustworthiness, what Miranda Fricker calls testimonial injustice, threaten a distinctly epistemic form of exclusion by preventing members of that group from having their testimony fairly considered. Norms of testimonial fairness are needed to preempt and “neutralize the impact of prejudice” by delineating the proper procedure for giving and considering claims (Fricker 2007, 92; see also Bohman 2000, 33–4; Christiano 2008, 58–62, 200–1; Peter 2021). Deliberation, in this sense, aims to secure the influence and consideration citizens are due as political equals by minimizing the contingencies of ordinary speech that so often lead to unequal voice. ⁶

This notion of how deliberation realizes equality has remained remarkably consistent even as deliberative theory has shifted focus away from deliberative practices toward deliberative systems. In the deliberative systems view, interlocking sites of imperfectly deliberative communication can work together to promote “mutual respect among citizens” and an “inclusive political process on terms of equality” in ways analogous to interpersonal deliberation (Mansbridge et al. 2012, 11–2). Recent work theorizes the conceptual ideal of “deliberativeness” realized by deliberative systems as fundamentally distinct from concrete practices of interpersonal deliberation (Bächtiger and Parkinson 2019; Scudder 2023). Nonetheless, these efforts define deliberativeness in terms of the flow of perspectives characteristic of deliberative exchange and the normative goods secured by it: mutual respect, equal influence, and universal inclusion.

But in emphasizing what goes on within deliberative exchange, theorists risk overestimating deliberation’s independence from its communicative context. Take the familiar critique that citizens are too saddled with cognitive biases to meet the herculean cognitive demands of deliberation. These arguments operate on the same level of giving and taking reasons as deliberation itself. Yet we are not just biased. We are biased toward certain positions and interpretations, and we typically share these biases with millions of others with whom we share a communicative context. For that reason, the content of shared biases is likely to privilege some positions and interpretations over others. Sociological work on “resonance” finds that the claims most likely to achieve ubiquity in discourse are those that, in Aristotelian fashion, merge seamlessly with widespread assumptions and accepted beliefs such that new positions appear “natural and familiar” to listeners (Gamson 1992, 135; Woodly 2015). For similar reasons, Young (2001, 685–6) argues the persuasive advantages of familiarity reinforce the status quo by favoring audiences’ presuppositions, making it more difficult for deliberators “to think critically” about them.

Consider how mainstream debate over criminal justice policy is so often carried out in terms of retributive punishment. While public support for capital punishment has declined, researchers attribute this decrease to the rise of “innocence frames” that oppose the death penalty on the grounds that mistaken courts will put innocent people to death (Baumgartner, De Boef, and Boydstun 2008). Innocence frames challenge the death penalty as a practice without unsettling deeper assumptions about its justice as a form of retributive punishment. Indeed, these frames accept the cultural primacy of retribution, questioning only how it is carried out. Yet the presumption of retribution, even when turned toward just causes, limits the kinds of claims that will be successful in public debate, making carceral logics intuitive compared to rehabilitative alternatives.

Such biases pose problems for grounding equality in deliberative exchange. James Lindley Wilson (2019, 150–1) worries unequal cultural resonance leads claims to land with unequal force by virtue of their content, independent of their speaker or argumentative strength. This results in objectionable “content-based inequalities” in persuasive force. Wilson points to the daylight between testimonially just deliberative exchange and the epistemic resources deployed within it. Citizens’ background understandings, like the naturalness of retribution, tilt the deliberative playing field toward familiar ideas, concepts, and interpretations in ways distinct from deliberative procedure. Wilson argues for recalibrating deliberative rules to counter content-based “differentials in consideration” stemming from “social and economic structures.” But it is not clear how deliberative rules can achieve this goal, at least not without a deeper “genealogy” of our epistemic resources and their influence on how we make and consider claims (Young 2001, 686).

EPISTEMIC RESOURCES AND HERMENEUTIC INJUSTICE

The “real world” may be nothing more than “clouds of swarming atoms,” wrote William James ([1890] 1905, 289; see also Dewey 1925; 1927, 324), but “the world we feel and live in will be that which our ancestors and we, by slowly cumulative strokes of choice, have extricated out of this, like sculptors, by simply rejecting certain portions of the given stuff.” Because persons who share a community rely on a common set of inherited concepts, “in my mind and your mind the rejected and selected portions of the original world-stuff are to a great extent the same,” allowing us to understand we live in this world and render claims about it mutually intelligible. Social and feminist epistemologists have expanded on the classical pragmatists to explain the development of our shared understandings, identifying epistemic resources with the “communal ways of organizing things” shared by a community of knowers (Nelson 1993, 139; see also Dotson 2014; Pohlhaus 2012). Epistemic resources are not constitutive of beliefs themselves, but are rather “building blocks” that may be fit together in any number of ways to describe our experience. ⁷

Shared resources at once expand and constrain our ability to seek and share knowledge. As Gaile Pohlhaus (2012, 717–8) argues, our conceptual tools “help us to understand, investigate, and know about specific parts and particular aspects of the world” but “do not help us to know all parts of the world equally or even all aspects of a given part of the world equally.” Rather, they pick out particular aspects of our undifferentiated experience as worthy of attention and make some inferential connections more obvious than others. All considerations are differential by virtue of our reliance on inevitably partial epistemic resources (Fricker 2007, 2–4; Harding 1991). For this reason, partiality cannot be objectionable generally. Rather, what makes instances of partiality objectionable is how the playing field is tilted and the Jamesian strokes which determine the tilt.

To explain this, I start from a stylized model of how epistemic resources emerge out of the “dialectical relationship” between persons’ social situatedness as individuals and their reliance on one another for knowledge (Pohlhaus 2012, 719; see also Medina 2013). We use a set of inherited concepts, methods, and interpretations to help us navigate our interactions with the external world. When frictions arise between resources and our experiences, we “recalibrate our epistemic resources or create new ones until the tension between our resources is alleviated.” Differences in persons’ situatedness and experience mean that resources that cause frictions for one person may still work well for another. To motivate a change to collectively held resources, persons must communicate tensions to those who may not recognize them on their own. Changes occur through iterated dialogue among differently situated knowers as they apply general concepts to their personal experiences and learn from others who are doing the same. ⁸

This idealized picture assumes all people participate equally in the iterative critique, contribution, and revision of epistemic resources and that each person’s contributions to that dialogue are taken up on their merits. It is characterized by “epistemic relational equality” (Fricker 2015, 77). The resources resulting from such a process are still partial, in that they prioritize some aspects of the world over others. But that partiality is not itself objectionable on egalitarian grounds as it reflects the equal participation of all in the preceding inquiry. ⁹ However, when knowers are regularly and arbitrarily denied the capacity or opportunity to participate in the revision of resources, it reflects a condition of epistemic relational inequality characterized by persons or groups’ systematically unequal epistemic participation. There may be innocuous and incidental sources of unequal epistemic participation, but my focus here is on inequalities of participation that track other non-epistemic forms of advantage or disadvantage across domains of social life. When groups’ contributions figure unequally into inquiry, the resources that emerge are not just partial but partial toward the epistemically (and socially) advantaged.

In such cases, inquiry will tend to favor the perspectives and interests of the epistemically advantaged and yield “lacunae” around the points of frictions faced by the epistemically disadvantaged characteristic of hermeneutic injustice (Fricker 2007, 158). Advantaged knowers develop pervasive and self-reinforcing ignorance concerning the disadvantaged (Medina 2013, 29–40; Mills 2007), which grows to the extent advantaged and disadvantaged knowers differ in their social positions and their shared experiences. The resulting gaps in shared epistemic resources further reinforce disadvantaged knowers’ epistemic marginalization as they are forced to interpret experiences and make claims intelligible within frameworks from which their perspectives have historically been excluded. As marginalized knowers become aware of these lacunae, they may contest them or, together with others, develop alternate epistemic systems counterposed with dominant ones. ¹⁰ However, such contestation faces serious barriers to uptake, among them suppression “by prevailing knowledge validating processes” (Collins 2000, 254), preemptive dismissal (Medina 2013; Mills 2007; Pohlhaus 2012; Young 2000, 55), and co-option by better-off groups (Collins 2000, 270; Davis 2018). Even if knowers can receive a fair hearing, intelligibility may come at the cost of relying on resources that are unreflective of their experiences and that may legitimize their unequal status (Catala 2015; Dotson 2014).

As unequal epistemic participation in political inquiry yields flawed resources over time, it creates a hermeneutic environment inhospitable for deliberation’s egalitarian aims. The problem here is not partiality generally but partiality that reflects and reinforces disadvantaged groups’ persistent marginalization in public discourse. Consider the earlier example of retributive punishment. The incarcerated are, in a very literal sense, blocked from participating in mainstream debates around criminal justice, which primarily reflect the values, interests, and emotions of those outside the carceral system (Medina 2021). To maximize their intelligibility and persuasiveness, incarcerated or formerly incarcerated speakers must articulate their claims within a field of understanding characterized by the exclusion of their viewpoints, and which tends to assume incarceration to be a natural, desirable, or unavoidable feature of political life. Persons in or coming out of the carceral system may face an uphill battle not only to be taken seriously as individuals (due to testimonial injustice) but also to make their claims about carceral logics intelligible and persuasive, even to well-meaning audiences abiding by deliberative norms.

In cases like this, deliberative exchange may itself be fair while still subjecting some participants to a discursive situation that blunts the force of their claims. Hermeneutic injustice produces similar consequences to testimonial injustice (and other forms of internal exclusion) while operating at a different temporal and relational register. Where testimonial injustice is inflicted by deliberators within ongoing give and take, hermeneutic injustice stems from past silences that generate exclusions within the conceptual terrain of present exchange. And where the former sees the relation of interest as between speakers and listeners, the latter sees it as between speakers and the epistemic community to which they belong. ¹¹

Background inequalities in epistemic participation undermine deliberation’s egalitarian potential in at least three ways. First, the hermeneutically marginalized suffer a deficit of deliberative influence because their audiences often lack the terms needed for them to make certain claims intelligible, let alone persuasive (Fricker 2007, 157). Second, marginalized speakers are denied the respect of being taken up on their own terms. Being forced to communicate in a conceptual vocabulary that excludes one’s experiences and self-conceptions saps the epistemic basis of self-respect, akin to Du Bois’s (1903, 7; see also Mills 1997, 33) sensation “of measuring one’s soul by the tape of a world that looks on in amused contempt.” Third, speakers are less likely to have their interests considered equally in decision-making, even when they are included in deliberation. Inquiry will tend to proceed from the starting point of dominant groups’ conceptions of problems faced by marginalized groups, rather than the conceptions of those groups themselves.

Furthermore, these dynamics limit deliberation’s capacity to generate political legitimacy and good decisions. Deliberation confers legitimacy by ensuring each speaker or viewpoint is granted fair consideration or equal opportunity to influence decisions (Beitz 1989; Christiano 2008; Viehoff 2014), while its capacity to harness dispersed information depends on each relevant voice getting heard (Estlund 2008; Landemore 2012). In both cases, disadvantaged speakers’ inability to contribute to deliberation with equal force threatens to skew the outputs of deliberation toward the advantaged, undermining those outputs’ democratic and epistemic quality.

AN INSTITUTIONAL ACCOUNT OF EPISTEMIC PARTICIPATION

Deliberation’s egalitarian potential depends not only on what happens during debate but also on the conceptually and temporally prior process of collective knowledge-making in which the raw materials of deliberation come about. I argue that in epistemic communities composed of millions of strangers, this process is best understood as occurring not only through interpersonal contribution but also through the institutions that sustain public inquiry at scale. To see this, it is worth juxtaposing my account with an alternative conceptualization of epistemic resources, which is as byproducts of vast networks of interpersonal interactions—the simple model described earlier scaled up. Fricker (2007, 155–6) emphasizes that gaps in collective epistemic resources often emerge from “non-participation in professions that make for significant hermeneutical participation (journalism, politics, law, and so on).” Nonetheless, she characterizes persons’ capacity to contribute to shared resources in distinctly interpersonal terms. Epistemic contribution, Fricker writes, is akin to “offer[ing] someone a cup of tea,” a concrete interpersonal act “between individuals or small groups… in the home, on the street, or a place of work” (2015, 75–6; 2007, 160–1, 169–75). We might think, then, that our “pool of shared epistemic materials” comes about through countless interpersonal acts of expression and uptake over time.

There is much to like about this “network” model. For one, it offers a straightforward picture of epistemic equality and inequality rooted in the iterative processes described in the previous section. And in doing so, it offers a clear explanation for how hermeneutic flaws emerge. Systematic gaps in collective resources at the structural level reflect unequal participation on the transactional level, often grounded in testimonial injustices and reciprocally reinforced by flawed resources (Catala 2015). Institutions matter in this view because they organize patterns of interaction, such as who interacts on a day-to-day level (Anderson 2012, 170–2) or what can be legally communicated (Fricker 2015, 83–5). In response, theorists like Fricker suggest we can challenge hermeneutic flaws by specifying transactional norms and virtues to minimize inequalities in our exchanges and designing social arrangements to protect and encourage exchanges of that sort (2007, 86–108, 169–75; 2015, 82).

But the network model is less plausible on closer inspection. Emphasizing exchanges between individuals comes at the expense of the idea that shared resources “are the kinds of things that stand outside or beyond any one individual” (Pohlhaus 2012, 718). In this view, our contributions are incorporated into collective understandings primarily as listeners internalize them and spread them to others in an ongoing chain of interaction. Yet, how would a particular concept become so widespread within a community that it could be mutually intelligible among people who have never met and have no direct or indirect social connections? Additionally, if all we have is iterated interaction, it is not clear how an idea might become sufficiently concrete to be identifiable as a shared resource across its innumerable applications. The network picture is one of dizzying conceptual diversity and churn. Moreover, it is particularly unsatisfying given how much of our normative concern with flawed communication is motivated by stereotypes and misconceptions that are depressingly common and entrenched in mass discourse. One might argue that as ideas ripple through our networks, they achieve equilibrium in relation to the interpretive needs of dominantly situated knowers. But even so, we would need an explanation of how that churn locks into stable configurations that are taken up and circulated by a multitude of individual knowers.

While the network account recognizes that our epistemic interdependence emerges out of the conjoint nature of social activity (Dewey 1927, 250), it shares deliberative theory’s fixation with interpersonal interaction. Compare it to that offered by John Dewey, who emphasized the impersonal interdependencies brought into being by modern capitalism and communication technology. “[R]ailways, mails, and telegraph-wires… influence more profoundly those living within the legal local units than do boundary lines,” Dewey wrote (301–2). Our “face-to-face associations” are increasingly conditioned by “remote and invisible organizations” that structure the flow of communication across them (296). Given that we cannot all contribute directly to each other, Dewey recognized the role of mass communication in bridging our spatially limited social networks. “[T]he rapid and easy circulation of opinion and information” across vast geographic and social distances brought into being epistemic communities at the scale of mass polities, “far beyond the limits of face-to-face communities” and even the most dispersed personal connections (307). What defined these far-flung publics as a “new form of political association” was not massively iterated dialogue but the reliance on common inputs prior to individual communicative activity.

Dewey suggests a more plausible view of how epistemic resources come about: out of the reflexive interplay of mass and interpersonal registers, in which resources circulated through authoritative channels to many at once are taken up and refined in interpersonal conversation before being reciprocally taken up again by institutions. This is not to say that grassroots interactions do not matter but that agency resides with both publics and institutions. Epistemic resources depend not only what we hear and learn from others in our direct conversations but also what we hear about the lives of the millions of citizens we will never meet through institutional channels like news media, the entertainment industry, political rhetoric, and advocacy groups. These institutions function as intermediaries of epistemic participation when citizens’ perspectives, experiences, and contributions inform the stories and interpretations packaged and circulated to mass audiences. Intermediation occurs on the demand side, such as when audience interest leads a journalist to seek out more information on a particular topic, and on the supply side, when “deliberative entrepreneurs” draw media attention to submerged issues and advance new frames in subsequent coverage (Barvosa 2018, 43). “Hybrid” interactions between traditional mass communication and decentered online discourse have generated new opportunities for this kind of responsiveness, like when grassroots voices gain viral attention on social media and are taken up and amplified by mainstream elites (Chadwick 2017). For example, the online circulation of videos of police brutality and the digital mobilization of the Black Lives Matter movement forced complacent media and political elites to reckon with long ignored racial injustices. While the evolution of elite rhetoric on race and racism has been halting and marked by backlash, Deva Woodly (2022, 161, 170–80) argues that movement efforts have successfully introduced “new concepts into the political lexicon” that set the groundwork for further change.

The impersonal relations between subjects and audiences that this hybridity brings into being are not readily captured by a model of epistemic participation as interpersonal giving. Making sense of those relations requires a closer look at structured processes of intermediation: how citizens contribute (or are blocked from contributing) to the public flow of information that media institutions make possible and the epistemic resources to which those flows give rise. Institutions function according to reasonably stable rules that give rise to regular, predictable patterns in their behavior as intermediaries. These rules may constrain the openings for deliberative entrepreneurship and its likelihood of success. They may favor some entrepreneurs over others. Explaining how institutions facilitate or frustrate participation will require looking closely at the ways institutional logics determine patterns in whose contributions shape what is communicated to publics and to which publics institutions are most responsive.

To illustrate this, I focus on the example of how market incentives within commercial media give rise to systematically unequal epistemic participation. The business side of American news media has always been unforgiving. But massive technological and economic changes have placed its traditional advertising-based revenue model under unprecedented strain. Tightening margins for legacy and digital outlets alike have ratcheted up pressure to secure stable revenue streams. I point to three ways these pressures shape social groups’ relative influence on news. First, audiences’ unequal resources affect outlets’ attention. Advertising hitches the value of information to its potential to attract the high-income consumers most desired by advertisers, a reward structure that encourages outlets to focus on issues and adopt frames that appeal to better-off audiences (Hamilton 2006). As outlets pivot to digital subscription models to compensate for lost advertising revenue, they face equivalent pressures to prioritize audiences able and willing to pay for online news. These audiences tend to be wealthy, white, and highly partisan (Usher 2021). Second, incentives to attract consumers refract underlying inequalities among audiences. The size and homogeneity of different groups within a market encourage publishers to tailor content to appeal to the largest, most homogenous among them, with generalist appeals often crowding out the voices of political and racial minorities (George and Waldfogel 2003). Finally, markets shape who reports, edits, and publishes news. While journalism has never been terribly lucrative, the reporting jobs remaining after decades of layoffs are highly competitive with relatively low compensation and dim long-term prospects. Many outlets, particularly digital outlets, have shifted risk to journalists by embracing the low overheads of freelance work (Pickard 2020, 84–6). This precarity makes recruiting and retaining members of disadvantaged groups that much harder, resulting in digital and traditional news industries that are demographically unrepresentative of the general public (Usher 2021, 46–51). ¹²

News media’s incentives to favor the presumptions and interests of advantaged audiences exacerbate patterns of hermeneutic marginalization. ¹³ As journalist Wesley Lowery (2020) argues, “the mainstream has allowed what it considers objective truth to be decided almost exclusively by white reporters and their mostly white bosses… calibrated to avoid offending the sensibilities of white readers,” all while restricting “coverage of black and brown neighborhoods to the crime of the day.” These structural biases lead to the circulation of widespread stereotypes and misperceptions. TV news so consistently overrepresents African Americans in coverage of crime that, all else being equal, frequent local news consumers are more likely to hold false beliefs about the level of crime in their communities, the percentage of racial minority perpetrators and victims, and racial minorities’ “natural” tendency for violence (Dixon 2008). The same incentive structure constrains Black citizens’ ability to amplify their claims through mainstream news media, as reporting often frames Black speakers negatively by default. News coverage of non-white protestors, for example, tends to use vocabulary “associated with fear and anger” more often than coverage of white protestors (Gause, Moore, and Ostfeld 2023, 448).

Such structural dynamics have little to do with direct interpersonal communication. Rather, they primarily concern how institutions incorporate citizens’ claims and experiences on the ground into the epistemic resources distributed to wider publics. Internal logics that lead journalists and editors to take up persons’ contributions unequally result in a systematically flawed vision of the world projected through their institutional megaphone. Of course, such biases cannot be attributed solely to institutional features. Media professionals do their jobs in a social–epistemic environment warped by inequality and oppression. But my point is that, even against such a backdrop, institutions themselves contribute to the silencing of marginalized voices at a scale far greater than any one individual. Even when the professionals acting within those institutions are epistemically virtuous, they face strong institutional incentives to favor some perspectives over others. In such cases, institutions generate flows of communication as if the people within them were prejudiced. ¹⁴

EQUAL EPISTEMIC PARTICIPATION AS A REGULATIVE IDEAL

If our epistemic resources come about through channels of mass communications, realizing democratic equality will depend on securing the right kind of communication system. Thus far I’ve focused on how media institutions perpetuate hermeneutic injustices that undermine deliberation. This section develops a positive account of how media institutions can facilitate equal epistemic participation to promote the background conditions of hermeneutic justice needed to sustain egalitarian communication. I argue this goal is best pursued as a regulative principle that commits us to identifying and correcting systemic sources of unequal epistemic participation among social groups as they appear, rather than theorizing perfectly ideal institutions in the abstract.

To start, we need to ask what a community of equals would look like as an epistemic community. Perhaps one might think that in an egalitarian society, differences in social position and thus systematic differences in experience would shrink, such that widely available epistemic resources worked equally well for each person. No doubt a more egalitarian society would see a significant narrowing of systematic differences across group experience related to social and material inequality. But short of each person being cognitively identical and occupying the same social and geographic position, human difference keeps open the possibility that one will experience frictions others have not. Though entangled in practice, material and epistemic equality operate on separate tracks.

Given the possibility of difference, what matters is each person’s capacity to participate in ongoing processes of critique, revision, and redescription of shared epistemic resources. Resources should come to reflect, to the fullest extent possible, the perspectives contributed by community members, considered as equals. Fricker (2015, 80) rightly argues that participation in this sense is “emphatically not a matter of securing acceptance or agreement, or of having others adopt the contributor’s particular interpretive habits,” but rather of ensuring that contributions are not arbitrarily discounted or silenced and that adequate mechanisms for participation are in place. A community of equals will not and should not converge on a consensus about how to interpret the world. Rather, under conditions of equality, the resources we rely on would incorporate the widest range of publicly salient experiences and perspectives among members of the polity, bounded by their equal status. Put another way, it matters less that all epistemic resources perfectly reflect the input of all contributors than that resources are responsive to diverse contestation and free of pernicious stereotypes and gaps indicative of identity-based hermeneutic exclusion. ¹⁵ Achieving this goal requires identifying and removing sources of unequal epistemic participation (Catala 2015; Fricker 2007, 153). Deliberative rules accomplish this at the level of interpersonal exchange. But if epistemic participation also occurs through the touchpoints between interpersonal exchange and institutional intermediaries, then it requires addressing barriers to equal participation at those touchpoints as well.

A tempting conclusion would be that, given the risks of exclusionary gatekeeping and structural bias, we should do away with large-scale institutional intermediaries altogether. Social media suggests the possibility that autonomous and horizontal communication networks could replace traditional elite intermediation in facilitating epistemic participation on a mass scale. Indeed, online counterpublics like Black Twitter have emerged as critical sites for the formation and transmission of new epistemic resources in face of their members’ exclusion from mainstream media (Lee-Won, White, and Potocki 2018). Equal epistemic participation undoubtedly requires citizens to have access to a wide variety of diverse discursive spaces. Hyper-centralized mass media has historically impeded that goal, and social media has arguably advanced it (Cohen and Fung 2021). But the proliferation of discursive spaces is not itself sufficient for securing equal epistemic relations. Our interest in equality rests on how our experiences and perspectives factor into the understandings of others unlike ourselves. ¹⁶ The fragmentation of the public sphere into countless isolated sub-publics hinders our ability to share knowledge across lines of difference unless combined with mechanisms to link sub-publics back together through common flows of communication. Institutions of mass communication have traditionally served this function but have rarely done so in a way that supports the interest in equality that motivates it.

As new technologies diversify democratic communication environments, it is even more important that democracies seek out intermediaries capable of curating widely accessible flows of public information in ways that reflect persons’ claim to equal voice. And our interest in equal participation provides reasons to prefer some types of intermediary institutions over others based on the epistemic resources they produce and how they produce them. The institutions that best incorporate publics’ voices into shared resources on footing as equals will not necessarily grant all persons an equal soapbox. An intermediary may produce resources that better reflect key patterns in relevant citizens’ interests than thousands of unfiltered voices, or at least do so in a way more amenable to audiences’ limited time and attention. They may also better serve other goals interlinked with democratic equality, such as the acquisition and application of technical knowledge.

Similarly, we should not expect all intermediaries to support equal epistemic participation in an identical way. Just as individually non-deliberative institutions can contribute to deliberative values as part of a deliberative system (Bächtiger and Parkinson 2019, 85–6; Mansbridge et al. 2012, 2), how a particular institution supports epistemic participation will depend on its place in a broader division of communicative labor. Intermediaries targeting the general public, like national political campaigns and broadcast news networks, should aim to bridge diverse groups through epistemic resources that reflect each groups’ equal contribution. But other organizations, such as advocacy organizations or Spanish-language news, may intentionally prioritize the perspectives of specific identity, interest, linguistic, or geographic communities. Nonetheless, these group-specific institutions can still promote equal epistemic participation in at least two ways. First, such organizations serve as intermediaries within groups, aggregating and crystallizing the grassroots claims of individual members into forms of solidaristic consciousness rooted in shared experience (e.g., Collins 2000). For this reason, they must also be attuned to potential inequalities in their treatment of group members facing overlapping forms of marginalization. Second, group-based intermediaries serve as links in longer chains of intermediation. Amplification of group-specific resources in the institutional public sphere increases their likelihood of being taken up by mass intermediaries. However, the risk that those resources will be coopted by or lose their specificity among mass audiences points to the enduring need for intermediaries serving particular communities.

In general, I suspect the institutional intermediaries most defensible on egalitarian grounds will be those most responsive to the engagement and contestation of diverse publics, whatever their intended scale. To that end, institutions at every level of the communications ecosystem bear the burden of proving their democratic credentials. The positive task for egalitarians, then, is to articulate principles for institutional design, to identify which kinds of institutions can meet this standard, and to balance citizens’ claims to participation with other epistemic interests. The negative one is to identify and remove mechanisms that perpetuate unequal participation. On both fronts, equal epistemic participation refines our normative criteria for iterative and ongoing institutional experimentation.

To make this concrete, consider the example of unequal market pressures described in the previous section. Evidence of the deleterious effects of the collapse of the local newspaper industry on issues from polarization to municipal finance has renewed policy interest in protecting local journalism where it survives and reviving it where it has disappeared (Pickard 2020). To this end, “link taxes” and other efforts to modernize outlets’ monetization strategies look to enhance publishers’ market power within the existing media landscape while avoiding legitimate concerns about government discrimination between high- and low-value news. But such policies do little to work against, and may even exacerbate, market pressures toward better-off audiences. A commitment to equal epistemic participation pushes us to seek out structural reforms to stabilize local news while simultaneously minimizing unequal market incentives. Examples of such policies might include payroll tax credits for locally owned outlets, grant support for nonprofit news, programs for recruiting reporters from underrepresented backgrounds, or making journalists eligible for public service loan forgiveness and other forms of direct aid. Critically, such policies avoid making government the arbiter of good journalism. Rather, they aim to promote structural conditions conducive to the production of equality-enhancing news while leaving the question of what that news looks like in the hands of democratic publics and their journalistic intermediaries.

CAN DELIBERATION OVERCOME HERMENEUTIC FLAWS?

I conclude by returning to the deliberative argument. No doubt most deliberativists are also deeply concerned with the distorting effect of pre-deliberative inequality and the institutions that perpetuate it. Yet the argument that institutionally abetted hermeneutic injustice requires looking beyond deliberation may still give some pause. When deliberation occurs against an unjust epistemic backdrop, why not recalibrate deliberative processes to counter the effects of background inequalities post hoc? Work responding to racial and gender-based internal exclusions often takes this tack. One approach defends deliberative norms specifically attuned to speakers’ unequal starting positions (Davis and Finlayson 2022; Drake 2023, 106–8; Scudder 2020). While Wilson (2019, 158–65) acknowledges the need to address unjust communication structures at the root of content-based inequalities of consideration, he prioritizes principles of “deliberative triage” to correct for those inequalities within deliberative exchange, such as assigning greater priority to the judgments of disadvantaged speakers and instituting a “ceiling” on consideration of advantaged ones. Alternately, many deliberativists advocate deliberative procedures designed to amplify disadvantaged voices, among them recruitment strategies that overrepresent disadvantaged groups, communication formats that favor cooperation over argument, and rules that give marginalized speakers priority in speaking order and time (e.g., Beauvais and Bächtiger 2016; Karpowitz, Mendelberg, and Shaker 2012). Both approaches are valuable means of protecting deliberation’s egalitarian core in social–epistemic contexts of inequality. But they aim to accomplish that goal by compensating for hermeneutic flaws’ downstream consequences with special consideration or affordances, rather than by addressing flaws directly. For that reason, I am skeptical such efforts can succeed on their own.

This is because epistemic resources are not frameworks imposed on our judgment but are constitutive of judgment itself. Deliberation proceeds by giving reasons from within a system of epistemic resources shared by deliberators. It excels at reconciling inconsistencies or contradictions internal to that system using the resources already provided. Kristie Dotson (2014, 118) calls this type of inquiry a “first-order change” to our epistemic system: efforts “to make one’s behavior reflect one’s beliefs and values” while taking those beliefs and values as a given. This is fine if we assume the beliefs and values deliberators bring into deliberation are right for the task. But if they are products of unequal epistemic participation, two problems follow. Even the best arguments are prone to begin from misleading premises or follow pernicious logics that reaffirm those inequalities. And disadvantaged speakers must still argue within a framework that is inadequate for explaining or may even obscure their experiences.

One could argue deliberators have an obligation to critically assess and revise their epistemic resources upon challenge and to cultivate habits of hermeneutically virtuous listening (Fricker 2007, 169–70; Scudder 2020, 101–6). But Dotson points to how difficult this can be (see also Anderson 2012, 167–70). Second- and third-order changes target, respectively, gaps within an epistemic system (insufficient epistemic resources) and flawed systems themselves (inadequate epistemic resources). However, using resources from within a flawed epistemological system “may thwart one’s ability to make significant headway in becoming aware of [that system’s] limitations” (Dotson 2014, 132). This is because epistemological systems are resilient. When contestation is raised from and articulated within a system of flawed understandings, it runs the risk of becoming incorporated and reconciled into that system “without redefining its structure” (121). Moreover, reliance on flawed resources biases revision toward “what the system is prone to reveal.” This results in “a vicious loop,” where, in response to a challenge, an epistemic system changes on the margins while keeping its basic shape and its practical supremacy (132; Davis 2018, 715). The advantaged continue to see their everyday understandings as working reasonably well, rather than recognizing the flaws that permeate them. Innocence frames about the death penalty provide an example of this resilience in action: the prospect of putting innocents to death has led more citizens to oppose the death penalty for prudential reasons, without substantially altering attitudes toward its moral legitimacy.

Deliberative ideals of reasoned justification, rational revision, and open listening work well on the first order, identifying what rationally follows from our commitments within a given set of understandings and assumptions and ensuring persons’ contributions are given the consideration and uptake they are due. Adaptations to deliberative procedures work on this level by amplifying historically silenced voices to ensure their contribution to the exchange of reasons. But even then, there remains the risk that those contributions, once made present, will be subsumed into prevailing systems of shared knowledge, given their resilience to revision from within. In those cases, deliberation may promote positive first-order changes while reinforcing second- or third-order inequalities. Well-calibrated deliberation and well-designed fora can address these inequalities’ most glaring interpersonal consequences, but they are unlikely to rectify deeper flaws embedded in how deliberators consider claims and issues in the first place.

The argument that deliberation is too entangled with systems of power to neutralize background inequality is not a new one. But it would be a mistake to conclude we should give up on deliberation as a site of democratic equality. Rather, the account I have sketched is about where else equality and inequality reside in the public sphere. Deliberative theorists are right to identify interpersonal conditions necessary for equal opportunity for influence or consideration. But against a backdrop of deep structural inequalities of voice, even ideal deliberation will struggle to achieve these goals. It may even be counterproductive. Instead, deliberative and structural approaches to communicative equality should be seen as complements operating at distinct but interrelated social and temporal registers.

Taking hermeneutic flaws seriously provides guidance about the kinds of deliberative practices and frameworks we ought to embrace. In terms of practices, my account suggests the more one insists that deliberators adhere to broadly shared political ideas and values, the greater the risk that marginalized participants will suffer from unequal influence or consideration. For that reason, “wide” approaches to deliberation that embrace group-specific appeals, affect, and anecdote (e.g., Christiano 2008, 190–2; Young 2000, 47–51) are likely to better accommodate challenges to ossified epistemic systems (Woodly 2022, 14–8). Similarly, deliberative fora may be designed to better support the inclusion of submerged epistemic resources, such as by prioritizing open expression and cooperation over rational argumentation and creating opportunities for enclave deliberation prior to mixed discussion (Karpowitz, Mendelberg, and Shaker 2012).

As for deliberative systems, my argument pushes back against evaluative standards of deliberativeness modeled solely on interpersonal exchange. For example, André Bächtiger and John Parkinson (2019, 116) recognize the importance of narrative and symbols in deliberative systems but dismiss the “demanding” idea that narratives and symbols be themselves “constructed democratically, reflectively, and authentically.” Deliberativeness, in their view, requires only that persons have opportunities to hear and consider “representations of stories” from a diversity of citizens. Yet, that position fails to appreciate how deeply our stories structure the reception and consideration of claims. Rather, deliberative systems must be evaluated based on not only whether they secure patterns of deliberative communication but also whether they promote the background conditions of hermeneutic justice that make genuinely egalitarian deliberation possible.

CONCLUSION

Democratic equality in the public sphere cannot start and end with an ideal of how citizens talk to one another. The shared concepts and ideas that bind us together as a community of knowers are themselves loci of communicative equality and inequality. To be an equal, I have argued, is to have one’s perspectives and experiences factor into the understandings one shares with strangers in indirect but meaningful ways. Well-calibrated deliberation can limit the harmful consequences when we fall short of this admittedly demanding goal. But achieving communicative equality at the scale of mass democracy will depend as much on the sprawling, densely intermediated, irredeemably hierarchical, and explicitly non-deliberative media ecosystem that shapes deliberation as it will deliberation itself.

Recognizing the institutional public sphere as a site of equality and inequality empowers democratic publics in turn. As I have argued, the well-documented exclusions that linger within deliberation run deeper than traditionally understood, stemming not only from narrow rules and personal prejudices but also from the stuff of deliberation. Identifying the institutional sites where this stuff is produced and circulated creates new opportunities for democratic agency. Much criticism of deliberative democracy focuses on citizens’ nominally fixed cognitive biases. But my account calls attention to the social structures that give biases their content and determine their democratic consequences. These structures are anything but fixed. As Dewey (1927, 350) writes, “a subtle, delicate, vivid and responsive art of communication must take possession of the physical machinery of transmission and circulation and breathe life into it.”