New Trouble For Deliberative Democracy

In the past two decades, democratic political practice has taken a deliberative turn. That is, contemporary democratic politics has become increasingly focused on facilitating citizen participation in the public exchange of reasons. Although the deliberative turn in democratic practice is in several respects welcome, the technological and communicative advances that have facilitated it also make possible new kinds of deliberative democratic pathology. This essay calls attention to and examines new epistemological troubles for public deliberation enacted under contemporary conditions. Drawing from a lesson offered by Lyn Sanders two decades ago, the paper raises the concern that the deliberative turn in democratic practice has counter-democratic effects.

ISSN 1718-9977 (digital) Explore this journal Cite this article Talisse, R. (2017). New Trouble For Deliberative Democracy. Les ateliers de l'éthique / The Ethics Forum, 12 (1), 107-123. https://doi.org /10.7202/1042280ar Article abstract In the past two decades, democratic political practice has taken a deliberative turn. That is, contemporary democratic politics has become increasingly focused on facilitating citizen participation in the public exchange of reasons. Although the deliberative turn in democratic practice is in several respects welcome, the technological and communicative advances that have facilitated it also make possible new kinds of deliberative democratic pathology. This essay calls attention to and examines new epistemological troubles for public deliberation enacted under contemporary conditions. Drawing from a lesson offered by Lyn Sanders two decades ago, the paper raises the concern that the deliberative turn in democratic practice has counter-democratic effects.

NEW TROUBLE FOR DELIBERATIVE DEMOCRACY
ROBERT TALISSE DEPARTMENT OF PHILOSOPHY, VANDERBILT UNIVERSITY

ABSTRACT:
In the past two decades, democratic political practice has taken a deliberative turn. That is, contemporary democratic politics has become increasingly focused on facilitating citizen participation in the public exchange of reasons. Although the deliberative turn in democratic practice is in several respects welcome, the technological and communicative advances that have facilitated it also make possible new kinds of deliberative democratic pathology. This essay calls attention to and examines new epistemological troubles for public deliberation enacted under contemporary conditions. Drawing from a lesson offered by Lyn Sanders two decades ago, the paper raises the concern that the deliberative turn in democratic practice has counter-democratic effects.

I. INTRODUCTION
Deliberative democracy is an appealing yet elusive ideal. In its canonical versions, it is the thesis that democracy's value-its legitimacy, authority, authenticity-lies in its ability to base political decisions on citizens' reasons rather than simply on their preferences or votes. 1 To be sure, there is a great variety of views in currency claiming the name, and so the foregoing summary statement requires a good deal of clarification. Yet any version of the canonical view gives rise to a series of questions that help to explain deliberative democracy's elusiveness. What does it mean to base political decisions on citizens' reasons? How are reasons to be distinguished from preferences? Why do reasons confer legitimacy (or authority, or authenticity) on collective decisions? Does deliberative democracy rest upon the assumption that citizens already share a view about what reasons are? Hence extensive and rapidly growing literatures have emerged around these (and other) questions.
It is somewhat surprising, then, to find in a review essay published nearly twenty years ago James Bohman (1998) declaring that deliberative democracy has "come of age." Noting that the core idea of deliberative democracy had been a central and explicit theme in much democratic theory since at least the 1980s, Bohman was canvassing the then-recent theoretical developments aimed at showing that the deliberative ideal could be regarded as feasible. That is, by 1998, the central challenges to deliberative democracy concerned not its theoretical architecture, but rather its practical implementation. According to Bohman, deliberative democracy's principal theoretical commitments-including its normative superiority to aggregative, pluralist, and elitist models of democracy-had been widely accepted. By 1998, deliberative democracy had "come of age" in that it had established itself as one of the central frameworks for normative democratic theory. As Bohman presents it, the remaining task is that of enacting the theory.
Writing two years later, Samuel Freeman sounds a similar tone. Freeman observes that "deliberative democracy" had become "more than just another popular label"; it is, he contends, a distinctive family of views united by more than simply a common rejection of an entrenched opposing view (2000, p. 371). Freeman notes that, as it is a positive program within normative democratic theory, deliberative democracy is the site of several internal disputes. After carefully working through many of these, Freeman concludes that deliberative democracy provides a vision of the democratic ideal that is indeed superior to that offered by well-established non-deliberative theories. However, his enthusiasm is somewhat measured. In the end Freeman remains "sympathetic" to deliberative democracy, but he is not fully an advocate; he expresses the apprehension that deliberative democrats have yet to demonstrate that the deliberative ideal is practically feasible (2000, p. 418).
It is safe to say that deliberative democracy's popularity has only grown in the intervening decades. Today, deliberative democracy is arguably the predominant framework in normative democratic theory; hence, it is difficult to find a normative democratic theorist who does not embrace some version of the core deliberativist thesis that democracy's value (authority, legitimacy, justice) is owing to democracy's ability to shape political decision in response to the open exchange of ideas, reasons, and arguments of citizens. 2 Moreover, the loudest opposition to deliberativism tends to originate from theorists who are suspicious of normative accounts of democracy as such. 3 Still, the concerns over deliberative democracy's feasibility remain, and these challenges have led deliberative theorists to adopt increasingly stylized models of public deliberation. These models vary significantly over fundamental issues: Who deliberates? When, where, among whom, and for how long should deliberation occur? What questions are suitable for public deliberation? Is deliberativeness primarily a feature of interpersonal communicative interactions, or is it rather a property of certain systems of collective decision? Each of these questions is the focus of intense and ongoing debate. Hence, even though deliberative democracy dominates normative democratic theory, it remains a highly troubled framework.
I will not canvass these longstanding and intricate debates here. 4 Instead, I will revisit a kind of critique of deliberative democracy that targets its desirability under social conditions that currently prevail and should be expected to persist. To explain: in 1997, Lyn Sanders argued that the social dynamics of race, gender, and class significantly impact deliberative encounters in ways that replicate the patterns of exclusion, disadvantage, and marginalization that obtain in the society at large. Drawing on robust studies of jury behavior, Sanders noted that women, non-whites, and the economically disadvantaged speak less frequently, are almost never selected for the role of foreman, and are more likely to be interrupted and dismissed while speaking, as compared with fellow jurors who are white economically privileged males. Sanders argued that the jury studies suggest that, even under institutionally favourable conditions, deliberative democracy is likely to further entrench existing patterns of social inequality. 5 She thus urged caution in calling for the deliberativization of existing democratic practice. Sanders argued that, in order to play their intended role in enriching democracy, deliberative institutions and practices must operate against the background of broader egalitarian social commitments, and that these are commitments that deliberation alone cannot foster; furthermore, she showed that they are commitments that deliberation, when enacted under conditions in which they are insufficiently entrenched, can undermine. In a nutshell, then, Sanders's lesson is that we do not necessarily make progress towards the deliberative democrats' political ideal by adjusting existing democratic institutions so that they more closely approximate the ones prescribed in deliberative democratic theory. 6 The trouble is that in the past two decades democratic political practice has taken a deliberative turn. That is, contemporary democratic politics has become increasingly focused on facilitating citizen participation in public argumentation. Owing largely to advances in communications technology and social media developments, the public sphere is saturated with outlets, sites, and forums for public political discourse, from comments sections on news websites to feeds on Facebook and Twitter. Even nightly news programming is presented in a proand-con debate format, where viewers observe an exchange of competing reasons and then are primed to draw their own conclusions. Just as Sanders would predict, our politics has become increasingly divisive and uncivil; more importantly, the prevailing divides have become less a matter of disagreement among democratic citizens and more a power struggle among conflicting visions of what it means to be a democratic citizen, with each side condemning the others as fundamentally opposed to a proper political order. Under such conditions, civil disagreement is hardly possible, as the contending parties are apt to regard each other as peddling a distorted or perverted conception of democracy itself. In the US in particular, the increased emphasis on public argument has helped to bring about conditions under which no real political debate is possible; adherents of contending views do not debate, but merely challenge each other's competence, sanity, and fitness for citizenship. As Sanders warned, increased deliberativeness under existing conditions seems to have exacerbated underlying social divisions.
In this essay, I raise the worry that deliberative democracy in practice is unavoidably vulnerable to the kind of pathology to which Sanders called attention. However, the core of the specific concern I will raise differs importantly from what Sanders described. Sanders argued that, in order to be democratically enriching, deliberation needed to operate against a background of the kind of egalitarianism that could dismantle entrenched hierarchies of race, gender, and class. To be sure, Sanders's argument identifies a demanding precondition for deliberative democracy. The worry I will raise points to epistemological prerequisites for democratic deliberation that arguably are even more demanding.
In the next section (II), I will show that, although deliberative democracy is most commonly presented as a moral ideal, it nonetheless has a decidedly epistemological dimension, and thus places on democratic citizens distinctively epistemological requirements. This means that there are specifically epistemological ways in which deliberative democracy can falter. In the third section, I will review a familiar way in which public deliberation can fail epistemologically. Then I will argue, in the fourth section, that there are unique epistemological problems that arise from the fact that democratic deliberation is conducted not merely between contending parties, but among contending parties arguing in front of an onlooking audience. Once it is noticed that public deliberation is frequently conducted for the sake of the onlookers, new occasions for epistemological pathology arise. These third-party epistemological pathologies are difficult to counteract within a democratic framework-hence the "new trouble" announced in my title. The concluding section (V) will draw some admittedly bleak upshots of the foregoing analysis.

II. THE EPISTEMOLOGICAL DIMENSION OF DELIBERATIVE DEMOCRACY
Deliberative democracy is most frequently proposed as a centrally moral ideal. The idea is that, in collectively deciding how the coercive power of the democratic state is to be exercised, citizens owe to each other reasons for favouring a given policy over its alternatives. 7 The deliberativist claims that, when collective political decision is driven by activities of public deliberation, political policy emerges less as an imposition upon the democratic citizenry, and more as an expression of the popular will; political decisions preceded by public deliberation are thus said to realize the traditional ideal of collective self-government. Moreover, the deliberativist contends that public deliberation helps to legitimize collective decision by giving citizens access to the reasons behind public policy, reasons which can subsequently be challenged, revised, or overturned in ongoing public discourse. In this way, again, public deliberation is proposed as a means for making collective decisions that each citizen can regard as something more than a raw exercise of power; deliberativism regards the state and its policies as vulnerable to the reasoned contestation of democratic citizens, and, in this way, power is rendered accountable to the citizens. Finally, the deliberativist holds that processes of public deliberation manifest an attractive conception of citizenship in that when citizens deliberate, they must civilly give and receive reasons, acknowledge each other's points of view, and respectfully argue in ways that provide others not only with a chance to speak, but also with an opportunity to be heard. Hence deliberative democracy invokes a particular conception of the traditional democratic idea of an active and engaged citizenry. 8 To put these points together: the deliberative democrat's central contention is that public deliberation is necessary in order to realize the democratic ideal of collective self-government among morally equal and active citizens.
Articulated as such, deliberative democracy is clearly a moral ideal. But it also makes indispensable use of several epistemic concepts. 9 Deliberation itself is unavoidably epistemic; it is a process of discerning and evaluating reasons. And reasons are most certainly epistemic items, as they are (on anyone's view) considerations that count in favour of some conclusion. Note, moreover, that deliberativists hold that democracy calls for public deliberation precisely because there is disagreement among citizens over how the state's coercive power should be exercised. Disagreement is not deployed here as a merely descriptive term, indicating the fact that unanimity does not prevail. Rather, the deliberativist acknowledges that there is disagreement because there is a clash among reasons, and different reasons favour different policies. The task of deliberation is hence that of attempting to consider the full range of reasons and discerning their respective weight so that one could decide which policy outcome is best supported by the reasons. It is difficult, to say the least, for an individual citizen to survey the full range of reasons in play with respect to any given public policy, so citizens must deliberate together; they must share, exchange, and scrutinize each other's reasons. Accordingly, public deliberation is partly-perhaps largely-a process of public argumentation where citizens make the case for their favoured public policy to each other, consider cases made by others for alternative policies, and all stand ready to be challenged.
The ideal of a deliberatively engaged and arguing citizenry is undeniably demanding, and, again, some have criticized it on that ground. 10 My present point, however, is that, although deliberative democrats most frequently offer moral reasons to hold that public deliberation is necessary for proper democracy, the processes of democratic deliberation themselves cannot be identified except by reference to epistemic concepts. We might say, then, that the deliberative democrat proposes that citizens morally owe each other civil participation in a collective epistemic activity. Hence deliberative democracy involves a moral requirement and an epistemic requirement; citizens must interact civilly, and their interactions must rise to the epistemic level of deliberation.
Once we see that deliberative democracy is a both moral and epistemic proposal, we also see that the normative core of deliberative democracy is partly epistemological. It would be hard to imagine any democratic theorist endorsing deliberativism in the light of a demonstration that public deliberation, even when conducted civilly, always produces epistemically disastrous results. That is, part of the normative appeal of deliberative democracy lies in the presumed potential for civil public deliberation to yield epistemological benefits of some kind. Hence deliberativists often claim that public deliberation produces epistemically better collective decisions, more rational policies, better informed voters, more intelligent citizens, and the like.
I will not examine here the question of whether public deliberation actually yields epistemological benefits. 11 My point is that deliberative democracy can falter normatively in at least two ways: first, it falters morally when citizens are not able or are not inclined to engage each other civilly on political questions; second, it falters epistemologically when citizens indeed engage each other, and may even do so civilly, but engage in ways that severely fall short of the epistemic ideals of public deliberation (better-informed judgments and voters, betterreasoned policy decisions, greater public understanding of public policy, increased accountability, and so on).
To be sure, although these two kinds of failure are conceptually distinct, in practice they comingle. For example, it is common for incivility among deliberating parties to have its root in accusations of epistemic incompetence. And one especially potent form of incivility consists in the systematic impugning of others' epistemic capacities or credentials. Indeed, Sanders notes that many of the varieties of deliberative incivility and exclusion that she discusses have their root in a prior judgment that non-white non-males lack "epistemological authority" (1997, p. 349). Now, it would be optimistic to claim that deliberative democratic incivility always has its source in an unjustified negative assessment of the epistemic condition of one's opposition. Surely a considerable portion of democratic incivility is due to unadulterated bigotry and garden-variety intolerance. But distinctively epistemological failings of deliberative democracy are prevalent and have accordingly attracted a good deal of attention. Reviewing a familiar kind of epistemic pathology of deliberation will set the stage for a new kind of difficulty.

III. FAMILIAR TROUBLE: THE POLARIZATION DYNAMIC
Deliberative democrats hold that citizens should engage in public deliberation. In public deliberation citizens do not merely announce the reasons driving their political advocacy; rather, they participate in a collective epistemological activity that involves public political argument. Deliberative democratic citizens reason together; they present their arguments to each other for the sake of advancing the rational collective investigation into some public and political issue. This collective aspect of public deliberation provides occasion for a range of epistemological pathologies. I here focus on a common dynamic among at least three such pathologies, and the dynamic begins with a well-studied and common phenomenon known as group polarization.
Group polarization is the phenomenon where members of a doxastically homogeneous deliberative group predictably move, imperceptibly to themselves, towards a more extreme version of the view they held prior to deliberating. It is important to notice that the trouble with group polarization is that the doxastic shift is driven by group dynamics rather than by reason. When groups polarize, it is not due to the introduction of new information or better arguments favouring a more extreme position; polarization occurs simply as the psychological consequence of immersing oneself in what Cass Sunstein has described as an epistemic enclave, a cognitive environment of relative unanimity where one hears "louder echoes" of one's own voice (2007, p. 13). And as Sunstein has noted repeatedly (2003;2009), the technology that structures most of our political communication enhances individuals' ability to preselect the political valence of their interlocutors and even their news and information. As a result, discussion within epistemic enclaves is rampant, and group polarization prevails.
Now, the antidote to group polarization is doxastic heterogeneity amidst social norms that invite disputation and welcome dissent. In short, deliberating groups need to take steps to ensure that critical voices are encouraged and heard; they need to inoculate themselves against their tendency to construct echo chambers (Sunstein, 2003). It may seem an easy fix, yet such countermeasures are more difficult to implement than one might suppose. Consider: group polarization tends to encourage a closely related phenomenon, epistemic closure. 12 As groups polarize, they become less able to countenance the possibility of reasoned and sincere disagreement; opposing views come to sound like confused noise, critics begin to look craven and ignorant, and the view favoured by the group comes to be regarded as the only rational view there could be. As there's obviously no point in trying to argue with craven noisemakers, members of polarized groups become less able to deliberate with anyone who is not already within their fold. The strong sense of an epistemic in-group (and out-group) encourages yet another closely related pathology-namely, the epistemic marginalization of dissenting voices; this involves not only the tendency to decline to engage in deliberation with dissenting others, but also the denial of their epistemic capacities as such. The epistemically marginalized are not merely ignored; they are overtly regarded as incapable of knowing, or even of serving as sources of information. Such marginalization is obviously correlated with other forms of social disadvantage, including violations of democratic equality. 13 It's not difficult to see, then, that in real-world political deliberation group polarization, epistemic closure, and epistemic marginalization operate in a dynamic of mutual reinforcement. 14 The degree to which a group is polarized tracks the degree to which members fail to recognize their critics as even rational, let alone as possibly correct or even as sources of valuable information. Call this the polarization dynamic. It goes without saying that the polarization dynamic is poisonous from the perspective of deliberative democracy. Recall that deliberative democracy is premised on the idea that stark disagreement over public policy is possible among well-intentioned, sincere, and duly informed democratic citizens. The polarization dynamic not only dissolves civility, but also disables public deliberation by encouraging the idea among the citizenry that ultimately there is nothing to deliberate about because reasonable disagreement is in fact not possible. Accordingly, the directive to group members to welcome dissent and invite criticism might be useful for preventing polarization, but is of limited help in counteracting polarization once it has emerged within a deliberating group.
Sunstein's own prescription hence is to introduce legal measures that could limit a doxastic group's capacity to enclave. These proposals rely less on group members' inclination to welcome dissenting voices and more on institutional design aimed at making political echo chambers more difficult for groups to construct. Among his more notorious suggestions is that politically extremist websites should be legally required to carry links to opposing websites (Sunstein 2007, p. 204). Of course, the efficacy of this policy still depends largely on individual visitors' willingness to actually follow the opposing links and investigate the opposing viewpoints open-mindedly. And it is not difficult to imagine ways in which Sunstein's envisioned "must carry" laws could be subverted so that they contribute to group polarization. To see this, consider a politically progressive site that features dozens of opposing links, but only to the most unhinged and irresponsible conservative sites. This would serve to confirm the progressive group's favoured image of their opposition, and thus would contribute to their polarization. So maybe there is no failsafe against the polarization dynamic, but there still could be legal interventions to combat it.
In fact, one could argue that our current media environment is well suited to the task of combating group polarization. Since actual democracy has taken a deliberative turn, it is difficult in our day-to-day lives to escape the clash of political viewpoints, much less deny that there are clashes of this kind. News outlets, televised and online, are now almost entirely devoted to report-and-discuss formatting, where a host first presents a story, and then moderates a brief panel discussion among proponents of differing political perspectives. Viewers are explicitly tasked with weighing reasons and evaluating the arguments presented by the panelists. Online media allow for ongoing exchanges of views and arguments among citizens who otherwise would not interact. And, judging from the popularity of politically oriented news programming, online sites, and social media, citizens are largely interested in participating in public deliberation. From the perspective of our communications and media technologies, we should be living in a deliberative democrat's paradise. Yet, as we all recognize, public political discussion is horrendous, both morally and epistemically. What's going wrong?

IV. NEW EPISTEMOLOGICAL TROUBLE
A lot of work on deliberative democracy intentionally employs an avowedly simplified model of deliberation. Often, it is presumed that there is a single question under consideration, which admits only of a binary, yes-or-no response, and the deliberation is conducted by only two parties. The deliberating parties are taken to be addressing only each other, each evaluating the other's reasons while also proposing arguments of their own that are designed to move the interlocutor. Of course, no deliberative democrat is really committed to the idea that real world deliberative encounters are so simple. The typical models are intentionally simplified for purposes of theoretical manageability; everyone acknowledges that actual political deliberations will be far more complicated. But if public deliberation can't be made to look theoretically appealing under highly idealized conditions, there's no reason to think it worthwhile under more complex circumstances. Nonetheless, simplifying measures can sometimes omit too much, rendering a model unduly simplistic and hence unable to capture relevant phenomena.
What standing models of public deliberation seem to omit is that even when deliberation is indeed conducted between only two parties who are explicitly addressing only each other, deliberative exchanges are frequently nonetheless public performances enacted in the presence of an onlooking audience. In fact, in our current communications environment, public deliberation is most frequently conducted for the sake of the onlookers. That is, although the participants in the deliberative exchange might explicitly address each other, they are often implicitly addressing the audience as well, and it is the latter that is their central, yet only implicit, focus; the reasons entered into the deliberative exchange by the deliberators are commonly designed to move the audience rather than convince the interlocutor. Importantly, the audience typically comes to the exchange for the sake of gaining information about the issue under debate. Perhaps more commonly, the audience views the debate for the sake of seeing how their favoured view stacks up against its competition. They look on precisely because they want to see how their favoured view overcomes or prevails against the opposition; they watch the debate unfold as a means of learning about the dialectical situation that obtains among the positions in play. Even though the onlookers may have adopted a position with respect to the issue under discussion, they are, so to speak, as yet uninformed about the relative strengths and weaknesses of the options in play. They seek information about the relative merits of the competing views by watching those who have this knowledge engage in public argumentation.
Public political argument among interested and purportedly informed deliberators before an audience who may have formed opinions but do not yet know the comparative strengths and weaknesses of the competing views creates opportunities for a new kind of epistemological failure in public deliberation. In our collaborative work, Scott Aikin and I have developed the idea of a dialectical fallacy (Aikin and Talisse, 2014b). Dialectical fallacies are ways in which argumentation fails specifically with respect to the onlooking audience. They are distinctive failures in that a dialectical fallacy can be committed by an arguer who nonetheless does not commit a formal or informal fallacy against his or her interlocutor. Further, the deployment of a dialectical fallacy need not involve any incivility towards one's interlocutor. Thus, when an arguer commits a dialectical fallacy, he or she need not have thereby violated any of the standing moral or epistemological requirements identified by extant conceptions of deliberative democracy. And yet that arguer will have acted in a way that is objectionable from the point of view of deliberative democratic citizenship.
To get a better sense of what dialectical fallacies are, consider the contrast between the informal Straw Man Fallacy and the dialectical Weak Man Fallacy. 15 In the textbook version of the Straw Man, an arguer misrepresents his or her interlocutor's view so that it is easier to refute; the arguer then validly refutes the more flimsy version of this interlocutor's view, but presents himself or herself as having refuted this interlocutor. Note that, to describe the Straw Man, it is necessary to refer to an audience to whom the misrepresentation is projected. 16 The Weak Man also involves a misrepresentation projected to an audience, but is importantly distinct. An arguer who commits the Weak Man is one who seeks to discredit a view by engaging with an especially inept proponent of it; the arguer then validly refutes this proponent's actual argument, but presents himself or herself as having refuted the best the opposition has to offer. Unlike the Straw Man, the Weak Man need involve no mistreatment of the perpetrator's interlocutor; indeed, the Weak Man can be deployed with the utmost respect, fairness, civility, and epistemological integrity towards one's discursive partner.
Crucially, successful deployments of the Weak Man are formally sound and informally cogent; proper Weak Man arguments indeed refute one's interlocutor. The fallaciousness of the Weak Man occurs entirely at the level of the onlooking audience; the perpetrator has misrepresented not his or her specific interlocutor's argument, but rather has projected a distorted view of the dialectical situation that obtains between that interlocutor's view and its opposition. The perpetrator has presented the state-of-play in the dialectic as one in which his or her own view obviously prevails against its opponents, all of whom are at least as feeble as the one he or she has just refuted decidedly. The onlookers are thus left with the impression that there is but a single viable view in play, and the best opposition to it is easily shut down. When fully successful, the Weak Man creates the impression among the audience that there is no use in seeking out or listening to further opponents of the prevailing view, since the best of the opposition has been decisively rebuffed.
In the context of political argumentation, the Weak Man serves to shut down public deliberation as such; it overtly promotes the idea that there is ultimately nothing to deliberate about since there is but one responsible view in play and all other perspectives are weak and confused. The attempt to deliberate with those who disagree comes to be seen as a waste of time; what one's opponents need is liberation from their ignorance, not an exchange of reasons among equals. The Weak Man hence produces conditions ripe for the polarization dynamic.

Consider as a second example what Scott Aikin and I have playfully called
Modus Tonens (Aikin and Talisse, 2008). Modus Tonens is the tactic of restating an interlocutor's claim in an incredulous tone of voice. Now, assuming that proper argumentative exchange allows for some degree of biting and snarky engagement, incredulously restating what an interlocutor has said is not necessarily out of bounds; one is surely permitted to express exasperation and surprise when signaling to an interlocutor that one finds what he or she has said in need of clarification or more deliberate affirmation. But like the Weak Man, Modus Tonens is deployed for the sake of projecting to one's audience a particular conception of the dialectical situation between the views. More specifically, Modus Tonens is the attempt to project to the onlookers that one's interlocutor is dialectically subordinate, someone who needs additional prompting and special assistance in articulating his or her own views. In a successful deployment of Modus Tonens, the perpetrator presents himself or herself as the teacher of the interlocutor, the more intellectually mature party to the discussion who must hence enact and enforce proper norms of serious intellectual discussion. Again, the tactic does not necessarily involve any mistreatment of one's interlocutor, but it serves to project to the relatively uninformed onlookers the view that, of the positions in play, only one is worthy of serious discussion.
The Weak Man and Modus Tonens are but two kinds of dialectical fallacy. There are many others. But I will not labour the point by cataloguing them here. 17 The important thing to note is that when public argumentation occurs in the presence of an onlooking and as-yet uninformed audience, interlocutors are incentivized to implicitly address the onlookers in a strategic way. In such cases, arguers may address their reasons directly to their interlocutor, and in this they might violate no standard norm of civility or principle of proper epistemic conduct owed among deliberative partners. That is, interlocutors might trade only in mutually acceptable reasons, sustain a respectful and unaggressive tone, listen sincerely to each other, invite objections and questions, and so on, while nonetheless arguing with a view towards projecting to the onlookers a particular conception of the dialectical situation that obtains among the interlocutors and their respective positions. These projections can serve to miseducate the onlookers in ways that serve to disable deliberation among them. To return to the case of a successful deployment of the Weak Man: the onlookers will be convinced that there really isn't anything to deliberate about. They will hold that there is but one position that is well-informed and defensible, and all of its critics have been handily repudiated; and they will conclude that anyone who sees fit to engage the question any further must be badly misinformed and thus not worth arguing with. Thus, enclave deliberation is encouraged, and the polarization dynamic is set in motion.

V O L U M E 1 2 N U M É R O 1 H I V E R / W I N T E R
The crucial point bears repeating once again: when the polarization dynamic is initiated by the deployment of dialectical fallacies, no public deliberators need to have behaved uncivilly or in an epistemically improper manner toward their interlocutors. The perpetrators of dialectical fallacies can have clean hands from the perspective of their actual deliberative encounters. What renders them criticizable is something about their deployment of legitimate forms of dialectical exchange with their interlocutors. Perpetrators of dialectical fallacies opportunistically exploit the fact that onlooking audiences are often relatively uninformed about the issue under debate, and indeed are watching the debate for the purpose of learning about the respective merits of the disputing sides. The engagement with the actual interlocutor hence becomes incidental with respect to the actual aim of moving the audience by constructing for them a projection of the state-of-play among the disputants and the opposing views. In the cases most worth the attention of deliberative democrats, dialectical fallacies are deployed for the sake of projecting to an audience the view that there is but one intellectually responsible and defensible position to take on a given question, and thus all opposition is misguided, ignorant, or vicious. This kind of conception of the dialectical situation among the going positions in a dispute sets the polarization dynamic in motion, and thus undermines deliberative democratic practice by attacking the very presumption upon which it relies-namely, that there could be reasoned and sincere but stark disagreement among duly informed democratic citizens over important matters of public policy.

V. BLEAK UPSHOTS
Now, it seems to me that something like the account presented in the previous section is at work in a great deal of our current politics. We confront conditions where democratic citizens are not only increasingly enclaved, but also increasingly of the view that there could be no reasonable opposition to their political perspectives. As one would expect under such conditions, our public discourse is saturated with medicalized accounts of political disagreement; for many citizens, those with whom they disagree are not to be reasoned with but simply diagnosed as cognitively or morally impaired. The seemingly endless parade of political panels and debates operate as public spectacles that pantomime deliberation when actually serving only to confirm audience biases. In short, although our political practice has taken a deliberative turn, we are now living within a simulation of deliberative democracy, a context where our aspirations and attempts to realize them are systematically turned against themselves, resulting in a distortion of our democratic ideals and their further dissolution. However, I won't continue lamenting in this way; I don't think there's anyone who would vigorously dispute the claim that the contemporary state of democracy is worthy of serious concern. I will conclude instead by identifying a few dispiriting upshots of the foregoing analysis for democratic theory.
First, the account offered presents a supplement to Sanders's initial concerns. Recall that she cautioned against the deliberativization of democracy under political conditions where certain forms of inequality prevail; she argued that, when conducted amidst social inequality, deliberative democracy will simply reinforce those inequalities. The argument above suggests a similar lesson, but with an epistemological bent. Roughly: when conducted amidst a population that is epistemologically unequally situated, in that some are informed about the political issues under discussion whereas others are not, deliberative democracy creates distinctive opportunities for strategic arguers to create epistemic enclaves among the less-informed, thereby initiating the polarization dynamic and disabling public deliberation. To put this point in a different way, deliberative forums and episodes of public deliberation may be helpful when it comes time for antecedently well-informed citizens to evaluate the reasons in play with regard to a given policy issue. But public deliberation is not the right format for finding out what reasons there are; it is not the right way to set about informing oneself about an issue. It seems then that deliberative democracy requires robust nondeliberative institutions by which citizens can prepare to deliberate; it can function only against a social epistemic backdrop of nondeliberative but shared sources of reliable information.
The trouble, of course, is that this very suggestion seems to contravene much of the spirit of deliberativism. After all, it is a call for what looks like a return to oldfashioned media and news formatting, where a newscaster presents the day's stories of note, and those stories become the basis for subsequent deliberations. Familiar hazards abound with this model. And one suspects that when fully elaborated, the suggestion involves a subtle form of epistemic paternalism, where citizens' everyday political talk must be facilitated and curated by experts who supply the framework and parameters within which citizen deliberation is to occur. This always involves the risk of degrading into something decidedly nondemocratic. 18 Still, the reality remains that our media and communications technologies have already made the deliberative turn. Our politics is now increasingly conducted by means of pro-and-con argumentation and discussion performed purportedly for the sake of helping citizens to become informed and make up their minds about the pressing political issues of the day. The actual result of all the talking is that our politics have become increasingly argumentative and disagreeable, but far less reasoned and almost entirely devoid of actual disagreement. In fact, we now confront a media landscape of opposed "political realities," each with its own unique markers of epistemic reliability and norms of civility, sharing so little intellectual and moral ground that no discourse across such "realities" is possible. Amidst a permeation of pantomimed public deliberation, deliberative democracy is undermined.
The second upshot is that these conditions might not be remediable. Under existing conditions, democratic citizens need to be able to counteract the polarization dynamic. And this requires in many instances the ability to discern and diagnose dialectically fallacious performances of public argumentation. Yet, as we have seen above, the very idea of a dialectical fallacy is complicated. In fact, the entire enterprise of analyzing and assessing instances of argumentation calls for a family of robust and sometime unwieldy concepts, arguably an entire metalanguage not unlike the kind deployed in formal logic. For example, a few moments' reflection on the concept of hypocrisy demonstrates the need for a fairly robust menu of epistemological concepts: in order to say what hypocrisy is, one must introduce, at the very least, distinctions between saying and doing, and intending and not-intending that are philosophically far more slippery than is often noticed. Or, to consider the matter from a slightly different direction, it looks as if any account of hypocrisy will have to investigate a possible conceptual connection between instances of hypocrisy and instances of lying; but we know that it is surprisingly difficult to state clearly what it is to lie. 19 Matters get only more complex when one attempts to devise the requisite theoretical tools for assessing and evaluating speakers' performances in argumentative encounters.
Yet deliberative democracy requires such tools to be both forged and mastered. Prima facie, the requirement looks doomed; one should not build into one's conception of democratic citizenship the requirement to master a substantive theory of argumentation. It may be retorted that what is needed is not that citizens learn argumentation theory, but only that they gain competence in the metalanguage by which argumentative performances can be assessed.
This retort is correct as far as it goes. However, the current state of argumentation theory does not allow for a clean distinction between the concepts employed in the metalanguage and the theoretical apparatus designed to elucidate them. To put the matter starkly: it is difficult, even among the professionals, to keep the theory and the theorized phenomena distinct. What looks to one argumentation theorist as a garden-variety case of some specific fallacy will look to the proponent of a different theory of argument as no fallacy at all. One worries that in order to get deliberative democracy right, one must first complete the task of argumentation theory as an academic discipline. Such completion is a long way off. And democracy can't wait.
I conclude with a third and related upshot. No matter how things stand with respect to the demandingness of the theoretical apparatus needed to detect and diagnose fallacious argumentative performances, deliberative democracy remains a demanding proposition. As argued above, the deliberative ideal involves not only a set of moral requirements; there are epistemological requirements, too. In light of the discussion of dialectical fallacies above, we might say that deliberative democracy calls for a substantive epistemological ethic, a set of norms delineating what one owes, epistemically, to one's interlocutors and to one's audience. These norms will include prescriptions outlining when one must concede a critic's point, revise one's view, change sides, be silent, admit one's error, suspend judgment for the purpose of gathering more information, and much else. We all know how difficult it is to abide by such norms in the context of relatively low-stakes exchanges at academic conferences and departmental meetings. To expect citizens as such to adopt them, and enact them reasonably successfully, in the relatively high-stakes contexts of politics seems naïve.
Of course, deliberative democrats will concede that the deliberativist epistemological ethic is demanding. They will agree that existing citizens are unlikely to embrace the requisite norms. However, they will next add that deliberative democracy must be learned, cultivated, and practiced. Again, this is correct as far as it goes. The trouble is that, if the arguments above are roughly correct, it seems that we cannot learn good deliberative democratic epistemological habits by engaging in public deliberation. To repeat my variation on Sanders's lesson: deliberative encounters under existing conditions should be expected to initiate the polarization dynamic. And it is not yet clear to anyone how demanding epistemological norms can be reliably cultivated.
We seem to have reached an impasse. Or, perhaps more accurately, we seem to be caught between two commitments that don't ultimately sit well together. On the one hand, we tend to take our own political views to be competently reasoned and well-informed. Additionally, we tend to take ourselves to be politically fairminded, duly responsive to countervailing considerations, and welcoming of good criticism from formidable critics. Yet, on the other hand, we tend to see the vast majority of our political opponents to be short-sighted, less than rational, ignorant, unwittingly in the grip of various biases, ideologies, and illusions, unwilling to fairly engage with their critics, unable to respond to objections, and worse. We take ourselves to be fully invested in the democratic project, and often see that project as involving public discourse among citizens who staunchly disagree; yet we often find ourselves unable to formulate the most powerful objections to our own most cherished political commitments. In fact, in unguarded moments, we are inclined to flatly deny that there are any such objections to the view we most deeply hold. Accordingly, we tend to see democracy's present ills wholly as the result of others' failings. We think democracy would get back on track if only everyone else would see the light. Too often, we tacitly think that when others "see the light," they will adopt our own most cherished political beliefs; we attribute the facts that our preferred candidate lost the election and our preferred policy lost at the polls to the ignorance, gullibility, immorality, and selfishness of others. Crucially, we never attribute our political wins to those same forces; when our side wins in democracy, it is always due to a triumph of public virtue and good sense. Consequently, we tend to see democratic progress as requiring deeper and expanding levels of unanimity, and an increasingly diminishing field of matters about which there could be reasoned disputation. Perhaps the deliberative turn has failed us all.