Civic Virtue in the Deliberative System

The normative stability of a deliberative and democratic political order and the creativity and quality of the decisions its produces depend on citizens developing civic orientations and capacities through participation in deliberative events aiming at the cooperative solution of political problems. That, at least, is the claim made by critics of the systems approach to deliberative democracy, who argue that its proponents have lost sight of the educative function that respectful public reasoning plays for citizens. In this article I offer a response to this line of argument. There is no good philosophical reason to suppose that only unitary deliberation can perform an educative function for citizens. The kinds of informal and uncooperative public speech that occur in distributed deliberative processes can also develop participants’ civic capacities and civic virtue – and not merely through their systemic effects. This is an insight that should encourage us to rethink the design and facilitation of deliberative forums and pay more attention to citizens’ everyday deliberation. Author Biography Daniel Hutton Ferris is a PhD candidate at the University of Toronto, studying under Melissa Williams, Simone Chambers and Andy Sabl.

Lamentation about low quality public discourse is a fundamental driver of the deliberative democratic research programme.Many democratic reformers, including a number of contributors to this journal, suggest that part of the solution to our discursive deficit might be the multiplication of minipublicsdesigned temporary forums for deliberation between citizens about matters of public concern-in which people can learn how to engage in cooperative and respectful public reasoning (Fung, 2003;Smith, 2009).Since the 1970s they have helped to trigger a profound shift in political life around the world: minipublics have come to be sequenced into more and more policy processes and a public-engagement industry has emerged to design and facilitate them (Lee, McQuarrie, & Walker, 2015).And yet some of the political theorists who once championed cooperative public reasoning appear to have started losing interest in it.Many deliberative democrats now argue that reformers seeking to generate democratic legitimacy or sensitive decisions should be primarily concerned with the quality of the overall deliberative system, in which a wide variety of different modes of communication occur (Dryzek, 2010a;Parkinson & Mansbridge, 2012).Proponents of this 'systems approach' might judge a deliberative system healthy even if it is home to very little face-to-face deliberation between citizens that looks respectful, reasonable and cooperative.
One important criticism of this systemic turn is that is loses sight of the fact that reasonable communication plays a crucial role in educating citizens to civic virtue.A substantial body of empirical findings appear to demonstrate that citizens develop civic orientations and capacities through participation in what, following Goodin (2005) and Mansbridge (1983), we might call 'unitary' deliberation (e.g., Pincock, 2012).And David Owen and Graham Smith (2015) have recently argued that these orientations and capacities are necessary conditions for the normative stability of a political system or its capacity to produce sensitive decisions.They claim that by effectively abandoning the attempt to improve certain crucial kinds of civic virtue, proponents of a systems approach risk advocating political reforms that undermine the production of very basic political goods.
In this article I make the negative claim that, pace Owen and Smith (2015), there is no good philosophical reason to suppose that only unitary deliberation can perform an educative function for citizens.When deviations from unitary public reasoning occur in minipublics they can sometimes improve, rather than degrade, deliberative quality, because they are a way in which competent communicators can acknowledge and respond to the lack of unity that always exists between themselves and even sympathetic interlocutors.And minipublics that are less cooperative may sometimes cultivate civic virtue better than those with the cooperative interactions preferred by classical models of deliberative democracy.But if division can be as desirable as unity at the local level, there is no reason to expect civic virtue to be generated better by a unitary and cooperative deliberative process than by one networked across multiple forums or communicative channels and populated by adversarial as well as cooperative participants.Rather than giving up on civic virtue, the systems approach permits a more nuanced understanding of how it might be cultivated both in everyday talk between citizens and in the context of structured deliberative events.

I.
Owen and Smith's Argument against Distributing Deliberation.
Classical theories of deliberative democracy, for instance those articulated by Cohen (1997) and Rawls (1997), tend to implicitly take a 'face-to-face' process of public reasoning as the exemplary case of deliberation.This process is conceived of as 'unitary,' in the sense that the purposes of the various parties are treated as shared and hence their interaction can be viewed as a cooperative game in which there is no conflict of interest (cf.Goodin, 2005, pp. 183-187;Mansbridge, 1983, pt. I).Certain norms of public reason are viewed as appropriate for this kind of fully cooperative communication.Most importantly, interlocutors should sincerely discuss content which is mutually recognised to be both relevant and correctly characterised and draw on forms of reasoning and argumentation mutually recognised as valid.And classical theories of public reason tend to imply the 'unity of the deliberative virtues': that a wide variety of deliberative goods-for example inclusiveness, respectfulness and rationally-motivated consensus-must be realized simultaneously for any of them to be realized at all.In Habermas's model of communicative action, for instance, speech can only take a rational and respect-embodying form (interlocutors assume that validity claims can be redeemed with good reasons) if it refers to a certain shared content (a background of meaning contained in the lifeworlds of participants) and if interlocutors believe one another to be communicating sincerely (Habermas, 1998).
The exemplary case of deliberation on the systems approach, by contrast, is a semi-institutionalized process of decision-making distributed across a number of networked forums or and communicative interactions.A socially-distributed process of practical reasoning can be viewed as 'deliberative' because it involves "weighing the reasons relevant to a decision with a view to making a decision on the basis of that weighing" (Cohen, 2007, p. 219;cf. Chambers, 2012).The relation between deliberation in this sense and public reason norms is a complex one.In a distributed deliberative process it may be necessary to realise different deliberative goods in different locations.So, for example, Habermas' 'anarchic public sphere' is a more inclusive communicative arena than the parliamentary complex but may rightly involve a less careful making and interrogating of validity claims, since a healthy democracy may need to place more emphasis on broad participation in 'contexts of discovery' and more emphasis on an orientation to rational consensus in the 'contexts of justification' (Habermas, 1996).Those who advocate deliberative sequencing treat unitary deliberation as one means amongst many to achieve sensitive political decisions.And if we take the need to divide deliberative labour seriously then we might in principle judge the quality of deliberation occurring in a political system to be satisfactory without substantive public reason norms being realised anywhere within it.Owen and Smith (2015) argue that we can't realize the goods public reason was supposed to help provide without actually relying on generalised public reasoning between citizens to secure them.It is not possible, they argue, to separate the ends that deliberation would ideally realize-sensitive decisions that can be accepted as normatively legitimate by free equals-from processes of unitary deliberation in accordance with substantive norms of public reason.
Their argument depends on the importance of certain civic virtues that unitary deliberation has the capacity to develop: [A] 'distributed' approach to deliberative systems sets aside any concern for the deliberative capacities and powers of citizens.It thus loses sight of two important points.First, part of the political ideal of deliberative democracy is that its (normative) stability is generated by citizens being able intelligibly to conceive of (adopt a stance towards) themselves as equals engaged in a process of public reasoning oriented to a shared practical judgment, where such a process involves citizens reflectively taking up each other's standpoints.Second, part of the importance of actual (unitary) deliberation as a practice of public reasoning is that it is a creative process in which novel shared reasons can emerge within the activity of reasoning together as equals.The reasons to which rule is responsive are liable to be process-dependent and bound to whether citizens in such a process can conceive of themselves as reasoning together as equals.There is no reason to think that a 'distributed deliberation' process would track this aspect of unitary deliberation.(Owen & Smith, 2015, p. 219) It is worth reconstructing the two related claims that are being articulated in this highly condensed argument.Owen and Smith are arguing, first, that citizens would not support a more deliberative and democratic political order unless they could view themselves as being genuinely treated as equals by it-and that this could only occur if they think of themselves as included in a group with a united political fate which is determined together.The requisite sense of political life as cooperative on a deep level could only emerge if citizens were civic-minded (i.e., inclined to promote collective, as well as merely private, goods) and if they possessed firmly-rooted expectations of civic reciprocity and particular civic capacities (e.g.information about co-citizens and policies).And these orientations, expectations and capacities can only be developed broadly across society if citizens generally are well-versed in imaginatively standing in one another's shoes.Since participation in concrete instances of unitary deliberation can generate a sense of themselves as cooperating equals, it also reinforces commitment to a more deliberative and more democratic political order. 1 Owen and Smith's (2015) second claim is that unitary deliberation can generate creative and therefore high-quality decisions in a way that other forms of political speech cannot.A key attraction of deliberative theories of democracy is that, rather than viewing politics as an inevitably zero-sum game in which individuals compete to realize only contingently-compatible preferences, they pay attention to the fact that communication can sometimes crystalize or transform preferences in ways that produce resolutions to conflict situations which all parties can see themselves as benefiting from.Deliberation "can involve the search for creative alternatives that can meet the interests of all sides better than each initially thought possible" (Stevenson & Dryzek, 2014, p. 16).
Owen and Smith are concerned that a political theory which de-emphasizes unitary deliberation demands less of citizens than one which hopes for a realization of cooperative political debate in general and thereby valorizes an adversarial vision of democratic life that closes down opportunities for genuinely creative problem solving.Higher quality outcomes require opportunities for citizens to develop cooperative orientations through participation in unitary deliberative forums.
These arguments are both important and prima facie plausible.They cohere well with classic theoretical accounts of deliberative democracy.And a fuller statement of them could draw on a range of suggestive empirical evidence linking a cooperative orientation to interlocutors with the production of a variety of political goods: 1. Experimental research on social dilemmas.Face-to-face discussion is the single best way to improve cooperation between individuals facing social dilemmas (Sally, 1995).And cooperative behaviour is observed more often in situations governed by norms of reciprocity or when participants perceive that their counterparts are committed to achieving a cooperative outcome (Kopelman, Weber, & Messick, 2002, pp. 135, 145).

Experimental research on perspective taking.
There is evidence that individuals who engage practices of perspective-taking express more 1 While Owen and Smith (2015) do not tell us what they mean by the phrase 'normative stability', it seems likely that they are referring to something equivalent to what Rawls called 'stability for the right reasons'.The stability of a political order, for Rawls, is a measure of how likely citizens growing up within it are to comply with it and its normative stability is distinguished from a mere modus vivendi by being regenerated not because it pragmatically suits a certain array of social forces on whose support the regime depends, but because the members of different social groups growing up in that regime come to affirm its constitutional essentials as just and the laws it produces as legitimate by the lights of their most deeply held values (see Rawls, 2001Rawls, , pt. 5, 2005, pp. xlii-xliii, 140-150, 385-396), pp. xlii-xliii, 140-150, 385-396).
empathy for and altruism towards others, to engage in less stereotyping of others-even unconscious stereotyping-and are in a better position to ameliorate intergroup conflict practices (Galinsky & Ku, 2004;Moskowitz, Gollwitzer, Wasel, & Schaal, 1999).3. Survey data from deliberative polls.If we assume that communication in Fishkin's deliberative polls is largely cooperative (but see Azmanova, 2010) then research on their effects may support for the claim that unitary deliberation can help generate normative stability for the democratic system and appropriately-sensitive decisions.Participation seem to significantly increase peoples' political knowledge. 2It also appears to induce a significant number of participants to become more public spirited.For example the percentage of participants in a group of Texan deliberative polls who were willing to pay more for utilities in order to support conservation measures benefiting the community as a whole rose by 30% on average (Fishkin, 2009, p. 142).And participants in a deliberative poll from fifteen towns around New Haven became much more likely to favor revenue-sharing between the towns over the course of their discussions with one another.3 The unity of the deliberative virtues thesis, which informs most theories of public reason, is doing important work in Owen and Smith's argument.Talking together can generate sensitive decisions and normative stability only if it is respectful and inclusive and oriented to justification of truth claims (and so on).
The difference between communication instantiating just some deliberative virtues and that which abides by substantive public reason norms is one of kind, rather than degree, in terms of their capacity to generate civic-mindedness and civic-capacities.If this claim is true then deliberative systems theorists are misguided if they believe that a distributed deliberative democratic process can be substituted for a unitary one and may end up advocating a kind of democracy that is insensitive to normatively unstable and cannot produce creative decisions.

II. The Division of Labour in all Interpersonal Deliberation
While Owen and Smith (2015) are right to emphasize the positive effects of debate between citizens, they may overemphasise the extent to which those consequences can only be produced by unitary deliberation in which citizens share a common purpose and adhere to substantive public reason norms.What they fail to take seriously enough are the multiple productive divisions and distances between individuals which occur in every kind of real-world deliberation, 'unitary' as well as 'distributed'.Deliberation in the real world occurs between individuals with distinct experiences and understandings and, because it has a time-limit, is not perfectly open-ended.It combines the 'togetherness' characteristic of unitary deliberation with a division of deliberative labor.As participants in real-world deliberation we may well discover that there is no way we can in good conscience align our understanding and will with that of the group, that a fusion of horizons may not in this instance be possible, given the temporal resources available to us and our different hermeneutic starting points.If we are competent and reflective deliberators, then we know that, because of this distinct possibility, we are precluded from wholly 'pre-committing' to full cooperation with a deliberating group.We must always be ready, if often unwilling, to detach our purpose from that of the rest of the group in certain conditions-and thereby break the group's unity and cooperative orientation-if we feel unable to understand things as others do or to reasonably assent to what the other members of the group thinks should be done.
There are both theoretical and practical reasons for thinking that this 'failure to commit' does need not signal a strategic orientation to self-interest narrowly understood.Deliberators enter communication with different personality traits (Jennstål & Niemeyer, 2014) and with different experiences and understandings, which they are normally at least marginally better at interpreting than their co-citizens.That is not to say that I have special access to a beetle in my box (Wittgenstein, 2009, sec. 293) but merely to point out that we often do, as a matter of fact (and perhaps especially in the particularly muddy terrain of normative judgement), have a better insight than others do into how we feel and are therefore in a better position to articulate why something strikes us as reasonable or not.This is not to deny that others can sometimes be good at interpreting our experiences or even have insight about us and our needs that we ourselves lack.Parents, for example, may regularly articulate the needs of their small children more adequately that those children themselves.But respect for the dignity of co-citizens should make us wary of those who claim this sort of interpretive superiority in the context of reasoning together as citizens.
High quality political communication often requires participants to recognise disunity of purpose in the deliberating group and respond to it by sacrificing certain deliberative virtues for others and violating certain norms of public reasoning.As William Rehg (1997;cf. Ferrara, 2008;Zerilli, 2012) has pointed out, sensitive acts of judgement that permit us to weigh reasons properly are not just logical, mechanical or dialectical procedures, but also a matter of appropriate psychological disposition.This means that we may well have to present reasons in a way that inflects them with a certain emotional resonance if we want to encourage interlocutors to weigh them properly.We may want, for example, to steer our interlocutors between rash and over-cautious assessment of the reasons we offer by inducing fear or pathos.In doing so we might violate substantive norms of public reason and we might have to acknowledge a division in the purpose of the deliberating group uncharacteristic of unitary deliberation.But, by facilitating the sensitive judgement necessary for genuine communication, this may ultimately improve deliberative quality.
There may be a world of normative difference between unitary and distributed deliberation-and their relative capacity to generate civic virtue-considered at a very high level of abstraction.But treating the two kinds of discourse as classes of real-world deliberation, rather than abstract ideal types, reveals them to be continuous with one another in various ways rather than cleanly disjunctive.Unitary and distributed deliberation would not be able to serve as substitutes for one another if the former involved a realisation of all the deliberative virtues at once, while the latter necessarily involved sacrificing some of them.But fully cooperative public reasoning between interlocutors who share a single purpose is an ideal type, a counterfactual which only can be approached asymptotically under real world conditions.We should not necessarily expect there to be a categorical difference between the civiceducational effects of a more cooperative exercise in face-to-face public reasoning and a less cooperative deliberative process, even if the latter is distributed across multiple communicative exchanges and engages actors that sometimes take an adversarial orientation to one another.
There are three conclusions that might be drawn from these considerations.
First, those who design and facilitate deliberation in minipublics should understand the choice of how to structure deliberation as an optimisation problem that involves inevitable trade-offs between deliberative virtues.We do not face a choice about whether participants would ideally be take a wholly cooperative orientation to one another-a 'deliberative stance' (Owen & Smith, 2015)-or not, but a choice between how the inevitable and healthy interplay between unity and division should play out.Rather than asking themselves how they might approximate an ideal of public reason, institutional designers should ask themselves which public reason norms should be encouraged, and which relaxed, in a given deliberative process.
Second, since unitary deliberation is not an unmitigated good, we should not necessarily judge deliberative processes as failures if they do not approximate it very closely.A minipublic that was wholly cooperative might well be improved if participants became less willing to cooperate, for example if one participant started playing devil's advocate.Deliberation with some disunity in the group, or an outright division of labour, is not necessarily a normative second-best.Imagine a spectrum of communicative acts, ranging from wholly unitary and cooperative to wholly adversary and competitive.The ideal deliberative interaction for a given time and place does not lie at one end or the other, but, in a world of interpersonal pluralism and limited time for discussion, somewhere in the middle. 4hird, there is no principled reason to suppose that the virtues Owen and Smith identify with 'unitary' deliberation cannot also be produced by the other kinds of communication that structure the deliberative system.Their argument that a distributed deliberative process cannot serve as a substitute for a unitary one is grounded in the idea that the latter involves a kind of commonality of purpose that is absent in the former and that makes in decisively better at generating solidarity.As I have argued, however, all interpersonal deliberation involves salutary disunities and distances between participants.We should, then, not expect there to be a categorical difference the capacity of more and less unitary deliberation to generate valuable ethical motivations and orientations.Distributing deliberation across multiple institutions and communicative processes, each of which realize certain deliberative virtues at the expense of others (à la Goodin, 2005), might generate civic virtues (and therefore normative stability and creative decisions) just as well as, or better than, multiplying opportunities for citizens to engage in unitary deliberation.
Any serious consideration of real public speech quickly reveals the empirical import of these claims.Work on the political use of rhetoric has shown that violations of public reason may have positive effects (Chambers, 2009;Dryzek, 2010b;Medearis, 2005;Young, 1996;Zerilli, 2012).'Expressive' speech may embody certain deliberative virtues: it may attempt to present an authentic expression of the situation of some group; it may embody openness if it communicates insights of specialist discourses to a wider audience; it may be argumentatively structured around certain validity claims or some notion of a common good; and it may embody respect by employing, for example, elaborate forms of greeting.And yet rhetoric often realizes certain deliberative virtues at the expense of others.Someone using it may sacrifice sincerity in order to employ techniques such as irony or choose to characterize content controversially and make use of highly condensed, stylized or figurative forms of reasoning which would not be recognized as valid by just anyone.
Owen and Smith recognise this point to a certain extent.They are happy to concede it when framed as a trade-off between local relaxation of certain deliberative standards and positive effects at the systemic level (Owen & Smith, 2015, p. 221).Following Dryzek (2010b), they concede that Pauline Hanson's expression of extreme anti-immigrant views violates public reason norms in a profound way but may have contributed to an improvement of public debate in Australia by allowing certain marginalized discourses to crystalize and be taken into consideration at the political center.Owen and Smith are, however, much less inclined to accept the idea that deviations from the norms of public reasoning appropriate for cooperative discussion in unitary forums can have positive effects at the local level, rather than elsewhere in the deliberative system.They offer a challenge: "deliberative systems theorists need to show why non-deliberative processes [by which they mean those that do not conform to the unitary ideal] are to be favoured over potentially deliberative ones with an account that does not conduct its evaluation only at the systemic level" (Owen & Smith, 2015, p. 226).
While I have already provided the raw materials for fashioning a response to Owen and Smith's challenge, it will helpful to provide a concrete example of the kind of communication whose existence they doubt.Though it was not itself part of an exercise in unitary public reasoning, a speech delivered by Obama in 2016 serves my purposes.Surrounded on stage by victims of gun violence and their families, he reminded his audience that "every single year, more than 30,000 Americans have their lives cut short by guns-30,000" and that "hundreds of thousands of Americans have lost brothers and sisters, or buried their own children." A number of those people are here today.They can tell you some stories.
In this room right here, there are a lot of stories.There's a lot of heartache.There's a lot of resilience, there's a lot of strength, but there's also a lot of pain.
[…] Every time I think about those kids it gets me mad.And by the way, it happens on the streets of Chicago every day.
[…] I'm not looking to score some points.I think we can disagree without impugning other people's motives or without being disagreeable.We don't need to be talking past one another.But we do have to feel a sense of urgency about it.In Dr. King's words, we need to feel the "fierce urgency of now."Because people are dying.
[…] If you have any doubt as to why you should feel that "fierce urgency of now," think about what happened three weeks ago.Zaevion Dobson was a sophomore at Fulton High School in Knoxville, Tennessee.He played football; beloved by his classmates and his teachers.His own mayor called him one of their city's success stories.The week before Christmas, he headed to a friend's house to play video games.He wasn't in the wrong place at the wrong time.He hadn't made a bad decision.
He was exactly where any other kid would be.Your kid.My kids.And then gunmen started firing.And Zaevionwho was in high school, hadn't even gotten started in lifedove on top of three girls to shield them from the bullets.And he was shot in the head. 5his example is intended to illustrate that speech which recognises the disunity of purpose between the speaker and their interlocutors, the partial absence of a 'we, together', can nonetheless sometimes be preferable to that which does not, even when assessed at the 'local', rather than systematic, level.Obama seeks to persuade his audience not just by 'the orderly exchange of information and reasons' (Goodin, 2005, p. 40) that may be appropriate when the sharedness of a group's purpose is beyond doubt.Persuasion of this kind, for example the allusion to 'your kid' being 'shot in the head', recognises and responds to actual divergence in the aims of deliberators, even as its success depends upon their possible convergence.And yet this was a high quality deliberative act whether we assess it at a systematic or local level, whether we measure the quality of deliberation in terms of its capacity to generate respect, inclusion and epistemic sensitivity (like Mansbridge et al., 2012) or in terms of its effect on decisionquality and normative stability via civic-mindedness and civic-capacity (like Owen and Smith, 2015).Its success was due to, not despite, the fact that it exhibited an awareness of the kind of divisions in all deliberation and the concomitant need to emphasise certain deliberative virtues above others in any real act of discursive practical reasoning.
It is worth summarising my argument in this section.While the 'unity of the deliberative virtues thesis' may hold true at a very high level of abstraction, it has important limitations as a theory about real interpersonal deliberation.Even speech which is imperfect by the lights of public reason, or which recognises some disunity of purpose in a deliberating group, can have highly positive effects.It too can induce high quality decision making and generate stabilizing kinds of normative perceptions of and orientations toward the political system.There is no principled reason why a distributed deliberative process cannot be substituted for a unitary one.This is not to say that the normative situation does not change in important, complex and interesting ways when, say, a deliberating group decides to continue its deliberation by splitting into two subgroups who talk cooperatively amongst themselves and then engage in an adversarial debate.My point is a basic, but philosophically important one: there is no principled reason to expect such a process to be normatively second-best in terms of its capacity to facilitate creative and high-quality decisions or normative stability for the political system.

III. Some Objections.
Before assessing the practical import of this argument, I want to very briefly consider two further worries about the systems approach and deliberative sequencing, suggesting possible responses to them that build on the line of argument just developed.
The first, perhaps rather academic, concern about the systems approach is that it is difficult to consider it a theory of deliberative democracy in a very meaningful sense. 6It could be argued that many deliberative democrats have lost their faith in the more interesting and perhaps radical philosophical commitments of their youth.After a middle age spent trying to find space in their vision of democracy for the pursuit of self-interest (Mansbridge et al., 2010), 'agonistic enquiry' (Bächtiger, 2010) and perhaps even insincerity (Markovits, 2006), they have finally lapsed into conceptual incoherence and a more or less complete retreat from their ealier guiding principles.
A quick glance back at Habermas' work since at least the early nineties undermines this narrative.The fact that he draws a sharp line between communicative and strategic action does not mean that he is insensitive to the tension between 'we, together' and 'we, separately' which structures all communication.His repeated references to the 'decentred' or 'subjectless' nature of public communication (Habermas, 1994(Habermas, , 1996, sec. 7.1.2) , sec. 7.1.2)anticipate the systemic turn because they highlight the fact that, pace Fishkin (2005), we should not understand popular sovereignty in Rousseauian-republican terms of the 'general will' emerging in a particular instance of self-rule between free and equal individuals meeting in a face-to-face forum.Habermas (2003) is aware that total immersion in what Owen and Smith (2015) call a 'deliberative stance'-a orientation that inclines citizens to solve problems cooperatively or not at all-may make deliberators blind to strategic games their interlocutors might be playing with them: "even when taking a performative attitude," he acknowledges, "participants do not allow themselves to be fully consumed lock, stock and barrel by their engagement to the point of not being awareat least intuitivelyof much that they could know thematically by taking an observer's objectivating attitude" (p.107).
A second (and related) worry about the systemic approach it that it may entail a certain kind of elitism that was not present in earlier deliberative theory.There is good empirical evidence that, at least in the U.S., people are often deeply resistant to expressing their political opinions in public, especially when they are unpopular or the issue is a contentious one, but may be more willing to engage in an organized deliberative event.(Eliasoph, 1998;Wells et al., 2017).They might be more willing, however, to engage in an organized deliberative event (Neblo, Esterling, Kennedy, Lazer, & Sokhey, 2010).If deliberative democrats give up on the aim of multiplying opportunities for citizens to engage in high quality face-to-face public reasoning with one another then are they giving up on the project of putting actively engaged citizens 'at the centre' of democracy? 7I don't think so.In fact, we could turn on its head Owen and Smith's (2015) worry that the systems approach fails to put actively engaged citizens 'at the centre' of democratic theory.Rather than accepting or celebrating the rule of various kinds of elites, the systems approach could be seen as reconnecting with earlier participatory models of democracy (Barber, 2003;Pateman, 1970Pateman, , 2012) ) by re-valorising a broad range of participatory and discursive civic practices.
Refocusing from unitary to distributed deliberation need not imply an elitist disregard for the civic capacities of normal people, but rather an attentiveness to the way those capacities can be cultivated in the inevitably messy interactions of normal life, rather than primarily in the context of 'top-down civic projects' (Eliasoph, 2009) that occasionally stand in a somewhat ambiguous relation to democratic values (Lee, 2015;Lee et al., 2015).The systems approach might encourage researchers hoping to better realize democracy to pay attention to the way civic virtue can be cultivated by 'everyday talk' (or "everyday deliberation": Mansbridge, 1999, p. 228).Searing, Solt, Conover, and Crewe (2007) lament the fact that "despite Mansbridge's and Habermas's inclusion of ordinary citizens in their thinking about deliberative systems, these citizens have often been left out of the research picture because their talk seems to be so casual and not directed at producing publicly binding decisions".They suggest that one reason for a lack of research into the capacity of everyday talk to generate civic virtue is the belief that "its differences from structured deliberation, and certainly from ideal deliberation, may be more differences of kind than of degree" and that "it simply may not produce the benefits associated with ideal deliberation" (Searing et al., 2007, p. 611).In this article I have tried to undermine some of the philosophical basis for these commonly-held beliefs.And, as I point out below, there may be good empirical reasons to be wary of them too.Rather than evincing an elitist disregard for the speech and action of ordinary citizens, the systems approach may help to put it back at the centre of democratic theory if it encourages more research on the conditions and consequences under which civic virtues might be cultivated in a more robustly deliberative mass culture.

*
Owen and Smith (2015) are right to remind those who adopt a systems approach to deliberative democracy not to lose sight of the importance, for creative, highquality and normatively stable deliberative and democratic processes, of practices of respectful and cooperative public reasoning between citizens.My negative claim in this paper is that they fail to justify their stronger thesis that unitary and distributed deliberative processes have fundamentally different normative capacities and therefore cannot, even in principle, be substituted for one another.Division in deliberation groups and the violation of substantive public reason norms it necessitates can both be healthy and help educate us to Chambers (2012) implies a passive role for citizens by focusing on the democratic potentials of the use of public opinion surveys.
civic virtue-and not merely through their systemic effects.Sequencing deliberation and prioritizing certain deliberative virtues over others is not just a normatively second best yet 'pragmatic' response to unfortunate circumstances (pace Goodin, as well as Owen and Smith).
The relatively abstract claims made in this paper provide support for several recommendations about the design and facilitation of deliberative events and about the reform of the deliberative system more generally.
First, those who design and facilitate deliberative forums should not shy away from explicitly prioritizing the cultivation of certain civic virtues over others.Widespread acceptance of the unity of the deliberative virtues thesis may have encouraged participation professionals to be reticent about how they have had to make these trade offs.But it should not be viewed as a deviation from a deliberative ideal when designers of minipublics explicitly choose certain civic virtues to be prioritised.They can defend this decision by reference to the role that forum plays within a broader deliberative system that is likely to cultivate a variety of civic capacities and virtues-or by reference to the relative abundance or scarcity of purely local deliberative resources (including time and the amount of overlap between participants pre-deliberation experiences and interpretations).
Second, designers of deliberative forums may do well to recognize the value of disagreement and division as sources of certain kinds of civic virtue and civic capacity.Empirical evidence suggests that contestatory deliberative processes may be better than more consensual ones at stimulating more creative discussions or higher quality recommendations (Nemeth, Brown, & Rogers, 2001;Schweiger, Sandberg, & Ragan, 1986).And yet unitary deliberation in minipublics tends to involve much more agreement than disagreement, partly because expressions of disagreement violate politeness norms, but also presumably due to the expectations and preferences of organizers and sponsors.8When a minipublic is convened for largely epistemic reasons-for exploring all sides of an issue or finding solutions to complex strategic problems-it may be useful for deliberation to be structured as an adversarial debate, or for facilitators to play devil's advocate if dissent from a majority opinion is not forthcoming.Heightened disagreement may be particularly unpleasant and demotivating for some, but this effect could be reduced by facilitators who depersonalize it by portraying it as a valuable and routine feature of minipublics.
The benefits of a more adversarial minipublic may not be only systematic: expressions of dissent could sometimes increase citizens critical engagement with the topic under consideration and perhaps help improve their understanding of their co-citizens' perspectives and their society (Azmanova, 2010).
Finally, researchers seeking to understand and improve civic virtue and deliberation may do well to pay more attention to everyday talk and other forms of informal civic participation.Survey research suggests that people do have politically-relevant discussions quite a lot in informal settings, but that it is mostly with close friends and family. 9The majority of highly respectful political talk-'unitary' in the sense of being oriented to cooperation and consensuswill continue to occur in relatively private contexts or amongst people whose opinions do not radically diverge.Outside of those contexts political talk, even if relatively good-natured, will have a looser orientation to the achievement of shared political judgement.But there is some good empirical evidence that this kind of talk can critically engage citizens in public life and thereby contribute to the development of a wide variety of civic capacities.It has been found, for instance, that: • Citizens who attend either a formal or informal meeting to discuss a public issue are, because they attended that meeting, more likely to engage in volunteer work or in solving a community problem (Jacobs, Cook, & Delli Carpini, 2009).• Citizens who talk about politics in public settings are more likely than others to help their neighbours and feel part of the local community and are more likely to feel as though they are qualified to participate in politics, that the government is responsive to people like them and that people can only reach their full potential through civic participation (Searing et al., 2007).• In unmoderated small group discussions those who hold minority opinions can have important and enduring influences on the majority and thus set in motion productive deliberative contestation.This is especially likely if the group is oriented to encouraging originality or just having a discussion for its own sake (rather than explicitly being oriented to reaching agreement or reducing conflict) and if the minority accept fundamental group norms and share an identity with the rest of 9 Delli Carpini et al. ( 2004) estimate that about seven out of ten Americans have conversations about public issues at least a few times per month; about half of the (American) respondents contacted by Keeter, Zukin, Andolina, and Jenkins (2002) reported talking about politics at home and 60% of them reported talking about the current affairs very often; and Wuthnow (1994) estimates that a quarter of American adults are members of discussion groups, the majority of which discuss current events or political issues.But Conover, Searing, and Crewe (2002) estimate that "two-thirds of the Americans and 70 per cent of the British never or only rarely discuss political topics at social gatherings with people that they do not know very well" (pp. 33-35).
the group, as is likely to often be the case in everyday talk between friends or neighbours (Mendelberg, 2002).
A more robust culture of everyday deliberation could generate civic virtue, build deliberative capacity and ultimately exert pressure for democratization.It is not utopian to believe that citizens could become more willing to express dissent in public and productive ways as changing cultural norms and the reform of key institutions promote a more robust 'deliberative culture' (Sass & Dryzek, 2014;cf. Chambers, 2000).Multiplying minipublics is just one method for helping to achieve this.Researchers might therefore do well to spend more time analysing the way a variety of different institutions structure mass patterns of everyday political talk and how it might be improved.10