Estlund Reading Group Chapter 10

[David’s response to Harry on chap. 9 is now below this post, so don’t miss it — SCM]


In this chapter, Estlund seeks to identify the correct role played by an ideal deliberative situation in democratic theory. He argues that while in practice, democratic communication should not aim to resemble ideal deliberation, nonetheless the idea has an important function as a template through which to examine real-life instances of democratic communication and identify deviations from the ideal. Real deliberative practices and institutions should not aim to mirror the model deliberative situation because when epistemic distortions arise as a result of deviations from the ideal, it may be justified to employ further deviations to remedy these. This leads him to defend a model of wide civility for the informal political sphere, which makes room for sharp, disruptive and even suppressive forms of participation under certain circumstances. This wide version of civility is appropriate only for the informal public sphere, however. In formal political institutions such as the courts and legislatures the norms of narrow civility still apply. In summary then, it seems that there are three main arguments at work in this chapter: (1) that the appropriate way to think of the ideal deliberative situation is not as a set of prescriptions for citizens to aim at, but rather as an analytical tool for diagnosing and remedying failures; (2) that there might be good epistemic reasons to reject the narrow civility inherent in model deliberation in favour of a wider version; and (3) that while the use of countervailing deviations from the ideal might be appropriate in the informal political sphere, formal instances of political deliberation ought still to be governed by the requirements of narrow civility.

Estlund begins by addressing the question of whether the rules of conduct governing the ideal deliberative situation ought to be applied to all aspects of public communication, given that we consider these rules to be both morally and epistemically beneficial. As we saw in chapter 9, model deliberation incorporates a number of restrictions which are nonetheless compatible with freedom of speech. These restrictions can be defended in terms of fairness or individual rights, but they also bring epistemic advantages, in that they prevent the discussion from being infiltrated by inequalities in power. This might lead us to suggest that these institutional and moral norms of ideal deliberation ought to be applied to all instances of public communication. This position, which Estlund refers to as the wide mirroring doctrine, holds that all communication in the public sphere ought to aspire to the restrictions and norms of civility inherent in the ideal deliberative situation, insofar as this is possible. This would require that all participants have equal access to the deliberative forum, and that all orient their discussion towards questions of justice or the common good. Estlund demonstrates that the wide mirroring doctrine is impractical and undesirable, using examples of the poet who publishes his work independently, and thus enjoys access to publication others do not share; the philosophical anarchist, whose deliberations never address the common good; and the filmmaker, whose works influence many people by appealing to their emotions, rather than through reasoned arguments. Under the restrictions of model deliberation, all such forms of communication would be prohibited, as would a wide range of other expressions that we might think valuable. These examples go to show that in society generally, we should not wish for widespread adherence to the norms of ideal deliberation.

Next Estlund considers what he refers to as the informal political public sphere, which encompasses the political activity of nonofficials, and of officials outside of their formal capacities. This includes political speeches, candidate and citizen debates, opinion journalism, political advertising, political art and demonstrations. Estlund maintains that the norms constraining the informal political sphere are different and more restrictive than those governing the nonpolitical, while not as restrictive as those pertaining to the most formal kinds of political activity. The boundaries between the political and the nonpolitical domains are unclear, and imposing restrictions on informal political activity would risk having a ‘chilling effect’ on some aspects of the nonpolitical sphere. However, in clear cases of political expression, Estlund argues that if it were possible, it would be desirable for the ideal deliberative norms to govern these. The difficulty that arises is that real life instances of informal political activity are unlikely to come close to complying with these norms, and this leads to the problem of the second best – given that deviations from the ideal deliberative situation are bound to occur, it may be the case that it is no longer appropriate to attempt to approximate that model. There is no reason to believe that we are obliged to conform to the civility norms of ideal deliberation when noncompliance is widespread. The best way to respond to deviations may in fact be through the use of further deviations, and this claim forms the basis of Estlund’s rejection of the wide mirroring doctrine.

The wide mirroring doctrine posits that duties of civility require all participants to behave as they would under model conditions, and extends this to all forms of public communication. A narrower version would argue that this applies only to political communication, and the narrowest version, the one that Estlund supports, holds that this requirement only exists in formal instances of political deliberation. If we consider duties of civility to be a type of collective action duty, then, so Estlund argues, these duties are altered in circumstances of widespread failure to comply – this does not mean that it becomes no holds barred or that we have no duties at all, only that the content of the duties changes. In order to determine what our duties of civility might entail in a situation of general noncompliance, what is needed is a breakdown theory providing a principled account of our new duties. In formulating this breakdown theory, Estlund draws on the ideas of Herbert Marcuse, who, following Mill, suggests that the value of tolerant public deliberation is its truth promoting capacity – that is, a civil and tolerant exchange of views will have a tendency to encourage true, or at least objectively better quality, decisions to emerge. In circumstances where there has been a breakdown in the conditions necessary to realise this potential, Marcuse considers that wider standards of civility become justified, because more narrow standards can no longer fulfil their epistemic purpose. In order to remedy these defective background conditions, which are preventing deliberation from performing its truth promoting role, constrained transgressions become permissible.

The claim being made here is that when such background deviations from ideal deliberative norms occur, it is effectively a case of “power’s interference with reason” (p. 193), and in such situations, it might be that it is not enough to speak truth to power; we need to fight power with power, to try to remedy the power imbalance. An epistemic distortion has arisen, and countervailing power might be necessary to try to alleviate the effects of this, perhaps in the form of intentionally suppressing a viewpoint that is currently overrepresented, or introducing an underrepresented message via sharp transgressive forms of action. Endorsing Marcuse’s approach suggests that if we are unable to approximate real social conditions in which power does not interfere with the exercise of reason, then it might be required to use some type of remedial strategy “which would support our ability to infer from the imperfect real discourse to conclusions about what would have been accepted if it had been ideal” (p. 194). In this sense, the ideal deliberative situation, where reason is unpolluted by the distorting effects of power, still plays a central critical role, despite recognition that real life discourses cannot be expected to meet these standards. Thus, we have epistemic reasons to believe that those who hold viewpoints which are being suppressed in real life political communication may be justified in exercising a greater degree of power in promoting their messages, in order to countervail the initial distortion. The justification for these wider standards of civility are entirely remedial, and can only be extended to those whose points of view are currently being denied a fair hearing.

To support this argument Estlund examines a practical case in which we might consider that departure from the equality norm of ideal deliberation is preferable to conformity with it. In this instance it is not that countervailing distortions are necessary to remedy already existing inequalities, but it is still a problem of second best. Given that in real life it is impossible for all to have unlimited opportunities for input, it might be better to allow some inequalities if these increase the amount of input overall, rather than insisting on equality and thereby levelling downward. If everyone’s opportunities for input are equalised, but this can only be realised at too low a level to be effective, this may have epistemic losses which outweigh any gains. One possible institutional implication of this would be to implement a voucher scheme for purchasing additional political input, with the difference in cost between the cash value and the purchase price of the voucher being used to subsidise the price of vouchers, making them more affordable. Some citizens will be able to afford to purchase more vouchers than others, and will therefore be able to enjoy an unequal amount of political input. However this inequality is not problematic, Estlund argues, if it increases the amount of political input available overall, and with this increase results in epistemic gains in a way that a strictly egalitarian approach cannot.

Estlund suggests that there are at least two reasons why it preferable to have an ‘unruly’ informal political sphere that is not constrained by the ideal norms of equal time, access and power. First, the value of many of the ideas expressed in this domain will lie outside the realm of the political, and thus it is of less significance if they arose in conditions of equality. Secondly, political deliberation will benefit from taking place against a background that is rich, varied and spontaneous, and can draw on a wider range of original ideas and confrontations than we could expect to arise under the formal conditions of ideal deliberation. However, there is still a vital role to be played by the notion of the ideal deliberative situation, in that it can act as a template against which real life instances of democratic communication can be matched, and deviations identified. When deviations are identified, intervention might be required, not necessarily to try to bring practice closer in line with the ideal, but instead to “bring forces to bear, rational or not, that countervail the effect of the initial deviations so far as possible” (p. 200). The result will be a dynamic and unruly informal political sphere, shaped by deviations from the ideal and countervailing responses to these, which will strengthen our conviction that the outcomes reached could have been agreed upon under ideal deliberative conditions. This approach, Estlund argues, has the benefit of putting “the ideal speech situation in its place”, recognising its appropriate theoretical role while making room for sharp and transgressive forms of political activity (constrained by the boundaries of civility more widely construed). These countervailing insertions of power will not neutralise the original distortion or entail that deliberations now resemble the ideal. However, they can shift the balance of power and enable us to achieve the same outcome as we might have expected to arise in the ideal speech situation.

Estlund’s final task in the chapter is to demonstrate why formal political activity, that which takes place by officials in formal political institutions, ought to be governed by the norms of narrow civility – that is, to show that we may not fight fire with fire there, even though we may do so in the informal political sphere. This is the case because formal politics is more likely to approximate the ideal than is political communication in other settings, and because there are other contexts outside of this domain where views can be expressed without such narrow restrictions. Although much of the discussion that takes place within formal institutions such as the representative assembly will not resemble the ideal very closely, such norms ought still to be aspired to. This will not be excessively restrictive or impose unbearable epistemic costs, as formal political institutions exist alongside the informal public sphere, which is governed by wider and more permissive rules of civility. Estlund gives four reasons why formal politics should aim for the narrowly civil norms of ideal deliberation, and therefore may not respond to breakdowns with countervailing deviations. Firstly, whereas the boundaries between the informal political sphere and the domain of the non-political are uncertain, we can be more clear about what counts as internal to a formal political institution and therefore ought to be governed by norms of narrow civility. Secondly, given the relatively small portion of life encompassed by formal political activity, these restrictions will only apply to a fraction of the public communication that takes place. Thirdly, we are more easily able to monitor the behaviour of participants in formal institutions, which is related to the fourth point that misconduct in elected assembles and other official institutions is likely to be perceived by the public, and punished accordingly. For these reasons, within the confines of formal political institutions, the model deliberative norms of narrow civility ought to be adhered to.


I found myself in agreement with much of what Estlund has to say in this chapter, and find the concept of a breakdown theory for situations of noncompliance with ideal deliberation a valuable and important component of a philosophical framework for democracy. It would however be good to have David elaborate a bit more on what kinds of disruptive behaviour he thinks ought to be allowed under the wider norms of civility of the informal public sphere. It seems likely that certain types of unruly or transgressive conduct may not have the epistemic benefits we would hope for, and might be more detrimental than strict adherence to the norms of ideal deliberation would be (even under conditions of general noncompliance). For instance, is it permissible for politicians to lie in their campaign speeches and advertising, if the purpose of doing so is to counterbalance existing distortions by power? Surely this would have negative epistemic consequences, and so we might think that even in situations of widespread insincerity, we have a duty to be as truthful as possible. It would be interesting to know in a bit more detail just what forms of disruptive behaviour David would wish to sanction and what he would prohibit.

The other question I have concerns the ideal deliberative situation itself, and whether even under these hypothetically ideal conditions, there might be some distortions caused by inequalities or imbalances of power that the template of ideal deliberation fails to notice. It might be the case that even if the hypothetically perfect conditions of equality and fairness laid down by the model of ideal deliberation could be realised, some inequalities between participants would persist which could skew the decision reached, and this might give us reason to doubt that the eventual outcome is the one that best tracks justice/the common good. The reason I argue this is because even if all participants are given equal access to the deliberative forum, equivalent chances to speak, as much time as they wish to speak, equal attention and recognition, etcetera, there will still be differences between individuals which mean that some are better equipped to make use of these resources than others, and will therefore stand a greater chance of persuading others. The skills, expertise and self-confidence required to engage effectively in rational argumentation are not possessed in equal amounts by all people. The capacity to participate in reasoned argument is not innate to all individuals, a talent all are equally blessed with, but comes with experience, practice and education – things some individuals are fortunate enough to have greater access to than others. These people will be more adept at expressing themselves according to these norms, and may therefore be advantaged in the deliberation. Their points of view may be accorded more credibility or their messages taken to be more persuasive than those of others, simply because they are better able to use the tools of rational argument to articulate them. Here I am thinking of the problems raised by people such as Lynn Sanders and Iris Marion Young in relation to deliberation – that even under seemingly ideal conditions, natural and social differences which exist between people will bias the process in favour of privileged individuals and distort the eventual outcome.

If this is true, and even under ideal conditions of equality and fairness we can expect distortions to arise through inequalities of deliberative capacity, then might it not be the case that we would have to make room for the kinds of unruly and transgressive behaviour that Estlund wants to allow under nonideal conditions? We could argue that the kinds of conduct Estlund advocates only as remedial strategies, such as protest and emotional appeals, ought to be incorporated into our model of ideal deliberation, rather than being merely a last resort when ideal deliberation breaks down. This is a particularly interesting question to me as these are the sorts of issues I look at in my own research – whether there might be some deliberative inequalities that ideal deliberative models fail to notice, and how we might remedy these. I suppose that one obvious line of response to this would be to say that in the situation where some participants are comparatively lacking in the skills, expertise and experience to engage in deliberation as equals, then this is itself a deviation from ideal deliberation, and we have moved from the hypothetical ideal to the real-life nonideal case. However, it appears to me that there is nothing in the ideal deliberative situation as Estlund articulates it in chapter 9 (or indeed in most theoretical accounts of model deliberation) which stipulates that in addition to having equal access and chances to speak, participants must all have equal deliberative capacities. It seems possible that even in the ideal case, we might have epistemic reasons to allow for a wider range of expressions than the norms of narrow civility suggest (at least in the informal political sphere – I would agree that narrow norms of civility ought to apply within formal institutions).

Be Sociable, Share!
This entry was posted in Posts, Reading Group. Bookmark the permalink.

6 Responses to Estlund Reading Group Chapter 10

  1. Hi Rebecca – thanks for the great post.

    Like you, I would be very interested to hear more from David regarding what kinds of departures from the ideal he thinks might be justified by the breakdown model. David doesn’t actually offer any specific examples of what behaviour might be permissible under the breakdown theory (the voucher system is not a response to non-compliance and so raises different issues). One suggestion David does make is that when power interferes with reason ‘the remedy might sometimes be the conscientious suppression of an overrepresented message’ (p. 194). In particular, I was wondering if the breakdown theory would permit people to speak insincerely about political issues?

    Suppose the government is deliberating over a new health care plan, and there are two main choices, X and Y. Let’s say X already has the support of powerful interest groups such as the health insurance industry, and so the arguments in favour of X are overrepresented in the informal political arena whereas arguments in favour of Y are underrepresented. A writer at a prominent political magazine (something like The New Republic) happens to believe X is the best proposal, and that its main competitor, Y, has serious defects which render it unsuitable. Nevertheless, the writer is aware of David’s breakdown theory, and so is committed to levelling the playing field with regard to the debate. The writer was going to submit to the magazine a very well-researched analysis piece which carefully lays out the merits of X and Y, and strongly suggests that Y suffers from fatal difficulties. But, in the interests of levelling the playing field, they decide not to offer that analysis piece, and instead they publish a long essay extolling the virtues of Y, even though they don’t believe the essay’s conclusions are sound.

    If the writer’s action is permissible under the breakdown theory, this raises a couple of worries. One is that it suppose citizens’ duties of civility (such as the duty to be sincere) are justified instrumentally, in terms of their ability to deliver the epistemic goods. I’m not sure this accords with my sense of what justifies the various duties of civility I might be under (e.g. sincerity, reasonableness, willingness to listen to others, etc…). Though these duties might have epistemic benefits under certain conditions, I’m not sure those benefits are the primary grounds for the duties. Second, if everyone knows that in circumstances of non-compliance the duty to be sincere no longer applies, then surely this might have very serious epistemic costs? Once we can’t believe what anybody is saying, how will it be possible to have any kind of fruitful dialogue in the informal political arena? Micah Schwartzman raises some of these issues in a terrific paper, ‘The Sincerity of Public Reason’.

    The broader point I want to make here is one that is structurally similar to the very familiar critique of act-utilitarianism. If everyone knows that everyone else is going to behave like an act-utilitarian, then act-utilitarianism might be self-defeating when construed as a theory to regulate individual behaviour. We will fail to maximize aggregate utility if we all behave like act-utilitarians. I’m worried that a similar problem might afflict David’s breakdown theory, and the sincerity example was only meant as an illustration of the problem. If everyone knows that everyone else is directly aiming to achieve the epistemically best result, we may fail to achieve the epistemically best result. If we all know that each individual is taking epistemic matters into his or her own hands by suppressing certain ideas, overrepresenting other ideas, and generally trying to correct for perceived distortions in power using whatever means are at their disposal, we might (rationally) lose faith in the results of our political dialogue. Truth, like utility, might be one of those things that is best attained by having individuals follow second-order rules (e.g. always speak sincerely) even when non-ideal conditions obtain and departures from those rules look necessary from an individual point of view.

    But maybe I’ve not got the right sense of what kinds of behaviour would be permitted by David’s breakdown theory.

  2. Ben Saunders says:

    Thanks from me too. Great summary.

    I agree that pretty much all David says here sounds very sensible although, thinking about it, I’m not so sure why we can’t ‘fight fire with fire’ even in the formal sphere. I agree that more restrictive rules are appropriate here, for the reasons David gives. But then, it’s a separate question what to do when those rules are broken.

    David claims it’s not that rules lapse in the face of non-compliance, they simply have different content, but I’m not sure how they differ – is it only in how to respond to the non-compliance of others? For instance, suppose we have a norm against lying. Would that norm, in the formal sphere, effectively be ‘don’t lie and, if others lie, don’t lie’ but in the informal sphere ‘don’t lie, but if others lie you can lie to redress the balance’? Or would the first part of the norm alter too?

    Finally, Jonathan’s question interested me, because his example of the magazine writer supposes someone deliberately setting out to redress the balance, even though it means promoting a cause contrary to their own interests or what they think right. I’d assumed that it was the silenced minorities that were justified in breaking deliberative norms (like suffragettes or the civil rights movement).

    Does David think that third parties can act, in the way Jonathan suggests, to correct distortions? And, if so, would he say they are not only permitted to but have a duty to do so?

  3. I’ve now posted a reponse to the comments up to this point.

  4. The move to wide civility responds to a problem of the second best. Given that some conditions of the ideal speech situation are unsatisfied, it can be a mistake to adhere strictly to the others. One condition that seems especially hazardous to uphold is the stricture against self-interested voting. Political scientists register surprise that majoritarian politics in the U.S. tolerates such huge disparities of wealth: their “median voter theorem” predicts otherwise. In other words, if Americans voted self-interestedly, there should be less inequality of wealth and, to that extent, a closer approach to distributive justice and to the conditions of ideal speech. Wide civility needn’t reject self-interested voting the way it should lying. Lying has no general tendency to correct “power’s interference with reason.” For that matter, neither does “sharp and disruptive political activity.” But self-interested voting (bitter or otherwise) might.

  5. Ben Saunders says:

    People might be voting self-interestedly but irrationally. It’s commonly said that the majority of people think they’re above average, though of course this isn’t true (and can’t be if average is median rather than mean).

    My conjecture is that when people hear about something like inheritance tax hitting the top 15% say, a lot more than 15% of voters think it may affect them. This is aggravated by two further factors: i) The ‘American Dream’ – people aspire to upward mobility so, even if aware that they’re not in the top 15% now, a lot of people think they – or their children – might one day be so. ii) Higher rates of participation (voting and in other forms) among the rich, which leads politicians to favour their interests (See Lijphart APSR, 1997, who uses this as an argument for compulsory turnout).

    I think these considerations go some way to explaining why we don’t see more redistribution, though voters vote self-interestly (or, at least, try to).

  6. I think the empirical evidence weighs against the suggestion that voters tend to vote self-interestedly. (There’s a big literature, but one place I’ve seen some of it cited recently is in Caplan’s recent book, The Myth of the Rational Voter.) They tend to vote for what they take to be good for the nation, or in the common interest. Of course, there are exceptions. My account of civility doesn’t address the second-best question in the case of voting, but Bill is right there is a question there. If self-interested voting were plausibly thought to be disproportionately pushing in one direction on the issues then, if there were no way to restore it to common-interest voting, there would be a case for countervailing self-interested voting in order to partly remedy the epistemic distortion. I wonder if there’s any reason to think that certain identifiable points of view are more likely to vote self-interestedly in an irremediable way. Nothing of that sort comes to mind.

    Bill’s point is a little different, I think. He seems to think that people don’t generally vote self-interestedly (I think he’s right) but thinks that they should. That asks us to compare general self-interested voting to general “sociotropic” voting and ask which is epistemically better. Some people think that self-interested voting is likely to do better because people know their own interests better than those of others. But, whether that’s true or not, the idea that my interests only get addressed if there are enough others with my same interests is patently inhumane. I discuss related issues in this week’s chapter on the democracy/contractualism analogy. If people voted only in their own interests the small percentage of people whose well-being is normally at stake in the threat of a famine would not have a voice. If, instead, others noticed the danger to those few and registered, in their vote, the obvious barbarity of letting them starve, they would be less likely to starve (as Sen famously argues with empirical evidence).

    Bill’s right, though, that if the rich are few, and the poor are many, self-interested voting would tend to harm the position of the rich, potentially increasing distributive justice. But that’s just the flip side of the coin I’ve been discussing. If the very poor are few and the less poor are many then self-interested voting would leave the very poor out in the cold.

    This doesn’t settle the question of the epistemic comparison between general self-interested voting and general sociotropic voting. But it looks to me far from obvious that self-interested voting would do better. In the case of ideologically concentrated self-interested voting their might be a case for retaliatory self-interested voting (or, even better, voting in the common interest of all except the class that is voting self-interestedly). However, it’s not clear that self-interested voting takes that particular form.

Leave a Reply