Let me begin by thanking Danny for commenting on my work. Danny and I overlapped in graduate school for three years, and he’s offered perceptive comments on my work ever since. So I’m especially grateful to him for his lengthy engagement with my work.
Danny has two big concerns about the second chapter of my book: (i) that we might ground restraint in the value of publicity and (ii) that my objection to divisiveness-based arguments for restraint is unsuccessful. Together, (i) and (ii) may vindicate restraint despite the objections in the chapter.
A quick note before I begin my replies. Even if Danny’s objections are successful and restraining divisiveness and promoting publicity both support restraint, restraint is not thereby vindicated (and Danny does not say as much). Just to be clear: to vindicate restraint, we must show that both objections override the integrity and fairness objections. We do not yet have an argument for that position.
All public reason liberals value publicity, the social state where the rationales for state coercion, power and laws are publicly known and recognized. That includes me. But unlike Danny, I do not think publicity-based arguments provide an independent argument for restraint but rather an argument that we should adopt an accessibility or shareability requirement on justificatory reasons based on a generic commitment to public justification. I spend time in Chapter 4 rebutting the case for these requirements, and there I discuss whether a publicity grounding of a shared reason requirement is successful (p. 110, p. 123). But we’re not to chapter 4 just yet.
But that aside, Danny’s publicity-based case for restraint goes as follows:
If we take the Public Justification Principle to demand publicity as well as sufficient reasons held by all citizens, then it will seem relatively straightforward that the reasons underlying political arrangements, and used in their construction and interpretation, will need to be accessible/shareable.
Danny’s argument assumes that publicity could support restraint by supporting shareability and accessibility requirements (though you can adopt restraint without the requirements and adopt the requirements without restraint). But I can’t see how this argument works. The question of whether publicity supports restraint is derivative from which sort of reasons we want to make public. If we think the reasons that justify coercion must be shared, then we will want shared reasons to be public. But if we think the reasons that justify coercion need merely be intelligible, then we will want intelligible reasons to be public. So a publicity-based grounding of restraint must have already settled on a criterion of justificatory reasons. Accordingly, it is hard to see how we can get from publicity to a conception of justificatory reasons, and then from there to restraint.
I think Danny is incorrect because he’s not making a distinction that almost no one else makes either. There is a difference between whether a reason is publicly recognized as a reason and whether that reason is publicly recognized as a reason for everyone. These are not the same thing. We could see value in making arguments public that we reject but that others accept. So supporting publicity is not the same thing as supporting shared or accessible reasons requirements. As a result, I can’t see a clear move from a concern with publicity to a conception of justificatory reasons.
Now, Danny could claim that while operating under a shared reasons requirement, we might find publicity easier to realize, given that it may be less difficult to publicize arguments everyone already accepts. But even so, we already need to know what sorts of reasons we want to make public.
In this case, it is difficult to see how publicity considerations could support restraint, much less override the integrity and fairness objections.
In my view, the only way publicity supports restraint is by supporting a shared reasons requirement, and then adding an argument for the third premise of the Master Argument. But we’ll see in Chapter 4 that this doesn’t work.
One of the more confusing parts of the public reason literature are the hand-wavy attempts to vindicate restraint on the grounds that keeping private reasons private reduces social divisiveness. What is not clear in the literature is why divisiveness is a problem. Typically liberal theorists prize rigorous disagreement and think divisiveness is sometimes a political necessity for realizing justice. So what sort of divisiveness is problematic? And why is it problematic?
The argument seems to be that the use of religious reasons, by being divisive, undermines social stability in some sense. I distinguish between an empirical conception of stability and a normative conception of stability in order to clarify which form of stability is at issue. The divisiveness argument supports restraint either because restraint produces real-world political stability or because it supports stability for the right reasons. I argue that both arguments for restraint fail, and I call my objection to these arguments the divisiveness objection, as it responds to divisiveness arguments (I think another name for the objection might have been clearer, like “the objection to divisiveness arguments” but that was too unwieldy).
I interpret the argument from divisiveness for restraint as a defense of the third premise of the Master Argument (connecting exclusion to restraint). That is, I understand it as an argument that restraint will help political order be based on genuinely normative considerations, like good public reasons. If restraint helps exclude dangerous, destabilizing arguments and reasons from entering into the political process, then presumably it will tamp down on both empirical and normative instabilities.
Danny notes, fairly, that it seems like divisiveness is an independent argument for restraint. And I admit that treating the argument from divisiveness as part of the Master Argument might feel like a stretch. But we need to explain why public reason liberals are concerned with divisiveness and if divisiveness has no negative impact on the quality of public justification and does not undermine publicly justified law, it is not clear why public reason liberals should care about it. Perhaps instability of some sort is just plain bad, but what would explain what it is bad on public reason liberalism is that it would undermine a publicly justified order which, in virtue of being publicly justified, will be stable both empirically and normatively.
Danny next turns to a rather hard-hitting criticism of my use of the divisiveness objection. I claim that there is no empirical evidence that a failure to comply with principles of restraint in the literature (or principles similar to them) will help prevent religious conflict and violence. I suggest instead that the religious violence that is sometimes blamed on a lack of restraint is caused by other factors, such as attempts by religious groups to co-opt state power to use for their own ends.
Danny thinks my claim is “astonishing” and severely underdocumented (though the documentation I reference is an extensive attempt to address the empirical claim). Danny thinks the historical record,
at least seems quite clear in demonstrating that: a) societies that have been embroiled in sectarian conflict have overcome those conflicts after embracing certain norms of restraint; b) societies have intentionally embraced norms of restraint in response to perceived conflicts between sects, and lower levels of divisiveness have followed; c) members of societies that have embraced norms of restraint appeal to the importance of avoiding sectarian conflict to defend adherence to those norms; and d) there are other areas of life where we know that taking divisive issues off of the table for negotiation can enable processes of deliberation and collective decision-making to go forward.
I think what Danny misses is that I’m critiquing divisiveness arguments for principles of restraint in the literature, not any sort of deliberative restraint at all in any context. It may well be that discussing religion at certain times and in certain contexts rather than others has helped to reduce conflict and violence, but the even the empirical evidence even for this claim is feeble, despite the presence of a weak correlation. What we lack any evidence for is that the principles of restraint defended by public reason liberals have played a role in reducing divisiveness on net.
This may come as a surprise to readers, but I agree that much of what has been accomplished by way of increased stability is at least connected to repressing certain kinds of religious talk. But the form of restraint that has historically done so are norms that prevent people from talking about religious issues with just about anybody ever. We have strong social norms against talking about religion with people we don’t know. I think these norms may have benefits, but they have also been destructive in various ways (a la Mill, people don’t know what they believe or why in many cases). But that’s not the sort of restraint public reason liberals are focused on.
III. Addendum on Assurance Mechanisms
Danny worries that my and John Thrasher’s game theoretic analysis of restraint’s role in generating stability for the right reasons does not show that restraint is unnecessary to promote assurance. All we show, Danny claims, is that restraint is not sufficient to generate stability. But here I think Danny misses the structure of the dialectic. Justifying restraint requires showing that it has a concrete benefit. Otherwise, it is pointlessly illiberal to restrict public discourse. If we can show that restraint does not promote stability for the right reasons or even clearly contribute to it, then this would show that one reason to support restraint (the promotion of stability for the right reasons) fails. Thrasher and I argue that the use of public reasons does not reliably contribute to assurance due to the phenomena of cheap talk and informational drift. Since it does not contribute, then the stability-for-the-right-reasons basis for restraint fails.