Again, I really appreciate all the excellent and challenging discussion so far. I am glad to have the chance to reply. Sorry this is rather lengthy, and still doesn’t address nearly all the points.
I assert and employ what I call a “qualified acceptability requirement” in a schematic form, leaving largely open which views count as qualified (although I fill in bits of it at crucial points). I don’t go deeper into moral philosophy in order to provide a moral argument for it. Many people object to the whole approach, which is very similar to that of Rawls’s political liberalism. What I do by way of defense is to defend it against several lines of objection. Several commentators have wished I gave a “direct” argument for it. To say, however, that the theory “needs” to do so seems to require an awful lot. Is a theory not entitled to any beginning premises? If there are objections to the premises, that’s one thing. I defend against what I take to be the more important ones, and no doubt there are others that would need to be answered. But apart from the evaluation of specific objections, the theory doesn’t seem much damaged by noting that the premises are not themselves supported by deeper argument. What theory could be immune from that objection?
Several comments worry that I am committing the very same expert/boss fallacy that I identify and criticize elsewhere. The move of mine that is suspect is my holding that a view is not qualified unless it has the right view of which views are qualified—a kind of “expertise.” As I say in comment 2, QAR empowers no one, so there’s no real issue of bosses. Still, counting some points of view as disqualified because they are mistaken about certain things might seem similar. And so it is a good question whether it is also defeated in a similar way.
Despite a certain parallel structure, the two are on very different moral footings. The expert/boss fallacy (as I call it) seeks to empower some over others, and it is thus subject to the QAR. The QAR, since it is a doctrine used in justifications that would empower some over others, is also subject to the QAR (subject to itself). However, since it does not seek to go from expertise to authority (but only to “qualification”) it does not fail in the way the expert/boss fallacy does.
The suggestion against me seems to be that, if I am consistent, points of view should not be counted as disqualified on the basis of having false views. On that alternative approach an objection to a political justification stemming from the view that women are animals and not fully human must be counted as qualified and so as sufficient to defeat the proposed justification. Rather than embracing that conclusion (I take it) my critics argue that my own approach cannot coherently avoid it and so my whole approach must be flawed.
This objection grants for the sake of argument that QAR is an appropriate constraint on claims to legitimate authority, and asks whether expertise can consistently be blocked at that level but not at the level of determining which points of view count as qualified. So if the question is about consistency we have to see whether I am using a double standard across those two levels. At the first level, where we ask whether knowers have legitimate authority, the standard I use to block that move is to say that their status as knowers is not acceptable to all qualified points of view. So what happens if I were to consistently use that same test against my own proposal to count those points of view that lack certain correct views as disqualified? The test required by consistency, again, is whether the identification of the supposedly correct points of view is itself acceptable to all qualified points of view. Do all qualified points of view agree with the proposition (whatever it might be), disagreement with which I am using to disqualify points of view? I agree that this sort of consistency is required.
The particular proposition that prompted the objection was “C is the class of qualified points of view,” where C is stipulated by us to be the true boundaries of the qualified set. So the question forced on me by consistency is whether all qualified views accept the correct set of qualified points of view, including any respect in which the boundaries are determined by whether the point of view is correct or incorrect on certain matters. The objector’s worry is that I must grant that there is qualified disagreement about even this.
As I note in the chapter, it is certainly logically possible for there to be qualified disagreement about who is qualified. And if there is such disagreement (which is not an empirical, but a moral question as I briefly explain in Comment *** on the discussion of Chapter 1) then the whole project would collapse. No class of the qualified would be available, not even the true one, since none would be acceptable to all qualified points of view.
“Qualified,” recall, is a term of art. Readers cannot infer anything about its content from what that word means in ordinary language. So if an objector wants to argue that there is, indeed, qualified disagreement on which points of view are qualified, we would want some argument based on an account of the content of “qualified.” I don’t mean that they bear some special burden of proof. But if there is no real argument for that position, then the theory should be free to go either way. And since it needs to go the other way, it should.
There is, so far as I can see, nothing implausible about holding that one of the disqualifying characteristics is having a false view of which points of view are qualified. That represents error on a moral matter of special importance. Some such errors will be disqualifying and some won’t. To take a different example of a disqualifying moral error, having erroneous views about which beings have moral standing at all would be disqualifying. If my move is blocked by an expert/boss problem, then so would that one. But I see no reason to think either of them is blocked. I claim that objections stemming from those particular erroneous views should not be thought to have a certain kind of moral significance: defeating proposed political justifications. I simply put on the list of disqualifying errors, erroneous views about another dimension of moral status, erroneous views about which points of view are entitled to justifications they can accept—about which points of view are qualified. It seems perfectly sensible to deprive points of view that are mistaken on that moral matter a veto over proposed justifications.
So, to return to the objection, has consistency been respected? I believe I am applying the same test at both levels. The expert/boss fallacy is blocked by saying there is qualified disagreement about who is expert. The view that only points of view with the correct (“expert”) view of who is qualified are themselves qualified is put to the very same test. But it passes. There is no qualified disagreement about that.
As I say in the book, my method is not to propose a general theory of the qualified (or the “reasonable”), defend it in a general way, and then apply it to cases. Rather, it is to assert only a few things about that boundary, at crucial places in my argument. They are not themselves defended to any great extent, but that is never an interesting objection to a claim. The interesting question would be whether there is some reason to think I am not entitled to them. If you prefer, you can see me not as asserting those things, but as saying that if they were genuine implications of the correct view of the qualified, look at what we could do with democratic theory. Of course it is logically possible that they are not correct, but what reason is there to think so?
Turning to another matter: Bill Edmundson writes, “Let me suggest one POV [point of view] that an epistemic-proceduralism-friendly QAR would –wrongly– have to exclude. Let this be POVx, which holds that “Any true POV, whatever it may be, meets the QAR.” Let’s call this the view that all true views are qualified. He argues, then, that the qualified acceptability requirement must deny that all true points of view are qualified, at least if it is to be of any use to epistemic proceduralism. I take it that his argument for this claim is in the following passage:
“Let me call POVs that aim at truth and employ generally accepted modes of reasoning “expertise-claiming.” Of course, it will be controversial which is the true POV. Therefore, to accept POVx will mean either a) to allow as qualified all expertise-claiming objections, or b) to disqualify all expertise-claiming objections that do not command unanimity within the circle of qualified POVs. Rawls seems to take route b) in Political Liberalism; but I take Estlund to tend rather to option a). The problem with responding to POVx by taking route a) of course is that not all POVs now qualified will rank democracy above truth. Moreover, it is far from certain that democracy rates higher epistemically that all other expertise-claiming POVs (this in fact is the central problem, as Estlund has explained). The QAR will not do the winnowing it needs to do unless it responds to POVx along line b), which in turn draws the objections to Rawls that Micah refers to.”
I find this difficult to understand, so I may have it all wrong, but let me try. Qualified points of view disagree about which is the true point of view, of course. So, if one accepted that true points of view must be qualified, it will be controversial which ones get so qualified. What I don’t see is how this leads to a problem or a dilemma. Bill says that because it’s controversial which views get qualified by being true, holding the view that true views are qualified requires one to choose between two choices. a) Count as qualified some different and less controversial class of views (for example, all those that claim to be true). (Here Bill actually switches from “points of view” to “objections,” but I’ll take him to mean objections based on expertise-seeking points of view.) He says that this counts too many views as qualified for epistemic proceduralism’s purposes, which I’m sure it does. In particular, he worries that it qualifies points of view that deny democracy’s epistemic superiority. I will return to this specific point about epistemic superiority at the end of my comments. Even apart from that, I agree that granting that all “expertise-claiming” views as qualified would be fatal. That would qualify lots of views that reject all kinds of things I regard as crucial to qualification, including basic moral equality of human beings, etc. But I would not count that larger class as qualified.
The only other option, Bill says is b.) to count as disqualified any point of view (or objection stemming from such a point of view) that meets with reasonable objection. I’m puzzled, because he says this seems to be the route Rawls takes. But Rawls does not say that a point of view or an objection stemming from a point of view is disqualified just because there is reasonable disagreement about it. That would mean disqualifying a theistically based objection to an atheistic political justification, on the ground that there is reasonable disagreement about theism. But Rawls counts the objection as reasonable…because there is reasonable disagreement about theism. I’m sure Bill agrees with this reading of Rawls, so I’m pretty sure I haven’t gotten his argument right. I’d welcome any clarification.
In general though, suppose we had to grant that true points of view are qualified. I just don’t see how this upsets the apple cart in any way. If a true view is not being treated as qualified by the other views then they are mistaken in a morally important way. Of course, it will not be available to public reason to count any view as qualified because it’s true, since that is open to qualified controversy. So, instead, each qualified point of view brackets the question of which is true, and counts certain ones as qualified on other grounds (often including commitments to certain truths, as I’ve suggested above). Now if this way of identifying qualified views happens to leave out a true point of view, then they have misidentified the class of qualified views, a morally important mistake. (Rawls at some point conjectures that the true view seems pretty likely to meet the criteria for “reasonable,” but there is no guarantee.)
A final point about this. On my own view, defended in this chapter, we must take the qualified acceptability requirement itself to be true. If the true view is qualified (Bill’s principle), then the true qualified acceptance criterion will include it (not qua true, but on other grounds), or else it wouldn’t be the true criterion. So the danger that the true view might not be counted as qualified is avoided.
On the question of epistemic superiority and qualified disagreement: Several people raise good questions about exactly what is being asserted about the epistemic value of democracy. And this is relevant to Bill’s objection. My response to the comments on Chapter 2 is relevant here. Recall, I want to argue that arrangements that propose greater extents of authority have a presumption against them, and must rebut the presumption subject to the qualified acceptability requirement. So let me bring that point to bear here.
Suppose a certain epistocracy is actually epistemically better than democracy (a possibility I readily grant). There will be qualified objections to this claim, as everyone here seems to grant. But here’s an objection to my view (close to Bill’s objection, I hope): consider the claim that democracy is epistemically better. If the true point of view includes the view that epistocracy is better, then mustn’t we grant that this provides a qualified objection to the claim that democracy is better?
I argue in my remarks last week that if there is reasonable objection in both directions democracy wins, because as a proposal of a lesser magnitude of authority it enjoys a presumption which is not overcome. So I don’t deny that the true view might include that superiority of epistocracy, nor do I deny that this produces a qualified objection to claims that democracy is superior to epistocracy. Rather, I claim that there are qualified objections in both directions, leading to a potential stand-off, which is resolved by the presumption against authority. So democracy does not win on the grounds that it is superior to epistocracy. It wins that contest by default, in the face of qualified dispute in both directions. What it needs to beat on epistemic grounds would be any arrangement that does not propose a greater magnitude of authority. One example would be random choice of policies. I need it to be beyond qualified disagreement that some democratic arrangement is epistemically better than that.
The question of markets vs. democracy is an interesting one in this context, but I have to be brief. I regard a democratically authorized market economy as a feature of a democratic arrangement, so that’s not the relevant competitor. Consider some sort of anarcho-capitalist arrangement, where the market is not authorized or controlled by the state, but is held to be the final arbiter of authority. This does not obviously propose asymmetrical authority relations in the way epistocracy does. And yet it is not, I think, democratic (which is not yet any argument against it). So, at the moment I believe that epistemic proceduralism can only win its case for democracy by arguing, in a way that is beyond qualified disagreement, that it is epistemically better than deep anarcho-capitalism. I mention this here only to suggest that a random procedure might not be the only one that is on all-fours with democracy with respect to the magnitude of the authority that some are proposed to have over others. I don’t have a clear enough idea of what the anarcho-capitalist arrangement involves to know whether there really is a well-defined alternative there. But if there is, democracy’s case depends on the epistemic comparison, within public reason, with anarcho-capitalism. This is mainly illustrative, and not something that strikes me as a serious threat, but I’m not arguing the matter here.