I think I understand Jon’s objection as you understand it, so we’re starting from the same point. You go on to say that you’re not sure that “the deficiency in ultimate justification that Arthur believes characterises Benedict’s comprehensive argument for L is a good argument for thinking that consensus justification is necessary”. I take this to be an elliptical summary of Jon’s argument, so just for clarity I set that argument out more fully as follows:

1. Arthur believes there to be a deficiency in Benedict’s ultimate, comprehensive doctrine-based justification for endorsing L—which justification would, on a convergence model, serve as part of a public justification of L (along with a bunch of other justifications based in other comprehensive doctrines).

2. Therefore Arthur can’t sincerely believe that L is justified to each person.

3. Every (reasonable) person has to believe sincerely that L is justified to each person if she is to endorse L herself (that is: it doesn’t suffice that her own comprehensive doctrine alone should give her grounds to endorse L).

4. Therefore Arthur can’t endorse L.

5. Since Arthur can’t endorse L, L isn’t justifiable to each person (to be justifiable to each person, presumably, a law must be such that each person can endorse it).

6. Therefore the convergence model of public justification fails to show that L is justifiable to each person; whereas a consensus model focused on public reasons, which will not fall foul of the sincerity requirement, may succeed.

Your first reason for thinking that even on a convergence model, Arthur can sincerely believe that Benedict has no grounds to object to L is that L is at least compatible with what Benedict reasonably believes—even though Arthur takes what Benedict reasonably believes here to be unjustified. That Arthur takes what Benedict believes here to be unjustified, you say, “does not imply that L is not reasonably acceptable to Benedict”. I agree with you about that. But I don’t think the question is about whether L is reasonably acceptable to Benedict. It’s about whether Arthur can see L as justifiable to each person. The contention is that he can’t if he takes of Benedict’s grounds for endorsement of L to be unsound in some sense.

You then offer (in the same paragraph) what I think is a different reason to suppose that convergence doesn’t threaten justifiability to each person, viz., that the sincerity requirement itself is too strong. You say:

Why should Arthur care whether Benedict does have a cogent ultimate justification for L though? One could think that what should matter to Arthur is whether Benedict has standing to object to L. If L is reasonably acceptable to him, arguably he doesn’t have grounds to object. Any further failure of ultimate justification is Benedict’s problem, not Arthur’s. Arthur doesn’t have a problem of sincerity with his view there because he doesn’t have to have a view there.

Funnily enough, this is the sort of thing that I’d be inclined to say with regard to ultimate justification for the fundamental political values and ideas. But I would be more reluctant to say it about public justification for laws and policies, because such justification needs shared frameworks for discussion and debate. It needs such shared frameworks because the public sphere is meant to be a place where society forges its path as a shared project, rather than a kind of huge ongoing compromise, and because of its role in shaping future generations. As I understand your picture of the convergence model, it doesn’t offer any such shared framework: rather, proposals are submitted and then vetoed or not vetoed, and that’s more or less the end of it (although presumably it’s permitted to point out when someone’s doctrine implies acceptance of some proposal although that person herself fails to see this).

It seems to me that roughly this objection to the convergence model is precisely what you offer at the end of your ‘more political liberal in spirit’ argument (from “Our real problem” onwards). That confuses me, because as I understand the first part of that argument (from “Look…” to “…do we not want to go there”) it’s entirely opposed to the picture that Jon and I are offering.

As for your official second reason for rejecting the anti-convergence argument—i.e. the worry about disdain—I’m not sure what to say. It’s just an aspect of the fact of reasonable pluralism that people who adhere to one reasonable comprehensive doctrine will take those who adhere to another to be mistaken, even if they’re reasonable. (That’s how an atheist, I take it, feels about a theist.) But if both sets of people are reasonable, that shouldn’t imply any disdain: the burdens of judgment explain how people can reasonably differ even though only one of them can, in many cases, have got the matter right. The problem with convergence justification is not that it involves public justification in terms of doctrines that are disdained by others, but that involves public justification in terms of doctrines that people who subscribe to competing doctrines just can’t see as capable of doing the justificatory work that’s required. To see them as so capable would be, in effect, to abandon the competing doctrines.

How does that sound?

As soon as we add much content to the story, I think we are going to end up specifying a regime of rights, e.g. “B is not the one using coercion because she’s the one who owns the book.”* That’s fine, it’s just that I don’t see why that particular regime of rights is not subject to the same demands of stringent public justification as any other regime. It shouldn’t get to win just because every other conception fails a public justification test. Hence I don’t think it’s odd to think of non-agreement as the default. Saying this does not imply that anyone has natural authority over anyone else at all.

*i.e. in short, what Jon said.

PL
1. A has a deeply held moral conviction that Xing is wrong.
2. People should not be forced to do things that contradict their deeply held moral convictions. Therefore:
3. A should not be forced into Xing.

Both of these premises can be shared by everyone. One states a fact about the world, in particular a fact about someone’s beliefs, and the other states a moral conviction that, arguably at least, could be part and parcel of a political liberal conception of liberty of conscience.

This PL argument is quite different from an argument that A might present:
A
1. Xing is wrong.
2. People should not be forced into doing things that are wrong. Therefore:
3. I should not be forced into Xing.

The reason why we should not cut out the middle man and go straight to the religious reason is because premise 1 may very well be false. If so, A’s a bad argument. (Not bad as in believed to be bad by the majority, rather just plain bad). But PL can still be a good argument.

1. In the inoculation case, it’s true that the exemption imposes risks on third parties. But presumably, the religious believer would say: “I didn’t create the risk. It’s a natural risk of disease, and I’m simply unwilling to contribute to rescuing other people from it. I’m unwilling because it violates my deepest convictions to undertake medical care of this kind.” So I suppose I have two questions: (1) Does refusing to prevent risk to others count as coercing them? (2) If a public justification is victorious only when it defeats all other considerations, how can it do that here, when presumably nothing could defeat the religious objection, at least in the mind of the believer?

If the response to the second question is that religious believers must publicly justify imposing risks on others (or coercing them), what does the justificatory liberal say to the religious believer who rejects the Liberty Principle or the Principle of Public Justification (PPJ)? Although I’m not sure about this, stating that question leads me to think that justificatory liberalism presupposes something equivalent to an overlapping consensus on the Liberty Principle (and PPJ). There must be a consensus among all reasonable citizens on at least one shared premise, namely, that infringements on liberty must be publicly justifiable.

2. The second example is parental choice of education (which is suggested on p. 18-19 of your paper, including the reply to Macedo). I’m still not entirely sure how religious defeaters are supposed to factor into reasoning about this case. Why can’t parents raise religious objections to the state’s account of what is in their children’s best interests? Presumably, the state has a public justification for teaching children the skills they will need to participate in a democratic society, function in the market, etc. But on the asymmetry thesis, public justifications can be defeated by religious objections. If that’s incorrect, or if it’s displaced here by the demand for public justification when there is parental coercion against children, then I’m not sure where your paper was headed in the discussion of education. Maybe the thought is: we at least have to recognize the nature of the objection religious parents are making. We can’t claim they are being coerced for reasons they share, since they don’t share those reasons. They are being coerced for reasons that, on their view, are defeated. But that doesn’t seem like much consolation for them. Nor does it seem all that different from Macedo’s claim that those who would impose their religious views on others (including on their children) have to face up to the diversity of reasonable views in their society. And they have to do that by giving public justifications for their attempts to coerce others, including their children.

Let’s begin this comment with Andrew Lister and Simon May’s concern. Here’s Lister:

The fact that I have a particular religious conviction may ground an exemption from a general law, but the religious conviction itself doesn’t justify the exemption.

I am not sure why we should think this. These reasons can play a defeater role if we relax symmetry and consensus, so I do not see the problem. But nonetheless, I am worried about the way you are individuating reasons. Suppose an Amish man claims:

My religion prevents me and my community from participating in the Social Security system. We see it as an affront to our way of life. We take care of our own.

And then you say that this isn’t itself a defeater reason but instead “grounds an exemption” from a general law. So in other words, it grounds or justifies the use of another defeater reason. How is this any different from holding that the religious reason defeaters the law? On your view, it is playing a defeater role by means of a more general reason. But it is still playing a defeater role.

Further, I don’t see why we shouldn’t ‘drop the middle-man’ reason and go straight to the religious reason. What is wrong with that?

Let’s take up Lister’s second comment:

If there are reasonable religious views and reasonable non-religious views, then all laws must have both religious and non-religious rationales; the religious are constrained to support only those laws that reasonable non-religious people support, and the non-religious are constrained to support only those laws that reasonable religious people support. Reformulating p.17, couldn’t we say ‘even if a religious rationale is necessary in our society for a publicly justified law, it can be defeated by a reasonable secular conviction without any religious backing’?

A’s rationale is not defeated by B’s reasons. A’s proposal is defeated by B’s reason. And yes, a religious proposal can be defeated by secular reasons and vice versa. I hope that helps!

Finally, let’s turn to Micah Schwartzman. He wonders how Jerry and I would deal with two proposed cases of religious defeaters. I can only speak for myself. But let’s review the cases anyway:

Case #1: Suppose the state requires inoculation against some disease. A significant number of citizens seek exemption from this policy on religious grounds. If the exemptions are granted, there will be a significant increase in the risk of disease in the general population.

This is the case where a religious exemption has a negative externality, i.e. producing disease in others. I think this can rightly be seen as a case of imposing upon others. So that imposition must be justified. The difficulty is that we have two cases of coercion: the state coercing the religious, and the religious imposing significant risk of disease on everyone else.

I think the state’s first priority should be to do whatever it can to find a way to not coerce at all in this case, to find a solution whether members of this religion can isolate themselves from the rest of the public without forcing them to be treated. If this cannot be done, then I think the imposition can be justified. But the case has a lot of complexities so my view is tentative.

Case #2: The state requires that all children pass standardized tests for reading, writing, etc., through the age of 16. Some religious believers seek an exemption for girls, ages 12-16, on the ground that additional education will weaken their commitment to traditional gender roles, which are religiously mandated.

This is a case that involves children. And they’re not full agents. I don’t know what to think about children yet. Then again, I’m not sure anyone who does not base rights directly on a perfectionist account of interests has a good theory. Obviously we have to balance parental rights against some reasonable account of the child’s interests. Of course I think that forcing these girls not to be educated is wicked and can be prevented by law. I don’t think the coercion can be justified. The problem in this case is that again we have double coercion. The parents want to coerce the child, but the state wants to coerce the parents. Since I think the coercion of the child is unjustifiable, then I think the state can coerce the parents to stop. But what if the child doesn’t want to be educated? What then? I am not sure what to say in this case unless one could successfully make the case that at the right level of idealization she has an overriding reason to endorse going to school. I am not sure yet how to answer this question.

I will now reply to a few worries about the status of the presumption against coercion. Simon May argues that we might do without a presumption altogether. Instead:

In contractualist stories, we’re not left with the state of nature as a baseline, but non-agreement.

I think this is odd. Suppose you and I have a disagreement about whether you should control me. If we begin a contractualist story with the assumption that no one has natural authority over others and that all are free and equal vis-à-vis one another, then this example of simply beginning with non-agreement seems very implausible. I take it that the central liberal idea is the notion of natural equality of persons. Due to this, I take it that there is a natural asymmetry between the status of my claim over me and your claims over me. This is why Jerry and I believe there is a natural presumption against interference and coercion (coercion is just the forms of interference used in politics) because it expresses a commitment to regarding people as free and equal.

Onto Jon Quong’s concerns. Quong worries that justificatory liberals must make use of moralized concepts because until norms are justified we cannot know what counts as coercion. He writes:

Consider the following simple example. Until we know who owns that bit of property over there, we cannot know whether my statement ‘I’m going to destroy that bit of property unless you do as I say’ is a coercive threat or not. I agree that we must rely on the idea of public justification to determine who owns that bit of property, and I also agree that if the statement turns out to be a coercive threat, we must declare there is a presumption against the threatened action unless there is some further decisive public justification that can be provided.

Jerry and I have talked about this a bit, so it is cool you brought it up. In The Metaphysical Elements of Justice, Kant talks about how we move from a state of mere possession of private property to ownership and he believes this requires devising a rule of distribution that all rational persons can assent to. So we have a state of nature, but coercion only applies as a concept to bodies and holdings that one is actually holding. For instance, on Kant’s view, if I turn around and you snatch my bookbag in the state of nature, that isn’t coercion because we haven’t publicly justified property titles yet. Only once we’ve publicly justified property titles, can we specify stealing as coercion because we’ve justified property norms that count property as extensions of ourselves. But this does not mean I can’t apply the concept of coercion at all in the state of nature. Rather my range of application for the concept is restricted until property rights are justified. So I disagree that the concept of coercion has no range of application until public justification occurs. Instead its range of application is simply sharply circumscribed.

Hi everyone. I am so pleased by the discussion. I will post three replies. The first is to worries about convergence. To shorten my post, I will accept Jon Quong’s view in his most recent comment that Porter adequately captures his concerns. So I will address Porter and Lister directly and Quong by default.

First, I will address Andrew Lister’s concerns. Lister worries that on the convergence view that:

Just because some reasonable comprehensive doctrine opposes L, we can’t vote for L, on the convergence view, even though there may be a good case for L based on premises that all qualified views accept.

It depends on why a person with a reasonable comprehensive doctrine opposes L. If the person is reasonable and has overriding reason of her to reject L, then L cannot be publicly justified. So yes, if only shared reasons are allowed, then L may be justified. But remember, even on Rawls’s view, this is possible in the full justification stage. In this way, the convergence view can reduce the number of regimes and laws that can be publicly justified. But notice that it also increases the number of regimes and laws that can be justified as well, because it allows more reasons to support a regime or law. This helps to answer another one of your criticisms:

Maybe the problem with convergence, as Kevin and Jerry articulate the idea, lies in its dependence on the existing range of views. Even if all existing reasonable doctrines support L, it may not be the case that all possible reasonable doctrines support L. Particularly if we’re designing the basic structure of society, which shapes aspirations as well as setting incentives, we may not be satisfied that no reasonable citizen currently objects; we may want to insist that no reasonable citizen could object, and I think this demand goes in the direction of the consensus interpretation that Jon favours.

It’s not clear whether this is a worse problem for the consensus theorist or the convergence theorist because convergence both increases the opportunities for public justification by increasing the number of permissible reasons to propose and decreases the opportunities for public justification by increasing the number of permissible reasons to reject. Consensus views, therefore, may have as much of the problem you raise.

Incidentally, I think it violates the spirit of justificatory liberalism to require that all laws are such that no possible reasonable person could have a reason to reject them. This would make political society impossible, I think. Not only that but it would run against the grain of constructing principles that are based on our actual considered judgments.

Onto Porter’s concerns:

Michael Porter worries that:

… An overlapping consensus from comprehensive premises would not be good enough as a justification of a conception of justice to each person because it’s not the case that each person in the overlapping consensus could sincerely believe that many of those premises justified the imposition of the principles of the conception of justice on others, given her own comprehensive views.

To be honest, I still cannot see why this should be. The sincerity principle’s application in these conditions will depend on your theory of public justification and epistemic justification. And I cannot see a case for a reasonable principle of sincerity that rules out convergence because I think regarding another reasonable citizen as having her own justified support for a doctrine is fairly easy to get, particularly if you buy Jerry Gaus’s account of justification in JL. I don’t agree with it in all of its details, but as far as the appropriate conception of epistemic justification for politics, I think it is close.

You also note later that the concerns you raise may apply to the standard version of justificatory liberalism articulated by Rawls. I think you’re worried that given that convergence plays a role in the full justification/overlapping consensus stage that this will make the third stage of justification – public justification – hard to achieve.

I have a chapter on how public justification proceeds in my dissertation. I think that on the convergence view, as Jerry and I argued, reasons are to some extent hidden from public view and that we need institutions that will aggregate and transmit these reasons to our political institutions and to citizens in order to enable public justification to proceed. But my view is that we simply need good information transmitting institutions to accomplish this goal, so I spend some time in my last chapter talking about how justificatory liberals should focus on developing a good media culture to achieve this goal.

This is just a very brief note to say that Tom has provided an excellent, and extremely helpful, re-statement and clarification of my reply to Simon.

Thanks Tom!

I was a bit sloppy and should rephrase my statement of Jon’s position. It isn’t that Arthur thinks that Benedict’s comprehensive premises are false, per se, but that he thinks they are unjustified for Benedict to hold. Since Arthur thinks Benedict is not justified in believing Rb, Arthur thinks Benedict is not ultimately justified in believing anything on the basis of Rb:

In order for A’s comprehensive doctrine CDa, to generate a sound justification for A to accept law L, it must be the case that A is justified in believing CDa. If CDa were unjustifiable for A to believe (suppose CDa were ‘I should always do whatever the little birdie outside my window tells me’), then it would also be unjustifiable for A to rely on CDa in justifying his acceptance of L. The same will hold for person B with regard to his doctrine CDb. Thus, in order for the convergence model to be consistent with PJS, it must be the case that all the members of the justificatory constituency view one another’s comprehensive or nonpublic doctrines as justifiable for their adherents to believe.

Since Arthur does not think that Benedict is ultimately justified in endorsing L on the basis of Rb (or “CDb” in the above), and since there is no other relevant consideration, the claim is that Arthur cannot sincerely believe that L is appropriately justified to each person. Instead Arthur must think that there is a good justification in terms of public reasons. Once there is some such justification, then Arthur can believe that Benedict is ultimately justified in endorsing L because Arthur can sincerely believe that each person is justified in endorsing the political values of public reason. This is my understanding of Jon’s objection.

The point I was making was not one about whether Benedict has good reason to accept the political values that serve as premises in a public justification. That’s a separate issue, albeit an interesting one. The point I was making concerns whether the deficiency in ultimate justification that Arthur believes characterises Benedict’s comprehensive argument for L is a good argument for thinking that consensus justification is necessary.

I’m not sure it is, for two reasons:
a. Arthur can sincerely believe that Benedict has no grounds to object to L, because L is perfectly compatible with, or even implied by, what Benedict reasonably believes. That Arthur thinks that Benedict is not justified in believing his wacky premises, and hence is not ultimately justified in endorsing L for those wacky reasons, does not imply that L is not reasonably acceptable to Benedict. Why should Arthur care whether Benedict does have a cogent ultimate justification for L though? One could think that what should matter to Arthur is whether Benedict has standing to object to L. If L is reasonably acceptable to him, arguably he doesn’t have grounds to object. Any further failure of ultimate justification is Benedict’s problem, not Arthur’s. Arthur doesn’t have a problem of sincerity with his view there because he doesn’t have to have a view there.

b. To defend a consensus account by appeal to the view that Arthur cannot sincerely think that Benedict has an ultimate justification for endorsing L because Arthur does not think that Benedict is justified in believing his comprehensive premises may presuppose a (disdainful) comprehensive view about the extent of justification within other comprehensive moral doctrines. I’m not sure about this reason though. It could be that the argument rests not on that disdainful view but simply on the claim that a reasonable person could endorse it, which is certainly possible. Nevertheless, it seems to me more political liberal in spirit to advance an argument that just says:

“Look, we have a bunch of reasonable doctrines around our society. Within the public sphere, it’s really none of our business whether other citizens endorse their comprehensive doctrines for justifiable reasons or whether they really are just plain wacky. That’s an after hours, private sphere sort of discussion. Instead, in the public sphere, we leave that issue off the table entirely, and simply ask whether any reasonable person, assuming their reasonable comprehensive moral doctrines as a given, is in a position to object to the laws or fundamental political arrangements of our society. If these laws and arrangements are implied by their comprehensive moral doctrines, then they are not. So an overlapping consensus from comprehensive doctrines would in principle suffice for stability for the right reasons, even if it does not thereby guarantee something that it is not our present business to worry about, i.e. each person’s ultimate philosophical justification for their endorsement of those laws. Boy, do we not want to go there. Our real problem is that since reasonable pluralism is so broad and since we are talking about an enduring arrangement here, we can only sincerely believe a stable overlapping consensus is possible if there is a public justification cashed out in terms of political values that we can expect reasonable citizens to endorse. Anything else would be a matter of luck, and luck will run out sooner or later.”

I’m not claiming that this sort of view is compatible with justificatory liberalism. Perhaps JL does involve the more demanding standard, in which case there may well be a sincerity problem there.

Doesn’t it make a difference whether B thinks the belief that Ra justifies L is unreasonable, or simply false? If A thinks B is unreasonable to think Ra justifies L, I can see a legitimate concern about convergence justification. Yes, we both agree on L, but A badly misunderstands the implications of his own view, and I (B) shouldn’t exploit that unreasonableness to put in place my own view. However, if B simply thinks {Ra justifies L} mistaken, but not unreasonably so, it would seem that in refusing to support L B is engaging in a kind of paternalism, a kind of substitution of judgment that A could object to, as Simon argued, along the lines of “It’s none of your business to decide for me whether my fundamental concerns support L. I think they do, and that should be good enough for you.”

Maybe the problem with convergence, as Kevin and Jerry articulate the idea, lies in its dependence on the existing range of views. Even if all existing reasonable doctrines support L, it may not be the case that all possible reasonable doctrines support L. Particularly if we’re designing the basic structure of society, which shapes aspirations as well as setting incentives, we may not be satisfied that no reasonable citizen currently objects; we may want to insist that no reasonable citizen could object, and I think this demand goes in the direction of the consensus interpretation that Jon favours.

What Jon has been contending, if I understand him, is that convergence models of public justification fall foul of a sincerity requirement (as expressed, for example, in the principle of justificatory sincerity). The requirement that public justification be articulated in terms of political values is not, then, grounded in the following supposition: that an overlapping consensus from comprehensive premisses would not be good enough as a justification of a conception of justice to each person because the falsity of some of those premisses would preclude their having justificatory power (this is the claim you attribute to Jon). It’s grounded instead as follows: that an overlapping consensus from comprehensive premisses would not be good enough as a justification of a conception of justice to each person because it’s not the case that each person in the overlapping consensus could sincerely believe that many of those premisses justified the imposition of the principles of the conception of justice on others, given her own comprehensive views.

Public reasons, intimately bound up with the “certain fundamental political values and ideas (citizens as free and equal, society as a fair system of social cooperation, and the burdens of judgement)” that Jon mentions, can by contrast provide justifications to each person, since each person in the overlapping consensus can sincerely believe that such reasons justify the imposition of principles of the supposedly justified conception of justice on others. Anyone offering such reasons in support of some conception is offering premisses which others take them to be justified in holding. This is so because the fact of the overlapping consensus guarantees that the fundamental political values and ideas which give rise to the conception of public reason are themselves shared by each person.

Now, you might worry that even if this solves the problem here, some version of the same problem will recur when we come to ask how the fundamental values and ideas are justified, since consensus members’ answers to this question will vary from comprehensive doctrine to comprehensive doctrine. How can one member of the consensus see another’s membership itself as justified? Another way to put the worry would be to say that it’s not clear that we can insulate justifications which start from the fundamental values and ideas from the justifications which get us to them in the first place.

I’m not sure what to say about these worries (I take Jon’s comment in #13 that this “is not … a matter of public justification, but rather a matter for what Paul Weithman calls ‘comprehensive public philosophy’” to be gesturing at one possible response), but in any case the position I’ve just outlined seems not to be at least obviously vulnerable to your charge of demandingness or to the objection that it presupposes mutual disdain among citizens.

It’s worth noting also this position is actually compatible with your first answer to the question why public justification in terms of public reason is necessary. It’s necessary because there can’t be the appropriate sort of stability (enduring stability for the right reasons) without it, as you say. But that seems to leave open a question about whether public reason should be conceived in terms of consensus justification or convergence justification. The point I take Jon to be making is that in fact this question isn’t left open: the need for public reason is a need for consensus justification, given the sincerity requirement. The convergence conception of justification doesn’t in fact give us a conception of public reason which will do the job at all.

The question as I see it is: what is it that makes public justification of a conception of justice in terms of public reason necessary in political liberalism? Why not settle for an overlapping consensus on a conception justice based on comprehensive premises? There are at least two different explanations of this necessity.

1. The first explanation is that there can’t actually be an enduring overlapping consensus from comprehensive premises on a conception of justice unless that conception of justice can also be justified in terms of public reason. Reasonable pluralism is just too broad and open-ended. We can always imagine some or other reasonable comprehensive moral doctrine that lies outside the proposed convergence, and sooner or later some reasonable citizen will espouse that doctrine. Since the conception of justice is to endure as the regulative conception of justice for a society across generations, we cannot rest content with a temporary convergence of the doctrines that happen to exist at the moment. One function of an argument for the conception within public reason is to demonstrate that all reasonable citizens have some reason to accept the conception, whatever their specific comprehensive doctrine may be.

I think this is a pretty good argument. If reasonable acceptability to each reasonable citizen is a requirement, if the asymmetry argument is not accepted, and if we are talking about a conception of justice that is to endure indefinitely, then I think we would need some sort of public justification cashed out in terms of something like public reason. (I don’t accept the first premise, but I think the rest of the argument works.)

2. Your explanation of the need for public justification in terms of political values is that an overlapping consensus from comprehensive premises would not be good enough as a justification of a conception of justice to each person. It would not be good enough since, even though they are reasonable beliefs, a whole bunch of those comprehensive premises would be false. Since they are false, they would not actually support the conclusion. Hence citizens who accept the conception of justice on the basis of those premises would not really be justified in doing so, and that triggers your sincerity concern. So a purely comprehensive overlapping consensus is in principle inadequate.

This second explanation is more demanding than the first, since it requires that there be actual, all-the-way-down, 100% genuine justification for each person to accept the conception of justice, and not simply that it be reasonably acceptable to reasonable citizens, whether or not they are genuinely justified, all the way down, in endorsing the conception.

But in addition to the grounds for objection issue I mentioned above, there is also the consideration that it may not be desirable to conceive of citizens motivating the project of public reason by appeal to the idea that the comprehensive doctrines of other citizens don’t really perform justificatory work. This seems to involve a presumption of mutual disdain at the level of comprehensive moral doctrines. This premises public reason on an essentially comprehensive moral claim, and one which will itself be controversial between comprehensive doctrines (imagine an ecumenical view that says that every reasonable comprehensive doctrine’s path to a liberal theory of justice is equally and entirely valid, and provides perfectly adequate justification for endorsement — I think you want to deny this view, but I don’t know if you can do that qua political liberal).

That being said, I’m a bit puzzled at how this section fits with the general idea of convergence. Page 17: “Even if a secular rationale is necessary in our society for a publicly justified law, it can be defeated by a reasonable religious conviction without any secular backing”. The first part of that sentence refers to p.15: “[G]iven the contingent facts of contemporary Western society, if citizen Alf proposes L on purely reasonable religious grounds, for Alf to legitimately endorse L in the public sphere he must believe that there are grounds that plausibly justify L to reasonable non-religious members of the public” Isn’t the same true vice versa? If there are reasonable religious views and reasonable non-religious views, then all laws must have both religious and non-religious rationales; the religious are constrained to support only those laws that reasonable non-religious people support, and the non-religious are constrained to support only those laws that reasonable religious people support. Reformulating p.17, couldn’t we say ‘even if a religious rationale is necessary in our society for a publicly justified law, it can be defeated by a reasonable secular conviction without any religious backing’?

Very interesting paper. I’m intrigued by the final section, the one that suggests that the public justifiability of laws is a criterion of institutional design but not a rule we want citizens themselves to apply. Do we have a name for this yet? indirect neutrality? Government house Kantianism? A bit disheartening, though, to discover yet another dimension of possible variation in conceptions of public reason.

I have a couple of comments, but I’m going to separate them into different posts. This comment is about consensus vs. convergence.

I’m not clear on the relationship between these two modes or models of public justification. Page 11: “To be sure, there is nothing wrong with a public justification built on consensus - that is one way a law might be publicly justified to all. But the aim of consensus tends to be frustrated…” This statement seems to suggest that convergence is the necessary general condition, and consensus the fortunate but not necessary special case. But that’s not right. Neither condition nests the other, because there can be consensus without convergence and convergence without consensus.

The way I see it, each approach makes a plausible objection to the other. The convergence objection to consensus focuses on a case in which we have convergence but no consensus. This is a case in which all qualified points of view converge for their various reasons on law L, but if one looks at the set of supporting premises they all accept, one can’t find a set of shared (i.e. public) reasons sufficient to justify L. So all reasonable people accept L, but because they do so for conflicting reasons, they can’t vote for L, on the consensus approach. That strikes me as counter-intuitive, despite Jon’s account of sincerity.

The consensus objection to convergence focuses on a case in which we have consensus but no convergence. That is to say, some qualified points of view reject L, but if one looks at the set of supporting premises all qualified views accept, there is a set of shared reasons sufficient to justify L. This strikes me as a counter-intuitive feature of the convergence view. Just because some reasonable comprehensive doctrine opposes L, we can’t vote for L, on the convergence view, even though there may be a good case for L based on premises that all qualified views accept.

Thanks again for your detailed reply to my initial commentary. I have a few very brief comments in response, but I’d like to start with a small apology. We’ve just begun the first full week of term here in Manchester, and so I might not be able to find the time over the next few days for continued participation – I just didn’t want you or anyone else to think I was ignoring what has been a terrific conversation thus far. I’m also going to use this as my excuse for the very brief and tentative comments below. I’ll take your points in order.

1. First, you strongly caution against moralizing the notion of coercion. But I don’t think I’m the one moralizing the notion of coercion: you and Jerry are doing that. If we declare there is a strong presumption against some form of action, X, then X is a moralized concept since we’ve declared that it’s something not to be done except under unusual circumstances. I don’t think one can say there is a moral presumption against doing X, and maintain that X is not a moralized notion.

You also warn that if we moralize the concept of coercion then we are undermining the project of public justification. Public justification, you point out, is supposed to justify moral and political norms, but if we rely on already moralized concepts, then it seems like something other than public justification is playing a determining role in setting the moral and political norms. I don’t disagree with this point. As I see it, public justification does determine (at least) political rules. Only once we have these prima facie rules in place can we know what constitutes coercion. Any act that then appears coercive would stand in need of justification. Consider the following simple example. Until we know who owns that bit of property over there, we cannot know whether my statement ‘I’m going to destroy that bit of property unless you do as I say’ is a coercive threat or not. I agree that we must rely on the idea of public justification to determine who owns that bit of property, and I also agree that if the statement turns out to be a coercive threat, we must declare there is a presumption against the threatened action unless there is some further decisive public justification that can be provided.

2. I suppose I don’t think we can defend a presumption in favour of liberty on the grounds that it makes political decisions easier. If there is a presumption in favour of liberty in some domain or with regard to some set of actions, it must be because we have already weighed all the reasons that will usually be pertinent to that domain and decided that liberty looks to be the decisive value. You can’t get away from the problem of trade-offs or weighing values by stipulating a presumption in favour of one value since that presumption must itself have been justified by weighing all the relevant reasons.

3. You worry that if we abandon a unique presumption in favour of liberty then we might not have a liberal theory at all, but I don’t think we need to worry about that. We could have, for example, a theory where liberty, equality, and reciprocity were all important moral values, and that doesn’t mean the resulting theory won’t be liberal, rather it might be a form of liberal egalitarianism. To say that there must be a unique presumption of liberty in order to have a liberal theory sounds too strong to me. I think Rawls’s theory is egalitarian, for example, but there is no unique presumption in favour of equality in his theory.

4. Finally, with regard to Jerry’s position on the presumption in favour of liberty. I’ve heard Jerry talk about this, but I confess I’m still not sure I fully understand the distinction he’s driving at (this in no way implies that my lack of understanding is caused by Jerry’s lack of clarity. I’m sure my lack of understanding is my own fault!). Since I’m not sure I’ve yet got my head round the distinction, I think I’ll say very little about it right now. The only thing I will say is this. The presumption in favour of liberty does seem to work as a presumption within morality in your paper. The liberty principle, as it functions in your argument in defence of religious defeaters, means that a religious person can legitimately refuse to do what would be a demand of justice for a non-religious person, provided that religious person does in fact have a defeating reason within their belief-set. The presumption works in a way that means the moral demands on the religious citizen (Alf in the example from my commentary), are different (weaker) than the demands on a non-religious citizen who lacks any defeating reason. So I’m puzzled. But as I said above, I am also certain I’ve not yet grasped the full force of Jerry’s point here, so I’m sure there is more to this than I can see.

Thanks again for the thoughtful replies, and sorry that I might be out of touch for a couple of days.

***

Simon: thanks for your further thoughts on Rawls and the overlapping consensus. I’d like to briefly respond.

A key difficulty, I think, is that we have very differing views regarding the role of the overlapping consensus within Rawlsian political liberalism. I have to run off to class, so I’m afraid this will be far too brief, but I think the difference between us is that you are supposing the project of public justification applies to the task of securing an overlapping consensus on a political conception of justice. I, on the other hand, see public justification as applying only to the political discussions that occur between citizens who are already committed to certain fundamental political values and ideas (citizens as free and equal, society as a fair system of social cooperation, and the burdens of judgement). It is a commitment to those ideas (free, equal, fair, and burdens) that, for me, then activates the project of public reason within Rawlsian liberalism, and thus there is no public justification in the normal sense when we ask whether people should endorse those values in the first place. How people decide to become members of an overlapping consensus on those fundamental political values is not (in my perhaps unorthodox version of political liberalism) a matter of public justification, but rather a matter for what Paul Weithman calls ‘comprehensive public philosophy’ (see his paper ‘Liberalism and the Political Character of Political Philosophy). That stuff is, for me, something that falls outside the domain of public reasoning or justification within Rawlsian political liberalism. Obviously things are different in Jerry and Kevin’s theory, and that’s why they need to worry about whether people’s comprehensive doctrines are justified, but the sincerity requirement doesn’t apply in the same way within the Rawlsian framework.

Sorry, I realize that I probably haven’t managed to clarify my position there, but I have to run. Hopefully we can talk about it more later this week.

Now this is not an essential commitment of liberalism at all. A liberal theory must include, somewhere along the line, an important role for liberties, and accord those liberties a great deal of weight relative to other kinds of values (whether that amounts to lexical priority or not). But the liberty principle is much stronger than this.

Instead of the liberty principle serving to specify a default presumption, we could argue that nothing acts as a default presumption at all. Whatever the final set of arrangements, whether anarchy or command economy socialism or something in between, that set of arrangements must satisfy whatever the operative standard of public justification is set to be. Thus, no alternative should win because all the other alternatives lose. If every conceivable arrangement fails to satisfy the standard of public justification, then either we relax that standard of justification or we are left with an indeterminacy thesis of some sort. In contractualist stories, we’re not left with the state of nature as a baseline, but non-agreement.

To recap, Jerry and I argue in the paper that when we relax the consensus requirement on public reasons we can more easily see our way to rejecting the error of symmetry, or the view that the same standards apply to reasons to reject as reasons to propose.

As a result, we can allow religious reasons to play the traditional liberal role as defeaters but not as justifying coercion based on religious reasons, something we think is off the table.

There are ways to go after the connections between these concepts. Micah Schwartzman has openly worried about how Jerry and I will handle some cases and Simon May is concerned that convergence isn’t directly connected to the role of religious reasons as defeaters. I hope to reply to these concerns, but for now I will address Jon’s argument.

I take it that Jon’s main argument is that the religious reasons as defeaters position relies crucially on the liberty principle that Jerry and I assume in the paper. I will not address all of Jon’s concerns here. Instead, I will make three brief comments about Jon’s response and present a general reply.

(1) Jon argues that we may need to moralize the notion of coercion in order to determine whether an act is coercive. He argues on page 9: “Determining whether or not someone is acting within their rights (not acting unjustly) thus appears to be at least one of the variables necessary to decide if an action is an act of coercion.”

I want to strongly caution against this move. One of the points of justificatory liberalism is to justify moral and political norms. If we employ concepts that are already moralized, we defeat in part the point of being justificatory liberals because we are using moral and political concepts that already have substantive moral content. For what moral theory will be the source for the content of these concepts? I think it is better to try not to moralize the concepts if it can be avoided (and I think it can be). Myself, I prefer something like Nozick’s definition which only appeals to norms of expectation in fleshing out its definition but I will not defend that view here. Of course, Jerry has outlined his view in detail elsewhere.

(2) One difficulty with abandoning a unique presumption in favor of coercion is that justificatory liberalism must not decide on trade-offs between distinct presumptions. The reason Rawls wants to give liberty (near?) lexical priority is in part because it makes a contractarian decision procedures much easier to employ. If we abandon a unique presumption in favor of liberty, then probably we are committed to abandoning a unique presumption in favor of any particular value. But then the problem of making trade-offs becomes incredibly difficult.

(3) It is not clear that if we abandon a unique presumption in favor of liberty that we have a liberal theory of justice at all. And that’s what I thought the justificatory liberal project was.

General Comment:

Jerry has decided that the role of the presumption in favor of liberty has not been laid out clearly enough in his work. We’ve talked about this extensively, particularly last term when two papers at the Public Reason Conference here at Arizona (Steven Wall and Andrew Lister’s papers, specifically) raised the same concern that Jon Quong has. The concern is simple:

“Ok, sure there’s a presumption in favor of liberty. But there are also presumptions in favor of all kinds of other stuff, like need or against harm, etc. Why privilege the presumption in favor of liberty? What norms justify such a presumption?”

Jerry’s view is that the presumption against interference/coercion or in favor of liberty is not a presumption within morality. Instead it functions more as a baptizer of moral and political norms. Following Scanlon, Darwall, Feinberg, Benn and Rawls, Jerry thinks that a putative norm can only be a bona fide norm if we can show that others have overriding reason of their own to acknowledge the force of the norm. And this means that some presumption in favor of individual autonomy and against interference lies at the heart of moral theory and (he thinks) liberal political theory generally.

When we make moral demands on others and especially when we coerce them, we place a demand on them. But what makes this demand normative in the first place? We could explain the moral force of the demand in terms of well-being or utility. But the deontological nature of justificatory liberal political theory weighs against taking such a position. Instead Jerry thinks we must answer this question by appealing to something like the second-person standpoint.

So again, the presumption against coercion or interference is a gatekeeper for valid moral claims and not itself a presumption within morality.

I think that Jon Quong’s arguments against the liberty principle presupposes, as I think Wall and Lister’s papers did, that Jerry conceives of the presumption as one within morality. But then the concern is obvious