Public reason as solving a fundamental disagreement problem

One strand in the “public reason” approach to political justification, stated in a very general form, might go something like this: in the context of disagreement about which shared, public norms to codify and enforce, when actual consensus is not present, the obvious or evident nature of important, relevant objective reasons, the recognition of which would rationally tend to lead to consensus (this situation constituting a certain kind of hypothetical consensus) suffices to make for permission to enforce.  One way to characterize this general approach is by saying that there is a “disagreement problem” specific to the relevant context that is “solved” by taking steps toward consensus (with enough truth in it) on the matter as against simply seeking the truth. Here is a specific proposal along these lines, a first shot in the right direction with this kind of approach, to play with and evaluate:


Group G is permitted to enforce norm N on the members of G iff (a) no members of G reject, on the basis of quite strong and evident reasons, G enforcing N and (b) no members of G fail to reject, on the basis of ignorance of strong and evident reasons, G enforcing N.

I’ll note a couple of things about this proposal/formulation and what I’m trying to do with it.  

First, with Estlund, I’m using the acceptance/rejection/acquiescence attitudinal stances here because belief does not seem most salient.  For example, I might believe that G enforcing N is not what G should do here – maybe I think G enforcing N* is quite a bit better and an available alternative – but nevertheless acquiesce in G enforcing N (in a not-super-unreasonable way) and this seems to be what matters.  

Second, with the two conjuncts here, I’m allowing both (i) passive non-rejection (acquiescence, non-consideration) to do normative work subject to only weak normative conditions (roughly, this stance not being super-unreasonable in anyone) and (ii) rejection to do normative work only when stronger normative conditions (roughly, not being super-unreasonable) are met.  This is because it seems that you have to do both of these things in order to solve the problem at hand (roughly, that of achieving normative conditions for fair consensus, the actual achievement of such being unrealistic). At a more intuition-driven level, you need to include these sorts of elements in order for something like a theory of legitimacy (in Estlund’s sense) to do the work that we think it should do.  

Third, both conjuncts incorporate a theory of reasonableness that requires some contact with reality (knowledge, correct valuing including the value of basic norms or models of fairness in relevant circumstances).  As a result, I’m not so concerned about the extent to which false reasons or conclusions are allowed to do work, though the model is biased against this and might, with more precisification, become more or less thus biased.  

Fourth, relatedly, this proposal seems to be different, in many ways that I am not fully tracking right now, from how Estlund sometimes seems to think of public reason simply as a response to something like this sort of disagreement (it is certainly quite different from how, in a different mood, he thinks of public reason as a “tolerance response” to certain “moral worldview” beliefs that might well be and often are false or quite unreasonable).  What jumps out at me is that, by rejecting the religious tolerance model and incorporating objective normative reasons (facts, ignorance of them) in the relevant benchmarks of reasonableness, my approach is much less truth-phobic and much less concerned to accommodate false and unreasonable religious or similar moral-worldview beliefs (i.e., in objections/rejections or reasons for such).


37 thoughts on “Public reason as solving a fundamental disagreement problem

  1. Haven’t had time yet to engage with this post in a substantive way, but absolutely could not resist copying and pasting this example of public irrationality as solving a fundamental disagreement problem. From a Facebook group that non-ironically calls itself “Open Group on Objectivism”:

    Alas, it has become necessary to write this.

    This is an open group. That is: it’s not restricted to members of a private club. It is a group on Objectivism. This means it’s for discussion of Objectivist philosophy and other topics from an Objectivist point of view.

    It does not mean that we are open to debate on what Objectivism is or that Objectivism is wrong. You don’t have to be an Objectivist to participate here, but don’t tell us that Objectivism is wrong, and don’t misrepresent what Objectivism is.

    It’s an open group on Objectivism, but the truth of Objectivism can’t be questioned; the moderators reserve the right to tell you want counts or doesn’t count as Objectivism, but you can’t dispute their interpretation. This makes Rawlsian public reason seem sensible by comparison.

    Liked by 1 person

    • That doesn’t strike me as quite so ludicrous as you suggest. What doesn’t seem ludicrous to me is the idea of an open group — a group that does not restrict participation to those who accept a particular set of claims — that is nonetheless not open to debate about the truth of its core claims or their correct interpretation. After all, that’s how many religious groups operate. You don’t have to be Catholic to go to Mass — though you do have to be Catholic to participate fully — or even to come to any number of less formal, non-sacramental get-togethers. But you do not get to openly question Catholic doctrine in Mass, nor would you be welcome to challenge it or to debate the correct interpretation of Catholic doctrine in, say, a less formal Catholic prayer group, or a reading group, or an ‘exploration’ group. I see nothing wrong with that in principle; the groups and their activities are open to people who don’t buy this stuff, whether they’re curious and open to it or just there out of anthropological curiosity, but debating the truth of Catholicism is not part of what those groups or activities are about, so we’re not welcome to go do that there. That of course makes perfect sense at Mass, because Mass is not a discussion group. But it makes equally good sense in certain sorts of discussion groups; come to the group on Catholic Social Teaching if you like, but don’t show up and start arguing that the problem of evil disproves the existence of God, that scripture is inconsistent with magisterial authority, that the doctrine of the Immaculate Conception is bogus, or any of that jazz. Of course it would be ridiculous for Catholics to refuse ever to listen to any arguments against Catholicism in any forum whatsoever, but one can exclude critiques of that sort from certain venues without being a close-minded dogmatist.

      What makes the policy of the Objectivist ‘open group’ seem less sensible is that Objectivism is supposed to be a sort of philosophy, and philosophy is supposed to be about reason and argument and stuff like that. It’s definitely not supposed to be like religion, and especially not anything like Catholicism.

      Alas, it seems to me that in some respects it would be better off if it were more like Catholicism; after all, there are some Catholics who are first-rate philosophers, but I can’t think of a single Objectivist that I’d put in that category.


      • Well, I was assuming that the thing I was describing as ludicrous wasn’t a worship service–as I think the “discussants” involved would be the first to insist. I’d love to hear what they’d have to say at having their Objectivist discussion group compared to a Catholic mass, but then, I’d have to accept their terms to ask the question, and that’s out of the question.

        What makes the policy ludicrous is not just that it’s a discussion group purporting to discuss philosophy (though there’s that), but that it’s an “open” philosophy discussion group of a philosophy that lacks a canonical statement of its doctrines. There is no canonical account “within Objectivism” (whatever that phrase even means) of either the core or the non-core doctrines of Objectivism, of the criteria for “corehood,” or frankly, of a single actual doctrine, much less of the justification for one. I can almost conceive of someone innocently failing to see that forty years ago (almost), but I can’t even begin to get my mind around a mentality incapable of seeing it thirty years after the Kelley-Peikoff dispute (which was about this). (I won’t belabor the point, but contrary to popular belief, Leonard Peikoff’s own candid admission, his Objectivism book is not and logically cannot be a canonical statement of the doctrines of Objectivism, even apart from the disputability of the claims in it.)

        So it’s not just that a philosophy discussion group isn’t a worship service, or that Objectivism claims not to be a faith. Catholicism has had centuries to work out and explicitly formulate its doctrines (and has done so). Objectivism has had none of that, and done none of it. So it’s unsurprising that in practice the “philosophy discussion group” in question just consists of bitching about Alexandria Ocasio-Cortez. There’s not even a pretense of doing anything that resembles philosophy. The model is neither a philosophy discussion nor Catholic mass (which outsiders can at least enjoy on aesthetic grounds), but a bunch of drunk people railing at one another in a bar.


        • See, that’s just why Objectivists, like Protestants, can’t rely on a doctrine of sola scriptura and at the same time insist on doctrinal purity: there’s no canonical statement of doctrines in the scripture. Plainly they’re working with a kind of apostolic succession: Rand -> Peikoff -> some successor that Peikoff will name. Perhaps he should create a college of cardinals to name his successor instead?

          I’d think that, given that there is no Objectivist magisterium, it’d obviously be impossible to refuse to allow for diversity and disagreement in the interpretation of Objectivism, no matter what status is given to this or that text by someone other than Rand. Texts don’t interpret themselves and their plain meaning does not answer every question that one might reasonably have about them. So, while it’s not entirely silly to refuse to allow certain sorts of polemical ‘debate’ in an ‘open group,’ it is pretty silly to insist on some particular interpretation and to refuse to discuss differences. But that’s what happens in many Protestant churches — it is, in a way, the entire history of Protestantism — so it’s not at all surprising to see it happening here.

          The main point is, of course, that Objectivism mirrors religion in ways that its proponents ought to find uncomfortable.

          Liked by 1 person

          • Peikoff & Co. get sola scriptura and doctrinal purity by insisting that Rand’s texts are sui generis: they are too clear to require interpretation, and they address every legitimate question worth asking, unanswered questions being either illegitimate, pointless, or relatively trivial.

            Peikoff’s position on apostolic success is remarkably Muslim, or least Sunni. The Sunni Muslim view is that there is no deity but God, and Muhammad is God’s prophet; Muhammad is chronologically the last of the prophets, bearing Divine Revelation 3.0, the last and best version of the Abrahamic Program; no improvements are needed or permissible (except, of course, minor adjustments to account for the minor differences in context between, say, 7th century Arabia and other times and places). Of course, even Muhammad was fallible, so that while God’s direct word is infallible, Muhammad’s applications and commentaries may–despite his exemplary status–have involved some screw-ups (peace be upon him). This is not to invite frivolous skepticism about Muhammad; few have the standing for such skepticism. The point is that not even Muhammad is God.

            Yes, there needs to be a caliphate after Muhammad, but that’s a practical political matter which Muhammad himself did not bother to address or solve. I mean, he unified Arabia. What more do you want?

            Peikoff’s position is Sunni Objectivism. There is no deity but Objectivism, and Peikoff is its prophet, the last in line, etc., But Peikoff’s writing Objectivism is analogous, not to the divine revelation (or even a commentary on it, which would be presumptuous), but to Muhammad’s conquest of Arabia. It was a practical rather than theoretical accomplishment. It gathered up the doctrines–as Muhammad gathered up the tribes of Arabia–and unified them. It produced nothing novel, Rand forbid. Not even the organizational structure is novel. Paradoxically, the content of the book is not Objectivist doctrine, having passed through mortal hands.How to characterize it? Divine-revelation-passed-through-exemplary-but-fallible-hands.

            Yes, there needs to be an Objectivist caliphate after Peikoff, but he’s not going to bother to appoint one. He wrote OPAR. You couldn’t ask for more.

            But in case you did, feel free to waste the next 35 minutes of your life listening to Prophet Peikoff (peace be upon him):

            I can’t wait for his fiction to come out. Can you?

            Liked by 2 people

            • I read 35 minutes as 5 minutes and was going to click – thus breaking my long-standing abstinence from all things officially Objectivist. Happily, I corrected my error. Also I want Munch’s The Scream in emoji form.

              Liked by 1 person

  2. I think I still struggle with this version of the public reason approach for the same basic reason I struggle with others: we’ve got to say what counts as reasonable, and in the process of doing so we’re going to end up taking some controversial positions, so we might as well just abandon the approach to begin with. I wonder, though, whether I might misunderstand part of what you say. We get this:

    Group G is permitted to enforce norm N on the members of G iff (a) no members of G reject, on the basis of quite strong and evident reasons, G enforcing N and (b) no members of G fail to reject, on the basis of ignorance of strong and evident reasons, G enforcing N.

    How am I to read (b)? Every member of G rejects the enforcement of N out of ignorance of good reasons? Or: if any member rejects the enforcement of N, she rejects it on the basis of ignorance of good reasons? Presumably the latter.

    Still, it strikes me as impossible to offer any account of what the “quite strong and evident reasons” are that will be both (i) sufficiently robust to ground political legitimacy and (ii) endorsed by everyone that we would recognize as reasonable on independent grounds. Why not just say something like the following: whether or not the state or the law or what have you is genuinely authoritative and/or legitimate is a question of substantive moral philosophy about which reasonable, informed, honest, and decent people will give different answers or give the same answer for different reasons; theoretical consensus is not realistically achievable, but the achievement of such consensus is not a pressing practical matter; the pressing practical matter is for practical consensus about the framework in which we live together and negotiate our various disputes, and the kind of overlapping consensus we need is less likely to be achieved by policing the sorts of reasons that we can offer each other and thereby ruling most of our opponents’ views out of bounds than by (a) insisting that we offer each other reasons when possible rather than resorting to force and (b) allowing people a considerable space of liberty to live out their alternative and rival commitments. This response of course does not solve any hard problems by itself, but neither does public reason liberalism; it creates fewer problems by encouraging us to seek points of agreement with one another and work from there rather than demanding, in the name of tolerance, that people set aside their deeply held views when they think about what laws and governments should do.

    But how different, ultimately, is a proposal of that sort from what you’re sketching out here? It seems quite different than the idea of public reason as applied not narrowly to what this or that public official says in an official capacity, but to public discourse more broadly and even to philosophy itself. On my alternative, we don’t waste time trying to formulate some sort of relatively controversy-free theory of public rationality; we reason however seems best to us, and then we engage in public argument with people who disagree with us, starting from wherever we find ourselves, necessarily appealing to different considerations in different debates with different sorts of people. We may not get anywhere, but we’ll at least have discarded the burden of first trying to convince people that their beliefs don’t pass some test that they don’t accept as a criterion for public acceptability.

    Liked by 1 person

    • What’s the difference between theoretical and practical consensus?

      I agree with both of your criticisms of Michael’s proposal (the first being the one about the unclarity of conjunct [b], the second being the one about the need for uncontroversial criteria for “strong and evident reasons”). But I think your proposal brings a version of Michael’s principle in through the back door.

      the kind of overlapping consensus we need is less likely to be achieved by policing the sorts of reasons that we can offer each other and thereby ruling most of our opponents’ views out of bounds than by (a) insisting that we offer each other reasons when possible rather than resorting to force and (b) allowing people a considerable space of liberty to live out their alternative and rival commitments.

      The “less likely to be achieved,” “when possible,” and “considerable space of liberty” provisos all leave the door open to some version of Michael’s principle. In other words, they leave open the possibility that consensus still requires us to police some sorts of reasons, to abandon reason-giving after a certain point, and to allow people considerable but not unlimited space to live out their commitments, where some of them will just have to be shut down without much explanation.

      Maybe the issue here is that some version of (something like) public reason is unavoidable. If what we’re talking about is formal political deliberation or discussion, some kinds of reasons are going to be excluded on any reasonable account. The problem with the specifically Rawlsian/Estlundian conception of public reason is Rawls’s front-loading the account against perfectionist conceptions of ethics and politics, moral realist conceptions of meta-ethics, and alethic conceptions of justification. That’s too narrow, constrained, and tendentious a conception of public reason (and Estlund seems to follow Rawls in the relevant respects). But a less narrow, constrained, and tendentious version of public reason is probably unavoidable, at least for the narrowly and formally political context Rawls intends.

      Maybe this argues for our reading something by Kevin Vallier and/or Paul Weithman at some point, both Rawlsians with Catholic commitments trying to broaden Rawlsian strictures to accommodate faith commitments like theirs. You probably have to go a few steps beyond Vallier and Weithman to get the right account, but their views may at least be a step (or two) in the right direction.


      • “The “less likely to be achieved,” “when possible,” and “considerable space of liberty” provisos all leave the door open to some version of Michael’s principle. In other words, they leave open the possibility that consensus still requires us to police some sorts of reasons, to abandon reason-giving after a certain point, and to allow people considerable but not unlimited space to live out their commitments, where some of them will just have to be shut down without much explanation.”

        Agreed, with the qualification that there’s no obvious reason why we have to shut things down without much explanation. We can explain all we want. Of course some claims that we reject will be so wild that we won’t bother, but that’s true for public reason political liberals, too.

        What I’m not so clear about is whether we get all the way to Michael’s principle here, or whether we get to anything that we should call by the same name as the views we get from Rawls, Estlund, Vallier, et al. I’m inclined to agree that we should consider reading some Vallier and Weithman. My limited reading of Weithman left me with the impression that he is just a Rawlsian political liberal, but it was pretty limited and pretty long ago.


        • Actually, I was thinking about “shutting down” points of view less on grounds of craziness than on grounds of dangerousness: there are going to be times when you shut a view down without explanation because it’s more urgent to shut it down than explain why you have to. I was thinking of cases like involuntary commitment of suicidal people on psychiatric grounds, or cases like the Branch Davidians at Waco, where a lifestyle is taken to be dangerous to the surrounding population. In the nature of the case, you can’t explain yourself in an psychiatric emergency, or during the execution of a search warrant. But you may still have to act. And you may, in these cases, not bear any burden of having to explain yourself retrospectively, either. It might be superfluous in the case of the psychiatric emergency (the patient no longer needs it, having emerged from the emergency), or might be overtaken by later events (the people in question are now murder suspects rather than interlocutors). I agree that that might be the case either on your view or on public reason; I regard the two as equivalent in that respect.

          I hesitate even over these cases, though, and over the idea that some views are too crazy to require engagement. Yes, some views are crazy (psychiatric emergencies involve craziness by definition), but the more coercively you deal with someone, the more of a burden you bear (it seems to me) to produce an explanation/justification for your use of coercion that’s responsive to their desire not to be coerced. Obviously, if the situation is urgently dangerous, the explanation/justification can’t be offered at the moment in real time. But arguably, an explanation/justification is owed retrospectively unless the party in question waives the need for it. If you involuntarily commit a mentally ill person, you may owe them an account that justifies your actions once the psychiatric emergency is over. Arguably, the Clinton Administration was forced by Congressional Republicans to offer a justification for the ATF/FBI raid at Waco whether or not that justification was addressed to the Branch Davidians themselves, or responsive to their theology (which it wasn’t). My point simply that the “shutting down” issue is tricky and involved.

          Both Vallier and Weithman have books out on our topic. Here’s Weithman’s:

          And here’s Vallier’s:

          I helped Paul put this conference on; the intro is a nice, concise, Weithmanly account of the book and of his views:

          I know Paul’s views more from talking to him (twenty years ago) than reading him. I have no reasonable faith that I can summarize them accurately, so I’ll leave things there.

          Peikoff is a different matter.

          Liked by 1 person

          • The requirement of retrospective justification or explanation seems either trivial or too strong. It’s trivial if it means that we have to offer a good reason for what we’ve done; if we were justified in doing it, there was a good reason, so we can just say what it was. It’s too strong if it has to be a justification that the person to whom we’re offering the justification will in fact accept regardless of what his beliefs are; that just makes justification hostage to irrationality and stupidity.


            • I don’t think it’s trivial to say that we have to offer a good reason for what we’ve done after the fact. The non-trivial requirement is that you have to go through the process of actually giving a retrospective justification that is responsive to the (somewhat rational) concerns of the person in question. The alternative might be to do what you were going to do with justification in hand, and treat doing it as a fait accompli once it was done without any further requirement.

              The model I have in mind is a warrantless search or seizure in criminal procedure. On some views, if you execute a warrantless search on grounds of exigency, you have to go back after the fact and produce the justification you had at the time for executing the warrant (i.e., go through some equivalent of the warrant process involving a sworn affidavit, etc.). It’s essential, actually, that the justification be literally retrospective. It isn’t kosher to execute the warrant and then produce a justification based on what you’ve found through the execution. The retrospective justification has to appeal only to facts known prior to its execution (which is easy to fudge in practice, but the distinction is clear enough conceptually).

              The justification need not be hostage to irrationality or stupidity, but it has to be responsive to the particularities of an actual person. For instance, if you psychiatrically confine someone, once they’re out of a psychotic state, you may still have to respond to particular fears they happen to have about psychiatric treatment. From the perspective of an experienced clinician, many of these fears might appear (or even be) irrational or stupid. But you have a duty to respond to them anyway, at least within limits. It’s understood within clinical settings that people are especially liable to what we would ordinarily call irrationality, but the whole point of clinical training is to be able to deal with that partly on its own terms.


              • The requirement for warrantless searches seems trivial; I mean, if you perform a search without a warrant, you’d better have a good reason, and so you should be able to give it. So too, if you psychiatrically confine someone, you’d better have a good reason, and so you should be able to give it. If the person you’re confining or whose property you’re searching has to accept your reason, then the demand is too strict, but if it just has to be a good reason — specified in part, in the relevant instances, by laws — then it seems hard to see why it would be hard to meet, provided that you actually have a good reason for what you’re doing. Of course people often fail to have good reasons for what they do, but that’s irrelevant. If you have good reason to do a thing — and I’m not interested in defending any cases in which people don’t — then you should be able to give that reason. I worry that I’m misunderstanding what you have in mind, because I’m having a hard time seeing why anybody would want to say otherwise, or why this or that particular political philosophical view will have a hard time meeting this requirement.


      • Oh, and what’s the difference between practical and theoretical consensus? Roughly that in practical consensus, we agree about what to do, while in theoretical consensus (about practical matters), we agree about why we should do it. We don’t have theoretical consensus about why it is wrong to murder 5 year olds, but we have an overwhelming practical consensus that it is wrong to murder 5 year olds (and a consensus that goes beyond trivial implications of the word ‘murder,’ too, since most people will agree on a wide range of central cases no matter how much disagreement we find in certain borderline cases). More relevantly to our recent discussions, I take it that we — Americans, at least, and a lot more besides — have a strong practical consensus in support of democracy qua universal suffrage, and hence against Brennanite epistocracy. But there’s obviously no theoretical consensus on why we should have democracy rather than Brennanite epistocracy. We don’t need that theoretical consensus, and in a way the case for democracy might even be strengthened by the lack of theoretical consensus about it — we might think that if we get a good case for democracy on so many different, competing political views, we have pretty good reasons to support democracy even if the case for each pro-democratic view seems independently somewhat weak. That’s of course not to say that Brennanite epistocrats should not be permitted to argue against democracy, or even that they’re mistaken; maybe they’re right and we need to work toward implementing epistocracy. In any case, we need some strong practical consensus about it, or else we’ll end up in something approaching the condition of civil war.

        I think the main challenge for a view like mine is to articulate it more fully in a way that avoids either collapsing into a quasi-Hobbesian sort of view that treats all acquiescence, even out of mere fear, as establishing practical consensus, or building so many moral constraints into the idea of consensus as to end up with a view broadly like Rawls or Estlund’s — the sort that says that since people would agree if they accepted my theory, they can be justly coerced. I think I’m closer to the Hobbesian horn of that dilemma.


        • I guess I’m a little skeptical that your distinction between practical and theoretical consensus can stably be maintained in political life. It’s very difficult to get practical consensus without some degree of theoretical consensus. There are, to be sure, degrees of theoretical consensus corresponding to the comprehensiveness of the sort of justification you’re seeking, or seeking to give. But the distinction doesn’t work as a sharp one, and since it’s not sharp, I’m not sure how much work it will do once applied to a range of topics.

          A relevant book to read on this topic would be Tom Tyler’s Why People Obey the Law, an empirical account of adherence to the law in the United States.

          I don’t mean that Tyler literally resolves your dilemma, but his view suggests that actual adherence arises from people’s (minimal) grasp of the moral rationale for a law. Assuming that a law has a justifiable rationale, and that rationale is either competently conveyed or widely understood, people tend to obey because it seems moral to do so not because they fear the sanctions of punishment.


          • I think I just don’t see the truth of the claim that it’s very difficult to get practical consensus without some degree of theoretical consensus. If we read ‘some degree’ weakly enough and treat claims like ‘people should not φ᾽ as theoretical, then sure, we need some degree of theoretical consensus. But I’ve already given two examples where we have considerable practical consensus — we should have a democratic political system and we should prohibit the murder of young children — even though there is nothing at all like theoretical consensus about why we should do these things. Of course theoretical differences about practice will tend to yield practical disagreements, too, but it’s important to see that they often don’t yield differences that matter for political purposes, and that the prospects for practical consensus about the broad features of a political system are much, much higher than the prospects for theoretical consensus about the right philosophical account to give.


            • My skepticism arises from the thought that your analysis only looks plausible if we confine ourselves to very simple uncontroversial examples, and bracket issues regarding the counterfactual stability of the consensus in a non-homogeneous population, and also bracket how interested the relevant population is in the issue, i.e., what kind of stake they take themselves to have in it. But take somewhat more complex cases, introduce some controversy, make sure that large swatches of the population have a stake in various policy outcomes, and test for counterfactual stability of any given consensus. My hypothesis: as you vary these variables, “practical consensus” becomes paper thin and counterfactually unstable; the greater the instability, the greater the need for theoretical as well as practical consensus.

              Take, for instance, a war demanding wide participation. It will not be sufficient to get “practical consensus” on the proposition “Let’s go to war against our enemies.” That may suffice for awhile, but the longer the war lasts, and the more it demands of people, the less sufficient a merely practical consensus will be. Counterfactual stability can only be purchased through credible answers to justificatory questions. And that requires theoretical consensus. My suspicion is that most live issues are like this. The murder of young children is not a live issue for anyone. That’s why it works as an example.

              War aside, I think there are going to be lots and lots of cases where you get the illusion of practical consensus not because people agree but because it never occurred to them to think about the issue. In the absence of a theoretical consensus about the subject matter, you get the illusion of “practical consensus,” followed by one change that upsets the apple cart, and then the realization that there was no real consensus but just ferment concealed by self-deception and hypocrisy. Sexual ethics is the perfect example of this. When you lack a theoretical consensus on the aims of human sexuality, sexual norms (proscriptions) may look like a case of practical consensus. (To quote a friend of mine out of context: “I mean, you don’t date your cousin. [Pause, then weakly] Right?”) And maybe there is some practical consensus there, but the degree of it is very easy to over-estimate. All you need is one vocal minority who says, “Hey, what about…?” and within short order the “consensus” evaporates, or is revealed to be a series of empty taboos without justificatory basis.

              I guess I think that lots of things in life are like war and sex.


              • I think the real problem here is that ‘practical’ and ‘theoretical’ (and ‘consensus’) need to be elaborated. War strikes me as a case straightforwardly in my favor and not yours. When in American history has there ever been theoretical consensus about war? Never so far as I’m aware. Yet we’ve been to war, even back in the days when it had to be declared and all that. Of course often there isn’t what I would regard as consensus about it — I’m thinking of something considerably stronger than a winning vote — but my basic thought applies to those cases, too; if we think that we should stay out of war in a particular case, it matters more whether enough people agree with us about that than it does whether they agree with us about why, and it will be easier to get agreement about avoiding or going to war than it will be about why.

                Keep in mind, too, that my initial remarks were directed at a fairly high level of abstraction. The sort of theoretical consensus that I’m proposing we can do without is political philosophical consensus, broadly construed to include ordinary people’s beliefs on philosophically relevant matters, and what I’m proposing we don’t need that consensus for is for a largely stable, peaceful society.


    • I like your approach, David (in your initial response to my post). In particular, it does seem promising that the right response to lack of consensus on what first-order public norms are best and worthy of being enforced is shifting the debate to that of what procedures and institutions for determining what first-order public norms will end up being enforced will lead to outcomes that, even if not always correct, are acceptable to all (and thus tend to lead to consensus). My suggested approach of appealing to there being sufficient and obvious-enough reason for all to come to the same conclusion about the merits of a norm or its enforcement regime, seems less promising precisely because it does not squarely address the shift in relevant task from normative-truth-determination to coming to some consensus that is good enough for all (but that might yield norms and enforcement regimes that are not best or even pretty bad).

      However, because a wicked or otherwise way off-base consensus (regarding procedures and institutions for determining which norms will be enforced) will not do, this does put us in the position of saying just what acceptability standards are and how they relate to correctness standards. We might cash out acceptability standards in terms of norms or their enforcement being good-enough (though not best) or in terms of one having only less-than-definitively-justified belief that the norm or enforcement regime is not best (or even pretty far down on the list of options). However, referencing reasonableness is also pretty intuitive. If so, we are back in one of the thickets that you want to avoid. In any case, potential problems in defining reasonableness aside, I think you face questions about non-actual consensus that is acceptable but rendered non-actual by the wrong, bad sorts of objections. Why (or why not) would this yield (or not yield) permission to enforce?

      I don’t think that having these sorts of acceptability standards requires the kind of “policing” of reasons that the public-reason approach is associated with. Though I think a good political process would have to have some orientation toward consensus (and what I am calling acceptability standards) – not just correctness – especially at the level of everyday informal political debate I don’t see why banning correctness-orientation altogether would make any sense just because consensus (and acceptability) is so central. Relevant as well: the same good reasons might speak to both correctness and acceptability – and so “getting them in there” through correctness-oriented debate might help realize acceptable proposals.

      Liked by 1 person

      • I don’t think it’s a thicket that I want to avoid. If I understand you, the thicket is determining what counts as sufficiently unreasonable to be excluded, and in one way I’m all about getting into the thicket; it seems to me that the attempt to avoid it is where a public reason approach of any ambitious scope will go wrong, because there is no hope of avoiding it by identifying some non-trivial standards of reasonableness to which all relevant parties will agree, since one of the things we’ll reasonably disagree about is what is and isn’t subject to reasonable disagreement. The project I want no part of is the project of trying to say what the standards of reasonable disagreement are in some kind of abstract, principled way that is prior to actual debate. The thicket I’m happy to enter into is the thicket of arguing about what sorts of views don’t pass muster — though I also think we ought to be prepared to treat some views as so beyond the pale that we don’t even argue about them, in principle we can still do that, and in practice it will sometimes be better to do so (I think, for example, that some folks on the left these days wrongly treat certain sorts of views as so unreasonable that they aren’t worth arguing about, not because I think those views are reasonable after all, but because it seems to me that honest, intelligent people of good will who hold them or might be tempted to hold them will be more likely to moderate or change their views if met with reasoned argument rather than righteous denunciation, or at least mere righteous denunciation).

        The idea of public reason might be two things: (1) the idea that we should exclude some sorts of reasons from public debate (however we construe ‘public’) because those reasons do not pass tests of minimum rational acceptability; (2) the idea that we should exclude some sorts of reasons from public debate (however we construe ‘public’) even though they might be perfectly rational, because they are subject to reasonable disagreement. I don’t want to reject either (1) or (2) wholesale. But we can’t defend (2) by focusing on reasons of the sort excluded by (1) but not (2). Rawls, Estlund, and pals are definitely defending a version of (2); your proposals seem to be limited to a version of (1). Is that right?


        • Addendum: it sometimes seems like defenders of (2) really regard the views they exclude as falling under (1), but just want to avoid that kind of debate. Informally, I’d propose that anyone who wants to defend a version of (2) offer examples of views or reasons that they themselves accept but that would be excluded by (2) and that would not in effect be treated by their preferred version of (2) as defaults. Some people could meet that demand easily: e.g., so-and-so believes that abortion is or very likely is unjust in many cases, but defends a version of (2) on which the sorts of moral and metaphysical considerations that play into her judgment are excluded.


        • I’m not sure how relevantly “bad” objections and relevantly “bad” reasons are related. If an objection is null or disqualified (so that it, unlike qualified objections, is no obstacle to, say, generating permission to enforce some norm collectively), this is presumably in virtue of the grounds or reasons for the objection. (I’m thinking of the reasons/grounds here as psychological, but I suppose they could be factual or out-in-the-world in the sense of what reasons/facts are available to support the objection.) It is clear that null/disqualified objections imply some reasons (or sets of reasons) that are “bad” in the sense of contributing as they do to the nullifying work. But I don’t see how an objection being null in virtue of being based on reason R implies that R is not to be used (or somehow has no force) in some (or any) context of debate about what the state is permitted to do (or about anything else). I could be confused or missing something. That said, translating from forbidden reasons to null objections, yes, a specific, limited version of (1) is closer to my view (such as it is) than is (2).


          • It seems easier to show that an objection is subject to reasonable disagreement than to show that it does not rise to minimal standards of rationality. Take abortion as an example: if I object to laws permitting late-term abortions, it is easier to show that my objection (and any reasons or arguments that I offer to support it) is subject to reasonable disagreement than it is to show that it does not rise to minimal standards of rationality. Of course many supporters of abortion rights want to present opposition to abortion as though it could be based only on the sorts of considerations that do not meet minimal standards of rationality, but I for one think that’s plainly mistaken. But perhaps an example of that sort doesn’t fit well in this context because your model would treat non-legislation on that issue as the default. So perhaps income taxation for social security programs would be a better example; it is easy to show that objections to such taxation are subject to reasonable disagreement, hard to show that they do not meet minimal standards of rationality. Some progressives want to treat opposition to income taxation for social security as though it could be based only on the sorts of considerations that do not meet minimal standards of rationality, but I think that’s plainly mistaken (I think the justice of income taxation for social security is subject to reasonable disagreement even though I am firmly in favor of income taxation for social security). A view of sort (2) above would be under pressure either to declare income taxation off limits or to hold that it is not in fact subject to reasonable disagreement. A view of sort (1) would not face that pressure; it can let us argue about it and legislate about it because neither support for nor opposition to income taxation for social security fails to admit of defense by appeal to minimal standards of reasonableness.

            A view of sort (1) seems like it will be fairly minimal, though, in what it will exclude. Will it really have much bite?

            It occurs to me in reflecting on your response here that public reason views might be understood very differently in a way that I hadn’t fully considered. My way of thinking about this has been focused on what the laws require us to do or not do. But perhaps one could think that the point of standards of public reason is not, in the first instance, to limit what our laws can and cannot require, but just to limit the kinds of considerations that can be offered in favor of them; within those limits, any law or policy supported by reasons is fair game, whether it requires me to do something that I can reasonably disagree with or not. The idea here would not be that we cannot legislate on matters about which there is reasonable disagreement, but simply that we cannot appeal to certain sorts of reasons; that limitation will, in turn, rule out laws and policies that can only be supported on the basis of considerations that are excluded from public reason, but it won’t tell against laws that require me to act in ways with which I deeply disagree. On the former sort of view, laws prohibiting abortion would be out because abortion is subject to reasonable disagreement; on the latter sort of view, laws prohibiting abortion are fair game, we just can’t support them with certain sorts of arguments (ones appealing to premises from revealed religion, say), but if other rationales are available, then so be it. This latter sort of view would not point in a highly libertarian direction. Rawls and Estlund both strike me as not clearly favoring it over the former sort, but perhaps they do, or are best read as so doing. I take it that your own ideas here are not along those lines, though?

            Liked by 1 person

            • Right. My vaguely public-reason-y thought goes like this: some proposals (for what we all ought in some sense to do, to be enforced) are acceptable even if they are not (or are not known to be) correct and this implies that some objections to a proposal do not imply that it is unacceptable to the objector (or anyone else). These would be the objections that do not make sufficient appeal to strong enough reasons weighing in favor of the proposal being unacceptable (at least to the objector, but probably also to others, at least similarly-situated others).

              Writing this, it occurs to me that what matters here is not precisely rationality (e.g. minimal standards of rationality) or reasonableness in the objection, but rather the objection having the right relationship to the right reasons – those that determine acceptability. We might say something like this: a proposal is acceptable if it is good enough and no competing proposal is known to be better (or believed to be better to a high degree of confidence). That’s not going to be exactly right, but I like that ballpark. In any case, the relevant reasons, broadly speaking, would seem to be the same (good) reasons that would go into determining which proposal is best or correct. But asking ‘is that objection reasonable enough’ or ‘does that objection meet minimum standards of rationality’ is not going to get at whether an objection has the right relationship to reasons that suffice to make the proposal unacceptable (or, in Estlund’s terminology, would make the objection a qualified objection).

              None of this has much to do with what reasons to disallow in our debating which proposal we are to go with (though of course some reasons will be bad or irrelevant). Some bad reasons will be immoral, grossly insensitive to important moral concerns, etc. That might or might not imply that we should police the debate against raising or considering them (this is just a different issue). And objections that are based on particularly bad reasons (whether the “basing” relation is psychological or some relation to facts that one is, or should be, in touch with) – including immoral reasons – would not seem to be in any special category with regard to the acceptability of the proposal. Many objections will be “unqualified” simply because they do not reference sufficiently weighty reasons against the proposal being acceptable (apparently, in the typical case at least, against the proposal being good or best, sufficient to make the proposal unacceptable).

              Regarding your (1) and (2) – or how these might “translate” from ruling out reasons in public debate to making objections irrelevant to defeating acceptability. Proposals are subject or not subject to reasonable disagreement – i.e., reasonable objection. But such disagreement or objection (to proposals) themselves being subject to reasonable disagreement or objection (or not) should not be a relevant thing. The relevant thing for objections is that they are reasonable (or not) or that they meet minimal standards of rationality (or not). But, to me, this is a distinction without a difference (or at any rate without a clear difference until and unless relevant terms are cashed out explicitly). Intuitively, yes, it is easy for objections to be subject to reasonable objections and harder for objections to fail to meet minimal standards of rationality. But I don’t think the former property is supposed to render objections such that they defeat proposals. In any case, on the way I am thinking about acceptability, none of this really hits the nail on the head for determining which objections correlate with proposals failing to be acceptable to all (or which reasons or sets of reasons, in a context, make for a proposal being acceptable or unacceptable).

              I’m hoping this makes some sense. I’m aware that I’m thinking about acceptability (as against correctness) in a way that is quite different from the way that Estlund does. I’m feeling the need to go back and reread the relevant sections of Estlund to get straight on what these differences are (and make sure I’m not off-track or confused in some way).


              • I think at this point I’m confused enough about how the approach you’re trying to develop differs from Estlund’s that I might need to start over. For instance, you say: “Proposals are subject or not subject to reasonable disagreement – i.e., reasonable objection. But such disagreement or objection (to proposals) themselves being subject to reasonable disagreement or objection (or not) should not be a relevant thing.” But it would seem that it can’t fail to be a relevant thing if the first thing is a relevant thing; the problem, as I see it, is that there will be reasonable disagreement about what counts as a reasonable disagreement; if what matters is that proposals are subject or not subject to reasonable disagreement, then clearly it matters whether the objections are reasonable, and there will be reasonable disagreement about that in some cases. It’s a separate question whether that level of disagreement makes approaches like Estlund’s hopeless, but I don’t think he can deny that it will arise as soon as we try to specify some criteria for reasonable or qualified disagreement.

                You also say: “Intuitively, yes, it is easy for objections to be subject to reasonable objections and harder for objections to fail to meet minimal standards of rationality. But I don’t think the former property is supposed to render objections such that they defeat proposals.” But against that, we have Estlund saying quite explicitly: “It is a matter on which there will be reasonable disagreement, and that is fatal to the proposal to use either position in justifying political arrangements.” On its face, this is ambiguous between what we might call a results-focused view (the law or policy requires something to which there are qualified objections) and a justification-focused view (this rationale for the law or policy appeals to some claims to which there are qualified objections), but in either case we get defeat — in the former, defeat for the law or policy no matter what can be said for it; in the latter, defeat for the law or policy adopted on these grounds — and we get defeat not on the grounds that the proposal or its justification is deeply irrational, but simply on the grounds that people reasonably disagree about it. The difference between that and a view of sort (1) is that they exclude a very different range of things for different reasons.

                The question is: what are we proposing to exclude, and why? The interest of the question comes in the thought that we might exclude certain proposals without taking a stand on their truth. That would seem to push against a view of sort (1), since holding that a proposal is not minimally rational entails that it isn’t worth taking seriously as a candidate for truth. But views of sort (2) seem to face the problem that arises when we try to say what is and isn’t a reasonable disagreement.

                Liked by 1 person

  3. Because I can’t multi-task, I’m going to write this comment just responding to your post, Michael, but without reading what David’s written just above. So forgive any cross-commenting, as they say.

    First, a question: do you want your account to be general for all groups, public and private, or do you want to restrict it to the political context?

    Second, I don’t really understand conjunct (b).

    no members of G fail to reject, on the basis of ignorance of strong and evident reasons, G enforcing N.

    I think the triple negative is confusing: “no member fails on the basis of ignorance.” Maybe it should be: “every member who rejects the norm, does so in a way that expresses ignorance of the strong and evident reasons…”

    Third, I wonder whether you’re conflating an account of bad reasons for rejecting a norm with good reasons for enforcing one. Doesn’t your principle just say that in any context in which we urgently have to come to a consensus, we can enforce urgency-requiring norms on anyone who lacks a reasonable basis for dissent?

    Even if we bracket the criteria for reasonableness (or “strong and evident reasons”)–a big “if”–there are going to be contexts so urgent that in the absence of consensus, we have to enforce a norm even when someone has awesomely good reasons for dissenting. Matters might just to be too urgent for us to worry about things like dissent. Urgency is a reason for enforcing a norm over dissent; the greater the urgency, the more obvious it is that dissent should be ignored. So in some contexts at least, urgency’s being a good reason for enforcing a norm in the absence of consensus seems partly independent of the issue of “strong and evident reasons,” whether expressed or not. The reasons for enforcement are purely a function of urgency. I agree that in other contexts, there has to be sensitivity to reasons (so that your principle will apply). I guess I just mean that one principle won’t fit all contexts.

    This is a vague comment, but I’ll make it anyway. I guess I find the underlying motivation for the whole exercise excessively truth-phobic. That comes out in this formulation:

    One way to characterize this general approach is by saying that there is a “disagreement problem” specific to the relevant context that is “solved” by taking steps toward consensus (with enough truth in it) on the matter as against simply seeking the truth.

    Why the need for a contrast between “taking steps toward consensus” and “seeking the truth”? The contrast seems to presuppose that there is some deep, inherent conflict between consensus-seeking and truth-seeking such that we always face a conflict, and truth has to lose out. Why can’t we aim at consensus in such a way as to aim simultaneously at truth? Even supposing that there was a conflict between those aims, why not think that the conflict was local and context-specific, so that truth-seeking remains the larger global aim even in contexts in which it’s temporarily subordinated to the demands of consensus or accomplishing some practical task?

    Too much work here seems to be done by the word “simply,” which draws an artificially sharp contrast between practical and theoretical aims. It’s as though you were contrasting a practical aim, consensus, with a somewhat fanciful theoretical aim, that of maximizing the number of truths spoken in political discourse on an act-consequentialist conception of speech acts. I think we need a more nuanced account of how truth-seeking operates in practical contexts. The Rawlsian/Estlundian model involves too simple a picture of what that’s supposed to be, then ditches truth-seeking in favor of a truth-phobic conception of consensus. My worry is that you’re taking too much truth-phobia on board.

    To bring this down to earth a bit, I couldn’t help laughing out loud when I read this item in The New York Times. It’s a piece explaining what powers the president has to declare a national emergency to build a wall with Mexico. They trot the reader through all kinds of legal mumbo jumbo, then finally get to the crux of the matter:

    Does it matter if there is no true emergency?
    Probably not.

    Right. Because why would truth even matter? Of course, if there aren’t enough votes to override, we still get consensus without truth.


    • That NYT piece (the first one) does seem to illustrate, without quite highlighting, a tricky constitutional problem. Presumably we do need an executive with the power to declare national emergencies; if we had to wait for Congress to agree to declare one, we could lose the ability to respond to real emergencies in an appropriately timely way. So too, presumably we do not want that power to be completely unchecked; if the executive can declare an emergency for any reason whatsoever and can’t be overridden, we could end up with serious abuse of power. Yet the existing restriction on Congress — the unavailability of ‘legislative veto’ — seems well grounded not only in constitutional law (as I understand it, the deepest reason why the court declared the legislative veto unconstitutional is that it violated the Presentment Clause, and it seems to do that) but more fundamentally in the principle of checks and balances. After all, it’s not like the executive’s power is unchecked; Congress can override the president’s veto in the same way it can override any presidential veto. But clearly in the present arrangement, it turns out not to matter whether there’s really an emergency. What matters is whether 2/3 of Congress agrees that it isn’t really an emergency and that the declaration of emergency should be overridden — which, given the realities of partisanship, is not going to happen. This seems like an unhappy result, but all the steps that led to it seem acceptable and maybe even conclusively reasonable.

      I’m really trying to avoid ending up with the bland classical conclusion that nothing can prevent wicked and stupid people from making a complete mess of things if they get into office, but that’s about the best I can do at the moment.

      Liked by 1 person

      • I have trouble reading articles like that online, and don’t have access to a printer at home, so I can’t claim to have read Goldsmith with all due care. But all of what he says in the first half strikes me as an elaborate missing of the point. And even what he says in the second half seems inessential to the real issue.

        The real problem is that we live in a constitutional republic that has exactly the same structural problem as the Weimar Republic: it gives the executive emergency powers but fails to define “emergency.” That one fact strikes me as the single most dire problem we face. Goldsmith’s tedious belaboring of how past presidents have invoked these similarly elastic and undefined emergency powers doesn’t address that, and neither do his criticisms of the special features of Trump’s invocation.

        What this means is that the following obvious question has no legal standing whatsoever in our law: for any x declared to be a national emergency, is it really one? I could direct your attention away from this problem by telling you that the Governor of New Jersey declared a statewide emergency last week in the face of two or three inches of snow, but that seems a case of diversion rather than an attempt to look the real problem squarely in the face. The real problem is that we’re dealing with a president engaged in yet one more Orwellian misuse of language while lawyers deftly point out that, well, everybody does it.

        We’ve gotten to the point where I think most people think that there is no fact of the matter about whether something is an emergency. We can, evidently, use the word “emergency” any way we want to. So if we want to call a bunch of Guatemalans trying to cross our southern border an “emergency,” well then, it is one. And if we want to call it an “invasion,” it is one. And if our president claims that our “national security” is threatened, then it must be.

        We need an executive to declare national emergencies, to be sure, but an executive constrained by the rule of law ought to have to declare one on the basis of an explicit, written definition of “emergency.” Police officers have to make executive decisions under much tighter constraints than those imposed by a slow-moving pseudo-invasion, but what they’re executing laws where the most crucial terminology is explicitly defined. A police officer cannot decide–on the fly and pandering to the whims of public opinion–that as of Monday, February 18 at 5:20 pm, the elements of armed robbery or aggravated assault will be x,y, and z–and then just arrest anyone who satisfies those arbitrarily-dreamed-up elements. It wouldn’t help to discover that we lived in a regime in which his predecessors did the same thing, or did a similar thing. Evidently, that’s the kind of regime we live in–a populace in desperate search for a Fuhrer, and a Fuhrer in desperate search for his Reichstag fire. We’re all just sitting here, waiting for all of the pieces to fall into place.


        • I agree that Goldsmith’s appeal to previous presidential uses of emergency powers is no defense of those uses. I think, though, that he’s responding to the suggestion that Trump’s use here constitutes some kind of constitutional crisis. In response to that worry, it is indeed relevant to note that previous presidents have done this without toppling our constitution. That’s a limited point, to be sure, but it’s a response to hyperbolic worries that are really out there.

          I don’t think it’s quite true that the president’s emergency powers are undefined. On the contrary, whenever the President declares an emergency, he has to specify the provisions of law under which he will act; it’s true, apparently, that it’s an emergency if he declares it to be, but what he can do in light of the emergency is pretty heavily constrained, and it seems likely that some of the legal challenges that he’ll face will come in the form of arguments that the statutes he’s appealing to don’t in fact support what he’s proposing to do.

          Trump’s lawyers are, for instance, relying primarily on the following section of the U.S. code:

          “[i]n the event of a declaration of war or the declaration by the President of a national emergency in accordance with the National Emergencies Act (50 U.S.C. 1601 et seq.) that requires use of the armed forces, the Secretary of Defense, without regard to any other provision of law, may undertake military construction projects, and may authorize the Secretaries of the military departments to undertake military construction projects, not otherwise authorized by law that are necessary to support such use of the armed forces.

          It is, to say the least, not obvious that the ’emergency’ requires the use of the armed forces or that the wall would be a military construction project at all, let alone one required to support the use of the armed forces. In any case, while he can declare the emergency without having to justify his claim that there is an emergency, the powers he gets via declaring the emergency are constrained.

          So I’m not convinced that there is a dire constitutional problem here. Nor am I bothered by ’emergency’ taking on a specialized legal meaning distinct from its ordinary language use — an emergency is any circumstance deemed by the executive to warrant such-and-such extraordinary powers — as such; that’s a pervasive feature of law. I’m not sure it’s quite right to say that the word can be used any way we want, either. The substantive question is whether it ought to be given a definition rather than left to the executive’s discretion. To really assess that, though, we need candidate definitions; what definition would not be unduly narrow, potentially constraining efficient and timely executive action, and yet not so broad as to effectively allow anything to count as an emergency?

          I’m not a legal scholar, so I have no business forming a strong view about whether there should be a definition and, if so, what it should be, prior to doing considerably more legal research than I’ve done. But it certainly does seem advisable to have a definition if we can get one that does all and only the work it’s needed to do. That said, I think I’m with Goldsmith insofar as his idea is that Trump is not really threatening to undermine our institutions, but is doing considerable damage to informal norms.

          Liked by 1 person

          • The point I’m making is not primarily one about constitutionality. My point is that regardless of the legalities, it is extremely dangerous for a country to grant emergency powers to an executive but leave the concept of “emergency” either unspecified or under-specified. The entire discussion of Trump’s invocation of emergency powers seems to me to involve a series of red herrings that skirt the real issue.

            The real issue is: if emergency powers are not legally, bindingly constrained by a plausible definition of “emergency,” they are excessive and easily liable to immense abuse. At a bare minimum, it has to be recognized that emergencies are temporary conditions, that “temporary” has to be defined, and that the exigencies involved in an emergency have to be genuinely extraordinary, where “extraordinary” has to mean something more than that “It is likely that more-than-the-average number of people will die as a result of a given condition.” Even a 3″ snowfall meets all but the last criterion–temporary, exigent, even “extraordinary” in a given winter–but a snowfall is still absolutely not an emergency by any semantically plausible conception of what an emergency really is. In New Jersey, every snowfall has become an “emergency,” and the governor has emergency powers to control people’s movement and in principle to declare a curfew during one. The fundamental problem is the decades-long refusal to define the concept of “emergency” while according the executive immense power to act in “one.” All of these other back-door methods to “constrain” executive authority strike me both as evasions and as recipes for failure.

            You say that Trump is pretty “heavily constrained,” but if that’s so, how is it that he’s declared a national emergency? If he was genuinely constrained, there would be a constraint on that very action. But there isn’t: that’s why he’s acted, and that’s we face the problem we face. And of course, the deeper issue is not whether Trump is constrained, but whether the presidency is. If the U.S. presidency was genuinely constrained, what should we make of the whole first half of Goldsmith’s article, which gives us one example after another of the successful use of emergency powers by a string of presidents going back a few decades? Surely the relevant question in each of those cases is: were they all really emergencies? I would say: almost none of them were. Regardless, at this point, all of those decisions are taken for granted as legitimate exercises of executive power in genuine emergencies. That doesn’t strike me as a constrained executive. It looks more like an executive run wild.

            That the President faces legal challenges doesn’t mean that he’s “heavily constrained.” For one thing, he may end up prevailing in court. Some experts regard him as having a good chance. And why wouldn’t he prevail, if all he was doing was contesting “informal norms”? The Court has no jurisdiction over informal norms.

            But whether he prevails in court or not, he can act essentially without constraint until the final judicial decision is made–which may be a long time. If he was really “heavily constrained,” he would face a legal situation such that he wouldn’t think of undertaking his current course of action: his lawyers would have told him not to proceed because the law was solidly against him. He’s proceeded as he has precisely because the law is not entirely against him. If “emergency” were properly defined in this context (as it is, not perfectly but to a greater extent in ordinary law) he would be constrained in the genuine sense of being literally unable to declare a national emergency in the first place.


            Nor is Trump’s claim of emergency power outlandish—at least by the standards of past presidential practice.

            The relevant issue is not whether Trump’s claim is outlandish “at least by the standards of past presidential practice.” By Goldsmith’s own account, the essential feature of “past presidential practice” is the absence of standards involved. What makes Trump’s claim outlandish is his invocation of emergency powers in the absence of anything that qualifies as a genuine emergency. The same charge of course could be made of Bush, Clinton, Obama, and so on. So Goldsmith’s argument here is a tu quoque followed by a diversion: everybody does it, so let’s do an end-run around the real issue, and focus on other things.

            Consider the logic of this argument:

            Many charge that Trump is declaring an emergency when there is no emergency. But this begs all the relevant questions. The relevant statute on which Trump relies does not define the term “emergency.” Presidents have always—really, always—had discretion to decide if there’s an emergency. And presidents have often declared emergencies under circumstances short of necessity, to address a problem that does not rise to an “emergency” as defined in common parlance to mean “a serious, unexpected, and often dangerous situation requiring immediate action.”

            This is a misconstrual of all of the relevant issues. I charge that Trump is declaring an emergency where there is none. An emergency, I would say, has to involve extraordinary danger under temporary circumstances. The “invasion” in question meets neither criterion–it’s not that dangerous, and obviously not temporary.

            What question have I “begged”?* Maybe I need to justify my account of the semantics of an emergency. Fair enough. But it’s surely plausible to think that the paradigm examples of emergencies have the features I’ve mentioned. If you call 911 for anything short of that, you may well be charged with a crime, but whether you are or not, the use of 911 for non-emergencies is widely recognized as a serious problem. That’s what led to the development of 311, 511, etc. How is it that lawyers and politicians are permitted such semantic latitude? In one context, emergencies have one set of features and are legally actionable on the basis of those features, but if you switch contexts, you’re “begging the question” if you suggest that they have the same features?

            That Presidents have always had “discretion” to declare emergencies is exactly my complaint. That they have misused the term to cover things that aren’t emergencies even by the loosest colloquial understanding is a bug, not a feature. But the real problem here is a kind of fanatical positivism that regards it as uncontroversial that if a term isn’t yet defined in a statute, it’s “no big deal” to give someone the power to give his own private meaning to the term, and then employ the might of the armed forces to enact his will. But that is literally what Goldsmith is saying. The question he fails to address is: why is it that the executive needs emergency powers that are unconstrained by a definition of what an emergency is? Yes, there may be other constraints, but what justifies the absence of that one? If we agree that the executive must be constrained, and we agree in other contexts that it is highly problematic to conflate emergencies with non-emergencies, why should the executive not be constrained in the use of its emergency powers by a definition of “emergency”?

            You quote the part of the U.S. Code on which Trump is relying, and say

            It is, to say the least, not obvious that the ’emergency’ requires the use of the armed forces or that the wall would be a military construction project at all, let alone one required to support the use of the armed forces. In any case, while he can declare the emergency without having to justify his claim that there is an emergency, the powers he gets via declaring the emergency are constrained.

            Nothing in the statute requires that anything be “obvious.” The whole point of a grant of discretion to the executive is to permit opaque reasoning involving considerations that are totally non-obvious. Goldsmith has already told us that the modal-looking terms (required, have to…) don’t have to be taken literally, either. So the permitted latitude accorded the president is vast. It’s not obvious that the work of the Army Corps of Engineers has to be done by the army (i.e., the armed forces), or that the building of dams or the “provision of outdoor recreation” requires the use of the armed forces, and yet all of this is taken for granted in American law. But if the Army Corps of Engineers is “the Nation’s environmental engineer,” and the Wall requires environmental engineering, and the Army’s construction projects have to meet certain security requirements (the engineers have to be protected by armed troops against threatening invaders), and we are facing a foreign “invasion” (the President has decided that an uninvited influx of foreigners is an invasion, and that an invasion of foreigners is a foreign invasion), then once we grant the existence of a national emergency, Trump gets everything he wants without trying. The powers he has aren’t constrained in any substantive way; they’re essentially unconstrained. Yes, eventually he will face the courts and have to answer for some of this, but until then, it’s smooth sailing. And it might be smooth sailing afterwards, as well.

            I don’t think we need an actual candidate definition of “emergency” in hand to realize that having a serviceable definition is better than having no definition. And I’m the last person to object to “emergency” taking on a legal meaning distinct from the ordinary language usage; the ordinary language usage is probably too imprecise for legal use, and incapable of constraining the kind of lunatic we currently have in office. My point is precisely that we need a legal definition of emergency suitable for national security purposes, and that our lacking a legally explicit one is a serious political problem. It may not be a constitutional problem, but that’s beside the point: constitutional problems don’t exhaust the universe of serious problems. It’s telling that the question you ask at the end of your penultimate paragraph is not taken seriously by legal scholars. Raise it with them, and they’ll simply change the subject. It’s one task to answer that question, and another to argue that that question should frame the debate. So far, I’ve just been arguing for the latter.
            *This thought occurred to me later, but isn’t it the responsibility of the party making a claim to defend it, not his critics? If Trump asserts that the situation on our southern border is an “emergency,” he’s the one who bears the burden of proof for arguing that it is. So it seems to me illegitimate of Goldsmith to ignore the fact that Trump hasn’t met that burden (so that Trump is the one begging all the relevant questions), and then construe the critics’ question, “What emergency?” as an assertion involving a “begged question” against Trump. In the following conversation, it’s Smith, not Jones, who is begging the question:

            Smith: This is an emergency!
            Jones: I don’t see it. How?
            Smith: Well, it is!
            Jones: Come on. It doesn’t satisfy any plausible definition of emergency.
            Smith: Prove it! You’re begging the question!

            Smith is employing a clever and common tactic for evading his actual burden of proof: he begs the question, then half engages in an argument from ignorance to demand that Jones satisfy his, Smith’s, burden of proof as the price for posing an objection. Even if we construe Jones’s “come on” statement as a positive statement requiring justification, the fact remains that in this dialectical context, it’s Smith who owes the account of “this” as being an emergency to Jones, not Jones who owes Smith a definition of “emergency,” followed by a justification of that definition. It’s not legitimate to assert that p and then demand that your interlocutor give a justification of ~p as a preliminary to a discussion of the merits of p.


            • “You say that Trump is pretty “heavily constrained,” but if that’s so, how is it that he’s declared a national emergency?”

              What I wrote was: “it’s true, apparently, that it’s an emergency if he declares it to be, but what he can do in light of the emergency is pretty heavily constrained…while he can declare the emergency without having to justify his claim that there is an emergency, the powers he gets via declaring the emergency are constrained.”

              Your response proceeds as though by declaring an emergency the executive can just do whatever he claims is necessary in light of the undefined emergency. In fact what he can do is what some particular statutes say he can do. In this case, what the statute says he can do is order some military construction projects to support the armed forces. On its face, it would seem that if he tries to order some other projects that aren’t covered by the statutes, he can be legally challenged and, in principle, prevented. Perhaps I’m misreading the statute when I take it that the wall would not be a construction project that supports the use of the armed forces. Even if the order is not open to challenge on that front, there is a significant constraint here, even if it’s not the constraint that we should want. To compare this to the emergency powers granted to Hitler seems a bit hyperbolic.

              I agree that it would be better to have a serviceable definition of ’emergency.’ It seems like it shouldn’t be a tall order to come up with one that doesn’t have either of the flaws that I identified, but I’ve never written a law and don’t know all the details, so I’m not prepared to offer one.


    • In reply to your substantive response to my post, Irfan…

      (1) The context is meant to be general, but public in a certain sense. It is meant to apply to moral communities as well as political communities, to informal as well as formal legal enforcement.
      (2) Yeah, the second conjunct is messed up. David’s second interpretation is the one that I had in mind. The idea is supposed to be that there are conditions (of being out of touch with relevant reasons) that render consensus about a norm being a good one to have adequate general compliance with null (such that the consensus does not generate permission to collectively coercively enforce). This is analogous to consent being nullified if there are threats, inadequate information, etc.
      (3) The issue is permission to enforce, not reasons to enforce. We can get this, on my principle, if there is no consensus (about the norm and its enforcement regime being good, best, useful, etc.) when the dissent (or lack of consensus) is unreasonable in the “objective” way that I suggest, with obvious, strong reasons that would definitely favor the view in question being ignored or somehow culpably not registered. This is analogous to non-consent being nullified (or rendered to be the same as non-nullified consent). And also analogous to certain kinds of implicit consent that are actually weak forms of hypothetical consent (you would agree, you just don’t realize this thing about the situation that you for some reason don’t have good access to).
      (4) Yes, extreme urgency could get this normative result. But I’m not thinking that it is the urgency that does the work all on its own. (Maybe what is more important is that we are all stuck together and have to decide on some norms of obligation and permission and some enforcement regime for this.)
      (5) I agree that truth-seeking and consensus-seeking are not in conflict in the way that the standard public-reason approach would have it. The better way of putting my motivation is this: when we cannot (in this context) agree on which norms (or norms + enforcement regimes) are best, we will only be able to agree on norms (or norms + enforcement regimes) that are acceptable. I don’t think acceptability (as against correctness) should be construed in a truth-phobic way. As suggested in my response to David, a norm (or norm + enforcement regime) might be acceptable to some group member X in virtue of X regarding that norm as not-best but good enough or in virtue of regarding it as not-best but not with a lot of confidence. On this sort of view, retreating to acceptability might work only because we all share a common perspective and truth-orientation (even though circumstances are not propitious for coming to consensus about which norms are best).


  4. Here’s a live example, from Israel, of the issue Michael was talking about in his original post:

    Kach was excluded from the Knesset in 1988 (and generally, from Israeli political life, at least within electoral politics and within Israel proper) on public-reason-like grounds. Their policy proposals were thought to express racist incitement, which was treated as a non-qualified view in Estlund’s sense.

    I wonder whether public-reason-like views have somewhat greater plausibility in parliamentary systems that lack written constitutions, but where specific parties are associated in a very specific and determinate way with particular rights violative policies, e.g., Kach with deportation, segregation, and anti-miscegenation. The rationale for excluding a given party is to provide the functional equivalent of a Bill of Rights in the absence of one: precisely because Israel lacks an American-style Bill of Rights, it has to safeguard rights by policing access to the electoral system, lest a rights-violative party win office, and pass the relevant policies. Obviously, this gambit can either fail or be abused–fail when a given majority refuses to exercise the exclusion option, or be abused in a case where a benign coalition is excluded on ideologically tendentious grounds.

    In any case, the exclusion of Kach seems to fit Michael’s principle, where “G” is Israelis, and “N” is “excluding Kach and affiliated parties.” If we assume that Kach is racist, and assume that there are no strong and evident reasons in favor of any racist policy (of the Kach-type), or any objections to exclusion that aren’t racist, then Israelis can exclude Kach.

    Having said all that, I think it’s equally clear that someone could in principle split hairs on all of those points: they could challenge the judgment that Kach was racist; they could challenge the claim that all the reasons against racism of the Kach variety are strong and evident; they could insist that the reasons for including Kach do not flout any strong and evident reasons.

    Once you look at the public reason issue while focusing on specific issues, I think it starts to seem as though invocation of “public reason” has become a substitute for the use of informal logic. The relevant logical issue concerns burdens of proof. Public reason seems problematic in part because it systematically seems to shift the burden of proof away from defenders of public reason and onto everyone who rejects the strictures of public reason. I wonder if critics of public reason might do well to produce worked-out analyses of how the norms of informal logic apply to recurring dialectical situations in political discourse (as in the work of Douglas Walton and others).

    Liked by 1 person

    • For this value of ‘N’, the relevant thing would be that, if Kach or others like him were to object on racist grounds (but all others were to agree), we would have nullified non-consensus – which has the same normative upshot as non-nullified consensus. We could permissibly enforce N. But the same would be true for any other value for ‘N’. The excluding of individual or reasons from actual debates, legislative bodies, etc. seems different from the issue of which grounds for objection render an actual objection (or rejection) otiose so that it could be as if you had (non-nullified) consensus about what N should be (so that N could be permissibly collectively enforced). However, racist reasons (or someone who mainly offers racist reasons) might sensibly be excluded from the process of deciding about what our collectively enforced N’s will be – that would be consistent with our coming to consensus on the value of any given N in light of good moral and practical reasons.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s