Summary of chapter 4
The goal of chapters 3 and 4 is to explain that the “Public Justification Principle” (PJP) does not entail the ‘principle of restraint.’ This involves showing that there is no necessary relation between the PJP and an ‘accessibility’ or ‘shareability’ requirement on justificatory reasons. Chapter 3 identifies two desiderata for evaluating different conceptions of justificatory reasons: (1) respect for personal integrity, and (2) respect for the fact of reasonable pluralism. Chapter 4 argues that the convergence account of public reason, which does not include the principle of restraint, fulfills these desiderata more successfully than rival consensus accounts.
Vallier begins by distinguishing between three increasingly stringent requirements on justificatory reasons: intelligibility, accessibility, and shareability. Here is his statement of each:
“Intelligibility Requirement: A’s reason RA can figure into a justification for (or rejection of) a coercive law L only if it is intelligible to all members of the public.” (P.106.)
“Accessibility Requirement: A’s reason RA can figure into a justification for (or rejection of) a coercive law L only if it is accessible to all members of the public.” (P.108.)
“Shareability Requirement: A’s reason RA can figure into a justification for (or rejection of) a coercive law L only if it is shared by all members of the public.” (P.110.)
Vallier also mentions a “Symmetry requirement”: “The same evaluative standards must hold for reasons to propose law L as the reasons to reject law L” (p.111). He claims that this requirement is presupposed by most advocates of consensus accounts of public reason (though he also notes, “so far as I can tell, no one defends it” (p.111)).
In section II, Vallier argues against the accessibility requirement. There are two main arguments against this requirement: (1) it fails to achieve its own aims, and (2) it is “excessively restrictive of integrity and diversity and so should be rejected” (p.112). According to (1), the accessibility requirement cannot rule out all religious reasons from public justification, since some religious arguments are accessible (i.e., rely upon common evaluative standards of reasoning and evidence). Vallier discusses two kinds of accessible religious reasons: those of natural theology, and those drawn from religious testimony. Argument (2) is that the accessibility requirement, by denying justificatory force to some intelligible reasons, fails to respect integrity and diversity (reasonable pluralism) as much as the intelligibility requirement.
Section III focuses on the shareability requirement. Vallier argues against the main reasons that have been advanced in support of this requirement. First, he claims that the principle of respect for persons does not obviously require shareability. Here he refers back to his argument in chapter 2 that this claim concerning the nature of the principle of respect for persons presupposes the shareability requirement, rather than justifies it. Second, he contends that the claim that reasons are ‘inherently shareable’ is either false or trivial. Third, he asserts that publicity does not require shareability. (I should note, in connection to Danny Shahar’s comments on chapter 2, that Vallier’s comments here turn on the claim that publicity is important for stability, a claim that Vallier takes to have refuted earlier. However, I do not think that the most compelling justification for publicity is stability; rather, I interpret it as resting primarily on an ideal of shared political autonomy. Given how brief Vallier’s comments are on this here, though, I will not pursue this issue.)
The most substantive discussion in this section concerns the idea that public reasons must satisfy a ‘sincerity requirement.’ (Such a requirement is part of Rawls’s account of public reason, and has been defended in recent years by Micah Schwartzman and Jonathan Quong.) “Public reason liberals,” Vallier notes, “often hold that justificatory reasons must be offered in a sincere fashion” (p. 123). A sincerity requirement may spell trouble for a convergence account of public reason because it seems insincere for citizen One to offer argument X for law L to citizen Two that One herself does not believe (as she rejects the comprehensive doctrine that underlies argument X, but advances argument X only because she knows that citizen Two endorses the comprehensive doctrine in question, and she wants Two to support law L for her own reasons). Such cases of insincerity, it would seem, are avoided if one insists on the shareability requirement for public reasons. However, Vallier argues that “most sincerity requirements decompose into a reasons requirement,” and thus “sincerity discussions do not mark off a distinct domain of normative concern” (p. 124). The argument seems to be something like this: we care about sincerity because we are committed to a principle of respect for persons. This principle also underpins the PJP. What satisfying the PJP involves depends on our account of what is required of persons when they justify coercive laws to others. And here we simply have different accounts of what such justifications require, namely, the intelligibility, accessibility, and shareability requirements. So, “if only intelligibility holds, then what it means to be sincere is to offer arguments for proposals that one believes are justified according to the evaluative standards of one’s discussant and one’s own evaluative standards” (p.124).
In section IV Vallier makes a positive case for the convergence account. He notes that the convergence account will still prevent religious citizens from imposing their preferred laws on others unless those laws can somehow satisfy the PJP, by being ultimately justifiable to others in terms that they can accept. Nonetheless, such restrictions are less severe than those imposed by the consensus accounts, and impose no normative restrictions on the voting behaviour of average citizens (p.127).
Vallier also revisits the distinction (introduced in chapter 2) between ‘indirect’ and ‘direct’ approaches to public justification: “a direct approach to restraint requires that citizens themselves restrain their activities, whereas an indirect approach relies more on institutional design to prevent private reasons from generating publicly unjustified law” (p.128). Interestingly, he notes that consensus views can overcome the integrity objection if they adopt an indirect approach. Average citizens could ‘vote their consciences,’ so to speak, and thus retain their integrity. But an accessibility or shareability requirement on the justifications for laws nonetheless could be satisfied via institutional mechanisms and the behaviour of public officials. Only the right kinds of justifications would survive the process of shaping laws. Since Vallier concedes that a principle of restraint does apply to legislators, judges, and other public officials directly involved in the shaping, interpretation, and implementation of laws (pp. 127-128), it seems that the integrity objection can be overcome by all accounts of public reason so long as an indirect approach to public justification is employed. In reply, Vallier holds that the convergence view remains superior as it avoids the integrity objection on both the direct and indirect approaches to public justification, whereas consensus views avoid the integrity objection only by adopting the indirect approach (p. 129).
In section V, Vallier addresses the objection that the convergence account is too ‘permissive’ since its version of the PJP is incapable, by itself, of picking out which laws among the eligible set are the most justified (the ones that society should adopt). We end up with a surfeit of possible laws and acceptable justifications (the set is “stuffed”). He interprets Rawls’s original position as a way of dealing with this problem: the original position enables us to identify a single ‘best’ conception of justice (which, in turn, can guide us in shaping society’s ‘constitutional essentials’ and ‘matters of basic justice’). Vallier’s alternative strategy consists of two steps: “First, members of the public list undefeated but eligible principles. Second, they select a decision procedure to choose a proposal from the set.” (P.131.) So long as the decision procedure itself satisfies the PJP, the problem of too many potential laws is overcome.
The flip-side of this worry is that the convergence account is too restrictive, as its interpretation of the PJP would allow “a dysfunctionally small number of laws to be justified” (p. 130). In reply to this worry, Vallier notes that in many cases allowing dissenters to laws to be exempt from them (a strategy of accommodation) can enable laws to be enacted rather than defeated in the face of deeply held opposition on the part of some citizens. More generally, Vallier claims that convergence, given its openness to citizens’ diverse reasons, might add proposals to the “eligible set,” rather than remove them.
Finally, in section VI a recent argument in favour of the consensus account of public reason advanced by Stephen Macedo is addressed. I will not delve into this exchange except to highlight two things that Vallier says in reply to Macedo that I found interesting. First, he tries to address the worry that the convergence view itself is partisan on questions of justice, as it is “simply too libertarian” (p.137), that is, the objection that it inherently ‘tilts’ in a classical liberal or libertarian direction (as Gerald Gaus argues). Consequently, the convergence view, despite its claims to respect persons and their integrity, neglects the interests and life-prospects of the least advantaged. In reply, Vallier notes that libertarian regimes require the coercive enforcement of (a certain interpretation of) private property rights, and thus must satisfy the PJP. Consequently, non-libertarians could ‘defeat’ such justifications for coercion (p.137). Second, with respect to what coercive measures can satisfy the PJP, much depends on how “fine-grained [the] individuation of coercive actions” is (p. 138). If each individual law must satisfy the PJP, then “perhaps more opportunities for defeat will present themselves” (p. 138). In contrast, if, say, complete conceptions of justice (applying to constitutional essentials and matters of basic justice) are to be justified, then the scope for limiting legislative action because of vetoes from members of the public will be much smaller (the entire ‘package’ is voted up or down, so to speak). Vallier correctly notes that the issue of individuation is independent of the differences between the consensus and convergence accounts of public reason (p.138), that is, convergence accounts are not necessarily committed to the fine-grained individuation of coercive actions.
As my summary indicates, this is a very rich and dense (in a good way!) chapter. I found Vallier’s discussion consistently stimulating. I’ll limit my comments to 5, even though there is much more that I could say.
1. The sincerity requirement
I’m afraid that I was not satisfied with Vallier’s treatment of the sincerity requirement (pp. 123-124). I don’t think that this requirement simply collapses into a ‘reasons requirement.’ Rather, I think that it is an independent normative requirement. In particular, I do not think that simply satisfying the intelligibility requirement for reasons suffices for sincerity.
Here is a scenario that might illustrate why I was dissatisfied with Vallier’s reply to Quong. Imagine that Yvonne endorses law L1 because of Ry (a reason found in Yvonne’s comprehensive doctrine). She knows that Zachary adheres to a different comprehensive doctrine, and thus Ry will not convince him to support L1. But Yvonne wants Zachary to support law L1 nonetheless. So Yvonne provides Zachary with an argument, Rz, that she thinks (correctly, let’s assume) is drawn from Zachary’s comprehensive doctrine, and which supports L1. But Yvonne also knows that Rz, while supportive of L1, in fact better supports a rival law, L2. However, since she wants L1 and not L2, she refrains from mentioning this fact to Zachary. It seems that Yvonne is being insincere. She is not treating Zachary with respect (or at least with inadequate respect; she is still providing him with a justification that he can accept). Yet she also seems to be satisfying the intelligibility version of the PJP.
2. The principle of restraint: citizens versus public officials
This issue may very well be dealt with by Vallier later (in chapter 6), but I would like to bring it up now. Vallier concedes that a principle of restraint does apply to political officials: “Political officials are not permitted to use coercion to further their sectarian, religious ends in the face of defeater reasons affirmed by members of the public.” In contrast, “restraint does not apply to citizens in their capacity as voters” (p.127). One reason for this seems to be, I think, that political officials choose to pursue their positions, with the responsibilities that they entail, and so cannot later object on grounds of integrity that they have to exercise restraint in justifying their (coercion-shaping) political decisions. Citizens, in contrast, are subject to the laws of their society irrespective of their career choices. Another reason is that public officials directly participate in the formation, implementation, and interpretation of coercive laws, whereas average citizens normally do not.
Nonetheless, there is something puzzling about this distinction. If public officials fail to respect the principle of restraint — if, say, legislators pass laws based upon their particular religious beliefs despite the existence of defeater reasons — then do voters not have some kind of duty to hold them accountable for this failure? I would be inclined to think so. Voters should ‘punish’ (electorally) their representatives for failing to live up to their professional responsibilities. (This is, of course, Rawls’s view. He holds that the duty of civility applies primarily to public officials acting within the ‘public political forum,’ and that citizens fulfill their duty of civility by holding their officials to account, including especially when voting.) In short, I’m sceptical that if public officials have a duty to comply with a principle of restraint this duty has no normative implications for citizens in general.
A related question that I have concerns instances when all voting citizens are legislators. Laws sometimes are decided by referenda (in some societies, such as Switzerland, this is quite common). When a law is decided in this way, by the citizens themselves through their vote, is Vallier’s view that they should be constrained by a principle of restraint? That is, does everyone become a public official? If this is the case, would this constitute a reason for not employing referenda in law-making, given the burdens with respect to integrity that this would impose on voting citizens? Or are citizens nonetheless free to vote however they like even in referenda on laws? (This obviously is more a question of clarification than an objection.)
3. The Discussion of Rawls’s Original Position
Vallier presents the original position as (at least in part) an attempt by Rawls to identify which conception of justice is the most justified. The original position thus enables us to overcome the problem of indeterminacy, given the wide range of available conceptions of justice by identifying which one is the best. (Pp.130-131.)
Vallier is indeed correct that the role of the original position is to demonstrate the justificatory superiority of one conception of justice, ‘justice as fairness,’ over other conceptions. But the role of the original position is not to serve as a kind of ‘decision procedure’ for citizens. Rather, it is a ‘philosophical device,’ one way for reasonable citizens to try to satisfy the ‘criterion of reciprocity’ in their relations with others, and thereby identify the most reasonable conception of justice. “[E]ach of us must have principles and guidelines to which we appeal in such a way that the criterion of reciprocity is satisfied,” Rawls writes. “I have proposed that one way to identify those political principles and guidelines is to show that they would be agreed to in … the original position.” (Rawls 2005, pp.xlviii-xlix.) Once we have this conception of justice in hand, though, citizens still have to participate in the political decision-making processes of their society. That is, they still have to make public reason arguments in favour of the laws and policies that they take justice as fairness to support or require, and they still have to vote, run for office, and so forth, in order to implement those laws and policies. “[D]emocratic decisions and laws are legitimate, not because they are just but because they are legitimately enacted in accordance with an accepted legitimate democratic procedure” (Rawls (2005), p.428).
The original position, in other words, is a philosophical device by means of which citizens can explain (rigorously) why they are committed to the conception of justice as fairness, and thus endorse the political proposals that they think follow from that conception. When citizens defend their political positions in this way, they do so as free and equal members of the public, not as privileged decision-makers. “In justice as fairness there are no philosophical experts,” Rawls writes, “Heaven forbid!” This is not to say, of course, that philosophers cannot contribute to the public political culture of democratic society. “Citizens must, after all, have some ideas of right and justice in their thought and some basis for their reasoning. And students of philosophy take part in formulating these ideas but always as citizens among others.” (Rawls (2005), p. 427.) So the role of the original position is to help citizens identify the most reasonable conception of justice, by means of which they subsequently can ‘orient’ their political activity. We still need publicly justified decision procedures (democratic institutions), though, to decide the laws themselves.
Another concern that I had with Vallier’s discussion here has to do with his characterization of the ‘second stage’ of political justification, what Rawls calls “full justification.” Within full justification, citizens determine how the conception of justice as fairness ‘fits’ within their respective comprehensive doctrines (justice as fairness is a ‘module’ that can be ‘embedded’ within different comprehensive doctrines). Through full justification citizens achieve “reflective equilibrium.”
Vallier claims that “[f]ull justification is a convergence conception” (p. 131). Yet one aspect of Vallier’s account of a convergence justification seems different from Rawls’s method of reflective equilibrium. For Vallier (as I read him), political proposals must be acceptable to (reasonable, moderately idealized) members of the public, given their comprehensive doctrines (religious views, etc.) as they are. Citizens’ broader systems of values and beliefs are ‘given,’ so to speak, and political proposals must adapt themselves to them. There is no adjustment to be done on the part of the comprehensive doctrines themselves. (Or at least I read Vallier’s account this way, given what he writes about citizens’ “belief-value sets” and their relation of political justifications in chapter 1.)
However, one feature of Rawls’s method of reflective equilibrium is that achieving equilibrium may require changes in citizens’ comprehensive doctrines, that is, adjusting elements of citizens’ broader values to cohere with what they take the most justified conception of justice to be. There obviously are limits to what kinds of changes can be expected of citizens, but Rawls mentions that many elements of comprehensive doctrines are open to different interpretations, and are silent, vague, or internally conflicted with respect to some questions of basic political justice. Thus there is room for adjustment within (many) comprehensive doctrines to justice. An overlapping consensus (a state of full public justification), then, is something to be achieved, on Rawls’s view, and this achievement can involve changes within comprehensive doctrines, not simply finding political principles (or proposals) that are acceptable to citizens’ comprehensive doctrines as they exist pre-political deliberation.
4. Rough-grained individuation: from convergence to consensus public justification?
Vallier mentions briefly that what coercive measures can survive the PJP depends much on how “find-grained [the] individuation of coercive measures” is (p.138). If each individual law must satisfy the PJP, then “perhaps more opportunities for defeat will present themselves” (p.138). If a number of laws are justified as a ‘package deal,’ though, citizens can defeat the proposals only by finding the overall package unacceptable.
I wonder if there is not an argument to be made for a very ‘rough-grained’ individuation of coercive measures, namely, that what is to be justified via the PJP is a conception of justice that is to apply to society’s ‘constitutional essentials’ and ‘matters of basic justice,‘ understood as an overall, interdependent system of law. This is, of course, essentially Rawls’s view: the entire conception of justice is to satisfy the PJP. So if we insist on a very rough-grained individuation of coercive measures, we end up with (what Vallier calls) Rawls’s “consensus-convergence hybrid view of public reason” (p.130).
One reason why such a view might be attractive is that we might be concerned with the overall coherence of our society’s laws and institutions. That is, if we regard the basic structure as a unified system of social cooperation, we naturally would want it to make sense as a system. Among other things, we would want to avoid laws and institutions that are in tension with, or even undermine, one another. Moreover, whether a particular law is acceptable to (some) citizens might depend significantly on what other laws exist. Hence we might have good reasons for wanting a very rough-grained individuation of coercive laws, something like Rawls’s ‘package’ of conceptions of justice for constitutional essentials and matters of basic justice. And if this is so, then it seems we can end up with a very Rawlsian view, if not Rawls’s own view, by beginning with a convergence account of public justification, and combining that with a concern with ensuring the overall coherence of society’s basic laws.
5. The coerciveness of libertarianism and classical liberalism
Finally, I found Vallier’s comments on convergence public reason liberalism and libertarianism provocative. (Within philosophy, ‘classical liberalism’ often is distinguished from ‘libertarianism.’ I assume that Nozick-style ‘natural rights’ libertarianism is off the table on any public reason view, as it clearly is a ‘comprehensive’ theory. So by ‘libertarian’ I infer that Vallier has in mind classical liberalism. To avoid confusing myself, given the ideas that I typically associate with ‘libertarianism,’ I’ll just refer to classical liberalism here.)
As I understand it, Gaus’s argument for convergence liberalism’s ‘classical tilt’ is that both classical liberals (CLs) and egalitarian liberals (ELs) will agree (converge) on a core set of state functions, namely, those performed by a classical liberal regime (CLR). This is so even if CLs and ELs disagree over the relative coerciveness of CLRs and egalitarian liberal regimes (ELRs). (Both CLs and ELs have ‘reasonable’ views about coercion, but they disagree over the relative coerciveness of their preferred regimes.) Convergence liberalism ‘tilts’ towards CLRs because CLRs are at least acceptable to ELs (even if they find them highly coercive), whereas ELRs are not acceptable to CLs.
Yet, as Vallier correctly points out, enforcing classical liberal property rights requires state coercion. So why assume that a ‘full-blown’ CLR would be acceptable to ELs? More specifically, why think that ELs would converge with CLs in supporting the full scope of coercion exercised by CLRs with respect to property? To the extent that there is any convergence among CLs and ELs, it only would be with respect to certain core liberal rights, and not extensive coercively-enforced CLR property rights. (Insofar as there is convergence on a right to property, it likely would only exist with respect to personal property, not widespread private ownership in the means of production, intellectual property, inheritance laws, etc.).
So if CLs are justified in rejecting ELRs because they object to the coerciveness of (‘full-blown’) ELRs (given CLs conception of coercion), so too are ELs justified in rejecting (‘full-blown’) CLRs because they object to the coerciveness of CLRs (given the EL conception of coercion). The cases are symmetrical. Thus we have convergence only on core liberal rights (including a right to personal property); beyond this core there is no agreement in either direction. The only recourse (to avoid anarchy beyond the liberal core) is to a mutually acceptable decision procedure. In short, there is no classical tilt to convergence public reason liberalism.
I would be (pleasantly) surprised if Vallier actually agrees with this. If he does not, though, I would be curious to know where I have gone wrong.
Okay, that’s it! Thanks again to Kevin for writing such a thought-provoking book. I’m really enjoying reading it.