The Tyranny of the Ideal

by Gerald Gaus

  It is important to stress that stating that a society forgoes a moral rule on this matter is not to say that it has no rule whatsoever about it. It is common to distinguish injunctive from descriptive social rules (or norms).133 On Cristina Bicchieri’s influential analysis, a descriptive norm is solely focused on a group’s empirical expectations of what others in the group will do. An example of such a norm is a pure convention, say of walking on the right side of a sidewalk or footpath. If one expects others to walk on the right side, then it is typically in one’s interest to also walk on the right. In the case of walking on footpaths (unlike, say driving cars on a busy road), one may sometimes deviate from the norm (say one is going to a next-door shop, and it is easier to walk against the current), and others are unlikely to be significantly disadvantaged by one’s deviation. Thus descriptive norms are not enforced; the sole motivation by Betty is to best promote her interests and goals by aligning with the actions of others. Still, overall, groups can achieve large-scale coordination simply by knowing what most others are likely to do. In contrast, injunctive social rules carry not simply empirical, but normative expectations: others expect one to follow them, and are apt to hold one responsible for violations. They can form a practice of mutual responsibility (§IV.2.1.2). Norms of fair division are such injunctive rules: subjects tend not only to follow fair division norms, but to blame and punish those who do not comply.134

  While Alf’s core concern is whether a practice of responsibility regulating some matter is justified given his perspective, he also recognizes that a practice of accountability requires that other perspectives have reasons to comply: thus his concern in justifying a moral rule is whether all have adequate reasons to endorse injunctive rules. This is a stronger claim than saying all have an interest in conforming to a social rule. When thinking about the “no moral rule” option we should not conceive of it as precluding a descriptive norm about this matter. This is important, for the question whether to endorse some rule Ri is not simply whether we seek to achieve some coordination in some area of social life (a mere convention can do that), but whether this area is of such a nature and of sufficient importance that we must regulate it via (justified) injunctive rules—a system of shared normative expectations and demands, and quite possibly punishment. This is why a chaotic moral state of nature is an inappropriate benchmark for a contractualist theory, as if the absence of a moral constitution is no coordination at all. From the justificatory point of view, the absence of a justified injunctive rule about some type of social interaction may well be simply a descriptive norm or convention about it.

  That said, injunctive moral rules often have great advantages over purely descriptive norms or conventions when social cooperation needs to be secured. Descriptive norms can be unstable: since they are sustained only by empirical expectations of general conformity, as soon as one begins to doubt whether others are actually conforming to the norm, one is apt to withdraw one’s compliance.135 Thus “trembling hands”—noncompliance with the norm caused by errors in applying it—may lead others to abandon the norm as they conclude that general conformity is breaking down (“Look at the people violating the rule; it isn’t much of a convention!”). In contrast, injunctive social rules are upheld by a system of normative expectations and demands and thus are designed to respond to noncompliance. Consequently, while the absence of an injunctive rule about some matter does not imply the absence of social cooperation, it often does imply imperfect and fragile cooperation. The simple game of the Stag Hunt provides a good example (figure 4-4).136

  The best for both parties is to hunt a stag, but for that they must cooperate. While Alf cannot catch a stag alone, he can capture a hare alone, though the meat is much less than half a stag. It would seem that, since cooperating on hunting a stag is best for both, a descriptive norm (“We hunt stag around here”) would suffice. But note that if either Alf or Betty suspects that the other will not act according to the convention, he or she will be tempted to defect and hunt hare; in this sense the Hunt Stag/Hunt Stag equilibrium is fragile. An injunctive norm (“Hunt Stag!”) could change the game, allowing players to demand that the other Hunt Stag, and punish those who fail to. A practice of responsibility for hunting stag may well be to the advantage of both, even on a matter of simple coordination. When Alf and Betty seek to cooperate in situations with a Prisoner’s Dilemma–like structure (where each does best under universal cooperation rather than universal defection, but one does yet better if others cooperate when one defects), the case for an injunctive rule is obviously even more compelling.137

  Figure 4-4. A simple Stag Hunt game

  Thus whether a perspective ranks an injunctive moral rule above no such rule is the critical distinction for our contractors: any rule injunctive Ri ranked below the “z” option by a perspective is not eligible for that perspective; the set of rules that are eligible for all perspectives constitutes the socially eligible set. Now within this socially eligible set, some rule injunctive Rx could be judged by all contractors (i) as better than “z” but (ii) as a worse option than some other member of the socially eligible set. In more technical language, this rule would be Pareto dominated by another member of the socially eligible set. The set of rules that are (i) in the socially eligible set and are (ii) not Pareto dominated by any rule in the socially eligible set is the socially optimal eligible set.138 Clearly our contractors would prefer to select a rule from the socially optimal eligible set; to abide by an eligible rule that was, in the eyes of all, worse than another member of the socially eligible set would be unfortunate for all. They might feel a bit like asses—from everyone’s perspective on justice, they could do better. However, for our purposes the critical issue is the conditions under which a society could possess a socially eligible set, and so create a public moral constitution all can live with (whether or not it is optimal).

  3.3 Abandoning the Optimizing Stance

  Recall that our interest in a socially eligible set is that, even if all the perspectives in our contract cannot all converge on a single best rule for regulating some area of social life, they might converge on a set of injunctive rules, which all rank as better than no moral injunctive rule at all on this matter. For such an eligible set to obtain it must be the case that our contractors do not insist on what might be called the “optimizing stance”—that the only rule that is acceptable is one’s top-ranked rule. Here is where we must break with a great deal of moral and political philosophy, which identifies moral thinking with an optimizing perspective: thinking through the issue from his own perspective on justice Alf decides what is the best injunctive rule, or best principle, and so he concludes that, and that alone, is what morality requires. As Kurt Baier says, “Where interests conflict, there are many possible regulations dealing with the conflict. The directive embodying the regulation would not be properly moral (as opposed to being legal or conventional) unless it purported to be the best possible way of regulating such a conflict.”139 If Alf and Betty, who do not share a normalized perspective on justice, insist on the optimal choice as given by his or her perspective—each can see only as part of an acceptable moral constitution what he or she sees as the best rule—they will almost surely fail to endorse a set of eligible moral rules. Diverse perspectives will disagree about what is best because they disagree about justice. If we accept the optimizing stance, the idea of a contractarian theory of justice is either an illusion (because the contract, which was supposed to justify a conception of justice in the face of diversity, really presupposes a fully normalized perspective on justice) or a dead end (because no agreement will be achieved). But perhaps this simply shows the folly of a contractarian approach. Why shouldn’t a person insist on following his or her personal optimizing commitments? What good reasons could one have for not doing so?

  The fundamental reason is that the optimizing stance precludes a great good—one which that very perspective on justice almost surely endorses—of a social life of widespread moral acco
untability (§IV.2.1.2). Suppose that Alf, occupying his perspective, insists that only his optimal rule is acceptable as a basis for the moral regulation of some important interaction (say, about the rights of bodily integrity), while Betty insists that only her (incompatible) optimal rule could qualify as the basis of a practice of moral accountability. When they meet each other Alf will not be able to suppose that a goodwilled Betty, who reasons about justice as well as he can expect of another, will see that she ought to conform to his preferred rule. Employing her capacity for rational moral reflection and agency, she has constructed a reasonable perspective on justice; given it, she simply cannot come to the conclusion that his proposal is the best rule (as we have seen, her perspective categorizes and evaluates the world differently). But if Alf recognizes all this, he cannot infer ill will on her part, for he too embraces the optimizing stance, and he sees her perspective on justice as a reasonable one. Thus their reactive attitudes at the heart of the practice of moral accountability are undermined—and thus also the practice itself.

  The orthodox moral philosopher will insist that this is just too bad, but true justice is true justice, and Alf cannot renounce the conclusions of his perspective that his rule best tracks true justice.140 The Open Society, however, does not ask him to do so; indeed, if his perspective is to be part of a system that is morally improving (§IV.4), it is critical that he not renounce the conclusions of his perspective. Moral improvement depends on Alf (and others) seeking the best answers and, hopefully, convincing others that his is a better answer. What he must renounce is his claim that the optimal rule as identified by his perspective on justice is the sole acceptable basis of a system of moral accountability—that it must be part of our moral constitution. As we have seen, it is this insistence that undermines the practice of accountability—a practice that his own perspective on justice surely values.

  Some moral philosophers are, as they say, willing to “bite this bullet” and accept that their optimizing stance precludes relations of moral responsibility.141 Strawsonians will be deeply skeptical that they really can leave behind their participant perspective in favor of the theory of morality endorsed from what they see as the objective perspective. Their insistence on the correct, objective point of view and their formal pronouncements are forgotten when they reenter the participant perspective of an agent among others and react to the slights, lack of respect, and ill will that they attribute to others when they conclude that their moral rights have been neglected. But even from the objective perspective of the moral philosopher examining a system of relations from the outside, a system of rules to which goodwilled rational moral agents cannot hold each other accountable must be a terribly flawed system, for we have good formal and empirical evidence that without accountability and punishment, systems of moral rules are invaded by defectors, undermining the very basis of moralized, stable, social cooperation.142

  The optimizing stance is also self-defeating, once we take a more inclusive view of optimization. Our analysis in chapter II concluded that an individual perspective on justice will almost surely be unable to find its ideal; being confined to a neighborhood, the identification of its own ideal will be elusive. However, as we saw in chapter III, other perspectives can uncover parts of the landscape beyond one’s ken; revealing features of the social world that are not salient on one’s own view, they can help bring one’s own ideal into closer view. But this requires, we saw, a network of interconnected communities of inquiry—what we termed “republican communities of moral inquiry.” It is precisely the framework for such interactions that the moral constitution of the Open Society provides. Adopting the optimizing stance toward the moral constitution precludes one’s perspective participating in this framework for inquiry; relations of accountability, and shared empirical and normative expectations between one’s own perspectives and those from whom one could learn, will be undermined. To participate in shared moral inquiry with other perspectives on justice one must provide the foundations for moral relations of accountability in one’s interactions with them. To wish to learn from other views, while insisting that only one’s own view is correct and that all must live by it, is hardly a basis for a community of shared inquiry. By seeking to optimize in this way one forgoes optimizing in the sense of better understanding one’s own commitments regarding justice.

  It might be wondered why, if one cannot insist that one’s ideal be institutionalized in the face of disagreement by other perspectives, one should care about improving one’s understanding of ideal justice. It would seem that the Open Society allows one to better know one’s ideal, but simultaneously instructs one not to pursue it. Why, then, care about the ideal? Those who believe that inquiry into ideal justice is more like theoretical than practical inquiry143 will not see this as a pressing worry, but those who accept that recommendations are critical to justice (§I.1.5), will indeed wonder what is the point of such moral “inquiry.” We shall return to the idea of moral improvement (§IV.4), but this much is manifest: the Open Society invites all to share their ideals, and show other perspectives how our common life can be made more just. Especially given the polycentric nature of the Open Society, ideals of a more just society can spread from one network to another, often deeply changing the public moral constitution. Sometimes this consists in the elimination of unjust categories such as racial or gender classifications, sometimes in the development of new rules of personal privacy and bodily integrity. That the public moral constitution denies the claim of any perspective to implement its controversial ideal of justice by no means implies that deeper understandings of different visions of the ideal do not help shape our common life together. The benefits of exchange of ideas among different communities of moral inquiry can be widespread even if none of the ideas is accepted as a blueprint for our shared moral life.

  We can generalize this idea beyond inquiry into justice to more diffuse benefits. In a recent analysis of the conditions of tolerance and the gains from interacting with those with different ideological perspectives, Ryan Muldoon, Michael Borgida, and Michael Cuffaro develop a model based on Ricardo’s account of trade, which indicates that

  as tolerance is a measure of how willing to engage with others with different ideological commitments one is, tolerance increases one’s chances of discovering a satisfactory tradeoff between material gains and ideological purity. This suggests, then, that if one has an interest in material gain, then one has a corresponding reason to become more tolerant. On this model, increased tolerance results in increased reward, even when the potential discomfort of engaging with someone with different ideological preferences is taken into account. From this model, we see that rational actors ought to choose to become as tolerant as they can, with the expectation that this tolerance will be amply rewarded. Tolerance, then, can be thought of not just as a liberal duty to others, but as a rational duty to oneself: to promote one’s own self-interest best, one ought to be more tolerant of others. Rationality thus motivates individuals to see others who are different as potential partners in exchange, rather than as merely potential sources of ideological conflict.144

  Those with different perspectives—on justice, but on other matters too—provide opportunities for productive exchange based on comparative advantage; the more one fruitfully interacts with diverse others (even granting that this might offend some of one’s sensibilities about justice), the more opportunities for advantageous trades (or, more broadly, productive interactions). Again it is important that the Open Society arose in the great trading cities, in which not only different skills but different cultures and moral perspectives interacted. Thus satisfaction of a variety of goals typically strongly inclines against the optimizing stance. When people insist on their optimum choice as the only acceptable common rule, they preclude widespread acceptance of a practice of accountability that provides a framework for a wide-ranging system of cooperation among diverse agents, which provides innumerable opportunities for mutual benefit.

&nbs
p; 3.4 The Social Space of the Open Society

  Rawls tells us that the aim of his account of public reason is to arrive at a public moral constitution “that all can live with.”145 In the terms of our analysis, the fundamental claim is that in a society with a number of perspectives on justice N, versions of the institutions of the Open Society (§IV.2) will be in the socially eligible set for a group of perspectives coming near to N. This is not to say that all, or indeed many, perspectives will rank rules consonant with the elements of the Open Society as the best rules by which to live, but they will be rules that a maximally large subset of perspectives can live with. Yet, as Rawls also observes:

  No society can include within itself all ways of life. We may indeed lament the limited space, as it were, of social worlds, and of ours in particular; and we may regret some of the inevitable effects of our culture and social structure. As Isaiah Berlin long maintained (it was one of his fundamental themes), there is no social world without loss: that is, no social world that does not exclude some ways of life that realize in special ways certain fundamental values. The nature of its culture and institutions proves too uncongenial.146

  Some perspectives will conclude that no moral rules at all are better than the moral constitution of the Open Society: they place little importance on a cooperative social life supported by a practice of accountability, little weight on the improvement of their own perspective’s ability to better locate its ideals, and little importance on a framework of mutual exchange. All this is possible—and no doubt occurs. But to conclude that the space of the Open Society is inadequate in the face of perspectival diversity is to retreat into the social world of one’s own perspective, forgoing the great benefits of a practice of accountability and productive interchange with diverse communities of moral inquiry.147