The Tyranny of the Ideal

by Gerald Gaus

  101 Gaus and Nichols, “Moral Learning in the Open Society.”

  102 This experiment was partly inspired by the work of Cummins, “Evidence for the Innateness of Deontic Reasoning,” and “Evidence of Deontic Reasoning in 3-and 4-Year-Olds.”

  103 In other treatments, with compatible results, subjects were given a seven-point scale, which allowed them to sit on the fence by choosing the middle. Interestingly, subjects trained on permissions seemed far more likely to sit on the fence (not knowing what to do) than those trained on prohibitions.

  104 Federal Communications Commission, Report and Order on Remand, Declaratory Ruling, and Order, http://transition.fcc.gov/Daily_Releases/Daily_Business/2015/db0312/FCC-15–24A1.pdf.

  105 See Benn, A Theory of Freedom, pp. 289–91.

  106 See Royal Society of Chemistry and American Chemical Society, The Discovery and Development of Penicillin, 1928–1945.

  107 Ibid.

  108 The analysis presented here concerns the distinction between, on the one hand, prohibitions/requirements and, on the other, permissions. This should not be confused with Hayek’s case for prohibitory rules over requirements. Hayek stressed that it was a fundamental mistake to follow Hobbes and J. L. Austin in seeing laws as commands. Although he recognized a continuum, Hayek emphasized that at one end is a quintessential command, which seeks to drastically reduce the feasible options open to the agent (at a limit to simply one act), and a prohibition that does not seek to prescribe specific actions in specific circumstances (and so reduce the feasible set to a singleton) but rather prohibits certain specific conduct, allowing complying individuals to choose their course of action from a slightly pruned set of options. For Hayek, the crucial difference between the two extremes is that prohibitory rules allow individuals to use their information to form plans and respond to novel situations while requirements do not. There is more to this point than most philosophers think, but developing it would take us too far afield. See Hayek, The Constitution of Liberty, p. 150. Compare Adam Smith, Theory of Moral Sentiments, pp. 78–82.

  109 See Page, Diversity and Complexity, chap. 1; Waldrop, Complexity. See also §II.2.2.

  110 The Order of Public Reason, §18. See also my Justificatory Liberalism, pp. 199ff.

  111 D’Agostino, Incommensurability and Commensuration, p. 104.

  112 Ibid., p. 105. Paragraph break deleted.

  113 Ibid. See also The Order of Public Reason, pp. 374–80.

  114 Gray, Post-Enlightenment Liberalism, p. 314.

  115 Wiles, Economic Institutions Compared, p. 40.

  116 For an excellent analysis, see Mueller, Public Choice III, chaps. 23–24.

  117 For a more thorough analysis of this question, see Van Schoelandt and Gaus, “Political and Distributive Justice.” I have also considered these matters in The Order of Public Reason, pp. 509–29.

  118 See Satz, Why Some Things Should Not Be for Sale.

  119 For a study of people’s different understandings of “taboo trades” and their relation to ideology, see Tetlock, “Coping with Trade-Offs.”

  120 For a discussion of this matter, see my essay “The Egalitarian Species.”

  121 Bicchieri, The Grammar of Society, p. 105. For a helpful overview see Güth and Tietz, “Ultimatum Bargaining Behavior.” See also van Damme et al., “How Werner Güth’s Ultimatum Game Shaped Our Understanding of Social Behavior.”

  122 Henrich and Smith, “Comparative Evidence from Machiguenga, Mapuche, and American Populations.” The Machiguenga and the Mapuche are small-scale societies; the other results are from urban university students in the United States, Israel, and Indonesia. The Ultimatum Game does not show significant variance in play on the basis of age or gender.

  123 Young, Responsibility for Justice.

  124 See Mackie, “Effective Rule of Law Requires Construction of a Social Norm of Legal Obedience.” My thanks to Gerry Mackie for discussions about this matter.

  125 In addition to ibid., see Bicchieri, Norms in the Wild; Bicchieri and Mercier, “Norm and Beliefs”; Platteau, Institutions, Social Norms and Economic Development; Stuntz, “Self-Defeating Crimes.”

  126 For a short description of this experiment, see Mockus, “Building Citizenship Culture in Bogotá.” For an in-depth treatment, see Mockus, “Bogotá’s Capacity for Self-Transformation and Citizenship Building.” See also http://www.corpovisionarios.org.

  127 Hadfield and Weingast, “What Is Law?” For an application of this model to broader themes concerning public reason, see Hadfield and Macedo, “Rational Reasonableness.” Van Schoelandt demonstrates the superiority of an alternative application (to that proposed by Hadfield and Macedo) in “Rawlsian Functionalism and the Problem of Coordination.”

  128 Rawls, A Theory of Justice, p. 106.

  129 Sen, The Idea of Justice, p. 135.

  130 The difficulty in further formalizing this idea is to distinguish R-type rules that are inconsistent with other R-type rules from other types of rules (Q-types) that may yield deontic prohibitions, requirements, or permissions that in some specific case are inconsistent with those yielded by the relevant R-type rule, and so are in that sense inconsistent with the R-type rule. It is, for example, the difference between assigning inconsistent property rights (a clash of alternative R-type rules) and property rights clashing with a right to assistance (a Q-type rule). If one holds that all rights must be compossible this problem will not arise, but then one must choose over entire schemes of rights (entire moral constitutions) rather than over individual rules. For a more thorough analysis of the idea of alternative social rules, see The Order of Public Reason, pp. 267ff.

  131 On a plausible contractarian account, the evaluation of proposed rules will be what Sen calls “comprehensive”: the perspectival representatives will consider how well the rules do on both procedural- and outcome-justice. See §I.2.2.

  132 Sen, “Maximization and the Act of Choice,” p. 184.

  133 See, for example, Cialdini, Kallgren, and Reno, “A Focus Theory of Normative Conduct”; Bicchieri, The Grammar of Society, chap. 1; Mackie, “Effective Rule of Law Requires Construction of a Social Norm of Legal Obedience.” I am employing the idea of a “descriptive norm” widely here, to include conventions.

  134 See Fehr and Fischbacher, “Third Party Punishment and Social Norms.”

  135 Bicchieri, The Grammar of Society, pp. 29–43. For a different analysis of how conventions become injunctive rules, see Sugden, “Spontaneous Order.”

  136 See Bicchieri, The Grammar of Society, pp. 39–40.

  137 It is because injunctive rules can solve such problems that, as the Ostroms have shown, it simply does not follow that the state and law are necessary for solving “tragedy of the commons” problems—where resource use has Prisoner’s Dilemma–like properties. See Ostrom, Governing the Commons.

  138 In Sen’s terms, the optimal socially eligible set is the joint choice set for our contractors. In The Order of Public Reason the optimal eligible set was defined in a less demanding way, requiring only what Sen calls a “maximal set” rather than a choice set (the elements of the choice set are always part of the maximal set, but not vice versa). I employ the latter, more demanding idea here because the logic of choice is more manifest and the problem simpler. See Sen, Collective Choice and Social Welfare, pp. 1–20.

  139 Baier, “Moral Obligation,” p. 226. Emphasis in original.

  140 See, for example, Mack, “Peter Pan Strikes Back”; Arneson, “Rejecting The Order of Public Reason.”

  141 See, e.g., Wall, “On Justificatory Liberalism,” pp. 143–44.

  142 On these matters see The Order of Public Reason, chaps. 3 and 4.

  143 See here Estlund, “Utopophobia,” pp. 132ff.

  144 Muldoon, Borgida, and Cuffaro, “The Conditions of Tolerance,” p. 330.

  145 Rawls, “Kantian Constructivism in Moral Theory,” p. 306.

  146 Rawls, Justice as Fairness, p. 154. The passage also occurs in Political Liberal
ism, p. 197.

  147 Rawls, Justice as Fairness, p. 128.

  148 See, e.g., D’Agostino, “The Orders of Public Reason.”

  149 Hayek, Law, Legislation, and Liberty, vol. 2, The Mirage of Social Justice, p. 146.

  150 The idea is wonderfully explained by Moehler, “The Scope of Instrumental Morality.”

  151 There is reason to doubt this. Unless Betty is committed to injunctive norms, she is apt to keep the bargain only if threatened with punishment. She will be tempted to play “snatch” rather than “exchange.” See Schwab and Ostrom, “The Vital Role of Norms and Rules in Maintaining Open Public and Private Economies.”

  152 This distinguishes the present analysis from that of some followers of Rawls, who seem to be committed to the impossible project of always justifying coercion. For an excellent analysis, see Van Schoelandt’s “Justification, Coercion, and the Place of Public Reason.”

  153 Vanderschraaf, “The Circumstances of Justice,” p. 330. Vanderschraaf’s overall analysis differs from that presented here; on his view justice appears to exclude settling on a Nash equilibrium in pure strategies, leading instead to a correlated equilibrium in which everyone gives up something. I have argued elsewhere that such a “compromise requirement” is not appropriate in the context of rule selection, especially once dynamic considerations are taken into account. The Order of Public Reason, pp. 403ff.

  154 For a general analysis, see Page, Diversity and Complexity, pp. 109–10, 138–40.

  155 Some, such as Donald Saari, have recently argued the Borda count is manifestly the best. For a critical analysis, see Risse, “Why the Count de Borda Cannot Beat the Marquis de Condorcet.” For Saari’s reply, see “Which Is Better: The Condorcet or Borda Winner?”

  156 Nozick, Anarchy, State and Utopia, p. 98.

  157 Schotter and Sopher, “Social Learning and Coordination Conventions in Intergenerational Games,” p. 507.

  158 See my The Order of Public Reason, chap. 7.

  159 Bicchieri and Chavez, “Behaving as Expected.” See also Bicchieri and Chavez, “Norm Manipulation, Norm Evasion.” In an attempt to replicate the first experiment’s results with children it was found that young Responders, though they reported the coin toss to be fair, rejected the outcome when they ended up with a low offer. See Castelli et al., “Fairness Norms and Theory of Mind in an Ultimatum Game.”

  160 Bicchieri and Mercier, “Self-Serving Biases and Public Justifications in Trust Games.”

  161 Hayek, The Constitution of Liberty, p. 63.

  162 This is, alas, a fairly widespread view today. See, for example, Cohen, Rescuing Justice and Equality, part 2; Estlund, “Human Nature and the Limits (If Any) of Political Philosophy.” For a sustained, characteristically intemperate, critique of stability, see Barry, “John Rawls and the Search for Stability.”

  163 “Other things equal, persons in the original position will adopt the most stable scheme of principles.” Although the “criterion of stability is not decisive,” if the parties find that a conception is unworkable, this would force a reconsideration of their initial choice. Rawls, A Theory of Justice, pp. 398–99, 472, 505.

  164 Ibid., p. 401; Weithman, Why Political Liberalism?, p. 45. On the idea of a justice equilibrium, see also my “A Tale of Two Sets.”

  165 I make this argument in some detail in “The Turn to a Political Liberalism.”

  166 I am following Page, Diversity and Complexity, pp. 149–50.

  167 See Boyd and Richerson, The Origin and Evolution of Cultures, chap. 9.

  168 See, for example, ibid., chaps. 1 and 2; Weisberg and Muldoon, “Epistemic Landscapes and the Division of Cognitive Labor.”

  169 Page, Diversity and Complexity, chaps. 6 and 7.

  170 This result relies on several theorems, which show the benefits of averaging performance over a wide variety of circumstances. See ibid., chap. 6. Recall how averaging across diverse perspectives increases predictive performance (appendix B).

  171 Page, Diversity and Complexity, chap. 7.

  172 Various caveats are necessary here, of course. If reformist and conservative citizens tend to come into conflict and so destabilize the constitution, then the benefits of diversity may be swamped. See Page, Diversity and Complexity, pp. 194ff. My aim here is not to show that diversity never poses problems for the Open Society, but that diversity of perspectives on justice has critical, and almost always overlooked, benefits.

  173 Haidt, The Righteous Mind, chap. 5.

  174 Ibid., p. 161. “Very conservative” respondents rely more on authority and loyalty. In the discussion referred to here, Haidt was relying on an earlier version of this theory, which specified only five foundations; some of the foundations are differently characterized in this earlier version.

  175 Ibid., p. 290.

  176 Ibid., p. 293.

  177 Ibid., p. 294. Emphasis in original.

  178 Ibid., p. 277. Emphasis in original.

  179 Mill, On Liberty, p. 253.

  CHAPTER V

  Advancing from the Citadel

  It is as if I had performed a strategic retreat into an inner citadel—my reason, my soul. … I have withdrawn into myself; there, and there alone, I am secure.

  —ISAIAH BERLIN

  1 RECOUNTING THE JOURNEY

  OUR INQUIRY INTO IDEAL THEORY BEGAN WITH IDENTIFYING UNDER what conditions ideal theory is inherently distinct from a theory of moral improvement. Few doubt that a political philosophy can rank social orders in terms of their justice but, as Sen pointed out, that does not require specification of the ideally just social world. Without claiming to have identified necessary and sufficient conditions for a theory to be sensibly described as “ideal” (a hopeless task), I argued for a sufficient and, I think, enlightening condition: a theory of justice makes ineliminable reference to an ideally just condition if it specifies two distinct dimensions along which judgments of justice are made: how inherently just a social world is (the Social Realizations Condition), and how similar that social world is to the ideally just social world (the Orientation Condition). Here we have an interesting class of theories in which understanding of the ideal is absolutely necessary for our judgments as to whether one social state is more just than another. It may be less inherently just, but closer to the ideal, or more inherently just, but further from the ideal.

  Chapter II developed a more rigorous analysis that sought to better understand these conditions and how they relate to each other. Critical to this analysis was the idea of a perspective on ideal justice. A perspective is, as it were, a formally complete specification of an ideal theory. It takes a set of evaluative considerations (e.g., liberty, equality, desert) and evaluates a variety of social worlds—the features of those worlds that are relevant to justice—determining how just that world is. Because the ideal is necessary, a perspective also judges the similarity of the basic features of social worlds to the ideal and how far they are from the ideal. One of the real benefits of thinking more formally about the ideal—developing a model of ideal theory—is that we can appreciate that such ideal theory is a useful and distinctive enterprise only when the problem of securing justice is “moderately rugged.” If similarity of features and inherent justice are perfectly correlated, reference to the ideal is unnecessary; if they are perfectly uncorrelated the problem becomes chaotic. These moderately rugged landscapes are characterized by neighborhoods, in which the justice of near social words is correlated but, outside some area, the justice of other social worlds is not correlated with our present world. Critical to my analysis was the Neighborhood Constraint: we know far more about the inherent justice of social worlds within our neighborhood than of far-flung worlds outside of it. I stressed that what is in our neighborhood and what is feasible to institute—in the sense of bringing about what we expect to bring about1—are by no means identical notions, but we should expect them to be reasonably well correlated. Assuming that the ideal is not in our current neighborhood, we are liable to mistake what, and
where, it is. As we approach an ideal, we have good reason to suspect that it will not be what we expected, and we now can see a better social state.

  The very essence of ideal theory is that it confronts us with The Choice. In our neighborhood we have better grasp of the justice of possible social worlds, and so we can locally optimize—seek the best world in our neighborhood. If local optimization always put us on a path to the ideal, the ideal would not be necessary. Sen’s constant improvement model would suffice. If the ideal is necessary, sometimes it must tell us that the Mount Everest of justice lies in a different direction than local optimization. And thus The Choice. Do we make our world more like the ideal, and so making it less like the most just social state in our neighborhood? But we do not know the ideal well, as it lies outside our neighborhood. When should we choose modifications moving toward the ideal and not make our world more similar to a more just world we can know reasonably well?

  Ideal theory always confronts The Choice. However, if an ideal theory could expand its knowledge of the landscape of justice, it could come to better know its ideal. It would still have to make The Choice, but at least it would have a more certain ideal target. Chapter III was devoted to exploring ways to do this. As recent research in other fields has shown, a diversity of perspectives can be an amazingly effective way to solve the sort of rugged optimization problem posed by ideal theory. While we identified ways in which diversity can increase knowledge of the ideal, we saw that as diversity increases, and so diverse teams can better explore the entire landscape, the perspective breaks up into competing theories of the ideal. Thus rather than exploring different perspectives on an ideal, we end up with competing theories of the ideal. These theories can still benefit from each other’s searches in important ways, but it will often be well-nigh impossible for them to share their insights. Sometimes one can learn from some aspects of one perspective’s search, other times from another, and sometimes perspectives will combine and sometimes fission. The result will be a complex problem of partially overlapping, and shifting, “republican” communities of moral inquiry.