The Tyranny of the Ideal

by Gaus, Gerald;

  We thus have arrived at:

  The Fundamental Diversity Dilemma: As we increase diversity of perspectives we can bring the ideal closer to our world, but as diversity increases we disagree about the justice of alternative social worlds, including that of the ideal.

  A community drawing on multiple perspectives has the potential to increase the effective size of its neighborhood, as well as to bring the ideal closer to it. But perspectives on the ideal are integrated. A perspective has a certain similarity ordering because it sees certain institutional and other features as relevant to justice; these features are the basis of its evaluation of the overall justice of a social world.43 When a perspective applies its mapping relation (including its predictive models) and its evaluative standards to these features it arrives at an overall justice score. Small variance in any of these may not make much of a difference, but substantial diversity in one element reverberates throughout the perspectives. We might call this diversity contagion: diversity is introduced into one element of a perspective, it induces diversity in another, and this in turn produces more diversity. We have seen that the sort of robust independence of elements of a perspective (ordering, features, overall evaluation) supposed by Page’s insightful model of perspectives largely (though not entirely) avoids this contagion;44 however, it is an inappropriate conception of a perspective for the pursuit of the ideal. On the more appropriate model we have developed, if two perspectives on utopia differ much in one respect they substantially differ in others as well. When substantial and systematic differences set in, perspectives end up searching “different landscapes.”45 Thus, instead of seeing themselves as diverse teams exploring the same evaluative core, diverse philosophers create deeply different perspectives and so different landscapes of justice.

  The upshot of our analysis is that it is very difficult to “manage” evaluative diversity,46 if this means allowing “just the right” amount of diversity to solve an optimization problem (say, by admitting views that orient the search for justice in helpfully different ways) without introducing deeper diversity (as, for example, about what worlds are the most just). These differences fracture the optimization landscape into multiple landscapes with people searching for divergent ideals, and so our searches often have little interest to others. Ensuring that large groups are diverse enough to effectively find new solutions to shared problems is difficult indeed. Authoritarian regimes have often learned this lesson the hard way. Stuck in the mire of approaching problems of reform in the same orthodox way (employing the orthodox perspective), such regimes have often sought to allow diversity of perspectives within in the limits of the orthodox ideology. Glasnost and Perestroika were the shining examples of the late twentieth century. In the twenty-first century Chinese communistic capitalism is another attempt—its efforts to manage diversity include the Tiananmen Square Massacre. Of course none of this is to say that such efforts cannot succeed, but they seem inevitably to rely on force, intimidation, and oppression.

  3 THE BENEFITS OF DIVERSITY

  3.1 The Fundamental Diversity Insight

  I have been canvassing some of the problems arising from diverse perspectives, problems that I believe pose insurmountable obstacles to effective use of the Hong-Page theorem in a theory of ideal justice. But that this formal approach is not as helpful as it initially appeared by no means implies that diverse perspectives are not a great boon for a society that seeks greater justice, or better pursuit of the common good. The companion to the Fundamental Diversity Dilemma is

  The Fundamental Diversity Insight: Any given perspective Σ on justice that meets the Social Realizations and Orientation Conditions is apt to get caught at poor local optima; other perspectives can help by reinterpreting the problem or applying different predictive models, showing better alternatives in Σ’s neighborhood.

  The Fundamental Diversity Insight is the positive lesson to be learned from our examination of Page’s pathbreaking work. While we must be skeptical that the formal Hong-Page theorem is of direct relevance to our problem, an underlying lesson remains. Yes, different perspectives can have great difficulty communicating, especially in ideal theory, where it may be very hard to know whether we are talking about the same social worlds at all. And yes, as perspectives differ, normative disagreement arises, so that in the end the perspectives are exploring different landscapes (optimization problems). A society that embraces deep perspectival diversity will be one of deep normative diversity. But none of this shows, as Fred D’Agostino puts it, that the different perspectives cannot ever “get it together.”47 It shows, rather, that getting it together is the really difficult task; diversity’s benefits are by no means automatic, but neither are they always beyond our reach.

  Of course we learn in multitudinous ways from each other. Other perspectives uncover features that our way of looking at the world neglected. In the last generation feminist perspectives certainly have played a fundamental role in alerting almost all theories of justice to features of the social world that were all-but-invisible to them. Feminists explored features of the social world, such as the family and language, that simply did not register as critically important to most traditional political theories. Here, Landemore and Page are certainly correct that the force of the better argument can produce convergence on parts of perspectives.48 Some claims of feminist perspectives have been accepted by liberal, conservative, and other perspectives, while others could not be integrated, and the perspectives remain distinct, pointing to different views of justice. All of this is as important as it is familiar—I shall not seek to rehearse these considerations. Let us focus on those benefits highlighted by our model.

  3.2 The Deep Insight of Hong and Page’s Analysis

  Recall Popper’s and Elster’s fundamental insight that a reasonable utopian theory must admit that its ideal is constantly revisable (§II.3.4). However, we cannot suppose—as perhaps Wilde did—that once “humanity” lands at Σ’s utopia, Σ adherents will be able to see from there the “better country” for which they should now set sail. After all, Σ supposed that it was landing at an island with the highest peak. It may be so lucky as to constantly see further and higher, but there is no reason to expect so. The critical feature of Hong and Page’s pass-the-baton dynamic is that, when Σ has done as well as it hoped, some other perspective may share enough while also differing enough with Σ that it can, as we put it, see a way upward (§II.1.1.2). Suppose a traditional socialist perspective saw socialism as the abolition of “capitalism,” understood as “markets-with-private property.” And suppose that socialism thus understood was secured (as it was in much of the world in the early and middle twentieth century). Now enter another perspective, which sees “private property-with-markets” as not a single feature (or two that always march hand-in-hand), but two quite different features, and so “state socialism” could be improved by “market socialism.” In a stylized history of ideas, we might understand this as a claim of the perspective of Tito’s Yugoslavia.49 It shared enough with Soviet state socialism to make intelligible to the Soviets the claim that “according to our understanding, ‘socialism’ should distinguish much more sharply markets and private property, and the improved socialist ideal lies in accepting the former but not the latter.”

  A perspective does not need to “land” at its utopia to understand this. What can be done after arrival can be done before—deliberating with allied perspectives to see what properties it has overlooked, or what properties it has mistakenly bundled. And what holds for the highest peak also holds for all peaks from which it cannot really see the “next destination.” It is other, differing but related perspectives, that are most likely to see overlooked superior alternatives—ones that the original perspective can appreciate it has overlooked. This is the great insight of the handing-off-the-baton dynamic, even when the conditions for the formal proof are not met. It is a deep and important idea that we must not overlook, though remaining skeptical of more ambitious attempts to use the
Hong-Page model.

  3.3 Modular Problems

  This is another case where we tend to learn the most from those whose perspectives have more in common with ours. D’Agostino, though, has fruitfully analyzed the social epistemology of more radical forms of perspectival disagreement. Even perspectives that differ on all five of our elements can, perhaps with modification, adopt solutions from other perspectives to help solve problems. To borrow an idea from D’Agostino, when a problem is of a modular nature, its solution by one perspective can be fitted into that of others.50 One of the noteworthy features of social democratic perspectives in Western Europe and the United States in the 1990s and the first decade of this century was concern about improving the quality of delivery of social services. While traditionally many social democratic perspectives saw the relation between providers and recipients in resolutely antimarket terms, a number of social democratic governments and parties adopted efficiency criteria from more traditionally liberal or libertarian perspectives to improve delivery of public services, such as health care and education. Although these social democratic perspectives did not simply adopt the evaluations of health care and education arrangements of more free market–friendly perspectives, they did see that for some particular problems central to the social democracy, features stressed by free market perspectives (such as aligning agent incentives closer to those of principals) were also features relevant to justice as understood by the social democratic perspective.

  As I have stressed earlier, institutions can have important interdependencies (§II.2.1), and so what we take as a “modular” solution may be more embedded in other commitments than we first realized, leading to much more radical changes than initially contemplated. Whether a problem is modular is something we discover. Consider, for example, the evolution of the Labour Party in the United Kingdom, from its early twentieth-century founding to, say, its perspective at the turn of the twentieth-first century. By taking over efficiency and market considerations as part of its perspective, it came to the point of explicitly rejecting what was a fundamental commitment—“the common ownership of the means of production”—as one of its official aims.51 Learning from other perspectives and revising one’s own perspective are deeply intertwined activities. Of course this sort of dynamic is to be expected if we accept Popper’s (and my) basic point—that we learn what our ideal is as we seek it (§II.3.4).

  3.4 Recombination

  Different perspectives can recombine to form new perspectives. At times this process is carried out radically and quickly, when diverse political perspectives recombine into a new one. One remarkable instance of this was the emergence in 1950s America of “fusion” conservatism, fusing conservative traditionalism with classical liberalism, most notably in the work of Frank S. Meyer.52 Meyer insisted that traditional conservatism “was far too cavalier to the claims of freedom, far too ready to subordinate the individual person to the authority of the state.”53 Whereas traditional American conservatives such as Russell Kirk upheld James Fitzjames Stephen’s critique of Mill’s defense of liberty, Meyer defends Mill: “The only alternative to the moral rule of liberty is to enthrone the sad tendency of human history as right, to glorify with James Stephen ‘the man of genius who rules by persuading an efficient minority to coerce an indifferent and self-indulgent majority.’ … Liberty is the political end of man’s existence because liberty is the condition of his being. It is for this reason that conservatism, which in preserving the tradition of this truth, is only consistent with itself when it is libertarian.”54 It was not only Mill—Meyer insisted that Adam Smith and the Austrian economists such as Carl Menger, Ludwig von Mises, and Hayek must be integrated into an adequate American conservatism for the twentieth century.55 “Fusion conservatism” had a profound effect on American conservatism in the latter part of the twentieth century and continues to this day. “Extremism in the defense of liberty is no vice. … Moderation in the pursuit of justice is no virtue!”56 is not the rallying cry of a traditional conservative.

  This is but one example. Liberalism has recombined with versions of socialism, producing an egalitarian liberalism, distinctly ambivalent about private property and the market. In the first part of the twentieth century, this “new liberalism” sought to merge the socialist ideas of the United Kingdom labor movement with traditional liberal concerns to produce a “Liberal Socialism.”57 To L. T. Hobhouse, one of its chief architects, liberal “individualism, when it grapples with the facts, is driven no small distance along Socialist lines.”58 In the United States, John Dewey pursued a similar recombination project, more Marxist than Hobhouse’s.59 Recently we have witnessed fusions of feminism or environmentalism with various liberal, socialist, and conservative worldviews. And, as our example of “bleeding heart libertarianism” in §II.2.1 indicates, products of recombination such as egalitarian liberalism themselves can be combined with doctrines such as libertarianism.60 The diversity of perspectives has led to yet more new perspectives, each with their own distinctive understanding of the ideal—ones which, at least to their proponents, are more adequate than the perspectives out of which they arose.

  3.5 Improving Predictions

  Throughout the last two chapters I have stressed the heavy reliance of ideal theories on predictive models. To evaluate possible social worlds we rely on predictive models as to how their institutions operate—when the worlds are far off, I have argued, we must rely on these predictive models and very little else (§II.3). When the concern is the interaction of the institutions in social worlds very dissimilar from our own, the accuracy of these models is, to be generous, not high. We have already seen how predictive diversity is critical in improving our predictive models (§II.4.2, appendix B). As Page notes, diversity contagion, which led to the breaking up of perspectives—producing different classifications of the relevant features of the world, different understandings of similarity, and so different optimization problems—also encourages perspectives to employ different tools for understanding their (differing) social worlds, with their different problems.61 As our perspectives differ, we develop new tools for modeling and predicting, and diversity of the tools is itself an important force in helping all perspectives better model their landscapes. As has been discussed (§III.2.2), whereas some features of Page’s analysis depend on agreement in perspectives, diversity of predictive models helps everyone better search their own justice landscapes.

  4 ESCAPING THE TYRANNY OF THE IDEAL

  4.1 The Tyranny of The Choice

  Let us pause to take stock of some of our conclusions thus far. Recall first that an interesting ideal theory must meet the

  Social Realizations Condition: T must evaluate a set (or domain) of social worlds {X}. For each social world i, which is a member of {X}, T evaluates i in terms of its realization of justice (or, more broadly, relevant evaluative standards). This must yield a consistent comparative ranking of the members of {X}, which must include the present social world and the ideal, in terms of their justice.

  I claim that this is a condition for an “interesting” ideal theory in the sense that only an ideal theory that meets it identifies an ideal social world that allows us to compare its justice to our world and to intermediate social worlds. It is certainly conceptually possible to have a theory of the ideal that is like a dream (§I.1.4), from which, when we awake, we are uncertain how this dream world compares to our world, or whether there might be other worlds that fall short of the dream world but still are admirably just. But we have set dreaming aside. The Social Realizations Condition also requires that an interesting theory of the ideal be able to give us a prediction of how different worlds, structured by certain institutions and practices, will work out in terms of their overall justice. And that is because we wish to aim at the ideal—even if we cannot actually achieve the ideal “down to the least detail,”62 we can achieve approximations of it (§I.1.5). Recall Rawls’s conviction that “by showing how the social world may realize the features of a realisti
c Utopia, political philosophy provides a long-term goal of political endeavor, and in working toward it gives meaning to what we can do today.”63 But if we are to actually seek out the ideal—if it is to guide our quest for justice—we must know how this set of social and political institutions will work out under whatever constraints the theory deems morally relevant and fundamental to the human condition.

  I added to this:

  The Orientation Condition: T’s overall evaluation of nonideal members of {X} must necessarily refer to their “proximity” to the ideal social world, u, which is a member of {X}. This proximity measure cannot be simply reduced to an ordering of the members of {X} in terms of their inherent justice.

  If the Orientation Condition does not hold, the ideal is unnecessary to orient our search for greater justice—when it is not met the pursuit of justice can be understood in terms of Sen’s climbing model (§I.1.3). If we could just climb up the Social Realizations ordering we would not need to orient ourselves by locating the global optimum. When the Orientation Condition is fulfilled we cannot simply move up the ordering to more just social states, eventually arriving at (or at least near) the global optimum in {X}. Increase in inherent justice and proximity to the ideal are distinct dimensions. On the model I have advanced, “proximity” is understood in terms of basic structures of the worlds; we have seen that the much-discussed idea of feasibility is unsuitable for the orientation function (§II.1.3).

  In addition to these two conditions I have insisted that any ideal theory confronts the Neighborhood Constraint (§II.3): we have better knowledge of the social realizations of near worlds than of faraway worlds. Remember, the Orientation Condition requires that there be a dimension of similarity (the x-axis) that tells us something about the difference in the underlying structures of social worlds; the Neighborhood Constraint insists that our knowledge of justice is not uniform across this structure, but is, in general, a decreasing function of the distance a social world is from our current one. Now, to be sure, many ideal theorists will reject much of the rest of my analysis by simply denying this constraint: they will claim that we have as firm knowledge of the social realizations of institutions in worlds radically different from our own as of those that are very similar to the world we live in. One can say this, but since the Social Realizations Condition requires that we employ predictive models to judge how an interacting set of institutions will function, and since these models are very imperfect, predictions that rely on them alone will be of dubious accuracy. Roughly, in our neighborhood the justice of our current world is correlated with those that are close to it, and so small changes in the basic structure of our social world can be expected to have modest effects on justice within this vicinity. Outside of this area we are relying solely on predictive models that have a marked tendency to decrease in accuracy as we move away from observed conditions. Because any perspective’s understanding of the ideal is based on incomplete information and predictive models of uncertain reliability, its location of the ideal and the estimation of its justice are always subject to revisions, perhaps quite radical ones (§II.3.4). Popper was absolutely correct: “it is not reasonable to assume that a complete reconstruction of our social world would lead all at once to a workable system”64—much less the ideal system—because we simply do not know enough about the ways such a completely different social world would work. If a theory of utopia denies this then, as Popper says, it replaces inspiration for sound social science: as has often been the case, it is more of an exercise in fiction and imagination than an analysis that should seriously inform our recommendations about how to make our world more just.