“Gaus’s choice”: The Open Society as an ideal

Abstract

Gaus has recently offered a novel argument for an open society, connecting it to recent methodological debates regarding ideal and non-ideal theory. The argument, briefly put, is that political philosophies based upon the search for an ideal to anchor their critical and prescriptive ambitions, so-called “ideal theories,” are bound to fail. The best possible case for doing ideal theory should lead us to recognize that such endeavors are conceptually incoherent or lead to morally perverse outcomes. Consequently, the best way to approximate a knowledge of the optimal institutional state of affairs is to organize institutions around the goal of maximizing perspectival diversity, what Gaus calls the “Open Society,” not according to an ideal blueprint. I argue that precisely because ideal theory has the problems that Gaus skillfully illuminates, we should be skeptical about the picture he paints of the Open Society.

Gaus has recently offered a novel argument for an open society, connecting it to recent methodological debates regarding ideal and non-ideal theory (Gaus, 2016). The argument, briefly put, is that political philosophies based upon the search for an ideal state of affairs to anchor their critical and prescriptive ambitions, so-called “ideal theories,” are bound to fail:

under the conditions of human existence, we cannot know what such an ideal would be…only those in a morally heterogeneous society have a reasonable hope of actually understanding what an ideal society would be like, but in such a society we will never be collectively devoted to any single ideal. (Gaus, 2016: xix)

Thus, the argument is composed of two smaller claims. The first, the “Ideal theory claim,” is that ideal theory cannot conceptually live up to its billing: the best possible case for doing ideal theory should lead us to recognize that such endeavors are conceptually incoherent or lead to morally perverse outcomes. The second claim, the “Open Society claim,” is that the best way to approximate a knowledge of the optimal institutional state of affairs is to organize institutions around the goal of maximizing perspectival diversity, not according to an ideal blueprint. Put together, Gaus claims that ideal theory leads to irresolvable dilemmas, and that the best way to make good on what ideal theory strives for is to not strive for ideal theory at all.

The problem for Gaus, I argue, is that the “ideal theory claim” is so well established that it undermines his “Open Society claim.” Precisely because ideal theory has the problems that Gaus skillfully illuminates, we should be skeptical about the picture he paints of the Open Society. I begin this short commentary by rehearsing Gaus’s arguments about ideal theory and the Open Society, and then show that the Open Society suffers the same pathologies as ideal theory, which I illustrate through a discussion of the current debate surrounding free speech.

Ideal Theory and “The Choice”

Gaus (2016: 41) argues that an ideal theory aiming to have any political import must satisfy two desiderata: the “social realization condition,” which says that an ideal theory must establish a set of evaluative criteria that can be used to “evaluate the justice of various social worlds, which realize to different extents, and in different ways, the relevant principles or standards of justice”; and the “orientation condition,” which says that these evaluative criteria cannot only refer to the inherent justice of a social world, but also to a social world’s proximity to the ideal. Ideal theory is both practical ¹ and necessary when it can tell us how just a state of the world is, and whether moving toward some particular state of the world will move us in the direction of the optimal state of the world.

The “orientation condition” is where the action is in Gaus’s argument. What distinguishes an ideal theory from a more pragmatic, problem-solving approach, or Sen’s “climbing model,” is that an ideal theory is oriented toward something beyond the moral problems of the day, or mere normative evaluation. Making the claim “state of affairs X is intolerably inegalitarian or illiberal or inefficient” does not require imagining an ideal state of affairs U where equality, liberalism, or efficiency is perfectly achieved. Whereas Sen or a pragmatist will advocate “climbing” the mountain we find ourselves upon with the evaluative criteria at hand, ideal theory, if it’s adding anything other than a rhetorical flourish to a set of evaluative criteria, must also be in the business of orienting us toward the ideal mountaintop.

The reason why this matters is because the evaluative criteria and the ideal orientation can pull in opposite directions. Because institutions are complex and interdependent, particularly when it comes to their relationship to justice, “the justice of an institution, practice, or policy can be dependent on what other institutions or policies are in effect” (Gaus, 2016: 63). A committed social democrat, for example, may very well think that, all things being equal, a more just society would be one where universities have no tuition, since that would make higher education accessible to people of all socio-economic backgrounds. However, this conclusion will likely be conditional on other institutions and policies obtaining: relative equality in secondary education, equal access to information about universities, and publicly provided SAT prep programs. Absent these pre-conditions, pursuing a tuition-free university could be counterproductive from a social democratic perspective, as it would essentially be subsidizing richer students’ access to university. While we may move closer to the more just institutional arrangement by implementing tuition-freeness, our society is made less just because we have not secured the background conditions necessary to make that institutional innovation justice-promoting.

As Gaus puts it, institutional complexity and interdependency create a “rugged optimization landscape,” where institutional similarity is not well-correlated with justice. The landscape is rugged because the route to the “peak” optimal institutional arrangement will involve innovations resulting in “valleys” of less just states of affairs. Creating the institutions necessary to secure optimal justice, in other words, will involve “worse-before-it-gets-better” movements en route to the optimum. We cannot assume that getting closer, in terms of institutional design, to the ideal will be justice-promoting, and we cannot assume that justice-promoting improvements move us closer to the ideal institutional configuration. Consequently, a theory that makes use of an ideal model will produce evaluative criteria and an orientation that pull in different directions: there is no reason to assume that making things more just here and now will orient us toward the ideal.

The ideal theorist’s response to this is simple enough: sacrificing small improvements in justice is worth the end-result of reaching optimal justice. After all, it’s an optimum! This response, however, assumes a heroic amount of knowledge and confidence in our assessments: it ignores the fact that we know much more about social worlds that are like our own, that are in our “neighborhood” (Gaus, 2016: 76), than hypothetical “yet-to-be realized” worlds (Gaus, 2016: 78). When presented with a choice between an institution that would incrementally improve the justice of our society, and one that would lead us toward the ideal, we have good reason to be more confident in our assessments of the former than the latter. The complexity, interdependency, and novelty of the ideal institutional configuration conspire to rob us of our confidence in it.

Because of this, ideal theorists invariably face “The Choice”:

in cases where there is a clear optimum within our neighborhood that requires movement away from our understanding of the ideal, we often must choose between relatively certain (perhaps large) local improvements in justice and pursuit of a considerably less certain ideal, which would yield optimal justice. (Gaus, 2016: 82)

This is the crux of the ideal theory claim: ideal theorists must be committed to an orientation condition of their theory; however, because of institutional interdependency and our better knowledge of more proximate social worlds, local justice and optimal justice will often pull in opposite directions, with our confidence in achieving the former being much greater than the latter. The Choice, which results, is the unpalatable consequence of being committed to articulating ideals and using them as guides in our normative analyses.

The Open Society

Given the problems that ideal theory faces, Gaus offers an argument for an Open Society. The Choice arises because of our inability to be confident in the ideal, our lack of knowledge of its shape and functioning, relative to our understanding of more local improvements. So, a way of alleviating the problem is by increasing our knowledge of what we want to achieve in society and our confidence in its achievability and workability. Gaus contends that our ability to make accurate predictions is increased by (1) the existence of perspectival diversity and (2) the ability for these diverse perspectives to communicate with one another. Gaus’s Open Society is meant to achieve both. First, the Open Society requires the existence of secure, stable, and recognized institutions: by establishing institutions, an open society creates common ontological categories for citizens and “common sources of interpreting those categories” (Gaus, 2016: 178), which can serve as the basis for, and conduits of, communication amongst them.

Second, and most important, is that in an open society, the design of these institutions is not based on a pre-decided ideal of what we want out of social cooperation. Instead, institutions ought to be designed with the aim of accommodating extant diverse perspectives and allowing the introduction and development of new diverse perspectives (Gaus, 2016: 174–176), since such diversity is our best asset for answering the question of which ends we want our social institutions to be oriented towards. Put differently, perspectival diversity can contribute to discovering the knowledge of the ends that ideal theory tries to find; however, achieving that diversity requires giving up attempts to organize our institutions around such knowledge.

So, which institutions are required for an “open society”? Gaus offers three basic categories, and implies a fourth. First, law and morality ought to be organized around moral prohibitions, not permissions (Gaus, 2016: 195–198). Because an open society seeks to achieve maximal perspectival diversity, permissiveness should be the default; in the Open Society, the basic idea is “Do it unless we say not to” and not “Don’t do it unless we say you can.” Going the other way, thinking that an action is permitted only when a moral rule permits it, will unduly limit the sorts of lifestyles, perspectives, and ideological viewpoints that populate the Open Society.

Second, the Open Society requires that there be “jurisdictional rights,” in order to mitigate the complexity that a maximally diverse society will necessarily engender. Because a public order based on the actually-existing perspectives of its members will be constantly in flux, as views and perspectives come and go, we must provide individuals with the ability to establish “jurisdictions” within which they have control over what happens. Private property rights, for Gaus, are the quintessential jurisdictional rights, allowing people to make “small social worlds” for themselves without negotiating with, or concerning ourselves with, others’ perspectives (Gaus, 2016: 201). More generally, the basic rights familiar to liberal society (association, religion, privacy) are jurisdictional rights, in that they establish protected individual and associational realms in which one’s perspective can hold court.

The third institution follows from the emphasis placed on private property: markets, the institution par excellence of facilitating cooperation without thick agreement (Gaus, 2016: 202). Markets can do this because they focus our requisite agreement on relatively less controversial things (like what is being traded and the terms on which it is being traded) without requiring agreement on the metaphysical nature or moral worth of things. A society wishing to avoid imposing values upon others, like the diversity-oriented Open Society, must be a market society of some sort.

Fourth, though stated less explicitly, the Open Society requires some brand of democratic politics. The basic idea of the Open Society—that we ought not, and cannot, know or decide upon what we ought to strive for a priori, and therefore must allow for a diverse and pluralistic society to figure such things out themselves—presumes some mechanisms for translating these diverse perspectives into public policy, and procedures to enact and pursue some measures and not others. Gaus suggests that this is the domain of democracy (2016: 202). While Gaus is clear that specifying one of the many democratic choice procedures in the abstract is impossible (2016: 225), a democratic procedure of some sort is still necessary to produce a social choice.

This is the Open Society claim. Perspectival diversity is the best bet for achieving the knowledge presupposed by ideal theory, yet achieving this diversity requires us giving up designing institutions according to some ideal. Instead, we should seek to create the institutions of an open society, designed to maximize diverse perspectives, and facilitate communication between them: permissive moral-legal codes; jurisdictional rights; markets; and democratic procedures and institutions.

The Open Society as an Ideal

The question that suggests itself is this: If “the ideal theory claim” is accurate, is Gaus entitled to offer the “Open Society claim”? Gaus wants to say that because its content is not pre-ordained, that its institutions are designed to make that content an ongoing debate, the Open Society does not fall victim to the pathologies of ideal theory. In ideal theory, the moral principles set the stage for deriving institutions; in the Open Society, the institutions set the stage for deriving moral principles. And yet, according to Gaus, a commitment to openness and perspectival diversity does direct us toward certain institutional configurations. We can say that the Open Society displays the pathologies of ideal theory if its institutional configuration entails institutional interdependency such that getting one aspect of the Open Society correct without others could be less desirable than an otherwise sub-optimal alternative. Or, put differently, it has the trappings of ideal theory if it gives rise to The Choice. Even with a commitment as minimal as the Open Society, I argue, we still find ourselves facing The Choice; pursuing the Open Society entails foregoing local improvements to justice, requiring something like an article of faith when weighing it against other alternatives.

Gaus’s insistence upon markets as an institutional core of the Open Society is important to note here, since markets are complex, requiring several interdependent institutions. As Rodrik (2008: 156) notes, markets require various components: property rights, regulatory institutions, stabilizing institutions, institutions for providing social insurance, and institutions for conflict management. This is true of the institutions necessary for the Open Society more generally: the protection of liberal “jurisdictional” rights requires a variety of effective governance structures, legal codes with roughly-accepted interpretations, and so forth. All of which is to say: the institutions of the Open Society, as minimal as they are, are not inherently a “package deal.” One can have property rights without the institutions necessary to secure that exchanges are happening voluntarily or fairly; the ability to exchange property rights can be secured without establishing competitive markets; we can have corporations without the bureaucratic and legal bodies necessary to effect legitimate corporate governance regimes, etc. Furthermore, one can have these things without having other aspects of a liberal society that an open society seems to require: a society can have competitive and smoothly running markets with secure private property rights, but also not have established institutions protecting free speech or association (see Chang, 2002: 71–110).

This becomes even more complicated when we consider the democratic institutions presumed by the Open Society. Again, Gaus is agnostic on the precise nature of these democratic politics. Still, we note that even a minimal conception of democracy has several institutional prerequisites: regular and free elections; universal suffrage; protection of civil liberties; and so forth. Each of these can be had without the others, resulting in authoritarianism, not democracy (Levitsky and Way, 2010). Democratic procedures—like the permissive codes, the jurisdictional rights, and the market economy, which are part and parcel of the Open Society—exhibit high levels of institutional complexity and interdependency: there is no reason to think that moving toward this institutional configuration in part, but not in total, is an inherently positive thing.

Faced with moving toward this Open Society or investing time and resources into, say, industrialization or growth, one faces precisely “The Choice” that Gaus warns of ideal theory. We must choose to pursue this Open Society despite the institutional interdependence that will render the pursuit a “rugged optimization landscape.” This means potentially forgoing surer local improvements to achieve a grander picture of society, the preferability of which is based on what looks an awful lot like an ideal theory, Gaus’s protestations to the contrary notwithstanding.

The fact of this complexity and interdependence makes sense of the actual historical development of such institutional configurations. As we know, private property, competitive markets, liberal rights and institutions, and democratic government were not established ex nihilio, but against a background of local norms, customs, and specific practices of exchange and ownership. Thus, even if we approve of the liberal societies that came later, the fact is that they required a large amount of “worse-before-it-gets-better” moves to develop. Private property required breaking up and disrupting established patterns and relationships of land-tenure and production (Moore, 1993; Scott, 1998), causing hardship and pain for those accustomed to such practices, only to get an initially fledgling, unequal, and imperfect system of property rights in return. Establishing competitive markets has required the displacement and disciplining of large swaths of population in order transform them from peasant workers to wage laborers (Polanyi, 1944). This, again, was not simply a matter of moving incrementally toward the market, but rather disrupting extant relationships and imposing costs upon people with the promise of a better institutional configuration in the future.

Similarly, the rights and procedures associated with liberal democracy have often not been developed spontaneously and without harm, but through violent and colonial means. Armed with ideas about the preferability and righteousness of liberal democratic institutions and norms, colonialists have been all too willing to subject peoples to hardships in the pursuit of an ideal—even when the ideal is liberal openness and inclusivity (e.g. Mehta, 1999). Now, no doubt Gaus is aware of this history, and is certainly no proponent of such actions. One may think we ought to pursue the Open Society, and still be critical of how we have historically gone about pursuing it. Still, this history should give us pause as it illustrates the more general point: the Open Society is not something that can be created spontaneously and all at once. Instead, choices must be made to enact it, choices which will involve breaking up past ways of doing things and foregoing other potential alternatives. This makes the force and violence underlying transformations to liberalism look less like an unfortunate historical hypocrisy, and more like the consequences of pursuing a set of institutions far from the neighborhood in which one finds oneself.

In this light, maximizing perspectival diversity, and the institutions necessary to achieve it, looks more like an ideal that, if we heed its call, will orient policy and institutional design away from local improvements in justice. A society that finds itself in a traditional sort of gift economy, or organized around norms that privilege group-oriented modes of consensus or elder rule, may eventually be improved if it were to develop markets, “jurisdictional” rights, permissive legal/moral orders, and democratic procedures. However, moving closer to this institutional configuration does not mean we are approximating or moving closer to the freedom and perspectival diversity such a configuration promises: instilling property rights without competition or exchange will likely create more power imbalances than it will mitigate; creating “markets” without recognized and legitimate legal regimes to enforce property rights can give moral cover to coercive market “exchanges”; establishing electoral procedures without universal suffrage or fair competition is not obviously an improvement upon the local and traditional means of decision-making; and so forth. Put differently, the decision to create such institutions will look much like “The Choice” that generically plagues ideal theory, because such institutions are complex, interdependent, and imperfectly correlated with the outcome they strive to achieve.

In attempting to deal with the problems raised by ideal theory, Gaus inadvertently demonstrates their ubiquity.

The Case of Free Speech

To put all this another way: a principled commitment to openness is in inherent tension with the conservatism (cf. Estlund. 2017) inherent to Gaus’s principled resistance to ideal theory. By way of illustration, consider the current debate surrounding the protection of hate speech. Park’s (2017) New York Times op-ed is exemplary of a growing attraction to the idea that speech ought to be restricted because of its ability to injure or harm. Park argues that the ACLU’s current policy of protecting all free speech claims regardless of its content—for example, protecting the speech rights of the KKK or neo-Nazis—is misguided because it ignores the current context in which such speech takes place: “the question [the ACLU] should ask itself is this: could prioritizing first amendment rights make the distribution of power in this country even more unequal and further silence the communities most burdened by histories of censorship?”

I discuss this stance here not to endorse it, but to point out that there is a Gausian logic to Park’s rather un-Gausian argument. Here, the ACLU looks like an organization enraptured by the ideal of free speech; pursuing this ideal leads the ACLU away from a more local optimum, where extant injustices are taken into account and inform our protection of public speech. Allowing the KKK or neo-Nazis to march results from The Choice to pursue the open ideal instead of a more local improvement where hatemongers are not empowered to spread their vitriol.

Undoubtedly Gaus would demur. Perhaps there are good reasons to restrict speech in some circumstances; however, we would only have epistemic confidence in the justice of such an intervention given the initial achievement of a perspectivally diverse Open Society. To know whether the ACLU ought to be restricted, something like the current ACLU position would have to prevail. But Park has a handy response: why should we have confidence that such a state of affairs is achievable and desirable now? As Wiens (2018: 25) puts it: “Gaus has yet to explore the causal mechanisms by which we could bring about the Open Society given the mechanisms that are operative at the status quo.” ² Not allowing Nazis to speak seems to allow us to reach a more local optimum, where minorities and the oppressed don’t have to suffer the indignities of hate speech (Waldron, 2014). It’s not obvious why we should forego that optimum to pursue the less proximate, and therefore less sure, condition of perspectival diversity.

The dilemma turns on what “neighborhood” we think we’re in. If we are in the neighborhood of the Open Society then it would make sense to follow the current ACLU policy. If, however, we are far off from such a neighborhood, we would have no reason to follow such a policy, and it would appear prudent to heed Park’s advice. Now, maybe we have some reason to believe that we currently are or aren’t in the neighborhood of the Open Society. But for Gaus’s claim to work—for the Open Society to be a “nonideal” and thereby escape The Choice—we’d need to think that it is always optimizing to pursue openness, that pursuing it would never require foregoing a surer improvement to justice. Otherwise the Open Society looks more like any other hypothetical institutional configuration, the wisdom of pursuing which will be a function of our proximity to it.

Conclusion

While my discussion in this short memo has been focused specifically on Gaus’s arguments against ideal theory and for the Open Society, by discussing the problem of hate speech, my hope is to suggest a more general point. I have sought to challenge the idea that by minimizing the standards of one’s normative theory, one is divesting oneself of the pathologies of ideal theory. A commitment to openness as a normative ideal is still a normative ideal. If one wishes to articulate an integrated normative political theory (see Simmons, 2010: 21–24) with any sort of institutional detail, then, given existing norms and practices, pursuing a “nonideal” like the Open Society will involve many of the same dilemmas that pursuing the ideal society raises. In light of this, perhaps ideals are necessary not because we can always act on their behalf, but precisely because we will unavoidably face difficult practical choices that force us to stray from them. If The Choice is ubiquitous, then perhaps we need an ideal not for purposes of orientation, but, as Carens (2013: 126) has put it, so that we can keep in mind “the difference between what we should embrace and what we should only endure.”