Political Liberalism,Western History,and the Conjectural Non-West

I. Decency beyond Ethnocentrism

In The Law of Peoples—first published in 1999—John Rawls advances the category of “decent” peoples. Decent peoples are those peoples who, although not liberal, should be tolerated by liberal ones. He lists two criteria that societies should fulfill to be counted as decent. ⁵ The first criterion is that the society not have “aggressive aims” against other societies. The second criterion is that the society secure basic rights for its members in accordance with its “common good idea of justice,” that the duties it imposes are such as to be seen by citizens as “fitting with their common good idea of justice and . . . not . . . as mere commands imposed by force,” and that its officials sincerely believe that “the law is indeed guided by a common good idea of justice.” ⁶ Furthermore, Rawls argues that a “decent consultation hierarchy,” which is “one kind of decent people” (leaving the possibility open that there be other kinds), ⁷ is ruled by a comprehensive (religious or philosophical) doctrine. ⁸ Rawls then gives an example of such a decent consultation hierarchy, an “idealized Islamic people” that he calls Kazanistan; I return to Kazanistan later in this essay. ⁹

Although the judgment about whether nonliberal societies are decent or not is made from within the “point of view of [liberal societies’] own political conceptions,” Rawls submits that the Law of Peoples should not therefore be considered ethnocentric. ¹⁰ He argues that whether it is “ethnocentric or merely western” turns not on its “time, place, or culture of origin” but on its “content” and whether this content “satisfies the criterion of reciprocity.” ¹¹ Rawls is thus keen to describe decent societies in such a way as to satisfy this criterion, providing conditions minimal enough for liberal and decent societies to be different from each other but demanding enough for them to respect (and abstain from interfering in the internal affairs of) each other.

Although the question of ethnocentrism is not front and center in The Law of Peoples, his notes and correspondence leading up to the publication of the book reveal that it had been on Rawls’s mind for a long time. While he writes, in the preface of the published work, that he had been thinking of developing the account provided there since “the late 1980s,” in a letter to Thomas Pogge, dating from the summer of 1994, ¹² Rawls writes that he had “left the subject since the mid-seventies because [he] hit a problem then that [he] didn’t know how to resolve.” ¹³ He further states that he saw the Law of Peoples as “connected with the problem of whether justice as fairness is historicist or relativist,” which is the reason why he “started on it again several years ago and tried out various ways of coping with the earlier problem,” hoping “to fix things up, yet that will be difficult.” ¹⁴

The difficulty of the task and Rawls’s dissatisfaction with his initial attempts to describe decent societies, to which I return below, are recurrent themes in the correspondence. Interestingly, the language of (avoiding) relativism and historicism, which comes up a few other times in the notes and letters, ¹⁵ drops out of the published version; only the language of “universal-in-reach” is maintained. ¹⁶ Indeed, it is possible to read The Law of Peoples as reflecting a toned-down version of Rawls’s ambition of addressing the question of ethnocentrism, in response to the pushback he received (this pushback is well-illustrated in the archives; the idea of decent societies was viewed as too liberal by some and not liberal enough by others ¹⁷ ). As he writes in a letter to Arthur Applbaum, dated November 16, 1994, Rawls was “torn between two possibilities.” ¹⁸ The first possibility involved viewing “the liberal conception of justice and the hierarchical conception” as “equally just.” The second possibility is that “they are not equally just” but rather the latter is “sufficiently just for those societies to be recognized as reasonable, or at least as decent members of peoples in a society of peoples.” He describes the second as the easier position to defend and writes that he “will probably end up going in that direction.” This is ultimately what happened; decent societies are, as just mentioned, evaluated from the point of view of liberal societies and are therefore to be viewed as part of the discussion of the appropriate foreign policy for a liberal democracy, and not of global justice more broadly. ¹⁹

The reduced ambition of the account of decent societies did not, however, eliminate the problems with it. Indeed, the published work elicited responses similar to the ones generated by earlier versions. ²⁰ I argue in what follows that the challenge remains to get the “content” right, whether the concern is global justice or more simply the appropriate foreign policy (and the limits of toleration) for liberal societies. More specifically, I contend that this content must remain faithful to political liberalism’s institutional, “realistic” approach. The rest of this essay is devoted to showing how and why Rawls fails at this challenge.

II. Pluralism and Institutional Bifurcation

“As a political doctrine,” liberalism, Rawls writes, “supposes that there are many conflicting and incommensurable conceptions of the good.” ²¹ He glosses the existence of conflicting conceptions of the good as “the fact of reasonable pluralism.” In liberal democracies, pluralism leads in turn to the development of two cultural spheres. The first, the “public political culture,” includes “the political institutions of a constitutional regime and the public traditions of their interpretation (including those of the judiciary), as well as historic texts and documents that are common knowledge.” ²² This sphere is distinct from the sphere to which “comprehensive doctrines” apply, which Rawls identifies as the “‘background culture’ of civil society.” ²³ The latter is “the culture of the social, not of the political. It is the culture of daily life, of its many associations: churches and universities, learned and scientific societies, and clubs and teams, to mention a few.” ²⁴

This distinction is crucial to Rawls’s Political Liberalism (1993), for it delimits the realm to which liberal principles, including the ideal of public reason, apply—namely, only to the public political culture. The political conception of justice in liberal democracies is independent from comprehensive doctrines; the first should be “freestanding” in relation to the second. As Rawls puts it, a political conception of justice draws all of its “essential elements” from the “category of the political” and not from any comprehensive doctrine that “always extends beyond the category of the political.” ²⁵

Compare this account with the account of decent societies; as mentioned above, in the case of a decent society, “its religious or other underlying doctrine is assumed to be comprehensive and to have influence on the structure of government and its social policy.” ²⁶ In other words, “the category of the political” in a decent society is in fact coterminous with the comprehensive doctrine animating it. Or perhaps it is more accurate to say that the category of the political simply does not apply to decent societies at all. In any case, it is clear that decent societies are not characterized by the dualism between the public political culture and the background culture.

Working backward in the comparison between the two types of society, what this lack of dualism suggests is that decent societies also lack pluralism. In the introduction to Political Liberalism, indeed, Rawls argues that the “dualism” between the public political culture and background culture is “not a dualism originating in philosophy” but rather one that originates in the “special nature of democratic political culture as marked by reasonable pluralism.” ²⁷ Pluralism is—to complete the picture—a “reasonably favorable historical condition” ²⁸ for the emergence of this “special nature” of democracy, because it “allows a society of greater political justice and liberty.” ²⁹

Rawls does not want to attribute pluralism to the so-called decent societies despite the fact that decent hierarchical societies are “associationist” in form, meaning that “the members of these societies are viewed in public life as members of different groups, and each group is represented in the legal system by a body in a decent consultation hierarchy.” ³⁰ On the one hand, the significance of this associationist aspect of decent societies should not be exaggerated, because it is, as Forrester suggests, intrinsic to Rawls’s conception of society as such; Forrester argues that, in A Theory of Justice, “voluntary associations” rather than social classes are viewed as the main units of society and described in “naturalized terms.” ³¹ That Rawls glosses the relevant groups in decent societies as “associations, corporations, and estates,” ³² and not classes, or even religious or ethnic groups, may say more about his general social presuppositions than about his ideas of the nonliberal, non-Western world. On the other hand, it is telling that, while voluntary associations in liberal democracies are precisely the spaces in which distinct comprehensive doctrines develop and affirm themselves, they all share, in decent societies, the same comprehensive doctrine. Here Rawls was inspired, as he writes in letters and earlier drafts, by Hegel and the legal scholar Philip Soper. ³³ But he was also thinking of Islam. In a letter to Fernardo Tesón (probably from 1994), ³⁴ he writes that, while he took the idea of a just consultation hierarchy from Hegel, a “German friend” told him that the term “is also used by present day democratic movements in Islamic societies.” “An interesting (to me) coincidence, if true,” he comments.

In any case, besides the existence of various associations in decent societies that, contrary to associations in liberal democracies, share the same comprehensive doctrine, decent societies also harbor minorities without thereby undermining their unity around a comprehensive doctrine; as he describes it, Kazanistan is “marked by its enlightened treatment of the various non-Islamic religions and other minorities,” which, he continues, “have been living in its territory for generations.” These minorities originate “from conquests long ago or from immigration which the people permitted.” ³⁵

Again it is surprising that conquests and—however limited—immigration should not lead to pluralism in decent societies; for Rawls otherwise argues, in relation to the notion of “common sympathies” that characterize societies, that “Historical conquests and immigration have caused the intermingling of groups with different cultures and historical memories.” ³⁶ While these groups “reside within the territory of most contemporary democratic governments,” he also attributes, in the same passage, “groups with diverse ethnic and national backgrounds” to both liberal and decent polities. Furthermore, as part of a discussion on reasonable pluralism in liberal democracies, Rawls stipulates that reasonable pluralism obtains even if “in other historical ages . . . people within a domestic society were united . . . in affirming one comprehensive doctrine,” and then comments, between parentheses, that “perhaps they never really have been.” ³⁷

What could lead those people who become liberal, and not those who might become decent, to turn the existence of a multiplicity of national and ethnic groups into pluralism properly so-called (the kind of pluralism that is associated with the dualism between the public political culture and the background culture)? Put differently, what would explain the institutional divergence between liberal and decent societies, despite their both being subject to conquests, immigration, and thus at least a modicum of diversity? The schematic account of Western history that Rawls offers in a few passages of Political Liberalism provides a telling, albeit brief, answer to the preceding questions—to anticipate, the answer revolves around the Reformation. More specifically, I will show that Rawls wavers between institutional and noninstitutional explanations; while he draws mostly on an institutional history, focused on issues of distribution of power arising from the Wars of Religion, he also invokes a culturalist account in which liberalism is intimately tied with Protestantism. My contention is that, on the institutional account, there should be no divergence between liberal and decent societies. The noninstitutional account does explain the divergence but at the high price of undermining political liberalism’s core, and distinctive, commitment.

III. Western History, the Reformation, and the Emergence of Liberal Democracies

In Political Liberalism, Rawls writes that “three historical developments deeply influenced the nature of moral and political philosophy”: the Reformation, the “development of the modern state with its central administration,” and modern science. ³⁸ The Reformation, on Rawls’s account, “fragmented the religious unity of the Middle Ages and led to religious pluralism,” which in turn “fostered pluralisms of other kinds.” ³⁹ Of the modern state, Rawls writes that it was “at first ruled by monarchs with enormous if not absolute powers.” ⁴⁰ In terms of modern science, Rawls says that it includes “the development of astronomy with Copernicus and Kepler and Newtonian physics” and “the development of mathematical analysis (calculus) by Newton and Leibniz.” ⁴¹

This reads like a standard account of the emergence of modernity. Yet what initially appears as a Westphalian account ⁴² of the rise of the modern state and its tribulations is actually meant to function as a story about the origins, not of the modern age but of liberal democracies specifically. This is where the Reformation comes in (I will return to the modern state and modern science below). It is the Reformation that leads to pluralism, which in turn leads to the “dualism” between the public political culture and background culture—that is, the sociological fact that makes liberalism both plausible and possible. It might thus be tempting to simply conclude that, since the Reformation occurs in (parts of) the Christian world, its mere occurrence explains, for Rawls, the institutional divergence between liberal states and nonliberal ones.

But Rawls must be careful, in his account of the Reformation, to avoid resting his explanation of the emergence of liberalism on Protestantism—doing so would entail that the freestanding conception of justice depends upon a (particular) comprehensive doctrine. Indeed, what Rawls highlights about the Reformation is its effect on the distribution of power between various religious groups, not its ideational content. To put it differently, he gives explanatory weight to the Wars of Religion rather than the Reformation per se. Consider, for example, his discussion of the idea of an overlapping consensus in Justice as Fairness. Rawls gives the example of “Catholics and Protestants in the sixteenth century” who both hold the view that “it was the duty of the ruler to uphold the true religion and to repress the spread of heresy and false doctrine.” The “principle of toleration” between them arises initially based on “various historical contingencies” and in the form of “a mere modus vivendi.” ⁴³ The two groups accept toleration “at first reluctantly” and only go on to affirm it with time. ⁴⁴ Crucially, Rawls’s explanation for the evolution of their attitudes is not located in the realm of values; rather, he says that “What is essential for an overlapping consensus is stability with respect to the distribution of power.” ⁴⁵ It is this equilibrium, brought about by the Wars of Religion, that leads the various conflicting groups within Europe to recognize that it is unlikely that one of them will ever dominate and gradually pushes the citizens of each country to affirm a shared (political) conception of justice, “irrespective of the political strength of their comprehensive view.” ⁴⁶ On this account of Western history, then, the rise of the conditions for pluralistic life can be traced to institutional, not cultural, factors. Pluralism is tied to the Reformation in a contingent, rather than necessary, manner.

As Roy Mottahedeh puts it, toleration developed in the West after the wars of religion “not because Christian theology convinced the combatants of the evils of intolerance but because intolerance seemed on the verge of tearing society apart and was too costly.” ⁴⁷ This passage is drawn from Mottahedeh’s article entitled “Toward an Islamic Theology of Toleration,” contained in Rawls’s “Islam and consultation hierarchy” folder in his archives; Rawls marks this passage and jots down “yes” in the margin. The passage continues that it took “many more generations” before the emergence of Locke’s theory of toleration, let alone the institutionalization of toleration, which came about in the nineteenth century. ⁴⁸

On this reading, then, what matters about the Reformation is the way in which it led to the rise of different conceptions of the good life—comprehensive doctrines—and, in the longer run, as a product of the Wars of Religion, to “stability with respect to the distribution of power” of these various comprehensive doctrines. But if this is true, this account implies that such stability, obtaining in any state, should be favorable for the development of pluralism, dualism, and therefore liberal democracy. It also implies, conversely, that the absence of such stability is conducive to conflict and/or domination. There is no third, decent option.

To propose that what matters about the Reformation, and therefore for the rise of liberal pluralism, is not mere stability in the distribution of power, but something else, requires playing up the distinctiveness of its values, not only its institutional effects. It would require that Rawls present the Reformation as bringing about not simply pluralism as balance of power but pluralism as a Lockean doctrine of toleration, not to mention one that percolates through the general population (since, it should be recalled, Kazanistan is also characterized by toleration, though only presumably at the level of its rulers). For example, in “Justice as Fairness: Political, Not Metaphysical,” Rawls writes that “the historical origin of this liberal supposition [that there is a multiplicity of conceptions of the good] is the Reformation and its consequences,” explaining the causal relationship to involve “the development of various doctrines urging religious toleration.” ⁴⁹

This suggests that Rawls might have ultimately viewed Protestantism as ushering in a new ethos of tolerance, not simply because of its being enmeshed in the power politics of the Wars of Religion, but rather because of deep-seated values within Protestantism—for example, the way that Protestantism approaches the question of the mediation between believers and God. Despite his explicit rejection of religion, Rawls was clearly worried about Church power. In his essay “On My Religion,” written in 1997, he reports that, “In the early years after the war,” he was interested “in the history of the Inquisition and how it had developed.” He continues, “the history of the Church includes a story of its long historical ties to the state and its use of political power to establish its hegemony and to oppress other religions,” noting that he came to think “of the denial of religious freedom and liberty of conscience as a very great evil” and that this denial “makes the claims of the Popes to infallibility impossible to accept.” ⁵⁰

The idea here then is that Protestantism—Protestant values—held a deeper import for Rawls’s philosophy than he recognized, ⁵¹ or even perhaps more generally, that liberalism, and particularly its commitment to secularism, is at base a distinctively Protestant doctrine, as critics have charged. ⁵² On this interpretation, the presence of a distinctively Protestant attitude (as in the rejection of Church power) is a necessary condition for the rise of liberal democracy.

But Rawls explicitly disavowed any view tightly linking Protestantism and liberalism. He argues, for example, that Protestants share the persecuting zeal with Catholics ⁵³ and that Muslims were as likely to be tolerant as both Protestants and Catholics. ⁵⁴ Indeed, the cost of making too much of the Reformation is a rather high one for Rawls; tying political liberalism to a religious worldview would contravene the whole point of Rawls’ proposal, which is to separate political liberalism from religious and other ideological content. For, recall that the three historical transformations adumbrated above form an important part of the introduction to Political Liberalism, the book that sets out the argument for a political conception of justice that is independent of any comprehensive doctrine. If Protestantism is built in as a condition for the possibility of pluralism and of liberal democracy, then the project of political liberalism is doomed from the get-go. Either, then, the history Rawls gives is institutional and therefore applies to all entities sharing its effects, or it is cultural, in which case political liberalism is not so political (institutional) after all. Something has got to give.

It is true that the history Rawls offers is very brief; it is nonetheless very important. Rawls writes that political philosophy shows “us the way in which [our society’s] institutions . . . developed over time as they did to attain their present, rational form.” ⁵⁵ The three historical developments explain how and why the institutions of modern, liberal democracies developed as they did, becoming more and more pluralistic and tolerant and exhibiting the institutional dualism described above. They contribute to Rawls’s aim to bring about “reconciliation” with “our” society and its history. ⁵⁶ This history, brief as it is, combined with Rawls’s sociological observations about the workings of liberal democracies (the institutional dualism between a public political culture and background cultures), arguably takes in Political Liberalism the place occupied by the original position in A Theory of Justice; it motivates the normative conception in the former just as the original position is the basis for the theory of justice laid out in the latter.

In any case, the point of the preceding is less to critique Rawls’s history than to reveal the way in which political liberalism relies on—whether Rawls states it outright or leaves it implicit—a story about the conditions for the possibility of the rise of liberal democracies and the fact that these conditions have to be seen as institutional for political liberalism to live up to its claim of distinctiveness from classical liberalism. Once this is recognized, then it follows that the realm of imagination, when it comes to decent societies, must be restricted to the same logic of institutions and their effects. In the next section, I will demonstrate how the appeal to the “hypothetical” nature of decent societies papers over an unjustified asymmetry in Rawls’s approach to Western and non-Western societies.

IV. Realistic Utopia and the Conjectural Non-West

Rawls writes that his “remarks about a decent hierarchical society are conceptual,” which he clarifies to mean that they involve thinking through whether such a society could be imagined at all (and, if so, whether it ought to be tolerated). ⁵⁷ Indeed, he appeals to the conceptual nature of decent societies to defend against the argument, “stated as an empirical fact supported by historical experience” that “full democratic and liberal rights are necessary to prevent violations of human rights”—that is, that decent societies (which respect human rights without being democratic) are impossible. ⁵⁸ Rawls also emphasizes that, just as he does not presuppose the existence of “actual decent hierarchical peoples,” he does not “presuppose the existence of actual reasonably just constitutional democratic peoples” either. ⁵⁹

While this line of defense was developed early on, ⁶⁰ it is also accompanied, in Rawls’s letters, by expressions of frustration at not being able to provide an adequate account of decent hierarchical societies. For example, in his letter to Tesón mentioned earlier, Rawls describes not being able to “give an account of a just consultation hierarchy” as “one of the most serious faults” of his original “Law of Peoples” essay. ⁶¹ In a letter, dated from March 1998, ⁶² to his colleague at Harvard and scholar of Islamic history, Roy Mottahedeh, cited above, Rawls writes of “having trouble” meeting the demand of readers to “give a description or example of a decent people,” and that he feels that his project “collapses” if he cannot do so.

Enter Kazanistan. Kazanistan appears in the final, book version but not in the earlier versions of “The Law of Peoples.” ⁶³ Perhaps Rawls was inspired to come up with this Islamic example by the remark of the German friend, mentioned previously, who suggested that the idea of a consultation hierarchy was used by “present day democratic movements in Islamic societies” (crucially, however, dropping the “democratic” part). In any case, in the letter to Mottahedeh, he appends a new addition to “The Law of Peoples”—a section on Kazanistan—in which he draws on Mottahedeh’s article “Toward an Islamic Theology of Toleration,” also cited previously. In the text of the letter, Rawls describes his account of an Islamic society as “imaginary” and asks Mottahedeh whether the rulers of Kazanistan might view his description “as an idealized form of what they tried to do in their best moments.” He also asks whether there are “Islamic countries that either now, or in history, might be illustrated by Kazanistan.” “Have there been any actual examples?” ⁶⁴

In the published The Law of Peoples, Rawls describes the example of Kazanistan as “hypothetical.” ⁶⁵ But he also mentions in a footnote that the doctrine of toleration he attributes to its rulers is “similar to the one found in Islam centuries ago,” citing the Ottoman Empire. ⁶⁶ He also writes that the notion of spiritual or moral jihad (as opposed to its military variants) that he associates with this example “was once common in Islamic countries.” ⁶⁷ He argues that “even though it is only imagined,” he does not think it “unreasonable that a society like Kazanistan might exist, especially as it is not without precedent in the real world.” ⁶⁸ Finally, he writes:

Readers might charge me with baseless utopianism, but I disagree. Rather, it seems to me that something like Kazanistan is the best we can realistically—and coherently—hope for. ⁶⁹

Kazanistan is a product of the approach that Rawls otherwise describes as conjecture, which involves “arguing as best we can that the social world we envision is feasible and might actually exist, if not now then at some future time under happier circumstances.” ⁷⁰

The first question that arises about the preceding is why Rawls felt such a strong need to provide an example, given the insistence on decent societies being conceptual. Why the need to make it so specific as to name it as “one of the ‘stans,’” as Anne Norton puts it? ⁷¹ Note that Rawls did not simply take the example of Kazanistan as illustrative; his project would “collapse” without it is what he says in his letter to Mottahedeh. Rawls wanted to have it both ways; decent societies had to be conceptual as a defense against “empirical fact supported by historical experience” and based on realistic examples (hence Kazanistan) as a defense against “baseless utopianism.”

This balancing act is not impossible to maintain in principle, but it requires being clear on what would make decent societies realistic and what would prevent utopianism from being “baseless.” As I have suggested throughout this article, the key here is to remain faithful to the institutional theory that is distinctive of political liberalism. To see how “conjecture” can succeed within this institutional approach, let us go back to Rawls’s argument, cited earlier, that liberal democracies, as he presents them, are also idealizations; Rawls abstracts away all of the ways (racism, gender discrimination, religious persecution, etc.) in which, as a matter of fact, so-called liberal democracies do not conform to the ideals that they in principle uphold. Similarly, in the case of “peoples” more generally, Rawls, as mentioned above, arrives at them by taking existing states and abstracting away institutional features, relating to unrestricted sovereignty, that he finds problematic. ⁷² Rawls also abstracts away divisions and conflicts, in order to maintain the notion of common sympathies. ⁷³

When conceiving of decent societies, on the other hand, Rawls does not abstract away from existing nonliberal states, setting aside the ways in which they fall short of ideals they hold. In fact, he does not start from existing nonliberal states at all; he starts instead from religion, namely Islam. Furthermore, although Islam is public as much as it is private in its operation in Kazanistan, there is no indication that Rawls draws Islamic principles from public political institutions and documents. ⁷⁴ Compare this with the way that he approaches culture in liberal democracies; except for the hints at Protestantism mentioned in the preceding, Rawls is primarily interested in the “public political culture,” described at the beginning of this essay. This public political culture is by definition tied to a specific set of political institutions; the ideas it embodies can be culled from public documents such as constitutions. In other words, it is because public political culture is “public” and “political,” not because it is culture as such, that it is possible to derive “certain fundamental ideas” central to a “political conception of justice” from it, ⁷⁵ as well as to view it as inspiring “common sympathies.” ⁷⁶

On the one hand, then, Rawls draws on (nonpublic) cultural precepts to inform his idealized account of decent societies. On the other hand, he actually also implicitly builds into his account the three crucial institutional developments cited earlier. First, Rawls grants decent societies, just as he does liberal democracies, the status of “peoples,” entities capable of engaging in international relations. What sets “peoples” apart from states as conventionally conceived is simply their lack of “traditional rights to war and to unrestricted internal autonomy.” ⁷⁷ In other words, peoples are idealizations of existing states. This means that decent societies are equally affected by the major “historical development” that is the rise of the modern state. Second, he attributes to decent societies the capacity of waging modern warfare, a capacity that actually poses a threat to liberal democracies (i.e., they are not restricted to technologies like bows and arrows; it is otherwise difficult to make sense of the requirement that a decent society not have “aggressive aims” against other societies). This suggests that decent societies are affected by the scientific and technological developments that characterize the modern world. Finally, as I have shown earlier, decent societies involve diversity, however minimal. All of this implies that, for an account of a decent society—however idealized—not to be baseless, it has to be compatible with these assumptions; a decent society, whatever culture it ascribes to, is also characterized by diversity and organized within a modern state form. ⁷⁸

One could perhaps argue here that the basis for Kazanistan is not in fact Islam but the Ottoman Empire (which Rawls alludes to in a footnote, as mentioned previously), and that therefore Rawls does provide a realistic basis for Kazanistan that checks the boxes just listed. But the Ottoman Empire is, as the name suggests, an empire, not a state (or a people in the way that Rawls imagines it). The institutional differences between the two forms of organization are far from trivial. Without fleshing those out here, it might suffice to mention that they are taken to be important by scholars of Islamic history. For example, in The Impossible State, Wael Hallaq argues that the traditional Islamic conception of politics “provided for what John Rawls calls a well-ordered society,” ⁷⁹ but—and this is the critical point here—this Islamic political conception is incompatible, according to Hallaq, with the modern state; the sovereignty distinctive of the latter is incompatible with the divine sovereignty characteristic of Islamic orders. ⁸⁰

In any case, I have shown in the previous section that a society characterized by diversity and organized within a modern state form is bound, on Rawls’s institutional account of the dynamics of power and order, to develop in the direction of pluralism—except if one group overpowers the rest. It is not just that the Ottoman Empire is not a good example, then; it is rather that good examples of decent societies cannot, in fact, be found. It is no wonder that Rawls expressed such frustrations concerning his search for them. The only two possibilities open for modern peoples are, on Rawls’s own account, liberalism and illiberalism. No amount of appeal to (nonpublic) cultural values or religious ideas about the common good can succeed at making plausible a modern state that involves the mix between human rights and the lack of democratic procedures that Rawls attributes to decent societies. ⁸¹ Or, to put it differently, Kazanistan is impossible.

Thomas Christiano argues that, while it might be true that a decent society is a social and political “anomaly,” and therefore “very unlikely,” ⁸² this does not imply its “conceptual or metaphysical impossibility.” ⁸³ But “metaphysical” possibility or impossibility cannot be the standard here, given that Rawls’s political liberalism takes as its starting point the world as it is, and not possible alternative universes. ⁸⁴ As for “conceptual” possibility or impossibility, it should also be limited, as I have been arguing throughout this essay, by the same kind of institutional theory that is supposed to animate political liberalism as a whole, and to mark it out as different from classical liberalism. For, recall that the Law of Peoples is meant to partake of a “realistic utopia.” This “realism” involves acceptance of the “actual laws of nature” and “constitutional and civil laws as they might be,” but also “that its first principles and precepts be workable and applicable to ongoing political and social arrangements.” ⁸⁵ Rawls gives the example of “primary goods,” such as income, and argues that they should be “workable,” meaning that one’s share of them “is openly observable and makes possible the required comparisons between citizens.” This leads him to reject such “unworkable ideas as a people’s overall utility, or . . . [Amartya] Sen’s basic capabilities.” ⁸⁶ Note here that Sen’s capabilities approach is not problematic, for Rawls, at the conceptual level; it is simply “unworkable.” Why should the same not be said of Rawls’s own idea of decency? The forgoing also suggests that it would not have been possible for Rawls not to attribute the modern state form (and modern science) to decent societies. For this would have posed difficulties for his global justice theory not simply in what concerns decent societies, but also in what concerns his account of the Society of Peoples and even of the liberal peoples of which they are a part.

In short, even if decent societies are only imaginary, conjecture about them should be constrained by the commitments of political liberalism to be political and of political philosophy to a realistic utopia. No amount of imagination or “hypotheticals” can succeed in the dual task of offering an institutional account of the rise of liberal democracies and of the global state system (including trends like migration, power politics, and the formation of a “Society of Peoples”) while at the same time positing free-floating, non-Western alternatives.

V. Roads Foreclosed

Instead of pure philosophical imagination, it is history—as well as sociology, politics, and social theory—that should have informed political liberalism’s approach to the nonliberal, non-Western world, in the same way that it has informed, to the extent that it has, political liberalism’s approach to the liberal world. ⁸⁷ What if, for example, Rawls had mentioned colonialism as a fourth historical development that “deeply influenced the nature of moral and political philosophy”? ⁸⁸ Colonial ventures began during the same early modern period that interests Rawls and that gave rise to the modern state, modern science, and the Reformation. It is telling that Rawls writes that, “Unlike most Muslim rulers,” the rulers of Kazanistan “have not sought empire and territory.” ⁸⁹ This is telling not only because, as Norton puts it, expansion is posited here as “the Muslim norm” (from which Kazanistan deviates). ⁹⁰ It is also particularly noteworthy because “empire and territory” do not come up in Rawls’s account of Western history relayed above; the persecuting zeal mentioned in relation to Christianity relates primarily to intra-Christian conflicts.

Where the imperial ventures of European nations in the eighteenth and nineteenth centuries come up is not the discussion of Western history, nor the account of decent societies, but the section of The Law of Peoples on liberal peace. ⁹¹ Rawls also mentions there the U.S. interventions in Chile, Guatemala, Iran, and Nicaragua. ⁹² His main conclusion from these observations is simply that the United States and other European nations fall short as constitutional democracies. The relationship between these imperial ventures and the political forms of government available for nonliberal societies to adopt is not relevant to the discussion. ⁹³ The topic of imperialism thus ironically only arises as a matter relevant to the relationships among (quasi) liberal states, not between these and the rest of the world.

Capitalism is another fundamental historical development that can be mentioned here, as is globalization. ⁹⁴ But besides these longue durée historical developments, what should have been of direct relevance for understanding the nonliberal world is its actual, contemporary shape—which would have formed the basis for idealization. It is telling that there is no indication in his files that Rawls read about, or was interested in, the non-Western world beyond Islam. His “Hiroshima” folder, for example, is only concerned with the nuclear bomb and does not reveal any particular interest in the history or politics of Japan. The folders on “nationalism” and on “civil wars” only include material related to Europe and the United States. The “empire” folder concerns Russia and Britain. ⁹⁵

What Rawls read about and was interested in was Islam, particularly Islamic history and jurisprudence. ⁹⁶ This would be on par with Rawls’s interest in Protestantism—and with a general interest that Forrester identifies among theorists of international justice in the 1980s in a “distinctive realm of culture and value.” ⁹⁷ But to read about Islam alone, especially in its textual and historical form, in order to understand modern “Islamic” societies—let alone the East more generally—is to commit the mistake of cultural essentialism. ⁹⁸ What is especially important for my purposes in this article is that, as I have argued above, this mistake was not committed in describing liberal democracies, even in their idealized form.

Instead of searching in Islamic jurisprudence for ways to conceptualize the nonliberal world, Rawls might have more plausibly read about the contemporary political organization of non-Western societies and about political struggles as they unfolded in the non-Western world. ⁹⁹ This would have allowed him to remain faithful to his institutional approach even as he extended his gaze beyond Europe and North America.

***

In his letter to Applbaum, Rawls writes, commenting on his search for an adequate account of a just consultation hierarchy:

I worry that we are too likely to think that our view must be right and so there must be another equally sound alternative. Why shouldn’t there be more than one kind of just society? Is it a lack of our imagination? ¹⁰⁰

There is indeed a problem with liberal imagination when it comes to the non-Western world, but the problem is not, as Rawls identifies it, the lack of it. The problem may, in fact, be one of too much imagination, untethered to history and politics. To imagine nonliberal alternatives, one needs to understand the nonliberal world first—its practices, its institutions, its disagreements, and the histories of colonialisms that were central to its formation. This is an especially pressing requirement for political liberalism, as its claim to distinctiveness is that it starts not from abstract ideals, metaphysical assumptions, or cultural preferences but from the world as it operates; its utopia starts in the reality of existing social and political arrangements. Failing this, liberalism’s political variant appears no more political, and no less parochial, than classical liberalism.