PART THREE—
SUBSTANTIVE COMMITMENTS, PARTIAL INTERPRETATIONS, AND IMPERFECT JUSTICE

Chapter Seven—
Partiality and Comprehensive Pluralism:
Strategies against the Mutual Eradication of Individual and Community

Systems theory, discursive proceduralism, and pragmatism have all failed to justify proclaiming an end to interpretation. Accordingly, it now becomes imperative to return to interpretation in a last-ditch effort to determine whether the crisis in interpretation may be nonetheless overcome and whether justice according to law might be after all susceptible to reconciliation with justice beyond law. In this last part of the book, I explore whether there might be a path toward just interpretations that might steer clear of the many obstacles encountered thus far.

The challenge of clearing a path to just interpretations is daunting. Not only must a suitable way be found to close the gap between self and other without undermining the diversity that stems from the coexistence of a plurality of conceptions of the good; but this must be done in a way that fully accounts for the conclusions reached throughout the preceding analysis, including those concerning contingency, lack of a workable neutral moral standpoint, impossibility of successfully relying on pure proceduralism, and impracticability of completely severing law from ethics or politics. These obstacles notwithstanding, I believe that a path to just interpretations may be found through reliance on pluralism—or more precisely, on pluralism understood in substantive normative terms. I call the latter kind of pluralism "comprehensive pluralism." I define it and indicate how I believe it might open a path toward just interpretations in the remainder of this chapter. In chapter VIII, I close the inquiry into just interpretations by focusing on significant practical applications of interpretive norms derived from comprehensive pluralism.

I—
From Pluralism in Fact to Pluralism as Norm: The Road to Comprehensive Pluralism

Throughout this book, I have referred to pluralism to denote societies whose members adhere respectively to a multiplicity of diverse conceptions of the good. Such societies experience "pluralism in fact." The fact of pluralism taken by itself, however, does not tell us anything about how we ought to deal with conflicts arising within pluralistic societies, including conflicts regarding just interpretations. Depending on one's normative commitments, there are many possible ways of approaching conflicts within such societies. For example, one might prefer to tackle the problems of a pluralistic society in a way that maximizes freedom, or equality, or that best conforms to the fundamental tenets of a majority religion. There is, however, one plausible normatively grounded approach to these problems that stands out as different from the rest. This approach is predicated on the conviction that pluralism itself is desirable and should therefore be endorsed and promoted. According to this approach, which I call "pluralism as norm," the conflicts in a society that is pluralistic in fact ought to be handled in a way that is designed to preserve and enhance pluralism. Furthermore, the thesis I defend in this chapter is that normative pluralism—or, more precisely, comprehensive pluralism—affords the best possible means to deal with the conflicts within a society that is in fact pluralistic, consistent with an ethic of reconciliation between self and other and with an eye to minimizing the violence inflicted on the self or the other. This thesis, which can be encapsulated in the slogan "pluralism as norm is best for pluralism in fact,"^[1] can be summarized in its essentials as follows.

Legal interpretation cannot avoid reliance on substantive normative commitments, and that precludes genuine neutrality as between competing conceptions of the good. Comprehensive pluralism, in turn, is a conception of the good, but while admittedly non-neutral, it is different from other such conceptions inasmuch as it is dependent on the latter for its own survival. Indeed, the aim of comprehensive pluralism is to encompass, and to foster peaceful coexistence among, as many competing conceptions of the good as possible. Moreover, comprehensive pluralism must be prepared to accept the norms produced by other conceptions of the good, but only to the extent that such norms do not interfere with its encompassing design.

[1] Normative pluralism could be pressed further to extend even to homogeneous societies. Thus, one could claim that promoting the values of pluralism, tolerance, and diversity would be beneficial to a homogeneous society. Although the position I defend is not necessarily inconsistent with this claim, I remain agnostic as to whether it is desirable to extend normative pluralism beyond factually pluralist settings. In any event, given the broad conception of factual pluralism that I adopt, all modern constitutional democracies would qualify as being pluralistic in fact.

Accordingly, the normative apparatus associated with comprehensive pluralism's integrating mission can be conceived as consisting of second-order norms that are distinguishable from first-order norms—that is, all the other norms associated with one or more of the remaining conceptions of the good. In short, comprehensive pluralism's principal aim is to negotiate the tension between first-order and second-order norms without thereby compromising the latter, all the while remaining as inclusive as possible with respect to the former.

From the standpoint of legal interpretation, comprehensive pluralism requires the legal interpreter to be guided by second-order norms while urging those who submit claims calling for legal interpretation to remain as true to the first-order norms to which they are committed as is possible consistent with genuine acceptance of the dictates of legitimate judicial adjudication. Moreover, even allowing for the fact that every person can be both legal claimant and legal interpreter and that, as we shall see below, promotion of second-order norms is likely to have spillover effects on the evolving configuration of first-order norms, comprehensive pluralism maintains a split between demands for justice predicated on first-order norms and legitimate endeavors to dispense justice, which must look, above all, to second-order norms. In short, to the extent that comprehensive pluralism amounts to yet another conception of the good, it is clearly distinguishable from contemporary Kantian visions such as those of Habermas or of Rawls in A Theory of Justice . On the other hand, to the extent that comprehensive pluralism is more than yet another conception of the good, in that it unleashes an inevitable but ultimately unresolvable dialectic between first-order and second-order norms, it differs markedly from theories based on building an "overlapping consensus," such as that elaborated by Rawls in Political Liberalism, and from all theories ultimately reducible to mere relativism.

Before further exploring comprehensive pluralism, it is necessary to provide a somewhat fuller account of the distinction between "pluralism in fact" and "pluralism as norm." The picture drawn of pluralism in fact thus far is one of a society composed of diverse ethnic, religious, cultural, or ideological groups who do not see eye to eye concerning at least some fundamental ethical, legal, and political issues. To complete this picture, in terms of a phenomenology of pluralism—conceived as the product of any plausible division of the relevant policy along the axes of self and other—reference must be made to pluralism among individuals. Indeed, individuals within a polity may have different values and objectives, or, in the words of Rawls, different "plans of life" (1971, 92-94). Moreover, whereas an individual standing alone may not be able to develop a full-fledged conception of the good, he or she may certainly decide to opt out of an existing conception, thus lending an individual dimension to the conflict between

conceptions of the good. Consistent with this, we can distinguish between group and individual pluralism in fact or between "communal pluralism" and "individualistic pluralism." Furthermore, so long as both of these are incorporated within the definition of pluralism in fact, all contemporary constitutional democracies can be deemed to be in fact pluralistic.

Although communal pluralism and individualistic pluralism are conceptually distinct, they need not actually conflict with one another. Individuals do not customarily live apart from communities, and virtually never without significant communal ties. As a matter of fact, individualistic pluralism can often be protected through endeavors to safeguard communal pluralism and vice versa. For example, an individual's interest in practicing his or her religion may be adequately protected through group rights accorded to every religion within the polity. Conversely, a religious group may find sufficient protection in individual rights to freedom of assembly and freedom to worship. It follows from this, that the distinction between individualistic and communal pluralism only becomes crucial when group rights cannot be recharacterized as individual rights and vice versa, and that occurs most notably when the individual becomes alienated from his or her own group. For example, if a group's integrity and survival depended on maintaining a prohibition against divorce, and if a married member of that group wanted to divorce another such member, then either the group would prevail under a collective right to survival as a group or the individual would prevail under an individual right to divorce. In this case, however, it would be impossible to recharacterize the collective right involved as an individual right or vice versa. For one committed to promoting pluralism, that would mean that a choice would have to be made between furthering communal pluralism at the expense of individualistic pluralism or vice versa.

Communal pluralism can be conceivably destructive of individualistic pluralism and vice versa, but actual tensions in a pluralistic contemporary polity are more likely to be reflective of the tendency of individuals to have affiliations with a number of distinct groups, which may be in conflict with one another. Consider, for instance, the difficult group identification issues that may confront a contemporary African-American woman in the United States. As a woman, she may be drawn to feminism and be critical of all forms of sexism, including those prevalent among certain African-American men. As a member of a group that has long been the victim of violent and pervasive racial hatred, however, an African-American woman may be loathe to further fuel the fires of racism as a consequence of acting on her feminist convictions.^[2] On the other hand, a racist woman who also happens to be a feminist may be more prone to identify with men of her own race

[2] For an extended discussion of the dilemma confronting black feminists in the United States, see Crenshaw 1993.

than with women of other races when it comes to dealing with any issue in relation to which she regards race as a factor. More generally, the split into self and other, which constitutes an essential feature of pluralism in fact, should not be regarded as necessarily fixed or permanent. More probably this split is multiple, fragile, and context dependent. Accordingly, this split is better understood in relational terms as a consequence of alliances and divisions, prone to shift over time, and likely to intensify or abate depending on the political climate or particular issues involved (Rosenfeld 1994, 4,13-14).^[3] In short, in a contemporary setting, pluralism in fact is most likely to reflect the citizen's many evolving ties to a multiplicity of different types of communities as well as the sense that each citizen has that he or she is entitled to carve out a personal path to happiness and fulfillment.

In addition to there being a split between self and other between individuals, between the individual and the group, and between different groups, there are also likely to be manifestations of such a split within the individual (Lacan 1966, 655, 839-40; Rosenfeld 1995, 1056-59). Furthermore, consistent with pluralism in fact, individuals and groups may at times embrace contradictory value preferences or lack a coherent conception of the good. What is more, they may partially buy into many different conceptions of the good, or even be unaware of some of their allegiances or value preferences.

These difficulties arising at the level of the individual or at that of an actual group are likely to disappear at the level of society as a whole, and are certainly overcome in the context of counterfactual reconstruction. In other words, whereas it may be impossible at any given moment to determine who can be counted as an adherent to a particular conception of the good, or how much a person may be committed to the conception in question, it should be readily apparent which are the main competing conceptions of the good within the polity as a whole. Similarly, from the standpoint of counterfactual reconstruction, what is of foremost importance is to construct a picture that is representative of actual conflicts and potential avenues of resolution rather than a detailed depiction of the actual predicament of every person within the polity.^[4]

[3] This is not to deny that some splits along self and other can become rigid and endure over several centuries, as history has painfully demonstrated in several cases. The point is rather that the latter type of split ought to be regarded as the exception rather than the rule from the standpoint of a phenomenology of functioning constitutional democracies that are pluralistic in fact.

[4] Whereas no quantitative method is available to sort out what is purely idiosyncratic from what has broader intersubjective significance, the context should usually make matters sufficiently clear. Thus, for example, an individual's obsessive quest for simultaneous fulfillment and violation of a religious duty could be safely ignored in the process of counterfactual reconstruction. In contrast, a widespread tension among a significant segment of the popula-tion arising from a simultaneous partial adherence to a religious and to a secular ideology would certainly appear to warrant being taken into proper account.

In light of the preceding observations, a few further remarks concerning conceptions of the good, value preferences, perspectives, vantage points, and interests are required for purposes of rounding out the broad outlines of pluralism in fact. As will be remembered, a conception of the good is a particular ethic, and, at least from the standpoint of counterfactual reconstruction, it amounts to a comprehensive integrated perspective. A conception of the good may be primarily religious, ethnic, cultural, or ideological. Thus, for example, Catholicism, Zionism, Marxism, and John Stuart Mill's liberalism all yield their own distinct conception of the good. Moreover, the integrated perspective framed by a conception of the good prescribes certain value judgments and certain value preferences. As Habermas emphasizes, value preferences are intersubjective, and they derive from the ethical norms associated with the particular conception of the good of a given ethical community (1996, 255-56).

A value preference may be grounded in more than one conception of the good. For example, both a religious conception of the good and a capitalist ideology revolving around the family as the optimal unit of production and consumption may generate a value preference for the zealous preservation of traditional family life. The contours of a given value preference may be different, however, depending on which of the numerous conceptions of the good susceptible of affording it adequate grounding is relied on by particular proponents of that value preference. Moreover, differences in the contours of value preferences tied to different conceptions of the good are most likely to emerge in the context of conflicts or rankings among diverse value preferences. Two important consequences follow from this: first, the full import of an asserted value preference can only be adequately gauged from its relation to the conception of the good on which it is actually grounded; and second, particular value preferences can furnish areas of convergence or overlap for proponents of different conceptions of the good.

Like value preferences, ethical norms are susceptible of being grounded in more than one conception of the good. Actually, certain ethical norms, such as the prohibition against murder, appear to be embraced by all conceptions of the good. However, which homicides qualify as murder is likely to vary depending on one's conception of the good. Thus, for instance, certain abortion opponents consider abortions to be murder and the killing of physicians who perform abortions, justified homicide. Supporters of abortion, on the other hand, regard such killings of physicians as murder while maintaining that abortion in no way involves murder. Consistent with

this, moreover, even universal embrace of the prohibition against murder as an ethical norm does not convert the latter into a moral norm (transcending all conflicts among conceptions of the good) in any but a purely formal, and hence vacuous, sense. Moreover, agreement concerning an ethical norm, such as the prohibition against murder, coupled with disagreement over what the ethical norm in question actually proscribes, may well be the source of some of the most intense conflicts among proponents of rival conceptions of the good.

It is only when it is grasped from the perspective of the conception of the good to which it happens to be linked that the full import of an assertion of a value preference or ethical norm can be properly evaluated. Furthermore, whereas a developed conception of the good frames a comprehensive integrated perspective, to the extent that such conception promotes role differentiation, its adherents will most likely embrace respectively different vantage points. For example, even if labor and management both subscribe wholeheartedly to a capitalist ideology, the vantage point of labor would still differ in significant respects from that of management.

Within this framework, interests loom as flexible, adaptable, and transformable. Informed by needs and desires, interests can be molded by conceptions of the good and vantage points, through education and other intersubjective dealings. Moreover, particularly in dynamic pluralist settings characterized by a high level of intercommunal interaction and fluid boundaries between internal and external others, interests are also likely to assume an active role in the construction and transformation of conceptions of the good and vantage points. Within the dialectic dynamic of a constantly evolving pluralist setting, interests can serve both to guide their proponents' participation within established games with fairly well defined rules and to prompt their proponents to seek new games or to change the rules of existing games. In other words, in the absence of any neutral moral standpoint, the pursuit of interest presumably involves both maximizing one's advantage within the existing institutional arrangement and promoting institutional changes that would optimize one's chances to realize one's objectives. For example, consistent with the discussion of the feminist challenge to Habermas's discursive proceduralism in chapter V, in a setting with rules devised by men, feminists can strive both to better the lot of women within existing rules and to press for rules that are better suited to women's interests.

Although conceptions of the good, value preferences, vantage points, and interests are never likely to remain rigid or immutable within a setting that is pluralist in fact, their actual degree of elasticity ultimately depends on the kind of pluralist polity involved. Greatest elasticity is likely to occur under "melting pot" conditions in highly interactive societies typified by a

broad-based intertwining of individualistic and communal pluralism. At the other end of the spectrum, least elasticity is likely to occur in societies that are predominantly communally pluralistic and that experience a very low level of intercommunal interpenetration.

In the broadest terms, pluralism as norm, to which we now turn, stands for the proposition that pluralism in fact is good and that it ought therefore to be encouraged and protected. Normative pluralism is distinguishable from both monism—roughly defined as the view that there is a single conception of the good that is correct and that all value preferences are to be judged in terms of that conception—and relativism—the view that all value preferences are ultimately purely subjective and so contextually bound to the conception of the good from which they emerge that it would be meaningless to seek to gauge them from the standpoint of any other perspective.^[5] In other words, against monism, normative pluralism holds that the good extends beyond any single conception of the good; against relativism, that not all conceptions of the good are equivalent as the mere projections of contingent perspectives. As we shall see below, normative pluralism can be viewed as mustering relativistic tendencies against monism and monistic tendencies against relativism while carving out a path that remains distinct from both monism and relativism.

Before pursuing these matters further, there are a couple of distinctions that must be briefly addressed. The first of these is that between "methodological pluralism" and "substantive pluralism"; the second, that between "limited pluralism" and "comprehensive pluralism." Methodological pluralism can be characterized as a strategic tool of limited scope, whose main function is to combat the tendency to hegemony of conceptions of the good vying for dominance in settings that are pluralistic in fact. As a strategic tool, methodological pluralism may embrace any position that might help weakening the grip of the most dominant conceptions of the good. Accordingly, methodological pluralism may align itself with relativism or with particular conceptions of the good positioned as significant rivals of the dominant ones. In short, methodological pluralism encompasses all weapons and strategies that may be profitably put to use for purposes of undermining monism.

Substantive pluralism, on the other hand, embraces a particular conception of the good that prescribes inclusion and accommodation of as large a plurality of conceptions of the good as possible. Consequently, substantive pluralism is definitely antirelativistic inasmuch as relativism would be indifferent as between inclusion or exclusion of a plurality of conceptions of the

[5] For a more nuanced, thorough, and extended discussion of the relationship among pluralism, monism, and relativism, see Kekes 1993, 13-14, 34-35, 118-38.

good. In its antirelativism, however, substantive pluralism appears to come perilously close to becoming monistic in its own right. Indeed, like monistic conceptions of the good, substantive pluralism asserts its superiority over its rivals. Nevertheless, as will become clear below, substantive pluralism remains unlike its monistic rivals in that it both acknowledges the value of other conceptions of the good and cannot dispense with them.

Substantive pluralism may be pursued with moderation or systematically with unrelenting determination. In cases of moderate pursuit, we have what I refer to as "limited pluralism," whereas in cases of systematic pursuit, we have "comprehensive pluralism." Both limited and comprehensive pluralism embrace the normative objectives of substantive pluralism. Comprehensive pluralism, however, is much more radical than its limited counterpart, in that it seeks to level all existing hierarchies among conceptions of the good. Limited pluralism, on the other hand, is prepared to accept existing hierarchies, or to recommend relatively modest changes, while mainly preaching tolerance of a plurality of conceptions of the good and peaceful coexistence among proponents of different such conceptions. In short, comprehensive pluralism calls for equalization of all conceptions of the good, whereas limited pluralism mainly aims for acceptance of a plurality among such conceptions.

Actual constitutions in established democracies tend to promote limited pluralism inasmuch as they make room for tolerance without dislodging certain deeply entrenched traditions. Comprehensive pluralism, on the other hand, is unlikely to be embraced in actual constitutional practice owing to its radical implications but looms as an apt normative standard for counterfactual reconstruction. In other words, comprehensive pluralism affords a critical ideal that allows for a principled determination of how a polity that happens to be pluralistic in fact might be better stirred toward the objectives of substantive pluralism.

Consistent with the preceding observations, the thesis I promote—that pluralism as norm is best for pluralism in fact—can now be further specified as follows. Comprehensive pluralism as a critical counterfactual ideal yields the best possible normative criterion for the reconciliation of self and other within a pluralistic in fact society, in a way that maximizes the potential for justice while minimizing that for violence. Thus, equalization, and inclusion, of all conceptions of the good furnishes the normative yardstick against which the call for justice, the reach to the other, and the quest for just interpretations must be set. But before going any further, it is necessary to deal with an apparent contradiction that lurks within the very core of the ideal of comprehensive pluralism. Indeed, the requirement of equalizing all conceptions of the good seems squarely at odds with that of being all-inclusive. Or, put somewhat differently, how can substantive pluralism's

call for inclusiveness prevail over rival claims for exclusiveness when all different conceptions of the good are supposed to be placed on a strictly equal footing?

II—
The Dialectics of Comprehensive Pluralism

The apparent contradiction between pluralism's thrust toward equalization and its commitment to inclusiveness is reminiscent of the paradox of tolerance discussed in chapter VI. Just as tolerance of the intolerant ultimately jeopardizes tolerance, equalization of all conceptions of the good eventually casts pluralism and antipluralism as equally legitimate. Moreover, if that proves to be the case, then comprehensive pluralism dissolves, in the last analysis, into relativism. On the other hand, if comprehensive pluralism places itself above rival conceptions of the good—in ways that are analogous to Popper's suggested resolution of the paradox of tolerance through intolerance of the intolerant—then comprehensive pluralism may, in the end, prove to be but monism in disguise. In sum, either pluralism is limited or it seems impossible for it to be genuinely pluralistic.

Further inquiry reveals, however, that comprehensive pluralism need not be relegated to a monistic or a relativistic fate, provided it is properly understood in terms of the dialectic that it unleashes. Indeed, viewed as a counterfactual ideal rather than as a realizable end-state, comprehensive pluralism serves to launch a coordinated two-pronged attack against both hierarchy and exclusion. As already indicated, comprehensive pluralism figures, in part, as a conception of the good that claims superiority over its rivals, but only for the limited purpose of minimizing exclusion of other conceptions of the good. Conversely, comprehensive pluralism's systematic leveling of conceptions of the good does introduce some measure of relativism, but it is a limited and narrowly targeted one, whose only aim is to undermine the pretensions to superiority of certain conceptions of the good. Strictly speaking, therefore, comprehensive pluralism is not relativistic as between conceptions of the good; it is merely skeptical concerning any claim to a hierarchy among them.

To better grasp the dialectic that gives shape to comprehensive pluralism, we must refer again to the distinction between the norms emanating from comprehensive pluralism as a conception of the good and those deriving from other conceptions of the good. The former, as we have seen, constitute second-order norms, whereas the latter amount to first-order norms.^[6] Accordingly, the privileging of comprehensive pluralism as a conception of the good without falling into monism can be recharacterized as follows:

[6] This distinction is analogous to that drawn between first-order and second-order ends in chapter VI.

privileging comprehensive pluralism over other conceptions of the good boils down to affording second-order norms priority over first-order norms since that is a prerequisite to achieving equality among first-order norms.

As a counterfactual ideal, comprehensive pluralism becomes operative in settings marked by historical contingency and perspectival partiality. Thus, it is largely a historically contingent matter which first-order norms are favored and which disfavored or even banished at any particular time and place. Similarly, prevalent conceptions of the good and the nature of the interaction among them inevitably shape and constrain the scope of actual and potential perspectives capable of playing a relevant role in a given spatiotemporal setting. Consistent with this, moreover, the aim of comprehensive pluralism is to overcome the very contingency and partiality with which it is confronted, for purposes of clearing a path leading to equality among first-order norms. Specifically, comprehensive pluralism confronts the status quo through a dialectical process that involves two distinct logical moments that combine to delimit the quest for equality among first-order norms.

Set against a competition among a multiplicity of first-order norms vying for predominance, comprehensive pluralism's first logical moment is a negative one characterized by a strict refusal to endorse or favor any of the competing first-order norms. Thus, in its negative moment, comprehensive pluralism imposes strict equality and neutrality among all existing first-order norms and the conceptions of the good from which they derive.

Carried to its logical conclusion, however, comprehensive pluralism's first moment leads to self-destruction. If all first-order norms are completely neutralized through a leveling negation, then the very pursuit of pluralism would become meaningless. In the absence of a plurality of viable conceptions of the good, no first-order norms would remain for pluralism to protect. Accordingly, to avoid self-destruction, comprehensive pluralism must supplement its negative moment with a positive one. The object of that positive moment is to foster readmittance of previously leveled and equalized conceptions of the good into the pluralist universe.

In its positive moment, however, comprehensive pluralism must confront a major problem. Not all conceptions of the good excluded in the course of comprehensive pluralism's negative moment can gain readmission in its positive moment. For example, a crusading religion, for which conversion of the infidel, by force if necessary, is a sacred duty that admits of no exceptions, has no place under comprehensive pluralism. Moreover, even those conceptions of the good that can be slated for readmission cannot occupy the same position under comprehensive pluralism as they did prior to its deployment. Thus, religions that depend for their survival on radical intolerance can only be readmitted on condition that they pose no serious threat to other religions or to nonreligious conceptions of the good. This

could be accomplished, for example, by relegating readmitted religions to the private sphere.^[7]

Because comprehensive pluralism cannot equally readmit in its positive moment all the conceptions of the good that it has equally excluded in its negative moment, it inevitably falls short of its ideal of equal accommodation for all conceptions of the good. At best, comprehensive pluralism can undertake to better approximate equality among all first-order norms without ever reaching its goal of' providing for full equality among all first-order norms. Moreover, because of this, comprehensive pluralism, once unleashed, becomes engaged in a ceaseless dialectic marked by a constant succession of negative and positive moments, without ever reaching a final resting points.^[8] Indeed, since at the completion of every positive moment deriving from comprehensive pluralism some first-order norms are altogether left out and others included but displaced, there are bound to be calls for greater inclusiveness and equality, and hence a need for further negative leveling. Furthermore, the constant leveling and repositioning of first-order norms not only affect their location in relation to the center of gravity of the relevant normative universe but also make them more susceptible to internal alterations and reconfigurations.^[9]

There is a strong parallel between comprehensive pluralism and comprehensive justice in that they both underscore the ever present opportunities for amelioration within settings that can never fully escape from imperfection. Although comprehensive pluralism cannot eradicate uncertainty or anxiety, it does nurture hope of progress and of better and more mutually enriching avenues of reconciliation between self and other. But before attempting to make the case for comprehensive pluralism any further, it is necessary to focus on a couple of important consequences that follow from its logical structure.

The juxtaposition of comprehensive pluralism's two logical moments underscores the asymmetry between its negative and its positive moment. In its negative moment, comprehensive pluralism reaches for, and achieves, radical equality. Moreover, such equality is reached procedurally as existing hierarchies among first-order norms are routinely struck down. In terms of

[7] Cf. Marx 1967 (arguing that religious emancipation can only be obtained at the cost of relegating religion to the private sphere).

[8] It should be emphasized that the succession in question is strictly a logical one and not necessarily a temporal one. Logically, every negative moment must be followed by a positive one and vice versa. Historically, the leveling and reinstating functions may be amalgamated or out of logical sequence provided they remain susceptible to reconstruction in the proper logical order.

[9] The susceptibility of first-order norms, value preferences, and conceptions of the good to internal changes under the aegis of comprehensive pluralism will be explored in greater detail in the course of the next chapter.

counterfactual reconstruction, comprehensive pluralism, in its negative moment, pictures all first-order norms as strictly equivalent and through projection of this construct against prevailing hierarchical arrangements, fashions a powerful yet interpretively unproblematic means of critique. For example, in the context of a religiously pluralist society in which the state has singled out a particular religion for purposes of official endorsement, comprehensive pluralism's negative moment makes for a simple, immediate, obvious, and swift critique.^[10] Furthermore, comprehensive pluralism's achievement of complete (albeit purely formal and fleeting) equality in its negative moment affords a systematic way to characterize the difference between limited and comprehensive pluralism. Indeed, unlike its comprehensive counterpart, limited pluralism fails to fully level all first-order norms in the course of its negative moment. In other words, limited pluralism fails to distinguish consistently and systematically between first-order and second-order norms. In contrast, in its positive moment, comprehensive pluralism aims at equality, but equality at the end always eludes its grasp. Accordingly, the best that can be hoped for is that in its positive moment, comprehensive pluralism will eradicate some inequalities while leaving others in place. Or more precisely, taking into account that repositioning first-order norms may lead to changes within their internal configuration, comprehensive pluralism is poised in its positive moment to wipe out certain inequalities but only at the cost of triggering others. Furthermore, since reintegration of first-order norms in the course of its positive moment is not automatic, but instead contingent on their compatibility with second-order norms and with other first-order norms, comprehensive pluralism's positive moment, unlike its negative moment, is not reducible to proceduralism. Also, not only must comprehensive pluralism in its positive moment fall back on substantive norms, it cannot avoid reliance on contestable interpretive practices. Indeed, application of substantive criteria to determine which first-order norms ought to be readmitted is not reducible to purely formal or quantitative procedures.

Since comprehensive pluralism in its positive moment can neither fully reconcile legal equality with factual equality nor obviate the problem of just interpretations, it seems fair to ask whether it is altogether worth pursuing. A positive answer to this last question seems reasonable, provided one makes a persuasive case concerning the priority of second-order norms

[10] This does not mean that comprehensive pluralism automatically condemns state endorsement of a particular religion regardless of the circumstances. Ultimately, any legitimate verdict under comprehensive pluralism must fully account for both its negative and its positive moment. Although the circumstances that would warrant this might be exceedingly rare, it is not inconceivable that in its positive moment comprehensive pluralism would reinstate a state religion to the position it held prior to its demotion in the course of comprehensive pluralism's negative moment.

over first-order norms and provided reliance on the dialectic promoted by comprehensive pluralism significantly increases the chances for greater inclusiveness and reduced inequities. In short, going through comprehensive pluralism's positive moment seems justified so long as it leads to a more inclusive and more equitable accommodation of diverse first-order norms. Moreover, the inner reconfiguration of first-order norms in the context of comprehensive pluralism is only objectionable if it limits rather than expands opportunities, and if it constrains rather than broadens the scope for self-fulfillment and reconciliation with the other.

Comprehensive pluralism must rely in its positive phase on substantive norms and contestable interpretations, but these need not produce interpretive hurdles leading to arbitrariness or excessive indeterminacy. Indeed, the priority of second-order norms mandates certain clear-cut results, such as the repudiation of actively antipluralistic conceptions of the good and limited acceptance of less activist ideologies that are nonetheless intolerant. In addition, in other cases, the priority of second-order norms can provide guidance without fully determining the most desirable outcome.

More difficult situations arise when the pursuit of one conception of the good that is not inherently objectionable impedes the pursuit of another such conception. However, even in those cases, satisfactory solutions may often be available, through the deployment of a reversal of perspectives along the lines contemplated by Habermas.^[11] Thus, after being considered successively from the standpoint of each of the perspectives involved, conflicting pursuits could be ranked according to their respective importance in relation to the particular conception of the good that provides them with normative grounding. In some cases, such rankings would establish clear orders of priority that should prove equally acceptable to all proponents of the reversal of perspectives test. Moreover, inasmuch as such a test leads to greater inclusiveness of the other without undue sacrifice of diversity, it ought to include comprehensive pluralists among its most loyal supporters. Because of the lack of a neutral perspective transcending all communal biases, the reversal of perspectives consistent with comprehensive pluralism cannot be expected to lead to a determinate resolution of all difficult cases. Nevertheless, by supplementing the priority of second-order norms, the reversal of perspectives enables comprehensive pluralism to chart a distinct course toward a better integration of an increasingly diverse range of first-order norms. Further discussion of the interpretive implications that follow from this will be postponed until the next chapter, but two important points deserve mention at this point. First, all interpretation deriving from comprehensive pluralism is bound to be intersubjective and dialogical rather than monological. Second, all such interpretation is not

[11] See chapter V, above.

only intersubjective but also intercommunal in its reach, as it takes into serious account all others and all communities within the polity.

To recapitulate: the promise of comprehensive pluralism is considerable. From the standpoint of its negative moment as a counterfactual construct, comprehensive pluralism furnishes a crisp and reliable means to criticize prevailing inequities. From the standpoint of its positive moment, on the other hand, comprehensive pluralism generates counterfactual models suggestive of ways in which existing inequities might be plausibly overcome and greater inclusiveness and diversity achieved. These characteristics certainly make comprehensive pluralism attractive. It remains to be seen, however, whether the claim that comprehensive pluralism offers the best hope for societies that are pluralistic in fact can be persuasively substantiated.

III—
The Case for Comprehensive Pluralism

To assess the virtues of comparative pluralism as compared to those of its most likely rivals, it is necessary to take a closer glance at what it stands for. Above all, by only claiming that comprehensive pluralism is best in the context of settings that are pluralistic in fact, it is not necessary to embrace the proposition that pluralism itself constitutes the ultimate good. This is certainly advantageous, for if pluralism itself were postulated as the summum bonum, then the case for comprehensive pluralism would become purely circular. This circularity is avoided, however, if the ultimate good is deemed to be the reconciliation of self and other in the least coercive and least confining manner possible. In that case, moreover, the strongest case for comprehensive pluralism would consist in proving that it is best suited to lead to the desired reconciliation.

By conceiving such reconciliation in the widely encompassing relational terms stressed throughout this book, it is possible to cast a far-reaching normative net that bridges most of the many divides long associated with polities that are pluralistic in fact. Indeed, inasmuch as self and other are conceived as cutting across individual and communal divides, as somewhat malleable and prone to evolve, and as capable of simultaneously expressing complex identities and multiple alliances, the traditional oppositions between the individual and the group and between liberals and communitarians tend to lose much of their grip. Because of this, as we shall see, comprehensive pluralism seems particularly well suited to harmonize the perspectives and vantage points of individual and group and of liberals and communitarians in a large number of situations.

Consistent with regarding the reconciliation between self and other as the ultimate good, autonomy, reciprocity, empathy, dignity, and diversity rank among the highest values and occupy the place of second-order norms within the perspective of comprehensive pluralism. Autonomy and dignity

are closely intertwined with the notion of self hood and occupy a prominent place in connection with the self's capacity for ethical choice and commitment. Reciprocity, on the other hand, cements the bonds between self and other, by prescribing that the other be recognized as another self (Hegel 1977, ¶¶ 178-96). Moreover, a distinction can be drawn between "mere reciprocity" and "reversible reciprocity."^[12] As a norm, mere reciprocity prescribes recognition of the other as possessing a perspective or vantage point, without concern for actual differences among perspectives or vantage points. In contrast, reversible reciprocity requires not only recognition that the other has a perspective but also empathy for the other, based on a proper recognition of differences in perspectives. Finally, diversity emerges as a key value to the extent that it makes for less confining and less coercive avenues toward reconciliation between self and other. Without diversity, reconciliation would only be possible through self-constraint to the point of yielding to the identity of the other, or through coercing the other to fit within the mold of the self's own image. With diversity, on the other hand, not only is there greater room for more satisfactory reconciliation between self and other, but every self can rely on a much wider array of choices toward self-fulfillment.

Comprehensive pluralism confronts two different kinds of rivals, which may be characterized respectively as "metaphysical" and "postmetaphysical." Much like Habermas's discourse ethics discussed above, comprehensive pluralism rejects the position of metaphysical rivals typified by comprehensive religious conceptions of the good. Specifically, comprehensive pluralism opposes its metaphysical rivals' claims to the truth and to hegemony, and subsumes them under the dictates of its second-order norms. From the perspective of comprehensive pluralism, metaphysically grounded religious and ethical systems are reduced to the level of first-order norms subject to exclusion to the extent that they are antipluralistic. Thus, metaphysical perspectives cannot be accepted on their own terms and are reduced to becoming one among many vehicles of self-expression and self-fulfillment. In other words, within the framework of comprehensive pluralism, metaphysical perspectives lose their grip on the ethical to become absorbed into the rich and varied aesthetics that gives expression to pluralistic ethics.

Comprehensive pluralism thwarts the intercommunal aspirations of metaphysical perspectives but does not seek to interfere with their intracommunal pursuit. Nevertheless, by uprooting all hierarchies among first-order norms, and by promoting the spread of avenues of intercommunal interaction, comprehensive pluralism vastly increases the pressures for internal reconfiguration of metaphysical perspectives. Accordingly, even if

[12] For an extended discussion of reciprocity and of the progression from mere reciprocity to reversible reciprocity, see Rosenfeld 1991a, 242-49.

proponents of metaphysical perspectives are willing to withdraw from the public square, they are unlikely to prevent external influences from penetrating within their own community and from eventually prompting some of their fellow members to seek internal changes. For example, a religion with a clergy open only to men may withdraw from the affairs of the state and still face pressures from within its ranks for admitting women to its clergy as a consequence of the spread of feminist ideas. In short, for the militant proponent of a metaphysical conception of the good, comprehensive pluralism poses an unacceptable double threat: it relativizes a cherished and revered conception of the good, and it also exposes such conception to internal erosion.

Whereas there is very little common ground between comprehensive pluralism and its metaphysical rivals, matters seem quite different when it comes to the relationship between comprehensive pluralism and its postmetaphysical rivals. What unites all major postmetaphysical perspectives, including comprehensive pluralism, is the search for a reconciliation of self and other without reliance on metaphysics. Consistent with this broad definition, postmetaphysical perspectives span a spectrum that extends from liberalism to republicanism and to communitarianism and find expression in the views of such diverse thinkers as Locke, Rousseau, Kant, Mill, Marx, Habermas, Rawls, Sandel, Rorty, Raz, and Dworkin. Some of these views have already been examined and found wanting from the standpoint of just interpretations in the course of the preceding analysis. Accordingly, comprehensive pluralism can be deemed superior to the perspectives respectively emanating from the latter views, provided it leads to a reasonable resolution of the problem of just interpretations.

Pluralism has been frequently associated with liberalism, but the two are by no means coextensive.^[13] Although liberalism encompasses a wide variety of different views, and lacks any commonly accepted definition (Raz 1986, 1), it nonetheless seems inextricably linked to limited pluralism. Beyond that, however, at least some versions of liberalism rely on certain overriding values, such as liberty (ibid., 2), equality (Dworkin 1978, 125), or justice (Rawls 1971, 3-4), and thus appear incompatible with comprehensive pluralism.^[14] Furthermore, in contrast to liberalism, republicanism and communitarianism are not customarily linked to pluralism, and even, on occasion, are deemed altogether inconsistent with pluralism.^[15] This notwithstanding,

[13] See Kekes 1993, 199-217, for an excellent discussion of the relationship between pluralism and liberalism. The following discussion relies significantly on his analysis.

[14] Cf. Kekes 1993, 199 (arguing that "there are good reasons for supposing that pluralism and liberalism are incompatible").

[15] See, e.g., Sunstein 1993, 26-27, 38-39 (arguing that deliberative representative democracy predicated on republicanism is incompatible with "interest-group pluralism").

as we shall see, in the context of a pluralistic in fact polity, neither republicanism nor communitarianism is inherently incompatible with pluralism. Consistent with this, I will now argue that, in the presence of pluralism in fact, comprehensive pluralism is better suited to further the respective principal objectives of liberalism, republicanism, and communitarianism than the overriding values customarily associated with any of the three.

A—
Comprehensive Pluralism and Liberalism

In spite of the significant differences noted above, Anglo-American liberalism encompasses a family of conceptions of the good, which, for our limited purposes, can be distilled to the following essentials. The individual is the subject of moral choice, and the purpose of society is to optimize the opportunities for individual self-realization by striking an appropriate balance between individual autonomy and individual welfare. Within this framework, moreover, individual interests ultimately prevail over group interests, and groups—including the polity taken as a whole—conceived as associations or aggregations rather than as organic units are justified in terms of their suitability for advancing individual interests. To put it in terms of the distinctions drawn above, for liberalism, the polity is individualistically pluralistic in fact, and its principal objective is to promote the most extensive possible equal opportunity for individual self-realization.

From the standpoint of the present discussion, the most important division among proponents of liberalism is over how best to strike a balance between autonomy and welfare and how best to secure equal opportunity for individual self-realization. Thus, for example, libertarians insist that liberty is paramount, and trust that by maximizing liberty, individual self-realization will naturally follow.^[16] Liberal egalitarians, on the other hand, tend to stress that equality is paramount (Dworkin 1978, 115, 125). More generally, whatever a particular proponent of liberalism deems paramount as a function of personal theoretical convictions and prevailing contextual considerations, he or she is likely to postulate as an overriding value. As already mentioned, liberty, equality, and justice have been advanced as overriding values, to which can be added fundamental human rights (Berlin 1969, 165). Furthermore, the face of liberalism is likely to change depending on which values it embraces as overriding, and the possible variations are numerous, since liberalism may treat as overriding any one of these values, or certain combinations of them.

[16] The libertarian position is rooted in the philosophy of Locke (see Locke 1960) and counts among its best-known contemporary exponents, Robert Nozick (see Nozick 1974). For a more extended discussion of the libertarian belief that liberty leads to an optimal balance between autonomy and welfare, see Rosenfeld 1991a, 58-59, 224-25.

From the perspective of comprehensive pluralism, liberalism's key weakness lies in its need to lock in any of the above-mentioned values as overriding. Leaving aside any pluralistic critique of liberalism's individualistic bias, pluralism proves to be better suited to advance liberalism's chief objectives than any combination of liberal values enshrined as overriding. This may seem paradoxical at first but should become plain on further inquiry. Thus, one can recharacterize the liberal ideal as the reconciliation of the individual self and the individual other in a way that best maximizes their respective opportunities for self-realization. Moreover, it seems fair to assume that conceptions of one's individual good and of one's self-realization are likely to vary from one individual to the next, and that, consistent with the preceding analysis, no neutral Kantian moral perspective is available. Under these circumstances, a flexible approach—that is, the ability to appeal to different values to varying extents, depending on the material conditions and the conceptions of self-realization actually involved—would seem far preferable in terms of the requisite reconciliation than would unbending commitment to certain values taken as overriding. The liberal may object that overriding values are required as a prophylactic to prevent the self from trampling on the other. The pluralist would reply, however, that a pluralist mindset is the best guarantee against intentional interference by the self against the other. Thus, even if the same values happened to be stressed by pluralists and liberals, the pluralist approach would still be preferable, both because of its greater adaptability and because of its sharper focus on reconciling self and other in the most mutually satisfactory manner possible. Moreover, as conceptions of autonomy, of welfare, and of plausible ways to strike an acceptable balance between them are likely to evolve significantly over time, the superiority of the highly flexible pluralistic approach over its liberal counterparts should become increasingly apparent.

B—
Comparing Liberalism, Republicanism, and Communitarianism from a Pluralistic Perspective

The cases for pluralism's superiority over republicanism and communitarianism bear certain resemblances to but also certain significant differences from the case concerning the preferability of pluralism over liberalism. As in the case of liberalism, both republicanism and communitarianism encompass a broad array of different views, and respectively lack any commonly accepted definition. Unlike liberalism, which seems inextricably tied to limited pluralism, however, republicanism and communitarianism need not be pluralistic in any way. Thus, for example, classical republicanism based on the paramountcy of civic virtue was thought to "flourish only in small communities united by similar interests and by a large degree of ho-

mogeneity" (Sunstein 1993, 20). Similarly, communitarianism can invoke the image of a tightly woven organic whole encompassing a single community with a unanimously shared conception of the good.^[17]

Republican and communitarian visions steeped in conceptions of the polity as essentially homogeneous need not concern us to the extent that they take no account of pluralism in fact. That leaves only those republican and communitarian conceptions that purport to account for heterogeneity. Although critics have expressed skepticism (Fallon 1989, 1717-18, 1725-33), republicanism and communitarianism are not inherently incompatible with some degree of pluralism.^[18] Indeed, one could imagine a version of republicanism in which citizenship and civic virtue remain the principal focus, but in which the common good would account for differences in perspectives rather than remaining monolithic. In other words, the republican ideology under consideration would still place the civic duties of the public citizen ahead of the interests of the private person but would allow for differences over which civic duties and which configurations of public citizenship would best serve the common good of the polity. Furthermore, assuming communitarianism is ultimately distinguishable from liberalism and individualism because of its commitment to the primacy of the community over the individual, then communitarianism should be compatible with communal pluralism even if it precludes individualistic pluralism. Consistent with this, within a communally pluralistic communitarian setting, solidarity could function as the overriding value, provided it encompassed intracommunal bonds as well as intercommunal ones.

For purposes of considering the advantages of comprehensive pluralism over republicanism and communitarianism, it suffices to bear in mind the most salient features of the latter two. Accordingly, I gloss over differences among various republican and communitarian perspectives and concentrate on those among their respective principal features that accentuate their contrast with pluralism. Moreover, I treat republicanism and communitarianism as distinct, even though they often seem linked. Or, more precisely, whereas it is obvious that not all communitarians are republican, all contemporary versions of republicanism are arguably ultimately communitarian in nature. Be that as it may, the following discussion is predicated on acceptance of the possibility of noncommunitarian republicanism.

[17] Cf. Sandel 1982, 151 (contrasting the unity of community in a constitutive sense to the plurality envisioned by Rawlsian liberalism).

[18] There is a revival of republicanism among certain contemporary constitutional theorists in the United States. These theorists have sought to reconceptualize republicanism so as to take pluralism in fact into proper account. See Michelman 1988; Sunstein 1988.

C—
Comprehensive Pluralism and Republicanism

At the core of the republican vision is a conflict between private man (woman) and public citizen. This conflict shifts the center of gravity of the struggle between self and other from the realm of interpersonal relations to that of each person's inner universe. This shift, moreover, emerges particularly clearly in Jean-Jacques Rousseau's republican conception of self-government driven by adhesion to the general will. Rousseau's republicanism is rooted in his conception of the social contract as a pact between each individual qua individual, on the one hand, and society as a whole (of which each individual member is but a part), on the other.^[19] Furthermore, as a consequence of having been a party to Rousseau's social contract, every individual assumes a dual role: on the one hand, the individual is one among the many who are governed; on the other hand, the individual is an integral part of society as an indissoluble whole in its capacity as governing sovereign (Rousseau 1947, 16-18). Within this vision, the divide between the governed and the governors is correlated to the split between the individual interests and particular will of the private person and the common good embodied in the general will that informs the civic virtue of the self-governing public citizen (ibid., 14-16).

The notion of the general will propounded by Rousseau remains somewhat mysterious and does not always seem fully consistent. Rousseau characterizes the general will as the sum of differences between all the individual wills, or as the "agreement of all interests" that "is produced by opposition to that of each" (ibid., 26, n. 2). Thus, the general will is not reducible to any individual will, nor is it the will of the majority envisioned as a mere aggregate of individual wills. Nevertheless, under at least one plausible reading, Rousseau's general will accounts for pluralism in fact inasmuch as it takes clashing individual wills into account in the course of its formulation.

Consistent with these remarks, it is now possible to see how the individual comes to occupy the center stage with respect to the conflict between self and other in the context of Rousseauian republicanism. From the vantage point of the private person, the self is embodied in the particular will, and the other, in the general will. Conversely, from the vantage point of the public citizen, the self becomes identified with the general will, and the other, with the individual interests that inform the particular will. It follows from this that self-identity either as a private person or as a public citizen necessarily involves some measure of self-alienation—some degree of incor-

[19] See Rousseau 1947, 16-17. For a discussion of the contrast between Rousseau's conception of the social contract and those of Hobbes and Locke, see Rosenfeld 1985, 863-67.

poration of the other as part of oneself and some degree of repudiation of one's self (identity) for purposes of severing (part of) the self that must be cast away as other. Moreover, although within this perspective the focus of the conflict between self and other becomes predominantly internal, it also has important external repercussions. Indeed, definition of the self from the vantage point of the public citizen requires building bridges toward others in their capacities as fellow citizens in order to ferret out the "agreement of all interests" that is constitutive of the general will. At the same time, the public citizen must repudiate not only his or her particular interests (inasmuch as they cannot be incorporated into the general interest) but also the particular interests of others. Conversely, from the vantage point of the private person, the other is not only the internalized communal identity of oneself as public citizen but also all of one's fellow citizens in their capacities as private persons with different particular interests. In short, the self of the republican public citizen is built on a combination of alienation from oneself and of internalization of the other.

Viewed from the standpoint of the relationship between self and other, there is much that is common between Rousseauian republicanism and comprehensive pluralism. Indeed, by subordinating first-order norms to second-order norms, comprehensive pluralism conditions self-realization on accepting some degree of alienation from one's own first-order objectives. Just as the Rousseauian individual must sacrifice purely particular self-interest for the sake of seeking reconciliation with the other under the guidance of the general will, so too the proponent of particular first-order norms, under comprehensive pluralism, must give up pursuing them on their own terms in order to reconcile his or her own pursuit of self-realization with those of others.

Focus on these similarities brings out two issues around which the key differences between republicanism and comprehensive pluralism revolve. These two issues, which are closely related, concern respectively the possibility of reaching agreement among all interests and the extent to which self-sacrifice might be warranted for purposes of seeking to reconcile the self and the other. Regarding the possibility of reaching an agreement, there seems to be a significant difference depending on whether one believes that there are universal interests shared by all members of the polity. If there were universal interests, and particularly if these were predominant, then emphasis on civic virtue and on adhesion to the general will could be justified as consistent with equity and with a tolerable degree of self-sacrifice. If, on the other hand, consistent with the conclusions reached above, no universal interests—other than perhaps the interest in the survival of the polity—are likely to be found in settings that are pluralistic in fact, then civic virtue and conformity to the general will would either be

contrived or require an unconscionable degree of self-sacrifice and self-denial. In Rousseau's own case, moreover, republican virtue seems to extol much too high a price in terms of self-denial, as evinced by his famous dictum "Whoever refuses to obey the general will shall be compelled to it by the whole body: This in fact only forces him to be free" (ibid., 18).

Even if they were equal in all other respects, pluralism would still be preferable to republicanism, by refusing to enshrine civic virtue as an overriding value. Also, although contemporary republicanism may be less harsh than Rousseau's, inasmuch as it remains genuinely republican it must still adhere to the primacy of the public citizen over the private person. Comprehensive pluralism, in contrast, harbors no preference among the two and is thus better suited to less coercive means of reconciliation between self and other. In other words, by not requiring any a priori choice between public citizen and private person, comprehensive pluralism is less likely to exact excessive personal sacrifice for the sake of the self-governing polity's cohesiveness.

D—
Comprehensive Pluralism and Communitarianism

Communitarianism, for its part, assigns overriding status to the values of solidarity and group loyalty. Moreover, as already pointed out, communitarianism need only be considered as a plausible rival to comprehensive pluralism inasmuch as it makes room for communal pluralism. Consistent with this, before it is possible to indicate why pluralism is preferable to communitarianism, it is necessary to take a closer look at a conceivable communitarian vision that might be compatible with the preservation of heterogeneity.

The kind of communitarianism that might rival pluralism would make room for polities composed of a plurality of diverse communities. Furthermore, a distinction must be drawn among those polities made up of essentially self-enclosed separate communities with a minimum of intercommunal dealings and those that include communities largely open to one another and linked together through a multiplicity of intercommunal undertakings. Polities consisting of largely self-enclosed communities are not much different from those constituting a single tightly knit community. Accordingly, only the communitarian model that encompasses a multiplicity of communities open to one another is relevant for present purposes.

An important question that arises in relation to the latter model is whether solidarity can play a significant role in intercommunal dealings. It seems reasonable to answer that question in the affirmative, since it is easily conceivable that intercommunal bonds calling for solidarity between members belonging to different communities would play a non-negligible role

in a communally pluralistic polity. For example, intercommunal bonds of solidarity could develop among members of different religions who unite to promote the spirit of religion in the face of a mounting tide of secularism.

Given the possibility of intercommunal solidarity, the question becomes whether communitarianism in a communally pluralistic polity would be better off if solidarity is made into an overriding value. Returning to the example of intercommunal solidarity among members of different religions, it should be noted that each such member must have dual loyalty and solidarity with two different communities that do not completely overlap. Intracommunally, such member must achieve solidarity with those who adhere to his or her own religion. Intercommunally, on the other hand, that member must establish solidarity with adherents to other religions. Moreover, even if all the religions involved are in full agreement when it comes to dealing with secularism, there may well be certain matters over which these various religions differ sharply. In the latter case, the two loyalties involved would come into conflict with one another, and intracommunal and intercommunal solidarity would tend to pull in opposite directions.

Assuming solidarity and group loyalty are considered overriding values, then the above conflict and tension seem bound to strain the unity and cohesiveness of the communities involved. To ease this situation, these communities would seem to have to move in the direction of greater self-enclosure. Thus, a person with dual commitments that come into conflict with one another may only be able to maintain membership in the two communities involved by tempering his or her loyalty to each of them. To avoid diluting loyalties and weakening solidarity, therefore, there might be no other workable solution but to confine one's commitments to a single community.

In multicommunal settings where significant intercommunal dealings are virtually unavoidable, the communitarian ideal may be better served by a proliferation of overlapping communities than by a retrenchment into completely separate self-enclosed communities. Consistent with this, moreover, the communitarian ethos would seem best served by extending the reach of loyalty and solidarity to new frontiers, even if that would result in a diminution in the latter's intensity. Or, put more accurately, the communitarian ethos in communally pluralistic polities requires some loosening of loyalty and solidarity in order to allow for an expansion of the reach of communal interaction.

It follows from the preceding observations that pluralism is better suited than communitarianism—understood as prescribing solidarity and group loyalty as overriding values—to advance the case of the communitarian ethos in communally pluralistic settings. Indeed, proliferation of overlapping communities and multiplication of diverse communal commitments depend on the availability of a degree of flexibility and diversity largely

lacking in the context of strict adherence to loyalty and solidarity as overriding values. Pluralism, on the other hand, is more finely calibrated than communitarianism for purposes of fostering an optimal reconciliation of self and other without undermining the predominance of communal relationships or unduly diluting communal bonds. As a matter of fact, pluralism is not inherently inconsistent with promotion of high degrees of loyalty and solidarity, and because of its flexibility it can broker a viable integration of overlapping communities by forging better distributive channels toward an optimal apportionment of solidarity and group loyalty.

E—
The Superiority of Comprehensive Pluralism

Thus far, in building the case for the superiority of comprehensive pluralism over liberalism, republicanism, and communitarianism, I have largely taken the latter's chief objectives at face value and on their own terms. From the perspective of reconciling self and other in the least coercive or confining manner possible, however, these objectives hardly impose themselves as self-evident. Actually, as we shall now see, these objectives cannot be legitimately considered to be overriding from the standpoint of reconciling self and other. Moreover, as a consequence of this, the case for the superiority of comprehensive pluralism turns out to be much stronger than may be inferred from the preceding discussion.

Given the broad characterization of self and other that has emerged throughout this book, there appears to be no need for postulating any fundamental primacy of the individual over the group or of the community over the individual. Similarly, whereas self-government figures as an important value in the context of enhancing the mutual autonomy and dignity of self and other, this hardly justifies elevating civic virtue into an overriding value. From the standpoint of reconciling self and other, all principal liberal, republican, and communitarian values appear to have a significant role to play. Accordingly, liberty, equality, human rights, justice, civic virtue, solidarity, and group loyalty are all worth pursuing, the only question being how they ought to be harmonized and combined to better reconcile self and other.

There is no single answer to the last question, because the right interplay among the above-mentioned values depends in part on contingent and contextual factors. Thus, for example, pursuit of liberal values ought to predominate in settings where communities are powerful, repressive, and rather self-enclosed. Conversely, communitarian values should be strongly stressed in the face of rampant individualism with consequent loss in social conscience and collective cooperation. More generally, as needs, interests, and circumstances vary, none of the values under consideration can claim any set entitlement to priority. Because of this, and because it molds the

relationship among these values to suit the constantly evolving quest for reconciliation between self and other, comprehensive pluralism emerges as far superior to liberalism, republicanism, and communitarianism. Actually, because of its flexibility and broadly encompassing sweep, comprehensive pluralism can make use of what is best in liberalism, republicanism, and communitarianism while staying clear of their respective drawbacks.

IV—
The Nexus between Comprehensive Pluralism, Interpretation, and the Division of Labor among Law, Ethics, and Politics

Liberalism, republicanism, communitarianism, and comprehensive pluralism each have interpretive consequences that can be assessed in terms of the quest for just interpretations. These interpretive consequences may not always be easy to pin down, as they tend to vary depending on the particular version of the comprehensive perspective from which they are drawn. For example, both Posner's law and economics supplemented by liberal individualistic constitutional constraints discussed in chapter VI and Dworkin's expansive liberal egalitarian conception of substantive constitutional rights supplemented by his theory of law as integrity mentioned in chapter I emanate from a liberal perspective. Posner and Dworkin part company, however, due in large measure to Posner's embrace of liberty and welfare maximization as overriding values as contrasted with Dworkin's commitment to equality as paramount. Nevertheless, distinctions within any of these three pale in comparison with the clear-cut differences between liberal, republican, and communitarian interpretations. Thus, the boundaries of fundamental rights are bound to shift depending on whether these rights are assessed from a liberal, republican, or communitarian standpoint. A Rousseauian republican conception of liberty, for instance, insists on freedom through the polity or positive freedom,^[20] in contrast to liberal conceptions of liberty, which reserve a prominent place for freedom from the polity or negative freedom.^[21]

As already mentioned, comprehensive pluralism derives distinct interpretive consequences from the operation of its two logical moments identified above. Because of comprehensive pluralism's refusal to embrace any first-order norm as overriding, however, the distinctness of its interpretive products is not always readily apparent. Indeed, to the extent that comprehensive pluralism draws on some of the values espoused by other compre-

[20] For a discussion of Rousseau's conception of positive freedom, see Rosenfeld 1985, 870.

[21] For a classical statement of the nexus between liberalism and negative freedom, see Mill 1859.

hensive perspectives—without ascribing the same weight to these values as competing perspectives do—there are inevitable interpretive overlaps between comprehensive pluralism and rival perspectives. Nevertheless, even when outcomes coincide, reconstruction pursuant to the dictates of its two logical moments is bound to reveal comprehensive pluralism's distinct interpretive imprint.

I have already briefly described the contrast between comprehensive pluralism's negative moment—a rather transparent process consisting in the systematic leveling of hierarchies among first-order norms—and its positive moment—a constructive effort at reintegration that falls inevitably short of its mark because of unavoidable contingency and limitations. The time has come now to inquire into how comprehensive pluralism might contribute to the quest for just interpretations. More particularly, it is necessary to focus on how comprehensive pluralism's logical split between its negative and its positive moment and its normative distinction between first-order and second-order norms may make use of the divide between law, ethics, and politics to generate just interpretations.

It should be emphasized from the outset that comprehensive pluralism's rejection of overriding values should not be misconstrued as meaning that it can dispense with justice. Indeed, so long as there is pluralism in fact, comprehensive pluralism is concerned with striking some balance among competing perspectives and must thus incorporate justice among its constellation of second-order norms. That does not mean, however, that comprehensive pluralism necessarily endorses any particular conception of justice or that it singles out justice as a first-order norm—that is, justice as it emerges from within the perspective of a given conception of the good vying for supremacy in the arena of first-order norms—for special treatment. Similarly, whereas comprehensive pluralism incorporates equality as part of its second-order negative leveling of all first-order norms, it does not inherently embrace any particular conception of equality or necessarily favor equality as a first-order value as against other such values. More generally, comprehensive pluralism contemplates that every value and (first-order) conception of the good will vary depending on whether it is considered from the standpoint of second-order objectives or from that of first-order ends.^[22]

Inasmuch as selves tend to be more deeply anchored within the realm

[22] Strictly speaking, the second-order conception of the good—that is, comprehensive pluralism—also varies depending on which of these two standpoints is embraced. Thus, from within any first-order perspective, comprehensive pluralism appears to be no different than any other rival (first-order) conception of the good. I do not pursue this any further here as it has no bearing on the search for the interpretive consequences that follow from the endorsement of comprehensive pluralism.

of first-order norms, the first-order norms that one identifies with will undoubtedly be much more internalized than the second-order norms that one has agreed to honor. Accordingly, notwithstanding that prescriptively comprehensive pluralism requires internalization of second-order norms, phenomenologically valued first-order norms will tend to appear as internal, and second-order norms, as external. But once the process of internalization required by comprehensive pluralism takes hold, then internalized first-order norms must be first recast as external. Indeed, I cannot adopt the perspective of comprehensive pluralism unless I first gain some distance from my communally grounded beliefs. Moreover, the negative moment of comprehensive pluralism fosters detachment from one's first-order commitments. On the other hand, after having carried such detachment far enough, comprehensive pluralism, through its positive moment, requires reintegration of communally grounded norms to the extent required to maximize diversity while minimizing confinement and coercion. Reintegration, in turn, requires a reinternalization of first-order norms, but such reintegration needs to be partial rather than complete. Otherwise, the gains made through internalization of second-order norms would altogether dissipate. In short, under comprehensive pluralism, the interplay between first-order and second-order norms unleashes a dialectic revolving around internalization and externalization. This dialectic engages the essential components of each participant's shifting normative vantage point precariously anchored in a commonly shared perspective that remains in a perpetual state of construction.

Its constant need for internalization and externalization suggests that comprehensive pluralism could make good use of the division between ethics—the locus of internal normative relationships—on the one hand, and law and politics—characterized by external normative relationships—on the other. As discussed in chapter III, emphasis on the external links forged through law can defuse ethical conflicts among proponents of competing conceptions of the good, and therefore provide the breathing room necessary for peaceful coexistence among the members of an ethically divided polity. Consistent with this, comprehensive pluralism could presumably appeal to law to deal with first-order norms in the least intrusive manner possible—that is, by limiting constraints on first-order norms to the realm of external relationships. From the standpoint of comprehensive pluralism's negative moment, law should provide for external relationships in a way that avoids, as much as possible, privileging certain first-order norms over others. From the standpoint of comprehensive pluralism's positive moment, moreover, law should furnish the space necessary for ensuring that the sacrifices required for reintegration of first-order norms are largely confined to the realm of external relationships. Ideally then, from the perspective of comprehensive pluralism, law would promote equality among competing

first-order perspectives through provision of a neutral area for external relationships, and liberty by circumscribing the space for constraints to an intercommunal sphere reserved for external relationships.

As just described, the ideal of comprehensive pluralism seems virtually identical to that of liberalism, with the one difference that unlike liberalism, comprehensive pluralism does not appear to be biased in favor of individualism. However, because actual neutrality cannot be achieved except in a purely (negative) formal sense, and because in any complex contemporary constitutional democracy the private preserve of intracommunal dealings cannot be neatly separated from the public sphere of intercommunal relationships, this ideal lacks sufficient counterfactual critical bite to yield sound interpretive criteria. In particular, this ideal is deficient in that it overemphasizes form over substance and essence or end-state over process. Instead of focus on neutrality and equality, the emphasis should be on neutralization and equalization; instead of on liberty, on liberation; and instead of on internal and external, on internalization and externalization.

The liberal ideal is insufficient for purposes of advancing the interpretive enterprise consistent with comprehensive pluralism, but that does not mean that this ideal ought to be altogether discarded. The problem with the liberal ideal is that it is one-sided. But so are the republican and the communitarian ideals. Indeed, the ultimate, unattainable, ideal of comprehensive pluralism maps out a vision of full integration through self-realization without sacrificing full differentiation. In accordance with this ideal, the self would not be merged into the other, but each would become transparent to the other, and they would jointly devise and maintain a unified and integrated communal setting in which each self could freely invent, develop, and pursue a conception of the good. This ideal enterprise would thus function much like a self-governing artists' community devoted to optimizing individual and collective opportunities for aesthetic creativity, satisfaction, and diversity. Moreover, within this ultimate ideal of comprehensive pluralism, the liberal, republican, and communitarian ideals would on one level mutually limit one another to sustain differentiation, while on another level they would become subsumed within the unified construct designed to lend support to integration. In other words, within the ultimate ideal of comprehensive justice, the respective ideals of liberalism, republicanism, and communitarianism loom as both antagonistic and complementary.

The ultimate ideal of comprehensive pluralism has much in common with the concept of comprehensive justice. Actually, the vision promoted by this ideal amounts to a representation of the state of affairs that would obtain if comprehensive justice could be fully achieved. Accordingly, just as comprehensive justice should figure in the determination of just interpretations, so too should its corresponding ideal. However, since both of

them are impossible to achieve, their contribution to establishing just interpretations must be limited to providing a sense of direction . Thus, faced with an ever-changing process of confrontation and accommodation, the seeker of just interpretations must refer to these two impossible objectives to determine whether a proposed resolution to a conflict is likely to lead in the right direction (or where many different resolutions are proposed, whether any of them is more likely than the others to lead in the right direction). Moreover, this sense of direction should inform the backdrop against which criticism of an actual state of affairs is launched as well as the backdrop against which concrete proposals for change are articulated.

Set against the two impossible objectives discussed above are the two poles that must be reconstructed as if they were part of a progression toward greater perfection. These two poles are the actual state of affairs torn by a conflict calling for actual resolution and the plausible horizon of possible concrete resolutions of that conflict. Both of these poles, moreover, involve a significant element of contingency: the conflict itself is historically contingent, and the plausible horizon for its resolution is bound to be partially determined by the particular nature of the conflict as well as by cultural, technological, and institutional contingencies that constrict the range of available options. Beyond these contingencies, however, within contemporary constitutional democracies, conflicts and their plausible resolutions revolve around the antagonistic and complementary relationship among the liberal, republican, and communitarian strains associated with comprehensive pluralism. Thus, the conflicts between self and other which call for just interpretations seem amendable to recharacterization. They can be recharacterized as involving either infringements on liberty or equality, or frustration of self-realization, or the undermining of communal solidarity, or finally the placement of impediments on the convergence of integration, differentiation, and self-realization.

Just interpretations are thus circumscribed by the logics of comprehensive pluralism, on the one hand, and by the dialectics fueled by the concurrent antagonism between, and complementarity of, the liberal, republican, and communitarian tendencies within comprehensive pluralism, on the other. Within this framework, the division of labor among ethics, law, and politics might at first be attempted as follows. Ethics would preside over the internalization of both intracommunal and intercommunal norms, or, in other words, of both first-order and second-order norms. Politics, for its part, would provide external harmonization of intracommunal and intercommunal objectives through self-government. Finally, law would both secure the necessary conditions for the proper functioning of ethics and politics and provide interstitial regulation to cover gaps unfilled by ethics or politics. Moreover, this division of tasks would, by and large, place law under

the liberal ideal, ethics under the communitarian ideal, and politics under the republican ideal.

The above division of labor is not only sharp and clear but also seems to allow the tensions between law, ethics, and politics to provide an expressive outlet for the antagonism between the liberal, communitarian, and republican strands of comprehensive pluralism. By the same token, this division apparently allows for expression of the complementarity of these three strands through the reconciliation and harmonization of law, ethics, and politics, or, in other words, of justice according to law and justice beyond law. Unfortunately, however, as will now be demonstrated, in spite of its great appeal, the above division must be rejected as inadequate. If ethics merely internalized first- and second-order norms, this would not lead to communitarian pluralism but rather to confusion. As we have seen, second-order norms are not on the same plane as first-order norms. Moreover, from a phenomenological standpoint, inasmuch as intracommunal norms are more deeply entrenched than intercommunal norms, genuine internalization of second-order norms only seems possible if accompanied by externalization of first-order norms. Such externalization, however, cannot be achieved by ethics alone, but requires recourse to law and politics. Indeed, unlike in a metaphysical perspective where ethics, law, and politics merely complement each other, from the perspective of comprehensive pluralism, law and politics are also, in part, antagonistic to ethics. In particular, law binds together the outer framework of a polity committed to comprehensive pluralism, through the leveling of all hierarchy among first-order norms and through reintegration of such norms in a way that precludes them from gaining readmission on their own terms. Law's leveling thrust introduces a contradiction between the self's internal commitment to certain first-order norms and the self's external commitments based on the proposition that these same first-order norms are no better than any others. To put it in terms of the operative dialectic, the self may be initially drawn to law in order to place the other's first-order norms at a sufficient distance to create the space necessary for the self to attend to his or her own first-order concerns. But once the wide net of law is cast, the self's own first-order norms will also become caught in it, with the consequent alienation of the self from his or her own first-order norms. Finally, in yet another dialectic reversal the very distance that separates the self from his or her own first-order commitments allows for greater tolerance of the other's first-order norms. On the other hand, law's concentration on reintegration yields a cluster of first-order norms that is different—if not in configuration, then in weight and in its relationships to other norms—from those previously internalized by the self. To cope with this difference, the self must mediate its ethical commitments through law, and this involves essen-

tially two separate steps: first, law's negation and displacement of the self's cluster of first-order norms allows for sufficient externalization of the latter to make room for internalization of second-order norms; and second, by mapping out the (external) norms designed to govern intercommunal dealings, law leaves space for intracommunal pursuits and thus makes for (partial) reinternalization of first-order norms.

It follows from this that within comprehensive pluralism, ethics is dependent on law, and because the two are antagonistic as well as complementary, they cannot be collapsed into one another. Moreover, just as ethics is dependent on law, so too is law dependent on ethics. Because law cannot achieve neutrality and equally accommodate all first-order norms in its positive thrust, it must remain open to ethical criticism and influence. Specifically, ethics is bound to call on law to create conditions for greater and better accommodation of diverse first-order norms. Furthermore, depending on the circumstances, the ethical call on law may be primarily steeped in the communitarian or in the liberal strand of comprehensive pluralism.

From the standpoint of ethics, politics plays a role similar to law. Broadly speaking, politics provides the arena for the democratic setting and pursuit of intercommunal objectives and thus fosters antagonism and complementarity regarding communal (first-order) norms. In its own way, politics tends to level all first-order norms, by forcing them to compete for approval by the majority.^[23] On the other hand, political solutions in a pluralistically diverse polity are unlikely to satisfy many of the holders of different clusters of first-order norms whose value preferences become significantly frustrated. Inasmuch as politics fulfills an integrating function by pursuing a common base for intercommunal dealings, it can serve as a point of departure for internalization of second-order norms; inasmuch as it leads to division and fragmentation, it may foster externalization of first-order norms as well as retreat within communal bounds shielded from intercommunal strife.

In terms of just interpretations, it is the relationship between law and politics that warrants further exploration. Ideally, politics should provide an external medium for the reconciliation of self-government and self-realization, of public citizen and private or intracommunal person. In actuality, however, republican self-government and individual or communal self-realization are often at odds with one another. Moreover, even if the

[23] Presumably, if a particular cluster of first-order norms is already endorsed by the majority, there might be no real competition, and the antagonism between ethics and politics might well be minimized. Majority endorsement, however, need not spell the end of competition. It is conceivable that a majority might be convinced that it would be politically advantageous to accommodate certain value preferences held by a minority, to ensure overall more advantageous and more harmonious intercommunal dealings.

sacrifices required to abide by the republican ideal were deemed legitimate no matter how severe, no "general will" in the equal interests of the entire citizenry could ever be devised. Accordingly, recourse to ethics for purposes of internalizing civic virtue, while necessary to prevent runaway fragmentation due to unmediated clashes of interests, would be inadequate to bring about a political balance between self-government and self-realization.

On the level of politics, pursuit of self-government may become reduced to attempting coercion of the other, through use of political leverage to force the other unduly to sacrifice self-realization in the name of a "common good" that leaves that other at a distinct disadvantage. On the other hand, in the relentless pursuit of self-realization, the political clash of self-interests may lead to such excessive fragmentation as to threaten the cohesion of the polity. Against these two dangers, law offers two kinds of safeguards: antagonistic counterweights and complementary relief.

Law's mediation can furnish a counterweight against excessively coercive politics,^[24] through constitutional constraints. More specifically, constitutional constraints can play both an enabling and an antagonistic limiting role with respect to politics. On the one hand, constitutional provisions accorded the force of law function as constitutive rules that afford a measure of institutional stability to the game of democratic politics. On the other hand, constitutional constraints carve out part of the domain of external relationships in order to place it beyond the reach of everyday politics. Thus, constitutional constraints may antagonize politics by thwarting the latter's reach toward some of its objectives.^[25]

Law also complements politics to the extent that the realm of external relationships is apportioned among the two of them, largely as a matter of convenience. It will be remembered that in chapter III a distinction was drawn between two distinct approaches toward a division of labor between law and politics. The first of these was the structural approach, which, roughly speaking, institutionalizes the divide between constitutional constraints, on the one hand, and politics—broadly understood to include laws inasmuch as these are the products of the will of a legislative majority—on the other. The second was the thematic approach, according to which certain matters are left to law and others to politics, primarily as a matter of convenience. Now, it is in the context of the thematic approach that law

[24] I use the word "excessively" advisedly, as majoritarian politics, no matter how considerate, necessarily involves some measure of coercion.

[25] It can be objected that constitutions are as much a subject to politics as ordinary laws and all other issues ordinarily dealt with in the political arena. Suffice it for now, in reply, to point out that acceptance of constitutional politics as a fact of life does not preclude asserting that counterfactually, and within the perspective of comprehensive pluralism, constitutional constraints on politics are more—or more precisely, other—than merely political. I shall discuss these issues in greater detail in the course of the next chapter.

and politics can be regarded as complementary. To the extent that outcomes in politics are more volatile and unpredictable than outcomes in law, and that a complex pluralistic polity lacks the resources to deal thoroughly and competently within the realm of politics with all issues conceivably amendable to political resolution, it makes sense to turn to law to unclog the channels of politics. Moreover, although the fact of apportionment may be a matter of convenience, the way in which such apportionment is achieved is often not a matter of indifference. Thus, depending on contextual factors, maintaining as much stability as possible in certain areas may be a prerequisite to viable politics within the constraints of comprehensive pluralism. In short, if law as complementary could be likened to frozen or decelerated politics, then its legitimacy would hinge on its ability to shield the political engine from overheating.

To recapitulate: the ethics of comprehensive pluralism is communitarian, but instead of addressing a single community; it seeks to provide the internal glue for a community of communities. This, however, cannot be achieved directly; it requires the mediation of law. Law makes for the externalization of certain norms, which, in turn, allows for the internalization of others, all of which is necessary to transform the priority of second-order norms into a working ethical principle. As indispensable to the survival of the ethics of comprehensive pluralism, law is ethical. But as necessarily other than ethics—remember it is law's propensity for generating external relationships that renders it indispensable—law must, on some level, remain distinct from ethics. On the other hand, from the standpoint of law, law must remain independent from ethics as the leveler of first-order norms but must fall back on ethics to make up for its failure to achieve neutrality and equality in the course of its reintegration of first-order norms. Moreover, to the extent that law must act independently from ethics, it clearly tends toward the liberal ideal. Nonetheless, because it needs to rely on ethics to compensate for its failure to achieve neutrality, law must remain open to communitarian influences. Finally, politics emerges as a medium for external reconciliation of self-realization and self-government. Inasmuch as such reconciliation can never be fully achieved, however, politics needs to be constrained by law and ethics to avoid the excesses of self-government to the exclusion of self-realization, and of self-realization for some at the expense of others. Furthermore, insofar as law and politics are properly regarded as complementary, the two remain on the same plane, and law can be subsumed, by and large, under the broader objectives of politics. Consistent with this, politics bears a close connection to law and ethics, but there remains, all the same, a limited area within which (democratic) politics ought to remain supreme. That area covers those instances in which all available options would neither materially advance nor set back self-realiza-

tion. Indeed, in those instances, as all other relevant things remain equal, self-government looms as clearly superior to the alternative, and politics as the appropriate vehicle for self-government becomes paramount. Accordingly, with respect to the instances in question, politics ought to be driven by the republican ideal. Otherwise, politics should accommodate a combination of liberal, republican, and communitarian influences.

The preceding remarks concerning the division of labor among law, ethics, and politics, within the framework of comprehensive pluralism, touch on, in broad outline, some of the salient requirements for just interpretations. A more specific examination of the interpretive consequences of this division of labor will be addressed in the next chapter. However, there is one issue with very significant interpretive consequences that must be briefly examined now, before the close of this chapter. That issue concerns the nexus between counterfactual reconstruction of the division of labor among law, ethics, and politics and what actually happens to be embodied in law.

It is obvious that what is actually encompassed within law does not necessarily correspond to what ought to be counterfactually consistent with comprehensive pluralism. For example, actual constitutions may constrain actual politics in ways that seem unwarranted counterfactually, or actual laws may fail to complement ethics or politics as they should. In such cases, a question arises concerning the interpretive implications of the discrepancy between the actual and counterfactual domains of law.

Where a discrepancy is over the content rather than the domain of law, then reference to the appropriate counterfactual content seems entirely legitimate for critical purposes. Moreover, provided there is room for a range of plausible interpretations, then reliance on the relevant counterfactual seems fully warranted to interpret the actual law in ways that bring it as close as possible to its counterfactual counterpart. For example, if an antidiscrimination law that omits any reference to discrimination on the basis of homosexuality can be plausibly interpreted as either permitting or prohibiting such discrimination, and if the latter is counterfactually called for, then commitment to just interpretations would clearly seem to require interpreting the antidiscrimination law in question as prohibiting discrimination against homosexuality.

With discrepancies concerning the domain of law, in contrast, it is far from obvious that juxtaposition of the actual and the counterfactual would be productive with regard to legal interpretation . Undoubtedly, such juxtaposition would be highly useful from the standpoint of criticizing the actual law involved for trespassing beyond the legitimate bounds of legal relationships. But what interpretive use could possibly be derived from the conclusion that the matter addressed by an actual law ought legitimately be left

to ethics or politics? Arguably, under those circumstances, no plausible legal interpretation of the actual law would in any way advance the cause of just interpretations.

On closer examination, that last conclusion is only warranted in some of the cases in which discrepancies concerning the proper domain of law are present. Indeed, if counterfactually a matter ought to be entrusted to politics, and if actual politics or the plausible horizon of possible politics call properly handle the matter in question, then the above conclusion holds. However, if counterfactual politics can in no way be approximated by actual politics (or possible politics within the relevant horizon of plausibility), then actual law (or possible law within the horizon of plausible law) might well turn out to be the best available means to approximate counterfactual politics. Consequently, in that case juxtaposition of the actual and the counterfactual would yield significant consequences for legal interpretation notwithstanding discrepancies regarding the proper domain of law. Moreover, in such a case, it is the counterfactual division of labor among law, ethics, and politics, rather than its actual (or possible within the relevant horizon of plausibility) counterpart, that determines which values ought to inform the optimal legal interpretation under the circumstances. For example, if, counterfactually, a matter ought to be left exclusively to politics and be governed by the republican ideal, but if actual and foreseeably possible politics are so fragmented and self-centered as to preclude genuine republican self-government, then actual or foreseeably possible law interpreted in terms of republican values might well represent the best available approximation to just interpretations.

In conclusion, comprehensive pluralism has definite interpretive consequences, which, I have argued, are better suited than those of its rivals to lead polities that are pluralistic in fact to just interpretations. These interpretative consequences are, in part, procedural and, in part, substantive, and they hinge on juxtaposition of the counterfactual, the actual, and the possible within a plausible horizon of change. The interpretive consequences of comprehensive justice, moreover, depend on complex uses of the division of labor among law, ethics, and politics. Because of these intricacies, and because of the significant role of contingency, comprehensive pluralism rarely leads to direct or obvious interpretive outcomes. As a matter of fact, the interpretive consequences of comprehensive pluralism have proven to be above all relational . That is not surprising since comprehensive pluralism is primarily concerned about the relationship among diverse perspectives and between the actual, the foreseeably possible, and the counterfactual. Finally, although comprehensive pluralism's interpretive consequences loom as primarily relational, as I seek to demonstrate in the next chapter, it does not follow that they cannot be concrete or determinate.

Chapter Eight—
In Pursuit of Meaning Amid Partial Subjects, Elusive Others, the Open Texture of Law, and Imperfect Justice

I—
Micro-Interpretive Versus Macro-Interpretive Consequences of Comprehensive Pluralism

In the last chapter, I provided an overview of the interpretive consequences of comprehensive pluralism. Here, I seek to round out my examination of the interpretive implications flowing from comprehensive pluralism, by shifting the focus to salient particulars and reconstructing the quest for just interpretations from the "bottom up." In the last chapter, I dealt with "macro-interpretive" issues arising in connection with comprehensive pluralism; here, I start from a select number of "micro-interpretive" issues, with a view to determining how comprehensive pluralism might lead to sufficiently concrete and determinate meanings in its quest for just interpretations.

From a micro-interpretive standpoint, the starting point is the confrontation between self and other as mediated through law. Moreover, although the confrontation in question is multifaceted, given the present focus on legal interpretation, phenomenologically, the optimal starting point is the citizen's encounter with law—or, more precisely, with a particular law. To the extent that such a law is external to the citizen—that it imposes external constraints on him or her—it represents the other as against the citizen as a self. Conversely, to the extent that the citizen regards that law as the externalization of his or her will—or, in other words, the citizen views himself or herself as the author or coauthor of that law—the law in question becomes an expression of the citizen's self. Thus, the citizen finds the confrontation between self and other embedded within laws as well as permeating all intersubjective attempts at ascribing commonly shared meanings to laws.

The citizen is bound to encounter law in many different capacities, in-

cluding those of lawmaker, law administrator, law enforcer, adjudicator, and law-abiding person. Also, the need to ascribe meanings to laws arises no matter in which of these capacities the citizen happens to approach the law. For our purposes, however, it suffices to concentrate on the quest for meaning from the standpoint of the law-abiding person and from that of the adjudicator.

The law-abiding person must focus primarily on the confrontation between self and other within the law, in contrast to the adjudicator, who must concentrate above all on the conflict between self and other over what meaning ought to be ascribed to a law implicated in a legal dispute between them. Although the focus of the adjudicator within the perspective of comprehensive pluralism is distinct from that of the law-abiding person, as we will see, their respective viewpoints are by no means unrelated. Actually, the standpoint of the law-abiding person must be incorporated as a building block in the construction of a vantage point suited to the needs of the adjudicator. Consequently, our inquiry into the micro-interpretive implications of comprehensive pluralism must begin with a quick glance at the interpretive predicament of the law-abiding person.

The citizen who takes the vantage point of a law-abiding person within the perspective of comprehensive pluralism is called on to indicate why he or she ought to be bound to obey a particular law. Ideally and in the most general terms, the answer should be that the law in question encapsules a legitimate reconciliation of self and other. Furthermore, even at this high level of generality, the boundaries of the duty to obey laws emerge rather sharply. If a given law lacks any plausible trace of either the self or the other, then the law-abiding person does not strictly speaking have a duty to obey it. Indeed, on the one hand, a legal duty that is exclusively owed to oneself is no duty at all, whereas, on the other, a law that completely lacks any trace of the self is not only patently unjust but also purely oppressive. Thus, in any pluralistic setting, any law reducible to a mere projection of the self (which from the vantage point of the other amounts to a mere extension of the other) is but an illusion or an instrument of tyranny.

Between these two extremes, every law can be construed as embodying some reconciliation between self and other, and the question becomes whether such reconciliation should be deemed legitimate by the law-abiding person. Again, in the most general terms, those laws should count as legitimate which can be reconstructed as aiming at a reconciliation wherein the self meets the other halfway. And that could be accomplished in one of two ways. First, a law would be legitimate if the law-abiding person could be viewed at once as its author and as a person properly subjected to it—or, in other words, could be regarded as the law's author notwithstanding the degree to which that law incorporates the aims of the other and at the same time considered properly subjected to that law, even after acknowledging

that the law's content bears a significant imprint of the other. Second, a law would be legitimate if it could be reconstructed as seeking to apportion rights and duties in a way that provides equal consideration to both self and other.

From the vantage point of the law-abiding person within the perspective of comprehensive pluralism, then, laws can be roughly divided into three different kinds. First are purely oppressive laws imposed by a foreign occupier, or by an enemy within the polity bent on annihilation or complete subjugation, which properly call for rebellion or resistance rather than obedience. Second are laws that can be construed as striking the optimal possible balance between self and other, which must be obeyed, in part, because they are self-imposed and, in part, because by treating them as binding, one extends to the other the recognition needed for reconciliation through law. And third are laws that apportion rights and duties in a way that evinces equal consideration of both self and other, which must be obeyed either because they facilitate reconciliation between the two or because they block a path to greater subordination of one to the other.

There is likely to be much overlap between the second and the third kind of law described above, but the two must be kept distinct, for there are laws of the third kind that do not satisfy the requirements for laws of the second kind. Indeed, as a consequence of comprehensive pluralism's rejection of the viability of a neutral Kantian moral perspective discussed above,^[1] there may be situations in which self and other are unable to find sufficient common ground for purposes of arriving at a mutually acceptable reconciliation. In such situations, a law that favors neither the self nor the other, but which might not be voluntarily embraced by either of them, might nevertheless be justified. The basis for such justification, moreover, could be either that the law in question prevents mutual eradication of self and other or that it preserves a status quo that precludes domination by one over the other pending potential overcoming of the current impasse.

Whereas the paradigm cases corresponding respectively to the three kinds of laws identified above demarcate counterfactual ideals around which legal discourse seems bound to revolve, actual laws will inevitably fall short of these ideals. Accordingly, the interpretive challenge consists, above all, in gauging how actual laws stack up against these ideals, with a view to either legitimating the status quo or justifying calls for changes. Moreover, from a micro-interpretive standpoint, the principal tasks are to endow the relevant ideals with semantic content suited to the particular context involved; to furnish credible depictions of the relationship between the challenged status quo and plausible alternatives; and to harmonize viable arguments for or against departure from the status quo with pertinent semantically

[1] See chapter V, above.

cohesive depictions of the relevant counterfactual ideal(s) while remaining within the normative constraints of comprehensive pluralism.

From the vantage point of the law-abiding citizen, a particular law open to many different interpretations should, if possible, be construed as co-authored by the sell and the other. The next best possibility would be to construe it as authored in part by the self and in part by the other, with both of them agreeing to be mutually bound to the extent that the law is traceable to the other. Finally, there is a third, less desirable, possibility that nonetheless allows for legitimate sell-justification and for treating the law in question as binding.^[2] According to this third possibility, if the law under consideration could be interpreted as being equally alienating from the standpoints of self and other, it could still be worthy of obedience to prevent intensifying discord among the two.

The kind of joint authorship relevant in the first of the three above-described cases involves an overlap of interests or value preferences between self and other.^[3] For example, if self and other belong to different religions, but if each of these would be better served if the state withdrew from the private sphere, then self and other should welcome the opportunity to coauthor a law confining state intervention to the public sphere. Although different particular aims and conceptions of the good would ultimately stand behind the respective support accorded by self and other to their coauthored law, so long as the operative overlap remains firmly in place the law in question should not give rise to significant interpretive disputes. This is hardly surprising, for in these cases self and other are assumed to have the same legal aims notwithstanding that they differ in their extralegal objectives. In a complex pluralistic polity, however; these cases are likely to be exceedingly rare, thus primarily serving as useful counterfactual markers.

The second kind of collaborative authorship mentioned above, on the other hand, is likely to be much more prevalent, but it is also bound to generate its fair share of interpretive disputes. The kind of coauthorship involved here is perhaps best exemplified by the writing of a private con-

[2] In terms of prudential considerations, the scope of self-justification for abiding by a law would be much broader than when viewed from the perspective of comprehensive pluralism. Thus, submission to even a very oppressive law could be justifiable for prudential reasons if failure to comply could lead to dire consequences.

[3] Cases in which more than an overlap underlies coauthorship, such as when self and other within a polity join forces to combat an external other, are best reconstructed as the product of a self rather than as the coproduct of a self and another. In such cases of self-legislation the resulting law is not reducible to a mere illusion, however, for the fragile self that binds fellow citizens together against a foreign other is most probably in constant danger of dissolution. In any event, so long as this precarious self holds together, interpretive difficulties are not likely to arise. Accordingly, there is no need for further examination of these cases.

tract to govern an exchange between two parties. As pointed out in chapter I, such a contract is all expression of the joint will of self and other which is to a certain extent contrary to the individual will of either of them. Although the contract is a joint writing, and the signatories to it adopt the same legal objective, they do not ordinarily share the same legal aims at the time of entering into contract negotiations. Thus, as already noted, a prospective buyer typically wants to pay as little as possible, while a prospective seller typically wishes to charge as much as possible. Initially then, buyer and seller are likely to bring somewhat conflicting legal aims to their contract negotiations. Consequently, contractors carve out a common scheme not because their interests and value preferences happen to overlap but because they decide that compromise affords the best available means toward their respective objectives.

A contract is thus a jointly authored law based on compromise. In the context of ordinary commercial transactions, moreover, there is no harm in seeing both parties as coauthors who assume joint responsibility for the actual text of their contract. In carrying the contract analogy to coauthorship based on compromise within the realm of intersubjective dealings among proponents of competing conceptions of the good, however, coauthorship is best conceived as several rather than joint. Indeed, out of concern not to be swept within the perspective of the other, self and other might be viewed as agreeing to string together a mutually acceptable text in which fragments authored by each of them are blended with fragments contributed by the other. As thus conceived, moreover, whereas the law represented by the contract is binding on both parties, it is, in part, binding as self-imposed self legislation and, in part, as mutually coercive.

The contract analogy thus leads to the conclusion that, consistent with comprehensive pluralism, law is binding because it is self-imposed and because it is coercive—a conclusion that seems much like those reached by Habermas and by Hart (1961, 165-67). The justification regarding law's bindingness under comprehensive pluralism differs, however, from the respective justifications offered by Habermas's discourse theory and Hart's positivism. Unlike under these other theories, under comprehensive pluralism, self-legislation and coercion do not figure as alternatives, and neither of them could ultimately become subsumed under the other. Actually, under comprehensive pluralism, it is the very opposition between self-legislation and coercion that justifies adherence to laws that fall within the paradigm demarcated by the contract analogy. Paradoxically, within the perspective of comprehensive pluralism, coercion is not only constraining but also liberating: it allows the self to engage in a joint enterprise with the other without either of them being swept into the perspective of the other. In other words, so long as the contract analogy holds, law's coercive side allows the self to become enlisted in the pursuit of the other's objectives

without having to express allegiance to the latter's conception of the good while, at the same time, assuring the other's cooperation in one's own projects without threatening to compromise the other's inner commitments.

The legitimacy of the paradigm modeled on contract hinges not only on opposition but also on balance between self-legislation and coercion. For self and other to meet halfway, each must find a balance between the gains and sacrifices resulting from their mutual compromise. Ideally, this means that the gains and sacrifices of each will be equivalent to those of the other and that each one's gains will be proportional to his or her own sacrifices. Failure to satisfy the latter requirement, however, does not preclude the self from meeting the other halfway. Thus, even if neither self nor other derives any positive gain from a law, such law could still be legitimate consistent with comprehensive pluralism if the sacrifices required of each of them were equivalent. In that situation, moreover, the law in question would fall under the third kind of case identified above.

The third kind of case is the micro-interpretive equivalent to the first logical moment of comprehensive pluralism, which as we saw in chapter VII consists in the leveling of all first-order norms. Indeed, in some cases a standoff between self and other may be the only way for them to meet halfway and thus the best alternative short of dissolving the unity of the polity.^[4] Furthermore, the second kind of case, modeled on the contract analogy, corresponds, by and large, to the confluence of both logical moments of comprehensive pluralism and gives expression to the interplay between law and ethics, externalization and internalization, which is set in motion by attempts to harmonize competing first-order norms under the aegis of second-order norms.

Both of these latter cases, unlike the first case, are likely to raise vexing interpretive issues. Indeed, how can we tell whether self and other meet halfway, when there is no common measure to gauge their respective perspectives? Similarly, how can we compare their respective gains or sacrifices? Furthermore, are not these difficulties likely to be compounded as the lack of common measure may encourage self and other to argue that, in spite of appearances to the contrary, a particular law imposes a relatively heavier burden on them than on their counterpart?

In keeping with the contract analogy, many of these interpretive difficulties may be circumvented or minimized. As long as a contract represents the embodiment of the joint will of both contractors, then arguably the

[4] Given that comprehensive pluralism is agnostic as to whether a polity that is not pluralistic in fact ought to be committed to normative pluralism (see chapter VII, above), it has nothing to say on the issue of whether secession would be preferable to continued association among conflicting communities. In any event, there are cases in which external factors, such as economic or political conditions, make secession impractical.

procedural constraints inherent in contract formation would appear to guarantee that self and other would meet halfway. As we have seen in the course of our discussion of proceduralism in chapter V, however, contract alone does not ultimately guarantee pure procedural justice. For one thing, background conditions and substantive concerns must be taken into account. For another, consistent with the fluid conception of self and other advanced throughout this book, it is by no means assured that the needs and objectives of self and other will remain essentially unchanged throughout the life of the contract. Thus, whereas a party to a contract may have meant one thing on entering into the contractual relationship, that party may in good faith subsequently believe that he or she meant quite a different thing. In short, the respective identities of self and other, which figure prominently in the determination of whether a particular contract allows them to meet halfway, may evolve sufficiently during the projected life of their mutual undertaking to raise serious doubts as to whether their contract amounts to joint self-legislation.

The interpretive difficulties confronting self and other as contractors are compounded in the case of the adjudicator charged with resolving contractual disputes. To a large extent, the adjudicator must put himself or herself in the shoes of the contractors to determine their respective contractual intent, and that task is complicated by epistemological difficulties above and beyond those confronted by the contractors themselves. But, in addition, the adjudicator is also responsible for factoring background constraints and substantive concerns into the reading—or rewriting—of the contract before him or her. In that latter capacity, moreover, the adjudicator must either directly apply substantive norms or counterfactually reconstruct the intent of each contractor as if the latter had acted as a fully cognizant and fully responsible agent operating within the perspective of comprehensive pluralism.

The contract analogy to self-legislation has thus helped identify three different interpretive tasks likely to figure in the quest for just interpretations consistent with comprehensive pluralism. First, an actual intent for self and other must be imputed. Second, their counterfactual intent must be reconstructed so as to conform with the dictates of comprehensive pluralism. And third, substantative norms derived from comprehensive pluralism must be readied for application through interpretation. Furthermore, the second and third of these tasks are to a significant degree overlapping, as substantive norms can be frequently, but not always, factored in counterfactual reconstruction.

The search for reconciliation between self and other in terms of the three tasks mentioned above suggests that, from a micro-interpretive standpoint, legal interpretation is likely to remain incomplete and tentative. Nevertheless, these three tasks and particularly counterfactual reconstruction

and implementation of substantive constraints can contribute to a significant narrowing of available options, and thus to channeling legal interpretation in the direction of greater justice and more suitable reconciliation of self and other. In what follows, I briefly spell out how these tasks may be harnessed in the pursuit of just interpretations within the perspective of comprehensive pluralism. First, I explore from a micro-interpretive standpoint how counterfactual reconstruction might deal with antagonistic positions predicated on competing conceptions of the good, including clashes between proponents of different religious ideologies. Second, I examine the nexus between comprehensive pluralism and internal transformations of self and other. Third, I investigate how the three tasks above might contribute from a micro-interpretive standpoint to harmonizing competing first-order norms within the space made available for them by second-order norms. Finally, I focus on how interpretive insights deriving from comprehensive pluralism might contribute to resolution of difficult cases without leading to pathological suppression of difference.

II—
Making Interpretive Sense of Clashes among Partial Selves and Elusive Others: On the Road to Justice as Reversible Reciprocity

Confronted with the task of interpreting a law open to many different meanings, the adjudicator must, on the contract analogy, seek to approach the law in question as a piece of coauthored self-legislation. Viewed on the surface, this seems rather banal, as it is a well-established judicial practice to elucidate the meaning of a statute by reference to the legislative intent of parliament, and that of a constitutional provision by reference to the intent of the framers of the constitution. Viewed in greater depth and in the context of comprehensive pluralism, however, the adjudicator's task is not only more difficult than merely ascertaining legislative intent but also rather different. Indeed, leaving aside the inherent difficulties in deciphering the intent of the parties to a contract from a reading of the coauthored writing that embodies their agreement, the adjudicator confronts partial selves, elusive others, and a series of extralegal constraints, all of which play crucial roles in the quest to settle on legitimate meanings. Moreover, whereas these additional concerns in some sense clearly complicate the search for just interpretations, paradoxically, in another sense, they can ultimately contribute to simplifying it. On one level, reconciliation of self and other in ways that fully account for all relevant identities and differences is a never-ending project fraught with significant interpretive uncertainties. Nevertheless, these very interpretive uncertainties, when properly embedded in a setting animated by partial selves and elusive others, call for gap-filling measures as a prerequisite to establishing any coherent meaning.

Furthermore, these gap-filling measures must be justifiable in terms of the dictates of comprehensive pluralism to be legitimate, and on counterfactual reconstruction, rather than on unavailable or unreachable factual conclusions, to be workable. On the level of gap filling consistent with comprehensive pluralism, therefore, the interpretive task is measurably simplified as a manageable counterfactual reconstruction based on a stable normative foundation replaces a never-ending factual search hampered by insurmountable interpretive hurdles.

For all their stabilizing potential, the normative constraints and counterfactual reconstructions associated with comprehensive pluralism cannot purge the interpretive enterprise of contingency. Actually, the very starting point in the quest for just interpretations is mired in contingency, as the self, the other, the law that purports to bind them together, and the plausible avenues to better reconciliation between self and other are all to a significant degree dependent on historical contingencies and thus bound to vary from one setting to the next. In contrast, the framework of normative constraints and the architectonic principles of counterfactual reconstruction linked to comprehensive pluralism are essentially invariable and thus furnish a set mold transportable from one setting to the next. Consistent with this, the interpretive enterprise circumscribed by comprehensive pluralism tends to organize the relevant material in sets of recognizable relational clusters. Furthermore, it is the very recurrence of these relational clusters that allows for the simplifying steps that render the interpretive enterprise manageable.

A—
Comprehensive Pluralism's Interpretive Baseline: Equal Worth and Mere Reciprocity

Lying at the very core of comprehensive pluralism is a normative commitment to the equal worth of self and other.^[5] Because of this, for a law to make it possible for self and other to meet halfway, it must consistently uphold their equal worth. Otherwise, a law would have to be considered oppressive. Furthermore, assessment of whether a law complies with the equal worth imperative would in some cases remain independent of the imputed objectives of the law-abiding person, while in other cases, it would have to factor in such objectives. In other words, on the contract analogy, the optimal interpretation of a law is likely to depend, in part, on extracontractual normative constraints and, in part, on historically contingent contractual objectives.

At a minimum, equal worth depends on what I have referred to in chap-

[5] This commitment is derived from comprehensive pluralism and does not, therefore, imply any reliance on overriding values.

ter VII as "mere reciprocity" or recognition of the other as an other self. Some laws or legal interpretations, such as those that would make for subordination of a person or group to the will of another, would obviously fail to meet the criterion of mere reciprocity and ought therefore clearly be fought or avoided. Moreover, whereas repudiation of such laws or legal interpretations may often be recharacterized as originating in self-legislation or contract, this need not be the case. At bottom, arguments against unlimited subordination are ultimately buttressed by normative constraints inherent in comprehensive pluralism rather than by arguments from contract or self-legislation.

Beyond repudiating subordination, the interpretive implications of mere reciprocity defined as recognition of the other as another self remain largely unclear. Thus, consistent with that definition, mere reciprocity may equally plausibly be interpreted as merely requiring formal equality or, on the contrary, as imposing an impossible Kantian requirement to treat the other exclusively as an end. To avoid this kind of indeterminacy, it is useful to further specify mere reciprocity as requiring recognition that the other has—or is capable of having—a perspective or a vantage point of his or her own.^[6] Consistent with this, the equal worth of self and other implies mutual recognition as possessors of a perspective or vantage point. This mutual recognition is essential to collective self-legislation and to the legitimacy of contract as a vehicle suited to the common need of self and other to meet halfway but is not itself a by-product of self-legislation or contract. It figures instead as a prerequisite to self-legislation and contract and lies beyond their respective bounds. Hence, mutual recognition as possessors of a perspective furnishes both the foundation to and the limits of self-legislation and contract. Furthermore, it follows from this that, ideally, such mutual recognition should be cast as a constitutional principle that adjudicators ought to use in shaping the course of law and legal interpretation in the direction prescribed by comprehensive pluralism.

Positing mere reciprocity as a constitutional principle leads to certain determinate consequences but proves insufficient to carry out the normative prescriptions of comprehensive pluralism. Adherence to mere reciprocity thus forecloses slavery and conscious subordination^[7] and requires the self to recognize that the other is equally entitled to any right claimed by the self as the possessor of a perspective or vantage point. For example,

[6] To avoid the difficulties presented by exceptional cases such as those involving the profoundly retarded or by very young children, I understand the capacity to have a perspective counterfactually. Accordingly, every human being and every identifiable group should be deemed normatively entitled to be treated as if he or she possessed the capacity to form and express a perspective.

[7] For a more extended discussion of the implications of mere reciprocity, see Rosenfeld 1991a, 246-48.

if the self asserts entitlement to freedom of expression rights as the possessor of a perspective, then mere reciprocity would compel the self to recognize that the other is equally entitled to such rights. Mere reciprocity, however, is too limited from the standpoint of comprehensive pluralism, for it fails to account for differences in perspective. Because of this, mere reciprocity stresses identities at the expense of differences and does so at the highest levels of abstraction where self and other are reduced to mere possessors of a perspective or vantage point. Mere reciprocity thus promotes bare tolerance but cannot prompt self and other to meet halfway as it lacks the means to constrain a dominant self to acknowledge the needs and aspirations rooted in the perspective of the other as opposed to those traceable to the other possessing a perspective.

B—
Linking Mere Reciprocity to Reversible Reciprocity

To avoid a regime tailored to dominant perspectives coupled with bare tolerance which would at best satisfy the minimum requirements of limited pluralism, mere reciprocity must be supplemented by reversible reciprocity. As mentioned in chapter VII, reversible reciprocity requires recognition of and empathy for the perspective of the other and thus makes for proper acknowledgment of differences and for accommodation of diversity. Reversible reciprocity relying on the reversal of perspectives allows for evaluation of conflicts from the standpoint of each perspective and vantage point involved and thus opens the way to solutions that fairly account for differences in perspective and vantage point (Kohlberg 1979; Rosenfeld 1991a, 249-50).

Whereas mere reciprocity implies equality among subjects as possessors of a perspective or vantage point, reversible reciprocity targets equality among perspectives. If reversible reciprocity could guarantee full and equal accommodation of all existing perspectives within a polity, then it would prove interpretively sufficient for purposes of achieving the objectives of comprehensive pluralism. However, as demonstrated with respect to the logics of comprehensive pluralism in chapter VII, equal accommodation of all competing conceptions of the good in a positive sense is impossible to achieve. Accordingly, reversible reciprocity may be perfectly suited for purposes of equally leveling all conceptions of the good in the negative moment of comprehensive pluralism but is, standing alone, inadequate for purposes of partial reintegration of first-order norms during comprehensive pluralism's positive moment.

To overcome this limitation, it is necessary to link mere reciprocity to reversible reciprocity. Moreover, this link, which is meant to cast mere reciprocity and reversible reciprocity as antagonistic yet complementary, is essential. Indeed, strict and exclusive adherence to reversible reciprocity

would give no weight whatsoever to the fact that one is a proponent of a particular perspective and would require everyone to treat his or her own perspective as if it were that of another. Thus, the self would be forced to treat himself or herself as an other rather than being merely expected to treat the other as another self. In the last analysis, self-effacement to relate to oneself as an other seems equally likely to upset the equilibrium between self and other as the failure to acknowledge anything beyond that the other is the possessor of a perspective.

By pitting mere reciprocity against reversible reciprocity one maximizes the chances that self and other will be able to meet halfway. More specifically, it is the dynamic born out of the tension between pursuing one's own perspective within the bounds of mere reciprocity and deferring to the other out of empathy following from reversible reciprocity, which seems to lead self and other to meeting halfway. From the standpoint of the self, as a consequence of constantly alternating between self-assertion and empathy, mere reciprocity and reversible reciprocity emerge as complementary. Furthermore, such alternation closely tracks comprehensive pluralism's demand for allegiance to second-order norms coupled with its prompting proponents of previously leveled first-order norms to press for their reintegration within the prevailing normative order.

From the standpoint of the adjudicator, the dynamic tension between mere reciprocity and reversible reciprocity produces important interpretive consequences. These consequences relate both to the reconstruction of the respective positions of partial selves and elusive others and to the reconciliation of these positions. For purposes of reconstruction, the adjudicator must seek to harmonize the various partial selves of the law-abiding person as counterfactual contractor or self-legislator into a cohesive perspective or vantage point that the person in question could plausibly endorse as his or her own. In the course of carrying out this interpretive task, the adjudicator should accentuate the common elements that emerge from a conjunction of the various partial selves involved at the expense of the contradictions and inconsistencies encountered along the way. Moreover, the adjudicator should also identify the common core of various partial selves through a focus on the contrasts between the positions attributable to the latter and those espoused by those who constitute the relevant other. In short, the adjudicator can simplify the interpretive task by stressing the commonalities linking together partial selves and the oppositions that fuel the antagonism between self and other.

Mere reciprocity figures prominently in the adjudicator's efforts at reconstruction, but reversible reciprocity dominates the latter's quest for reconciliation. Indeed, for purposes of reconciling the respective positions of self and other, the adjudicator must develop equal empathy for the clashing perspectives and vantage points involved in the relevant conflict. Ideally,

having given equal weight to all the perspectives and vantage points involved, the adjudicator would resolve the conflict at stake in a way that accommodates as much as possible as many as possible among the relevant perspectives and vantage points, without losing sight of the intensity with which each of the parties involved adheres to his or her position.

Rejection of Habermasian notions of dialogically redeemed universal morality, or of solutions in the equal interests of all, implies that an adjudicator's reliance on reversible reciprocity can never lead to complete justice or incontestable outcomes. Even if an adjudicator were to fully capture the positions of self and other, and to grant an absolutely equal weight to the respective perspectives and vantage points of all involved, the resulting adjudications would nonetheless most likely have a disparate impact on the respective interests in self-legislation and self-realization harbored by self and other. Conceivably, in some cases, adjudication could come very close to being in the equal interest of self and other. One such case would be where self and other largely share the same interests against a common external threat. Such cases are of limited interest, however, as they vastly differ from the large number of cases in which clashing interests give rise to demands for competing interpretations. Furthermore, at the other end of the spectrum, there may be cases in which the conflicting interests are so radically antagonistic that no plausible adjudication could conceivably carve out any common ground. For example, a conflict between proponents of two ideologies exclusively bent on destroying their rival could only be resolved, consistent with the dictates of reversible reciprocity, by completely frustrating all the parties involved.

Between these two extremes, it seems fair to expect that there would be room for certain overlaps, certain possibilities for compromise, certain resolutions dependent on mutual adjustments or on limited measures of self-restraint, and certain concessions based on relative orders of priority. Accordingly, the task for the adjudicator would be to identify these and to vie for solutions consistent with equal respect for all law-abiding persons involved and with equal consideration toward all represented perspectives and vantage points. Moreover, after having ascertained who stands for what in relation to the conflict at hand, the adjudicator should seek to craft a just resolution of the contested issues, primarily through reliance on the normative precepts of comprehensive pluralism and through submission of conflicting claims to the test of reversible reciprocity.

In principle, all first-order norms, interests, objectives, life plans, and self-realization aspirations stand on an equal footing from the standpoint of comprehensive pluralism. In the course of resolving conflicts, however, adjudicators need to parse out the clashing value preferences advanced by the various parties to the controversy. This has to be done in accordance with the norms imposed by comprehensive pluralism and with the precepts

of reversible reciprocity. Significantly, after completion of this process, not all value preferences involved are likely to fare equally well. Consistent with adherence to comprehensive pluralism, second-order norms have priority over first-order norms, with the consequence, for example, that an intolerant (first-order) conception of the good is assured to fare much worse than one that is not intolerant. Thus, the ideal of giving equal weight to all perspectives is made subject, in the course of adjudication, to the normative constraints of comprehensive pluralism.

These normative constraints are non-negotiable and they limit self-legislation and the paths to self-realization.^[8] Moreover, they include constraints that derive from adherence to mere reciprocity as well as others necessary to prevent subordination to any particular set of first-order norms. Ideally, these normative constraints ought to be constitutionalized to the extent that they are needed to carry out the objectives of comprehensive pluralism in the realm of external relationships. Indeed, because they transcend the bounds of self-legislation, these normative constraints ought to be elevated above the constant give-and-take of ordinary everyday politics.

The first task of the adjudicator, therefore, is to evaluate the conflicting positions before him or her in terms of their compatibility with second-order norms. In performing this task, however, the adjudicator should be vigilant against inhibiting the pursuit of first-order norms more than is necessary to uphold the priority of second-order norms. For example, while the pursuits of a crusading religion would have to be substantially curbed, comprehensive pluralism by no means automatically justifies complete frustration of all its pursuits. Thus, if such a religion's coercive proselytizing could be separated from its other pursuits, such as worshiping and spreading other aspects of its religious dogma among existing adherents, then only the former pursuit ought to be frustrated. To do more than that would actually run counter to the fundamental tenets of comprehensive pluralism as it would single out the pursuit of certain first-order norms not inherently incompatible with second-order norms for discriminatory treatment or eradication.

A far different situation is presented when the adjudicator confronts a conflict in which the pursuit of one set of first-order norms not inherently incompatible with second-order norms inevitably tramples on the pursuit

[8] Alternatively, these constraints may, through recharacterization, become incorporated into self-legislation. To the extent that these constraints frustrate the actual objectives sought through self-legislation, however, it is preferable to view them as external to the process of self-legislation. Indeed, whereas for a proponent of comprehensive pluralism modifying one's objectives to conform with second-order norms may constitute an act of self-legislation, this would not be the case for those who reject comprehensive pluralism.

of another such set of first-order norms. This latter situation, which arises after the adjudicator has rejected those proposed alternatives that proved inherently inconsistent with second-order norms, calls for recourse to reversible reciprocity. Specifically, it requires an evaluation of the conflicting claims at stake according to what I have called elsewhere the principle of "justice as reversible reciprocity" (Rosenfeld 1991a, 7, 254-82).

C—
Justice As Reversible Reciprocity

Building on the notion of reversible reciprocity, justice as reversible reciprocity allows for a ranking of conflicting claims not inherently incompatible with comprehensive pluralism in a hierarchical order, and requires that claims that rank higher in the hierarchy be given priority over those that rank lower. Such hierarchical ranking, moreover, is twofold: on the one hand, interests, value preferences, and objectives must be ranked in the order of importance that they have within a particular conception of the good; on the other hand, interests, value preferences, and objectives issuing from different conceptions of the good must be ranked based on a comparison of their relative importance within the conception of the good to which they are attached. Accordingly, in case it is impossible equally to accommodate two conflicting interests, value preferences or objectives, those that rank higher within the perspective of their proponents ought to prevail over those that rank lower within the perspective of theirs. For example, if two communities compete over a scarce resource but one demonstrates that the resource in question is essential to its very survival while the other concedes that it can survive without it, then justice according to reversible reciprocity requires that the resource be awarded to the first community. In short, when confronted with clashing claims that are not inherently incompatible with comprehensive pluralism, justice as reversible reciprocity prescribes satisfying the highest-ranking interests within each of the contending perspectives before turning to the next highest ranking interests within any of them. Finally, when it comes to equally ranked interests, justice as reversible reciprocity calls for satisfying first those that are most widely shared, thus placing reliance on the principle of majoritarian democracy.

From the standpoint of the adjudicator, the interpretive consequences derived from the normative constraints of comprehensive pluralism and from the dictates of justice as reversible reciprocity boil down essentially to three. First, for purposes of interpretation, laws ought to be treated as contracts or instances of self-legislation only to the extent that they do not contravene the priority of second-order norms over first-order norms. As already mentioned, ideally, the pertinent normative constraints will be con-

stitutionalized, and the adjudicator will strike down nonconforming contracts and self-legislation as unconstitutional. In other cases, the adjudicator will interpret laws as far as possible as consistent with pertinent normative constraints notwithstanding any contractual or self-legislative intent to the contrary.

The second and third interpretive consequences for the adjudicator, on the other hand, derive from justice as reversible reciprocity. The second corresponds to those cases in which justice as reversible reciprocity leads to an order of priority among the interests, value preferences, and objectives involved; the third, to those cases in which justice as reversible reciprocity yields no such priorities. In the second case, the adjudicator must ascertain what each party to the controversy seeks, compare all the perspectives represented in the controversy, establish an order of priority among conflicting claims, and vindicate that order of priority in the course of settling the meanings of laws. Moreover, ideally, respect for the requisite order of priorities ought to be constitutionalized. This could be accomplished, for example, by shaping existing constitutional rights as requiring respect for proportionality in ways that assure that the relevant priorities will be respected.

In the third case, all interests, value preferences, and objectives involved are in themselves equal in importance, with the consequence that the adjudicator should be guided by the principle of majoritarian democracy. Accordingly, laws emerging from a majoritarian process ought to be treated as instances of self-legislation, and the adjudicator should scrupulously track the legislative intent (or reconstruct it to the extent that the actual intent is difficult to discern) behind them in the course of articulating their meaning. Furthermore, to the extent that laws falling within this third category are best analogized to contracts, the adjudicator should rely heavily on the actual or reconstructed intent and objectives of those who are properly considered to be the contractors. Finally, laws within this third category are best reconstructed as the products of ordinary everyday majoritarian politics, and distinguished from mere politics by stressing their potential for stability and predictability.

To recapitulate: ideally and counterfactually, laws in the first category should be constitutionalized, those in the second category shaped by constitutional constraints, and those in the third category left to the will of the majority as extensions of politics that remain on the same plane as ordinary politics. On closer scrutiny, the second category encompasses a gray area between constitutional and ordinary legal norms. This results, in part, because the second category is not as fully constitutionalized as the first and, in part, because the divide between what should be subjected to hierarchical considerations and what should be left on an even plane is often blurred and contestable.

D—
Factoring the Difference between Constitutional Politics and Ordinary Politics in Reconstructive Adjudication

Based on these distinctions, it is now possible to draw a systematic distinction between constitutional politics and ordinary politics.^[9] Indeed, constitutional politics is confined to the first of the three categories of law identified above and to part of the second category. Ordinary politics, on the other hand, underlies all laws falling in the third category and has some influence on those within the second category but ought to have none within the first category.

Before exploring the implications of this proposed systematic division between constitutional and ordinary politics any further, however, it is necessary to focus briefly on the possible role of constitutional politics within the context of comprehensive pluralism. At first, it may seem that there is nothing political about constitutionalizing the normative constraints inherent in comprehensive pluralism—or, in other words, about translating the priority of second-order norms over first-order norms into a constitutional principle. On further reflection, however, even if all concerned remain unanimous in their commitment to implement the priority of second-order norms, they may still disagree on how best to achieve their commonly shared objective. Specifically, to what extent the priority of second-order norms ought to become embedded in constitutional norms (as opposed to purely ethical norms) and what precise form the embodiment of second-order norms in constitutional principles should take loom as matters about which reasonable proponents of comprehensive pluralism might disagree as a matter of constitutional politics. Furthermore, constitutional politics also seem bound to figure in connection with the determination of the degree to which laws in the second category ought to be molded by constitutional concerns. Finally, constitutional politics and ordinary politics might be expected to overlap in the gray area within the second category wherein the blurred boundaries between constitutional and ordinary legal norms must be set. In contrast, laws in the third category ought to fall exclusively within the grip of ordinary politics.

The above division of labor between constitutional politics and ordinary politics is systematic and driven by normative considerations emanating

[9] This distinction figures prominently in the work of Bruce Ackerman. See Ackerman 1991. Ackerman's account, however, is more intuitive than systematic. For him, constitutional politics occurs at extraordinary moments when the people become mobilized in pursuit of the common good. In contrast, in the context of ordinary politics, Ackerman argues that the people are motivated by narrowly drawn self-interest. Unlike Ackerman's distinction, the one offered here does not depend on particular historical events or on the mindset of the relevant actors within the polity. Instead, the distinction in question is normatively grounded in the perspective of comprehensive pluralism and systematized through counterfactual use.

from comprehensive pluralism. Moreover, the distinction between constitutional and ordinary politics is particularly important from a critical counterfactual standpoint. On the one hand, this distinction is useful for purposes of assessing constitution making;^[10] on the other hand, for purposes of critically evaluating actual lawmaking and legal adjudication. Furthermore, the very acknowledgment of constitutional politics presupposes that there is some room for contingency in the context of normative assessments of constitution making. Wherever counterfactual reconstruction of constitutional constraints consistent with comprehensive pluralism leaves room for constitutional politics, available alternatives can be profitably narrowed down. This can be accomplished by reference to the design of the actual constitution makers and to relevant historical narratives that link together the constitution's origins with its current interpretive profile. In sum, by blending relevant contingent events into the mold carved out by comprehensive pluralism, the adjudicator can find meaningful ways of bridging the gap between partial selves and elusive others and measurably narrow down the legitimate interpretive alternatives. Also, the adjudicator seems poised to achieve all this without losing sight of the importance of maintaining a cogent demarcation between the proper realm of constitutional constraints and that which ought to be left to majoritarian legislation.

Two serious objections may be raised at this point against relying on justice as reversible reciprocity to adjudicate claims among proponents of conflicting conceptions of the good in accordance with the requirements of comprehensive pluralism. The first objection is that conceptions of the good should be regarded as cohesive and indivisible wholes rather than as mere aggregates that might be readily disassembled and reassembled to suit needs for a hierarchical ranking of interests, value preferences, and objectives. The second objection is, in turn, that even if the first objection could be successfully addressed, comparisons cutting across different perspectives are either impossible to carry out or highly unlikely to uncover enough common ground to lead to fair resolutions of conflicts.^[11] I address these objections in order, starting with an examination of the case of clashing religions to determine whether it is necessary to treat conceptions of the good as indivisible wholes consistent with adherence to comprehensive pluralism.

[10] To the extent that constitutions are living experiments subject to evolution and constant adaptation, constitution making need not be understood in the narrow sense as merely referring to the work of actual constitutional assemblies. It can also be profitably interpreted in a broader counterfactual reconstructive sense as encompassing significant steps leading to important evolutionary adaptations.

[11] This last charge is reminiscent of the feminist challenge to Habermas's discourse theory of law discussed in chapter V. As we shall see below, however, comprehensive pluralism allows for a different response to this charge than does Habermas's discourse theory.

III—
The Case of Clashing Religions and the Limits of Justice as Reversible Reciprocity

Imagine a polity composed of two religious communities that are diametrically opposed to one another, not only on matters of religious dogma but also on all significant economic, social, and political matters. Imagine further that each of these religions holds that its beliefs, precepts, and practices are bound together in such an indissoluble bond that failure to abide by any of them constitutes mortal sin. Finally, imagine that neither of these two religions preaches the eradication of the other—thus not qualifying as a crusading religion—but that each requires that its precepts be fully carried by all its adherents in the public sphere as well as in the private sphere. For example, one religion may prescribe self-reliance and a free market economy while the other may command solidarity and a socialist economy. In such a situation, the best alternative would be to split the current polity into two and give each of the two religious communities complete control over its own destiny. But if for some reason that were not to prove a viable option, then justice as reversible reciprocity would indeed seem to lead to an impasse. Since, by definition, all interests, value preferences, and objectives within each of these religions are on an equal footing, and since all of them are of equal paramount importance for both religions, a reversal of perspectives fails to reveal any plausible course of action within the bounds of comprehensive pluralism.

On further thought, there might be a solution consistent with justice as reversible reciprocity, provided the two religious communities involved are not of the same size. If the majority of the citizens of the relevant polity belong to one of these religions, and the remainder to the other, then, arguably, majority rule ought to prevail, as all relevant interests, value preferences, and objectives are by stipulation on the same plane. Moreover, the consequences of this argument are far-reaching, for in every polity with a plurality of competing conceptions of the good, it seems fair to assume generally that proponents of one conception of the good are, by and large, as committed to it as proponents of other such conceptions are to theirs. But if this is true, then comprehensive pluralism seems essentially collapsible into majoritarianism.

In spite of its superficial appeal, the above argument is clearly wrong, as comprehensive pluralism, which is predicated on the inherent equal worth of all first-order perspectives as well as that of all persons, is certainly not reducible to majoritarianism. Indeed, as will be remembered, justice as reversible reciprocity is not supposed to be unlimited in scope but is instead made subject to the normative constraints imposed by comprehensive pluralism. Moreover, chief among these constraints is steadfast commitment to the priority of second-order norms over first-order norms, which requires

placing great value in protecting the integrity of (first-order) perspectives. In other words, comprehensive pluralism prescribes mutual respect between self and other, to be understood as respect for the integrity of each other's perspective. Accordingly, from the perspective of comprehensive pluralism, for majoritarianism to be legitimate, it must be consistent not only with justice as reversible reciprocity but also with preservation of the integrity of all the (first-order) perspectives involved. Consistent with this, to return to our example, since majoritarianism would lead to complete capitulation of the minority religion to the majority religion, it would patently fail to uphold the integrity of the minority religion and would hence be clearly illegitimate.

The limitation of justice as reversible reciprocity by the requirement to preserve the integrity of the perspectives respectively embraced by self and other has interpretive implications that will be addressed briefly in section V below. For the moment, I wish to concentrate on two of the most salient consequences that follow from the rejection of majoritarianism as the means to resolve the clash between the two religions involved in our example. First, paradoxically, the only way to resolve the conflict between these two all-encompassing religions consistent with the prescriptions of comprehensive pluralism is by preventing both of them from gaining control over the realm of intercommunal dealings. This follows from the very logic of comprehensive pluralism, for any other solution would result in subordination of one first-order perspective to another. Now, if there were no room whatsoever left for intracommunal dealings, the proposed solution would lack any practical consequence. Also, if because of the comprehensiveness of the religious ideologies involved, it were impossible to map out an institutional framework for intercommunal dealings that would not be fully subsumed under either of the two clashing religions, then the proposed solution would prove impossible. But otherwise, comprehensive pluralism would require resolving the clash between our two all-encompassing religions by expelling them both from the public sphere. Furthermore, the very nature of this solution makes it plain that comprehensive pluralism is firmly grounded in a substantive normative vision rather than being merely reducible to a process-based perspective.

The second salient consequence that follows from the rejection of a majoritarian solution to the clash between the two religions of our example is that, even if our minority religion counted with but a single adherent, comprehensive pluralism would still call for refusing to regulate intercommunal dealings in conformity with the dictates of either of the two religions. This may seem an absurd result, but it is nonetheless clearly mandated by the fundamental normative tenets of comprehensive pluralism. Indeed, from the standpoint of the equal preservation of the integrity of

each perspective within the polity, it makes no difference whether any such perspective is shared by nearly everyone or whether it is merely embraced by a single person.

If the last result is difficult to accept, it is not because it is at odds with the ethos of comprehensive pluralism but rather because recourse to comprehensive pluralism may be questioned under the highly rigid and artificial circumstances of our example. As already mentioned, I am agnostic as to whether it is desirable to promote the ethos of comprehensive pluralism in polities that are not in fact pluralistic. Moreover, even if we were to assume that comprehensive pluralism is intrinsically desirable in all circumstances, the case for it could still be much less compelling in some contexts than in others. Accordingly, the situation under consideration involving a completely homogeneous society saddled with a lone dissenter certainly presents close to the least compelling case for recourse to comprehensive pluralism.^[12]

If we broaden our focus beyond the artificial example discussed above, however, we are bound to notice that the appeal of comprehensive pluralism is likely to increase as we move to polities that are increasingly pluralistic in fact. Thus, as we reach a typical contemporary pluralistic polity with many different religions that are hardly monolithic; with a wide array of partial selves connected to various degrees to several perspectives; with wide-ranging overlaps of value preferences; and with a significant measure of fluidity among communal groupings; comprehensive pluralism leads to outcomes that are likely to be considered appealing by vast majorities, including many who reject comprehensive pluralism as a persuasive conception of the good. For example, the Judeo-Christian heritage that has been so influential in Western civilization is hardly monolithic in several relevant respects. First, all three of the major religions linked through that common heritage share common elements; second, none of the three is internally monolithic, as each encompasses a plurality of different views concerning several religious issues; third, none of them is all-encompassing in the strong sense of dictating particular courses of action in all human endeavors;^[13] and fourth, none of them rigidly places all its interests, value preferences, and objectives on the same footing.^[14] Furthermore, many citi-

[12] Actually, the minimum conditions necessary for comprehensive pluralism to make any sense are those of a completely homogeneous polity in which the possibility of opting out or dissenting from communal norms is at least acknowledged. If no such possibility were ever conceived, then the very notion of comprehensive pluralism would be altogether meaningless.

[13] For example, each of these three religions seems relatively open to a wide array of options in the realms of economic and social policies ranging from laissez-faire capitalism to socialist interventionism.

[14] Proponents of all three religions would thus presumably agree that adherence tomonotheism ranks as a higher-order value than abiding by even the most minor of religious prescriptions. See, e.g., Peters 1990, 201-11, 356-57.

zens in a contemporary pluralist polity are likely not to identify with any religion, and even among those who do, a sizable number are likely to identify with religion only in part. Yet others are likely to opt out of the religion of their ancestors, or to convert from one of the religions within the polity to another. In short, under such circumstances, some kind of separation between church and state may seem quite wise and highly advisable, not merely from the standpoint of comprehensive pluralism, but also from that of other perspectives, such as liberalism.^[15]

Complex and multifaceted ideologies seem clearly better suited than monolithic ones to offer fertile grounds for comprehensive pluralism. Moreover, the conjunction of respect for the integrity of perspectives with justice as reversible reciprocity is more likely to yield workable solutions if there is a fair degree of fluidity among the perspectives that contribute to the definition of the respective identities of self and other. In other words, there appears to be an increase in opportunities for mutually acceptable intercommunal regulation if communal boundaries are open and flexible rather than static or rigidly entrenched. But for the requisite openness and fluidity to be present requires that self and other be capable of internal transformation. Accordingly, I now briefly examine how comprehensive pluralism is poised to contribute to such transformation.

IV—
Comprehensive Pluralism and the Internal Transformation of Self and Other

As a consequence of its stress on the equal worth of different perspectives and on accommodating as many of them as possible, comprehensive pluralism sharpens our awareness of the plurality of available conceptions of the good. Moreover, adherence to the priority of second-order norms is bound to distance us somewhat from our own first-order value preferences while at the same time getting us closer to those espoused by others. This, incidentally, is not without danger, for the insecurity resulting from alienation from one's own value preferences many well lead to intolerance and

[15] In a religiously pluralistic society, the separation of church and state may well count as a desirable objective from within the perspective of at least some of the religions involved. It is therefore hardly surprising that some religious leaders have been among the most avid supporters of the separation of church and state in the United States. See, e.g., Testimony of J. Brent Walker, General Counsel and Assoc. Dir., Baptist Joint Comm. on Public Affairs Before the Senate Comm. on the Judiciary, Federal Document Clearing House, 25 October 1995, available in LEXIS, Nexis Library, CURNWS File; Curtis 1995; "Dissenting from the Right—Religious Right," Christian Century 112:631, 21 June 1995.

outright repudiation of normative pluralism. But, by the same token, commitment to the priority of second-order norms affords opportunities for self-enrichment and for better reconciliation between self and other, through greater interpenetration between their respective perspectives. In short, openness to each other's perspective makes for an internal transformation of self and other in ways that may be both more fulfilling and more apt to carve out mutually acceptable common paths without having to abandon cherished differences.

The propensity of comprehensive pluralism to lead to internal transformation may be best captured through consideration of an example. Imagine, therefore, a society that proclaims allegiance to comprehensive pluralism but that bans homosexuality and is completely intolerant of homosexual lifestyles. To reconcile its embrace of comprehensive pluralism with its intolerance of homosexuality, this society would have to articulate a reason why conceptions of the good promoting homosexual lifestyles should not be tolerated in the same way as are other nonmajoritarian conceptions of the good. Imagine further that such a society argues that homosexuality is abnormal and deviant and, as such, a threat to the preservation of the social fabric.^[16] If such an argument were accepted, then a society devoted to comprehensive pluralism would presumably be able to justify intolerance of homosexuality in the same way as it could justify intolerance of crusading religions.^[17]

To overcome the belief that homosexuality poses a threat to society, and to make acceptance of homosexuals possible, requires a multifaceted approach to combat prejudice and alter perceptions. For our purposes, however, what is crucial is the need for recharacterization, leading in this case to a higher level of abstraction. Instead of focusing on homosexual practices and on traditional attitudes toward them, the aim should be to stress the need of every person, heterosexual and homosexual alike, for intimate sexual relationships and to insist that all adults should be equally entitled to

[16] Cf. Chief Justice Burger's concurring opinion in Bowers v. Hardwick (1986) (referring to homosexuality as a "crime against nature" and as a "heinous act").

[17] There is, of course, one crucial difference between the two that ought to be recognized even by those who claim that homosexuality is abnormal. Whereas crusading religions preach a duty to proselytize by force if necessary, partisans of homosexual lifestyles are most unlikely to seek converts among the heterosexual population. Nevertheless, from the standpoint of maintaining a viable comprehensively pluralistic society, it is indifferent whether a perceived threat to its integrity is voluntary or involuntary. For example, if a religion forbade its adherents from being vaccinated, and the latter's refusal would result in the spreading of a deadly disease, the pluralist state would be justified in disregarding their conception of the good to the extent necessary to order them to be vaccinated against their will, even though they had no intention to harm or influence fellow citizens who do not share their religious faith.

privacy to pursue consensual sex with another adult.^[18] Obviously, the move to a higher level of abstraction in this case is designed to shift the focus from what is different about homosexuals to what heterosexuals and homosexuals share in common. But less obvious, perhaps, is that if the move is successful it will not merely result in adding the pursuit of homosexual lifestyles among the multiplicity of existing legitimate first-order conceptions of the good. The move will also undoubtedly have consequences within the latter conceptions of the good, such as further weakening beliefs in any necessary nexus between sex and procreation, or sex and marriage. Or the move may lead to calls for legitimizing homosexual marriages, causing a shift in attitudes toward marriage, and thus leading to transformations within groups adhering to conceptions of the good that place great value on traditional notions of marriage. Furthermore, increasing acceptability of homosexual lifestyles is also likely to have repercussions for particular religions traditionally opposed to homosexuality, as homosexual adherents become emboldened to seek changes within their religious community.^[19] Finally, assuming that homosexuals succeed in their quest for initial acceptability, they might well push for additional changes in pursuit of broader equality capable of encompassing differences and thus lead to further internal transformation within groups adhering to more traditional conceptions of the good.^[20]

One should not lose sight of the possibility that a push for greater acceptance of homosexuality might actually result in a hardening of positions and a backlash. Nevertheless, so long as the spirit of comprehensive pluralism prevails, moves toward greater inclusivity are likely to produce greater openness toward alien perspectives as well as internal transformations of self and other in enriching new directions. Indeed, such internal transformations may significantly broaden the possibilities for viable reconciliation of self and other in ways that remain protective of the integrity of all the relevant perspectives involved.

[18] This is the position embraced by the dissenters in the U.S. Supreme Court's 5-4 decision in Bowers v. Hardwick (1986), at 218-19, in which the Court refused to recognize a constitutional right to homosexual sex among consenting adults.

[19] See, e.g., Sullivan 1995 (arguing that the Catholic church should permit homosexual sex on the analogy of permitting heterosexual sex in the case of married infertile couples).

[20] In the pursuit of equality-as-identity, homosexuals may search for official recognition of same-sex marriages. In the pursuit of equality-as-difference, in contrast, homosexuals are more likely to pursue public alternatives no longer geared to favoring traditional marital relationships and family values. Compare Sullivan 1995 (arguing for assimilation of homosexuals) with Bersani 1995 (arguing against assimilation and emphasizing implications flowing from affirmation of gay sexuality).

V
The Interpretive Quest to Accommodate the Priority of Second-Order Norms and to Protect the Integrity of Perspectives

Based on the preceding observations, accommodating diverse perspectives while protecting their integrity need not lead to undue fragmentation of conceptions of the good. This is, in part, because in complex pluralistic societies perspectives are not likely to be monolithic or rigidly bounded; and, in part, because external reorderings of existing conceptions of the good to achieve greater inclusiveness are likely to be accompanied by internal transformations driven by the search for unity and continuity within the self. Actually, it is the partial self confronting the elusive other who seems most vulnerable to increasing fragmentation. On the other hand, the self's gesture of reconciliation toward the other, through more inclusive reconstruction and internal transformation, aims at (re)unification and cohesion. In other words, fragmentation is the by-product of pluralism in fact, whereas unification and cohesion loom as the objectives of comprehensive pluralism.

When internal transformation and the search for cohesion through reconstruction are placed in their proper context, it becomes apparent that rankings of interests, value preferences, and objectives, in accordance with justice as reversible reciprocity, do not necessarily lead to the destruction of the identity or integrity of conceptions of the good. As we have seen in the last two sections, conceptions of the good can be sufficiently elastic and adaptable to withstand frustration and reorientation of relatively less important objectives, as well as some measure of internal transformation, without fundamental loss of identity or integrity. Moreover, consistent with the dynamic environment characteristic of complex pluralistic societies, the identity of conceptions of the good is perhaps best analogized to that of a person who evolves throughout his or her entire life without losing the sense of self hood—that is, the sense that the same self somehow endures throughout all the changes.^[21] Accordingly, the first objection raised at the end of section II—that conceptions of the good must be regarded as indivisible wholes that cannot be partially accommodated based on a ranking of their priorities—ultimately misses the mark, for it is predicated on a misconception of the relative positions of self and other within a complex and constantly evolving pluralistic setting.

The second objection raised at the end of section II—that comparisons across different perspectives are either too difficult to carry out or unlikely to uncover common ground—however, seems more troubling. On the one

[21] See Ricoeur 1990 for a thorough and penetrating analysis of self-identity based on the contrast between identity derived from selfhood and identity derived from sameness.

hand, the transformability of self and other and of their respective perspectives appears to increase significantly the prospects for establishing common grounds. But, on the other hand, in order to achieve this, it is necessary to unleash potentially destabilizing interpretive strategies that raise at least two additional problems. First, as the example of the quest for inclusion of homosexual lifestyles discussed in section IV indicates, the kind of transformation required to open new common ground depends on reinterpretation of the implications of homosexuality. Such reinterpretation, moreover, relies on a shifting of levels of abstraction that upsets the clustering of identities and differences associated with the interplay of those perspectives granted accommodation within the status quo targeted by proponents of homosexual lifestyles. And, as already pointed out in chapter I, shifts in levels of abstraction tend to lessen the constraints that keep meanings relatively determinate and thus threaten to throw the interpretive enterprise into disarray. In particular, if comparisons across various perspectives tend to be complicated even without major shifts in levels of abstraction, then in cases involving wide fluctuations in such levels, fruitful use of the comparisons in question would seem much more difficult.

A—
Intraperspectival Commitments, lnterperspectival Comparisons, and Shifting Levels of Abstraction

The second problem, which is also related to the need for significant shifts in the levels of abstraction, is that proponents of change are prone to unleash transformations that they do not seek, or that they may even fear, but that they cannot control. Thus, challengers of the status quo may unwittingly destabilize settled meanings in ways that frustrate their own quest for inclusion or that provoke inclusions or exclusions they oppose. To illustrate this, let us return briefly to the example concerning homosexuality discussed in section IV. As indicated there, the key argument offered in support of tolerance for homosexuality is that decisions regarding intimate sexual relations among consenting adults should be left to the exclusive discretion of those concerned. This argument is pitched at a level of abstraction designed to draw attention to what heterosexuals and homosexuals share in common. However, this argument may also advance the purposes of others with whom proponents of tolerance for homosexuality might not wish to identify, including proponents of incest or of sadomasochistic sex. As a matter of fact, proponents of rights for homosexuals may be eager to differentiate homosexuality from incest or sadomasochism to the extent that the latter, which continue to convey connotations of deviance, had traditionally been lumped together with homosexuality.^[22]

[22] See Justice Harlan's concurring opinion in Griswold v. Connecticut (1965), at 499 (spe-cifying that the constitutional right to marital sexual intimacy should not be misconstrued as extending to homosexuality or incest).

These two problems, while real enough, are not insurmountable. Departures from the status quo cannot be achieved without some prior loosening of established meanings, but inasmuch as desired changes point to definite new directions and are subjected to the constraints imposed by comprehensive pluralism, they appear more prone to reorienting meanings rather than merely rendering them overly indeterminate. Furthermore, any claim for or against greater inclusivity necessarily involves some shifts in levels of abstraction and some unintended consequences in the course of pursuing an optimal mix between relevant identities and differences. It follows from this that all moves involve certain risks and that progress is by no means assured. For example, as discussed in chapter III, women who have achieved equality-as-identity may wish to improve their lot and reach a more equitable equal footing with men by aiming for equality-as-difference. But in calling attention to differences, women may unwittingly rekindle suppressed support for inequality based on difference and thus jeopardize already acquired gains while struggling for additional gains. Notwithstanding these uncertainties, claims do not arise in a vacuum, but rather in a context of relatively settled meanings. Moreover, in most cases only a limited horizon of plausible changes from the status quo would arguably lead to improvements from the standpoint of comprehensive pluralism. In short, meanings are not fixed, but it does not follow that anything goes. So long as we do not lose sight of the relevant context and of the direction in which comprehensive pluralism points, meanings can become sufficiently determinate to indicate which interpretations are likely to bring us closer to justice.

Looking at this from the micro-interpretive standpoint of the adjudicator, the key to success lies in the ability to carry out reliable comparisons across diverse perspectives. As already pointed out, three tasks confront the adjudicator: imputing an actual intent for self and other regarded as contractors or self-legislators; reconstructing their counterfactual intent as shaped by comprehensive pluralism; and interpreting the substantive constraints emanating from comprehensive pluralism so as to make them applicable to legal disputes. The second of these three task depends on the above-mentioned comparisons, for the counterfactual reconstruction at stake is designed to best approximate what self and other would have intended had they been aware of, and determined to abide by, the relevant order of priority encompassing the various interests, value preferences, and objectives of all involved. Moreover, this second task is crucial because actual intents must be canceled out to the extent that they prove incompatible with their counterfactually reconstructed counterparts, and because limitations pursuant to the substantive constraints of comprehensive pluralism

tend to weigh heavier on interpretive outcomes on failure by justice as reversible reciprocity to yield concrete answers.

Establishing proper intraperspectival and interperspectival comparisons thus represents important yet difficult undertakings for which the adjudicator must take responsibility. With respect to intraperspectival comparisons, the adjudicator must rely, in the first place, on the input of the proponents of the perspective involved. Moreover, the adjudicator must be mindful that such proponents may have an incentive to exaggerate or distort the importance of certain values within their own perspective in the course of pressing their interests in conflicts with interperspectival implications. To counter this potential for distortion, the adjudicator can test intraperspectival claims for consistency, for integrity over time, and against the claims of other proponents of the same perspective. Although claims based on individual idiosyncrasies may be particularly difficult to evaluate, it seems reasonable to assume that a vast majority of intraperspectival claims are communally rooted and thus shared by fairly identifiable groups of individuals. Accordingly, in most cases, the adjudicator will be able adequately to carry out intraperspectival comparisons for purposes of interpreting laws consistent with the dictates of comprehensive pluralism.

Interperspectival comparisons, on the other hand, loom as more delicate. Nevertheless, there are certain distinctions that can help to make the adjudicator's application of the criterion of justice as reversible reciprocity more manageable and more reliable. Three such distinctions deserve further elaboration: that between primary and secondary goods; that between primary and secondary values; and that between subject-exclusion and benefit (burden) -exclusion.

B—
Primary Versus Secondary Goods

The distinction between primary and secondary goods is derived from Rawls (1971, 90-95; 1993, 178-87), but as I conceive it in the context of comprehensive pluralism, it differs in many significant respects from the conception developed by Rawls. As understood here, primary goods are goods that inhere in the second-order conception of the good, or goods considered as such within each of the various first-order conceptions of the good present in the relevant polity. Primary goods deriving from the second-order conception of the good shape the substantive constraints of comprehensive pluralism and need not therefore be subjected to interperspectival comparison. In contrast, the remaining primary goods—whether they inhere within each relevant first-order conception of the good as a matter of necessity or of contingency—do figure in interperspectival comparisons and have priority over all other goods, which are thus relegated to the status of secondary goods.

It may be objected, however, that according automatic priority to primary goods is not always warranted because the fact that a given good is valued within each of the perspectives involved does not automatically warrant that such good is equally important for each of these perspectives. Suppose, for example, that all relevant perspectives regard achieving a minimum of subsistence for each individual as an important good, but that such good ranks as the highest priority from the standpoint of certain perspectives but not from that of others. Suppose further that pursuant to one of the conceptions of the good involved, abiding by certain religious precepts is a greater good than securing enough food for all the members of the faith. Under these conditions, if the only way to feed all members of the polity is by producing food using means that contravene the religious precepts identified above, and if no other means would be adequate for purposes of feeding all but those who feel bound by the religious precepts in question, then it would appear that giving priority to primary goods would violate the precept commanding that the highest-ranking goods within each perspective be pursued before addressing the next highest good within any perspective.

C—
Primary Versus Secondary Values

To avoid the problems posed by cases in which a primary good proves to be one of relatively low priority for some of the perspectives involved, it is helpful to supplement the distinction between primary and secondary goods with that between primary and secondary values.^[23] Primary values are those that rank the highest within a particular perspective, whereas secondary values are those that rank lower within that same perspective. Now, where primary goods and primary values tend to converge, there is a very strong case for the priority of primary goods. Moreover, whereas this is not logically required, it stands to reason that the two should often overlap. Thus, it would hardly be surprising if all the perspectives involved considered freedom from hunger and freedom from torture as goods, and if most of those perspectives considered these goods to be among their most highly valued ones. Consequently, if a primary good is involved and if it counts as a primary value for a vast majority of the perspectives involved, the adjudicator should be entitled to adopt the presumption that the primary good in question is entitled to priority.

This last presumption should be rebuttable, but only on a showing that, under the particular circumstances involved, giving priority to the primary

[23] This distinction is also made by John Kekes (1993, 18-19), but I do not follow his definition, which comes close to equating primary values with what I have described as primary goods.

good involved would only marginally advance the objectives of those perspectives for which that good represents a primary value while disproportionately disadvantaging those other perspectives that view pursuit of that good as a secondary value. Take, for example, the issue of whether euthanasia ought to be legally permissible in a polity in which the protection and preservation of life is a primary good and a primary value for most but not all existing perspectives. In this case, there ought to be a presumption against the legality of euthanasia. But suppose further that the proponents of euthanasia specify that they only support it in cases in which responsible adults in the possession of their full mental capacities freely choose it for quality-of-life reasons, and that they believe euthanasia to be inextricably linked to the primary value of individual control over his or her destiny. Moreover, a reversal of perspectives would not settle this issue, for it seems safe to anticipate that opponents of euthanasia would be as intensely committed to its prohibition as would proponents of its legalization. Proponents may further argue, however, that whereas prohibition of euthanasia would deprive them of something essential from the standpoint of their own conception of the good, legalizing euthanasia would by no means have a like impact on their opponents. Indeed, the opponents may find euthanasia profoundly ethically repulsive, but they would remain free to rule it out for themselves and to seek to persuade others to do likewise. Accordingly, proponents of euthanasia could plausibly argue that the harm to them following from its prohibition would be much greater than that to its opponents in case of legalization.^[24]

This last argument may seem very similar to the liberal argument made by John Stuart Mill that society has no business regulating the individual's purely self-regarding acts (1859, 73-74). In the context of comprehensive pluralism, however, the Millian distinction between self-regarding and other-regarding acts is generally considered to be untenable. This is consistent with the view that perspectives are fluid, open to external influences, anti susceptible to internal transformations and that partial selves and elusive others are in a constant state of dynamic interaction. Accordingly, so long as any intracommunal practice is not devoid of any conceivable intercommunal consequence, it may not be deemed self-regarding. Viewed in this light, legalizing euthanasia is certainly likely to have repercussions on the perspectives of its opponents and on the membership of opponent groups. That does not mean that communities that oppose euthanasia should be granted a veto over the decision of other communities within the

[24] See Cruzan v. Director, Missouri Department of Health (1990), at 314 (Brennan, J., dissenting) (the state's "general interest in life" is outweighed by a person's "particularized and intense interest in self-determination" in determining whether or not to continue receiving life support).

same polity to legitimate euthanasia for their own members, but comprehensive pluralism requires that this conflict be settled on the basis of something beyond the distinction between self-regarding and other-regarding acts.

D—
The Distinction between Subject-Exclusion and Benefit-Exclusion

Before exploring any further how comprehensive pluralism might settle this conflict concerning euthanasia, it is time to introduce the third distinction mentioned above, namely that between subject-exclusion and benefit-exclusion. This latter distinction is primarily helpful when it comes to determining whether a law is consistent with respect for the equal worth of self and other, or, more precisely, whether a law complying with formal equality requirements derived from mere reciprocity is thus consistent.^[25] All laws classify and as a consequence are prone to benefit members of certain classes in ways that members of other classes are not benefited. For example, a law that provides that no one under eighteen years of age can drive an automobile excludes a sizable portion of the population from the benefit of driving. Benefit-exclusions are not objectionable per se, particularly when they relate to matters properly left to majoritarian politics consistent with comprehensive pluralism. But some benefit-exclusions may become tantamount to subject-exclusion—that is, to exclusion from membership in the relevant community, or, in other words, to exclusion from equal membership in the polity in ways that are inconsistent with the equal worth of self and other.^[26]

A further distinction must be drawn between a per se subject-exclusion and a comparative subject-exclusion. The former may either involve an outright denial of membership in the relevant community or a benefit-exclusion involving a primary good that is also a primary value. The latter, on the other hand, only involves a benefit-exclusion, but because others who are similarly situated are not thus excluded, it becomes an indirect means to achieve a subject-exclusion. For example, in a wealthy polity in which every citizen has the means to educate his or her own children, enactment of a law barring public expenditures for education would be reasonably deemed to produce benefit-exclusions without thereby becoming the

[25] Although the distinction between subject-exclusion and benefit-exclusion is relevant in both these situations, cases involving a failure to meet formal equality requirements are more efficiently dealt with by means of direct reference to the dictates of mere reciprocity.

[26] Not all departures from equal membership in the polity necessarily involve violations of the equal-worth principle. For example, at least arguably, denying short-term foreign transients full voting rights does not amount to a violation of their right to equal worth. On the other hand, denying women full voting rights would obviously be contrary to upholding their equal worth.

source of any subject-exclusion. Furthermore, if the state restricts free public education to those who could not otherwise pay for their children's education, the benefit-exclusion that would only affect the wealthier members of the polity would still not amount to subject-exclusion. But if free public education were only denied to a persecuted ethnic minority, then the resulting benefit-exclusion would be tantamount to subject-exclusion—or, more specifically, to comparative subject-exclusion—inasmuch as it is the fact of being singled out for different treatment than that reserved for others who are similarly situated, rather than the benefit-exclusion itself, which results in the denial of equal worth. From an interpretive standpoint, per se subject-exclusions can be weeded out through implementation of the mere reciprocity criterion. And so can some comparative subject-exclusions, such as, for example, those based on racial discrimination. But others cannot be thus eliminated, nor can they be brought to light through implementation of the reversal of perspectives in accordance with, justice as reversible reciprocity. Moreover, the above example concerning euthanasia appears to fit within the latter pattern. Indeed, the conflict between proponents and opponents of euthanasia cannot be resolved at the level of mere reciprocity as they mutually recognize each other as possessors of a perspective, and as neither of them are bent on eliminating or suppressing the perspective of the other but are only determined to pursue their own conception of the good in a fair and open competition. Also, as we have seen, the conflict cannot be resolved at least with respect to value preferences by appealing to justice as reversible reciprocity since protagonists on both sides of the conflict advance primary values.

It is conceivable, however, that justice as reversible reciprocity might resolve the conflict through focus on the respective interests involved. Thus, pursuant to a reversal of perspectives, opponents of euthanasia might conclude that their interests in preventing others from choosing euthanasia because of fears about repercussions on the polity as a whole is somewhat less central vis-à-vis their own perspective than the proponents' interest in freedom to choose euthanasia happens to be in relation to theirs. But it seems equally plausible that the opponents of euthanasia would conclude that their own interests are just as strong from the standpoint of their perspective as the interests of the proponents of euthanasia is from theirs.

If justice as reversible reciprocity cannot settle the issue, it seems that we have reached an impasse: either the perspective of the proponents of euthanasia must prevail at the expense of that of its opponents or, conversely, the latter must be vindicated at the expense of the former. Moreover, unlike in the case of the clashing religions examined in section III, the conflict at hand cannot be settled by relegating the antagonists to the precincts of some walled-off private sphere. But also, unlike in the case of the clashing

religions, the antagonists in the case of euthanasia are not otherwise completely at odds with each other.

Building on the commonly shared values by proponents and opponents of euthanasia, and relying on the asymmetry between their respective positions, the adjudicator can overcome the apparent impasse mentioned above through reliance on the distinction between benefit-exclusion and comparative subject-exclusion. As postulated above, all involved agree that protection of life is a primary good but disagree as to whether it is a primary value. If the views of the proponents of euthanasia are embedded in legislation applicable to the polity as a whole, opponents of euthanasia can still vindicate their primary values intracommunally though they are prevented from doing so intercommunally. On the other hand, if euthanasia becomes illegal, its proponents will be prohibited from implementing their primary values even within the limits of their own community. Also, without questioning the genuineness of the opponent's fear that recognition of euthanasia as legitimate might lead to erosion of the polity's commitment to the protection of life, the fact that the proponents of euthanasia embrace the protection of life as one of their value preferences seems to mitigate the dangers of the dreaded erosion. Under these circumstances, therefore, it seems fair to characterize frustration of the objectives of euthanasia's opponents as involving a benefit-exclusion and, in contrast, frustration of the objectives of euthanasia proponents as, in comparison, amounting to a subject-exclusion.

There may be other ways to establish that outright prohibition of euthanasia under the circumstances of the above example would be violative of the duty to uphold the equal worth of self and other. That does not detract, however, from the usefulness of the distinction between benefit-exclusion and subject-exclusion to the adjudicator charged with resolving legal conflicts in ways that are consistent with the precepts of comprehensive pluralism. More generally, the latter distinction together with those between primary and secondary goods, and between primary and secondary values, with the normative constraints emanating from comprehensive pluralism, and with the criterion of justice as reversible reciprocity, furnish a fairly comprehensive set of interpretive devices enabling the adjudicator to arrive at just interpretations. In many cases, this means that the adjudicator will be led to a determinate resolution of the conflict at hand, in a way that best promotes justice under the circumstances without ever fully reaching it. In other cases, the available interpretive tools may not enable the adjudicator to settle on a determinate solution but may point to a particular direction that provides the best means, under the circumstances, to lessen injustice and move closer to the reconciliation between self and other.

The actual solutions reached through deployment of the interpretive

apparatus emerging out of comprehensive pluralism may often be similar or even identical to those that emanate from monistic conceptions of the good, such as liberalism. Thus, for example, the solution pursuant to comprehensive pluralism to the conflict concerning euthanasia could also be justified by reference to the Millian distinction between self-regarding and other-regarding acts. Nevertheless, it is crucial not to be lulled by the similarity of certain solutions into blurring the divide between comprehensive pluralism and other perspectives. For one thing, in spite of similarities, solutions under comprehensive pluralism may differ in scope; for another, solutions under comprehensive pluralism are context dependent and hence turn on the particular contingencies involved in ways that monistic solutions do not. Finally, solutions under comprehensive pluralism also differ from those under relativism, in that the latter are contextualized through and through, whereas the former are contextualized only to the extent consistent with adhesion to the mold carved out by comprehensive pluralism.

In the last analysis, the interpretive enterprise charted by comprehensive pluralism is characterized by the processing of an ever-changing flow of contingent inputs through a steady and fairly thoroughly defined mold. The functioning of this interpretive enterprise, moreover, is perhaps best captured by reference to the analogies between semantic value in intertextual exchanges and economic value in market exchanges discussed in chapter I. Focusing more particularly on market exchanges, the contingent inputs molded by the various perspectives and vantage points found throughout the polity are analogous to the contingent desires that motivate individuals to participate in market transactions. Furthermore, there is an analogy between the invisible hand mechanism, which allows the market to channel clashes between actors who seek to satisfy self-interest into the production of the common good, and the structured bundle of interpretive means, which allow the adjudicator to resolve conflicts in ways that lead to greater justice and better reconciliation between self and other. Of course, this last analogy is far from complete, as the adjudicator, unlike the rules of market competition, must wield a very visible hand charged with imposing substantive constraints as well as procedural ones. The fact remains, nonetheless, that so long as the adjudicator remains faithful to the interpretive enterprise defined by comprehensive pluralism, there is every reason to believe that he or she will be able to resolve conflicts in the best possible way. This conclusion is also reinforced if we realize that, in spite of all the contingencies in play over time, the plausible alternatives available to a good faith adjudicator at any particular moment in time are likely to be rather limited in number. Indeed, on the analogy to the values emerging from market exchanges and to the semantic values issuing from intertextual exchanges discussed in chapter I, the interplay between contingent factors and the

enduring mold carved out by comprehensive pluralism will most likely yield a narrow range of legitimate semantic path openings and path closings, hence greatly reducing the number of potentially acceptable solutions. Accordingly, although meanings constantly change under comprehensive pluralism, such changes are likely to be both incremental and limited when gauged in terms of the passage from one moment to the next. Because of this, the conscientious adjudicator bears but a relatively small risk of being unwittingly led astray.

VI—
Difficult Cases and Pathologies

A—
Revisiting Abortion

That the amount of uncertainty or indeterminacy in ordinary cases may amount to less than meets the eye does not sufficiently indicate how comprehensive pluralism might fare in difficult cases. Moreover, the example involving euthanasia discussed above does not qualify as a truly difficult case, because it was carefully limited to consenting adults in full possession of their mental capacities making decisions in their own case. Abortion, however, does constitute a genuinely difficult case, if not the difficult case par excellence. Indeed, as pointed out in the course of the brief discussion of abortion in chapter V, neither universal morality, nor an ethics predicated on the reversal of perspectives, nor politics, can authoritatively resolve the profound and intense conflicts that abortion arouses in pluralist societies. The question that remains is whether comprehensive pluralism is suited to fare any better. Also, consideration of this question is likely to provide a fair sense of the limits of comprehensive pluralism and, in particular, of the outer bounds of its interpretive potential.

What makes abortion such a difficult issue to handle is that it combines religious and secular aspects, and, depending on one's perspective, it is either central to the equal worth of women and to the integrity of certain conceptions of the good or violative of the most elementary duty to the other as it inevitably results in the annihilation of an other. Certain opponents of abortion base their position on religious grounds and confront antagonists who embrace a secular perspective. Presumably, in such a case, the conflict over abortion cannot be resolved like that between clashing religions considered in section III. Other opponents of abortion may offer support for their position without relying on religion and thus trigger a secular conflict that seems no more amenable to legitimate resolution pursuant to comprehensive pluralism than that in which religious views are pitted against secular ones. Thus, secular opponents of abortion may claim that fetuses must be deemed to be persons from an ethical, constitutional,

and legal standpoint^[27] and that consequently abortion is inadmissible as violative of the minimum requirements of mere reciprocity. Against this, secular proponents of abortion would argue that personhood is contingent on birth and that, accordingly, prohibiting abortion signifies denial of the equal worth of women and disregard of the integrity of the perspective of those who view the right to abortion as essential for purposes of maintaining control over one's body and one's destiny.

Ronald Dworkin has proposed a more unusual approach to the problem centered on treating differences over abortion as the product of a clash among antagonistic religious ideologies (1993, 30-178). According to Dworkin, the clash over abortion is not primarily about whether the fetus ought to be considered as possessing the attributes of personhood but rather about which of two diametrically opposed conceptions of the sanctity of life is the right one. Opponents of abortion, claims Dworkin, view the sanctity of life in terms of reverence for the life-generating processes of nature (ibid., 11, 92). Proponents of abortion, on the other hand, conceive of respect for the sanctity of life as requiring a commitment to upholding a certain quality of life (ibid., 33, 97). Accordingly, when a fetus is defective with but a short and painful life ahead for it, or when carrying the pregnancy to term would have a significantly adverse effect on the quality of life of the mother or of her loved ones, then abortion would be consistent with respect and concern for the sanctity of life (ibid., 33, 97-100). What is particularly attractive about Dworkin's approach is that it allows the conflict over abortion to be resolved much like that between clashing religions considered in section III. As Dworkin sees it, inasmuch as the conflict over abortion is a religious one, the state ought to stay away from it, and allow each person to follow his or her conscience (ibid., 26, 164-65). Consequently, from a conceptual standpoint, the state should favor neither the proponents nor the opponents of abortion and should remove the issue of abortion from the agenda of the public intercommunal sphere. From a practical standpoint, however, prohibiting state intervention when it comes to abortion is tantamount to a constitutional injunction against banning abortions and thus appears to hand a clear victory to the proponents of abortion.

Even if the conflict over abortion were reducible to a clash among an-

[27] This position is contrary to that adopted by the majority on the Supreme Court in Roe v. Wade (1973) in which it was decided that the fetus was not a "person" for constitutional purposes. Subsequent to the Roe decision, however, certain members of Congress advocated a constitutional amendment that would provide that the fetus must be considered to be a person for constitutional purposes, on conception. See S. 158, 97th Cong., 1st sess. (1981); H.R.J. Res. 62, 97th Cong., 1st sess. (1981); see also Wesffall 1982, 97-102.

tagonistic religious perspectives, Dworkin's approach proves, on further consideration, to be ultimately unsatisfactory. This is because casting irreconcilable differences over abortion as a religious conflict does not eliminate the residual issue concerning the legality of abortion. Consistent with comprehensive pluralism, neither of the two religious perspectives involved ought to influence the determination of whether abortion should be legal or illegal. Moreover, if no relevant secular considerations were pertinent, then it would seem that comprehensive pluralism would require some determination of the legal issue, but would be indifferent as between making abortion legal or illegal. In any event, the residual issue would not vanish and would necessarily call for some kind of resolution.

To illustrate this consider the following example. A polity is equally divided among two religious groups, one of which requires the sacrifice of the firstborn child on his or her first birthday; and the other of which places the sanctity of life above all, and considers it a paramount divine command to intervene for purposes of saving the lives of those who are unable to protect themselves. It is obvious that, under these circumstances, the state must intervene to prevent a religious civil war and that regardless of the actual nature of such intervention, one of the two religious groups will be incidentally favored and the other frustrated. Thus, even something as seemingly neutral as imposing a strict physical separation between the two groups would favor the group that sacrifices babies and frustrate the other group.

Returning to abortion, not only is the legal issue unavoidable, but the matter is not essentially reducible to religion, notwithstanding Dworkin's argument. Dworkin seeks to bolster his position by noting that opponents of abortion, who nonetheless are prepared to lift the ban on abortion in cases of rape or incest or to save the life of the mother, can only justify their position by appealing to the sanctity of life, rather than to the proposition that personhood attaches at conception (ibid., 94-97). It is true that if the fetus is deemed entitled to full personhood rights, the fact that the pregnancy originated in rape or incest would not, in itself, justify altering the rights of the fetus. But if rights to personhood are not considered absolute, or if, in a universe of partial selves and elusive others, actual positions need not be expected to be consistent, then it does not follow that opponents of abortion must rely on the sanctity of life on the pain of logical or practical contradiction. In any event, there are opponents of abortion who allow for no exception, even saving the life of the mother, and who would thus remain consistent with the position that the fetus is entitled to the full rights of personhood even on Dworkin's own terms. Furthermore, there are arguments on both sides of the abortion issue that may be supported without any appeal to religion or religious aspirations. In short, in the context of a

contemporary pluralistic polity, no matter how deeply the abortion debate happens to be immersed in religion, there are bound to remain residual secular arguments for and against abortion, as well as the inescapable residual secular issue concerning its legality.

Abortion drives comprehensive pluralism to the limit, because it cannot be confined within the realm of first-order norms. Inevitably, to the extent that it raises the question of whether personhood ought to be extended back to cover the unborn fetus on grounds that are arguably exclusively secular, abortion triggers second-order normative issues. Specifically, abortion problematizes the question of membership within the polity, which is a critical question on which the entire edifice of comprehensive pluralism rests. Indeed, as noted already, the chief objective of comprehensive pluralism is to lead to the best possible reconciliation between self and other in ways that minimize violence and subordination. Because of this, the integrity of comprehensive pluralism hinges on avoiding bias in determining who shall count as a self for purposes of implementing second-order norms and of according equal consideration to all first-order perspectives that happen to have a proponent within the polity. Moreover, the requisite integrity arguably depends on the determination that every person—understood broadly as encompassing individuals and communal groups—within the polity qualifies as a self. But as soon as personhood becomes contested, and especially if those involved advance conceptions of personhood predicated on particular first-order perspectives, the integrity of comprehensive pluralism becomes severely and even potentially fatally compromised. For if proponents of a first-order perspective can impose criteria for membership within the normative space carved out for the polity, they become empowered to subordinate or exclude other first-order perspectives and thus undermine the fundamental precepts of comprehensive pluralism.

To the extent that the conflict on abortion turns on the plausible secular limits of personhood, comprehensive pluralism cannot advance its resolution. Although it may seem on the surface that extending personhood to the fetus makes for greater inclusivity and ought therefore be prescribed by comprehensive pluralism, on closer inspection, this does not follow at all. Because of the unique dependency of the fetus on its mother, granting personhood to the fetus cannot be simply equated with the mere addition of yet another self or first-order perspective to the previously acknowledged total within the polity. Erroneous extension of personhood to the fetus would have second-order, not merely first-order, consequences for the mother, as it would deprive her of her fundamental right to equal worth. Consequently, comprehensive pluralism cannot provide an answer to the question whether personhood ought to extend to the fetus. And because this question exceeds its limits, the best that comprehensive pluralism can do is to deal with issues relating to abortion after the decision concerning

personhood has been made. Or, in other words, from a logical standpoint, comprehensive pluralism enters the scene after the decision concerning extending personhood to the fetus has been made.

In view of this, it would seem preferable, as an interpretive matter, to enshrine a polity's initial decision on abortion in its constitution. If the constitution explicitly permits or prohibits abortion, the adjudicator is spared the need to confront an interpretively impossible task that would inevitably lead him or her to an arbitrary decision. Enshrining a provision dealing explicitly with abortion in the constitution, however, does not solve the problem from the standpoint of constitutional justice. Indeed, regardless of what the relevant constitutional texts actually provide, the problem posed by abortion cannot be definitively settled at the level of counterfactual reconstruction.

In spite of the ultimate impossibility of resolving the problem of abortion under present material and ideological conditions,^[28] and in spite of the semantic indeterminacy to which this impossibility leads, there remains nevertheless an important interpretive contribution with respect to the conflict over abortion which can be made within the confines of comprehensive pluralism. Regardless of how the ultimate decision concerning whether to extend personhood to fetuses is made, there are at least two crucial interpretive tasks that are well within the purview of the adjudicator working within the bounds of comprehensive pluralism. The first of these is to weed out religious from secular arguments concerning abortion; the second, to explore to what extent divisions might be plausibly mended regardless of the outcome of the ultimate conflict.

With respect to the first of these tasks, it is imperative to sort out the religious from the secular grounds of arguments concerning abortion because only the latter may have second-order normative implications. To the extent that secular and religious considerations only relate to first-order norms, they stand on the same plane. But because of comprehensive pluralism's refusal to ground its normative universe on metaphysical foundations, only secular considerations may be taken into account for purposes of determining the scope of second-order norms. Consequently, notwithstanding that comprehensive pluralism cannot settle whether personhood should be extended to the fetus, it can, and does, rule out adopting any religious perspective on abortion and making it binding for the polity as a whole.^[29] Conversely, within the relevant operative secular constraints, the

[28] It bears emphasizing that to a significant extent the problem of abortion is contingent on the limits of available technologies. Thus, if acceptable means to end dependency of the fetus on its mother were in place, granting rights of personhood to the fetus would presumably not entail any threat to the mother's equal worth.

[29] It follows from this that the adjudicator must weed out religiously grounded positionson abortion, but such positions should only be excluded so long as they cannot be independently supported on the basis of secular arguments.

adjudicator should endeavor to ensure that proponents of a religious perspective should be enabled, as far as possible, to act on their convictions about abortion within their own communities.^[30]

Concerning the second task identified above, the adjudicator ought to take responsibility for ironing out contradictions and inconsistencies, as well as for locating potential areas for compromise, notwithstanding existing divisions. Moreover, in performing this latter task the adjudicator should be aided by comprehensive pluralism's rejection of monism and its rigid insistence on postulating overriding values or objectives (Kekes 1993, 19). For example, even against the most extreme opponents of abortion who refuse to recognize any exception whatsoever, the adjudicator may legitimately carve out an exception for cases in which the mother's life is endangered. Indeed, neither mere reciprocity nor justice as reversible reciprocity requires sacrificing one's own life to save another life. More generally, submitting conflicting claims to justice as reversible reciprocity may often result in narrowing the gap between the antagonists, regardless of whether rights of personhood have been extended to the fetus in the first place.^[31] In short, comprehensive pluralism cannot resolve the conflict over abortion, but it can reduce the tensions that surround it.

B—
Pathologies

Conflicts can be defused, either by avoidance or as a consequence of increased understanding and empathy. The preceding review of interpretive practices associated with comprehensive pluralism revealed that the latter makes use of both these approaches. Avoidance, however, proves to be a double-edged sword. On the one hand, in certain cases, such as that of the clashing religions explored in section III, avoidance appears to be the best available alternative. Indeed, when two religions are on a collision course, mechanisms of avoidance, such as relegation of each religion to intracommunal bounds, may provide the best, if not the only, means to prevent em-

[30] In this connection, opponents on religious grounds to abortion would, save in the most extraordinary circumstances, never be obligated to have an abortion. Religious proponents of abortion, on the other hand, cannot be given equivalent assurances, for if abortion were banned on secular grounds, they could not be exempted from such prohibition.

[31] In this connection, it is interesting to note the contrast between the ways in which the U.S. Supreme Court and the German Constitutional Court have approached the controversy over the constitutionality of abortion. Whereas both courts have reached fairly similar results from a practical standpoint, the U.S. Supreme Court has departed from the privacy and liberty interests of the pregnant woman, see Roe v. Wade (1973) and Planned Parenthood of Southeastern Pennsylvania v. Casey (1992), and the German Constitutional Court, from protection of the life of the unborn, see Abortion Case (1975) and Abortion Case 2 (1993).

barking on a course toward mutual destruction. But, on the other hand, if avoidance extends well beyond where it is necessary, it seems bound to lead to pathologies. Specifically, excessive avoidance would tend unduly to curtail intercommunal dealings and to foster retrenchment within the walls of one's own community. Moreover, with respect to the remaining intercommunal encounters, a policy of avoidance would presumably lead to an excess of caution and indifference.

Understanding and empathy, in contrast, potentially draw communities closer together, and ought therefore to be fostered wherever a policy of avoidance is no longer needed to maintain the peace. However, although this may seem less obvious than in the case of avoidance, understanding and empathy can also lead to pathologies if pushed beyond certain limits. As already noted, exclusive reliance on the reversal of perspectives is unsatisfactory because it may lead to an excess of self-denial. Moreover, overemphasis on understanding and on empathy is likely to stunt the drive toward self-expression and self-realization and to transform the polity into a bland yet repressive environment in which controversy and difference become largely muted. Indeed, if taken too far, the fear of offending the other may become completely paralyzing.

Warding off the alienation resulting from excessive avoidance and the dull conformism stemming from an understanding steeped in disproportionate self-suppression requires recourse to certain prophylactic interpretive measures. From the standpoint of the adjudicator, it is imperative to search for interpretive paths that reduce reliance on avoidance to the necessary minimum and that emphasize the desirability and potential of diversity as much as possible. From the standpoint of the state, on the other hand, it is highly advisable to forgo any pretense to neutrality and to actively promote through education, policies, and laws the virtues of diversity and tolerance, mutual respect for self and other, and the numerous attractions of the value preferences inherent in normative pluralism.

VII—
Concluding Remarks

The fear that there is nothing beyond or behind interpretation gave particular urgency and poignancy to the quest for just interpretations. The pursuit of that quest, moreover, has taken us on a long and varied journey from the unsettling landscape of deconstruction to the more promising terrain of reconstruction. This journey has been marked by tales of unraveling followed by efforts to recombine the resulting seemingly disparate fragments into cogent narratives suggestive of ways in which law, ethics, and politics might be harmonized. Many reasons, such as misunderstandings, faulty reasoning, failures in perception, or biases linked to vantage points, to name a few, may be responsible for disagreements on the meaning of

legal norms. In societies that are pluralistic in fact, however, destabilization and problematization of meanings is traceable, above all, to the conflict among competing conceptions of the good. Due to loss of metaphysical certainty and erosion of confidence in the authority of reason and in the belief in the inevitability of progress, which were driving forces behind the emergence of modernism, the postmodern polity has experienced a seemingly insurmountable split between justice according to law and justice beyond law. Accordingly, the destabilization of legal meanings goes hand in hand with law's detachment from its traditional normative moorings. Furthermore, as a consequence of this, law loses its aura of authority and legitimacy and becomes equated with raw, unmediated power. In sum, law seemingly becomes purely political when, in fact, the destabilization of meaning proves the result of the fragmentation of its ethical base.

Under these postmodern conditions, legal interpretation emerges primarily as a source of anxiety and uncertainty. Moreover, deconstruction, which eventually proves to have paved the way to the road to recovery, looms initially as the main culprit in the undermining of seemingly stable and traditionally rooted legal meanings. With meanings uprooted, plausible interpretations greatly multiply but increasingly fail to provide any comfort as they leave matters more and more unsettled. It is no surprise, therefore, that serious efforts to solve the contemporary crisis in interpretation, as we have seen, have led back to the appealing strictures of formalism and prompted various different attempts at reinstating the authority of law by putting an end to interpretation.

Neither formalism nor the end of interpretation, however, can ultimately furnish a viable and legitimate solution to the contemporary crisis in interpretation. As our inquiry has revealed, so long as one looks to overcoming pluralism and to transcending the subjective element in interpretation as the means to overcome this crisis, all efforts to reach a satisfactory solution are bound to end in frustration. Contrary to expectations, it is only by accepting pluralism and by acknowledging that it is impossible to arrive at objective interpretations, or at some neutral and objective ground lying somewhere beyond interpretation, that we can work our way out of this crisis. More precisely, it is not only by accepting the fact of pluralism but also by embracing comprehensive pluralism as a normative imperative, and by realizing that interpretation is ultimately neither purely subjective nor objective but intersubjective as the joint product of self and other, that the path to recovery from the crisis can be located.

Pluralism, understood broadly, is thus both the source of the crisis in interpretation and the key to its solution. Moreover, deconstruction, which seems at first synonymous with the destabilization of meaning, in fact plays a key role in the process of recovering meaning under postmodern conditions. Indeed, deconstruction, understood in its broader ontological and

ethical sense rather than in its narrowest purely methodological sense, is not reducible to the absolute opposite of construction or reconstruction. Instead, when properly considered, deconstruction and reconstruction make up the two sides of the same coin. They complement one another in ridding the polity of obsolete entrenched meanings derived from discarded metaphysical visions, thus clearing the way for the elaboration of intersubjectively negotiated meanings aimed at narrowing the gap between self and other.

Comprehensive pluralism turns the fact of pluralism into a normative imperative, while its correlate, comprehensive justice, transforms the impossibility of justice into the launching pad for the necessary and unending quest for greater, yet imperfect, postmetaphysical justice. Awareness of the impossibility of justice is dangerous, for it can easily mislead one into believing that all possible ways of relating to the other, no matter how oppressive, are ultimately normatively equivalent. But when such awareness is coupled with commitment to the ethos of comprehensive pluralism, it is potentially reassuring and liberating. Although this is initially bound to seem paradoxical, it is the very impossibility of justice, as revealed through the conjunction of comprehensive pluralism and comprehensive justice, that leads to the realization that not all actions are normatively equivalent and not all outcomes equally unjust. Accordingly, justice may never amount to more than a hope, but that hope renders the postmetaphysical search for greater justice or lesser injustice meaningful. Moreover, consistent with this, justice according to law can neither be fully severed from nor fully reconciled with justice beyond law. Instead, as we have seen, the two are inextricably related to one another. The postmetaphysical temptation may be to make justice according to law autonomous, but the best solution turns out to depend on it remaining at once complementary and antagonistic to justice beyond law.

Finally, in the context of comprehensive pluralism and comprehensive justice oriented toward the reconciliation of self and other, and of the consequent dual relationship between justice according to law and justice beyond law, there is room for just interpretations, but not for anything beyond interpretation. In other words, within the universe delimited by comprehensive pluralism, the two senses of just interpretations converge. Because there is nothing behind or beyond interpretation, just interpretations—in the sense of the right interpretations—are ultimately no more than just interpretations—in the sense of mere interpretations. Indeed, in the absence of a metaphysical realm beyond that of our daily existence, and of an objective reality behind the plane of our intersubjective dealings with others, interaction between self and other is through interpretation, and the optimal possible reconciliation between self and other through the best

possible interpretation. In sum, all intersubjective meaning is derived from interpretation, and law, ethics, and politics, through which self and other seek to make the best of their common predicament, are at bottom but interpretation. At the end, the best we can do is to aim at a justice that depends on interpretation. That may not seem to be much, but it is everything.