“Is toleration a political virtue?” The question sounds rhetorical. Toleration is usually considered the fundamental, even constitutive virtue of liberalism, and its characteristic playground is the political. What can it be other than a political virtue? In this chapter, I will attempt to answer this allegedly rhetorical question in the negative and to argue that toleration is neither political nor a virtue, at least in the strict sense that I will try to elaborate. This statement certainly sounds odd, especially to political scientists and legal theorists. But then, provocative statements are often made by philosophers only to be later tempered and qualified, which is exactly what I will try to do after arguing for a non-political and non-aretaic concept of toleration.
As everybody familiar with the vast literature on toleration knows, the major obstacle in the philosophical analysis of the concept is characterizing what it is not. Two methodological approaches for such a characterization suggest themselves: the broad view, which tries to do justice to the large variety of contexts and linguistic uses with which the concept is and has been associated, and the narrow view, which delineates the contours of the concept in the light of its theoretically distinguishing features. The first method is historical or sociological in nature, whereas the second is philosophical or normative. The historical view is liberal enough to include under the title of toleration political and social phenomena that were either not called by that name in the past or are no longer treated as cases of toleration. The philosophical view is more restrictive, filtering out those phenomena that do not satisfy certain theoretical conditions even if they are in many respects similar to toleration.1
In effect, neither of these two methods should be followed in a pure and exclusive way. To put it in Kantian terms, an historical study of toleration with no theoretical guidance is blind; a philosophical-normative analysis of the concept with no regard to its actual evolution is vacuous. A purely historical survey would risk the pitfalls of anachronism and the incommensurability of the phenomena investigated. Abstract theoretical analysis of the idea of toleration that ignores the way the idea has operated in political rhetoric runs the risk of becoming irrelevant, since toleration is not a theoretical concept in the strict scientific sense. So although my approach to the question will be basically philosophical, I shall start with a few comments on the way the historical evolution of the idea of toleration transformed it in ways that are compatible, or even supportive of the normative analysis proposed in the rest of the paper. But I admit that my argument is only partly corroborated by the ordinary language of toleration, and that it is just one conceptualization of a highly heterogeneous idea that cannot by its nature be given a historically adequate account that will also be theoretically coherent. From the point of view of legal theory or political science, my “distilled” concept of toleration will certainly appear artificial and abstract. But I believe that a normative theory of toleration must start with concepts whose contours are theoretically well-defined even at the expense of doing justice to all our intuitions.
The argument of this paper is threefold: toleration is a moral rather than a political concept; toleration is not a virtue in the narrow sense but rather an attitude or a mode of judgment; and toleration is not obligatory but supererogatory. These three claims are interrelated and interdependent.
Unlike the concepts of the good or the just, toleration has a relatively short history and one that is mostly confined to one civilization. Being a “thick” concept, it is much more dependent on particular normative and cultural circumstances than its universal moral cognates. Although the political arrangements within the Roman empire and the New Testament parable of the wheat and the tares (Matthew 13) are often cited as origins of political and religious toleration, the concept itself appears only in the early modern period, and even then, in the beginning, not under the title “toleration.” The two contexts in which the modern idea of toleration gradually emerged were religion and royal grace. In light of the question raised in this paper, it should be emphasized that neither of these is “political” in the strict sense of the concept.
Take the religious context first. For Erasmus of Rotterdam, a typical example of the early thinkers on toleration, the highest goal is pax or concordia, that is to say the preservation of the harmonious unity of the Church, even at the cost of relinquishing some traditional Christian practices and declaring them “things indifferent” (adiaphora). The “tolerant” acceptance of unorthodox beliefs and practices is not based on the recognition of differences but on the distinction between what is religiously essential and what is merely doctrinal, between the inward effort to save Christian unity and the outward indifference to other religions.2 Toleration consists of both “sufferance” and “comprehension,” that is to say, the patience with nonconformist religious views is ultimately justified by typically inclusive reasons concerning the integrity of the religious community.3 Erasmus’ ideal of accommodation is religious and its justification pragmatic.
The second source of the idea of toleration is grace. From medieval times, the king or the ruler enjoyed the privilege of showing leniency towards communities or individuals under his jurisdiction. When shown to individuals, this “tolerant” attitude is closely associated with mercy, but with regard to groups, primarily religious communities, its effect is similar to our notion of toleration. The existence and some practices of Jews were “put up with” by Christian or Muslim rulers in their respective jurisdictions as a matter of sheer benevolence or pragmatic accommodation.4
A common feature of these two origins of the modern idea of toleration, which is of particular theoretical value for the conception advocated in this paper, is caritas. Charity or grace is the fundamental motive behind religious toleration as it is conceived by humanist Christians like Erasmus as well as by the merciful ruler. The endurance of differences or deviations from orthodoxy is not grounded in respect for the other, let alone for his rights, but in either love or a sense of power. Toleration, whether shown to Christian sects or to non-Christian religious minorities in a Christian polity, is primarily understood in terms of indulgence. Normatively speaking, this indulgence is supererogatory, modeled on the religious ideal of imitatio Christi, that is the adoption of Jesus’ charitable attitude. Like other supererogatory acts, this idea of toleration is not based on principle, but rather on benevolence; not on justice, but on a higher moral standard.
Both religious toleration and grace-based tolerance of minorities are decidedly of much political significance, but they are not political in their ultimate justification. However, in the course of the sixteenth century there was a growing awareness, for example among the so-called politiques in France, that tolerant practices should be adopted for purely political purposes, primarily co-existence and the maintenance of the unity of the state (rather than that of the Church). In the course of the seventeenth century, this typically political understanding of toleration gradually gained a theoretical guise as well as a linguistic title. Toleration became a principle grounded in a specific view of the state and its partial separation from religion and in the emerging concept of individual citizens having inalienable rights as individuals (against each other and against the state). In John Locke’s Letter, toleration is no longer conceived as either a purely religious ideal for the preservation of the unity of the Christian community or a personal favor granted by the sovereign. It has now become a duty of the state towards its citizens, a state whose function is strictly separated from the function of the church.5 Toleration became political in the strict sense by being transformed into a universal principle, applied to (almost) all citizens of a polity and exercised not as a matter of personal favor but as a duty, not as a personal discretion of the power of the ruler but as a constitutional principle of the law.
Religious tolerance and royal grace do not amount to political principles in the sense that the authority of the church or the king to decide matters of religious practices and beliefs was not challenged. But from Locke to Mill the authority of the state is systematically restricted to public matters and subordinated to universally applied laws. In that respect, toleration becomes political in essence, losing its supererogatory and paternalistic dimension. The public-private divide, which has been the major ground for liberal toleration from Locke to Rawls, is not just a religious, pragmatic, or epistemological distinction, but a principled definition of the realm and scope of the political. Thus, for Pierre Bayle, a tolerant political regime is only the second best option, to be gradually replaced by a completely neutral state that is totally indifferent to religious differences in society.
With the establishment of modern liberal democracy, Bayle’s vision became a reality. The successful career of the idea of toleration paradoxically led to its own decline, or at least made it superfluous in its traditional political form. In the second half of the twentieth century religious, ethnic, and sexual minorities have become more and more impatient with the status of being tolerated. In a multicultural society, the demand for recognition supersedes that of toleration. The state is expected to be neutral rather than restrained in its treatment of conflicts of value or religion. Pluralistic conceptions of value call for acceptance rather than toleration, which is often considered patronizing and condescending. As Bernard Williams pointed out, toleration may prove to have been an “interim value,” a political necessity along the path from a persecuting to a fully pluralistic society.6 Indeed, toleration had a crucial role in restraining the forces of persecution and intolerance and the gradual creation of a culture of either indifference or respect with regard to unorthodox beliefs and practices. But then, equality before the law and respect for the rights of individuals and minority groups tend to make toleration politically redundant. This does not mean that toleration has lost its meaning in contemporary liberal society. But, as I shall argue in the rest of this chapter, it means that the core of the concept should now be captured in more moral and personal terms, that is to say as applying to the realm of interpersonal and intercommunal relations rather than to the state, the law, or the constitutional structure of society.
This is a very rough skeletal survey of the genealogy of the idea of toleration; it is by no means intended to be understood as a history of the concept. It aims only to uncover a certain dialectical nature implied in the historical unfolding of the idea. A concept, which in its inception was typically religious, gradually transformed into a political one. When it lost its political role, it became (again) a personal or intercommunal value. To put it alternatively, toleration, originally conceived as a “negative” necessity, became in a second stage one of the “positive” values of the liberal state. Finally, it might turn out to be redundant in a truly pluralistic society. What started as an idea of grace or charity developed into a principle of political duty, only to become again a matter of charitable attitude that is supererogatory. In seventeenth-century England, toleration was a way to deal with intrareligious strife; in Mill’s nineteenth-century England, toleration lost its religious acuteness and was relegated to secular differences; in present-day England, toleration has regained a religious role but now applies primarily to the interreligious relationship between the majority and the religious minorities in society. This dialectic evolution of toleration does not bring us back full circle to the early modern period, but it does unravel certain tensions inherent in the very concept of toleration. We shall turn now to an analysis of the concept itself, which will be normative rather than historical, although informed by the genealogical account.
The idea of toleration evolved side by side with modern notions of rights, respect for individuals, separation of state and church, state neutrality, value pluralism, and skepticism. It was also instrumental in their entrenchment in the political culture of constitutional democracy. But once these ideas have become firmly established, the role and scope of toleration itself became hard to define. Thus, the analytical literature on toleration consists of a long list of what distinguishes toleration from: compromise, peace or co-existence, indifference, skepticism, recognition, acceptance, indulgence, open-mindedness, patience, endurance, condonation, charity, respect, pluralism, and more. Consequently, it is by no means easy to articulate what is left as a distinctive feature of toleration.7
This difficulty has led philosophers, like Williams, Walzer, Rawls, and Gray, to argue that toleration must be understood as a political practice rather than a moral virtue. I would like to argue for the opposite position, namely that the only way to mark the distinctive character of toleration is by regarding it as non-political. Bernard Williams contends that toleration cannot be a moral virtue since its motives are obscure and varied; it is rather a practice motivated by skepticism or the aspiration to peace than by a systematically moral attitude such as respect for autonomy. This contingent nature of toleration is exactly what makes it for Williams a transitory value, important in our time, but not necessarily beyond it.8 Michael Walzer states that his interest in toleration lies in its political dimension since any other view would not be able to do justice to the rich history of the concept.9 Toleration characterizes “regimes” and institutionalized social arrangements of coexistence. John Rawls also insists on the specifically political nature of toleration, which belongs strictly to the sphere of “public reason” rather than to a moral (comprehensive) doctrine. Toleration describes the way in which different but “reasonable” moral conceptions are mutually accepted within the framework of a just political society.10 John Gray takes a further step by claiming that toleration is not a principled political arrangement but rather a modus vivendi between people and groups who are not necessarily tolerant themselves, that is to say, a concept which applies to coexistence in non-liberal societies that lack an “overlapping consensus.”11
As I see it, the main problem with the political account of toleration is that for both analytical and normative reasons we do not want nowadays to ground liberal democracy on the idea of toleration. The main business of the liberal state is to respect and protect the rights of both individuals and groups, to establish justice and equality between its citizens, to secure the rule of law. The state is an embodiment of an impersonal constitutional structure which derives its validity from universalizable principles. In that respect it is neutral, at least with regards to its citizens, even if not with regards to values or moral doctrines. Unlike a medieval sovereign, the state is an impersonal institution which cannot be described as “suffering” in having to reconcile itself with beliefs and practices to which “it” does not subscribe. Hence, it cannot be said to overcome or endure its wish to undermine or interfere with them. In other words, the state cannot be engaged in toleration. The law either permits or prohibits certain practices and activities. The prohibited act cannot be tolerated by the law and the permitted practice cannot be said to be endured as a matter of charity or restraint.
Thus, for example, the issue of Muslim female students wearing headscarves in French state schools is not really a matter of toleration but a question of the correct interpretation of constitutional principles and of the idea of the separation of state and religion.12 As we shall see, only the way in which this religious practice is viewed by non-Muslim French citizens (rather than the state) may involve toleration. Or, to take another example, Will Kymlicka criticizes Rawls for his model of toleration as applied to individuals’ freedom of conscience, arguing that such a freedom has become a “fundamental human right.” He suggests an alternative analysis of tolerance, which applies to minority groups or communities.13 But once we go beyond his example of the Ottoman Millet system (which, being patronizing and pragmatic, could be described as “tolerant” in the traditional sense) and discuss present-day dilemmas of the treatment of minorities, then Kymlicka’s own critique points to the irrelevance of toleration. The legal status of minorities and their authority over their individual members is a matter of rights rather than of toleration by the state, of justice towards collective entities which struggle to maintain their identity. A final illustration of my point may be found in the value of freedom of expression. An individual might be appreciated for her toleration of repugnant or offensive speech by another individual. But the state must respect freedom of expression as a fundamental right. This right may be justified in terms of skepticism, personal autonomy, communicative reason, etc., but not as a matter of indulgence or endurance. If a particular expression goes beyond the permissible limits, then the state must interfere with it rather than tolerate it.
State neutrality and the protection of rights does not, therefore, leave room for state tolerance. But can a perfectionist view of liberalism allow for a strictly political analysis of toleration? According to Joseph Raz, the state is not neutral between values; it promotes only those ways of life that advance personal autonomy. But then, Raz himself believes that the state should not tolerate those practices that undermine autonomy. The object of state toleration is thus restricted to the (competitive) plurality of “good” options, those which although incompatible with each other cultivate personal autonomy.14 However, within that restricted domain of worthwhile alternatives, the state must remain neutral, at least in the sense that it should not prohibit any of these alternatives. It may promote this or that practice (for all kinds of reasons which have to do with democratic choice), but this does not mean that it can be described as “tolerant” towards those ways of life that are not at the top of its priorities. The analysis of toleration suggested here does not necessarily rely on a neutralist conception of the state. It is indeed true15 that modern liberal states are not neutral in the traditional “night-watch” sense and that they pursue substantive social goals and values. But the active promotion of communal identity, for instance, or the commitment to policies of affirmative action cannot, in my view, be considered as “tolerant” to the beneficiaries of these aims. They should rather be conceived as political duties of the state, or maybe even as the rights of those beneficiaries. Thus, within the framework of pluralism, toleration is an attitude of individuals (or groups) towards each other, exercised in their attempt to achieve their competing goals, rather than a norm of state action or a constitutional principle.
For the same reasons, toleration is not an attitude that can be shown by any state organ or institution. The court operates on the basis of the law and has no values of its own which can be overcome or restrained. On the one hand, it is the duty of judges to ignore their personal moral views rather than to manifest toleration of other, incompatible views. On the other hand, the court should not tolerate violations of the law, even if the judge personally feels she could tolerate the offensive act. The same applies to political authorities, officials, and institutions. Even the police acting leniently against law breakers should be better described as restrained rather than tolerant, since, as we shall see, the reasons and motives for its indulgent enforcement of the law are different from those typical of toleration. The courts or the police do not do us, individual citizens, any favor by letting us freely criticize the government or express controversial views that sound obnoxious to others. Public officials should definitely exercise discretion in carrying out their public duty; they may even be expected to show equity, i.e., go beyond the strict letter of the law; but this does not mean that by that they display a tolerance to the citizens.
For John Rawls, toleration is a constitutive virtue of political liberalism. Toleration characterizes the way we view comprehensive moral doctrines that are different from ours but are nevertheless recognized as “reasonable.” This crucial property of reasonableness is for Rawls derived from the idea of public reason which allows for an irreducible plurality of moral and religious values. Public reason in Rawls’ eyes operates on the political level of justice.16 Thus, toleration, as I understand it in Rawls, should better be seen as a bridge between the moral and the political. It belongs to neither: from the moral perspective, a rival moral view or practice cannot and should not be tolerated; from the political perspective, it ought to be fully accepted as reasonable and legitimate, rather than just tolerated. Toleration is the willingness to suspend the comprehensive moral point of view in favor of the narrowly political. But the reason for this suspension is of a practical nature, namely the achievement of social stability and peaceful coexistence in a deeply divided society.
So, although for Rawls toleration is constitutive of political liberalism, the ultimate reason to adopt toleration as a value is pragmatic. Kant, from whom the idea of public reason is derived, offers a more principled basis for toleration. The public use of reason is the condition for the operation of reason, its progress and perfection. In his famous essay “What Is Enlightenment?” the term “toleration” is mentioned only once and in a negative tone, describing it as “presumptuous” or patronizing. According to Kant, it is the duty of the prince (rather than an act of tolerance) to allow his subjects to freely exercise their own reason in matters of conscience.17 Thus the term “toleration” carries for Kant a pejorative meaning, associated with its traditional identification as grace. But, in her seminal article on Kant and toleration Onora O’Neill discusses the way Kant uses the concept of toleration, rather than the word. Toleration is justified not in terms of respect for the autonomy of the individual but as a constitutive condition of the free use of reason. It is a value that applies in the public domain rather than in the private (as has been the case in the long tradition of liberalism).18
Is this a “political” concept of toleration? It is, in the sense that it amounts to the duty of the prince to allow the free communication of ideas among rational persons in society so as to promote the process of enlightenment. Toleration here means the abstention of the political authority from censorship and intervention in the critical dialogue concerning religious issues and other matters of conscience. However, toleration is not a distinctly political virtue for Kant in the sense that the political is exactly the realm of the private use of reason, i.e., what he refers to as the rationality of the exercise of authority. In that sphere, Kant insists that citizens owe absolute obedience. Their freedom of thought and communicative action does not extend to practice and behavior. Toleration, therefore, applies strangely enough only to the “republic of letters,” only to communication within the “community of scholars.” From our contemporary point of view this is a very limited concept of toleration. Furthermore, the political abstention from censorship amounts at most to a negative concept of toleration, and Kant is therefore justified in treating it as the ruler’s duty. The positive value of toleration as the intrinsic condition of reason as such (as it is manifested in the community of scholars) is for its part typically non-political. It relates to the virtues of critical dialogue rather than to the way state authorities control our lives. In that deep sense of the condition of public reason, toleration is not a political virtue but a universal imperative. It seems that Kant was right in denigrating toleration in the literal sense of patronizing and presumptuous charity shown by the prince to his citizens. Toleration in this negative sense will become more and more superfluous the closer the private use of reason approaches to its public and universal use.
This does not mean that there is no political dimension in toleration, either in Kant or in general. An enlightened prince as well as a modern liberal state can and should promote the value of interpersonal toleration in society. The government has the power to inculcate standards of toleration by education, the support of institutions in which reason is freely exercised, and even to use its authority and capacity to enforce practices that advance communication and narrow the gap between the public and the private use of reason. But this political concept of toleration is, in contradistinction to O’Neill’s argument, instrumental precisely in the sense that once this gap is overcome, namely in the Kingdom of Ends, no room will be left to toleration, indeed not even to politics as the private use of reason.
It must therefore be emphasized that although I have tried to argue that toleration is not in its essence a political concept, I do not mean to deny that it has an important role in politics. Although the state cannot be said to be tolerant, either towards its citizens or towards other states, the interrelations between communities, religious or other, within society can be characterized in terms of tolerance. In that respect, the tolerance of individuals towards each other may often have political significance when the object of toleration is a political issue such as abortion. Or consider the demand of orthodox residents in Jerusalem to block traffic on the Sabbath in their neighborhood. Even if the court or the municipality prohibited such road blocks on the grounds of the freedom of movement on main traffic arteries of the city, individual secular citizens could be expected to show tolerance for the orthodox residents by voluntarily avoiding these roads on the Sabbath. Furthermore, we shall see in the next section that toleration tends to raise the level of solidarity and hence may lead to the strengthening of social cohesion and communal bonding. If justice promotes the values of liberty and equality, toleration upholds fraternity.
Another political aspect of toleration is associated with the implications of a tolerant attitude to third parties. Unlike forgiveness, for example, which has no effect on people other than the forgiver and the forgiven, toleration often has social costs.19 By tolerating an undesirable practice, I might weaken the ability of others who are or will be offended by it to fight against it. I would then be refraining from intervening in behavior from which others may suffer and who have not expressed any wish that I should so refrain. In that respect, my choice of toleration should be politically sensitive. It may call for a joint decision on the part of many people who stand to lose from the tolerated attitude. In the same way as I cannot forgive someone for an offense done to another person, I should not tolerate a behavior that is harmful primarily to other people.
States or nations, then, can be tolerant only in the derivative sense, namely in consisting of tolerant individuals (or communities of individuals). States can only indirectly promote moral norms that encourage tolerant attitudes in interpersonal relations. Cultures may be described as generous, forgiving or tolerant, but states or institutions as such cannot.20 The state cannot give generously although it can establish tax deductions for voluntary donations as a way to cultivate personal generosity in society. Although the implementation of rights is different from toleration, the two are closely related. Historically, toleration has led to the creation of a system of human rights (both individual and communal). But respect for rights may also foster a tolerant attitude since both require a capacity to separate between the act and the agent, as we shall shortly see.
This section was concerned with the negative characterization of toleration, attempting to show that it is not political in its essence. It is now time to proceed to a more positive account. Toleration in many cases amounts to refraining from insisting on our rights and to acting indulgently towards others who are wrong. In that sense, it goes beyond the political into the moral.21
A tempting way to approach toleration as a uniquely moral phenomenon is by describing it as a virtue. There is a sense in which it is difficult to deny that toleration is a virtue. Rawls says that justice is the primary virtue of social institutions. Similarly, one may say that toleration is the virtue of liberal society. This is the sense in which toleration is a good, a desirable trait or property, typical of, or even essential to a liberal constitutional system. This does not mean, however, that toleration is a virtue in the more strict, dispositional sense, traditionally associated with Aristotle. According to my analysis it is not. Although it is a personal attitude (rather than an institutional or political arrangement), it is not a naturally based trait of character. It does not have as its basis an inborn disposition. It is not acquired by habituation and conditioning. It benefits other individuals and society at large rather than the agent. It is hard to see toleration in terms of self-realization or the actualization of a human potential. Toleration never comes naturally to us, since it involves the subject’s reconciliation with a moral (or other) wrong or failing. Its absence from traditional lists of “the virtues” is not accidental since it does not belong to a general theory of human nature or to moral psychology.
Aristotle defines virtue as the mean between two naturally given extremes. Toleration cannot be subjected to such an analysis. Historically it falls, indeed, between persecution or intolerance and indifference or full acceptance. But this historical process does not refer to natural human dispositions but rather to religious and ethical norms of changing political cultures. Furthermore, Aristotle views virtue as the manner in which an action is performed: the courageous act is that piece of behavior as it is performed by the courageous individual, who has acquired the right disposition in the face of risk and danger. In toleration, it is the motive or the intention of the particular act that defines its value and the tolerant disposition is at most derivative of such particular acts.22 Once I choose to restrain myself from interfering in your wrong conduct it does not matter how I do so. For instance, the ease and smoothness with which the act is performed, which Aristotle believes are essential indicators of a virtuous act, are of no relevance to toleration.
The denial of the status of virtue to the concept of toleration should be qualified in two ways. First, the modern usage of the term “virtue” is wider than the Aristotelian. I mentioned justice as the virtue of social institutions. We may add fairness as the virtue of citizens in a just society, or truthfulness in the world of scientific (or other) communication. Toleration may be viewed in that sense as the virtue of citizens and communities in a multi-cultural or heterogeneous society. Secondly, although toleration is not a virtue in the strict Aristotelian sense of a character trait or a natural disposition, it is closely related to certain psychological dispositions that may enhance or impede it. Patience, indulgence, and temperance are natural propensities that make it much easier for people to show toleration. But they do not constitute it and may often promote indifference or compromise rather than toleration. Alternatively, a religious fundamentalist may be of a very kind and patient character, but for ideological reasons choose intolerance in all matters pertaining to competing religious practices. Thus, toleration is more than restraint or self-control, although these personal capacities are highly instrumental in its exercise.
After having proposed that toleration is neither political, nor a virtue in the strict sense of the terms, I turn now to a more positive analysis of its nature. I suggest that toleration be understood as a supererogatory attitude. This view relies on the common distinction in theories of toleration between agent and action, or in Augustinian terms, between “sinner and sin.” Despite the close relation between acts and their agents and the way they reflect on each other, philosophers have correctly argued that judgments of acts and judgments of actors can, and sometimes should be separated. Respect for the autonomy of the other or the attitude of forgiveness are two examples (which are pertinent to toleration) of the judgment of individuals that is independent of the judgment of their action. And there are of course judgments of actions or beliefs that are independent of judgments of their subjects, typically in court decisions or in the evaluation of scientific theories. However, if we wish to argue that toleration is a matter of the separation of the impersonal judgment of the act or the belief from the personal judgment of the agent or the subject, we must explain the mechanism through which this separation is made and the moral justification for doing so.
I suggest that toleration requires a shift from the impersonal judgment of actions to the personally based judgment of the agent. This shift is, as I have argued elsewhere, of a “perceptual” nature.23 It involves a Gestalt switch from one legitimate perspective to the other. The two perspectives are basically valid, yet incompatible in the sense that they cannot be adopted simultaneously (like the famous rabbit/duck image). From the impersonal view point, an action or a belief may look patently wrong, but from the personal it may be treated as understandable, tolerable, or forgivable due to the motive for its performance or the way it was adopted by the subject. The two perspectives are mutually exclusive. Thus, when we engage in moral or legal assessment of a type of action, we intentionally ignore the personal circumstances of the agent and the way he was led to act in the way he did. On the other hand, when we tolerate a person’s behavior or beliefs, we make ourselves blind to the negative features of the behavior and the wrongness of the beliefs. As in perception, we can switch from one perspective to the other (with varying degrees of effort, having to do with moral training), and the alternative perspective always remains in principle available to us. But the adoption of the one necessarily means the temporary suppression of the other. Structurally, this perspectival shift is analogous to the suspension of disbelief, traditionally associated with the aesthetic experience of a theater spectator: we can either see the events on stage as the movement of actors who are making their living, or as the dramatic deeds of fictional heroes; but we cannot enjoy the magic of the play while reflecting on the actor’s personal life.
The perceptual analysis of toleration explains why toleration is not a virtue. The shift from the impersonal judgment that an act is wrong to the personal tolerant acceptance of the agent despite the act is not a matter of a general disposition or a character trait. It is an intentional choice freely made by an individual in a particular case. It is more of a decision than a predisposition. Although the capacity to make the tolerant switch is facilitated by certain dispositions like patience and restraint, its constitutive conditions are of a cognitive kind, namely the capacity to abstract action from agent, or a belief from the subject holding it.24 Consequently, contrary to common wisdom, toleration does not consist of a “non-judgmental” disposition or blindness to the failings and defects of others, but rather of the capacity to alternate from one mode of judgment to another. But this capacity is neither a mere behavioral practice, a habit, nor a psychological feature of the agent. Showing toleration is at its core a deliberate choice based on reasons. The Gestalt switch from judging the action to tolerating the agent is undertaken from a specific motive that must be transparent to the tolerant person.25 Unlike Aristotelian virtuous action, which is typically performed with ease, as a “second nature,” tolerating wrong actions and beliefs has a price and takes an effort. It should thus be emphasized that toleration is an active attitude, to be clearly distinguished from passive mind-sets like indifference, acquiescence, condonation, or resignation.26 The personal dimension of the tolerant attitude means then that both the tolerator and the object of toleration must be persons, which explains the previously discussed claim that the state cannot strictly speaking show toleration. Institutions cannot engage in the sort of perspectival shift of judgment that constitutes toleration. Similarly, despite common usage, practices cannot be the object of toleration, but only the individuals taking part in those practices. We may be confident in our belief that female circumcision is morally wrong, in the sense that we have no reason to accept it as such or to approve of it. But we can nevertheless tolerate the individuals or communities practicing it on the grounds that we can understand, or even respect, the way the practice evolved in their culture and the central role it plays in their overall faith and way of life.
However, this analytical description of the idea of toleration, even if it makes sense phenomenologically speaking, requires a normative complement. If the two perspectives, the act-oriented and the agent-oriented, are equally valid, why and when should one be substituted for the other? What kind of reasons could support the renouncement of condemnation of an objectionable action in favor of a tolerant restraint from interfering in it? The two sets of respective reasons are valid but of a different, even incommensurable kind. One set of reasons has to do with the autonomy of the individual, with respect for her authentic commitment to certain values, and with the personal integrity in which these values are pursued. The other set of reasons concerns the wrongness of the act, the cognitive error in the beliefs underlying it, or the harm caused by the action to others. Thus, it is not the case that for the pro-lifer the reasons for tolerating abortions are simply stronger or more weighty than those for persecuting women and doctors who perform them. From this point of view, they cannot be overriding, let alone conclusive. But they do have an appeal that may create a switch in perspectives towards a tolerant acceptance of the agent, rather than the acts.
To see how this can take place, consider the case of forgiveness, which in many interesting respects is analogical to toleration, and precisely in the way a perspectival change is justified.27 When we are harmed or offended we are fully justified in responding with a hostile attitude. Justice requires that offensive actions be punished and their offender suffer the cold shoulder shown to him by the offended party. But then the offended person may adopt the alternative approach, trying to understand the other, hoping to restore the broken friendship, wishing to open a new page. Showing forgiveness does not go against justice, it goes beyond it. Forgiveness is supererogatory, that is a morally valuable attitude, though it is not required as a matter of duty or justice. Forgiving is beyond the call of duty exactly in the substitution of the personalized evaluation of the circumstances of the offense for the impersonal assessment of the offensive act. The attitude of forgiveness is moving just because it is a voluntary, optional renunciation of justified hostility and vindictiveness.
Similarly, the second-order reason for ignoring the force of reasons for interfering with the wrong behavior of others does not create a duty, nor is it called for as a matter of justice. Toleration is a supererogatory option that is morally valuable because it lies beyond the call of duty. I cannot interfere with the way my neighbor decorates his home, since it lies within his protected rights (even if his taste is repugnant). But I may call the police if this neighbor holds a noisy party after midnight. When I nevertheless tolerate the neighbor’s behavior, I withhold my judgment, or at least do not act on it, although it is within my rights and interference would be justified. Or, for an example from the sphere of relations between groups or communities in society, consider again the non-observant Israelis who are entitled to drive along a major thoroughfares crossing orthodox neighborhoods during the Sabbath but restrain themselves from doing so. The basis for this considerate approach does not consist of appreciation of the religious norms as such (which the non-observant do not share), but of good will towards the potentially offended orthodox neighbors whose sincere faith might be offended by the act.
The reason for adopting a tolerant attitude is, therefore, typically moral, based on good will, on the good intention of putting the agent before the act.28 Strangely, we return back full circle to the origins of toleration as grace! But unlike grace, the motive of toleration is conceived here as impelled by a special concern for the tolerated person and personality rather than for the self-image or the sense of power of the tolerating party. Hence, this concept of toleration does not involve any haughty or humiliating attitude and is inclusive rather than exclusive. Although the analysis offered here does not regard toleration as a patronizing attitude, nor does it restrict toleration to royal privilege, it shares with grace the discretionary, supererogatory deontic status. And in that respect it supports the genealogical dialectic of toleration, which started as a personal attitude, went through a political phase, and seems to end up nowadays as a matter of inter-personal or intercommunal relations.
Even though toleration is not strictly speaking a virtue and is not essentially political, it is undeniable that it has deep political importance, as is primarily, though not exclusively, manifested in liberal societies. Unlike forgiveness, which is a personal attitude between individuals, toleration is also shown by and towards groups, or rather between individuals as members of groups or communities. Unlike forgiveness, toleration can be exercised in an anonymous way, that is, towards unidentified individuals who belong to a particular group. This lends toleration a specifically political value. Contrary to forgiveness, which aims at restoring a broken personal relationship like friendship or love, toleration creates social solidarity, a sense of unity among people belonging to a common world even if they do not know each other personally. Toleration strengthens social bonding and trust, since it demonstrates good will, respect, and understanding towards individuals beyond their behavior and opinions. Forgiveness is ad hoc in nature, i.e., shown on a one-time basis, to a particular individual. Toleration, in contrast, is shown either to an individual or to a group of individuals for a whole spectrum of actions of a certain type. Thus, avoiding driving through an orthodox neighborhood on the Sabbath on a one-time basis is not tolerance, nor is the selective or haphazard selection of the particular roads in which the “tolerator” avoids driving. Although toleration is optional, it creates a kind of promise to refrain from interference not only in a present objectionable action but also in behavior of the same kind in the future, either of the same agent or of others belonging to the same group. This gives toleration a political dimension that is absent from forgiveness.29
Furthermore, although I have taken pains to distinguish between toleration on the one hand and the respect for rights, peace, and co-existence on the other, it must be stressed that toleration is highly instrumental in promoting these specifically political values. For, after all, respect of another person’s rights requires exactly the same capacity to separate the actor from her action and respect her freedom to engage in action that is deemed objectionable. It is true that this separation is obligatory in the case of rights, while it is supererogatory in the case of toleration; but the two are nevertheless mutually reinforcing. Social solidarity advances political stability and enhances the conditions of the communal co-existence that is of crucial importance in multi-cultural and pluralist societies. Forgiveness, even if it does not render interpersonal duties and obligations superfluous, tends to reduce the level of appeal to these norms in regulating personal relations. Similarly, toleration cannot be expected to serve as a substitute for legal norms and a system of enforceable rights, but it does ease political tensions and decrease the level of litigation in society.
Toleration is particularly called for in heterogeneous societies. The social cohesion of a tribal society, for example, is based on the large extent to which values and beliefs are shared by individual members. But, in pluralistic societies, this cohesion can be achieved only by appealing to other sources. Pragmatic considerations may lead to unity based on compromise. A principled conception separating agent from action establishes toleration. Although it is true that we show tolerance to begin with only to people to whom we feel close in some way, the tolerant attitude reinforces the sense of fraternity. In the absence of a substantive shared system of values in pluralist societies, this feature of toleration adds an important value. It also explains why historically speaking, although a tolerant attitude to other individuals has always been a value, the specifically political ideal of toleration was articulated only in early modern Europe, with the rise of religious pluralism and inter-religious strife within previously homogeneous societies.
So again, even if, as I have suggested, a tolerant society is not a society whose laws or institutions are tolerant, it is a society whose individual members and groups adopt a certain measure of supererogatory restraint in not insisting on their full rights. Toleration is not a political matter in the sense that it does not belong to the realm of constitutional arrangements, the rule of law, or the institutional relations of power and authority. But it may have a great political value since, as the old Talmudic saying reminds us, “Jerusalem was only destroyed because judgments were given strictly upon biblical law and did not go beyond the requirements of the law.”30
1. Michael Walzer and Joseph Raz are typical contemporary proponents of these two respective approaches to the study of toleration. Michael Walzer, On Toleration (New Haven: Yale University Press, 1997), and Joseph Raz, “Autonomy, Toleration and the Harm Principle,” in Susan Mendus, ed., Justifying Toleration (Cambridge: Cambridge University Press, 1988), 155–75. As the reader will immediately realize, both commentators to this chapter take the typically broad view of toleration: Kathryn Abrams for empirical reasons relating to the way the concept is used in current discourse; Andrew Sabl for epistemic (and maybe normative) reasons associated with the legitimacy of the diversity of concepts of toleration.
2. Mario Turchetti, “Religious Concord and Political Tolerance in Sixteenth and Seventeenth-Century France,” Sixteenth Century Journal 22 (1991): 15–25.
3. Gary Remer, Humanism and the Rhetoric of Toleration (University Park: The State of Pennsylvania University Press, 1996), 43–54.
4. For the idea of tolerance as grace, see Yirmiyahu Yovel, “Tolerance as Grace and as Rightful Recognition,” Social Research 65 (1998): 897–919, particularly the opening section.
5. Admittedly, Locke appeals also to religious arguments about the un-Christian nature of persecution, but it seems that he makes these as an ad hominem challenge to the proponents of intolerance rather than as independent positive support for the principle of toleration. See particularly Jeremy Waldron, “Locke: Toleration and the Rationality of Persecution,” Susan Mendus, ed., Justifying Toleration (Cambridge: Cambridge University Press, 1988), 62–63.
6. Bernard Williams, “Toleration: An Impossible Virtue?” in David Heyd, ed., Toleration: An Elusive Virtue (Princeton: Princeton University Press, 1996), 26.
7. Kathryn Abrams, for example, associates toleration with the virtues of curiosity, open-mindedness, and humility. Even if these virtues may prove to be conducive in some contexts to the development of a tolerant attitude, they are by no means constitutive or essential to it. Actually, toleration is typically the attitude of a person who is strongly committed to and confident in the values she holds.
8. Bernard Williams, “Toleration, A Political or Moral Virtue?” Diogenes 44 (1996): 36.
9. Walzer, On Toleration, chap. 1, and particularly note 3.
10. John Rawls, Political Liberalism (New York, Columbia University Press, 1993), 59, 194–95.
11. John Gray, The Two Faces of Liberalism (New York: The New Press, 2000), chap. 1.
12. Similarly, unlike Andrew Sabl, I do not consider the restraint from demanding Jews in the U.S. Army to uncover their heads indoors as a case of toleration. It is an issue of the balancing military codes with religious practices which must have an either-or answer: does a Jewish soldier maintain the right to wear a yarmulke while in active military service?
13. Will Kymlicka, “Two Models of Pluralism and Tolerance,” in David Heyd, ed., Toleration: An Elusive Virtue, 81–105.
14. It should be noted that for Raz toleration is a matter of interpersonal relations rather than of the political regulation of the acts of the state towards its citizens. On that point I follow his non-political approach to toleration although my analysis of the concept is different. See, “Autonomy, Toleration and the Harm Principle,” 162–65.
15. As argued by Kathryn Abrams and Andrew Sabl in their comments to this article.
16. Political Liberalism, 62.
17. Immanuel Kant, “What Is Enlightenment?” in H. Reiss, ed., Political Writings (Cambridge: Cambridge University Press, 1996), 58–59.
18. Onora O’Neill, “The Public Use of Reason,” Onora O’Neill, ed., Constructions of Reason (Cambridge: Cambridge University Press, 1989), 28–50.
19. I owe this reflection to Thomas Pogge.
20. As in the case of toleration, we metaphorically refer to certain states as generous in having a developed system of social benefits. But high unemployment payments or long maternity leaves are expressions of a conception of just distribution and social priorities rather than of a giving disposition or largesse.
21. I will put aside other non-political forms of toleration, such as religious, epistemological, cultural, and aesthetic toleration, all of which, I believe, are derived from the moral core of the concept.
22. See Glen Newey, “Tolerance as Virtue,” John Horton and Susan Mendus, eds., Toleration, Identity and Difference (London: Macmillan, 1999), 54. Although I do not consider toleration as a virtue, I agree with Newey’s thesis that a tolerant act cannot be fully reduced to a description of the tolerant agent.
23. David Heyd, “Introduction,” Toleration: An Elusive Virtue, 10–17.
24. These cognitive capacities are lacking or only partly developed in children. Hence their tendency to ad hominen arguments on the one hand and intolerance on the other. For a more elaborate presentation of the educational problems in inculcating tolerance in young people, see my “Education to Toleration: Some Obstacles and Their Resolution,” in Catharine MacKinnon and Dario Castiglione, eds., The Culture of Toleration in Diverse Societies (Manchester: Manchester University Press, 2003), 196–207.
25. Iain Hampsher-Monk, “Toleration and the Moral Will,” in John Horton and Susan Mendus, eds., Toleration, Identity and Difference (London: Macmillan, 1999), 17–37. For a similar view see also Robert P. Churchill, “On the Difference between Moral and Non-moral Conceptions of Toleration: The Case for Toleration as an Individual Virtue,” in Mehdi Amin Razavi and David Ambuel, eds., Philosophy, Religion, and the Question of Intolerance (Albany: State University of New York Press, 1997), 189–211. Churchill characterizes toleration as “voluntary forbearance on the basis of reasons,” but stops short of categorizing it as supererogatory. A closer approach to toleration as supererogation (“the deliberate suspension of moral entitlement” can be found in Peter Johnson, “As Long as He Needs Me? Toleration and Moral Character,” in John Horton and Peter Nicholson, eds., Toleration: Philosophy and Practice (Averbury: Aldershot, 1992), 146–64.
26. I take issue with MacKinnon’s view that toleration can be manifest in people who simply “mind their own business,” since this is an attitude that is too close to indifference. Tolerating another person, according to the analysis advocated here, means an active effort to understand the action to which one objects in terms of the agent’s motives, views, and circumstances. This involves what MacKinnon calls “engagement,” although she clearly distinguishes it from toleration. Catharine MacKinnon, “Toleration and the Character of Pluralism,” in The Culture of Toleration in Diverse Societies, 58–59. In this active aspect of the tolerant attitude my analysis lies closer to what Kathryn Abrams calls “engaged toleration” (which she presents as an alternative to my view). For in order to respect and “accept” the agent, tolerator has to understand not only her values and beliefs as such but the way they have been formed and the manner in which they cohere with other values and beliefs of that particular individual.
27. See Hagit Benbaji and David Heyd, “The Charitable Perspective: Forgiveness and Toleration as Supererogatory,” Canadian Journal of Philosophy 31 (2001): 567–86.
28. For a good presentation of the view of toleration as a moral, rather than political, virtue, see Anna E. Galeotti, “Toleration as a Moral Virtue,” Res Publica (2001): 273–92. However, Galeotti does not agree with the perceptual model outlined here. She believes that the moral conception of toleration is too abstract to support toleration as a social practice. The approach of this article is to leave the regulation of social behavior in the context of race and gender to legal norms and a system of political rights and promote tolerant attitudes only through educational means.
29. Forgiveness and promise are held by Hannah Arendt to be two conditions of action: forgiveness overcomes the irreversibility of the past, whereas promises overcome the unpredictability of the future. Toleration, according to my description, could be understood as a promise of forgiveness: by tolerating your present behavior I announce that I will also refrain from interfering in it in the future. In this Arendtian sense toleration is definitely of a political nature. Hannah Arendt, The Human Condition (Chicago: The University of Chicago Press, 1958), secs. 33–34.
30. Babylonian Talmud, Tractate Baba Mezia, 30b.