157
7
Five Regimes of Toleration
Michael Walzer
Multinational Empires
The oldest arrangements are those of the great multinational empires – beginning, for
our purposes, with Persia, Ptolemaic Egypt, and Rome. Here the various groups are
constituted as autonomous or semiautonomous communities that are political or
legal as well as cultural or religious in character, and that rule themselves across a
considerable range of their activities. The groups have no choice but to coexist with
one another, for their interactions are governed by imperial bureaucrats in accordance
with an imperial code, like the Roman jus gentium, which is designed to
maintain some minimal fairness, as fairness is understood in the imperial center.
Ordinarily, however, the bureaucrats don’t interfere in the internal life of the
autonomous communities for the sake of fairness or anything else – so long as taxes
are paid and peace maintained. Hence they can be said to tolerate the different ways
of life, and the imperial regime can be called a regime of toleration, whether or not
the members of the different communities are tolerant of one another.
Under imperial rule, the members will, willy nilly, manifest tolerance in (most of)
their everyday interactions, and some of them, perhaps, will learn to accept difference
and come to stand somewhere on the continuum that I have described. But the
survival of the different communities doesn’t depend on this acceptance. It depends
only on official toleration, which is sustained, mostly, for the sake of peace – though
individual officials have been variously motivated, a few of them famously curious
about difference or even enthusiastic in its defense. These imperial bureaucrats are
often accused of following a policy of “divide and rule,” and sometimes indeed that
is their policy. But it has to be remembered that they are not the authors of the
divisions they exploit and that the people they rule may well want to be divided and
ruled, if only for the sake of peace.
Imperial rule is historically the most successful way of incorporating difference
and facilitating (requiring is more accurate) peaceful coexistence. But it isn’t, or at
least it never has been, a liberal or democratic way. Whatever the character of the
different “autonomies,” the incorporating regime is autocratic. I don’t want to
idealize this autocracy; it can be brutally repressive for the sake of maintaining its
conquests – as the histories of Babylonia and Israel, Rome and Carthage, Spain and
the Aztecs, and Russia and the Tatars amply demonstrate. But settled imperial rule is
often tolerant – tolerant precisely because it is everywhere autocratic (not bound by
the interests or prejudices of any of the conquered groups, equally distant from all of
them). Roman proconsuls in Egypt or British regents in India, for all their prejudices
and the endemic corruption of their regimes, probably ruled more evenhandedly than
any local prince or tyrant was likely to do – in fact, more evenhandedly than local
majorities today are likely to do.
Imperial autonomy tends to lock individuals into their communities and therefore
into a singular ethnic or religious identity. It tolerates groups and their authority
structures and customary practices, not (except in a few cosmopolitan centers and
capital cities) free-floating men and women. The incorporated communities are not
voluntary associations; they have not, historically, cultivated liberal values. Though
there is some movement of individuals across their boundaries (converts and
apostates, for example), the communities are mostly closed, enforcing one or another
version of religious orthodoxy and sustaining a traditional way of life. So long as they
are protected against the more severe forms of persecution and allowed to manage
their own affairs, communities of this sort have extraordinary staying power. But they
can be very severe toward deviant individuals, who are conceived as threats to their
cohesiveness and sometimes to their very survival.
So lonely dissidents and heretics, cultural vagabonds, intermarried couples, and
their children will flee to the imperial capital, which is likely to become as a result a
fairly tolerant and liberal place (think of Rome, Baghdad, and imperial Vienna, or,
better, Budapest) – and the only place where social space is measured to an individual
fit. Everyone else, including all the free spirits and potential dissidents who are
unable to move because of economic constraint or familial responsibility, will live in
homogeneous neighborhoods or districts, subject to the discipline of their own communities.
They are tolerated collectively there, but they will not be welcome or even
safe as individuals across whatever line separates them from the others. They can mix
comfortably only in neutral space – the market, say, or the imperial courts and
prisons. Still, they live most of the time in peace, one group alongside the other,
respectful of cultural as well as geographic boundaries.
Ancient Alexandria provides a useful example of what we might think of as the
imperial version of multiculturalism. The city was roughly one-third Greek, one-third
Jewish, and one-third Egyptian, and during the years of Ptolemaic rule; the
coexistence of these three communities seems to have been remarkably peaceful.
Later on, Roman officials intermittently favored their Greek subjects, perhaps on
grounds of cultural affinity, or perhaps because of their superior political
organization (only the Greeks were formally citizens), and this relaxation of imperial
neutrality produced periods of bloody conflict in the city. Messianic movements
among Alexandria’s Jews, partly in response to Roman hostility, eventually brought
multicultural coexistence to a bitter end. But the centuries of peace suggest the better
possibilities of the imperial regime. It is interesting to note that though the
communities remained legally and socially distinct, there was significant commercial
and intellectual interaction among them – hence the Hellenistic version of Judaism
that was produced, under the influence of Greek philosophers, by Alexandrian
writers like Philo. The achievement is unimaginable except in this imperial setting.
158
The millet system of the Ottomans suggests another version of the imperial
regime of toleration, one that was more fully developed and longer lasting. In this
case, the self-governing communities were purely religious in character, and because
the Ottomans were themselves Muslim, they were by no means neutral among
religions. The established religion of the empire was Islam, but three other religious
communities – Greek Orthodox, Armenian Orthodox, and Jewish – were permitted
to form autonomous organizations. These three were equal among themselves,
without regard to their relative numerical strength. They were subject to the same
restrictions vis-à-vis Muslims – with regard to dress, proselytizing, and intermarriage,
for example – and were allowed the same legal control over their own members. The
minority millets (the word means religious community) were subdivided along ethnic,
linguistic, and regional lines, and some differences of religious practice were thereby
incorporated into the system. But members had no rights of conscience or of
association against their own community (and everyone had to be a member
somewhere). There was, however, further toleration at the margins: thus, Karaite
sectarians within Judaism were accorded fiscal independence, though not full millet
status, by the Ottomans in the sixteenth century. Basically, again, the empire was
accommodating toward groups but not toward individuals – unless the groups
themselves opted for liberalism (as a Protestant millet, established late in the
Ottoman period, apparently did.)
Today, all this is gone (the Soviet Union was the last of the empires): the
autonomous institutions, the carefully preserved boundaries, the ethnically marked
identity cards, the cosmopolitan capital cities, and the far-flung bureaucracies.
Autonomy did not mean much at the end (which is one reason, perhaps, for imperial
decline); its scope was greatly reduced by the effect of modern ideas about
sovereignty and by totalizing ideologies uncongenial to the accommodation of
difference. But ethnic and religious differences survived, and wherever they were
territorially based, local agencies, which were more or less representative, retained
some minimal functions and some symbolic authority. These they were able to
convert very quickly, once the empires fell, into a kind of state machine driven by
nationalist ideology and aimed at sovereign power – and opposed, often enough, by
established local minorities, the great beneficiaries of the imperial regime and its last
and most stalwart defenders. With sovereignty, of course, comes membership in
international society, which is the most tolerant of all societies but, until very
recently, not so easy to get into. I shall consider international society only briefly and
incidentally in this essay, but it is important to recognize that most territorially based
groups would prefer to be tolerated as distinct nation-states (or religious republics)
with governments, armies, and borders – coexisting with other nation-states in
mutual respect or, at least, under the rule of a common (even if rarely enforced) set
of laws.
International Society
International society is an anomaly here because it is obviously not a domestic
regime; some would say that it is not a regime at all but rather an anarchic and lawless
condition. If that were true, the condition would be one of absolute toleration:
anything goes, nothing is forbidden, for no one is authorized to forbid (or permit),
even if many of the participants are eager to do so. 1n fact, international society is
not anarchic; it is a very weak regime, but it is tolerant as a regime despite the
intolerance of some of the states that make it up. All the groups that achieve
statehood and all the practices that they permit (within limits that I will come to in a
moment) are tolerated by the society of states. Toleration is an essential feature of
sovereignty and an important reason for its desirability.
Sovereignty guarantees that no one on that side of the border can interfere with
what is done on this side. The people over there may be resigned, indifferent, stoical,
curious, or enthusiastic with regard to practices over here, and so may be disinclined
to interfere. Or perhaps they accept the reciprocal logic of sovereignty: we won’t
worry about your practices if you don’t worry about ours. Live and let live is a
relatively easy maxim when the living is done on opposite sides of a clearly marked
line. Or they may be actively hostile, eager to denounce their neighbor’s culture and
customs, but unprepared to pay the costs of interference. Given the nature of
international society, the costs are likely to be high: they involve raising an army,
crossing a border, killing and being killed.
Diplomats and statesmen commonly adopt the second of these attitudes. They
accept the logic of sovereignty, but they can’t simply look away from persons and
practices that they find intolerable. They must negotiate with tyrants and murderers
and, what is more pertinent to our subject, they must accommodate the interests of
countries whose dominant culture or religion condones, for example, cruelty, oppression,
misogyny, racism, slavery, or torture. When diplomats shake hands or break
bread with tyrants, they are, as it were, wearing gloves; the actions have no moral
significance. But the bargains they strike do have moral significance: they are acts of
toleration. For the sake of peace or because they believe that cultural or religious
reform must come from within, must be local work, they recognize the other country
as a sovereign member of international society. They acknowledge its political
independence and territorial integrity – which together constitute a much stronger
version of the communal autonomy maintained in multinational empires.
Diplomatic arrangements and routines give us a sense of what might be called the
formality of toleration. This formality has a place, though it is less visible, in domestic
life, where we often coexist with groups with which we don’t have and don’t want to
have close social relations. The coexistence is managed by civil servants who are also
domestic diplomats. Civil servants have more authority than diplomats, of course,
and so the coexistence that they manage is more constrained than that of sovereign
states in international society.
159
But sovereignty also has limits, which are fixed most clearly by the legal doctrine
of humanitarian intervention. Acts or practices that “shock the conscience of
humankind” are, in principle, not tolerated. Given the weak regime of international
society, all that this means in practice is that any member state is entitled to use force
to stop what is going on if what is going on is awful enough. The principles of
political independence and territorial integrity do not protect barbarism. But no one
is obligated to use force; the regime has no agents whose function it is to repress
intolerable practices. Even in the face of obvious and extensive brutality,
humanitarian intervention is entirely voluntary. The practices of the Khmer Rouge in
Cambodia, to take an easy example, were morally and legally intolerable, and because
the Vietnamese decided to invade the country and stop them, they were in fact not
tolerated. But this happy coincidence between what is intolerable and what is not
tolerated is uncommon. Humanitarian intolerance isn’t usually sufficient to override
the risks that intervention entails, and additional reasons for intervening – whether
geopolitical, economic, or ideological – are only sometimes available.
One can imagine a more articulated set of limits on the toleration that comes with
sovereignty: intolerable practices in sovereign states might be the occasion for
economic sanctions by some or all of the members of international society. The
enforcement of a partial embargo against South African apartheid is a useful if
unusual example. Collective condemnation, breaks in cultural exchange, and active
propaganda can also serve the purposes of humanitarian intolerance, though
sanctions of this sort are rarely effective. So we can say that international society is
tolerant as a matter of principle, and then more tolerant, beyond its own principles,
because of the weakness of its regime.
Consociations
Before I consider the nation-state as a possibly tolerant society, I want to turn briefly
to a morally closer but not politically more likely heir to the multinational empire –
the consociational or bi- or trinational state. Examples like Belgium, Switzerland,
Cyprus, Lebanon, and the stillborn Bosnia suggest both the range of possibility here
and the imminence of disaster. Consociationalism is a heroic program because it aims
to maintain imperial coexistence without the imperial bureaucrats and without the
distance that made those bureaucrats more or less impartial rulers. Now the different
groups are not tolerated by a single transcendent power; they have to tolerate one
another and work out among themselves the terms of their coexistence.
The idea is attractive: a simple, unmediated concurrence of two or three
communities (in practice, of their leaders and elites) that is freely negotiated between
or among the parties. They agree to a constitutional arrangement, design institutions
and divide offices, and strike a political bargain that protects their divergent interests.
But the consociation is not entirely a free construction. Commonly, the communities
have lived together (or, rather, alongside one another) for a very long time before
they begin their formal negotiations. Perhaps they were initially united by imperial
rule; perhaps they first came together in the struggle against that rule. But all these
connections are preceded by proximity: coexistence on the ground, if not in the same
villages, then along a frontier only roughly defined and easily crossed. These groups
have talked and traded, fought and made peace at the most local levels – but always
with an eye to the police or army of some foreign ruler. Now they must look only to
each other.
This isn’t impossible. Success is most likely when the consociation predates the
appearance of strong nationalist movements and the ideological mobilization of the
different communities. It is best negotiated by the elites of the old “autonomies,”
who are often genuinely respectful of one another, have a common interest in
stability and peace (and, obviously, in the ongoing authority of elites), and are willing
to share political power. But the arrangements the elites work out, which reflect the
size and economic strength of the associated communities, are dependent thereafter
on the stability of their social base. The consociation is predicated, say, on the
constitutionally limited dominance of one of the parties or on their rough equality.
Offices are divided, quotas established for the civil service, and public funds allocated
– all on the basis of this limited dominance or rough equality. Given these
understandings, each group lives in relative security, in accordance with its own
customs, perhaps even its own customary law, and can speak its own language not
only at home but also in its own public space. The old ways are undisturbed.
It is the fear of disturbance that breaks up consociations. Social or demographic
change, let’s say, shifts the base, alters the balance of size and strength, threatens the
established pattern of dominance or equality, undermines the old understandings.
Suddenly one of the parties looks dangerous to all the others. Mutual toleration
depends on trust, not so much in each other’s good will as in the institutional
arrangements that guard against the effects of ill will. Now the established
arrangements collapse, and the resulting insecurity makes toleration impossible. I
can’t live tolerantly alongside a dangerous other. What is the danger that I fear? That
the consociation will be turned into an ordinary nation-state where I will be a
member of the minority, looking to be tolerated by my former associates, who no
longer require my toleration.
Lebanon is the obvious example of this sad collapse of consociational
understandings; it has guided the description I have just given. But in Lebanon
something more than social change was involved. In principle, the new Lebanese
demography or the new economy should have led to a renegotiation of the old
arrangements, a simple redivision of offices and public funds. But the ideological
transformations that came with social change made this very difficult to achieve.
Nationalist and religious zeal and its inevitable concomitants, distrust and fear,
turned renegotiation into civil war (and brought the Syrians in as imperial
peacemakers). Against this background, consociation is clearly recognizable as a preideological
regime. Toleration is not out of the question once nationalism and
religion are in play, and consociation may still be its morally preferred form. In prac-
160
tice, however, the nation-state is now the more likely regime of toleration: one group,
dominant throughout the country, shaping public life and tolerating a national or
religious minority – rather than two or three groups, each secure in its own place,
tolerating one another.
Nation-States
Most of the states that make up international society are nation-states. To call them
that doesn’t mean that they have nationally (or ethnically or religiously)
homogeneous populations. Homogeneity is rare, if not nonexistent, in the world
today. It means only that a single dominant group organizes the common life in a
way that reflects its own history and culture and, if things go as intended, carries the
history forward and sustains the culture. It is these intentions that determine the
character of public education, the symbols and ceremonies of public life, the state
calendar and the holidays it enjoins. Among histories and cultures, the nation-state is
not neutral; its political apparatus is an engine for national reproduction. National
groups seek statehood precisely in order to control the means of reproduction. Their
members may hope for much more – they may harbor ambitions that range from
political expansion and domination to economic growth and domestic flourishing.
But what justifies their enterprise is the human passion for survival over time.
The state these members create can nonetheless, as liberal and democratic nationstates
commonly do, tolerate minorities. This toleration takes different forms, though
it rarely extends to the full autonomy of the old empires. Regional autonomy is
especially difficult to implement, for then members of the dominant nation living in
the region would be subjected to “alien” rule in their own country. Nor are
corporatist arrangements common; the nation-state is itself a kind of cultural
corporation and claims a monopoly on such arrangements within its borders.
Toleration in nation-states is commonly focused not on groups but on their
individual participants, who are generally conceived stereotypically, first as citizens,
then as members of this or that minority. As citizens, they have the same rights and
obligations as everyone else and are expected to engage positively with the political
culture of the majority; as members, they have the standard features of their “kind”
and are allowed to form voluntary associations, organizations for mutual aid, private
schools, cultural societies, publishing houses, and so on. They are not allowed to
organize autonomously and exercise legal jurisdiction over their fellows. Minority
religion, culture, and history are matters for what might be called the private
collective – about which the public collective, the nation-state, is always suspicious.
Any claim to act out minority culture in public is likely to produce anxiety among the
majority (hence the controversy in France over the wearing of Muslim headdress in
state schools). In principle, there is no coercion of individuals, but pressure to
assimilate to the dominant nation, at least with regard to public practices, has been
fairly common and, until recent times, fairly successful. When nineteenth-century
German Jews described themselves as “German in the street, Jewish at home,” they
were aspiring to a nation-state norm that made privacy a condition of toleration.
The politics of language is one key area where this norm is both enforced and
challenged. For many nations, language is the key to unity. They were formed in part
through a process of linguistic standardization, in the course of which regional
dialects were forced to give way to the dialect of the center – though one or two
sometimes managed to hold out, and thus became the focus of subnational or
protonational resistance. The legacy of this history is a great reluctance to tolerate
other languages in any role larger than familial communication or religious worship.
Hence the majority nation commonly insists that national minorities learn and use its
language in all their public transactions – when they vote, go to court, register a
contract, and so on.
Minorities, if they are strong enough, and especially if they are territorially based,
will seek the legitimation of their own languages in state schools, legal documents,
and public signage. Sometimes, one of the minority languages is in fact recognized as
a second official language; more often, it is sustained only in homes, churches, and
private schools (or is slowly and painfully lost). At the same time, the dominant
nation watches its own language being transformed by minority use. Academies of
linguists struggle to sustain a “pure” version, or what they take to be a pure version,
of the national language, but their fellow nationals are often surprisingly ready to
accept minority or foreign usages. This too, I suppose, is a test of toleration.
There is less room for difference in nation-states, even liberal nation-states, than
in multinational empires or consociations – far less, obviously, than in international
society. Because the tolerated members of the minority group are also citizens, with
rights and obligations, the practices of the group are more likely than in multinational
empires to be subject to majority scrutiny. Patterns of discrimination and domination
long accepted – or, at any rate, not resisted – within the group may not be acceptable
after members are recognized as citizens (I will consider some examples in Chapter
4). But there is a double effect here, with which any theory of toleration must reckon:
though the nation-state is less tolerant of groups, it may well force groups to be more
tolerant of individuals. This second effect is a consequence of the (partial and
incomplete) transformation of the groups into voluntary associations. As internal
controls weaken, minorities can hold their members only if their doctrines are
persuasive, their culture attractive, their organizations serviceable, and their sense of
membership liberal and latitudinarian. In fact, there is an alternative strategy: a rigidly
sectarian closure. But this offers hope only of saving a small remnant of true
believers. For larger numbers, more open and looser arrangements are necessary. All
such arrangements, however, pose a common danger: that the distinctiveness of the
group and of its way of life will slowly be surrendered.
Despite these difficulties, a variety of significant differences, especially religious
differences, have been successfully sustained in liberal and democratic nation-states.
161
Minorities often, in fact, do fairly well in enacting and reproducing a common culture
precisely because they are under pressure from the national majority. They organize
themselves, both socially and psychologically, for resistance, making their families,
neighborhoods, churches, and associations into a kind of homeland whose borders
they work hard to defend. Individuals, of course, drift away, pass themselves off as
members of the majority, slowly assimilate to majority lifestyles, or intermarry and
raise children who have no memory or knowledge of the minority culture. But for
most people, these self-transformations are too difficult, too painful, or too
humiliating; they cling to their own identities and to similarly identified men and
women.
National (more than religious) minorities are the groups most likely to find
themselves at risk. If these groups are territorially concentrated – like the Hungarians
in Romania, say – they will be suspected, perhaps rightly, of hoping for a state of
their own or for incorporation into a neighboring state where their ethnic relatives
hold sovereign power. The arbitrary processes of state formation regularly produce
minorities located in this way, groups that are subject to these suspicions and very
hard to tolerate. Perhaps the best thing to do is to pull in the borders and let them
go, or to grant them a full measure of autonomy. We tolerate the others by
contracting our state so that they can live in social space shaped to their own needs.
Alternative solutions are more likely, of course: linguistic – recognition and a very
limited degree of administrative devolution are fairly common, though these are
often combined with efforts to settle members of the majority in politically sensitive
border regions and with periodic campaigns of assimilation.
After World War I, an effort was made to guarantee the toleration of national
minorities in the new (and radically heterogeneous) “nation states” of Eastern
Europe. The guarantor was the League of Nations, and the guarantee was written
into a series of minority or nationality treaties. Appropriately, these treaties ascribed
rights to stereotypical individuals rather than to groups. Thus the Polish Minority
Treaty deals with “Polish nationals who belong to racial, religious, or linguistic
minorities.” Nothing follows from such a designation about group autonomy or
regional devolution or minority control of schools. Indeed, the guarantee of
individual rights was itself chimerical: most of the new states asserted their
sovereignty by ignoring (or annulling) the treaties, and the League was unable to
enforce them.
But this failed effort is well worth repeating, perhaps with a more explicit
recognition of what the stereotypical minority member has in common with his or
her fellows. The United Nation’s Covenant on Civil and Political Rights (1966) takes
this further step: minority individuals “shall not be denied the right, in community
with other members of their group, to enjoy their own culture, to possess and practice
their own religion, or to use their own language.” Note that this wording still falls
within the nation-state norm: no recognition is accorded to the group as a corporate
body; individuals act “in community with”; only the national majority acts as a
community.
In time of war, the loyalty of national minorities to the nation-state, whether or
not the minorities are territorially concentrated or internationally recognized, will
readily be called into doubt – even against all available evidence, as in the case of
anti-Nazi German refugees in France during the first months of World War II. Once
again, toleration fails when the others look dangerous, or when nationalist demagogues
can make them look dangerous. The fate of Japanese-Americans a few years
later makes the same point – their fellow Americans imitated, as it were, conventional
nation-statehood. In fact, the Japanese were not, and are not, a national minority in
the United States, at least not in the usual sense: where is the majority nation?
American majorities are temporary in character and are differently constituted for
different purposes and occasions (minorities are often temporary too, though race
and slavery together make an exception; I shall consider the exception later on). It is
a crucial feature of the nation-state, by contrast, that its majority is permanent.
Toleration in nation-states has only one source, and it moves or doesn’t move in only
one direction. The case of the United States suggests a very different set of
arrangements.
Immigrant Societies
The fifth model of coexistence and possible toleration is the immigrant society. Now
the members of the different groups have left their territorial base, their homeland,
behind them; they have come individually or in families, one by one, to a new land
and then dispersed across it. Though they arrive in waves, responding to similar
political and economic pressures, they don’t arrive in organized groups. They are not
colonists, consciously planning to transplant their native culture to a new place. They
cluster for comfort only in relatively small numbers, always intermixed with other,
similar groups in cities, states, and regions. Hence no sort of territorial autonomy is
possible. (Though Canada is an immigrant society, Quebec is an obvious exception
here; its original settlers did come as colonists, not as immigrants, and were then
conquered by the British. Another exception must be made for the Aboriginal
peoples, who were also conquered. I will focus here primarily on the immigrants. On
the Quebecois and Aboriginals, see the section “Canada” in Chapter 3; on American
blacks, imported as slaves, see the section “Class” in Chapter 4.)
If ethnic and religious groups are to sustain themselves, they must do so now as
purely voluntary associations. This means that they are more at risk from the
indifference of their own members than from the intolerance of the others. The
state, once it is pried loose from the grip of the first immigrants, who imagined in
every case that they were forming a nation-state of their own, is committed to none of
the groups that make it up. It sustains the language of the first immigration and,
subject to qualification, its political culture too, but so far as contemporary
162
advantages go, the state is, in the current phrase (and in principle), neutral among the
groups, tolerant of all of them, and autonomous in its purposes.
The state claims exclusive jurisdictional rights, regarding all its citizens as
individuals rather than as members of groups. Hence the objects of toleration, strictly
speaking, are individual choices and performances: acts of adhesion, participation in
rituals of membership and worship, enactments of cultural difference, and so on.
Individual men and women are encouraged to tolerate one another as individuals, to
understand difference in each case as a personalized (rather than a stereotypical)
version of group culture – which also means that the members of each group, if they
are to display the virtue of tolerance, must accept each other’s different versions.
Soon there are many versions of each group’s culture, and many different degrees of
commitment to each. So toleration takes on a radically decentralized form: everyone
has to tolerate everyone else.
No group in an immigrant society is allowed to organize itself coercively, to seize
control of public space, or to monopolize public resources. Every form of
corporatism is ruled out. In principle, the public schools teach the history and
“civics” of the state, which is conceived to have no national but only a political
identity. This principle is, of course, only slowly and imperfectly enforced. Since
public schools were founded in the United States, for example, the schools have
mostly taught what English-Americans conceived as their own history and culture –
which extend back to Greece and Rome and include classical languages and literature.
There was and still is considerable justification for this standard curriculum,
even after the immigrations of the mid-nineteenth century (when Germans and Irish
arrived) and the turn of the century (when Southern and Eastern European peoples
came), for American political institutions are best understood against this
background. In more recent times (and in the course of a third great immigration,
which this time is largely non-European), efforts have been made to incorporate the
history and culture of all the different groups, to ensure a kind of equal coverage and
so to create “multicultural” schools. In fact, the West still dominates the curriculum
almost everywhere.
Similarly, the state is supposed to be perfectly indifferent to group culture or
equally supportive of all the groups – encouraging, for example, a kind of general
religiosity, as in those train and bus advertisements of the 1950s that urged
Americans to “attend the church of your choice.” As this maxim suggests, neutrality
is always a matter of degree. Some groups are in fact favored over others – in this
case, groups with “churches” more or less like those of the first Protestant
immigrants; but the others are still tolerated. Nor is church attendance or any other
culturally specific practice turned into a condition of citizenship. It is relatively easy,
then, and not at all humiliating, to escape one’s own group and take on the reigning
political identity (in this case, “American”).
But many people in an immigrant society prefer a hyphenated or dual identity, one
differentiated along cultural or political lines. The hyphen joining Italian-American,
for example, symbolizes the acceptance of “Italianness” by other Americans, the
recognition that “American” is a political identity without strong or specific cultural
claims. The consequence, of course, is that “Italian” is a cultural identity without
political claims. That is the only form in which Italianness is tolerated, and then
Italian-Americans must sustain their own culture, if they can or as long as they can,
privately, through the voluntary efforts and contributions of committed men and
women. And this is the case, in principle, with every cultural and religious group, not
only with minorities (but, again, there is no permanent majority).
Whether groups can sustain themselves under these conditions – without
autonomy, without access to state power or official recognition, and without a
territorial base or the fixed opposition of a permanent majority – is a question still to
be answered. Religious communities, of both sectarian and “churchly” sorts, have
not done badly in the United States until now. But one reason for their relative
success might be the considerable intolerance that many of them have in fact
encountered; intolerance often has, as I have already suggested, group-sustaining
effects. Ethnic groups have done less well, though observers eager to write them off
are almost certainly premature. These groups survive in what we might think of as a
doubly hyphenated version: the culture of the group is, for example, AmericanItalian,
which means that it takes on a heavily Americanized form and is transfigured
into something quite distinct from Italian culture in the home country; and its politics
is Italian-American, an ethnic adaptation of local political practices and styles.
Consider the extent to which John Kennedy remained an Irish “pol,” Walter
Mondale is still a Norwegian social democrat, Mario Cuomo is still an Italian
Christian Democratic intellectual-in-politics, and Jesse Jackson is still a black Baptist
preacher – each of them in many ways similar to, but in these ways different from,
the standard Anglo-American type.
Whether these differences will survive into the next generation or the one after
that is uncertain. Straightforward survival is perhaps unlikely. But that is not to say
that the successors to these four exemplary figures, and to many others like them,
will all be exactly alike. The forms of difference characteristic of immigrant societies
are still emerging. We don’t know how “different” difference will actually be. The
toleration of individual choices and personalized versions of culture and religion
constitutes the maximal (or the most intensive) regime of toleration. But it is radically
unclear whether the long-term effect of this maximalism will be to foster or to
dissolve group life.
The fear that soon the only objects of toleration will be eccentric individuals leads
some groups (or their most committed members) to seek positive support from the
state – in the form, say, of subsidies and matching grants for their schools and
mutual aid organizations. Given the logic of multiculturalism, state support must be
provided, if it is provided at all, on equal terms to every social group. In practice,
163
however, some groups start with more resources than others, and then are much
more capable of seizing whatever opportunities the state offers. So civil society is
unevenly organized, with strong and weak groups working with very different rates
of success to help and hold their members. Were the state to aim at equalizing the
groups, it would have to undertake a considerable redistribution of resources and
commit a considerable amount of public money. Toleration is, at least potentially,
infinite in its extent; but the state can underwrite group life only within some set of
political and financial limits.
Summary
It will be useful here to list the successive objects of toleration in the five regimes (I
don’t mean to suggest that they mark a progress; nor is the order in which I have
presented them properly chronological). In the multinational empire as in
international society, it is the group that is tolerated – whether its status is that of an
autonomous community or of a sovereign state. Its laws, religious practices, judicial
procedures, fiscal and distributive policies, educational programs, and family
arrangements are all viewed as legitimate or permissible, subject only to minimal and
rarely strictly enforced (or enforceable) limits. The case is similar in the consociation,
but, now a new feature is added: a common citizenship more effective than that of
most empires, one that at least opens up the possibility of state interference in group
practices for the sake of individual rights. In democratic consociations (such as
Switzerland), this possibility is fully realized, but rights will not be effectively
enforced in the many other cases where democracy is weak, where the central state
exists by mere sufferance of the consociated groups and is mostly focused on holding
them together.
Nation-state citizenship is more meaningful. Now the objects of toleration are
individuals conceived both as citizens and as members of a particular minority. They
are tolerated, so to speak, under their generic names. But membership in the genus
(in contrast to citizenship in the state) is not required of these individuals; their
groups exercise no coercive authority over them, and the state will intervene
aggressively to protect them against any effort at coercion. Hence new options are
made available: loose affiliation with the group, nonaffiliation with any group, or
assimilation to the majority. In immigrant societies, these options are widened.
Individuals are tolerated specifically as individuals under their proper names, and
their choices are understood in personal rather than stereotypical terms. Now there
arise personalized versions of group life, many different ways of being this or that,
which other members of the group have to tolerate if only because they are tolerated
by the society as a whole. Fundamentalist orthodoxy distinguishes itself by its refusal
to take this general toleration as a reason for a more latitudinarian view of its own
religious culture. Sometimes, its protagonists oppose the immigrant society’s regime
of toleration as a whole.
164
7
Practical Issues
Michael Walzer
Power
In ordinary speech, it is often said that toleration is always a relationship of inequality
where the tolerated groups or individuals are cast in an inferior position. To tolerate
someone else is an act of power; to be tolerated is an acceptance of weakness. We
should aim at something better than this combination, something beyond toleration,
something like mutual respect. Once we have mapped out the five regimes, however,
the story looks more complicated: mutual respect is one of the attitudes that makes
for toleration – the most attractive attitude, perhaps, but not necessarily the most
likely to develop or the most stable over time. Sometimes, indeed, toleration works
best when relations of political superiority and inferiority are clearly marked and
commonly recognized. This is most obviously the case in international society, where
ambiguous power relations are one of the chief causes of war. The same proposition
probably holds with regard to some domestic regimes, like the consociation, where
uncertainty about the relative power of different groups may lead to political turmoil
and even to civil war. In immigrant societies, by contrast, the same uncertainty works
in an opposite way: if people are unsure where they stand vis-à-vis others, toleration
is obviously the most rational policy. Even here, however, questions about political
power regularly arise – though perhaps not the single big question, who rules over
whom? Instead a series of smaller questions regularly pose themselves: Who is
stronger most of the time? Who is more visible in public life? Who gets the larger
share of resources? These questions (the big one too) can hardly be understood without
reference to the discussions still to come in this chapter about class, gender,
religion, and so on; but they can also be asked independently.
In multinational empires, power rests with the central bureaucrats. All the
incorporated groups are encouraged to regard themselves as equally powerless, and
hence incapable of coercing or persecuting their neighbors. Any local attempt at
coercion will produce an appeal to the center. So Greeks and Turks, for example,
lived peacefully side by side under Ottoman rule. Were they mutually respectful?
Some of them probably were; some were not. But the character of their relationship
did not depend on their mutual respect; it depended on their mutual subjection.
When subjection isn’t an experience shared equally by all the incorporated groups,
toleration among them is less likely. If one group feels a special affinity with the
imperial center and is able to form an alliance with its local representatives, then it
will often try to dominate the others – like the Greeks did in Roman Alexandria. In
the imperial case, power is most effective in promoting toleration when it is distant,
neutral, and overwhelming.
In this form, imperial power is clearly most helpful to local minorities, who tend
therefore to be the most loyal supporters of the empire. The leaders of national
liberation movements commonly express (and exploit) resentment toward these same
minorities, who are identified now as collaborators with the imperialists. The
transition from imperial province to independent nation-state is a critical moment in
the history of toleration. Often minorities are harassed, attacked; and forced to leave
– as in the case of the Indian traders and artisans of Uganda, who were driven into
exile soon after the withdrawal of the British (and who mostly followed the latter to
Britain, bringing the empire home, as it were, and creating a new diversity in the
imperial center). Groups of this sort sometimes manage to turn themselves into
tolerated minorities, but the path is always hard, and the endpoint, even if it is
successfully reached, probably represents a net loss of security and status for the
minorities. This is one of the common costs of national liberation, though it can be
avoided, or at least mitigated, if the new nation-state is liberal and democratic.
Consociation probably requires something like mutual respect at least among the
leaders of the different groups – for the groups must not only coexist but also
negotiate among themselves the terms of their coexistence. The negotiators, like
diplomats in international society, have to accommodate each other’s interests. When
they can’t or won’t do that, as in Cyprus after the British departure, consociation will
fail. But individual members of the different communities need not accommodate
each other, except when they meet and bargain in the marketplace. In fact,
consociation is probably easiest when the communities don’t have much to do with
one another, when each of them is relatively self-sufficient and inwardly turned. Then
power is expressed – populations counted and wealth put into play – only at the
federal level, where communal leaders argue about budgetary allocations and the
composition of the civil service.
In nation-states, power rests with the majority nation, which uses the state, as we
have seen, for its own purposes. This is no necessary bar to mutuality among
individuals; in fact, mutuality is likely to flourish in liberal democratic states. But
minority groups are unequal by virtue of their numbers and will be democratically
overruled on most matters of public culture. The majority tolerates cultural
difference in the same way that the government tolerates political opposition – by
establishing a regime of civil rights and civil liberties and an independent judiciary to
guarantee its effectiveness. Minority groups then organize, assemble, raise money,
provide services for their members, and publish magazines and books; they sustain
whatever institutions they can afford and think they need. The stronger their internal
165
life and the more differentiated their culture is from that of the majority, the less they
are likely to resent the absence from the public sphere of any representations of their
own beliefs and practices. If minority groups are weak, by contrast, their individual
members will come increasingly to adopt the beliefs and practices of the majority, at
least in public, and often privately as well. It is the intermediate positions that
generate tension and lead to constant skirmishing over the symbolism of public life.
The contemporary French case, as I described it in Chapter 3, provides ample
evidence for the last of these possibilities.
The case is similar early on in the history of immigrant societies, when the first
immigrants aspire to nation-statehood. Successive waves of immigration produce
what is, in principle, a neutral state, the democratic version of imperial bureaucracy.
This state takes over and sustains – for how long no one can know – some of the
practical arrangements and some of the symbolism of its immediate predecessor. So
each new immigrant group has to adjust to, even if it also transforms, the language
and culture of the first group. But the state claims to be above the fray, with no
interest in directing the course of these transformations. It addresses itself only to
individuals and so creates, or tends over time to create, an open society in which
everyone, as I have argued, is engaged in the practice of toleration. The much heralded
move “beyond toleration” is presumably now possible. It remains unclear,
however, whether significant group differences will remain to be respected once this
move is made.
Class
Intolerance is commonly most virulent when differences of culture, ethnicity, or race
coincide with class differences – when the members of minority groups are also
economically subordinated. This subordination is least likely to occur in multinational
empires, where each nation has its own full complement of social classes.
Multinationalism commonly produces parallel hierarchies, even if the different
nations do not share equally in the wealth of the empire. International society is
marked by the same parallelism, and so the inequality of nations produces no
toleration problems (whatever its other problems). State elites interact in ways determined
entirely by differences of power, not of culture; and the elites of dominant
states learn very quickly to respect previously “inferior” cultures when their political
leaders suddenly appear in the council of nations with new wealth, say, or new
weapons.
Ideally, consociations take the same form – the different communities, unequal
within, are roughly equal partners in the country as a whole. But it often happens that
one community that is culturally different is also economically subordinate. The
Lebanese Shi’a provide a useful example – of not only this double differentiation but
also the political disenfranchisement that is its common consequence. The process
works the other way too: when government officials discriminate against the
members of such a group, the hostility these members encounter in every other area
of social life is legitimized and intensified. The worst jobs, the worst housing, the
worst schools: this is their common lot. They constitute an ethnically or religiously
marked lower class. They are tolerated in some minimal sense – allowed their own
houses of worship, for example – but they are strictly on the receiving end of this
toleration. Consociational equality, and the mutual recognition it is supposed to
generate, are both undercut by class inequality.
National minorities in nation-states sometimes find themselves in a similar
position, and sometimes for the same reasons. Whether the causal sequence starts
from cultural stigma or economic or political weakness, it regularly encompasses all
three of these. But it can also happen that relatively powerless national minorities, the
Chinese in Java, say, are well-off economically (though never as well-off as demagogues
rallying the majority against them suggest). Retreating empires often leave
successful minorities dangerously exposed to the intolerance of the new rulers of the
nation-state. This intolerance can take extreme forms – as we have seen in the
example of the Indian settlers in Uganda. Visible prosperity is certain to put a
national minority, especially a new national minority, at risk. Invisible poverty, by
contrast, brings less danger but greater misery, making for radical nonrecognition and
a kind of automatic, unreflective discrimination. Consider the “invisible” men and
women of minority groups (or lower castes) who provide society’s streetcleaners,
garbage collectors, dishwashers, hospital orderlies, and so on – whose presence is
simply taken for granted and who are rarely looked in the eye or engaged in
conversation by members of the majority.
Immigrant societies regularly include groups of this sort – the newest immigrants
from poorer countries, for example, who bring their poverty with them. But longlasting
poverty and cultural stigma are less often the lot of immigrants (who are, after
all, the paradigmatic members of an immigrant society) than of conquered indigenous
peoples and coercively imported groups like the black slaves and their descendants in
the Americas. Here the most radical kind of political subordination goes along with
the most radical kind of economic subordination, with racial intolerance playing an
important role in both cases. The combination of political weakness, poverty, and
racial stigma poses enormously difficult problems for the regime of toleration that
the immigrant society is supposed to be. Stigmatized groups usually do not have the
resources to sustain a strong internal life, so they cannot function like a corporately
organized religious community in an imperial setting (though conquered natives are
sometimes allowed the legal forms of such a community) or like a territorially based
national minority. Nor are their individual members allowed to make their own way,
following in the upward bound footsteps of the immigrants. They form an
anomalous caste at the very bottom of the class system.
Toleration is obviously compatible with inequality whenever the class system is
reiterated, more or less similarly, in each of the different groups. But this
compatibility disappears when the groups are also classes. An ethnic or religious
166
group that constitutes society’s lumpen proletariat or underclass is virtually certain to
be the focus of extreme intolerance – not, indeed, of massacre or expulsion (for the
members of such groups often play an economically useful role that no one else
wants to take on), but of daily discrimination, rejection, and debasement. Other
people are, no doubt, resigned to their presence, but this is not the kind of
resignation that counts as tolerance because it goes along with a wish for their
invisibility. In principle, one could teach respect for people of the underclass and
their roles – as well as a wider toleration for all sorts of people doing all kinds of
work, including hard and dirty work. In practice, neither specific respect nor wider
toleration is likely unless the connection between class and group is broken.
The purpose of affirmative action or reverse discrimination in the admission of
students to universities, the selection of civil servants, and the allocation of
government funds is to break this link between class and group. None of these
efforts are egalitarian so far as individuals are concerned; individuals are merely
moved up or down the hierarchy. Affirmative action is egalitarian only at the group
level, where it aims to produce similar hierarchies by supplying the missing upper,
professional, or middle class to the most subordinate groups. If the social profile of
all the groups is more or less the same, cultural difference is more likely to be
accepted. This proposition doesn’t hold in cases of severe national conflict, but
where pluralism already exists, as in consociations and immigrant societies, it seems
plausible. At the same time, the experience of the United States suggests that
privileging the members of subordinate groups, whatever its useful long-term
consequences, reinforces intolerance in the short term. It causes real injustice to
particular individuals (usually members of the next-most-subordinate groups), and it
breeds politically dangerous resentments. It may well be the case, then, that wider
tolerance in pluralist societies requires a wider egalitarianism. The key to success in
these regimes of toleration may not be – or may not only be – the reiteration of
hierarchy in each group, but also the reduction of hierarchy across the society as a
whole.
Gender
Questions about family arrangements, gender roles, and sexual behavior are among
the most divisive in all contemporary societies. It is a mistake to think that the
divisiveness is entirely new: polygamy, concubinage, ritual prostitution, the seclusion
of women, circumcision, and homosexuality have been argued about for millennia.
Cultures and religions have marked themselves off by their distinctive practices in
these matters – and then have criticized the practices of the “others.” But a virtually
universal male domination set limits to what could be argued about (and who could
join the argument). Today, widely accepted ideas about equality and human rights call
those limits into question. Everything now is open to debate, and every culture and
religion is subject to a new critical scrutiny. This sometimes makes for toleration but
sometimes, obviously, for its opposite. The theoretical and practical line between the
tolerable and the intolerable is most likely to be fought over and eventually drawn
here, with reference to what I will call, summarily, questions of gender.
The great multinational empires commonly left these questions to their
constituent communities. Gender was taken to be an inherently internal affair; it
didn’t, or it was not supposed to, involve any kind of communal interaction. Strange
commercial customs were not tolerated in the common markets, but family law
(“private” law) was left entirely to the traditional religious authorities or the (male)
elders. Customary practice was also in their hands; imperial officials were unlikely to
intervene.
Consider the extraordinary reluctance with which the British finally, in 1829,
banned the suttee (the self-immolation of a Hindu widow on her husband’s funeral
pyre) in their Indian states. For many years, the East India Company and then the
British government tolerated the practice because of what a twentieth-century
historian calls their “declared intention of respecting both Hindu and Muslim beliefs
and allowing the free exercise of religious rights.” Even Muslim rulers, who had,
according to this same historian, no respect whatsoever for Hindu beliefs, made only
sporadic and half-hearted efforts to suppress the practice. Imperial toleration
extends, then, as far as the suttee, which – given British accounts of what the practice
actually involved – is pretty far.
It is at least conceivable that consociational arrangements might produce a similar
toleration, if the power of the joined communities was in near balance and the
leaders of one of them were strongly committed to this or that customary practice. A
nation-state, however, where power is by definition unbalanced, would not tolerate
customs like the suttee among a national or religious minority. Nor is toleration at
that reach likely in an immigrant society, where each of the groups is a minority
relative to all the others. The case of the Mormons in the United States suggests that
deviant practices like polygamy won’t be tolerated even when they are wholly
internal, when they involve “only” domestic life. In these last two cases, the state
grants equal citizenship to all its members – including Hindu widows and Mormon
wives – and enforces a single law. There are no communal courts; the whole country
is one jurisdiction within which state officials are bound to stop a suttee in progress
in exactly the same way that they are bound to stop a suicide attempt if they possibly
can. And if the suttee is coercively “assisted,” as in fact it often was, the officials have
to treat the coercion as murder; there are no religious or cultural excuses.
That, at least, is what follows from the nation-state and immigrant society models
as I have described them. But reality sometimes lags behind – as with another ritual
practice involving women’s bodies: genital mutilation or, more neutrally,
clitoridectomy and infibulation. These two operations are commonly performed on
infant girls or young women in a large number of African countries, and because no
one has suggested humanitarian intervention to stop them, we can say that they are
tolerated in international society (tolerated at the state level, but actively opposed by a
167
number of organizations working in international civil society). The operations are
also performed in African immigrant communities in Europe and North America.
They have been specifically outlawed in Sweden, Switzerland, and Great Britain,
though without any serious effort to enforce the ban. In France, the classic nationstate
(which is now also, as we have seen, an immigrant society), some 23,000 girls
were said to be “at risk” in the mid-1980s. How many of these were actually operated
on is unclear. But there have been a number of highly publicized trials (under a general
law against mutilation) of the women who perform the operations and of the
mothers of the girls. The women have been convicted, and their sentences then
suspended. In effect, the practice (as of the mid-1990s) is condemned publicly but
tolerated in fact.
The argument for toleration has to do with “respecting cultural diversity” – a
diversity conceived, as in the standard nation-state model, to follow from the choices
of stereotypical members of a cultural community. Thus a 1989 petition against
criminalizing what the French call “excision”: “Demanding a penal sentence for a
custom that does not threaten the republican order and that there is no reason not to
assign to the sphere of private choice, like circumcision, would demonstrate an
intolerance that can only create more human drama than it claims to avoid, and that
manifests a singularly narrow conception of democracy.” As with the suttee, it is
important to get the description right: clitoridectomy and infibulation “are
comparable… not to the removal of the foreskin but to the removal of the penis,”
and it is hard to imagine circumcision in that form being treated as a matter of
private choice. In any case, the infant girls are not volunteers. And the French state,
one would think, owes them the protection of its laws: some of them are citizens,
and most of them will be the mothers of citizens. They are in any case residents of
France and future participants in the social and economic life of the country; and
although they may remain wholly confined within the immigrant community, they
also (this is the advantage of living in France) may not. With regard to individuals like
that, toleration surely should not extend to ritual mutilation, any more than it does to
ritual suicide. Cultural diversity at that extreme is only protected against interference
when boundaries are much more firmly drawn than they are or can be drawn in
nation states or immigrant societies.
In other sorts of cases, where the moral values of the larger community – the
national majority or the coalition of minorities – are not so directly challenged, the
excuse of religious or cultural difference (and “private choice”) may be accepted,
diversity respected, and nonstandard gender practices tolerated. Thus the case of
narrowly constituted or sectarian minorities like the American Amish or the Hasidim,
to whom state authorities are sometimes ready to offer (or the courts to mediate) one
or another compromise arrangement – the separation of the sexes in school buses
and even in classrooms, for example.
But similar concessions won’t so readily be offered to larger, more powerful (and
more threatening) groups even in relatively minor cases – and the standing
compromises can always be challenged by any sect or group member who claims her
citizen rights. Imagine that an arrangement is worked out (as it surely should be)
allowing Muslim girls in French public schools to wear their customary headdress.
This would be a compromise with the nation-state norm, one that would recognize
the right of immigrant communities to a (modestly) multicultural public sphere. The
laicist traditions of French education would continue to govern the school calendar
and curriculum. Imagine now that a number of Muslim girls claim that they are being
coerced by their families to wear the headdress and that the compromise arrangement
facilitates this coercion. Then the compromise would have to be
renegotiated. In the nation-state and the immigrant society, though not in the
multinational empire, the right to be protected against coercion of this kind (as, more
obviously, one would be protected against the far more severe coercion of
clitoridectomy) would take precedence over the “family values” of the minority
religion or culture.
These are matters of extraordinary sensitivity. The subordination of women –
manifest in seclusion, bodily concealment, or actual mutilation – is not aimed solely
at the enforcement of patriarchal property rights. It also has to do with cultural or
religious reproduction, of which women are taken to be the most reliable agents.
Historically, men have entered into the larger public life of armies, courts, assemblies,
and markets; they are always potential agents of novelty and assimilation. Just as
national culture is better preserved in rural than in urban settings, so it is better preserved
in private or domestic than in public settings – which is to say, in the standard
cases, among women rather than among men. Tradition is transmitted in the lullabies
that mothers sing, the prayers they whisper, the clothing they make, the food they
cook, and the domestic rites and customs that they teach. Once women enter the
public sphere, how will this transmission be effected? It is because education is the
first point of entry that questions like the wearing of traditional headdress in public
schools are so fiercely contested.
This is the form the argument takes when a traditional culture or religion
encounters the nation-state or the immigrant society. “You are committed to
tolerating our community and its practices,” the traditionalists say. “Given that
commitment, you cannot deny us control over our children (and particularly our
female children) – else you are not in fact tolerating us.” Toleration implies a right to
communal reproduction. But this right, if it exists, comes into conflict with the rights
of individual citizens – which were once confined to men and were therefore not so
dangerous, but are now extended to women. It seems inevitable that individual rights
will win out in the long run, for equal citizenship is the basic norm of both the
nation-state and the immigrant society. Communal reproduction will then be less
certain or, at least, it will be realized through processes that yield less uniform results.
Traditionalists will have to learn a toleration of their own – for different versions of
their own culture or religion. But before that lesson is learned, we can expect a long
168
series of “fundamentalist” reactions that are focused most often on questions of
gender.
The abortion wars in the United States today suggest the character of this
reactionary politics. From the fundamentalist side, the moral issue is whether society
will tolerate the murder of infants in the womb. But the political issue, for both sides,
has a different focus: who will control the sites of reproduction? The womb is only
the first of these; home and school come next and are already, as we have seen, in
dispute. What cultural differences will remain to be tolerated once these disputes
have been resolved, as they eventually will be, in favor of female autonomy and
gender equality? If the traditionalists are right, nothing will remain. But they are
unlikely to be right. Gender equality will take different forms in different times and
places, and even in the same time and place among different groups of people, and
some of these forms will turn out to be consistent with cultural difference. It may
even happen that men will play a larger role in sustaining and reproducing the
cultures they claim to value.
Religion
Most people in the United States, in the West generally, believe that religious
toleration is easy. They read about religious wars near to home (in Ireland and
Bosnia) or far away (in the Middle East or Southeast Asia) with incomprehension.
Religion in those places must be contaminated by ethnicity or nationalism, or it must
take some extreme, fanatical, and therefore (as we understand things) unusual form.
For haven’t we proved that freedom of worship, voluntary association, and political
neutrality work together to reduce the stakes of religious difference? Don’t these
tenets of American pluralism encourage mutual forbearance and make for a happy
coexistence? We allow individuals to believe what they want to believe, to join freely
with fellow believers, to attend the church of their choice – or to disbelieve what they
want to disbelieve, to stay away from the church of their choice, and so on. What
more could anyone want? Isn’t this the model of a toleration regime?
In fact, of course, there are other actual or possible regimes: the millet system was
specifically designed for religious communities, and consociations commonly bring
together different religious or ethnic groups. But the toleration of individual
believers, as this was first worked out in England in the seventeenth century and then
carried across the Atlantic, is the dominant model today. And so it is necessary to
look closely at some of its complications. I want to consider two issues of historical
and contemporary importance: first, the persistence at the margins of modern nationstates
and immigrant societies of religious groups that demand recognition for the
group itself rather than for its individual members, and second, the persistence of demands
for “religious” tolerance and intolerance that extend beyond association and
worship to a great variety of other social practices.
One reason that toleration works so easily in countries like the United States is
that the churches and congregations that individuals form, whatever their theological
disagreements, are, mostly, very much like one another. Seventeenth-century
toleration was first of all a mutual accommodation of Protestants. And in the United
States, after an early effort to establish a “holy commonwealth” in Massachusetts, the
expanding toleration regime tended to protestantize the groups that it included.
American Catholics and Jews gradually came to look less and less like Catholics and
Jews in other countries: communal control weakened; clerics spoke with less
authority; individuals asserted their religious independence, drifted away from the
community, and intermarried; fissiparous tendencies well known from the first days
of the Reformation became a general feature of American religious life. Toleration
accommodated difference, but it also produced among the different groups a pattern
of accommodation to the Protestant model that made coexistence easier than it
might have been.
Some groups, however, resisted – Protestant sects determined to escape the
“dissidence of dissent” (the ground, so to speak, in which they had originally taken
root) and orthodox factions within the traditional religious communities. I will
continue to refer to the examples of each mentioned earlier: the American Amish
and the Hasidim. The toleration regime accommodated these groups too, though
only at the margins. It permitted them their isolation, arid it compromised with them
on critical questions like public schooling. The Amish, for example, were for a long
time permitted to educate their children at home; when they were finally required,
first by the state of Pennsylvania and then by the Supreme Court (with reference to a
Wisconsin case), to send the children to public schools, they were allowed to
withdraw them at an earlier age than that stipulated in the law. In principle, what was
tolerated was a series of individual choices, made in successive generations, to join
the Amish congregations and to worship in the Amish manner. In practice, it was the
Amish community as a whole and its coercive control of its own children (which was
only partly mitigated by public schooling) that was, and is, the real object of
toleration. For the sake of (this kind of) toleration, we allow Amish children to
receive less of an education in citizenship than we require of American children
generally. The arrangement is justified in part by the marginality of the Amish, and in
part by their embrace of marginality: their deep commitment not to live anywhere
except on the margins of American society and not to seek any influence beyond
them. Other similarly marginal religious sects have maintained a similar control over
their children largely unchallenged by the liberal state.
The most interesting feature of early American toleration was the exemption from
military service of the members of certain Protestant sects well known for their
pacifist convictions. Today, conscientious objection is an individual right, though the
sign of conscientiousness that the political authorities are most ready to recognize is
membership in those same sects. In its origins, however, objection was effectively a
group right. Indeed, claims of conscience over a wide range of social issues – the
refusal of oaths, of jury service, of public schooling, of taxes; the demand for
polygamous marriage, animal sacrifice, ritual drug use, and so on – gain whatever
169
legitimacy they have, even today, because they are religious practices, features of a
collective way of life. These practices would have no legitimacy at all if they were put
forward on a purely individual basis, even if the individuals insisted that their
understanding of what they ought to do, or not to do, was a co-knowledge (conscience)
shared between each of them and his or her God.
Minority religious practices and prohibitions, beyond association and worship, are
tolerated or not depending on their visibility or notoriety and the degree of outrage
they arouse in the majority. A great variety of practical accommodations are available
in both nation-states and immigrant societies. Men and women who tell the
authorities that their religion requires them to do this or that may well be permitted
to do it, even if no one else is, especially if they do it quietly. And communal leaders
who tell the authorities that their coercive power is necessary to the survival of the
community may well be permitted to exercise that power, subject to certain liberal
constraints. But pressure is steady, if only intermittently forceful, toward the
individualist model: the community conceived as a free assembly – entrances and
exits open, with little claim and little capacity to shape the everyday life of the
participants.
At the same time, this regime of toleration is under pressure in the contemporary
United States from groups within the (Christian) majority who have no quarrel with
the freedom of assembly or worship but fear the loss of social control. They are
prepared to tolerate minority religions (they are advocates, therefore, of religious
liberty), but they have no tolerance for personal liberty outside the house of worship.
If sectarian communities aim to control the behavior of their own people, the more
extreme members of religious majorities aim to control everyone’s behavior – in the
name of a supposedly common (Judeo-Christian, say) tradition, of “family values,” or
of their own certainties about what is right and wrong. This is surely an example of
religious intolerance. It is a sign of the partial success of the regime of toleration,
however, that antagonism is not directed against particular minority religions but
rather against the ambience of freedom that the regime as a whole creates.
No doubt toleration flourishes in this ambience – and even reaches what I have
described as its most intensive form – but religious toleration, at least, doesn’t
depend upon it. Extensive restrictions on personal freedom such as a ban on
abortion, the censorship of books and magazines (or of texts in cyberspace),
discrimination against homosexuals, the exclusion of women from certain
occupations, and so on, even if they are the products of religious intolerance, are
entirely compatible with religious toleration – that is, with the existence of many
different churches and congregations whose members worship freely in many
different ways. The contradiction is not between toleration and restriction; it lies
deep within the idea of religious toleration itself, because virtually all the tolerated
religions aim to restrict individual freedom, which is, for liberals at least, the
foundation of the idea. Most religions are organized to control behavior. When we
require them to give up this aim, or to give up the means necessary to its
achievement, we are requiring a transformation whose end product we cannot yet
describe.
Entirely free religious communities already exist, of course, but they don’t seem
satisfying to all, perhaps not even to most, believers. Hence the recurrence of sectarian
and cultic religiosity and of fundamentalist theologies, which challenge the
prevailing regime of toleration. Assuming that the challenges are overcome (the same
assumption that I have made in previous sections), what then? What will be the
staying power and organizational strength of a purely voluntary faith?
Education
Schools have already figured significantly in this essay – most particularly in the
discussion of gender and cultural reproduction. But there is an important issue that I
must address here (and again in the section on civil religion), which has to do with
the reproduction of the regime of toleration itself. Doesn’t the regime have to teach
all of its children, whatever their group memberships, the value of its own
constitutional arrangements and the virtues of its founders, heroes, and current
leaders? And won’t that teaching, which is more or less unitary in character, interfere
with or at least compete with the socialization of children into the various cultural
communities? The answer, of course, is affirmative in both cases. All domestic
regimes have to teach their own values and virtues, and this teaching is certainly
competitive with whatever else children are taught by their parents or in their
communities. But the competition is or can be a useful lesson in (the difficulties of)
mutual toleration. State teachers must tolerate, say, religious instruction outside of
their schools, and the teachers of religion must tolerate state organized instruction in
civics, political history, the natural sciences, and other secular subjects. The children
presumably learn something about how toleration works in practice and – when
creationists, for example, challenge state instruction in biology – something also
about its inevitable strains.
Multinational empires make the most minimal demands on the educational
process. Their political history, which consists mostly of wars of conquest, is unlikely
to inspire feelings of loyalty in the conquered peoples, and so it is best left out of the
official curriculum (it is more likely to figure in communal stories of heroism in
defeat). Loyalty to the emperor, portrayed as emperor of all his peoples, is more
often taught. The emperor, rather than the empire, is the focus of official education,
for the latter often has a clear national character whereas the individual leaders can at
least pretend to rise above their national origin. Sometimes, indeed, they aim at
radical transcendence, deification, which frees them from any particularist identity.
But it is nonetheless an example of religious intolerance when the deified emperor
demands to be worshipped by his subjects – like those Roman rulers who tried to
bring statues of themselves into the temple in Jerusalem. The school is a better venue
170
for the imperial image, which can look benignly down upon children studying
anything at all, in any language, under any local or communal auspices.
Consociations can also teach a minimalist curriculum, one that is focused on an
often sanitized history of communal coexistence and cooperation and on the
institutions through which these are realized. The longer the coexistence has lasted,
the more likely it is that the common political identity will have taken on cultural
content of its own – as Swiss identity clearly has done – and will have become fully
competitive with the identities of the different communities. Still, what is taught, in
principle at least, is a political history in which these communities have a recognized
and equal place.
The case is very different, obviously, in nation-states with national minorities,
where one community is privileged over all others. This kind of regime is far more
centralized than are empires and consociations, and so it has a greater need
(particularly if it is democratically organized) for citizens – men and women who are
loyal, engaged, competent, and familiar with the style, as it were, of the dominant
nation. State schools will aim to produce citizens of this kind. Thus Arabs in France,
for example, will be taught to be loyal to the French state, engaged in French politics,
competent in the practices and expressive modes of French political culture, and
knowledgeable about French political history and institutional structures. By and
large, Arab parents and children seem to accept these educational aims; they have
sought, as we have seen, to assert their Arab or Muslim commitment only through
the symbolism of dress, not to alter the curriculum. They are, or they seem to be,
content to sustain their own culture in nonstate schools, in religious settings, and at
home. But French citizenship is a weighty matter, with resonance far beyond the
narrowly political sphere. Its integrative and assimilative power has been demonstrated
over many years and must appear to many parents, if not to their children,
as a cultural threat. The more countries like France become (like) immigrant societies,
the more this threat will be resisted.
What form that resistance is likely to take can be seen in the curricular wars of an
immigrant society like the United States. Here children are taught that they are
individual citizens of a pluralist and tolerant society – where what is tolerated is their
own choice of cultural membership and identity. Most of them, of course, are already
identified, because of the “choices” of their parents or, as in the case of racial identities,
because of their location in a social system of differentiation. But they are, as
Americans, entitled to make further choices and required to tolerate the existing
identities and the further choices of their fellows. This freedom and this toleration
constitute what we can call American liberalism.
The schools teach children from all of America’s ethnic, religious, and racial
groups to be liberal in this sense, and so to be Americans – much as children in
French schools are taught to be republican and therefore French. But American
liberalism is culturally neutral in a way that French republicanism cannot be. This
difference seems to fit the two political doctrines: republicanism, as Rousseau taught,
requires a strong cultural base to sustain high levels of participation among the
citizens; liberalism, which is less demanding, can allow more room for private life and
cultural diversity. But such differences can easily be exaggerated. Liberalism is also a
substantive political culture that has its origins, at least, in Protestant and English
history. The recognition that American schools in fact reflect this history, and can
hardly be neutral with regard to it, has led some non-Protestant and non-English
groups to call for a multicultural education – which presumably requires not the
subtraction of the liberal story from the curriculum but the addition of other stories.
It is commonly and rightly said that the point of multiculturalism is to teach
children about each other’s culture, to bring the pluralism of the immigrant society
into its classrooms. Whereas the earlier version of neutrality, which was conceived or
misconceived as cultural avoidance, aimed to make all the children into Americans
simply (which is to say as much like English Protestants as possible), multiculturalism
aims to recognize them as the hyphenated Americans they are and to lead them to
understand and admire their own diversity. There is no reason to think that this
understanding or admiration stands in any tension with the requirements of liberal
citizenship – though it is important to stress again that liberal citizenship is more
relaxed than that of a republican nation-state.
But multiculturalism is also, sometimes, a program of a different sort, one aimed
at using the state schools to strengthen threatened or devalued identities. The point
isn’t to teach other children what it means to be different in a certain way, but to
teach children who are supposed to be different how to be different in the right way.
Hence the program is illiberal, at least in the sense that it reinforces established or
presumed identities and has nothing to do with mutuality or individual choice. It
probably also entails some form of educational separation, as in the theory and
practice of Afrocentrism, which is a way of providing for black children in the state
schools what the Church provides for Catholic children in private schools. Now
pluralism exists only in the system as a whole, not in the experience of each child,
and the state must step in to compel the various schools to teach, whatever else they
teach, the values of American liberalism. The Catholic example suggests that an
immigrant society can make do with this arrangement, at least as long as the bulk of
its schoolchildren are in mixed classrooms. Whether liberal politics could be
sustained if all children received some version (their “own” version) of a Catholic
parochial or Afrocentric education is more doubtful. Success would then depend on
the effects of education outside the school: the everyday experience of mass communication,
work, and political activity.
Civil Religion
Think of what is taught in state schools about the values and virtues of the state itself
as the secular revelation of a “civil religion” (the term is Rousseau’s). Except in the
case of the deified emperor, this revelation is religious mostly by analogy, but the
analogy is worth pursuing. For here, as the school example makes clear, is a
171
“religion” that can’t be separated from the state: it is the very creed of the state,
crucial to its reproduction and stability over time. Civil religion consists of the full set
of political doctrines, historical narratives, exemplary figures, celebratory occasions,
and memorial rituals through which the state impresses itself on the minds of its
members, especially its youngest or newest members. How can there be more than
one such set for each state? Surely civil religions can tolerate each other only in
international society, not within a single domestic regime.
In fact, however, civil religion often makes for intolerance in international society
by encouraging parochial pride about life on this side of the border and suspicion or
anxiety about life on the other side. Its domestic effects, by contrast, can be benign,
because it provides everyone (on this side of the border) with a common basic
identity and so makes subsequent differentiation less threatening. Certainly civil
religion, like state education, is sometimes competitive with group membership: thus
the case of French republicans and French Catholics in the nineteenth century – or
of republicans and Muslims today. But because civil religions commonly have no
theology, they can also be accommodating of difference, even or especially of
religious difference. Despite the specific historical conflict of the revolutionary years,
then, there is no reason why a believing Catholic cannot also be a committed
republican.
Toleration is most likely to work well when the civil religion is least like a . . .
religion. Had Robespierre, for example, succeeded in binding republican politics to a
fully elaborated deism, he might well have created a permanent barrier between
republicans and Catholics (and Muslims, and Jews). But his failure is emblematic:
political creeds take on the baggage of genuine religious belief at their peril. One
might say the same thing about the baggage of genuine antireligious belief. Militant
atheism made the communist regimes of Eastern Europe as intolerant as any other
orthodoxy would have done – and politically weak as a result: they were unable to
incorporate large numbers of their own citizens. Most civil religions wisely make do
with a vague, unelaborated, latitudinarian religiosity, one that is more a matter of
stories and holidays than of clear or firm beliefs.
Of course, it may be just this latitudinarianism that orthodox religious groups
object to, fearing that it will make their children tolerant of religious error or secular
disbelief. It is hard to know how to respond to anxieties of that sort; one hopes that
they are justified and that the public schools and the stories and holidays of the civil
religion will have exactly the effects that orthodox parents fear. Parents are free to
pull their children out of the public schools and to escape the civil religion through
one or another form of sectarian isolation. But it makes no sense to argue that
respect for diversity bars an immigrant society like the United States from teaching
respect for diversity. And it is certainly a legitimate form of such a liberal education
to tell stories about the history of diversity and to celebrate its great occasions.
In nation-states, the stories and celebrations will be of a different sort: they will
come out of, and teach the value of, the historical experience of the majority nation.
So the civil religion makes further differentiation possible within the majority – along
religious, regional, and class lines – but provides no bridge to minority groups.
Instead, it sets the standard for individual assimilation: it suggests, for example, that
to become French you have to be able to imagine that your ancestors stormed the
Bastille or, at least, that they would have done so had they been in Paris at the right
time. But a national minority with a civil religion of its own can still be tolerated, so
long as the rites are celebrated privately. And its members can become citizens, can
learn the ways of, say, French political culture, without any imaginative investment in
Frenchness.
The common identity fostered by a civil religion is especially important in
immigrant societies where identities are otherwise so diverse. In multinational
empires, obviously, identities are even more diverse, but there, beyond the unifying
figure of the emperor and the common allegiance he claims, commonality is less
important. Contemporary immigrant societies are also democratic states, and they are
dependent for their political health on some degree, at least, of commitment and
activism among their citizens. But if the local civil religion is to enhance and celebrate
these qualities, it must accommodate not only other religions but also other civil
religions. Its most enthusiastic protagonists, of course, will want to replace the
others: that was the point, for example, of the Americanization campaigns of the
early twentieth century. And perhaps, indeed, that will be the long-term effect of the
American experience. Perhaps every immigrant society is a nation-state in the
making; and civil religion is one of the instruments of this transformation.
Nonetheless, a campaign on its behalf is an act of intolerance, an act likely to
provoke resistance and to multiply divisions among (and also within) the different
groups.
It turns out, in any case, that a civil religion like Americanism can live fairly
comfortably with what might be called alternative civil religious practices among its
own participants. The stories and celebrations that go along with, for example,
Thanksgiving, Memorial Day, or the Fourth of July can coexist in the common life of
Irish-Americans, African-Americans, or Jewish-Americans with very different stories
and celebrations. Difference here is not contradiction. Beliefs come into opposition
far more readily than stories do, and one celebration doesn’t deny, cancel, or refute
another. Indeed, it is easier to watch the private communal or familial celebrations of
our fellow citizens if we know that they will also be celebrating publicly with us on
some other occasion. So civil religion facilitates the toleration of partial differences –
or it encourages us to think of difference as only partial. We are Americans but also
something else, and safe as something else insofar as we are Americans.
No doubt there are, or there might be, minority civil religions, ideologically or
theologically elaborated, that contradict American values, but these have not been
much in evidence in American public life. Similarly, it isn’t difficult to imagine a more
intolerant Americanism, one, for example, defined in Christian terms; connected
exclusively, even racially, to its European origins; or given some narrow political
172
content. Americanisms of this sort have existed in the past (hence the notion of “unAmerican
activities” developed by the anticommunist right in the 1930s) and continue
to exist, but none constitutes the dominant version right now. It isn’t only in
principle but also in reality that American society is a collection of individuals with
multiple, partial identities. Of course, religions have often involved denials of such
realities, and civil religions can attempt a similar denial. It may even be true that the
pattern of difference in the United States and in other immigrant societies is unstable
and impermanent. Even so, a Kulturkampf against difference is not the best response
to this condition. Civil religion is more likely to succeed by accommodating than by
opposing the multiple identities of the men and women it aims to engage. Its object,
after all, is not full-scale conversion but only political socialization.
Tolerating the Intolerant
Should we tolerate the intolerant? This question is often described as the central and
most difficult issue in the theory of toleration. But that can’t be right, because most
of the groups that are tolerated in all four domestic regimes are in fact intolerant.
There are significant “others” about whom they are neither enthusiastic nor curious,
whose rights they don’t recognize – to whose existence, indeed, they are neither
indifferent nor resigned. In multinational empires, the different “nations” are perhaps
temporarily resigned; they accommodate themselves to coexistence under imperial
rule. But were they to rule themselves, they would have no reason for resignation,
and some of them certainly would aim at ending the old coexistence in one way or
another. That might be a good reason for denying them political power, but it is no
reason at all for refusing to tolerate them in the empire. The case is the same in
consociations, where the whole point of the constitutional arrangement is to restrain
the likely intolerance of the associated communities.
Similarly, minorities in nation-states and immigrant societies are and ought to be
tolerated even if it is known that their compatriots or fellow believers in power in
other countries are brutally intolerant. These same minorities cannot practice
intolerance here (in France, say, or in America), that is, they cannot harass their
neighbors or persecute or repress deviant or heretical individuals in their midst. But
they are free to excommunicate or ostracize deviants and heretics, and they are
equally free to believe and say that such people will be damned forever or denied a
place in the world to come – or that any other group of their fellow citizens are living
a life that God rejects or that is utterly incompatible with human flourishing. Indeed,
many of the Protestant sectarians for whom the modern regime of toleration was
first designed, and who made it work, believed and said just such things.
The point of separating church and state in the modern regimes is to deny political
power to all religious authorities, on the realistic assumption that all of them are at
least potentially intolerant. Given the effectiveness of the denial, they may learn
tolerance; more likely, they will learn to live as if they possessed this virtue. Many
more ordinary believers obviously do possess it, especially in immigrant societies,
where daily encounters with both internal and external “others” are unavoidable. But
these people too need separation, and they are likely to support it politically as a way
of protecting themselves, and everyone else, against the possible fanaticism of their
fellow believers. The same possibility of fanaticism also exists among ethnic activists
and militants (in immigrant societies), and so ethnicity too has to be separated from
the state, for exactly the same reasons.
Democracy requires yet one more separation, one that is not well understood: that
of politics itself from the state. Political parties compete for power and struggle to
enact a program that is, let’s say, shaped by an ideology. But the winning party,
though it can turn its ideology into a set of laws, cannot turn it into the official creed
of the civil religion; it cannot make the day of its ascension to power into a national
holiday, insist that party history be a required course in the public schools, or use
state power to ban the publications or the assemblies of other parties. That is what
happens in totalitarian regimes, and it is exactly analogous to the political
establishment of a single monolithic church. Religions that hope for establishment
and parties that dream of total control can be tolerated both in liberal democratic
nation-states and in immigrant societies, and they commonly are. But (as I suggested
at the beginning of this essay) they can also be barred from seizing state power, and
even from competing for it. What separation means in their case is that they are
confined to civil society: they can preach and write and meet; they are permitted only
a sectarian existence.
In: On Toleration, New Haven and London: Yale University Press, pp. 14-36, 52-82.