The transition from caste to class-based identifiers could scarcely be made quickly and completely in any society -- some kind of transitional device or marker is needed. Such a device enables some portion of a population to move into a new space where rules of equality apply, leaving the rest of the population in the old space where rules of inequality prevail. Admittedly the concept of a nation was never conceptualized in this way, but that was its hidden agenda. In order for the concepts of nationalism to work, they had to be taken at face value. Unless members of a nation truly believed that they were privileged to belong to a special community entitled to sovereignty and self-government, they would scarcely have accepted the equalitarian norms that granted all of them the right to move out of inherited occupations to embrace new ones, and to choose their own marriage partners rather than permit their families and social pressures to dictate who they would marry. The latent function of nations, therefore, was to provide a platform on which to achieve the transition from caste to class, while excluding the rest of the population who remained subject to the traditional constraints imposed by family and occupation.

In a sense, nationalism, like ethnicity, is an ancient phenomenon. There have always been distinctions between "we" and "them," between the "in-group" and the "out-group," the "endocentric" and the "exophobic." Nation comes from a Latin root, natus, meaning "to be born." Related derivatives are natal, native, nascent, nature, all pointing to origins and original form. It came to refer to a people who share common customs, origins, history, and language. Comparable notions are expressed by society, which comes from the Latin socius, meaning ally or companion. Its derivatives include social, sociable, associate, and association. In one sense society refers to an upper class, but in sociological usage, it has come to mean all those who share institutions, culture, and interactions. Traditionally, those who belonged to a nation or society thought well of themselves and looked down on outsiders. Residues of these ancient attitudes persist to this day: for example, Eurocentrism continues to mark the way Westerners view the world, as does the Sinic view of the world which has long seen China as the "middle country," the "jung quo".

Terms used to refer to outsiders are typically derogatory. The word, pagan, for example, comes from pagus which referred, in Latin, to a district or village. From this root comes peasant, a word that retains its rural connotation, while pagan came to mean a non-believer in a very pejorative sense. Savage shares a root with sylvan, coming from the Latin form, silva, signifying forests, and hence forest dwellers. Barbarian comes from the Greek, barbaros meaning non-Greek.

These ancient concepts had no connotations of ethnicity or political rights. A well-known test case can be found in the Jewish Exodus from Egypt. The ancient Hebrews can be seen, anachronistically, as an ethnic nation struggling to escape from slavery and establish its own state. A more authentic interpretation, I believe, is provided by the book of Exodus which attributes the flight to supernatural revelations that led Moses to lead his followers in search of the "Promised Land" Jehovah had set aside for His "Chosen People." Moses did not claim any inherent right of the Jewish nation to equality and sovereignty. Rather, he relied upon sacred authority as he perceived it. The Crusades were also wars predicated upon competing sacred rights, not ethnic differences.

In the great majority of cases, however, armed conflicts between rival rulers or overlords lacked legitimation by sacred claims or ethnic differences. Instead, dynastic rivalries or simple struggles for power between greedy rivals provided the main explanation. Leaders called upon their followers to support them. Nationalism in this form can be understood as simple ethnocentrism, the idea that one's own group is superior to others. Perhaps more realistically, it involved rival factions, defined as a typically caste-based alliance of patrons and clients who contend with rival factions for wealth, land, or privileges. Such factions consisted of families, clans, tribes, or kingdoms -- their conflicts involved greed, pride, or vengeance, but not ethnic claims to sovereignty.

#Modern Nationalism

By contrast, modern nationalism has evolved as a new form of ethnocentrism based on profoundly different motives that have been generated by modernity. At the societal level, the transition from caste-based inequalitarianism to class-oriented equalitarianism has simultaneously mobilized marginalized communities to demand more equitable treatment and also activated defensive reactions by more privileged communities, thereby angering lower status groups and prodding them to react. This is a classic form of circular causation in which A provokes B to provoke A in an endless cycle. Its basic premise is the threat/promise of equality that moves upper and lower status groups to contend with each other. Such conflicts were not generated in caste-based societies where status inequalities were accepted as normal and inescapable.

At the political level, the transition from caste to class structures leads to the emergence of states in which representative government replaces kingship, thereby changing the basis for legitimacy of social stratification. The traditional system attributed political authority to sacred forces that legitimized inequality as both unavoidable and just, whereas the new order repudiated these beliefs, replacing them with the notion that sovereignty rests in the right of individuals to govern themselves or elect representatives to act on their behalf. The promise of equality among citizens and equitable policies enforced by representatives in government changed the ground rules.

Traditional nationalism hinged on the criteria of royal legitimacy -- subjects felt honor-bound to support their kings, to offer tribute to them and to support their wars. That was the basis for thinking of a regime as a nation. It is well symbolized by the struggle of Joan of Arc to create a viable monarchy and defend its interests against the outsiders.

Modern nationalism replaces the king by the people as the source of sovereignty and the focus of loyalty. Patriotism is re-defined as a willingness to fight for one's country rather than one's king. A deep problem arose in making this transition, however, because the political revolution involved in transforming a monarchy into a republic did not automatically produce patriotism among all the residents of a country. Many would not easily abandon their loyalty to the dethroned king, and others would see themselves as outsiders, alienated from the new republic. Retaining a king under the guise of a constitutional monarchy might ease this transition, yet it could also create ambivalence -- even today, it is not clear in such regimes whether officials are "His Majesty's Servants" or servants of the people. In England, the House of Lords has only recently abolished the right of Lords to inherit their seats, a residue of the caste system. Today the Hereditary Lords are being replaced by Life Lords selected by the regime to enhance support for its policies. As for the House of Commons, it has achieved the transformation to popular representation after many reforms that began two centuries ago.

From the beginning, it has never been quite clear who, exactly, are the members of a republic with the right to be represented in its governance. Oligarchic forces seeking to protect their property and establish a ruling elite could not easily establish boundaries to protect themselves from outsiders seeking the privileges and status of people as constituents of the regime. The American Constitution opens with the claim that We, the People of the United States...do ordain and establish this Constitution... The word, people, comes from the Latin term, populus, from which other words like popular, populace, population, and public are also derived. Although sharing a common etymology, their connotations have diverged. If we use population to mean all humans in the world, or within any territorial boundary, then we can make a distinction between people as those in a population who count, and populace as those who do not. Since the boundaries are never sharp, this distinction invites efforts to members of every populace to become people and, reciprocally, it temps the people to resist and seek to preserve their special status.

In the American case, the notion of who "the people" are was qualified from the beginning by the exclusion of indigenous tribes and slaves, to say nothing of poor people, women, and others who would not be enfranchised. Eventually, the notion of the people was replaced by the idea of the nation, a term that did not gain widespread currency until after the Civil War in the mid-nineteenth century. Until then, the republic was conceptualized as a constitutional system of government for the people, or a union of sovereign states. After the deep trauma of the Civil War, it became apparent that a more cohesive identity was needed and the victors drew on the ancient term, nation, to replace the increasingly ambiguous concept of the people.

#The Nation-State

At the same time, an equation was created between the nation and the state. The core idea is expressed by the term, nation state* {C2} . This phrase eases the transition from a kingdom to a democracy. In both cases the state is a focus of patriotic fervor, but in the former its identity hinges on the ruler as sovereign, whereas in the latter, the nation is sovereign.

Although this was not an easy transformation, the basic concepts seemed to be clear enough. In its ideal form, a nation consists of everyone over whom a state exercises jurisdiction and the members of this nation have the sovereign right to choose those who will represent them in governing the state. To represent this idealized notion, I use the phrase, national state* {C1}. In such an idealized state, all citizens share a common ethnic identity and everyone who has this identity is a citizen. Clearly no state in the world today is a national state -- perhaps Iceland comes nearest to approximating this idea. Israel has embraced the notion by giving all Jews the right to come to Israel as citizens. Although it denies the suffrage to Jews residing outside the state, they automatically have the right to vote if they come to Israel to cast their ballots.

Realistically speaking, none of the sovereign states of the world are truly national states. Instead, they all have multi-ethnic populations and, in varying degrees, they exercise jurisdiction over non-nationals* {C5} who cannot or will not count as members of the dominant nation. Of course, one can rationalize the phrase, "nation-state" as just a convenient way to distinguish an independent state, like the United States, from its sub-states, like Virginia and Hawaii. In this usage, nation is just a synonym for state, as it is in the name of the United Nations. Fortunately, we can use other phrases like independent state or sovereign state* {C2} to refer to polities whose sovereignty is recognized in the world system of states.

This usage enables us to use nation unambiguously, in its modern sense, to refer to any ethnic community that claims sovereignty, and we may refer to all members of a nation as its nationals. I shall distinguish two kinds of nations: dominant Nations {B5} and ethnic nations {B4}. The socio-political sense of the word used here needs to be distinguished from its special legal sense. In some states, national is defined legally to refer to non-citizens who are not aliens -- these are people under a state's jurisdiction, perhaps in conquered territories, who have limited rights but not those of full citizenship -- residents of the U.S. Virgin Islands or American Samoa are examples. I shall avoid using national in this sense. Rather, I shall always mean by National* {C3} a member of a Nation, and by national* {C6} a member of an ethnic nation.

Moreover, every such nation, by definition, either dominates a state or aspires to become a state. This criterion identifies two different kinds of nations and it is very important to be able to distinguish clearly between them. I use state nation* {C4} to refer to states that seek to become national states by converting non-nationals to nationals, or by eliminating them. Between these polar extremes, many intermediate and changing policies can be found. State policies that encourage the assimilation of non-nationals are liberal whereas those that discourage such assimilation are illiberal, and states may follow mixed policies that encourage assimilation by some while repelling others.

By contrast with state nations, we may speak of ethnic nations* {B4} as ethnic communities that claim sovereignty and statehood -- I use ethnonation as a synonym. Again, there are important differences. Ethnic nations living within enclaves* {D2} experience greater obstacles to statehood than those living in exclaves, territorially separated from the state exercising jurisdiction over them. Enclave nations may have to settle for autonomy as a form of self-rule without full independence. The distinction between state nations and ethnic nations enables us to speak more clearly about the rise of modern nationalism and its implications for ethnicity and ethnic conflict.

Most ethnic communities are not ethnic nations because they do not aspire for independence, either because they want to become members of an existing nation, or they are so dispersed and weak that they know claims of sovereignty will be hopeless. Consequently, they are not nations, as I have defined this term, although in popular usage they may be called "nations" to avoid using a pejorative. This usage, however, is consistent with contemporary scholarly practice in which a distinction between ethnicity and nationalism is made. See, for example, NATIONALISM AND ETHNIC POLITICS (the journal of the Research Committee on Politics and Ethnicity of the International Political Science Association). Indeed, we may view ethnicity as a property of communities that do not constitute the dominant nation of a state. In many, perhaps most, states there is such a Nation* {B5} -- I am capitalizing the word to distinguish dominant nations from stateless nations, the ethnic communities (nations) that aspire to attain statehood. Instead of defining ethnicity by such markers as language, religion, ancestry or race, therefore, it may be simpler and more useful to define ethnicity as any property that excludes individuals, permanently or temporarily, from membership of a Nation. They may, however, identify themselves as ethnic nations, or nations* {B4}. By underlining this word we may be able to distinguish easily between the two kinds of nation: Nations and nations.

This distinction also enables us to say that the population of any state may be divided between its Nationals and its non-Nationals. Capitalization of National permits us to distinguish this concept from nationals in a legal sense. Communities whose members do not identify themselves with a Nation are classed as ethnic, but it is important to make a distinction between two kinds of ethnic: those who are permitted to become nationals and those who are not. The term ethnic is often used more narrowly for the former sub-category and the latter, under various headings, are denied opportunities to become nationals. In traditional civilizations, where caste-like distinctions were pervasive, differences of religion, language and ancestry marked separate communities and preserved their separateness. In modern nations, however, these distinctions lost their power to the degree that they could not be invoked to prevent national assimilation. A new distinction arose to provide a barrier that could not be overcome, namely the idea of racial differences.

#Race as a Marker

The idea of race as a criterion for exclusion is relatively modern, although it can be traced to the Columbian era. Before Christopher Columbus sailed to the New World, racism was not a significant factor in European consciousness, nor did it spring up quickly during the 16^th century. There was a transitional period during which Europeans sought to explain and justify their intrusion and conquest of indigenous peoples by relegating them to some kind of sub-human category. The natives were viewed as belonging to nature and hence to be captured and handled like animals rather than to be treated with the respect that evolving concepts of international law stipulated for all "civilized" peoples. [Henningsen's paper]. Under the mantle of this mythology, all kinds of barbarities, genocide and theft were perpetrated by Europeans upon the conquered peoples, even those who, as in Mexico and Peru, had achieved levels of civilization that equaled or even surpassed those known in Europe.

When the American Constitution was promulgated, the indigenous peoples still fell under the rubric of natives, and those of African descent, as slaves, were viewed as sub-human property. Following the Civil War (1861-65), however, when American nationalism surfaced and the slaves were emancipated, these fictions lost their plausibility. In their place, the myth of racial differences flowered, and it spread to encompass all non-Caucasian communities. In addition to Red (Indians) and Black (Africans), the categories of Yellow and Brown (Asians) became popular. They provided a basis for permanently excluding some ethnic communities from membership in the Nation.

The notion of racial inferiority received some support, in the popular mind, from Darwin's theory of Evolution -- his Origin of Species appeared in 1859. Evolutionary theories were invoked to make racial distinctions credible on the premise that, since humans had evolved genetically, there must be significant differences between human races, just as there were between different animal species. Although humans could be distinguished from animals, different stages in human evolution were associated with different levels of human consciousness and, therefore, equalitarianism could be ascribed to some humans (i.e., the Whites) whereas other humans (the non-Whites) were viewed as inherently inferior and, hence, not deserving of equal treatment. The conquers were classified as "White," and non-Whites were lumped together under such racist categories as "Black," "Brown," "Yellow" and "Red." The most conspicuous symbol of American racism during the following century could be seen on public toilets in the South where "White" and "Colored" signs marked the distinction. Parallel distinctions were made in South Africa under the guise of Apartheid.

It is useful to make a clear distinction between race as a socially constructed category that has little to do with genetic differences, and racism as a mental construct that uses racial stereotypes as a basis for excluding communities from Nationhood. In the context of the creation of Nations as a way of extending equalitarian norms to the dominant community of a state, racism provided a plausible substitute for language, religion and ancestry as an inescapable criterion for excluding some minorities. Its strength is reflected in the continuing use of "race and ethnic" as a set phrase. Although no clear line is usually drawn between race and ethnicity, the important distinction arose between those who could and those who could not Nationals.

Following the removal of racial discriminations in American immigration laws during the 1920s and the success of the American Civil Rights movement culminating in the ban on racially segregated schools imposed by the U.S. Supreme Court, in 1954, in the landmark Brown vs. Board of Education decision, racism lost its formal recognition in the United States. However, simultaneously it became an important rallying cry for ethnonationalists. Well aware of the injustices their ancestors had suffered in the name of racism, they adopted a counter-racist ideology. In protest against past injustices, they proclaimed their sovereignty and demanded the right to establish their own states and create their own nations. Although racism became more of a shibboleth in the New World where new nations were created by diverse immigrant communities, the notion has also spread to Old World countries even though ancient barriers of language and religion still retain their sting as grounds for distinguishing Nationals from non-nationals.

#The Ambiguity of Ethnicity

The distinction between state Nations, and ethnic nations is polar, and leaves open a large marginal zone populated by ethnic communities whose members may or may not want to become Nationals -- and for whom assimilation to a Nation may or may not be possible.

Traditional nationalism had no such ethnic character. Modern nationalism, by contrast, is loaded with ethnic ambiguities* {B2}. The basic problem arises, I believe, because of the fluid character of the transition from caste to class orientations. Essentially, nations consist of members who have accepted equality and class mobility as norms while those they exclude are viewed in residual caste-like terms as inherently unequal and therefore condemned to social immobility. So long as ancient caste markers based on language, religion, or ancestry remained active, they provided criteria that enabled ruling communities to dominate marginalized communities. However, these markers eroded in countries where industrialization and democratization led to the creation of Nations in which equalitarian norms would prevail. Even so, a large residual category of non-Nationals remained. In the older Afro-Eurasian countries, the ancient distinctions still remained strong enough to arm the ancient cleavages reflected in many current ethnonational conflicts. However, in much of the New World, as noted above, these distinctions lost their potency and were replaced by the newer racial criteria.

Regardless of the criteria, we can visualize the transition from caste-based inequalitarianism to class-oriented equalitarian norms in terms of a spectrum that places state Nations at one extreme and ethnic nations at the other. At these extremes, they invite confrontations and also international cooperation or networking. Just as the states of the world increasingly join forces to defend existing boundaries and protect the status quo, so ethnic nations are rallying to help each other achieve their revolutionary aspirations, especially to gain independence or autonomy and to achieve the benefits of statehood. Some ethnic nations have organized themselves into a network, making use of the INTERNET: their names and information about them can be found on the Web Site of the Unrepresented Nations and Peoples Organization at: UNPO.

Most of these ethnic nations are unfamiliar to us, but a few have attracted extensive media coverage and are very well known. A conspicuous example of state Nationalism vs. ethnic nationalism can be seen in Kosovo where the Serbian Nation has attempted to unite all Serbs in one state and to oust non-Serbs from its territory. By contrast, the Slovak community has already achieved independence at the expense of its Serbian minority, and the Kosovar Albanians, after having been ruthlessly attacked, are now, with international support,, seeking to create their own national identity. Although such conflicts receive a great deal of media coverage, little effort is made to clarify the long-term dynamics of the processes that have produced them.

Between the extremes of state and ethnic nationalism, there is a wide zone of ethnic ambiguity in which three rough categories may be identified. Near the pole of ethnic nationalism we could find communities whose members reject assimilation to the Nation yet lack the cohesion and territorial foundation needed to become nations. A good example in the U.S. is the "Nation of Islam," a dispersed Black community that has elected to adopt a religious distinction to distinguish itself from the dominant Nation. One might refer to them as proto-nations* {B3}. They can be distinguished from ethnic nations by the lack of a territorial base and ancestral traditions. Instead, they are recruited on the basis of profound alienation to join a movement that promises a sense of "national" identity and a solution to their acute problems. A more familiar example can be found among the Roma (Gypsy) peoples who are scattered in many countries and have experienced persecutions and marginalization in many of them.

Toward the opposite end of the spectrum, we find ethnic communities whose members have been largely assimilated to a Nation yet continue to feel some degree of discrimination. A good example can be found in the "Americans of Japanese Ancestry" (AJA) community who insist that they are not hyphenated "Japanese-Americans" although they do want to keep many of the cultural practices of their ancestors. The same is true in less conspicuous ways of many European communities whose Americanized members retain cultural practices of their homelands. Many Scottish, Irish, German, Polish, Italian and French immigrants to the United States, for example, retain much of their cultural heritage and love to celebrate their historical, culinary, musical, and artistic achievements. For all practical purposes, however, they have fully assimilated to the American cultural scene and view themselves as patriotic citizens.

This phenomenon is sometimes discussed under the heading of ethnic tokenism, or symbolic ethnicity. It is viewed as having a folkloric interest, but no political or economic importance, perhaps because it generates little conflict and much pleasure. Actually, I think, this is an important manifestation of ethnicity and deserves to be treated more seriously. Perhaps if we used another term for it, like virtual ethnicity* {B1} , the subject would be treated more seriously. I believe it showcases a desirable outcome of ethnic diversity which not only enriches the experience and lifestyle of many Nationals, but also offers many others cultural opportunities that can make their lives more interesting and gratifying.

Between virtual ethnicity and proto-nationalism lies an intermediate and very problematical experience that lies at the core of ethnic ambiguity* {B2} . Much of the literature on ethnic diversity* {A4} focuses on such communities. Their members may or may not become assimilated to the Nation for reasons that arise on both sides of the boundary. Prejudices, harassment and barriers imposed by Nationals make them feel unwelcome, leading many ethnics to reject the ideas of assimilation or integration. Instead, they may see ethnic nationalism as a preferred solution for their problems. However, because they are dispersed geographically, and many members of these communities have successfully assimilated -- especially because of mixed marriages that enable them to claim membership in more than one community -- they are vulnerable to internal tensions and factionalisms that hamper their ability to establish clear identities and find satisfactory solutions to their ethnic dilemmas. It would be useful to have a term that could be used to identify ethnics experiencing this kind of ambiguity. I suggest ethnic marginals. This term would enable us to think of them as the most important subjects of ethnic diversity. Broadly speaking, ethnic diversity refers to all three intermediate categories: including virtual ethnicity, proto-nationalism, and also, especially, ethnic marginality.

Of course no clear lines can be drawn between these intermediate categories, and identifying three of them is arbitrary. However, it will help us to think about how constitutional arrangements can help to achieve better solutions to the problems posed by ethnicity and nationalism. Figure 1 may provide a useful way to visualize and remember this spectrum.

#A Colloid Society

We tend to think of countries, states and nations as homogeneous entities, and we reify them just as we draw boundary lines between states. We see modern societies as class-based communities where social mobility is encouraged and equalitarianism is the norm. In such a society, everyone has an opportunity to advance economically and socially and also to share sovereignty as a citizen eligible to vote and to be represented in government. By contrast, we see traditional societies as places where inequalities prevail on the basis of sacred forces under the authority of monarchs whose right to rule has supernatural legitimacy. Such societies have a caste-like social structure based on the inheritance of occupations and the prevalence of endogamy. Although we typically use Caste to refer to the ritually elaborated Hindu social system, we need to understand that its core features of occupational inheritance and endogamy are pervasive in traditional civilizations.

No doubt both of these images are exaggerated ideal types. More importantly, in today's world I think both principles are usually mixed together in the same country, though in varying proportions. Moreover, during the period of industrial imperialism, states that were democratically accountable at home ruled arbitrarily, as a master race, in conquered territories -- in these empires democratic equalitarianism may have prevailed at home while un-democratic inequalitarianism was enforced in the imperial possessions. ² More importantly now, following the demise of these empires, we see that in most state nations, while equalitarian norms flourish among Nationals, the non-nationals in these countries are subjected to caste-like subjection. Precisely because of the wide-spread acceptance of equalitarian norms, however, these conditions are experienced as far more onerous than they would have been traditionally: they generate anger and often provoke ethnic protest movements.

It might be useful to speak of any society where class-based equalitarian practices are intermixed with caste-like inequalitarian sub-structures as colloidal* {A3} . A colloid, physically speaking, is a substance distributed throughout a different substance -- metaphorically, a colloid society might be defined as one that accepts equalitarianism, class and social mobility for members of a dominant Nation but rejects these principles for folks belonging to communities that are marginalized on the basis of such criteria as their race, religion, language or ancestry. Such cleavages are understood as having an ethnic character in modern societies, but in traditional civilizations, parallel divisions are experienced between caste members who respect each other's differences, and outcaste groups whose beliefs and life-style cannot be reconciled with those of caste communities.

In a colloidal state, the dominant Nation recognizes all people as sovereign, entitled to fair treatment and equal opportunity in a class-based social system. However, it also condemns non-people to subordination and inequality. These non-people may carry the residual onus of a traditional status as outcastes, or they may be burdened by new tensions based on conquest, migration, or artificial barriers based on racist myths.

Such racial and ethnic minorities may be officially barred from membership of a the Nation or, if they are eligible, they may not yet have become Nationals. Reacting against such barriers, some marginalized communities mobilize to claim that they, in fact, are nations entitled to the privileges of self-government and sovereignty. When successful, they would become Nations and might well, in turn, impose barriers against ethnic and racial minorities under their domination. Alternatively, they may demand the full rights of citizenship and integration with the dominant Nation. Between these options, as noted above, in the limbo land of ethnic ambiguity, members of some ethnic communities vacillate between the options of Nationhood and secession.

The colloidal status is inherently unstable and subject to increasingly urgent demands for change, not only from ethnic communities but also from Nationals who have become alienated by the uncertainties that equalitarianism produces. Some members of a dominant Nation react against the insecurities posed by modernity, especially the self-doubts they experience as a result of failed expectations. The equalitarianism of a class-oriented society not only produces opportunities, but it raises hopes that are often not fulfilled. Lower caste communities feel no guilt since they can blame their adversities on Fate or Karma. By contrast, when secular, equalitarian notions prevail, people in the lower classes have to blame themselves for having failed, or look for scapegoats. During a period of transition, lower caste communities (whose members previously had no option but to accept their status) begin to demand and expect new opportunities for social mobility and increasingly protest, even resorting to violence, when these expectations are not met.

By contrast, upper caste communities respond to the new class-based ideas with fear and ambivalence. They want to protect their privileges and resist changes that open up opportunities for outsiders or lower caste groups. They may well support neo-traditionalist ideas that give them a sense of self-righteousness which compensates for their failures in the secular society, and rationalizes their scorn for ethnic and racial minorities who are viewed as nonbelievers or heretics. The neo-traditionalists fantasize that a return to the certainties and family values of the past will, somehow, create a righteous Nation that can justly marginalize non-believers and legitimize the sacred inequalities of an imagined past.

Hostility to minority communities and the desire to enforce their subordination, therefore, plays an important part in the neo-traditionalist fantasy world. Although neo-traditionalists do not think of the past in caste-like terms, in fact they yearn for a social system in which they could, proudly and without guilt, play the role of a ruling caste, lording it over lower caste communities. Consistently with the ideology of caste-based societies, neo-traditionalist Nationals invoke divine authority to legitimize their claims to superiority and their right to oppress those they view as unrighteous misbelievers.

Between these two social extremes, we find middle-caste communities whose responses to change may well be highly ambivalent. They welcome new opportunities for social mobility that may enable them to raise their own status (income, prestige and power) but they also worry that growing pressures from lower status people will lead to their own marginalization. Caught in a squeeze between these forces, they can ambivalently support fascist authoritarianism or liberalizing democratization.

As for the non-nationals, their responses to marginalization are predictably strong. Some proclaim their willingness to accept citizenship and seek membership in the Nation. Barriers to acceptance, however, provoke their anger and impatience, perhaps leading them to a defiant interest in their own ancestral heritage. The problems and inequities generated by this form of ethnic has become the focus of much of the literature on ethnic diversity. It focuses on injustices to be rectified and prejudices to be overcome as they are experienced by the ethnic marginals. The underlying vision is that those on all sides of the prevailing ethnic divisions will accept the possibility of achieving justice and equality within the framework of existing states, and the hope that non-Nationals can become integrated with the dominant Nation without having to sacrifice their own tural identity.

Other non-nationals reject such ideas as utopian and unjust because past injustices and oppression can never be conciliated. They, therefore, reject the status quo, proclaim their sovereign rights, and demands independence, or at least autonomy. The struggles led by ethnic nationalists, as reflected in the projects outlined by the unrepresented nations referred to above, are a result and they will, continue to engender conflicts in many countries during the coming years. It is helpful to use ethnic cleavage* {A5} rather than diversity to characterize this relationship. Research on ethnicity often mixes these two forms together, but I think we need to distinguish clearly between them, even though there can be strong overlaps, and many individuals are torn between the two. Under the heading of a colloid society, we may well expect both diversity and cleavages to co-exist and individuals will undoubtedly vacillate between the goals of integration and cleavage.

How members of an ethnic community see themselves is, assuredly, affected by the policies of the state where they live, and the attitudes expressed by its Nationals. The more supportive and tolerant they are, the more ethnics will be drawn toward the ideals of diversity, seeking harmony as citizens of the state nation. By contrast, the more exclusionary and hostile the disposition of the state and its Nationals, the more likely it is that ethnics will become nationalists, demanding a separate political status.

Many if not most modern states have embraced the ideal of a national state. The strategies they adopt to achieve this ideal vary from liberal and to illiberal. The former seeks to assimilate non-nationals by non-coercive means, while the latter follows harsher policies designed to root out non-nationals, including such extreme measures as the Nazi Holocaust. Between these extremes we find ambivalence and vacillation. We can also find, hopefully, signs that some states support multi-cultural policies that encourage ethnic diversity and view it as desirable. Ideally, they extend equality of opportunity to everyone and enable aliens to become citizens without compelling them to assimilate to the dominant nation.

Indeed, we may well anticipate that the idea of a Nation as the basis for legitimizing governance in a republic will pass. In its place there is emerging the idea of a multi-national state. Switzerland is perhaps the best prototype but it is more of a confederacy in which each canton is based on a kind of sub-Nation. However, such a multi-national state need not be colloidal -- it may provide equal rights and protections for all citizens even though they belong to different nations.

An alternative model that I would personally favor might be called a civic state. Instead of recognizing a congeries of nations as the basis for legitimizing representative government, one would think of all humans as people entitled to representation and justice in the modern state. Their sovereign rights would depend on their humanity rather than on membership of any nation. Such a transformation would, of course, require the elimination of all caste-based notions that marginalize the non-Nationals in a state. Instead, the distinction between Nationals and Non-Nationals would vanish. More significantly, the whole idea of nations and nationalism would become obsolete. It would be replaced by citizenship or residence, concepts that have no ethnic connotations. Individuals would have a normal place of residence that entitles them to be represented in the governance of that place -- such places already exist as villages, cities, sub-states as well as independent states. The ultimate citizenship of all humans would be vested in some kind of world order.

No doubt to the implementation of such a vision would involve the comprehensive elimination of caste-based practices and the extension of class-based concepts to a global system. That is surely a long-term futurist dream, but it is worth contemplating as an ultimate possibility. In such a world order, citizenship would depend on place of residence and willingness to be obey the law rather than ancestry, language, religion or race. Individuals could easily transfer their citizenship from one state to another because the transitional status of belonging to a nation would lose its force. Something like this may be happening today in the European Union where the old identities based on language and religion are weakening under the influence of a growing sense of European civic identity devoid of national sentiments. Parallel changes could well occur in the future, not only in regional contexts, but even in a global one where, idealistically speaking, all humans would vest their claims to equal opportunity on their status as world citizens. Collloidal mixtures as a caste/class transitional phenomenon would disappear, but social integration would not mean cultural homogenization. Instead, a world order should provide space for many different cultural communities to maintain their identity, perpetuate valued traditions and share their difference. Life should become enrichingly variable and secure. In order for that to happen, however, democratic practices that enable responsible and representative government to protect the interests of all citizens need to become established throughout the world. For that to happen, however, we need to re-think the structures of democracy ³ that can and will work to cope with serious problems, including the need to overcome ethnic conflicts. In Part II of this analysis, I will present some thoughts on these possibilities.

It has become increasingly important, for the sake of world peace and security, to help all countries find ways to handle ethnic tensions within their borders -- not only those fueled by ethnonationalism but also, and increasingly, the growing dissatisfaction of dispersed non-National minorities living scattered throughout many countries. Ethnic dissidents are suppressed by ruthlessly efficient one-party states, but in the long run such regimes aggravate the repressed discontents which become truly explosive if and when the regimes collapse. Far more often, in today's world, there are weak authoritarian states, often dominated by military groups but sometimes by elected politicians, who abuse their authority. Such regimes are unable to solve many important problems, including those posed by antagonized ethnic communities. These are the countries where ethnic nationalism and violent conflicts are most prevalent.

To deal effectively with these problems, it seems clear that responsible governments with popularly elected leaders are necessary (Riggs, 1995) . Only when those who are affected by a regime's policies have some meaningful way to influence the way they are treated can we assume that their interests will be taken into account and respected. However, the mere fact that a government is elected does not assure its responsiveness to the concerns of all those who live under its authority -- democratic government by itself does not automatically resolve ethnic conflicts. An important reason is surely the inability of some residents to vote for office holders. Since the right to vote for political leaders is normally exercised only by citizens of a state, the interests of aliens are often ignored. However, aliens often get diplomatic support from the country of which they are citizens, they are not truly helpless. By contrast, stateless persons and ethnic minorities who are classed as non-Nationals often suffer the most because of their lack of political representation. Their lot may be improved if they can become naturalized citizens -- or, more importantly, if non-Nationals can become Nationals. That is because citizens who are not Nationals often find that their votes carry little weight, a fact that raises important questions about the way democratic constitutions actually work.

These considerations lead to questions about the willingness and ability of regimes to open their doors so that excluded people can actually secure political representation and real protection for their interests. It is not enough just to write clauses into a Constitution that guarantee human and civil rights. When only Nationals are well represented in a government, and many residents in a state are not Nationals, even if they are citizens, we may expect public policies to be so biased that many ethnic communities will be alienated and some of their members will resort to violent protests and disruptive acts. To avoid these difficulties, the victims of injustice need to be brought into the political process. Different constitutional mechanisms, I believe, affect the ability and willingness of regimes to do that, and in the analysis that follows I shall discuss the influence of these mechanisms, paying particular attention to the way power is distributed among the elected office-holders of a state.

#Democracy Is Bifacial

First, however, we need to recognize the fact that modern democracies are caught in a bifacial* {D1} dilemma. Precisely because they permit political opposition and protect civil liberties, they enable discontented people, including ethnic minorities, to organize and press for reforms. Their ability to act is enhanced by modern organizational methods, leadership techniques, access to powerful weapons and enhanced information and communication technologies, including the INTERNET. Moreover, because marginalized minorities embrace the modern norms of equality and mobility based on class structures and national identities -- rejecting traditional casteism which consigned them to inescapable subordination -- they are sure to become increasingly dissatisfied with the status quo. They will embrace projects designed to promote radical changes that improve their status.

The other face of democracy opens because of its ability to identify its critics and adopt policies designed to handle their problems. Consequently, although ethnic movements can flourish in democracies, it is also possible for democratic governments to understand grievances and take steps to resolve ethnic protests by peaceful means. Because they can respond to public opinion and deal more realistically with serious problems, democracies are able, in principle, to identify the causes of ethnic unrest and achieve acceptable solutions to these problems. Thus democracies can respond to the concerns of people who may not be able to exercise the rights of citizenship. The design of a constitutional system may well affect its ability to do this. An excellent discussion of the problems of diversity and appropriate policies for managing them can be found in Inglis (1996).

The bifacialism of democracy is also reflected in the contradictory responses it provokes among its citizens. The beneficiaries of inequality, those who are rich and prominent, tend to resist changes that might undermine their high status, while the poor and unfortunate, blame others and reject their marginalization -- they organize to demand justice and better opportunities. Whereas ethnic (ascriptive) differences, based on language, religion, race or ancestry (often mixed together in various combinations) made inequalities seem inescapable and just in traditional (caste-oriented) societies, (achieved) differences in income, prestige, occupation, and places of residence are acceptable in a modern (class-based) society only when everyone, in principle at least, has an equal opportunity to acquire these benefits.

Democracies, therefore, face the challenge of making equal opportunity not just an empty slogan but, indeed, the working basis for life. They need to open the doors of opportunity for everyone and deal with the genuine disappointment of those who feel cheated, especially when their grievances are based on ascriptive characteristics, such as ethnic differences. In countries under authoritarian or arbitrary rule, the traditional foundations for sacred authority and legitimate inequality have crumbled but the modern basis for secular authority and equalitarian mobility have not been institutionalized. Under the arbitrary and oppressive rule of dictators, movements for democratization are increasingly active and it is, therefore, highly relevant and important to understand which constitutional designs will be most likely to bring peace and order.

Adding Other Variables

To assess the comparative value of different constitutional designs, we need to be clear about the differences between the ethnic problems to be solved and this compels us to take into account a number of environmental variables: historical, geographical and demographic.

The Demographic/Geographic Variable: in Figure 1, I have identified a spectrum that ranges from National identity to ethnonational resistance, with intervening categories of virtual and proto-ethnicity, and ethnic ambiguity. These attitudinal differences are profoundly affected by one's place of residence. Those who live widely dispersed among members of different ethnic communities are more likely to become citizens and accept a National identity, whereas those who live concentrated in enclaves that have clear boundaries can more easily reinforce their separate identity and regard themselves as a distinct ethnic nation. Between these polar types, however, most ethnic communities find themselves in intermediate situations.

Sometimes, members of an ethnic community are grouped in scattered pockets like villages, urban ghettoes, or zones that lack clear boundaries but do enable members of an ethnic community to mingle most of the time with their own fellow-ethnics. Mattei Dogan (1997, p.73) uses containment to speak of this type of situation. I suspect that the emergence of ethnic ambiguity and proto-nationalism is likely to be associated with demographic containment. By contrast, Dogan refers to ethnic intermingling which occurs, he writes, when smaller ethnic pockets occur well mixed with pockets of a different ethnicity. We might speculate that virtual ethnicity and ethnic ambiguity are more likely to arise under such conditions.

Although we need to visualize these possibilities, no clear lines can be drawn between them. Moreover, continuous population movements and cultural changes make them quite fluid. My own experiences of urban ethnicity in America suggest the co-existence of different patterns among members of the same ethnic community. For example, many Chinese Americans live and work in a "Chinatown" district whose shops and restaurants attract many non-Chinese visitors, but most Chinese in America live widely dispersed throughout the whole population. My guess is that all patterns of ethnic containment, intermingling and dispersion can be subsumed, for political analysis, under the heading of ethnic diversity.

It is only in the more extreme cases, where substantial enclaves exist and self-identification as an ethnic nation can thrive, that one can think realistically of ethnic cleavages. A classic example occurred in former Czechoslovakia where the large and homogeneous Slovak region was able to gain a non-violent agreement for partition. However, as the cases of Kosovo and East Timor have recently shown us, the resistance of citizens attached to the surrounding country can make partition extremely difficult. When a secessionist nation constitutes the majority in a federal republic, it can break away more easily, but there may be a high price because of the ethnic diversity that will prevail in the new state -- Croatia and Bosnia-Herzegovina provide good examples.

The Historical Variable: differences in the sequence of events or the causes that have generated multi-ethnic societies also need to be considered. In principle, ethnicity always results from migration as one community comes into contact with others. Massive flights and invasions have taken place for hundreds, even thousands, of years and these events, like converging tectonic plates, have generated far-reaching socio-political consequences.

In modern times, the most obvious case involves the movement of European settlers into the New Worlds (the Americas, Australia and New Zealand). They displaced indigenous peoples, appropriating their lands and transforming their life-styles. These displaced peoples have not only been denied admission to the status of Nationals, but they have rejected assimilation and, increasingly, demanded autonomy as sovereign peoples. From this process ethnic cleavages have evolved.

The migrating Europeans were not culturally homogeneous. When settlers came to power they created new Nations with a dominant cultural character. However, they were more or less willing to accept other Europeans as members of a dominant Nation. In the process, however, cultural distinctions remained and often became a basis for privileging some and marginalizing others -- hence the pheonomenon of ethnic diversity.

Other Europeans migrated as merchants and conquerors, creating modern empires and a global economy. Even after de-colonization, some of them remain as expatriates in foreign lands. More importantly, the imperial dynamics they launched led to many more migrations of workers, traders, and functionaries coming from China, India, the Middle East and many other places. They have become ethnic pariahs, persons without a homeland.

Although typically marginalized politically (though not economically) it is true that in some countries they have become the dominant majority and in others they have become affluent minorities. Lacking indigenous roots nor a home enclave, however, they are usually unable to assimilate to a Nation nor able to create their own nation -- they survive by various expedients ranging from revolutionary violence to corrupting collaboration.

Some have been displaced and become international refugees, while others have been able to assimilate and become fully accepted citizens in dominant Nations. J. S. Furnival has well described this phenomenon under the term, plural society but unfortunately this phrase has been stretched to refer to any kind of ethnic diversity, or even interest-group pluralism. We need a more precise term to identify a highly variable but globally important ethnic phenomenon. I have suggested poly-communal society.

This term also points to the retention of tribal enclaves in some of the new states born from the collapsed empires. In such societies it may be very difficult to generate a new Nation: instead, one of the tribal communities is likely to dominate the state and other tribes predictably resist assimilation, asserting their own ethnonational claims to statehood. They become commingled, to use Dogan's term. In such a context the ethnic pariahs, often intermingled (in Dogan's sense) among all the tribes, become a convenient scapegoat -- victims that members of all the indigenous communities can easily persecute -- though not with impunity, because they themselves suffer if they destroy these guest workers or merchants whose services have become economically indispensable. To solve such problems, some form of power sharing among the elites of rival communities seems to be a necessary condition for the avoidance of anarchic fragmentation.

Recognizing the important differences between types of modern ethnicity paves the way for a consideration of how different constitutional regimes can handle the problems they create for themselves and for others. Most importantly, I think, parliamentary regimes are more likely to find acceptable solutions for these problems than presidentialist regimes. A general discussion of the relative problems and merits of these two types of constitutional system can be found in (Linz, 1990 and 1994). I shall not review these structural differences here. Let me say, however, that I use presidentialism* {D4} to identify a constitutional system based on the separation-of-powers principle. The propositions offered here are largely based on theoretical considerations rather than empirical data, although the readily available examples strongly suggest that parliamentary systems are inherently more capable of resolving ethnic conflicts than presidentialist regimes. I believe that case studies and statistical data are both needed to test the validity or limitations of this theory.

Let me focus first on the problems generated by ethnic diversity. To cope non-violently with these problems, democratic states need to adopt the most appropriate governmental structures. As noted above, I am persuaded, contrary to much conventional thinking, that parliamentarism is more likely to facilitate the accommodation of ethnic demands than presidentialism.

The easiest way to accommodate the concerns of ethnic minorities whose members see themselves as citizens is to permit them to organize politically and secure representation in a legislative assembly. This is the normal situation involving ethnic diversity. In parliamentary regimes, executive power is exercised by a cabinet that is continuously subject to review by parliament and can, therefore, be expected to be sensitive to the needs of minorities, especially those that have parliamentary representation.

By contrast, in presidential systems, the executive power is exercised by one person, a president elected independently of the congress and not vulnerable to discharge by a parliamentary vote of no confidence. No doubt, during electoral campaigns candidates for the presidency often seek support from minority communities and promise to adopt policies favorable to their interests. However, once in power, it is difficult to sustain this sensitivity and presidents are likely, I think, to be more interested in major issues that concern the majority communities, especially the most powerful interests whose financial support they need in order to win elections.

A widely held myth holds that a president, as head of state, serves a unifying function, representing the polity as a whole, not just factions or parties. Even if it were possible for presidents to commingle reliably the ceremonial and executive functions, the fact that non-nationals are effectively excluded from the electoral process means that they cannot perform this function as well as hereditary monarchs ⁴ A striking example is the King (or Queen) of England whose unifying role for the Crown Fiefdoms and Dominions of the British Commonwealth survives -- it is inconceivable that an elected "president" could ever match this achievement. Another concrete example occurred in Japan during its postwar Occupation: General Douglas MacArthur and his staff recognized that, despite their strong hostility to the Emperor, he could not be replaced by an elected president without truly disruptive consequences.

No doubt in presidentialist democracies, it is possible for minority communities to organize themselves as interest groups, to lobby elected representatives and make political contributions. However, impoverished minorities can scarcely expect to win the same benefits for themselves that having their own elected representatives could provide. Moreover, even when minority members are elected to a legislature, they are likely to find that their concerns and demands are ignored. Without having the kind of voice in the executive branch that coalition cabinets provide in parliamentary regimes, minority members of the legislature in presidentialist systems are likely to become frustrated and embittered. By contrast, parliaments are much more able both to represent minority interests and to serve as the unifying symbol of a state.

Proportional Representation

Ethnic minorities in all countries encounter serious obstacles to success in their efforts to secure political representation, especially when only one candidate per district can be elected. Even when the major political parties earnestly seek minority votes and promise benefits in return, they are typically unwilling to nominate members of small ethnic communities to stand for election. Understandably, they expect members of the majority community to be more likely to succeed, especially in single-member districts. ⁵ Moreover, successful candidates necessarily give priority to the interests of the majorities whose support made their victories possible -- this gives them little incentive to pay much heed to the claims of minority communities.

No doubt the outcome can be different when one of the various forms of proportional representation (PR) is used. There are many ways to manage PR systems, and some are more conducive to desired outcomes than others, but all such schemes offer minorities the possibility of winning some political representation, especially when electoral districts have a substantial number of seats to fill.

Of course, this is not an automatic outcome. If members of a community are divided internally and vote for rival candidates, their ability to gain representation can be seriously jeopardized. Moreover, when ethnic solidarity is weak because members of a community give higher priority to non-ethnic criteria, they also lose power. Demographics also affects the outcome -- for example, if a community numbers only ten percent of a district's population, it has little chance of winning a seat if the district can elect less than ten representatives. This liability can be overcome in part if two or more communities form a coalition that can secure a large enough vote to elect a member, or an ethnic party broadens its appeal to attract support from members of other ethnic communities. Such possibilities may favor conciliatory candidates or parties willing to make concessions in order to broaden their electoral base, but the result may also be that central concerns of an ethnic minority have to be compromised.

An unintended benefit of PR may arise from the incentives it gives small ethnic communities to form alliances with other minority groups. When members of these groups are not citizens but are eligible for naturalization, an activist minority may help them seek citizenship in the hope that they will then be able to achieve a large enough vote to secure representation in the legislature. Thus it seems reasonable to assume that PR will not only increase the political power of minority communities but also enlarge the body of citizens recruited from other minorities.

The Constitutional Context

For proportional representation to provide an effective means for empowering ethnic minorities, however, we need to take into account the dynamics of the constitutional system. Clearly, I think, parliamentary regimes can benefit from electoral proportionality. This applies especially to states that do not follow the Westminster model in which single-member districts prevail, leading normally to a two-party system. Although third parties may play an important role in such systems, they normally require support from a large percentage of the voters, thereby weakening the voice of dispersed ethnic minorities.

In multi-party parliamentary states, however, where no single party can control a majority of votes in parliament, coalition governments are normal. In such governing coalitions, minority parties often hold seats. Because their continuing support is necessary in order for a cabinet to stay in power, even a small minority party can exercise substantial influence. Of course, this by no means gives any ethnic minority the assurance that its special needs and concerns will be satisfied. However, it does provide political leverage and it can lead members of an ethnic community to feel safe and to gain support in the making of public policy. The prospects for power-sharing or consociationalism are much greater in parliamentary regimes with multi-member electoral districts and proportional representation than they are in those with single-member districts and two-party systems. An additional distinction may be made between power-sharing systems on the basis of the degree to which minority groups exercise a veto power. In some usages, consociationalism involves a strong form of power-sharing in which veto rules apply -- one manifestation of this principle can be found in the U. S. Senate where individuals strongly opposed to a measure have the right to delay action by a filibuster that, in effect, can block passage -- of course, the most familiar example of how the veto power can be exercised is in the United Nations Security Council. These distinctions raise questions that I cannot discuss here. Moreover, these are just theoretical arguments, but they do provide a basis for expecting that, in parliamentary regimes, many of the problems generated by ethnic diversity can be solved through the normal processes of democratic government.

It is more difficult, I believe, to achieve such solutions in democracies governed by a separation-of-powers constitution. That is because proportional representation is likely to be counter-productive in such systems, and I doubt that real power-sharing or consociationalism is possible under presidentialist rules. In these systems, the president chooses a cabinet composed of non-members of the assembly, and minority members in a congress are rarely included in any Government. Presidents must enhance their prospects for success by nominating cabinet members likely to win support from the dominant majorities in congress. Because they are not themselves members of the legislature, they need to rely on persuasion and external pressures to win legislative support. Consequently, representatives of permanent minorities in a congress may find that all they can do is to protest and vote "no". By doing so, however, they weaken the ability of a congress to reach agreements. Moreover, by increasing the size of the mainstream majority that is required in order to enact a law, they undermine the legislative power in its relations with the president. Consequently, having representatives of a permanent minority in the congress increases the risk of executive/legislative gridlocks. Although this might lead to almost permanent deadlocks, my guess is that more often the risks of such a problem lead presidents, with strong military and elite support, to enance their powers and ability to dominate the congress, often by means of shrewd exercise of patronage or through a hegemonic party.

The validity of these propositions can be tested empirically since there are quite a few presidentialist regimes, especially in Latin America, where various PR ⁶ systems have been adopted. In all of them, I believe, catastrophic regime breakdowns have occurred, leading to authoritarian rule, a result that can scarcely be advantageous for minority communities. Although single-member districts and two-party systems by no means assure democracy in a separation-of-powers regime, I believe the data will show that success in such regimes is much more likely when there are two major parties.

However, even a two-party system may fail if it is highly polarized -- which is to say, if each party has extremist goals that are incompatible with the aspirations of its political rival. The prospects for success in two-party systems increase when centripetal forces are strong, leading each party to move toward the political center, seeking a middle road. When both parties agree on their main goals and differ only about policies that concern instrumental means, they may be willing to accept bi-partisan compromises. The likelihood that a two-party system will be centripetal is enhanced when there are no deep splits between rival groups in society -- in short, when profound social conflicts based on class or ethnicity do not exist.

The American Exception

The success of presidentialist regimes depends, therefore, on the existence of considerable ethnic homogeneity. The American experience may be mentioned as an apparent exception. Despite ethnic diversity, most immigrants experience Americanization and, indeed, have sought to join the mainstream of middle-America as soon as possible. The caste to class transformation that led, historically, to the definition of the American Nation as composed of Whites at the expense of non-Whites was, I think, reinforced by the political reality that the viability of the separation-of-powers constitution hinged on the centripetal forces that pulled the two major parties toward the political center. I do not know if members of Congress were aware of this dynamic, but resistance to the acceptance of racial minorities as Americans surely facilitated the survival of a highly precarious constitutional system. In virtually all other countries where the presidentialist system has been established, constitutional break-downs have occurred, sometimes quite often. I attribute this to the fact that where PR was used, a multiplicity of parties truly hampered the ability of Executive-Legislative agreements to be reached, and even in the countries with only two parties, the centrifugal forces were so strong that agreements within the Congress itself were also seriously obstructed.

A related factor arises from the tendency of many eligible citizens in the U.S. not to vote. Although this alarms idealists who think a high turnout at the polls is necessary for democracy to work, it may actually contribute to the survival of the American system. It means that those who do vote are "middle Americans," they are middle class people who share most of the values of "Americanism." In other presidentialist regimes electoral turnout has often been much greater, sometimes enhanced by compulsory voting. The result, however, has been greater polarization of politics and increased centrifugalization. Although this has increased representation for minorities and the poor, it has hampered the ability of Congresses to make decisions and provoked more severe inter-branch conflicts.

By contrast, parliamentary regimes can more easily accommodate themselves to ethnic heterogeneity. This is especially true where multi-party systems prevail, provided they are not excessively polarized. Having a moderate number of parties as participants in coalition governments provides an opportunity for several kinds of cleavage to co-exist, including significant ethnic differences. This means that parliamentary systems are able to use proportional representation and multi-party districts in a way that enhances the prospects for survival of democracy and does not undermine it, as PR does in a presidentialist regime. Statistical evidence for the ability of parliamentary systems to outlast presidentialist regimes is provided by Przeworski (1996)⁷

If this is true -- and more research is needed to establish its validity -- then we may conclude that presidentialist regimes are more vulnerable to the controversies due to ethnic diversity than parliamentarist regimes, and also less able to manage them effectively. It follows that new states with multi-ethnic populations that want to establish viable democratic governance are more likely to succeed if they adopt parliamentary rather than presidentialist institutions. However, if this choice is foreclosed because a state is already committed to presidentialism, then it needs to pay special attention to the rules and institutions that can handle ethnic controversies under more unfavorable conditions. To understand this relationship we need to look at some of the most important variables, including those that produce unexpected results.

Some clarification of these conditions can be obtained by a deeper analysis of the way the American presidentialist system has worked. On the basis of the arguments offered above, apparent success of the United States seems hard to explain. Why, indeed, has the U.S. system been able to succeed not only in the sense of surviving for over two centuries, but also in its capacity to handle continuing, perhaps even growing, problems caused by ethnic diversity? I believe there are plausible and specific reasons but to discuss them here would require more space than is available. However, in an earlier essay, I have spelled out some of these unique circumstances that make the American experience understandable. Riggs, 1994a

The other exceptional case is that of the United Kingdom where, under the Westminster model, a two-party system has prevailed. Undoubtedly the Westminster model, as adapted to the circumstances of many countries, both inside and outside the Commonwealth, that have been deeply influenced by the British experience, also needs to be critically examined, but I cannot do that here. However, there is a substantial literature that looks at the way the Westminster model has evolved outside the UK that could be studied to test prepositions about why this model has worked as well as it has in one country and what revisions in that model were adopted to enable it to work better in deeply divided multi-ethnic societies. However, it goes beyond the scope of this paper to explore this question here.

Mobilizing ethnonational communities in many countries are, increasingly, demanding the right to secede and govern themselves. Conflicts provoked by ethnic cleavages differ so much from those due to ethnic diversity that they call for a radically different response. The latter involves finding ways to bring ethnic communities into the main stream of democratic politics, whereas the former requires a state to release control over a community and giving it the right to govern itself. The difference between releasing and embracing a community has radical implications for the structure of a democratic polity.

Moreover, I'm persuaded that it is easier for parliamentary systems to do both, but the case is not so clear-cut for handling ethnic cleavages as it is for responding to ethnic diversity. The democratic response to ethnic diversity, as shown above, involves the application or rules, like those for proportional representation, that empower ethnic minorities within the established political system. It is much easier for a parliamentary democracy to do that than it is for a regime based on the separation-of-powers. By contrast, the problem facing a regime confronting revolts based on ethnic nationalism involves letting go, permitting the community to govern itself, even though that may involve making decisions that contravene established principles of the host country.

I say host country here because, to the extent that ethnonationalists occupy enclaves located within the boundaries of a state, they resemble guests whose activities are bound to affect their hosts. It differs significantly from the situation of nationalists living in exclaves outside the boundaries of a country -- for them to become independent creates far fewer problems for the state from which they are separated, and they are able to manage their own defense and foreign policy. By contrast an autonomous enclave (I shall call it an autonomy) can cause serious problems for its host country when it tries to make its own defense and foreign policy decisions. That is why, in most cases, the best solution involves autonomy rather than independence.

Advantages of Autonomy. Establishing an autonomous enclave within a state can actually be quite advantageous. It can save money and, at least potentially, prevent the violence, costly litigation, and bitter controversies that accompany ethnonational movements. An autonomy can administer its own domestic affairs and make all the important decisions that affect the life of its people, something that can significantly reduce the administrative and legal costs that are otherwise borne by the state.

Autonomy may also reduce costs attributable to the increasingly global character of ethnic communities. since the rise of the INTERNET, be actively involved in global communications. Its citizens may travel freely and be able to trade internationally by virtue of agreements covering freedom of movement and trade, and of course they can become active in polital movements that cross state lines -- as seen in the Unrepresented Nations and Peoples Organization UNPO, mentioned above. Allies of each ethnic nation located abroad may also be able to bring pressures to bear on a state accused of abusing one of its minority communities. Permitting ethnic autonomies to govern themselves may well mitigate these pressures and, again, be advantageous for the host country.

Some Examples. Democratic regimes differ, historically, in their willingness and ability to accept autonomous regimes within their territorial boundaries. Looking at some leading examples, it strikes me that the successful cases involve parliamentary systems, while continuing disputes persist in presidentialist regimes. Among the parliamentary regimes, consider the following cases:

An early example is the North America Act of 1867 whereby the British Parliament granted dominion status and full self-government to Canada, and this example was followed when Australia, New Zealand and the Union of South Africa also received dominion status with full self-government. In Finland, the Swedish-speaking Aland islands enjoyed expanding autonomy, culminating in 1988 in majoritarian parliamentary self-rule. The Danish government granted autonomy to Greenland in 1979, and the Faroe Islands have also become self-governing -- both of these island territories also have their own languages. Admittedly these decisions involved exclaves rather than enclaves, but they were all accomplished without violence. No doubt other imperial possessions had to fight for their independence, but these examples who that it was possible for a parliamentary regime to surrender power to a dependency without violence.

As for enclave nations, we may point out that, in Canada itself, in 1991, Parliament approved a law granting the Inuit peoples self-government, subject to acceptance of the plan by the residents of the new Nunavut Territory. They endorsed the arrangements by a referendum held the following year. In Spain, autonomy statutes have been approved, following regional referenda, in 19 regions, including notably the Basque and Catalan provinces. Each has its own elected assembly which chooses a regional president.

Belgium has created three self-governing regions to accommodate rival ethnonational aspirations in Dutch-speaking Flanders, French-speaking Wallonia, and bilingual Brussels. In Scandinavia, Sweden has established self-government for its Saami community, and Norway has a Saami assembly for its Laplander population. The leading example of an East European democracy confronting ethnonational claims for independence was Czechoslovakia which authorized partition of the country in 1993.

Viewed in terms of arrangements for autonomy within a country, the British case is exceptionally interesting and complex. Most surprisingly, the crown fiefdoms of the Channel Islands and Isle of Man have self-government outside the jurisdiction of Parliament, under the jurisdiction of the sovereign. As fiefdoms, their status is old and long antedates the rise of modernity. Ten or so small island colonies remain in the British empire, and virtually all of them have self-government. More importantly, the major components of Great Britain itself, have moved toward autonomy, with the Scottish Parliament. Established by the Scotland Bill, December 1997 and the next year Parliament passed the Government of Wales Act 1998, which established the National Assembly for Wales. An apparent exception might be Northern Ireland where inter-communal strife has prevailed for many years. However, in this case the obstacles to peace have arisen at the local level between contending parties within Ulster -- both the Irish and British regimes have agreed to a compromise settlement which appears now to be starting to work.

Presidentialist Obstacles. By contrast, presidentialist regimes have found it much more difficult to accept arrangements which grant autonomy to ethnonational enclaves. Although many indigenous peoples live in Latin America and a variety of more or less extreme nationalist and guerrilla movements have risen among them, it has been extremely rare to find cases where autonomy has been authorized. In Mexico, the Zapatista movement, in the state of Chiapas, has waged a continuing struggle with the regime since 1994. Despite a number of cease-fire agreements, a stable resolution of this conflict has not been reached.

In the Philippines, the most notable example of ethnonational movements can be found among the Moros in Mindanao. A less known struggle is taking place in the Mountain Province where the Cordillera Peoples Alliance seeks to unite the Igorot people to fight for self-rule. In all these cases, however, it seems to be extremely difficult to reach agreements that legitimize any kind of autonomy for these minority nationalities.

In Russia, problems involving the Chechen demand for independence or full autonomy remain unresolved. The presidentialist regime in Georgia has been unable to reach agreements for autonomy in the separatist regions of Abkhazia and South Ossetia. In Romania, the Democratic Alliance of Hungarians (DAHR) seeks to achieve local government, cultural and territorial autonomy and the right to self-determination, but without success in this presidentialist regime. However, the Hungarian Democratic Union was able to gain representation in the government.

The American Exception Again. By contrast, in the United States many "first nations" have been recognized and exercise autonomy on the basis of peace treaties entered into after they were defeated in war or made peace to avoid war. They administer self-governing enclaves within the boundaries of the United States. However, the fact that their status depends on treaties provides a telling clue to explain why these arrangements are exceptional in presidentialist regimes.

It would, I think, have been very difficult to secure Congressional support for legislation authorizing the creation of autonomous enclaves within the U.S. The pattern of colonization of the country in which many wars were fought with different indigenous communities led, at various times, to treaties in which the sovereignty of the tribes was recognized and they were allotted tracts of land.

Without pretending that these settlements were just or that the resulting pattern of self-government within "reservations" is satisfactory to the communities concerned, it is true that they provided some degree of autonomy and self-government. Moreover, a point that is usually ignored is the constitutional basis for these arrangements. Self-government by indigenous peoples in America was made possible by the constitutional provision, Article VI, #2, which recognized treaties, made by the President with the support of a two-thirds Senate vote, as part of the country's "supreme law."

A decisive historical preconditions was the recognition of the sovereignty of natives tribes before the treaties were signed. Of course, communities whose sovereignty was not recognized could not make treaties and therefore did not benefit from this practice. As a contemporary example, consider the situation of the Hawaiian community, peoples who are descendants of a monarchy that had international recognition during the 19^th century. It was overthrown in 1893 by a coup led by settlers who, five years later, were able to persuade the U.S. government to annex the islands.

Many Hawaiians today are active in a sovereignty movement that has not yet gained official recognition. The fact that non-Hawaiians now constitute a majority of the population in the islands, and the complexity of the presidentialist system, both conspire to keep the U.S. government from approving a form of autonomy that would, I think, have been quite feasible under a parliamentary regime. The Canadian decision to establish Nunavut may be mentioned as an example. The State of Hawaii has moved to recognize Hawaiian claims by establishing an Office for Hawaiian affairs that has significant influence and resources -- it's Board of Trustees is elected by citizens having Hawaiian ancestry -- for details see: OHA . Ironically, a non-Hawaiian is challenging before the Supreme Court -- Rice vs. Cayetano -- the constitutionality of restricting the right to vote for members of the OHA to ethnic Hawaiians -- see: Hawaiians Thus an agency that offers some kind of limited self-administration to an ethnonation in the U.S. is now under attack in the name of equalitarianism.

On its face, the cases mentioned above provide empirical evidence that parliamentary regimes have been more able than presidentialist ones to offer autonomy to ethnonational enclaves. Is this just an historical accident, or are there logical reasons for this difference? I believe that there are and will mention several possibilities: the constitutional consent formula, the locus of decision-making, the prospects of confederalism and the bureaucratic factor. Let me take up each in turn.

The Constitutional Consent Formula

To exempt any part of a country's domain from the normal political processes of any democratic polity requires some kind of constitutional decision. This is not just a matter of making a law, but involves enabling those living within an enclave to be immune from the laws of the land, and to create their own special rules. In a presidentialist regime, this involves gaining support for such an exception not only from the chief executive but also from the legislature, the supreme court, and perhaps from the population at large by means of a constitutional amendment. These are high hurdles -- the establishment of zones of autonomy for indigenous communities in the U.S. was possible, I believe, only because such decisions could by-pass normal political procedures because they were covered by the special status accorded to treaties.

In parliamentary regimes, by contrast, parliaments have the ultimate authority to make decisions, including fundamental laws that have constitutional significance. Although there may be restrictions on how such fundamental laws can be enacted, the fact is that it is normally much easier to achieve basic constitutional changes in a parliamentary system than in a regime based on the separation-of powers.

The Locus of Decision-Making

In presidentialist regimes the separation of powers empowers the legislative branch to micromanage public policy making. By contrast, under parliamentary rule, the fusion of powers results in the exercise of these powers by a cabinet, subject to parliamentary approval. While holding the executive accountable, this process also means that it is possible to process a large number of decisions in a coordinated way, linking powerful senior career officials with politically responsible cabinet members in a way that can place the national interest above the particularistic interests of locally-elected legislators.

With reference to decisions affecting the establishment of self-governing enclaves, it seems clear that under presidentialist rules, members of congress from districts or provinces that might be adversely affected by such decisions will be attracted to congressional committees authorized to evaluate them, whereas, of course, ethnonational minorities will be unable or unwilling to elect representatives to a congress to participate in such committees. The presidentialist system, therefore, is structurally driven to give the decision-making power affecting the fate of unrepresented ethnic nations to committees that are, inherently, biased against their interests. Most legislative committee members will oppose self- government for indigenous peoples because they abridge the opportunities for patronage that would otherwise be available to them.

By contrast, in parliamentary systems, a cabinet supported by generalist bureaucrats is more likely to adopt a National rather than a purely local perspective, and it may even see that it is advantageous for the Nation to allow non-Nationals to govern themselves. In exchange for the surrender of direct authority over an enclave, the regime avoids the political costs of perpetual conflict with its people and reduces the administrative expenses involved in efforts to enforce unacceptable laws on a rebellious minority. Moreover, the surrender of direct rule does not necessarily mean the loss of influence. The people living in any enclave will surely want to engage in external relations involving trade, travel, and communication and they will need help in many ways from their host country. In exchange for this external assistance the host state may legitimately make reciprocal demands that promote harmony between the two jurisdictions.

The Prospects of Confederalism

Although the distinction between federalism and confederalism has been blurred in popular usage, an important point needs to be stressed here. In a confederal system, political units that retain their autonomy unite in order to achieve shared go