Document Details
Conference Paper
Publication Year:
Subtheme Paper: Autonomy and Diversity

Subtheme Paper
Autonomy and Diversity

Johanne Poirier


While closely related to collective rights, autonomy in a diverse society is about empowerment. It relates to the capacity to adopt one’s own (“auto”) norms or laws (“nomos”). This paper offers a brief reflection on the meaning of autonomy and on the groups which may legitimately and realistically claim forms of institutional autonomy. It then explores how the concept of “autonomy” can be translated into various forms of institutional design in contemporary states “blessed” with the asset of diversity. Advantages and disadvantages of various options are canvassed. The importance of trust-building mechanisms is also emphasised as a counter-balance for the segmentation which can result from otherwise legitimate schemes meant to promote the autonomy of certain culturally distinct groups.

1. Introduction

In the enlightening introductory paper to the theme of “Building on and Accommodating Diversities”, we are reminded that “[d]iversities are not to be considered as a burden but as an asset that states can build upon”. While this very conviction is not—by any stretch of the imagination—universally shared, it is fortunately gaining in legitimacy around the world, in academia as well as in political discourse and practice. Translating this moral and political imperative into specific practices and institutions is a major challenge, however. As they say: “the proof is in the pudding”. And the same pudding cannot be served to everyone, on every occasion.

This subtheme paper explores how the concept of “autonomy” can be translated into various forms of institutional design in contemporary states “blessed” with the asset of diversity. It offers a few recipes for the pudding. The objective is to encourage participants to confirm, challenge, test, and complete some contemporary theories concerning the accommodation of diversity through the recognition of autonomy.

Before turning to the “plumbing” of institutional autonomy (section 4), however, the concept of autonomy is explored (section 2), as well as the question of who is entitled to institutional autonomy (section 3). The final section offers some concluding thoughts on the importance of trust-building mechanisms, including the effective participation of groups who enjoy degrees of normative autonomy, in central institutions (section 5).

2. What is Autonomy?

Over recent decades, through a limited number of binding norms, and a greater range of soft law instruments, international law has gradually widened the classic concept of self-determination to incorporate that of internal self-determination. The former was traditionally understood as the right of colonized or severely oppressed people to create an independent nation state. The latter includes measures of independent control over areas deemed to be crucial to the group’s well-being within the existing state. Similarly, some constitutions recognize, either explicitly or through unwritten constitutional principles that the protection of minorities—and aboriginal populations—can entail a certain degree of organizational autonomy, that is “self-management” or “self-government”.

The concept of autonomy is closely connected to the notion of collective rights, although they are not coterminous. While individual rights are held by all individuals simply by virtue of being human, collective rights refer to the rights of cohesive groups to have their collective identity and interests protected. The exact reach of collective rights is controversial, and will depend on the collectivity’s particular circumstances and standing in the state. They do not necessarily entail a form of self-government. Let us presume, for instance, that a group that has been historically disadvantaged obtains redress in the form of various quotas (as is the case of the “scheduled castes” in India). This arguably is a group right. However, this does not inevitably translate into institutional autonomy.

The etymology of the word “autonomy” underlines the importance that groups (or individuals, for that matter) should be able to adopt their own (“auto”) norms (“nomos”). Autonomy is not about having rights—though this is crucially important. Autonomy is about having some space (both physical and political) in which to exercise the right to decide about one’s rights. Autonomy is about empowerment. The paradigm shifts from “rights” to “power”.

3. Who is Entitled to Autonomy?

Some well-accepted contemporary theories, notably developed by political philosopher Will Kymlicka, distinguish between “old” or “national minorities” on the one hand, and more recent ones, on the other. The first category refers to groups which have enjoyed a long-standing history in a particular country, including a degree of political autonomy, prior to their forceful or voluntary annexation within the larger entity that transformed them into minorities or largely disenfranchised aboriginal peoples. Groups belonging to this category often wish to maintain specific and distinct institutions to protect their specificity.

By contrast, these theories postulate that groups belonging to the second category, mostly made up of recent immigrants, seek to integrate in their host country. This alleged preference, combined with that country’s legitimate concern with maintaining a certain degree of cultural and linguistic stability, means that “new minorities” will be entitled to the protection offered by enshrined fundamental rights (notably, freedom of expression, of religion, the right to equality and non-discrimination). It is assumed that their wish is to fit in and that consequently, they will not claim institutional autonomy. For its part, the host state has no moral or legal imperative to alter its structure so as to grant self-government to such communities.

This dichotomous distinction between “national minorities” and other minority groups does bring clarity to the debate, but it leaves a number of grey zones. It probably applies more adequately to “countries of immigration”, rather than those in which recent diversity results notably from the incoming of international refugees or internally displaced persons, or the arbitrariness of some international border drawing. Adapting this typology requires substantial empirical and contextualized analysis. It is by reflecting on some hard cases that our collective can best be enriched.

As a way of fostering comparative deliberation, let us consider in some detail a particular case which is difficult to classify according to the “either/or” scheme. A million French-speaking Canadians live outside Québec, the sole constitutive unit in Canada in which French-speakers form a majority and thus command a provincial state apparatus through the simple application of majority rule. The other French Canadians are scattered throughout Canada, though more concentrated in some areas than in others. They constitute minorities in each of the other provinces (only one of which, New Brunswick, is officially bilingual). These French-Canadians have a long-standing history in the country. They are not immigrants (except in the extended sense that all people of European descent in the New World could be considered immigrants). Few of the French-speakers outside Québec—if any—ever enjoyed a degree of prior political autonomy before becoming minorities in English-speaking provinces.

Over the last forty years, the linguistic rights of this population have been greatly strengthened (e.g. rights to schooling, school management, bilingual services, etc.). However, apart from the right to manage the minority language school system, they do not enjoy special rights of representation in federal or provincial institutions, nor do they enjoy autonomous institutions. They are rights holders, but they have very little control or input on the definition of these rights, or the resources which are to be allocated to their protection, for instance. It is not clear on which side of the fence they would fall according to the binary approach. Are they entitled to autonomy —including measures of self-government? And if not, why not?

In the end, resolving the issue of who is entitled to autonomy may come down to a mixture of corrective justice for past injustice

(e.g. previous attempts at assimilation, oppression, acculturation,exclusion from one’s homeland in the case of many aboriginal peoples, etc.) as well as realpolitik. Those who can claim autonomy, are those who either pose a substantial threat to the survival of the overall state by being in a position to consider secession, or groups which may not have that capacity, but whose vitality is considered paramount by the state for some ideological or political factor. They can mobilize to negotiate their claims with the state. Hence, the “French factor” clearly partakes of Canada’s very identity, and of its distinctiveness from the United States. In that context, it is arguable that French-Canadians outside Québec could claim a form of autonomy, which immigrant groups, even long-standing ones, cannot, both in terms of corrective justice and by reason of their political weight in the context of Canadian society.

Iris Young has challenged Kymlicka’s dual classification and posits a continuum on which various minority groups will be positioned, not necessarily in a stable and permanent fashion. She gives the example of African Americans who clearly did not come voluntarily to North America, but who, as a rule do not seek autonomous institutions. Unfortunately, what the continuum concept gains in diagnostic potential, it loses in terms of the ability to identify distinct solutions (e.g. temporary measures to promote substantive equality vs. permanent institutional design to allow for the maintenance of difference). To return to our previous example, it may be that French-Canadians outside Québec fall somewhere between “national minorities” and “immigrants”. But that does not answer the question of their entitlements: individual and collective rights solely? Or measures of self-government?

As a way of partly filling the continuum, let us consider a multiple pronged typology (adapted from those elaborated by Paul Chartrand and Alain Cairns). In a complex polity, we can posit at least four categories of citizenship status: “citizen minus”, “citizen equal”, “citizen plus” and “citizen plural”.

Every member of the state has right to equal citizenship and should never be treated as a citizen “minus”. “Citizen minus” conveys repression and likely forced assimilation. Aboriginals who have been deprived of basic rights of citizenship—such as the right to vote—would fall in that category. This is surely easily discarded in a contemporary democratic context.

Next, comes the “citizen equal” model. Here, the assumption is that all citizens are entitled to the same rights and services. Fundamental rights which should not be curbed, even on grounds of accommodating diversity, would provide the foundations of this category (we return to this question in the last section of the paper). In some cases, equality could entail temporary measures to redress systemic disadvantage, such as affirmative action. This is a basic citizenship regime for any member of a multicultural—and even multinational—society. These measures do not amount to institutional autonomy in the sense considered here.

Institutional autonomy would be reserved to the “citizen plus” or the “citizen plural” categories, who in addition to their “citizen equal” status, can also claim measures of self-government. These two categories vary in terms of the degree of autonomy which each group can claim (both from a normative and a realpolitik perspective). It may be that the distinction between the “plus” and the “plural” revolves around the political power, degree of political mobilization and conception of self apart from the mainstream. For Chartrand, First Nations ought to be considered “citizen plural” because of their inherent entitlement to self-government.

I proposed this typology as a working hypothesis for our subtheme discussions. No groups should ever belong to the “citizen minus” model. Some groups in a complex society can legitimately claim accommodation according to the “citizen equal” model (including temporary quotas, etc.). Others, by virtue of their history, social cohesion, or political power within the overall state, may qualify as “citizen plus or plural”. The degree of institutional autonomy to which they can aspire would be reflected by this qualification. In that context, and to conclude with our prior examples, Québec and aboriginal populations in Canada could claim “plural status”—and thus a very high degree of institutional autonomy. French Canadians outside Quebec could claim a “citizen plus” status, which would entitle them to certain attenuated forms of self-government. Other diversities can claim special accommodation measures— short of autonomy—to enable them to be “citizens equal”.

Obviously, the requirements for accommodating diversity will always depend on a particular country’s circumstances, including its specific diversity mix (e.g. nature of diversity, territorial distribution, respective proportions in the over-all population, etc.). Each multinational state will be faced with these types of conundrum, to which there is no ready-made answer. Again, contextualized discussion would benefit from exploring this issue in a variety of settings, including in developing or emerging federations or quasi-federations. It may also be that discussion grounded in concrete examples—particularly different hard cases—will lead to a clarification or rejection of this typology.

4. Institutional Options to Promote Autonomy

“Autonomy” can take the form of “territorial” or “non-territorial” solutions, or of a combination of both. Institutions can be set up at the central, regional or local level. In other words, it is not only at the national or central level that adaptations have to be made, but also, in some cases, within constitutive units. Asymmetrical solutions may also offer appropriate responses to the heterogeneity of diversities found in a particular polity.

Every option carries with it a number of advantages and disadvantages. It is worth considering who wins and who loses with each of these options. The idea is not to insert a form of Manichean accounting, but to underline that preserving both diversity and some form of cohesion, necessarily entails compromises.
4.1 Territorial Solutions

Short of full-fledged independence, one of the most advanced forms of autonomy is a federal system which transforms a particular minority into a majority within one or several of the constitutive units. To the extent that competences and fiscal resources allow for effective decision making and implementation of policies, autonomy flows from an application of majoritarian politics within that unit. The distinct group can then develop policies adapted to its cultural preferences, without having to constantly negotiate with the federal majority.

Some federations were at least partly built on that principle from their inception: that is the case of Switzerland or Canada, for instance. Others have gradually chosen that route as a means of accommodating diversity and reducing tension within a former unitary State. Since the 1970s, Spain, South Africa, Ethiopia and Belgium have gradually adopted forms and shades of territorial federalism (although in some cases, the “F-word” is not formally used, for historical and political reasons). Devolution in the United Kingdom follows a similar logic, particularly with regards to Scotland.

In some cases, the territorial federal option will entail the redesigning of borders between units, or the creation of additional constitutive entities. Switzerland and India offer relatively recent examples of such restructuring. The objective is to seek a better fit between relatively homogeneous and geographically concentrated minorities and the political borders which define the territorial competence of legislative and executive organs.

Hence, under the pressure from a catholic French-speaking minority, the canton of Bern was split in 1975 to give rise to the new canton of Jura. This was accomplished through a remarkable series of public consultations with the affected population (both minorities and majorities). Since the late 1950s, 14 new states have been created in India, in order to provide the institutional advantages which flow from territorial federalism to linguistic/ethnic/ religious minorities.

While territorial solutions offer an effective mechanism for accommodating diversity for geographically concentrated minorities and aboriginal populations, they also raise a number of concerns. Let us consider three.

First, in highly diverse societies, a significant degree of cultural homogeneity within constitutive units will rarely be attainable, short of an ever increasing number of very small units. In many cases, and notably in major cities, the intermingling of diverse groups is simply unavoidable. Nigeria has moved from 3 to 36 units, in an attempt to appease tensions between its 250 or so ethnic groups. This appeasement has only been partially successful. The Nigerian example invites us to reflect on the risk of extreme fragmentation that can come with constant restructuring.

Second, fear is sometimes expressed that territorial solutions may foster secessionist tendencies. Empirically, such occurrences are rare. An appropriate federal structure and spirit are not absolute antidotes against secession. However, they may be the most effective means of limiting the risk that it occurs, while also reducing the threat of oppression of the group which seeks this ultimate form of autonomy.

Third, recent history has shown the devastating effect of some searches for autonomy based on cultural homogeneity. Ethnic cleansing or major involuntary displacement is a particularly perverse way of generating conditions for a diverse group to be geographically concentrated in a particular territory. The war in Bosnia represents a good and difficult case in point.

The 1995 Dayton Agreement divided that territory into two main entities. At present, the Republika Srpska can claim a relatively high level of ethnic homogeneity, compared with the heterogeneity that characterized the region until the early 1990s. Building a permanent federal solution partly on today’s socio-demographic reality is vigorously opposed by Bosniaks who were displaced (or whose families were eliminated) during the war. On the other hand, a return to a more unified and mixed Bosnia is opposed by Serbs of the Republika Srpska, who wish to maintain and strengthen the autonomous institutions which they obtained through the Dayton Agreement. These institutions are notably conceived as measures of cultural protection, given the minority status of the Bosnian Serbs, which resulted from the explosion of Yugoslavia and the independence of Bosnia. They are also viewed as a shield against reprisals for war exactions.

I certainly have no solution to propose to this particular conundrum. I simply raise it in order to encourage a contextualized discussion on the pros and cons of territorial federal options.

Apart from these risks associated with territorial solutions, some other concerns can be raised by specific groups which are affected by territorial solutions: a majority at the central level, minorities within the minority and those that are left out.
4.1.1 What’s in it for the “Majority”?

The advantages of a territorial solution for a geographically concentrated minority are obvious. But why—as a participant in a civil society seminar in Sri Lanka candidly and perceptively once asked—should the majority agree to this?

Indeed, agreeing to institutional autonomy for the benefit of distinct groups is not necessarily without cost. The majority may fear the lack of consistency in a number of policy areas. It may also resist practices in some constitutive units which it deems to be contrary to fundamental rights (the controversy regarding a particularly harsh application of the sharia by certain Nigerian constitutive units is a good illustration). Some may reject asymmetrical arrangements as an assault on the principle of equality between constitutive units. There may be concerns that solidarity between citizens of the overall state will be threatened by such territorial restructuring. The majority may experience a sense of loss regarding a unified nation which it perceives—rightly to wrongly—to be inclusive of every citizen in the state, over and beyond the various types of sociodemographic and cultural differences. These concerns ought to be taken seriously.

This being said, there are advantages in federal arrangements, even for the majority. Territorial federalism may offer advantages that are similar to those found in classic non-multinational federations such as Germany, Australia or the United States. Federalism multiplies levels of entry for citizens as well as checks and balances between centres of power. It brings power closer to the people, while allowing joint action in some cases. It counterbalances management from afar in geographically vast countries. The fact that some units are composed of a minority—which is transformed into a majority within a constitutive unit—should not represent a hurdle in that context.

More prosaically, the majority may simply not have a choice, if it wants to live in a peaceful, stable and democratic state. Oppression is neither legitimate nor effective. Attempts at minimizing differences by appealing to a single nation are likely to be counterproductive. In some cases, it is only by conceding “citizen plus” or “citizen plural” status to a segment of the population, that the risks of violence and instability will be reduced.
4.1.2 “Minorities Within the Minority”

A side effect of territorial federalism is that even with a meticulous drawing of borders, a constitutive unit will rarely contain an entirely homogeneous population. And even if substantial homogeneity could be attained, it may not resist the test of time. People move and diversity has a tendency to erupt at every level of most societies. In other words, there are likely to be minorities within the minority-transformed-into-a-majority at the level of a federated entity. The debate concerning the adequacy of various options then shifts from the overall state to that entity.

Again, the “citizen minus” status ought to be avoided, and all members of the constituent unit should at the minimum enjoy the status of “citizen equal”. In some cases, this could imply some form of affirmative action or temporary quotas. Arguably, in other cases, however, the “minority within the minority” may be entitled to “citizen plus” status (with guaranteed linguistic rights, for instance) or even to the status of “citizen plural” (with decentralized autonomous institutions). For instance, in Switzerland, control over schools can be delegated to municipalities within some unilingual cantons, so that a geographically concentrated minority at that level can exercise control over a vital institution. This, of course, falls short of “full institutional autonomy” (we cannot multiply parliaments ad infinitum). Finally, some mechanisms to guarantee effective participation of a minority at the overall federal level can be implemented in favour of the “minority within the minority” to ensure that their voice, and the asset that their very existence represents within their constitutive unit, is not ignored.
4.1.3 And “Those Who are Left Out”

Drawing borders means some people are in, some people are out. “Minorities within the minority” are incorporated sometimes against their will, or their interests. Similarly, territorial solutions have the effect of leaving some people out. To illustrate this other side effect of a territorial solution, let us return to the Canadian example. The territorial solution in Canada has been beneficial for the French-speakers in Québec. But it has further marginalized the French-speakers who live in the other provinces. Federal (central) institutions, function in a bilingual mode. However, apart from the right to manage their own educational institutions (and much more timidly, the right to have their interest considered with regards to health institutions), these various minorities do not enjoy any form institutional autonomy either in their province, or at the federal level.
4.2 “Cultural Autonomy” and “Personal Federalism”

When territorial autonomy is not a viable option in a particular state, it may be adequate to provide for a form of “personal or cultural autonomy” which is “de-territorialized”. This will notably— but not exclusively—occur when a group linked on the basis of some “personal characteristic” (e.g. religion, language, ethnicity, etc.) is geographically dispersed.

For instance, a number of federal (India being one) and non-federal countries recognize the right of religious communities to be governed by their “personal law” particularly with regards to family matters, regardless of where they live. Such laws are not necessarily determined by democratic institutions, although, in theory, they could be.

“Cultural autonomy” refers to a more institutionalized version of the personality principle. Developed by theorists in Eastern Europe, it has known some partial implementation in that region as a means of managing diversity in highly heterogeneous contexts. In some cases, it is also favoured despite the existence of geographically concentrated minorities, as an alternative to territorial subdivisions. Cultural autonomy translates into various shades of self-government for certain communities, with regards to matters considered to be crucial to their vitality. Again, the purpose is not solely to guarantee collective rights to those groups, but also to institute forms of normative autonomy: the right to determine one’s own law. In other words, we are not only in the “citizen equal” paradigm, but also in some versions of the “citizen plus” or even “plural” models.

For instance, non-territorialized “cultural councils” have been instituted in Estonia, Slovenia or Romania, in favour of certain “historical minorities”. With some variations, these institutions exercise a degree of control over education or certain social services. Some of these councils can levy taxes. They are either elected or nominated by their constituency.

The most advanced form of personal autonomy would posit the existence of actual constitutive units based on linguistic, ethnic, religious—or even racial—criteria. In theory, these units would enjoy normative autonomy to be exercised by distinct legislative and executive organs. Membership in such a “community” carries with it a number of rights (e.g. to vote, to obtain services, to participate in the institutions, etc.), as well as obligations (e.g. to respect the laws of the community, to pay taxes). The logistical problems generated by people simultaneously or successively belonging to different communities can be daunting. Partly for this reason, such a high degree of non-territorialized autonomy remains largely elusive.

But the most critical obstacle facing this type of institutional option lies with the “connecting factors” or criteria for belonging to the group. It is generally agreed that there should be a significant degree of self-identification. Furthermore, to avoid false intrusions within the group, a certain degree of group control over membership is also generally considered legitimate. The fundamental right of association and the principle of individual self-determination entail that individuals should be entitled to “opt in” the minority system, be governed by the “general law”, or belong to a “neutral” group, were the entire society to be divided into “cultural segments”. Moreover, given the objective which is to nurture diversity, it is important to allow for multiple, superimposed and shifting identities.

Given these major difficulties, it is not surprising that there are no actual examples of full-fledged “personal federalism”. As we shall see, Belgium has instituted an attenuated form of personal federalism. The closest example might be Lebanon, where the entire population of this consensus-driven but officially unitary state is divided into 18 distinct “religious” groups. While a “neutral” category was originally envisaged when the regime was set up in 1943, it was never put into place. Half-way between “personal law” and “cultural pluralism”, the actual workability of a non-territorial form of autonomy is difficult to assess from the Lebanese example given the weakness of that country’s democratic institutions. In any event, Lebanon may represent a better example of normative power-sharing at the centre—a version of consensual democracy—than of normative autonomy for non-territorial constitutive units.
4.2.1 Who Wins ? Who Loses ?

Who wins and who loses in such a (still rather theoretical) system? On the positive side, distinct groups that are geographically dispersed can aspire to democratic self-government, while the majority need not fear territorial disintegration. On the negative side, the risk of consolidating systemic mono-identities and of forcing people into fossilized ghettos is undeniable. This is likely not the best way of fostering the peaceful sharing of the same physical and political space. Moreover, it is difficult to imagine how such a system can work for matters which are not closely connected with language, culture or religions. Yet, the interests of minority groups may be related to territorially-based issues, such as natural resources. Creating self-governing institutions could deflect from the real issues which require that their concerns be taken seriously by the overall state or the constituent unit in which they live.

This being said, in a world in which the geographic concentration of culturally distinct groups is no longer the norm, the idea that some degree of normative power can be held by groups of individuals on criteria other than those of nationality or place of residence, deserves consideration.
4.3 A Mixture of Territorial and “Personal” Solutions

Some intermingling of territorial and non-territorial options has actually been attempted in particular instances as a means of drawing on the advantages of each, while minimising their drawbacks.

This is the case in certain forms of aboriginal self-government. In Canada, aboriginal self-government (still in its infancy) translates into institutional autonomy for groups who reside on certain traditional lands. However, given the high proportion of aboriginal people who now live in urban settings, contemporary schemes introduce extra-territorial law-making that draw upon principles of cultural autonomy. Concretely, this means that self-governing institutions can legislate with regards to individuals who reside on the group’s homeland, but also with regards to those who are connected to the community but who do not habitually reside in the community. Criteria for determining membership in the community are outlined in specific agreements and legislation elaborated by the community itself. Services can thus be developed by those autonomous institutions, but offered to members who live away from the traditional territory.

The Belgian federation offers another example of this mixed configuration. The country is divided into three Regions, which have competences over matters related to territory (land use, agriculture, environment, transport). The same territory is also divided into three Communities, which are responsible for matters in which culture and language are deemed to be of particular salience. A person is thus simultaneously a citizen of Belgium, of a Region, and of a Community. The distribution of competences is divided threefold, rather than between two orders of government. While Regions have directly elected Parliaments, Communities have indirectly elected legislative organs. All have their own government and civil service. We are clearly in a three-way federal paradigm.

The logic of the system would lead one to think that any person belonging to the Flemish Community, for instance, could receive services in Dutch anywhere in Belgium. This is not the case. For a number of historical and political reasons, Communities only exercise their competences in specific areas. In other words, this “personal federalism” is—paradoxically—also “territorialized”. The main advantage of the system is that it allows the two major Communities to exercise some competences simultaneously in Brussels, the capital city.

To avoid the permanent labelling of people, an original solution was designed. In bilingual Brussels, both the French and the Flemish communities can organize services in either language, to which anyone can have access. In other words, no one need declare an indelible identity. Concretely, this means that any resident of Brussels can send his or her children to French or Dutch-speaking schools, join a choir financed by either Community, or receive certain types of social services organized by one or the other Community. The freedom of choice discussed above is thus largely respected and the fossilization of identities is partially avoided.

The system is eminently creative and has been studied in a large number of conflict resolution settings. It has drawbacks, however, notably in terms of social cohesion and distrust. Hence, while no one is forced to choose a single identity, the Communities largely function in isolation, creating ever increasing schism between citizens, rather than creating bridges between them.

5. Conclusion

As was mentioned earlier, adapting state structures so that certain groups enjoy a degree of institutional autonomy may be the best insurance for peaceful living together. This being said, such autonomy can also lead to isolation, polarization and a lack of solidarity between peoples who self-govern separately. This is where federal principles, which need not be incarnated in actual federal institutions, come into play. Some balance between self-rule and joint-rule must be found. The balance requires a certain degree of trust, or at least a leap of faith that the “Other[s]” can be trusted. If the proper balance is reached, trust will be consolidated.

Diversity as an asset implies that the interest of diverse groups be duly considered at the federal level, and not merely relegated to autonomous institutions. To avoid excessive fragmentation, polarization and isolation, it is crucial that groups which enjoy institutional autonomy also partake in various forms in common institutions. For one thing, it is likely that a number of policies developed at the central level would impact on them. Their concerns ought not to be set aside on grounds that their major concerns are being dealt with elsewhere. Proper participation through consultation, guaranteed seats in legislative assemblies, quotas in the executive, the civil services, the judiciary, the police, may be justified to ensure that the voices of diversity are heard. However, such measures also carry a risk of segmentation and the promotion of mono-identities, which can contradict the very objective of promoting diversity.

The risk of segmentation inherent in autonomous institutions and guaranteed rights of participation should be met by deliberate bridges built into the system. If every citizen of Bosnia can only elect their own member of the three-person-Presidency, by declaring their mono-identity, collective action may be illusory. If every Belgian pupil follows a course solely designed by a single community with very little connection with the others, while their parents only watch television in a single language, there are risks not only of ignorance of the other communities, but also of prejudice and of the demonization of the others.

While autonomy may be the best guarantee against oppression on the one hand, and territorial disintegration on the other, it can also have the effect of erecting a “Berlin wall” between people. The metaphor is admittedly strong. But mechanisms for ensuring that children, students, all walks of civil society, politicians, artists, get to know people behind the self-governing diversity are vital. They can include exchanges, common projects and effective communication. At the institutional level, they can lead to limiting or prohibiting mono-ethnic political parties and to promoting cooperation between policy makers. While such trust-building strategies are surely beneficial in any federation, they would appear to be even more so in “post-conflict” situations. Only then can diversity be appreciated as an “asset”.