Between a republican and a Bengalee state: Confronting exclusionary constitutionalism in Bangladesh

Abstract

The constitutional design of Bangladesh is characterized by an ambivalent choice: it aspires to establish a republican yet a Bengalee state by putting itself in the conflicting terrain within the demos–ethnos binary. This article aims to examine the implication of this problematic choice along all three axes of the constitution’s elemental parts: its identity, rights and structure. While the identity element of the Bangladesh Constitution embodies the ethno-nationalist vision of the Bengalee state that transforms demos into ethnos, its rights and structural aspects reflect its republican promise to transform ethnos into demos. Contemporary scholarship seeks to confront the exclusionary dimension of the ethno-nationalistic choice in Bangladesh but ends up accepting ethnos as a politically superior value. Such an approach brings us to the politics of difference and, with that, undermines the integrationist potential of the republican constitution. In response, this article defends the republican promise of the Bangladesh Constitution while arguing that what we need in Bangladesh is the ‘de-ethnicization’ of the republic, one that can be achieved by transforming ethnos into demos and not the other way around.

I. Introduction

In The Inclusion of the Other, German philosopher Jürgen Habermas argues for maintaining a distinction between the two figurations of peoplehood: ethnos and demos. ¹ Habermas’s defense of this distinction is directed against the idea of ethnonationalism that collapses the distinction between an ethnos, a pre-political community of ‘shared descent organized around kinship ties’ and a nation constituted as a state.² This collapse suggests that ‘the demos of citizens must be rooted in the ethnos of nationals (volksgenossen) if it is to stabilize itself a political association of free and equal legal consociates’. For Habermas, there is no a priori reason why a constitutional democracy should rest itself on this logic. Instead, he believes that the republican concept of citizenship may better serve the purpose of constitutional integration. In other words, the progressive extension of citizenship to the whole population provides the state not only with a new source of legitimation but also with a new level of abstract, legally mediated social integration.³

From a Habermasian perspective, a republican constitution founded on the voluntary association of the demos is therefore more appropriate than the ethno-nationalist or even the liberal-communitarian conception of the nation and democracy.⁴ This is because it allows us to disrupt any convergence – symbolic or historical – between republicanism and nationalism that existed, for Habermas, only as a transitional and historical constellation. As he argues, ‘republicanism is neither conceptually nor practically dependent on nationality, and the twentieth century, in particular, has provided grotesque examples of the dangers of emphasizing the relationship between ethnos and demos’.⁵ In this article, we wish to argue that the constitutional architecture of Bangladesh is caught up in the same danger that Habermas finds in the collapse between republicanism and ethnonationalism. This is evident in Bangladesh’s ambivalent choice in that it aspires to establish a republican yet a Bengalee state, thus placing itself in the conflicting terrain of the demos–ethnos binary.⁶

The purpose of this article is therefore to examine the implication of this problematic choice along all three axes of the constitution’s elemental parts: its identity, rights and structure. In order to do so, we draw on and situate our findings in three different families of constitutional strategies – assimilation, accommodation, and integration – all of which find their place in the constitutional design of Bangladesh.⁷

First, we identify that the original Constitution of Bangladesh adopted the ethno-based Bengalee identity for all, regardless of ethnic difference and diversity.⁸ This ethno-nationalistic paradigm of identity formulation falls prey to being categorized as ‘assimilative constitutionalism.’⁹ In assimilative strategy, unity is achieved through a coercive ‘acculturation’ that ‘involves one community adopting the culture of another and being absorbed into it’.¹⁰ In that sense, assimilative strategy is only a form of exclusionary constitutionalism. Second, while ethnicizing constitutional identity, which aims to transform ethnos into demos, the Constitution adopted the demos-centric arrangements for other elemental parts: the architecture of rights and electoral and representative structure of the state. This is because it provides for a republican architecture of rights and a unitary governance structure based on territorial constituency rather than ethnic representation. This arrangement of individual rights and de-ethnicized representative structure fits into the ‘integrationist’ approach to constitutionalism,¹¹ which promotes a common public identity without demanding ethnocultural uniformity or assimilation.¹² From this perspective, the integrationist approach comes closer to the promise of the abstract and legally mediated form of civic integration defended by Habermas.

Finally, in the subsequent constitutional arrangements, there has been an oscillation between these two approaches. During the first military regime,¹³ the Constitution was amended (the Fifth Amendment) to redefine the constitutional identity by replacing the ethno-centric Bengalee nationalism with Bangladeshi nationalism, a demos-centric construction of territorial nationalism.¹⁴ However, the recent Fifteenth Amendment¹⁵ has reinstalled, in the name of restoring the original constitution, the previous arrangement of ethno-centric identity of the people, and with that returned to the assimilationists framework. Interestingly, the same amendment has brought a change in relation to its rights arrangement by inserting a new provision for cultural rights of the ethnic groups.¹⁶ Between the time of the Fifth and the Fifteenth Amendments, the government also signed the Peace Accord to grant certain forms of regional autonomy to the Chittagong Hill Tracts (CHT) people, thus initiating a structural change towards ethnic accommodation.¹⁷ With these changes, the Constitution seems to adopt an ‘accommodationist’ strategy,¹⁸ which requires the recognition of ethnic difference and ‘adjustment to the special interests and needs of groups’.¹⁹

Such accommodationist demands were first raised by Manabendra Narayan Larma, an influential ethnic leader who represented the CHT people at the Constituent Assembly of Bangladesh. Larma demanded that the constitution be designed around ethnic difference and his proposal was more comprehensive, as he argued not only for the recognition of ethnic identity but also for the inclusion of enforceable group rights and a separate legislative council for the CHT people.²⁰ However, this demand has not been addressed in the Constitution, although the Peace Accord and the Fifteenth Amendment have already initiated some accommodationist strategy. This brings us the accommodationists’ critique that, despite its recent move to plurality-consciousness, the Constitution is not sufficiently inclusive.²¹ There is also a more forceful argument that the Constitution, with the changes made by the Fifteenth Amendment, has maintained the relations of domination through assimilation of minority identity into the dominant culture.²² From this perspective, critics argue for reform along accommodationist strategy while formally opposing coercive assimilation by the majority.

While we share similar concerns about assimilationism to those raised by the accommodationists, we differ in terms of solution. In this article, we defend republican-integration against both assimilationist and accommodationist strategy.²³ Its main targets are the accommodationists’ response to the failure of the current constitutional model. As we argue, the accommodationist approach seeks to confront ethno-nationalism but ends up accepting ethnos not only as a morally superior value but also as a site of political mobilization. This article responds to these problems while arguing that a truly inclusive constitution in Bangladesh would entail the reversal of such logic. That is, it should involve the de-ethnicization of the Constitution by maintaining neutrality to ethnic difference, which resonates with the republican promise of transforming ethnos into demos. This defence for the republican-integrationist strategy is grounded on the following justifications.

The first justification is contextual. The integrationist approach claims that it functions better when the societies are not deeply polarized and ‘when there is already extensive heterogeneity, hybridity, and mixing’.²⁴ Demographically, Bangladesh has only 2 per cent of people identified as ethnic minorities as opposed to the originated majority. The ethnic minorities composed of as many as eleven communities live predominately in the hilly region of Bangladesh (known as Chittagong Hill Tracts), but some part of these communities also inhabit in the plain land.²⁵ Moreover, with increasing Bengalee migration to the CHT region,²⁶ a demographic shift has taken place, resulting in mixing and hybridity. The implication of this shift is that the multiple ethnic groups living in CHT cannot ‘realistically’ aspire and maintain ‘either territorial autonomy or consociation’.²⁷ However, the question may remain whether Bangladesh is still deeply divided along ethnic lines, despite the relatively small size of its ethnic minorities and the demographic mixing in the CHT region. It is important here to emphasize that a society may be ethnically diverse yet not deeply divided. As Benjamin Reilly suggests, for a society to be ethnically divided, its ethnic community must have politically salient cleavages around which interests are organized for political purposes or political mobilization.²⁸ Such political mobilization occurs, at an empirical level, through ethnically characterized electoral politics, where political parties respond by organizing themselves on the basis of ethnicity and ethnic individuals cast votes only for their own ethnic political party.²⁹ In Bangladesh, ethnic divisions display no such political salience in the polity’s party system.³⁰ This indicates that Bangladesh is not deeply divided, at least along an ethnic axis, although a fairly deep division can be found to exist along ideological lines (between Islamists and secularists) or on religious lines (between Muslims and Hindus). Given that such ideological or religious division is beyond the scope of this article, we accept that problem of ethnic exclusion in Bangladesh can be addressed by an integrationist approach that functions better in an ethnically diverse yet not divided society.

The second justification comes from the transformative promise of republican-integration. The republican-integrationist approach shares the goal of unity with blindness to ethnic difference without denying the social existence of ethnic diversity. In this way, it differs significantly from the assimilationists approach, which eliminates difference by establishing one ethnic identity. In contrast, the integrationist approach achieves unity by disestablishing all the ethnic identities, including the majority’s own. This de-ethnicization expands ‘we-ness’, which then underpins the demos-centric, trans-ethnic solidarity within the framework of republican citizenship. At this point, someone may legitimately ask how this promise of republican-integration is to be translated in political terms. To address this concern, we wish to clarify that our aim is not to offer an empirical account of the transformation from ethnos to demos and the ways in which such transformation could make a qualitative difference in the current socio-political milieu. Instead, our prescription is founded on the proposition that constitutions can transform the polity by reducing the gap between the norms and social facts. Therefore, a careful and purposive constitutional design can result in changes in political behavior and practice.³¹ Moreover, as we show in this article, the republican promise of the Bangladesh Constitution prescribes the removal of cultural and socio-economic obstacles. Such prescription will help us to understand the ways in which the Constitution can deal with the difference-sensitive claims in the real world while maintaining a demos-centric, difference-blind normative foundation.

The remainder of this article is developed in the following way. Part II offers a critic of the assimilationists promise as reflected in its adoption of Bengalee nationalism. It argues that the Constitution has not just declared Bengalee nationalism as a rhetorical value, but also sets it as a catalytic force for transforming demos into ethnos. Parts III and IV respond to the accommodationists’ arguments for constitutional reform by way of defending an integrationist approach that meets the republican promise of the Bangladesh Constitution. In Part III, we argue that the recent constitutional amendment has had the effect of shifting to an ethno-centric arrangement from the difference-blind approach of rights provided in the original Constitution. Part IV shows how political contestation over the issue of ethnicity has led to the problematic plan for structural reform that masquerades as a solution. It argues that the ethno-centric reform of representative structure will produce minorities within the minority, thereby bringing us to circular logic of exclusion. The article concludes with a demand for reform along the republican promise of difference-blindness.

II. ‘Bengalee and democratic state’: Exclusion by assimilation?

Joseph Raz once criticized the juxtaposition of Jewish and democratic state as ‘morally indecent’.³² The constitutional characterization of Bangladesh as a ‘Bengalee’ yet a democratic state invokes an analogy with such criticism of the Jewish state advanced by Raz. In this section, however, we argue that the idea of Bengalee state is more problematic than what Joseph Raz called ‘morally indecent’. We develop this argument by showing that the current constitutional arrangement entails coercive assimilation through the exclusion of the ethnic ‘other’. In doing so, we first examine the moral relevance of Bengalee state, then move to show how it comes to be the site of assimilative politics. To begin with, the project of assimilative politics has been fortified by the constitutionalization of four political ideals – nationalism, democracy, secularism and socialism (NDSS) – with a definition of nationalism in article 9 that enforced the superiority of Bengalee identity.³³ The enumeration of these ideals was a much-debated issue from the very moment of drafting the original Constitution. The debates were folded around the reconcilability of two sets of ideals that came in pairs: ‘democracy–socialism’ and ‘socialism–nationalism’. In the Constituent Assembly, it was raised that ‘democracy’ cannot be reconciled with the ideal of socialism.³⁴ The other line of critique came to address the problematic relationship between ‘nationalism and socialism’. For example, Serajul Islam Chowdhury has criticized the mixing of nationalism with socialism while identifying the danger that it ‘could result in the production of what has come to be known as Nazism’.³⁵ Chowdhury goes further to identify the exclusionary dynamics of Bengalee nationalism while criticizing it in the following constitutional terms:

in their enthusiasm they were oblivious of two ground realities. Firstly, that there were non-Bengali small nationalities living within the territory of Bangladesh, and secondly, that in the modern world a state with a single nationality is not a viable proposition. There is, however, an unconscious display of nationalist chauvinism in the idea advanced in the constitution to the effect that all citizens of the state would be called Bengalis. ³⁶

This criticism reminds us of Joseph Raz’s ‘moral thesis’ against the idea of a ‘Jewish and democratic state’.³⁷ In Raz’s view, any ethnic or national values to be attached to the state are true but only signify ‘false values of national self-aggrandizement and chauvinism’.³⁸ Raz argues that such Jewishness of the state gives a ‘false consciousness’ to the people who are non-Jews, but the citizens of that Jewish state.³⁹ Hence, democracy cannot be retained by one state simultaneously while declaring its identity as a Jewish state alongside being a so-called neutral state. For him, the principle of democracy, being a true signifier of universal value, does not necessarily require the state to customize or localize that universalism into its own (state in concern) language. In other words, there is no need to declare or identify a particular state with an ideal to be a democratic and moral state, even if it is for the sake of any symbolic meaning.⁴⁰ It can simply act on being the moral state – otherwise, it cannot be a home for other people who do not belong to that particular symbolic ideal.⁴¹ This leads Raz to claim that morally decent countries need not declare their nationalism to show their character and traditions to the whole world.⁴² Such argument advanced by Raz reveals that ‘the Jewish and democratic elements are morally odd, or at least in tension with each other.’⁴³

From this perspective of Raz’s moral thesis, it may appear that the vision of being a ‘Bengalee state is morally odd with the idea of democratic state’. But we will argue that the implication of establishing a Bengalee state is somewhat more sinister than the case of privileging Jewish nationalism in the context of Israel. This is because the Constitution of Bangladesh does not just privilege Bengalee nationalism over other nations, as in the case of Jewish nationalism; rather, it also provides for coercive acculturation of other ethnic communities by denying the possibility of other identities.⁴⁴ The original Constitution institutes this logic of acculturation by declaring that ‘citizens of Bangladesh shall be known as Bengalees’.⁴⁵ That means it imposes ‘Bengalee nationalism’ for all of its citizens irrespective of the distinctive ethnic and cultural background of many other communities.

Interestingly, there was a strong resistance to such constitutional arrangement at the moment of constitution-making. In the Constituent Assembly, Larma, the lone representative of the CHT people, raised his concern against the constitutionalization of Bengalee nationalism in the following words:

You cannot impose your national identity on others. I am a Chakma, not a Bengali. I am a citizen of Bangladesh-Bangladeshi. You are also Bangladeshi, but your national identity is Bengali … They [tribals] can never be Bengali.’⁴⁶

In response, Sajeda Chowdhury, the female member of the Constitution Drafting Committee, argued that the inclusion of tribes as Bangalees placed the hill people in a more dignified position as it recognized them as a nation rather than sub-nation.⁴⁷ The argument advanced by Chowdhury echoes the position of Sheikh Mujibur Rahman, the country’s founding leader, who expressed that the non-Bengali ethnic communities should assume the Bengali identity by shunning their own.⁴⁸ Such abrupt denial of ethnic identity is thought to be founded on the concept of ‘assimilative constitutionalism’.⁴⁹ It performs an inherently exclusionary function in the name of acculturation that amounts to the negation of the existence of other ethnic communities.⁵⁰

This situation assumes a more complex character with the recent enactment of the Fifteenth Amendment,⁵¹ which has replaced the word ‘citizen’ (as used in the original Constitution) with the word ‘people’, and stipulated that ‘the people of Bangladesh shall be known as Bangalees as a nation’.⁵² This replacement of ‘citizen’ by the ‘people’ reinforced the ethno-nationalist promise of conceiving Bangladesh as the ‘Bengalee state’. This conflation of people as Bengalees is nothing but a forcible transformation of demos into ethnos. Seen from this perspective, the concept of the Bengalee state finds its striking correspondence with one of Haiti’s earliest constitutions of the Black state, where all Haitian citizens were legally defined as Black, regardless of skin colour or prior racial categorization. By securing Haitian citizenship, a person became Black in the eye of the constitution of Haiti;⁵³ in a similar way, a non-Bengalee citizen of Bangladesh becomes a Bengalee in the eye of the Bangladesh Constitution.

While intensifying its assimilationist strategy by the Fifteenth Amendment, the Constitution has, however, provided for the ‘soft recognition’ of the existence of communities other than Bengalees. This comes in the form of non-justiciable cultural rights provided under the newly inserted article 23A, which aims to protect and develop the cultural tradition of ‘the tribes, minor races, ethnic sects, and communities’.⁵⁴ In Part III, we will return to article 23A in order to examine its implications for the republican architecture of rights. For now, we wish to engage with the scholarly response to the issue of recognition of ethnic diversity as mentioned above.

One criticism that seems most dominant across disciplines is that the Constitution of Bangladesh is defective in terms of addressing ethnic diversity.⁵⁵ For example, Ridwanul Hoque, coming from the perspective of inclusive constitutionalism, argues that the non-recognition of indigenous peoples in the original Constitution is a ‘genetic defect’⁵⁶ or a ‘grave mistake’.⁵⁷ This echoes the recent acknowledgement from the chairman of the Constitution Drafting Committee that the ‘sense of fulfilment’ for Bengalee nationalism has led to the constitutional exclusion of the other ethnic communities living in Bangladesh.⁵⁸ Hoque, however, appreciates the recent change made by the Fifteenth Amendment as a positive move from non-recognition to an explicit ‘attempt of inclusion’.⁵⁹ For a truly fuller recognition of the ethnic peoples, he argues that Bangladesh needs to ‘shift clearly to an accommodationist approach to these peoples’ distinct identity’.⁶⁰

In contrast, Mohammad Shahabuddin sees the recent amendment as the reinforcement of the dominance and hegemony of Bengalee nationalism through which the hill people of the CHT have become ‘constitutional outcasts’.⁶¹ He argues that, ‘While this provision finally acknowledges the existence of communities other than Bengalees, it nonetheless underscores that unique cultures of these communities fall outside the “national culture” (defined in line with Bengalee nationalism).’⁶² This argument stems from the fact that, under the current Constitution, the ethnic communities are termed and classified as ‘tribes’, ‘minor races’ and ‘ethnic sects’, not as ‘indigenous’, ‘aboriginal’ or ‘adivasi’, as they demanded or deserved.⁶³

The responses advanced by Hoque and Shahabuddin seem to fall into the accommodationist camp, as they insist on the recognition of ‘multiple public (constitutional) identities’ to secure the coexistence of different communities within the same state.⁶⁴ In this sense, they seek to resolve the problem of ethnic exclusion through emphasizing ethnos itself. So, we would suggest, in the name of accommodation, they endorse the identity-based politics, or what Charles Tylor calls the ‘politics of recognition’.⁶⁵ One obvious implication of such politics of difference is that it involves ‘the positive validation of ethno-cultural difference’.⁶⁶ This is why the effectiveness of such strategy is questioned by both the liberal and republican thinkers. For example, David Miller, who is considered a liberal nationalist,⁶⁷ identifies the danger of identity politics and argues that it can be counter-productive (self-defeating) in many ways.⁶⁸ According to him, the advocate of difference would destroy the conditions under which disparate groups in a culturally plural society can coexist with a common goal of social justice. In this sense, identity politics is ‘potentially damaging to the interests of the groups it is meant to serve’.⁶⁹ This led him to defend ‘republican citizenship’, which, according to him, ‘is better able to respond to cultural diversity than these other versions’. It can do so, he argues, ‘by virtue of its ability to draw groups who initially have very different priorities into public debate, and to find compromise solutions to political issues that members of each group can accept’.⁷⁰ In other words, republican citizenship can resolve the problem of ethnic exclusion in a better way than the strategies suggested by the accommodationist (multiculturalist) camp. For Miller, this defence of republican citizenship is, however, linked closely with the defence of nationality because republican virtues ‘are likely to be cultivated only within national borders’.⁷¹

At stake in Miller’s account is that it speaks of a ‘moderate nationalism’ that is different from race-based or ethnicity-based form of nationalism.⁷² It allows the formation of an autonomous self-constituting political nation based on a territorial rather than ethnically determined figuration of peoplehood.⁷³ As we suggest, this idea of territory-based nationalism is consistent with the republican idea of equal citizenship that combines ethnic neutrality with civic solidarity.⁷⁴ For the purpose of this article, we may call it ‘republican nationalism’ as distinct from liberal nationalism.⁷⁵ Our purpose here, however, is not to define or defend Miller as a republican theorist, but rather to draw on the fact that the marriage between republicanism and nationalism represents a defensible alternative to ethno-nationalism.⁷⁶ One advantage of republican nationalism is that it allows a demos-centric integration without assimilation. We will discuss shortly that such demos-centric integration in turn involves ‘disestablishment’ and non-domination, which marks its distinction with liberal neutrality.⁷⁷

Having identified its demos-centric implications, we wish to suggest that Bangladesh should adopt republican nationalism to replace its ethno-nationalist model of assimilation. It is important to mention here that republican nationalism is not a new idea for Bangladesh. A model of republican nationhood was adopted by the constitutional Fifth Amendment, which was recently invalidated by the Supreme Court.⁷⁸ By this amendment, the Constitution introduced the concept of ‘Bangladeshi nationalism’. It changed article 6 of the Constitution by providing that the citizens of Bangladesh shall be known as Bangladeshis instead of Bengalee. While untying the peoplehood with ethnicity, the same amendment did, however, replace secularism with a firm belief in the Islamic faith, forming the part of fundamental principles of state policy. This kind of idealization reduced the inclusionary promise of Bangladeshi nationalism, and leads the critic to dub it an ‘Islamic-nationalist’ political project.⁷⁹ However, such criticism seems to undermine the distinction between constitutional identity defined by article 6 and the fundamental principles articulated in Part II of the Constitution, given that article 6 has greater normative force than the Part II principles, which are non-justiciable in nature.⁸⁰ As we suggest, with the adoption of Islam as a fundamental principle, the Constitution simply engendered the same problem of what Raz refers to as a ‘morally indecent’ state in reference to Israel. Yet such problem of value endorsement should not deter us from appreciating how the Fifth Amendment de-ethnicized the identity of the constitutional subjects. That is, it stripped constitutional identity of any ethnic ingredients, and in this way transformed ethnos into demos. ⁸¹ Looked at from this perspective, Bangladeshi nationalism introduced by the Fifth Amendment was a better choice because it resulted in a ‘fair compromise’ to the issue of ethnic identity. This is because, it did not recognize non-Bengalee identity, but nor did it declare the supremacy of Bengalee ethnicity. This neutrality to ethnic difference corresponds to the republican promise, which involves ‘trans-ethnic solidarity’ through the idea of equal citizenship.⁸²

There are at least two aspects that would allow us to distinguish the republican dimension of Bangladeshi nationalism from the idea of a ‘homogenic national state’.⁸³ First, while ‘national state’ seems insensitive to diversity and seeks to establish homogeneity through assimilation, Bangladeshi nationalism remained indifferent to ethnic diversity, and symbolized integration without assimilation. Second, ‘national state’ involves the strategy of domination as it grounds national solidarity on the common or shared culture or heritage, which is already always characterized by the predominant historical narrative of nationality. In contrast, Bangladeshi nationalism disestablished the logic of ethno-nationalist assimilation, and with that reflects the policy of non-domination that lies at the heart of republican integration.

At this point, we should elaborate the relationship between ‘disestablishment’ and republican integration. To do so, it seems useful to reflect on the debate concerning the use of laïcité (the principle of secularism) because it involves the issue of republican impartiality and is seen as a yardstick forms of republican integration.⁸⁴ One possible advantage of this is that it will help us respond to the charge raised by the radical advocate of difference (such as Iris Marion Young), who argues that republicanism involves the imposition of oppressive norms in the name of impartiality.⁸⁵

The principle of laïcité was developed as a legacy of the struggle by the French republic to institutionalize the separation of the Catholic Church and the state.⁸⁶ The 1905 republican law separating the church and state articulates the principle of laïcité in a twofold way: first, it guarantees the free exercise of religions; and second, it declares that ‘it [state] neither recognizes nor subsidizes any religion’.⁸⁷ In its original juridical formulation, laïcité requires the neutrality of the state and public authorities, which means no more than the application of this neutrality to religious affairs.⁸⁸ The implication of this principle is that it neither promotes nor combats particular religious practice, and it permits every individual to have or not to have any religion. In this respect, it was different from a principle that aims at competition between religious affiliations and citizenship but was similar to establish a ‘civic religion’.⁸⁹ There is, therefore, a point in questioning the difference-blindness of laïcité regime, insofar as it demonstrates its ‘moral supremacy’ over all religions.⁹⁰ Moreover, from a historical perspective, it also invokes the charge of privileging European cultural majorities while rendering this privilege invisible.⁹¹ Such criticisms, however, become stronger with the recent use of the laïcité principle as a republican justification for the ban of hijab for Muslims in France. It substantiated the claim that republican difference-blindness can be used to discriminate against those religious minorities whose religious practice is more visible than that of others.⁹² Therefore, it shows that the laïcité principle has undergone a radical shift from its original republican articulation in the 1905 law.

What is striking in this bifurcation of laïcité regime within the republican promise is that it slightly shifts from ‘disestablishment’ of religion to the express imposition of ‘obligation’ at the individual level. The obligation-generating function of the ‘new’ laïcité principle sets out to encourage or discourage a particular practice, such as prohibiting hijab. But “disestablishment” would condemn this kind of policy, as Patten has rightly argued that a disestablishmentarian response to cultural diversity ‘would condemn policies that consciously set out to encourage or discourage particular forms of cultural life’.⁹³ Looked at from this perspective, we wish to note that the idea of republican neutrality should not be equated with the new laïcité regime that demands uniformity in public and associational life. Rather, it is the ‘disestablishing’ effect that forms the kernel of republican neutrality, and laïcité can be considered a yardstick of republican integration only if it is taken to imply the condition of disestablishment.

This is why many scholars have attempted to distinguish republican laïcité from the new regime of civic laïcité. ⁹⁴ In a carefully written work, Cécile Laborde advances a similar argument about laïcité as she argues that a truly republican laïcité must focus on where actual neutrality remains unrealized in order to address the privileges of some religions and conceptions of the good life over others. Interestingly, such understanding involves a positive prescription towards the reform of what she calls “official republicanism”. This allows her to offer a distinctively critical vision of republicanism that calls for a strategy of non-domination by way of removing socio-cultural obstacles to minority incorporations. As we suggest, this strategy of non-domination echoes the core idea of republican liberty advanced by Phillip Pettit.⁹⁵ Laborde, however, shows how the strategy of non-domination can work through the arrangement of positive rights and disestablishment.

In Part III, we take on and situate Laborde’s argument in the context of Bangladesh. For now, we should reassert to the claim that Bangladeshi nationalism was reflective of republican neutrality that seeks to disestablish the predominant historical narrative of Bengalee nationality. This was warranted by the ‘we-perspective’ used in the preamble of the original Constitution.⁹⁶ At stake in this ‘we-ness’ is that it not only requires de-ethnicization of nationality but also a shift of its foundation from ‘cultural properties’ to ‘civil rights’.⁹⁷ Before the changes made by the Fifteenth Amendment, the Bangladesh Constitution was almost there. But, as shown above, Bangladesh has returned to the idea that the dominant Bengalee majority alone constitutes the people, a construction that transformed demos into ethnos. ⁹⁸ Having identified this, we now turn to discuss how this ethno-centric (re)turn has affected other elemental parts of the constitution: rights and structure.

III. Rights-based pluralism and the politics of ‘inclusive exclusion’?

As we noted above, the original Constitution has counterbalanced the ethno-nationalist promise of assimilation by adopting a republican-integrationist architecture of rights that prioritizes demos over ethnos. In this part, we examine how this demos-oriented architecture of rights is contested, reasserted and is being directed to an ethno-centric turn. Interestingly, the contestation over the rights issue exists from the founding moment of constitution-making. While debating in the Constituent Assembly, Larma proposed for a right-based accommodation of the non-Bengalee ethnic communities. He expressed his discontent against the constitutional non-recognition claiming that the Constitution ‘did not reflect the hopes and aspirations of the tribal population’.⁹⁹ But when his recognition-demand was rejected by the assembly leaders, he advanced a stronger demand for special group rights. That is, he demanded special constitutional protection of the social, political, economic and religious security of the ethnic communities.¹⁰⁰ No such arrangement was made in the original constitution to protect the rights of the ethnic communities; rather, the Constitution has adopted the individualistic model of rights, with commitment to equality and non-discrimination.¹⁰¹

On the dawn of long-standing political demand for Indigenous recognition,¹⁰² the government of Bangladesh recently brought a constitutional reform. In 2011, the Constitution was amended to introduce a new provision – article 23A – that obliges the government to protect and develop the unique local culture and tradition of the non-Bengalee communities.¹⁰³ Therefore, the Constitution now provides for ‘cultural rights’, but the striking point is that such rights have been included in the form of principles (directive) and not as fundamental rights, as originally demanded by Larma.¹⁰⁴ The implication of this seems profound. This is because, in the context of the Bangladesh Constitution, fundamental rights have stronger justiciability than the fundamental principles (directives) which are ‘contra-judicative’ in nature.¹⁰⁵ From this perspective, article 23A implies only a ‘weak’ form of constitutional protection of cultural rights. There is, however, an option to consider whether article 23A constitutes a ‘moral tool’ that promotes constitutional ‘negotiation’ among culturally diverse people.

This point was picked up by Tarunabh Khaitan, who sees article 23A as an example of political constitutionalism.¹⁰⁶ While writing for an entirely different purpose (that is, to offer a perfectionist justification for constitutional directives), Khaitan emphasizes the importance of article 23A in terms of promoting ‘constitutional polyvocality’, which he builds on the Jacobsohnian balance of ‘disharmony.¹⁰⁷ According to him, article 23A illustrates constitutional directives’ ability to allow ‘unresolved contestations over identity to be reflected in the constitutional text, thereby endorsing value pluralism’.¹⁰⁸ Interestingly, such an argument corresponds to the demands of the ethnic communities, which feels a moral or strategic need to continue a political dialogue. For example, Raja Debasish Roy, an influential intellectual figure from the ethnic communities in Bangladesh, argues that without ‘even effective negotiations with governments and others, meaningful autonomy will remain as elusive as ever’.¹⁰⁹

What is striking about such an approach is that it seeks to institute the space of constitutional dialogue within the framework of constitutionalism. We may find a similar argument in the work of James Tully, who emphases constitutional negotiation as a precondition of cultural accommodation. But Tully disagrees that modern constitution can offer such space. In his exemplary work, Strange Multiplicity: Constitutionalism in the Age of Diversity,¹¹⁰ he argues that modern constitutionalism is inherently homogenizing, and is unsuitable for the accommodation of diversity. In order to accommodate diversity, the post-imperial constitutional project must acknowledge and act upon the premise that the constitution is ‘not monologue’ but rather a ‘dialogue’ between different groups.¹¹¹ In other words, the accommodation of diversity presupposes the ‘constitutional negotiation’ of diverse groups. As he argues:

A contemporary constitution can recognise cultural diversity if it is conceived as a form of accommodation of cultural diversity. It should be seen as an activity, an intercultural dialogue in which the culturally diverse sovereign citizens of contemporary societies negotiate agreements on their ways of association over time in accord with the conventions of mutual recognition, consent and continuity.¹¹²

The polyvocality argument that Khaitan situates in relation to article 23A aligns with Tully’s invocation for constitutional dialogue, despite their distinctive approach to modern constitution.¹¹³ Conceiving article 23A as a political formula of negotiating difference, Khaitan sets out to defend ethnic difference as a moral site of political mobilization. Seen in this light, Khaitan’s reading of article 23A reflects accommodationist (or multiculturalist) approach, endorsing the logic of identity politics or what Charles Taylor advanced as the politics of difference.¹¹⁴

Such logic of difference comes to be more problematic if it is emphasized for the purpose of legal constitutionalism as opposed to political constitutionalism. The advocacy for legal constitutionalism can be found in the account of those who seek to protect ethnic minority under the legally enforceable regime of group-differentiated rights. For example, Shahabuddin argues that the omission of any specific guarantee for minority rights in the Bangladesh Constitution reduces the minority groups to the position of individual citizen, thereby assimilating the minority identity into the dominant culture.¹¹⁵ He contends that the ‘liberalist-individualist’ architecture of rights is responsible for this problem because ‘the individualist notions of equality and non-discrimination are not merely inadequate for minority protection but are indeed the modus operandi of assimilation and the extinction of group identity’.¹¹⁶

Following this argument, Shahabuddin identifies how this assimilative project was reinforced by the court’s adherence to the individualist principle of equality and non-discrimination as enshrined in the Constitution of Bangladesh.¹¹⁷ The case of Mohammad Badiuzzaman v. Bangladesh and Others ¹¹⁸ has been used as an example in this regard. In that case, the CHT Regional Council Act of 1998 was challenged on the ground that such a special arrangement violated fundamental rights to equality and non-discrimination.¹¹⁹ The court upheld the framework of equal rights against the idea of special group rights while indicating the need for political settlement towards progressive and innovative constitutional reform. This signifies, for Shahabuddin, the court’s failure to offer ‘legal approval’ to measures necessary for peacebuilding in the CHT region.¹²⁰ In response, Shahabuddin argues for a difference-sensitive arrangement of rights which ‘requires that group rights be accommodated in one form or the other’.¹²¹ Patten provides a useful account of such rights-arrangement responsive to differentiated citizenship:

a politics of difference extends to all citizens a basic package of standard liberal rights, plus a set of difference- sensitive policies designed to reach out to members of cultural minorities and provide acknowledgement, accommodation, and assistance to their ways of life. The difference model does not abandon the idea of individuals as autonomous seekers of their own conceptions of the good, but it adds to this view of individuals the idea that they are also bearers of a cultural identity that they do not share with all other citizens.¹²²

If this ‘difference model’ is what Shahabuddin has in mind, then his critique of a liberal-individualist framework of equal rights provides no good reason to believe that his position was anti-liberal.¹²³ This is because such an account of differentiated-rights has successfully been generated by the liberal-multiculturalists, advocating for reconstructing the liberal concept of rights¹²⁴ and, more radically, the idea of equality itself, as we will see shortly. So, the problem here is simply not that Shahabuddin’s account of differentiated rights offers no greater protection than does the liberal-individualistic framework of rights. Rather, the major problem of his argument lies in the following aspects: first, he fails to recognize the republican promise of Bangladesh Constitution and the ways in which such promise characterised its arrangement of rights;¹²⁵ and second, and most importantly, he subscribes to the danger posed by multiculturalism, which can be described by taking note of the following argument from Habermas:

in the case of multiculturalism, discrimination takes place within the framework of a broadly legitimate constitutional state and takes the more subtle form of domination by a majority culture that has merged with the general political culture. However, against Charles Taylor’s communitarian proposal, I argue that a ‘politics of recognition,’ which is supposed to ensure the equal right of different subcultures and forms of life to coexist within a single republican polity, must reject collective rights and survival guarantees.¹²⁶

Habermas’s response to the multiculturalist argument stems from his republican justification for equal citizenship, which entails the concept of equal rights. Therefore, it raises the legitimate question of whether such an idea of equality denies difference.¹²⁷ Such debate around equality and difference forms a recurring theme across feminism, race theory and, more generally, discrimination theory.¹²⁸ For the purposes of this article, it seems useful to respond briefly to some charges made, especially by the accommodationist critics of equality. Critics coming from the difference theorists mainly attack the idea of formal equality in which ‘equality is equated with sameness’.¹²⁹ For example, Cristine Littletone charges that such sameness-based vision of equality is an exercise of power by and for a white male elite.¹³⁰ Marion Young, while advocating for group-differentiated rights, advances her argument on the ground that equal citizenship institutes sameness: ‘citizenship for everyone, and everyone the same qua citizen’.¹³¹ If equality is conceived of sameness, for her it adds two different meanings to the universal idea of citizenship: first, it empathizes generality rather than difference (what citizens have in common as opposed to how they differ); and second, it entails universal application of law and rules that are blind to group and individual difference.¹³² Citizenship as generality tends to enforce homogeneity while universality as equal treatment promotes oppression and disadvantages.

The implications of these claims have been profound, for they instigate reform proposals from both the liberal and non-liberal theorists. Some liberal theorists advocate restructuring the individualistic framework of rights, while the difference theorists translate their demand into the idea of ‘true equality’, which is grounded not in sameness but rather in difference.¹³³ For example, Sheila Foster offers a powerful account of such ‘true equality’ with respect to difference which sets that ‘the goal of diversity should be to affirmatively include individuals from systematically excluded and disadvantage groups’.¹³⁴ In other words, true equality would enable diversity to erase the negative aspect of difference by eradicating institutional processes that translate into perpetual and systematic disadvantages for individuals with difference. In this respect, equality with respect to difference means no more than equality with the option of affirmative action.

This jurisprudence of positive discrimination, we argue, is consistent with the republican promise of equal rights and disestablishment. Habermas recognizes the potential for a gap between norms and facts – between the formal guarantees of equal rights and their actual worth. This tension requires him to set conditions for overcoming the inherent moment of inertia and inequality and, with that, to insure fairness and stability of political dialogue. For him, such conditions can be fulfilled by taking welfare measures for economic and social equality. More crucially for our purposes, Laborde defends the idea of affirmative action as part of what she calls ‘critical republicanism’. While counteracting the demand for differential treatment along ethnic and cultural lines, she argues that socio-economic disadvantage, rather than ethnic origin, is far preferable to focus. This is because the removal of socio-economic obstacles will facilitate the integration of ethnic minority, without requiring that they may be preferentially promoted.¹³⁵ In this respect, we endorse Laborde’s view and wish to add that the logic of removing social obstacles corresponds with the idea of disestablishment that forms the ethical precondition of republican unity.

At stake in such republican understanding of equality is that it indicates a shift in the meaning of ‘equality’ itself. That is, it draws on its descriptive meaning – ‘the same’ – and operates through its prescriptive meaning. ¹³⁶ The equality in its prescriptive meaning involves the application of rightful rules to appropriate characteristics of the individual being ‘adjudged equal’.¹³⁷ In other words, prescriptive equality is comparative and relational than absolute.¹³⁸ We suggest that the republican idea of equal rights embodies both the descriptive and the prescriptive meaning of equality. In this sense, republican equality differs from Young’s understanding of equal citizenship, which builds only on the descriptive meaning of equality. In contrast, republican equality starts with equal status of citizenship but does not end with equal treatment to those with difference – it remains indifferent to ethnic difference but not to socio-economic disadvantages that originate from such difference.¹³⁹

Turning to the arrangement of the Bangladesh Constitution, we can see that it embodies the republican idea of equality as discussed above. The provision of article 28 sets out the principle of equality and non-discrimination along different categories of identity – race, caste, sex, religion and even place of birth. But the same article provides for affirmative action for the advancement of any backward section of citizens.¹⁴⁰ In a similar way, the Constitution allows the making of special provision in favour of any backward section of citizens to ensure adequate representation in the service of the republic.¹⁴¹ The court of Bangladesh has been faithful to this republican idea of equality, as in the case of Mohammad Badiuzzaman. As shown above, Shahabuddin has been critical of Badiuzzaman, and of the arrangement of rights in general, despite the fact that the provisions of equality address much of the concerns raised by the difference theorists. He fails to acknowledge that the problem of the Bangladesh Constitution is not its demos-centric architecture of rights; rather, it is the ethnos-centric construction of nationality that we need to challenge and disestablish.

Far from this promise of disestablishment, the present Constitution, with the insertion of article 23A, has tilted its balance towards ethno-centric nationalism. This turn is reflected in the shifting language of the court on the question of ethnic identity and regional autonomy. Before the insertion of article 23A, the court was reluctant to recognize the ethno-centric arrangement of rights, particularly the right to autonomy under the CHT regulation.¹⁴² However, the opposite direction towards the ethno-centric interpretation can be found in a recent case of Wagachara Tea Estate Ltd. v Muhammad Abu Taher and Others (2014).¹⁴³ In that case, the court has, by reinterpreting the status of ethnic people and their CHTRC,¹⁴⁴ recognized the status of ethnic minorities from ‘tribal or ethnic sects’ to ‘Indigenous people’. Moreover, it accepted, reinforced and recognized the ‘special status’ of CHT that makes it a ‘distinct’ region from other parts of Bangladesh as a republic, the point to which we will turn in the next section.

For now, we wish to claim that these arguments offered by the court stem directly from the language of the current Constitution, particularly article 23A. This has legitimized the promise of emphasizing ethnos as enforced by the original Constitution. It did not guarantee non-exclusion, but results in a more ‘subtle form of domination’, often in the form of what can be called ‘inclusive exclusion’.

IV. Structural reform for autonomy: the circularity of exclusion?

In the previous part, we have discussed how the architecture of rights is shifting from demos to ethnos-centric turn. This part examines the structural aspect of the Constitution with a focus on the question of regional autonomy and a special legislative council for the CHT people, a long-standing demand raised by their representative. In doing so, we engage with the two important structural models suggested by the accommodationists: ‘consociationalism’ and ‘centripetalism’.¹⁴⁵ The purpose of this is twofold. First, it shows how the demand of regional autonomy rests on the concept of differentiated citizenship, and therefore reinforces the ethno-nationalistic aspirations rather than the republican promise; second, it will help us prescribe the ways in which the Constitution can better respond to the minority demand for structural reform without compromising its republican promise.

On this point, it will be useful to briefly introduce the ideas of ‘consociationalism’ and ‘centripetalism’. The consociation model is advanced by Arend Lijphart, who advocates the establishment of an ‘ethnic federation’.¹⁴⁶ This model argues for a ‘cross-community power-sharing executive’ by which major elite representative from different communities can jointly work for conflict eradication.¹⁴⁷ However, to achieve consociation, Lijphart prescribes that three important elements need to be present: coalition – complete or concurrent;¹⁴⁸ proportionality in public sectors – legislative, executive and judiciary;¹⁴⁹ and autonomy or community self-government.¹⁵⁰ This way, consociation guarantees a society that will produce homogenous ethnic constituent units though complete ethnic representation. On the other hand, centripetalism is equivalent to words such as ‘convergence’, ‘centrism’ and ‘bringing together’. Major proponents of this model are Donald Horowitz¹⁵¹ and Benjamin Reilly.¹⁵² This model, like consociation, focuses on the institutional accommodation of minorities and how they can contribute towards a more deliberative democracy. However, the centripetal model differs from the consociation model in claiming a unique institutional design to manage democracy in a divided society by deinstitutionalizing the existing ethno-centric representation. That institutional design does not opt to ‘simply replicate existing ethnic divisions in the legislature and other representative organs’;¹⁵³ rather, it aims at depoliticizing ethnicity ‘by putting in place institutional incentives for politicians and their supporters to act towards the accommodation of rival groups’.¹⁵⁴ Along these lines, Horowitz argued for non-ethnic federalism in which powers would devolve from the central to the local authority. Having identified these differences, let us see what model is adopted in Bangladesh and where it is moving.

From the time of drafting the original Constitution, the demand for structural safeguards was pressed by the representative of the CHT people, both within and outside the Constituent Assembly. On 15 February 1972, Larma led a delegation to meet the key architect of independent Bangladesh, Sheikh Mujibur Rahman, and raised a series of demands for structural autonomy that include the establishment of a special legislative body, continuation of the offices of the tribal chiefs, and constitutional entrenchment of the CHT Regulation 1900.¹⁵⁵ It seems that such demands by the ethnic leader for a strong regional council echoed the spirit of the consociation model: to formulate an ethnic federation based on the ‘rule by the minority over itself in the area of the minority’s exclusive concern’.¹⁵⁶ These demands for structural safeguards were rejected on the ground that they were ‘parochial’.¹⁵⁷ Consequently, the ethnic aspirations for regional autonomy did not find a place in the Constitution. In contrast, the Constitution has established a demos-centric parliament while providing that ‘parliament shall consist of three hundred members to be elected in accordance with law from single territorial constituency by direct election’.¹⁵⁸

One way to make sense of this territorially distributed electoral structure is to view it as showing ‘difference-blindness’ with regard to ethnic diversities. From this perspective, it was consistent with the republican promise of the Bangladesh Constitution. But this difference-blind aspect of the constitutional structure is overshadowed by the ethno-nationalistic promise of the Bengalee state. This is why the constitutional denial of regional autonomy is often seen as part of the chauvinistic project of subsuming the CHT people through forced assimilation to Bengalee nationalism.¹⁵⁹ Such understanding is also responsible for the ongoing ethnic conflicts that began in the wake of the constitutional refusal to recognize the regional autonomy of the CHT peoples. The discontent ultimately culminated in the ‘macro-nationalist’ claim for separate nationhood to be known as the ‘Jumma nation’.¹⁶⁰ By resorting to insurgency, the hill people continued the power-sharing demands, including for regional autonomy. In response to this, a political settlement was reached and the CHT Peace Accord was signed between the CHT region and the government of Bangladesh in 1997.

This Accord provides for the Chittagong Hill Tracts Regional Council (CHTRC), giving certain regional autonomy to the communities living only in CHT regions. It mandates two-thirds ethnic representation and also guarantees an ethnic chairman.¹⁶¹ This arrangement, in particular the formation of CHTRC, has been appreciated as a ‘paradigmatic improvement in the style of indigenous people’s participation’ and ‘a means of tacitly recognizing their separate cultural and political identity’.¹⁶² If we situate the Accord in relation to the models suggested by the accommodationists, we may find it shares the features of both the consociation and centripetal models. On one hand, it provides certain forms of autonomy to a group of territorially concentrated ethnic minorities. However, in terms of representative arrangements, it provides for a cross-ethnic electoral constituency by making it subject to the centralized process of deliberative democracy.¹⁶³

Interestingly, this ‘hybrid arrangement’ provided by the Accord was challenged in the case of Mohammad Badiuzzaman on the ground that the alleged accord violated the basic structure of the Constitution. In this case, the petitioner argued:

the cumulative effect of the various clauses of the impugned acts and the creation of the Regional Council is not only that of violating various provisions of the Constitution as mentioned hereinabove, but also of destroying one of the basic structures of the Constitution, namely, the unitary character of the state by supporting to create a territorial unit which eventually may claim the status of a federating unit.¹⁶⁴

This argument involves the court deciding on the question of structural reform for regional autonomy. Focusing on the regional aspect of the CHTRC, the court accepted the petitioner’s view. In its judgment, the court was satisfied that promoting the interests of people living in a particular territory would destroy the very fabric of the unitary republic. Moreover, the court emphasized the constitutional principle of equality and non-discrimination principles, as shown in the previous part. The heavy implication of such reliance is that the court did not aspire to accept any constitutional structure that may potentially threaten the republican nature of Bangladesh as a state. Instead, such an approach of the court endorses the ‘difference-blindness’ which is consistent with the concept of republican nationalism that we are defending in this article. However, coming from an accommodationist camp, Ridwanul Hoque criticizes such approach of the court:

In its analysis, the Badiuzzaman court failed to appreciate that indigenous self-determination in a unitary State such as Bangladesh can be achieved through ‘a range of possibilities of institutional re-ordering other than the creation of new states,’ or without undermining state sovereignty’. The CHTRC and the transfer to the CHT district councils of certain indigenous-specific powers are definitely innovations of the type that is within the fold of a unitary constitutional order.¹⁶⁵

For Hoque, the CHTRC and the new model of enhanced Indigenous participation through the district councils could alternatively be seen as an innovative institutional reordering. However, according to him, this ‘innovative’ arrangement should have been approved from the normative perspective of inclusive constitutionalism, ensuring both the recognition and participation-based inclusion of the CHT people. A similar argument is advanced by Mohammad Shahabuddin, who argues that having confined itself to the state’s unitary character, ‘the court failed to offer legal approval to measures that parties to the Accord accepted as crucial elements of peacebuilding in the region’.¹⁶⁶ His criticism was therefore directed against the structure of the Constitution that ‘conceives of Bangladesh as a unitary national state, thereby limiting the scope for accommodating ethnic differences’.¹⁶⁷ From this perspective, the court’s disapproval of the consolidation of indigenous self-governance was nothing but the negation of the ‘rival approach [that] recognizes the right of tribal people[s] as distinct peoples’.¹⁶⁸

This brings us to the question of whether the approval to the regional council for the CHT people would then promote the politics of difference? If so, could it solve the problem of exclusion along ethnic lines? One interesting way of responding to this question is to view the operation of the regional council as reproducing the ethnic minority within its own framework. For example, the Bengalee people living in the CHT region may be viewed as an ethnic minority if seen from a demographic point of view. Even the member of a non-Bengalee community may find that they are in a marginalized community if they do not belong to the ethnic sect having majoritarian control in the political, cultural and economic activities of the region.¹⁶⁹ Therefore, regional autonomy would not resolve the problem of exclusion; rather, it would create the circularity of exclusion by producing a ‘minority within a minority’. In other words, Bengalee people and other ethnic sects would form new minority groups in the face of a new majority, where ethnicity would never be a vanishing point. Interestingly, the idea of the Jumma nation has already replicated that same logic and the risk of ethnic exclusion among them that they themselves confront in the name of ethnic difference.¹⁷⁰ In addition, there are other limitations associated with the concept of regional autonomy – that is, it does not include the autonomy of small ethnic communities living on the plain land.¹⁷¹ Seen in this light, the demand for regional autonomy contradicts its own logic of non-exclusion.

This circularity of exclusion along the ethnic line helps us understand the limits of ‘ethnic federalism’ as advanced by the advocate of consociation model. The problem with this model is that it validates the use of ethnicity not only as a morally superior value but also as a subject of political mobilization. The implication of this may be counter-productive. It may be detrimental to the condition of Indigenous people, making them more vulnerable than before. An example of this can be found in the case of Malaysian federalism, as explored by Andrew Harding. By carefully analysing the power, structure and intensity of the newly created autonomy (under the tenets of traditional federalism) of Malaysian Indigenous peoples (Sabah and Sarawak), Harding has demonstrated how ethnic federalism in Malaysia brings discontent and division to an ethnically divided society. He concludes that ‘their status within the federation (both constitutional and political), and federal interventions in state politics that have eroded state autonomy’.¹⁷² Thus, the demand for regional autonomy in the form of ethnic federalism is unable to resolve the problems of exclusion by the means of structural reform.

It does not, however, rule out the possibility of non-ethnic federalism advanced by Horowitz. For Horowitz, federalism is not only a means of distributing power among majority and minority peoples, but also an effective way of removing absolute ethnic majority in a particular region.¹⁷³ This is why non-ethnic federalism is particularly attractive compared with the more overtly consociational features conventionally proposed.¹⁷⁴ As we suggest, Horowitz’s argument about federalism comes closer to the republican form of federalism supported by Habermas. As acknowledged by Habermas, federalization may work as a possible tool to safeguard cultural autonomy by decentralizing state power.¹⁷⁵ Unlike ethnic federalism, it does not require the virtue of formally designing state institutions on the basis of ethnic identities. Such a de-ethicized model of federalism is closely linked to the concept of difference-blindness. Insofar as its structural reform is concerned, Bangladesh can therefore adopt this strategy of non-ethnic federalism.¹⁷⁶ This choice is crucial because it will settle what the Constitution ultimately prioritizes: ethnos as demanded by the advocates of difference or demos as entailed by its vision of becoming a republic.

V. Conclusion

This article argues for de-ethnicizing the Constitution as a response to the problem of ethnic exclusion in Bangladesh. By using the ethnos–demos binary as an explanatory framework, we argue that the real problem of the Bangladesh Constitution lies not in its structural elements or in the architecture of rights – both of which are based on the idea of demos as opposed to ethnos. Instead, it is the ethno-based formulation of identity that sets out its exclusionary paradigm by transforming demos into ethnos. Therefore, a truly inclusive constitution in Bangladesh would entail reversing such logic. Towards that end, we offer the following suggestion. For the identity part, we call for a complete disavowal of Bengalee nationalism and its replacement by the idea of Bangladeshi nationalism; for the rights part, we seek status quo ante; and finally, for the structure part, we only need the status quo of the original constitutional arrangement or, at best, the adoption of non-ethnic federalism.

This demos-centric prescription has been defended in this article not only against conservative ethnonationalism but also against the accommodationist (or multiculturalist) approach of constitutionalism. On one hand, it offers a critical foil for the ethno-nationalistic paradigm of the Bengalee state that amounts to exclusion in the name of assimilation. On the other hand, it distances itself from the accommodationist approach that confronts ethnic assimilation by relying on the politics of difference. Despite sharing similar concerns about assimilation, we do not agree with the accommodationist claim that difference-blindness amounts to forced assimilation and exclusion. Such an understanding is too simple to do justice to the integrationist potential of a difference-blind constitutional arrangement. This is what we take as the point of our departure: we defend the republican promise of difference-blindness. As we believe, it is the vision of the Bengalee state – not the republican arrangement of rights and structure – that stands between us and a truly inclusive constitution in Bangladesh.