Civil society and the contemporary threat to religious freedom in Bangladesh

Against the international backdrop of rising religious tensions, this article explores contemporary civil society views on religious freedom in Bangladesh. It uses critical frame analysis of the corpus of civil society organizations’ (CSOs) submissions to the United Nations’ third cycle Universal Periodic Review (UPR), 2013–18. It provides a timely assessment of Bangladesh’s fulfilment of international obligations on religious freedom, and shows how the politicization of religion and the resultant conflict between ‘secularism’ and ‘extremism’ have been fuelling inter-communal tensions and religious intolerance. In particular, CSOs’ UPR submissions present powerful accounts of the principal human rights pathology affecting the country today, religious-based violence. This is accompanied by a narrative of police malpractice, judicial failings, discrimination, oppression and incitement. A further key finding is ‘situated knowledge’ or first-hand accounts of legal restrictions and government repression of civil society organizations. Consonant with the classical work of liberal theorists, we argue that unprecedented importance now attaches to safeguarding civil society criticality in order to defend religious freedom and uphold human rights in the Republic.


Introduction
Bangladesh is a country with marked faith-based tensions. These have been recently thrown into stark relief following the Fifteenth Amendment to the Constitution in 2011 that confirmed Islam as the State religion of the Republic. This has supported the rise of a range of Islamic extremist groups, led to religious radicalization and 'attacks and threats against religious minorities' (HRW, 2018, p. 2). Whilst communal tensions are nothing new, of late, they have escalated to unprecedented levels. The reasons for this are not just weak governance or the absence of a strong state, but also the clash of secular and religious ideologies. As Khan (2017, p. 191) argues, Bangladesh 'is currently facing an incremental growth of radicalization … Over time, political violence, ideological clashes between secular and right-wing ideologies, and weak governance have created conditions for the growth of radical Islam'. This rise of radical Islam has threatened constitutional values such as democracy and secularism. 1 Furthermore, it has also violated the country's obligations under a raft of UN human rights treaties that uphold the principle of religious freedom.
Taking this into account, this article provides a timely assessment of the issues and challenges in fulfilling Bangladesh's international rights obligations on religious freedom. It uses discourse analysis of the corpus of civil society organizations' (CSOs) submissions to the third cycle (2013-18) Universal Periodic Review (UPR), the United Nations' (UN) official five yearly monitoring mechanism. Specifically, it explores the 'issue-salience'or, level of attention afforded todifferent human rights 'pathologies' 2 (in other words, rights violations and implementation failings) related to the upholding of religious freedom identified by CSOs in their UPR submissions. Furthermore, in order to provide an in-depth understanding of inherent meanings, lived experiences and sentiments around contemporary religious freedom issues, it uses textual analysis to examine CSOs' use of languageor, 'framing' in the discourse.
Such a methodology emerges fromand, at the same time helps to inform us ofthe challenges posed by seeking to apply the universalist norms underpinning UN rights agreements in specific governance contexts with their own cultures, histories and governing traditions. In short, this 'governance view' of rights implementation reflects the reality that local human rights practices are contingent and place-based. In the case of Bangladesh such specificities include governmentalitynot least, periods of authoritarian rule, party politics and political patronage. As the following analysis reveals, such factors underpin a series of contemporary pathologies such as restrictions on civil society, corruption in public administration (in turn, contributing to authorities' failure to uphold religious freedom including police malpractice) and government failure to respond to earlier UPR recommendations. In addition, inequalities and discrimination in the pluri-religious state explain the prevalence of violence, threats and incitement.
The current 'governance view' of rights implementation (meaning that we not only examine the exercise of executive power and public administrationbut also civil society perspectives) is consonant with the classical work of liberal theorists (Hobbes, 1651;Rousseau, 1968), as well as later thinkers on the role of civil society (e.g., De Tocqueville, 1835Tocqueville, /1956). This is apposite because key importance attaches to civil society criticality in order to uphold religious freedom and address prevailing human rights pathologies.
Accordingly, the remainder of the article is structured thus: the next section provides a theoretical discussion on the relationship between religious freedom, secularism and civil society. This is followed by an outline of the research context; attention then turns to the study methodology. The study findings follow -first, in relation to the nature and salience of different pathologies, and second, CSOs' use of critical framings. The implications of the analysis are outlined in the conclusion.

Religious Freedom, Secularism and the Civil Society
Religious freedom is a complex and contested concept (Cf. Sullivan et al., 2015); one that is considered central to upholding values of secularism. The latter is a broader concept that originated with the idea of the separation of the church from the state or the immanent from the transcendental (see Taylor, 2011). Over the years, scholars have defined it in different ways. For example, advocates of secularization theory define it in terms of privatization of and/or decline of religion (see Berger, 2009). Calhoun (2011, p. 22), however, argues that defining secularism as religious privatization is misguided, because 'religion has never been essentially private'. Casanova (2011) too argues that the secular and the religious are mutually constitutive and there are multiple ways of experiencing the secular. 3 As a result of these interpretations, different countries have adopted contrasting models of secularism. Compared to the European 4 and American context where secularism refers to the 'wall of separation' 5 ; in the Indian context it referred to dharma nirapekshata (religious neutrality) and sarva dharma samabhava (religious equality). The latter terms broadly mean 'twin tolerations' (see Stepan, 2011), or what Bhargava (2011) has called 'principled distance' which refers to peaceful coexistence of different religions that are equidistance to the state. The Bangladeshi context is similar to India for secularism refers not to the separation of religion and politics, but the peaceful coexistence of diverse religious communities and the exercise of religious freedom.
Secularism thus creates the essential conditions for religious freedom in a particular political context. Whereas in religious thinking there is an emphasis on the idea of 'community', in secularism primacy is afforded to individuals and citizenship rights. According to this view it is the fundamental duty of the state to ensure and protect these individual rights, especially the right to religious freedom. Although religious freedom can mean many things, in this article we define it as tolerance of different theological systems of belief (and, crucially, non-belief), and freedom of individual action to follow one's own beliefswhether faith-based, atheist or agnostic. As Sajó and Uitz (2012, p. 136) explain, 'the formulation of freedom of religion and conscience as an individual right stems partly from placing religious choices in individual conscience … In a legal sense it is the scope of protection a legal system is willing to provide to individuals and groups claiming that their religious liberty has been restricted'.
In contrast, religious extremism aligns itself with the notion of religious particularism (Glock & Stark, 1966). While religious particularism may not always result in extremism or violence, what is observed is that religious particularism is often associated with intolerance of 'other' viewpoints. It denotes 'a belief in the exclusive authenticity of one's own religious tradition. Religious particularists are likely to believe that salvation is confined to rather narrowly defined groups of believers' (Swatos, 2003, p. 79). 6 Furthermore, extremism is often driven by a sense of past injustice and a desire for revenge.
In the context of rising religious extremism, civil society can play a positive and constructive role in fostering tolerance between communities and religious traditions (Freedman, 2009). However, first we need to consider the questions 'what is civil society' and 'how can it advance religious freedom'? As with secularism and religious freedom, conceptions of civil society also vary widely (Chaney, 2019, p. 532;Jensen, 2006, p. 39;Sahoo, 2013, p. 24). 7 While liberal theorists like De Tocqueville define civil society as a strong, vibrant associational life that acts as the watchdog of democracy, neo-Marxist scholars like Gramsci see it as a site of hegemonic and counter-hegemonic struggle (see Kumar, 1993). Other scholars such as Habermas define it as the 'public sphere' where private individuals transcend their ascribed identities and come together as a public and form an interactive body of citizens engaged in a rational critical discourse (see Calhoun, 1992). In this article, we follow the liberal, De Tocquevellian view of civil society as associational life wherein CSOs generally work to uphold democracy. According to this view, civil society is neither governed by the logic of acquiring state power (like political parties) nor by the logic of maximizing profit (as with private enterprises following market forces). In contrast, civil society's objective is to act as a 'sphere of solidarity' (Alexander, 1997) and protect public interest and democratic values. When the state tries to violate or undermine such values, civil society mobilizes to remind the state of its obligations. However, in authoritarian or unaccountable state contexts, civil society is often oppressed and the state ignores its commitments to democratic values and citizenship rights. 8 At present, this is primarily the case in Bangladesh where the state undermines pluralism and the diversity of civil societyand ignores its international commitments to religious freedom and rights of minorities. As the UN Office of the United Nations High Commissioner for Human Rights warned: 'the Government should protect the vibrant civil society and pluralistic society in Bangladesh by addressing the existing anxiety expressed by members of religious minorities and indigenous peoples who feel vulnerable in the face of rising religious extremism and acts of violence perpetrated against them … civil society should continue to claim its space, with the support of international stakeholders, to express dissenting views and voice concerns as appropriate' (OHCHR, 2016, p. 20, para 104).

Research Context
This section discusses the historical background and contextualizes the study's findings. It has two parts. The first ('Background'), deals with religion and demography, followed by an overview of the interplay of politics and religion in the latter part of the twentieth and early twenty-first Century. The next part ('Current Challenges') discusses the political turmoil, the rise of extremism and challenges facing civil society.

Background
The last official state-wide census 9 revealed that, of Bangladesh's 152.5 million people, Sunni Muslims constitute 90% of the population and Hindus make up 9.5%. The remainder is predominantly Christian (mostly Roman Catholic) and Theravada-Hinayana Buddhist. Other minorities include, Shia Muslim, Bahais, animists and Ahmadiyya Muslims. Estimates of the latter vary widely (Hasan, 2017, p. 314). In the wake of the post-2015 humanitarian crisis (for a discussion see Mahbubul Haque, 2017), there are approximately 30,000 registered and up to 600,000 unregistered Muslim Rohingya refugees in the southeast around Cox's Bazaar.
In terms of history, the Bengali-speaking Muslims of East Pakistan revolted against the Urdu-speaking Muslims of West Pakistan and created the new nation of Bangladesh. Soon after independence in 1971, this ethno-linguistic nationalism ('we are Bengalis, we want a nation of Bengali-speaking people') shifted towards religious nationalism (Karim, 2005, p. 5), manifesting itself in increasing Islamization. Although there had always been a tussle between the secular and the religious forces, the Bangladeshi state in the early years of independence maintained a strong secular orientation. However, things changed quickly and the state secularism came under pressure partly because of 'foreign policy compulsions'economic aid from oil-rich Islamic countries forced the founding leader Mujibur Rahman to embrace the views of Islamic religious forces (Datta, 2002, p. 146). As a consequence, 'despite the official ban on religious political parties, Mujibur's secular socialist policies led … to the introduction of "multi-theocracy in the name of secularism"' (Datta, 2002, p. 146).
Subsequently, as Siddiqi (2006, p. 2) notes, there was 'a gradual but sustained mainstreaming of Islam in public political life, in the representational practices of the state as well as in national policy and constitutional principles, began in the mid-1970s … By the mid-1990s, Islamic symbols and idioms had become part of everyday political vocabulary'. This transitional period began in 1975 with the assassination of Sheik Mujib, the country's first political leader. For the next decade-and-a-half the political landscape was dominated by leaders drawn from the army: General Ziaur Rahman (1975-81) andGeneral Hossain Ershad (1982-90). In order to legitimise their hold on political power both offered political concessions to potential allies. Moreover, they quashed political dissent, and at the same time introduced constitutional reforms. These actions gave greater legitimacy to the application of Islamist principles in public life. This culminated in 1977, with the government of General Zia that brought in a new constitution; one that removed the commitment to secularism 10 and emphasized the Islamic legacy of the nation. During this period the ban on religious-based political parties was also removed. In a further step-change, in 1988, General Ershad amended the constitution and made Islam the official state religion.
In 1990, a mass movement led by the main opposition parties in association with the principal Islamist political forcethe Jamaat-e-Islamiforced President Ershad out of office. With the restoration of democracy in 1991, it formed a ruling coalition with the Bangladesh Nationalist Party (BNP). 11 Throughout subsequent years Islam has continued to be an important aspect of mass politics with the two largest political parties (Awami League and BNP) keen to highlight their Islamic credentials and uphold Islam as an integral part of political life and national identity. Although this intermixing of religion and politics resulted in several instances of violence, Bangladesh, up until early 1990s, was relatively free from communal violence. However, this changed significantly when: The country witnessed the first major outbreak of communal violence in November 1990 following tensions in India in the wake of the Ayodhya controversy. [Later …] attacks on Hindus increased dramatically after the 1992 Babri Masjid destruction incident. 12 The growth of religious tension in India had a spill over effect and Bangladesh for the first time witnessed large scale violence perpetrated on the Hindus. (Datta, 2002, p. 316)

Current Developments
At the turn of the twenty-first century, Bangladesh witnessed increasing religious violence and radicalization of youth, especially in the form of armed militants. Scholars have noted that madrasas may have played a major role in this regard. As Huque and Akhter (1987, p. 204) note, 'the madrasa is one of the principal educational institutions contributing to the process of socialization into the Islamic way of life'. The background to this lies in the period 1991-1996, during which, as Rahid (2012, p. 25) observes, 'there was remarkable expansion of Madrassa education (religious education) with state support. The growing number of unskilled Bangladeshi workers in the Middle Eastern countries made Bangladesh's economy increasingly dependent on foreign remittances. At the same time, Saudi funding to various Islamic NGOs began to proliferate in Bangladesh starting their operation'. Bhardwaj (2011, p. 3) concurs with this view and notes, to understand 'the history of the Islamisation of Bengal, [it is] essential to understand the roots of Islam in Bengal … the resurgence of Islam after the 9/11 attacks in the US, and the growth Middle East-funded Madrasa schools are important factors'. For its part the UN has observed that, While perhaps half of the madrasas follow the national curriculum, a large number of Qaumi madrasas, with their own curriculum continue to operate outside of the Ministry of Education. Members of religious minorities have frequently expressed their fears that the madrasas, in particular Qaumi madrasas, promote extremist views, such as stigmatizing all non-Muslims as "infidels". The spread of madrasas, in particular those not operating in line with the national curriculum, seems to be a main source of anxiety among religious minorities. (OHCHR, 2016, para 69) In addition to the impact of the madrasas, there has also been continued political party support to extremist groups. For example, Khaleda Zia 13 included the Jamaat [-e-Islami Bangladesh] in her cabinet, which some scholars referred to as 'playing with fire' (Lintner, 2004, p. 433). Subsequently, as Riaz (2008, p. 33) explains, 'Islamists in general and Jamaat-e-Islami in particular, are exercising influence beyond their support base in society, and have been forcing the majority onto the defensive'. However, in early 2010, Sheikh Hasina's support to secularists brought her into conflict with Islamic extremists.
Notwithstanding this tumultuous political backdrop, in 2011, the Fifteenth Amendment to the Constitution 14 (the version that is valid today) confirmed Islam as the State religion of the Republic. Yet crucially, in a more liberal vein, it also gave legal recognition to the fact that other religions may be practised in peace and harmony (Article 2A). Moreover, the revised Constitution asserts the aim of eliminating communalism in all its forms. However, in a most contradictory fashion it proscribes the granting by the State of political status in favour of any religion and the abuse of religion for political purposes. It also deems discrimination against, or persecution of, persons practising a particular religion unconstitutional (Article 12). Whilst it might constitute a more pluri-religious framework, the confirmation of Islam as the official state religion means, contrary to some claims (Salehin, 2013); it does not amount to restoration of secularism. With hindsight, perhaps this is unsurprising. As Anis (2015, p. 17) reflects: secularism has had a beleaguered existence in Bangladesh. For many it has never meant a full separation of religion and state, but rather a peaceable side-by-side existence of different religious communities. In a region wracked with horrific communal violence, this is no small concern. As a result, secularism has never really meant the freedom to think or speak on religious matters.
More recently, the UN Human Rights Council echoed this point: some of the measures taken by the Government in the interest of upholding secularism seem to lead to the paradoxical result of shrinking the very space that secularismlike democracy is supposed to provide … In other words, while the Government may be fighting the instrumentalization of religion, it could at the same time be seen as using religion to achieve political goals. This may erode the credibility of the Government's profession of inclusive secularism. (OHCHR, 2016, p. 8, para 35) The reality is that the Fifteenth Amendment does not represent an abrupt shift from a hitherto secular state. Rather, it reflects the latest chapter in Bangladesh's history whereby, 'Islam and politics are inextricably interrelated and intertwined. The amalgamation of Islam with politics, which is in other words called Islamism or political Islam, has been a characteristic feature of Muslim society since the dawn of Islam's culture and civilization' (Islam & Islam, 2018, p. 17). As Hasan (2012, p. 69) explains, 'the complexity of the secular-religious problem that Bangladesh has faced since its establishment … is a national problem, with a national context that has become global'.
However, aspects of this historical narrative sit uncomfortably with the Republic's international treaty obligations. As the OHCHR (2016, p. 4, para 11) notes, 'Bangladesh is party to almost all core human rights treaties'. These include the International Covenant on Civil and Political Rights (ICCPR, acceded to in 2000) and the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD, acceded to in 1979). Article 18 of the former states, 'Everyone shall have the right to freedom of thought, conscience and religion. This right shall include freedom to have or to adopt a religion or belief of his [sic] choice, and freedom, either individually or in community with others and in public or private, to manifest his religion or belief in worship, observance, practice and teaching'. Whilst Article 5 of the latter asserts, 'the right to freedom of thought, conscience and religion'. 15 Despite Bangladesh's (albeit ambiguous) constitutional guarantees and international treaty obligations on religious freedom, a fundamental shortcoming is that protection under domestic law is limited. For reformers it was hoped that the Discrimination Bill 2015 would change matters, but it has yet to be enacted, leading to criticism from the UNHRC. It asserted that, 'the State party should ensure that the anti-discrimination bill, 2015, protects against direct and indirect discrimination in the public and private sphere based on a comprehensive list of grounds for discrimination, including colour, descent, caste, national or ethnic origin, religion … Furthermore, the State party should: Protect the safety and security of persons belonging to minority religious groups and ensure their ability to fully enjoy their freedom of religion and to worship without fear of attack' (UNHRC, 2017, p. 3, para 12a).
Not just failing to protect minority citizens from discrimination and violence, the state has also failed to protect the autonomy and diversity of civil society. In fact, the state has been fearful of civil society's questioning of its inability to maintain law and order and protect rights and freedom of citizens, especially ethnic and religious minorities. As a consequence, it has developed deep distrust towards civil society and prevented its involvement in 'political' and 'sensitive' affairs. Given this, civil society has become depoliticized and instead of promoting state accountability and raising issues that are vital to democratic governance, civil society's task has now generally been limited to apolitical delivery of welfare services.
Subsequently, in the wake of the political turmoil of 2015, 16 the context for CSOs has yet worsened, and the human rights agenda has been forced into the background (See Feldman, 2015;Haque & Ahsan, 2014;Mohajan, 2014;Saidul Islam, 2011). Recent accounts attest to 'civil society groups fac[ing] pressure from both state and non-state actors, including death threats and attacks from extremist groups, and escalating harassment and surveillance by security forces' (HRW, 2018, p. 2). As Husain (2016, p. 291) notes, 'it is obvious that present-day Bangladesh shows very clear tendencies of intolerance on the basis of social, political and religious identity. Such intolerance, however, is a phenomenon, which contradicts the traditions of the country spanning over a thousand years where tolerance and coexistence with different groups were cornerstones of its civilisational ethos'.

Methodology
In methodological terms, the present analysis offers a transferable discourse-based approach to studying human rights implementation. As noted, the UPR allows for nongovernmentalor, civil society organizations to submit formal written submissions to the Human Rights Council. These constitute a rich and unique dataset. Founded in the critical, 'situated knowledge' (Stoetzler & Yuval-Davis, 2002) of civil society it reveals the multiple and inter-connected issues facing human rights compliance in a manner that complements legal analyses and quantitative rights indicators.
The discourse analysis was operationalized by examining 27 reports submitted by civil society organizations for the third cycle UPR covering the years 2013-18 (see Appendix 1). The aforementioned number of CSO submissions under-reports the breadth of civil society input into the UPR for many are joint submissions authored by broad coalitions of standalone organizations (one, for example, has input from 20 CSOs, and another is co-authored by a network which is made up of over 1000+ groups and 4000+ individuals from 150 countries). 17 The discourse analysis was undertaken using a grounded theory approach (Urquhart, 2012) that combined inductive and deductive coding techniques (Fereday & Muir-Cochrane, 2006). First, in order to generate initial codes of human rights pathologies related to religious freedom, a formative exploratory reading was undertaken of the submissions. Subsequently, the texts were re-read and all references to violations and implementation issues around religious freedom and human rights ('pathologies') were copied into a database. This was then carefully and systematically coded. Not only did this reveal the nature and breadth of the (non-discrete) rights pathologies in the CSO discourse, it also showed the 'issue-salience' or level of CSO attention to (and prioritization of) the different issues raised by CSOs in the third cycle UPR corpus. During the coding process appropriate software 18 was used to divide the UPR reports into 'quasi-sentences' (or, 'an argument which is the verbal expression of one term, idea or issue', Volkens, 2001, p. 93). Sub-dividing sentences in this manner controlled for long sentences that contained multiple references to religious freedom issues. To ensure accuracy the coding was repeated by a research assistant. In the small minority of cases where coding differed 19 such instances were re-checked and a final coding agreed. All textual extracts were logged into a database that enabled descriptive statistical analysis of issue-salience and subsequent examination of framing.
Framing derives from the classic work of Erving Goffman (1974). It refers to the language used by policy actors. Effectively it is a 'schemata of interpretation' (Goffman, 1974, p. 27) that is concerned with the inherent meanings, sentiments, messages and criticality in relation to social and political communication (Heine & Narrog, 2015). As the following reveals, framing deepens the analysis for it moves beyond the level of attention to different types of rights violationor pathologyto consider the way that language is expressed (Druckman, 2001). The frames were inductively generated through systematic reading of the corpus of religious freedom extracts in the database. For example, 'attack', 'threat', 'hurt/ pain', 'injustice' and 'fear'. The level of attention to the different frames in the corpus was determined by content analysis of electronic versions of the CSOs' UPR submissions (Krippendorff & Bock, 2008).

Contemporary Pathologies in Upholding Human Rights and Religious Freedom
In this section the study findings are presented in relation to the pathologies highlighted in the civil society organizations' UPR discourse. These are summarized in Table 1. Each is discussed under one of eight sub-headings below.

Religious-based Violence
Reflecting a wider global trend (Chaney, 2019;Juergensmeyer, 2003;Lewis, 2011), religious-based violence is the lead pathology in the CSOs' third cycle UPR discourse, constituting almost a third (31.2%) of all quasi-sentences. Notwithstanding the UN's Second Cycle 'concern at the losses of life, injuries and damages to religious sites and property in the most recent confrontations', and official recommendation that the government 'ensure the effective investigation and sanctioning of all cases of violence against religious minorities' (UNHRC, 2013, para 73 and recommendation 129.93), as the following analysis reveals it remains the principal rights violation. Its cause lies in the politicization of religion and the resultant opposition of 'secularism' and 'radical extremism' fuelling longerstanding inter-communal tensions and religious intolerance. The CSO discourse outlines how violence against religious minorities, especially, Hindu priests, Christians, Buddhists and Ahmadiyas has increased significantly over the last few years. Moreover, it shows how the state has failed to listen to civil society and protect the human rights defenders. In consequence, several working to protect of freedom of religion have been killed (Human Rights Forum Bangladesh, 2017, p. 4). The CSO discourse also explains how some groups are more susceptible to violence than others. For example, reports note that children have often been victims; 'they have been subjected to various forms of violence, including killings, torture and rapes not only at homes but also at schools, communities and work environments. Furthermore, it is observed that children from ethnic and religious minority communities are more vulnerable to violence and torture than others' (Child Rights Advocacy Coalition in Bangladesh, 2017, p. 5). Despite the prevalence of such violence, the Government of Bangladesh has taken little effective action against extremists. Blogging has found itself at the heart of the rise in religious violence. Several bloggers who had written critical posts on Islamic religious extremism have been killed. For example, Rajib Haidar, a 30-year-old architect who wrote seeking a ban on Jamaat's fundamentalist politics, was stabbed to death in February 2012. Others have suffered a similar fate. Considering this, a significant strand of the CSO discourse underlines the 'democratic pathology' of the violence. For example, one submission observed that 'self-censorship in the media is growing as a result of the endemic violence against journalists and media outlets, and the systematic impunity enjoyed by those responsible' (Reporters Without Borders, 2017, p. 3). Whilst another opined, 'the killings of bloggers and publishers who wrote on issues of religion, secularism, atheism, justice for war crimes, and sexuality, by Islamic fundamentalist groups, have significantly contributed to a climate of fear and self-censorship among secular writers and in the broader artistic communitythis fear contributes to the shrinking space for civil society' (Freemuse, Drik, PEN International and PEN Bangladesh, 2017, p. 5, emphasis added).

Restrictions on Civil Society
Notwithstanding the Second Cycle UPR Recommendation that the Government of Bangladesh 'take steps to ensure that civil society can operate without intimidation' (UNHRC, 2013, recommendation 129.103). 'Restrictions on civil society' was the second-most prominent pathology in the discourse (17.1% of quasi-sentences). It can be seen as part of a wider international trend as governments have used legal and coercive means to undermine the independence of civil society (e.g., Chaney, 2018aChaney, , 2018bJefferson & Jensen, 2009;Thornton, 2013). One of the key instruments in the current government clamp down on civil society is Section 57 of the Information and Communication Technology Act (2006), which has been frequently used to detain and silence dissidents on flimsy grounds. It states: If any person deliberately publishes or transmits or causes to be published or transmitted in the website or in electronic form any material which is fake and obscene or its effect is such as to tend to deprave and corrupt persons who are likely, having regard to all relevant circumstances, to read, see or hear the matter contained or embodied in it, or causes to deteriorate or creates possibility to deteriorate law and order, prejudice the image of the State or person or causes to hurt or may hurt religious belief or instigate against any person or organization, then this activity of his will be regarded as an offence. 20 Moreover, repressive amendments to the Act in 2013 make alleged breaches of Section 57 'non-bailable and non-referable' offences, raise the sentence for offences from 10 to 14 years, and grant law enforcement officials the power to arrest without a warrant. Unsurprisingly, much of the CSO discourse is coruscating in its criticism of Section 57. Many highlight the vague nature of the notion of 'hurt to religious sentiment'. For example, one CSO noted that, 'Laws that make it an offence to hurt religious sentiment have a stifling impact on the enjoyment of freedom of expression and religion and limit a healthy dialogue and debate on religion'. This account proceeds to underline how the domestic legal code breaches the country's international treaty obligations: 'Prohibitions of displays of lack of respect for a religion […] are incompatible with ICCPR, except in the specific circumstances envisaged in article 20, paragraph 2, of the Covenant', according to the Human Rights Committee's General Comment No. 34.52 Section 57 of the ICT Act does not meet the criteria for limitation, as stated in article 20 paragraph 2 of ICCPR' (Freemuse, Drik, PEN International and PEN Bangladesh, 2017, p. 14). This shows that, from a civil society perspective, the purpose of such laws is seen as being to silence critics and undermine religious freedom. In another example, a CSO complained that, 'members of [… a named] human rights organization were charged under Section 57 of the Information and Communication Technology Act for publishing a report on alleged extrajudicial executions by security forces during a protest rally in May 2013' (Amnesty International, 2017, p. 2).

The Human Rights Incompatibility of Current Laws
Allied to the foregoing, the widespread incompatibility of the Penal Code with human rights on religious freedom is the third most-cited pathology in the discourse (13.2% of quasi-sentences). The discourse is typified by the observations of one CSO: 'A variety of provisions in the Penal Code seek to protect the feelings of religious believers, including by prohibiting insults against a religion or religious beliefs as such: . Section 295 criminalizes defiling "any object held sacred by any class of persons with the intention of thereby insulting the religion of any class of persons or with the knowledge that any class of persons is likely to consider such […] defilement as an insult to their religion […]", providing for up to 2 years' imprisonment and/or a fine; . Section 295A criminalizes "outraging the religious feelings of any class of the citizens of Bangladesh" "with deliberate and malicious intention" and by insulting the religion or the religious beliefs of that class, providing for up to 2 years' imprisonment and/or a fine; . Section 298 criminalizes "uttering words […] with deliberate intent to wound religious feelings", providing for up to 1 year of imprisonment and/or a fine' (ARTICLE 19 and Research Initiatives Bangladesh, 2017, p. 8).
As the CSO discourse underlines, these provisions in the Penal Code are incompatible with Article 19(3) of the ICCPR, which does not permit limitations on freedom of expression to protect religions or belief from criticism, or to shield followers of a religion or belief from offence or insult. 21 The civil society criticism of the current legal framework is echoed by the UN Special Rapporteur on freedom of religion or belief (OHCHR, 2016, p. 17, para 87), who noted that: It is a truism that freedom of expression is not absolute and must sometimes be limited. However, the decisive point is that any limitations deemed necessary must be clearly and narrowly defined and should meet all the criteria contained in constitutional law and international human rights law. Already section 295A of the old Penal Code fails to meet these criteria … The offences mentioned in these two sections are only vaguely circumscribed and thus remain vulnerable to highly subjective invocations and arbitrary applications.
The CSO discourse was also critical of the constitution of Bangladesh. For example, one complained that it recognises the right to freedom of speech and expression, which includes freedom of artistic expression. However, the Constitution itself … conditions the exercise of this right to the adherence to moral, political and religious codes, [It is] not consistent with international law ratified by Bangladeshincluding ICCPR and ICESCR. (Freemuse, Drik, PEN International and PEN Bangladesh, 2017, p. 8) A number of CSOs highlighted the incompatibility of the 15th Amendment to the Constitution of 2011 that fell short of restoring secularism as a fundamental principle of State policy, but retained Islam as a state religion. Echoing Hobbesian European Enlightenment thought (that argued when authority asserts power over matters of faith and doctrine it invites discordsee Hobbes, 1651), they argue that the Amendment undermines attempts to protect freedom of religion or belief found elsewhere in the constitution (e.g., Article 2A, with guarantees against religious discrimination and Article 28(1) on belief) and is incompatible with Article 18 of the ICCPR. A further strand of the CSO discourse centres on personal laws and discrimination against women and girls. For example, a CSO noted that, 'in Bangladesh, discriminatory personal laws affecting all religious communities and women have fewer marriage, divorce, and inheritance rights than men, which increase their socioeconomic insecurity. No steps have been taken for the withdrawal of reservations to Article-2 and 16(1) (C) of UNCEDAW convention' (Steps Towards Development, 2017, p. 4). The prominence of this pathology again evidences government failure to respond to civil society criticality for the UN's Second Cycle UPR to the Government of Bangladesh concluded that it needed to urgently: 'enhance its law reform through partnership and collaboration with religious leaders, lawyers and civil society organizations including women's nongovernmental organizations' (OHCHR, 2016, para 98).

Authorities' Failure to Uphold Religious Freedom
This was the fourth-ranked pathology (12.9% of quasi-sentences). The CSOs express dissatisfaction with the actions of authorities from government downwards, notably the police and the judiciary. Typifying the widespread frustration with the ruling elites, one CSO noted that the government has formulated no effective strategy to investigate and prosecute perpetrators of violence against religious and ethnic minorities. As a result, violence against such communities have increased significantly since the second UPR cycle. The wider literature suggests (e.g., Haque & Mohammad, 2013;Paksha, 2010) that corruption is a major factor in the malaise. As one account notes, 'Bangladesh's current endemic corruption situation is deeply rooted in its recent history and the political process it has gone through … the political parties have given way to the military governments and military-backed civil government, who ignored people's participation and transparency' (Rahman, 2018, p. 315). It is a view supported by the UPR discourse. For example, one submission complained that, The Bangladeshi authorities' public response to the wave of violent attacks against secular activists has been marked by a troubling unwillingness to condemn the killings and appears to shift the focus onto the victims. The police are often reluctant to file charges or investigate influential politicians or businessmen for threats or violence against media workers, thereby creating a culture of impunity. (Amnesty International, 2017, p. 4) As Inglehart's (2009) international analysis reveals, there is a co-relationship between state capacity and rights violation. Citizens feel unsafe and human rights violations occur frequently in states that are weak and lack proper institutionalization. In contrast, stronger state apparatuses provide safety and security to their citizens as well as protect them from being subjected to violence or discrimination. The present analysis supports this. The UPR discourse reflects widespread civil society concern over limited state capacity (inter alia, measured by political will, human capital, financial resources, staffing levels and legal powers)to address religious freedom and communal violence issues. For example, one CSO complained, The National Human Rights Commission (NHRC) is far from independent and effective due to gaps in its founding act; absence of provisions regarding investigations of allegations of rights violations committed by the military and law enforcement agencies LEAs; and lack of authority to prosecute perpetrators. The Commission's institutional capacity remains limited. The process for selection of Commissioners does not fully comply with the Paris Principles. (Human Rights Forum Bangladesh, 2017, p. 4) To prosecute perpetrators and to protect citizens' rights the government needs to strengthen its law and order mechanisms, especially the police. As Boateng and Buckner (2017, p. 1) argue, in order to maintain state legitimacy, the role of the police force needs to be rethought; they should be active and effective not just in eliminating violence and discrimination, but also in promoting peace and democratic values of society. However, the third cycle discourse underlines that, in the case of Bangladesh, there is significant work to be done before this happens. For example, one CSO asserted that it was, 'deeply concerned that between February and October 2015, seven secular bloggers were killed in a series of attacks against media workers. On 26 February 2015, atheist blogger Avijit Roy was hacked to death and his wife was critically injured in an attack in Dhaka. While eight people were subsequently detained 29 in connection with Roy's murder, no charges have been filed against suspects' (Civicus, 2017, p. 3)

Threats/incitement
This is the fifth most alluded to pathology (11.8% of quasi-sentences). Here it should be noted that Bangladesh has been involuntary host to several covert extremist religious groups who have been responsible for inciting violence within the country. The Third Cycle UPR discourse attests to how this 'hosting' role has had dire consequences. For example, one CSO account recalls, 'how it met members of the Buddhist, Christian and Hindu communities who expressed a deep sense of fear and insecurity at the rising number of targeted killings and the growth of Islamic militancy. Many raised concerns about the lack of effective action by the government and the police to provide protection against and prevent human rights violations … [and] violent attacks against religious minorities' (Christian Solidarity Worldwide, 2017, p. 3).
A raft of other examples allude to the prominent case in 2017 regarding human rights activist Sultana Kamal. The radical Islamic group, Hefazat-e-lslam, deliberately misinterpreted her comments in a television show and threatened her with violence. Hefazat Vice President, Junayed A -Habib, even proclaimed that 'we will break every bone in her body!' (Front Line Defenders, 2017, p. 4). Furthermore, as the following example of CSO discourse explains, social media are at the centre of current extremist threats and religious intolerance: Bloggers and online activists … attracted the attention of extreme religious groups mainly after their lead on the Shahbagh Movement held in 2013. Then with a view to politicize the Shahbagh movement, the issue of the religious beliefs of bloggers and online activists were brought to the forefront of public scrutiny. A smear campaign spread throughout the nation alleging that the bloggers are atheists and an attempt to frame the movement as a struggle between the faithful and the nonbelievers. As a result of this, bloggers and online activists became the target of religious extremists and further created a misconception among the common people in Bangladesh … These killings are part of an alarming trend of violent intolerance towards freedom of religion. (Center for Social Activism and Association for Progressive Communications, 2017, p. 11)

Government Failure to Respond to Earlier UPR Recommendations
This was the sixth-ranked pathology (8.7% of quasi-sentences). It aligns with the wider international literature on the non-enforceability of international treaty obligations if they are not enshrined in domestic law (Bowman, 2007). The civil society discourse is condemnatory of the state's failure to act. For example, referring to the country's second UPR in 2013, one CSO noted that: 'Bangladesh accepted 185 recommendations from the UN Committeewith ten of them concerning freedom of religion. The government has since then taken some action against perpetrators of violence against minorities and tried to ensure that freedom of religion and belief is maintained. Although in some cases the perpetrators of violence have been brought to justice, in general, the state has not been able to contain or stop such extremist violence; frequency of threat and attacks continue to occur, threatening the security and wellbeing of ethnic and religious minorities in different parts of the country' (Christian Solidarity Worldwide, 2017, p. 3).
Furthermore, during the 2013 UPR, the Bangladeshi government also agreed to ensure the freedom and autonomy of civil society. However, the CSO discourse asserts that evidence suggests the contrary. It cites restrictions on freedom of expression, a shrinking of the civic space and increasing state oppression. Thus for example, one CSO noted that, 'overall, we find that despite Bangladesh accepting many UPR recommendations … minimal progress has been made on their implementation. Of particular concern are the continued existence of laws, together with legislative proposals, that impose impermissibly broad limits on free expression online and offline; increasing attacks against journalists, bloggers and human rights defenders and on-going impunity; and the continued targeting of religion and belief minorities' (ARTICLE 19 and Research Initiatives Bangladesh, 2018, p. 2).

Discrimination, (In)Equality and Oppression
This was the seventh-ranked pathology (4.7% quasi sentences). Amongst a range of issues, the CSOs highlighted sex discrimination affecting women from across difference religious communities. For example, one pointed out the discriminatory nature of personal laws that negatively affect women from all religious communities, especially on matters related to marriage, divorce and inheritance rights. As a consequence, 'women experience more socio-economic insecurity than men' 22 (Steps Towards Development, 2017, p. 2). In addition to gender discrimination, several CSOs also highlighted discrimination against religious minorities. For example, 'the Jummas, an indigenous community in the Chittagong Hill Tracts (CHT) region, have constantly been harassed and prevented from practicing their religion. Even sometimes, the security forces destroyed their worship places and disrupted the observance of religious rituals' (Unrepresented Nations and Peoples Organization, 2017, p. 3).
Furthermore, intersectional equality issues were a prominent trope. For example, one CSO highlighted that, 'Visible LGBT activists are under physical threat from religious extremists. Government officials and police do not have transparent processes for holding individuals and groups accountable. The murder of Xulhaz and Tonoy and the lack of a strong government response demonstrates the lack of legal protection for LGBT people' (Roopbaan, 2017, p. 9).

Conversion Issues
Earlier work has highlighted how religious minorities, the poor and vulnerable have been particularly susceptible to forced conversion to Islam. As Uddin (2016, p. 326) explains, 'over the years, many incidents of … forced religious conversion … took place in the river-valleys, often committed by the security forces and settlers against the Khyoungsa indigenous people'. The issue of forced conversion is particularly prevalent in the north of the country, 'as part of a nationalist purification process, the Bangladesh government subsequently encouraged the migration of plains dwelling Bengali Muslims to the CHT. Tribals were increasingly vulnerable to … forced conversion to Islam by the new settlers' (Jhala, 2013, p. 118). More recently, the UN Special Rapporteur on freedom of religion or belief found that, Feelings of insecurity exist in communities from which people have converted. Smaller minority communities in particular have expressed concerns that they would in the long term lose their members to the predominant Islam or to Christian missionaries, which fosters suspicion of other communities. Furthermore, rumours and unrealistic projections can damage the general harmonious relations between the followers of different religious groups. (UNHRC, 2016, para 59) The Rapporteur's account also points to an economic dimension to forced conversions; 'Trying to convert others likewise falls within freedom of religion or belief, as long as that persuasion is free from any coercion. This is an important caveat. However, its application in practice is not always easy, and there are lots of grey zones in this field. Missionary activities should certainly not exploit the vulnerability of people, for instance, those living in extreme poverty' (OHCHR, 2016, para 61). The Bangladesh Government is currently pursuing a policy of discouraging missionary activities, with a view to avoiding religious conflicts and forced conversion. There have been restrictions on the issuing of visas for international co-religionists. According to the UN, 'Religious minority communities deplored the fact that inviting fellow believers from abroad can become quite difficult, which also may infringe upon their freedom of religion or belief. Members of Christian communities reported they felt exposed to the unfounded suspicion that they would engage in systematic proselytism' (OHCHR, 2016, para 62).
Although the issue of forced conversion attracted limited attention in the CSO discourse (0.4% of quasi-sentences) the scale and nature of the alleged conversion incidences is troubling. One CSO reported that, 'A convert to Christianity from Islam, Hossain Ali, was also hacked to death on 22nd March 2017 by three Islamist assailants in his hometown of Garialpara, Kurigram in northern Bangladesh, who approached him on motorbikes, threw a Molotov cocktail at him, and attacked him with sharp weapons, killing him on the spot' (ADF International, 2017, p. 6). According to the CSO discourse, children have often been taken out of their homes and forcibly converted to Islam. Although some of these children do return to their homes, many of them go missing. For example, one CSO noted that over the past seven years 72 children have been taken from their families under the guise of a better life, and then forcibly converted to Islam against their will and the will of their parents (Cultural Survival and AILC, 2017, p. 5). Such instances not only violate individual rights to freedom of religion and belief but also contravene Bangladesh's obligations under the UN Convention on the Rights of the Child.
Finally, it is worth reflecting on how the foregoing pathologies might be thought of in relation to each other. The relationship between them emerges from the 'governance view' that this study employs. In other words, our analytical framework is linked to the tensions that emerge when seeking to apply universal values set out in UN treaties in specific governance contexts with their own cultures, histories and governing traditions. It reflects the reality that local human rights practices are contingent and place-based. The different rights pathologies in this study can be seen to fall into two broad groupings. The first relates to governmentalitynot least, periods of authoritarian rule, patronage and the exercise of executive power. The pathologies in this group include: restrictions on civil society, corruption in public administration (in turn, contributing to authorities' failure to uphold religious freedom including police malpractice) and government failure to respond to earlier UPR recommendations. The second is concerned with inequalities and discrimination in the pluri-religious state. Specific pathologies in this group include violence, threats and incitement.

The Framing of the CSO UPR Discourse on Religious Freedom
Essentially, the following analysis reveals that the framing of the civil society discourse (in terms of 'attack', rights-denial, threat, hurt/ pain, defence, (in)justice, and fear)reflects the emotions and experiences stemming from the basic opposition of the two remaining frames -'secularism' and 'radical extremism'as well as longer-standing inter-communal tensions and religious intolerance exacerbated by the rise of extremism.
Civil society organizations' identification of violence as the principal human rights pathology in the UPR corpus is reflected in the framing of the discourse on religious freedom. The most prominent frame used by civil society organizations is the notion of 'attack' on minorities and religious freedom (Table 2). Together with associated tropes (see below), it accounts for almost a half (45.7% of all quasi-sentences, N = 756). Secularists are frequently the target and extremism is often cited as the cause of the attacks. For example, one CSO report described how 'seven secular bloggers were killed in a series of attacks against media workers between February and October 2015' (Civicus, 2017, p. 3).
The denial of human rights was the second most cited frame (17.5% of quasi-sentences). A core strand of the discourse using this frame alluded to widespread restriction and oppression of human rights defenders. For example, 'human rights defenders, secular activists, and others have suffered restrictions on their right to freedom of expression. They have faced repressive tactics, received threats or been subjected to physical attacks (some of which have proved fatal) by armed groups and other non-state actors' (Amnesty International, 2018, p. 4).
Reflecting the negative psychological consequences of the current situation in Bangladesh 'threat' was the third-ranked frame (12.2% of quasi-sentences). Frequently, it was associated with the actions of extremist groups. For example, 'in May 2015, four masked men killed a citizen-journalist in broad daylight who wrote for Mukto-Mona ('Free Thinking'). He had been getting threats from Islamist extremists for months in connection with his blog posts' (Reporters Without Borders, 2018, p. 2).
'Hurt/ pain' was the fifth-ranked frame. Its principal use was in reference to the notion of 'hurting religious sentiments' in relation to 'Section 57' restrictions. For example, 'the Information Communications Technology (ICT) Act, which was passed in 2006 and amended in 2013, poses a serious threat to freedom of expression of LGBT communities. It criminalizes 'hurting religious sentiments' and dissenters as well as members of minority communities have been imprisoned under this law' (Roopbaan, 2018, p. 2). The defence of individual and group rights was the sixthranked frame. For example one account described how law enforcement officials harassed Christians noting that 'on 24 March 2017, four plain-clothes police officers near Dhaka raided the home of a Christian woman without a warrant. When the neighbours tried to defend her, around 30 policemen attacked and beat the villagers' (Christian Solidarity Worldwide, 2018, p. 6).
(In)justice was the eighth-ranked frame. CSOs emphasized the barriers to justice for victims and their familiesincluding 'cases of enforced disappearances, extrajudicial killings, and torture' (Solidarity Group for Bangladesh, 2018, p. 5). Finally, 'fear' was the ninth-ranked frame. CSOs noted how state oppression, Islamic extremism, religious/ ethnic discrimination, attack on bloggers and several other instances of everyday violence have created a climate of fear in society. As a consequence, several secular writers and blogger have left the countrywhilst, the discourse highlights, those who remain have decided to resort to self-censorship.

Discussion
The foregoing analysis paints a troubling picture of religious intolerance and violence in today's Bangladesh. Drawing on 'situated knowledge' in the form of civil society organizations' accounts of the current situationas submitted to the latest UPR, the findings reveal the pathologies, emotions and experiences stemming from the basic opposition of 'secularism' and 'radical extremism'and, how in turn it is fuelling long-standing inter-communal tensions and religious intolerance. One of the major factors responsible for this malaise is the politicization of religion. Specifically, the sustained mainstreaming of Islam in public political life that began in the mid-1970s, such that by the mid-1990s, Islamic symbols and idioms had become part of everyday politics. Another is the critical juncture that came in 2011 with the Fifteenth Amendment to the Constitution, which overturned earlier secular aspirations and confirmed Islam as the State religion of the Republic. The present analysis also reveals powerful accounts of the principal human rights pathology affecting the country today, religious-based violence. It is a narrative associated with police malpractice/ judicial failings, discrimination, oppression and incitement. The analysis is founded in the situated knowledge of legal restrictions and government repression of civil society organizations. In public policy terms, as the literature on agenda-setting (Cobb & Ross, 1997) highlights, ranking the violations identifies which issues CSOs think are most in need of OHCHR condemnation and government action. In another regard, the analysis tells us that we need to understand threats to religious freedom in Bangladesh as being more than a simple narrative of the rise of emergent 'political Islam'. Rather, as the diverse issues and framings in the civil society discourse suggest, the threat to religious freedom is multi-faceted, contingent and place-based. It is grounded in Bangladesh's culture, history and governing traditions. By taking a 'governance view' of rights implementation we see that the non-discrete pathologies fall into two broad groupings. One concerned with governmentality (inter alia, periods of authoritarian rule, party politics and political patronage) that includes restrictions on civil society, corruption in public administration (in turn, contributing to authorities' failure to uphold religious freedom including police malpractice) and government failure to respond to earlier UPR recommendations. The second is associated with inequalities and discrimination in the pluri-religious state. These include violence, threats and incitement.
Analysis of the framing of civil society discourse tells us about the underlying emotions and meanings attached CSOs' accounts. These apply to the breadth of rights pathologies identified in the first section of this article. In ranking the frames according to their prominence in the UPR corpus we follow established scientific method in discourse analysis (Goffman, 1974;Singh & Swanson, 2017;Snow, 2013). As Snow (2013, p. 136) explains, 'frame analysis is rooted in the symbolic interactionist and constructionist principle that meanings do not naturally or automatically attach themselves to the objects, events, or experiences we encounter, but arise, instead, through interpretive processes'. He continues, 'like picture frames, they focus attention by punctuating or bracketing what in our sensual field is relevant and what is irrelevant, what is "in-frame" and what is "out-of-frame,"'. Thus, by considering the emphasis or relative level of attention we place on different frames 'they function as articulation mechanisms in the sense of tying together the various punctuated elements of the scene so that one set of meanings rather than another is conveyed, or, in the language of narrativity, one story rather than another is told'. In addition, frame analysis humanizes the research data rather than merely categorizing the pathologies identified in the first part of our analysis. In this way they illuminate the personal and group experience of oppression and denial of religious freedom.
At this juncture it is apposite to reflect upon how CSOs continue their work in the face of the ongoing threat to religious freedomand the ways they compromise, resist, or collaborate with each other and with government. The context to this issue is set out in a growing literature on government restrictions on civil society (Feldman, 2015;Haque & Ahsan, 2014;Mohajan, 2014;Sahoo & Chaney, 2020;Saidul Islam, 2011). This repression has had a wide-ranging impact on the ability of exogenous interests to criticize government. This includes the curtailing of CSOs' ability to collaborate with international NGOs and fundersas explained in the following UPR submission: Under the Foreign Donations (Voluntary Activities) Regulation Act, CSOs wishing to receive funds from foreign sources or collaborate with international actors must register annually with the Non-Governmental Organisation Affairs Bureau (NGOAB) and seek permission for planned activities prior to receiving funds. After legislative amendments in 2016, the NGOAB gained wide-ranging discretionary powers [… this has led …] to a situation where CSOs can be financially enfeebled and unable to operate, pending permission from the NGOAB … the strengthening of the Foreign Donations (Voluntary Activities) Act has stymyied operation of independent civil society groups in Bangladesh. (Civicus, 2017, p. 4) Others have alluded to how: 'The government has used draconian legislation against a wide spectrum of civil society in an attempt to silence them' (Amnesty International, 2017, p. 2); and how CSOs are, 'concerned that the government continues to restrict freedom of expression and is failing to provide adequate protections to human rights defenders campaigning for freedom of religious belief' (Christian Solidarity Worldwide, 2017, p. 4).
The civil society discourse also reveals how government repression is curtailing cooperation and co-working between CSOs. For example, one referred to how 'the killings of human rights defenders and the subsequent government inaction also led to the disintegration of activist networks and loss of community trust' (Front Line Defenders, 2017, p. 3). For its part the UN has said the legal restrictions have 'undoubtedly ha[d] a chilling effect on civil society organizations, human rights activists and members of religious minority communities. It contributes greatly to the perception of a shrinking space for frank public discourse' (OHCHR, 2016, para 86). CSO resistance is illustrated by the words of another CSO which said, 'the Bangladeshi government does not take kindly to criticism of its constitution or the State religion, Islam. Journalists and bloggers who resist censorship or self-censorship on these subjects may face life imprisonment, the death penalty, or murder by Islamist militants, who often issue online calls for the deaths of outspoken secularist bloggers and writers' (Reporters Without Borders, 2017, p. 6).
In terms of how CSOs' are continuing their work in the face of government repression, the discourse reveals the value of UPR as a mechanism to attract international attention, criticize and call for government to change its ways. For example, one CSO said: we 'call on the Government of Bangladesh to create and maintain, in law and in practice, an enabling environment for civil society, in accordance with the rights enshrined in the ICCPR, the UN Declaration on Human Rights Defenders and Human Rights Council resolutions 22/ 6, 27/5 and 27/31' (Civicus, 2017, p. 2).
The overwhelming narrative of the CSO discourse is one of dissent and resistance to repression and government failure to uphold religious freedom. Accounts of coworking with government are absent from the corpus of UPR submissions. 23 As noted, consonant with the classical work of liberal theorists (Hobbes, 1651;Rousseau, 1968) as well as later thinkers on the role of civil society (e.g., De Tocqueville, 1835/1956), we argue that the present findings on the suppression of civil society and growing intolerance and religious-based violence means that unprecedented importance now attaches to safeguarding civil society criticality in order to uphold religious freedom and address prevailing human rights pathologies in the Republic.  (1985) comparing the human body and the body politic using the language of medical pathology to investigate the disorders and challenges governments experience in making and implementing policy, including imperatives such as upholding human rights. In metaphorical terms, it also fits with the present use of corpus analysis of CSOs' discourse. 3. Casanova (2011) has distinguished between the secular (modern epistemic category), secularization (modern world-historical process) and secularism (worldview and ideology). Especially for him, secularization could mean: (a) institutional differentiation between the sacred and the profane, (b) decline of religious beliefs and practices, and (c) privatization of religion. 4. There is a good deal of variation within Europe; it is not a homogeneous category. Different countries in Europe have followed different model of secularism. While countries like Spain, Poland and Italy have privileged one religion; other countries like France have followed Laicite. 5. It refers that the state shall not be associated with any particular religion and permit freedom of conscience, belief and religion for all citizens without any discrimination. 6. See Swatos (2003, p. 79). 7. While Sahoo (2013) discusses the liberal-pluralist, the Neo-Marxist and the Communitarian understandings of civil society, Jensen (2006) discusses the pluralist conceptions of civil society through an analysis of the sphere concept, the Lockean concept and the Scottish concept. 8. For an excellent discussion on the relationship between civil society and the authoritarian state, see Lewis (2013 11. The BNP that was created by General Zia-ur-Rahman as a counter to the secular Awami League positioned itself as party in tune with religious beliefs 12. On 6 December 1992, a crowd of Hindu Kar Sevaks (activists) demolished the sixteenthcentury Babri Mosque in Ayodhya, Uttar Pradesh. The demolition occurred after a violent political rally at the site. 13. Prime Minister of Bangladesh from 1991 to 1996, and again from 2001 to 2006. 14. Three developments prepared the ground for the Fifteenth Amendment. The murderers of Sheik Mujib were executed. This dealt a body blow to the counter-revolution and a legal ruling declared the Fifth Amendment to the constitution 'unconstitutional'. Finally, the movement of the Sector Commanders' Forum's (2007-08) demands for the trial of war criminals prompted a popular response, particularly from young people, calling for the revival and re-establishment of Mukti Judhyer Chetonaor, the restoration of the Constitution. In electoral terms, these factors also saw the victory of the Awami League-led Grand Alliance and the defeat of the Islam-pasand Four-Party Alliance, led by the BNP in the 2008 parliamentary election. 15. However, Bangladesh has made reservations to article 14 (3) (d) of the International Covenant on Civil and Political Rights. 16. This is an ongoing state of political turmoil between the two main political parties, the Awami League (AL) led by prime minister Sheikh Hasina and the opposition Bangladesh Nationalist Party (BNP) led by Khaleda Zia. In the wake of the controversial 2014 general election, the BNP raised several demands for a second election under a neutral caretaker government. In January 2015 their demands remained unmet and the BNP initiated countrywide protests. Violent clashes ensued, the AL called the BNP terrorists and Khaleda Zia was forcefully confined to her office. 17. Human Rights Forum, Bangladesh (HRFB). The network is Namati. 18. UAM Corpus Tool 3. 19. 15 instances. 20. https://www.icnl.org/research/library/files/Bangladesh/comm2006.pdf. 21. Ergo, because such expression does not, per se, intend to have the likely consequence of advocating hatred against a group. If the intention was such it would constitute incitement to hostility, violence or discrimination proscribed in Article 20(2) of the ICCPR. 22. Bangladesh has reservations to Article-2 and 16(1)(C) of CEDAW 23. However, it should be noted that there are those that do have strong links with government. The background to this is that government crackdown on civil society organizations' political activism led to a split within the NGO umbrella body ADAB (Association of Development Agencies in Bangladesh). NGOs wary of political activism left forming a new body, the Federation of NGOs in Bangladesh (FNB). As Rahman's (2006, p. 464) instructive account highlights, this can be seen as 'another attempt by the state to penetrate and fracture a civil society group along partisan lines'. The result was that FNB organizations now mainly focus on service delivery and advocacy functions rather than mobilization (Chowdhury et al., 2017). It is in this context that, 'large NGOs like BRAC or Grameen Bank enjoy cooperative relations with the government and widespread public support among the Bangladeshi elite' (Rahman, 2006, p. 464 ORCID Paul Chaney http://orcid.org/0000-0002-2110-0436