Governing Islam in French cities: defining ‘acceptable’ public religiosity through municipal consultative bodies

ABSTRACT In recent years, European states have institutionalised relations with Islamic groups in the form of national Islam councils. Similarly, municipalities have set up more or less comparable bodies to address issues related to urban religious diversity. However, rather than being restricted to Muslim representatives, municipal consultative bodies usually incorporate a variety of religious actors. This contribution analyses three such bodies in the French cities of Rennes, Bordeaux, and Toulouse. Adopting a governance perspective and drawing on qualitative fieldwork, I argue that by providing concrete advice on how to address religious issues, these bodies define what are considered ‘acceptable’ and ‘unacceptable’ public religious expressions, ultimately influencing normative ideas about laïcité. Moreover, I argue that the history of relationships between religious and municipal authorities and the political culture of the cities, among other factors, shape these local processes, thereby emphasising the distinct role of cities and urban actors in governing religion.


Introduction
Since the early 2000s, in an attempt to assert their national sovereignty, many European democracies have set up representation bodies to nationalise Islam and make it fit into their institutional frameworks of church-state relations. In France, the UK, Germany, the Netherlands, and Spain, governments have opted for this type of national body to institutionalise relationships with Muslim organisations. However, such bodies do not only offer Islamic communities symbolic recognition and official representation before the state. They also work as regulatory bodies by which the conditions for the exercise of religious freedom and the practice of religion are established (Laurence 2012;Zeghal 2005).
More recently, multiple municipal governments in countries across Europe have set up similar consultative bodies to address issues related to religion and religious diversity in their cities, albeit more limited in remit, size, and competences. Although officially these local bodies focus on the public practice of religion in general rather than exclusively on Islamic religious practice, they can be understood as similar institutional developments emerging at local level.
Three such municipal consultative bodies in France, the Rennes Consultative Committee on Secularism (Comité consultatif laïcité de Rennes), Bordeaux Sharing (Bordeaux Partage), and Toulouse Fraternity -Council on Secularism (Toulouse Fraternité -Conseil de la laïcité) form the empirical focus of this contribution. These urban consultative bodies are instruments of institutionalised participation in which individual citizens or representatives of social groups are invited to discuss salient policy issues. In the context of policies on religion, these are usually participatory bodies to which municipalities invite representatives of diverse religious groups and other civil society organisations to discuss the city's approach to religious issues. These governance bodies are particularly intriguing in the case of France, where interactions between the state and religious actors are highly contested. In this contribution I analyse these three local bodies as public policy instruments for the governance of religious diversity and, specifically, for the governance of Islam.
Thus far, research has identified the widespread adoption of public consultation as a policy tool in the governance of religion, and in the governance of Islam in particular (Fourot 2015;Torrekens 2012). These studies have paid attention to the composition of such local consultative bodies, highlighting the contested nature of the selection of religious representatives, in particular Muslim ones, the exclusion of certain actors, and the power imbalances between state and religious representatives (Duthu 2009;Fournier 2009;Lamine 2004a). However, we lack a more thorough understanding of how these bodies affect the regulation of religion.
In this contribution I analyse the concrete ways in which the work of these bodies contributes the guiding ideas about 'acceptable' and 'unacceptable' forms of public religiosity that ultimately inform broader normative and political understandings of laïcité and the place of religion in the public sphere. 1 Based on the analysis of three cases in France, I argue that these municipal bodies are more than spaces for dialogue, the sharing of knowledge, and the promotion of mutual understanding between religious groupsall of which are explicitly stated aims. They are also spaces for the 'localisation' of Islam, mirroring processes of its 'domestication' at national level. In other words, in their negotiations of religious diversity, these bodies serve as venues where the boundaries of 'acceptable' and 'unacceptable' public (Islamic) religious practices are drawn and adjusted to specific local conditions. Moreover, with their 'localisation' of broader debates about the place of religion in public life into urban contexts, these bodies contribute to the production and rooting of normative ideas about laïcité, and serve as spaces for the (re-)socialisation of various actors, in particular Muslim local leaders, into this republican principle.
Unlike previous studies, whichwith a few notable exceptions (Duthu 2009;Körs and Nagel 2018;Lamine 2004a)have mostly analysed a single municipal body (Torrekens 2012) or compared bodies across countries (Griera, Giorda, and Fabretti 2018), I compare three such bodies within one national context. In doing so I briefly examine how local factorsincluding the history of relations between local political and religious authorities, the political culture of the city, the political affiliation of the ruling party, and local violent eventsinfluence the ways these bodies work, their implications for policymaking, and the normative definitions of acceptable religious expressions derived from the discussions in such bodies.
Moreover, the study contributes to a better understanding of how the meanings of the notion of laïcité vary as the principle is put into practice locally by a variety of actors. In doing so, it considers cities, alongside central states, as sites of the governance of religious diversity. While often omitted in the sociological literature on the governance of religious diversity, cities are increasingly receiving attention from scholars in this field (Martínez-Ariño 2018). This is partly because, as this contribution shows, cities can shape policies on religion in significant ways. This finding challenges the commonly held assumption that policies on religion in France are almost exclusively an affair of the central state.
This contribution draws on the analysis of qualitative empirical material collected in Rennes, Bordeaux, and Toulouse between November 2015 and January 2017. These cities were selected based on some of their similarities and differences. They are medium-sized cities and regional economic centres in France. In all three cases the percentage of foreignborn populations was below 15% in 2014 (Insee 2014), and electoral support for the National Front (Front National, since 2018 Rassemblement National) in municipal elections did not exceed 10%. However, they are ruled by different political parties: the Socialist Party (Parti Socialiste) in Rennes since 1977, conservatives (since 2015 Les Républicains) in Bordeaux since 1947, and both in alternation in Toulouse. Moreover, while Rennes has a long history of strong Catholic predominance and good relations between the city's Catholic milieu and the ruling Socialist Party, Bordeaux has a longer tradition of religious diversity in municipal political life (Malogne-Fer 2019), and Toulouse's social, political, and religious life was shaken by a direct terrorist attack in 2012.
This study combines semi-structured interviews, document analysis, and participant observation. In each city I interviewed politicians and staff of the municipal administrations, religious actors, secularist groups, other civil society associations and NGOs. I also collected documents produced by the municipalities and other relevant urban actors and participated in the monthly meetings of the consultative body in Rennes between November 2015 and June 2016.

Municipal governance of religious diversity: a public policy instruments approach
The place of religion in the public sphere and the relationship between states and religious organisations are mostly regulated at national level through constitutions and laws, and more recently through bodies such as national Islam councils. The later can be understood as neo-corporatist institutions, comparable to labour unions or other forms of group interest representation, which aim at submitting Islamic communities to the rule of the state by granting them official recognition and participation in policymaking (Laurence 2012).
Although for the most part research has examined how different state-church arrangements result in different approaches to dealing with religious diversity (Soper and Fetzer 2007), there have been calls for greater attention to be paid to other levels of regulation, including cities and the role of urban actors (Bowen 2007;Itçaina 2015, 2016). Studies have examined a variety of municipal administrative practices and their relevance to the governance of religious diversity (Martínez-Ariño 2018). These include issues related to urban planning and its impact on the spatial presence of religions (Gale 2008), the banning of certain religious symbols in urban spaces, such as the burqa (Burchardt and Griera 2019), and religious dietary provisions in municipal school canteens (Binet 2016;Papi 2012). Research has also examined the increasing role of Muslim actors in governance at different levels, from the local to the national (O'Toole et al. 2013), with an emphasis on their institutionalisation in urban settings (De Galembert 2006). All these studies emphasise the relevance of local conditions, such as conceptions of public space and the role of municipal bureaucracies, in understanding the governance of religious diversity. Nonetheless, more research on the municipal level is needed because, as Giorgi and Itçaina (2016, 6) have argued, 'local politics often become the place of experimentation with innovative policies' that operate within a particular set of constraints and opportunities.
Two conceptual frameworks constitute the general background to this contribution: the governance framework and the policy instruments framework. In relation to the first, scholars agree that a transformation in the traditional ways of steering societies has led to a new situation in relation to governance, one in which non-state actors, alongside governments, increasingly take part in policymaking and implementation (Klijn 2008). I analyse the consultation bodies in order to capture how both state and non-state actors are involved in policymaking. Secondly, in my analysis I draw on the literature on public policy instruments, understanding them not only as 'concrete and specified operational forms of intervention by public authorities' (Bemelmans-Videc, Rist, and Vedung 1998, 4), but also as tools that condense a certain 'form of knowledge about social control and ways of exercising it' (Lascoumes and Le Galès 2007, 3).
In the current context of governance, instruments of 'participatory governance' (Cornwall 2008, 271) are increasingly considered appropriate to the task of addressing different policy issues. One such instrument is what Cornwall (2004) refers to as 'invited spaces', that is, intermediate institutions provided by the government in which only a small group of 'articulate elite community members' (6) is chosen to represent the interests of larger population groups. These instruments serve to make policy decisions and legitimise others that have already been taken (Aarsaether, Nyseth, and Bjørnå 2011), as well as to work as forums for civic education. Their development is affected by contextual factors, such as the attitudes of the ruling powers towards participation and the political culture and cultural practices of decision-making in each context (Cornwall 2004).
One such type of invited space, local consultative bodies, has become a common policy tool whereby cities deal with policy issues such as the integration of immigrants, cultural diversity, gender equality, and public housing. Similar governance instruments are also becoming common in religion policies. While some are single-issue participatory processes, like the case reported by Conti (2016) of discussions over the building of mosques in Italian cities, in other cases, like those in my study, such bodies are longterm initiatives not restricted to just one issue. Similarly, interfaith bodies have often served as policy tools for the governance of religious diversity in urban contexts: they promote intercultural understanding and social cohesion, disseminate knowledge about religion, and may prevent conflicts from emerging (Dick and Nagel 2017;Griera 2012;Griera and Forteza 2011;Griera and Nagel 2018).
Following Lascoumes and Le Galès (2007), in this contribution I analyse the three municipal consultative bodies as public policy instruments. They are comparable devices commissioned by the respective municipal governments to address religious issues, which define a certain relationship between the government and the governed and carry normative understandings of what is acceptable religiosity in urban public spaces. The three committees are just one example of institutionalised mechanisms of consultation that work as a form of invited participation and select members according to predefined religious categories (Burchardt 2018).
After commenting on the composition of each of the three bodies and the rationale and motivations underlying their existence, the analysis below focuses on the ways in which these bodies may influence policymaking. Ultimately, I am interested in understanding how institutionalised participation may shape normative definitions of public religiosity and understandings of laïcité.

Religion in consultative bodies: a tale of three French cities
In the aftermath of the Charlie Hebdo and Bataclan terrorist attacks, the French authorities reinforced the state's commitment to 'republican principles', including laïcité. A stronger emphasis on this notion, alongside stricter policies in the realm of national security, came to the fore in public and political debates. At city level too, municipal authorities felt compelled to intervene urgently in order to prevent conflicts arising in their cities. However, municipal measures to address religious issues were in place in French and other European cities long before these attacks.
Despite France being a centralised country, the decentralisation policies of the 1980s significantly changed its territorial structure, giving more power to cities (Thoenig 2005). While cities in France do not have explicit competencies in respect of religious issues, policies that affect religious practicesuch as on urban planning and school canteensfall under the competence of municipalities. In parallel, there is a trend towards promoting mechanisms of deliberative and participatory democracy, which municipal authorities use, among other things, to improve management and fight communautarisme (Jouve 2005). 2 Despite France's legal separation of church and state, participatory devices have also spread in the field of religion, and religious actors are increasingly conceived as governance partners because they provide urban political authorities with important resources (De Galembert 2006;Lamine 2004b). The three participatory bodies in this study constitute 'invited spaces' in which the municipality invites different urban actors, including religious actors, to participate.

Rennes
The first case is the Rennes Consultative Committee on Secularism (Comité consultatif laïcité de Rennes), set up by the mayor of the city in 2015. Rennes, capital of the Brittany region, has witnessed the arrival of ethnically and religiously diverse populations. The municipal government (led by the Socialist Party for over forty years) has a strong public interventionist character, also in the realm of religion policies, and a strong tradition of Christian democracy, influenced by progressive Christian organisations (Le Galès 1993). This situation, connected to a dense fabric of civil society associations, is crucial in understanding the rather strong participatory character of its local political culture and the composition and modus operandi of the consultative body.
The Rennes Consultative Committee on Secularism was part of the 2014 municipal electoral programme of the Socialist Party. However, it was not established until after the Charlie Hebdo and Hypercacher attacks. Formed by some forty local actors that are deemed to represent different community interests, the committee meets on a monthly basis and works as an advisory body for the municipal government. Alongside religious actors, representatives of the municipal government and members of municipal groups in the city council, city officials, representatives of secularist groups, freemasons, NGOs, experts, and a school representative sit on the committee. While this selection is broad, it is often criticised as including only actors likely to be sympathetic to the political stance of the municipality. Moreover, it is notable that the local branch of the National Federation of Free Thought, a small atheist association (La libre pensée), declined an invitation to participate because its members felt that by inviting religious representatives to the committee the municipality was violating the separation between church and state established in the 1905 law. As one of its members put it during our interview: 'We are ready to be interviewed by the mayor, but not to discuss living together with priests'.
The main and most immediate aim of the body in Rennes, after exchanging information about the practice of laïcité in the cityunderstood in this case as state secularism and the regulation of religionand monitoring its application, was to draft a local charter of laïcité. Once approved by the municipal council, the charter would be implemented in the municipal administration and was also intended to serve as a useful resource for other actors. The municipal authorities thought that this would be an appropriate tool providing guidance in a simple and clear way, like the 'Charter for secularism in school' (Charte de la laïcité à l'école) published by the Ministry of National Education in 2013 in the form of a poster to be displayed in all public schools. Local actors in Rennes were involved in different stages of the process, from discussing the issues at stake to gathering relevant information from various stakeholders and finally offering concrete advice in support of action by the municipality. Interviews were held with local actors to establish the concerns of the population and to complement issues to be addressed proposed by the municipality itself.
This participatory process, which is part of a citywide plan to include citizens in local policymaking called the Citizen Factory (La Fabrique Citoyenne), is justified by drawing on the city's long history of citizen participation. Participation was seen by many of my interviewees as 'the way of doing things in Rennes', as well as a way of 'utilising' the resources already available in the city's civil society infrastructure. Moreover, this tradition of relating to non-state actors also includes religious actors, mainly as a result of the city's long history of Christian democracy in the twentieth century characterised by smooth relations between urban politicians and Catholic actors. Including religious actors in this 'laïcité partnership' was justified as a natural step, in harmony with the city's political culture.

Bordeaux
The second case is the series of yearly interreligious conferences called Bordeaux Sharing (Bordeaux Partage), running since 2010. The historical presence of a large representation of Protestants and Jews among the foreign commercial bourgeoisie in the city of Bordeaux, capital of the Gironde region, is reflected today in the close relationships between the mayor and some religious leaders. Similarly, the political culture in Bordeaux is very much centred on strong (conservative) mayors, as evidenced by the duration of their office (between 1945 and 2019 there were only two mayors in Bordeaux, with a brief temporary mayor in post between 2004 and 2006).
The mayor of Bordeaux established the body for interreligious dialogue, in which different topics are approached from the perspective of a selection of the major religions in the city, in order to further mutual knowledge and dialogue. In this case, in contrast to developments in Rennes and Toulouse, no direct link exists between a specific violent event and the setting up of this governance structure. However, in this period recurrent debates over the building of a large mosque in the city (Joffrin 2014) led to a scrutinising of the mayor's support for the project. Accused in social media of being too permissive with radical Islam and too close to the main imam of the city, Alan Juppé probably saw the creation of this body as a way to regain his own legitimacy by showcasing the support he enjoyed from the leaders of the main religious groups in the city.
In Bordeaux, the composition of the body is limited to representatives of religious groups, the mayor, and a deputy mayor. Visually the arrangement consists of the mayor surrounded by prominent religious figures, some of them with high media profiles, evidence of a more personalistic style of governing. In the mayor's own words, Bordeaux Sharing is 'a small body where I gather the cardinal, the rabbi, the Protestant pastor, the imam and the representative of the Buddhists to share the republican values we hold in common' (emphasis mine). 3 While at some point members discussed whether secular actors (Freemasons and atheists) who had sought to participate should be accepted, they finally rejected that option.
The rationale behind the establishment of such a body in Bordeaux was to offer an open forum where representatives from various religious groups are able to discuss topical issues with each other from the perspectives of their respective religions. Thus far, topics of discussion have included the role of women in religion, the role of religion in conflict and violence, and the role of religious groups in welcoming foreigners. The body is conceived as an interreligious forum moderated by the mayor and concerned to institutionalise already existing bottom-up initiatives of interreligious dialogue. One explicit aim is to show the contribution that religions can make to the process of living together. 4 Another goal is to show that religions can live well with one another, that they can talk to each other, and that there is no conflict between them. Moreover, religious actors participating in this body are seen by the municipality as messengers both from and to their communities. From a more critical standpoint, however, this body can be seen as a strategic tool the mayor may use to foster the political legitimacy of his decisions among certain populations. Unlike the case of Rennes, no explicit mention is made regarding policy advice or participation in decisionmaking in the description of this body. In this sense, the rationale behind it is more one of 'displaying' a specific, normative view of the presence of religions in the city's public sphere, and of the contributions they make, than one that stresses their role in informing policymaking.

Toulouse
The third case is Toulouse Fraternity -Council on Secularism (Toulouse Fraternité -Conseil de la laïcité). Toulouse, capital of the Occitanie region, has a long history of immigration. In contrast to the other two cities, Toulouse has seen frequent political alternation between left-and right-wing parties in its municipal government, and its political culture has been strongly affected by the terrorist attack on a Jewish school perpetrated by a local inhabitant in 2012.
Toulouse Fraternity replaced an earlier consultative body installed in May 2013 after this attack. Set up as an emergency response to the attack and in pursuit of a strategy to prevent tension escalating in the local community, the aim of the earlier Toulouse City Council on Secularism (Conseil de la Laïcité de la Ville de Toulouse) was to promote the principle of laïcité and advise the municipality on issues related to religion. In December 2014 the new conservative municipal government replaced it with a body called Toulouse Fraternité -Conseil de la laïcité. Although this is presented as a continuation of the previous council, the name emphasises the new body's task of promoting fraternité. This shift, together with the stronger emphasis put on the role of religious groups, was strongly criticised by the socialist opposition. The composition of the Toulouse council closely resembles that of the committee in Rennes in including different civil society actors alongside religious representatives. The list of religious traditions included in all three bodies, namely Buddhism, Catholicism, Islam, Judaism, Orthodox Christianity, and Protestantism, is seen as exhausting the possibilities of religious representation. When one city official explained the composition of Toulouse Fraternity to me, he said 'all religions are there'. This indicates that there is a clear understanding of what does and does not constitute a religion from the perspective of state actors setting up urban governance networks. The religious groups involved reflect the tradition of bottom-up interfaith dialogue in the three cities. Pentecostal churches, Jehovah's Witnesses, less established branches of Islam, Hindus, and other smaller groups are not even considered. Of particular concern to state actors is which Muslim actors should be deemed 'appropriate' and 'representative' of the whole group of Muslims. In Toulouse as well, the local branch of the National Federation of Free Thought association adopted a very critical stance towards the consultative body.
The setting up of the initial Council on Secularism (laïcité) may be understood as an effort to calm the situation after the terrorist attack on the Jewish school. Following the attack the municipal authorities perceived the prevention of potential local conflicts as needing to come from within the religious field as well as from the state. The initiative was also an attempt to show that the issue was not a matter of conflict between local religious groups. Thus, in this case, the existence of a violent event in the city was a critical juncture that very much influenced the body's creation and direction. Its more explicit tasks were 'to advise the mayor in matters related to the application of laïcité', such as the management of public space in relation to religious demonstrations, to promote dialogue to fight communitarianism, and to organise events at the mayor's request. Since being reconfigured in 2014, the emphasis has been on promoting the value of fraternity and on recognising the different spiritualities that 'constitute the soul of the city', according to the minutes of the municipal council.

Implications for policymaking and the definitions of 'acceptable' religiosity
In Rennes, drafting the 'Rennes Charter on Secularism' (Charte rennaise de la laïcité) was the main and immediate goal of the body, which acted proactively rather than reacting to pressing issues. Discussions in the committee were goal-oriented, informed by the pragmatic aim of providing responses to specific situations: after taking stock of the state of affairs in the city regarding particular issues, such as menus in school canteens and the provision of plots for religious groups in municipal cemeteries, members discussed potential challenges, gathered knowledge, and made recommendations as necessary. In cases where agreement among members was not possible, the different positions were recorded. The final document, containing the debates and very detailed recommendations, was presented to the municipal council, which immediately agreed to commission the committee to pursue concrete measures regarding, for instance, the provision of various menus (one without meat, one without pork, etc.) 5 in school canteens. The city council also agreed that the committee would continue to put forward advice at the request of the municipal government until the end of the council's term in office in 2020.
Along with analysis of the final document, my participation in the meetings of the committee allowed me to observe the degree of detail in the discussions held. I will illustrate this with two examples. The first one relates to requests by municipal employees for permission to pray during working hours. The committee indicated in its recommendations that the neutrality of public services prohibited prayer at work. However, this position assumes that prayer is noticeable by others, otherwise how could it be prohibited? The committee's discussions concerned Muslim workers. Some committee members stated that the fact that some employees had requested permission to pray during working hours indicated that they had not understood that prayer may be done privately in silence, without anyone else noticing. A preference for privatised and invisible forms of religiosity was shared by most of the committee members. Interestingly, secular and secularist actors drew explicitly on Catholic understandings of religious practices (individual silent prayer) as the default template in assessing certain situations and providing recommendations. Without much conscious awareness, theological considerations were brought into discussions in a state-led setting.
The second example refers to another intensively discussed issue, namely whether women wearing the veil could accompany school outings. This discussion emerged following debates triggered by a 2012 circular of the Ministry of Education which assumed that these supervisors had to comply with the neutrality rules of the education system. The committee did not reach a consensus on this but the majority of its members favoured allowing veiled mothers to take part in school outings. The fact that no (veiled) Muslim woman was invited to take part in a discussion affecting them directly was very telling of how these women are often seen as passive subjects. The impossibility of listening to the voices of the women potentially affected compelled some non-Muslim religious leaders to dismiss the debate as outrageous and as curtailing freedom of religion generally.
In the case of Bordeaux, the body does not generate explicit outcomes that shape political decisions and policy measures addressing religious diversity, nor is it expected to. The fact that relations between religious groups and the political authorities assume a more personalised shape (Malogne-Fer 2019), as is also evident in Bordeaux Sharing, goes hand in hand with the existence of fewer regulations and administrative procedures regarding religion. Bordeaux Sharing works instead as a source of legitimacy for the mayor and his decisions, as a resource the mayor can mobilise at different moments, especially moments of 'crisis' such as after a terrorist attack, in order to prevent misunderstandings and confrontations, 6 and as a model for harmonious relations between groups. As presented on the city's website, this shared aim of living well together 'has allowed us to prevent external conflicts disrupting the life of the neighbourhoods'. 7 Public events of this type of body can be seen as 'ceremonial areas in which participation [is] ritualised' (Cornwall 2004, 5), rather than actual settings for active citizen participation. The staging of 'interreligious understanding' as a public performance, then, suggests a different way of dealing with religious issues. In this case, it is more a politics of big names, where the mayor and the most prominent representatives of the different religious groups gather in a highly visible event. It is thus by establishing a visible model (Brighenti 2007, 334) of good religion, and good relations between religions, that the municipality tries to influence the behaviour and attitudes of the wider population. As presented in the 'Calendar of living together', published annually by the municipality in collaboration with the members of Bordeaux Sharing, this interfaith body 'is a valuable tool for living together due to its symbolic value'. Moreover, by discussing the contribution of religions to specific social issues, such as the reception of immigrants, the body displays an understanding of laïcité that emphasises and recognises the positive role that religions may play in public life.
Toulouse Fraternity, like the committee in Rennes, aims to offer advice to the municipal government on matters related to religion. Since its conception, it has generated outcomes in the form of reflections and recommendations that have informed subsequent political decisions. One such example is the 'Charter of republican principles and values' (Charte des principes et valeurs républicaines), which the municipality wanted all associations receiving public funding to sign as proof of their commitment to these values, including references to laïcité. Before the decision that all associations financed by the city had to sign this charter was made and approved, the municipal government requested that Toulouse Fraternity reformulate the text. In drafting it, this body contributed to establishing a 'soft' way of regulating the behaviour of local associations in relation to religion (e.g. preventing proselytising) and the principle of laïcité. Although the work of Toulouse Fraternity had a less clearly defined outcome than was the case with the Rennes charter on secularism, it has also had immediate implications for policymaking in relation to the regulation of religion. While it has not provided a long catalogue of detailed recommendations as in the Rennes document, Toulouse Fraternity issued the guide 'Secularism and public space' (Laïcité et espace public). This document refers to the legal framework regarding issues such as religious celebrations in public spaces and the wearing of the so-called 'burkini' on public beaches, and suggests ways to improve specific local situations. This body has also offered advice on other matters, for example in relation to dietary provisions (i.e. offering a menu without meat) in school canteens. Interestingly, when communicating the new model of menus to families, the emphasis was put on nutrition rather than religious matters, thereby preventing potential controversy from emerging. Briefly, while the degree of detail and intervention is significantly lower than in the case of Rennes, and while the work of this body did not start out from an all-encompassing approach to religious issues in the city, it also has a regulatory nature.
Moreover, beyond the final outcomes in the form of recommendations, policy documents, and official statements, discussions within Toulouse Fraternity reveal the different understandings of laïcité among its members. 8 While for some political actors on the left laïcité should be understood exclusively as freedom of conscience and the neutrality of state institutions, for members of the current government laïcité is understood as 'inter-cultuel' (interfaith) and the promotion of the value of fraternity and living together. Beyond political differences in the meanings attached to laïcité, which are not exclusive to this particular local context but reflect differences in partisan affiliation, other discussions reflect particularities of the local context that affect ideas about public religion and the meanings attached to laïcité. In particular, the impact of the terrorist attack in Toulouse is reflected in the emphasis put on measures aiming to reinforce knowledge of and commitment to so-called republican values, and the capacity of different groups to live together peacefully. Ultimately, this shows that the meanings given to the notion of laïcité, as this principle is put into practice on the ground by various actors, are varied, which challenges homogenising understandings of state secularism in France.
This analysis has shown that, similar to national Islam councils, local consultative bodies play a role in defining 'acceptable' religion. For instance, they define which options are preferable in relation to the provision of menus in school canteens, the regulation of religious practice at work, and the different modalities available to address and support or restrict the construction of places of worship. Albeit in different ways, they all display and promote both a rather privatised form of religious practice, considered compatible with republican ideals, and non-conflictual relations between groups.
While the main concerns relate to Islam, including a variety of religious groups and nonreligious associations in these bodies rather than limiting the selection to Islamic actors (as in the case of national councils) avoids establishing too strong a focus on Islamic groups. It also serves to ensure the neutrality of public engagement with religious groups (Lamine 2004a). Furthermore, the presence of different religious groups mediates state regulation of Islam, thereby sometimes softening potential attempts to limit certain (Islamic) religious expressions (as the example of the veiled mothers in Rennes illustrates). Ultimately these bodies provide practical examples of how religious groups themselves, and other state and non-state actors, become active in regulating religion and promoting a 'pedagogy of laïcité', which aims to make this notion understandable and accessible to the public. Next to their explicit outcomes, these participatory devices can be interpreted as 'schools of democracy' that shape the civic culture of citizens and 're-socialise' community leaders into republican virtues. In this sense, these municipal councils partly resemble neo-corporatist state-society governance arrangements, as suggested by Laurence (2012) in his examination of national Islam councils. By institutionalising consultation and selecting representatives of different religions deemed to be attuned with the political agenda of the municipalities, these local councils reduce the potential for conflict and mediate the demands of religious groups, particularly Islamic ones, at the local level.
However, despite these similarities, these bodies also differ in some respects. In what follows, I will show how some local specificitiessuch as the relationship between municipal and religious authorities, political constellations, and the political cultureshape their functioning and outcomes.
The compositions of these three bodies, which differ slightly across cities, reflect the different history of relations and roles attributed to religious groups. This influences the resulting visions of laïcité and normative ideas about public religiosity. The Rennes Consultative Committee on Secularism and Toulouse Fraternity adopt a broad approach to the selection of actors, recognising that not only religious but also secular actors should be involved in issues related to religion and the deployment of laïcité, whatever that may mean in each context. Moreover, as the text of the 'Rennes Charter on Secularism' states, provided that neutrality is respected the inclusion of religious representatives indicates an assumption that laïcité evolves and can recognise and cooperate with religions (for similar arguments, see Lamine 2004a). The inclusion of religious groups is facilitated in the case of Rennes by the long history of amicable relationships between the civil and Catholic authorities in the city and its strong civil society infrastructure. In Bordeaux, this is possible due to the history of bilateral interreligious dialogue in the city and the close personal relationship between the mayor and main religious representatives (Malogne-Fer 2019).
Moreover, because Bordeaux Sharing is set up as an interfaith group, secular actors are considered to be at odds with the activities of the group and are therefore excluded from it.
The reorientation of Toulouse Fraternity after the new municipal government took office shows that political constellations differ in their visions of laïcité and of the relationship between state and religious representatives. The turn to a conservative government in Toulouse is reflected in its understanding of laïcité. The new body aims to 'favour the recognition and respect of all religions that irrigate French society [. . .] and to grant free religious expression' (proposal for the creation of Toulouse Fraternity 12 December 2014), a position which differs significantly from how the previous body and the Rennes committeeboth of which were set up by socialist mayorsapproached the matter.
Finally, the political culture of the city regarding citizen participation affects the structure, functioning, and outcomes of these bodies. This is very clear in the cases of Rennes and Bordeaux, and less so in the case of Toulouse because of the extreme circumstances produced by the terrorist attack. In Rennes this policy instrument reflects a relationship between the state and stakeholders in which both parties are active players. This is the result of a local political culture that has traditionally been grounded in strong citizen involvement, proven by the local plan to promote citizen participation. In contrast, the more personalistic governing style of the mayor in Bordeaux, linked to a history of strong mayors, is reflected in the fact that Bordeaux Sharing was not set up to inform municipal policies on religious matters and it does not include representatives of secular civil society associations. Rather it serves as a stage on which a symbolic representation of mutual understanding between religions, orchestrated by the central figure of the mayor, is performed. As the municipal politician in charge of Bordeaux Sharing put it in an interview conducted in December 2016, 'We have relations with religions to send a symbol that religious actors talk to each other, talk to the mayor, and respect each other' (emphasis mine). This approach emphasises the symbolic and performative aspects, which can serve as 'role models' for the wider population. In the case of Toulouse the terrorist attack geared measures towards 'educating' the public, in particular the youth, in topics related to republican values and principles and the city's history of migration, as a way to fight the so-called 'identity closure' of certain groups.

Conclusion
In this contribution I have analysed the role of cities in the governance of Islam and of religion more broadly. Focusing on three municipal consultative bodies set up to address issues of religious diversity, the contribution has shown that policies on religion, even in centralised countries like France, are not the exclusive monopoly of the central state. From the perspective of public policy instruments, I have shown thatby treating these bodies as policy instrumentswe can grasp how they define a certain relationship between government and governed (religious and non-religious actors), in which both are considered actors of the governance of religion. These bodies epitomise the notion of 'governing of and through religion', where religious actors participate in urban policymaking and the regulation of religion.
Moreover, the analysis shows how these bodies draw on, and produce, knowledge and expectations about public religious expressions that do not simply reproduce national discussions and discourses on laïcité. Next to national legal and political frameworks, these bodies are crucial in setting the parameters that define 'acceptable' public religiosity as opposed to that which is considered illegitimate. By drafting charters, giving 'expert' advice, and performing mutual understanding, they all contribute to producing certain ideas of good/bad, acceptable/unacceptable religiosity, thereby contributing to the 'formatting' of diverse religious practices and expressions according to established (Christian) patterns (Roy 2013). 'Acceptable' religiosity, as explicitly or implicitly prescribed by these bodies, is generally religiosity that is performed in private and is not too visible, and does not feature traits of 'communitarianism' deemed incompatible with republican citizenship.
In this sense, these bodies partly resemble national Islam councils in that they set the conditions for (Islamic) religious practice. However, by incorporating members of multiple religious traditions and secular associations, these bodies make the seemingly intractable issue of regulating Islam less controversial. Moreover, the presence of other religions can counterbalance attempts to regulate Islamic practices in ways that would not be done with other religions.
These consultative bodies are thus not simply instrumental devices that provide pragmatic recommendations for practical problems, nor are they merely implementing a top-down mandate from the central state. These 'invited spaces' develop normative definitions of accepted public religiosity that may ultimately affect broader political ideas of laïcité. As the study also shows, contextual factors matter, and shape these urban responses to religious diversity in different directions. This emphasises the need to put cities at the centre of our analyses. Local factors such as the history of the relationship between municipal and religious authorities, the political culture of the city regarding citizen involvement, political constellations, and local violent events may all play a role in relation to the setting up of these bodies, their composition, tasks, and orientations, thereby ultimately affecting the resulting definitions of laïcité.
Finally, these results show that the ways in which the principle of laïcité is interpreted and applied to govern religion are varied and changing. While the national legal and political frameworks set the basic parameters for the handling of religion, urban actors in cities can also shape religion policies in significant ways. This practical instantiation of the principle of laïcité may ultimately also make it evolve in varying directions, challenging homogenising ideas of state secularism in France, and beyond. Notes 1. The notion of laïcité is widely used both in academia and in French public discourse, referring to the right to freedom of religion and conscience, the separation of church and state, and the neutrality of the state towards religious organisations in France. In this contribution I treat it as a 'category of practice', used by multiple actors and requiring analysis, rather than as a 'category of analysis' (Brubaker 2013). I therefore do not adopt an analytical definition but instead examine the varied meanings attached to the term by different actors when they apply it in practice. 2. The term communautarisme is often used in public debates to refer to a perceived threat posed to social cohesion by the formation of intermediate forms of sociality and claimsmaking by minority groups that are seen as challenging universalist understandings of the Republic. From a critical perspective, the discourse of communautarisme is understood as a strategy to delegitimise requests for recognition made by ethnic or religious minority groups (Montague 2013). 3. My own translation from http://www.al1jup.com/rassembles-unis-soudes/. 4. The notion of 'living together', or vivre ensemble in French, is widely used in discussions on religion in public life. It is both a legal construct and a term of common use. In its judgement regarding the 2010 law prohibiting the concealment of the face in public spaces (case S.A. S. v. France), the European Court of Human Rights refers to the so-called 'French principle of living together' to support the argument made by the French legislature that covering one's face impedes human relationships, necessary for living in society. This was not uncontroversial because of the perils of using general principles to limit religious freedom (Adrian 2017;Tsevas 2017). Moreover, vivre ensemble is also commonly used, but only seldom defined, by a variety of social and political actors when discussing laïcité. It often refers to social cohesion, some sort of peaceful coexistence between people of different religions and an attitude of respect and civic behaviour. It is in this broader sense that the term is used in Bordeaux Sharing (Bordeaux Partage). 5. Provision of halal and kosher products in public school canteens in France is rare if not nonexistent because the purchase of such products by schools would be considered as state funding of religious groups and thus in contravention of the 1905 law. 6. http://www.al1jup.com/bordeaux-partage/. 7. Author's translation from http://www.bordeaux.fr/p81905/dialogue-inter-religieux. 8. I would like to thank Frank Laporte for sharing his research report 'Toulouse Fraternité-Conseil de la laïcité: Quelles politiques publiques municipales autour de la laïcité et de la citoyanneté' with me.