Daily Life Spatialities and Temporalities of Religion in Ottoman Tunis: Reflections on the Complexity of Urban Religious Landscapes.

On September 26, 1861, an angry crowd of protesters gathered inside the Ezzituna Mosque of Tunis. They were upset by the consequences of the modernizing reforms the local government of this Ottoman province was implementing. They were reacting against the loss of power for a group of local notables that the reform of the courts of justice was implying. They were also denouncing foreign interferences, as the local implementation and negotiation in Tunis of the Ottoman reforms happened in the context of growing pressures by European consuls ([90]). The gathering then evolved into a street demonstration that took the direction of the Palace of the Bardo, the residence of the Bey outside of the city walls ([20], 761 – 762).

What could be interpreted as the mere protest of a group of conservative religious notables against modernity and its ambiguities, in fact hides numerous layers of complexity as far as identity, but also time and space are involved. These dimensions can be better understood through an attention to the relationship between temporalities, spatialities and religion. The starting place of the demonstration, indeed, itself has a multimillennial history, that links it to the symbolic of power and legitimacy. The course of the demonstration, also, covered places that embedded precise meanings and participated in the building of the message that was delivered to the Bey. The very timing and rhythm of the demonstration was highly meaningful in a context in which all actions were meant to have a strong impact. After their prayers, regulated by a codified time, the demonstrators did not directly walk to the Bardo palace. Before exiting the city walls, indeed, they stopped, in the core of the inner city, the medina, at a shrine and mausoleum dedicated to a local saint, Sidi Mahres (951 – 1022), patron of the city since the Middle Ages ([11]). This pause, in a tense moment, in contrast with the rush towards the seat of the political power, also had a strong political and spiritual meaning. The cult of Sidi Mahres is linked to a millennial history, the location of the shrine having been the object of veneration since ancient times. Passing through this place and explicitly pausing meant for the protesters that their demonstration had acquired a religious, but also civic, legitimacy and force: their use of time and space converged in building meaning.

Time and space contributed to the symbolic charging of their claims. Demonstrators grabbed symbols of the saint, among which his banners, that they carried in the demonstration (Ibn Abī Dhiyāf, vol. 5 – 6, 89). From that moment on, the rhythm of the demonstration turned to more frantic, expressing a sentiment of anger. After the failure of the negotiation with Prime Minister Khaznadar (Giorgios Kalkias Stravelakis) in front of the palace, they broke the banners of the saint. This sign meant that a riot could start. In 1836 a revolt against the generalization of conscription had already used the rituals linked to this saint: the protesters had gathered at the mausoleum and had drunk together from the same jar the water of the saint's basin in sign of solidarity and dedication to their cause ([20], 759).

It is the object of this paper to investigate, as an echo to Ibn Khaldun's (Tunis, 1332–Cairo, 1406) invitation ([143]), these dimensions of complexity in the relationship between time, space, society, and religion.

The methodological approach chosen for this study, that examines situations pertaining to the nineteenth century in the mirror of longue durée considerations, pertains to a vision of urban history inspired by the historical anthropology of space ([135]), by the historical anthropology of urban time ([70]; [56]) and by the micro-study of daily life practices ([33]; [98]; [118]; [126]; [133]). The period under consideration belongs to the Ottoman era in Tunisia, that began in the sixteenth century and ended with French colonization in 1881.

The posture of this research is also inspired by the method and tradition of Alltagsgeschichte ([103]). Studying daily life ([62]) is intended, in contrast to mechanist interpretations, as a way to unearth traces of spatialized practices of religion or of elements pertaining to religiosity under diverse temporalities. The intent is also to avoid interpreting history as the interaction of undiscussed blocks (space, time, identity, religion). Focusing on all the interstices inside and between such notions (spatial, temporal, identity-related, confessional) is seen as a way to complexify the theorization of the interaction between time, space and religion.

Following seminal suggestions by Robert [80], 160). The main sources for this study are chronicles dating back to Ottoman times. Chronicles represented the annals and minutes of the deliberations of the local city assembly ([86], 236). They were written by an urban notable, the secretary of the assembly and mirrored the spirit and nature of local urban order: civic life in the hands of the notables of the guilds and of the confessional communities. In addition to the minutes of the meetings, they also contained lists of notables as well as a narration of all the events regarding the city. In such civic chronicles (yawmiyāt, ḥawādith), all events pertaining to the city were described on a daily basis. This source provides precious access to the understanding of how people of different confessions interacted in the public space or on how religious practices or elements and manifestations of religiosity participated in the structuration of the urban space, with a different meaning according to the times of the day and of the year. Chronicles also allow one to reflect on the notion of sacred time and its various meanings.

Urban history now has a long history as far as the problematization of the definition of space is concerned ([124]). Revising conceptualizations that tended to define space as the passive receptacle and decorum of social phenomena and insisting on the importance of interactions and mutual formations is one of the main stakes of contemporary research ([109]; [149]). For that sake, new attention to the question of religion and religiosity has been growing. In contrast with previous studies that tended to elude this dimension, focusing more attention on institutions, power and trade, a series of studies has underlined the importance of religion in the structuration of urban space ([57]). This comprised methodological proposals aiming at refining the spatialization of religious life at the scale of urban neighborhoods ([104]) and at complexifying not only the definition of space, but also that of religion, not to be seen as an undiscussed block, as well as the interaction between both elements ([29]).

In other words, the study of the mutual formation of urban space and religion is a way not only to articulate the relationship in a more complex way, but also to reflect on the very nature of each element ([112]; [128]). Complexity studies, partly in the wake of a suggestion by Henri [97]), but also around Edgar [111]; [2]) developed since the 1970s, in the scholarly context of geography, and in that of sociology and history ([122]; [37]). Authors have insisted on the necessity to refine visions of the nature of space, to reflect on the question of the identity of actors ([46]). All this constitutes the context of what has been called the spatial turn, a paradigmatic evolution that has had important consequences as far as the exercise of historical research is concerned ([144]). Authors have also insisted on the existence in social sciences of a geographical turn ([100]). The study of urban temporalities has also been the object of stimulating theoretical efforts, following seminal impulses by Alain [42]) and Maurizio [69]). Attention was dedicated to the diverse nature of urbanity and urban life according to the times of the day or of the year, with for example studies on the urban night ([49]; [94]; [71]). The articulation between time and space was also the object of specific attention, notably following the creation of the journal EspacesTemps in 1975. More recently, scholarship insisted on the notion of urban rhythms ([102]) and has focused on the hypothesis of a possible temporal and rhythmic turn in the social sciences ([72]). In the wake of theoretical impulses initiated in France and Germany, reflections on temporalities are now one of the main stimulating issues of historical studies at the international scale ([34]; [63]).

In the context of biblical exegesis introducing further elements of complexity in the narration of the birth of monotheism ([127]), the study of religion has been the object of important revisions in the last decades too ([129]). Developing a dialogue with anthropology, historians have shifted from attention overwhelmingly dedicated to institutions, dogmas and norms towards everyday life practices. The study of popular forms of religiosity was already one of the main focuses of microstoria. Both Giovani Levi, with studies on exorcism in seventeenth century Piedmont ([99]) and Carlo Ginzburg, with studies on witches and agrarian cults in sixteenth and seventeenth century Friuli ([67]) have illustrated how, not only older forms of religiosity persisted in European societies, but also on how attention to such practices and beliefs invites to revise both the methodology of historical research and the very concepts of religion.

In general, the attention to the longue durée persistence of forms of religiosity of pagan origins in Europe has severely nuanced the image of a block a Christianity or even of Christianity as a dogmatic and practical block. Historians have shown how old forms of religiosity cohabited in Europe with Christianism throughout the Middle Ages and the early modern era, how Christianism accepted and included some of these beliefs and practices under the form of the cult of the saints or of shrines structuring rural and urban landscapes and on how such cults were repressed by the Inquisition and later by the normative apparatus of the Protestant Reform and Catholic Counter-Reform. Scholars have shown the importance in rural medieval England of beliefs and practices linked to the 'walking dead' ([68]) or to the power of female witches ([32]) for example. Some of these cults were more or less convincingly Christianized, some other remained hidden. Even if the use of the notion of syncretism requires careful methodological precautions ([30]), Christianism, too, is partly the result of various layers of such mixings, some scholars have argued ([48]). Methodological reflections on the study of popular forms of religiosity were part of the founding issues the Ecole des Annales has been dealing with since the time of Marc Bloch and Lucien Febvre ([85]).

This focus induced, too, renewed visions of spatiality. Urban morphogenesis belongs to such revisions involving stimulating ways of conceiving the relationship between space and religion, through attention to the anthropological dimension and to the complexity and temporal inertia of phenomena of spatialization. An entry through temporality allows, in this case, to articulate longue durée perspectives and various layers of time. Such interpretations have been developed in the case of Paris, with a focus on the importance of pagan rites in the historical structuration of the urban space in the Middle Ages and early modern era ([51]) and of Ancient Rome ([52]). The spatialization of theological impulses around the notion of morphogenesis also became a fruitful field of reflection ([53]). The notion of religious landscape, such as developed in the early twenty-first century, includes this kind of perspective ([154]; [134]). It considers the complex historicity of shrines and places of cult or devotion, as well as the complex identity of the individuals and groups using them. It insists on the dimension of constant negotiation a religious landscape is the result of:

L'ensemble de ces signes, de ces repères, forme ce qu'on appelle désormais un paysage religieux, entendu à la fois dans sa matérialité visible et métaphoriquement comme le spectre d'identités religieuses multiples et négociées. La notion de paysage religieux naît de la constatation que le culte et les rites n'existent qu'en tant qu'ils sont ancrés dans l'espace, que ce soit de manière stable ou provisoire. Les temples, les sanctuaires, forment l'armature religieuse d'un territoire (de la même manière que les géographes parlent d'armature urbaine). (Scheid and Polignac 2010 , 431)

Articulating this research posture with the question of temporalities is the object of the present article. In the framework of cities and regions of the Islamic world, indeed, an effort at deconstructing static concepts and visions of the interaction between religion, space and temporalities has also been promoted by scholars. Nelly Amri, for example, proposed research postures aimed at avoiding interpreting religion as just a block of norms, thanks to an attention to historical anthropology at the micro-scale. Her seminal studies on female sainthood in Tunis belong to the key theoretical works on religion, space, temporalities, and society in general, beyond the only horizon of studies on the Islamic world ([10]; [12]). Her familiarity with original medieval sources helped her shape research methods fruitful for all periods and places. In this context of a stimulating scholarly landscape in the region, the very concept of Islamic city has also been deeply nuanced and discussed as for its role in existing reifications of the interaction between religion and society ([150]; [89]). Urban history in the region has been the object of stimulating methodological and interpretive proposals that challenged existing narratives on the very nature of cities ([4]). Visions of religions as civilizational invariants have been challenged, and an attention to negotiations, mediations and accommodations has developed. The aim here is to examine, around the case of Tunis in late-Ottoman times, how religion, temporalities and space were defined by daily life practices whose complexity allows one to interpret the nature of all three notions.

Tunis was in Ottoman times (1574 – 1881) a city inhabited by people of different religions and backgrounds ([139]; [1]; [130]; [19]; [19]). Populations of different identities, professions and confessions cohabited. The main religion was Islam. Its architectural and social manifestations marked both the urban space and the rhythm of time, from the ubiquitous presence of mosques of different types to signs revealing the presence of Islam in daily life: regular calls for prayer shaping the soundscape, individual practices, constant reference in the institutional order. Another important local religion was Judaism, with the presence of numerous synagogues in the old city, each one corresponding to different types, forms of communal life and religiosity.

The space and temporality of the city was also deeply marked by this Jewish component, from the built dimension to questions of identity. Jewish life, just like Muslim manifestations of religiosity, contributed to the definition of urban rhythms. The city also counted various churches, corresponding to different Christian communities (Sicilian, Sards, Venetian, Genoan, Maltese, Greek, Armenian...): these places contributed to the shaping and identity of their respective surroundings and urban temporalities tended to be organized around the cult calendar of each community. The geography of sainthood Christian churches referred to entangled various scales from local to references to places of origins. For example the church of the Genoans was dedicated to San Lorenzo, in reference to the Cathedral of Genoa ([20], 277). In 1877 the British consul obtained a licence for the construction of a Protestant Church on Sidi Mendil street ([20], 255). Beyond the cultual architectures, the shaping of space by religions took various forms: festivities, processions, identity markers, cemeteries.

The object of the present reflections is both to study the respective influence of all these elements on urban space and temporalities, and to explore some fascinating interstices between them, in order not to consider religions as blocks and the urban space as the mere projection of such blocks onto the city.

The medieval diversity of the population, that already comprised Muslims, Jews and Christians ([136]; [76]; [7]; [125]), as well as other religious practices, including Black African religions ([110]) and various forms of syncretism, had been reinforced after 1492 by an influx of Jewish and Muslim refugees that had been expelled from the Iberian Peninsula (together with some Christians who refused the new order). Jews, Muslims, and Christians themselves had different origins ([87]; [147]), traditions and cultures. The Jewish communities for example were constituted of people originating either from the Sahara, from the Iberian Peninsula or from local Berber tribes.

In the framework of the Ottoman empire, many complex identities were also represented in Tunis: converted Greek (like Prime Minister Khaznadar in the episode narrated in the introduction), Georgian or Serbian Ottoman administrators, Albanian soldiers, Turkish (of different origins and identities) civil servants, Arab civil servants from other cities and regions, converted Sard, Corse, Sicilian or Maltese navigators. The idea here is thus to refute interpretations of the interaction of religion, time and space that would be based on mechanical visions, and instead to decipher the complexity of interactions based on nuanced visions of identity and individual expressions of religiosity.

It is important too to have in mind the complexity of temporalities themselves, between longue durée constants or evolutions and the shorter duration of individual lives or of single events. Temporalities are also to be considered as complex for their characteristic of containing various ideas of time and as interpretations of various layers of pasts. Confessional identity, for example, was complex. In Ottoman times, many Muslims converted from Jewish or Christian families, or lived in Muslim families in which the memory of conversion had been kept throughout the generations. As for Christians, there were numerous Sicilians and Maltese, as well as Greeks ([81]). Many former captives in the framework of the Mediterranean economy of piracy had converted, numerous formerly Christian traders or members of guilds had also converted, and many inhabitants of Tunis counted a Maltese, Sardinian, or Sicilian ancestor. A very significant share of Muslim families came from Jewish families. Family names were often common between both confessions. It was not rare to have families with branches in both confessions.

The difference between religions should also not be read according to present-day European categories and clichés. Not only there are strong theological common features between Judaism, Christianity, and Islam, but also it was not rare to count a converted member of a local family of Rabbis among the circle of the Muslim religious notables (Ulamas). Mixed couples were common practice. The influence of Andalusian philosophy reinforced the theological space of a common ground for monotheist religions. The difference between Shia' and Sunni Muslims was not theorized as a matter of personal identity and of dogmatic blocks, it was rather a matter of theological erudition. Ibadism was strong among the Muslims of Tunis, particularly among merchants and traders originating from Djerba Island ([8]; [74]). What it meant to be Muslim, Jewish, or Christian was a matter of nuance and in no way a block of invariant identities. It is crucial to have this in mind when reflecting on the dimension of spatialization and to avoid projections on the past of present-day categories, themselves often inherited from the colonial period ([84]).

The existence of various layers of common believes and practices reinforced this proximity. It is not only that, on a theological point of view, Muslims and Jews had a lot in common. It is also that some common beliefs and practices dated back to their common North-African prehistoric and Ancient religious features, many of which had been integrated into local forms of Judaism and Islam after the conversion of North-African tribes respectively to Judaism and Islam (in some cases with a passage through Christianism or various forms of this religion later labelled as heretic, like Arianism) ([101]). Such roots, common to all populations of the region whatever the later conversion of their tribe, include features, on the importance of which historical anthropology has insisted since colonial times, like the omnipresence of Fatma hands as a propitiatory and protective sign ([121]), the belief in the protective power of the representation of eyes ([92]), the cult of trees ([26]) and in the invisible presence of Djinns and supranatural creatures of various natures ([55]). Complex processes like the conversion of entire North-African tribes to Judaism, Romanization, Christianization and Islamization included phases of accommodation and acceptation of such features into the new framework. Some historians have underlined the importance of syncretism in the formation of North African Judaism and suggested also that cults and beliefs inherent to the Punic religion at the time of the conversion of numerous North African Berber tribes helped the rapid spreading of this religion ([78]; [138]; [96]; [41]). Judaism in North-Africa is thus not a block, it integrated various elements of previous religious beliefs of the converted populations.

Comparable processes occurred with Christianism, either the dominant form or other forms like Pelagianism or Arianism ([43]) and with Islam. Historians, thus, insisted on the existence of numerous continuities between Ancient and Medieval configurations ([142]). None of these processes in '-ation' was, indeed, the importation of a block of beliefs: it was a process of adaptation, mitigation, syncretism, and superposition, that participated in the complex construction of religious beliefs in Ancient, medieval and Ottoman North-Africa. Colonial studies about the religions of Ancient North Africa ([117]) and about the prehistoric and ancient religion of the Berbers have provided numerous elements of information. They have also insisted on the echoes in early twentieth century's popular religiosity of such heritages ([23]). This represents an important corpus of indications. When using the data given by colonial science, however, one must have in mind the warnings suggested by Yvon Thébert:

L'histoire coloniale tourne autour de quelques thèmes qui tous, au fond, visent à fonder la légitimité de la conquête française et à la charger d'une mission civilisatrice. Dans cette vision largement ethnique 'l'indigène' est expulsé de l'avant-scène et réduit à une masse de manœuvre [...] la plupart des spécialistes de l'antiquité participaient à l'élaboration de cette vision [...] chacun se piquait d'ethnologie [...]. Le recours à la notion de 'berbère' dont le contenu semble assez clair pour se passer d'explication, continue à accroître les incertitudes. On risquerait une fois de plus de céder à une vision raciale, fort peu explicative, de l'histoire de l'Afrique du Nord. (Thébert 1978 , 64 – 65)

Always having in mind this crucial warning, one can investigate on syncretism and the enduring of old forms of religiosity. As far as Tunis is concerned, various elements have been studied, from the importance of magical practices ([55]) to the persistence of a prehistoric "mythical substrate" ([152]) that people of various confessions shared. There is for example the centrality of cosmic trees as elements of the religious landscape and as structuring elements of popular practices by Jews and Muslims ([115]). There is also the importance of some shared propitiatory rites, that echo in Judaism and Islam practices of previous religions that these populations had in common. Rain invocation counts among such practices, with common formulas and places ([120]). Rites of possession are part of this substrate too, as well as the sociability that surrounds them ([60]).

This popular religion, and the social organization that was attached to it, was the object in colonial times of a specific surveillance by the police, as it represented the potential basis for insurrections ([114]). It is also probably why, under French colonization, a militant of anticolonial resistance like Frantz Fanon, who had been a psychiatrist at Manouba hospital in Tunis, underlined how European visions of such practices were at the core of colonial prejudices ([59]). The popular practices of religiosity in Tunis, inserted in Judaism and Islam, were even part of the very nature of these religions, which should not be seen as constant blocks: they were built in a process of accommodation too. In Tunis, not all features of prehistoric or ancient origin can be summed up as Berber or pertaining to a local ethnicization of identity: there were multiple links with religions from the rest of Africa, due not only to the important presence in Tunis of populations of Sub-Saharan Africa but also to a longue durée circulation of beliefs and practices ([93]; [110]). These elements play an important role in the structuration of the local religious landscape and geography. They articulate in their very nature various temporalities too. For most of the places of cult, archaeology, epigraphy, or philology have illustrated links to previous shrines or sacred spaces. There are strong continuities in the localization of cults along millennia. There is also a form of inertia: some locations that were apparently not the object of an organized devotion during a certain period could very well be reactivated in a new context under a form that evoked their previous functions. Some places were also kept secret among certain circles (religious notables, women, guilds, inhabitants of a certain neighborhood, members of a brotherhood). Some places had a religious or spiritual meaning and were part of the religious landscape without even being marked by a built element or a visible sign. Some built structures whose function might seem not specifically religious were in fact places of religiosity too, like hammams (baths).

This geography was (is) known by all. This theme constitutes a stimulating horizon of contemporary historical anthropology ([6]; [39]; [15]).

The cult of the saints too, is the object of important studies. It was already one of the focuses of studies in colonial times ([50]) and has been, on a methodological point of view, at the core of reflections on the way to decolonize anthropological knowledge about such dimensions ([107]). Renewed approaches, inspired by the social sciences, have been proposed ([145]; [83]; [14]; [17]). One of the main achievements of this trend of research is to interpret this kind of religiosity as fully part of society and not as marginal practices ([10]). It also led to attention being paid to individual religiosity, its spatialization in the context of reflection on the nature of the cultural and religious landscape:

Fruit d'un mystérieux échange entre plusieurs sensibilités religieuses accumulées dans un espace en partage, cette identité est caractérisée au Maghreb musulman par un paysage cultural rassemblant dans ses recoins les plus reculés des lieux de sépulture, des mausolées, des mosquées et des zaouïas [...] Les us et les coutumes tribales ont été mêlées aux préceptes religieux, facilitant la rencontre entre personnes appartenant à des groupes ethniques et confessionnels différents. (Aïssa 2011 , 15 – 16).

Reflections focus on the complexity of the relationship between history, beliefs, rites, and their temporalities (Amri 2008) as well as on a relativization of the dichotomy between the world of religious notables and popular religion ([38]): a scholarly consensus has now been established around the fact that beliefs once stigmatized as popular were (are) in fact shared by all social strata. Historians also insisted on the importance of the sociology of the marabouts (literally the people with the power to establish connection between the visible and the invisible) in the structuration of property in Tunisian towns ([54]). Magical practices had a status of performativity in the real world. As Dalenda Larguèche states:

Notons que l'Islam n'était aucunement hostile à certaines de ces pratiques fondées sur la magie. Bien plus, sous l'influence maraboutique, l'Islam a su accaparer quelques-unes des notions de magie en matière d'incantations et de talisman. (Larguèche 2000 , 216)

The religious landscape of Tunis partly derived from its nature in ancient times: it was characterized by the presence of numerous cemeteries and shrines but had long been, as for urban functions, in the shadow of neighbouring Carthage. In Islamic times, Tunis developed as a city. It became however the capital city of Ifrīqiya only in 1229. This new function was mirrored in the construction of numerous mosques and places of cult ([44]), among which the most important was the Ezzituna mosque (Zitouna, Al-Zaytūna, meaning Olive in Arabic) ([153]). The mosque was built in the eighth century, with massive re-use of Ancient architectural materials from Carthage ([61]; [131]), on the site of a Christian shrine dedicated to Saint Olivia (448 – 463), itself built upon a prehistoric shrine. Olivia of Palermo is said to have lived at the time of the Vandal Kingdom of Carthage as a hermit in Tunis and to have had medical magic powers. She is said to have been killed in the context of the fight between the Arianism form of Christianism the Germanic Vandals were following and Nicene Christianity ([151]). The seventeenth-century chronicle of Ibn Abi Dinar provides information about the way this history was integrated into the Islamic narrative of the city. So does the nineteenth-century chronicle by shaykh al-Islām M. Bayram al-Khāmis:

The al-Zitūna mosque is the first jāmi' built in the capital (al-hāḍira). Its construction was completed in the year 141 h./758, according to what is written on the arches of the room dedicated to prayer. The date is engraved on the arch facing the holder of the Quran. This mosque is the largest in the city and brings baraka (good fortune) to its inhabitants (...). The mosque of Al-Qasr was a church (Kanīsa) before the Conquest. (Bayram t.2, 346)

In the religious landscape of Tunis, and even in the whole Islamic world, the Ezzituna mosque remained in Ottoman times a major referent and the seat of a very prestigious Islamic school and university, teaching religious scholars ([38]). Its nature thus invites to refute too strong dichotomies between the popular religion and the religion of the religious notables: it is in the nature of Islam, as chronicler Bayram himself underlines several times in his writing, to accommodate both forms of religiosity and in Ottoman times the religious scholars and notables of the Ezzituna mosque were actively participating in the popular religiosity embodied and spatialized by shrines and the cult of the saints. There was even in Ottoman times a "mystical geography" ([20], 456) inside the Ezzituna mosque, with regular appearances of saints in different angles and behind certain pillars, including appearances of the Khadr, the syndic of the Guild of the Saints and 'Abd al-Qādir al-Jīlī, the founder of the Qādiriya brotherhood in Baghdad. The old city of Tunis comprised numerous other mosques, including, as Bayram states, seven of Hanafite rite (Bayram, vol. 2, 346). All of them, as well as the itineraries between them, constituted the elements of a magical geography, whose nature is key for the understanding of the very nature of the local religious landscape.

The location of synagogues in the medina of Tunis also responded to such logics ([105]; [28]; [65]). In addition to the great medieval synagogue of the Hara (Jewish neighborhood), numerous smaller synagogues, linked to the religiosity of specific guilds or communities of origin (Djerba, Al-Andalus, Sahara, Livorno...) constituted the elements of the local Jewish religious landscape. Numerous elements of the local magical geography were shared with people of other confessional affiliation.

In general, in the city, the cult of the saints represented a complex geography that linked people's religiosity to strong forms of spatialization (Amri 2008; [14]; [17]). Each sanctuary, or place, was the object of seasonal visits (mawsim) and processions, that participated in the very definition of space:

La paix, impliquant la neutralité de l'espace et du temps, œuvre à la création des " Mousim ", sortes de rassemblements religieux et festifs donnant l'occasion de visiter le mausolée du saint, d'activer le négoce et de se réapproprier, à travers les processions, le pèlerinage. Et c'est aux prodiges du saint, indiquant ses multiples interventions (de) sécuriser l'espace d'inviolabilité. (Aïssa 2011 , 16)

The very definition of space should take this dimension into account, as all places were connoted in a way or another by a such dimension. Temporalities, too, were entangled, at different temporal scales: for many such places there were forms of continuity from Ancient to Ottoman times ([66]), the uses of the places were regulated by a precise calendar. Reverence to the saints, and to the previous spiritual manifestations of their existence and recognition could take the form of mausoleums (zaouia) ([13]; [18]), mosques, small shrines, or even unmarked spaces whose imperious meaning everybody was aware of. Bayram lists zaouias and insists particularly on those dedicated to "Sīdī Mahres, Sīdī ben 'Arūs, Sīdī 'Alī ben Zyād, Malek ben Anīs, Sīdī Mansūr". He also notes that "one could add at least one hundred more zaouias to this list" (Bayram, vol. 2, 346).

Some spaces that could be seen as voids in the urban fabric were in fact full of spiritual meaning. Among the most important places in this spiritual, but also very concrete, geography of the city, there is for example the shrine of 'Sīdī al-Halfāwī, a saint known in the city for his therapeutic virtues and particularly revered by women. The zaouia of Sīdī al-Halfāwī was in Ottoman times a major place of devotion for people in search of healing ([95], 221). The main place of urban devotion to the saints, however, was the mausoleum of Sīdī Mahres. In addition to his powers of protection of the city, he was said to have instituted an official protection over the Jews and their right to own property, build houses, public buildings and commercial facilities and have their own confessional liberty. The actions of the saint could be called as arguments in disputes over construction in Ottoman times: at a certain point, an extension of the Jewish neighborhood was planned. Some argued that this extension went beyond the limits of the power of the saint, who was supposed to have indicated the location of the Jewish neighborhood by launching a stick. A winning argument was to say that there was no limit in the launching power of the saint and thus no limit for Jews to build ([36]). This illustrates the legal value of beliefs or mythology in real life, but also their role as an instrument of social and confessional mediation. It illustrates too how temporalities had blurred frontiers. Interpretive suggestions by Paul [148]).

Another major cult, with highly significant forms of spatialization, was that of Sayyda Mannūbiyya, a female saint of the thirteenth century ([31]). Her shrine constituted a major landmark in the religious and social landscape of the city. Values of the feminine Baraka, a capacity to be linked to the world of spirits and to transmit good luck, were associated with the frequentation of this shrine. Scholars have insisted on the importance of such cults and beliefs in Tunisia ([5]) and on the centrality of the feminine cults of saints in popular religiosity ([22]). The general framework is that of a definition of Sufism which included such dimensions ([16]). Feminine cults of saints and propitiatory rites linked to fertility, healing, family life, luck and wellness were also the basis of important forms of sociability, whose spatialities were central structuring elements at the micro-scale of neighborhoods and streets and whose temporalities provided strong elements to the rhythms of social life.

Sisterhoods, like the tijaniyya in Tunis were an important element of sociability too ([106]) and in this framework feminine practices were a key element in the definition of space and of gendered regimes of spatiality and temporality ([35]; [95]; [94]; [40]). Various women being venerated as saints in the same shrines by both Jewish and Muslim women (Amri 2008), probably in echo to common Berber ancient practices and beliefs, had their shrines, whose networks constituted a distinctive element of the religious landscape. Feminine rituals invite to rethink the relationship between individuals, gender, space, and religiosity ([146]). Other saints, like Ahmad b. 'Arūs, were the object of precisely spatialized cults in the city too, under the form of shrines that people visited (ziyara) on precise occasions ([45]; [9]).

Shrines and places of rituals constituted protected places for all. Some old places of devotion could be reactivated, as in the case of the shrine of the Shadhili order of Sunni Islam, built on top of the Djellaz hill by Prime Minister Khaznadar under the Patronage of Sadoq Bey ([132]). Khaznadar also built in the 1850s the mausoleum (zaouia) of Sidi Ali Chiha in the Halfaouine neighborhood of the medina. This act of piety, spiritual and political, was aimed at demonstrating his local belonging and at inserting his own actions into the framework of this religiosity. Giving to a zaouïa was an act of charity and euergetism, as poor people were generally fed in the premises.

The religious geography of shrines is also to be understood through processions and festivals, that expressed their spatiality in relation to the whole territory of the neighborhood and whose yearly rhythm was key in the very definition of the local calendar. Each neighborhood had its own shrines, like the shrines of the saints Sidi Ali el-Gorjani, Sidi Mohammed el-Chérif or Sidi Hacen el-Séjoumi. Displaced shrines, as in the case of the mausoleum of Sidi Kacem el Jellizi, built in the fifteenth century and expressing the religiosity of Andalusian refugees, acquired an important value in the urban fabric. In this form of religiosity, a shrine can be displaced when the population is displaced (an echo to this dimension can be found today, from Marseille to Sarcelles, [119], in the religious landscape of populations who settled in other lands). This zaouia was enlarged in the seventeenth and eighteenth centuries. Outside of Tunis, a major place of devotion, on a hill over Carthage, location of the lighthouse since ancient times, but also probably of various cults, was the shrine of Sidi Bou Said (1160 – 1231). The sociology of the frequentation of these shrines invites to deconstruct rigid categories of identity. Some shrines were common to various confessions, some shrines expressed gendered differences, some expressed values of coexistence and cohabitation.

All these elements are constitutive of the religious landscape of Tunis. They suggest interpretations of spatiality that nuance visions based predominantly on power, norms, institutions, property, or trade. They also suggest, as for temporality, an attention to the mixing of various layers of time and to the definition of the present in intimate relationship with perceptions and narrations of the past. When reflecting on the definition of religiosity, attention to Ḥisba, this negotiated and spatialized implementation of rules of moral and religious inspiration, suggests defying from rigid and normative interpretations ([77]; [3]; [64]; [123]; [90]).

The study of urban religiosity through the geography, anthropology and sociology of shrines also reveals the importance of alternative logics in the organization of space and in the distribution of symbolic values on the territory. It suggests, too, to insert this dimension in the very definition of space and time. Through an attention to shrines, living-together patterns can be understood outside of normative visions or of visions of society as made of adjacent blocks. What the micro-scale reveals is that the spatialization of confessional identities did not respond to a fixed geography. This opens to a critical discussion of sometimes too rigid paradigms describing the spatial inscription of communal identities or of the very nature of such communal identities.

This also invites to reconsider the relationship with temporalities. An examination of the territorial and temporal modalities of the communalization of religion ([75]; [87]) is thus crucial. In the case of Ottoman Tunis, the existence of mixed quarters (Jewish and Muslim) and of mixed shrines is a key element to be considered to avoid static visions of this complex reality. A study of daily interactions, passages, circulations, mediations, conflicts, and regulations through the spatialization of complex forms of religiosity allows one to develop nuanced visions. Daily life, and the importance of common festivals and celebrations, the centrality of commensality and cohabitation, suggests defying from exclusive paradigms. This vision of the interaction between space, time and religion also invites to revise the very definition of urbanity ([140]). The spatialization of the self was often more complicated than just the projection onto the urban space of pre-defined identities ([47]) that colonial biased reifications and their contemporary echoes have contributed to impose.

For who wishes to understand aspects of daily life in late-Ottoman Tunis, two major sources are available: Ben Abī Al-Ḍiāf's chronicle and Bayram's writings, which are constituted of both a chronicle and various reports, letters, treaties, and essays. These two men were members of the urban elite and had important administrative functions. Ben Abī Al-Ḍiāf (1804 – 1874) was a member of various local councils as well as a secretary of the Bey. Writing the urban civic chronicle was part of his official duties. Bayram V (1840 – 1889) belonged to a lineage of important figures in the religious life of the city. Among his ancestors are indeed numerous naqīb al-ashrāf (a figure at the head of the group of notables whose genealogy claims a prophetic ancestry), mufti (a jurist specialized in Coranic and hadith studies, as well as islamic theology and jurisprudence), muḥtasib (a civic and religious function in the context of the concept of good governance – Ḥisba), and shaykh al-Islām (higher authority among the religious scholars -'ulamā'), a function that he himself embodied. He belonged to the circle of late-Ottoman reformists in Tunis. After the French colonial occupation of 1881, he fled to Istanbul and then finished his career in Cairo.

The reading in terms of micro-spatiality and temporality of the writings of these two men teaches a lot about the complex entanglements the religious sphere was the object of in urban Tunis. It also allows one to decipher not only the urban manifestations of social phenomena having to do with religion, but also to explore all the interstices and blurred dimensions between aspects of social life that sometimes tend to be analysed in normative terms.

The city of Tunis was intimately marked by manifestations of Muslim faith, which not only used existing spaces but also gave them their substance and meaning. Muslim religiosity was also a major element in the definition of the city's calendar and of its rhythms at various temporal scales (from year to day and from season to precise moments, from Fajr (sunrise) to Dhur (Zenit) and 'asr (sunset) and 'Isha (midnight). The Friday prayer constituted one of the main moments of expression of this faith. Many believers gathered in the various mosques of the city. This created moments of devotion in the neighbouring streets, as well as moments whose nature was defined by religiosity. The five daily recurrences of the prayer, with the call to prayer and joy (hayy 'al-ṣalāt, yahh 'alā-al-falāh) by the various muezzins of the city's mosques, also participated strongly in the construction of urban temporality and of the soundscape of the city. Chronicles and sources show how every disturbance in this rhythm was the sign that something grave happened.

When reflecting on the spatiality and temporality of religious principles in urban life and on the role of normative influences, one must also consider the conditions of implementation of Ḥisba, this collection in the form of treaties of norms of good Islamic governance and behaviour ([90]). Sources like the chronicles provide indeed indications on the debates in society about the implementation of such normative texts, themselves the object of practical considerations by their writers and the results of numerous phases of negotiation and adaptation. One example, about the year 1823 (1239H.) in Al-Ḍiāf chronicle (Ben Abī Al-Ḍiāf, vols. 3 – 4, 144 – 146) is very telling. On the day of the celebration of Mawlid al-Nabawī indeed, the chief of the police (al-mizwār) ([88]) executed a woman and two Christians on Kasbah square in repression of what he considered their bad behavior. The civic chronicler dedicates long pages to this tragic event, explaining how the police was wrong and how their reference to Ḥisba as a justification was biased.

Muslim religious life was also marked by the yearly recurrence of specific festivities, that shaped both temporality and space. For Eid al-Fitr (celebrating the end of the month of Ramadan during which Muslims fast from sunrise to sunset) and Eid al-Kebīr for example, processions (mawākib) were organized in the streets of the city. They mirrored the social organization at the scale of neighborhoods (Bayram, vol. 2, 723). The yearly calendar was rhythmed too by the recurrence of Mawlid, celebrating the birth of the Prophet of Islam. This festival, on the nature of which theological debates are numerous ([137]), was marked in Tunis by both celebrations at the court, at Ezzituna mosque and by popular celebrations. As Ben Abī Al-Ḍiāf's chronicle illustrates, every inflexion in the urban geography of celebrations had an institutional, spiritual, and political meaning, as on Friday March 4, 1814 (rabi'I 12, 1229), when the prayer of the members of the court

happened for the first time in the al'Wazīr Yūsef Sāhib al-Tāba' mosque in the Halfaouin neighbourhood. It was the first prayer to be held in this mosque. The Bey and his ministers attended it, as well as the people of the majlis al-shar'ī. The first preacher was our shaykh Al-'Alāma A. Mohammed b. Husīn Bayram. (Ben Abī Al-Ḍiāf, vols. 3 – 4, 59)

The chronicle provides a list of the persons who attended, with their respective functions. The whole ceremony was a form of institutional and theological pedagogy in this popular neighborhood. The aim was to teach the crowd how to celebrate according to the conception of religion embodied by the power and its different elements, in contrast with spontaneous popular celebrations that tended to be less formal. In this year, as the chronicler explains on the next page (Ben Abī Al-Ḍiāf, vols. 3 – 4, p.60 – 61), there was also an important geopolitical and theological tension at the scale of the whole Ottoman empire: the Wahhabi had just conquered "al-Haramayn al-Sharīfayn" (the two sacred Mosques) of Mecca and Medina, imposing a vision of religion that rejected elements considered concessions to paganism. The celebration of a festivity, sometimes considered as inherited from paganism, by the court in a popular neighborhood might thus also express a message of solidarity of Tunis with the Ottoman government in its rejection of Wahhabism. In the chronicle, the narration of these events is followed by a whole theological dissertation on the meaning of the various forms of cult and devotion in Islam.

This entanglement between religion and geopolitics, the one shaping the other, and vice-versa, is also due to the importance in Tunis, just like in many cities of the region, of the local networks of the Mecca Pilgrimage. The organization of the yearly pilgrimage not only provided, with all the phases of its preparation, an important rhythmic feature in religious and urban life, but also participated in the definition of the identity of numerous notables.

Other recurring festivities are also mentioned in the sources. On the occasion of a circumcision, for example, specific temporalities and phases contributed to inscribe celebrations into the public life of the neighborhood. This included songs and manifestations of joy, as well as very precisely defined moments of ritual (Bayram, vol. 2, 724).

The month of Ramadan was also one of the main temporal markers of Muslim religiosity in the city. Fasting, prayers, expressions of charity and kindness, as well as collective meals after the sunset shaped the geography and rhythm of the city. Ramadan was also an occasion for games, music, dance, and profane activities. Shaykh al-Islām Bayram, while rhetorically regretting that this sacred time not be fully dedicated to prayers and devotions, describes people's habits to go to specially built temporary amusement parks inside the old city during the nights of the month of Ramadan:

There are no amusement parks with games in the city, except during Ramadan. These are places for boys, where, during the night, they play. There are also shadow theatre plays behind a curtain [...]. It's actually someone who makes these puppets talk and the place is not very clean. Only boys go in. In general, only boys who have no manners go to these places to pass the time, have fun, and laugh [...]. These places of amusement keep them awake until the early hours of the morning during Ramadan nights [...]. It is the same as for those who play music, chess, or checkers in the cafés. In fact, they and those who work in these amusement parks do not devote themselves to worship during the nights of Ramadan. They pursue ghosts (ghūl) and the devil (shītān). (Bayram, vol. 2, 687 – 688)

Some amusement parks were also situated inside the Frank neighborhood, outside of the walls of the medina, where most of the Christian inhabitants of the city lived in Ottoman times:

There are in this neighbourhood some European amusement facilities. There are also some European cafés and some hotels that respectable people avoid. (Bayram, vol. 2, 688)

The writings of the shaykh al-Islām illustrate how, during the late-Ottoman era, Ramadan time leisure activities were gendered: the specific religious calendar reinforced the gendered spatialization of some activities. But the strongest difference was social, as other remarks by Bayram illustrate:

Notables have the habit to celebrate Ramadan nights at home or at some friend's place. (Bayram, vol. 2, 688)

No indication about if they prayed more than the rest of the population or if they simply hid their profane activities beyond private walls. What notables did publicly instead, the chronicles confirm, is to provide food for the poor people of their neighborhood. The sources also provide indications about local elements of Islamic religiosity whose roots probably go back much further in history: prayers for the rain; cult of the saints, presence of shrines, regular ceremonies, and invocations in front of or inside these shrines, linguistic formulas and gestures with propitiatory or protective functions. The civic chronicle justifies the importance of such cults for the inhabitants of the city with a Coranic reference (sūrat 2 – 114) justifying the devotion to local saints of the past (Ben Abī Al-Ḍiāf, vols. 3 – 4, 68 – 69). Here again, different layers of historicity interact in the description of the religious landscape of the city.

These dimensions were often shared with local Jews, whose religion also integrated earlier elements. The forms and moments of devotion were often common. There were also specific Jewish spaces and rhythms, mostly around the celebrations in the numerous synagogues of the old city. During the Sabbat, the neighborhoods with a strong Jewish presence adopted a specific rhythm. In general, Jewish festivities marked the local calendar and space.

The same logic applied for Christian spaces and times around the Sunday Mass and the Holy Week mostly. One should not have of these confessional differences, however, the image of a city functioning according to separate blocks. Temporalities and spaces were intertwined, as most neighborhoods were mixed. Confessional difference also allowed the existence of blurred interstices: Muslims enjoying alcohol in Christian or Jewish owned taverns, or Jewish or Christian women converted to Islam to escape the rigidity of family law for example. As for the consumption of alcohol, chronicler Al-Ḍiāf narrates how it was common for Muslims to escape social control in their neighborhood in frequenting taverns outside, with the consequence for inhabitants of frequent encounters with drunk people (Ben Abī Al-Ḍiāf, vols. 3 – 4, 93 and 119 – 120; [88]).

Interconfessional contacts also occurred at the scale of neighborhoods, streets, and homes, often under the form of shared meals, or of exchanges of dishes. As anthropologists have illustrated ([21]), food in North Africa is not only a contextual element of celebrations on the occasion of religious festivities: it is a cult in itself. In other words, food is religion (this is not a metaphor) and commensality has a religious dimension, that chronicles and sources abundantly illustrate (Bayram, vol. 2, 721 – 722 for example). On many occasions, inviting neighbors was part of the celebration. Finishing a rite with a shared meal was part of the rite itself. There were specific recipes for all religious celebrations of the Muslim (Eid al-adhā, Eid al-Fitr, 'Āshūrā', mawlid al-nabawī), Jewish and Christian calendars as well as for all the events of confessional life (circumcision ...) (Bayram, vol. 2, 723 – 724). People of different faiths used to eat together and exchange dishes and desserts. Such exchanges were not only signs of sympathy and living together friendliness, they were not only responding to cultural traditions: they were part of the cultual dimension. Wedding banquets were also moments of interconfessional sociability. Not only people from the neighborhood, of all confessions, were invited, as well as members of the extended family, sometimes of different confessions too, but also the organization of the festivity traditionally required confessional mixity. Musicians, indeed, were mostly Jews, including at Muslim marriages. Shaykh al-Islām Bayram, expressing conservative opinions that other members of the elite, and even less of the ordinary people, did not necessarily share, presents the organization of weddings under the lens of degrees of morality:

During the night of the wedding, at the banquet (al-walīma), prostitutes ('āhirāt) are singing. As for divorced women, they never sing outside of the presence of their husband. Most musicians are Jews, as Muslims avoid this profession. [...]. On the seventh night of the wedding, when senior religious notables (kubār al-'ulāma) are present, music stops. (Bayram, vol. 2, 728)

What these historical explorations illustrate is how, in matters of religion, temporalities are entangled: references to far away pasts participate in the structuration of less far away other pasts, the interpretation of which constitutes a key element of a present time that itself soon becomes a past to which reference can be made. Thus, the pertinence of a reading of such social constructions of time in terms of regimes of historicity ([73]). Temporality can be seen as the complex interaction of different modalities of conceiving time, different perceptions of time and different rhythms. In this context, religion both shapes time and is shaped by it. The complex entanglement of temporalities religion is the mirror, the object, and the vector that also creates a certain atemporality: the time of religion, although always precisely inserted into the social context, is often presented in the sources as a transcending time.

In the years and decades that followed the events presented in this article however, the Ottoman province of Tunis experienced new configurations, in which conceptions of religious time were deeply challenged. The end of the Ottoman period, indeed, was marked by growing foreign pressures (by France and Great Britain mainly) of always more insistent colonial nature. This induced strong re-interpretations of confessional identities and of the urban spatialities and temporalities attached to them. Reformism, conservatism, and references to the past were increasingly inserted into rhetorical frameworks determined by this aggressive context.

With colonisation by France in 1881, religions and confessional identities were the object of the often violent imposition of a new order. This induced the imposition of new spatialities ([113]) as well as new interpretations of the past and of the rhythms of religions. With independence in 1956, processes of re-appropriation of the past and re-interpretation of identities were also key in the definition of the new national framework ([82]), in the context of the exile of a large share of the local Jewish population ([108]) as well as of the Italians ([116]).

The religious landscape of places of cult, shrines, and places of worship as well as the religiosity that is attached to it, still has, however, in contemporary Tunisia a central role in the definition of space and in the articulation between religion, space, time, and society. The Jewish presence in the city of Tunis, which remains strong and constitutive of important features of local identity, is today less important than in medieval, Ottoman and colonial times. Common festivals and pilgrimages still exist, however. Some shrines of Tunis and Tunisia were also recreated elsewhere. After the migration of Tunisian Jews to France and to post-Ottoman and post-British Mandate Palestine between the 1960s and 1970s, numerous local Tunisian shrines were recreated there ([27]; [119]). Muslims migrants also relocated some places of devotion abroad ([24]). During the 2010s, various attacks against shrines were conducted by factions inspired by religious visions hostile to this form of religiosity ([79]). They included the partial destruction of the shrine of Sidi Bou Saïd in 2013. Béji Caïd Essebsi (1926 – 2019), who became President of the Republic in 2014, in the context of post-revolutionary reconfigurations of power, insisted in his memoirs that he was born inside the Sidi Bou Said Mausoleum, under the benediction of the Saint Sidi Bou Said El-Béji, of whom he was given the surname ([58]). This illustrates how the aura of protection conferred by saints still has a performative value in present-day mentalities and politics.

In general, the interpretive entry proposed here might thus be useful for the understanding in current anthropology of the complexity of religious identities as well as of the complex value of references to complex pasts: all this invites to have in mind that religious temporalities are the object of constantly redefined entanglements.

Anonymous 1843. Kitāb al-ḥisba [The Book of Ḥisba]. Al-Maktaba al-Wataniyya al-Tunsiyya (National of Tunisia, BNT), Manuscript number 23870.

Bayram Al-Khāmis, Muhamed 2000². ṣawfat al-I'tibār bi-mustawda' al-amṣār wa al-aqtār, tahqīq [The Most Important Study on What is Contained in Capitals and Countries] (text in Arabic edited and annotated by 'Alī ben Tāhar al-Shannūfī and Riyād al-Marzūqī and 'Abd al-Hafīz Manṣūr. Carthage: Beit al-Hikma. 6 vol.

Ben Abī Al-Ḍiāf, Ahmad 1965 (2004²). Ithāf ahl al-zamān bi akhbār mulūk Tūnis wa 'ahd al 'amān [Celebration of People of the Past through Reports on the Rulers and on the Fundamental Pact]. Tunis: Al-Maktab al-Tārīkhiya/Al-Dār al 'Arabiyya lil-Kitāb. 9 vol.

Ibn Abī Dhiyāf, Ahmed 2004. Ithāf al-zamān bi-akhbār mulūk tūnis wa 'ahd al-amān [Celebration of People of the Past through Reports on the Rulers and on the Fundamental Pact]. Tunis: Dār al-'Arabiyya lil-kitāb. 5 vol.