The Orisha religion in a transnational perspective

Afro-Atlantic religions and transnationalism

Yoruba-derived religions in the Americas have been more or less transnational since at least the latter half of the nineteenth century. Yoruba religious networks have been forged through travels of Afro-descendants and more recently by the hosting of international conferences, pilgrimage tourism, and online exchanges between initiates on the two sides of the Atlantic Ocean. In a previous work (Capone, 2004a), I highlighted the link between the field of Afro-American studies and the transnational approach, suggesting that, from the beginning, the Afro-American universe had been structured according to a ‘transnational’ logic and not simply by the forced displacement of enslaved Africans. In the case of Afro-Atlantic religions, the transnational approach is indeed imposed by the ethnographic materials. Today, their study must also consider the ritual networks between Africa and Brazil, or between the different centres preserving African traditions on American soil, which have a direct impact on local contexts. We have therefore replaced the comparative approach underlining the cultural continuity between Africa and the Americas with a new one which takes the space of circulation of social actors, religious practices, symbols, and knowledge into account.

The points that I will raise in my keynote lecture are the fruit of a long personal journey going back to the beginning of the 2000s when I undertook the study of the transnational networks of Afro-Atlantic religions – particularly between Brazil, Cuba, and the United States – through the analysis of the re-Africanization processes within Brazilian Candomblé and Cuban Santería or Regla de Ocha. ⁴ Since the 1970s, we have been challenged by the spread of the Orisha religion that obliges us to study these religious phenomena in a network, linking different localities of the ‘Afro-religious’ Atlantic. Today, we can no longer understand a ‘local’ phenomenon without having a more ‘global’ vision of its developments and its confrontations with other religious contexts, in a dialogue that is often tense between the ‘diaspora’ and the ‘motherland’. This led me to create a research group in the early 2000s on the transnational networks of Afro-American or Afro-Atlantic religions and, from 2008 to 2011, to coordinate a research team on the transnationalization of Afro-Atlantic religions in the Americas and Europe, within the international project RELITRANS, funded by the French National Research Agency (ANR). ⁵ My reflection was thus nourished by a collective work ⁶ that contributed to the development of research on transnationalism in France, including several doctoral theses on the transnationalization of Afro-Brazilian religions in Europe (Portugal, France, Italy, Switzerland, Austria, and Germany). Today, we have thus gathered enough in-depth research, carried out in different national settings, to allow us to identify some recurrent dynamics within these processes.

Transnationalism or globalization

We are all aware that, in the context of what is called ‘cultural globalization’, religions are undergoing significant transformations. On one hand, there is an intensification of the translocal circulation of followers, symbols, and beliefs that once belonged to a religious practice linked to a specific historical and geographical context. On the other hand, local religious fields are opening up to new practices and new representations. Globalization has had an incomparable impact on the displacement of certain religious practices which, until then, had remained deeply rooted in specific traditions, territories, and social groups. The use of the term ‘globalization’ reflects this level of integration and interconnection, which today is expressed by the individual’s empirical perception of belonging to a ‘global world’, beyond any real territorial attachment. Thus, being initiated into one of the Afro-Atlantic religions implies the perception of being part of a whole that goes beyond the limits of one’s own daily experience in local worship communities, thanks mainly to social networks that bring together practitioners from different countries. However, this religious globalization often implies a reverse process, from South to North, from peripheries to metropolitan centres, or, in our case, from ‘diaspora’ to ‘motherland’ (Capone and Mary, 2012). Unlike the ‘globalization’ of the great monotheistic religions, these shifts are now taking place through polycentric transnational networks, which do not depend on missionary logic. As Csordas (2015) reminds us, these reorientations constitute the outline of a ‘global geography of the Spirit’.

In our research, we have chosen to stress the concept of transnationalization, which we felt was more relevant than that of globalization. In the early 1990s, in a pioneering work, Basch et al.(2000 [1994] 7) proposed the notion of transnationalism as a new analytical field for understanding migration, ‘to emphasize that many immigrants today build social fields that cross geographic, cultural, and political borders’. A few years later, Hannerz (1996: 6) stressed the importance of using the term ‘transnational’ instead of the ‘prodigious’ use of the term ‘globalization’ to refer to any process that crosses national borders, a term ‘in a way more humble, and often a more adequate label for phenomena which can be of quite variable scale and distribution, even when they do share the characteristic of not being contained within a state’. ⁷

The same unease with the use of this term has also been highlighted by Csordas (2009), for whom the influence of globalization is often thought to be fundamentally unidirectional, going from a globalizing centre to a periphery that would passively receive the global flows.

However, the studies we have carried out on the transnationalization of Afro-Atlantic religions show a reality that is diametrically opposed to this type of dynamic. For the most part, these are practices that generate multidirectional flows of people, goods, and religious values. These flows – and the networks they generate – can change, their intensity can increase or decline, and their scope of action can widen or narrow according to the various rearrangements of the religious modalities involved. Thus, while the term ‘transnationalism’ occupies a central place in the Anglo-Saxon literature, we prefer to replace it with the notion of ‘transnationalization’ which allows us to emphasize the processes, instead of underlining a sort of intrinsic quality of certain social phenomena.

Moreover, ‘cultural globalization’ is obviously not a new phenomenon. We know that world history has seen other phases of globalization, such as those driven by colonial expansion. Several authors have emphasized the historical depth of these phenomena, since even among immigrants to the United States at the turn of the nineteenth century, transnational links between the society of origin and the host society could be observed. Transnationalism is therefore not a new phenomenon, but a ‘new perspective’ (Portes, 2003: 874). The transnational approach – ‘the transnational lens’ – helps us to capture contemporary phenomena as the current ‘sense of simultaneity’ (Levitt and Glick-Schiller, 2004) of inhabited worlds or the feeling of living in a ‘global world’. This ‘sense of simultaneity’ is due not only to what has been called the ‘death of distance’ which characterized our pre-COVID era, but also – and above all – to the rise of social networks, among them Facebook and Instagram, and of new communication tools, such as WhatsApp, which enormously facilitate contacts between initiates of different nationalities with the help of Google Translate. The multiplication of these global interconnections makes it possible for initiates in the Orisha religion to develop the feeling of living in shared worlds that are being built on both sides of the Atlantic.

This new perception brings several changes in Afro-Atlantic religious practice. First, the multiplication of links between the various centres preserving African traditions, which are not only in Africa, but also in the Americas. Today, different traditions of the Orisha religion are present in the Brazilian religious field, confronting Brazilian Candomblé with ritual practices from Cuba and Nigeria. This has contributed to the creation of an ‘ecumene’, to use the notion proposed by Hannerz (1989), for whom the Greek term oikoumene designates ‘a region of cultural interaction and exchange’, a notion which also evokes that of the black Atlantic, elaborated by Gilroy (1993). This transnational space, this ‘Yoruba ecumene’ – Yoruba culture being the origin claimed by many Afro-Atlantic religions – is then a co-construction not only between Africa and the Americas, but also between the various centres producing ‘traditional’ discourses and practices on the American continent, such as Salvador de Bahia in Brazil, or Havana and Matanzas in Cuba.

Furthermore, the study of Afro-Atlantic religions questions the very opposition between the bond to a territory (the ‘local’ or the ‘national’), which would produce ‘pure’ and ‘authentic’ cultures, and the deterritorialization associated with the transnational, which would, in turn, put forward ‘hybrid’ or ‘creolized’ cultures (Werbner, 1997). In reality, at least for these religions, transnationalism does not prevent essentialist discourses, in which culture, despite its obvious transformation and adaptation, is still thought of as ‘pure’ and ‘traditional’. Transnationalization does not necessarily lead to phenomena of hybridization or creolization, as is shown in Candomblé’s ritual re-Africanization (Capone, 2016a).

Transnationalized practices often reinvest in their localities of origin by reviving cultural traditions, which are promoted to the rank of vectors of universality, as in the case of the Ifá cult, a divinatory system organized around two ‘national’ traditions, the Nigerian and the Cuban. I will therefore focus my analysis on the new configurations generated by the unprecedented encounter between different models of tradition within the Orisha religion.

Before presenting some ethnographic data, I propose to examine two issues that are at the very heart of my research on transnational fields: the weight of the nation, which is not automatically erased by transnational processes, and the possibility of being transnational ‘without moving’.

The weight of nation

In our research, we have worked with two different concepts: ‘transnationalization’ and ‘translocalization’. However, the terms ‘translocal’ and ‘transnational’ do not refer to the same processes or the same scales of analysis. While it may be appropriate in certain situations, the term ‘translocal’ evacuates the weight of the nation as a receptor or exporter of religious practices, which are often conceived as part of a national cultural heritage, as for Afro-Cuban religions that bring with them a strong nationalist component. Indeed, the weight of ‘national imaginaries’ makes Afro-Brazilian and Afro-Cuban religions – based on traditions emphasizing a logic of dual affiliation – both endogenous and exogenous religious practices. They are of African origin but are also – and above all – Brazilian or Cuban. These ‘national’ identities are inscribed at the very heart of religious identities. For example, when a Candomblé follower is initiated into Ifá in Brazil according to Cuban tradition, he will learn ‘Cuban style’ Spanish in order to be able to read Cuban Ifá treatises or other sacred texts, he will adopt Cuban prayers and invocations, and he will seek contacts with Cuban or Cuban–American initiates, while negotiating the place of this new religious identity alongside his previous ritual affiliations. This also involves the use of a national language that becomes the ritual language in processes of religious transnationalization. For example, the Orisha manifestation in the United States is often subject to ‘truth regimes’ originated from its context of origin, namely, Cuba, from which Orisha worship arrived in the 1960s. The use of Cuban-style Spanish by the embodied Orisha then becomes the proof of the veracity of the initiate’s possession.

A ‘real’ Orisha will then speak Spanish in Lucumí ceremonies in Miami or New York, as a ‘real’ spirit will speak Portuguese in Umbanda rituals in Paris or Rome. In the implantation of Afro-Brazilian and Afro-Cuban religions in the United States, Argentina, Uruguay, or Europe, the use of language thus acquires a ‘sacred’ character. Communicating in Portunhol, a mixture of Portuguese and Spanish, has become the sign of an ‘authentic’ possession in the countries of the Latin American Southern Cone (Frigerio, 2011).

However, the mastery of ritual language is also accompanied by the production of multilayered identities, since in these religions the notion of conversion does not apply. Indeed, when a practitioner undergoes new initiations, he does not abandon his previous religious practices but rather accumulates them with the new ones. Sometimes, these stratified religious identities also generate ritual changes that can lead to an ‘ethnicization’ or even ‘racialization’ of Afro-Atlantic religions, as in the North American case of Oyotunji Village (Capone, 2005; Clarke, 2004). This revives the tensions between an idealized African origin and a diversified American religious practice that has long included whites as well in countries that are still suffering from structural racism and social inequalities. Therefore, the ritual rearrangements within the Orisha religion do not abolish the different national scripts, but remain indebted to the national histories of each country (Frigerio and Oro, 2005).

The term ‘transnational’ should not then automatically refer to the erasure of the nation-state, since it often represents a redeployment of its prerogatives. Today, social actors can also embody their ‘national culture’ or a certain ‘national ethos’ in transnational settings. As Clifford (1997) reminds us, translocal phenomena are always embedded in particular geographies and histories, which are very often those of nation-states. The enduring presence of the ‘national’ in the ‘transnational’ is one of the most stimulating challenges in transnational studies.

Transnational (im)mobilities or how to be transnational ‘without moving’

The second point I would like to emphasize is that religious transnationalization is not necessarily linked to migration. The theme of (im)mobility is gaining increasing attention in the field of migrations (Easthope, 2009; Moret, 2020; Schewel, 2020), highlighting the mutual constitutive relationship between mobility and immobility (Capone, 2010b; Kaell, 2021; Rocha and Castro, 2021). We know that it is in the field of international migration studies that the theoretical frameworks of transnationalism have been drawn up. This new approach emphasizes the links between ‘here’ and ‘there’, staging a ‘circulatory territory’ (Tarrius, 1999) in which migrants develop an awareness of belonging to two worlds at once, to their land of origin as well as to their host country. In the early 1990s, Basch et al. (2000 [1994] 6) proposed the term ‘transmigrant’ to refer to the multiplicity of relationships that immigrants maintain with their country of origin.

However, while it is true that the activities of religious ‘transmigrants’ – namely, the pais and mães de santo (the leaders of Afro-Brazilian religions) who cross borders to take care of their foreign initiates – have been fundamental in the process of implantation of Afro-Atlantic religions, our research has shown that the processes of expansion of these religious practices can also take place without a strong presence of immigrants and beyond any missionary enterprise. In reality, movement is not a prerequisite for any transnational action. While some migrants periodically cross borders, there are also a large number of individuals whose lives are deeply embedded in the host society. Nonetheless, these immigrants are still inscribed in networks that link them – through flows of people, goods, or information – to their country of origin. They may not move physically, but they live their lives in a context that has become ‘transnational’, imagining themselves as members of a group that is constituted across space.

Individuals who do not move maintain social relations across borders through various forms of communication. As Levitt and Glick-Schiller (2004: 1009) point out, the actions and identities of those with weak connections to their society of origin are not less influenced by transnational dynamics. Hence, migrant communities are not the only ones affected by religious transnationalization. Transnational networks can also be formed while remaining at home, thanks mainly to the new means of communication, such as the Internet. The concept of (im)mobility can then be a tool for thinking about how some people are able to fully explore transnational networks, while others continue to practice their faith on a local basis in a world where digital technologies have become crucial to sociality and the construction of religious identity. Salazar (2011) has stressed the centrality of imaginaries in providing the cultural material to be used for the creation of translocal connections. Initiates who navigate in transnational networks accumulate what she calls ‘cosmobility capital’, resources, knowledge, and abilities that facilitate social, geographical, and religious mobility. This issue has accelerated with the COVID-19 pandemic, but previous works have demonstrated that mobility is a contested ideological construct involving much more than mere movement. It will be then interesting to explore the intersection of (im)mobility and new configuration of belonging in a digital world.

Transolorisha: agency and power in transnational networks

In a book in preparation (Capone, forthcoming) on the religious diasporic connections between Brazil and Nigeria, I propose the term ‘transolorisha’, where olorisha designates an initiate in the Orisha religion (o + ní + òrìsà = the one who possesses the Orisha), for the initiates circulating between American and African countries. ⁸ A transolorisha is thus an initiate in one of the variants of the Orisha religion in Brazil or elsewhere, who travels to Nigeria (or to another centre of Orisha tradition in the Americas) to carry out new initiations that will integrate her into new networks of social and spiritual relationships. The term ‘transolorisha’ captures this relationship to an elsewhere that does not necessarily imply a continuous movement between two localities. Transolorishas are not (trans)migrants but travellers between two cultures, always in search of a more direct contact with the ‘source’ of the ‘African tradition’ in a space that is being constructed on both sides of the Atlantic.

In our research, we conceived it as a space of relationships, as a space of circulation of people, objects, practices, symbols, and ideas. In other words, beyond the national and regional contexts in which these practices are deployed, there is a broader space that encompasses the multiple connections woven by the religious actors, and where a particular perception of belonging is experienced, the ‘feeling of translocality’ described by Falzon (2004) in his ethnography of the transnational networks of Hindu traders, or the ‘sense of belonging’ proposed by Levitt and Glick-Schiller (2004). But this feeling can also be produced on a virtual level. Indeed, the Internet, and especially social networks, have become a complementary but indispensable research field, which helps us to reconstruct the networks of individuals and groups interconnected by virtual links (Capone, 1999b).

While the transnationalization of Afro-Atlantic religions is made possible by a process of interaction between ritual specialists and individuals of different nationalities who move back and forth between different countries, one can also be transnational without moving. Indeed, many priests and priestesses of Afro-Atlantic religions do not necessarily travel, but have to handle religious knowledge and ritual protocols that can be different from those of their local context. Without physically travelling, they can remain connected to other places (through visits, telephone, especially WhatsApp, and social networks), sharing cultural codes, ways of thinking, or implicit references that allow for ‘conversation’ (Comaroff and Comaroff, 1991) with their ‘brothers in religion’, in Nigeria or elsewhere.

Research on religious transnationalization is thus not only about migration dynamics or the movement of individuals, but also about the consciousness of belonging to a ‘globalized’ world and the imaginary of being part of two worlds at once. The anthropologist can even do ‘multi-sited ethnography’ (Marcus, 1995) while remaining in one place, by analysing, for example, the connections, exchanges, borrowings, and tensions between different belief systems or between different regional traditions. This is the case, for example, of my fieldwork in Rio de Janeiro, where I work with specialists in Ifá divination, who follow several traditions: the Nigerian, the Cuban, and also, more recently, the Brazilian, a sort of synthesis of these two ‘national’ traditions.

Brazilians undergo initiations with Cubans and Nigerians, and must master the cultural codes and rituals of each tradition. This mastery, more or less accomplished, plays a central role in the processes we study. Practitioners of Afro-Atlantic religions can then develop a feeling of being members of transnational ritual networks, ‘inhabitants’ of a multi-territorialized space of relations, without necessarily ‘moving’. Indeed, initiates always emphasize their identification with ritual lineages, stressing the links with the ‘land of roots’, a relation that draw new strength from the gigantic weave of virtual links connecting initiates on the Net. However, mastering this plurality of ritual idioms also implies a form of reflexivity about one’s own identity, redefining the boundaries of the religious group and the different ‘geographies of power’ that link it to other religious traditions, as well as the possibilities of a dialogical relationship with other national traditions. The relationship between agency and power is at the heart of religious transnationalization processes. The implications of these processes of reconfiguration of ‘traditional’ power and prestige within transnationalized religious practices are, however, issues that have not yet received the attention they deserve in transnational studies. Some examples from my research on Ifá transnationalization in Brazil will show how a transnational context can profoundly alter the delicate balance between Afro-Atlantic religions, disrupting their power relations and hierarchical structures.

Power, gender, and possession

One of the most significant changes in the Afro-Atlantic religious field since the 1960s–1970s has been the diffusion of religious practices of African origin beyond ethnic and national borders and their implantation in new countries that did not necessarily have a long religious tradition comparable to those of the countries that saw the birth of Afro-Atlantic religions (Brazil, Cuba, and Haiti). In recent decades, these religions have also profoundly modified their image, occupying the public space in unprecedented ways and claiming their cultural specificity. From being ‘poor, black’ religions, they have been transformed, especially in Brazil, into religions whose practitioners today come from different social backgrounds, including foreigners, who import these religions into their own countries. Today, transnational circulation has thus led to an unprecedented confrontation between different models of tradition that must deal with a deeply stratified and fragmented universe. Indeed, while certain religious modalities, such as Ifá and Brazilian Candomblé, all refer to the same Yoruba origin, this ‘pan-Yoruba’ identity is never unique. On the contrary, it is characterized by its multiplicity, by different national religious identities and by their interaction, which is often conflictual.

In previous works, I have analysed the religious transnationalization brought about by the reintroduction of Ifá in Brazilian Candomblé temples. ⁹ I will just resume some points which disclose the impact that these processes are having on the Afro-Atlantic religious field through structural changes in religious hierarchy. Among the religious practices of Yoruba origin, the Ifá divination system occupies a central position. For the first time, the spread of the Ifá in the world and its knowledge inscribed in the corpus of odù, the divinatory ‘signs’, offers initiates into Afro-Brazilian religions the outline of a sacred Book that can be the basis of their religious practice. This special position allows the babalawo (the priest of Ifá divination) to develop hegemonic aspirations over the other local variants of the Orisha religion.

Recently, some religious leaders have even begun to claim World Religion status for what they call ‘Ifaism’, a religious variant based on the worship of Ifá/Orunmilá and his sacred scriptures, and which is often presented as a kind of monotheism. In Brazil, the Ifá cult – which is part of the Orisha religion since it is placed under the tutelage of the god Orunmilá – was revitalized in the 1980s, after falling into oblivion following the death of the allegedly last Brazilian babalawo in the 1940s. Yoruba language courses, which were quickly transformed into divination classes according to the Ifá system, prepared for the arrival of Nigerian babalawo in Brazil and, from the early 1990s onwards, Cuban babalawo who defend their own Ifá tradition (Capone, 2010a: 233–254). Indeed, Candomblé initiates seek to tirelessly expand their knowledge of Yoruba culture, complaining that ‘the elders’ did not transmit all the knowledge they had to the new generations. It is this unfinished transmission of sacred knowledge that has led to the transnationalization of other Afro-Atlantic religious modalities and their implantation in Brazil.

Issues of gender and possession are at the very heart of the rearrangements brought about by this religious transnationalization. In my work, ¹⁰ I have shown how the presence of Cuban babalawo in Rio de Janeiro allowed the initiates of Candomblé to become familiar with new models of tradition, where men, and in particular, the ogan – a ritual office reserved in Candomblé for heterosexual men who do not enter into a trance – found a new way to access the highest positions in the religious hierarchy in initiation into the Ifá priesthood. In fact, the babalawo priesthood, according to Cuban tradition, is restricted to heterosexual men who must not be possessed by the gods. Women can be initiated into the Ifá cult, but they occupy a lower place in the hierarchy, becoming iyapetebí, the babalawo’s assistant who cannot perform divination with the opelè (the chain of divination that is an attribute of the babalawo). The only possible consecration for women in the Cuban Ifá tradition is therefore the ceremony of Kofá or Ikofá, which corresponds to the first level of initiation for men, called Awo Fakan, a level to which homosexuals are confined.

On the contrary, in Brazilian Candomblé, religious power is concentrated in the hands of women (mães de santo) and of men (pais de santo) who are often homosexuals. All of them embody their deities in ritual ceremonies. The candidates for initiation who do not enter into a trance – that is, the ogan – occupy high positions in the hierarchy of the Candomblé temples, but they will always remain subject to the authority of their initiator. Thus, an ogan cannot initiate other people, because one of the necessary conditions for the reproduction of religious lineages in Candomblé is the direct experience of trance and the development of mediumnity, that is to say, the capacity to embody the deity of which the initiate is considered the spiritual ‘son’ (filho de santo). Without experiencing and mastering the trance of the gods, no one can initiate a novice in this religion.

In this encounter between Afro-Cuban and Afro-Brazilian traditions, linked by the same claims of a Yoruba origin, the main points of tension between the babalawo and the Candomblé initiates are then the importance given in Candomblé to feminine power, the role played by homosexuals, and the centrality of possession. In fact, in Cuban Santería, as in Nigeria, possession is not a prerequisite and an indispensable condition for initiation, as it is in Candomblé for the filhos de santo. The ogan thus become the best candidates for initiation into the Ifá cult, since they escape any suspicion of simulation that the Cuban babalawo put on Candomblé initiates, accusing them of not embodying gods, but spirits of the dead who can therefore be chased away. The centrality of possession as the basis of ritual practice in Candomblé is therefore deeply questioned by the initiates in the Ifá cult.

Moreover, the integration of Ifá priesthood into Candomblé also entails a real inversion in religious hierarchy, since, according to Cuban tradition, the babalawo is considered ‘superior’ to the olorisha, as Orunmilá (the god of divination) is ‘superior’ to the other Orisha. Thus, a babalawo who has just been initiated will automatically become the ‘elder’ of an olorisha with 30 years of initiation. This openly goes against the hierarchical organization of Candomblé, which is based on a strict principle of seniority.

In recent years, this opposition between Ifá and the Orisha worship has turned the universe of Afro-Atlantic religions upside down, triggering a struggle for religious supremacy in which the Nigerian babalawo play a fundamental role (Capone and Frigerio, 2012). By emphasizing the omniscience of Orunmilá, the Witness of Destiny, the babalawo use their knowledge, supposedly more ‘rational’, to impose their supremacy over the initiates into the Orisha religion, sending the olorisha back to a form of knowledge supposedly inferior to the imo jinlè, ‘the deep truth’ conveyed by Ifá. This vision has triggered acute opposition between initiates into the Orisha religion and babalawo, both in Brazil and in Nigeria. The babalawo defend the superiority of their cult, affirming that all the Orisha acknowledged it by being initiated into the Ifá cult. Yet the olorisha are far from accepting the tutelage of the babalawo and controversy has broken out in 2019 and still continues on social networks, pitting traditional Orisha families and babalawo associations in Nigeria against each other. This controversy has also direct effects on the ‘diaspora’, fuelling debates about which tradition is closer to ‘African roots’.

The multiple differences between the Orisha worship in Brazil, Cuba and Nigeria are always interpreted by the babalawo as being the consequence of a fundamental loss of religious knowledge, which would have produced this gap between matrices of meaning. For them, the differences between ritual practices in this transnational space do not call into question the strength of Yoruba culture, but are the product of ‘holes’ in the African collective memory (Bastide, 1970) that can be filled today by reinjecting cultural content and practices that had fallen into oblivion in Brazil or elsewhere (Capone, 2007). Exchanges between ‘sister religions’, such as Brazilian Candomblé and Cuban Santería, thus aim at re-establishing a common belief system, in which elements of different Afro-Atlantic religions, all claiming the same Yoruba origin, can be combined in different ways.

Productive misunderstandings and transnational imaginaries

The irruption of Ifá into Candomblé provides new forms of legitimization of ritual practice. Ifá is thus becoming the major place of production of meaning and articulation of religious knowledge, giving rise to new transnational imaginaries (Appadurai, 1996). These imaginaries are constantly being questioned, since, as we have seen, they modify the relations between genders in the division of religious work, transforming them into a real ground of negotiation between ‘local’ and ‘global’ cultures.

Nevertheless, the inscription in the local is not only done through the congruence of imported practices or a tuning in with pre-existing cognitive frameworks. In these ‘diasporic conversations’, the discourse displayed has not always to be coherent and in congruence with the beliefs and values of the targeted individuals. In the transnationalization of Ifá cult in Brazil, I stressed the importance of fuzzy and ambiguous discourses, eliciting a multiplicity of interpretations and appropriations. Adaptation strategies can therefore play as much on what is said as on what is not said, as much on negotiation at the ritual level as on misunderstandings, which can be more or less productive (Sahlins, 1985).

Misunderstanding thus becomes a form of interpersonal and intercultural strategy that is likely to prepare, produce, and give consent to the encounter between different belief systems (Capone, 2011). Whenever a consensus has to be established in this ‘dialogue’ between religious actors from different traditions, resort is taken to what Bourdieu (2001: 64) calls a ‘neutralized language’ and Galison (1997) calls a ‘trading zone’. Consensual signifiers will then be mobilized – terms, images, or objects whose simplicity seems to be agreed upon – but behind which each actor will put a different meaning. It is in these ‘zones of awkward engagement’ (Tsing, 2005) that the dynamics of cultural interaction at the global level become intelligible. It is in the tensions – Anne Tsing would call them ‘frictions’ – generated by the encounter between different modalities of worship, all of which are part of the same ‘Yoruba tradition’, that the ‘black Atlantic’ is configured as a community of sensibilities and destinies. Each friction, each clash in the cultural encounter reconfigures this imagined community (Anderson, 1983).

The confrontation between multiple visions of the tradition and the controversies that ensue thus constitute key moments that produce configurations of the Orisha religion at the transnational level which are always unstable. These new configurations are made possible by the work of misunderstanding (Capone, 2011) or by a ‘productive confusion’ which, according to Tsing (2005: 247), is ‘the most creative and effective form’ of cultural collaboration. However, if misunderstanding lays the foundations for establishing dialogue, it also makes it possible to manage the boundaries between belief systems that are at once close and opposed. Indeed, the Orisha religion, despite its unifying discourse, mobilizes universes that are not always compatible, as we saw in the opposition between Ifá and Candomblé. The transnational social space, connecting the African motherland to the diaspora, as well as the various centres of the African tradition on American soil, operates an entanglement of imagined worlds that poses the question of boundaries between modalities of worship in new terms.

This transnational social space is neither egalitarian nor homogeneous, but the result of power structures and internal boundaries – historical, political, and ritual – that cut across the networks of Orisha practitioners. Ritual lineages thus constitute social and symbolic boundaries that must constantly be renegotiated when new religious practices are grafted onto pre-existing ones. Conflict is then often the result of the disruption of the boundaries between these ‘communities’; it is the result of tensions or ‘frictions’ between the different religious practices involved.

This encounter between traditions, claiming the same Yoruba origin, proves that religious globalization cannot be thought of as a homogenizing factor. On the contrary, despite the appeals to a transnational Yoruba imaginary, elevated to the rank of a matrix of meaning by incessant references to the Ifá sacred literature preserved by the babalawo, the Orisha religion today constitutes a conflictual space, built around a structural tension between homogenization and heterogeneity of cultural practices. Analyses of the processes of ‘religious globalization’ must therefore also take into account the tensions, conflicts, adjustments and frictions between the religious systems involved, as well as the new responses they produce in this confrontation between a globalized Yoruba imaginary and localized Afro-American traditions. While some religious modalities, such as Ifá and Brazilian Candomblé, all refer to the same Yoruba origin, this identity is characterized by multiple national religious identities and by their interaction, which is often antagonistic. While there is a ‘sense of belonging’ (Levitt, 2004) to an ‘imagined community’, such as that of the practitioners of the Orisha religion, there is also an awareness of the sometimes-insurmountable differences between the multiple national versions of the ‘Yoruba tradition’.

In the opposition between variants of the Orisha tradition, all considered ‘traditional’ like Ifá and Candomblé, it is around the multiple notions of tradition that ritual negotiation takes place. Religious knowledge is therefore a scarce resource, since the discourse of Candomblé members tirelessly emphasizes the preservation of an ancestral cultural and ritual heritage that is incomplete because of the unfinished transmission of ritual knowledge from initiator to initiate. In the Orisha religion, knowledge is the basis of religious power and, for this very reason, must remain limited.

Moreover, the effort to recollect ancestral knowledge must be inscribed in a territory; it must be ‘localized’. Any ‘deterritorialization’ must then be followed by a new ‘reterritorialization’, whether real or symbolic (Capone, 2004a). However, the old divisions between regional or national traditions are never erased, resurfacing in the processes of memorialization of a tradition that wants to be millennial. What has changed today is that, instead of defining cultural ‘niches’ or territories of resistance (Bastide, 1960), this geography of memory no longer has a single referent, but a complex network that points to the many traditional centres of the Orisha religion: Ilé-Ifè or Oyó in Nigeria, Havana or Matanzas in Cuba, Salvador de Bahia in Brazil or, more recently, Oyotunji Village in the United States. All these places delimit the perimeter of the same symbolic community that does not exclude conflict and segmentation as a form of religious reproduction. This is the main challenge facing the Orisha religion today, torn between the aspirations of becoming a universal religion and the constraints of its inscription in particular national histories.