Navigation – Plan du site
Dossier : Les Amériques noires (I)

The birth of the Yoruba hegemony in post-abolition candomblé

Luis Nicolau Parés
p. 139-159


La naissance de l’hégémonie yoruba dans le candomblé depuis l’abolition de l’esclavage. Les données historiques montrent que des pratiques religieuses jeje et nagô, originaires d’Afrique occidentale, étaient déjà implantées à Salvador (Bahia, Brésil) dans les années 1860, ce qui laisse supposer qu’elles remontent à la période de la Traite. Même si l’ethnogenèse yoruba et le nationalisme culturel et racial de la « Renaissance lagosiène » des années 1890 peuvent avoir indirectement contribué à la « nagoïsation » du candomblé bahianais, l’article révèle que le développement de l’hégémonie religieuse de la « nation » nagô fut surtout le résultat de micro-politiques locales de compétition entre les créoles afro-brésiliens.

Haut de page

Entrées d’index

Géographique/ethnique :

Brésil, Bahia, Jeje, Nagô, Yoruba

Thématique/disciplinaire :

Anthropologie religieuse, Histoire
Haut de page

Notes de la rédaction

Manuscrit reçu en janvier 2005, accepté pour publication en mai 2005

Texte intégral

1« Candomblé » is the name given to the regional development of Afro-Brazilian religion in the state of Bahia. Originated in religious practices brought into Brazil by African slaves, candomblé is a spirit-possession cult involving divination, initiation, sacrifice, healing and celebration. Candomblé congregations worship a series of spiritual entities, often associated with forces of nature, who receive periodic ritual offerings in their shrines and who periodically « possess » selected devotees during drumming-dancing public ceremonies. Regular interaction with the gods is supposed to bring fortune to the religious group and to defend it against misfortune.

  • 1 During the slave-trade period Nagô was an ethnonym used to designate Africans from the Yoruba-speak (...)
  • 2 There are also the caboclos, although cult houses exclusively worshipping these Brazilian spiritual (...)

2Beyond these shared liturgical and conceptual aspects, cult groups often resort to the discourse of « nations » to negotiate, construct and legitimate their ritual differences and collective identities. Today, candomblé cult houses generally claim to belong to one of the three main « nations », Nagô, Jeje and Angola which are characterized by the worship of different kinds of spiritual entities1. The Nagô worship the orixás, the Jeje the voduns, and the Angola the enkices2. Each of these groups of African deities are usually praised in the corresponding ritual language (Yoruba, Gbe and Bantu-derived dialectal forms) and has its own ritual particularities (drum rhythms, dances, food offerings, etc.). Hence, despite the creative eclecticism and movement of values and practices across « nation » boundaries, certain key ritual features are considered important as diacritical signs of a real or imagined continuity with a distinct African past and religious tradition.

  • 3 Ketu is the name of one of the Nagô (Yoruba) « sub-nations » and refers to the precolonial Yoruba k (...)

3In contemporary candomblé the Yoruba or Nagô nation is paramount and the orixá pantheon and its associated liturgy have become distinctive elements of the religious institution. A recent survey conducted in Salvador (Mott and Cerqueira 1998, p. 13) suggests that cult houses self-identifying as belonging to the Nagô-Ketu or Ketu « nation » constitute 56.4% of the total (against 27.2% of the Angola and a mere 3.6% of the Jeje)3. To a great extent this Ketu predominance is due to the social visibility and prestige of three particular cult houses.

  • 4 The date of foundation of the Ilê Iyá Nassô is uncertain. Some authors speculate it could be the en (...)

4In religious circles and in Bahian society at large it is widely assumed that the Ilê Iyá Nassô (also known as Engenho Velho or Casa Branca) – founded by Yoruba-speaking women (some of them allegedly from the city of Ketu) – was the first candomblé in Bahia (and implicitly Brazil). From this original terreiro or religious congregation two others were issued: the Gantois (Ilê Iyá Omin Axé Iyamassé) in mid-nineteenth century and the Ilê Axé Opô Afonjá in 19104.

5These three most famous cult houses claim to belong to the Ketu or Nagô-Ketu nation and they occupy a long-established leading position in the religious field. Because of the agency of its leaders and participants (starting in the late 19th century), together with the influence of intellectuals and the media (since the 1930’s and more noticeably since the 1970’s), these cult houses have been progressively identified as the true guardians of African religious « purity », « tradition » and « authenticity ». Also since the 1970’s they were used as important icons of resistance, pride and dignity in the Black Movement political agenda. By these means the Nagô-Ketu identity emerged as a visible mark of prestige and authority. Because of this and in an effort to legitimise their practices, many religious congregations with no spiritual or ritual affiliation with these « traditional » Ketu houses chose to self-identify as belonging to this « nation » and to follow its ritual « model ».

  • 5 Elsewhere, I have examined the « Nagôization » process since its beginning throughout the 20th cent (...)

6Although the predominance of the Yoruba-Nagô identity in candomblé has lasted throughout the twentieth century, I contend that this may not have been always the case and that during the slavery period the Jeje religious practices played a more central role than has been usually acknowledged. Therefore, the paper focuses on the initial moment of transition when the Nagô referents and its associated collective identity became more prominent in detriment of the Jeje ones, a process which I may call the « Nagôization » of candomblé, and whose first stage can be roughly dated in the late 19th century5.

7The analysis will hopefully demonstrate that contemporary Yoruba hegemony was the result of a complex historical process involving different factors. I argue that at this early post-abolition stage the Yoruba promotion was mainly due to the agency of local religious leaders combined with the critical influence of transatlantic travelers, all part of the tangled web of politics of prestige and legitimacy within the Bahian Black community. I sustain that the trans-Atlantic network of communication contributed to the imagination of a new idea of Africa, a greater awareness of the racialization of social relationships and the increasing visibility within candomblé of the Nagô-Ketu (Yoruba) identity. However, I also contend that these were already ongoing local social dynamics and that the influence of transatlantic travellers in the « recycling » or re-signification of religious knowledge could only be discrete and affecting a limited number of religious congregations.

Antecedents of the « Nagôization » process of candomblé

8Before I examine what I have called the « Nagôization » of candomblé it will be useful to demonstrate that throughout the slavery period, and especially by mid-19th century, religious practices of African origin were not particularly dominated by the Yoruba orixá cults and that the Jeje vodun cults exerted if not a more critical influence, at least an equally important one.

9In Colonial Brazil collective identities among Africans and their descendants were articulated and expressed by means of participation in institutionalised forms of social organization. « African nations », like Angola, Jeje and Nagô, were identities constructed in the context of the slave trade and they defined their social boundaries in a dialogical relationship, valuing particular diacritical signs to establish differences among them. This contrastive and often competitive dynamic found fertile ground in institutions such as work crews (cantos), Catholic brotherhoods and their feasts (folias), secular drumming-dancing gatherings (batuques) and African-derived religious congregations (candomblé).

10It was probably in African-derived religious practices where the Jeje, Angola and Nagô ethnic identities took on a critical function. Continuing with the African tradition whereby collective identity was built around the worship of particular spiritual entities, in Brazil too, despite the disintegration of kinship corporations (or may be because of it), religious activities allowed for the reshaping of new communal identities.

11By the mid-18th century, it is probable that the Angola, Jeje, and other West African groups like the Courana, had already set the basis for the future institutionalisation of candomblé, promoting the organization of religious congregations in domestic and, most importantly, extra-domestic spheres (see for example Mott 1986 and Reis 1988). As I have argued elsewhere, the latter form of religious organization – going beyond the mere healing and oracular activities and involving an ecclesiastical structure with fixed shrines, a priesthood hierarchy and complex processes of initiation – may have been greatly influenced by the Jeje religious experts who had previous experience in the organization of such convents or « mystical schools » (Nicolau Parés s.d.). A logical inference from this hypothesis is that when the Nagô arrived in Brazil in the late-18th century and early-19th century, despite sharing with the Jeje similar forms of religious worship, they may have already found an incipient form of religious institutionalisation involving the cult of multiple deities within the same temple as well as serial forms of ritual performance. It should be noted that this argument stands in contrast to the contemporary oral tradition, which claims that the first candomblé in Salvador was the Ilê Iyá Nassô or Engenho Velho of Nagô-Ketu nation.

12The record indicates a number of religious cult groups in the early-19th century and if my argument is correct we could further speculate that, despite the massive demographic superiority of the Nagô from 1820 onwards, the Jeje religious traditions were critical points of reference in the organization of ritual practice. This seems to be confirmed by the available historical data from the 1860’s, as well as by contemporary linguistic ethnographic evidence. For instance, to this day, in houses which claim to be « pure Nagô », the names to designate the members of the initiation group (dofona, dofonitinha, gamo, gamotinha, etc.), the initiation room (hunco), the shrine’s room (peji), the drums (hun) and other central ritual features, are Jeje terminology (i.e. of Gbe linguistic origin). The fact that these elements form a part of the « deep structure » of the cult points toward critical agency by the Jeje in the founding of the institution of candomblé (Lima 1977, pp. 72-73; Castro 1981, p. 75; Braga 1995, pp. 38-39, 56).

13Copies of the satirical journal O Alabama are available for the period 1863-1871. This journal was published in Salvador by pro-Abolitionist African descendants who launched a systematic campaign against candomblé. Despite its strong ideological bias, O Alabama documented names of participants, locations of terreiros, African terminology and various feasts and religious activities that were sometimes witnessed by the journalists, who provided quasi-ethnographic descriptions. This rich data makes it clear that by 1860 candomblé had already attained a level of institutionalisation with levels of ritual and social complexity very similar to those of the present-day. If this was the case in the 1860’s, there is ground to infer that this institutionalisation, albeit creative, was the result of a process rooted in the transatlantic slave-trade period, since it is difficult to imagine that such a degree of ritual consensus was reached in just one decade.

  • 6 O Alabama, 1863-1871. I have analysed in detail the O Alabama data (Nicolau Parés s.d.). Contrary t (...)

14The O Alabama data, while often referring to Africans involved in candomblé, gives no indication at all that the terreiros were identified with particular « nations ». Yet a quantitative linguistic analysis of its African terminology indicates a similar number of Jeje and Nagô terms, suggesting a Jeje-Nagô religious equilibrium, if not a slight Jeje-dominance. For instance, « vudum » was the common term to designate the African deities. The most famous term today, the Yoruba word orixá, only appears in the composite term « babaloixa », a title to designate the high priest. References to specific deities and other ritual terminology also indicate a relative equilibrium between voduns and orixás6.

15Furthermore, taking into account the ethnic origin of candomblé leaders (when reported) and the contemporary identification of the « nation » of some terreiros mentioned in O Alabama – such as the Gantois (Nagô) or Bogum (Jeje) –, a similar quantitative equilibrium between Nagô and Jeje cult houses is found, if not a slight predominance, once again, of the Jeje over the Nagô. Other examples could be added, but the main point is that, despite the significant demographic and cultural presence of the Nagô in mid-19th century Afro-Bahian society, within the institution of candomblé there is no clear evidence that they exerted a particular dominance or had more social visibility than other « nations », at least until 1870. This case would also demonstrate how cultural influence is not necessarily related to demographics.

16As already suggested, only at the end of the 19th century can the emergence and visibility of the Nagô tradition in candomblé be identified. In his posthumous work, Os Africanos no Brasil, Nina Rodrigues (1977) acknowledged that at the start of his studies, in the early 1890’s, he was not able to differentiate between the Jeje and Nagô mythologies given their « intimate fusion ». Nonetheless, he concluded « Today [c. 1905] the Ewe [Jeje] mythology is dominated by the Yoruba one ». Following Ellis, he explains the assimilation of the Jeje culture by the Nagô in terms of the latter’s linguistic dominance and the « more complex and elevated » nature of the Nagô religious beliefs (Rodrigues 1977, pp. 230-231). While the evolutionist argument is seriously questionable, the perception of a supremacy of the Nagô tradition over the Jeje (and implicitly over all other « nations ») was established for the first time in written form. This perception was reproduced by Edison Carneiro (1991, p. 33) and Arthur Ramos (1979, p. 13) in the 1930’s and 1940’s and persists until today (for more contemporary authors see, for example, Verger 1981 and Santos 1986).

  • 7 O Alabama, 4 January 1868; 29 December 1870; 31 December 1870; 24 November 1871. Among other activi (...)

17It could be argued that Rodrigues privileged the Nagô because he conducted his research in a Nagô congregation – the terreiro Gantois (Ilê Iyá Omin Axé Iyamassé), very successful and socially visible at the time. As reported in O Alabama this terreiro (then known as Moinho) was already quite well-known by the late 1860’s. Decades later, in 1896, still led by her African founder, tia Julia, and her daughter Pulcheria, Gantois continued to garner the attention of the Afro-Bahian community with feasts that attracted enormous crowds7.

18Thus, Gantois’ reputation had already been established well before Rodrigues started his research, which may have been one of the reasons that he approached it in the first place. His focus on this Nagô congregation probably minimized his perception of the internal diversity of candomblé in the same way that, when studying the work crews (cantos), he focused on the African ones ignoring the Creole ones. Yet it seems unlikely that an attentive observer like Rodrigues would have not noticed the use of the term vudum if it was still widely used among the religious community as it appears to have been in the 1860’s. This suggests that a significant change took place between 1870 and 1895 finally propelling Nagô referents – such as the term orixá, for instance – to the foreground.

19This period was indeed a time of profound social and political changes still poorly understood. Since the 1871 law that declared free all children henceforth born of slave mothers, slavery had entered a gradual decline, especially in the cities. In an inegalitarian society, fundamentally structured on the slave-master relationship, a growing class of free and freed people of color, most of them Brazilian-born, was marginalized. Although some blacks tried to assimilate the hegemonic values of the white elite, the big majority was deprived of real « citizenship », leading them to approach alternative socio-cultural spaces like candomblé. The religious institution continued to be perceived by the upper-classes as a threat to public security and social stability, as well as a challenge to « civilised » values. Yet the efforts to control and repress these practices were counterbalanced by the erosion of the elite’s authority caused by slavery’s progressive decline.

20After bitter struggles between the conservative slavocrats and urban abolitionists groups, abolition was finally declared in 1888. The next year the new Brazilian Republic proclaimed « order and progress » as the national motto, and a new constitution was written that separated Church from State. To an elite that understood progress in European terms, Bahia – with a proportion of non-whites of 74.4% (the highest in the country) and its persistence of African « barbaric » customs – « appeared to be marching in the opposite direction » (Butler 1998b, p. 163). The regime change dramatically reduced the province’s influence on national politics and the situation worsened with the collapse of the sugar industry.

21Despite the oligarchy’s project for a « whitened » Bahia, the crisis of the 1890’s compelled the black lower-class to search for new opportunities and, in a demonstration of autonomy deeply disturbing to the elite, many moved to Salvador and the Recôncavo towns (Kraay 1998, pp. 10-11). It was in this unstable social context that candomblé flourished with renewed force, offering religious entrepreneurs a possibility of upward social mobility, while favouring new processes of black identity and alternative sense of community based on a distinct set of values, beliefs and practices.

The African-Creole antagonism and the Yoruba « cultural nationalism »

22The late-19th century candomblé Nagôization process cannot be attributed to a single cause. Rather, it must be understood in the context of a complex interaction of a plurality of factors that can only be tentatively explored here. This paper’s hypothesis is that, once Africans began to disappear in Bahia, the long established antagonism between African and Creole was taken on and perpetuated by their Creole descendants in local struggles for prestige and power within candomblé. In this context, the imagination of an ideal Africa – origin and source of ritual « purity » – and association with an elite of black educated transatlantic travellers was used by a limited number of religious leaders to advance their dispute for legitimacy and visibility.

23The antagonism between « assimilated » Creoles and « foreign » Africans dates from colonial times and is explicit in 18th century compromises of Catholic brotherhoods as well as in other 19th century documents. Space limitations do not allow me for a detailed analysis of the subject. Suffice it to say that, even if Creoles and Africans were not homogeneous social groups and their relationships did not need to be always conflictive, some kind of antagonism persisted into the late-19th century. As reported by Rodrigues (1977, p. 101), the last Africans in Bahia preferred « to live with their country-fellows since they know they are feared for their reputation of sorcerers; the Creole population does not like them ».

  • 8 Local newspapers seem to confirm an increase of new candomblés between 1896 and 1905 (Rodrigues 197 (...)
  • 9 « O seu terreiro era de gente da Costa (africanos) e ficava no bairro de Santo Antônio; que o terre (...)

24This separatism found a particularly fruitful field of expression in the religious domain. Even if during the second half of the century there had been an increasing ethnic and racial mixture among candomblé practitioners, a few terreiros remained less permeable and still identified themselves as exclusively African. Furthermore, in the post-Abolition there was a significant growth of new cult houses founded by Creole women and not male Africans as had been the tendency in previous decades8. The African-Creole shift in leadership, aside from gender politics, seems to have been of some importance in establishing differences between terreiros. For example, an old African woman told Rodrigues (1935, p. 171) she did not dance in the Gantois because it was of « gente da terra » [i.e. Creole and mulatto] while her terreiro was of « gente da costa » (i.e. Africans)9.

25Yet the Gantois, despite becoming increasingly Creole, was founded around 1850 by the African tia Julia (Julia Maria da Conceição Nazaré), when she split from the also African-founded Ilê Iyá Nassô where she was initiated. One can hypothesize that because of its African antecedents, the Creole descendants of terreiros like Gantois could gradually claim an African identity, but only and precisely when African-born religious experts began to disappear. This was the case for at least two reasons. Firstly, African religious practices were traditionally perceived as more efficacious than those of the discriminated Creole, and secondly, the post-Abolition period coincided with a moment in which the image of an idealised Africa was used as diacritical sign for the articulation of a new sense of Black community. In the 1890’s, for instance, at least two Carnival associations of Creole members took the African identity, Pandegos da Africa and Embaixada Africana. Although African identity was latent in colonial and imperial times, particularly after the end of the transatlantic trade, it was in the post-Abolition when it became racialised and perceived by Creoles as constituent of an emerging Black identity. However, in candomblé, with its long-established tradition of inner ethnic-religious competitive differentiation, the « africanization » could not be generic or unifying. Instead it privileged the Nagô « nation » or, more precisely, the Nagô-Ketu, to the detriment of other African traditions.

26In mentioning the Jeje vodun Loko, Rodrigues (1977, p. 231) comments that « some black Nagô tried to correct me about the name Lôco […] saying it was just a Creole corruption of the true name Irôco ». In O Alabama all references to this important tree-deity appear under the Jeje version Lôco, but by the end of the century the Nagô practitioners themselves are imposing their own Yoruba version, dismissing the « traditional » Jeje African name as just a « Creole corruption ». This shows that an ethnocentric Nagô « purification » in opposition to « Creole corruption » was already underway, and demonstrates how the agency of the Nagô practitioners was instrumental in their social-religious promotion.

27A critical question underpinning the previous paragraphs is to what extent the late-19th century African revivalism and the « Nagôization » within candomblé were influenced by the contemporary Yoruba « cultural nationalism » happening in West Africa; and to what point the transatlantic exchange was responsible for an increased awareness of the Nagô identity and subsequent efforts to legitimate its « superiority »?

28As stated by Vivaldo da Costa Lima (1987, p. 52), « at the end of the 19th century, the trip to Africa by free Africans and their children was an important legitimising element of prestige and generator of knowledge and economic power. Even as they traded in a wide variety of merchandise brought from the Coast to Brazil, they also, in today’s language, “recycled” the knowledge of the religious tradition learned from “the elders” in the terreiros of Bahia ».

29Examining this Black Atlantic network of communication, Matory (s.d.a) has noted that « at the British-dominated crossroads of African/African-American interaction, the Yoruba acquired a highly-publicized reputation for superiority to other Africans. This reputation for superiority was useful in the 1880’s and 1890’s, as the bourgeois black Lagosians faced new forms of economic disadvantage and racial discrimination ». Matory (ibid.) further argues that values of black racial and religious purity associated with contemporary candomblé « seem to be rooted in the racial and cultural nationalism of the Lagosian renaissance in the 1890’s ».

  • 10 For further discussion on Yoruba nationalism see also Law (1997), Matory (1999b; s.d.b, pp. 102-110 (...)

30The 1890’s « cultural nationalism » (an expression initially used by Ade Ajayi) was closely linked to the Yoruba ethnogenesis and is generally described as a reaction to British Colonial social exclusion and racism. As Ade Ajayi (1961) and John Peel (2000) have convincingly argued, this nationalism had its roots in the missionary movement and it was essentially Christian (especially Protestant). Indeed, the first intellectual mentors of the Yoruba « cultural » identity were Christian priests, several of them educated in Sierra Leone, but at the turn of the century, in Lagos, this movement was carried out mainly by newspapers journalists in a campaign defending African or black racial purity. As a result, a black bourgeois elite – including many Aku, Nagô and Lucumi returnees from Sierra Leone, Brazil and Cuba, respectively – began to reassert their dignity as a « race-nation », cultivating the Yoruba language, adopting African dress, collecting ancestral knowledge in the form of proverbs, stories and poetry, compiling historical narratives from oral tradition and « even started to find merit in some aspects of traditional religion » (Peel 2000, p. 279)10.

31The sense of commonality first experienced by the Nagô, Lucumi and Aku in their respective diasporic countries greatly contributed to the Yoruba ethnogenesis. But to what extent did the emergence of a Yoruba identity have a reciprocal influence on the Diaspora? Did the disputes between English and French colonial powers in the 1880’s and 1890’s, each supporting new local ethnic identities – « Yoruba » and « Djedje », respectively – have some resonance for the Afro-Bahian community? (Matory 1999a, p. 64). The racialization of social relations in post-Abolition Bahia may have been reinforced by ideas of racial dignity promoted by the Yoruba nationalism. The religious arena with its latent ethnic-ritual division also provided fertile ground for « nationalistic » revivals, and the « Nagôization » process in candomblé may have been indirectly favoured by the Anglophile Yoruba nationalism. Some Yoruba « traditional » cosmological elements, compiled by Christian priests and published in clerical literature, may also have reached Bahia. However, the transformational effect of all this external influence on actual religious practice – subject to an internal logic of religious efficacy – is more doubtful.

The transatlantic exchange and its religious dimension

  • 11 Pierre Verger (1987, p. 633) reports 2.630 Brazilian passports expedited to freed returnees between (...)

32Let’s examine in some detail the transatlantic exchange of goods, ideas, and people in order to evaluate its potential impact on candomblé. The return of liberated slaves from Bahia to Africa had been underway since the late-18th century but it became significant from 1835 onwards, with hundreds of returnees settling along the Mina Coast each year, many of them in Lagos, contributing to the above-mentioned Yoruba renaissance, and others regularly travelling and doing business between both shores. The estimated numbers of returnees for the 19th century vary considerably from 3.000 to 8.000 according to different sources11. More important to our analysis are the travellers who returned from West Africa to Bahia.

  • 12 « Livros de Entrada de Passageiros no Porto de Salvador, 1873-1889 », Books 1-4, Arquivo Público do (...)

33In Lagos, Governor Maloney reported that between 1882 and 1887, 28 ships arrived in Lagos from Brazil, bearing a total of 457 passengers, while 27 ships left Lagos for Brazilian ports, bearing a total of 286 passengers, or an average of 47 per year (Maloney 1889, apud Matory s.d.b, p. 114). Thus for every three people coming from Brazil, about two people were going from Lagos to Brazil. Looking at the Bahian record between 1873 and 1888 I counted lower rates: a total of 42 ships arriving from Africa (Lagos or Costa d’Africa) bearing 580 passengers, approximately 66% Africans and 33% Brazilians, an average of 38 passengers per year12.

34These low figures suggest that the influence of the transatlantic exchange in Bahia may not have been as dramatic as some authors would like it to be. Of course, cultural change is not necessarily relative to demographics and a few significant travellers would have been enough to generate important transformations. Yet there is reason to be cautious.

  • 13 O Alabama, 1863-1871; Rodrigues (1977, p. 105); Verger, Anthony and Lünhing (1995, p. 128), apud Co (...)

35The importation of African goods included ornamental cloth, parrot feathers and skin cream (pano da costa, papagaio da costa, limo da costa) – termed « da Costa » to indicate their African origin – as well as kola nuts (obi, orobo), palm oil (dende), pepper, beans, beads, calabashes, mats, cowries, drums, seeds and plants like akoko or ogbo13. It is worth noting that all these goods are important ritual objects in candomblé, suggesting that practitioners attributed the greatest efficacy to African religious products.

  • 14 Other oral testimonies claim that in Cachoeira African religious objects came through Salvador (Cas (...)

36In fact, local traders of these goods in Bahia were often candomblé practitioners, providing them with an important source of economic power. For example, mãe Aninha, lead priestess of the Axé Opô Afonjá, used to sell African goods at the Mercado Modelo in Salvador (Pierson 1971, p. 320). A more questionable oral testimony (Wimberly 1998, p. 86) reports that in Cachoeira, Tia Julia Guimaraes, a member of the Boa Morte sisterhood and a candomblé practitioner, also traded with African goods sent by her brothers living in West Africa14.

  • 15 Already in 1903 the commerce seems to have been through Dakar.

37As Rodrigues (1977, p. 105) reminds us, until the end of the century most of these products came from Lagos, usually traded by Yoruba and English-speaking businessmen15. It thus follows logically that it was more likely that news would be brought from the Yoruba cultural area than from other regions. Furthermore, the Lagosian (Yoruba) monopoly of imported religious goods surely increased the reputation of Africa-ness attributed to the Nagô and the religious congregations claiming the Nagô (Yoruba) identity. In that sense, the transatlantic commerce was an additional factor in the « Nagôization » process.

  • 16 Matory (s.d.b, pp. 194-195). Following Lindsay (1994, p. 31), Matory (ibid.) also states that Branc (...)

38Beyond the goods-trade, the most decisive factor in the transatlantic exchange was the circulation of people and their ideas. Some businessmen could have close relationships with candomblé, for instance Joaquim Francisco Devodê Branco (1856-1924) – one of the few transatlantic travellers for whom we have some reliable historical record. He was a freed Mahi returnee, resident in Lagos and with business in Porto Novo; he exported obi and orobo to Bahia, importing dry meat, tobacco and cachaça. He made several round trips in the 1880’s and was friend of mãe Aninha and god-father of her successor Senhora in the Axé Opô Afonjá16. He may have brought ritual paraphernalia and current news from the Coast, yet, in his position as a businessman, his influence on candomblé practice could not have been more than tangential.

  • 17 See for instance the case of the African « pai de terreiro » Grato deported to the Coast of Africa (...)

39More significant must have been the influence of transatlantic travellers involved in religious activities or bearers of esoteric knowledge. There is some evidence of orixá priests repatriated or deported to Africa17, but very little of religious experts travelling back to Brazil. The O Alabama mentions what appears to be a rather exceptional case of an African woman named Costancinha, who came all the way from « the Coast » to Bahia looking for religious treatment, given the great reputation of a local healer named Mr. Granada, devotee of « Oxalá-lorum » and « Changó » (O Alabama, 24 december 1863, p. 2).

40Writing in the 1940’s, Lorenzo Turner (1942, p. 66) comments that « those Brazilians and their African-born children who lived in Nigeria for so many years and who are now living in Brazil not only speak Yoruba fluently, but, as leaders of the fetish cults, they used their influence to keep the form of worship as genuinely African as possible » [my emphasis]. He mentions the case of Isadora Maria Hamus who was born in Cachoeira in 1888 and went to Lagos with a relative at the age of six where she spent eight years. Fluent in Portuguese, Yoruba and English, she returned to São Felix and became a leading member of the local candomblé community (idid., p. 64).

  • 18 See Pierson (1971, pp. 278-279), Carneiro (1985, p. 112), Lima (1987, pp. 45-53), Verger (1981, pp. (...)

41The case of Martiniano Eliseu de Bomfim, is probably one of the best-known examples of « transatlantic religious expert » and has been widely commented on in Afro-Brazilian studies. In an interview by Pierson in 1935-1937, Martiniano stated that he had lived in Lagos between 1875 and 1886 and subsequently made two shorter trips as well. In West Africa he became initiated as babalawo and after his return to Bahia became an informant of Rodrigues, thus reinforcing the researcher’s already-Nagô-oriented gaze. Martiniano helped his close friend mãe Aninha to found Axé Opô Afonjá in 1910, and to establish the « ministers of Xangô » in 1937 (see below), when he was already famous and well respected. His charisma, religious zeal and Yoruba « purism » made him an early advocate of candomblé’s « re-africanization » or, more precisely, a critical agent of its « Nagôization »18.

42Martiniano’s most serious rival in Ifá divination was the younger and apparently more aggressive Felisberto Américo Souza. Aside from his religious activities as babalawo, Felisberto was engineer, contractor and businessman exporting tobacco and importing black soap from Nigeria. Due to his regular contacts with Lagos, where his father, an ex-slave returnee, was established, Felisberto anglicised his name into Sowzer. As characterised by Matory, he was part of « an impressive dynasty of Brazilian-Lagosian travellers and priests ». He was grandson of the diviner Rodolfo Martins de Andrade (Bamboxé Obitikó) – one of the alleged founders of the Ilê Iyá Nassô (see below) – and had several priestly grandchildren in Lagos, Bahia and Rio. Therefore, the Sowzer-Bambosé family is another significant example of late transatlantic travellers and « leaders of the fetish cults » responsible for « recycling » Afro-Bahian esoteric religious knowledge (Carneiro 1985, p. 112; Verger 1981, p. 32; Matory s.d.b, pp. 81, 291-292).

43This evidence indicates that by the turn of the century there was in candomblé a wide consumption of African products and circulation of a few Yoruba-Anglophone religious experts who had direct contact with « the Coast ». Since most of the imported goods and « transatlantic religious experts » came from Yorubaland, this exchange certainly helped to legitimise the Nagô cult houses, in particular the very few which had communication with these travellers. Consequently there is evidence to suggest that, to a certain extent, the transatlantic exchange contributed to the « Nagôization » process.

The imagined Africa and the new narratives

44What remains to be discussed is the nature of the religious change that arose as a result of all this transatlantic movement. The factual or imagined voyages of religious experts to Africa – acquiring « authentic » esoteric knowledge and thus recovering a « tradition » which had been lost during the traumatic experience of slavery – constituted symbolic cultural capital which greatly increased their social prestige, religious efficacy and authority.

  • 19 See also Carneiro (1985, p. 48). For a recent interpretation of the foundation of the Ilê Iyá Nassô (...)

45In this context, the journey back to « the origins » became a central narrative element in the myth of foundation of the « oldest candomblé of Brazil », the Ilê Iyá Nassô or Engenho Velho. Verger (1981, pp. 28-29) collected oral testimonies stating that Iyá Nassô – one of the founders of the terreiro – together with her spiritual daughter Marcelina da Silva (Obá Tossi) and the latter’s biological daughter Magdalena, travelled to Africa and spent seven years in Ketu. They then returned to Bahia accompanied by the African babalawo Rodolfo Martins the Andrade (Bamboxé) and Magdalena’s three newborn children. According to Bastide’s version (1986, p. 323), Iyá Nassô arrived in Bahia as a free person and founded the candomblé in the 1830’s; Obá Tossi also arrived as a free person but returned to Ketu for seven years before assuming the terreiro’s leadership after Iyá Nassô’s death19.

46Despite the possible and yet undocumented factual reality behind these contradictory and very difficult-to-date narratives, the mythical character of these journeys is apparent. They legitimate the myth of foundation of the Engenho Velho by reinforcing the idea of direct contact with « pure » African sources while stressing the free status of its actors (Capone 1999, pp. 248-250).

47It is also worth noting that the city of Ketu, after being destroyed by the Dahomeans in 1883 and 1886, was only reconstructed in 1896, and that news of these events may have reached Bahia at that time. Since the Ketu ethnonym is not documented in the context of Brazilian slavery, and in candomblé the Nagô-Ketu « nation » claimed by Engenho Velho and Gantois is only named in the 1930’s, one can hypothesise – in the need of further research to test it – that the Ketu pre-eminence in the foundation narrative, despite any other possible historical antecedent, was a late-19th century elaboration inspired by Ketu’s reconstruction. If so, it would be a good example of how the strategies of « invention of tradition » were intermingled with the transatlantic dynamic in the « Nagôization » process.

48Atlantic commerce also favoured the circulation of English publications. By 1903, Yoruba-English dictionaries and other didactic literature could be found in Bahia. Rodrigues mentions for example the Church Missionary Society text Iwe Kika Ekerin Li Ede Yoruba – The fourth Primer in Yoruba Language, brought to him and translated by Martiniano (Rodrigues 1977, p. 133; Peel 2000, p. 387; Matory s.d.b, p. 117). Rodrigues also had copies of the works of Colonel Ellis, The Ewe Speaking People and The Yoruba Speaking People. At the time of his research, Rodrigues probably had little direct influence on the religious community, but these texts also reached a reduced circle of Yoruba-Anglophile literate religious experts like Martiniano, and through them, by means of oral transmission, some new religious ideas could be further propagated.

49This phenomenon is most noticeable in relation to certain deities and mythology. Matory (s.d.b, p. 114), following Rodrigues (1977, p. 217), suggests that the popularity of Olorum (Olodunmare) as the high God in Bahia resulted from previous Christian promotion of this deity in West Africa. Rodrigues (1977, p. 224) remarks that in Bahia, local people of Yoruba origin who were under the instruction of English Protestant missionaries in Lagos (probably referring to Martiniano) criticized certain versions of myths which attributed a past human life to the thunder god Xangô. The euphemistic interpretation – attributing a human origin to deities – was one of the arguments used by Protestant evangelists in Lagos to convert and gain new adepts (Peel 2000, p. 296). The story of Xangô as the 4th king of Oyo, his suicide and later deification, was subsequently reproduced in clerical literature such as the above-mentioned Iwe Kika Ekerin Li Ede Yoruba or the work of Ellis, and despite initial criticism, in time it gained popularity in Bahia. This was also the case with a myth of Yemanjá in which the water deity is presented as the mother of all the Yoruba orixás. This version, first collected and published by Father Baudin in 1884, was apparently a late creation of missionaries (probably looking for a unified and hierarchical Yoruba pantheon) and was introduced in Brazil in the post-slave-trade period, yet it also captured the imagination and became very popular both in West Africa and Brazil (Rodrigues 1977, pp. 222-223).

50Therefore, as far as we can infer from these examples, most of the late-19th century Bahian innovation in Nagô mythology was filtered through the agency of Christian Missions in Yorubaland. These new and yet reduced number of narratives and myths – disseminated through clerical literature, translated and orally transmitted by a local religious elite – may have found fertile ground in Salvador, however they seem rather anecdotal or marginal, if compared with the rich body of traditions inherited and re-elaborated during the slave-trade period. Furthermore, these few contributions resulting from the late-19th century transatlantic exchange belong to the domain of ideas and values, while there is little evidence that they affected the domain of ritual practice.

51Only at the level of institutional organization do we have an example of a significant transformation mediated by transatlantic travellers. As several authors have shown (Lima 1966; Braga 1995, pp. 47-48; Capone 1999, pp. 260-267), the institution of the twelve ministers or obás of Xangô (male dignitaries who support the leadership of the high priestess) introduced by mãe Aninha and her close friend, the babalawo Martiniano Eliseo do Bomfim in 1937, was inspired by the political organization of the Oyo kingdom and the Yoruba logic of left and right division. In fact, the Brazilian institution of the obás was a creative adaptation of the organizational model of the Xangô priesthood in Oyo. Father Baudin (1885, pp. 73-74), whom Martiniano may have read, described it as consisting of a chief, the magba, attended by twelve assistants: « the first calls himself Oton (the right arm); the second, Osin (the left arm); the third, Eketu; the fourth, Ekerin, etc. The chief and his assistants live at Oyo ». Yet as constituted in the Brazilian religious cult the institution of the obás was a rather original arrangement which did not find any counterpart in Yorubaland.

52Conceived as legitimating an imagined African orthodoxy, the institution of the ministers of Xangô could be interpreted as a self-conscious attempt to invest a « disturbed past » – as Sidney Mintz (1989, p. 14, apud Palmié 1993, p. 93) qualifies the past of any Afro-American culture – with continuity and moral significance, and in that sense it provides another good example of the « invention of tradition ». Ultimately, the initiative served wide political goals of black self-determination and empowerment (similar to those promoted some decades earlier by the Lagosian « cultural nationalism »), but also served as a marker of difference and of status compared to concurrent religious congregations like the Engenho Velho. The « ideology of prestige » founded on the conceptual triad « Africa-purity-tradition » had been promoted within candomblé since its beginnings and was part and parcel of the institution. Direct contacts with West Africa provided strategic elements and additional resources in an otherwise local dynamic for legitimacy and authority.

53The effect of individual charisma on religious change has not been sufficiently stressed in Afro-Brazilian studies. Within the 19th century relatively small Afro-Bahian religious community, the idiosyncrasies of character of leaders like Aninha or Martiniano together with their ability to mobilize a wide social network greatly contributed to the prestige of their cult houses, and consequently, of their ritual practices. However, it is worth noting that a structural innovation like the ministers of Xangô remained limited to the Axé Opô Afonjá and was not replicated by other terreiros. This example suggests that innovation may also encounter resistance when faced with institutionalised forms of religious practice and that its diffusion depends on a complex set of variables.


54In summary, Matory’s conclusion (s.d.a) that « much that appears to “survive” of African religion in the Americas is in fact shaped by an African cultural politics that long post-dated the slave trade » seems to be exaggerated. Ethnographic and historical data show that the legacy of some African religious practices has persisted from the slave-trade period until the present-day, partially supporting the Herskovitsian « memory-retention » model. Concurrently, socio-cultural conditions like slavery and modernity produced changes in ritual elements, functionalities and meanings, hence supporting the Creolization theories and their associated processes of « invention of tradition ». Moreover, as we have seen, the transatlantic exchange was an additional variable, although it may have affected not so much religious practice per se, but the discourses, values, ideas and collective identities which turn around it.

55In this context, the late nineteenth century seems to establish the conceptual basis for a notion of Africa as the original locus of a « tradition » that, from a distance, had to be recovered, reinventing continuities in order to overcome a « disrupted past ». This African idealization also constituted an alternative to, and reaction against the growing assimilationist Creole culture. Synchronized with the heightening of Yoruba cultural supremacy in the black Atlantic world, this process of « re-africanization » emerged effectively as one of « Nagôization » which favoured the visibility and prestige of a small number of cult houses. Within some sectors of the religious community this « Nagôization » may also have been perceived as a strategy to achieve black political empowerment.

56Yet, it must be noted that the main power struggle in which the Nagô identity became a diacritical sign was the local confrontation between the African-founded congregations and a growing number of « Creole » congregations dedicated to the Caboclo worship. The alliance and « national » solidarity articulated among the self-claimed Nagô-Ketu houses was favoured by their common interest to oppose the increasing visibility of concurrent cults perceived as « corrupted » and led by charismatic religious entrepreneurs. In this process the Jeje and Angola cult houses and their respective « nation » identities were also marginalized.

57Hence the bases of the Nagô-Yoruba hegemony within candomblé were established in the post-abolition cultural cross-roads. The dispute for legitimacy and prestige continued and acquired renewed strength in the 1930’s, leading intellectuals such as Edison Carneiro (1985, pp. 96-98) and Ruth Landes (1940) to reify the ongoing antagonism into a conceptual polarity between the « pure », female-dominated, Nagô cults versus the « syncretic », male dominated, Caboclo cults (often associated with the Angola traditions). But this is already another chapter of the story. Suffice to say that since then the worship of voduns and enkices persisted in muted and invisible spaces, while candomblé became the religion of the orixás par excellence.

58Aknowledgments: A preliminary version of this paper was first presented at the Sephis International Workshop « The trans-Atlantic construction of the notions of “race”, black culture, blackness and antiracism: towards a new dialogue between researchers in Africa, Latin America and the Caribbean », held in Gorée (Senegal), the 11-17 November 2002.

Haut de page


Ajayi J. F. Ade
1961 « Nineteenth-Century origins of Nigerian nationalism », Journal of the Historical Society of Nigeria, 2, pp. 196-211.

Bastide Roger
1986 Sociología de la religión, Ediciones Jucar, Madrid [1960].

Baudin Paul
1885 Fetichism and fetish worshippers, Benziger Bros, New York.

Braga Júlio
1995 Na Gamela do Feitiço. Repressão e Resistência nos candomblés da Bahia, EDUFBa, Salvador.

Butler Kim
1998a Freedoms given, freedoms won. Afro-Brazilians in post-abolition São Paulo and Salvador, Rutgers University Press, New Brunswick.

1998b « Afterword: Ginga Baiana – The politics of race, class, culture, and power in Salvador, Bahia », in Hendrik Kraay (ed.), Afro-Brazilian culture and politics, M. E. Sharpe, Armonk and London, pp. 158-176.

Capo Hounkpati B. C.
1991 Comparative Phonology of Gbe, Foris Publications, Berlin and New York.

Capone Stefania
1999 La quête de l’Afrique dans le candomblé. Pouvoir et tradition au Brésil, Karthala, Paris.

Carneiro Edison
1985 Candomblés da Bahia, Ediouro, Salvador [1948].

1991 Religiões Negras e Negros Bantos, Editora Civilização Brasileira S.A., Rio de Janeiro [1936-1937].

Carneiro da Cunha Manuela
1985 Negros estrangeiros: os escravos libertos e sua volta à África, Brasiliense, São Paulo.

Castro José Guilherme da Cunha (ed.)
1996 Miguel Santana, EDUFBa, Salvador.

Castro Yeda Pessoa de
1981 « Língua e nação de candomblé », África, 4, pp. 57-77, São Paulo.

Cohen Peter
2002 « Orisha journeys: the role of travel in the birth of Yorubá-Atlantic religions », Archives des Sciences Sociales des Religions, 117 (janvier-mars), pp. 17-36.

Harding Rachel
2000 A Refuge in Thunder. Candomblé and alternative spaces of Blackness, Indiana University Press, Bloomington and Indianapolis.

Kraay Hendrik
1998 « Introduction: Afro-Bahia, 1790s-1990s », in Hendrik Kraay (ed.), Afro-Brazilian culture and politics, M. E. Sharpe, Armonk and London, pp. 3-29.

Landes Ruth
1940 « A cult matriarchate and male homosexuality », Journal of Abnormal and Social Psychology, 35 (3), pp. 386-397.

Law Robin
1997 « The Atlantic slave trade and the construction of African ethnicity: the case of the Yoruba » [unpublished manuscript].

Lindsay Lisa A.
1994 « To return to the bosom of their fatherland: Brazilian immigrants in nineteenth-century Lagos », Slavery and Abolition, 15 (1), pp. 22-55.

Lima Vivaldo da Costa
1966 « Os obás de Xangô », Afro-Ásia, 2-3, pp. 5-36.

1977 A família-de-santo nos candomblés Jeje-Nagôs da Bahia: um estudo de relações intra-grupais, EDUFBa, Salvador.

1987 « O candomblé da Bahia na década de trinta », in Vivaldo da Costa Lima and Waldir Freitas Oliveira (eds), Cartas de Edison Carneiro a Arthur Ramos, Corrupio, São Paulo.

Maloney A. C.
1889 « Cotton interests, foreign and native, in Yoruba and generally in West Africa », Manchester Geographical Society Journal, 5, pp. 265-276.

Matory J. Lorand
s.d.a « Afro-Atlantic culture: on the live dialogue between Africa and the Americas »,

s.d.b Black Atlantic religion. Tradition, transnationalism and matriarchy in the Brazilian candomblé, Princeton University Press, Princeton [The pagination corresponds to the draft version 24.2.1997, entitled « Man in the “city of women”: transnationalism, matriarchy and the rise of the Afro-Brazilian candomblé »].

1999a « Jeje: Repensando Nações e Transnacionalismo », Mana, 5, pp. 57-80.

1999b « The English professors of Brazil: on the diasporic roots of the Yorùbá nation », Society for Comparative Study of Society and History, 41 (1), pp. 72-103.

Mintz Sidney W.
1989 Caribbean transformations, Columbia University Press, New York [1974].

Mott Luiz
1986 « Acotundá. Raízes setecentistas do sincretismo religioso afro-brasileiro », Revista do Museu Paulista, nova série, 31, pp. 124-147.

Mott Luiz and Marcelo Cerqueira (eds)
1998 As religiões Afro-Brasileiras na luta contra a Aids, Editora Centro Baiano Anti-Aids, Salvador.

Nicolau Parés Luis
2005 « The Nagôization process in Bahian candomblé », in Toyin Falola and Matt Childs (eds), The Yoruba Diaspora in the Atlantic World, Indiana University Press, Bloomington and Indianapolis.

s.d.  Do Lado do Jeje: História e Ritual do Vodun na Bahia.

Palmié Stephen
1993 « Against Syncretism. “Africanizing” and “Cubanizing” discourses in North American òrìsà worship », Counterworks, pp. 73-103.

Peel John David Yeadon
1990 « The Pastor and the babalawo: the interaction of religions in nineteenth-century Yorubaland », Africa, 60 (3), pp. 338-369.

2000 Religious Encounters and the making of the Yoruba, Indiana University Press, Bloomington and Indianapolis.

Pierson Donald
1971 Brancos e Prêtos na Bahia: estudo de contacto racial, Ed. Nacional, São Paulo [1942].

Ramos Arthur
1979 « Introdução », in Nunes Pereira, A Casa das Minas. Culto dos voduns Jeje no Maranhão, Vozes, Petropolis [1947].

Reis João José
1988 « Magia Jeje na Bahia: A Invasão do Calundu do Pasto de Cachoeira, 1785 », Revista Brasileira de História, 16 (8), pp. 57-81, 233-249.

2001 « Candomblé in Nineteenth-Century Bahia: priests, followers, clients », in Kristin Mann and Edna Bay (eds), Rethinking the African Diaspora: The making of a Black Atlantic world in the bright of Benin and Brazil, Frank Cass and Co Ltd, London.

Rodrigues Nina
1935 O animismo fetichista dos negros baianos, Civilização Brasileira, Rio de Janeiro [1896].

1977 Os Africanos no Brasil, Companhia Editora Nacional, São Paulo [1906].

Santos Juana Elbeim dos
1986 Os Nàgo e a Morte: Pàde, Àsésé e o Culto Égun na Bahia, Vozes, Petropolis.

Silveira Renato da
2000 « Jeje-Nagô, Iorubá-Tapá, Aon Efan, Ijexá. Processo de constituição do candomblé da Barroquinha, 1764-1851 », Cultura Vozes, 6 (94), pp. 80-100.

Turner Jerry Michael
1975 Les Brésiliens: The impact of former Brazilian slaves upon Dahomey, Ph.D. thesis, Boston.

Turner Lorenzo D.
1942 « Some contacts of Brazilian ex-slaves with Nigeria, West Africa », Journal of Negro History, 27 (1), pp. 55-67.

Verger Pierre
1981 Orixás, Corrupio, Salvador.

1987 Fluxo e Refluxo do tráfico de escravos entre o Golfo do Benin e a Bahia de todos os Santo, Corrupio, São Paulo.

Verger Pierre, Ming Anthony and Angela Lünhing
1995 « À la recherche des plantes perdues: les plantes retrouvées par les descendants culturels des Yorubas au Brésil », Revue d’Ethnolinguistique, Cahiers du Lacito, 7, pp. 113-140.

Wimberly Fayette
1998 « The expansion of Afro-Bahian religious practices in Nineteenth-Century Cachoeira », in Hendrik Kraay (ed.), Afro-Brazilian culture and politics, M. E. Sharpe, Armonk and London, pp. 74-89.

Haut de page


1 During the slave-trade period Nagô was an ethnonym used to designate Africans from the Yoruba-speaking area, in contemporary southwest Nigeria. Jeje was an ethnonym to label enslaved peoples from the Gbe-speaking area (Capo 1991), occupying the southern regions of modern Togo and République du Benin and usually referred in the literature as Adja-Ewe. Angola was a generic ethnonym to refer to a variety of Bantu-speaking groups, shipped to Brazil in the western ports of Central Africa.

2 There are also the caboclos, although cult houses exclusively worshipping these Brazilian spiritual entities seldom claim to constitute a « nation ». Each of the three main « nations » is divided in subcategories related to specific African « lands » or « provinces ». Nagô comprises Nagô-Ketu, Nagô-Ijexá and Efon. The Jeje differentiate between Jeje-Mahi, Jeje-Savalu and Jeje-Mudubi. The Angola nation also includes the Congo. As discussed below, many cult houses self-identify as belonging to a combination of « nations » (i.e. Ketu-Angola, Jeje-Angola-Caboclo, etc.).

3 Ketu is the name of one of the Nagô (Yoruba) « sub-nations » and refers to the precolonial Yoruba kingdom of Ketu, located in modern République du Benin.

4 The date of foundation of the Ilê Iyá Nassô is uncertain. Some authors speculate it could be the end of the 18th century while more conservative hypothesis suggest the early decades of the 19th century. Regardless, oral traditions and Afro-Brazilian studies have regularly insisted on this « myth of origin » of candomblé which attributes to the Ilê Iyá Nassô, the privileged role of « the oldest terreiro in Brazil ». Variants of this myth of foundation are discussed below.

5 Elsewhere, I have examined the « Nagôization » process since its beginning throughout the 20th century (Nicolau Parés 2005). In this paper I focus in a much more detailed way on the first stage corresponding to the late 19th century.

6 O Alabama, 1863-1871. I have analysed in detail the O Alabama data (Nicolau Parés s.d.). Contrary to my results, João José Reis’ analysis (2001) of the same data suggests a slight Nagô predominance, may be because he used other nineteenth-century sources besides the O Alabama. In any case the difference is little and the relative equal proportion is of more concern.

7 O Alabama, 4 January 1868; 29 December 1870; 31 December 1870; 24 November 1871. Among other activities the Moinho congregation organized a « devotion » for Nossa Senhora da Conceição together with a feast for a « Mãe d’Agoa » (probably Oxum) in December, and the feast of the new yam in November. Diario de Notícias, 5 October 1896 (Rodrigues 1977, p. 239).

8 Local newspapers seem to confirm an increase of new candomblés between 1896 and 1905 (Rodrigues 1977, pp. 240-245). As already suggested, the perception of religious activity as a means of social upward mobility for blacks in a period as tumultuous as post-Abolition, may explain in part this growth. For an analysis of the ethnic-racial and gender composition of candomblé in the 19th century see Harding (2000, chap. V), Reis (2001), Nicolau Parés (s.d., chap. IV).

9 « O seu terreiro era de gente da Costa (africanos) e ficava no bairro de Santo Antônio; que o terreiro do Gantois era terreiro de gente da terra (creoulos e mulatas) ».

10 For further discussion on Yoruba nationalism see also Law (1997), Matory (1999b; s.d.b, pp. 102-110).

11 Pierre Verger (1987, p. 633) reports 2.630 Brazilian passports expedited to freed returnees between 1820 and 1868, while Jerry Turner (1975, pp. 67, 78) counted 1.056 passports expedited between 1840 and 1880, and estimates a rough total of 3.000 ex-slaves returnees. Manuela Carneiro da Cunha (1985, pp. 210-216) evaluates about 8.000 the number of libertos who arrived from Bahia between 1820 and 1889.

12 « Livros de Entrada de Passageiros no Porto de Salvador, 1873-1889 », Books 1-4, Arquivo Público do Estado da Bahia.

13 O Alabama, 1863-1871; Rodrigues (1977, p. 105); Verger, Anthony and Lünhing (1995, p. 128), apud Cohen (2002, p. 27).

14 Other oral testimonies claim that in Cachoeira African religious objects came through Salvador (Castro 1996, p. 27).

15 Already in 1903 the commerce seems to have been through Dakar.

16 Matory (s.d.b, pp. 194-195). Following Lindsay (1994, p. 31), Matory (ibid.) also states that Branco was born in 1856 and taken as a slave to Bahia in 1864 [after the end of the slave-trade?] where he was freed by his master. According to my own research, Joaquim Francisco Branco’s arrivals in Bahia from Lagos are dated 19th November 1884, 30th December 1885, 31st July 1886, and 5th August 1887: « Livro de Entrada de Passageiros no Porto de Salvador, 1883-1888 », Book 4, Arquivo Público do Estado da Bahia.

17 See for instance the case of the African « pai de terreiro » Grato deported to the Coast of Africa in 1859: Jornal da Bahia, 17th July 1859 (Verger 1987, p. 532). The sorcerer Gonçalo Paraiso was also deported in 1860 (Butler 1998a, p. 192). Also in the 1850’s the case of José Felipe Meffre, son of a babalawo and practicing babalawo himself returned from Brazil to Lagos with all his ritual objects. After his conversion to Protestantism, he and a fellow converted babalawo used their clout to win more converts, including a « great idolatress », also returned from Brazil (Peel 1990, pp. 338-369; 2000, p. 352). In 1851, the leading Egba chief Sagbua, commissioned charms to « tie war », that is, to restrain their Dahomean enemies, from a Muslim-Brazilian returnee (Peel 2000, p. 199).

18 See Pierson (1971, pp. 278-279), Carneiro (1985, p. 112), Lima (1987, pp. 45-53), Verger (1981, pp. 31-32), Braga (1995, pp. 37-58) and Capone (1999, pp. 250-252). In Bahia, I identified the arrival from Lagos of an Eleseo do Bomfim (freed African) in 26 September 1878, the arrival from Rio de Janeiro of an Elizio do Bomfim (freed, single, worker, 56 years old) in 27 October 1880, and an Elizeo do Bomfim (businessman, 40 years old) in 18 November 1880: « Livros de Entrada de Passageiros no Porto de Salvador, 1873-1883 », Books 1, 3, Arquivo Público do Estado da Bahia.

19 See also Carneiro (1985, p. 48). For a recent interpretation of the foundation of the Ilê Iyá Nassô see Silveira (2000). Capone (1999, p. 250) also attributes a character of myth of foundation to the alleged journey to Africa of Marcos Teodoro Pimentel, founder of the first egun cult in the Itaparica island (Bahia). Jeje oral tradition also claims that Ludovina Pessoa, the « first » priestess founder of the Jeje « nation », travelled each year to Africa (Nicolau Parés s.d.).

Haut de page

Pour citer cet article

Référence papier

Journal de la Société des Américanistes, 2005, 91-1, pp. 139-159

Référence électronique

Luis Nicolau Parés, « The birth of the Yoruba hegemony in post-abolition candomblé », Journal de la société des américanistes [En ligne], 91-1 | 2005, mis en ligne le 10 janvier 2010, consulté le 24 mars 2017. URL : ; DOI : 10.4000/jsa.2873

Haut de page


Luis Nicolau Parés

Universidade Federal da Bahia, Brésil

Haut de page

Droits d’auteur

© Société des Américanistes

Haut de page
  • Logo Latindex
  • Logo Centre National du Livre
  • Logo CNRS - Institut des sciences humaines et sociales