Agidai

I love rereading my old notes and my own research. Not because I think I got everything right the first time, but because I know how much slips out of memory over the years. I have shelves and folders full of notes, photocopies, scans, articles, and half forgotten observations. Every now and then I pull a random stack back into the light and read it again slowly. That is often when the real discoveries happen. Not new information exactly, but connections that were always there and only become visible with time.

One of those moments happened when I returned to a book that deserves far more careful attention than it usually gets. In 1992, Editorial de Ciencias Sociales in Havana published Guillermo Andreu Alonso’s Los ararás en Cuba: Florentina, la princesa dahomeyana. An English translation titled The Arará in Cuba: Florentina, a Princess from Dahomey appeared a few years later through Editorial José Martí. The book is frequently cited but rarely read properly, and even more rarely understood on its own terms.

This is not a conventional academic monograph. It does not follow the rigid structures or theoretical fashions of modern anthropology. What it does instead is record. It documents names, lineages, ritual groupings, songs, drumming patterns, spirit classifications, and fragments of memory gathered from elders who were already disappearing when Alonso conducted his fieldwork. The scholarship lies not in abstraction, but in fidelity. He records what people said and what people did, even when those things resist systematisation.

That is precisely what makes it such a serious ethnographic source. The book does not try to harmonise contradictions or iron out inconsistencies. It does not force Arará religion into a clean framework borrowed from Santería or Vodou. It allows fragments to remain fragments. In doing so, it preserves something that would otherwise have been lost entirely.

The Arará are the descendants of Fon and Ewe speaking peoples from Dahomey, present day Benin, who were enslaved and brought to Cuba primarily during the late eighteenth and early nineteenth centuries. In Cuba, they preserved a religious tradition centred on vodun rather than orisha. In Cuban usage, Arará refers both to the people and to their religious system, commonly called Regla Arará.

Historically, Arará ritual centres developed mainly in Matanzas Province, in towns such as Matanzas, Perico, Jovellanos, and Agramonte. There they survived alongside Lukumí or Santería and Palo, but remained distinct from both. This point matters more than many people realise. Arará religion is not a variant of Santería, and it is not a marginal offshoot of Yoruba practice. It is a parallel West African tradition rooted in Dahomean cosmology, with its own ritual logic, spirits, songs, and internal classifications.

Arará and Haitian Vodou share common African origins, not a shared New World history. Both descend largely from Fon and Ewe religious systems. That is why we see vodun rather than orisha, similar categories of spirits linked to earth, illness, crossroads, water, fire, death, similar naming patterns, and comparable ideas about possession and ritual containment. But these traditions developed independently in the Americas.

Haitian Vodou evolved under French colonial rule, Catholic framing, and the Kreyòl language. Cuban Arará evolved under Spanish rule, different Catholic pressures, and Cuban Spanish. The result is that the pantheons overlap structurally rather than nominally. Spirits who do similar work may carry different names. Some vodun are preserved in Cuba that are marginal or transformed in Haiti, and others survive more clearly in Haiti than in Cuba. Arará vodun are cousins of the lwa, not alternate names for them. Treating them as interchangeable is one of the most persistent and damaging misunderstandings in Afro Atlantic religious discourse.

If we look for the closest parallel to Cuban Arará, it is not Haitian Vodou but Candomblé Jeje in Brazil. Both are Dahomean derived, vodun centred, non Yoruba in structure, and preserved as minority traditions alongside dominant Yoruba forms. The difference lies in historical context. In Brazil, Jeje traditions were gradually absorbed into a Yoruba dominant Candomblé framework, while retaining vodun names and rites in some houses. In Cuba, Arará remained more visibly separate from Lukumí, even while exchanging influences.

Because of this, Arará vodun often resemble Jeje voduns more closely than Haitian loa. Naming patterns in Arará sometimes align more clearly with Jeje than with Vodou. Both traditions preserve less personified and more functional vodun that resist neat syncretism. Anyone familiar with Jeje will recognise Arará quickly, not as something exotic, but as something structurally familiar.

Over the twentieth century, Regla Arará gradually moved closer to Santería in practice, language, and ritual framing. This was not because Arará was derived from Santería, but because both traditions existed in the same neighbourhoods, often within the same families. Santería became the dominant Afro Cuban religion with more houses, more priests, greater public visibility, and a clearer initiation structure. Arará communities were smaller and more vulnerable. Lukumí terminology increasingly became the shared religious language.

Inter initiation became normal. Many Arará practitioners were also crowned in Santería. Lukumí concepts were used to explain Arará vodun. Arará spirits were mapped onto orisha categories for clarity. Santería feast days, colours, and saints became reference points. This did not erase Arará, but it reframed it.

Knowledge transmission also shifted. As fluent speakers of Fon and Ewe languages disappeared, Spanish and Lukumí vocabulary filled the gap. Names survived, but explanations increasingly relied on Santería logic. This created the impression that Arará spirits were derivatives rather than expressions of a distinct cosmology.

Another crucial factor is secrecy. Arará communities were deliberately closed even by Afro Cuban standards. Their religious life was never meant to be public facing. Knowledge was guarded carefully, passed through family lines and trusted relationships, and often withheld entirely from outsiders. Silence was a strategy for survival rather than a lack of substance.

This secrecy meant that Arará ceremonies were rarely documented or described in detail. While Santería developed a relatively public grammar through cabildos, printed books, and later academic attention, Arará practice remained inward looking and protective. When Arará did speak outwardly, it often did so through Santería language because that was safer.

The result was fragmentation. Without pressure to standardise or publish, different houses preserved different elements. Names, functions, and groupings survived unevenly. Later sources appear contradictory not because the tradition is incoherent, but because it was preserved in pieces.

This is why Alonso’s book matters so much. It does not try to make Arará legible or fashionable. It records what could still be reached before silence closed completely. It respects secrecy without turning it into mystique. It allows contradictions to stand. It preserves fragments without forcing coherence. It is a salvage record, and a careful one.

It also helps illuminate obscure developments within Santería itself. One of the most interesting examples is the orisha Agidai.

Agidai appears in the United States in the 1980s and quickly gained popularity. His origin in Cuba is somewhat obscure. Willie Ramos, in his On the Orishas' Roads and Pathways: Obatalá, Odúa, Oduduwá traces Agidai back to an individua by the name of Carlota Armenteros from the city of San Jose de las Lijas, near Havana, Cuba. In wider Santería discourse, Agidai is presented as an Orisha of afudashe, the ability to speak prophetically and with power. He becomes the "patron Orisha" of the Oriate, the master of ceremony and cowrie shell diviner in Santería/Lukumi. Agidai is described as an Orisha fun fun, which leads many to associate him with Obatala, or to note that he must be a camino or road of Obatala. His eleke is made from white beads alternated with the traditional patterns for Oshun, Yemanja, and Shango. He lives in a small white tureen, and is sometimes referred to as “the eyes of Orunmila”, emphasising his tole in prophecy and divination.

By the 1990s, receiving Agidai becomes fashionable for Oriates in the Santeria religion, and among Santería priests concerned with cowrie shell divination. Yet tracing Agidai back to Africa has proven difficult. This difficulty was usually explained away with claims that Africa forgot what Cuba remembered. But in 2026, that argument no longer holds. Orisha tradition in Africa never had the office of the Oriate. Knowledge is distributed among elders rather than centralised in a ritual specialist. While afudashe exists as a concept, divine inspiration in divination is attributed to Ela, not to a distinct Orisha governing prophetic speech through shells. There is no Orisha by the name of Agidai in Nigeria fulfilling this role.

So where did Agidai come from?

In Alonso’s book, we find a spirit listed under several names, including Aggidai. Not as an Orisha, or as a patron of divination or prophecy, but as a vodun of the Arará. Aggidai is noted as a manifestation of Dasoyi, who in Cuba became known as the "father" of Babalu Aye. He is described as a messenger and one of the most important manifestations of this earth and illness related spirit.

Alonso, on page 20 of the English translation of the book notes: “Dasoyi reveals himself in seventeen avatars. Aggidai, the messenger, is one of the most important.”

There is no mention of cowrie shell divination, Oriates, or afudashe. These are not qualities associated with either Dasoyi or Babalu Aye anyway. Just Arará religion recorded as it existed. What is quite interesting however is that Ramos notes Agidai as an Orisha that is invoked in times of epidemics and diseases - a clear connection to Dasoyi / Babalu Aye! The tool placed outside Agidai's vessel - a T-shaped staff with two arms and two legs hanging from the T - also indicates a connection to diseases, particularly those of the limbs.

Ramos also mentioned that in the city of Cienfuegos, in the Cabildo de Santa Barbara, a version of Agidai associated with the Orisha Ogun is honoured. That particular Agidai lives behind the doorway of the Cabildo. Interestingly, it is not uncommon to find representations of Babalu Aye (or his regional equivalent) near the main entrance of some Brazilian and African temples. So while the Agidai of Cienfuegos may have become associated with Ogun, we are still able to see the roots of Dasoyi / Babalu Aye, who at times lives by the main gate.

It is also worth noting that the Arará name Aggidai may likely preserve a phonetic and functional memory of Agdi, a vodun found within the wider Dahomean and diasporic Babalú Ayé family of spirits. Rather than suggesting a direct equivalence, this points to a shared register of illness, abandonment, and punitive earth power, carried into Cuba through Arará transmission and reshaped by local ritual conditions.

Taken together, this strongly suggests that what later becomes Agidai, understood in Lukumí contexts as an Orisha of prophetic speech, almost certainly emerges from an Arará vodun that was gradually reinterpreted through Lukumí grammar as traditions converged. A messenger vodun operating within the Dasoyi / Babalú Ayé family is reframed as an orisha of divination and inspired speech, eventually becoming the patron of a small and specialised group of ritual experts. The function shifts. The attributes change. The name survives. Seen this way, Agidai stops being an African mystery and becomes something far more interesting. He becomes a product of Afro Cuban religious history itself. Not forgotten Africa, but creative adaptation.

This is exactly the kind of insight that The Arará in Cuba makes possible. It does not give us answers wrapped in certainty. It gives us fragments that allow better questions. And for a tradition like Regla Arará, that is precisely what makes the book so brilliant.