Ẹsẹ Ifá: The Verses, Their Structure, and How the Corpus Is Held
The verses filed under each Odù are case reports in verse, built to a described eight-part structure, held collectively across practitioners rather than by any one person.
An ẹsẹ Ifá (ẹsẹ, literally foot or leg, here a verse or stanza) is a poem filed under one of the 256 Odù. Almost every one of them tells a story of the same shape: someone in the past faced a difficulty, consulted named diviners, was told to make a particular offering, either did or did not, and something followed. The diviner recites these until the client stops him at the one that fits. What is being retrieved is therefore a precedent case, and the corpus is a body of such cases, memorised, indexed by 256 addresses, and distributed across the entire profession rather than written down anywhere.
How many there are
Nobody knows, and the sources give different estimates because they are estimating different things.
Abimbola reports the tradition's own figure: "It is believed that each ese has a total of 600 poems associated with it" Read against his surrounding text he means each Odù has around 600 verses, which puts the corpus near 153,600.
UNESCO's documentation gives approximately 800 per Odù, notes that the exact number is unknown, and adds that it "continues growing" That yields roughly 204,800.
Both figures are traditional estimates rather than counts, and neither has been verified by enumeration, because no enumeration exists. What can be said firmly is that the corpus runs to the order of a hundred thousand verses or more, that no single person holds all of it, and that no published collection comes close. Bascom's Ifa Divination prints several hundred verses; Abimbola's Sixteen Great Poems prints sixteen long ones.
The growth claim matters. UNESCO says the number continues to increase, which means the corpus is open rather than closed. New verses enter. This is not a fixed scripture.
The eight-part structure
Abimbola's analysis, which is the standard account, is that "each ese Ifá has eight structural parts out of which four are obligatory and the remaining four are optional" The eight parts, as set out in his Ifá Divination Poetry, are:
- The name or names of the Ifá priests who performed the divination in the past case.
- The name or names of the client for whom it was performed.
- The reason for the divination.
- The instruction given to the client by the diviners.
- Whether the client complied.
- What happened to the client as a result.
- The client's reaction to the outcome.
- The moral drawn from the story.
Parts 1-3 and 8 are the obligatory ones; parts 4-7 are optional .
This division does a great deal of work, and it is the key to understanding how a corpus this large stays stable in memory across generations.
Abimbola's account of what obligatory means here is precise: the obligatory parts "must be chanted in the same form and, as much as possible within human mental capacity, in the same language," so that "all Ifa priests within the same Yoruba dialect area would render the obligatory parts with the same accent and the same style of presentation" The optional parts, by contrast, "are chanted by each Ifa priest in his own language," where the priest "has freedom of creation to employ his own language, style and imagery when narrating the story of each poem the framework of which he has learnt from his master"
So each verse has a fixed skeleton and a variable body. Abimbola names this directly: "There is therefore a static and at the same time a dynamic aspect to every Ifá poem" , the static element being the myth-making core and the dynamic being "the creative and imaginative genius of individual Ifa priests"
The engineering consequence is worth spelling out. Fixing the diviners' names, the client's name, the occasion, and the moral, while leaving the narration free, means the verse's identity and its conclusion are protected against drift while its telling can adapt to the audience and the performer. A fully fixed text would be harder to memorise and would fossilise. A fully free text would not survive as the same verse across lineages. The split solves both problems at once.
The diviners' names in part 1 are frequently not ordinary personal names but compressed descriptive phrases, sometimes puns on the situation of the case. They function as much like a case citation as like a character list.
Style
Two features dominate. Abimbola: "Two of the most important and characteristic features of Ifa style are repetition and word-play," and "Both of these stylistic features often occur in combination"
Word-play in Yorùbá is doing something English cannot easily reproduce, because Yorùbá is a tonal language and shifting tone alone changes meaning. A single string of syllables can carry several senses distinguished only by tone, so a line can mean two things at once in a way that has no English equivalent. This is why the translation notes in this file repeatedly say that something is lost: it usually is, and usually it is this.
Abimbola makes an argument about function that is easy to miss and worth taking seriously. Clients patronised diviners "not only because of their belief in the predictions and pronouncements of Ifa but also because of the aesthetic enjoyment which they derive from listening to the chanting" In a non-literate society, consulting a diviner was "an important medium of contact with such a highly developed form of poetry," serving a function he compares directly to that of modern novels, poems and films . Divination was among other things how ordinary people encountered literature.
The long verses have their own status. Those running to many lines are called Ifá Nlanla (great poems of Ifá) and are "regarded as the rare gems of the whole system," which supports a proverb about their exchange value .
How the corpus is held
No individual holds the corpus. This is the single most important structural fact about it, and it has consequences for authority, verification, and what can be lost.
The minimum for initiation is documented: after the trainee "has learnt at least sixteen verses from each of the two-hundred-and-fifty-six Odù together with the sacrificial items and interpretations that go with them, he can present himself for initiation" Sixteen verses times 256 Odù is 4,096 verses as an entry requirement, with their prescribed offerings and interpretations attached. Against a corpus of 150,000 or more, a newly initiated priest holds a few per cent of it.
Specialists hold more in their own area and less elsewhere. Post-initiation training sends the new priest to a renowned specialist elsewhere, often far away, to gain depth in one field . The result is a profession in which knowledge is distributed and partly redundant: many practitioners know the common verses, fewer know any given rare one, and the total exists only as the union across all of them.
This has three consequences the reader should hold onto.
Authority is diffuse. There is no canonical text, no council that fixes the corpus, and no single person who can be asked what a verse really says. Disagreement between practitioners about a verse is not necessarily error by one of them.
Verification is peer-to-peer. Because verses overlap between practitioners, a claim can be checked against what others hold, and file 06 describes the mechanisms. The check is real but partial, and it works best on common verses and worst on rare ones.
Loss is silent and permanent. A verse held by one elderly specialist who dies without transmitting it is gone with no record that it existed. UNESCO's safeguarding rationale rests on exactly this: practitioners aging, limited resources, declining youth interest under religious pressure . The corpus can shrink without anyone being able to say what was lost.
Verses in three-layer translation
A note on the texts and on my own role. The Yorùbá texts below are taken from Abimbola's published editions, which print the Yorùbá original with his facing English translation. The digitised copies available to me are optical scans in which subdots and tone marks are frequently dropped or garbled. I have restored diacritics to standard orthography where the word is unambiguous, and I say so here rather than presenting the result as a clean transcription. Anyone quoting these texts should work from the printed editions. The idiomatic English is Abimbola's own published translation and is attributed as such. The literal glosses are mine, produced from the Yorùbá, and are marked accordingly. Confidence on the glosses is medium; confidence on Abimbola's English is high.
1. The opening of Ìwòrì Méjì
This is the first of the sixteen long poems in Abimbola's anthology, and its opening lines show the conventional structure arriving in order: the diviners' names, then the client, then the occasion.
Àpá ní ó gbó kòkò igbó, Ó rú ìtànná ẹ̀rù àjẹ́. Òrúrù ní ó wọ ẹ̀wù ẹ̀jẹ̀ kalẹ̀. Ilẹ̀ ni mo tẹ̀ tẹ́ẹ̀ tẹ̀ Kí n tó topọ̀n. Ọ̀pẹ tẹ́ẹ̀rẹ́ orí òkè Níí ya sí ya búkà mẹ́rìndínlógún. A dífá fún Ọ̀rúnmìlà, Wọ́n ní bàbá ò níí bímọ sí ètu Ifẹ̀ yìí. Mo gbọ́ títí, Mo rín wọn, rín wọn.
- Àpá the apa tree ní ó gbó it is that which grows old kòkò igbó in the depth of the forest
- Ó rú it puts forth ìtànná flowers ẹ̀rù àjẹ́ the dread/terror of wizards
- Òrúrù the oruru tree ní ó wọ it is that which wears ẹ̀wù ẹ̀jẹ̀ a garment of blood kalẹ̀ down to the ground
- Ilẹ̀ ni mo tẹ̀ it was the bare ground that I pressed/printed on tẹ́ẹ̀ tẹ̀ pressed and pressed
- Kí n tó topọ̀n before I came to press the tray (tẹ ọpọ́n, to mark the divining tray)
- Ọ̀pẹ tẹ́ẹ̀rẹ́ the slender palm tree orí òkè on the hilltop
- Níí ya sí ya is the one that branches this way and that búkà mẹ́rìndínlógún sixteen hut-like heads
- A dífá fún Ọ̀rúnmìlà Ifá divination was performed for Ọ̀rúnmìlà
- Wọ́n ní they said bàbá ò níí bímọ father would not bear a child sí ètu Ifẹ̀ yìí in this city of Ifẹ̀
- Mo gbọ́ títí I heard and heard
- Mo rín wọn, rín wọn I laughed at them, laughed at them
Translator: Wándé Abímbọ́lá
It is the apa tree which thrives in the forest and produces fearful flowers of wizards. It is the oruru tree which wears a garment of blood from top to bottom. It was on the bare ground that I printed Ifá marks Before I started using the wooden tray for divination. It is the slender palm-tree on hill-top Which branches here and there and has sixteen hut-like heads. Ifá divination was performed for Ọ̀rúnmìlà. It was said that father would never have a child in this city of Ifẹ̀. When I heard, I laughed and laughed at them.
The first three lines are the diviners' names, not scene-setting. In ẹsẹ Ifá the practitioners in the past case are routinely named by descriptive phrases, and here the phrases are tree-images that read in English as nature description. Abimbola's translation preserves the images and lets the naming function be inferred, which is a defensible choice but one that obscures part 1 of the eight-part structure for a reader who does not know it is there.
Ilẹ̀ ni mo tẹ̀ ... kí n tó topọ̀n contains a technical pun that English cannot carry. The verb tẹ̀ is to press, and tẹ ọpọ́n (contracted to topọ̀n) is the standard term for marking the divining tray; the same verb in tẹfá names the act of divining itself and, in another register, of being initiated. So the line "I pressed on the bare ground before I pressed the tray" is simultaneously a claim of great antiquity, a technical description of the marking procedure, and a pun on initiation. Abimbola's "printed Ifá marks" carries the technical sense and drops the pun.
Ọ̀pẹ tẹ́ẹ̀rẹ́ ... búkà mẹ́rìndínlógún, the slender palm with sixteen hut-like heads, is doing structural work: sixteen is the number of the palm nuts and of the principal Odù, so the image encodes the architecture of the system. English readers see a tree; Yorùbá listeners see the number sixteen announced.
The last two lines shift abruptly to first person and to laughter. The narrator is not identified. This kind of unmarked shift is normal in ẹsẹ Ifá and is one reason the verses are difficult for readers accustomed to consistent narrative voice.
2. A complete short verse: the interpretation for Ìwòrì Méjì
The verse above is one of the long poems. The short verses that make up the bulk of the corpus look different, and this one shows what a diviner actually delivers. Abimbola prints this in his appendix as the Ifá priests' traditional interpretation attached to Ìwòrì Méjì.
Ẹni tí ó dá Ìwòrì Méjì, Ifá sọ wí pé ọmọ mẹ́jọ ni òun máa tí yóò bí; gbogbo àwọn ọmọ náà ni ó sì ní àlùbàríkà. Ṣùgbọ́n ti abíkẹ́yìn wọn ni yóò pọ̀ ju ti gbogbo àwọn yòókù lọ. Ṣùgbọ́n nígbà tí abíkẹ́yìn náà bá dàgbà tán, kò níí mọ bàbá rẹ̀ sí nkankan mọ́. ... Kí ó lọ fi ẹyẹlé mẹ́rìndínlógún ṣe sàráà pẹ̀lú ẹgbàá mẹ́rìndínlógún, eku mẹ́rìndínlógún, ẹja mẹ́rìndínlógún, ìgán aṣọ kan. Kí ó lọ fi àgbébọ̀ adìẹ bọ Ifá. Kí ó fi àgbọn mẹ́ta àti obì mẹ́ta bọ orí ara rẹ̀.
- Ẹni tí ó dá Ìwòrì Méjì the person who cast Ìwòrì Méjì
- Ifá sọ wí pé Ifá says that ọmọ mẹ́jọ ni òun máa tí yóò bí eight children is what he will bear
- gbogbo àwọn ọmọ náà all those children ni ó sì ní àlùbàríkà have blessing (àlùbàríkà, from Arabic baraka, blessing)
- ti abíkẹ́yìn wọn that of the last-born among them ni yóò pọ̀ ju is what will be more than ti gbogbo àwọn yòókù that of all the rest
- kò níí mọ bàbá rẹ̀ sí nkankan mọ́ he will not reckon his father as anything any more
- Kí ó lọ fi ẹyẹlé mẹ́rìndínlógún ṣe sàráà let him go and make offering with sixteen pigeons (sàráà, from Arabic sadaqa, alms)
- pẹ̀lú ẹgbàá mẹ́rìndínlógún with sixteen times two thousand cowries
- eku mẹ́rìndínlógún, ẹja mẹ́rìndínlógún sixteen rats, sixteen fish
- ìgán aṣọ kan one bundle of cloth
- Kí ó lọ fi àgbébọ̀ adìẹ bọ Ifá let him go and offer a mature hen to Ifá
- Kí ó fi àgbọn mẹ́ta àti obì mẹ́ta bọ orí ara rẹ̀ let him offer three coconuts and three kolanuts to his own orí
Translator: Wándé Abímbọ́lá
Ifá says that the person to whom Ìwòrì Méjì appears during the process of divination would give birth to eight children, and all the eight children shall be prosperous. Ifá says that the youngest of them shall, however, be the most prosperous of them all. But when the youngest son becomes an adult, he would cease to give due respect to their father. ... This person should perform sacrifice with sixteen pigeons, sixteen times four thousand cowries, sixteen rats, sixteen fish and a bundle of cloth. He should also offer a big hen to Ifá. He should also offer three coconuts and three kolanuts to his own Orí.
Two Arabic loanwords sit inside this supposedly pre-Islamic material: àlùbàríkà from baraka and sàráà from sadaqa. Their presence is evidence of long contact between Yorùbá religious vocabulary and Islam, and it complicates any claim that the corpus is a sealed pre-contact deposit. Abimbola's English translates both into neutral religious vocabulary, so the reader of the English cannot see them.
The offering to orí is the offering to one's own inner head or destiny, a concept belonging to the cosmology section. Abimbola capitalises Orí in English, which reads as a deity; the Yorùbá orí ara rẹ̀, his own body's head, is more intimate and more literal than the English suggests.
On the cowries: I gloss ẹgbàá mẹ́rìndínlógún as sixteen times two thousand because ẹgbàá is conventionally two thousand cowries; Abimbola's English gives sixteen times four thousand. I have not resolved the discrepancy and flag it rather than silently harmonising. It may reflect regional variation in the value of ẹgbàá or a scanning error. Confidence on this figure: contested.
Note the shape of the whole. There is a prediction, a warning, and a prescription with exact quantities. Sixteen recurs four times in the offering. The diviner must hold not only the verse but the precise materials attached to it, which is what Abimbola means when he says the trainee learns the sacrifices along with the poems .
3. On the value of the great poems
Ifá nlanla ni a fi ń gbá Ifá nlanla, Oògùn nlanla ni a fi ń gbóògùn nlanla.
- Ifá nlanla great Ifá poems ni a fi ń gbá is what one uses to receive/take Ifá nlanla great Ifá poems
- Oògùn nlanla great medicine ni a fi ń gbóògùn nlanla is what one uses to receive great medicine
Translator: Wándé Abímbọ́lá
One gets a great poem of Ifá only in return for another great poem of Ifá, One gets a highly potent medicinal preparation only in return for another highly potent one.
This is a statement about the economics of knowledge in the profession, and it is more pointed than the English lets on. The verb gbà is to receive or take, and the construction is one of exchange between equals: knowledge is bartered against knowledge of like value, not taught freely. It describes how the corpus actually circulates between initiated practitioners, and it explains part of why the record is partial. A researcher without equivalent knowledge to trade has nothing the holder of a rare verse wants.
The parallel construction between poems and medicines is exact in the Yorùbá and treats the two as the same kind of asset, which supports the account in file 09 of how deeply the pharmacological and the literary are bound together in this corpus.
What kind of text this is
Several classifications get applied to ẹsẹ Ifá, and each captures something while misleading if taken alone.
Scripture. It carries doctrine and is treated with reverence, but it has no fixed text, no closed canon, no single authoritative version, and it is growing . Comparisons to the Bible or Qur'an mislead on every one of those points.
Poetry. It has metre, formal structure, characteristic devices, and aesthetic standards, and it is performed . This is accurate but incomplete, because it omits that the poems are functional retrieval units.
Case law. The structural fit is genuine: named parties, stated facts, a ruling, compliance or non-compliance, an outcome, and a holding. File 09 develops the comparison and its limits.
Oral history. Some verses encode events, migrations, and dynastic claims. Extracting history from them is possible but requires the care described in file 09, since the verses are not chronological and were not composed as history.
Pharmacopoeia. Verses carry plant knowledge and preparations, and the connection is systematic enough that Verger's 744-page study of Yorùbá plant use organises material by its links to Ifá signs .
The honest answer is that it is all of these simultaneously, and that this is unusual. The corpus was not differentiated into genres because the society did not store its knowledge in separate compartments. Filing pharmacology, precedent, poetry, ethics, and history under the same 256 addresses is exactly what makes it an information system rather than a literature.