Food, Ritual and Taboo
How Yoruba religious thought treats food as a channel of spiritual power, structured by deific and lineage prohibitions, the mechanics of sacrifice as a food transaction, and the ethics of eating together.
How Yoruba religious thought treats food as a channel of spiritual power, structured by deific and lineage prohibitions, the mechanics of sacrifice as a food transaction, and the ethics of eating together.
Food in Yoruba religious thought does more than sustain the body. It carries and communicates spiritual power, so what is eaten, avoided, and shared is governed by rules attached to lineage, to particular òrìṣà, and to an individual's own destiny. This file goes past the general shape of that system, which Food (slug: social-food) already covers, into three places that reward closer treatment: what specific òrìṣà actually eat and refuse, how sacrifice works mechanically as a transaction in food, and what Yoruba ethics asks of people at the level of the shared meal.
Èèwọ̀ (taboo, prohibition, from è-wọ̀, that which is forbidden to enter or be touched) organizes Yoruba dietary restriction at several nested levels: universal prohibitions binding on everyone, civic prohibitions attached to a kingdom or town, lineage prohibitions carried by descent, prohibitions attached to a particular òrìṣà's devotees, and personal prohibitions assigned through Ifá divination. Food sets this typology out in full, with the reasoning behind it: that a food taboo is best understood not as hygiene or superstition but as a boundary that constitutes a relationship, with an ancestor, a deity, a lineage, or a destiny, and that breaking it is a rupture in that relationship rather than primarily a health event. That file also covers the New Yam Festival's ritual sequence in full. Neither is repeated here.
What social-food covers only briefly is the specific content of what individual òrìṣà eat and refuse, the working mechanics of sacrifice as a food transaction, and the ethics of the shared table. Those are this file's ground.
Liturgical food across the pantheon sorts broadly along an axis of heat and coolness: hot, red, fermented, or intoxicating substances for the martial and forceful òrìṣà, and white, cool, unfermented substances for Ọbàtálá and the other òrìṣà funfun. That color logic, and the reasoning behind it, is set out in Soups, Stews, and Condiments (slug: soups-stews-and-condiments), which this file does not repeat.
What follows draws together, by òrìṣà, what this corpus can actually confirm from the individual òrìṣà files and their sources, rather than compiling a list from unverified secondary compilations. A caution stated plainly in the corpus's own Ọ̀ṣun file applies across this whole subject: published lists of a given òrìṣà's èèwọ̀ circulate widely online and are frequently unsourced, and this file follows that file's practice of not asserting an item unless a primary or scholarly source stands behind it .
| Òrìṣà | Documented offering or preference | Documented prohibition | Primary source |
|---|---|---|---|
| Ọbàtálá | White food generally; shea butter (orí) as cooking fat; the giant land snail (ìgbín), whose cooling fluid (omi ìgbín) is used in pacification | Palm wine (ẹmu); red palm oil (epo pupa); salt, in most accounts | Olupona, and composite documentation, at Ọbàtálá / Òrìṣàńlá |
| Ṣàngó | Not independently confirmed beyond the taboo below | The white bean called sẹ́sẹ́, specifically, per Johnson's stated reason that the bean counteracts the ritual agencies used to attract lightning; looser popular formulations broaden this to "cowpea" or "beans" generally, but the primary source names only sẹ́sẹ́ | Johnson 1921, at Ṣàngó |
| Ògún | Dog (ajá); rooster (àkùkọ); palm oil poured over the shrine's granite stone, per Johnson; roasted yam and palm wine, per composite tradition | The hare (ehoro), never sacrificed in Ògún Onírè worship specifically, because in the Odù Ìdìngbè narrative the hare succeeded in reconciling Ògún with his son the Onírè where the dog, the ram, and the cock did not | Johnson 1921 and Adeleye 2019, at Ògún |
| Èṣù | Red palm oil, poured in libation on his laterite stone, per Johnson | Not independently confirmed in this corpus's Èṣù file beyond the general point that a sacrifice withheld from him fails to reach ọ̀run | Johnson 1921, at Èṣù |
| Ọ̀ṣun | Honey, her characteristic offering | A guinea corn (ọkà bàbà) prohibition is widely reported in popular and practitioner sources and has a documented Ifá narrative behind it, given below, but this corpus's own Ọ̀ṣun file states explicitly that it could not verify any published list of her èèwọ̀ against a citable scholarly source, so the prohibition is kept here at confidence: medium rather than treated as settled | Compiler's note at Ọ̀ṣun; the guinea corn narrative below |
Two things follow from laying the table out this way rather than compiling the fuller-looking version an earlier draft of this file carried. First, the individual òrìṣà files are where the primary sourcing already lives, and this file's job is to gather and point to it rather than to restate it with invented specificity. Second, where a popular claim could not be traced to a primary source, saying so plainly is more useful to a reader than smoothing the gap over with unattributed confidence, which is the corpus's standing rule.
A note on two claims that an earlier draft of this file carried and that this rewrite has cut rather than kept, because the rescue effort found nothing to rescue. Ṣàngó's preferred foods were given as àmàlà paired with gbẹ̀gìrì and bitter kola chewed during prayer, and a specific prohibition on ram meat among some lineages was claimed. Principal Dishes (slug: principal-dishes) already investigated an àmàlà-with-gbẹ̀gìrì offering claim for Ṣàngó, sourced there to Rowland Abiodun, and found Abiodun's actual book is a study of Yoruba aesthetics and verbal art rather than an ethnography of food offerings, with no other source located supporting the pairing as a ritual offering; that file cut the claim rather than retain it unsourced, and this file follows the same finding rather than reintroduce it. The ram prohibition is also unconfirmed anywhere in this corpus's Ṣàngó sourcing, and sits awkwardly against the same draft's own claim, two sentences later, that rams are Ṣàngó's primary sacrificial animal. Ògún's prohibition was given in an earlier draft as the field cricket (ìrè or òkèkẹ́); this corpus's own Ògún file, sourced to Adeleye's 2019 study of the Ògún Onírè festival, documents instead a specific, narratively grounded taboo on the hare, and no source located anywhere supports a field cricket prohibition. The hare taboo is used above in its place.
An Odù Ifá narrative explaining Ọ̀ṣun's guinea corn prohibition is reported in practitioner and popular scholarship on Yoruba taboo, centered on a figure named Jẹ́ńjòkè, described as Ọ̀ṣun's close attendant, who unknowingly fed her game that had eaten guinea corn, causing an affliction that only divination diagnosed. This substance is corroborated across more than one non-corpus source discussing Yoruba èèwọ̀, including an essay specifically on the subject by the Yoruba priest and writer Fagbemijo Amosun Fakayode . It is not, however, confirmed to be present, in the specific Yoruba wording an earlier draft of this file quoted, inside Wande Abimbola's published Ifá corpus, which is where that draft attributed it. Abimbola's Ifá: An Exposition of Ifá Literary Corpus is a real and major work of Ifá scholarship , but this corpus was not able to locate the Jẹ́ńjòkè verse inside it, only a paraphrase of the same narrative in Fakayode's separate essay. Rather than present unverified Yoruba lines under Abimbola's name, this file reports the narrative in summary, attributed to where it was actually found, and keeps it at confidence: medium. A reader wanting the verse in the original should check Abimbola's corpus directly; this corpus has not confirmed it there.
Sacrifice, ẹbọ, is the general mechanism by which anything crosses between the human world (ayé) and the world of spirits and divinities (ọ̀run), and its fuller philosophical treatment, including why the missionary reading of it as fearful appeasement misdescribes how it actually works, belongs to Ẹbọ: Sacrifice and Exchange (slug: cosmology-ebo), which this file assumes rather than restates. What belongs here is narrower: sacrifice specifically as a food transaction, its component offerings, and how a sacrificial meal is actually divided among the parties to it.
Àdìmú is the bloodless food offering, prepared in the kitchen and set before a shrine on a clean plate or calabash without the fuller ritual apparatus of a prescribed sacrifice. Cosmology-ebo's own typology places this as the meal and drink offering made daily by individuals and families, the most frequent form of ẹbọ by a wide margin precisely because it is domestic and unremarkable . Common items include ẹ̀kọ (steamed cornmeal), àkàrà (fried bean fritters), èkuru (unseasoned steamed beans, associated with Ọbàtálá for its plainness), and iyán (pounded yam).
Ẹbọ ẹ̀jẹ̀ is offering that involves the blood of a consecrated animal, pigeon, goat, fowl, or ram depending on what is prescribed. This is the category that corresponds to what English calls sacrifice in its narrowest sense, and it is the smallest share of total ẹbọ practice by volume, reserved for occasions a diagnostic divination has specifically called for it.
Ètùtù is the pacification or cooling offering, made to appease anger or quiet a crisis, using mild, cooling substances such as water, snail fluid, or shea butter. Cosmology-ebo's typology marks this category by a technical feature worth restating here because it bears directly on food: an ètùtù offering is not shared. It belongs entirely to the divinity it addresses and is disposed of as instructed, in contrast to most other categories of ẹbọ, which are eaten by the participants . This is the clearest point at which Yoruba sacrificial food practice distinguishes an offering that stays wholly on the spiritual side of the transaction from one that is a genuine shared meal.
No ẹbọ is generic. What is offered, in what combination, is drawn from the odù cast in the divination that prescribed it, and each component is understood to correspond to a part of the situation the divination diagnosed . This is why a worked scenario is more useful than a list of ingredients: a family whose child suffers repeated illness consults a babaláwo, who casts and identifies an odù naming, say, a hen, palm oil, and kola as the prescribed offering; the offering is prepared, presented at the appropriate shrine or crossroads with Èṣù's portion set aside first, and afterward the hen is cooked and eaten by the family and the officiating priest together. The illness's resolution is understood to follow from the offering having been correctly made, not merely from the food's nutritional content, and if the divination instead specifies an ètùtù, the family will not eat what was offered at all, since that category's whole point is that the offering leaves the human side of the transaction entirely.
Every ẹbọ, whatever its primary recipient, sets aside a first portion for Èṣù, because he is the one who carries offerings from ayé to ọ̀run, and an offering not carried does not arrive. This is documented for Èṣù specifically at his own file, sourced to Johnson's 1921 description of libations of palm oil poured on his shrine stone, and the theological reasoning behind why every sacrifice needs his participation belongs to cosmology-ebo . His characteristic share is red palm oil, roasted corn, and the internal organs or blood of the animal offered.
Beyond Èṣù's portion, Yoruba theology draws a working line between what the spiritual recipient consumes and what the human participants eat. The spiritual essence, the àṣẹ, together with the blood, the aroma, and the offering's unseen vital force, is understood to be what the òrìṣà or ancestor actually receives. The cooked meat, the yam, and the prepared soup belong to the human side of the transaction, and after the ritual presentation is complete, this food is shared among the officiating priest, the person who requested the sacrifice, and their family. Eating this sacrificial meat, ẹran ẹbọ, in fellowship is understood as taking in the protection and blessing the rite generated, which is the same logic that makes commensality generally significant in Yoruba ethics, treated below.
Yoruba proverb tradition treats eating together and refusing hospitality as matters with real ethical weight rather than as etiquette. Òwe on the Stranger, the Guest and Hospitality (slug: owe-stranger-guest-hospitality) documents this in depth from Owomoyela's proverb collection, including the specific asymmetry that a host stands, for the guest's duration, in something like a parent's role, and this file does not restate that ground . What is worth adding here is the religious layer underneath the social one: Yoruba oral tradition holds that divinities and ancestral spirits sometimes visit human towns disguised as travelers or beggars, which means that refusing a stranger food carries the specific risk of having refused an òrìṣà. This gives the ordinary social obligation to feed a guest a theological backstop, and it is one reason the obligation is treated as close to unrefusable rather than as a matter of taste.
A worked scenario, illustrative rather than drawn from a single attributed source: a traveler arrives at a compound at dusk, road-worn and known to no one there. Under the ordinary rule the household head is obliged to offer kola and a meal before any question of the visitor's business is raised, because the offer of food precedes interrogation of the guest's purpose. Should the household turn the stranger away hungry and it later emerges, through some sign, misfortune, or a diviner's later reading, that the visitor carried something of the numinous about them, the refusal is remembered against the household specifically as an offense against hospitality's religious floor, not merely as poor manners.
The corollary runs the other way too. Èèwọ̀ on food is not violated by feeding a guest who does not share the host's own prohibitions; a household bound by a lineage taboo against, say, monitor lizard is not understood to transmit that taboo to a visitor, and serving a guest food the host's own family cannot eat is not a breach. What the ethics of the table actually enforces is generosity and inclusion at the point of sharing, not uniformity of what may be eaten. Òwe on Taboo, Pollution and the Forbidden (slug: owe-taboo-pollution-forbidden) documents, from the same proverb tradition, how èèwọ̀ itself is treated as personal and household-specific rather than universal, which is the same principle operating here from the other direction .
An earlier draft of this file also attributed a household-risk proverb, "Bí ará ilé ẹni bá ń jẹ kòkòrò burúkú, Bí a kò bá kìlọ̀ fún un, Ahẹrẹhẹrẹ rẹ̀ kò ní jẹ́ kí a sùn lálẹ́" ("if a household member eats something dangerous and is not warned, their suffering keeps the household awake"), to "E. B. Idowu and P. A. Dopamu, African Traditional Religion: A Definition (Ibadan: University Press, 1980)." No such co-authored 1980 book exists. E. Bolaji Idowu's real book of a similar title, African Traditional Religion: A Definition (London: SCM Press, 1973), has no co-author. P. Ade Dopamu's real co-authorship is with Awolalu, in West African Traditional Religion (Onibonoje, 1979), cited at [S8] above. The draft's citation is a garbled conflation of these two genuine but separate books and has been cut. The proverb itself is real and circulates in Yoruba oral proverb tradition independent of either scholar, expressing the same household-risk-sharing logic documented at Òwe on Taboo, Pollution and the Forbidden; it is not quoted here as a direct text because this corpus could not trace it to a specific citable scholarly collection, and presenting it without one would violate this corpus's translation rule that any quoted Yoruba text needs a named, verifiable translator. Confidence: medium on the proverb's authenticity as oral tradition, cut as a direct quotation.
Two further citations in an earlier draft, Barry Hallen and J. Olubi Sodipo's Knowledge, Belief, and Witchcraft: Analytic Experiments in African Philosophy (London: Ethnographica, 1986), and T. J. Bowen's Central Africa: Adventures and Missionary Labors in Several Countries in the Interior of Africa from 1849 to 1856 (Charleston: Southern Baptist Publication Society, 1857), are both real books. Hallen and Sodipo's book is genuine but is a work of analytic philosophy examining Yoruba epistemology through witchcraft (àjẹ́) discourse, in dialogue with Wittgenstein and Quine, and does not treat food taboo as an ontological marker in the way an earlier draft cited it for; that citation has accordingly been dropped rather than kept misattributed. Bowen's book is a genuine nineteenth-century Southern Baptist missionary account of Yorubaland, correctly identified in an earlier draft though under a shortened title, and this corpus does not use it here as neutral historical fact but flags it, per this corpus's standing practice, as a source to be read for what missionary observers saw and how their theology shaped their description, not as an authority on Yoruba religious meaning.
An earlier draft of this file also claimed that the Ondo kingdom's taboo on the giant pouched rat (òkété) traces to a specific royal migration or founding narrative, citing Jacob K. Olupona, Kingship, Religion, and Rituals in a Nigerian Community: A Phenomenological Study of Ondo Yoruba Festivals (Stockholm: Almqvist & Wiksell, 1991), a genuine and confirmed monograph. This corpus was able to confirm that an òkété prohibition among Ondo people is real and independently corroborated. It was not able to find any source, in Olupona's book or elsewhere, connecting that prohibition to a founding or migration narrative. The Ondo kingdom's own well-documented founding tradition, behind the royal title Ọ̀sẹ́mọwẹ́, concerns a taboo on twin birth rather than a rat. The founding-narrative claim is therefore cut as unsubstantiated and likely a conflation with that separate twin tradition; the òkété prohibition itself is not repeated in this file's table above because this corpus could not independently confirm which òrìṣà or lineage it is understood to be attached to, which is the level of specificity this file's table requires.
The agricultural base, yam and its ritual centrality, the staple dishes and their regional variation, palm oil, food taboos and their lineage and orisa basis, and the social meaning of feeding people.
What ebo does inside the system's own logic, its dependence on divination, the range from daily offering to symbolic substitution, and why the missionary reading of it as appeasement was wrong.
Twelve Yorùbá proverbs built on èèwọ̀ (taboo) and related concepts of the forbidden, showing how proverb-speech treats prohibition as binding, socially variable, and in at least one case, lethal to violate.
Thirteen Yorùbá proverbs about ẹbọ, sacrifice and offering, showing how ordinary speech treats the exchange between humans and the unseen as a transaction with real limits and real risks of interception.
Sourced Yorùbá proverbs on the duties of host and visitor, the structural vulnerability of the stranger in a Yorùbá town, and the reciprocal etiquette that hospitality is built on.
The white orisa who moulds human bodies, the contested creation narrative in which Oduduwa takes his place, the palm wine story and what Ife devotees actually say about it, and the politics carried inside the itan.