Àjẹ́ and "Our Mothers"

Àjẹ́ names a power held by women. It is understood to be real, to be capable of building or destroying, to operate through means that are not visible, and to belong particularly to elder women. Its holders are addressed collectively and respectfully as àwọn ìyá wa, "our mothers", or ìyá mi, "my mother". English-language writing has almost universally translated àjẹ́ as "witch" and this translation is wrong in ways that have done real damage, both to understanding and, in the twentieth century, to actual women.

This is the most misrepresented subject in the tradition. It is handled here without sensationalism and without the opposite error, which is to sanitise it into a benign feminine principle and edit out the fear. The Yoruba record contains both the reverence and the terror, and the ethnographic literature documents both. A reader who wants only one of them is not going to find the tradition's own position.

Much of what is specifically known about àjẹ́ is initiate-held and not public. The public dimension, which is Gẹ̀lẹ̀dẹ́, is extensively documented. The internal dimension is documented thinly and mostly through informants speaking about a category they do not claim to belong to. That asymmetry runs through everything below and is stated rather than papered over.

Why "witch" is a bad translation

Start with what English imports.

The English word "witch" carries the whole apparatus of the European witch trials: malevolence as the defining characteristic, secret association as itself criminal, a pact with a devil, and eradication as the appropriate social response. The category was constructed in order to be prosecuted. The Yoruba category is structurally different. It names a power that is held, that is dangerous, that is also generative, and toward which the correct communal posture is propitiation and honour rather than extermination 12.

Sarah Mathews puts the translation problem precisely: within Western discourse "witch" tends to hold a negative connotation related to women who wield magical powers used to inflict harm, and the Yoruba sense of àjẹ́ "is not a direct synonym to the commonly held understanding of witch or witchcraft in Western discourse, which is rarely thought to be related to good intentions and benevolent power" 1 She notes that Margaret Drewal, Oyeronke Olajubu and Teresa Washington all make the same objection 1.

The Drewals' formulation is the standard one. Although àjẹ́ may be a derogatory term to the Yoruba, the holders are addressed as ìyá mi or àwọn ìyá wa "in recognition of their positive dimension as protective progenitors, healers... and guardians of morality, social order, and the just apportionment of power, wealth, and prestige" 3 Note the last item in that list. The mothers are held to enforce distributive justice. A power whose function includes punishing those who hoard is not doing what a European witch does.

Two further translation problems compound the first.

The naming taboo. Àjẹ́ is not the ordinary term of address. The respectful circumlocutions, ìyá mi and àwọn ìyá wa, are what is used, and they are used precisely because the direct term is dangerous or offensive. Translating the avoided word rather than the used one inverts the register: it takes the term the culture avoids and makes it the standard English label.

The tonal collision. Àjẹ́, the power, and Ajé, the òrìṣà of the market, of wealth and of profit, are distinct words with distinct tone patterns. They are routinely confused in English-language writing, which is not tone-marked, and sometimes deliberately conflated by authors arguing for a connection between women's economic power and their spiritual power 1. That connection is a real argument and is treated below, but the words are two words. The corpus keeps them apart and marks the tone. Ajé the òrìṣà is treated in The Smaller Òrìṣà.

Better renderings than "witch", none of them adequate: the mothers; women's spiritual power; the elder women's power. There is no English word. This corpus uses the Yoruba.

What the power is held to be and where it comes from

The definition in the literature

Margaret Drewal's is the most-cited formulation. Àjẹ́ is "the concentration of vital force in women, their àṣẹ, their power to bring things into existence, to make things happen, create extraordinary potential that can manifest itself in both positive and negative ways" 4 The key move is that àjẹ́ is not a separate substance but a concentration of àṣẹ, the general Yoruba concept of the power by which things are effected, treated in Àṣẹ. It is not a foreign or demonic power. It is the ordinary generative power of the cosmos, held in unusual concentration.

Àjẹ́ names both the power and the women who hold it 1.

The holders are also called ẹlẹ́yẹ, owners of birds, from the association of the power with birds and with night flight; the bird is the standard iconographic marker of the mothers in Gẹ̀lẹ̀dẹ́ headdresses and on the Ọ̀ṣùgbó and Ògbóni edan.

Who holds it

The general account is that all women hold àjẹ́ latently and that elder women, particularly women past menopause, hold it actively, the stated reason being their retention of blood 15. Apter records the associated theory of reproductive blood: positive in its capacity to mix with male sperm and create new life, negative in the menses, understood as blood ejected because it cannot create new life, and feared by men because contact with it can neutralise their most powerful medicines 6.

This is worth stating plainly rather than softening. The system attributes the power to a woman's reproductive physiology and treats the post-reproductive woman as the most powerful because the power is no longer being expended in reproduction. It is a theory of female power grounded in the body. Whatever one makes of it, it is not a system that ignores anatomy, and this is one of the sharper pieces of evidence against the strongest form of the argument that Yoruba thought did not categorise by anatomy at all. Mathews makes exactly this argument and it is the thesis of her study 1.

The origin ìtàn

Two accounts circulate in the Ifá corpus and they are not reconciled.

Odùduwà and the calabash. Olódùmarè created three òrìṣà, two anamale and one anafemale, and gave the two male ones special powers. The female, here Odùduwà, asked what her power was to be, and Olódùmarè handed her a closed calabash in the shape of the earth containing a special àṣẹ. Asked what she would do with it, she said she would use it to punish anyone who insulted her and would not hesitate to use it to benefit those who supported her. That power is àjẹ́ 7.

The moral structure of that answer is the whole subject in one sentence. The power is conditional and reciprocal from the moment it is granted. It is not malevolent and it is not benevolent. It responds.

Èṣù and the oath. A second Ifá verse holds that women went to Èṣù to ask for the power. He gave it but instructed them to go to Ọ̀rúnmìlà before taking it up, and Ọ̀rúnmìlà made them swear to honour certain materials and signs so that men would have means of protecting themselves against the power. To bind the agreement the women were sent to Ọlọ́run to swear the oath 8.

That version encodes something different: the power is prior, and what the men's institutions negotiated was not its abolition but a treaty with terms. Babatunde Lawal and Raymond Prince both record it 78.

Rank in the cosmology

The claim consistently reported is that the mothers stand above the òrìṣà, serving as intermediaries between Olódùmarè and the deities, and that no òrìṣà is more powerful than they are 137. Drewal and Drewal place them second in power only to the genderless Olódùmarè 3.

If that is what the tradition says of itself, then the position of women's spiritual power in Yoruba cosmology is not marginal, not a folk supplement to a male pantheon, and not a demonology. It is at or near the apex.

The ambivalence, stated without softening

The power is held to do both of these things and the same women are held to do them.

Generative. Protection of children and lineage, fertility, healing, prosperity in trade, the enforcement of just distribution, and the guardianship of moral and social order 3. Communities prosper when the mothers are favourable.

Destructive. Infertility and miscarriage, interference with the menstrual flow, impotence in men, debilitating disease, and death. Apter's formulation from Black Kings and Critics is that àjẹ́ "denotes the woman who consumes life and blocks the flow of reproductive fluids", making it the inverse of fertility and successful delivery 9. McIntosh documents nineteenth-century accusations that women poisoned or harmed their husbands through àjẹ́ 10.

The idiom throughout is consumption and flow. The mothers eat; they block; they withhold. The imagery is of a stoppage in circulation, whether of blood, of children, or of goods. This matters for the economic argument below, because the same idiom governs both.

Olajubu's fieldwork includes an account from a woman describing her own initiation, which is worth reporting because it is one of very few first-person accounts in the literature and because it does not fit a sanitised reading:

She was instructed that her left leg was ese awo (leg of secret), and on it a cut was made to drain the blood into one of the small native pots. Each member then tasted the blood. Incisions were made on her head and eyelids to make her tough, hard hearted, and able to kill as often as is required... a live bird, which she had to swallow at her initiation. 11

That is a former initiate's account, given to a Yoruba scholar of religion, and it describes a society whose stated purpose includes killing. It is presented here as testimony about what an informant said, which is what it is, and not as an established description of a practice. The corpus does not resolve it. It does note that any account of àjẹ́ which contains only protection and healing is not reporting the ethnographic record.

Gẹ̀lẹ̀dẹ́ as the institutional response

Gẹ̀lẹ̀dẹ́ is the public, annual, expensive, artistically elaborate institution by which the western Yoruba address the mothers. It is treated in full at Gẹ̀lẹ̀dẹ́ and only its logic as a response to àjẹ́ is given here.

The structure of the answer is propitiation, not suppression. Gẹ̀lẹ̀dẹ́ honours the mothers, names them, entertains them, and asks them to turn their power toward the community. The Ẹfẹ̀ night opens with invocation of the mothers before it moves into satire, and the daytime Gẹ̀lẹ̀dẹ́ is display in the marketplace, which is the mothers' own ground 12. UNESCO's inscription notes it as the only known masked society also governed by women 13.

Read as social technology, this is a serious solution to a difficult problem. A society that believes elder women hold decisive and dangerous power has two options: try to destroy the holders, or build an institution that acknowledges the power publicly and gives the holders a stake in the community's flourishing. Gẹ̀lẹ̀dẹ́ is the second. It converts a diffuse fear into a scheduled, funded, artistic obligation, and it puts the community in the posture of a supplicant to women rather than a prosecutor of them.

That is precisely why what happened in 1950 is significant.

The Atinga witch-hunt, 1950 to 1951

What happened

In 1950 and 1951 an anti-witchcraft movement called Atinga, coming out of the southern Gold Coast, moved eastward into western Yorubaland. Its practitioners went from town to town claiming to detect witches. Women identified or induced to confess were beaten, stripped of property, and in some cases killed 11014.

This is the single largest documented episode of violence against Yoruba women in the historical record, and it took place in living memory, in the same region and among the same communities that maintained Gẹ̀lẹ̀dẹ́.

Why it is a genuine analytical problem

State the puzzle sharply. If the tradition held that the mothers were more powerful than the òrìṣà, that they were to be honoured rather than eradicated, and that an entire annual institution existed to keep relations with them in good order, then why did large numbers of Yoruba communities suddenly participate in hunting and killing the women who held that power?

Two bad answers are available and both circulate.

The first is that the practice was always latent and Atinga merely revealed a suppressed misogyny. That does not explain the timing, the geography, or the movement's foreign origin.

The second is that it was simply imported: Christian and colonial demonology reframed àjẹ́ as satanic and licensed prosecution. Mathews raises this possibility explicitly, asking whether Western ideology about what defines a witch and how to deal with one influenced these actions 1. There is something in it, since Atinga arrived from outside and since mission Christianity did supply a translation of àjẹ́ into "witch" with the eradication logic attached. But it does not explain why the movement found such purchase in these particular communities at this particular moment when it had not before.

Apter's explanation

Andrew Apter's "Atinga Revisited: Yoruba Witchcraft and the Cocoa Economy, 1950-51", in Jean and John Comaroff's Modernity and Its Malcontents (University of Chicago Press, 1993), gives the explanation that has held up 14.

His core claim is that "the cocoa economy intensified structural contradictions that were already articulated by the logic of witchcraft" 14 Unpack that carefully, because it is doing several things at once.

The logic of àjẹ́ was already a language for talking about a specific kind of social wrong: hidden accumulation, blocked circulation, wealth taken out of the flow that ought to sustain the lineage and the community. A woman accused of àjẹ́ was, in the idiom, one who consumed and withheld rather than fed and released 9. Belasco's observation is that trade wealth among women routinely produced accusations, on the reading that they were hoarding goods and using the power to do it 1, and Apter's phrase for what the accusation names is the attempt "to block the flow of productive resources" 1

The cocoa boom then supplied exactly the conditions that idiom was built to describe, at unprecedented scale. Cash cropping produced rapid, visible, individually held wealth. It set husbands against wives commercially, since Yoruba women traded on their own account and cocoa money moved through hands that customary obligation could not easily reach. It set literate youth against elders. It created accumulation that was real, private and unaccountable to the older redistributive claims.

Apter's summary sentence is the one to hold: "if cocoa production stimulated greater commerce and trade, it pitted men against women, literate youth against elders... Atinga attacked the female body as icon and agent of commodity value; of false representation, of unbridled circulation, and of hidden accumulation" 14

So the answer to the puzzle is that Atinga was not a rejection of the traditional understanding of àjẹ́. It was that understanding applied, under conditions of economic transformation that made its central accusation, that someone is hoarding what should circulate, feel true of a great many people at once. The women attacked were disproportionately the ones who looked, in the idiom, like accumulators: traders, women with independent money, elder women.

Two things follow that are worth stating.

The economic reading does not exonerate anyone and Apter does not offer it as exoneration. It explains why the violence took the shape and target it took. It does not make it something other than the killing of women.

And the older institution and the violence are not opposites drawn from different worlds. Gẹ̀lẹ̀dẹ́ and Atinga are two responses to the same premise. The premise is that elder women hold real and dangerous power over the circulation of life and wealth. Gẹ̀lẹ̀dẹ́ answers it with honour; Atinga answered it with a hunt. A reader who wants the tradition to have contained only the first is asking the record for something it does not hold.

The economic dimension, and Karin Barber's observation

The link between women's wealth and àjẹ́ accusation is documented independently of Atinga and it is the most socially consequential feature of the whole category.

Karin Barber's finding from Okuku is the sharpest statement of it. "Big Women" in the Okuku area could not convert wealth into reputation the way "Big Men" did, by acquiring land, marrying wives and having many children, and a woman who flaunted her wealth as a man did would be accused of witchcraft 15.

That is a ceiling, and it is a precisely described one. The mechanism is not that women could not become rich, since they demonstrably did. It is that the conversion of wealth into public standing, which was the whole point of wealth in a Yoruba town, ran through a display repertoire that was closed to women on pain of accusation. The Ìyálóde title was one of the few legitimate conversion routes available, which is part of why it mattered so much.

Mathews and Teresa Washington both note that the power and wealth of Ẹfúnróyè Tinúbú were popularly attributed to her possession of àjẹ́ 116, and the same attribution attaches to Ẹfúnṣetán Aníwúrà, alongside the cruelty narrative examined in The Ìyálóde 1. In both cases a woman's commercial success and her destruction by men are linked through the accusation.

The general point is that àjẹ́ operated as a sanction on female accumulation. It was simultaneously an acknowledgement that women held real power and an instrument for punishing them when they used it too visibly. Both functions are the same category doing its work.

Where the scholars disagree

On whether àjẹ́ should be read primarily as a generative principle. Teresa N. Washington's Our Mothers, Our Powers, Our Texts: Manifestations of Àjẹ́ in Africana Literature (Indiana University Press, 2005) and The Architects of Existence: Àjẹ́ in Yoruba Cosmology, Ontology, and Orature (2014) argue for reading àjẹ́ as a fundamental generative force, recovering it from the witchcraft frame 16. Washington writes from an engaged position and her work is best read alongside the ethnographic accounts rather than instead of them. The ethnographers, particularly Apter, keep the destructive dimension central and treat the category as a discourse about power and its abuses rather than as a positive metaphysics 914.

On what the category proves about Yoruba gender. Mathews's thesis is that àjẹ́ and Ajé are decisive evidence against Oyěwùmí's claim that the category "women" did not exist in Yorubaland before European contact, on the grounds that a power theorised through female reproductive physiology and attributed specifically to elder women cannot be described without the category 1. Her argument on Atinga is that the threat of successful women traders in a growing economy shows that female power grounded in the body existed with great force before European involvement 1. This is a real argument and it is one of the more concrete objections to the strongest form of Oyěwùmí's thesis. The counter-position is set out in Gender.

On whether the power is female-exclusive. Most of the literature treats àjẹ́ as women's. Some accounts admit male holders under other names, usually oṣó, sorcerer, treated as a different and lesser thing. The relationship between àjẹ́ and oṣó is not consistently reported and the corpus does not settle it.

What a reader should take from this

Three things.

The category is not witchcraft and translating it that way imports a prosecution logic the Yoruba framework does not contain. The framework's own answer to female spiritual power was an institution of honour, which is a genuinely unusual thing for a society to have built.

The power is ambivalent by definition, and the fear is as documented as the reverence. Accounts that keep only the reverence are producing a modern devotional literature, not a description.

And the category had a social function beyond the metaphysical one: it marked the boundary of what a woman could visibly accumulate. When an economic transformation made accumulation general, that boundary was enforced violently against real women within living memory. Any account of Yoruba women's spiritual power that stops before 1950 is stopping before the hardest part of the evidence.