How divination functions as a diagnostic instrument and what kind of reasoning it supports, and Hallen and Sodipo's finding that Yoruba herbalists distinguish knowledge from belief by a criterion stricter than the English one.
Two things happen in a Yoruba consultation that a modern reader will find unfamiliar, and they are usually run together when they should be kept apart. The first is divination, which is a diagnostic instrument with a specific structure and a specific kind of reasoning behind it. The second is the practitioner's own standard for what he will say he knows, which turns out on investigation to be considerably stricter than the standard an English speaker uses when saying "I know". The first is anthropology. The second is philosophy, and it is the subject of the most important piece of work anyone has done on Yoruba medical thought.
Ifá's mechanics are 05-ifa, and this file does not repeat them. What matters here is what divination does in a medical encounter.
It answers the second question. File 01 sets out the two-register structure of Yoruba diagnosis: what is wrong, which is proximate and physical, and why this person and why now, which is about selection. Divination is the instrument for the second question. It is not usually a way of finding out what disease someone has, and a practitioner who can see the symptoms does not need Ifá to tell him what they are.
It produces a case, not a diagnosis. The output of a casting is an odù, and attached to that odù is a body of memorised verse, ẹsẹ Ifá. Each ẹsẹ is typically a narrative: a named figure faced a situation, consulted, was told to make a particular offering, did or did not do so, and a particular outcome followed. The diviner selects from the verses attached to the odù those that fit the client's situation, and the fit is the diagnosis.
The reasoning is case-based. This is the accurate description of the method and it is worth stating precisely, because it is neither random nor deductive. The system holds a large library of precedent cases indexed by a randomising procedure. The diviner retrieves a set of cases, matches the client's circumstances against them, and derives a prescription from the matched precedent. That is structurally the same operation as case-based reasoning in law or in clinical medicine before statistics: the practitioner's competence lies in holding a large corpus of precedent and in judging which precedent the present situation resembles.
Two consequences follow.
The randomising step does real work. Because the diviner does not choose the odù, he cannot simply retrieve the precedent that confirms what he already suspected. The system forces him to reason from a case he did not select. Whether this is epistemically valuable is arguable in both directions: it defeats a specific bias, and it substitutes an arbitrary selection for a considered one. It is at least not the same thing as the diviner making it up, and accounts that treat divination as pure improvisation misdescribe the procedure.
The interpretive step is where the expertise lives. The casting produces an address; the verses at that address are numerous and various; the client's situation is not stated to the diviner in advance in the strict versions of the procedure. Selecting which verse applies, and reading the client's circumstances into it, is a skilled act of pattern-matching. This is why babaláwo of different quality produce different consultations from the same odù, and why the training is long.
Ifá is not the only route. The aláṣọtẹ́lẹ̀ diagnoses without Ifá's formal apparatus . Ẹ̀ẹ́rìndínlógún, sixteen-cowrie divination, is used by òrìṣà priests including many women, and is treated in 05-ifa/08. Direct physical assessment is standard and unremarkable: a practitioner looks, palpates, asks about the stool and the urine and the appetite, and reaches conclusions in the ordinary way.
The point worth making is that a Yoruba practitioner is not obliged to divine. He divines when the case calls for it, characteristically when the illness is unexplained, protracted, recurrent, or has resisted treatment. That distribution is itself informative: divination is deployed where the physical account has run out, which is exactly where the selection question becomes pressing.
Knowledge, Belief, and Witchcraft: Analytic Experiments in African Philosophy was published in 1986 by Barry Hallen, a philosopher, and J. Olubi Sodipo, then Professor of Philosophy at what is now Obafemi Awolowo University, with a foreword by W. V. O. Quine . It remains the only sustained analysis of an African belief system conducted from inside Anglo-American analytic philosophy, and its method was to hold extended philosophical discussions with practising oníṣègùn as colleagues rather than as informants .
The method matters. Hallen and Sodipo's stated position is that learning from an oral culture requires meeting members of that culture, because oral information comes out of mouths . They approached the herbalists as fellow specialists in a body of knowledge and asked them what their terms meant, at length, in Yoruba. What emerged is not an anthropological description of Yoruba beliefs. It is a report of what a group of expert practitioners say about the conditions under which a claim counts as known.
The starting point is Quine's thesis on the indeterminacy of translation . The worry Quine raises is that when we translate an unfamiliar language we project our own conceptual scheme onto it, and cannot get behind the projection to check. Hallen and Sodipo apply this to a specific pair of words that every dictionary and every previous ethnography had treated as unproblematic.
The standard translation gives mọ̀ / ìmọ̀ as know / knowledge and gbàgbọ́ / ìgbàgbọ́ as believe / belief. Hallen and Sodipo's finding is that this mapping is wrong, and wrong in a way that matters, because the Yoruba pair divides the epistemic field at a different place than the English pair does .
Ìmọ̀ requires first-hand experience. What the onísègùn count as ìmọ̀ is what a person has seen with their own eyes and experienced at first hand, with the added condition that the experience must be consciously comprehended and not merely undergone . Walking past a shelf of books does not give you ìmọ̀ of what is on it. Propositions describing such directly experienced and comprehended states of affairs are the ones the onísègùn will call òótọ́, true .
Ìgbàgbọ́ covers everything received from others. Information that arrives through another person, through books, through the media or through oral tradition is ìgbàgbọ́ . Note the range. This is not a category of doubtful or superstitious claims. It contains almost everything anyone believes, including things that are perfectly correct, and including the entire content of the tradition the practitioner works within.
The conversion condition. Ìgbàgbọ́ is not permanently second class. If a piece of second-hand information can be tested and accordingly verified, it has the potential to become ìmọ̀ . The route from belief to knowledge runs through the believer's own verification, not through the reliability of the source.
The word gbàgbọ́ is itself instructive: it is built from gbà, to accept or receive, plus gbọ́, to hear. To believe is to accept-what-one-has-heard. The etymology carries the epistemology, and the contrast with mọ̀ is the contrast between hearing and having.
Three reasons, and each is worth separating.
It is not the English distinction. In ordinary English, "I know" is happily said of things learned from reliable testimony. I know the population of Lagos; I have never counted it. On the onísègùn's criterion that is not ìmọ̀ but ìgbàgbọ́, however confident I am and however reliable my source. The Yoruba scheme locates the knowledge boundary at personal verification, and the English scheme does not. Two languages that appear to have the same pair of concepts turn out not to, which is Quine's point demonstrated on a real case rather than a constructed one .
It inverts the standard charge. The received colonial-era and mid-century account of "traditional" thought held that such systems accept received tradition uncritically and treat inherited assertion as fact. Hallen and Sodipo's finding is precisely the reverse. Their formulation: the model of African thought produced by English-language culture typifies it as treating second-hand information, oral tradition and book knowledge, as though it were true, as though it were knowledge, and this is exactly what the Yoruba epistemological system outspokenly and adamantly refuses to do . On the criterion the herbalists actually apply, the content of oral tradition is not knowledge in the strict sense and becomes knowledge only when tested and verified . The system is more demanding about testimony than the English-speaking one, not less.
It bears on the justified-true-belief analysis. Anglo-American epistemology spent the second half of the twentieth century on whether knowledge is justified true belief and what has to be added to handle Gettier cases. The Yoruba scheme does not construct knowledge out of belief plus conditions at all. Ìmọ̀ is not a species of ìgbàgbọ́; they are different things with different sources, and no amount of justification converts one into the other without the experiential step. Whether that is a better analysis is a live question. That it is a different analysis, arrived at independently by non-academic specialists reasoning in a language with no philosophical literature, is the finding.
Apply the criterion to a working herbalist and the consequences are concrete.
A practitioner who has used a preparation and observed the outcome has ìmọ̀ of it. A practitioner who has been told by his master that a preparation works, and has not yet used it, has ìgbàgbọ́, and on the strict criterion is not entitled to say he knows. The transmission of the tradition therefore delivers, in the recipient's own terms, belief that awaits conversion. That is a rather remarkable thing for a system of inherited esoteric knowledge to say about itself.
It also supplies the verification norm described in file 02: a claim from another practitioner is not accepted on authority but held pending the hearer's own use. And it explains a feature of practitioner discourse that outside observers often read as evasiveness, namely the care with which experienced practitioners hedge claims they have not personally tested.
The limits should be stated with equal care. The criterion is about the individual's warrant, not about experimental design. Personal observation of a treated patient who recovers is uncontrolled, unblinded, and subject to every confounder in clinical epistemology, chief among them the high base rate of spontaneous recovery. A system in which each practitioner verifies personally, and in which no negative result is recorded or communicated, will accumulate confident first-hand "knowledge" of treatments that do nothing. The onísègùn's standard is genuinely more rigorous than casual English usage about the source of a claim, and it does not address the problem that controlled trials address. Both statements are true and neither cancels the other.
The book's title includes witchcraft, and the treatment of àjẹ́ is the application of the method to a harder case.
Hallen and Sodipo challenge the comparative frameworks used by earlier scholars, Geoffrey Parrinder among them, on the ground that European witchcraft categories were built substantially from trial records and other indirect evidence, and do not map onto what Yoruba practitioners actually say . They approach àjẹ́ through the Yoruba concept of ènìyàn, person, treating the àjẹ́ as a personality type within Yoruba thought rather than as an occult category, and asking what a systematic psychology built on that conception of personhood would have to address .
The wider treatment of àjẹ́ and the Ìyáàmi belongs to 06-orisa and 18-women. What belongs here is that the same epistemological result applies: claims about àjẹ́ made on the basis of what one has been told are ìgbàgbọ́ by the practitioners' own criterion, and they are frank about this in ways the older ethnographic literature did not record because it did not ask.
The work is not uncontested and the objection is specific.
Critics have argued that Hallen and Sodipo take an unduly restrictive view of testimony in Yoruba epistemic practice . The objection runs: if nothing received from another person can be knowledge until personally verified, then a great deal of what any society plainly does know becomes unavailable, and the account cannot explain how an oral culture transmits knowledge at all. Critics have therefore explored ways in which a culture like the Yoruba can treat testimony as a genuine source of justification without collapsing into the uncritical deference to tradition that such societies have historically been charged with .
There is a second, methodological question that a careful reader should raise independently: Hallen and Sodipo interviewed a particular group of specialists whose profession is built on esoteric, personally transmitted, personally verified knowledge. Practitioners of a craft in which first-hand competence is the whole of one's professional capital may hold a stricter line on the ìmọ̀ boundary than a Yoruba farmer or trader would. Whether the finding generalises from onísègùn to Yoruba speakers at large is not settled by the study's design, and Hallen's later work on Yoruba moral epistemology has itself been challenged on related grounds .
Neither objection touches the core result, which is that the Yoruba pair does not map onto the English pair and that the difference is systematic rather than accidental. Confidence on the generalisation from onísègùn to Yoruba speech at large: contested.
Yoruba diagnostic and epistemological practices trace back to early regional healing traditions that integrated systematic botanical experimentation with esoteric divination . During the rise of the Ọyọ Empire from the seventeenth through the eighteenth century, royal courts centralized healing and Ifá divination guilds, standardising diagnostic procedures and herbal formulas across regional territories .
The collapse of Ọyọ and the ensuing nineteenth-century civil wars fractured these lineage networks, dispersing practitioners into expanding military and commercial hubs such as Ìbàdàn and Abẹ́òkúta . This wartime dislocation intensified pragmatic exchange among herbalists, fostering innovative medical adaptations. Mid-nineteenth-century Christian missionary expansion brought direct biomedical competition and moral condemnation, portraying divination and indigenous epistemologies as irrational superstition .
During British colonial rule, legislation such as the 1900 Native Medicine Proclamation and later anti-witchcraft statutes systematically targeted traditional diagnostics and ritual healing . In response, Yoruba healers between 1922 and 1955 formed organized professional bodies, such as the African Herbalists Society, adopting formal licenses and case registers to defend their empirical legitimacy against colonial suppression .
After Nigerian independence in 1960, state ministries established traditional medicine boards to integrate indigenous practitioners into public health frameworks and institutional research . In the contemporary era, the oníṣègùn continue to operate across Nigeria within pluralistic healthcare markets, applying first-hand verification alongside clinical biomedicine . Across the transatlantic diaspora, including Lucumí in Cuba and Candomblé in Brazil, the core case-based reasoning of Ifá and sixteen-cowrie divination endures as an active diagnostic and therapeutic framework .
The scale of the documented Yoruba pharmacopoeia, how plants are classified and named, the tonal wordplay that makes a plant's name a mnemonic for its use, and a worked table of named plants against what pharmacological research has and has not established.
What the instruments are, how a consultation runs from the client's whispered question to the prescribed offering, and which decisions belong to the diviner and which do not.
Ifá is a Yorùbá system for reaching a decision by casting a signature, retrieving the memorised verses filed under it, and reasoning from the precedents they contain.
Yoruba categories for mental disturbance, what traditional treatment consists of including the restraint that is its worst feature, Lambo's village system at Abeokuta and why it changed community psychiatry worldwide, and the trial evidence on collaboration with healers.
Hallen and Sodipo's finding that the Yoruba distinction between knowing and believing is drawn in a different place from the English one, why that is a genuine philosophical result, and what it exposes about Western epistemology.
How truth is determined in Yoruba thought and practice, the etymology and relationship of otito and ooto, Hallen and Sodipo's imo/igbagbo distinction, and the standards of evidence the tradition applies to testimony and oath.