Witchcraft and Medicine

Witchcraft occupied the center of medical life in early modern Europe in ways that modern readers are likely to underestimate. Disease attribution, diagnosis, legal testimony, and the entire framework of who counted as a legitimate healer were all entangled with questions about whether illness could originate in human malice, demonic agency, or both. Keith Thomas’s Religion and the Decline of Magic (1971) is the foundational account for England; Ilza Veith’s Hysteria (1965) and Andrew Scull’s Madness: A Civilisation (2015) trace the parallel story in the history of mental illness. Together they show that medicine and witchcraft were not opposite poles but competitors within a shared explanatory field, and that the gradual displacement of one by the other was driven by social and institutional forces as much as by intellectual argument. E.E. Evans-Pritchard’s fieldwork among the Azande of central Africa, published as Witchcraft, Oracles and Magic Among the Azande (1937, abridged 1976), provides the single most detailed study of how a witchcraft system actually works: how it explains illness, how oracles diagnose it, how medicines combat it, and why the system is internally coherent enough to resist falsification from within. The 1976 edition is an abridgment by Eva Gillies that reduced the original to roughly half its length, omitting technical anthropological discussions while retaining all core ethnographic content.(Evans-Pritchard, E.E., 1976) The Azande numbered approximately one million at the time of fieldwork, governed by a conquering aristocracy called the Avongara.(Evans-Pritchard, E.E., 1976) Eva Gillies’s editorial assessment, with which subsequent scholars have broadly agreed, is that no other single society has been studied in such detail from the standpoint of witchcraft belief, and that the theoretical implications extend well beyond the Azande themselves.(Evans-Pritchard, E.E., 1976)

Thomas provides a foundational structural explanation for why magical healing persisted so robustly alongside orthodox medicine: the physical and social environment of early modern England — high mortality, epidemic disease, poor sanitation, and scarce medicine — created conditions of vulnerability that made resort to magical and religious healing a rational response to circumstances beyond conventional medical control.(Thomas, Keith, 1971) The Reformation’s destruction of sacramental healing made this worse: Protestant iconoclasm systematically dismantled the shrines, healing wells, and sacred sites that had provided the infrastructure of supernatural healing, removing institutional resources precisely when popular need remained constant.(Thomas, Keith, 1971) The suppression of official supernatural medicine paradoxically increased demand for informal alternatives, explaining the persistence and even expansion of cunning folk practice through the sixteenth and seventeenth centuries.(Thomas, Keith, 1971)

Ancient Concepts: Pharmakeia and Strix

The Greek word pharmaka meant both “drugs” and “poisons,” and this duality was inseparable from the figure of the practitioner who wielded them.(Riddle, 1985) The pharmakis — the woman who prepares pharmaka — occupied an ambiguous position in ancient Greek society, simultaneously valued healer and feared sorceress. Roman culture added the strix, a nocturnal creature (often a witch transformed) that sucked blood and caused sudden infant deaths, embedding harmful magic in the etiological structure of unexplained illness.

The Hippocratic On the Sacred Disease (c. 400 BCE) mounted an explicit early attack on magical disease explanation, arguing that epilepsy — often attributed to divine or demonic agency — had a “natural cause from which it originates like other affections.”(Andrew Scull, 2015) This naturalistic move defined a lasting professional identity for learned medicine: the physician who explains disease by bodily process rather than supernatural cause. But this identity remained contested for two millennia. Even learned physicians who rejected demonic disease causation as a general principle typically acknowledged that the boundary with supernature was difficult to draw with confidence in any particular case.

Medieval Entanglement: Church Medicine and Demonic Causation

The medieval Church did not simply oppose medicine and magic; it provided the institutional framework within which both operated. Saints’ cults organized supernatural healing: Thomas Becket for general ailments, Christopher for sudden death, Uncumber for troubled marriages — each saint a specialized therapeutic agent.(Thomas, Keith, 1971) Holy water, consecrated hosts, and bells were employed for healing sick cattle and crops as well as sick humans.(Thomas, Keith, 1971) This sacramental medicine was not “superstition” within the Church’s own categories; it was orthodox practice.

Veith’s survey of the medieval period notes that for roughly eight hundred years, from the fifth century to the thirteenth, medical knowledge effectively stagnated: clerical medicine was the only literate tradition, and prevailing concepts of illness and healing arguably regressed from those of earlier antiquity, with miraculous cure by saints and relics displacing the more naturalistic disease models of the ancient world.(Ilza Veith, 1965)

Augustine’s theology, however, planted the seeds of a different framework. By populating Christian cosmology with demons, incubi, succubi, and witches — adapted from pagan supernatural figures — Augustine gave later demonologists the tools for interpreting illness as the work of malevolent agents, human or demonic.(Ilza Veith, 1965) His distinction between natural illness (treatable by medicine) and demonic affliction (requiring miraculous cure) established the dual-register structure that physicians and clerics negotiated for centuries after him.(Andrew Scull, 2015)

The Malleus Maleficarum (1494), commissioned by Pope Innocent VIII and running to thirty editions within two centuries, codified this framework into judicial doctrine.(Ilza Veith, 1965) Kramer and Sprenger drew a specific distinction: possession by a demon was an accident, and Christ had healed demoniacs; but the witch had made a pact with the devil, committing heresy. Heretics were not patients but criminals.(Ilza Veith, 1965) This theological move transformed what had been a medical question — is this person sick? — into a juridical one: has this person sinned? Veith’s retrospective reading of the Malleus concludes that many of those tried as witches showed symptoms consistent with hysteria: partial anesthesia, mutism, blindness, convulsions, and sexual delusions.(Ilza Veith, 1965) The medico-legal implications were total: the Inquisition had occupied medical terrain.

The Witch Trials Era: Medical Testimony, Identifying Marks, Possession vs. Disease

English witchcraft took a distinctive form. Where Continental authorities emphasized the diabolical pact and the sabbath, English prosecutions concentrated overwhelmingly on maleficium — harmful acts causing disease, death, or property damage to neighbors.(Thomas, Keith, 1971) Approximately 1,000 people were executed for witchcraft in England between 1542 and 1736, far fewer than on the Continent, where common law procedure, the absence of judicial torture, and growing judicial skepticism all constrained prosecution.(Thomas, Keith, 1971) Across Europe as a whole, between 50,000 and 100,000 people were executed as witches in the early modern period, and witchcraft and demonic possession were articles of belief among educated people as well as common ones.(Andrew Scull, 2015) The Witchcraft Act of 1604, enacted under James I, broadened prosecution categories and reflected the king’s personal investment in demonological theory, illustrated by his authorship of Daemonologie (1597); royal patronage of witch-hunting ideology gave it legislative momentum that intensified judicial willingness to treat disease caused by malevolent supernatural agency as a capital matter.(Thomas, Keith, 1971)

Early modern English law and medicine recognized witchcraft and poisoning as conceptually distinct but practically overlapping categories, with the legal problem lying in the impossibility of separating supernatural causation from toxic substance: poison worked by natural means and left material traces, while witchcraft by definition left none.(Thomas, Keith, 1971)

The social logic of the accusation was specific and well-documented. Thomas reconstructs it from a verbatim account in Thomas Ady: “She was at my house of late; she would have had a pot of milk; she departed in a chafe because she had it not; she railed; she cursed; she mumbled and whispered; and finally she said she would be even with me; and soon after my child, my cow, my sow, or my pullet died, or was strangely taken.”(Thomas, Keith, 1971) The accused were disproportionately poor, elderly women living at the margins of community charity — the neighbors most likely to be refused, most likely to curse, most likely to be blamed when subsequent misfortune struck.(Thomas, Keith, 1971) Thomas specifies this demographic profile precisely: the typical accused witch was old, poor, female, and socially dependent, often a widow or single woman relying on neighbors for basic subsistence.(Thomas, Keith, 1971) Margaret Murray’s rival thesis — that witch prosecutions targeted a real, surviving pre-Christian fertility cult — cannot withstand scrutiny: the confessions Murray used as evidence were products of torture and leading questions, not genuine accounts of cult practice, and no coherent surviving pagan religion is detectable in the record.(Thomas, Keith, 1971) A further problem for the demonological framework is that many of those convicted could not recite the Lord’s Prayer correctly and knew little of formal Christian doctrine — a religious ignorance that exposed the gap between the learned model of the witch-as-Satan-worshipper and the actual women appearing in the courts.(Thomas, Keith, 1971)

Physicians were called to testify at witch trials about the nature of the alleged victim’s symptoms. Their position was structurally uncomfortable.(Thomas, Keith, 1971) Courts expected medical opinion on whether symptoms were natural or supernatural; physicians who privately doubted supernatural causation nonetheless faced pressure to perform deference in legal proceedings. A minority of physicians, notably Reginald Scot, John Oxenbridge, and Edward Jorden, explicitly resisted witchcraft diagnosis and argued for natural causes behind apparently supernatural symptoms, representing an early professional medical skepticism toward supernatural disease attribution.(Thomas, Keith, 1971) This skeptical tradition drew on a broader current in learned letters: Robert Burton’s Anatomy of Melancholy, Francis Bacon’s natural philosophical program, and Jorden’s own Briefe Discourse (1603) each located apparently supernatural afflictions within the domain of natural bodily process, collectively building the evidentiary case that physicians rather than judges should arbitrate questions of possession.(Thomas, Keith, 1971) The 1602 trial of Elizabeth Jackson for allegedly bewitching Mary Glover crystallized the problem: Edward Jorden, Fellow of the Royal College of Physicians, diagnosed hysteria, but the court ruled in favor of witchcraft and imposed punishment.(Ilza Veith, 1965) Jorden then published his diagnosis as A Briefe Discourse of a Disease Called the Suffocation of the Mother (1603), systematically refuting every symptom used as evidence of bewitchment — insensibility to pin pricks, convulsions, periodicity of fits, choking — by showing them to be standard features of a natural disease.(Ilza Veith, 1965) This was the first systematic English medical argument that demonic possession symptoms were explicable as natural illness.(Thomas, Keith, 1971) Once physicians could offer credible naturalistic diagnoses for possession symptoms, courts progressively lost confidence in supernatural attribution, making Jorden’s treatise a professional pivot point in the witchcraft debate.(Thomas, Keith, 1971)

The convulsive fits, sensory distortions, and visions characteristic of demonic possession cases may in some instances reflect ergotamine toxicity from contaminated rye grain, a naturalistic hypothesis proposed by later historians as one of several environmental explanations for possession epidemics in early modern England.(Thomas, Keith, 1971) Thomas engages cautiously with the more specific ergotism hypothesis — that contamination of rye supplies in wet years correlated with outbreaks of mass possession symptoms — noting its partial plausibility while acknowledging the complexity of the evidence.(Thomas, Keith, 1971) Freud’s 1923 essay “A Seventeenth-century Demonological Neurosis,” which analyzed a possession case as masked depression, established the template for psychoanalytic retrodiagnosis of historical spirit-possession; Thomas cites this approach without endorsing it, noting its methodological ambitions rather than its conclusions.(Thomas, Keith, 1971) The theory of ocular fascination — that a malevolent glance could cause disease by transmitting harmful influence through the eye — had genuine defenders among learned physicians who worked out physiological mechanisms involving “venomous effluvia” emanating from the gaze.(Thomas, Keith, 1971) The boundary between learned and popular disease explanation was permeable at this point. Witchcraft as a category of disease causation was not simply folk belief held by the ignorant; it had professional supporters who attempted to give it mechanistic grounding within humoral and pneumatic frameworks.

Cunning Folk and Medical Practice

Cunning folk — the village specialists in charms, divination, and counter-magic — occupied a central position in early modern English medical care. They were the diagnostic first contact for many people who suspected illness had a supernatural cause, frequently consulted before physicians and sometimes instead of them.(Thomas, Keith, 1971) When illness struck, the question “is this witchcraft?” preceded the question “what is the natural cause?” — and cunning folk answered the first question as a professional specialty.

Their techniques included urine-gazing (uroscopy), crystal-gazing, sieve and shears, and key and book.(Thomas, Keith, 1971) Uroscopy was shared with licensed physicians: the same diagnostic procedure appeared in learned texts like the “Iudycyall of Uryns” (c. 1527) and Thomas Brian’s Pisse-Prophet (1637), operating across the orthodox/magical boundary.(Thomas, Keith, 1971) They also treated reproductive complaints — magical abortion and contraception are documented in ecclesiastical court records — and ran a full service economy of healing that overlapped substantially with both licensed medicine and parish pastoral care.(Thomas, Keith, 1971)

Despite being technically illegal under the Witchcraft Acts of 1542, 1563, and 1604, cunning folk were rarely prosecuted. Essex Quarter Sessions data shows only 13 of 48 cases involved white magic; York ecclesiastical court data (1567–1640) records only 25 of 117 cases resulting in punishment, demonstrating effective social toleration.(Thomas, Keith, 1971) The Royalist government periodically moved against seventh-son healers as competition for the royal touch’s monopoly on scrofula cure,(Thomas, Keith, 1971) but cunning folk as a class survived the formal legal framework largely intact.

Cunning folk charged fees for their services, documented in Lancashire Quarter Sessions records, with amounts in some cases comparable to licensed physician charges.(Thomas, Keith, 1971) They served as a parallel medical system addressing needs that orthodox practitioners either refused or could not supply: identification of witchcraft-caused illness, diagnosis of the responsible witch, recovery of stolen goods, and love magic.(Thomas, Keith, 1971)

Richard Napier, the early-seventeenth-century clergyman-physician whose casebooks (preserved in the Ashmolean manuscripts) document some 2,000 mental illness consultations, provides a unique archive of the integrated magical-medical encounter in practice. Napier cast astrological charts, recorded whether patients attributed their symptoms to witchcraft, prescribed herbs, prayer, or referral, and noted cases where patients reported “temptation by spirits.”(Thomas, Keith, 1971) His casebooks at Ashbourne include numerous patients describing themselves as tempted by spirits or afflicted by diabolical agency, creating one of the earliest documented clinical populations at the intersection of mental illness and spirit belief.(Thomas, Keith, 1971) His records demonstrate that magical and medical illness attribution coexisted at the point of care without requiring resolution — a clinician could work with both frameworks simultaneously.

Decline of Witchcraft Belief and the Medical Role

The Reformation created unexpected consequences for the medical management of witchcraft. Protestant rejection of Catholic exorcism as superstitious removed the Church’s primary ritual tool for treating demonic possession, leaving a therapeutic vacuum that cunning folk, physicians, and later Methodists competed to fill.(Thomas, Keith, 1971) Canon 72 of 1604 forbade Anglican clergy from performing exorcisms without episcopal license, structurally pushing possession cases toward medical practitioners as the default institutional authority for symptoms crossing supernatural and somatic boundaries.(Thomas, Keith, 1971) The Darrell controversy of the 1590s foreshadowed this institutional shift: John Darrell’s high-profile dispossessions, followed by his prosecution by ecclesiastical authorities, forced the first sustained English institutional debate about whether possession symptoms were supernatural or natural in origin, and the outcome effectively discredited Protestant exorcism as a legitimate clerical practice.(Thomas, Keith, 1971)

The Protestant Reformation also created a sharp conceptual binary between “true religion” (based on faith and scripture) and “magic” (associated with ritual efficacy), which the medieval Church had never enforced clearly.(Thomas, Keith, 1971) This binary eventually worked against witchcraft prosecution: once magical causation was definitionally excluded from legitimate Christian practice, the intellectual basis for believing in witch-caused illness became harder to defend from within Protestantism itself.

Successful witchcraft prosecutions became increasingly rare after the 1660s well before the Witchcraft Act was repealed in 1736, showing that judicial practice shifted before legal reform — elite skepticism about supernatural disease causation changed practice from above.(Thomas, Keith, 1971) Richard Bentley identified the agents of change with precision: “What then has lessen’d in England your stories of sorceries? Not the growing sect [of free-thinkers], but the growth of Philosophy and Medicine. No thanks to atheists, but to the Royal Society and College of Physicians; to the Boyles and Newtons, the Sydenhams and Ratcliffs.”(Thomas, Keith, 1971)

An earlier challenge to medical authority had come from Paracelsus, who publicly burned Avicenna’s Canon and delivered lectures in German rather than Latin, violently rejecting Galenic-Arabic concepts and the natural philosophy underpinning them; his insistence that all mental diseases had natural causes, including hysteria, helped loosen the grip of medieval scholastic medicine even while replacing it with systems no less speculative.(Ilza Veith, 1965) Johann Weyer’s De praestigiis daemonum (1563) argued that many women accused of witchcraft were mentally ill and deserved pity and medical treatment rather than punishment, providing an early medico-legal defense of the accused from within a learned framework.(Ilza Veith, 1965) Reginald Scot’s Discoverie of Witchcraft (1584) had established the template for skeptical rationalism, arguing that witch beliefs rested on misidentified natural phenomena, the delusions of elderly women, and fraudulent imposture.(Thomas, Keith, 1971) John Webster’s Displaying of Supposed Witchcraft (1677) translated mechanical philosophy into direct challenge to witchcraft’s medical claims, arguing case by case that symptoms cited as evidence of bewitchment were better explained by natural causes.(Thomas, Keith, 1971) Hutchinson’s Historical Essay Concerning Witchcraft (1718) completed the learned refutation.(Thomas, Keith, 1971)

Thomas’s account resists the simple Enlightenment narrative of science defeating superstition. His central argument is that magical beliefs declined not because they were empirically refuted but because the social and intellectual conditions that sustained them changed: new welfare institutions, new explanatory frameworks, and improved therapeutic alternatives made magic less necessary without requiring any direct demolition of its claims.(Thomas, Keith, 1971) The new natural philosophy did not refute witchcraft belief directly but made it progressively unnecessary: as mechanical, chemical, and probabilistic explanations became available for disease, crop failure, and unusual atmospheric events, the range of phenomena requiring supernatural causation narrowed correspondingly, without demanding any direct intellectual confrontation with witch-belief.(Thomas, Keith, 1971) The decline of witchcraft belief resulted from multiple converging forces: expanded poor relief reducing the charity-refusal pattern that generated accusations; natural philosophy providing alternative explanatory frameworks; shifting elite judicial attitudes; the plague’s disappearance from England after 1665 removing a major driver of protective magical demand; and the growing institutional monopoly of licensed medicine over legitimate healing.(Thomas, Keith, 1971)

The expansion of parish poor relief under the Elizabethan Poor Laws reduced the moral charge that drove accusations: when institutional mechanisms mediated the relationship between poor petitioners and reluctant neighbors, the guilt of refusal — and the curse it provoked — became less acute, reducing the social conditions that generated witchcraft accusations at their root.(Thomas, Keith, 1971) Agricultural improvement contributed a parallel structural change: as better crop varieties, drainage, and farming techniques reduced catastrophic harvest failures, and as credit and insurance buffered losses, the experience of unaccountable agricultural misfortune that had driven so many accusations became less frequent and less existentially threatening.(Thomas, Keith, 1971) The medical professionalization Bentley praised was not only intellectual but institutional: the Royal College of Physicians, the Society of Apothecaries, and eventually the hospital system created formal professional monopolies that progressively excluded magical practitioners from legitimate healing, a process of institutional displacement operating alongside the intellectual critique.(Thomas, Keith, 1971) Magic was not defeated; it became redundant for those who had access to the new institutions.(Thomas, Keith, 1971)

The Azande: Witchcraft as a Complete Explanatory System

Evans-Pritchard’s fieldwork among the Azande of southern Sudan and the Congo-Nile watershed in the late 1920s produced the most analytically detailed account of a witchcraft system in the ethnographic record.(Evans-Pritchard, E.E., 1976) His central argument, refined over the book, is that Zande witchcraft beliefs are not primitive errors but a coherent explanatory framework that fills a specific logical gap: they answer “why me, why now?” rather than “how?”

Azande understand perfectly well why termites eat the wooden supports of a granary, why crops fail in drought, why people fall ill from infection.(Evans-Pritchard, E.E., 1976) What they ask is why these impersonal natural processes converge on a particular person at a particular moment. Evans-Pritchard demonstrated this through his famous granary example: when a man sitting beneath a granary is killed by its collapse, Azande acknowledge the physical cause (termites) but insist that something else — witchcraft — explains why that man sat under that granary at the precise moment it fell.(Evans-Pritchard, E.E., 1976) Evans-Pritchard called this “dual causation”: the natural and the mystical are not competing explanations but accounts of two different questions, and neither cancels the other.(Evans-Pritchard, E.E., 1976)

The Organic Substance: Mangu

The materiality of Zande witchcraft sets it apart from many European and Islamic traditions, which located malefic power in knowledge, pact, or will. For Azande, witchcraft is a physical substance — mangu — located near the liver, visible at autopsy as a dark reddish swelling.(Evans-Pritchard, E.E., 1976) This substance is hereditary and transmitted unilinearly: sons of male witches are witches, daughters of female witches are witches, but the trait does not cross sex lines.(Evans-Pritchard, E.E., 1976) The harm done by witchcraft is not caused by the substance directly but by its soul (mbisimo mangu), which leaves the witch’s body at night and attacks the victim.(Evans-Pritchard, E.E., 1976)

This organic model has significant social consequences. Because almost any Azande could in theory carry the substance — inheritance being widespread — what determines whether a person is accused is not biology but social context: enmity, prior quarrel, and a misfortune requiring a responsible agent.(Evans-Pritchard, E.E., 1976) The system does, however, exempt one class entirely: the Avongara ruling aristocracy are considered above witchcraft accusation. A prince is never said to be a witch; nobles neither bewitch one another nor are bewitched by commoners, placing them outside the ordinary economy of suspicion.(Evans-Pritchard, E.E., 1976) A person accused of witchcraft does not react with outrage; since the substance can operate without the witch’s awareness, the accused can sincerely deny intent while accepting that their substance may have acted.(Evans-Pritchard, E.E., 1976) A witch does not necessarily know that he is bewitching someone, yet is held morally responsible and addressed as though he could control his witchcraft if he wished.(Evans-Pritchard, E.E., 1976) Evans-Pritchard observed that even those identified by an oracle as witches typically accept the oracle’s general reliability while doubting its accuracy in their particular case — what he called situational logic.(Evans-Pritchard, E.E., 1976) The substance itself may be “cool” or inactive: Azande distinguished between active and dormant witchcraft, so a confirmed finding at autopsy did not automatically condemn all patrilineal kin.(Evans-Pritchard, E.E., 1976) Post-mortem examination of the abdomen was in fact practised by Azande to confirm whether a dead person had harboured witchcraft-substance, and a positive finding retroactively validated prior accusations against that person.(Evans-Pritchard, E.E., 1976)

Oracle Consultation as Medical Diagnosis

When illness or misfortune strikes, Azande work through a diagnostic hierarchy of oracles before reaching conclusions about cause and treatment.(Evans-Pritchard, E.E., 1976) The poison oracle (benge) stands at the top of this hierarchy and is consulted for serious matters; below it rank the termites oracle (dakpa), the rubbing-board oracle (iwa), and finally witch-doctors. The poison oracle operates by administering a strychnine-containing substance to fowls while binary questions are put to it; survival or death constitutes the answer, and a second fowl is dosed to confirm the result.(Evans-Pritchard, E.E., 1976)

The consultation follows a fixed social protocol. When illness is suspected to have a supernatural cause, the sick person (or his household) first consults a lesser oracle; if the result is serious, the poison oracle is convened.(Evans-Pritchard, E.E., 1976) If the oracle names a witch, the accused is not confronted directly; a deputy of a chief delivers a polite notification, the witch is asked to withdraw his witchcraft, and compliance is signalled by blowing water from the mouth.(Evans-Pritchard, E.E., 1976) After this notification a wing from the fowl that died in declaration of witchcraft is brought to the accused, who blows water onto it and declares ignorance of any bewitching; this ritual binds both parties to the oracle’s finding.(Evans-Pritchard, E.E., 1976) The Azande did not accuse in the abstract: a person was only named as a witch in connection with a specific misfortune, and only someone already known to be on bad terms with the sufferer was likely to be named.(Evans-Pritchard, E.E., 1976) The parallel to early modern English practice is striking: Thomas documents the same structure — prior quarrel, refusal, curse, subsequent misfortune — operating in a completely different cultural context.

The poison oracle is surrounded by strict taboos: the operator must abstain from sexual intercourse, avoid certain foods, and keep the oracle away from women; breaches of taboo are held to corrupt the oracle’s verdicts.(Evans-Pritchard, E.E., 1976) Women are entirely excluded from witnessing or participating in oracle consultations, and the benge must not be polluted by female presence, such that a woman approaching the seance would invalidate the entire proceeding.(Evans-Pritchard, E.E., 1976) Evans-Pritchard provides verbatim transcripts of oracle seances, showing the formulaic address (“Poison oracle, this man says that he is suffering from witchcraft… kill the fowl if he is suffering from witchcraft”), the binary framing of questions, and the social reasoning behind each query.(Evans-Pritchard, E.E., 1976)(Evans-Pritchard, E.E., 1976) Questions put to the oracle are structurally constrained so that a positive result and a negative result each map onto a practical course of action, preventing the oracle from being caught in outright contradiction.(Evans-Pritchard, E.E., 1976) A different, more extreme form of oracle consultation operated in pre-colonial times: in serious cases such as accusations of adultery or sorcery, the benge was administered to the human suspect himself; survival indicated innocence and death confirmed guilt.(Evans-Pritchard, E.E., 1976)

Princes keep their own poison oracle as an instrument of political and judicial power; the king’s oracle is considered beyond question and serves as the court of final appeal in witchcraft disputes.(Evans-Pritchard, E.E., 1976) Challenging a prince’s oracle verdict amounts to lèse-majesté; the political authority of the prince reinforces the oracle’s epistemological authority, making the two systems mutually sustaining.(Evans-Pritchard, E.E., 1976) Azande do not, however, conceive of benge as a natural poison in the ordinary sense. It is a mystical agent whose virtue is of a different order from ordinary substances, acting not through chemistry but through supernatural power.(Evans-Pritchard, E.E., 1976)

Minor Oracles and Dreams

Below the poison oracle in the hierarchy sit three minor oracles used for everyday questions. The termites oracle (dakpa) operates by inserting two sticks cut from different trees into a termite run; the oracle’s answer depends on which stick the termites have eaten more by morning.(Evans-Pritchard, E.E., 1976) The three sticks oracle (mapingo) works by balancing a small pile of three sticks and asking a question; if the pile holds the answer is positive, if it falls the answer is negative; it is regarded as less reliable than the rubbing-board and reserved for minor queries.(Evans-Pritchard, E.E., 1976)

The rubbing-board oracle (iwa) consists of a small wooden table and a wooden lid; the operator rubs the lid across the table while posing questions, and adhesion indicates a positive answer while free sliding indicates a negative one.(Evans-Pritchard, E.E., 1976) The board is manufactured from special woods with ritual precautions, anointed with plant juices, and requires taboo observance before use, though the constraints are less strict than those governing the poison oracle.(Evans-Pritchard, E.E., 1976) Evans-Pritchard proposed that the rubbing-board works through slight and unconscious muscular movements by the operator, whose expectations and social knowledge shape the outcome without his being aware of it — an early articulation of what anthropologists would later call operator-expectancy bias in divinatory systems.(Evans-Pritchard, E.E., 1976)

Dreams occupy the same ontological register as oracle consultations. A bad dream is not a symbol of witchcraft for Azande but an actual experience of it: in sleep the soul leaves the body and encounters the soul of witchcraft directly, making the dream a direct nocturnal event rather than a representation of danger.(Evans-Pritchard, E.E., 1976)(Evans-Pritchard, E.E., 1976) A dream of witches may therefore be followed by a rubbing-board or poison oracle consultation to identify the witch encountered, treating the dream as evidentiary material rather than as psychological content.

Witch-Doctors: Healers Between Systems

Zande witch-doctors (abinza) combine the roles that European medicine separated: diviner, magician, herbalist, and physician.(Evans-Pritchard, E.E., 1976) They identify witches through seances involving rhythmic dancing and drum music; they treat illness by extracting witchcraft objects from patients’ bodies; they administer plant medicines with both pharmacological and magical purposes.(Evans-Pritchard, E.E., 1976) Their social position is structurally ambiguous: respected for knowledge but also feared as possible sorcerers, since the same powers that combat witchcraft could theoretically be turned to cause it.(Evans-Pritchard, E.E., 1976)

Evans-Pritchard directly witnessed the bogwozu trick — a witch-doctor concealing a piece of charcoal in a poultice and later “extracting” it from a patient’s body as a witchcraft object — and challenged his informant Kamanga about it. Kamanga’s response is analytically important: he acknowledged the sleight of hand but argued that patients need visible evidence of extraction, and that the genuine plant medicines were doing the real work; the theatrical surgery was a concession to human psychology, not a fraud.(Evans-Pritchard, E.E., 1976)(Evans-Pritchard, E.E., 1976) This exchange foreshadows debates in medical anthropology about placebo, performance, and the therapeutic function of ritual that would occupy scholars for the next half-century.

Azande possessed an extensive empirical pharmacopoeia alongside their magical practices. Evans-Pritchard collected almost a hundred medicinal plants from a two-hundred-yard stretch of path, and Azande demonstrated genuine skill at identifying symptoms and providing prognoses.(Evans-Pritchard, E.E., 1976) The defining feature of Zande medicine is that Azande did not distinguish between plants with demonstrable pharmacological effects and those with only magical rationale: both were classed as ngua (medicines/magic) and operated in the same ritual manner.(Evans-Pritchard, E.E., 1976) Evans-Pritchard was careful to note that some medicines administered by witch-doctors may carry genuine pharmacological effects, but that Azande themselves make no such distinction: a drug and a charm are equally ngua, both operated in magical rites.(Evans-Pritchard, E.E., 1976) As disease severity increased, the balance shifted from drug administration toward oracle consultation and counter-magic — the more serious the illness, the greater the weight placed on mystical diagnosis and intervention.(Evans-Pritchard, E.E., 1976)

Initiation and Pharmacopoeia

Witch-doctor initiation involves a structured curriculum of plant-knowledge acquisition and ritual boundary-crossing. The novice eats communal meals of special plant medicines at the source of a stream, and his teacher guides him to plant specimens in the bush and in his homestead garden, teaching him their identification and use.(Evans-Pritchard, E.E., 1976) Part of the initiation involves a symbolic death: the novice lies in a shallow trench and is covered with leaves while medicines are applied, marking the separation of ordinary social identity from the witch-doctor identity being acquired.(Evans-Pritchard, E.E., 1976)

Scepticism about individual practitioners does not unravel the system. When a Zande sees through the tricks of a particular witch-doctor, he concludes only that this man is a fraud, not that witch-doctoring as such is false; the belief that witchcraft is real and that medicines combat it remains intact.(Evans-Pritchard, E.E., 1976) This is the same closed-loop structure Evans-Pritchard identified in the oracle system: each auxiliary belief protects the core, so local failures of evidence do not reach the foundational assumptions.

Secondary Elaboration and the Closed System

The most intellectually consequential part of Evans-Pritchard’s analysis concerns how the oracle system maintains its authority despite the inevitable occasions when its verdicts are contradicted by events. He identified eight secondary elaborations available to Azande when the oracle fails: the benge may have been bad; a taboo may have been broken; sorcery may have corrupted the oracle; a more powerful oracle may have overridden the result; and so on.(Evans-Pritchard, E.E., 1976) These explanations form a closed logical system in which every failure is accounted for by invoking another mystical factor, making the oracle immune to falsification from within Zande thought.(Evans-Pritchard, E.E., 1976) Azande could not step outside the mystical idiom to subject it to external critique; the concepts of witchcraft, oracles, and magic were so interlocked that each explained and validated the others.(Evans-Pritchard, E.E., 1976)

The same closure applies to magic more broadly. Evans-Pritchard enumerated twenty-two reasons why Azande cannot perceive the futility of their magic, among them: magic is primarily deployed against other mystical powers rather than directly at the natural world; the belief system is intellectually coherent so that each element explains and confirms the others; negative results are always explicable by invoking some auxiliary failure; and social pressure sustains practice even when results are absent.(Evans-Pritchard, E.E., 1976)

Evans-Pritchard was careful not to characterize this as irrational. His point was the opposite: Zande beliefs are internally consistent, and once the idiom is learned, each mystical idea follows on another as reasonably as one common-sense idea follows another in European society.(Evans-Pritchard, E.E., 1976) He reported that during fieldwork he came to act as though oracle verdicts were true — not because he believed them intellectually, but because participating in Zande social life was impossible without doing so.(Evans-Pritchard, E.E., 1976) This methodological reflection became foundational to debates about rationality, relativism, and the limits of cross-cultural comparison in the philosophy of social science.

Appended to the 1976 edition is a formal epistemological typology. Evans-Pritchard distinguished mystical notions (which attribute supra-sensible qualities to phenomena that cannot be derived from observation or logically inferred from it) from common-sense notions (derivable from empirical observation) and scientific notions (methodical, experimental, and revisable by evidence); Zande witchcraft falls unambiguously in the mystical category, but this classification is descriptive, not pejorative.(Evans-Pritchard, E.E., 1976) The typology became a reference point in subsequent philosophy of social science debates about whether mystical beliefs can be evaluated by universal rational standards or only by criteria internal to the culture that holds them.

Witchcraft, Magic, and Vengeance

Evans-Pritchard described witchcraft, oracles, and magic as three sides of a conceptual triangle: oracles diagnose witchcraft and identify witches; magic combats witchcraft or takes vengeance on its practitioners; and witchcraft is the primary affliction that the other two address.(Evans-Pritchard, E.E., 1976) The vengeance medicine (bagbuduma) was the most destructive and simultaneously the most honourable of Zande medicines — it acted as both judge and executioner, seeking the witch responsible for a death and, if no guilty party was found, returning to destroy the person who had deployed it wrongly.(Evans-Pritchard, E.E., 1976) This auto-corrective feature built a moral regulator into the vengeance system itself: using the medicine against an innocent person was dangerous to the user.

Good magic (wene ngua) was distinguished from sorcery not by its destructive potential but by its moral sanction: good magic acted against witches, thieves, and adulterers — persons who had broken social law — while sorcery was used against innocent victims.(Evans-Pritchard, E.E., 1976) When a death was attributed to witchcraft, the dead person’s kin were obligated to use oracles to identify the witch and either exact compensation or deploy vengeance magic; the oracle hierarchy was mobilised across months or years to confirm the identification and validate the eventual vengeance.(Evans-Pritchard, E.E., 1976) Death, Evans-Pritchard wrote, is the situation in which Zande witchcraft, oracles, and magic operate in their most coherent and integrated form — it is death that most fully reveals the logic of the whole system.(Evans-Pritchard, E.E., 1976)

The formal sequence of vengeance involves both an avenger boy (who observes taboos enabling the magic to act) and a magician who owns the vengeance medicine; the process may extend over months to years, with the oracle asked at each stage whether the identified witch has yet died as a result of the medicine.(Evans-Pritchard, E.E., 1976) Vengeance is formally concluded when the prince’s oracle confirms the kin’s oracle finding; the fowl wings from the consultations are then hung on a tree beside a public path, constituting formal public notification that vengeance has been accomplished.(Evans-Pritchard, E.E., 1976)

Magic spells (sima) in the Zande system are not fixed formulae. The magician addresses the medicines in ordinary language, describing their task; the virtue of the rite lies in the medicines themselves rather than in the words of address.(Evans-Pritchard, E.E., 1976) Medicines are classified both by botanical form (bulbs/ranga, parasitic plants/ngbimi, creepers/gire) and by mode of use (whistles, body-medicines, infusions, soot-oil mixtures), with different forms associated with different magical purposes.(Evans-Pritchard, E.E., 1976) Crucially, the aim of most Zande magic is not to produce direct effects in the natural world but to combat mystical interference: a farmer’s medicines are deployed to keep witches away from his crop, not to stimulate the eleusine’s growth.(Evans-Pritchard, E.E., 1976) The concept of the soul of a medicine (mbisimo ngua) parallels the soul of witchcraft: the medicine’s virtue is believed to travel out in steam and smoke when cooked, seeking its target in the same way the witch’s soul leaves the body to harm victims.(Evans-Pritchard, E.E., 1976) This structural parallel between witchcraft and counter-witchcraft medicine is an index of how deeply the mystical idiom organises the entire domain.

Non-Witchcraft Evil Agents

Witchcraft does not exhaust the category of supernatural danger among the Azande. Evans-Pritchard documented several non-witchcraft evil agents, the most feared of which are the adandara wild cats: creatures whose sight is considered fatal, and who are held to be born to women through supernatural congress with male cats. Unlike witches, adandara are not social agents embedded in human community relations; they are natural-supernatural creatures feared on sight.(Evans-Pritchard, E.E., 1976) Nocturnal birds and the irakorinde (a person who cut upper teeth first, considered ominous) also belong to this category of non-witchcraft evil. The glossary of Zande terms that Evans-Pritchard appended to the 1976 edition distinguishes the key concepts systematically: mangu (witchcraft-substance and witchcraft), ngua (magic and medicines), soroka (oracles), sima (spell), mbisimo (soul), atoro (ghosts), and mbori (supreme being), each naming a distinct ontological domain within the system.(Evans-Pritchard, E.E., 1976)

Colonial Disruption and the Mani Association

British rule over the Azande altered the practical operation of witchcraft beliefs without dismantling their conceptual structure. The government declared it illegal to execute an alleged witch; the old practice of slaying witches was suppressed and replaced by a system of compensation and social pressure. The beliefs themselves remained intact; only the punitive action associated with them was modified.(Evans-Pritchard, E.E., 1976) The decline in witch-doctor authority followed a different mechanism: the British prohibited witch-doctors from publicly naming the witches they identified in their seances, removing the primary social function that gave those seances their meaning and authority.(Evans-Pritchard, E.E., 1976)

The closed associations that proliferated among the Azande in the early twentieth century, including the Mani association, are products of colonial disruption rather than indigenous Zande tradition. Evans-Pritchard argued that they were not only introduced after European conquest but are functions of European rule and signs of the breakdown of traditional social structures.(Evans-Pritchard, E.E., 1976) Mani challenges established Zande hierarchies of sex, age, and status: women participate on equal terms with men and may hold leadership positions and sponsor male initiates, a configuration directly contrary to the social order governing the poison oracle and most public ritual life.(Evans-Pritchard, E.E., 1976)

Mani initiation rites involve crawling under hoops, anointing with boiled plant medicines, receiving a secret name, and wearing a creeper girdle that comes with food and sex taboos; the format parallels witch-doctor initiation, and both draw on a shared pharmacological repertoire.(Evans-Pritchard, E.E., 1976) Each Mani lodge is an independent local unit with no formal relations with other lodges, though members recognise one another across lodge boundaries.(Evans-Pritchard, E.E., 1976) The association posed a direct political challenge to noble authority: members settled minor disputes among themselves without reference to the prince’s court, controlling wives independently and acquiring magical sanctions outside the noble judicial system.(Evans-Pritchard, E.E., 1976) Princes and nobles accordingly opposed Mani, seeing in it an independent source of commoner authority.

The Anglo-Egyptian Sudan government declared closed associations illegal in 1919 under the Unlawful Societies Ordinance. Before European intervention, membership was known to everyone in a locality; the suppression converted what had been a publicly acknowledged association into a genuinely secret one, driving it underground without destroying it.(Evans-Pritchard, E.E., 1976)

Modern Anthropological Perspectives

I.M. Lewis’s cross-cultural analysis distinguishes witchcraft accusations from spirit possession as different social strategies: possession tends to be used by subordinates to press claims on superiors, while witchcraft accusations typically run between equals or by a superior against a subordinate.(Lewis, I. M., 2003) This structural distinction helps explain both the gender profile of witchcraft accusation (poor women accused by slightly more prosperous neighbors) and the class dynamics Thomas documents in England. Thomas himself situates English witchcraft within a wider comparative frame, noting parallel cross-cultural evidence from S.F. Nadel on Nupe spirit mediumship, Raymond Firth on Malay spirit belief, and Richard Lieban on Cebuano sorcery — documentation confirming that witchcraft and sorcery operate as widespread organizing concepts for illness attribution across otherwise unrelated societies.(Thomas, Keith, 1971)

Walter Cannon’s work on “voodoo death” — demonstrating that extreme fear following a death-curse could produce fatal physiological responses in certain cultural contexts — provides a naturalistic mechanism for understanding how the subjective experience of being bewitched could result in genuine illness or death, without invoking supernatural causation.(Thomas, Keith, 1971) The social function of witchcraft accusation was partly to supply what naturalistic medicine could not: an emotionally satisfying identification of cause, culprit, and moral meaning for suffering that would otherwise appear random and meaningless, transforming anonymous illness into a story of social transgression with a responsible agent.(Thomas, Keith, 1971)

The evil eye tradition, documented across the Mediterranean by Maloney (1976), shows that the specialist practitioner who treats magically-caused illness — the Italian mago, the Jewish counter-magic expert — persists into contemporary practice. In contemporary Italy, wizards who treat evil eye ailments advertise through newspaper classifieds.(Maloney, Clarence (ed.), 1976) This continuity undercuts any simple periodization that locates “witch medicine” in the early modern past.

Lewis’s broader argument — that spirit possession and witchcraft belief serve as idioms for expressing social conflict and protest — provides an anthropological framework for Thomas’s sociological findings. The witch accusation is not irrational; it is a culturally specific mechanism for assigning responsibility for suffering, explaining why this person was struck down at this moment, in ways that neither naturalistic medicine nor random chance could supply.(Thomas, Keith, 1971)

See Also

• magic-and-medicine
• folk-medicine
• hysteria
• possession
• cunning-folk
• evil-eye
• exorcism
• Edward Jorden
• Richard Napier

Sources

Primary evidence for this page comes from:

• Evans-Pritchard, E.E. (1976). Witchcraft, Oracles and Magic Among the Azande. Abridged by Eva Gillies. Oxford: Clarendon Press. [Source ID: evans-pritchard-witchcraft-1976] — Lead authority for Azande comparative material
• Thomas, K. (1971). Religion and the Decline of Magic. London: Weidenfeld & Nicolson. [Source ID: thomas-religiondeclinemagic-1971] — Lead authority for English history
• Veith, I. (1965). Hysteria: The History of a Disease. Chicago: University of Chicago Press. [Source ID: veith-hysteria-1965]
• Scull, A. (2015). Madness in Civilisation. London: Thames & Hudson. [Source ID: scull-madnesscivilization-2015]
• Lewis, I.M. (2003). Ecstatic Religion. 3rd ed. London: Routledge. [Source ID: lewis-ecstaticreligion-2003]
• Maloney, C. ed. (1976). The Evil Eye. New York: Columbia University Press. [Source ID: maloney-evileye-1976]
• Riddle, J. (1985). Dioscorides on Pharmacy and Medicine. Austin: University of Texas Press. [Source ID: riddle-dioscorides-1985]

HUMAN-NOTES

Editorial Notes

Gaps the encyclopaedia compiler flagged for future evidence work, collected from inline markers in the body and frontmatter.

Ancient Concepts: Pharmakeia and Strix