Spirit Possession

Spirit possession refers to the state in which a person is understood by their community to be inhabited, controlled, or spoken through by an external spirit or supernatural agency. The phenomenon appears across virtually every human culture from ancient times to the present. Some traditions treat possession as dangerous illness requiring exorcism; others cultivate it as a path to healing authority. The same person may move from the first state to the second within a single social career. Sociologists and anthropologists have documented consistent patterns in who gets possessed and when: possession falls disproportionately on the marginalized — women in male-dominated societies, conquered peoples, low castes — and functions as a means of voicing claims and pressuring superiors that more direct channels do not permit. At the same time, possession stands at the center of some of the world’s most authoritative religious institutions, where it upholds public morality and enforces communal norms. The distance between these two functions — possession as quiet protest and possession as moral backbone of society — is one of the central tensions the anthropology of religion has tried to explain.

Definition and Scope

The English word “possession” covers a broad range of phenomena that cultures themselves distinguish. Lewis’s working definition is deliberately cultural rather than physiological: if someone is generally considered in their own community to be in a state of spirit possession, they are possessed.(Lewis, I. M., 2003) This matters because trance — the altered physiological state commonly associated with possession — does not always coincide with cultural attribution of possession, and possession may be attributed to people who show no observable dissociation at all. Trance and possession are separable phenomena.(Lewis, I. M., 2003)

Trance itself is not exotic. Most normal people can be brought into altered states through a range of techniques — fasting, dancing, rhythmic music, hyperventilation, drug ingestion, sensory deprivation, or hypnotic suggestion.(Lewis, I. M., 2003) Endorphins, naturally produced in the human brain, are released by many traditional trance-induction methods including prolonged dancing, providing a direct physiological basis for both analgesia and euphoria without any chemical agent.(Lewis, I. M., 2003) What cultures do with these states — the meanings they assign, the social roles they create, the specialists they authorize — varies independently of the underlying physiology.

The cross-cultural record is extensive. From the Tungus of Siberia (from whose language the word “shaman” derives) to the Somali of East Africa, from Haitian voodoo to the Islamic zar cults of Ethiopia and the Sudan, from the Hausa bori spirit pantheon of Nigeria and Niger to medieval Italian tarantism, possession belief systems have been documented in nearly every region of the world across recorded history.(Lewis, I. M., 2003)(Lewis, I. M., 2003)(Lewis, I. M., 2003) Haitian voodoo is a particularly striking case: practiced by over ninety percent of the island’s population, it has functioned as an unofficial religion of the peasantry and a vehicle of nationalist resistance, with the loa spirits drawing heavily on West African cosmological heritage while the small Westernized elite clung to Catholic orthodoxy.(Lewis, I. M., 2003)

Most earlier scholars who addressed ecstatic religion, including historians, theologians, and psychologists, were absorbed in the theatrical elements of possession or preoccupied with judging its authenticity, and so failed to ask the sociological questions about its social conditions that would prove most illuminating.(Lewis, I. M., 2003)

The Peripheral and Central Distinction

The most influential analytical framework for the comparative study of possession is I. M. Lewis’s distinction between peripheral and central possession cults, introduced in his 1971 monograph Ecstatic Religion and developed through subsequent editions.

Peripheral possession involves spirits that are morally indifferent — they strike capriciously and without reference to the victim’s moral conduct.(Lewis, I. M., 2003) This amorality is not incidental but structurally important: it renders the possessed person entirely blameless for their condition and for the demands the possessing spirit voices on their behalf. Peripheral cults tend to recruit from social categories that lack direct access to power: women in patriarchal societies, ex-slaves, conquered peoples, low castes.(Lewis, I. M., 2003)(Lewis, I. M., 2003) The possessing spirit functions as a spokesman for claims that the afflicted person could not advance under their own authority.

Central possession religions, by contrast, involve spirits that are explicitly moralistic — they intervene in human affairs to punish wrongdoing, uphold communal norms, and adjudicate disputes.(Lewis, I. M., 2003) Shamans in these systems are not marginal figures but religious authorities in the fullest sense, presiding over confession, law, and the management of the community’s relationship with the natural and supernatural world. The Akawaio of Guyana, for example, use the shamanic séance simultaneously as healing session, law court, and public confessional.(Lewis, I. M., 2003) This authority to identify wrongdoers and exact confession belongs not to the shaman personally but to the spirits speaking through him, so the social control function is accomplished without the shaman assuming personal secular power.(Lewis, I. M., 2003) Among the Eskimo, the shaman manages what Lewis calls the “cosmic economy” — maintaining the goodwill of nature spirits through rituals of public confession after violations of hunting and sexual taboos, with human survival directly at stake.(Lewis, I. M., 2003)

The sociological evidence also shows that central possession cults persist most tenaciously in small, fluid social units exposed to exacting physical conditions or alien oppression, suggesting that ecstasy functions as a heroic communal response to overwhelming environmental pressure.(Lewis, I. M., 2003) The central morality possession religion, more generally, reappears in altered form whenever a small community faces severe environmental or political pressure that overwhelms secular coping mechanisms; the comparative record confirms this elective affinity between ecstasy and extreme collective stress repeatedly.(Lewis, I. M., 2003)

Peripheral cults do not merely complement central ones; they can also serve as transitional stages toward them. The Giriama “therapeutic Muslims” of the Kenya coast illustrate this process: Islamic spirits cause possession illness in non-Muslims, and cure is offered through Islamic religious incorporation, making spirit illness the entry point for conversion.(Lewis, I. M., 2003) The Vedda of Ceylon show the two cult types operating simultaneously within a single society: male shamans possessed by clan ancestor spirits uphold moral norms and hunting fertility at the center, while women maintain a separate peripheral spirit cult operating alongside.(Lewis, I. M., 2003) Among the Macha Oromo of Ethiopia, a social crisis in which traditional hereditary priests lost authority produced a rapid explosion of new achieved possession shamanism, demonstrating that central possession can arise quickly wherever a vacancy in moral authority opens.(Lewis, I. M., 2003)

Lewis presents this not as a binary but as a continuum.(Lewis, I. M., 2003) The same ritual apparatus can appear peripheral from one social position and central from another. And the relationship between the two types is historically dynamic: new prophetic movements often begin with relatively open access for women and subordinate groups, but as they become institutionalized, women’s participation is pushed toward the periphery and redefined as deviant or dangerous.(Lewis, I. M., 2003) The Montanist controversy in early Christianity illustrates this compression with particular clarity. More generally, established religions tend to suppress ecstatic phenomena once securely institutionalized, because the inspired individual who claims direct access to divine knowledge threatens ecclesiastical hierarchy.(Lewis, I. M., 2003)

Gender and Possession Cults

The relationship between gender and possession is one of the most thoroughly documented patterns in the anthropology of religion. Women in male-dominated societies are consistently overrepresented in peripheral possession cults, and the social logic of this pattern is not difficult to identify once attention turns from the theatrical elements of possession to the social conditions surrounding it.(Lewis, I. M., 2003)

Lewis’s fieldwork in Somalia documented the pattern in close detail. The prime targets of sar spirit possession among the Somali are married women, particularly those subject to neglect, polygynous conflict, or the threat of easy divorce in a society where men hold virtually all the formal instruments of social power.(Lewis, I. M., 2003) The possessing spirits demand costly gifts and ceremonies from the husband — luxury goods, fine clothing, special food. Crucially, these demands are voiced by the spirit, not by the woman herself; it is therefore possible for the husband to yield without formally deferring to his wife.(Lewis, I. M., 2003) The ideological face of male dominance is preserved while material concessions are extracted.

The system operates through structured ambiguity. Men generally acknowledge that sar spirits exist (the Quran provides scriptural warrant, assimilating them to jinn) while being skeptical that their own wives are genuinely afflicted rather than malingering.(Lewis, I. M., 2003) Women, for their part, have ready explanations for why the wealthy are attacked more often than the poor. The mutual incomprehension is functional: it allows the system to produce real material outcomes without either side having to acknowledge openly what is happening.

The zar/sar complex extends across a wide geographic arc — from Ethiopia (where the name zar appears to have originated) through Sudan, Egypt, North Africa, and into the Arabian Gulf.(Lewis, I. M., 2003) The Hausa bori spirit cult of Nigeria and Niger operates on parallel lines but with greater elaboration: a pantheon of roughly two hundred individually named spirits, each associated with specific symptom clusters, constituting a system that Lewis describes as both a religious institution and a medical dictionary.(Lewis, I. M., 2003) Janice Boddy’s feminist reading of the Sudanese zar holds that the spirits actively take responsibility for disrupted fertility — preventing pregnancies, causing stillbirths — thereby lifting from women’s shoulders a burden of reproductive blame that their socialization gives them no other way to manage.(Lewis, I. M., 2003)

Czaplicka’s Siberian records confirm that the female shamanistic vocation surges reliably during wars, famines, and epidemics, supporting the view that acute social stress is itself a generative condition for women’s entry into the shamanic role.(Lewis, I. M., 2003)

The pattern is not confined to Africa or to Islam. Lewis documents comparable structures across South Asia, Southeast Asia, and Japan, all involving women in structurally similar situations of subordination.(Lewis, I. M., 2003) Italian tarantism, though interpreted locally as spider-bite poisoning rather than demonic possession,(Lewis, I. M., 2003) attracted subordinate male peasants in earlier periods as well as women, supporting the broader claim that it is social marginality rather than female biology that explains possession’s demographic distribution.(Lewis, I. M., 2003)

Possession as Protest and Its Limits

The protest reading of peripheral possession must be understood with some precision. Lewis does not argue that possessed women are consciously manipulating the system, only that the system regularly produces outcomes that serve the interests of the possessed and that the structural conditions consistently predict who will be possessed.(Lewis, I. M., 2003) Possession is a strategy in the sociological sense — a pattern of action that recurs because it works — not necessarily a conscious plan.

The reach of peripheral possession extends beyond women to all subordinate social categories. In North Africa and Ethiopia, ex-slaves and their descendants maintained bori possession brotherhoods that provided credit, lodging, legal protection, and welfare for their members, and sometimes purchased the freedom of enslaved members when the institution still existed.(Lewis, I. M., 2003) The ritual insignia are telling: when ex-slaves danced in Somali possession rites, they carried whips — presenting themselves as masters rather than slaves, a ritual inversion at the heart of the ceremony’s meaning.(Lewis, I. M., 2003) In Malabar, low-caste Nayar servants possessed by the spirits of dead dependants used the ritual occasion to voice grievances and act aggressively toward their social superiors, who were obligated in that context to receive them respectfully and give gifts.(Lewis, I. M., 2003)

Lewis distinguishes peripheral possession from witchcraft accusations as complementary social strategies with different social trajectories. Peripheral possession runs upward — subordinates press claims against superiors. Witchcraft accusations typically run horizontally between equals or downward from superior to subordinate.(Lewis, I. M., 2003) The rare exceptions to this directional rule — when an inferior explicitly accuses a superior of witchcraft — function as direct challenges to the legitimacy of the status difference itself.

The limits of possession-as-safety-valve are also visible in the historical record. The Rwanda kubandwa cult served as a vehicle for suppressed Hutu grievances against Tutsi dominance, and the existing accounts emphasize its role in ventilating aggression. But when the suppressed resentment reached its limit, the Hutu abandoned the cult and mounted a direct political revolution, massacring thousands of former masters.(Lewis, I. M., 2003) The capacity of ritual to absorb political energy has a threshold.

The Shamanic Complex and Possession

A long-running scholarly debate concerned whether shamanism and spirit possession were the same phenomenon or different ones. The position associated with Mircea Eliade held that shamanism in its pure form involved the shaman’s soul leaving the body to travel in other realms — soul flight — rather than the entry of an alien spirit into the shaman’s body, and that the two phenomena should be treated as historically and logically separate. Lewis rejects this distinction on the basis of the ethnographic record in shamanism’s own locus classicus.

From a sociological standpoint, shamanism is best defined by the shaman’s functional role as master of spirits, rather than by specific cosmological features such as celestial soul-flight; this framing makes shamanism a cross-cultural phenomenon rather than a culturally specific one.(Lewis, I. M., 2003) Among the Tungus of Siberia — the people from whose language the word “shaman” is derived — shamanism involves controlled spirit possession.(Lewis, I. M., 2003) The evidence from both Eskimo and Chukchee shamans similarly shows possession by spirits as a central feature. Eliade’s separation, Lewis argues, cannot be sustained for the very cases on which it is supposed to rest.

The Tungus evidence further shows how the shaman’s career begins not with controlled possession but with the opposite: a traumatic, unwanted seizure by a spirit, culturally understood as illness and involving stereotyped hysterical behavior.(Lewis, I. M., 2003) This pattern is not specific to the Tungus: across the shamanic traditions Lewis examined, initial possession is typically an unwelcome intrusion that the candidate actively resists, and the more dramatic the resistance, the more authority the eventual shaman commands.(Lewis, I. M., 2003) The path to becoming a shaman runs through mastering what initially mastered you. The shaman who persists through the initial crisis reaches a state in which possession can be turned on and off at will in séances — the possessed now “possesses” his spirits as well as being possessed by them. This paradoxical language reflects the reality of the shamanic condition: mutual control between shaman and spirit, each holding the other.

The collective dimension of the Tungus séance is also significant. Shirokogoroff’s account describes how rhythmic music and dancing gradually draw every participant toward the threshold of trance — the entire assembly approaches the shaman’s state, held back only by the shared belief that when the shaman is present, the spirit may enter only him.(Lewis, I. M., 2003) The shaman’s mastery contains and redirects communal ecstatic potential.

When possession moves from peripheral to central — from the afflicted woman’s strategy to the community’s paramount ritual institution — it can become embedded in formal political structures. The Korekore Shona of Zimbabwe maintained a shamanic hierarchy whose most powerful figures were the spirits of former chiefs, exercising moral authority and regulating rain and agricultural fertility.(Lewis, I. M., 2003) During the colonial period, the Rhodesian government suppressed this cult precisely because it functioned as a vehicle for anti-colonial resistance. After suppression, it survived in degraded form among Zezuru Shona women, maintaining ancestral presence in peripheral séances when direct male political action was unavailable — a central cult regressing toward the peripheral pattern under political pressure.(Lewis, I. M., 2003)

The most elaborated form of central possession religion Lewis documents is the Kaffa kingdom of Ethiopia, where divine kingship was supported by a hereditary paramount shaman — the Dochay — whose ritual authority mediated between the king, nature spirits, and ancestors, upholding the cosmic order of the realm. This institution persisted even after Islamic conquest of the region, embedded too deeply in local social structure to be dislodged by external religious change.(Lewis, I. M., 2003)

Psychiatry’s Encounter with Possession

The relationship between spirit possession and Western psychiatry has moved through three recognizable phases: initial pathologization, ethnographic correction, and theoretical synthesis.

Pathologization. Early psychiatric and ethnographic observers read the shaman’s initial crisis as evidence of mental illness — neurosis, psychosis, or “arctic hysteria.” Bogoras reported that Chukchee shamans were “almost hysterical” and their cunning resembled that of a lunatic. George Devereux went further, arguing that shamanic societies were in some sense “sick societies” in which only the neurotic could successfully introject community norms.(Lewis, I. M., 2003) Julian Silverman drew structural parallels between shamanism and schizophrenia. These accounts shared an assumption that behaviors departing from Western norms required a Western pathological category.

Correction. Direct fieldwork challenged this systematically. Shirokogoroff described Tungus shamans as psychologically functional within their own cultural context. S. F. Nadel’s comparative study of four African shamanic traditions was categorical: no shaman in everyday life is an abnormal individual, a neurotic, or a paranoiac — if he were, he would be classed as a lunatic, not respected as a priest. Neither epilepsy nor insanity were considered qualifications for the role; they were simply diseases.(Lewis, I. M., 2003) Jane Murphy reported that practicing Alaskan Eskimo shamans were found by direct assessment to be “unusually mentally healthy.”

Lewis frames these positions as not actually contradictory. The shaman does undergo a genuine initial crisis — the evidence is consistent on this point. But he masters it, and it is precisely this mastery that constitutes the credential for treating others. The distinction between the initial crisis and the achieved mastery is what earlier observers missed when they focused on the former and ignored the latter.

The wounded healer. Lewis’s synthesis turns on this concept. The shaman is not defined by his subjection to disorder but by his victory over it.(Lewis, I. M., 2003) The more severe the original affliction that is mastered, the greater the authority it conveys.(Lewis, I. M., 2003) Nadel’s analysis of four African traditions showed that what matters is not private visionary experience — which separates the exceptional person from the community — but public mastery of disorder and its beneficial deployment for the community. Mastery is demonstrated, not reported.

Lewis draws an explicit analogy to psychoanalysis: the training analysis required of psychoanalytic candidates, in which the future analyst must undergo the disorder they will treat, is the Western institutional analog to the shamanic initiation crisis.(Lewis, I. M., 2003) Both institutions hold that personal encounter with disorder is the precondition for therapeutic authority over it. The parallel is imperfect — Lewis is not claiming all shamans are self-healed neurotics — but it has enough structural force to redirect the comparative question.

Spirits as hypotheses. Shirokogoroff provided the most searching theoretical account of how possession cosmologies work as conceptual systems. He proposed that “spirits are hypotheses” — cultural constructs that formulate observations about psychological life and provide mechanisms for regulating it.(Lewis, I. M., 2003) On his view, the Tungus analysis of psychological states in spirit-symbols was not less precise than Western categories like “instincts” and “complexes” but differently precise — and in some respects more practically useful, since the spirit hypothesis provided a vocabulary for collective action around individual psychological disturbance. Western psychiatry, following this logic, is an alternative framework for managing the same phenomena that possession cosmologies manage through different means.(Lewis, I. M., 2003)

Lewis pushes Shirokogoroff’s argument in two further directions. First, he maps Freud and Breuer’s hysteria model onto peripheral possession: hysterical conversion reactions — partial repression finding oblique, covert expression for forbidden desires — are structurally the same mechanism that peripheral possession uses in the social register.(Lewis, I. M., 2003) Second, and most provocatively, he inverts the conventional hierarchy: psychiatry and psychoanalysis are not a more advanced version of what shamanism is a primitive version of. They are, rather, limited and imperfect forms of shamanism — limited because their domain of practice is narrower (psychogenic complaints only), and imperfect because they have lost the collective dimension that makes the shamanic séance more than individual therapy.(Lewis, I. M., 2003)

The scope of shamanic practice. The shaman does not limit himself to psychogenic complaints. Where Western psychiatry is defined by its domain — psychological disorder — the shaman treats organic disease, interpersonal conflict, environmental misfortune, and psychiatric disorder together, in a single integrated practice.(Lewis, I. M., 2003) The séance functions as “danced psychodrama and group therapy”: abreaction of repressed urges, collective license for normally suppressed desires, and communal support — with each participant ideally achieving trance and collapsing in a state from which they emerge, as Lewis puts it, “purged and refreshed.”(Lewis, I. M., 2003)

Lévi-Strauss, comparing shamanic cure to psychoanalysis, argued they were exact counterparts with inverted elements: the analyst listens while the patient constructs an individual myth from personal history; the shaman speaks and delivers a social myth from outside.(Lewis, I. M., 2003) Lewis finds this contrast contrived and at variance with the ethnographic record. The actual structure of shamanic séances shows considerably more variation — and more active patient participation — than Lévi-Strauss’s schema allows.

Cultural relativism and the limits of diagnosis. Lewis closes his psychiatric discussion with a pointed challenge to ethnocentric diagnosis. In societies where possession is a pervasive explanatory framework, disbelief in spirits would itself be the striking abnormality — a bizarre rejection of shared reality roughly equivalent to claiming possession in a Western secular setting.(Lewis, I. M., 2003) The refractory cases in possession systems — those who cannot enter into collective trance, who do not respond to the cult’s therapeutic conventions — are the exceptions that prove the rule. They are the genuinely disturbed, and they are recognized as such within the system itself.(Lewis, I. M., 2003)

The Catholic Church and the Residuum of Possession

The narrowing of Catholic possession doctrine tracks the history of medicine’s absorption of formerly demonological territory. In the medieval period, a wide range of hysterical and trance-related behaviors were unhesitatingly diagnosed as demonic possession. By the modern period, what remains is a small residuum of cases that Catholic psychiatrists find impossible to explain in naturalistic terms — all other cases are officially reclassified as “pseudo-possessions.”(Lewis, I. M., 2003) The category did not disappear; it contracted to whatever medicine had not yet absorbed. In early modern Europe, physicians and clerics both acknowledged two kinds of madness: one rooted in humoral disorder and amenable to medicine, another requiring spiritual intervention. As medicine expanded its explanatory reach, the spiritual category shrank.(Andrew Scull, 2015) Between 50,000 and 100,000 people were executed as witches across Europe in the early modern period — a scale of violence that becomes intelligible only against a background in which educated people across confessional lines believed possession and witchcraft were real, and in which skepticism about such beliefs was tantamount to atheism.(Andrew Scull, 2015)

Return to the West: Multiple Personality Disorder

Lewis’s 2003 preface to the third edition of Ecstatic Religion opens with a striking observation: the phenomenon he had spent the preceding decades documenting in non-Western societies had returned to the West in new dress. Multiple Personality Disorder (MPD), which became a significant diagnostic and therapeutic industry in American psychiatry during the 1980s and 1990s, displayed structural parallels to possession cosmologies that Lewis found impossible to miss.(Lewis, I. M., 2003) The “alters” in MPD function as possessing spirit entities; there are specialist diagnosticians and therapists analogous to cult practitioners; and the standard therapeutic account — childhood sexual abuse by a father — mirrors the social dynamics of possession witchcraft accusations, in which a possessed victim imputes their illness to the malevolent acts of a kin figure.(Lewis, I. M., 2003) Lewis does not claim MPD is not real, but he does argue that Western secular culture had generated, in a medical idiom, a framework structurally equivalent to what it had displaced.

This cross-cultural persistence of possession ideology underscores what Lewis takes to be the core point: the forms possession takes are culturally variable, but the social and psychological dynamics that generate them are not. Whether the idiom is zar spirits, loa, bori, arctic séance, or the therapeutic language of dissociative identity, the underlying patterns — initial crisis, achieved mastery, communal ritual management of psychological disturbance, and the pressing of marginalized claims through oblique channels — remain remarkably consistent. Lewis also observes a structural dynamic common to all these cases: possession typically begins as traumatic illness, and where initial exorcism fails, the response shifts to accommodation and domestication of the spirit through adorcism — what begins in agony ends in ecstasy.(Lewis, I. M., 2003) Paradoxically, repeated exorcistic rituals in patriarchal religious contexts can covertly function as ecstatic cults in their own right, making the exorcism/adorcism distinction less absolute in practice than in theory.(Lewis, I. M., 2003)

Judeo-Christian and Islamic Possession Traditions

Michael Dols’s study of the madman in medieval Islamic society (Majnūn, 1992) situates Islamic possession belief within a longer Judeo-Christian framework that directly shaped it.

Biblical and Early Christian Roots

The Old Testament presents spirit possession as an instrument of divine action. Saul’s mental disturbance was produced by an evil spirit (ruach) sent by God as punishment for his disobedience; music was employed as therapy, with David playing the lyre so that “Saul was refreshed, and was well, and the evil spirit departed from him.”(Dols, Michael W., 1992) Hebrew scripture used three terms related to madness: the general term for raving (also used for prophesying); shiggayon (impulsive expression or madness); and meshuggaʽ (madman), which was also applied to prophets. A major criterion for judging insanity in ancient Israel was the occurrence of “impulsive, uncontrolled and unreasonable behaviour” requiring physical restraint.(Dols, Michael W., 1992)

The association of demons with mental disturbance deepened through Persian influence on Judaism in the fifth and fourth centuries BC. Demons became omnipresent, intimate, and powerful — not confined to “an underworld of superstition and hocus-pocus that was believed in by only the gullible and ignorant lower class” but rather “an integral part of everyday life in the Middle East.”(Dols, Michael W., 1992) Jesus of Nazareth’s exorcisms were central to early Christianity, and in the contemporary Jewish view “to heal, to expel demons, and to forgive sins were interchangeable synonyms.”(Dols, Michael W., 1992) The demonic interpretation of mental derangement became decisive in Christian theology through Origen’s view of lunacy as intrusive possession. Byzantine saints’ exorcisms differed from Jesus’ technique in that they “usually touched the demon-possessed individual, used invocations, and were usually remunerated in one way or another” — making them functionally closer to magical practice.(Dols, Michael W., 1992)

One of the earliest Christian institutions specifically for the mentally disturbed was the hesychasterion (place of quiet) established by St Theodosius near Jerusalem in the early sixth century, for monks who had lost their psychic well-being through excessive undisciplined asceticism. Treatment was a supervised communal life and spiritual exhortation — not medicine or exorcism.(Dols, Michael W., 1992) The Christian holy fool tradition (salos) provides a further complication: Simeon Salos in sixth-century Emesa deliberately feigned insanity — dragging dead dogs through the streets, pretending to be epileptic — “for the sake of Christ” while secretly leading an ascetic life. His behavior was indistinguishable from that of genuinely disturbed individuals until after his death.(Dols, Michael W., 1992)

The Islamic Jinn

The Quran is notable for what it does not contain: unlike the New Testament, it records nothing about miraculous healings or exorcisms by Muhammad, and the characteristic attribute of “healer” is absent from the ninety-nine names traditionally ascribed to God in Islam.(Dols, Michael W., 1992) Yet belief in spirit possession flourished. The jinn — created from smokeless flame, as opposed to mankind from clay and angels from light — became an integral part of the Islamic worldview. Medieval Muslim jurists worked out the legal status of jinn in astonishing detail, including the rules governing marriage between jinn and human beings.(Dols, Michael W., 1992) In popular belief, jinn lived in deserts, cemeteries, public baths, abandoned places, and dark rooms; they assumed the forms of animals, particularly dogs and snakes; and “most characteristically, the demons drove men mad.”(Dols, Michael W., 1992)

The term majnūn (“the possessed” or “madman”) was itself ambiguous in pre-Islamic Arabia: it could indicate a prophet, a diviner, a poet inspired by jinn, or a genuine madman. The term appears eleven times in the Quran, always in Meccan passages, always in accusations made against Muhammad by his opponents — and always in the context of the question of the source of revelation.(Dols, Michael W., 1992) Muhammad himself reportedly dreaded being thought a poet or the possessed, and the Meccan leaders described possession signs as “choking, spasmodic movements and whispering.”(Dols, Michael W., 1992)

The cult of Muslim saints developed under the influence of Christianity and pre-Islamic practices in conquered territories; Goldziher’s observation that “in no other field has the original doctrine of Islam subordinated itself in such a degree to the needs of its confessors… as in the veneration of saints” reflects the depth of this accommodation.(Dols, Michael W., 1992) Shared healing across religious boundaries was common: Muslims, Christians, and Jews visited each other’s sanctuaries, venerated each other’s saints, and wore each other’s amulets, producing what Dols calls a “dual spiritual economy” in which popular saint veneration was complementary, not antithetical, to official Islamic doctrine.(Dols, Michael W., 1992) Al-Khaḍder (St George) “enjoyed by far the greatest reputation of healing the mentally deranged” across Palestine, honored by Muslims, Christians, and Jews alike; at his shrine, the mentally disturbed were chained in the narthex and given bread and water while the priest attempted to drive out the devil by beating.(Dols, Michael W., 1992) Sufi tekkes (lodges) in Anatolia also provided institutional care for the insane; the most famous admission criterion required that patients show agitation — quiet individuals and idiots were refused — and treatment lasted forty days in conditions of strict diet and solitary confinement.(Dols, Michael W., 1992)

Islamic Magic and Exorcism

Therapeutic magic was sanctioned by hadith: there was no harm in magical incantations employed for healing as long as they were not polytheistic, and the Arabic word for medicine (ṭibb) often signified magic in the medieval period.(Dols, Michael W., 1992) Medieval Islamic jurisprudence distinguished two categories of practitioner: licit exorcists (al-muʽazzimūn) who invoked God’s names and operated by obedience to God, and illicit sorcerers (as-saḥara) who enslaved devils through forbidden acts — with illicit black magic believed to be centered in Egypt.(Dols, Michael W., 1992) Solomon was the legendary founder of licit Islamic magic, said to be the first to subjugate the jinn to his will; illicit magic was attributed to the offspring of Iblīs (Satan).(Dols, Michael W., 1992)

The classic account of Islamic exorcism defines the muʽazzim (exorcist): he begins by beseeching God’s aid, speaks with the evil jinn, and demands that the jinn depart from the body of the possessed. The term ʽazama (to decide, to invite, to enchant) captures how the meaning evolved from a statement of firm resolve, through a magical entreaty, to a theurgic demand that God do one’s will.(Dols, Michael W., 1992) Ḥājjī Khalīfa, the seventeenth-century Ottoman historian, held that licit exorcism was “possible and permitted by reason and [Islamic] law” but required perfect piety, complete abstinence, isolation from the world, and devotion to God.(Dols, Michael W., 1992)

The baraka (power) of Muslim saints was interpreted by Dols as essentially benign or sanctified magic — transmitted through physical contact via spittle, sweat, and semen. Ibn Khaldūn’s comparison of saints and sorcerers made explicit “their close kinship: the saint was more of a magician and the magician was more of a saint than is commonly acknowledged.”(Dols, Michael W., 1992) Islamic legal schools were divided on the reality of sorcery: three of the four founders accepted it as real while Abū Ḥanīfa denied it; all four unanimously prohibited it; three considered teaching it to be apostasy; and three said the sorcerer should be executed if he had killed by his magic.(Dols, Michael W., 1992) Jurists also showed ambivalence about magical exorcism of epileptics, with one scholar declaring: “God has only prohibited what is harmful, not what is useful. If it is possible for you to be useful to your brother, do it.”(Dols, Michael W., 1992)

The most famous Islamic exorcism attributed to companions of Muhammad involved the use of the Fātiḥa (the opening sura of the Quran) as an incantation twice a day for three days to heal a shackled madman; the Prophet retrospectively approved, declaring: “By my life, there are those who profited by a false incantation, while you have profited by a true incantation.”(Dols, Michael W., 1992) Caliph al-Muʽtaḍid summoned lunatics and exorcists to his palace in 897 to deal with a spectre: exorcists proposed casting a spell on one of the lunatics so that his jinn could be questioned about the spectre’s identity. The episode illustrates, as Dols notes, that belief in jinn and demonic possession was common to all strata of Islamic society.(Dols, Michael W., 1992)

Adorcism and the Zār

The logic of Islamic possession cults mirrors what Lewis documented in sub-Saharan Africa: in Morocco, when initial exorcism fails, the therapeutic approach shifts to accommodation of the jinn rather than their expulsion. In the language of Lewis’s framework, this is the shift from exorcism to adorcism.(Lewis, I. M., 2003) The verb saraʽ (“to overcome,” “to throw to the ground”) means both to exorcize and to throw down — indicating the close relationship between epilepsy (characterized by falling) and possession, which shares this sign. “To exorcise the possessed individual had the sense, then, of throwing down the jinn.”(Dols, Michael W., 1992)

The Ḥamadsha of Morocco — a religious brotherhood whose members slash their heads during trance dances to please their possessing jinn — exemplifies this accommodation logic. The ḥadra (ritual dance) aims to satisfy or appease the jinn, not expel them permanently; it creates a symbiotic relationship between patient and jinn rather than a resolution.(Dols, Michael W., 1992) The zār ceremony of Egypt, designed to heal those possessed by spirits causing psychic and physical disabilities, operates on the same principle: “Resorting to the zār, after all other healing methods have failed, is in a sense ‘an acknowledgement that the demons have won; the whole tone of the ceremony is one of propitiation and persuasion rather than coercion’.”(Dols, Michael W., 1992)

Moroccan taxonomy distinguishes two modes of jinn attack: the jinn may strike (madrub) its victim, producing sudden physical disability, or inhabit (maskun) the victim, producing possession signs including falling unconscious, convulsions, and speaking in tongues. Majnūn generally means anyone attacked by jinn but specifically the possessed.(Dols, Michael W., 1992)

(Dols, Michael W., 1992): Saul agreed, and David was brought to court in order to perform this service… ‘And whenever the evil spirit from God was upon Saul, David took the lyre and played it with his hand; so Saul was refreshed, and was well, and the evil spirit departed from him.’

(Dols, Michael W., 1992): It, therefore, appears that a major criterion for judging insanity in ancient Israel was the occurrence of ‘impulsive, uncontrolled and unreasonable behaviour’ and that physical restraint was common.

(Dols, Michael W., 1992): as a result of Persian influence on Judaism in the fifth and fourth centuries BC, the notion that the Devil was responsible for all moral and physical evil penetrated deeply into Jewish religious thought… They were not inhabitants of an underworld of superstition and hocus-pocus that was believed in by only the gullible and ignorant lower class; the spirits had become an integral part of everyday life in the Middle East.

(Dols, Michael W., 1992): the exorcisms of Jesus do not follow the usual Jewish pattern: there is no ritual… there is only a verbal command to the evil spirits to depart… in his time, ’ “to heal”, “to expel demons”, and “to forgive sins” were interchangeable synonyms’. This view of healing is echoed in the later Talmudic saying: ‘No sick man shall recover from his illness until all his sins have been pardoned.’

(Dols, Michael W., 1992): Mental derangement especially was clearly believed by the early Christians to be demonic possession, and Origen’s view of lunacy as intrusive possession appears to have been decisive for Christian theology… The exorcisms of the Byzantine saints… differ from Jesus’ in that the saints usually touched the demon-possessed individual, used invocations, and were usually remunerated in one way or another.

(Dols, Michael W., 1992): St Theodosius ‘established for them a place of quietude, so that their hesychasterion [place of quiet] was in the monastery as a second monastery for the crowd who were received’. And they were supplied with everything that they needed… When the inmates were in possession of their faculties, however, St Theodosius encouraged them, urging them to be patient with the evils of this world.

(Dols, Michael W., 1992): On a dung-heap outside the city he found a dead dog. He took off the cord belt that he wore and tied it to the dog’s foot. Dragging it behind him, he ran through the city gates… When the new moon appeared he stared at the sky and fell to the ground, trembling like an epileptic… Some thought that Simeon was an impostor, others that he was a saint, and still others that he was insane.

(Dols, Michael W., 1992): Unlike the New Testament, the Qurʼān says nothing about miraculous healings and exorcisms, except, ironically, for relating Jesus’ healing of the blind and leprous… Muḥammad was not a miracle-worker… it is remarkable, yet consistent, that the characteristic of ‘healer’ is not one of the ninety-nine names or attributes that are usually ascribed to God in Islam.

(Dols, Michael W., 1992): The jinn appear frequently in the early or Meccan portions of the Qurʼān… The Qurʼān says that the jinn were created of smokeless flame… The jinn, both good and bad, became an integral part of the Islamic world-view. Even the legal status of the jinn was worked out by medieval Muslim jurists in all respects and in astonishing detail, especially with regard to marriage between jinn and human beings.

(Dols, Michael W., 1992): They usually lived in deserts, cemeteries, village guest-houses, public baths and watery haunts, abandoned places or ruins, and dark rooms; they might assume the forms of animals, particularly dogs, snakes and other creeping things, or misshapen men… Most characteristically, the demons drove men mad.

(Dols, Michael W., 1992): the term majnūn, ‘the possessed’, was quite ambiguous… In the Qurʼān the term majnūn, always in the singular, functions sometimes as an adjective and sometimes as a noun. It occurs only in Meccan passages, only in the context of accusations made against Muḥammad by his opponents (and against earlier Messengers by their enemies), and only in contexts involving the question of the source and process of revelation.

(Dols, Michael W., 1992): ‘Now none of God’s creatures was more hateful to me than an (ecstatic) poet or a man possessed [majnūnin]: I could not even look at them. I thought, Woe is me poet or possessed—Never shall Quraysh say this of me! I will go to the top of the mountain and throw myself down that I may kill myself and gain rest.’

(Dols, Michael W., 1992): Ignaz Goldziher has asserted in his seminal study of the cult of the saints: ‘In no other field has the original doctrine of Islam subordinated itself in such a degree to the needs of its confessors, who were Arabs only in a small minority, as in… the veneration of saints.’

(Dols, Michael W., 1992): As with doctors, there was specialization among saints, some of whom were considered particularly effective in curing mental illnesses. Al-Khaḍder or St George enjoyed by far the greatest reputation of healing the mentally deranged… when the mentally disturbed were brought to this church, the priest would chain them in the narthex in front of the church.

(Dols, Michael W., 1992): Down to the twentieth century, there has been in the Middle East an interchange of religious healing between Muslims, Christians, and Jews in the form of visitations to each other’s religious sanctuaries, venerating each other’s saints, and wearing each other’s amulets and talismans… there is a ‘dual spiritual economy’, in which the marginal form of Islam is complementary, not antithetical, to official doctrine.

(Dols, Michael W., 1992): The most famous tekke for treating the insane was probably that founded by Shaykh Karacaahmed… If the disturbed person raged and screamed, he was admitted; quiet individuals, idiots, and many epileptics were not admitted. The insane were admitted for forty days and treated in the following manner: the patient was taken to the tomb of Karacaahmed or to that of Shaykh Eschref.

(Dols, Michael W., 1992): magic was a means of forcing supernatural powers to fulfil a supplicant’s desire, especially for healing. The use of such therapeutic magic by Muslims was sanctioned by ḥadīth: there was no harm in magical incantations that were employed for healing as long as they were not polytheistic… the Arabic word for medicine, ṭibb, often signified magic in the medieval period.

(Dols, Michael W., 1992): The exorcists and magicians [al-muʽazzimūn was-saḥara] assert that the devils, jinn, and spirits [ash-shayāṭīn wal-jinn wal-arwāḥ] obey and serve them, being directed by their command and their prohibition. The exorcists, who pretend to observe the sacred laws, claim that this [power] is because of obedience to Allāh, may His name be magnified… The [other] magicians [as-saḥara] assert that they enslave the devils by offerings and prohibitive acts… This is common practice in Egypt and the nearby regions; the books which are written there are many and extant.

(Dols, Michael W., 1992): Licit magic, the ‘praiseworthy method’ (aṭ-ṭarīkā al-maḥmūda) in Islam is usually traced back to Solomon, and illicit magic (aṭ-ṭarīkā al-madhmūna) to Iblīs through his daughter or his son’s daughter, Baydakh… Solomon was said to be the first to subjugate the jinn to his will.

(Dols, Michael W., 1992): Ibn Khaldūn’s comparison of the saint and the magician brings out forcefully their close kinship: the saint was more of a magician and the magician was more of a saint than is commonly acknowledged. The power of the saint, baraka… is essentially benign or sanctified magic; this view of baraka, rather than equating it with Christian terms like ‘blessedness’ or ‘grace’, helps to explain many of its peculiar characteristics, such as its preservation and transmission through physical contact (via spittle, sweat, and semen).

(Dols, Michael W., 1992): three of the founders of the Islamic law schools believed that sorcery was real; the exception was Abū Ḥanīfa, who, like the Mutaʽzilites, denied its existence. None the less, sorcery was unanimously prohibited. According to some legalists, this included consulting a diviner, teaching oneself its methods, or teaching it to others. Three of the legal authorities considered learning or teaching sorcery to be apostasy.

(Dols, Michael W., 1992): There was some reservation, however, about those who exercised their talents on epileptics, claiming that they could conjure up the demons and that the demons obeyed them. Ibn Qudāma, a Ḥanbalite, said: ‘Our doctors consider them to be practitioners of sorcery, but one reports that Aḥmad ibn Ḥanbal suspended his judgement on this subject.’ Saʽīd ibn al-Mushaīb was questioned about the matter, and he declared: ‘God has only prohibited what is harmful, not what is useful. If it is possible for you to be useful to your brother, do it.’

(Dols, Michael W., 1992): Ḥājjī Khalīfa says that exorcism, ʽazāʼim, is taken from al-ʽazm, i.e. resoluteness of opinion; thus, it is a command with a clear intention that is obligatory on others, so that one says ʽazamtu ʽalayka, ‘I adjure you’ or ‘I enjoin you.’… This [activity] is possible and permitted by reason and [Islamic] law. Whoever denies these two should not be listened to because what he says leads to the denial of God’s omnipotence.

(Dols, Michael W., 1992): The convert said that he did, and he used the Fātiḥa, or the opening verse of the Qur’ān, as an incantation twice a day for three days. The shackled man recovered… The Prophet then said that he should, in the name of God, take the sheep, and he declared: ‘By my life, there are those who profited by a false incantation [ruqya], while you have profited by a true incantation.’

(Dols, Michael W., 1992): On Saturday, Ramaḍan 7, 284 [Saturday, 8 October 897], lunatics and exorcists [al-majānīn wal-muʽazzimūn] were assembled and brought to al-Muʽtaḍid’s Thurayyā palace, because of the spectre that was appearing to him… one of them mentioned that they might cast a spell [yuʽazzimūna] upon one of the lunatics and, when he fell down, [he might] ask the jinnī what the spectre was.

(Dols, Michael W., 1992): the muʽazzimūn should be translated as ‘exorcists’, from the Arabic root ‘azama, which may mean ‘to decide’, ‘to invite’, or ‘to enchant’. An exorcist usually begins by beseeching God’s aid, speaks with the evil jinn, and, then, demands that the jinn depart from the body of the possessed… the muʽazzim was a particular kind of magician; while the common intention of magicians was to bend God’s will to their purposes, the exorcist also had the capacity to coerce the jinn to do his will, specifically by coming out of the possessed.

(Dols, Michael W., 1992): one can see the close relationship between ‘epilepsy’, characterized by falling to the ground, and insanity, which shares this sign; to exorcise the possessed individual had the sense, then, of throwing down the jinn. Crapanzano also observed in modern Morocco that ‘the verb saraʽ, to overcome, to throw to the ground, means also to exorcise. Its basic meaning is indicative of the permanent explusion of the jinn in exorcism.’

(Dols, Michael W., 1992): The confraternity is particularly well known as a healing cult, in which the members slash their heads during their trance dances; the self-mutilation is believed to be pleasing to their possessing jinn… The aim of the ḥadra or ritual dance of the Ḥamadsha is to satisfy or appease the jinn — a ‘cure’ of the devil-struck and the devil-possessed. But only a temporary cure, for the ḥadra creates a symbiotic relationship, to the obvious advantage of the cult.

(Dols, Michael W., 1992): The zār is also designed to heal men and especially women who are believed to be possessed by spirits, which cause psychic and physical disabilities. The zār ceremony appears to have originated in Ethiopia and to have been diffused outward to western and southern Arabia, the Sudan, and Egypt… Resorting to the zār, after all other healing methods have failed, is in a sense ‘an acknowledgement that the demons have won; the whole tone of the ceremony is one of propitiation and persuasion rather than coercion’.

(Dols, Michael W., 1992): ‘The person who exhibits the symptoms of a possession state is said to be “inhabited” (maskun) by a jinn, whereas the person who suddenly is paralyzed is said to have been struck (madrub) by a jinn… in only one instance is a category of the jnun-produced illnesses symptom-specific: matrush. The word matrush refers to a sudden, unilateral paralysis of the face, usually on the left side. It is believed to be the result of a slap in the face by a jinn.’ … Majnūn generally means anyone attacked by jinn but specifically the possessed.

Korean Shamanism and the Manshin

Kendall’s ethnography of a Korean manshin (shaman) provides a close-grained account of how possession operates within a single biographical trajectory. In Korean shamanic practice, the manshin provides divination by entering trance to determine the supernatural source of a client’s trouble, including illness or bad luck attributed to hovering ghosts.(Kendall, Laurel, 1988) The most elaborate ritual is the kut, in which gods and ancestors appear and possess costumed shamans, venting grievances, providing divinations, and bestowing blessings on family members.(Kendall, Laurel, 1988)

Restless ancestors with unfulfilled desires — a concept the Korean term han captures — are understood as especially dangerous. Grandparents who did not live to see grandchildren, first wives superseded by second wives, young men and women who died before they could marry: all carry han that can afflict the living.(Kendall, Laurel, 1988) Persistent illness or a cluster of simultaneous misfortunes signals deeper household malaise: “the ancestors are hungry and the gods want to play,” requiring a full kut.(Kendall, Laurel, 1988)

Divine guidance during crisis forms a recurring motif in Yongsu’s Mother’s narrative. During her captivity by the People’s Army, a white-haired grandfather with a long beard and twisted staff appeared to her in a dream, warning her to escape: the Mountain God in a form consistent with standard iconography of that spirit.(Kendall, Laurel, 1988) During a kut performed for her illness, she dreamed that a white-haired grandmother and grandfather massaged her and offered water from a gourd dipper, saying “if you drink this, you’ll survive”: a visitation she understood as the receipt of medicinal water (yaksu) from divine beings.(Kendall, Laurel, 1988)

The path to becoming a manshin follows the pattern Lewis documented cross-culturally. Yongsu’s Mother — the shaman whose life Kendall recorded — experienced compulsive pilgrimages to sacred sites, hallucinations, and illness-then-healing cycles before her calling was acknowledged.(Kendall, Laurel, 1988) Her divine possession occurred spontaneously while dancing in the Mountain God’s costume at another shaman’s kut; she seized the Spirit Warrior’s flags and began shouting divine pronouncements before running home, her heart “thumping wildly,” feeling she “just wanted to die like a crazy woman.”(Kendall, Laurel, 1988) The pattern is strikingly consistent with what Lewis described among the Tungus and others: involuntary seizure, resistance, and eventual mastery.

Precognitive dreams in Yongsu’s Mother’s circle also marked her shamanic destiny before it was socially acknowledged. Her sister-in-law, exhausted on the wedding night, dreamed that Yongsu’s Mother was sitting on the floor beating an hourglass drum with the gods’ costumes overhead: a vision that retrospectively confirmed the shamanic calling that had not yet been formally recognized.(Kendall, Laurel, 1988)

Korean shamanism illustrates how medical pluralism operates in practice. Yongsu’s Mother’s postpartum illness was diagnosed simultaneously by biomedical practitioners (who suggested appendicitis or retained blood) and by shamans (who identified “ghost play” — kwisin norum). Multiple doctors offered contradictory diagnoses; the shamanic diagnosis provided a framework for action when biomedical uncertainty left the family without direction.(Kendall, Laurel, 1988)

Rivers and the Comparative Framework

W. H. R. Rivers, writing from his fieldwork in Melanesia and Polynesia, provided an early comparative framework for understanding the relationship between possession beliefs and medical practice. He classified non-Western disease causation into three categories: human agency (sorcery), spiritual agency (ghosts, gods), and natural causes. Where disease was attributed to soul-loss, the treatment was recovery of the lost soul-substance; where it was attributed to possession by an alien spirit, the treatment was exorcism or propitiation.(Rivers, W. H. R., 1924) Rivers argued that these therapeutic systems were not superstitious confusion but internally rational: once the etiological premises were accepted, the treatments followed logically.(Rivers, W. H. R., 1924)

Rivers’s contribution to understanding possession was to insist that practices Western observers dismissed as magic were operating according to the same logical structure as scientific medicine — diagnosis determines treatment — with different premises about what causes disease. He also recognized, decades before Lewis, that the earliest therapeutic agencies employed by humanity were those that worked through the mind, and that the conscious recognition that remedies act through mental processes was “one of the most recent acquirements of medicine.”(Rivers, W. H. R., 1924)

Human Notes

See Also

• shamanism
• exorcism
• zar-cult
• witchcraft
• cross-cultural-psychiatry
• wounded-healer
• trance
• folk-medicine
• african-traditional-medicine

Sources

All primary claims draw on:

• Lewis, I. M. (2003). Ecstatic Religion: A Study of Shamanism and Spirit Possession (3rd ed.). London: Routledge. [Source ID: lewis-ecstaticreligion-2003]
• Scull, A. (2015). Madness in Civilization. Princeton: Princeton University Press. [Source ID: scull-madnesscivilization-2015]