Demons Only Virgins Can See: Divination with Child-Mediums as a Medieval Type of Clerical Child Abuse

Abstract

This article examines the use of child-mediums in divination and magic as a specific medieval understanding of child abuse. Medieval authors believed that children were used in this way by learned men, particularly churchmen. They believed the practice was abusive, causing physiological and psychological harm. Many also thought, for different reasons, that it could produce revelations. This topic provided medieval authors with an opportunity to theorise about a specifically clerical form of child abuse, and it is an example of the kind of ritual magic extant in clerics’ own social worlds that fuelled paranoid conspiratorial fantasies, such as witchcraft.

Readers have long been drawn, often only in passing, to John of Salisbury's vivid account (1156/1159) of his childhood experience of being used (briefly) as a medium for magical divination:

During my boyhood I was placed under the direction of a priest, to teach me psalms. As he practiced the art of specularia magica [divination in reflective surfaces], it chanced that he after preliminary magical rites made use of me and a boy somewhat older, as we sat at his feet, for his sacrilegious art, in order that what he was seeking by means of finger nails moistened with some sort of sacred oil or chrism, or of the smooth polished surface of a basin, might be made manifest to him by information imparted by us. And so after pronouncing names which by the horror they inspired seemed to me, child though I was, to belong to demons, and after administering oaths of which, at God's instance, I know nothing, my companion asserted that he saw certain misty figures, but dimly, while I was so blind to all this that nothing appeared to me except the nails or basin and the other objects I had seen there before. As a consequence I was adjudged useless for such purposes …. So propitious was God to me even at that early age.¹

John explained that with time he became ever more horrified by the practice he had survived as a child. He knew of many men who practised divination of this kind, all of whom were eventually deprived of their sight and subject to other miseries he chose not to describe. His former teacher was one of only two exceptions, forsaking that art to become a canon. The other was a deacon, who became a monk at Cluny.² John's teacher had acted in accord with a commonly held belief that children made especially effective intermediaries with the other world because of their sexual purity. As Gervase of Tilbury described in the thirteenth century (1211/1215), there were ‘some demons which are only seen by virgins; for untainted flesh possesses greater spiritual vision. For this reason, necromancers claim that in the experiments of the sword, the mirror, the finger-nail, or the circle, only a virgin's eyes are effective’.³

Medieval authors, like Gervase and John, regarded the use of children in divination as part of a hazy cloud of potential dangers faced by children put under the power of older churchmen. Medieval schoolboys, often separated from their families and homes, were subject to the whims and proclivities of their teachers in many ways. The medieval education of boys frequently involved violence, and corporal punishment was a regular feature of any educational setting.⁴ The intensity of this violence depended on the teacher and was always traumatic to some extent.⁵ There was the potential too for other kinds of victimisation, including sexual abuse. Awareness of these dangers appears in medieval sources in scattered ways, erupting into full view only in exceptional circumstances.⁶ John's account of his experience as a grammar student references this dangerous period in the lifecycle of an educated medieval man. It also illustrates an awareness, albeit obliquely, of the long-term consequences childhood trauma presented in later life among his peers.

While generally neglected by modern historians of childhood, scholars interested in magic have explored the use of children as mediums and the significance of their perceived purity in divination. The only book-length treatment of divination in polished or reflective surfaces, or catoptromancy, remains Delatte (1932), but Richard Kieckhefer has placed the practice into the larger context of medieval magic.⁷ Kieckhefer and other scholars, such as Marie-Thérèse d'Alverny, have also published some of the surviving sources with excellent commentaries to guide readers seeking further materials.⁸ Clare Fanger has given the most thought to the importance of virginity in medieval theories of divination.⁹ Beyond the description of specific practices or types of magic, Fanger examines contemporary philosophical and liturgical understandings of virginity's effectivity, applying neglected and seemingly esoteric sources to issues of larger historical interest.

This study examines the intersections between the utility of children's supposed sexual purity in divination and the harm some medieval authors believed children suffered during divination. As Kieckhefer has noted, the ‘conventional wisdom’ of these authorities regarded ‘such dealings with young children as a form of abuse’.¹⁰ The adult practitioners of this magic, which required literacy and an acquaintance with liturgy, were overwhelmingly churchmen.¹¹ There is ample evidence to demonstrate that the belief in child-divination's efficacy led to real acts done to prepubescent children, between the ages of seven and the early teens, a range in which boys were subject to learned men for education. Historical understandings of the magical power of children's sexual purity along with the harm utilising that power could inflict on them represent a compelling medieval understanding of clerical child abuse.

Studying the harm some medieval writers believed contemporary magical practices did to children is difficult for the very reasons it is timely. The topic intersects with modern concerns regarding the maltreatment of children. Some of these concerns are factual and others are the constructs of conspiratorial imagination. Child-divination is a topic in which paranoid fantasies collide with realities. Clerical child abuse is a reality of great concern in the late twentieth and early twenty-first centuries. Scholars, such as Dyan Elliott, have looked to the medieval past to trace the historical abuse of children and clerical responses to it.¹² In this conversation, child divination, as a widely known clerical practice that relied on the sexual attributes of children, deserves consideration.

In the realm of conspiratorial delusion, divination through child mediums is perilously close to the tropes of the recurrent fantasy that conspiratorial groups secretly gather to abuse and murder children.¹³ Such a delusion has inspired centuries of persecutions against Christians, supposed heretics, Jews and ‘witches’, and that to this very day encourages conspiracy theories which maintain (among many claims) that secret groups ritually abuse and often murder children.¹⁴ It may also appear sometimes intertwined with hateful and fictious stereotypes, such as the libellous and imaginary ritual abuse and murder of Christian children by Jews.¹⁵

The use of children for divination, as an intersection between fantasy and reality, gives us an example of the kind of elite ritual magic extant in clerics’ own social worlds that fuelled the fantasies they projected onto others.¹⁶ These fantasies, such as late medieval witchcraft, drew from the well of clerical magic that was a feature of learned culture. Children's mediumship also provided medieval authors with an opportunity to theorise about a specifically clerical form of child abuse and its effects on individuals. Concern with such magic is an underappreciated part of medieval discussions surrounding children's vulnerability and special natures.¹⁷ Theories of children's mediumship are part of the discursive history of child abuse as well as histories of magic.

The study will begin with an examination of the classical and medieval contexts for the use of child mediums in divination. Led by this history and its contextual assumptions, the argument will next examine medieval theories for the effectivity of divination using child-mediums and the specific kinds of psychological and physiological harm divinatory methods were believed to inflict. The final section turns from theory to practice, examining historical evidence for the employment of child-mediums, including surviving instructional texts of learned medieval magic.

Child-mediums in divinatory magic: classical and medieval contexts

Medieval Latin Europe inherited concepts of divination through child-mediums from the classical world. As Daniel Ogden has explored, child-mediums were a regular feature of Greek and Roman divination. They were used particularly for necromancy, as intermediaries for conversation with the dead. Virginal boys would stare into reflective surfaces, dancing flames or mirrors and describe what they saw to adult questioners.¹⁸ Children made good mediums because they were somehow closer to the other world. One enduring logic was connected to Plato's idea of the soul's recollection of a higher world. In the Meno, for example, Socrates explains the concept of recollection by questioning a boy, suggesting that such memories might be clearly manifested through the young whose spirits had seen the higher world most recently.

In the Roman tradition of child-divination, there was also a well-established notion of potential harm to the child.¹⁹ Stories involved or alluded to the sacrifice of children and the torture of children. A child sacrifice could work through different logics according to different writers, including hieroscopic sacrifice, the turning of the boy into a loquacious ghost, or the offering of the boy as a sacrifice to other pre-existing ghosts. Sources in antiquity attributed sacrifices of this kind to a range of characters.²⁰ Beyond divination, the idea that the torture or murder of children could function as a route to power for adults appears in contexts such as the torture of a child to make a love potion by the witch Canidia in Horace's Epode 5.²¹

Divination through child-mediums was linked to classical models of pederasty, and the practice remained sexualised in the Middle Ages. Classical boy-mediums often conformed to eroticised ideals, needing to be beautiful or perfect in body and virgins to heterosexual intercourse (specifically).²² Medieval authors, in their turn, understood the sexual implications of the practice in complex ways, attributing it to a particular period of childhood, after the beginning of education but before the onset of puberty. Perhaps most troublingly, medieval discourses on children's mediumship continued to explore the utility of a child's sexual identity and how it could be used by adults to realise their desires. In other words, the use of child-mediums in the Middle Ages remained sexualised, but it was sexualised in distinctively medieval ways.

We can glimpse something of the transition of this kind of divination between the antique and medieval worlds in the pseudo-Clementine Homilies and Recognitions. In these early Christian novels, Simon Magus, opponent of the Apostle Peter, attempted anti-Christ, and founder of heresies, confesses to murdering a child to acquire knowledge.²³ The soul of this boy ‘unsullied and violently slain’ provided his killer with knowledge. As Simon explains to his dismayed disciples, ‘the soul of man holds the next place after God, when once it is set free from the darkness of this body’. Uninhibited by flesh, ‘it acquires prescience: wherefore it is invoked for necromancy'.²⁴ The boy's innocence and purity enhanced his disembodied perception beyond what an adult's soul could provide. In an exercise of subtle eroticism, Simon kept an image of the murdered boy in his bedroom clothed in purple.²⁵ In truth, as the Apostle Peter reveals, a demon is impersonating the dead boy.²⁶ Simon's murdered child-familiar became known to Latin Europe through the translation of the Recognitions made by Rufinus of Aquileia.

Especially in the Recognitions, the rationale Simon offers for the child's usefulness resembles familiar models of antiquity, but Peter's reinterpretation of it anticipates what became the dominant medieval theories of magical power. Simon claims that the dead mediate between this world and higher things, reaching back to a pagan model of divination and magic that relied on the support of ambiguous supernatural assistants or familiar spirits, often called parhedroi, who were frequently associated with the spirits of the dead.²⁷ Peter rejects the entire paradigm of assumptions that sustained this pagan model of magic, insisting that all such mediating spirits were really demons, eager to deceive men by taking false appearances and playing to superstitious societal conventions. In Peter's reinterpretation of Simon's child-familiar as a demon, we find what became the medieval reinterpretation of necromancy itself: such magic communicates not with the dead but with angelic or demonic spirits. In this model, while children may not grant access to the dead, they could open conversations with devils.

In the medieval model of necromancy, children remained able to communicate with other worlds for much the same reason cited in the Recognitions: they were pure.²⁸ Their virginal purity offered a conduit to supernatural power accessible to others through the cults of child saints as well as magic.²⁹ This purity consisted of a particular group of attributes powerful enough to access heaven (Matthew xix.14 and xviii.3–4; Luke xviii.16–17), especially if marshalled by an adult capable of sin. The Gloss on Luke explained them: ‘a boy does not persevere in anger, he does not remember injury, he does not desire women, he does not think or speak evil, and so too for you, unless you have such innocence and purity of the soul, you will not enter into the kingdom of heaven’.³⁰ Central to this model of childhood purity was the assumption that a child's mind was less overshadowed by the darkness of the body. The notion endured as part of the vast range of Platonic, or neo-Platonic, ideas that remained intrinsic features of medieval Christianity.

As Wisdom ix.15 said, the ‘tainted body weighs down the soul’. Exactly how the flesh was corrupt and how it exerted a corrupting influence on the soul was variable, but the general point was not. Gender, of course, played a central role, and women were generally viewed as more tied to their bodies and susceptible both to external spiritual influences and to delusions than men. Men, supposedly more rational than women, were the most effective authenticators and interpreters of spiritual revelations.³¹ Yet, even among men, the flesh generally enervated the spirit. Thinkers often found explanations for the body's harmful influence in theories that relied on the assumption that an adult's attachment to worldly sensory experience blocks spiritual perception. For many twelfth- and thirteenth-century authors, it was through fallen affections that, as Ailred of Rievaulx explained, ‘the images of corporeal things obscure and suppress the intellect’.³² For Hugh of St Victor, the fallen soul clung to these images through misplaced love as a kind of spiritual corporeality. Even when freed from the body, a human soul took these affections with it, seeing in the afterlife the images of those worldly things it had loved in life.³³ It was in this way that a sinful soul could not perceive the spiritual world as it really was, perceiving instead an integument or covering that, at best, revealed the hidden things of God through a glass darkly.³⁴ The visionary literature of this period, likewise, frequently made the same assumption.³⁵ In the fourteenth century, Petrarch's inner Augustine explained to him how the images of visible things weigh down the soul and distract it from God.³⁶ His near-contemporary, Richard de Bury, citing the Phaedo, proclaimed that a philosopher's insight arose because he ‘frees the soul from communion with the body differently than other men'.³⁷ Through discourses like these, it was a basic mainstay of learned culture that worldly love inhibited spiritual insights.

Sexual desire often functioned as the prototypical worldly love. As it had for Augustine, it was in the domain of sexuality that the fallen nature of human affection articulated itself most forcefully.³⁸ It followed that the kind of attachment to the world that most inhibited spiritual insights was sexual desire. It was through this logic that a prepubescent child appeared as a superior mediator with spiritual things. Sexualised in this way, divination through child-mediums relied on the comparative innocence represented by a child's immature libido. As St Gilbert explained, the onset of puberty marked the turning point at which the corrupt body began most vehemently to oppress the soul.³⁹

Theories of effectivity and of abuse: William of Auvergne

William of Auvergne wove these threads together in his influential account of child divination. A product of the Parisian schools, canon of Notre Dame, Master of Theology and bishop of Paris until his death in 1249, William's work is both a gauge of the intellectual world in the generation before Aquinas and Bonaventure and an often-neglected influence on it and on subsequent centuries.⁴⁰ In the second part of his De universo creaturarum, dedicated to the spiritual universe, he explained the procedures, the cultural contexts and the effects of divination through child-mediums.⁴¹ In his discussion, William argued that child divination was a harmful practice that could theoretically yield actual revelations.

The procedures William cited are familiar, and some scholars have found his level of detail suspicious.⁴² Adult practitioners would ask children to stare into reflective surfaces, such as a mirror, the child's own nails, an egg, an ivory handle, a polished sword or a bowl. He explained that practitioners believed images would appear on these instruments (often aided by a coating of oil) in the imaginations of child-mediums.⁴³ These images would offer replies to adults’ questions regarding the revelation of ‘theft, or a robber, or something else dealing with hidden things’.⁴⁴ The instructions included in magical handbooks from the fifteenth (and later) centuries align with William's basic outlines almost completely.

Platonic recollection might explain the efficacy of child divination as historically practised.⁴⁵ In William's summary, Platonic recollection basically maintains that ‘no new knowledge arises in our souls through teaching, training, or experience’. Rather, there is only old, innate knowledge that is concealed ‘as if buried’. These inborn ideas are excavated from their mental tombs by training and experience, allowing us to perceive what ‘we did not see was with us before’.⁴⁶ In Plato's view, the joining of soul to body is like death or burial in which the spirit's greatest powers cease to work.⁴⁷ Focusing attention on a reflective surface is supposed to turn the foolish soul back on itself so that it will direct its attention into itself. Such an inversion should allow the soul to excavate better the hidden truths inside it. As an intellectual of his time, William rejected Platonic recollection, insisting that humans do acquire new knowledge through worldly sensory experience and instruction. He concluded that children's mediumship cannot function through innate ideas.

Mistaken Platonic doctrine has wrought crimes and absurdities. Prime among them was the murder of an innocent boy by Simon Magus. William recounted how ‘the most famous of the ancient mages among the Latins was driven insane by this error’. This magus murdered a flawless boy believing that the boy's soul would ‘possess knowledge of all present, past, and future things’. He was so deluded, William explained, that he thought the spell had worked and that he heard responses to his questions from the boy's soul. This crime presented an example of the kinds of demonic tricks most illicit magic involves. Surely something like this came from the ‘suggestion of an evil spirit’, and another evil spirit stepped in to provide the replies which the magus heard.⁴⁸

William paused to consider, briefly, how a magus might have a child's soul as a familiar, however unlikely it would seem, and in his digression reveals the impact of recently translated classical material on his thinking. Maybe, he wondered, as Plato suggests, the soul lingers around its body for a while during which time it can be contacted with magic.⁴⁹ William's wording is almost certainly a paraphrase of the Phaedo, available to Latin Europe after 1160.⁵⁰ He mused that it still would seem unclear why such a soul would converse with its murderer, unless it was such a perfect Christian soul that it truly did love its enemy.⁵¹ William left this line of inquiry there, choosing to refocus on larger issues in which the attributes of children were more revelatory.

While he did not accept many of the specific tenets of Platonic recollection, William accepted many notions associated with the divinatory practices he analysed. Divinatory ritual created an altered state. This state did not uncover knowledge buried in the soul but rather buried powers of the soul, opening it up to spiritual illuminations.⁵² Such illuminations appeared frequently as images in the mind, specifically the imagination which mediates between the corporeal and the incorporeal worlds.⁵³ Illuminations, natural, divine or demonic in origin, can be perceived more effectively by those whose minds are distanced from their senses. In other words, these practices, sometimes, worked.⁵⁴ For similar reasons, some of the ancients used to blind boys to make prophets, or even blinded themselves.⁵⁵ Acting on this logic, William admitted that in his youth he attempted unsuccessfully to free himself from attachment to the body in order to see divine things.⁵⁶ As Clare Fanger has argued, William here accepted some of the fundamental tenets and motivations of angel magic and similar esoteric practices.⁵⁷

It was the burial of the soul's greatest powers in sensuality that accounted for how rarely divination seemed to work. William claimed that most of the time there was scarcely one out of seven or even ten virgin boys who saw anything useful.⁵⁸ Souls, immersed in the world of matter, resist abstraction from the flesh. This resistance is so significant that the inspection of reflective surfaces yields divinations or revelations ‘partially and very rarely’.⁵⁹ Most boys were too closely tied to the flesh for the technique to work. It remained possible, none the less, because, unlike adults, children were innately less attached to sensuality in one prominent way.

Sexual desire was the greatest impediment to souls turning inward toward spiritual insights. William explained:

The teaching and the worship of Christians especially despises sexual desires, and the reason is because [such desires] principally take souls captive and bind them and immerse them in bodies. Moreover, they inebriate them, delude them, and turn them away from spiritual joys and indeed [spiritual] knowledge along with the contemplation of divine things.⁶⁰

Children possessed an advantage in the contest with sexual desires as the prototypical impulses that led adult men away from divine things. Virgin boys were used in divination because ‘they have not yet been seized or polluted by the vices, which are of adults’.⁶¹ William's linkage of sexual maturity and adult perceptions likely also broadly reflects a range of near-contemporary pronouncements identifying puberty with the full development of the human mind, the ability to sin and participate in the sacraments.⁶²

A child's purity could enable divination, but it also attracted the attention of demons, and no activity of this kind could ever completely be free of their involvement. William believed that practitioners of such magic tend to ignore that their practices often communicate with fallen angels. Wicked angels respond to conjurers’ corrupt curiosity, which is nothing more than the ‘desire to know unnecessary things’.⁶³ Demons, in their perverse emulation of the creator, also possess a special passion for the virginal purity of children.⁶⁴

While William could craft a theory explaining the utility of child-mediums, he regarded the practice as inherently psychologically and physiologically harmful. The forcible inversion of the soul upon itself was dangerous, and that was why, ‘those skilled in these works, right away when an operation of this kind is finished, close and hold tightly closed the eyes of the boys’. The adult needed to hold the child's eyes closed ‘until the soul has returned to its former state’, dripping back ‘into its body and its organs, and that is to say, until it has recovered the strength, and the organs, which it had abandoned, at least a little bit’. William elaborated on the specific potential damage to the child:

In general, there is a chance of bodily harm to the child, or maybe insanity, as seems the case with this observation, which I have discussed before [2.2.35]: an awful expression prominently remains persistently on the faces of boys after the completion of works of this kind. If the vision or apparition arises from demonic inspiration, the awfulness will remain on the child's face for much longer on account of the demonic presence, which had significantly disturbed the nature of the child. Nor does it ever penetrate the human body without leaving the vestiges of its awfulness in it. So, it should not seem wonderous to you if a lesser awfulness appears in the eyes of an observer that has operated with nature alone than when a diabolical substance, awful and unfriendly by nature, has mixed or added its own works to this kind of vision.⁶⁵

Such demonic works were always designed to lead to eventual ruin, even if they revealed bits of truth regarding hidden things. Finally, demons often magnified the damage shining bodies could do to human vision, injuring the eyes of diviners.⁶⁶

William's treatment of child-mediumship was widely cited and excerpted, and later discussions are frequently indebted to the De universo in some way. In the fourteenth century, Peter d'Ailly included large portions of it in his work on the discernment of spirits.⁶⁷ Likewise, in the fifteenth century, Pedro Garcia, bishop of Ales and later Barcelona, made use of large sections of William's text.⁶⁸ One of the works that appealed to William's authority, the De configuratione qualitatum of Nicole Oresme, is particularly significant for the way it reconceptualises the use of children in divination.

Theories of hallucination and abuse: Nicole Oresme

Writing around a century after William's death, Nicole Oresme (1320/25–82) treated child divination in a different context. Oresme is a renowned and complex figure in the history of science, ‘a name to conjure with’, frequently cited as an anachronistic precursor to various branches of ‘rational’ inquiry.⁶⁹ In his De configuratione qualitatum et motum, Oresme discussed various types of magic and how these arts appeared to produce results in certain contexts. While the account of child divination he offered will sound deceptively ‘modern’, it remains important to place it in context. Like William, Oresme accepted frequent demonic interference in the world, and the branch of magic called (vera) nigromantia operated through demons. Oresme also conceded that the future could be foretold, in theory, through prophecy and divine revelation, especially by those ‘of sober and peaceful life, whose souls are like clear and shining mirrors, clean from all worldly thoughts'.⁷⁰ Child divination, however, was different. It did not produce valuable information and belonged to a different branch of magic that did not operate through demons. He viewed this divination as an example of human techniques and practices that cause people to believe things that are not true and to perceive things that are not actually there.⁷¹ It is how these methods ‘work’ that inspired most of his discussion. In this context, children's mediumship was devoid of actual revelatory value and (mostly) demon-free, and so, while he cited William of Auvergne, the spirit of his discussion differed significantly.⁷² None the less, Oresme continued to regard the practice as harmful to the child.

Oresme explained the use of children in divination to arise from their suggestable nature and the power of their malleable imaginative faculties. It is inherently easier to convince the young that what is false is true, especially if they have devoted their ‘thoughts and imaginations to carnal desires’.⁷³ Divinatory techniques activate their ready imaginations to bring on hallucinations. Ritual gazing into reflective surfaces facilitates an ‘inward reclusion of the soul and the withdrawing of the sensitive animal spirits to the interior powers’.⁷⁴ This withdrawal is frequently described as a sudden increase in the instrument's apparent size.⁷⁵ The same basic effect happens in seizures when people claim to see things when their exterior senses stop working. Like William, Oresme believed that what the diviner saw was in their imagination. Unlike William, for Oresme, these inner visions mostly lacked any cosmic significance. The only hidden things they revealed were those already stored away in the mind. Children see what they are culturally conditioned to see, ‘solely as the result of the stories of old women which they have heard’.⁷⁶

The focusing of the mind's attention and the powers of sense inward creates physical side effects. It changes the sensitive powers and the ‘passion of the mind’ while disrupting the body's sensitive spirits.⁷⁷ These motions create hallucinations and change the body and mind. Lasting side-effects include problems with vision but also changes in appearance and general comportment in conjurers. The face may look leaner. The eyes may become darker, like those of a ‘menstruating woman’, and a conjurer may begin to behave like a madman.⁷⁸ Many of these alterations are permanent. Impaired vision, altered faces and madness were natural dangers attendant upon these practices, without any action of demons at all.⁷⁹ While demons might act during the inward reclusion of the soul, Oresme appears far less interested in that possibility than in how certain practices can naturally make seemingly unnatural things appear. He concluded his treatment of divination in reflective surfaces by reminding the reader that, demonic or natural, these practices were extremely dangerous and deserving of damnation.⁸⁰

Child-mediums and evidence of practice

Numerous other medieval references to child divination exist in different genres across a large span of time. Conjuring spirits, or ‘princes of the thumb’, on the fingernails of boys features frequently in Jewish sources. Among many examples, it was treated by Rashi (d. 1105) in his commentary on the Talmud and discussed by Eleazar of Worms (d. 1238) as a way of communicating with angels.⁸¹ The Gilbertine Robert de Brunne referenced making children divine in shining surfaces in his 1303 Middle English Handlyng Synne. ⁸² In his manual for inquisitors (1376), the fourteenth-century Dominican, Nicholas Eymeric, cited child divination (in reflective surfaces) as an example of demonic invocation that appeared to contain no overt acts of adoration or worship.⁸³ Fatefully, he concluded such acts still implied apostasy.⁸⁴ The collection of ghost tales associated with Byland Abbey (c. 1400) suggests and describes the use of child-mediums.⁸⁵ Nikolaus Jauer in 1450 treated the subject as part of his De superstitionibus. ⁸⁶ In the mid-fifteenth century, John Hartlieb described many uses for child-mediums, arguing that the secret words practitioners whispered into the ears of children held between their knees constituted pacts with demons.⁸⁷ The artist Benvenuto Cellini, in his sixteenth-century autobiography, recounted how, on the advice of a Sicilian necromancer-priest, he had used one of his shop boys around the age of twelve to conjure legions of devils in the Roman Colosseum.⁸⁸

In these accounts, there are hints that the close physical contact between the child and the master and the secrecy of their practice could be erotic. For example, in another of his works against sorcerers (1396), Nicholas Eymeric alluded to the possibility. He seemed less concerned about the harm done to the children than in demolishing their supposed purity, emphasising the harm apparently innocent children did to the adult men they sexually tempted. For Nicholas, pure and innocent virgin children have no special ability to see things in reflections. These ‘innocents do not please demons; rather, they are both harmful and perverse, and dragging men into impurity and unchastity, moreover, many of them are incubi and succubi’.⁸⁹

The practice occurs in records of legal and inquisitorial proceedings. An inquisitorial interrogation formula transcribed in the French Midi during the thirteenth century has questions regarding the use of children in magic circles to get replies from demons.⁹⁰ Alfonso x around 1265 condemned divinations done in the palm of a virgin's hand, which likely referred to child-mediums.⁹¹ The surviving fragment of Nicholas Eymeric's register (1357–80) records multiple proceedings against individuals accused of using children for divination to find lost goods, the whereabouts of prisoners and other hidden things.⁹² In 1440, the canons of Leicester accused their abbot of divining in a boy's thumbnail.⁹³ In the late fifteenth century, one of Charles vi's physicians, Jehan de Bar, before he was burned alive, confessed to having a little boy observe a demon inside a crystal, among other offences. This demon, which Jehan originally maintained was an angel, appeared to the child as a bishop.⁹⁴ There are many other condemnations of divinations through demons that make no explicit reference to children but likely referred to practices that involved children, such as the condemnations of John xxii in 1318 and 1326/27.⁹⁵

Maybe the most revealing sources for the actual use of children in divination are surviving magical texts. A formula from the second half of the twelfth century, associated with St Martial at Limoges, tells the reader to have a child gaze into the shoulder bone of a ram coated in oil. Interestingly, it is partially written in a weak cypher.⁹⁶ In the early thirteenth century, the great scholar and advisor to the emperor Frederick ii, Michael Scot, gave instructions for divination through water in a basin with a virgin girl between the age of five and seven as a medium.⁹⁷ There are many more later sources.⁹⁸ For example, the Commonplace book of Robert Reynes (1470–1500) provides instructions for questioning angels through a child-medium.⁹⁹

Perhaps our best window onto the use of child-mediums by practitioners of magic can be found in two surviving fifteenth-century handbooks of magic: Bavarian State Library, CLM 849, published by Richard Kieckhefer as Forbidden rites: a necromancer's manual of the fifteenth century, and Bodleian Library, Oxford, ms Rawlinson D. 252.¹⁰⁰ These manuscripts are some of the earliest and largest sources created by practitioners of medieval ritual magic, or necromancy, known to have survived in modern archive collections. They are miscellanies of rituals or experimenta, designed to produce specific results through the conjuration of spirits. Their authors, compilers or scribes are unknown but seem to have been men of some education and low to middling status with larger aspirations, perhaps collaborating or working along with men of some means. Many of the divinatory rituals in them call for child-mediums.

The detailed rituals, operations or experiments recorded in these magical handbooks involving child-mediums confirm many of the features mentioned in other texts. The instruments are familiar: mirrors, basins, crystals, children's fingernails, oiled bones, swords and even a candle's flame. In one case, the child's own hand is turned into a black mirror by coating it with soot and oil.¹⁰¹ This is likely the same practice outlawed by Alfonso x and referenced by John Hartlieb.¹⁰² One frequently finds protective circles, a regular feature of necromancy. The beings invoked are of all types, including angels, ambiguous spirits and demons. The experiments are overwhelmingly concerned with theft, as William of Auvergne suggested. The prepubescent age of the child is sometimes explicit: less than twelve in one ritual from the Munich manuscript, or handbook, and less than ten in an experiment in Bodl. Lib., ms Rawlinson D. 252.¹⁰³ Connected to their age, the child's sexual purity remains paramount, and almost every operation involving a child-medium specifies that the child must be a virgin.¹⁰⁴

Purity is a common emphasis in medieval magical texts and a virgin child is a particularly potent form of it. Ritual magic texts frequently instruct operators to bathe, wear clean clothes, use new or clean ritual instruments and abstain from eating certain foods and from sexual intercourse. The operations using child-mediums contain familiar instructions of this kind. Among these, the purity of the child appears especially effective, offering faster and better results than an adult working alone seems able to achieve. In purely instrumental terms, the child's purity is their main utility.

These texts’ ritual instructions frequently reference the boy's virginity, and its power is essential to success. The magister conjures the boy (almost as if he were a spirit) through the boy's baptism and sacred chastity, and the boy often conjures the spirits and demons that appear through his own virginity.¹⁰⁵ These conjurations frequently take the form of a long list of powerful virginities into which the child's purity fits.¹⁰⁶ In some cases, the child's virginity appears to be one of the main controls the master possesses over the summoned spirit and its behaviour.¹⁰⁷ In these examples, the child brings something to the ritual the adult master cannot. As Claire Fanger has argued, divinatory operations without a child in these collections often assume the ritual will fail and need to be repeated many times. In contrast, rituals involving children seem to assume success after only a few repetitions.¹⁰⁸

The child's purity takes an additional form through the frequent requirement that the medium be of legitimate birth.¹⁰⁹ Such a specification emphasises ritual purity, but it also intersects with social realities in compelling ways, limiting the type of child available. They cannot be the illegitimate offspring of clerics, to whom these churchmen would presumably have easy access, and they cannot be haphazardly sourced street children from the poorest of the poor. In other words, legitimacy of birth represents a barrier of access impacting on the social class of those involved and the relationship(s) between practitioners, mediums and their families. Educational relationships, such as John of Salisbury had described, would fit this criterion, and some texts specify that the child should be literate.¹¹⁰ Homosocial educational settings may also account for the frequent assumption that the child is a boy.

There are moments of concern for the child's welfare along with an awareness of potential danger. Ritual magic often features conjurations commanding spirits not to lie, to do no harm, and to appear in a form that allows communication without provoking terror. Operations with child-mediums have these common elements.¹¹¹ In one ritual, the operator conjures the spirit to appear on the boy's nail causing ‘neither fear nor terror nor harm to this boy’.¹¹² In an operation to conjure four kings in a crystal or a mirror, the master prays to God that the spirits invoked will do no harm to his body and the boy's body and his soul as well as the boy's soul.¹¹³ Elsewhere, a spirit is commanded to offer true replies without doing harm or impeding the child's sensus. ¹¹⁴ The child frequently holds a knife, which, like a necromancer's sword, can be seen as an apotropaic, or protective, gesture.¹¹⁵ The closing of operations can involve careful instructions for dismissing the spirits, closing the child's eyes, making the sign of the cross over the operator and over the child, and erasing the protective circle.¹¹⁶ In these moments, the operator's concern for the child seems akin to that for his own welfare.

The specific needs and characteristics of a child appear occasionally. One ritual in the Rawlinson manuscript suggests that the child should be of good character, not too talkative, and not too quiet.¹¹⁷ When a child conjures a spirit, he can use his mother tongue rather than Latin.¹¹⁸ One operation in the Munich Handbook instructs the operator along with his associates to tell the child not to be afraid when he sees the summoned demon dancing and rejoicing on his own nail.¹¹⁹ The instructions frequently feature the need for repetitions, asking the child if he sees very specific things and repeating the rite until he does.¹²⁰ The conjurations themselves, spoken by the master and by the child, often describe how the spirit should appear.¹²¹ It is not difficult for a modern reader to find such questioning suggestive. As Frank Klaassen has argued, in this way operations with child mediums may have produced apparent results very frequently because of a child's suggestible nature.¹²² At any rate, this is what Nicole Oresme believed in the fourteenth century.

Finally, we can find unsettling suggestions of dangers of other kinds to the child. There is sometimes an unnerving closeness between the child-medium and the master in these operations. This intimacy could appear erotic and has unsettled modern scholars.¹²³ The child is frequently at the master's feet and specifically between his legs or knees.¹²⁴ The master often whispers conjurations in the child's ear, holds his hands and ties slips of parchment or cloth around his fingers.¹²⁵ In the twelfth-century ritual from St Martial, the child even seems to be offered to the spirit as part of a symbolic sacrifice, standing on a new frying pan (sartago) during the ritual.¹²⁶ The scribe wrote sartagine backwards, suggesting that he recognised this detail was worthy of camouflage.¹²⁷ The symbolic sacrifice in the St Martial formula recalls a fanciful tale told by the eleventh-century churchman and rhetorician Anselm of Besate, inspired by Horace's Canidia, in which a child is used to cast a love spell.¹²⁸ In this medieval adaptation one finds a similar transformation of a literal child sacrifice into a symbolic one.

The employment of child-mediums in these late medieval handbooks of magic reveals some convergences with the polemic of hostile authorities as well as how some of these common assumptions could be interpreted differently. Practitioners of these arts recognised their rites as potentially dangerous, and if done improperly they could cause harm. There are worrisome elements correlative to what modern scholars regard as other forms of abuse. Fundamentally, the child was a useful instrument for the realisation of an adult's desires because of their virginity and (one assumes) prepubescent mind. In this fashion, the use of a child-medium appears to have been as sexualised in practice as in theory. It is important to remember, however, that the actual practitioners of these arts could have operated without any intention of harm or acceptance of a view that their arts were inherently harmful. They were participants in a contemporary cultural polemic. Ritual magic was often at its core a method of accessing and connecting with the numinous.¹²⁹ Even if it summoned demons, these fallen angels were commanded through the power of God. In such a view, the child-medium could be like the child in the Middle English Pearl, a divinely connected collaborator providing knowledge of a higher world.

Divination through child mediums was an historical practice through which, in the opinions of many contemporaries, learned clerics did harm to children. Medieval responses to the use of children in divination are an underappreciated part of larger discourses regarding clergymen, child abuse and sexuality. Discussions regarding children's mediation in magic marshalled a range of beliefs and assumptions regarding childhood and human maturation. These beliefs situated children as useful instruments for the realisation of adults’ desires. Children were useful because of their sexual status, especially their lack of an adult's sexual desire and its supposed deleterious effects on human cognition. While useful, many medieval voices believed that using a human as such an instrument was wrong. Hostile authorities, such as William of Auvergne, accepted many of the fundamental logics of children's mediumship, but argued that the practice harmed both child and adult. Others, such as Nicole Oresme, rejected its logics but still found it harmful. It could cause madness, changes in appearance, blindness and damnation. The diviners themselves, in some contexts, appear not to have recognised their practices as excessively harmful and in others to have engaged in clearly questionable practices that reduced children into (symbolic) sacrifices. While hostile authorities’ opinions do not tell the entire story, they appear to be interpretations of real practices.

The historical use of children for divination combines delusion, conspiracy, theology, philosophy and the ritual manipulation of children. These medieval divinatory practices represent an under-appreciated and understudied history for what might be more familiar occult activities, such as John Dee's early modern conversations with angels.¹³⁰ Similar practices have persisted, or perhaps been resurrected, in the modern era.¹³¹ While the most iconic ritual abuses of children in history are fantasy, such as the Witches’ Sabbath or the Blood Libel, the use of child-mediums in divination was an historical reality. It is the many-faceted, dangerous implications of the ritual use of children that make it more worthy of scholarly attention. The use of children in learned ritual magic offers a rare window onto the real practices that informed medieval paranoid fantasies. Attention to these details will allow us to see better the processes of adaptation, and indeed projection, that likely fuelled the conspiratorial fantasies churchmen projected onto others in the medieval and early modern periods.