Show Summary Details

Page of

Printed from Oxford Classical Dictionary. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 30 November 2022

magic, Greekfree

magic, Greekfree

  • Radcliffe G. Edmonds III


Greek magic is the discourse of magic within the ancient Greek world. Greek magic includes a range of practices, from malevolent curses to benevolent protections, from divinatory practices to alchemical procedures, but what is labelled magic depends on who is doing the labelling and the circumstances in which the label is applied. The discourse of magic pertains to non-normative ritualized activity, in which the deviation from the norm is most often marked in terms of the perceived efficacy of the act, the familiarity of the performance within the cultural tradition, the ends for which the act is performed, or the social location of the performer. Magic is thus a construct of subjective labelling, rather than an objectively existing category. Rituals whose efficacy is perceived as extraordinary (in either a positive or negative sense) or that are performed in unfamiliar ways, for questionable ends, or by performers whose status is out of the ordinary might be labelled (by others or by oneself) as magic in antiquity.


  • Greek History and Historiography
  • Greek Myth and Religion


The term magic, from its earliest roots, indicates something out of the ordinary, since the Greek terms, magikē or mageia, refer to the activity of magoi, the Greek word for certain Persian priests. The terms first appear in Greek texts around the time of the Persian war, but although some sources (such as Herodotus) seem to be referring to actual Persians, many of the earliest witnesses use the term to describe a Greek ritual practitioner whose extravagant claims to extraordinary power are viewed with suspicion.1 Sophocles’ Oedipus (OT 380–403) calls the diviner Tiresias a fraud and a magos when Tiresias reveals that Oedipus is the killer of Laius, and Heraclitus of Ephesus includes magoi among a list of dubious religious performers—night-wanderers, magoi, Bacchic initiates, Lenaian revelers, mystic initiates (Heraclit. Fr. 87 Marc. = B 14 DK) = Clem. Al. Protr. 2.22). Magic is thus an exotic practice, coming from alien lands or practiced by those outside the normal order of society, but this exotic practice may appear as either more powerful than normal Greek practice (cp. Plato, Alc. 1 122a) or the work of charlatans preying upon the superstitious.

Other terms are applied in Greek in similar ways and often to the same phenomena. Goētia, the work of the goēs, refers to extraordinary thaumaturgical power but usually has a negative connotation. Epaoidē is an incantation, a song or spell with performative efficacy. The word pharmakon is used to mean drug or poison, but also magic spell or incantation (that is, something that creates a powerful effect in an unknown way), and the masculine pharmakeus and (even more often) feminine pharmakis are terms for those who use magic spells to harm others. All these terms are used in Greek to label people and actions that fall, in the opinion of the speaker, outside the normal order. Such people and actions, whether explicitly so labelled or not, may thus fall under the modern rubric of magic.

Magic may thus be defined as a discourse (that is, not a thing, but a way of talking about things) pertaining to non-normative ritualized activity, in which the deviation from the norm is most often marked in terms of the perceived efficacy of the act, the familiarity of the performance within the cultural tradition, the ends for which the act is performed, or the social location of the performer.2 Greek magic is then the discourse of magic within the ancient Greek world, which differs in various ways from the modern discourse of magic, as well as other ancient discourses of magic: Hebrew, Egyptian, Mesopotamian, Phoenician, and even Roman.

Modern scholars, in distinguishing magic from the normative discourses of religion and science, have often characterized magic by criteria deriving from earlier theological and anthropological models (especially that of Frazer). Versnel 1991 sets out most clearly these criteria; magic is characterized by a coercive attitude to the divine powers, concrete intention, impersonal action, and an anti-social social evaluation.3 While such criteria fit modern uses of the term magic as a negative image of religion and/or science, they fail to distinguish the discourse of magic in ancient Greece. A coercive attitude toward the divine powers appears as a strategy in rites both normative and non-normative, and concrete intentions rather than abstract blessings are in fact generally the norm in ancient Greek prayers and rituals. The modern criteria treat magic as a wholly negative category, making it difficult to explain why some ancients would label their own activities in such a way.

Ancient sources tend to focus on different criteria when drawing the lines between magic and normative activity; the Frazerian criteria intuitively familiar to moderns are less valid than other cues. The weakness of these modern (etic) criteria for distinguishing the classifications that are significant within the ancient contexts provides the best reason for attempting to uncover the (emic) criteria used by the ancient Greeks themselves to draw the distinction between what is labelled “magic” and what is not. The most important criterion for ancient labels of magic seems to be extraordinary efficacy, whether superhumanly high powered or abnormally ineffective, miraculous power or superstitious nonsense. The familiarity of the performance—or rather the deviation from the familiar norm—is another distinguishing feature of magic in the ancient discourse. Exotic materials and incomprehensible words in rites performed in unusual ways are more likely to be labeled magic than familiar rituals performed in familiar ways. The ends for which the rite is performed do occasionally serve as distinguishing criteria; violating the boundaries of life and death (whether by killing the living or raising the dead) tends to be seen as magic, but so too is cheating in a contest. The social location of the performer, however, plays a larger role. Even otherwise normal acts may be labelled magic if performed by a marginal and alien person; and the old, ugly, Thessalian woman is the stereotype of the witch precisely because she combines so many modes of alterity.4

All of these criteria of extraordinary efficacy, performance, ends, and social location depend upon the perspective of the one using the label of magic. Magic is thus a discourse, a construct of subjective labeling, rather than an objectively existing category. Attention must therefore be paid to who is labelling whom and in what circumstances, since what seems extraordinary to one person at one time may seem routine to another in different circumstances. It is particularly important to distinguish in the evidence between labelling others as doing magic and labelling oneself as doing magic, since self-labelling is inevitably positive, whereas other-labelling is usually negative. One may intentionally deviate from expected norms to enhance the appearance of superiority to the transgressed norms, but descriptions of others’ deviations are usually critical, either warning about the socially disruptive transgressive power or denigrating the uselessness and dishonesty of the charlatanry.

Sources for Magic

As a discourse describing non-normative religious activity, magic thus exists in the ancient Greek world even before the term magikē or its cognates comes into use. Such magic before magic appears in the literature of the Greeks before the Persian war, and the discourse of magic as a way of labelling extraordinary ritual performance or activity goes beyond the uses of any one of the terms that are used to describe it.

The label of magic appears in a variety of evidence from the ancient Greek world, but the different kinds of evidence provide different kinds of self-labelling and labelling of others. Works of the literary imagination, from early epic to late novels, provide some of the richest and most detailed descriptions; these depictions, however, are not meant to be portrayals of the real world, but rather of the way magic works in the imagination. Other texts, such as histories or law court speeches, provide a more accurate depiction of how the label was applied in the real world, but they tend to be more limited in their details. The material evidence, including not only epigraphic and papyrological texts but also artistic representations in various materials, provide a more direct witness to what the ancient Greeks were actually doing, especially for the self-labelling of magic, but such evidence is always scattered, fragmentary, and difficult to interpret.

Literary evidence goes back to the earliest preserved Greek literature, in the descriptions of the magical potions of Circe and Helen and the healing charm sung over Odysseus when he was wounded by the boar (Od. 10.234–240, 4.219–239, 19.455-458). Demeter, disguised as an old nurse, claims to know various protective spells to protect a baby that are the kind of extraordinary lore that marginal old women know. (Hom. Hymn Dem. 227–230). The trick of drawing down the moon appears among the activities of witches, characters like Medea or the love-lorn Simaetha in Theocritus (Idyll 2). Tragedy provides some evidence, such as the binding song of the Erinyes in the Eumenides (328–332) or the necromantic ritual of the Persians (620–680), but tragedy, like early epic, tends to limit the presence of the extraordinarily strange. The later epic of Apollonius, by contrast, provides a good scope for displaying the magical powers of Medea, and the most extravagant depictions of magic come in the later novels, such as HeliodorusAethiopika, or the rhetorical exercises of the Roman empire, where magicians of nearly unlimited power provide paradoxes for the orators to grapple with (e.g., Ap. Rhod. Argo. 3.528–533; Heliod. Aeth. 6.14; [Quint.] Decl. Maj. 10; Libanius Decl. 41).

While in epic and tragedy, the extraordinary efficacy of magical power is valid, if dangerous, in comic texts, magic deviates from the norm of efficacy in a negative sense: it is a sham, a scam perpetrated by deviant charlatans. From Aristophanes’ joke about drawing down the moon through the references to necromancy in lost old comedies to the wicked satires of Lucian, those who make claims to extraordinary magical power are always fakes, trading on the credulity of their victims (Aristophanes Clouds 746–757; Lucian Philops. 13–15). Such credulity no doubt reflects the belief that many had in the efficacy of magic, but the incredulity of the comedians likewise reflects the suspicion of many that the abnormality of magic was its inferiority rather than superiority.

Historical accounts that purport to represent real life sometimes include references in passing to magical practices, such as the iunx, the whirligig device used for attracting lovers or drawing down the moon, in the conversation Xenophon recounts with Socrates and the courtesan Theodote (Xen. Mem. 3.11.16–17). Law-court speeches from 4th-century bce Athens show that someone could be charged with doing harm by means of a magic love potion or spell (the term pharmakon could cover either), and Plato mentions ritual practitioners who sell their services in creating binding curses and even wax figurines of their victims that are left at crossroads (Antiphon 1; Isaeus 9.37; Dem. 25.79–80, 46.14; Plato, Rep. 364be, Leg. 933ae). Likewise, the Hippocratic treatise On the Sacred Disease critiques ritual practitioners who claim to be able to draw down the moon, control the weather, and—most importantly for his purposes—cure diseases (de morb. sacr. 4.1–8). The polemical context of the medical and philosophical testimonies means that the labelling of magic is often more precise and detailed than in more casual references, but it is nevertheless important to remember that such a polemical definition comes from a particular context and may not be generalizable beyond that context.

Some of the evidence from scientific writings collects and systematizes the lore about the properties of plants, animals, stones, and other natural phenomena, and such writings can provide useful insights into what practitioners of magic thought they were doing in their procedures, especially those treatises that include recipes and instructions (e.g., Theophr. Hist. Plant. 9; Dioscorides de Mat. Med.; Kyranides; Damigeron de lap.; Orph. Lith.). Many gems with names and figures can be understood as being created for magical purposes on the basis of the correspondences with such treatises, although many more such gems survive in museum collections (usually without any secure provenance) that may well have had similar magical uses. Other kinds of talismans and amulets provide evidence that ancient Greeks were creating and using magical protections, while the corpus of curse tablets, metal lamellae (usually lead) inscribed with curses that bind or otherwise wish harm to a target, has been increasing in recent years. Such materials, often deposited in graves or wells, illuminate the anxieties and fears, the hatreds and rivalries, of strata of society that rarely make it into the histories, bringing insights into their personal lives and showing why some might choose to make use of practices labelled magic in their society.5

One body of evidence, however, provides the most outstanding source for the understanding of the practice of magic by those who labelled themselves as performing magic, the collection of papyri from Egypt, known (with debatable accuracy) in modern scholarship as the Greek Magical Papyri.6 The Greek Magical Papyri include a number of recipe books with extensive collections of recipes for magic spells, detailed instructions of the sort that appear nowhere else for the performance of rituals in the Greek tradition. These spell books seem to have been compiled by multi-lingual Egyptian scribes in the 3rd to 5th centuries ce, but the sources for the individual spells remain the subject of debate, since they clearly adapt and combine Egyptian, Jewish, Greek, and other kinds of rituals into fascinating works of bricolage.7 The Greek Magical Papyri provide instructions for inscribing curse tablets and carving magical gems, revealing elements of the rituals involved that leave no trace in the archaeological record—the spoken prayers, the accompanying sacrifices, the incense burnt, and other such ritual actions. Each type of evidence provides its own challenges for interpretation, but all provide different perspectives on the discourse of ancient Greek magic.

Objectives of Magic

The objectives of magical practice, as it appears in the evidence, can be as varied as the range of human wishes; anything for which someone might want extraordinary power may appear as the aim of magic. Although some scholars have imagined magic as primarily concerned with harmful effects for personal purposes (with a special category of “white” magic to account for the exceptions), it is not the ill will that marks magic in the ancient sources as much as the extraordinary nature of the performance or the abnormal social position of the performer. Above all, the discourse of magic applies in cases where the practice is seen as an extraordinarily efficacious means of obtaining the end, whatever that end might be.

One of the ends for which magic may be used, however, is indeed to bring harm upon another person, and the category of curses includes many that fall within the discourse of magic. Curses aimed at hindering a rival in a wide variety of contexts appear not only in the literary evidence, but also in the material evidence of curse tablets, most often lead lamellae inscribed with the text of a curse. The contexts may be the more implicit competitions of business or personal rivalries or the explicit contests of public performances such as the theatrical or athletic arenas or the law courts. The extraordinary efficacy of these magical curses appears as a form of cheating within the contest, and it is notable that such curses are never boasted of as the means by which a rival was defeated (in contrast to strength, speed, or even cleverness), although rivals might accuse each other of cheating by use of magical curses.8

Curses deemed magical are distinguished not only by their extraordinary efficacy, beyond the power that ordinary ill-wishing of a rival might have, but even more by aspects of their performance, particularly the timing and audience of the cursing ritual. While curses performed by duly appointed officials of the community at regular times for an audience of the community may resemble magical curses in many ways in their form (the invocation of some divine power, a wish for harm to the target, and often some symbolic representation of that harm in the ritual), magical curses are characterized by performance in secret for an audience of the divine powers alone. The curses performed before the Athenian assembly meetings (parodied in Aristophanes, Thesmoph. 335–372) and the ceremonial Curses of the Teians (Dirae Teorum = SIG3 37–38) are part of normative religious practice, even if the harm they call down upon those who violate community norms is more colorful and more dire than that in most magical curses. The ceremonial cursing of any Theran colonist to Cyrene who tries to return to the mother city involves the melting of wax images as a symbolic expression of the harm wished upon anyone who violates the plan, but the way it is performed in public by the whole community sets it apart from the rituals whose procedures are described in the Greek Magical Papyri.9 One spell for calling down the wrath of the moon goddess on a target involves a pre-made offering (compounded of weird elements and stored in a box), which can be used, not just on socially sanctioned occasions, but “whenever you want to perform the rite, take a little, make a charcoal fire, go up on a lofty roof, and make the offering as you say this spell at moonrise” (PGM IV.2463–2466). The isolated and secret location, the lack of other audience, and the fact that the rite can be performed whenever the magician wishes, rather than at the sacred times appointed by the community, all place this curse within the discourse of magic, rather than normative religious practice.

The contrast in mode of performance is particularly significant in the curses that have been dubbed “prayers for justice” or “judicial prayers,” in which the one cursing calls down the vengeance of the divine power on the target because of some previous wrong done by the target. Such “prayers for justice” more often include the name of the agent making the curse, a feature usually absent in the magical curses, where the one cheating does not wish to name himself. The “prayers for justice” seem often to have been deposited or even displayed in a public place like a sanctuary, rather than secretly buried in a grave or a well, the favored places for the deposit of magical curses.10 The effects requested in such prayers for justice and other public curses tend also to be more dramatic and spectacular, since the curse is often intended to have a deterrent effect or to express the agent’s anger at being victimized by the target of the curse. Secret, magical curses in agonistic contexts more often ask for just enough effect to allow the curser to win, asking the gods to trip up the horses of an opponent’s chariot, to bind his tongue in a law court so that he cannot speak eloquently, or to bind back and restrain his personal or professional activities in the case of a more private conflict (e.g., chariot–Gager #5 = SEG 7.213 (15.847); tongue—Gager #40 = DTA 107; work—Gager #60 = SGD 52).

The language of binding indeed seems to characterize the magical curses, and the Greek term katadesmos—binding down—is used in the ancient sources to designate the magical curse tablets. In addition to verbs of binding, verbs of handing over or registering appear, as the target is consigned to the power of the divinity invoked. It is notable, however, that such language of transfer also appears in the public curses, whereas binding less often appears in such prayers for justice or community curses.

As an arena in which wins and losses are particularly passionately felt and the outcome is notoriously hard to control, the competitive context of eros (love and sex) produces a number of agonistic, magical curses. Some of these are simply curses to restrain a rival’s performance in the competition for the affections of another (e.g., DT 68, SEG 30.353), but others directly target the beloved, seeking to obtain the beloved’s affections.11 Such obtaining spells (agōgai) seek to bring the target to the agent of the spell for sex, but they operate not by binding down the target’s will but rather by inflaming her desire (the overwhelming majority of such erotic spells have a male agent and female target, although the pattern of spell is the same for female agents seeking male targets and homoerotic relations). The spell inflicts the torments of love—loss of sleep and appetite, internal burning feelings, twisting and turning—on the target, often in very graphic terms that make it clear that the eros inflicted on the target is a curse.

Fetch Euphemia, whom Dorothea bore, for Theon, whom his mother Proechia bore, to love me with love and longing and affection and intercourse, with mad love. Burn her members, her liver, her female parts, until she comes to me, longing for me .|.|. do not let her eat, or drink, or find sleep, or have fun, or laugh, but make her run away from every place and from every house and leave father, mother, brother, sisters, until she comes to me. (SM 45.29–32, 45–8 = PGM CI; compare SM 40.12–21 = PGM LXXI)

The list of sex acts desired (or from which the target is bound from engaging with any other partner) is often even more explicit (e.g., PGM IV 351–355, SM 48 21–24). While agōgai seek to obtain a relationship, other erotic spells seem designed to retain a lover, binding his will and restricting his ability to interact with other partners (more of such curses have female agents and male targets, but the pattern is similar whatever configurations of gender between agent and target). Although the erotic curses focus on some other, whether lover or rival, as the target, other spells appear that aim to enhance the attractiveness of the agent. Such erotic magic may take the form of an object, like the magic girdle (or necklace or other adornment) of Aphrodite, the kestos himas, which Hera borrows in the Iliad to seduce Zeus (Iliad 14.197–210).12 Other amulets endow the wearer with irresistible charisma, personal attractiveness that transcends the erotic and works in all social situations, as one recipe boasts, “even against kings!” (PGM XXXVI. 35–68).

The extraordinary efficacy of such items brings them into the discourse of magic, and the same is true of various substances and preparations that promise enhanced performance or success in the erotic arena, whether a simple recipe in the Greek Magical Papyri for a lotion of honey and pepper to “put on your thing” in order to “copulate a lot,” an exotic herb that provides superhuman performance, or even amulets designed to open the womb and increase the chances of conception (PGM VII.191–2, VII.183–5; Theophrastus Hist. Plant. 9.18.9, 9.18.5, cp. the common “womb-key” amulets, which depict a uterus-shaped jar and a large key to signify the opening of the womb for conception, e.g., CBd 728 = BM 1986,0501.31). Other substances and preparations with extraordinary efficacy serve other objectives, providing cures not just for erectile dysfunction or difficulty in conception but for the whole range of the thousand natural shocks that flesh is heir to. The remedy may be no more than an incantation, like the charm Odysseus’ uncle sang over him when he received the wound in a boar hunt that produced his famously recognizable scar (Od. 19.455–58). Plants, stones, and other substances may also be imagined as having a natural potency to heal certain problems, or the affliction may be imagined as being caused by a personal agent, either the magical attack of another mortal or an assault by some hostile divine power, often referred to as a daimon. A papyrus amulet designed to repel fever from a little girl illustrates the idea.

I, Abrasax, shall deliver. Abrasax am I! ABRASAX ABRASICHO'OU, help little Sophia-Priskilla. Get hold of and do away with what comes to little Sophia-Priskilla, whether it is a Shivering Fit—get hold of it! Whether a Phantom—get hold of it! Whether a Daimon—get hold of it! I, Abrasax, shall deliver. Abrasax am I! ABRASAX ABRASICHO'OU. Get hold of, get hold of and do away with .|.|. what comes to little Sophia-Priskilla on this very day, whether it is a Shivering Fit—do away with it! Whether a Daimon—do away with it! (PGM LXXXIX 1–27)

The deity Abrasax is invoked to drive off the unknown daimon or other hostile entity that afflicts little Sophia-Priskilla.

In addition to remedies to treat problems after they occur, protective amulets and other magical phylacteries aim to prevent any such harm before it occurs. Amulets, whether of papyrus or lamellae of silver or gold or even gems engraved with special images or words, can ward off disease or snake-bite or magical attack when carried on the person or protect a house or fields from any harm that might come to it. The most famous warding magics, known as the Ephesia Grammata, claim such extraordinary efficacy in a 4th century bce tablet.

Whoever hides in a house of stone the notable letters of these sacred verses inscribed on tin, as many things as broad Earth nourishes shall not harm him nor as many things as much-groaning Amphitrite rears in the sea. (Getty 81.A1.140.2 2–5)

The same Ephesia Grammata were reputed to have saved Croesus from burning on his pyre and to have enabled a wrestler who wore them to defeat all opponents—when the amulet was removed, he was thrown thirty times in a row (Eustathius on Odyssey XIX.247 2.201–2).

Such magical healing or protection is distinguished not only by its extraordinary efficacy but also by the weirdness of the performance of the preparation or ritual and, perhaps most of all, by the social location of the performer. The old wives’ remedies and the lore of the root-cutters and itinerant healers fall into the discourse of magic because of the marginal status (old, poor, female, foreign, etc.) of those who tout such remedies. The peculiarity of the performance, the weird substances employed, the strange times and seasons in which the materials must be collected, or the unusual ways in which they must be used, mark these remedies as magic, in contrast with the normal kinds of healing procedures available to the ancient Greeks. It is such extraordinary performance or social location that differentiates magical healing from the miraculous healing obtained at the temples of healing divinities such as Asclepius. The efficacy of those cures is just as extraordinary, but the practitioners at the temple have established status within the community and the rituals follow familiar patterns of contact with the gods, such as votive dedication, sacrifice, and even incubation (sleeping in the sanctuary to obtain a dream message from the god).

A similar distinction by performance and performers appears in divination, between magical ways of obtaining communications from the divine and the normative religious methods. Many varieties of divination appear in the ancient Greek evidence, from omens in the flights of birds to reading the signs in the liver of a sacrificed animal to the most prestigious Oracle at Delphi, but only some forms at some times fall within the discourse of magic (cp. PGM I.327–331). Divination in general appears in two modes: interpersonal communication between deity and mortal (along the model of speech) or impersonal communication (along the model of writing) that involves the decipherment of signs (cp. Plato Phaedrus 244). All sorts of divination in both modes may at times appear as magical, especially if they lay claim to extraordinary efficacy, but a peculiar or unfamiliar kind of performance by a performer with marginal status is more likely to be labelled magic than the Pythian priestess speaking the oracles of Apollo at Delphi or an army’s diviner reading the entrails of a sacrificial victim on the field before a battle.13 Thus, divination by lots or through dreams, although fairly normal in ordinary circumstances, may appear as magic if the practitioner is a distrusted outsider or if the procedure involves weird elements as in some of the ritual recipes of the Greek Magical Papyri (e.g., PGM VII. 740–755). Only one kind of divination, necromancy, appears as magic in nearly every circumstance; the skull cup evocations of dead spirits in the Greek Magical Papyri and even Odysseus’ summoning of the shade of Tiresias in the Odyssey are extraordinary enough in their performance to be labelled magic—even if a later writer rewrote the scene from the Odyssey to make it seem more magical by adding alien elements of magical words and invocations of Egyptian deities (PGM IV. 1928–2005, 2006–2125, 2125–2139, 2140–2240; Od. 11. 23–50, Sextus Julius Africanus Kestoi 18 = PGM XXIII). In the literary imagination, necromancy—no longer simple communication with the dead but the violent reanimation of a corpse—becomes the ultimate magic act, performed nefariously by the most horrible of marginal figures, such as the witch in Heliodorus’ Aethiopika (6.14–15), who reanimates the corpse of her own son, only to hear him prophesy her own immediate doom.

Astrology, one of the most prevalent forms of divination, at least in post-Classical times, enters the discourse of magic in different ways. While the marginal status of its performer might taint an astrological reading with the label of magic, it is the claims to extraordinary efficacy, bolstered by the extreme systematicity of its methods, that characterize more of the astrological writings as part of the discourse of magic. Pliny (NH 30.1.1–2) sums up this characteristic attitude: “to complete its universal sway, magic has incorporated with itself the astrological art,” but the same idea appears less explicitly in many astrological manuals in Greek. The precise calculations that can determine the exact influences of the celestial powers upon the lot of mortals surpass the efficacy of any other form of divination, and a practitioner who can master all the extreme complexities thus distinguishes himself as an extraordinary performer, abnormal in a positive sense as a learned magician who can learn things beyond the ken of normal mortals.14

This rhetoric of systematicity also characterizes many of the alchemical writings from antiquity, especially the works of Zosimus of Panopolis, a Greek writer from 3rd–4th century ce Egypt. Alchemy consists of the transformations of the qualities of matter (from grey lead to gleaming gold, from clear crystal to purple amethyst, etc.), as well as, in some texts, of spirit or soul. In processes sometimes compared to the work of the creator god, the alchemist purifies his object (be it matter or soul) from undesirable qualities and then imbues it with new virtues. The extraordinary efficacy of the procedures involved in such transformations is further marked by the extraordinary complexity of the performances, involving specialized knowledge that only a learned magician might know. Much of the technical knowledge seems to come from Hellenistic systematizations of the secret lore about stones and metals, and the 5th century bce philosopher Demokritos is often credited as the original founder of the art, which he received from Persian magicians like Ostanes (PM 3 35–64 Martelli = CAAG II. 42.21–43.22).

Theurgy, too, involves the purification of both matter and spirit to facilitate the connection of the practitioner with the divine powers, either by leading the spirit of the practitioner up to the gods or by drawing down divine power and infusing it into matter. As with astrology and alchemy, the extraordinary efficacy of theurgical rituals comes from the extreme complexity and systematicity of the theological and cosmological framework that the theurgist manipulates to achieve his or her results. Although some philosophical sources insist that the union of the mortal and divine is the only aim of the theurgy, other evidence shows that theurgical procedures were used for a variety of ends, whether receiving an oracular revelation during the meeting with the god or imbuing a statue or amulet with divine power that could be used for a wide range of purposes, from erotic to healing and protection and beyond (Iamblichus de myst.; Proclus de sacr. et mag.).

Techniques of Magic

Since the objectives of magical practice are beyond the power of ordinary humans, the performer must somehow access extraordinary power to achieve such extraordinary results. The magician may call on a variety of non-human powers, ranging from the supreme lord of the universe through a number of gods who specialize in certain areas to the spirits of the restless dead. Often the power employed is not explicitly invoked, but the desired result is articulated in some kind of performative utterance—I bind so-and-so, his tongue and hands, ... etc. Although modern scholars have often taken the way that the divine power is invoked to mark the distinction between religion and magic, classifying supplications to the divine as religion and commands to divinities or straightforward performative utterances as magic, the ancient evidence shows that such a distinction was rarely significant. Many examples combine the forms, using supplicative subjunctives and direct imperatives in the same plea (e.g., DT 25.13-14, 16-18 = Gager 46), and the modern critique of magic as working automatically, ex opere operato, stems from later Christian theological debates, specifically Protestant critiques of Catholic ritualism. In the ancient evidence, it is the departure of the performance from the normal and familiar patterns that marks it as magical, what Malinowski refers to as the “coefficient of weirdness.” The strangeness of the performance may be analysed in terms of the rhetoric of its expression, the “poetics of the magic charm” as Versnel calls it, noting the use of devices such as metaphor and metonymy, repetition and emphasis, vivid imagery and poetic language.15

The weirdness may be seen both in the words spoken and the rituals enacted. Magical rituals are marked by their unusual materials, be it strange herbs and exotic incense, peculiar paraphernalia and abnormal animals for sacrifice, gems or metals used as writing surfaces for their symbolic resonances, or special inks composed of symbolically charged ingredients (e.g., PGM I. 244–247, II. 35–40, IV. 2005, 2142–2143, 3200–3205). Likewise, the actions performed may deviate from usual religious practice. A recipe from one of the spellbooks of the Greek Magical Papyri provides a good example:

Take a lead lamella and inscribe with a bronze stylus the following names and the figure, and after smearing it with blood from a bat, roll up the lamella in the usual fashion. Cut open a frog and put it into its stomach. After stitching it up with Anubian thread and a bronze needle, hang it up on a reed from your property by means of hairs from the tip of the tail of a black ox, at the east of the property near the rising of the sun. (PGM XXXVI. 231–255)

The materials specified for the request to the divinity are markedly unusual—bat’s blood over the inscription done by a bronze stylus on a lead sheet, rolled up and stuffed into a frog. This elaborate procedure is not the ordinary way to petition a divine power; its very weirdness indicates that it must have extraordinary efficacy. The body of the frog is explicitly manipulated for the symbolic resonances of the action, which are articulated in the prayer.

Powerful angels, just as this frog drips with blood and dries up, so also will the body of him, NN whom NN bore, because I adjure you, who are in command of fire, Maskelli Maskellō (add the rest, the usual).

The sufferings of the eviscerated frog become a metaphor for the wished for sufferings of the target, whom the performer wants to drip with blood and shrivel away just like the frog. Similar metaphorical manipulations appear in the creation of figurines which are bound or mutilated or pierced in a performance that symbolizes the effect desired, often with vivid rhetorical force.16 At times, the target is represented by means of synecdoche, when a part of the target—some hair or other bodily material (often referred to as ousia—essence) or even just a written representation of the name—is manipulated in the ritual. The synecdoche may be incorporated into the metaphor, as when a figurine is made of wax and hair from the target is incorporated before the whole is melted or stuck with pins (e.g., PGM IV.296–304).

The frog recipe does not just rely on metaphor, whether through ritualized action or the performative speech acts describing it. Certain divine powers are invoked to effect the desired result that has been so graphically depicted. The magician mentions “powerful angels” who will ensure that the target (whose name is to be supplied in the blanks left in the recipe) suffers, and the magician can expect the angels to make the effect happen because he invokes the powers under the secret name that begins with Maskelli Maskellō.17 Such magic words, voces magicae, often appear as the secret names or other tokens that attest to the magician’s extraordinary connection with the divine power.18 The rhetorical effect of these voces magicae is enhanced by their peculiar sounds (the jingle of Maskelli Maskellō) or appearance (palindromes like Ablanathanalba), and the frog spell also includes such words that are to be written on the lamella in a pyramid or rectangular form. The shape of the written words supplements the rhetorical effect of the words spoken as well as the physical manipulation of the frog, and the recipe even includes a figure to be drawn on the lamella for further effect.

The Extraordinary Discourse of Magic

The discourse of magic, this mode of talking about and classifying activities, may thus be used intentionally by ritual performers to mark their own activities as extraordinary, filled with magical words and exotic rituals that indicate the abnormal efficacy of the performance. Such self-labelling nevertheless uses the same kind of criteria of efficacy, performance, ends, and social location as the labelling of others. Someone might appropriate the stereotype of an aged Egyptian and make claims of extraordinary efficacy that are bolstered by an exotic performance in order to increase the confidence of a client, just as someone might accuse an elderly foreigner speaking prayers in an alien language of transgressive ritual activity.19 Likewise, an author creating a literary account of magic would use the same markers of the extraordinary to create a convincing account. Aristophanes, for example, refers (Clouds 746–757) to a Thessalian witch who could draw down the moon for the purpose of helping a sneaky old man avoid his debts, a reference that includes the extraordinary efficacy of a bizarre performance for illegitimate ends by a socially marginal performer. The efficacy would indeed be extraordinary in the positive sense if such a scheme could work, but the character proposing the scheme is a comic idiot, and his plans are dismissed as delusional nonsense—something of abnormally low efficacy for dealing with the situation. The passage thus neatly illustrates the discourse of magic in the ancient Greek world, including the ambivalence of the efficacy of magic as positive or negative. In contrast to modern criteria for magic that focus on a coercive attitude to depict magic in a negative way, moreover, the evaluation of magic in ancient Greek culture can be positive as well as negative, and the discourse of magic to describe extraordinary ritual activity can be used either for praise or blame.

Discussion of the Literature

The scholarship on Greek magic blossomed in the years following the 1986 collection and translation of the Greek Magical Papyri by a team under the direction of Betz, although Luck’s 1985 collection of primary texts in Arcana Mundi also helped stimulate interest.20 The 1991 collection of essays in Magika Hiera still provides fundamental studies of many of the main issues, while some later collections have explored the variety of phenomena in ancient magic.21 Graf’s Magic in the Ancient World provides an introductory overview, and this work opened the door to even more scholarly attention to magic, including the overviews by Collins and Dickie.22 The number of studies on particular areas of Greek magic has proliferated in recent years, with good work on divination, curses, erotic magic, astrology, and theurgy, and the advancement of digital imagery has enabled new work to be done on gem amulets and other material remains.23 Ogden’s sourcebook helped to open up further the study of ancient magic to Greekless readers, even as its problems draw attention to the need for better definitions of magic.24 The most significant work in this regard has been done by Jonathan Z. Smith and Richard Gordon, especially in Gordon’s “Imagining Greek and Roman Magic,” and some recent scholarship has attempted to provide more comprehensive views, both of Greek magic and of magic as a later discourse.25 Much work, however, remains to be done in updating the corpora of texts, especially the treatises on alchemy and astrology and the ever-growing body of curse tablets. The discourse of magic in ancient Greek culture continues to fascinate.

Links to Digital Materials


  • Addey, Crystal. Divination and Theurgy in Neoplatonism: Oracles of the Gods. Farnham, U.K.: Ashgate, 2014.
  • Barton, Tamsyn. Ancient Astrology. Sciences of Antiquity. London: Routledge, 1994.
  • Beck, Roger. A Brief History of Ancient Astrology. Malden, MA: Blackwell, 2007.
  • Betz, Hans Dieter, ed. The Greek Magical Papyri in Translation: Including the Demotic Spells. 2nd ed. Chicago: University of Chicago Press, 1992.
  • Bortolani, Ljuba. Magical Hymns from Roman Egypt: A Study of Greek and Egyptian Traditions of Divinity. New York: Cambridge University Press, 2016.
  • Brashear, William M. “The Greek Magical Papyri.” Aufstieg Und Niedergang Der Römischen Welt II 18 (1995): 3380–3684.
  • Bremmer, Jan N. “The Birth of the Term ‘Magic.’” In The Metamorphosis of Magic from Late Antiquity to the Early Modern Period. Edited by Jan N. Bremmer and Jan. R. Veenstra, 1–11, 267–271. Leuven, Belgium: Peeters, 2002.
  • Collins, Derek. Magic in the Ancient Greek World. Malden, MA: Blackwell, 2008.
  • Daniel, Robert, and Franco Maltomini, eds. Supplementum Magicum. Vols. 1 & 2, Papyrologica Coloniensis, XVI. Cologne: Westdeutscher Verlag, 1990, 1992.
  • Dasen, Véronique.“Probaskania : Amulets and Magic in Antiquity.” In The Materiality of Magic. Edited by Jan Bremmer and Dietrich Boschung, 177–203. Paderborn, Germany: Wilhelm Fink Verlag, 2015.
  • Dickie, Matthew W. “The Learned Magician and the Collection and Transmission of Magical Lore.” In The World of Ancient Magic: Papers from the First International Samson Eitrem Seminar at the Norwegian Institute at Athens, 4–8 May 1997. Edited by David R. Jordan, Hugo Montgomery, and Einar Thomassen, 163–193 (Bergen, Norway: Paul Aströms, 1999).
  • Dickie, Matthew. Magic and Magicians in the Greco-Roman World. London: Routledge, 2001.
  • Dieleman, Jacco. Priests, Tongues, and Rites: The London-Leiden Magical Manuscripts and Translation in Egyptian Ritual (100–300 CE). Religions in the Graeco-Roman World 153. Leiden, The Netherlands: Brill, 2005.
  • Edmonds, Radcliffe. Drawing Down the Moon: Magic in the Ancient Greco-Roman World. Princeton, NJ: Princeton University Press, 2019.
  • Eidinow, Esther. Oracles, Curses, and Risk Among the Ancient Greeks. Oxford: Oxford University Press, 2007.
  • Eidinow, Esther. Envy, Poison, and Death: Women on Trial in Classical Athens. Oxford: Oxford University Press, 2016.
  • Faraone, Christopher A. Ancient Greek Love Magic. Cambridge, MA: Harvard University Press, 1999.
  • Faraone, Christopher A., and Dirk Obbink, eds. Magika Hiera: Ancient Greek Magic and Religion. New York: Oxford University Press, 1991.
  • Gager, John G., ed. Curse Tablets and Binding Spells from the Ancient World. Oxford: Oxford University Press, 1992.
  • Gordon, Richard. “Imagining Greek and Roman Magic.” In Witchcraft and Magic in Europe: Ancient Greece and Rome. Edited by Bengt Ankarloo and Stuart Clark, 159–275. Philadelphia: University of Pennsylvania Press, 1999.
  • Graf, Fritz. “Prayer in Magic and Religious Ritual.” In Magika Hiera: Ancient Greek Magic and Religion. Edited by Christopher Faraone and Dirk Obbink, 188–213. Oxford: Oxford University Press, 1991.
  • Graf, Fritz. Magic in the Ancient World. Cambridge, MA: Harvard University Press, 1997.
  • Johnston, Sarah Iles. Ancient Greek Divination. Malden, MA: Blackwell, 2008.
  • Johnston, Sarah Iles. “Myth and the Getty Hexameters.” In The Getty Hexameters: Poetry, Magic, and Mystery in Ancient Selinous. Edited by Christopher A. Faraone and Dirk Obbink, 121–156. Oxford: Oxford University Press, 2014.
  • Love, Edward O. D. Code-Switching with the Gods: The Bilingual (Old Coptic-Greek) Spells of PGM IV (P. Bibliothèque Nationale Supplément Grec. 574) and Their Linguistic, Religious, and Socio-Cultural Context in Late Roman Egypt. Berlin: De Gruyter, 2016.
  • Martelli, Matteo. The Four Books of Pseudo-Democritus. Leeds, U.K.: Maney, 2013.
  • Meyer, Marvin W., and Paul Allan Mirecki, eds. Ancient Magic and Ritual Power. New York: Brill, 1995.
  • Mirecki, Paul Allan, and Marvin W. Meyer, eds. Magic and Ritual in the Ancient World. Leiden, The Netherlands; Boston: Brill, 2002.
  • Ogden, Daniel. Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds, 2nd ed. Oxford: Oxford University Press, 2009.
  • Otto, Bernd-Christian. Magie : Rezeptions- Und Diskursgeschichtliche Analysen von Der Antike Bis Zur Neuzeit. Religionsgeschichtliche Versuche Und Vorarbeiten 57. Berlin: De Gruyter, 2011.
  • Preisendanz, Karl, and Albert Henrichs. Papyri Graecae Magicae: Die Griechischen Zauberpapyri, 2nd ed. Stuttgart: Teubner, 1973.
  • Riess, Werner. Performing Interpersonal Violence: Court, Curse, and Comedy in Fourth-Century BCE Athens. Berlin: De Gruyter, 2012.
  • Schäfer, Peter, and Hans G. Kippenberg, eds. Envisioning Magic: A Princeton Seminar and Symposium Magic in the Ancient Greek World. Leiden, Belgium: Brill, 1997.
  • Smith, Jonathan Z. “Here, There, and Anywhere.” In Magic in History: Prayer, Magic, and the Stars in the Ancient and Late Antique World. Edited by Scott Noegel, Joel Walker, and Brannon Wheeler, 21–36. University Park, PA: Pennsylvania State University Press, 2003.
  • Stratton, Kimberly B. Naming the Witch: Magic, Ideology, and Stereotype in the Ancient World. New York: Columbia University Press, 2007.
  • Tambiah, S. J. “A Performative Approach to Ritual.” In Proceedings of the British Academy LXV, 113–169. London: Oxford University Press, 1979.
  • Tanaseanu-Döbler, Ilinca. Theurgy in Late Antiquity: The Invention of a Ritual Tradition. Göttingen, Germany: Vandenhoeck & Ruprecht, 2013.
  • Versnel, Henk S. “Beyond Cursing: The Appeal to Justice in Judicial Prayers.” In Magika Hiera: Ancient Greek Magic and Religion. Edited by Christopher A. Faraone and Dirk Obbink, 60–106. New York: Oxford University Press, 1991.
  • Versnel, Henk S. “Some Reflections on the Relationship Magic-Religion.” Numen 38, no. 2 (1991): 177–197.
  • Versnel, Henk S. “The Poetics of the Magical Charm: An Essay in the Power of Words.” In Magic and Ritual in the Ancient World. Edited by Paul Allan Mirecki and Marvin W. Meyer, 105–158. Leiden, The Netherlands: Brill, 2002.
  • Viano, Cristina. “Les Alchimistes Gréco-Alexandrins et Le Timée de Platon.” In L’Alchimie et Ses Racines Philosophiques: La Tradition Greque et La Tradition Arabe. Edited by Cristina Viano, 91–107. Paris: Librairie philosophique J. Vrin, 2005.
  • Zografou, Athanassia. Papyrus Magiques Grecs: Le Mot et Le Rite. Autour Des Rites Sacrificiels. Παράρτημα‎ 85. Ioannina, Greece: Panepistimio Ioanninon, 2013.