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: 13 April 2021

magic, Romanfree

  • Richard Gordon


Roman religion has conventionally been understood as a civic or “polis” religion in which the population performed the same rituals, attended the same festivals, and believed in the same divinities, an image conveyed by the extant Roman historians (including the Greek Polybius) and the antiquarian tradition. This convention has successfully obscured the fact that the range of religious activities in the City, to say nothing of the surrounding areas of central Italy, was in reality always far wider. More neutrally, we may view the religious field at Rome as a site of constant, if intermittent, conflict over effective means of relating to the other world and the legitimate use of religious knowledge, conflict that parallels in a different key the disputes over proper religious observance that took place within the ruling elite itself and its various priestly colleges. If the larger category of dismissal was superstition, the narrower and still more negative one was magical practice. There were however several sub-classes here, of which witchcraft and sorcery were but two. Over the thousand years of knowable Roman history, which saw a single city extend its political and extractive reach to a maximum of 4.4 megametres and then decline, the understanding of magic as malign (i.e., witchcraft/sorcery) altered in often dramatic ways, beginning with anxieties typical of agrarian communities, and culminating in Late Antiquity in charges of lese-majesty at court and routinized attempts at revenge by rival rhetors, to which we can add the deployment of allegations of magic by Christian hardliners in attacking paganism and heretics. A significant process in this history was the gradual appropriation over the last hundred and fifty years of the Republic of a term (magia) and its associated stereotypes from the Hellenistic Greek world, which together provided a medium, widely exploited in a variety of literary genres, for re-figuring the social disruptions that attended the violent self-destruction of the aristocratic régime and remained thereafter a powerful imaginative resource for constructing a variety of boundaries around a moral centre, claimed to be steady but in fact altering very considerably under shifting political, social, and religious conditions. Magic was thus not simply a medium for accusation but also a metaphor and social figuration; it thus played a significant role in the long-term legitimation of the self-styled dominant religious order. Moreover, since marvel, transformation, and the uncanny likewise belonged to the same semantic field, magic helped sustain the fecund irrationality indispensable to a polytheistic world-view.


It is a moot point in what sense a concept can exist without a corresponding denotation. At any rate, what seems to be the oldest expression for magical practice, veneficia/venena et carmina (vel sim.), “potions and incantations,” was focused upon concrete agency not upon a formal abstraction.1 This phrase lies behind the unclarity that attends the (much later) reports concerning the rulings of Twelve Tables (fictive date 450–449 bce) on witchcraft, an unclarity that has led some to think that acquiring others’ crops by venena must have been a different delict from fruges excantare, “spiriting crops away by incantation.”2 There was no higher-level term in the early Republic: since it was always uncertain how witchcraft had been performed, both incantation (in/ex/cantare) and substances (venena) could figure as a synecdoche for a range of supposed witchcraft practices, of which the most explicit and later, memorable, was fruges excantare, to “lift” a harvest from someone else’s field, in other words, causing a neighbour’s crops to fail.3 The explicit recognition of this practice in this early law-code underscores the anxieties of the agrarian sector regarding what might be classified as exceptional (implicitly un-natural) meteorological events, such as hail in May, and the inclination to blame human malevolence rather than any other of a range of alternative explanations, but also the envy aroused by differential prosperity. Accusations of malign magic are a typical strategy of social levelling in agrarian societies; hence spiriting away crops and causing meteorological havoc are two of the most common activities of the witch in the later stereotype, transferred from actual farmers to a female fantasy figure.4

The Semantic Field

The earliest recorded term in Latin for a “wise woman,” praecantrix, perhaps a transposition of Gk. epôdós, appears c. 200 bce.5 Saga is first recorded in 2nd-century comedy.6 An agentive term focused on potions, namely veneficus (m.)/venefica (f.), “a person who knows about/mixer of potions,” is found at the same time, again no doubt a transposition from Greek pharmakeus/is.7 The verb venenare is used once by Plautus to mean “bewitch,” and the neuter words venenum/veneficium (also in the plural) had evidently begun to be used to mean “witchcraft/sorcery” by the mid-1st century bce, if not earlier, even though their commonest sense always remained the concrete “poison.”8 The word amatorium in the special sense of “love-philtre” was used casually by the 1st century ce, possibly already by Catullus.9 Maleficus became a common term in late antiquity for “wicked wizard,” but is earlier found only as an adjective (note however Tac. Ann. 2.69.3: alia malefica, “other devices of black magic”).10 We should also include here terms such as hariolus and haruspex, which, like Greek mantis, were used to refer to a variety of male ritual specialists with a broad range of skills.

On its first introduction into Latin, the Greek word mageia (itself, as an abstract substantive, a relative late-comer into Greek), is found almost exclusively, until the 2nd century ce, only as an adjective magicus. Pliny the Elder, in the third quarter of the first century ce, still uses the form (ars) magice (= μαγικὴ τέχνη‎, magikē technē), magical knowhow.11 In this adjectival form, the term found widespread acceptance among the Latin poets, especially the elegists, from the last decade of the Republic to the end of the reign of Augustus, at least partly for its metrical value by contrast with the metrically problematic venefic- words, which are almost never found in poetry. Its reference in these contexts varies between the dread marvellous (e.g., the Gorgon’s head) and dread divinity (the gods of the underworld), between uncanny witchcraft and mere superstition, between the activities of Italic herbalists and wise-women and those of famous mythical figures such as Medea and Circe.12 This integration of Greek literary stereotypes with indigenous Italic traditions, and the resulting shift in perceptions, is a striking feature of the poetry of the post-civil war period, although it surely draws upon the practice of earlier translations/adaptations into Latin of Hellenistic poetry and drama, such as Ennius, Accius (both of whom composed a Medea; Ennius’ effort became a school-text by the Late Republic) in the early 2nd century bce, or Varro Atacinus’ Argonautae in the 1st (frg. 11).13 Greek mythic witches mark a decisive step in the othering of the Italic tradition of herbal practice and the establishment of an enduring stereotype of the misuse of religious knowledge. On the other hand, it is also worth noting the related Greek words that were never taken over into Latin, notably γόης, γοητεία‎ (góês, goêteía), “wizard, wizardry” see Magic, Greek, μαγγανεία‎ (manganeia), “hocus pocus,” and the family of θελγ‎ -words (thelgein, thelktērion), connoting the marvellous, positive, sensual aspects of the magical.14

Although Plautus has one of his characters threaten to take revenge on a Thessalus veneficus (referring to pseudo-Sosias—in fact Mercury in disguise), and Vergil fleetingly introduces a male practitioner named Moeris, a Greek name surely borrowed from an unknown Hellenistic source, the standard poetic stereotype of the magical practitioner in the Republic, and indeed through the Julio-Claudian period, is female.15 This stock-figure, or rather caricature, was created by the stasis system employed by the rhetorical schools.16 However, already at the very end of the Republic, we find the first references, albeit in prose, to an entirely different representation, the learned male authority on magic.17P. Nigidius Figulus (praetor 58 bce) is the first known Roman author to have made use of works by magi; Anaxilaus of Larissa, whose work likewise combined accepted medical herb-lore with the marvellous, is supposed to have been exiled from Rome by Augustus.18 It is Pliny’s “historical” sketch of magic (HN 30.1-11) that for the first time fills in the background to this development, namely the growing awareness of educated Romans of the existence of an entire world of learned magic, invented by (Zoroaster) and spread into the Greek-speaking world by “Ostanes.”19 This magic drew equally upon medicine, divination (above all the most powerful kind, astrology), and religion. In this literature, the marvellous, miracle-working aspect of magic comes to the fore: in the hands of the truly wise and learned, there are no limits to human exploitation of the powers of nature, or, later, the powers of spirits.20 It is through this version of exotic, alluring, but always potentially dangerous magic that the male magus, mainly as practitioner of suspect methods of divination, above all astrology and necromancy, became firmly established in the discourse of the early Principate and in the terminology of later legal measures. In the mid-2nd century ce, Apuleius claims: “The man in the street thinks a magus is a man who, having direct communication with the immortal gods, is capable of performing anything he wishes by dint of the unbelievable power of incantation”; while a 4th-century declamation, influenced by the neo-Platonist (and Christian) claim that magic is performed by wicked daimones, puts it more pithily: “A magus is a man who can go against nature.”21 Physiognomists might even claim to read the baseness of these men in their faces.22

Real-World Roles and Practices

The term magic (and its equivalents in other modern European languages) is thus, given its vagueness, a suitable headword for referring to such a dynamic set of representations (the frequent call for better definitions is doomed to remain unanswered: like Greek magic, Roman magic is a mere modern convenience fostered by the cultural turn, not a neat theoretical construct). The main disadvantage is that it prompts many to imagine that magical practice must involve a special set of beliefs or state of mind, the remnants of the traditional opposition between magic and religion.23 It is less clear that the agentive magician is equally appropriate, partly because it has no female equivalent and partly because it nowadays evokes mere skill or marvel rather than a knowledge-practice. It is preferable to think in terms of an entire spectrum of marginalized religious (or ritual) specialists who competed both with one another and, to a limited degree, with those religious specialists, primarily diviners, recognized by the public religious sector. Religious specialism here does not imply full-time or professional employment, simply a claim to possess effective religious knowledge that might elsewhere in the religious field be dismissed as superstition or worse. The Chronicle of 354 ce claims that, in the aftermath of the Libo Drusus affair in 16 ce, 45 male and 85 female venenarii et malefici were arrested, apparently in Rome; if so, this would provide some indication of the number of better-known and more successful small-time religious specialists in the City with some regular clients and perhaps their own stalls or shops, who could easily be located by the vigiles, and so, if need be, pronounced “witches and sorcerers” by the authorities.24 The most regular accusation from the self-styled centre is that all expected some kind of payment, in cash or kind, for services rendered, as though civic priests did not demand choice cuts from sacrifices; on the other hand none of these practitioners would have understood their art as magic, and none of them would, in all probability, have used of themselves any of the terms (even saga or praecantrix) current in the literary tradition. Rather, they saw themselves as healers, as helpers and advisers in difficult social and emotional situations. Since no surviving author in Latin provides the kind of list of distinct practices or métiers we have for Greek, it is preferable to take three representative figures from a much wider range of distinguishable roles.25

The Wise-Woman, Saga, or Praecantrix

The derivation of saga from sagax (keen-witted) itself acknowledges the element of skill and knowledge required for these roles.26 Such women were primarily self-styled experts in the properties of medicinal herbs and animal parts and the preparation of protective or curative amulets.27 Each woman had her own rituals—some no doubt adapted from those practised by experts in the same family—for a range of other needs: divination of past, present, and future events, averting agrarian catastrophe, cases of male impotence and female infertility, marital strife, requests for altering feelings, recovering lost objects, protection against witchcraft.28 It was not only the ambivalence of herbs between cures and poisons but also their use in amatoria (attempts to alter feelings) that evoked the suspicion and anxieties that led to the easy passage from “wise woman” to “witch” betrayed by many occurrences in poetry of the word saga.29 Moreover, subjectively defined “protection against witchcraft” might well be interpreted as “aggressive magic” by the target: there are no neutral positions here. Finally, it is highly probable that that there were in fact more wise-men that wise-women, but it is telling that no word sagus was ever coined prior to magus, while the terms herbarius and medicus were apparently specialized in a positive sense, to mean an author of a book on medicinal herbs and animal parts.30 However, some such writers, including the grammarian Pamphilus of Alexandria, did include extensive references to the practices and claims of working herbalists; and Pliny comments sharply on their habit of collecting more of a plant than necessary in case they considered they had been paid too little; in which case they would allegedly replant a specimen, thus causing the complaint to recur.31

Hariolus, Haruspex, Coniector, Sortilegus, Sacrificulus, Vates

Although all these terms primarily denote diviners of one sort or another, the fact that scrying was merely one aspect, albeit a prominent one, of the range of abilities claimed by marginalized ritual-specialists, their availability as terms can properly be understood as explaining the failure to coin the word sagus. It is striking, at any rate, that their main forms are all masculine and that all are derogatory—that is, the marginality and cupidity of such figures is constantly insisted upon—yet not to the same extent as words for “witch” (saga, strix). It was these features, as well as the use of divination for diagnosis of illness and witchcraft, or the identification of a thief, that led to the informal assimilation of these roles to magic, first via the Marsi of the Abruzzi, and later as targets of occasional token expulsions of “magicians and diviners.”32 Despite the frequently expressed scepticism on the part of representatives of public religion, private divination continued to be an accepted part of negotiation with the other world until the Christian Empire; the differentiation between marginal male diviners and marginal wise-women employed gender asymmetry to distinguish between folly and danger. It was necromancy, the evocation of the dead from the underworld, together with astrology, that was imagined to be the typical form of magical divination.33

Virtually nothing is known of small-time ritual practitioners in the provinces outside Italy. Apart from the Druids and female scryers in Gaul and Britain, there were female seers in free Germania, of whom one, named Veleda, acquired some prominence in 69/70 ce.34 In the early medieval period, there were men in Gaul named tempestarii, who managed the weather, and we can assume similar specialists existed in the Roman period.35 The Councils relating to the western Church are full of references to persistent pagan/magical practices: canon 2 of the 16th Council of Toledo (693 ce), for example, lists auguratores ... seu praecantatores, “diviners or sorcerers” among the problems besetting the Church in the Visigothic countryside.36

The High Cultures of the Ancient Near-East

Although it will already have become obvious that, from the late 3rd century bce one can hardly speak of Roman magic independent of Greek traditions, it is in connection with the appearance, in the Latin-speaking West, of distinctively different forms of magical practice that the idea becomes really problematic (see magic, Greek), unless by “Roman” we mean simply “in Latin.” The earliest of these, arriving in late-Republican Rome via translations into Greek, were divinatory texts (especially relating to meteorological phenomena) and lists of plants and marvellous stones and their properties in the šikinšu and uruanna traditions of the Babylonian temple, which were used extensively by writers such as Pliny the Elder in compiling Latin versions of ta physika (Hellenistic works on the marvellous powers of Nature).37 At least some of this mediated lore, as well as simple astrological observations, will have been selectively appropriated into their practice by herbalists, doctors, and private persons. The second, which first appears in the Latin-speaking West at the very end of the 1st century ce, though mainly a century later, is Graeco-Egyptian knowledge-practice, a composite, dynamic blend of mainly Egyptian temple-lore with some Greek features that was slowly formulated in the late Ptolemaic and early Roman periods. Although Carthage and Hadrumetum (modern Sousse), in North Africa and Rome, are the only significant locations in the West where such practitioners established themselves, textual amulets on gold and silver sheet (also copper, bronze, lead) in this tradition, almost all in Greek, are found thinly scattered in the western Empire, from Britannia to Dacia.38 Another product of this tradition, intaglio amulets on semi-precious stone, mainly for ring-settings, have likewise been found here and there over the same area.39 Although it cannot be excluded that many such amulets were in fact manufactured in the eastern Mediterranean, they do indicate a certain demand in the Latin-speaking area for magical objects created in the Graeco-Egyptian tradition; and at least the curse-tablets connected with the circus and the amphitheatre, together with those for erotic and malign ends, imply the presence, albeit in small numbers, of practitioners trained in this tradition.40 Faint traces of this knowledge-practice, mainly the use of magical words, filtered very slowly into non-specialist cursing practice in the West.41

Amateur Efforts

At all times, people attempted to help themselves in such areas without the aid of experts. Everyone knew some simples and their properties, also their marvellous properties; they also knew, and exchanged, recipes and modest charms, such as the Elder Cato’s M. Porcius Cato, for dislocations or “Tarquenna’s” for gout.42 Everyone, likewise, knew simple forms of divination and methods of guarding against witchcraft and the evil eye, for example by crying praefiscinī!, spitting into one’s chest, putting up a satyr, or planting asphodel in the garden (Petron. Sat. 73.6; Plin. HN 28.35; 19.50; 20.108). The wearing of protective amulets was common, especially for children.43 The younger Pliny’s bête noire, M’. Aquilius Regulus (d. late 106/early 107 ce), used to paint colour round his eyes and shift a white plaster from eyebrow to eyebrow in court to avert the evil eye.44 After his humiliating erectile dysfunction, Polyaenus/Encolpius set about restoring his vigour in the privacy of his bathroom (Petron. Sat. 130.7f.). Cicero mentions that his friend Appius Claudius dabbled in necromancy (Tusc. 1.16.37; Div. 1.132). Caligula’s mental instability was said to have been due to an amatorium given him by his wife Caesonia.45 The best-documented forms of self-help, however, are the various genres of curse-tablets in the vernacular (i.e., non-Graeco-Egyptian) tradition, though once again, their principals did not, for the most part, view themselves as resorting to magic, let alone witchcraft, but as attempting to protect themselves from injustice, wrong-doing, threat, or enmity on the part of others—or reducing the odds in high-risk situations, even if they accepted that they were sailing close to the wind in depositing a tablet in a tomb. Here again, though, the principal’s “self-protection” would inevitably be viewed by the target as a magical attack: in such cases, there is no neutral position. Depositing tablets in tombs was thus also an insurance against discovery.

Representations of Magical Action

Essentially Foreign

In many traditions, the world over, so-called real magic is represented as invasive, as coming from outside. The earliest form of this strategy of othering at Rome is the settled belief that the two tribes of the Marsi in the Abruzzi, who were early allies of Rome in the second Samnite War (324–304 bce), were naturally endowed with the marvellous ability to split snakes in two by means of chants.46 By the late 1st century bce, they were said to be descended from the Greek (Circe) and to dispose of all kinds of magical powers.47 As descendants of Circe, they could be integrated into the dominant form of othering in the later Republic, the Greek mythological witch. In Greece, this figure took two forms, negative and comic, both of which were appropriated in the process of assimilating Greek literary culture to Roman requirements.48 The home of magic was thus re-located either to Thessaly in northern Greece or to the Caucasus, where Prometheus’ blood-ichor had given rise to a magical herb required by Medea.49 The sheer indifference to the precise location makes clear that these are heterotopias, real spaces in which different rules obtain, spaces inhabited by women who go around naked, with unbound hair, and ululating as they harvest malignant herbs with bronze sickles. At the same time, the reference to Greece set up parallel asymmetries, of political power, between Greece after Pydna (168 bce) and Rome, and of ritual legitimacy, between religion and magic.

What is lacking in this heterotopic scheme is a narrative about time. At some point in the Hellenistic period, perhaps as early as Hermippus of Smyrna (3rd century bce), but more likely later, someone composed a narrative, based upon the existence of numerous pseudonymous works in the tradition of ta physika, but of course also upon the etymology of the word magia, according to which magic was invented by the founder of Persian religion, Zoroaster (Plin. HN 30.3–4).50 Zoroaster’s dates had already, thanks to the Hellenistic authors of Persian histories, been fixed in relation to Greek events. Magic thereby acquired not only an old-new home, Persia, but above all what religion itself could never have, namely a history —in some versions, magic was claimed to be far older even than civilization, indeed five thousand years earlier than the Trojan War. Pseudo-historicization served three main ends: It enabled a variety of magical hot-spots, such as Thessaly, the Jews, Cyprus, and the Druids in Gaul and Britain, to be plotted along a space/time-scale and thus united in diversity into a single historical phenomenon.51 It recognized as the core of this phenomenon a variety of texts, written both by foreign experts, such as Ostanes, the Copt Apollobex, the Phoenician Dardanus, and Jewish Moses, and by Greeks—not just the Homeric Odyssey and Orpheus, but also (pseudonymous) works attributed to Pythagoras, Empedocles, Democritus, and Plato.52 And it brought magic right up to the present, indeed into the very palace of Nero at Rome (Plin. HN 30.14).53 In this narrative, magic is male, learned and occultic: witches and humble practitioners are of no interest and by implication belong to the unsophisticated world of the credulous unwashed.54 The sophist Lucian claims that learned persons of this kind, educated Greeks, might be admitted to wealthy Roman households as experts in magic and divination.55

Fictive Witchcraft

As pointed out above, images of the witch at Rome were, to an unknowable but surely considerable extent, founded upon Hellenistic Greek stereotypes absorbed through translations, imitations, and intertexts. In his Erotopaegnia, Laevius (1st century bce) sent up this dependence in six clever iambic dimeters on the fanciful mixing of an amatorium (frg. 27):

philtra omnia undique eruunt; / antipathes illud quaeritur, / trochiscili, unges, taeniae, / radiculae, herbae, surculi, / saurae inlices bicodulae, hin/nientium dulcedines.56

All about, they grub up every aphrodisiac (herb); look out for that special antidote (to indifference), whirly-wheels, “wrynecks,” ribbons, rootlets, leaves, twiglets, split-tailed lizard enticements, allurements of whinnyers.57

The first words of each of the first three lines, and the fifth, are Greek or quasi-Greek, underscoring the cultural origin of the topos, followed by three lines mainly in Latin, the last two of which are playful gibberish, reminding us of Laevius’ most brilliant surviving phrase vítuperones súbducti supércili / cárptores, “critics—raised-eyebrow carpers” (frg. 7).

Fictive witchcraft in Latin was a rhetorical construction which no sophisticated reader—that is, no reader—was ever expected to take literally. Moreover, the rag-bag of motifs bore no close relation to any plausible set of beliefs even among the credulous rural inhabitants among whom such images were supposedly current.58 As illustration, we may take a passage Ovid put into the mouth of Hypsipyle, complaining that Jason has abandoned her for Medea (Her. 6.81–96). Here we find in quick succession:

Medea’s herbal knowledge and incantation alone have caused Jason to fall in love.

Medea brings down the moon and eclipses the sun.

Medea stops running waters from flowing.

Medea can move rocks and trees by magic.

Medea, in dishabille, collects bones from graveyard pyres while they are still warm.

Medea “catches” her victims from a distance by making poppets and driving pins into them.

The routinized nature of such a list, shifting as it does between love-magic, reversal of the order of the cosmos and nature, night-witch fantasies, and urban myth, as well as its overall irrelevance to the fictive situation, indicate that this game is really a dual one of “spot the intertexts” and “spot the omissions.” The conventional, frequently hyperbolic character of such claims makes it impossible to take them as evidence for specific historical beliefs: as an amalgam of items from previous texts they are a purely imaginary product, insulated from observable reality, which explains why such images are found, albeit increasingly infrequently, even in late antiquity.59 It is rather the literary context that indicates how the reader is supposed to take them. Roughly speaking, we can distinguish two groups of these: a) erotic contexts in Latin elegiac poetry, where they are usually undercut by the claim that human passion is far stronger than magic, or, in lyric and satire, by turning an apparently grisly situation into comedy; and b) necromantic scenarios, in which the reader is assumed to relish the accumulation of ghoulish details—even here, however, as often as not the consultation brings the enquirer no reliable information.60

We can nevertheless discern a few failed attempts to add to the stereotyped inventory and adaptations to the specific context. As for the first, Vergil found somewhere a scheme that included transformation into a werewolf, and included it in his brief allusion to Moeris, a male wise-man (Ecl. 8.95–99).61 Despite Vergil’s authority, however, Propertius alone thereafter alludes to the werewolf, but changes its sex to female (4.5.14). In relation to adaptation, Ovid brilliantly evokes not just the trope of wax poppets and pins to account for the bedswerving narrator’s erectile dysfunction in Am. 3.7.27–38, but also the themes of Italic agrarian malign magic—withered crops, dried-out wells, acorns, grapes, and apples dropping too early; in Propertius, a chagrined speaker dismisses all the babble about love-magic in a single alliterative pentameter, et bibere e tota toxica Thessalia, “you drink down the poisons of all Thessaly,” in warning off a madly jealous but inexperienced competitor (1.5.6). In the 2nd century ce, Apuleius transposes the horrible strix (were-owl) trope into a comic-erotic scenario, complete with the motifs of cosmic reversal and the use of magical substances, when his Thessalian witch Pamphile falls desperately in love with a pretty Boeotian and turns herself into an owl instead of taking a taxi (Met. 3.15–16).62 There are also cases of complete inability to control the material, such as [Vergil] Ciris 369–377. The Flavian-Trajanic vogue for ghoulish flourishes learned in the rhetorical schools is clear from Tacitus’ entirely fictional account of the death of Germanicus and the younger Pliny’s ghost-story.63 All this is typical of the freedom fiction provides from “the pragmatic dimension so essential to real-life situations.”64

Witchcraft Narratives

The lists compiled for the creation of fictive witchcraft had their basis (at some remove) in innumerable stories about strange events, misfortunes, and accusations. However, the surviving literature in Latin contains very few usable narratives.65 Three examples indicate the restricted tone of these: apart from Apuleius’ Apologia (158/159 ce), a much re-edited speech for the defence in a case of alleged administration of an amatorium, no classic accusation-narratives relating to the Latin-speaking West survive.66 The first is the sole known case of a prosecution for “crop-destruction” under Twelve Tables 8.4, which took place in the period of turmoil in Italy after the Second Punic War and is known solely because it was written up by the historian L. Calpurnius Piso (cos. 133 bce) as an example of Roman good sense and virtue.67 It concerns a freedman, C. Furius Cresimus who was unusually prosperous and brought to trial before the aedile for “lifting crops”; in his defence he brought his farm tools and fat oxen into the court, declared them to be his only veneficia, witchcraft—and was duly acquitted.68 The other two focus on striges, night-witches, and both fall into the ancient equivalent of the horror-comedy genre. In Petronius’ Satyricon (say 40–55 ce), Trimalchio tells a story of how, when he was a boy, a young slave died, and the striges began to howl for his vitals. A giant house-slave went out to attack them with a sword, but came back black and blue, and later went crazy and died; while everyone’s attention was thus distracted, the striges removed the dead boy’s vitals and replaced them with straw.69 In another of Apuleius’ stories, a man named Socrates is attacked and relieved of his heart by striges in Thessaly, but magically lives on with a sponge stuck in his throat to stop the bleeding. One of the witches, Panthia, utters a curse: Heus tu, spongia, cave in mari nata per fluvium transeas, “Sponge, you come from the sea; take care not to cross a river.” When Socrates and his friend get to a stream, he bends down to drink and at that moment the sponge drops out of his throat and he falls stone-dead.70 Again, no reader is supposed to believe any of this: these are parodies of “old women’s stories” (fabulae aniles), which figure the cultural gulf between the mass of the ignorant-superstitious and cultivated men of good sense.

Explaining Magic

Although we know of relatively numerous attempts by Greek authors to account for magical effects, there are only two in Latin from the Principate, both quite interesting.71 The earlier is a passage in Lucan’s Grand Guignol staging of Erictho’s powers, in which the implied narrator debates whether the gods are compelled to fulfil magical desires by incantation and potions (cantus herbasque), or do so willingly because magic amounts to another religion? Or are they induced to act by threats? Does incantation affect all gods, or is there just one supreme divinity who is affected and then forces the others to act? (6.492–499, cf.443–451). This the sole passage in Classical Latin that clearly seems to echo Graeco-Egyptian magical practice, of which threats, legitimation through the temple, speculation about voces magicae and the idea of a supreme cosmic deity were standard features. The other is Apuleius’ post-Platonist conception of daimones (though he avoids using the term), which he invokes as a possible explanation for divinationes cunctas et magorum miracula, “all the forecasts and wonders of magi” (Apol. 43.2): transformed into demons, these daimones later became the standard Christian explanation of the power of magic.72 In the same connection, Apuleius offers a naturalistic hypothesis to account for how a boy medium might be inspired by divine power. Elsewhere in the same speech, he invokes another Platonic idea, of a supreme divinity, totius rerum naturae causa et ratio et origo initialis, “the source, guiding principle, and prime mover of the whole of nature,” knowable through philosophy and religion but also through “true” magic, shorn of its negative associations.73 Neither of these attempts to subsume magic into a larger speculative framework enjoyed wide circulation: a major effect of the creation of an abstract term, magia, was to unify the practice it supposedly denoted and create of it an anti-institution.


Although the classification of cursing as magic remains problematic, Pliny evidently understood it as a form of terrorization: defigi quidem diris deprecationibus nemo non metuit, “there is no one who is not afraid of being ‘caught’ by a curse” (HN 28.19).74 Cursing as an open act of requiting a perceived wrong in a private matter naturally involved invoking deities as witnesses or agents75 and became a minor poetic genre already in the Hellenistic period—Ovid’s baroque Ibis is however the only surviving complete example in either language.76 The slow adoption from the mid-Republic of the technique of using lead sheet for a written curse, which exploited an analogy with the permanent dedication in a temple, had the disadvantage that the most effective addressees were felt be underworld deities, so that tombs were the preferred location for deposition—which in practice meant disturbing the dead without being discovered.77 This in turn tended to shift the written curse, as far as outsiders were concerned, into the category “malign magic,” despite the fact that subjectively the aim of the principals was mainly to avert a perceived or threatened injustice or right a wrong.78

Rather fewer than 630 Latin curse-tablets on lead have been found: the most recent survey lists 618 published texts including those in Oscan, Etruscan and Gaulish.79 About one-third of the total comes from just three temple-sites, Bath and Uley in southern Britannia, and Mainz in Germania Superior. Virtually all of those from North Africa, and many from Rome, were written by professional practitioners in the Graeco-Egyptian tradition, including many for circus races and beast-shows.80 Given that the time-range extends from the 2nd century bce to the late 4th century ce, and even allowing for massive losses, it is clear that writing vernacular curse-tablets was in fact, even in Italy, let alone in the Latin-speaking western and Danubian provinces, rather uncommon. Opinions differ over whether the original impulse came from Magna Graecia via Oscan or from the eastern Mediterranean via Delos. Some of the earliest examples in Latin are actually from Hispania.81 Given the small numbers involved and the fact that every curse was a personal composition and then was buried, it is best not to classify them by putative topic but to arrange the texts in terms of the principal’s ability to employ appropriate religious knowledge.82 Many curse-tablets in Latin have no divine addressee and consist simply of names; in a number of cases the principal could not write but only knew that the practice required the burial of a lead tablet (no doubt accompanied by an oral curse) in a tomb or under water or at an otherwise sacral spot. The majority of vernacular curse-tablets, however, can be plotted on a continuum ranging between a sketchy idea of how to write such an appeal, for example, by naming a divinity and using a word such as demando or rogo (“I request”) or devoveo (“I assign [to the underworld]”) up to a tiny number of really accomplished examples, from Rome and the temple of Mater Magna in Mainz.83 The background and thus the aim of most curses is not made explicit; of those cases in which the aim can be made out, one group attempts to influence an up-coming court-case, another concerns slaves involved in intrigues and mutual accusations in large households; a small number attempts to obtain divine help with difficulties in an intimate relationship.84 Thanks mainly to the finds at Bath and Uley, a very large group relates to thefts.85 The recently introduced term “prayers for justice,” though perhaps useful for the eastern Mediterranean, is hardly applicable to the Latin West: the explicit claim that one has been wronged is merely a different strategy of cursing, which helped relieve the principal of anxiety about whether he or she might not in fact be employing magic in writing a defixio—the shared use of lead sheet is decisive.86


Wording, date, and context of the Twelve-Table rules are all quite uncertain. If genuinely archaic, they may have been introduced in response to a spasm of accusations in the context of a period of agrarian distress. In the early-modern period, the introduction of witchcraft legislation encouraged further accusations by providing formal recognition of the reality of malign magic, but we do not know how far this was also the case at Rome. In the period after the war against Hannibal, we hear of a series of mass executions (2,000 in one case, over 3,000 in another) for veneficium in 184, 180–179, and 153 bce, conducted by Roman praetors in and around Rome.87 It is no coincidence that these events took place shortly after the similarly radical repression of the Italic cult of Dionysus-Bacchus (186 bce) Bacchanalia: in each case a policing measure developed its own uncontrollable dynamic. The main legislation on magic through the late Republic and Principate (there never was a lex Iulia on this topic) was the fifth chapter of the law on murder and poisoning passed by Sulla L. Cornelius Sulla Felix; insofar as this was applied to “witchcraft,” it assumed that it was primarily a matter of substances (venenum).88 This meant that the primary focus in trials for malign magic or witchcraft was usually upon the evidence for purchase or administration of a harmful substance (malum venenum), though the place of amatoria here ensured that strange behaviour might be taken as evidence that such a substance had been administered. It has been argued that, already in the Julio-Claudian period, malign magic was defined as the performance of impia nocturnave sacra, “wicked or nocturnal rites,” specifying the use of incantation, cursing, and binding either by a principal or through an agent.89 The likelihood of such specific language being employed so early rather than in the late 3rd century ce is small; rather, a succession of senatus consulta and imperial rescripts made it possible for delatores and accusatores (professional plaintiffs) to use a looser notion of magic—understood primarily as illicit divination in cases of alleged lese majesty by members of the Roman elite or by prophets forecasting the success of pretenders to the throne—impia sacra was a capacious category.90 The sheer vagueness of the notion of magic in the mid-2nd century is clear from the speech of Apuleius in his defence on a charge under the lex Cornelia: having no evidence for the administration of an amatorium, the formal charge, the prosecutors cast about for circumstantial evidence that might suggest that Apuleius was the kind of person who might have done such a thing, for example by using a boy as a medium—here again divination is assimilated to magic (Apul. Apol. 42.4–5).91 By the later 2nd century, however, it seems probable that there was pressure to make the very knowledge of how to perform magical rituals itself a crime, knowledge that was by the late 3rd century understood as the possession of recipes or a formulary.92 This is the focus of Constantine’s rescript of c. 321 ce (CTh 9.16.3), which at the same time explicitly permits weather magic to protect crops. Later legislation by his sons, and by their successors in the Christian Empire, however, was directed against private divination and magic, and this clause disappears.93 These rescripts were provoked by particular incidents, at least some of them on the fringes of the court itself, and they betray an apparent intensification of anxiety about control of divination.94 However, there is reason to believe that, for the most part, accusations of magic (as was the case earlier) could be easily parried by the rhetorically accomplished, whatever the case lower down the social scale.95 Gossip and rumour played a central role in arousing suspicion and inventing so-called facts. Apuleius recounts a tongue-in-cheek story about a witch who was supposed to have prevented a pregnant woman from giving birth to her baby for eight years—and would have been stoned by an angry crowd if she had not locked them all up in their houses by means of her necromantic rites (Met. 1.9–10). Joking aside, that was probably the fate of other small-time ritual specialists who, for one reason or another, exhausted the tolerance of their host community. But that fate remained exceptional: people needed their small-time specialists; if and when a group or a crowd could be induced to no longer tolerate a particular individual, they called them veneficus or maga and used “justice,” either a public accusation or informal lynching, to dispose of them.


If one simply reads Augustan and Julio-Claudian literature, one get the impression that Graeco-Italic witches were scurrying about all over the place; but this was simply a literary fashion that remotely had to do with the anxieties about the gods’ moral judgement about the civil war, but more immediately to do with the themes and characters of elegiac poetry. So far as is known, this kind of witch disappears as a literary figure during the course of the first century ce, having bored sophisticated readers, but was replaced by necromantic scenarios, body-parts, and ghost stories in the period between Nero and Trajan. We might see these as the legitimate nightmare of an elite class enjoying immense privileges at the price of constant exposure to the vagaries of absolutism.

The value of a pseudo-history of magic invading from outside, representing an anti-world simultaneously immensely powerful and utterly ineffective, lay in its graphic inversion of the positive norms that allegedly underpinned this increasingly unstable political formation, constantly in need of the divine legitimation expressed in the form of military victory. Magic had to be powerful enough to warrant fear, but not powerful enough to seriously threaten the legitimate order.96 To underscore the claim of the centre and its peripheral dependencies—the civic elites—to possess legitimate religious knowledge, it was advisable occasionally to execute or expel supposed representatives of illegitimate knowledge (nothing being more illegitimate than the knowledge of how to kill by ritual means, or cause unfruitfulness). Private accusations of magical practice helped in turn to objectify such dangers. The moral ambiguity of the claims of wise-men and wise-women, and of a variety of diviners—to say nothing of professional adepts of ceremonial Graeco-Egyptian magic—played perfectly into this scheme: they might be useful, but they were also potentially dangerous. The primacy of magic in this scenario ensured its partial assimilation of divination and astrology, which were useful but also potentially threatening, politically. This larger foil makes it insufficient to speak simply of a discourse of magic in the Empire.

Discussion of the Literature

The humanist revival of the study of Roman law, together with the early modern witch-hunts, have meant that the criminalization of artes magicae and the early Christian reinterpretation of magic traditionally dominated the discussion of magic at Rome.97 For their part, the German neo-humanists were frankly bored by passages in the Augustan poets devoted to witchcraft and magic, which offended their sense of poetic value. In the later 19th century, the ethnographic reconstruction of magic as primitive and irrational left a permanent mark on the field until the 1970s—much longer in some circles; the neo-Romantic vogue for astral bodies and apparitions made it easy for the guardians of the Classical inheritance to ignore the entire matter. Moreover, with publication of the Greek magical papyri, in the period 1880–1934, the study of ancient magic, insofar as it was undertaken at all, focused almost exclusively on Greek texts—with the exception of the curse-tablets in Latin found in the 1890s at Hadrumetum (Sousse) in Tunisia by French army officers on their colonizing mission.98 The brilliant essay on magic by Marcel Mauss and Henri Hubert (1902–1903) meant that, in France at any rate, the general topic continued to be taken seriously.99 Alfred Ernout’s work on Bks. 28 and 30 of Pliny’s Historia naturalis in the Budé series, published 1962–1963, prompted him to suggest that Anne-Marie Tupet work on magical themes in the Latin poets for her thèse d’état, which was the first serious philological study. Her premature death, however, prevented her from continuing beyond the late Republic.100 This coincided with the Annaliste discovery of both mentalités and early-modern witchcraft, which quickly produced the first serious study of magic as a social phenomenon in Rome, by Rafaella Garosi, who was killed soon afterwards by a terrorist bomb.101 The vogue for Lévi-Straussian structuralism in the 1970s, while tending to give theoretical underpinning to the traditional binarism between religion and magic, encouraged analysis of socially produced boundary formation.102 It was, however, the symbolic turn as a whole, especially the work of Foucault, that helped institutionalize cultural studies, a shift within the historiographic paradigm that was the necessary preliminary to the recognition of magic as a significant cultural issue rather than a small philological niche. In the early 1990s, John Scheid, then a directeur at the École pratique, Ve section, and a pupil of Robert Schilling, invited Fritz Graf to give a course of seminars at the Collège de France, which became the first modern book on (Graeco-)Roman magic. At the same time, Matthew Dickie was working on a volume also devoted to magic in both cultures that, twenty years later, still stands as a major achievement.

As often happens, several factors combined to stimulate further work. There were spectacular finds of curse-tablets at Bath (1979/1980, published 1988) and Uley, in Britannia (1977–1979, partly published 1993), at Mainz in Germania Superior (1999–2001, fully published 2012), and the Fountain of Anna Perenna in Rome (1999, published by 2010).103 L’Année épigraphique was reformed under Mireille Corbier in 1991. Senior scholars organised major conferences, such as those hosted by Marvin Meyer and Paul Mirecki at the University of Kansas at Lawrence in 1992 and at Chapman in 1998, or by Alain Moreau and Jean-Claude Turpin at the Séminaire d’Étude des Mentalités Antiques at Montpellier in March 1999.104 Book series such as Brill’s Religions in the Graeco-Roman World, and gender studies and women’s history, of which the first sign was Elisabeth Wallinger’s Hekates Töchter, had significant impacts.105 Morton Smith’s claim that Jesus was a magician has produced a certain interest in anciet magic, albeit primairly Greek.106 Religious Studies departments, especially in the United States, were strengthened by the arrival of PhD students looking for uncultivated fields, and by BA students demanding new types of courses.107

As a topic, however, with the exception of law, Roman magic attracts less attention than Greek.108 Fraser’s “Roman Antiquity” and a handbook piece by James Rives are good brief general sketches to start with.109 The most useful study of magical topoi in Augustan and 1st century ce literature is Fauth’s Carmen magicum. On the complex relation between belief, religion, and literature, there is much to be learned from Feeney, Literature and Religion, particularly on the dangers of reading high literature as any form of social history.110 On small-time religious specialists in the late Republic and early Principate, albeit under the rubric of “magicians” and “sorceresses,” see Dickie.111 More neutral accounts of wise-men and wise-women, under the rubric of “low medicine,” are also available.112 There are several useful studies of the language of charms and curse-tablets.113 Literary texts have been scrutinized for their representations specifically of women.114 The “material turn” has opened up alternatives to written sources, although the interpretation of archaeological finds without any verbal contextualization tends to be problematic and sometimes verges on arbitrary.115

The major current interest is in the re-edition and re-analysis of the curse-tablets and the possibilities they offer for a modicum of social history.116 Although the body has been a significant focus in Religious Studies for some time now, it has hitherto provoked little work in this field, and there is much still to be done.117 Emotion too remains virtually unexplored.118 A further theoretical issue is the extent to which appeal to magic, as well as to normative religious practice, should be viewed as enlarging the agency of individuals, not only by expanding options, but also by providing a sense of empowerment in contingencies.119 Although this article is not intended to cover early Christianity, it seems clear that, the claims of Christian apologists and bishops notwithstanding, magical practices continued unabated throughout late antiquity and the early medieval period.120

Links to Digital Materials

The data-base of curse-tablets in Greek and Latin TheDeMa (Thesaurus Defixionum Magdeburgensis) (see magic, Greek) is currently unavailable, since the department of History at Magdeburg has been closed down; the process of transferring it to the University of Hamburg under the supervision of Prof. Werner Rieß has encountered technical problems, which are, however, expected to be resolved by the end of 2020. However all of the readings of the Latin texts by Amina Kropp, Magische Sprachverwendung in vulgärlateinischen Fluchtafeln (defixiones) (Tübingen, Germany: Gunter Narr, 2008) can be found via the Clauss/Slaby Epigraphic Data-Base, assuming one knows the AE number or some other reference number (e.g., Tab. def. Audollent), or a specific word.

The great majority of Latin texts mentioned here can be found on the Internet:


  • Clerc, Jean-Benoît. Homines magici: Étude sur la sorcellerie et la magie dans la société romaine impériale. Bern, Switzerland: Peter Lang, 1995.
  • Dickie, Matthew W.Magic and Magicians in the Greco-Roman World. London: Routledge, 2001.
  • Fauth, Wolfgang. Carmen magicum. Frankfurt: Peter Lang, 1999.
  • Fraser, Kyle A. “Roman Antiquity: The Imperial Period.” In Magic and Witchcraft in the West. Edited by David J. Collins, 115–147. Cambridge, UK: Cambridge University Press, 2015.
  • Graf, Fritz. Magic in the Ancient World. Cambridge, MA: Harvard University Press, 1997.
  • Gordon, Richard L. “A Babel of Voices: Styling Malign Magic in the Roman World,” Magic, Ritual and Witchcraft 14, no. 2 (2019): 155–188.
  • Gordon, Richard L., and Francisco Marco Simón, eds. Magical Practice in the Latin West. Leiden, The Netherlands: Brill, 2010.
  • Kahlos, Maijastina. “The Early Church.” In Magic and Witchcraft in the West. Edited by David J. Collins, 148–182. Cambridge, UK: Cambridge University Press, 2015.
  • Niedermayer, Matthias. Die Magie in den römischen Strafrechtsfällen. Gutenberg, Germany: Computus, 2017.
  • Ogden, Daniel, ed. Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds. New York: Oxford University Press, 2002.
  • Parker, Adam, and Stuart McKie, eds. Material Approaches to Roman Magic. Oxford: Oxbow, 2018.
  • Piranomonte, Marina, and Francisco Marco Simón, eds. Contesti magici/Contestos mágicos. Rome: De Luca, 2012.
  • Sánchez Natalías, Celia. Sylloge of Defixiones from the Roman West. Oxford: Archaeopress, 2020.
  • Stratton, Kimberley B., and Dayna S. Kalleres, eds. Daughters of Hecate. New York: Oxford University Press, 2016.
  • Urbanová, Daniela. Latin Curse Tablets of the Roman Empire. Innsbruck, Austria: Innsbrucker Beiträge zur Kulturwissenschaft, 2018.


  • 1. The earliest extant occurrences (in the pl.) are Cic., Brut. 217 (c. 46 bce, but referring to the elder C. Scribonius Curio [cos. 76]), Hor. Sat. 1.8.19 (c. 35–33 bce); Tib. 1.2.60; 8.23; Ov. Ars 2.102; Rem. 290; and Met. 14.357; the phrase clearly lies behind Verg. Ecl. 8.64–67. Venenum as “herb for use in magical practice”: Plaut. Pseud. 870 (of Medea); Pacuv. trag. 401 Ribbeck; and Cic. Orat. 229 (again on Curio).

  • 2. See e.g., Michael H. Crawford, ed. Roman Statutes (London: Institute of Classical Studies 1996), 2: 677–679 on VIII.1 and 682–684 on VIII.4; also James B. Rives, “Magic in the XII Tables Revisited,” Classical Quarterly 52 (2002): 270–290; and Matthias Niedermayer, Die Magie in den römischen Strafrechtsfällen (Gutenberg: Computus, 2017), 43–49.

  • 3. Excantare: Plaut. Bacch. ap. Nonius Marc, p.102: credo cuivis excantare cor potes (metaphorically, of a prostitute’s wiles), “I believe you can bewitch anyone you like”; Hor. Epod. 5.45: quae sidera excantata voce … lunamque caelo deripit (of the witch Sagana, who) “bewitches the stars and hauls the moon down from the sky”; and Sen. QNat 4.7.2: ne quis alienos fructus excantassit, “(forbids) anyone to bewitch others’ crops” (interpreted as weather magic). Incantare: cf. Hor. Sat. 1.8.49f. incantata . . . vincula (ironic), “magical substances.” Cantus, cantio and cantamen are the relevant substantives. The rare word incantamentum, evidently a calque on incantare, is used by Plin. HN 28.10, in discussing whether prayers and charms have specific effects. The agentive incantator/rix, is not found until after c. 200 ce. On Latin words for “incantation,” see Eli Edward Burriss, “The Terminology of Witchcraft,” Classical Philology 31 (1936): 137–145.

  • 4. Transferring crops: Verg. Ecl.8.99 (Moeris); Tib. 8.19; Ov. Rem. 255; Petr. Sat. 134.12; on weather (excluding the heavenly bodies and running waters): Tib. 1.2.49–50; Ov. Am. 1.8.9–12; Met. 7.202; 14.369–370; Sen. Herc. O. 454–457; Med. 754; 759–761; 766–768; and Luc. 6.467–469. The classic study is George M. Foster, “Peasant Society and the Image of the Limited Good,” American Anthropologist 67 (1965): 293–315.

  • 5. Plaut. Mil. 692f.: da quod dem quinquatrubus/praecantrici, coniectrici, hariolae atque haruspicae, “give me something to give the wise-woman, the fortune-teller, the soothsayer, and the diviner because it’s Minerva’s day” (put into the mouth of a wheedling wife who constantly duns her husband for money). Some mss. read praecantatrix. OLD translates “enchantress, witch,” which is too strong for the context. On agentives, see Maxwell T. Paule, “Quae saga, quis magus: On the Vocabulary of the Roman Witch,” Classical Quarterly 64 (2014): 745–757.

  • 6. Turpilius frg. 8: non ago hoc per sagam pretio conductam, “I do not act here as a wise-woman for a reward.” Nonius Marcellus 23.4 took Lucil. frg. 271 M. = 7.5 Charpin ut saga et bona conciliatrix, “like a wise-woman and gifted go-between,” as an earlier case, which is possible (so OLD), though the word might equally be an adjective, “a clever and gifted go-between.”

  • 7. The first surviving attestations of a substantive veneficus/a (albeit as a term of abuse) are as early as Plautus, e.g., Rud. 1112; Epid. 221; and Ter. Eunuch. 825.

  • 8. Venenare: Plaut. Rud. 1302; venenum/veneficium as “witchcraft,” e.g., Cic. Off. 3.76 (of getting someone to change his will); Ov. Rem. am. 251: ista veneficii vetus est via (so as to cause someone to fall in love), “that’s an old sorcery trick”; Petron. Sat. 128.2: veneficio contactus sum (to account for Polyaenus/Encolpius’ impotence), “I’ve been bewitched,” cf. 138.7; Plin. HN 29.107 (remedies for mange/baldness caused bewitchment); also 28.102 (hyena kidneys as remedy against fertility impeded veneficio, by witchcraft) etc.

  • 9. Plin. HN 28.19 incantamentorum amatoria imitatio, describing his lost imitation of Theocritus’ Pharmaceutria (Id. 2). Apart from a deliberately exotic list by Laevius (see § Fictive witchcraft), the loan-word philtrum first occurs in Ov. Am. 2.105, and even there for the sake of its initial plosive within the alliteration of p (profuerint pallentia … puellis).

  • 10. Maleficus as agentive noun: e.g., Firm. Mat. Math. 6.17.3; Jer. Vit. Hilar. 11.5 and 12 Leclerc-Morales; Comm. in Dan. 1.2a.2 (CCSL 1.5 p.783–734); CTh 9.16.4 and 6. As adj., applied to night-witches (striges): Fest. p.41425 L.; to men: Apul. Apol. 51.10; but apparently as an agentive already in Apol. Met. 6.16: magna . . . et alta prorsus malefica, “a truly accomplished sorceress.”

  • 11. E.g., HN 30.7, 10, and 11; cf. 28.188: negari magices sacrificiorum usus, “(people with freckles) are debarred from practising the rituals of magic” (cf. 30.16); also Verg. Ecl. 8.66: magicis . . . sacris; Aen. 4.493: magicas … artis; Hor. Ep. 2.2.208: terrores magicos. Catull. 92 proves that the sense “Persian priest” was still available in the Late Republic; cf. Apol. Apol. 25. For a thorough survey of mag- words in Latin, based on the Thesaurus Linguae Latinae, see James Rives, “Magus and Its Cognates in Classical Latin,” in Magical Practice in the Latin West, ed. Richard L. Gordon and Francisco Marco Simón (Leiden, The Netherlands: Brill, 2010), 53–77.

  • 12. The word maga however, always extremely uncommon, first appears in Sen. Herc. O. 523 and 526 and (a century later) Apul. Met. 2.5.

  • 13. Perhaps even as early as Livius Andronicus’ translation of the Odyssey in Saturnians in the 240s. On the shift of attitude implied by this assimilation, see the remarks of Matthew W. Dickie, Magic and Magicians in the Greco-Roman World (London: Routledge, 2001), 124–128.

  • 14. Cf. Hugh Parry, Thelxis: Magic and Imagination in Greek Poetry (Lanham, MD: University Press of America, 1992); and Richard Buxton, Forms of Astonishment (Oxford: Oxford University Press, 2009) 29–48, 157–209. However, Sueton. περὶ δυσφήμων λέξεων‎, s.v. Τελχῖνες‎ (Telchines) averred that some derived the name of these magical figures from θέλγειν‎ (thelgein): C. Blinkenberg, “Rhodische Urvölker,” Hermes 50 (1915): 271–303 (277–278).

  • 15. Plaut. Amph. 1043; Verg. Ecl. 8.96, 98 (Moiris is recorded in LGPN 2 p. 319 [Athens, IVa]); Hor. Carm. 1.27.4.

  • 16. The elder Seneca mentions that several controversiae (practice declamations) by Ovid were still known in his day (Contr. 2.2.1 and 8; 9; 12). Quint. Inst. 7.8.2. gives as a specimen of syllogistic reasoning a case involving a woman who gave her violent husband an amatorium and then left him; when she refused to return, the husband hanged himself; is she guilty under the lex Cornelia? Similar issues crop up at 7.3.7: an carmina magorum veneficia?, “are magicians’ incantations “venena” (i.e., under the lex Cornelia)?”; 7.3.10; 30; 8.5.31; cf. [Quint.] Decl. min. §385.

  • 17. Cf. Matthew W. Dickie, “The Learned Magician and the Collection and Transmission of Magical Lore,” in David R. Jordan, Hugo Montgomery, and Einar Thomassen, eds., The World of Ancient Magic (Bergen, Norway: Norwegian Institute in Athens, 1999) 163–194. For the story of the Persian magus Zopyrus applying physiognomic forecasting to Socrates in Athens, see e.g., Cic. Tusc. 4.37.81; Fat. 5.10; Heracleides Ponticus (IVa) wrote a dialogue Zôroastrês featuring a Persian magus at the court of Gelon of Syracuse (Strab. 2.3.4, 98C).

  • 18. Nigidius: frgs. 67; 127–128 Swoboda; Anaxilaus: compare Plin. HN 25.154 with 28.181; exile: Jerome, Chron. Ol. 188.1; and cf. Dickie, Magic and Magicians, 170–175.

  • 19. A similar account is implied by Apul. Apol. 90.6; Arnob. Adv. nat. 1.52.1. However Varro evidently knew of a parallel narrative relating to divination: Quod genus divinationis (hydromancy [bowl divination]) … Varro a Persis dicit allatum, “Varro tells us that this kind of divination was introduced from Persia”; if blood is added to the water, one can evoke the dead (necromancy): RD frg. IV Book I (I p.36 Cardauns) ap. August. De civ. D. 7.35. Note also Varro’s account of the 160 prophetic verses uttered by a boy medium in Tralleis about the outcome of the Mithridatic War, which was inspired by a vision of Hermes scried through the water in the bowl (Apul. Apol. 42).

  • 20. Cf. Philippe Mudry, “Mirabilia et magica: Essai de définition dans l’Histoire naturelle de Pline l’Ancien,” in Olivier Bianchi and Olivier Thévenaz ed., Mirabilia. Conceptions et représentations de l’extraordinaire dans le monde antique (Bern, Switzerland: Peter Lang, 2004) 253–264.

  • 21. More vulgari eum isti proprie magum existimant, qui communione loquendi cum deis immortalibus ad omnia quae velit incredibili[a] quadam vi canaminum polleat . . . Apul. Apol. 26.6; . . . homo, cuius ars est ire contra naturam: [Quint.] Decl. mai. 10.5.

  • 22. Polemon, De physiog. (1 p. 13014–13016 Forster, his tr. from Arabic): peritum fuisse magiae et venenorum torporem et mortem adferentium et vilissimae cuiusque artis, “(a man from Lydia who was) an expert in magic, in potions causing paralysis and death, and every sort of utterly despicable knowhow.” Cf. in astrology, Firm. Mat. Math. 8.30.11: . . . erit marsus, magus, biothanatus, “will be a Marsian [i.e., seer], a magus, (or) die a premature death by violence”; and cf. Ptolem. Apotel. 3.14.32 Hübner.

  • 23. For a defence of “homeostatic cluster definitions” in preference to the commoner “polythetic definitions” (let alone “monothetic” or essentialist definitions), see Michael Stausberg and Mark Q. Gardiner, “Definition,” in Stausberg and Steven Engler, eds., The Oxford Handbook of the Study of Religion (New York: Oxford University Press) 9–32.

  • 24. See MGH 9 [Chron. min., 1] 14526–14528.

  • 25. E.g., Iulius Pollux Onom. 7.188 (2 p.103f. Bethe), s.v. ἀγύρται‎.

  • 26. Explicit too in the expression plus scia used by Petron. Sat. 63.9 (in a negative sense).

  • 27. Cf. Plin. HN 25.10: durat tamen tradita persuasio in magna parte vulgi veneficiis et herbis id cogi eamque unam feminarum scientiam praevalere, “many common people stubbornly believe that (eclipses) are caused by witchcraft, and that herbal lore is the sole area of knowledge in which women are dominant.”

  • 28. Perhaps the best idea of their activities is Petronius’ amusingly dismissive account of the old saga Oenothoe in Sat. 131, 133–138. The association of sagae with erotic magic is already clear in Lucil. 271 M. with Nonius’ definition: sunt feminae ad lubidinem virorum indagatrices, “(they are) women who are experts in male sexual desires.”

  • 29. E.g., Luc. 6.457–458: mens hausti nulla sanie polluta veneni, excantata perit, “even if the mind is not attacked by drinking an amatorium, it is overwhelmed by incantation.”

  • 30. Cf. Plin. HN 20.191; 22.147; 25.174; 27.67; Diomedes Gramm. 1 p.32610 Keil; cf. herbaria, “the science of herbs,” Plin. HN 7.196. Pliny also uses medici for such authors (19.85; 20.253) and medicamentarii for experts in medical pharmacology (19.110). However, he also applies the term herbarii to wise-men (e.g., 7.117; see also the following n.), and medici to the “magical” Druids (30.13).

  • 31. Pliny’s sources often reported methods of collection and incantations, e.g., against hail, diseases, and burns, but he generally omitted them from his work because readers might object (HN 28.29); on such rituals, see e.g., Maria C. Martini, Piante medicamentose e rituali magico-religiosi in Plinio (Rome: Bulzoni, 1977); and Patricia Gaillard-Seux, “La place de l’incantation dans les recettes médicales de Pline l’Ancien,” in Sergio Sconocchia and Fabio Cavalli, eds., Lingue techniche del Greco e Latino, 4: Testi medici latini antichi (Bologna, Italy: Pàtron, 2004), 83–98. Pamphilus: Galen, De simpl. med. 11 p. 292–298 Keil; herbarii: Plin. HN 21.144; 25.174; 26.24.

  • 32. Thessalis magus imagined as capable of breaking witchcraft: Hor. Carm. 1.27.21–22; Marsi: Marsum augurem: Cic. Div. 1.132; for the trials involving magi and diviners in the Julio-Claudian period, see Marie Theres Fögen, Die Enteignung der Wahrsager (Frankfurt, Germany: Suhrkamp 1995), 89–113; cf. Matthew W. Dickie, “Magic in the Roman Historians,” in Gordon and Marco Simón, 79–104; and Niedermayer, Magie in den römischen, 79–125. you cannot abbreviate the title like this! Either Magie alone as still in note 89!] or Strafrechstfälle [NOT Strafrechtsfällen, since that is the dative plural, which would sound strange]. The other 9 references to this work must also be altered.

  • 33. E.g., Cic. Vat. 14; Verg. Ecl. 8.98; Aen. 4.490; Hor. Sat. 1.8.28f.; Tibull. 1.2.47–50; Ovid Am. 1.8.17; Met. 7.206; Sen. Herc. Oet. 458–460; Med. 740; Luc. 507–825; Val. Flacc. Arg. 1.774–784; Stat. Theb. 4.406–645; Suet. Nero 34.4; Herod. 4.12.4 (Caracalla); and cf. [Quint.] Decl. mai. 10.7. Astrologers more or less the same as magi: e.g., Tac. Ann. 2.27.2 (of Libo Drusus, 16 CE): Catusad Chaldaeorum promissa, magorum sacra, somniorum etiam interpretes impulit, “Catus got (him) interested in the forecasts of astrologers, magicians’ rites, even dream interpretation”; Dio 67.16.2 with CCAG 8.4 p.100.27–101.2 (on Larginus Proculus, 96 ce); Ptolem. Apotel. 4.4.10 Hübner; and cf. Frederick Cramer, Astrology in Roman Law and Politics (Philadelphia: American Philosophical Association, 1954), 248–281 (without commitment to the supposed Augustan decree of 11 ce).

  • 34. Druids foretelling the fall of Rome: Tac. Hist. 4.54.2; female scryers on the island of Sena off Brittany: Pompon. 3.6.48. For what it is worth, female “Druid” seers figure in several lives in the Historia Augusta, e.g., Sev. Alex. 60.6; and Aurel. 44.4, etc. Veleda (the name means “prophetess”): Tac. Hist. 4.61.2; 65.3; 5.24.1; Germ. 8; and Stat. Silv. 1.4.89. Vitellius had a female seer from the tribe of the Chatti (Suet. Vit. 14.5). Another female seer named Ganna accompanied her king to Rome under Domitian (Cass. Dio 67.5.3). On the turibula of C. Verius Sedatus at Autricum/Chartres (AE 2010: 950–952) see e.g., Miranda Aldhouse-Green, “The Magician’s House: Druids, Prayers, and Magic in Roman Gaul,” in Celtic Religions in the Roman Period, ed. Ralph Haeussler and Anthony King, Celtic Studies Publications 20 (Aberystwyth, 2017), 325–338. No reliable translation of the handful of curse-tablets in Gaulish, notably that from L’Hospitalet-du-Larzac (Aveyron), has yet been offered: see Pierre-Yves Lambert, Recueil des inscriptions gauloises 2.2: Textes gallo-latins sur instrumentum. Suppl. Gallia 45 (Paris: CNRS, 2002), 245–308.

  • 35. Agobard of Lyons, De grandine et tonitruis 15 (Migne PL 104 p. 147 = p. 14 van Acker). On Greek and Latin phylacteries against hail, see Francisco J. Fernández Nieto, “A Visigothic Charm from Asturias and the Classical tradition of phylacteries against hail,” in Gordon and Marco Simón, Magical Practice, 551–600.

  • 36. Nicole Zeddies, Religio et sacrilegium (Frankfurt, Germany: Peter Lang, 1999), 187 n. 93.

  • 37. Cf. Marie-Luise Thomsen, “Witchcraft and Magic in Mesopotamia,” in [The Athlone] History of Witchcraft and Magic in Europe 1; Biblical and Pagan Societies, ed. Bengt Ankarloo and Stuart Clark (Philadelphia: University of Pennsylvania Press, 2001), 3–95 (59–61, 93f.); and Erica Reiner, Astral Magic in Babylonia (Philadelphia: American Philosophical Society, 1995), 25–42, 119–132.

  • 38. Roy D. Kotansky, ed., Greek Magical Amulets, Papyrologica Coloniensia 22.1 (Opladen, Germany: Westdeutscher Verlag, 1994). Add now e.g., AE 1996: 955 = 1997: 998 and 2004: 853 (both Norfolk), 2008: 778 (nr. Oxford); 2013: 946 (London); and SEG 53 (2003), 1105 and 1110 (Rome).

  • 39. Simone Michel, Die magischen Gemmen (Berlin: Akademie Verlag, 2004).

  • 40. Jan Tremel, Magica agonistica. Fluchtafeln im antiken Sport. Nikephoros Beihefte 1 (Berlin: Weidmann, 2004), nos. 21–50; (Hadrumetum), 51–68 and 93–100; (Carthage), 70–90; (Rome); and add now AE 2014: 213 (Greek athletes and wrestlers at Rome). All those from Rome are late-antique, as are all those from the eastern Mediterranean—there was no continuous tradition in this area.

  • 41. E.g., the use of charaktêres (unrecuperable signs) in the otherwise banal texts found in the circus at Carthage: AE 1996: 1716a-b; or the primitive voces magicae (unintelligible “words”) in CIL XIII 11069–11070 (Charente Maritime).

  • 42. Cato Agr. 160; Varro Rust. 1.2.27 (evidently cited from an agricultural work by the Hostilii Sasernae, which contained many such items). Given the availability of fairly specialized herbal lore in written form, which abounds in claims for aphrodisiac powers (e.g., Plin. HN 20.32; 126; 227), the answer is likely to be positive.

  • 43. E.g., Véronique Dasen, Le sourire d’Omphale (Rennes, France: Presses universitaires de Rennes, 2015). On amulets and remedies against the evil eye (effascinationes) see e.g., Plin. HN 19.19, 50; 28.22; 37.145; Damigeron-Evax 21–24, Halleux-Schamp; persons suspected of casting the evil eye: Plin. HN 28.59; Gell. NA 9.4.7–9 with Plin. HN 7.16–18; cf. Stat. Silv. 2.6.73–75 (Nemesis); August. Conf. 1.7.11; and cf. Antón Alvar Nuño, Envidia y fascinación (Huelva 2012).

  • 44. Ep. 6.2.2. SHA Diadumen 4.2 remarks that midwives used to sell human cauls to advocati as an aid in court.

  • 45. Suet. Calig. 50.2; and Juv. 6.615–620. The myth-historical model for women as poisoners is given at Livy 8.18.1–10 (allegedly 331 bc).

  • 46. Lucil. Sat. frg.575–576 M. = 20.7 Charpin; and Plin. HN 28.30–31 (with a list of other such peoples with innate magical powers, rationalized as insita repugnantia, “natural protective power”).

  • 47. Circe’s son: Plin. HN 7.15; Solin. 2.27–30 (also Angitia for healing; cf. Plin. HN 25.86, etc.); magical powers: Hor. Epod. 5. 75f.; 17.29; Sat. 1.8.29; Ov. Ars am. 2.102; Fast. 6.141f.; Pliny HN 28.19; Juv. 3.169; and Sil. Pun. 8.495–499, etc. The Marsi themselves claimed that only true-bred Marsian males had such powers (Gell. NA 16.11.1–2).

  • 48. Cf. Sander M. Goldberg, “The Early Republic: The Beginnings to 90 BC,” in A Companion to Latin Literature, ed. Stephen J. Harrison (Malden, MA: Blackwell, 2005), 15–30; and Michael von Albrecht, Geschichte der römischen Literatur, 12 (Munich: DTV, 1997) 3–63. The best account of Latin literary magic is Wolfgang Fauth, Carmen magicum (Frankfurt, Germany: Peter Lang, 1999).

  • 49. Thessaly: Theophr. Hist pl. 9.15.2 already calls Thessaly the φαρμακωδέστατον‎ (pharmakôdestaton), “herbiest/witchiest” place in Greece; one story was that the herbs had fallen from Medea’s casket on her flight (Schol. Arist. Nub. 749a α‎). Caucasus: e.g., Prop. 1.12.9f., a glancing allusion to Ap. Rhod. Argon. 3.844–866.

  • 50. Cf. briefly Marco Frenchkowski, s.v. Magie, RAC 23 (2009) cols. 857–957 (864–866).

  • 51. This topography was reinforced by the alleged sources of semi-precious stones considered to have marvellous properties by the “Magi”: Plin. HN 37.133–192.

  • 52. See e.g., K. Preisendanz, s.v. Ostanes no.8, Paulys Realencylopädie der classischen Altertumswissenschaft 18.2 (1942), 1610–1642.

  • 53. A similar account of magic lies behind the remarks of Strabo 16.2.39, 762C; Apul. Apol. 31 and 90.2; Arnob. Adv. nat. 1.52.1; and Isid. Etym. 8.9.1–5. Thanks to the expansion of Graeco-Egyptian practice outside Egypt from the later 2nd century ce, Egypt, already famous, in Homer, as a source of marvellous herbs (Od. 4.220–232), increasingly replaces Persia as the privileged heterotopia, e.g., Apul. Met. 2.28–30; Apol. 38.7; Flor. 15 (§56) (Pythagoras studying there); Lucian Philops. 7, 31, 34, 57; Arnouphis and the “rain miracle” of 170/1 ce: Cass. Dio 71.8.10; Jerome Vit. Hilarion 12 Leclercq-Morales; and Damigeron-Evax Proem 4–8 Halleux-Schamp; Ambrosiast. In 2Tim. 3.8; August. De Civ. d. 10.8.

  • 54. Cf. Firm. Mat. Math. 3.2.18: (Saturn) mag[n]os famosos faciet vel philosophos opinatos vel sacerdotes templorum in magica semper opinione famosos, “will make celebrated magi or renowned philosophers or temple-priests who have always been famous in matters pertaining to (the art of) magic” (i.e., in Egypt); cf. Ptolem. Apotel. 2.3.34 Hübner.

  • 55. Lucian, De merc. cond. 27; cf. Markus Hafner, Lukians Schrift, Das traurige Los der Gelehrten“ (Stuttgart: Steiner, 2017) 31–41; and Dickie, Magic and Magicians, 219–224, 319–320.

  • 56. Regarding unges, many modern editions prefer to read ung<u>es, “finger-nail (cuttings),” on the grounds that trochiscili means “little wheels” and is thus a synonym of iunx. I take it as another deliberate use of a Greek term (i)unges, “wrynecks”, which denoted both a species of bird and a magical device used in love-magic, here understood as something different from trochiscili.

  • 57. This use of ἀντιπαθές‎ (antipathés) seems to be rooted through the adjective ἀντιπαθητικός‎ (antipathêtikós), “opposing passivity”; and Edward Courtney, The Fragmentary Latin Poets (Oxford: Oxford University Press, 1993), 140 cites Lucian Am. 27 from LSJ for the view that it means “an ingredient arousing mutual feeling.” “Allurements of whinnyers” is an absurd periphrasis for the Greek hippomanes, an aphrodisiac allegedly found on mares (or foals), whose use by Greek wise-women already attracted the attention of Aristotle, Hist. An. 6.18, 572a9–22; 6.22.17, 577a7–12; 8.24, 605a2–8 [Loeb line nos.]). A Latin equivalent was never developed (cf. e.g., Columella Rust. 6.27.3), which probably indicates that the notion was unknown in Italy except within the literary topos (cf. Ovid Am. 1.8.8: virus amantis equae, “slime/secretion of a mare on heat”—itself a rationalizing interpretation).

  • 58. Cf. Hor. Epod. 5; 17.76–80; Sat. 1.8; Vergil, Aen. 4.478–498 (the Massylian witch) with Sil. Pun. 8.98–103; Tib. 1.2.43–62; 8.17–24; 2.4.55–60; Prop. 3.6.25–30; 4.5.6–18; Ov. Am. 1.8.5–18; Her. 6.83–92; Rem. 249–290; Met. 7.179–233; 14.341–396; Medea (lost tragedy: Tac. Dial. 12.6); Sen. Herc. Oet. 452–471, 519–533; Medea 754–769; Petron. Sat. 134.12 (hexameters); and Luc. 6.452–491.

  • 59. E.g., Claud. In Ruf. 1.145–160; Get. 230–238; Prud. c. Symm. 1.96f.; Luxorius, In magum mendicum (299 Happ = Anth. Lat. 294 Shackleton-Bailey) (ironic); and cf. Nemes. Ecl. 4.62–73. Since we possess virtually no elegaic poetry from after the death of Augustus, it is impossible to judge the extent to which the stereotype continued to be deployed. The last grand amalgam is Lucan’s Erictho (6.588–830) in the mid-1st century ce, which is clearly fused with the Grand-Guignol necromantic scenario.

  • 60. A sub-set features curses against procuresses who attempt to part literary lovers from their “girls” because another man has offered them more money: Tib. 1.5.49–58; 2.6.53–54; Ov. Am. 1.8.13–114, cf. already Plaut. Asin. 1.2.131–137. Lurid necromantic scenes: Hor. Sat. 1.8.26–36 (comic); Sen. Oed. 530–658; [Sen.] epigr. 16 Baehrens = Anth. Lat. 406 Riese; Luc. 6.588–830; Stat. Theb. 4.443–645; and Val. Flacc. Arg. 1.730–751;774–818.

  • 61. With regard to Latin imitations of Theocr. Id. 2, we should note that Calp. Ecl. (Neronian period) omits the magical scene entirely.

  • 62. There seems to be no proper English word for the Greek technical term ousia, substance or essence, which is always represented concretely, e.g., hair, nail-clippings, a personal possession of the victim.

  • 63. Tac. Ann. 2.69.5; cf. 14.10.5; Suet. Calig. 59; Nero 34.4, 46.1; Cass. Dio 61.14.1; Plin. Ep. 7.27.5–11. But see already the lurid-absurd allegations of Cic. Vat. 14.

  • 64. Wolfgang Iser, Prospecting (Baltimore: Johns Hopkins University Press, 1989), 270.

  • 65. See Gordon, “A Babel of Voices,” 172‒183.

  • 66. Compare the different views of Jean-Benoît Clerc, Homines magici: Étude sur la sorcellerie et la magie dans la société romaine impériale (Bern, Switzerland: Peter Lang, 1995), 172–192; Fritz Graf, Magic in the Ancient World (Cambridge, MA: Harvard University Press, 1997), 64–88; Francesca Lamberti, “De magia als rechtsgeschichtliches Dokument,” in Apuleius, De magia, Jürgen Hammerstedt (Darmstadt, Germany : Wissenschaftliche Buchgesellschaft, 20112), 331–350; and Niedermayer, Magie in den römischen, 130–144.

  • 67. Calpurnius Piso frg. 36 Peter = Plin. HN 18.41–43; cf. Graf, Magic in the Ancient World, 61–64.

  • 68. In view of the fact that the hearing was before the curule aedile, Rives, “XII Tables,” 277 and Niedermayer, Magie in den römischen, 47–48, suggest that by the early 2nd century bc, it was no longer considered appropriate to impose the death penalty in such cases. The reservations of Dickie, Magic and Magicians, 143–144 are overscrupulous.

  • 69. Petron. Sat. 63; on striges killing babies, see esp, Ov. Fast. 6.101–168 (the nymph Carna as wise-woman).

  • 70. Apul. Met. 1.11–19 (Aristomenes’ story).

  • 71. Pliny’s lucubrations over vis carminum, “the efficacy of prayer/solemn utterance/incantation” (HN 28.10–22) hardly count as explanation. Despite the Neo-Platonist and Christian claim that malign magic is the work of wicked daimones/devils, the late legal sources do not rely on this account.

  • 72. Above all in August. De doct. christ. 2.2.19–24, cf. Valerie I. J. Flint, “The Demonisation of Magic and Sorcery in Late Antiquity,” in History of Witchcraft and Magic in Europe, 277–348.

  • 73. Apul. Apol. 64–65, cf. Jacques Annequin, “Magie et organisation du monde chez Apulée,” in AA.VV., Religions, pouvoir, rapports sociaux (Paris: Les Belles Lettres, 1980) 173–208; and Nicole Méthy, “Magie, religion et philosophie au IIe siècle d.n.è.,” in La magie, 3: Du monde latin au monde contemporain, ed. Alain Moreau and Jean-Claude Turpin (Montpellier: Université Montpellier III, 2000), 85–107.

  • 74. Deprecatio is here used as a synonym of dira precatio in Sen. Ben. 6.35.4–5: execraris enim illum et caput sanctum tibi dira precatione defigis, “for you execrate him and (try to) ‘catch’ someone whom you ought to revere by means of a curse”; in the following sentence, this is clearly supposed to be a public act in front of witnesses (si aperte illi … imprecareris); cf. mala precari at Cic. Pis. 33. Defigere is the word often used for the act of cursing someone in a written text, cf. AE 1975: 449 = 1987: 455 (Cremona); Collingwood-Wright RIB no. 6 (London); 221 (Clothall). Until recently, it was believed that the substantive defixio was attested only in late antiquity, but the word has now turned up (spelled defictcsio) in one of a small group of curses from a military site on the middle Danube (Abusina/Eining, Bavaria) dating from the Flavian-Trajanic period: Jürgen Blänsdorf, “Die Verflüchungstäfelchen aus dem Kohortenkastell Abusina/Eining,” Bayerische Vorgeschichtsblätter 84 (2019): 229-242 (p.231 no. 1).

  • 75. E.g., Livy 39.51.12 (suicide of Hannibal); Sen. De ira 2.36.5.

  • 76. Lindsay Watson, Arae (Leeds, UK: Francis Cairns, 1991), 79–149. [Vergil] Dirae is not typical of the Hellenistic genre.

  • 77. See Silvia Alfayé and Celia Sánchez Natalías, “Magic in Funerary Spaces,” in Choosing Magic: Contexts, Objects, Meanings, ed. Richard L. Gordon, Francisco Marco Simón and Marina Piranomonte, 39-51 (Rome: De Luca, 2020).

  • 78. For the two dozen or so curses that explicitly call for the death of the target, see Francisco Marco Simón, “Tradite Manibus: Transferred Death in Magical Rituals,” in Formae mortis: El tránsito de la vida a la muerte en las sociedades antiguas, ed. Francisco Marco Simón, F. Pina Polo, J. Remesal Rodríguez (Barcelona, Spain: Universitat de Barcelona, 2009), 165–180. A handful of epigraphic funeraries claim that the deceased has been killed by malign magic. In Latin, during the Principate see: CIL I2 3358a = AE 1982: 326 (Saturnia, Etruria); CIL III 2197 (Salona); VI 19747; 20905 (both Rome); VIII 2756 (Lambaesis); possibly IX 3030 (Teate); note also XI 4639 (Tuder: curse laid by a public slave, but discovered: see Duncan E. MacRae, “The Freedman’s Story,” Journal of Roman Studies 108 (2018): 53‒73); and cf. Fritz Graf, “Untimely Death, Witchcraft, and Divine Vengeance: A Reasoned Epigraphic Catalogue,” Zeitschrift für Papyrologie und Epigraphik 162 (2007): 139–150.

  • 79. Celia Sánchez Natalías, Sylloge of Defixiones from the Roman West (Oxford: Archaeopress, 2020).

  • 80. Circus texts in Latin: Tab. Def. Audollent nos. 272–295; AE 1905: 171; 1907: 68–69; 1911: 6 (Hadrumetum); CIL VIII 12504 (Carthage) (all II-IIIp). These and one or two others are most conveniently consulted in J. Tremel, Magica agonistica, nos. 21–45; 47–52 (Hadrumetum); and 68 (Carthage). The only known texts against venatores in Latin are Tab. Def. Audollent nos. 247–248 and 250–251; 254; Latin in Greek characters: 252; bilingual: 249, 253 = Tremel nos. 93–94, 96–97, 100, 98, 95, 99 (all from the amphitheatre in Carthage). All the late-antique circus-texts from Rome are in Greek.

  • 81. E.g. AE 1993: 1008; 1999: 954a-b; 2012: 740; CIL II/7. 250; see Francisco Marco Simón, “Early Hispanic Curse Tablets in Greek, Latin – and Iberian?,” Religion in the Roman Empire 5, no.3 (2019): 376‒397.

  • 82. Note here e.g., Francisco Marco Simón, “La expresión epigráfica de la divinidad en contextos mágicos del Occidente romano: Especifidad, adaptación e innovación,” Cahiers Glotz 21 (2010): 293–304; Richard Gordon, “Do the ‘Vernacular’ Curse-Tablets from Italy represent a Specific Knowlege-Practice?,” Religion in the Roman Empire 5, no.3 (2019): 417–439.

  • 83. E.g., CIL I2 2520a-e (“Johns Hopkins”); AE 2007: 260 (‘Via B. Bompiani); Mainz: e.g., AE 2004: 1026; 2005: 1123 and 1124. None of the Mainz tablets are addressed to Isis, to whom the other twin-temple was dedicated.

  • 84. Antón Alvar Nuño, Cadenas invisibles (Besançon, France: Presses universitaires de Franche-Comté); idem, “The Use of Curse-Tablets among Slaves in Rome and its Western Provinces,” Religion in the Roman Empire 5, no. 3 (2019): 398–416.

  • 85. Roger Tomlin, “The Curse-Tablets,” in The Temple of Sulis Minerva at Bath, 2: The Finds from the Sacred Spring, ed. Barry Cunliffe (Oxford: Oxford University Committee for Archaeology, 1988), 59–277; Stuart McKie, “Enchained Relationships and Fragmented Victims: Curse-Tablets and Votive Rituals in the Roman North-West,” Religion in the Roman Empire 5, no. 3 (2019): 440–455. Tomlin is still working on the difficult Uley texts; a provisional account by him in The Uley Shrines, ed. Anne Woodward and Peter Leach (London: English Heritage/British Museum Press, 1993), 113–130.

  • 86. Cf. Martin Dreyer, “‘Prayers for Justice’ and the Categorization of Curse Tablets,” in Contesti magici/Contestos mágicos, 29–32; note also the response by Hendrik S. Versnel, “Response to a Critique,” in Contesti magici, 33–46.

  • 87. Livy 39.41.5; 40.37.4; 43.2–3; Perioch. 48 (the scale of these latter executions is unknown); see Niedermayer, Magie in den römischen, 61; 62f. In 179, the trials were extended to Rome itself.

  • 88. See Jean-Louis Ferrary in Michael H. Crawford, ed., Roman Statutes (London: Institute of Classical Studies, 1996), 2: 749–753; James B. Rives, “Magic, Religion and Law: The Case of the lex Cornelia de sicariis et veneficiis,”in Religion and Law in Classical and Christian Rome, ed. Cliff Ando and Jörg Rüpke (Stuttgart, Germany: Steiner, 2006), 47–67. The known cases prosecuted under this law, all of which concern more or less prominent individuals or members of prominent families, are discussed individually by Niedermayer, Magie in den römischen, 68–146.

  • 89. James B. Rives, “Magic in Roman Law: The Reconstruction of a Crime,” Cl. Ant. 22 (2003): 313–339 (329). The main evidence is the late 3rd-century ce text [Paul.] Sent. 5.23.15 … ut quem obcantarent, defigerent, obligarent, fecerint, faciendave curaverint, aut cruci suffiguntur aut bestiis obiciuntur, “(people) who perform or cause to be performed (rites) in order to bewitch, “catch,” or bind anyone are to be crucified or sent to the arena.” Assuming a later date for [Verg.] Ciris 377, the earliest occurrences of defigere in this sense are Sen. Ben. 6.35.4; Plin. HN 28.19 (in each case of a curse; even in curse-tablets the word is rare); the sole other occurrence of the word occentare in the sense “bewitch” is Apul. Apol. 84.1; as for obligare, the earliest examples occur in 3rd-century defixiones from Carthage: Def. Tab. Audollent nos. 219 a 3 (as oligo); 248a12; 250a2 and b12; 252.36 (in Greek characters), probably as a translation of καταδέομαι‎, katadeomai, I bind; otherwise this passage of [Paul.] is the earliest example given in TLL.

  • 90. Cf. Cass. Dio 57.15.7–9 (Libo Drusus affair of 16 ce). On the other hand, “such laws do not seem to have been intended to have permanent force”: David Potter, Prophets and Emperors (Cambridge, MA: Harvard University Press, 1994), 175. Compare Niedermayer, Magie in den römischen, 146–184. On the occasional expulsions of magi and astrologers from the City of Rome, see Cramer, Astrology, 233–248; and Niedermayer, Magie, 185–196.

  • 91. The same habit of mind projected necromantic practices upon “bad” or unfortunate emperors, e.g., Suet. Nero 34.4; Herodian 4.12.3–4; SHA Did. Iul. 7.9–10; and self-declared: Amm. Marc. 29.2.17 (“Numerius”: PLRE 1 Numerius 2).

  • 92. This very issue is the topic of the sole surviving genuine practice-declamation by Hadrian of Tyre (Adrianus, second half of 2nd century) [H. Hinck, ed., Polemonis declamationes (Leipzig, Germany: Teubner, 1873), 44–45]; and for the late 3rd century: [Paul.] Sent. 5.23.18. However Fögen, Enteignung, 78–79 suspected interpolatation here to make the text fit the more drastic legislation of the Christian Empire.

  • 93. CTh 9.16.4 (357); 16.5 (356); 16.6 (357); 16.7 (364); 16.8 (370); 16.9 (371); 16.11 (389); 16.12 (409). These successive rulings are best followed in Roland Delmaire et al., Les lois religieuses des empereurs romain de Constantin à Théodose II (312-438), 2: Code Théodosien I-XV, Code Justinien, Constitutions Sirmondiennes (Paris: Éditions du Cerf, 2009), 135–157; and cf. Almuth Lotz, Der Magierkonflikt in der Spätantike (Bonn, Germany: Habelt, 2005), 62–183.

  • 94. E.g., Amm. Marc. 19.13–14 (359); 28.1.5–35 (368 and 371/2); 30.5.11 (375); cf. Clerc, Homines magici, 204–237; and Dickie, Magic and Magicians, 252–257.

  • 95. See e.g., Almuth Lotz, “Libanius and Theodoret of Cyrrhus on Accusations of Magic: Between Legal Norm and Legal Practice in Late Antiquity,” Magic, Ritual, and Witchcraft 14, no.2 (2019); 211- 229.

  • 96. See e.g., Plin. HN 28.22–29, 33–39 for a range of “normal” superstitions that instantiate the immanence of the unpredictable divine.

  • 97. Penal law: e.g., T. Mommsen, Römisches Strafrecht (Leipzig, Germany: Duncker & Humblot, 1889; repr. Darmstadt: Wiss. Buchgesellchaft, 1955; several internet versions), 635–637 (venenum); 639–643 (magic), cf. 861–865 (divination); and Eliane Massonneau, Le crime de magie et le droit romain (Paris: Sirey, 1933). Christian reinterpretation: L. Gardette, s.v. Magie, Dictionnaire de théologie Catholique 9.2 (1927): 1510–1550.

  • 98. Tab. def. Audollent nos. 264–298 (a few in Greek or Latin in Greek characters); AE 1905: 171, etc. The Carthaginian curse-tablets were excavated by Father Louis Delattre, who founded the Musée Lavigerie.

  • 99. Marcel Mauss, “Esquisse d᾿‎une théorie générale de la magie,” in Sociologie et anthropologie, ed. Marcel Mauss and Claude Lévi-Strauss (Paris: Presses universitaires de France, 1950), 1–141. Repr. from L’Année sociologique 7 (1902–1903): 1–146; tr. by Robert Brain as A General Theory of Magic (New York: Norton, 1972).

  • 100. Alfred Ernout, ed., Pline. Histoire Naturelle 28 (Paris: Les Belles Lettres, 1962); 30 (Ernout, 1963); Anne-Marie Tupet, La Magie dans la poésie latine, 1: des origines à la fin du règne d’Auguste (Paris: Les Belles Lettres, 1976). Already in 1954, Peter Green submitted a fellowship dissertation at Cambridge entitled “Prolegomena to the Study of Magic and Superstition in the Natural History of Pliny the Elder, with special reference to Book XXX and its sources,” which was unfortunately never published.

  • 101. Rafaella Garosi, “Indagine sulla formazione del concetto di magia nella cultura romana,” in Magia: studi di storia delle religioni in memoria di Raffaela Garosi, ed. Paolo Xella (Rome: Bulzoni, 1976), 13–93.

  • 102. Note esp. Jacques Annequin, Recherches sur l᾿‎action magique et ses représentations (Ier et IIème siècles après J.-C) (Paris: Les Belles Lettres, 1973).

  • 103. Bath and Uley: see n. 82; Mainz: Jürgen Blänsdorf, Die Defixionum Tabellae des Mainzer Isis- und Mater Magna-Heiligtums (Mainz: Direktion Landesarchäologie, 2012); and Anna Perenna: see provisionally Rosanna Friggeri et al., eds., Terme di Diocleziano: La collezione epigrafica (Milan, Italy: Electa, 2012), 617–639: Sala IX.49.

  • 104. Marvin Meyer and Pail Mirecki, eds., Ancient Magic and Ritual Power (Leyden, The Netherlands: Brill 1995); Paul Mirecki and Marvin Meyer, eds., Magic and Ritual in the Ancient World (Leyden: Brill, 2002); and Alain Moreau and Jean-Claude Turpin, eds., La magie. 4 vols.: Du monde babylonien au monde hellénistique. La magie dans l’antiquité grecque tardive/Les mythes. Du monde latin au monde contemporain. Bibliographie générale (Montpellier: Université Montpellier III, 2000).

  • 105. Elisabeth Wallinger, Hekates Töchter: Hexe in der römischen Antike (Vienna: Wiener Frauenverlag, 1994).

  • 106. Morton Q. Smith, Jesus the Magician (London: Gollancz, 1978); and Howard C. Kee, Medicine, Miracle, and Magic in New Testament Times (Cambridge, UK: Cambridge University Press, 1986).

  • 107. Among dedicated research projects note the electronic data-base of curse-tablets (in Greek and latin), TheDeMa, formerly at Magdeburg, now under transfer to Hamburg; and a long-term one based at Zaragoza, Spain, whose publications include Gordon and Marco Simón, Magical Practice; Piranomonte and Marco Simón, Contesti magici/Contestos mágicos; Antón Alvar Nuño, Envidia y fascinación (Huelva, Spain: ARYS 2012); Silvia Alfayé and Francisco Pina Polo ed. , Dioses, sacerdotes y magos en la Antiguëdad (Madrid, Spain: 2019); Richard L. Gordon, Francisco Marco Simón and Marina Piranomonte ed., Choosing Magic: Contexts, Objects, Meanings (Rome: de Luca, 2020); and Sánchez Natalías, Sylloge of Defixiones).

  • 108. Marie Theres Fögen, Die Enteignung der Wahrsager (Frankfurt, Germany: Suhrkamp, 1995); Clerc, Homines magici; and Niedermayer, Magie in den römischen.

  • 109. James B. Rives, “Magicians and Astrologers,” in The Oxford Handbook of Social Relations in the Roman World, ed. Michael. Peachin (Oxford: Oxford University Press, 2011) 679–692; cf. too Gordon, “Babel.” Daniel Ogden, Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds (New York: Oxford University Press, 2002) is an admirable aid in teaching.

  • 110. Denis Feeney, Literature and Religion at Rome (Cambridge, UK: Cambridge University Press, 1998); Richard L. Gordon, “Magic as a Topos in Augustan Poetry: Discourse, Reality, and Distance,” Archiv für Religionsgeschichte 11 (2009): 209–228. John Winkler’s scattered remarks on the magical narratives in Apuleius’ Metamorphoses are still the best on offer: Auctor and Actor (Berkeley: University of California Press, 1985). Stavros Frangoulidis, Witches, Isis, and Narrative (Berlin: de Gruyter, 2008) cannot be recommended. On topoi relating to ghosts and necromancy, note Catherine Schneider and Céline Urlacher, “Rationnel et irrationnel dans le Sepulchrum incantatum du Pseudo-Quintilien,” in Conceptions et représentations de l’extraordinaire dans le monde antique, ed. Olivier Bianchi and Olivier Thévenaz (Bern, Switzerland: Peter Lang, 2004), 99–113.

  • 111. Dickie, Magic and Magicians, 162–201.

  • 112. E.g., John M. Riddle, “High Medicine and Low Medicine in the Roman Empire,” ANRW II.37.1 (1993), 102–120; Patricia Gaillard-Seux, “À propos des livres XXVIII–XXIX–XXX de l’Histoire naturelle de Pline l’Ancien,” Latomus 57.3 (1998): 625–633 (and many other papers by her); cf. too John Scarborough, “Pharmacy in Pliny’s Natural History: Some Observations on Substances and Sources,” in Pharmacy and Drug-Lore in Antiquity, ed. John Scarborough (Farnham: Ashgate, 2010), no. 9 (original: 1986). On the supposed techniques of such persons, classed as “witches and sorcerers,” see Anne-Marie Tupet, “Rites magiques dans l’antiquité romaine,” ANRW II.16.3 (1986): 2591–2675.

  • 113. Jürgen Blänsdorf, “Ein System oraler Gebrauchspoesie. Die alt- (und spät)lateinischen Zaubersprüche und Gebete,” in Metrik und Medien-Wechsel/Metrics and Media, ed. Hildegard Tristram, ScriptOralia 35 (Tübingen, Germany: Narr, 1991), 33–51; Alf Önnerfors, “Magische Formeln im Dienst römischer Medizin,” ANRW II.37.1 (1993): 157–224; Hendrik S. Versnel, “The Poetics of the Magical Charm: An Essay on the Power of Words,” in Mirecki and Meyer, Magic and Ritual in the Ancient World, 105–158; Paolo Pocetti, “Manipolazione della realtà e manipolazione della lingua: alcuni aspetti dei testi magici dell’antichità,” in Linguagio: Linguaggi. Invenzione: Scoperta, ed. Ruggero Morresi (Rome: “il Calamo,” 2002), 11–59 (mainly Greek examples); and Amina Kropp, “How does Magical Language Work? The Spells and Formulae of the Latin defixionum tabellae,” in Magical Practice, 357–380.

  • 114. Kimberley B. Stratton, Naming the Witch (New York: Columbia University Press, 2007), 71–105; and Stratton, “Magic, Abjection, and Gender in Roman Literature” in Naming the Witch, 152–180.

  • 115. Silvia Alfayé, “Nails for the Dead: A Polysemic Account of an Ancient Funerary Pratice,” in Magical Practice, 427–456; Piranomonte and Marco Simón, eds., Contesti magici/Contestos mágicos; Andrew Wilburn, Materia magica. The Archaeology of Magic in Roman Egypt, Cyprus and Spain (Ann Arbor: University of Michigan Press, 2012); Antón Alvar Nuño, Envidia y fascinación (Huelva, Spain: ARYS, 2012); and Adam Parker and Stuart McKie, eds., Material Approaches to Roman Magic (Oxford: Oxbow, 2018).

  • 116. Re-editions: Amina Kropp, Magische Sprachverwendung in vulgärlateinischen Fluchtafeln (defixiones) (Tübingen, Germany: Gunter Narr, 2008); Urbanová, Latin Curse Tablets; Sánchez Natalías, Sylloge. Social history: e.g., Richard L. Gordon, “Fixing the Race: Managing Risks in the Circus at Carthage and Hadrumetum,” in Contesti magici, 47–74; and Antón Alvar Nuño, Cadenas invisibles (Besançon, France: Presses universitaires de Franche-Comté, 2017); idem, “Magical Solutions to Gossip and Accusations in the Roman World,” Magic, Ritual, and Witchcraft 14, no. 2 (2019): 189–210; idem, “Ritual Power, Routine, and Attributed responsibility: Magic in Roman Households, Workshops and Farmsteads,” in Choosing Magic: Contexts, Objects, Meanings, Richard L. Gordon, Francisco Marco Simón and Marina Piranomonte ed., 75–89 (Rome: de Luca, 2020).

  • 117. E.g., Richard L. Gordon, “Gods, Guilt, and Suffering: Psychological Aspects of Cursing in the North-Western Provinces of the Roman Empire,” Acta Classica Universitatis Scientiarum Debreceniensis 49 (2013): 255–281; idem, “Diagnosing the Signs: The Body as Subject in Latin Defixiones,” in Choosing Magic: Contexts, Objects, Meanings, Richard L. Gordon, Francisco Marco Simón and Marina Piranomonte ed., 91–98 (Rome: de Luca, 2020).

  • 118. Irene Salvo, “Sweet Revenge: Emotional Factors in ‘Prayers for Justice’,” in Unveiling Emotions, ed. Angelos Chaniotis (Stuttgart, Germany: Steiner, 2012), 235–266.

  • 119. Cf. Janico Albrecht et al., “Religion in the Making: The Lived Ancient Religion Approach,” Religion 48, no. 4 (2018).

  • 120. David E. Aune, “Magic in Early Christianity,” ANRW II.23, no. 2 (1980): 1507–1557; Giulia Sfameni Gasparro, “Magie et magiciens: Le débat entre chrétiens et païens aux premiers siècles d.n.è.,” in Charmes et sortilèges. Magie et magiciens, ed. Françoise Aubaile-Sallenave and Rika Gyselen, Res Orientales 14 (Bures-sur-Yvette, France: Groupe pour l’Étude de la Civilisation du Moyen-Orient, 2002), 39–66; and Nicola Denzey Lewis, “Lived Religion among Second-Century ‘Gnostic Hieratic Specialists’,” in Beyond Priesthood, ed. Richard L. Gordon, Georgia Petridou, and Jörg Rüpke (Berlin: de Gruyter, 2017), 79–102.