Scholia Reviews ns 7 (1998) 18.

Fritz Graf (tr. Franklin Philip), Magic in the Ancient World. Cambridge, Mass. & London: Harvard University Press, 1997. Pp. 313. ISBN 0- 674-54151-0. UK£23.50.

Michael Lambert
Department of Classics, University of Natal, Pietermaritzburg

Fritz Graf's imaginative contributions to the study of myth and ritual are deservedly well known; in this work, Graf brings his own scholarship, and that of participants in a series of seminars at the École Pratique des Hautes Études in Paris, to bear on the hitherto rather neglected field of magic in antiquity.[[1]] The result is an accessible, clear and well- annotated guide to the complex world of the ancient magician, which serves both as a valuable introduction to the field and as an invaluable resource for further research and debate.

In his introductory chapter (pp. 1-19), after a brief discussion of the sources for the study of ancient magic (for example, literature, the magical papyri and the tabulae defixiones) and some restrained Quellenforschung into scholarship on magic, Graf skilfully sketches the major contours of the magic-religion debate within classical scholarship, ethnology, and anthropology--a debate heavily influenced by the Frazerian distinctions between magic, religion and science 'according to the agent's intention, rationality and autonomy' (p. 14). Frazer argued that, in the spheres of science and magic, the agent's autonomy characterises his/her relationship with the natural and supernatural worlds; magic and science share rational procedures and laws whereas religion is characterised by irrationality, the absence of a practical goal and by a humble submissiveness towards the supernatural, absent in the coercive attitude of the magician.

Research into the religious experiences and practices of cultures outside Europe has long since demonstrated that the distinctions made by Frazer can only be made from within a specific world-view (the Judaeo-Christian one) and that many cultures simply do not distinguish between magic and religion. Some acholars have thus proposed scrapping the use of the word 'magic' altogether (for, after all, one person's magic is another person's religion), whereas others have proposed retaining the distinction only if it is clearly made by the culture under study. Graf decides to adopt the latter course and to analyse how the Greeks and Romans, who devised the words mageia and magia for the category 'magic', deployed these and related terms. This avoids the confusing (and Frazerian) use of the word 'magical' to designate the primitive stages of Greek and Roman religion (when the ancients themselves had not invented the term) and apparently lets Graf escape from a hermeneutic minefield. Of course, this option 'implies the scrupulous analysis of the ancient terminology' (p. 19), which Graf handles with elegance and precision, even at times managing a few Teutonic drolleries, which make delightful reading.[[2]]

In chapter 2, 'Naming the Sorcerer' (pp. 20-60), Graf traces the development of the term magos and considers the possibility that it had acquired negative connotations by the end of the sixth century BC, together with the words for itinerant priests and diviners (agurtes and mantis). Interesting to note is that Graf argues, convincingly, that the distinction between magic and religion had already been made in Heraclitus and Plato, a distinction made possible by the development of a philosophical theology and of science, for which the Hippocratic text On the Sacred Disease provides crucial evidence (pp. 30-32).

Roman usages of magus and magia are also carefully explored. In contrast to the Athenians who did not legislate against black magic, the Romans used Sulla's lex Cornelia de sicariis et veneficis (81 BC) as the fundamental law for any legal action against magic. In an examination of the usages of veneficium and veneficus, Graf rightly draws attention to the fact that this legislation did not condemn magic per se, but was aimed at crimes which caused the sudden, inexplicable deaths of citizens, in contrast to deaths caused by violence. Missing from the Roman evidence, notes Graf, is the Greek goes-agurtes- magos, practising a combination of divination, healing, initiation and magic.

Evidence for the negative connotations attached to magia is found in Pliny the Elder's Natural History (30.1), where Pliny clearly distinguishes between medicina and magia (false medicine) and, as Graf interestingly observes, associates magic with healing and divination, rather than the black arts, which are conspicuously absent from his text.

In chapter 3, 'Portrait of the Magician, seen from the Outside' (pp. 61- 88), Graf, in an attempt to discover how the Greeks (rather loosely used here) and Romans defined the magician, and distinguished him from similar, but less hated figures, focusses on two Roman magic trials, that of the freedman C. Furius Cresimus in the first half of the second century BC, and that of Apuleius in the second century AD. Graf argues that Cresimus' trial, recorded by Pliny, demonstrates convincingly that a person accused of magic is the 'marginal' on the fringes of society threatening the society's structures.

The better known trial of Apuleius (at Sabratha in Africa), whose successful Apologia sive de magia is the only extant evidence for the trial and thus has to be handled with caution, focusses on the erotic magic which the young philosopher was accused of using to win over the heart and the wealth of the widowed Pudentilla. Graf's analysis of the speech reveals how an unsually intense interest in religion (a silent prayer or rituals held at night or the possession of ritual objects) could be construed by the community as constituting sorcery. As was the case with Cresimus, the marginal person (the foreign, apparently itinerant philosopher with an interest in exotically-named fish!) can become, in the society of Apuleius, a dangerous magician, threatening, in this case, to destabilize social institutions like marriage and inheritance.

In chapter 4, 'How To Become A Magician: The Rites Of Initiation' (pp. 89- 117), Graf, using as a springboard Lucian's famous tale of the sorcerer's apprentice (The Lover of Lies 34-36), who was taught by Isis how to become a magician after spending 23 years in secret chambers under the sands of Egypt, considers how this tale embodies fundamental aspects of the magician's initiation; secrecy, personal communion with a deity, prolonged study (e.g. a knowledge of hieroglyphics being a prerequisite) and subterranean katabasis. All these features are present in the mystery cults as well and Graf usefully examines the similarities and differences between the magician's initiation and intiation into a mystery cult.

In chapter 5, 'Curse Tablets and Voodoo Dolls' (pp. 118-174), Graf focusses on ritual binding, using as his sources the texts of lead tablets and an array of literary references which begin with Aeschylus and Plato, who provides us wth the first detailed information on binding spells. The purpose of binding was to subject another to one's will in almost every sphere of human existence in which competition was involved, as Audollent's categories reveal (judicial, erotic and agonistic spells, as well as those against slanderers, thieves and economic competitors).[[3]] Graf argues that ritual binding, performed at a time of intense crisis and uncertainty, was intended not for vengeance or to injure the other party, but to influence the outcome of the competition, thus giving the community and the individual some means of emotional mastery. Pursuing his linking of the magical with the marginal and its associated reversals, Graf believes that the places where the spells were buried (cemeteries, wells, springs, sanctuaries) suggest movement towards the nether world and thus a reversal of the movement upward, which characterises the traditional religion of the polis; furthermore the fact that the victim and sometimes the sorcerer are regularly defined by the name of the mother, suggests reversal of the normal use of the patronymic. From the papyri, Graf reconstructs the binding ritual itself, showing how the spell was recited and written at the same moment, thus making the spoken language permanent--a message for the dead to deliver to the world below; another group of tablets consists of formulas, assimilating the victim's name to the uselessness of the cold lead tablet on which the spell is written (a similia similibus formula). Determined to lay the ghost of Frazerian sympathetic magic to rest, Graf argues that the magicians in this case (and in instances of reverse writing) were exploiting a traditional practice (i.e. writing on lead) to forge new and unexpected meanings in the magician's marginal universe.

Detailed analyses of ritual sequences in the papyri lead Graf into a fascinating account of the katadesmos/defixio, the procedure of which differs from magical act to magical act: piercing tablets or a cat with iron nails, binding statuettes with bronze or iron, ensuring that some figurines are well and truly aversi, with their heads and/or feet turned rigidly to the back, engraving magic words on female figurines which are then pierced, tied to a tablet (which may then be bound with a number of magical knots) and then deposited, often in graves of the untimely dead, who harbour especial resentment against the living, whilst the magician recites a prayer, facing the setting sun (another reversal). Graf believes that in the cases of erotic magic, the performers did not intend to maim or kill the victims represented by the figurines which may have had ousia (nails, hair, fabric of clothing) attached to them; the object of the spell was the total submission of the woman to the man. Drawing the attention to the fact that the tablets were not necessarily found with voodoo dolls, Graf argues that their efficacy did not depend on a 'sympathetic' act performed with the help of figurines. Apart from Frazerian 'sympathetic homology', Graf attempts to exorcise the psychological interpretation of the motivation behind the spells as well, demonstrating that the rituals were clearly not spontaneous and cathartic outbursts of hatred, but time-consuming and complex rituals, performed by professionals not necessarily directly involved in the crisis situations, as caches of texts found written in the same hand suggest.

In chapter 6, 'Literary Representations of Magic' (pp. 175-204), Graf criticises those scholars who uncritically use literary texts as sources for magic rituals and practices in an effort to fill in the gaps created by the comparatively rare occurrence of curse tablets in the Hellenistic period, as compared with the classical and imperial epochs. Apart from literary questions of intertextuality, especially important in this period, Graf ably demonstrates that the binding spell, for example, in Theocritus Idyll 2.1-138, does not correspond, in its ritual detail and use of magical materials, with love spells found in the papyri. Graf believes that Theocritus here constructs 'a kind of superritual capable of activating in its readers all sorts of associations connected with magic' (p. 184), but a superritual which would not work in the world of 'real' magic. In his analysis of Lucan's Erictho and her resurrection of a cadaver for divination, Graf argues that Lucan's use of the language of religion constructs magic as an essential perversion of the civic religion, thus opposing magic and religion in a manner not found in the papyri.

Like John Winkler, Graf raises the question of why women are almost exclusively the practitioners of magic in the literary texts, whereas men dominate the epigraphic texts and the papyri. Graf takes the sociological model ('for the transferring of power and fortune through women') to task on the grounds that erotic charms were not aimed at the fathers, but at their daughters, thus suggesting that erotic magic was the man's response to a painful personal crisis, in which the object of his desires seemed completely out of reach. Literary versions, which erase men's use of magic, remove erotic magic from the world of men where it should not exist. For, argues Graf, 'they [these stories] reveal the real existence of this magic; however, although practiced by men, it is in reality a concern of women. That is why a man using magic steps over the borderlines of male behaviour . . .' (p. 189). Graf is particularly unclear here; why, in reality, is magic a concern of women? Whose reality? That of male writers like Theocritus and Vergil? The 'reality' of the world of the papyri would suggest that magic is very much the concern of men. Speaking of gender reversals in the literary texts and women's 'reality' is very perverse, when the very gender categories supposedly transgressed, together with women's 'reality', are creations of the same sources (i.e. male-produced texts). In this respect, Winkler's interpretation (the displacement of irrationality onto women by male writers) is more convincing.[[4]]

In the final chapter, 'Words and Acts' (pp. 205-233), Graf addresses the question of the special nature of magic ritual and offers some suggestions for further debate. In an interesting analysis of the origins of Frazer's notion of 'sympathetic' magic, Graf argues that Frazer derived the essence of this idea from the sumpatheia of the Greek Stoics, that 'intimate orchestration that connects the whole cosmos and the planets to our everyday life' (p. 206). Convinced that the notion of 'sympathetic' magic should be relegated to the academic archives among with 'la mentalité primitive', Graf turns to Tambiah's notions of 'performativity' (the coincidence of action and linguistic utterance, as in the binding spells) and 'persuasive analogy' for hermeneutic assistance.[[5]] Focussing on the grammar of the ritual acts and on their semantic value in context (e.g. the use of the wax voodoo doll which conveniently melts at low temperatures), Graf considers how ritual communicates and concludes that, in the case of the isolated magician, the sender and recipient of the message are the same person: the group is missing. In contrast, literary magic, addressed to a group of listeners and readers, corresponds, structurally, to communal oath- making, rather than magic ritual. Because the magician works alone, deprived of the group's memory, he employs figurines and texts to 'freeze' the memory of the ritual act. Graf stresses that this isolation, evident in healing texts as well, constitutes the fundamental difference between magic rites and those of the civic religion; furthermore, the magician's search for close communion with the divine parallels the importance attached to the vertical dimension in group rituals.

Noting that the submissive attitude (traditionally thought to characterise religion as opposed to magic) is found in the sorcerers' hymns, Graf considers the wide range of coercion found in the magical papyri, 'from the cruellest constraint to the most obsequious submission' (p. 225). Coercion does characterise magic, but it is only employed if more benign charms fail. Criticism of coercion began with the Hippocratic text On the sacred disease and with Plato, was perpetrated by Christian polemic, and inherited by scholars like Frazer and Festugière, who thus used this feature of magic to distinguish it from religion. However, magicians use a variety of traditional religious attitudes and positions in new and interesting combinations, usually characterised by reversals--for example, animals for sacrifice are killed in different ways (strangulation is preferred so that blood was not spilt); the sacrifice of roosters and donkeys instead of pigs, sheep and cows; the sorcerer and god dine alone; the search for contact with infernal divinities and the dead.

With respect to the latter, Graf comments on what he considers to be the most significant development in the history of ancient magic. Fixation on the realm below certainly characterised magic in the classical period, in which binding spells predominated, but the emergence of the realm above, in the form of the Supreme God, who can coerce demons, heroes and the dead, in the papyri of the imperial epoch, is consonant with political realities of the period, tendencies in neo-Platonic thinking and the development of the extremely hierarchical pantheon of the imperial era. Magic had changed from a technique used to outwit or injure one's adversaries in an agonistic context to a quest for knowledge and 'spiritual well-being' (p. 223), similar to that in gnosticism and neo-Platonism.

Graf's attempt in this work to demonstrate that magic, far from being opposed to religion, uses religious rituals and liturgies in unexpected ways to create new meanings, exposes the problems with the hermeneutic approach he adopts: Greek and Roman authors clearly distinguished between magic and religion (as Graf makes clear in chapter 2), but distinctions are not clearly made in the papyri which form the bulk of the evidence Graf uses. Consequently, Graf oscillates between emic and etic stances; he anxiously desires to show that magic is part of religion, but simply ends up finding new and significant differences between the two.

Furthermore Graf's attempt to cast out the demon of 'sympathetic' magic leads back to the definitional problems he raised in his opening chapter. If one writes about magic in, say, a study in Basle, peopled by the stern spirits of structural functionalism and the sensible wraiths of ritual theory, in a society in which not only god, but the individual subject, died some years ago, then a belief in 'sympathetic' magic must seem absolument ridicule; however, if one lives in an African society where old women are still being burnt as witches, where muti (medicine) shops exist within a bone's throw of the university, where there is a strong belief in a universe thronged by spirits and forces and demons, a field of forces, as it were, to which the sorcerer has access, where one's students testify to the existence of voodoo dolls (most definitely similia similibus), smeared with ousia, then rejection of the concept of 'sympathetic' magic, of the belief in a law of sumpatheia running through the cosmos like an electric charge, becomes more difficult to reject. In an African milieu, the binding spells and the voodoo dolls of the ancient magician make more sense within a framework of 'sympathetic' magic, than they do within the doubtlessly more fashionable framework of 'performativity' and 'persuasive analogy'.

I can accept that Frazer's world view and Father Festugière's were shaped by their Christian beliefs, that their distinctions betweeen magic and religion were based on many false dichotomies, yet in his efforts to demonstrate that magic in antiquity was a specific form of religion, which reverses many aspects of traditional civic religion, Graf perpetuates the construction of magic and the magician found in the Greek and Roman literary sources. In the process he does not exorcize Frazerian spirits, but, like a true magician, conjures them up in new and unexpected combinations, filtered through the theoretical gaze of the late twentieth century.[[6]]


[[1]] Magic in the Ancient World was originally published as Idéologie et Pratique de la Magie dans l'Antiquité Greco-Romaine (Paris 1994).

[[2]] For example, the dead man in his tomb is a kind of 'infernal postman', delivering messages to the divine or demonic addressees (p. 131); and, on the concept of sympathetic magic, 'there are always spirits to be exorcized, notably Frazerian spirits' (p. 145).

[[3]] See Auguste Audollent, Defixionum Tabellae (Paris 1904).

[[4]] See The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece (London 1990) 89-90.

[[5]] See Stanley Tambiah, Culture, Thought, and Social Action: An Anthropological Perspective (Cambridge, Mass. 1985) 123-166.

[[6]] As the original French edition of this work was unavailable, I am unable to assess the worth of Philip's translation. However, some of the phrases in the text read like clumsy literalese: for example, 'élite milieus of the ending republic' (p. 57); 'But we also find in the papyri a whole gamut of rituals that transform the status of a person who is already a magician to promote the person to a higher level of power' (p. 105); 'That is also why the masculine sexual attractiveness, to appear beautiful in the eyes of the world, is a good that can help the ancient male to acquire some social status' (p. 187); 'But there is one crucial difference towards the Greco-Roman material' (p. 227).