My contribution to this festive volume deals with untimely death and one specific ancient reaction to it, the accusation of sorcery.
Both victims and perpetrators were often women, although not always - but still, I regard this as a valid contribution to this conference on women’s religious roles in the ancient Mediterranean world.[661] [662] [663] Past research on magic in the Graeco-Roman world has almost exclusively focused on a few standard sources, on literary texts of all epochs that talk about magic, and on the lead curse tablets and GraecoEgyptian magical papyri that attest to its practice. A. Some years ago, my attention was caught by an inscription in the collection of antiquities in the French Bibliotheque Nationale in Paris. It is a dark stone with a long grave epigram; it was written around the year 100 b.c.e. in Alexandria. The text is composed in iambics, a rather unusual meter for grave epigrams that most often were written in elegiac couplets, and the stone has been inscribed by a somewhat unprofessional hand.[664] The text opens with the customary formulaic address to a deceased person, basically a greeting. Then, we hear two voices, the voice of the deceased, Thermis, a young wife and mother, and the voice of the surviving husband, Simalos. Thermis starts with a short prayer to the Underworld gods, asking to admit her to their realm; such a prayer is rare in these texts. At the same time, her introduction gives the necessary information about her name, family, and personal situation. Then she laments her fate, the terrible way of her death that she ascribes to 0dpaKa, poison or spells; she curses the person or persons responsible for her death and prays for blessings for her young children and her husband, Simalos. Both voices, the one of the deceased person and the one of the survivor, are conventional in Greek grave epigrams; but it is rare that they are combined into a complex dialog beyond the grave, as they are here. It shows the intensity of the feelings of the mourning husband who must have written this text himself rather than have it commissioned to a professional poet or scribe: given the disjunction between the poetic style and the mediocre lettering, the husband who set it up for his wife invested more care into formulating the text than in having it written on stone. He is very conscious of its poetic form, to the extent that he invokes the Muses in a way only a poet would. At the same time he does not use the high- style and widely used elegiac couplets or hexameters, but the simpler iambics. Iambics are somewhat easier to handle than elegiacs; more importantly, they are the meter of curse poetry, among other things, and thus might be thought apt for this text with its elaborate curse. The intense emotions that made Simalos write these verses must have been caused by the fact that Thermis died relatively young, with her children still well under age; but, more importantly, these emotions resulted also from the remarkable and cruel way she died, killed by a wasting illness that lasted for three long and terrible months. Thermis, however, does not simply talk about her death that has cut short a happy marriage: she retaliates by cursing whoever did this to her. The unknown perpetrators shall suffer the same terrible death, and their children shall die as well: the entire family shall be made to disappear from the face of the earth ‘with all their roots’ (v. 17). The divinities of her present realm, Persephone and her daimones, shall help this curse to become true - as, somewhat unexpectedly, these same divinities are also asked to help to fulfill the prayer for long life and happiness of Thermis’s husband and children. The epigram combines two topics that by themselves are seldom attested in antiquity, and more rarely combined, although they belong intimately together. One topic is the explanation of an unexpected death by an evil spell, the other is the curse that the victim puts on the (mostly unknown and even unknowable) person who sent such a spell, in order to take her revenge. B. Many pre-modern societies coped with the extraordinary disruption of life caused by an unexpected and early death of a person by accusing someone of having caused it through sorcery; the same mechanism was available for other instances of a sudden crisis. E. E. Evans-Pritchard has given an anthropological description and explanation of this mechanism among the Azande of Sudan that is still a classical study; Jeanne Favret-Saada did the same for the Bocage, a rural region of contemporary France.[665] In a much wider perspective that aimed at a historical phenomenology of human reactions to death, the Swiss folklorist and classicist Karl Meuli collected a wealth of evidence from European folklore on death customs and beliefs that contain many cases of the same mechanism.[666] From their reading of Roman literature, philologists know of two prominent cases, the death of the young prince Germanicus in the fall of 18 b.c.e., and the death of Apuleius’s friend and stepson Pontianus, the son of his wealthy wife; both died in their late twenties or early thirties. Pontianus died of a (presumably somewhat prolonged) illness when he was away from his home-town in Carthage, Germanicus in the middle of a power struggle with the governor of Syria. In Germanicus’s case, it was his friends and especially his widow, Agrippina, who accused the governor Piso and his wife of sorcery, and even produced a witch, Martina, who conveniently died on her way to the trial in Rome (Tacitus’s detailed and colorful account of the affair presumably goes back to the memoirs of Germanicus’s daughter, Agrippina the Younger).[667] In the case of Pontianus, the accusation was brought forward by his uncles as part of the overall accusations of magic. In both cases, the accusation of murder never made it into court. Apuleius’s accusers quickly dropped the charge and focused on another instance that should prove Apuleius guilty; Piso was accused not of magic but of insurrection before a senatorial court and committed suicide; the minutes of the court proceeding, found recently in an inscription in Spain, do not even mention the suspicion of sorcery and foul play.[669] Obviously, although accusations of sorcery were quickly made when someone young died of a mysterious illness, they remained inside the circle of family and friends and were not easily shared by the Roman system of law. C. Grave inscriptions from both the Greek East and the Latin West preserve more cases, albeit with less telling details than the two literary accounts. Grave inscriptions can reflect private attitudes to a much higher degree than most other ancient documents. I will attempt to give an overview and focus on a few select cases drawn from an exhaustive catalog I published recently. Most inscriptions from the Greek world formulate the accusations in the same vague way as the Alexandria epigram that accuses someone of having “directed the terrible Erinys of pharmaka against my entrails and my life” (verse 6). “In case someone gave him pharmaka” is the cautious expression in another inscription from Alexandria.[670] A short inscription from Side in Cilicia, for the grave of a girl or a young unmarried woman, “Hermione, daughter of Hermogenes, from Aigeai,” is only somewhat less ambiguous: Hermione is described as “the victim of pharmaka” (pepharmakeumene), that is “poisoned” or “bewitched.”[671] We lack a context, but the wording points to a routine suspicion of sorcery rather than to accidental or intentional poisoning. One Greek inscription only is somewhat circumstantial and attempts to define the pharmakon that killed the deceased. It is the grave epigram of a young man who belonged to a leading family on the small island of Andros.[672] The deceased - one Abaskantos who derived his lineage from the mythical king Aeacus and who was buried together with his child - must have been a professional athlete or gladiator; he did not die as Achilles by love potions, but like Ares “through unholy plants,” on/ ooiais ^OTavais. The myths to which this text alludes are unknown; if one wants to make sense of them, however, one has to assume that they elaborate on well-known stories. It is at least conceivable that Achilles’s infatuation with Polyxena that lead to his death, according to most late sources, was understood as caused by a love-potion; and one could explain the fact that Otos and Ephialtes were capable of capturing the mighty Ares by making them use a harmful drug. None of these possible elaborations, however, is attested in our ancient texts. Whatever the mythology is, the overall sense of the statement seems clear: a harmful potion, concocted from poisonous plants, caused the early death of Abaskantos. The family takes some care to dispel the notion that it was a love potion: this must have been part of the gossip surrounding this prominent death. The very precise terminology, philtra, points to a practice that could be blamed for the unexpected death of young men; the “unholy plants” by extension then belong to the same realm of poisoning/sorcery, pharmakeia. The deceased, after all, was young and healthy: he should have had admirers and lovers, not enemies. Similar accusations come from the Roman West. One text, a grave inscription for a four year old child, outrightly accuses a saga, a witch, for the death; it warns the parents “guard your children, lest grief will attach itself to your heart.”[673] Another text, from Roman Africa, is the grave inscription of a twenty-eight year old woman who, like Thermis, died from a slowly killing disease: “Bewitched by spells, she was lying ill for a long time, as life was forced out from her.”[674] The grave inscription of a young freewoman in Salona (Dalmatia) who died when twenty-three years old - “at a flourishing age,” florente aetate - after an illness of a year and five months, ascribes this painful death to sorceresses, veneficae, whoever they were.[675] [676] According to a late and presumably Christian text from Bulgaria, a young wife “died through sorcery,” per maleficia de secolo obit.1 There is no instance among these Latin texts where we would doubt that we are dealing with what in our thinking too would be sorcery. Accordingly, in the few epigraphical texts that directly refer to venenum, “poison,” we can assume that the authors really imagined that the death had been administered by harmful potent substances. But as in the case of Germanicus, the deceased were young - a three year old boy in one case, a young man of twenty-five who “defended himself against injustice” in another one.[677] We do not know in either case whether the relatives also prosecuted the suspected poisoner or whether they were content with publicizing their anonymous suspicion on a tombstone. D. The Andros case let us guess at some of the gossip that surrounded untimely death. Gossip and rumors seem important in this world; suspicions were not brought to court but circulated in the village as soon as something unexpected happened. Rumor also pointed to the person one suspected as the perpetrator - and since there was no official accusation, the defense itself was difficult; only the trained orator Apuleius was capable, and daring enough, to provoke a lawsuit against himself. Lesser people, when accused, could not do this, but they had recourse to another form of public self-defense. We know this from a few Hellenstic lead texts found in the sanctuary of Demeter and Kore in Knidos where they must have been affixed to walls or trees to be read by everyone. One invoked the goddess against “whoever says that I make a pharmakon against my husband”: this, then, is a rumor which the petitioner wanted to stop - if we can take the present tense literally, she was accused of still being at it, and her husband might have been ill, but he was still alive.[678] The second text is more serious, and more circumstantial: the writer called the wrath of the goddess upon herself, “if ever I have given Asklepiadas a pharmakon or intended something evil in my heart against him, or called a woman to the temple so that she, for three half-drachmas, would remove him from among the living.”[679] The defense is accompanied by a self-curse in case the defendant would be cheating, and the seriousness of this self-curse must mean that Asklepiadas had died, and that his death was ascribed, directly or indirectly, to the sorcery/poisoning of the speaker, with several rumored stories of how she went about her crime, either doing it herself or engaging someone else, a local sorceress. Slander and rumor again, then, but of a more insidious kind, destined to ruin the reputation and the social standing of a widow. A very different text comes from further East and is dated a few centuries later. In one of the so-called “confession inscriptions” from Eastern Lydia, a woman named Tatias is accused of having used a pharmakon against her son-in-law who suddenly had become insane.[680] She defended herself with a public confession of her innocence, invoking the local god, Men, in a self-curse. She promptly died: this proved both her guilt and the god’s power. Shortly afterwards her son had a fatal accident: this made the god’s point even more clear. The mechanism, then, is very similar to what we perceive in Knidos. Misfortune strikes a family; a suspicion of foul play arises and is directed against a woman (the wife of a deceased or ill husband, the mother-in-law of a mentally ill man). Rumors spreads, and the woman has to defend herself publicly by having recourse to the gods. E. Rumor is not the only way of seeking revenge against suspected sorcery. Two almost identical Greek inscriptions from late Hellenistic Rheneia, the small island off Delos, explain the untimely death of a girl through pharmakeia. Formally, they are not grave inscriptions but curses, but since Rheneia served as the cemetery of the island of Delos on which it was forbidden to bury humans, the two texts must come from two graves.[681] They date to the late second or early first century b.c.e. and were written by Hellenized Jews, presumably Palestinian traders who settled in the international trade center that the island of Delos had become in this period. The wording of both texts is identical, with the sole exception of the name of the victim (Marthina in one case, Heraklea in the other one), and they were inscribed at about the same time, to judge from the letter forms. They invoke Iahwe in a formula known from the Septuagint to take revenge “on those who have cunningly murdered or poisoned poor Marthina/Heraklea who died before her time, unjustly shedding her innocent blood.” A relief of two raised hands depicts the gesture of invocation that accompanied the curse prayer. Both girls died young, but the circumstances of their death and the reasons for the curse are unclear. They were not openly murdered but died in a way that again triggered the suspicion of foul play to which someone - the parents? the community? - reacted with the invocation of an all-powerful god to punish the unknown culprits; to inscribe it on stone guaranteed the permanence of the speech act involved. Since we possess only these two monuments, we cannot know whether something unusual caused this suspicion, or whether this was a standard way of dealing with the sudden death of young people in this specific community. We can only state that the way the texts describe the accusation insists on the stealthy character of the killing: the formula “those who have cunningly (SoXmi, “by secret action”) murdered or poisoned,” describes any secret way of killing, in an almost overly precise way. In pre-modern societies, a curse is a judicial instrument that is used when other legal steps to punish a culprit are impossible, for whatever reason. In such a situation, humans called upon their gods for just retribution, and they could make this curse public to make it more effective. An inscription from the island of Chios, written in the later sixth century B.C.E., illustrates this mechanism almost in a textbook way.[682] The text, promulgated by the political authorities of the island state, orders the authorities to set up border markers, presumably to mark land confiscated during one of the political upheavals of the epoch. In such a partisan atmosphere, the temptations to remove the markers must have been great. Therefore, the law stipulates a heavy fine for any person who would remove a marker, and it defines the enforcing magistrate and the magistrate who supervises the enforcement. Each supervising magistrate in turn is threatened with a fine for neglecting his duty, all the way up the hierarchy; each time, the next level enforces the fine, the following level supervises the enforcement. The highest magistrate, however, is treated differently, for good reasons. A fine cannot be enforced any more, since there is no higher political authority. Instead, he is cursed in case he would neglect to enforce a fine due to a removal of a border marker by an inferior magistrate. At the end of the ascending chain of political responsibility, we find the gods; they take over as the enforcers of retribution when humans are unable to do so, and they are called into action by a curse. In this public context, a curse is used to punish offenders because there is no human authority left to enforce a punishment. In private contexts, curses are more often used because one did or could not invoke the public authorities for help, either because there was no law available, because the accuser could not make a case that would stand in court, or because they were not wealthy enough for litigation. In both spheres then, the public and the private, a curse was a powerful instrument to enforce justice. Many among the grave inscriptions that contain curses fit into this same pattern. A typical case is a Roman grave inscription for a certain Grattius, a young man of twenty-three years who was killed in an unspecified act of street violence. The very nature of the crime makes it almost impossible to determine a culprit, and the Roman legal system that left investigation and prosecution to private initiative made a trial under those circumstances impossible.[683] A grave inscription from Imperial Egypt, set up for three sailors who were killed in a river port, invokes the god Sarapis as the avenger of this crime; the perpetrators remained unknown, and it is unclear whether anyone investigated the crime.[684] In a rare instance from Syria, the murderer was unassailable because he was a senior administrator, and the family of the victim (“killed for nothing”) understandably did not dare to bring him to court.[685] In another isolated case, the relatives of the victim did not curse the actual killer but another person involved in the case who, however, could not be reached by the law: in the grave epigram of a young man killed by the lover of his wife, the deceased, one Aphrodisios, curses his wife (“may Zeus ruin her!”).[686] The actual murderer might well have been brought to court for murder, but no ancient law would incriminate the wife as well (at least not without clear evidence of her help). But the relatives who had the stone put up obviously regarded her as as guilty as her lover. A more complex case that leads back to magic comes from Imperial Rome; it can be pieced together from the inscriptions on an impressive funerary altar for one Iunia Procula who died at the age of eight years.[687] The main inscription on the front side of the altar gives the name of the deceased girl, after the indication of her age (vixit annos VIII menses XI dies V) and the information that she left father and mother in grief. The name of the person who had the monument made not only for his daughter, but also for himself and another person, the freedman M. lunius Euphrosynus, occupies the bottom line of this side. The line has two lacunae. The first is small and accidental; it must have contained the abbreviated name of the patron who set him free. The second gap is the intentional erasure of the name of the other recipient of the monument, presumably the spouse of M. lunius Euphrosynus; thus, we deal with a rare case of private damnatio memoriae. A second, lengthy inscription on the back of the same monument explains the erasure and reveals the social drama that led to it. It is written in (rather shaky) iambics. Fittingly enough for this meter, it is a curse, as is the Alexandrian grave epigram: although at this time iambics could be used for all sorts of purposes, their function as medium of public blame and curse in earlier poetry could still be accessible. The curse was pronounced against the freedwoman Acte, venenaria et perfida, “a treacherous sorceress.” She cheated on the man who freed her without charge (presumably to live with her), and when he fell sick, she robbed him of everything, including some of his servants. Hymnus and Zosimus, who are cursed as an afterthought and whose Greek names indicate that they must be slaves, might well be the adulterer and the eloping slave boy. [Actae liberta]e or, more officially, [Iuniae M. l. Acta]e (nine letters) then is a possible restitution for the gap on the front side. We can guess the drama in the house of M. Iunius Euphrosynus. A Greek freedman himself, as his Greek cognomen shows, he fell in love with the Greek slave girl Acte and freed her, trading freedom for sex; his claim that he freed her without asking to be paid is true only in a financial sense. But she cheated on him, and when he fell ill after the death of his small daughter lunia Procula, presumably their common child, she left him and took his two slaves with her; Iunius did not survive the shock. His heirs suspected magic (love magic as the root of his infatuation, binding spells as the reason for his illness, or both), banned her from the family tomb, cursed her and, as an afterthought, added the slaves that followed her to the curse. The curse was inscribed on the tomb of Iunia Procula because she was a dead “before her time,” an ahoros in Greek terminology, and thus was thought to be an ideal messenger and helper with the Underworld gods to whom a curse was implicitly addressed. In theory, Euphrosynus’s heirs could have brought Acte to court, as did the relatives of the widow Apuleius married in African Oea. But presumably Acte had moved out and disappeared from Rome; and at any rate, litigation was costly and more of an upper-class matter. These curses are not substantially different from other curses we can read on grave stones; since ancient tombstones were not hidden away in cemeteries but lined the major roads outside a town, for every passer-by to read, this was tantamount to the publication of a private curse. In one text, a husband curses those to whom his deceased wife had made a downpayment, in case these persons would refuse to pay him back and thus exacerbate his heavy loss.[688] There were laws against such an embezzlement, but they need incontrovertible ancient proof, and that might have been difficult to obtain in a situation where one party to the payment was dead; furthermore, the grave inscription acted preemptively and must have intended to shame the debtors into compliance. From another text, it becomes clear an old man cursed his brother on the death-bed for having cheated him his entire life; he then had this curse inscribed on the tombstone as a public revenge.[689] F. To conclude, I want to highlight two results of my inquiry. One is the insight into the relatively small number of sorcery accusations that are attested in these records. No one ever has counted Greek and Roman grave inscriptions that are elaborate enough to give some indications of the death; any estimate would have to reckon with several thousand texts. Since the death of young people more often demands explanations than that of people deceased after a long life, their number in this corpus would be relatively high, certainly more than one thousand. In this entire corpus, I have counted only thirty-five relevant inscriptions, and another thirty that might perhaps be relevant. Given that this explanatory mechanism was so well established in many other cultures, this small number surprises, as does the gender imbalance: among the clear cases, we count eighteen women and girls as against twelve men and four very young boys. It seems that ancient societies, both in Greece and in Rome, did not easily yield to such accusations, but that they were more likely to do so for the less public and more private women and children: the accusations always remained on the level of suspicion, rumor, and gossip, and did not make it into the courtrooms. “Real” witches, that is, and formal accusations of sorcery and witchcraft in connection with an unexplained death were very rare in the real world of Imperial Greece and Rome, unlike what ancient and modern fiction would make us believe. The other insight that deserves to be stressed is that there is no clear genderization of the persons accused of having caused an early death through pharmakeia. Often, the person remains undefined, both as to gender and to identity. In a few cases, a woman is accused, such as the courtesan Theoris of Lemnos in a notorious case of love magic in fourth century Athens; the otherwise unknown Martina in connection with Germanicus’s death; the ‘evil’ liberta Acte or an anonymous saga in Imperial Rome, or unspecified veneficae in another Roman epigram.[690] More rarely, the accused is a man - the philosopher Apuleius, an unknown magus in a Clarian oracle from Western Asia Minor, the freeman Atimetus in Imperial Rome. Compared to the stereotype of the female witch that we find in Greek and Roman literature, the reality “on the ground” is much more complex. Works Cited Bean, George E. The Inscriptions of Side. Ankara, 1965. Bernand, Etienne. Inscriptions metriques de l’Egypte greco-romaine. Paris: Belles Lettres, 1969. Besevliev, V. Spätgriechische und spätlateinische Inschriften aus Bulgarien. Berlin, 1964. Bjorck, Gudmund. Der Fluch des Christen Sabinus. Papyrus Upsaliensis 8. Uppsala: Almqvist & Wiksell, 1938. Blümel, Wolfgang. Die Inschriften von Knidos. Inschriften griechischer Städte aus Kleinasien 41. Bonn: Habelt, 1992. Collins, Derek. “Theoris of Lemnos and the Criminalization of Magic in Fourth-Century Athens.” Classical Quarterly 51 (2001): 477-93. Cumont, Franz. “Il sole vindice dei delitti ed il simbolo delle mani alzate.” Memorie della Pontificia Accademia Romana di Archeologia, Nuova Serie 1 (1923): 65-80. Eck, Werner, Antonio Caballos, and Fernando Fernandez. Das Senatus Consultum de Cn. Pisone Patre. Vestigia 48. Munich: Beck, 1996. Evans-Pritchard, Edward E. Witchcraft, Oracles and Magic Among the Azande. Oxford: Clarendon Press, 1937. Favret-Saada, Jeanne. Les mots, la mort, les sorts. La sorcellerie dans le Bocage. Paris: Gallimard, 1977. Graf, Fritz. “Untimely Death, Witchcraft, and Divine Vengeance. A Reasoned Epigraphical Catalog.” Zeitschrift für Papyrologie und Epigraphik 162 (2007): 139— 50. Grubbs, Judith Evans. “Stigmata Aeterna. A Husband's Curse.” Pages 230-42 in Vertis in Usum. Studies in Honor of Edward Courtney. Edited by John F. Miller, Cynthia Damon, and K. Sara Myers. Munich; Leipzig: Saur, 2002. Jeffery, Lilian H. The Local Scripts of Archaic Greece. Oxford: Clarendon Press, 1990 (1961). Kaibel, Georgius. Epigrammata Graeca ex lapidibus conlecta. Berlin: Reimer, 1878. Reprinted Hildesheim: Olms, 1965. Le Bas, Philippe and William Henry Waddington. Voyage archeologique en Grece et en Asie Mineure, 3. Paris, 1870. Meuli, Karl. “Lateinisch ‘morior' - deutsch ‘morden.'” Pages 439-44 in Gesammelte Schriften. Edited by Thomas Gelzer. Basel: Schwabe, 1975. Nolle, Johannes. Side im Altertum. Inschriften griechischer Städte aus Kleinasien 44. Bonn: Habelt, 2001. Peek, Werner. Griechische Vers-Inschriften. I: Grab-Epigramme. Berlin: AkademieVerlag 1955. Petzl, G. Die Beichtinschriften Westkleinasiens. Epigraphica Anatolica 22. Bonn: Habelt, 1994. Ricl, Marijana. The Inscriptions of Alexandreia Troas. Inschriften griechischer Städte Kleinasiens vol. 53. Bonn: Habelt, 1997. Robert, Louis. Collection Froehner. I: Inscriptions grecques. Paris: Editions des Bibliotheques nationales, 1936. Solin, Heikki. “Analecta Epigraphica LXVII-LXXVIII.” Arctos 15 (1981): 101-24. Stephen, Michele. “Witchcraft, Grief, and the Ambivalences of Emotion.” American Ethnologist 26 (1999): 711-39. Tupet, Anne-Marie. “Les pratiques magiques a la mort de Germanicus.” Pages 345-52 in Melanges Pierre Wuillemier. Paris: Gallimard, 1980. Yorke, V. W. “Inscriptions from Eastern Asia Minor.” Journal of Hellenic Studies 18 (1898): 306-27. Zingerle, Josef. “Heiliges Recht.” Jahreshefte des Osterreichischen Archäologischen Instituts 32 (1926): Beiblatt cols. 50.