Hostname: page-component-76fb5796d-r6qrq Total loading time: 0 Render date: 2024-04-25T12:48:21.784Z Has data issue: false hasContentIssue false

CONTINGENT CATASTROPHE OR AGONISTIC ADVANTAGE: THE RHETORIC OF VIOLENCE IN CLASSICAL ATHENIAN CURSES

Published online by Cambridge University Press:  07 March 2022

Radcliffe G. Edmonds*
Affiliation:
Bryn Mawr College, USA
Rights & Permissions [Opens in a new window]

Abstract

The surprising absence of violent language from classical Athenian curses is best understood as a rhetorical strategy appropriate for getting the divine powers to enact the curser's desire to harm his or her enemies and to gain an advantage in the particular agonistic context. A contrast with the extravagantly violent language of other contemporary curses, which call for unmitigated catastrophe to befall their targets, shows that the fundamental difference between these curses is the audience that they primarily address, which shapes the nature of the request that is made in the imprecation. Whereas contingent curses primarily address the human community with highly intense rhetoric to deter potential violation, these agonistic curses against rivals request assistance in the rivalry from some power beyond the human community, limiting the extravagance of the request to improve the chance of fulfilment.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press on behalf of The Classical Association

I bind Dionysios the helmet maker and his wife Artemis the goldworker and their household and their work and their products and their life…Footnote 1

This curse, aimed at the lives and livelihood of some artisans in classical Athens, is typical in the simple and restrained language with which the malevolence of the curser is expressed, but the lack of violent rhetoric has surprised and troubled interpreters. Scholars have debated whether the curser wishes to hold back his business rivals in a non-violent fashion or whether this lead tablet pierced with a nail actually, as Riess puts it, displays ‘a considerable amount of violence under the surface of relatively tame language’.Footnote 2 Faraone, by contrast, has argued that the language of these curses, ‘clearly suggests that their main motivation was restraining or inhibiting, not destroying, the victim’ – a restraint that was itself, perhaps, motivated by a scruple against killing a fellow citizen.Footnote 3

I argue that the ‘relatively tame language’ of these classical Athenian curses is best understood, neither as a veil for hidden violence nor as a reflection of an anxiety about shedding the blood of rivals, but rather as a rhetorical strategy appropriate for getting the divine powers to enact the curser's desire to harm his or her enemies and gain an advantage in the particular agonistic context. A contrast with the extravagantly violent language of other contemporary curses, which call for unmitigated catastrophe to befall their targets, shows that the fundamental difference between these curses is the audience that they primarily address, which shapes the nature of the request that is made in the imprecation. Whereas other curses primarily address the human community to deter potential violation, these curses against rivals request assistance in the rivalry from some power beyond the human community.

Faraone has dubbed ‘agonistic’ the curses that seem to aim at restraining the activities of rivals, and, although other scholars have pointed out that the social dynamics of envy (phthonos) that motivate these curses are more complex than the basic model of someone who fears defeat turning to a curse to even the odds, the underlying model of competition and rivalry provides the best way of understanding the context in which such curses were produced.Footnote 4 These curses seem to have been produced and deposited in secret and, even more significantly, never boasted of as the means by which one has bested a rival. In short, they are a means of cheating, fairly widespread perhaps, but generally acknowledged as an illegitimate way to win, to gain advantage in the agonistic situation.

By contrast, contingent curses are those that invoke retribution upon the target if a certain condition occurs. Tomb curses, which threaten destruction on anyone who disturbs the grave, fall into this category, as do the ceremonial curses that accompany oaths or form part of civic ceremonies like the convening of the Athenian assembly. If someone violates the tomb or breaks the oath or aids the Persians against Athens, then the gods are asked to bring catastrophe, which is often described in violent and extensive terms.

A curse is essentially a wish for something harmful to happen to the target. Since humans tend not to be able to make things happen just by force of will, some sort of extra-human power is needed to make the wish efficacious. Some divine power must intervene to create the desired effect, doing a favour for the mortal who wishes that harm befall another. This wish for extra-human intervention can be as simple and automatic as an uttered word, a performative utterance such as ‘I bind’, or as complex and contingent as an elaborate cursing ceremony that targets potential wrongdoers who have yet to take an action that would incur the curse. Any such wishing for divine assistance, however, constitutes a ritual act whose performance can be analysed on two axes of communication: the vertical, from mortal performer to divine audience; and the horizontal, from mortal performer to mortal audience.Footnote 5

The audiences on both axes are simultaneously present for any ritual act, but the performer may be focusing the performance on one axis or the other, depending on the nature of the ritual. Civic parades, for example, may nominally honour some divine power (or the spirits of deceased heroes), but the main ritual performance is aimed at the audience of the community on the horizontal axis. The rhetoric of that performance, therefore – the way in which the ritual is performed – is designed to have its effect primarily on the community audience, communicating, for example, the might and power of the military and the wealth of the community leaders who have sponsored the procession. By contrast, a small token offering left at a shrine or a grave may have an audience on the horizontal axis of other worshippers who come by, but the primary audience of the ritual act is on the vertical axis: the deity or deceased spirit for whom the offering is made. The cakes or flowers may signal to others that someone has visited the site, but the ritual itself, with prayers spoken and other actions taken that leave no trace, is designed to communicate not with other mortals but with the divine.

Strategies of curse prayers

In order to understand the surprising lack of violence in the agonistic curses, I suggest that analysing curses as prayers that request a divine power to effect something that the mortal performer could not otherwise achieve enables us to focus on the particular rhetoric of the curse requests to distinguish different kinds of requests and the different audiences for which they are performed. Contingent curses, which primarily appeal to the audience on the horizontal axis, tend to describe more elaborate and extensive harmful effects for the target, whereas agonistic curses, which have a very limited audience on the horizontal axis and appeal primarily to the vertical, tend to be more limited in the effects for which they ask. Both types of curses, however, may employ rhetorical devices in their ritual performances to increase the emphasis of their statements, devices that range from marked language patterns to persuasive analogies acted out in the ritual.

The rhetoric of contingent curses is generally aimed at preventing the contingency, at deterring a potential target, rather than providing a means of dealing with the contingency when it occurs. The focus is thus on the horizontal axis, the audience of other mortals, rather than the vertical axis. The rhetoric tends to be vivid and expressive, painting a picture of what would happen in the most graphic terms, to emphasize the point that it should not happen. The harm can be more extensive and extravagant, since the request is still contingent, rather than actual. The self-cursing in oaths provides a clear example of such a contingent curse strategy, but other kinds of contingent curses may be found: both private, as in tomb imprecations or personal agreements, and public, as in the curses invoked in civic ceremonies.

The locus classicus for the oath curse is, of course, the scene in the Iliad where the Greeks and Trojans swear to abide by the outcome of the duel between Menelaus and Paris. Agamemnon provides a vivid picture of what should happen to those who break the promise:

He spoke, and cut the lambs’ throats with the pitiless bronze; and laid them down on the ground gasping and failing of breath, for the bronze had robbed them of their strength. Then they drew wine from the bowl into the cups, poured it out, and made prayer to the gods who are for ever. And thus would one of the Achaeans and Trojans say: ‘Zeus, most glorious, most great, and you other immortal gods, whichever army of the two will be first to work harm in defiance of the oaths, may their brains be poured out on the ground just as this wine is, theirs and their children's; and may their wives be made to serve other men.’Footnote 6

Not only should the oath-breakers suffer a brutal death themselves, but their wives and children should also suffer for what they have done.

Such an extended curse, with elaborate consequences for both the oath-breakers and their families, appears in the classical Athenian context in the oath of the Amphictyons from the First Sacred War that Aeschines quotes:

They were not content with taking this oath, but they added an imprecation and a mighty curse concerning this; for it stands thus written in the curse: ‘If any one should violate this’, it says, ‘whether city or private man, or tribe, let them be under the curse’, it says, ‘of Apollo and Artemis and Leto and Athena Pronaea.’ The curse goes on: That their land bear no fruit; that their wives bear children not like those who begat them, but monsters; that their flocks yield not their natural increase; that defeat await them in camp and court and market-place, and that they perish utterly, themselves, their houses, their whole race; ‘And never’, it says, ‘may they offer pure sacrifice unto Apollo, nor to Artemis, nor to Leto, nor to Athena Pronaea, and may the gods refuse to accept their offerings.’Footnote 7

Again, the curse calls for a large number of bad things to happen to the potential targets, utterly destroying their lands, their family structures, their ability to succeed in all competitive endeavours, their relationships with the gods, and ultimately the lives of them and all their kin.

The aim of the curse is not to petition Apollo and Artemis and the rest of the gods invoked to take matters into their own hands and deal out such punishments, but rather to convince all the participating Amphictyons that they should ensure that the land sacred to Apollo is not appropriated by someone else. The rhetoric of the curse is primarily targeted along the horizontal axis at the Amphictyons, with the vertical axis of the gods as merely a secondary audience whom, the performers hope, will never need to become involved.Footnote 8

Oaths are not the only context in which such vivid and elaborate contingent curses are performed for a public audience. The Athenian assembly began with a ritual cursing of anyone who might be working against the interests of the community, for example by plotting to establish a tyranny or to give aid and comfort to the Persians. Unfortunately, we have only allusions to how dire this curse was, along with a parody version in Aristophanes’ Thesmophoriazousai. Aristophanes throws in a few other types of potential wrongdoers for jokes (slaves who betray their mistresses’ affairs or old women who lure away lovers with expensive gifts), but the gods are asked to utterly overwhelm the wrongdoers – and, again, not just them but their families as well.Footnote 9 The people of Teos likewise invoked curses on those who violated the norms of the community as a way of trying to deter such violators.Footnote 10 Both Athens and Teos had legal measures in place for the mortal community to inflict punishment should anyone commit these violations, but the curse was an expression of the community's wish that no one should do those things, a ritualized communication on the horizontal axis to the other mortals in the community.

Such curses need not be performed by the whole community, but an individual may make a contingent curse in a public setting that targets the audience on the horizontal axis in order to deter violators. Most tomb curses appear in contexts other than classical Athens, but Demosthenes tells us of someone who added curses to his will to ensure that his bequests were carried out (and of course the fact that the subject comes up in a lawsuit tells us that they were not).Footnote 11

Such requests to the gods are generally far greater than the requests in agonistic curses tend to be, since the performer is not actually making the request in that performance but rather announcing his intention to ask for such vengeance. If the contingency should occur, then the performer would have a ready-made justification for asking the god for the favour. The violator had been warned not to take that action unless he wished to anger the god, and the fact that he or she acted anyway thus counts as a deliberate affront to the deity, who is now honour-bound to take vengeance.Footnote 12 Such a planned justification enables the performer to ask for a greater favour from the deity than someone who has no such justification, and so the rhetoric of violence and destruction may be far more exaggerated in a contingent curse.

The contingent curse is addressed primarily to the horizontal audience, so the harm in the curse is framed in extensive and exaggerated terms: utter destruction for the target and all those around. This requested effect, however, remains potential and contingent; the gods are witnessing that the performer might make such a request, but they are not being asked to act immediately. The immediate impact of the curse is the deterrent effect for the audience of other mortals, the horizontal axis of communication.

By contrast, the agonistic curses inscribed on lead tablets and buried in secret places like tombs or wells have an extremely limited audience on the horizontal axis, so their rhetoric tends to be focused on the vertical axis, the audience of divine (or infernal) powers invoked to carry out the curse. The rhetorical aim is to convince superhuman powers to inflict the harm on the target desired by the performer – not potentially in the future, but actually right now, so the prayer must be addressed to the deity or deities most likely to grant that favour. In the absence of other kinds of justification for the request, the performer is apt to ask for the minimum effect needed to achieve the desired result, rather than for an extensive miracle that utterly demolishes the target and all their relatives.

While later curse tablets expand the list of deities invoked, the curse tablets from classical Athens are addressed primarily to a limited set: first and foremost to Hermes and Hekate, with other underworld figures such as Persephone or the dead appearing as well. While some curses merely refer to the gods by their names, many include a significant epithet – Hermes the Restrainer (katochos) or Chthonic Hekate (chthonia). One curse names Hermes both as an underworld power and as the restrainer, along with other traditional epithets:

Gods, with good fortune. I bind and will not release Antikles, son of Antiphanes and Antiphanes, son of Patrokles, and Philokles and Kleochares (and Philokles) and Smikronides and Timanthes (and Timanthes). I bind all of these before Hermes of the underworld, the tricky, the restrainer, the powerful, and I will not release.Footnote 13

In another curse, Hermes the Restrainer is invoked along with the Praxidikai, a set of goddesses whose very name indicates their function as the executors of revenge, whereas another curse simply calls upon ‘the god who restrains’.Footnote 14 Of course, some curses do not specify the deity in writing, but the place of deposition in a grave or well indicates that some underworld power is being contacted, and the performer would no doubt have named the intended recipient orally while depositing the tablet.Footnote 15 Whether Hermes or Hekate, Persephone or the Praxidikai, or even the spirit of a restless dead, the divinities chosen are the ones whose nature is such that they are most likely to undertake the action the performer wishes.

Such a selection is important because most of these curses offer little in the way of incentive for the deity to grant the performer's request. Beyond epithets that highlight the chthonic and binding nature of the divinity addressed, the curse texts include very few of the laudatory epithets and descriptions that other prayers use to win the favour of the deity. Like other rituals that may be labelled ‘magic’, such curses involve a petition to the divine powers that focuses on the present moment, rather than appealing to the past reciprocal relationship of the performer to the deity or even promising reciprocation for favours in the future.Footnote 16 The curse performed by Chryses in the Iliad shows the contrast, since Chryses bases his appeal to Apollo to do harm to the Greeks on their long history, and all of the sacrifices and honours he has provided to Apollo over the years:

He went in silence along the shore of the loud-resounding sea; and then, when he had gone apart, the old man prayed earnestly to the lord Apollo, whom fair-haired Leto bore: ‘Hear me, you of the silver bow, who have under your protection Chryse and sacred Cilla, and who rule mightily over Tenedos, Smintheus, if ever I roofed over a pleasing shrine for you, or if ever I burned to you fat thigh pieces of bulls or goats, fulfill for me this wish: let the Danaans pay for my tears by your arrows.’Footnote 17

The lead curse tablets, by contrast, never make the argument that the divinity to whom the prayer is addressed should provide the favour of harming the target because of all that the performer has done in the past. Nor, as in the case of votives, is an offering usually promised to the god if the assistance is given when requested. One of the few exceptions that proves the rule is a lead tablet from Attica on which the performer promises to sacrifice to Hermes and the Praxidikai if they restrain his target, Manês.

Manês I bind and I restrain. And you, my own Goddesses of Vengeance, restrain that man; and Hermes the Restrainer, restrain Manês and the affairs of Manês, and make all the work that Manês is working at come out to contrary and backward results for Manês. To you, Goddesses of Vengeance and Hermes the Restrainer, if Manês fares badly, I will sacrifice thank offerings.Footnote 18

The rhetorical focus of the appeal, therefore, is not on past favours or even future promises, but on the present moment of need. Later magical recipes in the Greek Magical Papyri show that such appeals which are focused on the present and pressing necessity are often accompanied by sacrifices or other offerings performed along with the prayer. Archaeological evidence provides little help in confirming this pattern for the earlier agonistic curse tablets but, while there may have been libations or other small offerings that accompanied the deposition, most curse tablets provide no indication that offerings to the deity addressed were part of the strategy for inducing the divinity to grant the request.

A few curse tablets do provide other reasons that the divinity invoked should come to the performer's aid, arguing that, because the target has done wrong, he or she rightly deserves punishment. One curse consigns the target to Hermes and Persephone, because ‘it is just that they receive the penalty due’.Footnote 19 Another implores Gê to restrain the targets: ‘My own dear Earth, help me; and since I have been wronged by Euryptolemos and Xenophôn I bind them.’Footnote 20 These curses have often been classified by scholars as ‘prayers for justice’ because of their supplicatory tone, but only this slight shift in rhetorical strategy differentiates this lead tablet pierced with a nail and buried in the earth from others with the same kind of ritual deposition, the same focus upon the vertical – rather than the horizontal – audience.Footnote 21

In the absence of either arguments justifying the divinity's intervention or the kind of praise and offerings found in normative prayers, the rhetorical strategy of these agonistic curses relies on transferring the responsibility for action to the deity, shifting the agency along the vertical axis. The strategy of such curses also involves limiting the number and scope of specific requests for harm. Since the divinity invoked has little incentive to grant an extensive favour, the curse performer simply consigns the target to the divinity, relying on the deity to take care of the matter.

The suggested harm tends to be limited to preventing the target's success in some regard, just tipping the balance in the contest between the agent and the target rather than the wholesale destruction of the target and their family unto the generation of generations. Hermes the Restrainer is asked to restrain the target, sometimes in a particular regard: ‘Iphemuthanês and Androsthenês I bind, and Simias (and) Dromôn. To Hermes the Restrainer, the feet, hands, soul, tongues, business, and profit.’Footnote 22 In a curse aimed at breaking up a relationship, it is specifically the sexual activities that are targeted; the one who sits by the side of Persephone (presumably Hekate) is asked to make sure that Theodora remains unmarried, especially to Charias, and that Charias himself forgets his desire to have sex with Theodora.Footnote 23

Restraining and binding back are the most common requests in the agonistic curses, but often the action of the performer is simply to consign the target to the tender mercies of the underworld power invoked, ‘to Hermes Katochos’, ‘to Hekate Chthonia’, and so on. Riess compares the language of the curse tablets to the terminology used in legal and business transactions, further noting that, ‘In legal contexts, the accusative case is regularly used to refer to the magistrate who has jurisdiction over a case. This means that the agent of the curse renders the accursed person subject to the jurisdiction of the chthonic powers named.’Footnote 24 The registering – or, more often in the Attic tablets, the binding – of the target pros (‘over to’) the divinity serves to put the target into the power of that god, who is then expected to engage in his or her natural activities of restraining, punishing, or harming.Footnote 25

Only a few examples of curses show a bolder strategy of asking for more from the divinity, making a list of specific harms to befall the target and, in addition, all the target's associates.Footnote 26 Most agonistic curses pursue a comparatively limited rhetorical strategy of not making extravagant demands, since the performer has little to offer the divinity as inducement to do what he or she wishes at the immediate moment in which the deity's action is needed. The limits of the request's scope, however, do not preclude the use of rhetorical devices to enhance the force and power of the appeal.

The rhetoric of cursing

Although agonistic and contingent curses address different primary audiences for different ends, and thus employ different strategies to communicate their intent, they use a similar set of rhetorical devices to enhance the power of their expression. Certain verbal devices, such as repetition, pleonasm, or specialized language, are often complemented by ritualized actions that are likewise chosen for their expressive effect. The metaphoric or metonymic analogies of such acts are at times also expressed verbally in the curse itself, creating a multi-media rhetorical effect.Footnote 27 Although scholars have often understood these acts as examples of Frazerian ‘sympathetic magic’, such a classification obscures the more significant differences of audience and ends, making the rhetorical manipulation of material the criterion for magic, rather than factors such as the legitimacy of the performance or the sociopolitical position of the performer. As I have argued, the extraordinary nature of the performance or sociopolitical position are two of the most valid criteria within the ancient evidence for the classification of ritual activity as non-normative or magical.Footnote 28 Both normative ‘religious’ curses and non-normative ‘magical’ curses make use of a common set of rhetorical devices, both verbal and enacted.

The most common rhetorical enhancements are simple verbal devices like repetition, pleonasm, parallelism, and asyndeton, which intensify the expressiveness of curses both agonistic and contingent. As Gager notes, these devices are part of the basic repertoire of rhetorical expression, and are not particular to curses.Footnote 29 Other devices, however, appear specifically in agonistic curses and almost never in the contingent curses where the primary audience is on the horizontal axis. Some curses deliberately scramble the ordering of the lines or even the letters of the written words on the tablet. In the curse against Manês, for example, the lines are inscribed in the reverse order, whereas two fourth-century bce Attic curses spell out all the names of the curser's enemies backwards, inscribing only the last two lines from left to right.Footnote 30 This manipulation of the written verbal expression marks the unusual nature of that expression, further differentiating it from normal speech or writing.Footnote 31

The rhetorical devices that are solely verbal have provoked less consideration among scholars than the ritualized manipulation of material objects. Frankfurter has indeed suggested that a focus on the ‘object-agency’, ‘the capacity of things (especially ritually prepared things) to influence actions and sentiments’, is an essential characteristic of magic.Footnote 32 Following Tambiah, however, I would argue that the manipulation of material objects is simply a more exaggerated rhetorical effect, one that is intended to express more emphasis and garner more attention from the audience. The simile expressed in the wish ‘may their brains be poured out like wine’ has a more powerful effect if at the same time the Greeks and Trojans are pouring wine out of their cups onto the ground. In this contingent curse, the action of pouring the wine is not sympathetic magic that creates some mystical connection between the wine and the brains but a rhetorical gesture that enhances the effect of the simile.

The famous cursing ritual that accompanies the oath of the Cyreneans provides a good test case, since the melting of wax figurines has often caused this ritual to be classified as magic:

On these conditions a sworn agreement was made by those who stayed there and by those who sailed to found the colony, and they invoked curses against those transgressors who would not abide by it – whether those settling in Libya or those who remained. They made waxen images and burnt them, calling down the following curse, everyone assembled together, men, women, boys, girls: ‘The person who does not abide by this sworn agreement but transgresses it shall melt away and dissolve like these images – himself, his descendants and his property; but those who abide by the sworn agreement – those sailing to Libya and those staying in Thera – shall have an abundance of good things, both themselves and their descendants.’Footnote 33

Here we find the same enhancing of the verbal simile with the ritual action; the spoken wish that the person who violates the oath should melt like wax in the fire is given more rhetorical effect by melting wax figurines in the fire. The primary audience of this curse, however, is clearly on the horizontal axis; the text explicitly indicates that ‘everyone assembled together, men, women, boys, girls’ took part in the ritual. The audience on the vertical axis – presumably Apollo as the deity who was authorizing the colonial venture – is not even mentioned in the text. The requested effects are characteristically extensive and exaggerated: utter destruction not just of the ones violating the oath but even of their later descendants and their property. The wax figurines have a rhetorical effect that is not well explained either by a theory of sympathetic magic (à la Frazer) or by one of object-agency (à la Gell).Footnote 34 The wish remains extremely violent and emphatic, but the violence is not literally being enacted on the target, a distinction that the contingent nature of the curse makes clear.

A similar enhanced rhetorical effect appears in the manipulation of the cut-up pieces of animals used in the strongest oaths in the Athenian legal system. As Faraone has shown, the swearing of the oath and self-imprecation by the successful party in a murder trial is given additional force by the ritualized process of cutting up special animal victims and performing the oath while standing on the bleeding chunks of meat.

He must not treat this oath as an ordinary oath, but as one which no man swears for any other purpose; for he stands over the entrails of a boar, a ram, and a bull, and they must have been slaughtered by the necessary officers and on the days appointed, so that in respect both of the time and of the functionaries every requirement of solemnity has been satisfied.Footnote 35

Such emphasis marks the unusual nature of this situation, worthy of the strongest of oaths; the more trouble it takes to prepare the ceremony, the more significant the performance becomes.

This rhetorical device of augmenting the impact of a verbal simile with a material component appears too in the agonistic curses, even if the most emphatic and spectacular examples come from later periods than classical Athens. Some curses make use of the material of the tablet itself to enhance the power of the simile in the curse. ‘And just as this lead is worthless and cold,’ reads one tablet, ‘so let that man and his property be worthless and cold.’Footnote 36 The curser could simply have wished that Pherenikos become as worthless and cold as lead, but the verbal deictic ‘this lead’, accompanied no doubt in the ritual with a physical gesture, helps to emphasize the comparison.

Similarly vivid is the simile that likens the targets to the corpse with whom the tablet is buried: ‘[And just as] this corpse lies useless, [so] may all the words and deeds of Theodora be useless with regard to Charias and to the other people.’Footnote 37 The curser wishes that Theodora be as sexually unattractive as a corpse, all her words and deeds of seduction useless, so that her relationship with Charias may fail, but the specific connection with the particular corpse makes the simile more vivid, enhancing its rhetorical power.

The most dramatic examples of such a rhetorical move come later, as, for example, the tablets from second-century ce Aquitania that refer to the mutilation of a puppy:

I denounce the persons written below, Lentinus and Tasgillus, in order that they may depart from here for Pluto and Persephone. Just as this puppy harmed no one, so [may they harm no one] and may they not be able to win this suit; just as the mother of this puppy cannot defend it, so may their lawyers be unable to defend them, [and] so [may] those [legal] opponents | be turned back from this suit; just as this puppy is [turned] on its back and is unable to rise, so neither [may] they; they are pierced through, just as this is; just as in this tomb animals/souls have been transformed/silenced and cannot rise up.Footnote 38

However, even if the rhetoric is more exaggerated, the basic logic of the operation remains the same. The manipulation of the object (in this case, the helpless puppy) serves to emphasize the verbal similes – as harmless as a puppy, as undefended as a puppy, as pierced through and incapable of rising as a puppy.

Other kinds of ritual operations can augment the rhetorical force of the imprecations. A fourth-century bce curse tablet from near the Piraeus refers to binding the targets in blood and ashes:

Hekate Chthonia, Artemis Chthonia, Hermes Chthonios: | cast your hate upon Phanagora and Demetrios,| and their shop and their property and their possessions.| I bind my enemy Demetrios, and Phana|gora, in blood and in ashes, with all the dead. Nor will the next four-year cycle release you.| I bind you in such a bind,| Demetrios, as strong as is possible,| and into [your] tongue a kynoton I nail in.Footnote 39

It is not clear what kind of ritual would create the blood and ashes mentioned, but any sort of holocaust sacrifice where an animal is slaughtered, drained of blood, and then burned would provide the necessary materials.Footnote 40

This tablet, which was discovered pierced with a nail, also refers to the process of nailing the tongue, another form in which the verbal expression is emphasized by the enacted one. Many agonistic tablets have nails driven through them, a gesture which serves to enact the verb of binding and mark the violent force of the performer's wish. In some tablets, the manipulation of the tablet itself is marked in the words, as in a long Attic curse that targets a number of opponents in a legal case: ‘All of these I bind, I hide, I bury, I nail down.’Footnote 41 The performer makes an analogy between their actions with the tablet – hiding, burying, and nailing – and their wishes for the targets, and this analogy serves to give emphasis to their desire, not so much a persuasive analogy as an expressive one.

In all these cases, the manipulation of material objects functions as an extension or enhancement of the verbal rhetoric, rather than some sort of separate operation that works by different rules of causality. Such material manipulations are not limited to magical rituals; indeed, they appear more dramatically in normative rites like oaths at earlier periods. Rather, I would suggest, these rhetorical emphases are addressed primarily to the audience on the horizontal axis, the mortals rather than the gods. Even in the agonistic curses where the horizontal axis of communication is limited, the performer is communicating this emphasis to himself or herself, expressing the force of his or her own feeling through the rhetoric of the curse.

Conclusions

The rhetoric of violence in curses thus depends primarily on the audience addressed. Contingent curses pile up the violent effects for descriptive and expressive purposes in order to provide a deterrent for the people witnessing the curse: that is, the other mortals on the horizontal axis of communication. Agonistic curses, by contrast, keep the effects requested to a minimum in order to improve the chances of getting the favour granted by the divine powers who make up the primary audience – the vertical axis. The level of violence explicitly requested is thus less a reflection of the malevolence of curser than of the rhetorical strategy at work in each kind of situation. The contingent curse may describe hyperbolic violence for the purpose of emphasis, since there is no immediate request to the divine power to act, but the agonistic curse is making an immediate request for action that the deity has little or no incentive to grant, and so the request needs to be as modest and limited as feasible, often just transferring the target into the power of the god.

The focus on whether the ritual is addressing the audience on the horizontal or vertical axis also helps explain the agonistic curses where the rhetoric of violence is more exaggerated. The agonistic curses that employ such devices may not just reflect the expression of the curser's feelings to himself or herself, but they may actually have been performed by a professional, who was thus performing along a horizontal axis that included the client as well. The exaggerated rhetorical force thus becomes a way to impress the client with the power of the curse and justify the hiring of a professional. Just as a litigant in the courts might hire a speech-writer to provide a defence speech with extraordinary rhetorical power, rather relying on the eloquence they themselves could produce, so too an individual who wanted an effective curse might hire a professional whose expertise with the genre could produce an extraordinary curse. The exaggerated rhetoric of violence in the performance of the curse serves as a guarantee of the curse's extraordinary efficacy, along with further rhetorical elaborations of curse figurines or other material manipulations.Footnote 42

Analysis of curses as ritual acts of communication or expression thus illuminates the social contexts of these performances, providing a better understanding not just of classical Athens but of other times and places as well. Curses in literature, for example, always have an eye to the audience on the horizontal axis, both the internal audience within the story and the external audience of the reader, so they tend to be more elaborate. Likewise, the variations in the category of ‘prayers for justice’ can be better understood by examining the strategies of appeal and the choice of audiences. There may be moral justifications to strengthen the case, or rhetorical flourishes to heighten the intensity, and the curse may perform a secret appeal to the divine for vindication or a public appeal to the community for widespread recognition of the wrong. The choice between describing a contingent catastrophe, with extreme violence extending beyond the specific targets, or merely asking for an agonistic advantage, just enough intervention from the divine power to ensure success in the ongoing rivalry or conflict with the target, does not depend upon some sort of aversion to explicit violence in classical Athens, but rather represents varying rhetorical strategies in performances addressed to different kinds of audiences.

References

1 TheDeMa Nr 203 DTA 69 G63. References to epigraphic sources will generally provide the number in the Thesaurus Defixionum Magdeburgensis (TheDeMa) database, as well as the number in the collection of Gager 1992 (G), in addition to the publication in an epigraphic collection: IG = Inscriptiones Graecae; DTA = IG, vol. 3, pt 3, Appendix: ‘Defixionum Tabellae’ (Berlin, 1897); DT = Audollent, A. (ed.), Defixionum Tabellae (Paris, 1904)Google Scholar. Unless otherwise noted, translations are taken from Gager, J., Curse Tablets and Binding Spells from the Ancient World (Oxford, 1992)Google Scholar.

2 W. Riess, Performing Interpersonal Violence. Court, Curse, and Comedy in Fourth-Century bce Athens (Berlin, 2012), 183. See also 168: ‘Underneath the tame linguistic surface, quite a few curses may have been designed to kill the victim; moreover, further deliberations suggest that the potential of violence contained in binding magic was generally higher than scholarship has so far surmised.’

3 C. A. Faraone, ‘The Agonistic Context of Early Greek Binding Spells’, in C. A. Faraone and D. Obbink (eds.), Magika Hiera. Ancient Greek Magic and Religion (Oxford, 1991), 8.

4 Faraone (n. 3). E. Eidinow, ‘Why the Athenians Began to Curse’, in R. Osborne (ed.), Debating the Athenian Cultural Revolution. Art, Literature, Philosophy, and Politics, 430–380 bc (Cambridge, 2007), 57–60, questions whether simple competition provides a sufficient explanation. See also E. Eidinow, ‘Magic and Social Tension’, in D. Frankfurter (ed.), Guide to the Study of Ancient Magic (Leiden 2019), 754: ‘If spell-types provide an indication of the emotional dynamics behind the spell, then we are dealing with a situation of more complexity than can be described by evoking any single dynamic, such as competition.’ I would argue that the competitive dynamic covers a broad spectrum of conflicts, with a variety of emotional dynamics.

5 F. Graf, Magic in the Ancient World (Cambridge, MA, 1997), 214.

6 Hom. Il. 3.292–301, translation from A. T. Murray (trans.), Homer. Iliad, Books 1–12, rev. W. F. Wyatt (Cambridge, MA, 1924). See C. A. Faraone, ‘Molten Wax, Spilt Wine and Mutilated Animals: Sympathetic Magic in Near Eastern and Early Greek Oath Ceremonies’, JHS 113 (1993), 73–4, who contextualizes this scene with others, both Greek and Near Eastern.

7 Aeschin. 3.110–11, translation from C. D. Adams, Aeschines. Speeches (Cambridge, MA, 2014).

8 Indeed, Aeschines’ point in quoting the curse is to convince his contemporary Athenian audience that they need to act, lest they trigger the contingency of the curse. He is claiming that Demosthenes has already incurred the wrath of the gods for his actions, but he is not just waiting for the gods to smite Demosthenes; he argues that the Athenians need to condemn Demosthenes and all his allies to prevent the contingency of the curse from touching them. Cf. Faraone's argument about this text in comparison with Near Eastern loyalty oaths, where the message of the oath and curse is unmistakably that the weaker party must yield to the imperial overlord – or else! C. A. Faraone, ‘Curses and Blessings in Ancient Greek Oaths’, Journal of Ancient Near Eastern Religions 5 (2005), 142: ‘In the context of this typically unbalanced political relationship, it perhaps makes sense to see this kind of oath-curse as a vehicle for broadcasting the threats that are implied in them.’

9 Ar. Thesm. 295–371. Thucydides (8.97) also mentions that the government of the 5,000 pronounced a curse on anyone who took pay for serving in office.

10 Dirae Teorum in R. Meiggs and D. Lewis (eds.), A Selection of Greek Historical Inscriptions to the End of the Fifth Century bc (Oxford, 1969), no. 30 = SIG III.37–8. The performance of the curse thus serves to divide the audience into the legitimate and illegitimate members of the community: those who abide by the rules and those who transgress them.

11 Dem. 36.52: ‘For this Pasio was grateful to him, but you make no account of it. Nay, in defiance of the will and the imprecations written in it by your father, you harass him, you prosecute him, you calumniate him.’ The speaker in one of Antiphon's tetralogies threatens the jury with the wrath of the demons of vengeance if he is convicted unjustly; there is no formal curse, but the pattern is similar. Antiphon 4.2.8: ‘But if I am wrongfully condemned by you, then it is upon you, and not upon my accuser, that I shall turn the wrath of the avenging demons.’ For tomb curses, see J. H. M. Strubbe, ‘Cursed Be He That Moves My Bones’, in Faraone and Obbink (n. 3), 33–59. See also, from seventh-century bce Rhodes, IG 12.1.737: ‘This tomb I, Idomeneus, made, so that there may be glory (for me). May Zeus impose utter destruction on whoever damages it’ (σᾶμα τόζ’Ἰδομενεὺς ποίησα, ἵνα κλέος εἴη⋅| Ζεὺ (δ) δέ νιν ὅστις πημαίνοι, λειόλη θείη).

12 This divine response, however, is still not ‘automatic’, but more like a penalty clause in a legal agreement. If the terms are breached, the wronged person still has to sue, even if the case is very strong because of the prior agreement. Likewise, the efficacy of performative utterances is not ‘automatic’, even if the desired result is expressed not as a conditional request (illocutionary utterance) but as a perlocutionary statement, ‘I bind’. The human agent performing the utterance still knows that divine assistance is needed to make it happen.

13 TheDeMa Nr. 113 Gager 102.

14 TheDeMa Nr. 201 Gager 61 DTA 109; TheDeMa Nr. 119 DTA 95 Gager 39.

15 Comparisons with the later recipes in the PGM (e.g. PGM IV 296–466 or VII 429–58) suggest such a verbal invocation accompanying the inscription and/or deposition of the written tablet. As Faraone has argued (e.g. C. A. Faraone, ‘Text, Image and Medium: The Evolution of Graeco-Roman Magical Gemstones’, in C. Entwistle and N. Adams (eds.), Gems of Heaven: Recent Research on Engraved Gemstones in Late Antiquity c. ad 200–600 [London, 2011], 57), the tradition of verbal incantations long preceded the written forms and continued to be used alongside writing.

16 R. Edmonds, Drawing Down the Moon: Magic in the Ancient Greco-Roman World (Princeton, NJ, 2019), 169–74.

17 Hom. Il. 1.34–42, translation from Murray rev. Wyatt (n. 6).

18 TheDeMa Nr. 201 Gager 61 DTA 109 (the lines of the text appear in reverse order on the tablet itself). Cf. Gager 48 = SEG 37.673.

19 TheDeMa 118 DTA 103 Gager 38.

20 TheDeMa Nr. 223 DTA 98 Gager 83.

21 Versnel's category of ‘prayers for justice’ has been much disputed since its formulation in H. S. Versnel, ‘Beyond Cursing: The Appeal to Justice in Judicial Prayers’, in Faraone and Obbink, (n. 3), 60–106, but it does usefully distinguish those curses that rely on an argument of justifying the request from the larger group that include no such justification. These prayers for justice are not mere outlying exceptions, but examples that show how a better-grounded argument in the prayer for a god to harm one's enemy can embolden the one praying to ask for a more substantial and visible response from the god. Nevertheless, we can still distinguish those for whom the communication to the god on the vertical axis is more significant than the appeal to the community on the horizontal axis by such features as where they are deposited or displayed; and the correlation between appeal to the wider audience and rhetorical intensity of the language of the curse still seems to hold. The supplicatory tone in some of these texts, however, cannot be taken as an indicator of ‘religious’ rather than ‘magical’ action. Rather, the extent and tone of the appeal to the community audience provides a better indicator of the normative status of the ritual.

22 TheDeMa Nr. 207 DTA 86 Gager 67.

23 TheDeMa Nr. 104 DT 68 Gager 22: ‘I bind Theodora in the presence of the one [female] at Persephone's side and in the presence of those who are unmarried. May she be unmarried and whenever she is about to chat with Kallias and with Charias – whenever she is about to discuss deeds and words and business … words, whatever he indeed says. I bind Theodora to remain unmarried to Charias and [I bind] Charias to forget Theodora, and [I bind] Charias to forget … Theodora and sex with Theodora. [And just as] this corpse lies useless, [so] may all the words and deeds of Theodora be useless with regard to Charias and to the other people. I bind Theodora before Hermes of the underworld and before the unmarried and before Tethys. [I bind] everything, both [her] words and deeds toward Charias and toward other people, and [her] sex with Charias. And may Charias forget sex. May Charias forget the girl, Theodora, the very one whom he loves.’

24 Riess (n. 2), 208: ‘Terms used on the tablets, such as katagraphô (“I write down”), apographô (“I write away”), and paradidômi (“I hand over”), normally appear in judicial contexts to validate business transactions.’ Riess goes too far, however, when he claims (210) that this assignment of the target to the underworld divinities indicates that the target is being dragged in a downward direction into the underworld – and thus to death. The further metaphoric movement is brought out in some curses, but by no means many, and it should not necessarily be read into those in which the metaphor is not developed.

25 Ibid., 214 n. 246, makes the fascinating point that none of the early Sicilian curses use the language of binding. The language of written assignment is used instead, a metaphor that appears more rarely in the Attic material, but which occurs, for example, in the curse of Dionysophon from Thessaly (DTA 78). See also the essay by Lamont in this collection.

26 Cf. TheDeMa Nr. 204 DTA 55 Gager 64, which proclaims, ‘All of these I consign in lead and in wax| and in water and to unemployment and to destruction and to bad reputation| and to defeat and in tombs, |both them and all| the children and wives who belong to them’ (lines 18–21).

27 S. J. Tambiah, ‘The Magical Power of Words’, Man, 3.2 (1968), 202, notes this complementarity in his treatment of ‘persuasive analogies’: ‘All ritual, whatever the idiom, is addressed to the human participants and uses a technique which attempts to re-structure and integrate the minds and emotions of the actors. The technique combines verbal and non-verbal behaviour and exploits their special properties.’ While Tambiah here stresses what I call the horizontal axis, the same rhetorical effects may be deployed along the vertical axis of communication in seeking divine assistance in fulfilling a wish.

28 See Edmonds (n. 16), esp. 5–19. Contrast, for example, the classification of all analogical actions in curses as magical in L. Watson, Arae. The Curse Poetry of Antiquity (Leeds, 1991), 50–1: ‘These and comparable gestures rest ultimately on the universal principle of “sympathy” which underlies so much of ancient magical practice.’

29 Gager (n. 1), 13–14: ‘we may add a partial list of recurrent features in the language of discourse of the defixiones: repetition, pleonasm, metaphor and simile, personification, rhythmic phrases, exaggeration, threats, promises, prayers, and formal appeal. Once again, it must be emphasized that these features are not distinctive of defixiones, but instead mark them as part of the general culture of their time.’ Cf. Riess (n. 2), 186. As H. S. Versnel, ‘The Poetics of the Magical Charm: An Essay in the Power of Words’, in M. Meyer and P. Mirecki (eds.), Magic and Ritual in the Ancient World (Leiden, 2002), 131, notes, ‘No doubt, repetition or extension of a formula by variation (albula glandula) – in more general terms: pleonasm or redundancy – confers emphasis, that is power, to the expression.’

30 TheDeMa Nr. 138 DT 60 Gager 42; TheDeMa Nr. 413 DT 61. Gager ad loc n. 29 suggests that both were executed by the same hand.

31 The employment of voces magicae takes this use of extraordinary language to another level, raising what Malinowski would refer to as the ‘coefficient of weirdness’ and marking the ritual as magical because of its extraordinary performance (see Edmonds [n. 16], 18). However, voces magicae rarely appear in classical-era curse tablets; they are a rhetorical innovation that came into usage in the late Hellenistic and Roman periods.

32 D. Frankfurter, ‘Magic and the Forces of Materiality’, in D. Frankfurter (ed.), A Guide to the Study of Ancient Magic (Leiden, 2019), 661–2.

33 SEG 43.1185 (Meiggs and Lewis [n. 10], no. 5).

34 In addition to Frankfurter (n. 32), D. Collins, Magic in the Ancient Greek World (Malden, MA, 2008), argues for Gell's idea of understanding the material objects themselves as having a kind of agency. Riess (n. 2), 199–200, prefers a moderated version of Collins’ idea, because he wants to emphasize the violence inherent in the curses. By contrast, Tambiah (n. 27) notes the complementary effect of material and verbal elements in the symbolic rhetoric.

35 Dem. 23.68, translation from J. H. Vince (trans.), Demosthenes. Orations, Volume III (Cambridge, MA, 1935). Cf. Aeschin. 2.87. See also Faraone (n. 6); C. A. Faraone, ‘At the Limits of Efficacious Speech: The Performance and Audience of Self-Curses in Ancient Near Eastern and Greek Oaths’, in C. Ando and C. A. Faraone (eds.), Serments, vœux et construction rituelle des actes de parole efficaces, Mètis 10 (2012), 130–1.

36 TheDeMa Nr. 120 DTA 107 Gager 40.

37 TheDeMa Nr. 104 DT 68 Gager 22.

38 TheDeMa Nr. 190 DT 111–12 G53.

39 Lamont, J. and Boundouraki, G., ‘Of Curses and Cults: Private and Public Ritual in Classical Xypete’, in Vavouranakis, G., Kopanias, K., and Kanellopoulos, C. (eds.), Popular Religion and Ritual in Prehistoric and Ancient Greece and the Eastern Mediterranean (Oxford, 2018), 127–8Google Scholar, ΜΠ11948. (It remains unclear to what kynoton refers.)

40 Cf. the holocaust sacrifice of a falcon in PGM IV 3125–71. Lamont suggests that the resonances of the epic line ἐν αἵματι καὶ κονίαισιν (Od. 22.383; Il. 16.796) would further enhance the curse's power, perhaps drawing on an oral hexametric tradition (Lamont and Boundouraki [n. 39], 129–30): see Lamont, J., ‘A New Commercial Curse Tablet from Classical Athens’, XPE 196 (2015), 166Google Scholar.

41 TheDeMa Nr. 140 DT 49 Gager 44.

42 Faraone, C. A., ‘Curses and Social Control in the Law Courts of Classical Athens’, in Cohen, D. (ed.), Demokratie, Recht und soziale Kontrolle im klassischen Athen (Munich, 2002), 90Google Scholar, suggests that the more elaborate ensembles of curse tablets and figurines are likely to be the work of professionals.