JESSICA LAURA LAMONT A N EW C OMMERCIAL C URSE T ABLET
FROM
C LASSICAL A THENS
aus: Zeitschrift für Papyrologie und Epigraphik 196 (2015) 159–174
© Dr. Rudolf Habelt GmbH, Bonn
159
A N EW C OM M ERCI A L C U RSE TA BLET
F ROM
C LASSICA L AT H ENS 1
A lead curse tablet recently unearthed in a Classical cemetery just outside of the Athenian Long Walls, northeast of the Piraeus and likely within the ancient deme of Xypete, is of great interest. Stored in the Piraeus Museum ( 11948), the tablet can be dated on archaeological and orthographic grounds to the early fourth century BCE. It invokes Hekate, Artemis, and Hermes in their role as “chthonic” deities; it targets a male and female couple, Demetrios and Phanagora, who ran a – understood here as “tavern” – in Classical Athens. Aspects of the text prove striking and unusual, including the presence of a hexametric Homeric intertext, (line 5), and the mention of a limit to the period of binding, (lines 6–7). The curse tablet is significant not only because it emerged in 2003 from a secure and well-documented archaeological context, but also because it was part of a larger cache of five lead tablets, all of which were pierced with iron nails and deposited together as a group within the same grave.2 Of the five tablets, four were incised with text; the fifth was left blank. The four curses are remarkably similar, exhibiting parallel formulae and narrative structures. They target four husband-wife couples, all of whom are associated with .3 Only one tablet, 11948, is presented here, but it is representative of the other three narratives. Because the thread linking the curses seems to have been occupational and the tablets were composed together and deposited as an assemblage, commercial rivalry seems the most likely motive for their creation. 11948 and the tablets with which it was deposited reveal a great deal about commercial rivalries, social networks, and private ritual in a local Attic deme community. The Context (Fig. 1) Along with the four other lead tablets, 11948 was discovered in a Classical grave on Piraeus Street 131–133 (Fig. 1). The discovery came in 2003 during rescue excavations in the region of New Phaleron, undertaken by the Ephoreia for Prehistoric and Classical Antiquities, under the supervision of Maria Petritaki. Especially during the fifth century BCE, but also throughout the fourth, this region benefitted from proximity to the commercial activity of the Piraeus; traffic was steady passing through this area en route from the port to the city center, through the artery created by the Long Walls. Epigraphic and archaeological evidence suggests that the local population was indeed diverse, consisting of Athenian citizens, metics, slaves, and foreigners; nearby temples to Kybele hint that this area may have held particular significance for residents with ties to the East.4 Situated in this region, the Classical cemetery was dated to the late fifth through fourth centuries BCE on the basis of pottery styles, funerary assemblages, and tombs.5 The nekropolis has yielded a variety of grave types, all roughly contemporary and in close spatial proximity. So near to the Piraeus, this heter1 For assistance in funding the research needed to undertake this project, my gratitude extends to several institutions,
including the American School of Classical Studies at Athens, the Alexander Onassis Public Benefit Foundation, the Department of Classics at The Johns Hopkins University, and the Charles Singleton Center for the Study of Pre-Modern Europe. I thank Maria Petritaki for permission to study and publish 11948. The following individuals were also very helpful in bringing together various aspects of this article, and I thank them kindly: Georg Petzl, Alan Shapiro, Robert Garland, Joseph Day, Daniel Dooley, Jonathan Meyer, Georgia Boundaraki, and Simon Oswald. 2 The editio princeps of 11948 is provided here, with information about the other tablets pulled mainly from the excavation reports: Petritaki 2009, pp. 462–5; 2010, pp. 450–1. It is hoped that the other three incised tablets can be published in the near future. 3 Petritaki 2009, p. 465, fn. 49. 4 Papachristodoulou 1973, pp. 189–217; Petritaki 2009, p. 469. The presence of slaves and foreigners is further suggested by the Attic of 415 BCE, which – in auctioning off property in Xypete – record the sale of one Scythian and three Thracian slaves (IG I³ 422.187–200). 5 Petritaki 2009, p. 464; 2010, pp. 450–1.
160
J. L. Lamont
Figure 1. Map of Xypete and location of curse tablets (Grave VIII, Pyre II, Map by G. Boundouraki, with alterations by author
131–133).
ogeneous funerary landscape could reflect the region’s diverse population, with burial styles that varied according to demographic. Amidst marble funerary and ( 8178–9, 9312, 9331), the excavations unearthed a large circular burial monument and various other types of graves: a made of clay, six tombs with tiled roofs (three of which were in the form of a hut, and three of which were cut into the bedrock), a broken poros sarcophagus, and two funerary-pyre burials.6 Associated with the pyre burials were two areas for offerings, as well as architectural blocks; the latter were interpreted as the remains of a destroyed enclosure wall. One of the pyre-burials (Grave VIII, Pyre II) contained the cremated remains of a female.7 At a depth of 1.98 meters in a stratum of brownish sand came a deposit of five lead sheets, all of which were folded and pierced with iron nails. The cache reportedly came from the wall delimiting the grave, which suggests that the deposition of the assemblage occurred close to the time of the burial itself. It was not, in other words, found in isolated loose fill or in a later (higher) disturbed phase of use. The tablets could have been deposited by a mourner involved in mortuary ritual at the burial, perhaps a family member, a dependent of the household, or a relation otherwise aware of the cremation. Women, for example, were expected to maintain tombs associated with the , and were responsible for the performance of graveside rites for dead relations. They would have had access to the grave, and ample opportunities for visiting the cemetery. Perhaps we can envision – in matters of binding that dealt with the fiscal success of an entire – that husbands, wives, and potentially even dependents (slaves, children) worked together in such supernatural undertakings, with women occasionally charged with depositing the tablets. Curses like these, which target couples and businesses, could be seen as a household or workplace affair. Alternatively, should we prefer a clandestine midnight scenario in which a , , , or other religious ‘professional’ slipped quietly into the cemetery to bury his sinister assemblage, a pre-prepared pyre burial would have been a practical and easily accessible location for deposition.8 The tomb, with its 6 Petritaki 2010, pp. 449–51. 7 The grave in question lay immediately outside of the northwestern Long Wall, roughly 2.5 km. northeast of the Piraeus
and close to the river Kephisos (c. 0.5 km.). 8 This scenario seems less likely; it would assume that the deposition was made after the digging of the tomb, but prior to the actual funerary rites and the filling in of the pyre grave; the grave would have been filled in soon after the pyre was extin-
A New Commercial Curse Tablet from Classical Athens
161
associated area for libations and other offerings, suggests that graveside ritual was a familiar activity, which would have required a degree of accessibility for the living. In other words, of the many and different types of tombs found in this cemetery, the pyre burial with its area for depositing offerings made this grave an accessible and practical locus for supernatural exploitation. The Tablet (Fig. 2) The bluish-grey tablet is roughly rectangular (12.7 cm. in length, 5.6 cm. in width), with the entire right side and almost all of the upper edge cut to a straight line. There is considerable abrasion along the tablet’s left side, and in some parts the lead has broken off entirely; this obscures the beginning of several lines of text. There is also damage along the bottom edge, though only left of the center. A thin vertical tear (0.6 cm. in length) splits the upper edge, just right of the center, where the lead has been folded; though it splits the last word of the first line ( ), it does not obscure it. There is a hole (0.4 cm. in diameter) from where the iron nail was driven through the tablet, which effaced only one letter, nu, at the end of in line 5. There is oxidation from the iron nail immediately below the hole, and also on the lower right side of the tablet. This latter region of oxidation is the larger of the two (1.28 cm. × 1.66 cm.), and has collected encrustation that obscures half of the ultimate letter of the narrative, the omega in . On the right edge of the tablet, perforated ‘dash’ marks are visible, along which the lead strip was cut; they suggest that the lead band had to be physically cut away from a larger sheet or panel, before the curse could be inscribed upon it. A blank margin was deliberately kept on the right-hand side of the tablet, which ranges from c. 1.0 cm. to 4.6 cm. (line 6 to line 5, respectively). The scribe, in other words, did not crowd each line with as many words as could be fit, but rather made deliberate line breaks in the face of available space. It is possible that the left-hand margin tapers slightly toward the right after line 6, or possibly that the lead itself was cut on an angle, sloping toward the right as it appears today. In general, the tablet is very well preserved, and the few regions with surface damage and textual lacunae can easily be restored with text from the other three tablets, so similar and repetitive are the curses. The letters are carefully spaced and neatly inscribed, which suggests that the inscriber was seasoned in the craft of writing. Sophisticated vocabulary is employed throughout (e.g., lines 5, 6–7), and several unusual terms also prove of interest (line 9, ). The average letter height is c. 0.3 cm. The nail piercing the tablet was 3.0 cm. in length, its head 1.5 cm. in diameter. Diplomatic Transcription (Fig. 3): ..A
1 2 3 4 5 6 7 8
. .
9
...I
G
Y
N
V
guished, and for the cache to have emerged from the wall of the grave, it need have been deposited prior. It is possible that the hurried nature of the rescue excavations led to a less than careful scrutiny of the grave’s stratigraphy with regard to the cache of curses. If a looser pocket of soil enveloped the tablets, suggesting a disturbed context and a later (post burial) deposition of the assemblage, the soil would still have been relatively compact after the passage of time; this stratigraphic nuance may have been missed in the rush to excavate ahead of a looming construction project.
162
J. L. Lamont
Figure 2.
11948, from Grave VIII, Pyre II, 131–133, early fourth century BCE. Piraeus Archaeological Museum Inv. 11948. Photograph by author, with permission from
The text and my translation run as follows:
1 2 3 4 5 6 7 8 9
[
]
,
,
. ’ [] . [ ] , [ ] ,
.O ’ + +
l. 6: text reads l. 9: [ ] restored from other tablets;
1 2 3 4
,
()
,
. written faintly as superscript above
Hekate Chthonia, Artemis Chthonia, Hermes Chthonios: cast your hate upon Phanagora and Demetrios, and their tavern and their property and their possessions. I will bind my enemy Demetrios, and Phana-
in
A New Commercial Curse Tablet from Classical Athens
Figure 3. Drawing of
5 6 7 8 9
163
11948, by author
gora, in blood and in ashes, with all the dead. Nor will the next four-year cycle release you. I will bind you in such a bind, Demetrios, as strong as is possible, and I will smite down a kynotos n [your] tongue. Commentary
: beginning in the first line and continuing throughout Line 1: [ ] the text, the scribe struggles when incising round letters with the stylus. Curved letters thus appear angular; for example, both thetas and omicrons are teardrop shaped. All thetas have dots rather than bars in the middle. Hekate, Artemis, and Hermes are invoked in the vocative, as an explicitly “chthonic” or sub-earthly triad. Other curse tablets bind their victims (DTA 105; SGD 170); deities with underworld associations held natural appeal for use within binding curses, especially for tablets deposited in underground ‘conduits’ such as graves or wells. Of help in making sense of this chthonic threesome is IG I3 383 (429/8 BCE), which implicitly references a cult of “Hermes and Artemis-Hekate” at an unspecified site in Attica: [ ] |[ ] |[ ] |[ ] |[ ] (IG I3 383.124–9).9 It is possible that this , in which Hermes and the fused Artemis-Hekate shared cult, inspired the invocation of the three chthonic deities in the curse tablets. Otherwise, Hekate and Hermes are frequently paired on Attic curse tablets, yoked together as a chthonic couple (e.g., NGCT 2; DTA 105–7, with inscribed on the outside of DTA 107b, addressing the rolled-up tablet as if a letter). Artemis Chthonia, however, is rare in the extant corpus of Attic curse tablets (see below). [ ] : the upper left corner of the tablet is missing for a length of 0.9 cm. Accordingly the first two letters, epsilon and kappa, are entirely gone, and only part of the diagonal stroke of the alpha’s 9 IG I3 383: an inventory of currency and other dedications from various Attic shrines, compiled in 429/8 BCE by the
Treasurers of the Other Gods. The record lists the property – predominantly in coinage, but not exclusively so – from Attic sanctuaries such as that of Hera in Xypete (l. 65), Poseidon at Sounion (l. 106–11), and Olympian Zeus (l. 77–9). As the deities listed in the inventory seem to have had sanctuaries in which coins were lodged, this account suggests that Hermes and Artemis-Hekate also shared a cultic space somewhere in Attica; this was the (unlocated) temenos that had accumulated nondescript before 429/8 BCE. Because here is singular, it is tempting to understand it instead as something like “of the custom of old” rather than a reference to “[missing numeral] … of old style coinage”. This, however, also seems unlikely, as the inventory is concerned specifically with coins, or otherwise itemized votives (e.g., ).
164
J. L. Lamont
right leg is visible. Hekate often trafficked in the supernatural; she appears frequently in Attic curse tablets, sometimes with the “Chthonia” epithet (cf. DTA 105–7). She was a dangerous, liminal goddess. In Attica it is thought that worshippers sacrificed dogs to Hekate, a ritual gesture intended to satiate the mercurial goddess, keeping her at bay.10 She also received cult in southeastern Attica in the deme of Halai Aixonides, where there was a Hekataion at a crossroad.11 : the tau is very faintly inscribed. In Attica, Artemis was a goddess strongly associated with maturation rites and the protection of women and girls; yet these four curses appeal to another, destructive side of Artemis, one tied to the realm of the sinister and the threatening. Artemis could be similar in form and function to Hekate; indeed, Artemis-Hekate appears as a coalesced divinity by the year 429/8 BCE (supra, IG I3 383.124–9). Even earlier, Aeschylus’ chorus of suppliant maidens sang, | ,| | (Aesch. Suppl. 674–7). Artemis was also summoned just before Hekate in the chorus of Aristophanes’ Frogs (1356–64), suggesting that the two were natural associates to a fifth century Athenian audience; much later in Theocritus, Artemis appears with Hades in the underworld context (2.33–4).12 Finally, Artemis and Hekate could be closely aligned in fifth–fourth century vase iconography.13 : the theta is broken by a tear 0.6 cm. in length. Hermes was the deity most commonly invoked on curse tablets, under a wide range of epithets, including “Chthonios”: cf. DTA 83, 91, 101, 104–7.14 He received cult in Athens on the Chytroi day of the Anthesteria festival.15 On this day Athenians prepared a sort of gruel for Hermes Chthonios in his guise as an impure, sub-earthly divinity; no priests could eat of this mixture, and temples were closed to prevent the intrusion of . Line 2: : the epsilon is faint as the tablet’s left edge is abraded. The sigma was initially incised on too small a scale; the scribe re-incised it, but in such a way that the initial mistaken strokes intercept the upper bar of the second, larger sigma. The verb, , is a strong one, less common in curse tablets but quite at home in literature with the sense of “bearing a grudge against” (Hom. Od. 11.149; Aesch. Ag. 133; Hdt. 9.79); as an aorist imperative with an prefix, it takes no fewer than five datives. As the command is aimed at the three chthonic deities, we might expect a plural imperative, . The scribe opts instead for the singular form, which agrees with the proper noun immediately preceding it, ( ). : it is possible that Phanagora was used as an identifier to further distinguish Demetrios (that Demetrios, the one whose wife is Phanagora), much as a demotic or patronymic would; or perhaps she was seen as one of several aspects of Demetrios’ life that the agent sought to destroy, alongside Demetrios’ shop, income, and possessions.16 Yet Phanagora’s name appears before that of her husband Demetrios, and is repeated again in lines 4–5; furthermore, the presence of three additional women on the other tablets in the cache – one named (Krite) and two each identified as the “wife” of a named male – suggests that Phanagora, too, was the wife of Demetrios, and that this cache was deliberately targeting male-female couples who ran . Female tavern keepers are targeted in other Attic curse tablets; for example, DTA 30.10 10 See Parker 2011, p. 159, fn. 133, with Aristophanes fr. 209 (Sch. on Theocritus 2.11–12); Plut. Quaest. Rom. 68.280b–c, 111.290d; Athens NM 1695, depicting offerings for Hekate. 11 Lauter 1991, pp. 27–70; Parker 2005, p. 68. 12 By 440 BCE, a sinister and vindictive sort of Artemis was also known from Sophokles’ Ajax, Artemis , though referenced explicitly as an “Olympian” goddess; this same Artemis received cult at Halai Araphenides (Soph. Aj. 172–81, Doric: ; McInerney 2015). A later curse from Oxyrhynchos (SM I 49) also contains a verse inscription to “Artemis-Hekate”, suggesting that in other regions, too, this sinister aspect of Artemis was collapsed with that of Hekate. 13 See Simon 1985, pp. 271–84. 14 Other common epithets of Hermes in binding curses were “Katochos”, “Eriounios”, “Dolios” (e.g., NGCT 24; SGD 18). 15 Or on the Choes day, according to the manuscript readings: see Parker 2005, p. 296, with Ar. Plout. 594 with schol.; Dem. Or. 54.39; Plut. Mor. 708f; Apollod. FGrHist 244F 110. 16 This seems to be the case in SGD 3 and 4, two curse tablets aimed at a blower in a silver-working shop. He is identified by profession, and the text then curses whatever work he produces along with other aspects of his life, such as his wife and possessions.
A New Commercial Curse Tablet from Classical Athens
165
targets ( ), while SGD 11.8 mentions Myrtale, described as both a and a . Other Attic curses target couples in a way that seems related to their occupation; DTA 69, for example, binds the helmet-maker Dionysios and his wife Artemis (a goldworker), along with their work and products. It is likely that, in small-scale Attic industries, possibly run from the home, members of households worked alongside one another and were accordingly targeted as units. : as the main target of this curse, Demetrios’ name is introduced here and then repeated two additional times in the narrative (lines 4, 8). He is one of four males targeted in this cache of curses, along with Euthydemos, Euphilitos (= ), and Kephisodotos from the other three incised tablets.17 That all are associated with suggests that all were members of the same profession: tavern keepers. Demetrios was as common a name in ancient Greece as it is today, and appears in other Attic curse tablets (e.g., DTA 30, 51; SGD 44). We cannot responsibly say anything about this Demetrios, the , other than that he was one of eight targets in an assemblage of curses that sought to bind couples associated with . Line 3: : polysyndeton with the repetition of . The abundance of velars (with kappas and chi) gives this line a harsh, incantatory dissonance. : the kappa is faint as the tablet has corroded along the left edge. : the kappa has an extra stroke between its two upper arms, appearing almost like a psi. The lambda and epsilon are crowded together, nearly touching. The , and his female equivalent, the , are professions commonly mentioned in Attic curse tablets; this identifying term gives name to the place of work, the , which was the generic term for a low-level shop or business. Usually during this period, indicates a “tavern”.18 References to abound in Attic curse corpora: (DT/DTA 70), (DTA 75), (SGD 43), (DTA 87, which features a group of ).19 As such a familiar and widespread aspect of daily life, it is no surprise that they figure prominently in Attic curse tablets. The targeting of four , with no mention of other professions or workplaces, suggests that commercial rivalry was the primary motivating factor for commissioning this cache of curse tablets. : As nouns referring to property “in money and in possessions”, are paired at Pl. Leg. 728e, Isoc. 1.28, and Plut. Alex. 59.2. The agent targets more than just the shop of Demetrios and Phanagora, seeking to destroy all aspects of their lives, with the overall aim to ruin the four couples as thoroughly as possible. : the iota in is very faintly incised and drawn close to the right leg of the preceding alpha. Similarly, in , the eta and left bar of the following mu are very faintly incised. Just as the couple’s money is targeted, so too are their possessions, in a thorough showering of ill-will on the couple. appears independently on other Attic curse tablets, such as in the binding of Sosikleia and her possessions ([ ] , DTA 108a.1). Line 4: : the delta is faint as the tablet has corroded along the left edge; the lead here is blotchy with shallow cracks. With an abrupt and forceful change in person and mood, the new verb assumes the first person future indicative form of .
17 Petritaki 2009, p. 465, fn. 49. 18 In Classical Athens, taverns were a common feature within the urban landscape: Ar. Ec. 154, Lys. 427; Lys. 1.24; Isoc.
7.49. 19 Compare the
with the more specialized Attic curse tablets: DTA 74, DTA 84, DTA 75, commissioners of many curse tablets in Attica.
, or “workshop”, which also occurs frequently in DTA 68, DT/DTA 71 and SGD 124, [] [ ] DT/DTA 71, SGD 52. It seems that workplace affairs preoccupied the
166
J. L. Lamont
’: , expected assimilation with representing / / in standard orthography.20 The epsilon of is elided with that of . : this invective phrase often turns up in Attic curse tablets, as well as in Classical tragedy and oratory.21 With respect to Attic binding curses, cf. DTA 35.16–17, | , and DTA 108a.2–3, with the icy bid that the target | . | : the curse repeats the names of its targets; for Demetrios and Phanagora, see the remarks on line 2. Line 5: : only the horizontal top bar of the gamma is visible as the tablet is broken on the left edge; the lead here is blotchy with shallow cracks. : this hexametrical formula is one of the most striking aspects of the narrative. The phrase occurs in the grimmest passages of Homeric epic, such as the suitor scene in Book 22 of the Odyssey, where it also appears as the second colon of the hexameter: | , ’ , ’ | | (Hom. Od. 22.383–6, “but he saw them all, the men, in blood and dust | lying fallen, like fish, whom the fishermen | have taken in their net of many holes, and dragged out | of the grey sea onto the shore”). Additionally, are paired in the Iliad as two agents that defile and stain the armor of Achilles, just before the death of Patroklos: (Hom. Il. 16.796). It seems possible that the scribe was deliberately incorporating a Homeric verse within the curse, unless this hexametric phrase preserves a trace of an earlier oral binding tradition, in which curses were recited or performed in meter.22 By comparison, another Attic curse tablet was composed in dactylic hexameter, perhaps with an elegiac couplet (DTA 108); an oral recitation or performance surely underpinned the incised metrical text.23 A third interpretation is that related to the ritual – graveside or otherwise – that accompanied the cursing process; the act of binding Demetrios and Phanagora “in blood and in ashes” could point to the presence or use of both organic materials within the cursing ritual. It has been argued that some Attic curse tablets employ verbs that describe rituals performed with the lead, such as “nailing down” and “burying” (e.g., DT 49, [] [] ), and thus that sometimes the incised text referenced proceedings 24 in the larger ritual process. : the scribe includes the ephelcystic nu, which has been almost entirely obliterated by the nail; only part of the vertical bar of the nu on the righthand side is visible, and faintly. It has been preserved in the other three tablets, so its presence here is certain. The inclusion of the nu seems unnecessary; as it is followed by a consonant rather than a vowel, hiatus need not be avoided.25 Since the ultima here is an anceps, furthermore, the nu is not required for metrical purposes to create a long syllable. Often the movable nu is used at the end of clauses or verses, and it is possible that the scribe intended to end this clause with , rather than have the line spill over into line 6 and finish with . In other words, the scribe had ample space after to incise additional words (4.6 cm.), but chose instead to terminate the line at ; this was likely because the phrase ends a line of hexameter.
20 Assimilated spellings were common during the fourth century in public decrees, with a considerable decrease after 350 BCE: Threatte 1980 vol. 1, pp. 625–7 (48.0532). 21 Tragedy and oratory: [ ] (P. Mich. 6973; Eur. Kres. F 448a, 70); , , ; (Dem. 23.79). 22 Cf. Faraone 1985, 1989, for indirect references to binding spells in fifth century Athenian theater. In later centuries, short bits of text from Homeric epic were excerpted and used in incantations, spells, and amulets, though usually toward protective or curative ends (Collins 2008, pp. 211–36). 23 Cf. SGD 150 and NGCT 52, which also have metricizing, if not metrical, elements to their narrative. 24 DT 49; SGD 48; see Eidinow 2007, p. 145. 25 Smyth 1984, p. 33 (§134).
A New Commercial Curse Tablet from Classical Athens
167
Line 6: : , with the phoneme /n/ assimilating to /m/ before all labials. The left half of the sigma’s bottom bar is obscured by the tablet’s broken left edge. The vertical right-hand stroke of the mu extends upward into the space between lines of text; it almost touches the alpha immediately above in line 5. A vertical crack runs just right of the omicron in . The following iota and sigma appear faint and more shallowly incised, and the sigma is missing its bottom bar. : this phrase ultimately makes better sense if taken with the preceding clause, as a prepositional phrase that complements , thus binding Phanagora and Demetrios, and all that they have, “in blood and in dust, together with all the dead”.26 The line was likely broken after (with 4.6 cm. vacant margin) simply because it is the end of a hexameter; follows as the beginning of another dactylic element (a contracted hemiepes: – – – –), with only a poetic break in line, rather than sense. This works well with o then beginning the next clause. O ( ): . Here the scribe becomes a bit careless with incision; these letters appear faint, sloppier than the preceding ones, and angle slightly upward toward the right. Though legible, the lambda is covered by orange encrustation, iron oxide, from the nail. The vertical stroke of the upsilon touches the sigma beside it. Undoubtedly, this is the most difficult and evasive clause of the text; I transcribe as O ( ) for the following reasons: First, as the negative particle o : that the scribe readily employed O for OY within the text is clear from line 7, TOIOT for ( ) . This reading of o is strengthened by other instances – private, public, and contemporary – of O for OY in negative particles such as ( ), , , etc.: early 2 2 fourth century Attic grave : IG II 11780.12 ( = ); IG II 10780.1 ( ’ for ); a phratry decree of 396/5 BCE, IG II2 1237.36 ( for ); IG II2 1237.37 ( for ).27 I opt for the negative particle o , rather than a relative genitive (with the antecedent being the curse itself/enmity from the previous clause), because the spell’s harsh invective would likely strive to ensure that the binding curse not be released by the next four-year cycle. This, of course, begs the question of why the scribe would even mention a limiting period ( |[ ] ) within the binding curse, only to dismiss it. I suspect that, in fourth century Athens, pentaeteric religious festivals such as the Olympic or Panathenaic Games were such prominent public events that they helped structure conceptions of time and limits of public office, with future chronologies tangibly conceived in pentaeteric increments (cf. Arist. Ath. Pol. 54.7; IG II³ 348.13). It is also possible that the author was trying to avoid the next four-year cycle somehow ‘loosening’ the binding powers of the curse; the sacred truce enforced by the Olympic Games, for example, prohibited aggressions such as legal disputes and (the enactment of) death penalties, so perhaps there was a concern that the next could undo the spell. Thus nothing – not even the next fouryear cycle – was meant to release this curse. Another tablet, SGD 170 from Pantikapaion, addresses the topic of the ‘loosening’ of the curse and similarly urges against it; in binding its targets, SGD 170 lists an astonishing array of chthonic deities that, as a group, must not be allowed to loosen the curse under even the most tempting of circumstances: [] | | | | | () (SGD 170, col. B lines 4–9).28 Thus our four incised curses, like SGD 170, would be urging with a negative (o ; [ ]; ) against the release of the binding. Next, rather than an entirely feasible gnomic aorist (o , “nor does the first four-year cycle release you”), I have chosen to transcribe as ( ), a future form assuming here that E was used to represent the diphthong EI. In public inscriptions between 403/2–376/5 BCE, both E and EI were common-
26 In DT 50.4,
similarly ends the clause (though also the line). for [ou] and secondary [o:] only by c. 350 BCE in public inscriptions: Threatte 1980 vol. 1, pp. 178, 238–61, esp. 13.00–13.02. In pronunciation both had likely monophthongized to [o:] or [u:] in Attic Greek by the time the tablet was inscribed. 28 The text orders that none of the gods or daimons be able to loosen the curse, not even if Maietas begs this as a favor, or even if they offer (as a sacrifice) a portion of thigh meat. 27 On the late orthographic standardization of
168
J. L. Lamont
Figure 4. Closeup of with permission from
11948. Photograph by author,
ly used for secondary [e:];29 although our scribe had no problem writing out the diphthong in , for example, the curse traffics heavily in future tense verbs and ( ) would mind this precedent: (line 4), (line 7), (line 9). |[ ] : the tablet’s left edge is broken, and the first tau in |[ ] is entirely lost; in view of the parallel texts from the other tablets, however, |[ ] (with theta rather than tau) might also be restored here.30 This phrase refers to a set period of four years (literally “five” for the Greeks, who counted inclusively); it must be the subject of the clause as these are the only nominative forms. In Classical Greece, the term was commonly employed in the public sphere in relation to athletic and religious festival cycles, such as the Olympic or Greater Panathenaic Games, and relatedly cyclical terms of public office;31 it is unique within the corpus of Attic curses. Line 7: : in most public inscriptions in the pre-403 BCE Old Attic script, omikron was commonly employed as the grapheme for both [o:] and the diphthong [ou] – two similar sounds that by this time had likely monophthongized – in addition to the inherited long-o (later spelled ) and short-o (see supra).32 ’: the angled strokes of the sigma overlap with and touch the preceding iota. The scribe initially had a slip of hand in the third stroke of the sigma, incising downward from the top bar; he corrected it with two linear, deeply cut strokes that intersect and muddle the first attempted sigma. The curse now addresses Demetrios directly, with in the vocative in apposition to [ ] . : while still entirely legible, the gamma is cut by a crack from where the tablet was folded. The omega is angled rather than curved on the top, and the two short horizontal bars are not level with each other. The personal pronoun emphasizes the agency of the curser. : this pleonastic phrase occurs in other Attic curse tablets, e.g. [ ] [ ] (DTA 45.1–2), and … (DTA 108.3–6), with an earlier literary precedent at 29 For use of E as the grapheme for the original diphthong see, for example, IG II2 1237, a phratry decree of 396/5 BCE, featuring (IG II2 1237.69), but EI for [e:] elsewhere in the text. As expected, there was even more variation in the private sphere. See Threatte 1980 vol. 1, pp. 177–8, 299–301. 30 , with > via assimilation with (= ). 31 IG II2 1172.27; Pi. Ol. 10(11).57, Nem. 11.27; Hdt. 3.97.4, 4.94; Thuc. 3.104.2; Lyc. 1.102; Arist. Ath. Pol. 54.7; Pol. 1308b; Plut. Mor. 841b; IG II3 348.13 32 Threatte 1980 vol. 1, pp. 178, 238–61, esp. 13.00–13.02. However, inscriptions in the private realm and especially those on curse tablets allow for more variation in script and orthography; for example, the consistent use of the Ionic script and both OY and EI on SGD 1, a curse tablet found in the right hand of a skeleton in the Kerameikos, dates to the fifth-century on the basis of pottery within the grave assemblage (namely a squat of the mid fifth-century; SGD 1; Peek, Kerameikos 3, p. 91 no. 3).
A New Commercial Curse Tablet from Classical Athens
169
Hom. Il. 10.443: … . In this text the repetition of dentals, |[ ] , gives the curse an incantatory, chant-like feel, helping to season the language of the spell itself.33 : initially, the scribe incised a sigma instead of the delta; realizing his error, he cut the delta more deeply over the still visible sigma (Fig. 4). The horizontal left bar of the omega touches the vertical right leg of the mu. Line 8: [ ] : the tablet’s left edge is broken and jagged; the delta is entirely lost. Between the rho and the epsilon are what appear to be two parallel iotas, the first shorter than the second; this spelling was likely deliberate, incorporating the glide /j/.34 Demetrios is here referenced a third and final time, now by himself, and addressed in the vocative. : the lead is blotchy and cracked along the tablet’s bottom edge, making the lower half of the epsilon and rho difficult to read without the aid of raking light. Some of the most elegant Attic curse tablets employ analogies using ; see the chillingly deictic DTA 67.8–10, | [ | ] [ ] , with referring to the retrograde letters incised on the tablet.35 : here the letters are larger (0.45 cm. in height), sloppier, and more widely spaced than in the lines above; the script dips down for the rho, alpha, and tau, before inclining upwards for the rest of the word. As a superlative adjective, could ostensibly modify , though the scribe’s insertion of postpositive suggests that these rather belonged to two separate clauses. It might instead be possible to understand as functioning here in the same way as with the superlative ( ), thus reading (with [ ] ) as an adverbial phrase, “as strongly as is possible”. Line 9: [ ] : the tablet’s broken left edge destroyed all but the upper tip of the iota and the nu. The restoration of is certain based upon the other tablets. With the inclusion of later in this line, it seems that [ ] should be understood with the preceding clause, see supra. : , with scribal confusion likely precipitated by the similarities in pronunciation between [ei], [e:], and [e:i], as well as the transition from the Old Attic script (see supra).36 Many Attic curse tablets target the tongue of the victim; usually appears in the accusative, but here makes sense as the verb is no longer but , which should take a dative. Several other Attic curse tablets, which also concern themselves with commercial ventures (shops and workshops), target the tongue of their victim(s); it is striking how varied Attic spellings of could be: (DTA 74.3), [ ] ( ) (DTA 75b.7), ( ) (DTA 84a.2), (DTA 87.4), ; (DTA 97.3, 7–8), et al.37 Finally, that the scribe felt the need to insert the postpositive particle after this noun suggests that begins its own clause, separate from the preceding [ ] . : initially the scribe left out this important postpositive particle, but later went back and incised it in very faint, tiny letters (c. 0.1 cm. in height) in the narrow space between lines 8 and 9, below the pi and epsilon of , and above the kappa and upsilon of . The lead here is blotchy and cracked, and the can only be read with the aid of raking light. The inclusion of indicates that everything from through should be understood as its own clause, separate from the preceding … [ ] . 33 Versnel 2002, pp. 105–58. 34 Cf.
, I. Smyrna 797; Threatte 1980 vol. 1, pp. 210f.
35 Tambiah understood “magical” speech acts as marked by two sorts of figurative language: metaphor and metonymy
(1985b, pp. 41–3). This framework explains the presence of analogies with in Attic curse tablets, and also the use of in line 9, which was used metonymically for Demetrios himself. 36 The interchanging of eta and epsilon is known from other early fourth-century Attic curse tablets, including SGD 9 (dated stratigraphically to c. 400 BCE, SEG 21.1093), in which E is twice used for (with H = 6 times al.); Threatte 1980 vol. 1, p. 161. 37 See also DTA 97, DT 47, DT 52, SGD 75.
170
J. L. Lamont
: the lead is discolored and corroded along the bottom edge, and most letters are cracked and faint; these letters too are larger and more carelessly incised than those toward the beginning of the text. This penultimate word is arguably the narrative’s most perplexing; the literal translation of is “dog’s ear”. Acts of cursing certainly employed exotic organic materials, such as fish, birds, a snake’s tooth, a wolf’s beard, a powdered lizard, and even the eyes of a bat;38 the tool-kits of and kindred practitioners were wondrous and macabre. Yet it is more likely that this term comes from the realm of gambling, as a contemporary fragment of the playwright Euboulos mentions as the name for a throw of 39 dice, the lowest possible roll at that. As these four curses reference taverns, venues ripe for gambling, perhaps is best understood in this sense. is certainly the direct object of , a verb whose double prefix thoroughly emphasizes the motion of striking; as I understand it, the noun is entirely separate from and not modified by the superlative , as indicated by the presence of after . If this dice-throw was indeed unlucky, casting it upon an enemy’s tongue makes sense in a binding curse. Yet another possibility is that this term betrays a ritual that took place during the graveside deposition of the tablets; perhaps it indicates an actual throw of the dice, or , during the ritualized binding process. : + ; already has the force of “knocking in” or “striking in” (cf. Ar. Ves. 130), but the addition of a second prepositional prefix, , further emphasizes this motion (“I shall knock-in in addition”). This unique verb takes a dative, , and just as it threatens to knock or strike a into Demetrios’ tongue, the physical act of hammering a nail into the lead tablet would have ritually echoed this wished-for sentiment. By the end of the tablet, as evidenced in , the angular crispness of the incised letters gives way to sloppier, larger, and more widely spaced ones. The scribe initially incised another nondescript letter in the place of the omicron, only traces of which are now visible. Only the left half of the final omega is visible, the rest of the vowel is obscured by heavy encrustation at the end of the line from where the iron nail oxidized (Fig. 4). Discussion The curse begins by summoning three sub-earthly deities, Hekate Chthonia, Artemis Chthonia, and Hermes Chthonios; it commands them to resent a wife-husband couple, Phanagora and Demetrios, as well as their , money, and possessions. Using the first person, the curse binds Demetrios and Phanagora – whose names are repeated a second time – in blood and in dust, together with all the dead. Strikingly, the phrase is hexametric, appearing twice in Homeric epic (Od. 22.383–6; Il. 16.796). After this, the narrative is less clear; I offer an interpretation in which the curse prohibits the release of the binding after a period of five years, (counting inclusively, a four-year period by modern count). This potential ‘limit’ to the period of binding makes 11948 unique, along with the three nearly identical curse-narratives with which it was buried. Unavoidably, it is unclear why the scribe bothered to mention . The period could correspond to the four-year cycle of a large religious festival, such as the Olympic or Panathenaic Games. At the very least, it suggests that the pentaeteric religious cycles were common ways of construing (future) spans of time in Classical Athens; perhaps in crafting a lasting binding curse, the author desired the spell to outlast the next cycle of religious festivals. This term was likely drawn from the civic realm. Such echoes of civic or legal language could have bolstered the curses with an air of authority, though possibly these formulae were influenced by (and indeed, developed alongside) the growing use of writing in legal, commercial, and other municipal contexts. The curse finishes by addressing Demetrios directly. It threatens to bind him, and to strike upon his tongue a . Literally meaning “dog’s ear”, this term likely comes from the world of gambling; a was the name for the lowest possible throw of dice. By striking Demetrios’ tongue with this condemningly unlucky roll, the curse reveals that local taverns were not just sociable watering holes, but venues ripe for gambling and other unsavory activities in Classical Athens. 38 Apul. Apol. 1.3; Hor. Sat. 1.8; Theocr. 2; PGM IV. 39 An. Ox. 2.21; Eubulus Fr. 57 (but
in Poll. 7.205).
A New Commercial Curse Tablet from Classical Athens
171
Commissioning a curse tablet was a drastic measure; commissioning five betrays an even greater investment, and state of desperation, on the part of the curser. As discussed above, 11948 was but one in a cache of five lead tablets, all of which were pierced with nails and found together in situ, in a secure and datable archaeological context. The pottery from the cemetery, as well as the tomb styles and grave assemblages, date the cache to the late fifth or fourth century BCE; this coheres with the tablets’ orthography, especially the presence of etas and omegas, which in public Athenian inscriptions followed the new Ionic alphabet of 403/2 BCE. That four of these tablets contain nearly identical curses – with highly detailed narratives that target four husband-wife couples of the same profession – makes them even more remarkable. Over three-quarters of published Greek curse tablets carry mere lists of names, or are otherwise too brief to understand the reasons for their creation.40 Many curse tablets, furthermore, lack a provenance, and too often their dates are assigned on the shaky basis of script or letter-form. Even in tablets known to have come from a certain sub-region, like the Piraeus, the specific circumstances of excavation (and thus depositional context) are usually unknown. This new cache of curses, by contrast, reveals a great deal about private ritual, commercial rivalries, and social networks within a local Attic deme community in the late fifth or early fourth century BCE. First, this cache survives as a ritually-bound assemblage. It allows us to discuss not only Demetrios and Phanagora, but a total of eight targets across the four incised tablets. The commissioner thought that he (they?) could profit from stymying the work and property of these four couples, all of whom made a living in a narrowly similar way – as tavern keepers; the commissioner intended the four couples to suffer primarily in this capacity. That all targets were tavern keepers seems unlikely to have been coincidental, and suggests that the curser himself was involved in the world of the . In Classical Athens, taverns were a common feature of the urban landscape, and tavern keepers were sometimes known for their scheming or cheating tendencies.41 Many Attic curse tablets relate to the world of the tavern, a commercial venue filled with a motley clientele, from travelers to prostitutes to gamblers.42 Athenians excluded from elite symposia – that bulwark of aristocracy – would have found refreshment in a local , a space more 43 public and democratic than the private . The profession of tavern-keeper was one that demanded a low level of skill compared to that of a potter or bronze worker, but afforded a high degree of familiarity and visibility within the local community; that this cache of curses lists four male tavern-keepers by name, along with two named wives, suggests that the commissioner was indeed well acquainted with these four couples and their . But why were they made the targets of such elaborate curses? It is impossible to say with certainty, but because the thread linking the narratives seems to have been occupational, the tablets were likely composed on account of business rivalry; rival taverns – with which the commissioner was intimately familiar – posed a threat that someone sought to control with an involved group of binding spells. Curses targeting shops and businesses were traditionally grouped under the subgenre of tablets known as commercial curses, though recently this category has come under question.44 However, this new cache of tablets, which targets four and likely the husband-wife couples running them, bolsters the validity of this category, and reveals that commercial rivalries could indeed motivate binding curses in Classical Athens. 40 Eidinow 2007, p. 154. 41 Ar. Th. 347, Plout. 435–6, here
.
42 DTA 30, 68, 70, 75, 87, with more detailed discussions of
in the commentary on l. 3 above.
43 Davidson 1997. 44 Faraone 1991, p. 27 fn. 47 included the following curse tablets within the “commercial” category: DTA 68–75, 84–7; DT 47, 52, 70–3, 92; SGD 20, 52, 73, 75, 88, and 124. Gager (1992, p. 151) added DTA 55, 97, 108–9; SGD 3, 60, 44. Ogden (1999, p. 34) adds SGD 3, 48, 72, 124, and 170; DTA 12, 30. Lopez-Jimeno (1999) adds SGD 4, 11, 43, 81. Most recently, Eidinow (2007, pp. 191–205) has questioned the validity of this category, dismissing the categorization of several of the above tablets, and noting that “for a number, the explanation of commercial competition simply does not make sense … an explanation that turns on a simple statement about commercial competition cannot adequately explain the contents of them all, and that other explanations (concerning local rivalries) must be considered”.
172
J. L. Lamont
Though there is nothing quite like this cache of binding curses, several other Attic tablets target , as discussed above.45 A parallel can perhaps be seen in DT 70–2, and SGD 43, a “group” of four Attic curse tablets dating from the fourth century BCE. I caution that these tablets lack a provenance, and have found their way to different European collections; furthermore, the curses incised on the tablets are only alike in that they target some of the same individuals, although each tablet employs different language and syntax.46 DT 70 begins [ ] | [ ] | (DT 70.1–3). Like 11948 and the tablets deposited with it, this curse too is rooted in the world of the ; it targets two taverns – one of which was charmingly named “Olympos Tavern” – along with named individuals who appear again throughout tablets DT 71, 72, and SGD 43. That the same individuals and businesses appear across these four tablets suggests that they were likely conceived of and commissioned as a group. Another Attic tablet, also from the fourth century BCE and based in the world of the tavern, is DTA 87a: :
1 2 3 4 5
| DA. OH |
: :
. [ ]
:
[ ]
: 6 7
[]
[] :
:
[
]
.
() ( )
.
8 ( ). 9 10
( ) : .
R
This elaborate curse tablet was elegantly incised and, like our cache, targets a number of taverns and tavern keepers, including at least one married couple (lines 1–3). In this text, however, the tavern keepers appear alongside individuals from other professions (e.g., line 5, , the cloth seller). Also like our tablets, DTA 87a reveals a high degree of familiarity, if not proximity and frequency of interaction, with the named targets; indeed, the narrative reveals that the commissioner of the curse was a neighbor of the targets (lines 1, 7). Additionally, some of the taverns targeted in DTA 87a are described by the name and demotic of their owner (line 8), and like the “Olympos Tavern” targeted in DT 70–2 and SGD 43, expose a degree of intimacy between the commissioner of the curse and the target(s).47 While these comparanda illuminate the ways in which Attic curses could target local taverns and tavern keepers, none quite resemble the unique text of 11948. Taken together, however, they show how curses circulated within local community networks. In considering this cache as an assemblage, it is important to remember that one of the five lead sheets was left un-inscribed.48 This blank tablet was still pierced by a nail, suggesting the presence of ritual even in the absence of text. Some Attic curse tablets consist only of named targets in the accusative case (DTA 34), likely serving as direct objects for an assumed but absent verb of binding that was recited orally. 45 E.g., DTA 30, 68, 70, 75, 87; SGD 11, 43; commentary on line 3 above. 46 DT 70, 71, 72 (?): in the collections of the British Museum; SGD 43: location now unknown, though formerly in the Froehner Collection in the Cabinet des Medailles of the Bibliotheque Nationale, Paris. 47 Wünsch notes, in reference to (DTA 87.2), “aliter explicare nequeo nisi ut hoc nomen tabernae inditum fuisse putem”. It could thus be the proper name of the tavern, as per Wünsch (cf. , , DT 70.2–3), or perhaps – as I suspect – the commissioner of the curse tablet was simply describing the targeted as the “tavern of the bald man”, whose distinct appearance took the place of his proper name in ‘targeting’ the curse. 48 Petritaki 2009, p. 464.
A New Commercial Curse Tablet from Classical Athens
173
Perhaps the single blank tablet in this cache, pierced by a nail and interred with the four inscribed curses, should be understood as a lead strip over which a binding curse was recited but not inscribed. At the very least, the blank tablet pierced with a nail and interred in a pyre-grave shows that each lead strip was but one vehicle in a larger chain of private ritual activity. The process of composing, incising, folding, and depositing the tablet likely involved incantations and other rites that, unfortunately, have not been preserved in the archaeological record alongside the metal sheets and nails. For example, one Attic curse tablet mentions the presence of organic materials within the binding ritual such as wax and a liquid (water or wine), or possibly ‘magic thread’ if we accept [ ] over [ ] : “I bind them all in lead and in wax and in water/wine/magic thread and in idleness and in obscurity and in ill-repute and in defeat and among tombs” ( | [ ] | , DTA 55, lines 16–8).49 Additionally, the Greek Magical Papyri, though temporally and geographically removed from the Classical Attic curse tablets under consideration here, reveal an array of spoken incantations and ritual actions employed during the cursing process; these include chants, recited and repeated oral formulae, wax figures, yarn, flowers, and the presence of , bits of hair or clothing – the very essence or “being” – from the person being cursed (e.g., PGM IV. 296–335). In this vein, a tablet from the Athenian Agora contained several extant strands of hair, which likely assisted in ‘targeting’ the spell against the woman Tyche (Agora Inv. IL 1737). Such organic materials rarely survive in the archaeological record, but were important parts of the binding ritual; the act of incising text onto a lead tablet was just one part of the larger cursing process. These five lead tablets were endowed with the capacity to shape and transform the local deme community in which they circulated. Far from being fixed carriers of meaning, the lead tablets, nails, organic materials, and oral recitations involved in the binding process are best understood as dynamic ritual processes, in which both the narrative and the object actively shape and were shaped by the spaces in which they circulated; speech and ritual were efficacious acts in the binding process.50 Though this article largely focused on the text of a single tablet, I have tried – by way of presenting the cache within its archaeological, socio-cultural, and ritual contexts – to show that, as objects, the tablets were inseparable from the social processes in which they were embedded. In order to be meaningful, in other words, the texts should not be considered apart from their material carriers, or from the context in which they functioned socially. 11948 intended, in a very real way, to engage with and change the community in which it circulated. Along with its four nailed comrades, the cache sheds light on private ritual, commercial rivalries, and social networks within a local Attic deme community during a period historically dominated by war and shifting political alliances. Abbreviations [DT] Audollent, A. M. H. (1904) Defixionum tabellae quotquot innotuerunt tam in Graecis orientis quam in totius occidentis partibus praeter Atticas in Corpore inscriptionum Atticarum editas. [DTA] Wünsch, R. (1897) Defixionum tabellae Atticae. [LGPN] Fraser, P. M. – E. Matthews (1987–) A Lexicon of Greek Personal Names. [NGCT] Jordan, D. (2000) New Greek Curse Tablets (1985–2000), GRBS 41, pp. 5–46. [PGM] Preisendanz, K. – A. Henrichs (1973–4), Papyri Graecae Magicae: Die griechischen Zauberpapyri; 2 vols. [SGD] Jordan, D. (1985) A Survey of Greek Defixiones Not Included in the Special Corpora, GRBS 26, pp. 151–97. [SM] Daniel, R. – F. Maltomini (1990–1) Supplementum Magicum, 2 vols., Papyrologica Coloniensia 16.1–2.
49 For
[ ]
, see Parker 2005, p. 127.
50 The lead tablets were not, in other words, passive objects or mere recipients of the human agents and actions that creat-
ed them. As Tambiah emphasizes – applying Austin’s speech act theory to rituals – utterances were seen as performative acts that did something; the binding curses incised on tablets were real utterances, representing very real actions (Tambiah 1985a, pp. 78–80; Tambiah 1985b, pp. 30–2; Austin 1975 passim). For a larger theoretical framework, see Bell 1992; Bell 1997; Gell 1998 (esp. p. 6); Malafouris 2013.
174
J. L. Lamont Bibliography
Austin, J. (1975) How to Do Things with Words. Bell, C. (1992) Ritual Theory, Ritual Practice. – (1997) Ritual, Perspectives and Dimensions. Collins, D. (2008) The Magic of Homeric Verses, CP 103, pp. 211–36. Davidson, J. N. (1997) Courtesans and Fishcakes: The Consuming Passions of Classical Athens. Eidinow, E. (2007) Oracles, Curses, and Risk among the Ancient Greeks. Faraone, C. A. (1985) Aeschylus’ (Eum. 306) and Attic Judicial Curse Tablets, JHS 105, pp. 150–4. – (1989) An Accusation of Magic in Classical Athens (Ar. Wasps 946–48), TAPA 119, pp. 149–60. – (1991) The Agonistic Context of Early Greek Binding Spells. In C. A. Faraone – D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion, pp. 3–32. Gager, J. G. (1992) Curse Tablets and Binding Spells from the Ancient World. Gell, A. (1998) Art and Agency: An Anthropological Theory. Jordan, D. R. (1985) Defixiones from a Well near the Southwest Corner of the Athenian Agora, Hesperia 54, pp. 198–252. Lauter, H. (1991) Attische Landgemeinden in klassischer Zeit, Attische Forschungen 4. López J. – M. del Amor (1999) Nuevas tabellae defixionis Áticas. – (2001) Textos griegos de maleficio. McInerney, J. (2015) “There Will be Blood …”: The Cult of Artemis Tauropolos at Halai Araphenides. In K. Daly – L. A. Riccardi (eds.), Cities Called Athens: Studies Honoring John McK. Camp II, pp. 289–320. Malafouris, L. (2013) How Things Shape the Mind: A Theory of Material Engagement. Nachmanson, E. (1904) Laute und Formen der magnetischen Inschriften. Ogden D. (1999) Binding Spells: Curse Tablets and Voodoo Dolls in the Greek and Roman Worlds. In B. Ankarloo – S. Clarke (eds.), Witchcraft and Magic in Europe, pp. 3–90. Papachristodoulou, I. (1973) , Archaiologike Ephemeris, pp. 189–217. Parker, R. (2005) Polytheism and Society at Athens. – (2011) On Greek Religion. Peek, W. (1941/56) Kerameikos. Ergebnisse der Ausgrabungen, vol. III, Inschriften, Ostraka, Fluchtafeln. Petritaki, M. (2009) . In V. Vasilopoulou – S. Katsarou-Tzeveleki (eds.), . : . 18–20 2003, pp. 451–80. – (2010) : 131–133 137, 56–59 B1, pp. 449–51. Simon, E. (1985) Hekate in Athen, AM 100, pp. 271–84. Smyth, H. (1984) A Greek Grammar. Benediction Classics. Tambiah, S. (1985a) Form and Meaning of Magical Acts. In S. Tambiah (ed.), Culture, Thought and Social Action. An Anthropological Perspective, pp. 60–86. – (1985b) The Magical Power of Words. In S. Tambiah (ed.), Culture, Thought and Social Action. An Anthropological Perspective, pp. 17–59. – (1985c) A Performative Approach to Ritual. In S. Tambiah (ed.), Culture, Thought and Social Action. An Anthropological Perspective, pp. 123–66. Threatte, L. (1980) The Grammar of Attic Inscriptions, vol. I, Phonology. Traill, J. S. (1975) The Political Organization of Attica: A Study of Demes, Trittyes, and Phylai, and their Representation in the Athenian Council. Hesperia Supplements 14. Versnel, H. S. (1991) Beyond Cursing. The Appeal to Justice in Judicial Prayers. In C. A. Faraone – D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion, pp. 60–106. – (2002) The Poetics of the Magic Charm: An Essay in the Power of Words, in P. Mirecki – M. Meyer (eds.), Magic and Ritual in the Ancient World, pp. 105–58. Wünsch, R. (1900) Neue Fluchtafeln I, Neue Fluchtafeln II, RhM 55, pp. 62–85, 232–71.
Jessica Laura Lamont, The Johns Hopkins University, Department of Classics, 113 Gilman Hall, 3400 N. Charles St., Baltimore, MD 21218
[email protected]
