1991 Binding and Burying With Cover Page v2

Accelerat ing t he world's research.
Binding and Burying the Forces of

Evil: The Defensive Use of "Voodoo
Dolls" in Ancient
Christopher Faraone
Cite this paper Downloaded from Academia.edu 
Get the citation in MLA, APA, or Chicago styles
Related papers Download a PDF Pack of t he best relat ed papers 
“Clay Hardens and Wax Melt s: Magical Role-Reversal in Vergil's Eight h Eclogue” Classical Philol…
Christ opher Faraone
“T he Wheel, t he Whip and Ot her Implement s of Tort ure: Erot ic Magic in Pindar Pyt hian 4. 213-19” Clas…
Taking t he "Nest or's Cup Inscript ion" Seriously: Erot ic Magic and Condit ional Curses in t he
Binding and Burying the Forces of Evil: The Defensive Use of "Voodoo Dolls" in Ancient
Greece
Author(s): Christopher A. Faraone
Source: Classical Antiquity, Vol. 10, No. 2 (Oct., 1991), pp. 165-205, 207-220
Published by: University of California Press
Stable URL: http://www.jstor.org/stable/25010949
Accessed: 19-01-2017 18:45 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/25010949?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
University of California Press is collaborating with JSTOR to digitize, preserve and extend access to
Classical Antiquity
This content downloaded from 128.135.100.111 on Thu, 19 Jan 2017 18:45:44 UTC
All use subject to http://about.jstor.org/terms
CHRISTOPHER A. FARAONE
Binding and Burying the Forces of Evil:

The Defensive Use of "Voodoo Dolls" in
Ancient Greece
THERE IS MUCH evidence that the ancient Greeks used binding magic at
public ceremonies to protect their cities and in more private rituals to resolve
interpersonal problems in their day-to-day lives.' The large number of inscribed
lead defixiones attests to the popularity of private binding magic as early as the
fifth century B.C. in Sicily and Attica.2 In this essay I survey still another rich
1. A portion of this article was presented at the 119th Annual Meeting of the American
Philological Association in New York City in December 1987, and a great deal of it appears as
chapter four of my Stanford Ph.D. dissertation, "Talismans, Voodoo Dolls, and Other Apotropaic
Statues in Early Greece" (1988). I should like to thank W. Brashear, W. Burkert, M. Edwards, F.
Graf, M. Jameson, D. Jordan, R. Kotansky, J. Scurlock, H. S. Versnel, J. Winkler, and others for
their encouragement and comments on earlier drafts. I should also like to thank the National
Endowment for the Humanities, whose Summer Stipend for 1989 allowed me to complete work on
this article. The errors that may remain are mine alone.
The following abbreviations will be used throughout: ANET, J. B. Pritchard, ed., Ancient Near
Eastern Texts Relating to the Old Testament3 (Princeton, 1969); DT, A. Audollent, Defixionum
Tabellae (Paris, 1904); DTA, R. Wiinsch, Defixionum Tabellae Atticae, IG 3.3 (Berlin, 1897); KAR,
E. Ebeling, Keilschrifturkunden aus Assur religiose Inhalts, 2 vols. (Leipzig, 1915-23); KBo,
Keilschrifttexte aus Boghazkoi (Berlin, 1916-); MMG, J. Scurlock, "Magical Means of Dealing with
Ghosts in Ancient Mesopotamia" (diss., Chicago, 1988); PGM, K. Preisendanz and A. Henrichs,
eds., Papyri Graecae Magicae: Die griechischen Zauberpapyri2 (Stuttgart, 1973-74); SGD, D. R.
Jordan, "A Survey of Greek Defixiones Not Included in the Special Corpora," GRBS 26 (1985) 151
97. Unless otherwise indicated, the number that follows any of these abbreviations refers to the
number of the text in the collection, and not to a page number.
2. DTA and DT are the basic collections. For bibliography and description of material pub
lished subsequently, see SGD. Aside from the prolegomena to the above-mentioned corpora and
surveys, comprehensive discussions include: E. G. Kagarow, Griechische Fluchtafeln, Eus Supple
menta 4 (Leopoli, 1929); S. Eitrem, "Bindezauber," RAC 2 (1954) 380-85; K. Preisendanz
? 1991 BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIA
166 CLASSICAL ANTIQUITY Volume 10/No. 2/October 1991
source of information-the literary, epigraphical, and archaeological evidence

that the Greeks in all periods used bound or buried images to control inimical
gods and demons as well as ghosts and other more human enemies.3 In doing so I
hope to show that the aim of these rituals, whether performed on behalf of the
state or an individual, was to control but not destroy the source of evil, and that
they were regularly perceived as forms of defensive magic. The essay falls into
four discrete sections. I begin by examining the abundant testimony to defensive
rituals that involve binding or burying the effigy of Ares; this material reveals the
close correlation between early Greek myth and much later ritual, and allows us
to see the continuity of beliefs about the bound god over a long period of time. In
order to set the material about Ares into a larger cultural context, I devote the
second section to a detailed survey of the binding and burial rituals (both public
and private) used by the Egyptians and the Assyrians. In the third section of the
paper I turn away from the gods and focus on the use of effigies in Greek magic
to protect people against the attacks of ghosts. After a detailed examination of
the role of the problematic xooaoooi in the so-called cathartic inscription from
Cyrene, I discuss how Greek ghost-banning combines funerary ritual and sympa
thetic magic in a unique way. In the final section I examine the archaeological
evidence for the private use of "voodoo dolls"4 in the Greek world as early as the
seventh century B.C., arguing that, in light of the other evidence collected here,
these artifacts can now be seen as merely one colorful band of a wide spectrum of
defensive magical rituals used to control the forces of evil in the ancient world.
BINDING ARES IN GREEK MYTH AND RITUAL
In the fifth book of the Iliad Aphrodite collapses into the lap of her mother,
Dione, and complains bitterly about the wound that she has just received from
Diomedes; when Dione tries to comfort her with a list of the indignities that
other gods have suffered at the hands of overweening mortals, she begins with
the god of war himself:5
"Fluchtafel (Defixio)," RAC 8 (1972) 1-29; 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) 3-32.
3. For some recent treatments of myths and rituals concerned with binding the gods in ancient
Greece, see K. Meuli [and R. Merkelbach] "Die Gefesselten Gotter," in K. Meuli: Gesammelte
Schriften 2 (Stuttgart, 1975) 1077-78; F. Graf, Nordionische Kulte, Bibltheca Helvetica Romana 21
(Rome, 1985) 81-96; H. S. Versnel, "Greek Myth and Ritual: The Case of Kronos," in J. Bremmer,
ed., Interpretations of Greek Mythology (London, 1987) 121-52.
4. I use the term "voodoo doll" advisedly throughout simply as the closest and most familiar
American equivalent of the German Rachepuppe or the French figurine d' envoatement, without
implying or suggesting any connection whatsoever to the religious practices of the Afro-Caribbean
people of Haiti and elsewhere.
5. The Greeks of later antiquity seem to have recognized the magical act reported in these
lines; the opening line (385) is twice quoted in a magical papyrus as an efficacious charm (PGM IV
474, 830). Although these quoted lines lack the rubrics that accompany most of the other Homeric
FARAONE: "Voodoo Dolls" in Ancient Greece 167
Ares had to endure it when strong Ephialtes and Otos,

sons of Aloeus, chained him in bonds that were too strong for hi
and three months and ten he lay chained in the brazen cauldron
and now might Ares, insatiable of fighting, have perished,
had not Eeriboia, their stepmother, the surpassingly lovely,
brought word to Hermes, who stole Ares away out of it
as he was growing faint and the hard bondage was breaking him
(II. 5.385-91, trans. R. Lattimore)
The mention of the two heroes Ephialtes and Otus points to a Boeot
for this myth, but a papyrus chrestomathy dated to the second cen
suggests that a similar myth was told about Enyalius (Ares) in Thrac
Binding is one of the most popular weapons with which the g
each other in Greek myth: Cronus is bound by Zeus (Hes. Erga 173a
threatened with the same fate at the hands of Poseidon, Apollo, an
1.399-401); and Hephaestus binds Ares and Aphrodite in bed (O
99), and even his own mother (Pindar, fr. 283 Snell; Plato, Rep. 378d
this phenomenon is a logical corollary to the fact of the gods' imm
since they cannot kill one another, they can only defeat their
imprisoning them, binding them, or similarly restricting their mo
Compare for example, Hephaestus's treatment of Ares in Demodocus
the summary execution of an adulterer caught in flagrante delicto
Athens (Lys. 1.49, Dem. 23.53). The myth about Ephialtes and Ot
Ares, however, is even more curious, because although it can no
linked to episodes in the life of the Etruscan Mars,7 it does corresp
verses cited in the magical papyri, it seems an obvious inference that the verse (and
handbook this may simply be shorthand for the entire passage) was to be used in a b
Compare the quotation of 11. 8.424 ("Will you dare to raise your mighty spear ag
accompanied by the rubric "To Restrain Anger," or I/. 10.193 ("Let's . . . seize, lest
joy to our enemies"), which is to be used "To Get a Friend." The Geoponica (10.87.6)
attaching verse 5.387 of the Ares passage ("and three months and ten he lay chained
cauldron") to a tree to prevent it from prematurely casting its fruit; i.e., it "binds" the t
to its fruit until the correct moment in its annual cycle.
6. POxy 1241 col. 4.20-24. Thrace is Ares' homeland (II. 13.301, Od. 8.361); see L
The Cults of the Greek States 5 (Oxford, 1909) 399-403. A similar myth seems to have
located) at the Cretan city of Biennos, whose name, according to Stephen of Byz
commemorated the Pie ("act of violence") that had befallen Ares, "whom they say [wa
by Otus and Ephialtes, the sons of Poseidon, and even now they give Ares the sacrif
Hekatomphonia." The Hekatomphonia appears to have been peculiar to the Messenians
of Biennos, and perhaps the Athenians; see A. Tresp, Die Fragmente der G
Kultschriftsteller, Religionsgeschichtliche Versuche und Vorarbeiten 15.1 (Giessen, 1
for a discussion of the pertinent texts, which all seem to go back to Diophantos of Spart
7. See Farnell (supra, n. 6: 407) and E. Simon, "II dio Marte nell'arte dell'Italia
46 (1978) 135-47, for the traditional interpretation, countered by H. S. Versnel, "Ap
One Hundred Years after Roscher," Visible Religion 4/5 (1985/86) 147-48. For poss
ern parallels to the brazen cauldron in the Boeotian myth, see H. A. Hoffner, "Hittit
Hebrew Teraphim," JNES 27 (1968) 65-66, who discusses the myth of the Hittite wea
Lihnzina, in which hostile demons and abstracted evils (e.g., "Bloodshed," "Tears") ar
nicely to some actual Greek rituals that were apparently performed in the
later periods.
Perhaps the most interesting testimony to these rituals is a first-century-B.c.
hexametric inscription that appears to be a text issued by the Panhellenic oracle
of the Clarian Apollo. In it, the people of Syedra, plagued by the incessant
attacks of pirates, are advised to erect a statue of Ares bound in the "iron chains
of Hermes" and supplicating a figure of Dike:8
Pamphylians of Syedra inhabiting a land in the bordering territory of the

people of mixed race, set up an image of Ares, the blood-stained slayer
of men, in the midst of your town and perform sacrifices beside it, while
holding him [sc. the statue] in the iron bonds of Hermes. On the other
side let Dike giving sentence judge him, while he himself is like to one
pleading. For thus he will be peacefully disposed to you, after he has
driven out the unholy mob far from your native land, and will raise up
much-prayed-for prosperity. But also you yourselves together put your
hand to the hard task, either chasing them away or manacling them in
unloosable bonds, and do not give to delay the terrible vengeance on the
plunderers. For thus you will escape from impairment.
(trans. H. W. Parke)
Despite the array of pronouns whose reference is not always certain, one can
easily make out the ceremony: the creation of a statue of the bound Ares kneel
ing9 before a menacing statue of Dike10 is expected "sympathetically"1 to re
in copper kettles sealed with lead and then deposited in the sea. Hoffner compares these with the
similarly sealed copper kettles that appear in the underworld in other Hittite myths (they are said to
contain "Wrath, Anger, Sin, and Rage"), a female statuette representing "evil" sealed in a jug with a
lead lid (Zech. 5:5-11), and Pandora's Jni0og in Hesiod.
8. G. E. Bean and T. B. Mitford (Journeys in Rough Cilicia in 1962 and 1963, DenkschrWien
85 [Vienna, 1965] 21-23 no. 26) and F. Sokolowski ("Sur l'oracle de Claros destine a la ville de
Syedra," BCH 92 [1966] 519-22) date the inscription to the first century B.C. L. Robert (Documents
de l'Asie Mineure meridionale [Geneva, 1966] 91-100) and J. Wiseman ("Gods, War and Plague in
the Times of the Antonines," in D. Mano-Zissi and J. Wiseman, eds., Studies in the Antiquities of
Stobi 1 [Belgrade, 1973] 174-79) attempt unsuccessfully to redate the inscription to the reign of
Lucius Verus using numismatic evidence; see E. Maroti, "A Recently Found Versified Oracle
Against the Pirates," AAASzeged 16 (1968) 233-38, for a good refutation of Robert, an article of
which Wiseman was apparently unaware.
9. H. W. Parke (The Oracles of Apollo in Asia Minor [London, 1985] 250 n. 20) and Graf
(supra, n. 3), follow Robert's suggestion (supra, n. 8) that the tableau contained three standing
figures. The expression in the Syedran inscription ("while restraining him with the iron bonds of
Hermes") need not mean that the god himself was actually depicted, but only that he was believed to
have invented the shackles or that they were within his traditional sphere of activity (cf. Veneris
vincula, Verg. Ecl. 8.78; surely we need not suppose that Venus made an epiphany during the spell).
Hermes often acts as an agent of binding in the texts of Attic defixiones in his guise as Katochos
("Restrainer"), as well as in some obscure myths about binding Charon and Prometheus (see Serv.
ad Aen 6.42). For the interpretation followed here, see C. A. Faraone, "Clay Hardens and Wax
Melts: Magical Role-Reversal in Vergil's Eighth Eclogue," CP 84 (1989) 296-97.
10. Sokolowski (supra, n. 8: 521) suggests that the statue of Dike in Syedra was depicted in an
aggressive stance like the Dike brandishing a da36bog and throttling Adikia on the Chest of Cypselus
strain and humble the hostile force of the pirates. Ares' relation
pirates is a bit unclear, but we do find evidence in the Greek novels
where for a popular belief that brigands and highwaymen were thoug
special devotees of this god.12 This ritual is much more elaborate th
binding an image of Ares; it creates a scenario in which both the inten
(Ares and the pirates) and the performers of the spell (the people o
represented by the figure of Dike) appear. This kind of complicated r
however, unique; it can be paralleled in a recipe preserved in the Gre
papyri, as well as in descriptions of similar erotic spells in Vergil and
The other explicit testimonia for the magical binding of effigies
all much later, but seem nevertheless to be variations of the same typ
The first witness is an anonymous epigram in the Palatine Antholog
EiS; oriXklv "AQEcog EX?waO?VTiV Ev OEQa@Xx
Eiloxe OVQ6Lo g otog ETl T r XOovL xeXXLTaL "AQSg,

oiTOTe O Qr'ixiESg E'3io0ETaLt e0vea rF6TO(v.
[Inscribed] on the base of an Ares [statue] that lies buried in Thrace:
As long as fierce Ares here has been laid low upon the ground
the Gothic peoples shall never set foot upon Thrace.
It is unclear exactly how the image has been treated; the rubric states tha
base was buried, but the epigram itself uses ambiguous language (xEx
"laid low"), a term that could very well suggest that Ares had been
made subservient.14 In any case, the ritual seems to have involved th
(Paus. 5.18.1-2), a scene that appears on a late sixth-century vase painting; see J. Fre
Adikia," in L. Varel, R. F. Willetts, et al., eds., Geras: Studies Presented to Geo
(Prague, 1963) 95-98, pl. 1.
11. I use this term advisedly throughout. S. J. Tambiah, "Form and Meaning of M
A Point of View," in R. Horton and R. Finnegan, eds., Modes of Thought (London, 19
rightly dismisses the common view that "sympathetic magic" is based on poor observatio
cal analogies. He distinguishes instead between the operation of "empirical analo
modern scientific discourse to predict future actions) and "persuasive analogies" (use
traditional societies to encourage future action). Cf. G. E. R. Lloyd, Magic, Reason an
(Cambridge, 1979) 2-3, 7. In the ritual described in this inscription, the binding of
expected to persuade Ares and the pirates to be restrained in similar fashion.
12. In the novel of Xenophon of Ephesus, for instance, Hippothoos and his band
(.norai) garland a wooden statue of Ares and prepare a peculiar form of sacrifice
tossing a javelin at a suspended victim (13.1). G. Palmeyda, Xenophon d'ftphese: L
(Paris, 1962) 32 n. 2, suggests that the motivation for this type of worship lies in
Thracian origins (2.1). In Apuleius's Golden Ass, the Thessalian bandits pour a liba
hymns in honor of Mars (4.22) and sacrifice an old male goat to him on an altar of green
scene in which they are addressed as deo Marti clientes. Ares is regularly connected with
ans to the north; see, e.g., Eur. Ba. 1330ff. (infra, n. 110), or two epigrams from
celebrate that city's escape from the siege of the Gauls in 262-63 A.D. (J. Fontenr
Apollo's Oracle, Cult, and Companians [Berkeley and Los Angeles, 1988] 23-24).
13. Faraone (supra, n. 9) 294-300.
14. Paton (Loeb) translates "rests on the ground," while Waltz/Saury, Anthologie
(Paris, 1974) ad loc. (who date the epigram to the late third / early fourth century on ac
tion and burial of an effigy of the god in hopes of warding off a hostile enemy;
here Ares appears to represent the warlike Goths, who were a continuous threat
to the Thracian border in the late antique period.
A historical account preserved by Photius describes a similar form of binding
ritual, which may have been aimed at averting this same menace from Thrace:15
The historian says that he heard from a certain Valerius, a man of high
rank, about silver statues which were consecrated to ward off barbari
ans. He says that during the reign of Constantius the Emperor, when
Valerius was governor of Thrace, it was reported that a treasure had
been found. Valerius visited the site and learned from the locals that it
was sacred and that the statues had been consecrated there according to
an ancient rite. This he then reported to the Emperor and received a
rescript empowering him to take possession of the objects reported.
When the spot was excavated, three silver statues were found deposited
there, of barbarous style, with hands bound, dressed in the embroidered
clothing of the barbarians, with long hair and inclining toward the
North, that is toward the land of the barbarians. As soon as the statues
were removed, a few days later the whole Gothic nation poured over
Thrace and shortly afterwards the Huns and the Sarmatians were to
invade Illyricum and Thrace also. For the site of the consecration lay
between Thrace and Illyricum and to judge from the number of statues,
they had been consecrated against the whole of barbary.
(trans. R. D. Blockley)
The historian Photius refers to is Olympiodorus, an occasionally tendentious

Egyptian priest from Thebes who searches everywhere for evidence of the effi
cacy of traditional pagan rituals. Despite such shortcomings, his narrative of the
"eyewitness account" of the governor of Thrace with his imperial rescript rings
true, and we are probably justified in accepting his report of the discovery of
some kind of bound and buried artifacts, while dismissing his interpretation of
the catastrophic outcome of their removal (especially since this causes grave
chronological problems with the rest of his narrative).16 Some of the details may
"inspiration paienne"), print couches " but suggest in a note that "appuye" might be closer to the
sense of the Greek. For the use of the verb XLVELV in magical texts to mean "make subservient," see
LSJ s.v., 1.5 (citing PGM IV 1718). The more familiar verb in this context is the intensive form
xacTcxXivetv, e.g., PGM IV 2076, 2450, 2497; SGD 141. For a detailed discussion of this verb in the
context of magical rituals, see F. Maltomini, "I papiri greci," SCO 29 (1979) 102. Perhaps the statue,
which depicted Ares helplessly sprawled on the ground (and bound as well?), was erected and then
subsequently buried in the ground to restrain the god further. An anonymous reader for the journal
suggests yet another possibility: an upright statue of the god was "toppled" during a ritual and then
buried with its base.
15. Phot. Bibl. cod. 80 p. 177 = Olympiodorus of Thebes, FHG 4.63.27 = frag. 27 Blockley.
16. On the chronological problems with Olympiodorus's account of this incident, see J. F.
Matthews, "Olympiodorus of Thebes and the History of the West (A.D. 407-25)," JRS 60 (1970) 96;
R. D. Blockley, The Fragmentary Classicising Historians of the Late Roman Empire (Liverpool,
1981) 164 n. 20.
also be suspected as fanciful embroideries drawn from the author's

background, in particular the differentiation of the three figures by thei
native hairstyles and costumes.17
Thus we see a pattern of a very old Greek myth and some much la
that concern the constraint of hostile armies-armies that must be m
bound if the northern borders of Greece are to be defended. We occ
find hints that similar beliefs were current in other parts of the Gre
well, and at a much earlier date. Pausanias, for instance, describ
image of the war god Enyalius at Sparta:
Opposite this temple [sc. of Poseidon] is an ancient image of Enyal

fetters. The belief of the Lacedaemonians about this statue is the
as that of the Athenians about their Wingless Victory; they believ
Enyalius will never escape by fleeing, while the Athenians think
Victory will always remain in that place because she has no wings
upon.
(3.15.7)
To Pausanias's mind, at least, the statue represents Sparta's continued excel
lence in the craft of war, which is chained down to ensure that it never deserts
her. The Greek god Ares/Enyalius was not, however, a god of martial excellence
like the Roman deity Mars, but rather the reckless, dangerous force of slaughter
and death on the battlefield. It is, of course, easy to understand how such an
explanation could develop, especially in the Imperial period, when Sparta, her
military prowess, and her peculiar social and religious institutions were popular
with the Roman intelligentsia, who came to view Lacedaemonian culture as
some Hellenic prefiguration of their own.18 This kind of confusion is facilitated
by the familiar ambiguity (as in the Syedra inscription, above) between the
encouragement of good fortune (i.e., keeping excellence in war on our side) and
the binding and warding-off of bad fortune (i.e., preventing "war" from hurting
us). Because of the obvious similarities between this Spartan statue and the
Boeotian myths discussed above, many scholars believe that it is part of a cult
17. Nearly all the extant Greek "voodoo dolls" are nude (see the Appendix at the end of this
article for a full survey and n. 94). Olympiodorus may be injecting some of his native Egyptian lore here
to fill in the details. See below for the more elaborate Egyptian "bound captive" effigies, which depict
national headdresses and exact racial characteristics. Compare also the account in Pseudo
Callisthenes, The History of Alexander 1.2 (a third-century-A.D. work written in Egypt), where
Nectanebo, the last of the Egyptian pharaohs, is said to destroy enemy armies by manipulating wax
figurines that depict soldiers in the national dress and arms of their country. 0. Weinreich, Antike
Heilungswunder: Untersuchungen zum Wunderglauben der Griechen und Romer, Religionsgeschicht
liche Versuche und Vorarbeiten 8.1 (Giessen, 1909), 167 n. 2, explains the rationale behind the
number of the effigies reported by Olympiodorus by pointing out that three as well as one is a common
pars pro toto number in Greek ritual (he gives several examples).
18. E. N. Tigerstedt, The Legend of Sparta in Classical Antiquity 2 (Uppsala, 1974) 95-225
passim.
borrowed at an early period from the Boeotians that involved binding effigies of
Ares/Enyalius to prevent him from harming the community'9
It is important to note how these bound statues are treated. The statue at
Sparta, like the one at Syedra, is not housed in a temple, but stands in an open
space in the city, opposite the temple of Poseidon. At Syedra sacrifices are to be
performed at the side of the Ares statue, but it is not clear whether these offerings
were to be performed only once or at regular intervals. A quip attributed to the
famous sophist Herodes Atticus (ca. 100-179 A.D.) suggests that sacrifices were
also traditionally offered to bound statues of Ares in Boeotia. Herodes scoffs at
misers who lock up their wealth in "prisons" and consider it worthy to be offered
sacrifice; such men, he said, should be called "Aloadae," because they sacrifice to
Ares after they have bound him (taTE& xT 6floaL arz6v, Philostratus, VS 2.547). It
is unclear whether this is a purely literary reference to a more detailed version of
the myth referred to in Iliad 5 (discussed above), or whether Herodes is alluding
to a traditional ritual performed in historical times by priests drawn from the clan
of the Aloadae, a ritual that may still have been performed in Herodes' time. The
anecdote does, however, attest to some kind of regular sacrifice to the bound god,
since any mention of sacrifices in a literary source (i.e., some local Boeotian epic
or hymn) was probably part of an aetiological myth explaining the particulars of a
local cult of the bound Ares. The mythical account of the Aloadae quoted above
from Iliad 5 limits the duration of Ares' incarceration to thirteen months, a figure
that has prompted the undoubtedly correct inference that the myth is to be
connected with an annual ceremony in which the god was released and then
rebound each year.20 Thus it appears that in some areas of Greece, at least, bound
statues were erected at specific points of crisis to ward off a particular military
threat, and that despite such hostile treatment they could be offered sacrifice and
treated with deference like any other cult statue.
BOUND STATUES IN EGYPT AND THE NEAR EAST: SOME COMPARATIVE

DATA
In Egypt "voodoo dolls" were deployed at the highest levels of stat

from the earliest times.21 In the daily rituals performed to ensure th
19. The antiquity of the Spartan cult seems to be assured by Pausanias's descrip
bound Enyalius as an ayakXja &Xadtov; there was also a temple of Enyalius at Geron
3.22.6) and on the road between Sparta and Therapnai a ieQoV a&xaLoTaTov of Ares Th
3.19.7). The image there was allegedly imported from Colchis, a mistake (according
Lakonische Kulte [Leipzig, 1893] 149-50) for Euboean "Calchis." Farnell, (supra, n. 6: 40
G. Frazer (Pausanias: Description of Greece [Cambridge, 1898] ad loc.) argue that do
Ares, a practice usually associated with Thracian and other non-Greek forms of a
purificatory ritual, is another indication that the Spartan worship of Ares is borrowed fro
20. In connection with the myth about the binding of Ares by the Aloadae, W. Bur
Religion, trans. J. Raffan (Cambridge, Mass., 1985), 169, suspects some similar annual
licence, a plausible interpretation with some nice Roman parallels, especially the Satur
21. The best general discussions of the Egyptian "voodoo dolls" are M. J. Rav
Egyptian Magic and Symbolism," OMRL 64 (1983) 7-47 (on the wax effigies); G. P
Egypt, the Egyptian gods were aided in their struggles with their enemi
creation and manipulation of wax images depicting their rivals. Egyptian
papyri dating for the most part to the earliest years of the Ptolemaic
provide us with the most detailed information about these effigies, whi
bound, abused, and then destroyed in a variety of ways. Such ritua
temple of Amon-Re in Karnak included the manufacture of red wax figu
the god's enemy Apophis, which were bound back to back with an effig
of the pharaoh's human enemies, mistreated, and then consumed by fi
similar fate was regularly reenacted in the temple of Osiris at Abydus fo
wooden effigies of Seth, whose name was inscribed on their chests; bo
and foot with the sinew of a red cow, these figurines were spat upon, tr
cut up, and then burnt.23 Effigies of Seth and other anonymous "foes of
were likewise stabbed and then burned on a brazier in a number of dif
ceremonies performed in the temple complex at Edfou.24 In each case the
Egypt is assimilated to Re, Osiris, or Horus, and his enemies (both fore
domestic) are correspondingly represented either explicitly or implicitly
bound figurines. Although most of the testimonia for these rituals is
Egyptian dynastic standards, there is much evidence that the Ptolemai
that preserve the instructions for these rites stand at the end of a long
tradition of manuscripts that were heavily guarded state secrets until t
periods of foreign domination.25 The pose that these wax figures are f
assume prior to their destruction is another indication of the antiquity of th
depicted kneeling and bound hand and foot-a statue type usually referr
the "bound captive" (Fig. 1)-they are identical to the kneeling an
figures (marked as foreigners by inscriptions, distinctive national dress,
features) that were regularly depicted in monumental reliefs cowering at the
of the pharaoh, who threatens them with a mace or an axe. They a
quently found in the form of free-standing statues and appear in subs
positions on royal furniture and even on the soles of royal sandals. In ea
the symbolism is twofold: these figures are meant to commemorate past
empreintes magiques de Gizeh et les Morts Dangereux," MDAI(Kairo) 16 (1958) 252-7

"Achtungstexte," Lexikon der Agyptologie 1 (1975) 67-69 (on the clay and stone dol
not been able to lay my hands on his most recent publication, Cinq figurines d'envo
Bibliotheque d'Etude 101 (Cairo, 1987), for which see the short review in BO 46 (198
See also R. Ritner, "The Mechanics of Ancient Egyptian Magical Practice" (diss., Chic
92-183.
22. R. O. Faulkner, "The Bremmer-Rhind Papyrus: The Book of Overthrowing Apep," JEA
23 (1937) 166-86; 24 (1938) 41-54. The papyrus appears to be a copy of a text that probably dates as
early as the Middle Kingdom; it seems to have been copied by a temple priest for his own personal
use. See Ritner (supra, n. 21) 210 n. 735.
23. The ritual is described in the "Book of Overthrowing Seth," which was probably composed
in the New Kingdom. See Raven (supra, n. 21: 10) for a full bibliography.
24. Ritner (supra, n. 21: 205-7) briefly describes the rites of Soker-Osiris, the "Play of Horus,"
and other ceremonies at Edfou that contain similar execration rituals, as well as rites in which fish are
labeled with the names of state enemies and then trampled and burned.
25. Ritner (supra, n. 21) 203 n. 713
of the pharaoh over his foreign enemies and at the same time to encourage
(through sympathetic magic) future victories.26
Smaller clay, mud, and stone versions of the free-standing "bound captive"
figure seem to have evolved artistically into a flattened, nearly abstract shape
that provided a wide and level writing surface for long lists of intended victims.
Unlike the monumental versions, these smaller effigies were not created for
public view; nearly all the extant examples were found buried in large caches
(frequently in sealed clay pots) in cemeteries or near mortuary temples, and, in
contrast to the wax dolls discussed above, it is often unclear whether they have
been intentionally abused or accidentally broken over the passage of time.27
They can roughly be divided into two categories: small figurines that are labeled
with the name of a single individual, and statuettes two to three times larger
carrying much longer texts that curse entire nations, their princes and soldiers,
hostile gods or demons, as well as Egyptian individuals who were probably
inimical to the regime. The earliest examples of both types are from four closely
related deposits in the necropolis at Giza, which date to the Sixth Dynasty (ca.
2300 B.C.).28 It is to the Middle Kingdom (2065-1785 B.C.), however, that the
bulk of the evidence belongs. The larger statuettes are particularly prominent
here,29 and bear texts that are nearly identical to those inscribed on simple
earthenware pots and bowls used in other execration ceremonies.30 The necropo
lis at Saqqara yielded many examples of smaller dolls, nearly all of which are
26. Ritner (supra, n. 21: 94-97 and passim) gives the most complete and up-to-date discussion,
repeatedly emphasizing the fact that such images did not simply reflect the past triumphs of the
pharaoh, but were magically designed to encourage further victories over his enemies.
27. See Posener's survey article (supra, n. 21) for a complete discussion. Ritner (supra, n. 21:
111 n. 376) supplies a full bibliography subsequent to 1972. G. Jequier ("Statue de prisonnier," in J.
Capart, ed., Documents pour servir a l'etude de a'drt egyptien 1 [Paris, 1931] 21-23) and J. P. Lauer
("Les statues de prisonniers du complexe funeraire de Pepi Ier," BIFAO 51 [1969-70] 37-45) discuss
the large (? life-size) free-standing statues of bound prisoners discovered in the mortuary temples of
Pepi I and II.
28. H. Junker, Giza VIII, DenkschrWien 73.1 (Vienna, 1950) 30-38, with pls. 6b, 7; Posener,
MDAI(Kairo) (supra, n. 21). All these dolls were discovered together sealed in a single large vase.
A. Abu Bakr and J. Osing ("Achtungstexte aus dem alten Reich," MDAI[Kairo] 29 [1973] 97-133)
and J. Osing, "Achtungstexte aus dem alten Reich II," ibid. 32 [1976] 133-85) publish three other
deposits of the same date and provenance, two of which were sealed in clay jugs.
29. G. Posener, "Nouveaux textes hieratiques de proscription," in Melanges Syriens offerts a
M. Rene Dussaud 1 (Paris, 1939) 313-17, publishes five large alabaster figurines (Cairo J 63955-59)
probably from the necropolis at Saqqara, which date to the first half of the 12th Dynasty. He also
discusses (in Princes et pays d'Asie et de Nubie: Textes hieratiques sur des figurines d'envoutement du
Moyen Empire [Brussels, 1940] 15-20) another, larger find from the same necropolis of slightly later
date, which includes both the large (31-34 cm) and smaller types (6-7 cm). The larger figurines have
texts very similar to those mentioned above and holes drilled under the arms to allow the attachment
of threads for binding. The texts are all written in red ink. Three large limestone dolls (14.5 cm) and
several hundred mud figurines, all dating to the 12th Dynasty and inscribed with very similar texts,
were discovered near a border fortress at Mirgissa in the Sudan. See A. Vila, "Un d6p6t de textes
d'envoutement an Moyen Empire," JS (1963) 135-60; and especially Ritner, (supra, n. 21) 134-70.
30. G. Posener, "Les textes d'envoutement de Mirgissa," Syria 43 (1966) 277-87.
inscribed with the names of individual Egyptians who are described

that is, in English parlance, "the late so-and-so."31 Two other cac
smaller type of doll come from the later periods and also bear the n
individual "dead" Egyptians.32 In every case, this peculiar designatio
refer to the potentially dangerous ghosts of recently deceased individ
still hovered about the place of burial.33
As discussed above, it is important to note that from the Old
onward the "bound captive" was a standard motif in the large stone r
lined the walls of public buildings. The ideological reasons that und
public monuments are clear: the Egyptian reliefs showing the king b
weapons over his enemies (portrayed in the classic "bound captive" p
at once an advertisement of past good fortune and an assurance for
good fortune. The pairing of a triumphant figure and a bound, h
recalls the statues of Ares and Dike described above in the Syedra ins
similar pairing occurs on smaller, apparently private Egyptian talism
the Egyptian and Nubian border numerous clay disks and one crudely
stele have been found that depict in relief a bound Nubian captive bei
single armed Egyptian soldier (Fig. 2).34
There are, then, three well-developed types of Egyptian rituals t
the protection of the king and country: (1) the burning of inscribed w
of hostile gods, demons, ghosts, and mortal enemies; (2) the shatterin
crockery inscribed with the names of a similar range of superhuman
enemies; and (3) the binding and burial (usually in graves or mortuary
of clay and stone figurines inscribed with the same range of names. T
ship between public and private binding rituals has occasioned some
31. Posener, Princes etpays (supra, n. 29) 15-20.

32. Nearly 100 small clay statuettes (4.5-8.2 cm.) were found at Lischt and date t
Intermediate Period (1785-1580 B.c.); they are inscribed in red ink, and some were f
in a small clay coffin. See "The Egyptian Expedition 1932-33," BMMA Suppl. (Nov
with fig. 32. Thirty-nine terra-cotta plaques (mass-produced by pressing clay into a m
relief the same "bound captive," except that he is shown in profile-just as in the mon
reliefs; see Posener, MDAI(Kairo) (supra, n. 21) 252-70, pls. xxvi, xxvii.
33. Posener, MDAI(Kairo) (supra, n. 21) 265-68; Abu Bakr and Osing (supra, n.
A section of the Bremmer-Rhind papyrus (supra, n. 22) refers to effigies made of "e
living and dead." For the power of ghosts in Egyptian popular belief, see A. Gardiner
Egyptian Letters to the Dead (London, 1928) 1-12, who point out that even the m
include "every dead man and every dead woman" among the possible causes of di
(supra, n. 21: 116-17) argues that the designation of these victims as mwt, "dead," cou
an optative or proleptic formulation wishing the early demise of the person so nam
however, the similar designation of a hostile ghost simply as "a dead person" in Assy
banning texts (infra, n. 63).
34. T. Sove-Soderbergh, Agypten und Nubien (Lund, 1941) 132-35, suggests th
commemorative of some past victory. Posener, MDAI(Kairo) (see n. 21) 254, is caut
not see any close connection with the "voodoo dolls." The ten disks found at Arieka, h
buried in the outer defensive wall of a fort, i.e., a good place for a talisman but a rather
for commemorative medallions of absolutely no material worth (i.e., they are made o
gold or any other material worth hoarding).
One rather tempting explanation is the "trickle-down effect": Junker, struck by

the fact that the smaller clay figurines all seem to come from the vicinity of
nonroyal tombs, proposed that they represented a democratization of the royal
ritual; just as the royal funerary rituals were gradually co-opted by the king's
nobles and courtiers, so too the protective rituals for saving all of Egypt were
gradually taken over by private individuals on a much cheaper scale. Similar
parallels between royal and nonroyal rituals in Assyrian magic suggest, however,
that such public ceremonies could just as well be imitating much older private
rites designed for the same purpose.35
With perhaps one exception,36 clearly identifiable "voodoo dolls" have not
survived intact in other parts of the Near East prior to the Hellenistic period, but
this is true for many kinds of artifacts; the dry, stable climate of the Nile valley is
well known as a unique preservative. From the surviving literary evidence, how
ever, one is able to piece together a similar pattern of a widespread use of such
effigies at both the personal and civic levels. A Babylonian cuneiform tablet, for
example, prescribes a series of sacrifices that are to be performed before setting
out for battle; at its conclusion, an image of the foe is fashioned from tallow, and
its face is bent backwards with a cord.37 Another ritual text also refers to a
military context:38
[You fashion] seven figurines of asphalt, seven of sheep tallow, and

seven of wax. Twist their arms behind their backs . . . and then the army
of the king will triumph over the army of the enemy.
These simple recipes recall the bound silver statues of Gothic tribesmen de
scribed above by Olympiodorus; in both cases images of human soldiers are
bound or manipulated in hopes of impeding their performance in future battles.
Elsewhere there is much evidence for more private ceremonies, in which
images are used in "medical" exorcisms to cure the sick, by binding, destroying,
or sending away the demon, ghost, or human sorcerer who has attacked the
patient. The Assyrian incantation series Maqlu (lit. "Burning [Rituals]") offers
35. Junker (supra, n. 28) 32-33. A similar relationship was once suggested between the
monumental figures of animals and demons that stood beside the doors of Assyrian palaces and the
very similar but cheaply manufactured clay figures that originally were only found in the founda
tions of private houses in Assur. Subsequent discoveries, however, prove that both kinds of image
were used simultaneously to protect royal buildings. See C. A. Faraone, "Hephaestus the Magician
and the Near Eastern Parallels for Alcinous' Dogs (Od. 7.91-94)," GRBS 28 (1987) 266-74, esp.
271-72. For the similar evolution of the royal Maql2 rituals from traditional private rites, see infra,
n. 40.
36. A sixth-century-B.c. bronze statuette from Syria and now in Copenhagen is depicted in the
same peculiar stance prescribed in a Neo-Assyrian ritual text, infra, n. 89.
37. H. Zimmern, Beitrage zur Kenntnis der babylonischen Religion (Leipzig, 1901) no. 173; see
M. Elat, "Mesopotamische Kriegsritualen" BO 39 (1982) 22-24, for discussion.
38. Elat (supra, n. 37) 5-25.
the most abundant testimony for the rite of burning effigies of demons
and living enemies as well, who are usually described as hostile sorcerer
superhuman targets and the destruction by fire recall the quotidian E
rituals discussed above. The rituals and incantations in the Maqlu series
have been borrowed from earlier private rites (which were originally
formed in times of medical crisis) and elevated in the Neo-Assyrian peri
status of a national, calendrical ritual in which the private enemies of t
were regularly destroyed in effigy.40 In the Maqlu series, as in Egypt, th
of potential evil is not limited to hostile deities or demons; at 2.146-51
figurines of anonymous human magicians are melted down as well. Ef
alleged human sorcerers and sorceresses could also be defiled and ph
abused (Maqlu 2.172-80). In the case of these unidentified magicians we s
the effigies cannot be literal portraits; the regular practice of binding and d
ing figures of both sexes was a precautionary measure, since the pract
usually did not know whether it was a man or a woman who was attac
patient.
In some healing rites the effigy was not destroyed, but simply bound or
otherwise restrained. A ritual against witchcraft from another cuneiform series
called Namburbi provides a very detailed treatment of the binding and burial of
images of hostile witches and warlocks:41
31 That sorcery, witchcraft, magic, [and] evil spells, whether of a man

or of a woman, may not approach a man:
32 You make two images of clay, two images of dough, two images of
wax, two images of tallow, [each pair consisting] of male and fe
male, and
33 you write their names on the left hips, you turn their hands behind
them.
34 You bind their feet.
(trans. R. Caplice)
39. G. Meier, Die assyrische Beschworungssammlung Maqlu, Archiv fur Orientforschung 2

(Berlin, 1937). For the incineration of effigies in a ritual not associated with the Maqh2 series, see E.
Ebeling, Tod und Leben nach der Vorstellung der Babylonier (Berlin, 1931) 118-19, no. 28, 11. 5-7.
40. These rituals were not, however, used against foreign enemies like the Egyptian rites; see
T. Abusch, "Mesopotamian Anti-Witchcraft Literature; Texts and Studies, Part 1: The Nature of
Maqld, Its Character, Divisions and Calendrical Setting," JNES 33 (1974) 251-62; idem, "The
Demonic Image of the Witch in Standard Babylonian Literature: The Reworking of Popular Concep
tions by Learned Exorcists," in Religion, Science, and Magic: In Concert and Conflict (Oxford, 1989)
27-60; idem, Babylonian Witchcraft Literature: Case Studies, Brown Judaic Studies 132 (Atlanta,
1987) 13-41.
41. R. I. Caplice, "Namburbi Texts in the British Museum IV," Orientalia 39 (1970) 134-41 no.
40, 11. 31-34 (translation quoted here) and rev. 1-4 (translation quoted below). Namburbi spells are
designed to avert evil omens; for a good introduction and overview of the ritual, see R. I. Caplice,
The Akkadian Namburbi Texts: An Introduction, Sources from the Ancient Near East 1.1 (Los
Angeles, 1974).
After a short prayer, which states that "these are the images of my adversary, my
prosecutor," the end of the ritual follows:
rev. 1 ... and you take the combings of their heads. You clothe them
in the combings.
2 You bind them [i.e., the statues] together in a skein. You put
them in a half-sila container ...
4 ... you bury them in the ground.
(trans. R. Caplice)
In this spell, the magical action is directed toward the victims by inscribing a name
and by placing strands of hair from the victim on the doll.42 The binding action is
multiple: the arms are twisted behind the figurine's back; the feet are bound; the
two dolls are tied together in a coil of string and then buried in a container.
Another type of very similar Assyrian ritual aims at the banishment of
"Appearing Ghosts."43 Aside from the presentation of a variety of food and
drink offerings, several texts prescribe the fashioning of images of ghosts. Of
particular interest are these instructions from Nineveh; after a number of prayers
and incantations, we find at the end of the tablet the following ritual:44
13 If a [dead] ghost appears to a living person [ . . . a figurine] of clay

you shall make, its name on its left hip write,
14 place it into a gazelle horn, [cover?] its face, [at sunset place it
either in the] shade of a caper tree or in the shade of a thorn bush.
15 You shall dig a hole and bury it.
(trans. G. Castellino)
As in the case of the Egyptian ritual discussed above, a figurine is inscribed with
the name of the (dead) person, placed in a container, and then buried in some un
cultivated land. Another recipe prescribes the distortion and abuse of the effigy:45
. . .knead clay, make a figurine

7 [of the ghost]; its name on the left hip you shall write, its feet twist
[ . . .] throw it; a dog's tooth
8 [as a] peg in its mouth fix;
(trans. G. Castellino)
42. For the use of human hair in Greek binding or erotic magic, see the summary of the
evidence in D. R. Jordan, "Defixiones from a Well near the Southwest Corner of the Athenian
Agora," Hesperia 54 (1985) 251-55.
43. For general discussion, see G. Castellino, "Rituals and Prayers against 'Appearing
Ghosts,' " Orientalia 22 (1953) 240-74; J. Bottero, "Les morts et 1'au-dela dans les rituels en
accadien contre I'action des 'revenants,' " Zeitschrift fir Assyriologie 73 (1973) 153-203, esp. 179
96; and J. Scurlock, MMG pp. 125-36.
44. K 2175 rev. 13-14 = MMG 20. See R. C. Thompson, "An Assyrian Incantation for Laying
a Ghost," Proceedings of the Society of Biblical Archaeology (1906) 219-27; Castellino (supra, n. 43:
265-69) for translation and discussion.
45. KAR 234.6-8 = MMG 22. See Castellino (supra, n. 43) for discussion.
In contrast to the treatment of images of demons or sorcerers in the

ceremonies, the figurines of ghosts were rarely destroyed; indeed, the
rites (which could last as long as three days) usually involved typical f
libations and offerings, and seem to represent a "second burial" of sort
at permanently removing the ghost from the world of the living.46 In
these images of ghosts were made of clay, inscribed with a name on the left
dressed, and then given both funeral offerings and further provisions f
journey back to the land of the dead.47 After the ritual was over, the figur
be bound, have its feet twisted, its mouth blocked with a peg, or its anus an
covered; such coercive measures (mild in comparison to those foun
Maqlu series) were usually reserved for ghosts of strangers not family m
and were accompanied by threats or oaths forced from the ghosts, wh
made to swear that they would never return.48 Finally the image was inc
in a jar, a traditional form of coffin in the Neo-Assyrian period, and b
sunset on the steppe or some similarly desolate area.49
It is clear that the destruction or burial of effigies was a common f
ritual used by the Egyptians and the Assyrians in both public and privat
There are hints, as well, that such rites existed in the less documented c
of the Levant and Anatolia. The Hittites employed clay, wax, tallow, d
and wooden effigies to restrain or injure their adversaries, both pu
46. Of the 88 Assyrian ghost-banning recipes collected by Scurlock, only 17 (MMG

23, 56-58, 61, 62, 65-68, 70, 71) explicitly prescribe the manipulation of figurines; they
the Neo-Assyrian period, with the exception of one text discovered in a third-century-B.c.
Uruk. Scurlock, MMG pp. 49-64, provides a thorough discussion of the use of effigie
where (pp. 29-30 with n. 127) notes the unusual length of these rites, which frequently last
days (e.g., MMG 22, 23, 55-58).
47. Ingredients: MMG 15, 20, 22, 58, 65, 66, 71 (clay); 15, 67 (wax); 56 (dirt, ox
(horse urine, flour); 61 (clay mixed with tallow and wax); 71 (excrement, straw, donkey
recipes do not mention the medium: 18, 21, 23, 62, 68, 70. Inscriptions on left side of doll: M
21, 57, 61, 66, 67. Dress: MMG 21 (white or red garment); 56 (lionskin); 62, 70 (gird
("makeshift" garment); 71 (red garment); 80 ("clean clothing"). Offerings given to figuri
21 (libation); 56 (nine dishes of barley gruel and libations); 58 (libation of hot broth, food of
65 ("food portion"); 67 ("You make funerary offerings for them. You pour out hot brot
beer"); 70 (beer, wine). Provisions given to figurine for return trip to underworld: MMG
sions); 56, 62 (filled waterskin, malt flour); 58 (image placed on a sailboat with travel pro
(groats, malt, beerwort, dried bread); 71 (travel provisions, copper axe).
48. Coercive measures: MMG 15 (eyes covered with bark); 22 (dog tooth inserted in
feet twisted); 57 (peg of cornelwood placed in mouth; "put a copper chain on it"); 62 (ha
behind back); 65 (bound to tamarisk). Threats: MMG 57. Forced oaths: MMG 18, 56, 57, 6
In some cases the binding seems designed to force an oath from the effigy (e.g., MMG
elsewhere it may simply be a way of inhibiting the ghost's ability to return and haunt the p
Scurlock, MMG pp. 58-59, for discussion.
49. Containers: MMG 18, 56, 61 (jar); 57, 71 (copper cup); 20 (gazelle horn). Burial:
(family tomb); 18 (locale unspecified); 20, 23 (in the shade of a camelthorn or acacia b
hole "in the west"); 56 ("abandoned waste"); 57 (steppe); 66 (canal bank). Scurlock, MM
283, points out the similarities between the ghost-banning rites and burial customs. For a pa
Ritner (supra, n. 21: 163-65), who discusses the assimilation of funerary ritual to Egypt
tion rites.
private.50 In the purification rite of the old woman Tunnawi, for example, the
practitioner abuses effigies of a pair of hostile sorcerers fashioned from wax,
flattens (or perhaps melts) them, and says: "Whatever wicked persons are mak
ing him/her unclean, let them be flattened [or perhaps "melted"] in this way."51
The Hittite "Soldier's Oath" is accompanied by a similar execration: "Just as
this wax melts, and just as this mutton fat dissolves, whoever breaks these
oaths ... let him melt like wax, let him dissolve like [mutton fat]."52 Parallels
are preserved in Aramaic inscriptions dating to the eighth and seventh centuries
B.C., for example the treaty of Matti'el, a minor king who ruled in the neighbor
hood of modern Aleppo: "As this man of wax is blinded, thus Matti'el shall be
blinded [sc. if he violates this treaty]."53
CYRENEAN KOAOXZOI AND EARLY GREEK GHOST-BANNING RITUALS
Two fourth-century-B.c. inscriptions from Cyrene attest to a similar use of

effigies (called xoXooool) in "sympathetically" magical ceremonies. Both texts
appear to contain material copied or paraphrased from much earlier documents,
a fact that allows us with some confidence to date the rituals contained therein to
the Archaic period in one case, and at least as early as the Classical period in the
second. The first inscription is the well-known description of the oath of the
seventh-century Theran colonists who founded the city of Cyrene in Libya; while
wax images (xokooooi) are burned, the participants say:54 "May he who does not
abide by this agreement but transgresses it melt away and dissolve like the
50. A. Goetze and E. H. Sturtevant, The Hittite Ritual of Tunnawi (New Haven, 1938) 72-75;
D. H. Engelhard, "Hittite Magical Practices: An Analysis" (diss., Brandeis, 1970) 172-78.
51. See Goetze and Sturtevant (supra, n. 50: 78) for this translation. The key verb shallanu, "to
make flat," is also used in another recipe (KBo VI 34 I 41-42) to describe the effect on wax and
tallow effigies placed in a cooking pot, prompting J. Friedrich, (infra, n. 52: 162-63) to suggest "she
melts them and says: '. . let them melt' " as a better translation. Either procedure can be paralleled
in Assyrian recipes.
52. KBo VI 34.40-rev. 5. For translation and commentary, see J. Friedrich, "Der hethitische
Soldateneid," Zeitschrift fur Assyriologie 35 (1924) 161-92; A. Goetze, ANET p. 353; N. Oettinger,
Die militarischen Eide der Hethiter, Studien zu den Bogazkoy-Texten 22 (Wiesbaden, 1976) 6-17,
72-76.
53. F. Rosenthal, ANET pp. 659-60. For dating, detailed discussion, and bibliography, see
J. A. Fitzmyer (The Aramaic Inscriptions of Sefire, Biblica et Orientalia 19 [Rome, 1967] 52-58) and
J. C. L. Gibson (The Textbook of Syrian-Semitic Inscriptions II: Aramaic Inscriptions [Oxford 1975]
18-43), who discuss similar stipulations in two Akkadian treaties.
54. For text, bibliography, and the current agreement on the archaic dating of the oath, see R.
Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions (Oxford, 1969) 5-9, no. 5, 11. 44
49. The translation is that of A. J. Graham, Colony and Mother City in Ancient Greece (Manchester,
1964) 226. C. Picard, "Le rite magique des ib6oXa de cire brfils, attested sur trois steles arameennes
de Sfir6," RA (1961) 85-88, argues that the Sefire inscription suggests a Near Eastern source for
Greek magic involving wax figurines, but he strangely does not mention the Cyrenean oath, which he
discusses at length in an earlier article (infra, n. 70). For a detailed analysis of the magical rituals
involved in such oaths, see C. A. Faraone, "Molten Wax, Spilt Wine and Mutilated Animals:
Sympathetic Curses in Near Eastern and Early Greek Oaths," JHS 113 (1993) forthcoming.
images, himself, his seed, and his property" (trans. A. J. Graham).

Cyrenean inscription that mentions xoXoooo is a much-discussed lex
consists of a series of purification rituals allegedly given to the city over
by the Delphian Apollo,56 including one for the removal of an evil
uncertain identity from a private house. Although the instructions f
have been roughly treated and abridged by the fourth-century reda
outline of the ceremony is well-enough preserved to reveal some star
larities with the Assyrian prescriptions for "laying" a ghost:57
Visitant sent [by spells] from afar.
If a visitant is sent against the house [and] if he [sc. the househo

knows from whom it attacks, he shall name him by proclamatio
three days.
If he [sc. the visitant, or perhaps the sender] is dead and buried in the
earth or has in some other manner perished,58 if he [sc. the householder]
knows his name, he shall make a proclamation by name [sc. for three
days?].
But if [in either case?] he does not know his name [he shall address him]:
"O avOQwejrog, whether you are a man or a woman," and having made
male and female xooooooi either from earth or from wood he shall
55. SEG 9.72.111-21; LSS no. 115; C. D. Buck, Greek Dialects (Chicago, 1955) no. 115; F.
Solmsen and E. Fraenkel, Inscriptiones Graecae ad Inlustrandas Dialectos Selectae 4 [Stuttgart, 1966]
no. 39; cf. SEG 20.717. For full commentaries, see S. Ferri, Notiziario Archaeologico del Ministero
delle Colonie 4 (1927) 93-144; G. Luzzatto, La Lex Cathartica di Cirene (Milan, 1936); H. J. Stukey,
"The Cyrenaean Hikesioi," CPh 32 (1937) 32-43; J. Servais, "Les suppliants dans la loi de Cyrene,"
BCH 84 (1960) 112-47; and most recently (with English translation) R. Parker, "The Cyrene Cathar
tic Law," in Miasma: Pollution and Purification in Early Greek Religion (Oxford, 1983) 332-51. The
translation closely follows Parker except where indicated.
56. The opening words of the inscription ("Apollo decreed. ..") have been taken by all
commentators to refer to the Delphian god because of the close connection between Cyrene and the
oracle (cf. Hdt. 4.150-58, 161.1).
57. W. Burkert, Die orientalisierende Epoche in der griechischen Religion und Literature,
SHAW (Heidelberg, 1984) no. 2, pp. 68-71, was the first to point out the parallels. The recent
appearance of MMG with its translation and discussion of numerous parallels has encouraged my
detailed discussion here, which corroborates and greatly expands Burkert's treatment.
58. Parker (supra, n. 55) prefers Fraenkel's minor emendation of the line and translates: "If
(the sender of the suppliant) has died in the land or perished anywhere else." In either case, there are
good parallels for this sort of all-inclusive (nearly legalistic) language designed to cover every contin
gency. For similar sets of contrasting pairs, see the description of ghosts in the Neo-Assyrian spells:
"whether you be a [sc. man's] buried ghost or a [man's] unburied ghost" (MMG 11); "whether it be
the ghost of my kith and kin, or the ghost of one who was killed in a military defeat, or a roving
ghost-this is he; this [i.e., effigy] is a representation of him" (MMG 62); or "because of the ghost of
my relative, which was set on me, or a strange ghost or a robber or a murderer [which] day and night
is bound after me and continually pursues me" (MMG 67). The contrast in all of these examples
seems to be between properly buried ghosts (with whom the patient is familiar) and strange ghosts
who roam the earth on account of the violent circumstances of their demise (in military defeat or
execution as a criminal) or because of some defect in their burial.
entertain them and set beside them a portion of everything. When you59
have done the customary things, take the xokooooL and [their] portions
and deposit them in an unworked glen.
This ritual is designed to prevent the attack of a LXEoLOg eraxto;g, a hostile

visitant, probably a ghost or a demon.60 The redaction of the recipe is unfortu
nately compressed to the point of obscurity. Three situations are envisaged,
which correspond to the divisions in the translation above. The first scenario
deals with a visitant who has been sent by a living enemy (compare the ghosts
sent by Assyrian sorcerers) or an angry corpse,61 while the second directly in
volves a dead man, probably the uneasy ghost itself operating of its own will
from the grave.62 The ritual in both cases is simple if you know the name of the
living magician who sent it or of the dead person whose spirit it is: simply invoke
the name for three days.63
In the third stipulation, we are given instructions about what to do if the
identity of the sender or the ghost is unknown.64 This section of the inscription is
most relevant to our discussion, as it reports the construction of a pair (male and
59. The verb changes here without warning or any apparent reason from the third person to
second. See infra, n. 116, for one possible explanation.
60. Burkert, Epoche (supra, n. 57) 68-71, and Parker (supra, n. 55) ad loc. support the earlier
interpretations of P. Maas, Epidaurische Hymnen (Halle, 1933) 139, and Stukey (supra, n. 55), who
were the first to argue that these stipulations concerned the banning of a 6ailcv or a spirit sent to do
harm. The seemingly contradictory use of the same word, ixoatog, to mean both daemonic attacker
and suppliant can be paralleled by the puzzling semantic range of ntakCavacog or &adoXaoT, which
can mean a murderer, the bai!ov sent by a murder victim against the murderer, and also a suppliant
polluted with blood (Parker [supra, n. 55: 108 n. 13]). This same usage of ixEotog is confirmed by
Hesychius (infra, n. 61).
61. Cf. Plutarch's description (infra, n. 76) of the Thessalian y&qoeg, who like Assyrian sorcer
ers have the power to send et6oXa against someone or send them away. For the use of ghosts in
Greek defixio rituals, see Jordan, SGD pp. 152-53. Stukey (supra, n. 55) adduces Xen. Cyr. 8.7.18
and Clytemnestra in the Eumenides as examples of murder victims sending out (the verb is
EJLCtEzYjTELv) C6auoveg from the grave. See also Hesychius s.v. &vtaict, a word glossed ixEoLog and
6aiowv; he goes on to explain, "and they call Hekate "Antaia" because she sends out [EjLZ?tEPeuLV]
these things."
62. For evidence that the angry ghosts themselves came out to do the haunting, see, e.g., the
bizarre ritual of [taoaXl.tou6g (Soph. El. 445 and Jebb's note and appendix ad loc.), or the piercing
of Oedipus's feet (infra, n. 103), which seem to be preemptive rituals designed to cripple a ghost in its
efforts to gain revenge. See infra my discussion of the angry ghosts of the Spartan regent Pausanias
and of Actaeon, both of which necessitated exorcism.
63. The Neo-Assyrian spells betray a similar emphasis on knowledge of the ghost's identity,
e.g., "If you know it, write his [sc. the ghost's] name" (MMG 17); "If a man continuously sees [dead
person]s, (either) one who is kn[ow]n to him, or not known to him, to keep them away, you make
figurines of the dead persons" (MMG 23); and "the dead persons whom I know (and) many whom I
do not know" (MMG 26). For the three-day ceremony, compare the Assyrian rites (supra, n. 46).
64. I follow Stukey's suggestion (supra, n. 55) that the xoXoooot were to be used in both
situations, taking my cue from the Assyrian spells that use male-female pairs of effigies to attack
anonymous living sorcerers (e.g., in the Maqlu series) as well as to ban anonymous ghosts (MMG
87.27-30: "You are the ghost of nobody ... whose name nobody knows .... Whether you be a
man who is like a (living) man or whether you be a woman who is like a (living) person"; cf. MMG
23, 58, 67, 69 for descriptions of pairs of male and female dolls).
female) of wood or mud statuettes (xoXo0aoi) ;65 after providing a ritual

them, the householder is to deposit them in some uncultivated spot.66 T
larities with the Assyrian materials quoted above are obvious; a pair of
male and female, are fashioned out of cheap materials and equated
anonymous attacker. After offerings are made to the statues, they ar
away and deposited in the wilds together with their "portions," that is, t
offerings. In the Assyrian ghost-banning texts, libations and other funerary
ings of food are also frequently presented to effigies, as well as provisi
their return journey, which are in some cases taken along when the eff
carried off and buried in the steppe or some other abandoned area like t
&aeyo6 mentioned in the Cyrene decree.67
The parallels with the Assyrian ghost-banning rites are intriguing, b
are not quite complete. The Cyrenean xokooooL are neither bound nor
ated, and it is unclear whether they are to be "deposited" in a hole (i.e.,
or simply set up in some uncultivated spot.68 There are, moreover, so
similarities between the ritual prescribed in the inscription and the ver
Greek practice of erecting cenotaphs for people who had died in far-off
or for those whose bodies were irrecoverable because of some disaster.
earliest known cenotaph is the Mycenaean chamber tomb discover
Midea, which contained two stone effigies (usually called menhirs),
dance of grave goods, and not a single indication of human remains.70 A
65. E. Benveniste, "Le sens du mot xokooo6g" RPh 6 (1932) 118-35, and Fraenkel's
tary on Aeschylus, Agamemnon 416, for a correction of the chronic misperception that
xoXooo6s always refers to gigantic statues. For refinements of Benveniste's approach and
to his assertion that xokooo6g originally meant "double" or "ritual replacement," see
"Qu'est-ce qu'un xokooo6s?" REA 62 (1960) 5-40; J. Ducat, "Fonctions de la statue dan
archaique: Kouros et kolossos," BCH 100 (1978) 239-51.
66. S. Ferri, "Note d'epigrafia cirenaica," Historia 3 (1929) 399, argues that if we
Wilamowitz's restoration Qe[it]oaT (as most recent editors do), it must mean "to depo
ground]," i.e., "to bury". Parker and most editors translate it simply as "to set up." G. D
"Le decretali di Cirene" RFIC 5 (1927) 205-6, originally restored the aorist infinitive as @
"to hang," and suggested some oscillation ritual connected with Minoan tree cult. In
Eastern voodoo-doll rituals quoted above, the dolls are always buried in the necropolis (i
supra, nn. 27-33) or some out-of-the-way place like the steppe (in Assyria; supra, n. 49).
67. Supra, nn. 46-49, for the funerary offerings given to the Neo-Assyrian effigies
indications of a "second burial."
68. Supra, n. 66.
69. E. Rohde, Psyche3, trans. W. B. Hillis (London, 1925), 42-43, 57 nn. 87-88; J. B
and D. Kurtz, Greek Burial Customs (Ithaca, 1971) 247-59.
70. The menhirs, carved from poros limestone, are different sizes, the larger being a
m tall and 0.64 m wide, while the smaller is half as tall and measures 0.52 m wide. A. W
"The Cenotaph of Midea," in The Royal Tombs at Dendra, Skrifter Utgivna av Kungli
istiska Vetenskapssamfundet i Lund 15 (Lund, 1931) 108-17, discusses the find in det
attention to the Homeric references to cenotaphs (Od. 1.289-92, 4.584). C. Picard, "Le
de Mid6a et les 'colosses' de Menelas (ad Aeschyl. Agamemn. v. 414 sqq.)" RPh 7 (1933
first argued for the approach taken here, pointing out that the ceremony involving xokooo
Cyrenean lex sacra, the description of the xokooooi in Aeschylus, and the menhirs in
cenotaph were all part of the same "Doric" (I would say "Peloponnesian") tradition o
of seventh-century-B.c. date discovered on the island of Thera (the so-called

Schiff's Grave) contained a large number of traditional grave goods, but there
was no trace of inhumation or cremation; here too, however, excavators found
two rather crudely carved stone statuettes, which were apparently placed in this
cenotaph as substitutes for the bodies of people who had died elsewhere.71 In
Italian Locri excavators found a fifth-century-B.c. "pot burial" in which the usual
remains of the deceased were replaced by a clay bust of a woman, again appar
ently as a replacement for a missing body.72 Literary evidence suggests that
placement of such an image was not always necessary: in the Odyssey Athena,
disguised as Mentes, tells Telemachus that if he learns that his father is dead, he
should erect a tomb for him and offer many sacrifices (1.289-92), and Menelaus
reports that he heaped up a funeral mound for Agammemnon in Egypt, since he
believed him lost at sea (4.584). In both cases the individuals were kings and may
have received special treatment: Herodotus reports that at Sparta it was custom
ary to bury e6woka of kings whose bodies were lost because they died in battle
far from home (6.58).73
A similar rite probably lies behind conflicting stories about the trouble that
arose for Sparta after Pausanias was starved to death in the temple of Athena of
the "Brazen House" (XaXxiolxog).74 The Delphic Oracle ordered the Spartans
to rebury Pausanias's body in the entranceway of the sanctuary and "to give back
two cLtbaca to the Goddess of the Brazen House in payment for one," a riddle
that they solved by dedicating to the goddess two bronze statues "as a replace
ment for [avTi] Pausanias" (Thuc. 1.134-35). The Spartans, although they be
lieved that they had handled Pausanias in strict accordance with their laws,
suffered from a state of pollution that could only be lifted by the reburial of
banning. Picard also discusses a rather early legend concerning Alcmene, who came from this same
city of Midea: when Alcmene died, Zeus sent Hermes to snatch her away and to replace her body
secretly with a stone. When the sons of Heracles discovered the stone, they set it up in a small wood
(as in the Cyrenean decree), a place that became a lQOov for Alcmene (Pherecydes FHG 2.82 =
Ant. Lib. 33; cf. Plut. Rom. 28.7, who reports a similar myth).
71. The two figurines (heights 19 cm and 18.3 cm) are roughly carved from non-native lime
stone (perhaps from Crete). The details of the bodies are only vaguely indicated, but the shoulder
length hair, eyes, ears, nose, and chin of the heads are clearly delineated. F. Hiller von Gaertringen,
"Das von A. Schiff entdeckte Grab," in Theraische Graber, Thera Untersuchungen: Vermessungen
und Ausgrabungen 2 (Berlin, 1903) 291-307, esp. 304-6, figs. 492-93, argues that the long hair and
pronounced buttocks suggest that both were meant to be females. He also believed that the grave
once held a body that was washed out of the tomb. The view is countered by Boardman and Kurtz
(supra, n. 69: 179, 258 fig. 34), who suggest that the stone statuettes were indeed substitute images,
which were placed in a cenotaph.
72. Boardman and Kurtz, ibid. 259 fig. 56.
73. J. Schafer, "Das Eidolon des Leonidas," in K. Schauenberg, ed., Charites: Festschrift E.
Langlotz (Bonn 1957) 223-33, gives a complete discussion, arguing that the term E'6oi)ov here
means substitute image, like the use of xoXooo6g in the Cyrenean texts.
74. See J. Fontenrose, The Delphic Oracle (Berkeley and Los Angeles, 1978) 129-31, for a full
discussion of all the sources and for the approach taken here, that the incident must have involved
the haunting and removal of a ghost.
Pausanias within Athena's temple. A similar situation seems to have occur

southern Italy early in the Classical period. The armies of Croton and M
tum overran and obliterated the neighboring city of Siris, and at the height
carnage they slew fifty youths crowded as suppliants about the altar of A
Later, a pestilence struck the two victorious cities, and again at the prom
the Delphic Oracle both communities erected statues of the youths at the
the massacre. The people of Croton had fifty life-sized statues made, as w
statue of Athena. The Metapontians, who heard about the oracle later o
their haste to fulfill its commands made fifty smaller statuettes out of stone
the manes" and gave sacrifice to the goddess.75 The images of the supp
youths of Siris and the statues of the Spartan Pausanias reveal how inti
religious scruple for proper burial is entwined with fear of the dead, espe
cases where suppliants are slain. In both incidents, the historical account
size the anger of Athena, within whose sanctuaries the murders occ
result, the dedication of the images is interpreted as a repayment to the g
The passing mention, however, that the second, smaller set of statues a
were "for the manes" suggests another interpretation, as do other, later
of the death of Pausanias.
Plutarch twice supplies us with a variant account of the events follo
Pausanias's death, in which it is revealed that the Ei6owov of Pausan
scaring people out of their wits as they entered the Brazen House. The
mention occurs in a scholion to Euripides' Alcestis, at the point where A
upon seeing his resurrected wife worries that she is some Q)ayota from t
Heracles replies defensively that since Admetus has made him a guest-f
(~evog), he, Heracles, could not possibly be a pIvXaywycog, a term t
scholiast explains as follows:76
qvXXaycyo yoi tVE y6 e qEg e tv OexTaXik ovTc xxako ievol, oittveg

xaot(lagToig Xto xa yoqTzeialS Ta& e'oL6a eTatyovoi TE xai E/dtyovol
oOg xci Adxoveg e?tezEtpavTo, vILxa To HnCuoavLaou eL6Woko
ceedTagaE Troig tQooLovtag TO va(O Trig Xakxtolxov, ()5 i(OTOQ
nHkozagQxog ev Taig 'OrtqQlxalig tEkexcZTaLg.
Some sorcerers in Thessaly are in this way called vuXayoyoi, who w

certain purifications and magic spells send ghosts out [i.e., to attack]
well send them away. Even the Spartans sent for them, when the ghost of
Pausanias terrified people as they were entering the temple of the Go
dess of the Brazen House, as Plutarch says in his Lectures on Homer
The comment is doubly instructive. First of all, it explains why Heracles

aback at Admetus's comment-because sending up ghosts from the g
haunt someone is the act of an enemy, not a friend. The scholiast also pr
75. Justin, Pompei Trogi Epit. 20.2.

76. Schol. Eur. Alc. 1128 = Plut. Mor. frag. 126 (Sandbach).
us with a Plutarchan fragment that mentions a fact Thucydides apparently left

out of his narrative (the haunting of the temple), while at the same time corrobo
rating an important detail in Thucydides' narrative; Plutarch says the ghost was
seen by people going into the temple-that is, at the very spot where (according
to Thucydides) Pausanias gave out his last breath, and where the oracle stipu
lated the body be interred. In another passing reference to these pvcaywyooi
Plutarch informs us that they were summoned by the Spartans at the command
of the Delphic Oracle:77
And in a similar way, when the Spartans, too, were ordered by the
oracle to appease the spirit [4vXT|] of Pausanias, the pvXaycyoi were
summoned from Thessaly,78 and after they made sacrifice, they drew the
ghost [E'i6wXov] from the temple.
Aside from his rather general reference to purifications and magic spells, Plu
tarch fails to specify what rituals were performed to exorcise the ghost. The
author of the Letters of Themistocles, probably a near contemporary of Plutarch,
testifies to the same tradition of a haunted temple when he refers obliquely to the
fact that Sparta was plagued by the nakacvaiog or &kaXTiTog (both terms can
mean "ghost") of Pausanias, which was in fact exorcised "by means of bronze
statues" (Epist. 5.15, XaXXEioLg a&v6QLaov aJLo6bloioJZtoPLOLRov), surely a refer
ence to the brazen effigies that all accounts agree were made at the bidding of the
Delphic Oracle. It is clear that in the second century A.D., at least, there existed a
tradition according to which the Spartans brought in Thessalian magicians who
performed rituals that seem to have included a rite of (re)burial of the body, as
well as the manufacture of two statues.
It is important to note, however, that this second tradition of the haunted
temple and the subsequent exorcism of the angry ghost does not (as far as we can
tell) necessarily contradict Thucydides' explanation of pollution followed by
purification by means of compensation to an offended deity; in his account of the
exorcism, Plutarch actually corroborates (e.g., in the detail of the place of death)
and augments (e.g., giving the occasion for the oracle) the Thucydidean narra
tive of Pausanias's death without directly confuting it. The widely differing inter
pretations of the same events can be attributed, in part, to the fully articulated
daemonology characteristic of the second century A.D.,79 as well as to a tradi
77. De sera num. vind. 17 = Mor. 560e-f. The dialogue is set in Delphi, and Plutarch raises the
story of Pausanias's ghost in connection with Delphic oracles in which the god's reply is based on an
assumption of the post mortem existence of the soul.
78. The manuscripts read e 'ITaitcag but most editors print /i OeoaXaoct to make it consistent
with the fragment quoted above. Mittelhaus, RE XIX 2084 s.v. "Phigaleia," suggests reading et
cItyaxag;, citing Paus. 3.17.9, where we learn that the same Pausanias went to Phigalia to consult
pvXctayyoi in order to appease the ghost of a young girl he accidentally killed in Byzantium.
79. W. Burkert, "'OH.: Zum griechishen 'Schamanismus,' " RhM 105 (1963) 48-49 (esp. n.
67), in his exegesis of Plato, Laws 909b (the use of state-hired magicians), discusses Plutarch's report
of these Thessalian yo6reg in detail, and points out that the silence in Thucydides' account (about the
tional split in the ancient Greek understanding of the exact operations

tion and purification.80 In addition, Plutarch is probably using a traditi
phic source here, which may reflect the Spartan understanding of these
better than does Thucydides, with his notoriously rationalizing and de
Attic viewpoint.81
There is one instance in Greece where the effigy of a ghost is bound
9.38.5):
With regard to Actaeon the Orchomenians say the following: A ghost
was running amok and ravaging the land.82 When they inquired at Del
phi, the god told them to recover the remains of Actaeon and bury them
in the earth. He also commanded them to make a bronze image of the
ghost and fasten it to a rock with iron. They also offer chthonic sacrifices
[evayciovotv] every year to Actaeon.
The similarities between this account and the latter traditions about Pausanias's
ghost are obvious: a horrible death, an Ei6bOov haunting the scene of the mur
der, and a Delphic oracle containing a two-part exorcism: (1) the (re)burial of
the body, or whatever was left of it, and (2) the construction of bronze statues.
The existence of the statue of Actaeon is alluded to in another source.
Apollodorus offers a different and a patently less plausible aition for a statue that
seems to have existed on Cithaeron, but one that nevertheless retains the pattern
of disruption and then return to order by the erection of a statue (3.4.4): he
writes that after they finally "came to their senses," the dogs of Actaeon began to
occasion of the oracle) can reasonably be filled in from the "ghost story" in Plutarch's version,
provided that one realizes that the highly elaborate daemonology of the latter only amplifies certain
tendencies in the Greek popular tradition that were already evidenced in the plays of Aeschylus and
elsewhere. He mentions the ei6&oov of Clytemnestra and the Erinyes as an alternate formulation to
the pollution that besets Orestes.
80. L. Moulinier (Le pur et l'impur dans la pensee des Grecs d'Homere d Aristote [Paris, 1952]
215-42) and Parker (supra, n. 55: 104-15) discuss the apparent inconsistencies in the Athenian
conception of pollution, separating them into two sources: (1) [Antiphon's] Tetralogies and the
tragedians, who consistently portray a "daemonic" interpretation of pollution, in which the soul of
the murder victim or the avenging spirit(s) that he or she sends out can do actual harm to the
murderer, his family and associates, and all that live in his city; (2) actual forensic orations, as well as
the works of the historians and the comedians, which make no mention of bai[toveg arising from
pollution, and speak only of the impurity of bloodstained hands and the legal remedies of exile,
expiation, and purification rites. Both Parker (ibid.) and J.-P. Vernant ("The Pure and the Impure,"
in Myth and Society in Ancient Greece, trans. J. Lloyd [Atlantic Highlands, 1980]) offer some
sensible corrections to Moulinier's exaggerated positivism and his outright rejection of the "dae
monic" interpretation of pollution as mere poetic fancy.
81. Fontenrose (supra, n. 74: 129) suggests that Thucydides knowingly suppressed the ghost
story current in the "folk" tradition because he did not consider it to be "true history." Rohde (supra,
n. 69: 173-75) singles out Boeotia and Laconia as two areas where deep religious conservatism led to
the exaggerated reverence for the souls of the dead and a strong belief in their continued efficacy
amongst the living.
82. Reading, with most editors, Hitzig's emendation JtEQTIQTEOV instead of the MSS' AETecXv
EXov.
howl inconsolably and searched vainly for their master until they came to the
cave of Chiron, who fashioned an e6coxov of Actaeon (there is no mention of
binding) to assuage their grief and, presumably, to quiet the uproar.83
As the foregoing discussion reveals, in terms of motivation, function, and
expected result, the symbolic (re)burial of the lost or angry dead by the erection
of cenotaphs or substitute images is remarkably similar to rituals in which the
hostility of an angry ghost is averted by the binding, the burial, or simply the
erection of such a substitute. As mentioned above, similar parallels exist be
tween the Assyrian and Egyptian ghost-banning ceremonies and their own indige
nous burial rites. It is also important to recall that some of the terminology of the
Cyrenean ghost-banning rite is similar to that used to describe Greek funerary
ritual.84 The pairs of stone statuettes from the cenotaph in Midea and the so
called Schiff's Grave on Thera also deserve special notice in connection with the
stipulations of the Cyrenean lex sacra, in that they each seem to constitute a
couple, as is prescribed by the inscription.85 The dates and provenances of these
artifacts vaguely suggest some Peloponnesian tradition dating back to the earliest
periods. Bronze Age Midea provides the earliest attested cenotaph in the Greek
world. The seventh-century-B.c. effigies from Schiff's Grave give us a glimpse of
a similar ritual performed on Thera roughly at the same time when her colonists
founded Cyrene, presumably bringing along their native cults and burial tradi
tions, including (it seems) a ritual exorcism of ghosts that employed the use of
83. A. Casanova, "II mito di Atteone nel catalogo Esiodeo," RFIC 97 (1969) 31-46, esp. 33,
connects this version of the etowAov story with POxy 2509 (which mentions only the grieving dogs
and Chiron) and suggests that Hesiod is the ultimate source. W. Burkert, Homo Necans: The
Anthropology of Ancient Greek Sacrificial Ritual and Myth, trans. P. Bing (Berkeley and Los An
geles, 1983), 112, in accordance with his overarching theory of sacrifice as a form of restitution argues
that the statue here is a symbolic restoration of Actaeon to the dogs that "sacrificed" him. The
"Orchomenian" coin shown in Daremberg-Saglio, DAGR s.v. "Actaeon," fig. 84, and described as a
fourth-century-B.c. illustration of this statue, is apparently a mirage; B. V. Head, Historia Numorum
(Oxford, 1911) 451, demonstrates that the provenence of the coin is Arcadian Orchomenos and that
the figure is that of the mortally wounded Callisto, not the bound Actaeon.
84. De Sanctis (supra, n. 66: 204-5) points out that the description of the meal offered to the
figurines uses terminology traditional to funerary offerings. Benveniste (supra, n. 65) argues that the
original and basic meaning of xokooodg is "image of the dead." The timing of the invocation in the
Cyrene inscription admits two different interpretations, "for three days," a solution favored by most
editors (e.g., De Sanctis, Wilamowitz, Vogliano, Stukey, Servais) and the one followed here, or
"thrice during the daytime," the reading adopted by Radermacher and Olivero. The three-day time
period recalls the length of some Assyrian ghost-banning rites, which are themselves patterned on
funerary cult (supra, n. 46). A triple call in one day, however, is very much like the call used by
Greeks to summon a dead soul back to its homeland for some kind of substitute burial; see Od. 9.65
67 with Eustathius's comments ad loc. (pp. 1614-15); Verg. Aen. 6.505-8; and Rohde's discussion
(supra, n. 69) 173-79.
85. The cenotaph at Midea (supra, n. 70) contained a pair of menhirs of different heights, but
no demarcations of gender; Schiff's Grave (supra, n. 71) also contained a pair of stone statuettes
which betray no clear indication of gender (pace Hiller von Gaertringen [supra n. 71]). One of the
statuettes is, however, slightly larger and pierced through the neck; compare the lead voodoo dolls
from Delos (Appendix no. 12), where the males alone were impaled with iron nails (infra, n. 95).
xoXoooot as substitutes for reburial.86 It is in fact interesting to note th

the Greek ghost-banning stories discussed above, only the effigy of Act
explicitly said to be bound, a variation one might expect, given the nati
tian traditions of binding effigies of Ares.
PRIVATE "VOODOO DOLLS" IN THE GREEK AND ROMAN WOR
There is widespread evidence that throughout the Greek and Roman

lead, clay, and wax figurines were regularly bound and buried in private ritu
bind one's personal enemies.87 A survey of all extant Greek, Etrusc
Roman "voodoo dolls" known to me appears at the end of this articl
Appendix. The earliest Greek examples seem to be two Archaic bron
rines, but this has been disputed: a Tegean example (no. 8 in the Ap
holds its arms aloft in a curious dancelike gesture and has its head
twisted completely round, and a bronze figurine from the island of Cep
(no. 10; Fig. 3) has its head, feet, and one arm completely reversed. As
the very real problem of dating uninscribed early bronzes once they ha
removed from their archaeological context,88 scholars have been reluctan
past to ascribe malevolent intent to these dolls because they do not exhi
of the more overt signs of magical activity, such as an inscription of the
name or limbs that are actually bound. They prefer instead to see these a
as "grotesques," ugly, contorted images that were used as a form of pr
"white magic" to ward off the evil eye.89 More recent discoveries, howe
86. Picard (supra, n. 70) argues that the Theran and Cyrenean data are related to the
menhirs and when taken together give evidence for a "Doric" tradition of ghost-laying ritua
87. For the most recent survey of the Attic dolls, see D. R. Jordan, "New Archa
Evidence for the Practice of Magic in Classical Athens," Praktika of the 12th Internationa
of Classical Archaeology, Sept. 4-10, 1983 (Athens, 1988) 273-77. For the older treatment
bibliography in the Appendix to this article.
88. E.g., P.-F. Perdrizet, "Les pieds ou les genoux au rebours," Melusine 9 (1898-99
dates the Cephalonia figurine (see Appendix, infra, no. 10) to Hellenistic or Roman time
K. A. Neugebauer, Katalog der statuarischen Bronzen im Antiquarium 1 (Berlin, 1931) 66
it as early Archaic.
89. Perdrizet (supra, n. 88: 41) and G. Q. Giglioli ("Curiosita archeologiche," SE 3 [
31) argue in one way or another that these dolls are grotesques, like the Greco-Roman minia
the Egyptian god Bes, the dwarfs, and other figures in shocking or obscene poses that we
ward off the evil eye in later antiquity. In his discussion of the position of the hands covering
of the lead figurine found in the Ilissus River (see Appendix, infra, no. 2), Perdrizet declares
just another "typical" form of the grotesque, like many a nude bronze found in Italy (both m
female) that covers the mouth with one hand and the anus with another (C. Q. Giglioli, "II r
Bianchini" BollComm 56 [1929] 5-51, provides a detailed description of a cache of these f
that was found in a ruin near Rome in the seventeenth century, sealed in a bronze box place
large terra-cotta vase). One should compare, however, the treatment of a ghost's effigy
57.41-43: "You make a representation of the ghost or anything evil that has seized hi[m].
its name (on it). You have it hold its mouth with its right hand and its rear end with its
effigy is then deposited in a copper container and buried. Compare also a sixth-century-B
figurine from Syria (Nat. Museum of Copenhagen, inv. no. 1989; see Giglioli, ibid. 50-51),
depicted in just such a pose.
similar dolls encourage the inclusion of these early bronzes in the category of
"voodoo doll."90 Four cast-lead dolls (three discovered in Attica, one in Sicily)
all have their heads distorted in a manner identical to the early bronzes, and in
addition also betray at least one other indication of malevolence: the doll found
in a disturbed fifth-century Athenian grave (no. 1) has its hands twisted behind
the back as if bound, two Attic dolls of unknown date (nos. 2, 3) have their arms
twisted behind the back, and the Sicilian figurine (no. 16) carries a late fifth
century inscription that lists the names of ten men. The last example, moreover,
has its arms raised in the same peculiar dancelike gesture as the Tegean figurine.
All the early figurines discussed above were cast from metal, and conse
quently their production required some modicum of technical skill; five of the
examples (nos. 4, 8, 13, 14, 17) have even been admired for their high artistic
quality. At no point, however, do we see, even in the best-crafted examples, any
attempt at realism. Thus it is clear that these images were never meant to be
portraits, and that the sympathetic action of the doll was directed by the inscrip
tions that appear on some of them, or by the formulas that were undoubtedly
uttered over them in the process of the rite.91 As time goes on we see a marked
decline in the quality of the images. The first examples are the related group of
figurines found in two different graves in the Ceramicus (nos. 5, 6; see PI. 5); all
dating to about 400 B.C., they seem to have been the work of the same person or
persons.92 These hand-molded lead figurines are squat and ugly, with oversized
feet and genitals; the hands are bound behind the back with a thong of lead.
They all appear to have been employed as part of a judicial curse aimed at
restraining the success of an opponent at law; each doll is inscribed on a leg or an
arm with a single name and placed within a lead box, which is inscribed with a
judicial curse.93 Four cast-bronze figurines found in a Hellenistic house on Delos
(no. 11; see Pls. 4b, 6-9) are similar in their smaller size and poor execution, but
they are uninscribed. Five cast-bronze statuettes of unknown date from Aloni
stena (no. 9; see Pls. 2, 3, 4a) are closely related to the Delian figurines in size
and treatment, and are probably contemporaneous; the hands of all five are
twisted behind the back and bound, and the feet of each are bound together with
90. For many other arguments that the twisted heads point to binding rituals, see L. Mariani
("Osservazioni intorno alle statuette plumbee Sovanesi," Ausonia [1910] 39-47) and D. Comparetti
("Figurina plumbea iscritta," 194-97) in Miscellanea di studi sicelioti ed italioti in Onore di Paolo
Orsi, Asso 28 [Catania, 1921].
91. For a general discussion, see Tambiah (supra, n. 11). C. Daxelmuller and M.-L. Thomson,
"Bildzauber im alten Mesopotamian," Anthropos 77 (1982) 27-64, esp. 53-58, argue that effigies
used in Mesopotamian magic were not meant to be portraits. In both the Greek and the Assyrian
traditions, it is the addition of the victims' names (or, less frequently, hair or some other kind of
ovoia) that allows one to pinpoint the target of the spell. The many uninscribed dolls must have been
the object of spoken incantations that similarly made some connection with the name of the victim.
92. Jordan (supra, n. 87).
93. Ibid. For a discussion of the wider cultural context of judicial curses in the Classical era, see
C. A. Faraone, "An Accusation of Magic in Classical Athens (Ar. Wasps 946-48)," TAPA 119
(1989) 149-61.
a thick band. Their heads are all twisted unnaturally, in one case co
around.
In the much later periods clay, lead, and (in Egypt) wax figurines show up
wherever Greek culture exercised its influence: for example, in Italy (nos. 16
24), North Africa (nos. 25, 26), Egypt (nos. 27-31), and Palestine (no. 32).94 Of
the later examples, I limit my discussion to the appearance of females, which are
somewhat rare in the earlier periods. The earliest example is the lead "ginger
bread" figure from Euboea (no. 15; see PI. 11), whose gender is only indicated
by the woman's name inscribed upon it. The earliest bound figurine with identifi
ably female features is the figurine of cast lead with arms twisted behind its back;
it was found together with a male of identical pose in the entrance of a late sixth
century-B.c. Etruscan tomb (no. 18; see Pls. 12, 13). They were apparently
placed there much later, for each doll is inscribed on its left leg with a name that
can be dated by epigraphical evidence to the late fourth or early third century
B.C. Two uninscribed couples roughly hewn from small slabs of lead (no. 12; P1.
10) were found buried in the supporting wall of the first-century-B.c. temple of
Zeus Hypsistos on Mount Cynthus on Delos. The method of binding seems to
have been gender-specific in this case: the two male figures have nails driven into
their eyes, ears, and mouths (the mouths are then covered with the right hands),
while the females are constrained only by a thick choke collar.95 There are also
four pairs of inscribed clay figurines (no. 20; Fig. 4) dating "no later than the first
century B.C." that like the Delian couples were carved from small slabs, but of
clay not lead; they were found in Puteoli jumbled together amidst the remains of
what appears to be a cremation burial. And finally, there is the pair of elongated,
Modigliani-like dolls found in Roman Morocco (no. 25); their attenuated legs
are unnaturally twisted, and the male holds his hands crossed before his belly as
if they are bound.
If we exclude the late antique Greco-Egyptian effigies used in erotic magic
(nos. 27-31), all but three of the extant female dolls appear in a pair with a male;
this may, of course, be only an accident of discovery, but the Italian provenance
94. The dolls found in Egypt and in the Near East are treated as Greek because they are
inscribed in that language or because they were discovered in situ with inscribed Greek defixiones. As
I have argued in the section "Bound Statues in Egypt and the Near East" above, these rituals are
shared by most cultures in the eastern Mediterranean; thus figurines found in Hellenized Africa or
Syria are undoubtedly the result of a syncretism of various Greek and non-Greek binding rituals.
Two of the Greco-Egyptian dolls (nos. 27, 29), for example, seem to recall the native Egyptian
tradition, both in their kneeling "bound captive" position, and in the care taken to depict the
hairstyle and jewelry (native dress?) of the victim. See Ritner (supra, n. 21) 93-94.
95. The treatment of the males indicates an effort to bind their cognitive and verbal faculties:
nails are driven into their ears and mouth, and between the eyes. These are often the targets in a
judicial curse; see Faraone (supra, n. 93: 58-60); idem, "Aeschylus' hymnos desmios (Eum. 306) and
Attic Judicial Curse Tablets," JHS 105 (1985) 151-52. The subsequent covering of the mouth with
the hand of the male dolls (see detailed description in the Appendix) reminds one of the binding spell
cast on Libanius, which involved a mutilated chameleon whose front left foot had been inserted in its
mouth. See Faraone (supra, n. 2: 21-22 n. 3) for discussion.
of two of the caches and the strong Roman influence on another (Morocco) may
point to a specific epichoric variation.96 This suggestion is further supported by
the strikingly similar configuration of the so-called human sacrifice of the Gauls
and Greeks that the Romans performed thrice during the great crises of 228, 216,
and 114/13 B.C.97 In each case the Romans buried alive a Greek couple and a
Gallic couple in the Forum Boarium; it was apparently a magical ritual designed
to protect the city from an attacking enemy.98 This use of "living images" can be
paralleled by binding rituals in which small animals (in one case bound) were
apparently buried (in lieu of a "voodoo doll") with lead defixiones as part of
some rather elaborate binding spells.99 On account of the anachronistic use of a
Greek couple (Greeks were never enemies of Rome during the years in ques
tion) it is usually argued that the burial ritual is an old Etruscan rite that still
expresses an earlier schematic view of all their enemies (i.e., Gauls to the north
and Greeks to the south are symbolic equivalents of "any and all enemies").100
Although some of the Italian figurines are inscribed with what appear to be
96. The only non-erotic female voodoo dolls, which do not appear as part of a couple, are: the
flattened lead "gingerbread" figure from Carystus (no. 15), which carries a defixio text that aims at
binding a woman named Isias; the lost terra-cotta doll inscribed "Noema" (discussed in the Appendix
parenthetically after no. 7); and the bronze female figurine from Smyrna (no. 33). The Delian
effigies date to a period of strong Roman influence, when the island was a major center of the slave
trade. See, e.g., M. Nilsson, "Roman and Greek Domestic Cult," Opuscula Selecta 3 (Lund, 1960)
277-85, who must rely in part on the abundant evidence from Delos for his discussion of the typical
Roman customs involving the dii penates and the lares. For the unique character of the Egyptian wax
voodoo dolls (Appendix nos. 27-31) and aphrodisiac binding spells in general, see Faraone (supra,
n. 2) 14-15.
97. Plut. Marc. 3.4 (225 B.c.); Livy 22.57.2-6 (216 B.C.); Plut. QR 83 = Mor. 284c (114 B.C.).
For the revival of these rites sometime in the Julio-Claudian period, see Pliny, NH 28.3.12, and
Suetonius, Gaius 29.2. The precise chronology of the three Republican episodes is fixed by C.
Cichorius, "Staatliche Menschenopfer," Romische Studien (1922) 7-16. D. Briquel, "Les enterres
vivants de Brindes," in Melanges J. Heurgon (Rome, 1976) 33-40, discusses a legend preserved by
Tzetzes (ad Lycophron Alex. 602) that tells of Greeks burying Aetolians alive (men, not couples) to
avert a crisis, and argues for a Greek origin to the rite, which when Romanized was expanded to
include females.
98. A. M. Eckstein, "Human Sacrifice and the Fear of Military Disaster in Republican Rome,"
AJAH 7 (1982) 69-95, reviews the large bibliography. He demonstrates that the scandal over the
priestesses was actually considered a prodigy that portended some greater, future disaster and that
the burial of the Greeks and Gauls was an act designed to ward off a catastrophe signaled by the
scandal as well as a series of other evil omens.
99. Faraone (supra, n. 2: 21-22 n. 3) discusses the use of a rooster, puppies, a chameleon, and
a fish in binding rituals.
100. F. Schwenn, Die Menschenopfer bei den Griechen und Romern, Religionsgeschichtliche
Versuche und Vorarbeiten 15.3 (Giessen, 1918) 148-50; K. Latte, Romische Religionsgeschichte
(Munich, 1960) 256-57; and most recently B. MacBain, Prodigy and Expiation: A Study in Religion
and Politics in Republican Rome, Collection Latomus 177 (Brussels, 1982) 62. The recent thesis of A.
Fraschetti, "Le sepolture rituali del Foro Boario," in Le delit religieux dans la cite antique, Table
ronde, Rome 6-7 avril 1978 (Rome, 1981) 51-115, that the rite is Roman from the earliest periods
does not seem to be tenable; see D. Briquel, "Des propositions nouvelles sur le rituel
d'ensevelissement de Grecs et de Gaulois au Forum Boarium," REL 59 (1981) 30-37. In the revival
under the Julio-Claudians the nationality of the victims seems to have been changed to reflect the
current realities; see Pliny, NH 28.3.12
personal Greek or Etruscan names, we might understand in the pairing

ples that some similar pars pro toto magic is at work, except on a smalle
and that all of these couples are somehow supposed to represent larger
groups-that is, whole families, clans, or perhaps towns. In any eve
interesting that the Romans or the Etruscans apparently felt that both
to be bound to restrain an enemy, while the Greeks were only concern
the males (cf. the three male silver statues in Thrace discussed above).
other hand, the pairing of male and female dolls of anonymous ghosts o
ers in the Cyrenean lex sacra and in the Assyrian ghost-banning rites m
vide an alternative and equally attractive explanation, along with some
ing parallels in the details of the rite. Each figure of the lead couple f
Etruscan tomb, for instance, was inscribed on the left side of the body
buried in a tomb, much like the instructions in the Assyrian ghost-banning r
(MMG 15, 20, 21). Furthermore, the nails pressed into the mouths of t
male Delian dolls recall the dog's tooth or cornel-wood peg that is inser
the mouth of an Assyrian ghost's effigy (MMG 22, 57).
Despite the variety of ways in which the effigies collected in the A
were treated, I think that at least in the case of the early examples we ca
say that the ultimate goal was the containment or the restraint of the enem
necessarily his or her demise. The same is true for the larger civic sta
cussed in the first section of this article. The symbolic logic of some of the
of binding is not, however, always clear to the modern observer. The na
appear in several dolls (e.g., nos. 7, 12, 21) are a case in point; like the na
pierce the rolled-up defixiones, they most probably aim at "nailing dow
nents and making them immobile, not at piercing their flesh or w
them.101 This is especially clear from the only extant detailed narrati
person's escape from a "binding spell."102 Sophronius, a Christian author
late sixth century, reports how an Alexandrian paraplegic named Theoph
visited in a dream by two saints and told to go down to the harbor and
large sum of money to the fishermen there to buy their next catch, sight u
He did as he was told, and when the nets were hauled up, amongst the str
fish a small box (xLI3P0TOV) was discovered, which contained a bronze s
with nails driven into its hands and feet. As each of the nails was withdr
paralysis in the corresponding limb ceased immediately.
Although certainly no one would vouch for the historicity of this pa
incident, the need for verisimilitude in the details of such miracle stories su
that the underlying assumptions about the paralyzing but non-fatal ef
101. Only a few early defixiones refer even vaguely to piercing or perforation. (e.g., XE
in DTA 96, 97, "sting" or "stab"). The nails are usually driven through the head or c
probably aim at "binding" the internal organs or abstract faculties, which are often the
defixiones (e.g., heart, soul, vision, speech, memory).
102. Narratio Miraculorum Sanctorum Cyri et Joannis, PG 87.3 col. 3625; the relevant
of the Greek text appear in DT, pp. cxxii-cxxiii.
"voodoo dolls" were still common knowledge, at least in the early Byzantine
period. Certainly the description of a naked doll imprisoned in a "coffin" and
deposited in a body of water is in complete agreement with much of the archaeo
logical evidence surveyed in the Appendix (esp. nos. 4-6, from Attica). The
sympathetic effect of the nails was merely to "transfix" the limb in order to make
it immovable. We must allow for a similar latitude in the interpretation of those
figurines that have their heads and feet twisted completely around (e.g., nos. 1
3, 8-10). This position most probably aims at the utter confusion of the victim's
limbs and not literally at the breaking of his neck or strangulation, the natural
result if this action were actually applied to a human body.103 In this regard it is
interesting to note Pausanias's description of the decoration of the famous Chest
of Cypselus (5.18.1):
A woman is represented as carrying a white boy asleep on her right arm;

on her other arm she has a black boy who is like the one that sleeps. The
feet of both boys are turned in different ways [6iLE(oxJTQaC vovg]. The
inscriptions show, what is easy to see without them, that the boys are
Death and Sleep, and that Night is the nurse of both.
(trans. J. G. Frazer)
Some scholars point to the reversed feet on many of the Greco-Roman voodoo
dolls discussed above (e.g., nos. 3, 8, 10, 16) and suggest that the participle used
by Pausanias to describe the feet of the boys (6eoxTQaE[t[tvovg) refers to the
same action and ought to be translated "completely twisted around" instead of
"turned in different ways."104 Death and Sleep are of course popularly thought to
come and fetch the dying (cf. esp. I/. 15.454-55, the death of Sarpedon), so it is
understandable that they should be depicted in a way that would hinder their
prompt arrival.
The foregoing discussions of literary and archaeological evidence have fo
cused on different aspects and contexts of binding magic in the eastern Mediterra
nean. Taken together, the evidence suggests that the ancient Greeks (and the
103. There is a good parallel in the similia similibus formulas found on defixiones that encour
age the victim to become analogous to the retrograde direction of the writing, e.g., the formula found
on a late Classical Attic defixio: "Just as these things [i.e., letters] are backwards [avevjtakLtv], so too
may all words and deeds be backwards for him" (SGD 40). The notorious mutilation of Oedipus's
feet may have been similarly motivated to prevent his ghost from returning to haunt his parents; such
babies were commonly thought to be atwotl ("untimely dead"), and since they had to wait for a long
time before entering Hades, they were thought to be good prospects for a haunt; see J. ter Vrugt
Lentz, Mors Immatura (Groningen, 1960) 32-51.
104. Pedrizet (supra, n. 88) and Giglioli (supra, n. 89) both discuss the Pausanias passage in
light of the voodoo dolls with twisted feet. H. Shapiro, "Personification of Abstract Concepts in
Greek Art and Literature to the End of the Fifth Century B.C." (diss., Princeton, 1976) 67-69, points
out that aside from the canonical and much-copied depictions of the death of Sarpedon, Death rarely
appears in Greek art, a fact that he attributes to the same "superstitious" caution that caused people
to avoid using his name.
Etruscans and Romans as well) were accustomed to bind or bury effigies

enemies in an attempt to inhibit their hostilities or at least their ability to h
As we have seen above, these enemies could be dangerous deities (e.g
Enyalius), ghosts (e.g., Pausanias, Actaeon), or mortal men (e.g., th
depicted by the silver statuettes in Thrace). Like those of the defixion
targets of the extant "voodoo dolls" surveyed in the Appendix are gen
assumed to be living human rivals. Near Eastern parallels suggest, howev
some of those figurines inscribed with human names may very well refe
or women who had already died, and that they were images used to bind
ghosts like those envisioned in the Cyrene decree, in the bound prison
scribed with the names of "deceased" Egyptians, or in the Assyrian
entitled "Against Appearing Ghosts."105
We must allow, too, for the possibility that some of the uninscribe
found in Greece may have even represented inimical gods or demons.
case, in particular, the effigy (no. 19; Fig. 5) holds the accoutrements
Etruscan Heracles; its head, however, is twisted completely about, as in
of the other dolls discussed above, and forces us to recall that the god/he
appears as an unpredictable force of death and destruction.106 It is not
reason that someone might wish to bind such a deadly superhuman po
lead voodoo doll from Sicily (no. 16) has an ugly, contorted face, w
rightly been described as "demonic." A similar interpretation has been a
above for the depictions of the gods Sleep and Death with twisted feet
Chest of Cypselus. In each case, the effigy is contorted to restrain the d
divinity represented. A further example seems to have come to light rece
late tenth-century-B.c. shaft grave at Lefkandi in Euboea-a painted te
centaur (height 36 cm.) broken in several places. Although normally suc
of disrepair would not provoke any suspicion of magical activity, ther
dence that the figurine was broken prior to its burial, as its severed h
found in a different tomb dating to the same period.107 In light of the
Etruscan, and Near Eastern practices discussed above, I suggest that the
may have been purposely decapitated and broken up, and then bur
protective measure against a demon envisaged as a centaur, a figure th
Heracles) often appears as a disruptive, destructive force, much like Pa
105. As far as I can tell, there is no evidence that Greek defixiones were used to bin
106. The locus classicus for the deadly Heracles is Od. 11.605-12, where he appear
about with his bow forever drawn-a deadly force with a hair trigger. He alone shares Ar
rITolJOQog (Hes. frag. 25.23 M-W) as the traditional destroyer of Troy and Oechalea.
107. V. R. d'A. Desborough, K. V. Nicholls, and M. H. Popham, "A Euboean Centa
65 (1970) 21-30, pls. 7-11, argue that the peculiar circumstances of the burial suggest tha
prized family heirloom, which was purposely broken and carried to the grave by two differ
members because neither wanted to give it up. This interpretation is followed in the final
report; see M. H. Popham, L. K. Sackett, and P. G. Themelis, Lefkandi 1, The Tombs
Contents (Oxford, 1980) 168-70, 344-45 with pls. 251 and 252.
Sileni, and other half-human beasts.'08 Burying the head in another place would
be a particularly effective way to prevent the statuette from being reassembled.
CONCLUSION
In Assyria continuously from the second millennium B.C. down through to the
Persian conquest and in Egypt well into the Ptolemaic period images were bound,
buried, or burned in preemptive defensive rituals that sought to destroy or at least
inhibit both personal and public enemies, whether mortal or superhuman. In
ancient Greece, myths alluded to in Homeric poetry and artifacts dating to the
Bronze Age and the Archaic period signal the existence of ceremonies of similar
intent and scope. The binding of Ares, for instance, and his imprisonment in a
bronze cauldron is reminiscent of the Assyrian and Egyptian procedures discussed
above for incarcerating bound images of ghosts and other enemies in clay or metal
pots deposited in a necropolis or on the steppe, activities that also seem to lie
behind Hittite myths about demons and personified evils imprisoned in bronze
pots in the underworld or the sea (supra, n. 7). The Greek "voodoo dolls" from the
Athenian Ceramicus (ca. 400 B.C.) repeat the pattern: bound lead effigies were
enclosed in inscribed lead boxes and then buried in a graveyard (nos. 4-6). The
more elaborate Syedran tableau of Dike standing over the bound and suppliant
Ares seems to have its parallel in Egyptian monumental reliefs, which so often
depict the pharaoh threatening a kneeling and bound foreign prince.
There are of course many notable differences in how these rituals were
performed. In Egypt, many of the national apotropaic rites were enacted on a
regular, if not daily, basis by a small coterie of literate priests working from
traditional ritual handbooks; it is only in the last periods of Egyptian dynastic
history that the priesthood appears to be making private copies of such texts and
performing these rituals for their own use.109 In Assyria and Greece binding
ceremonies appear both as part of annual ceremonies and as special rituals
employed in times of crisis. The annual repetition of the Maqlu ritual and the
apparently annual (re)confinement of Ares in Boeotia are good examples of
calendrical rituals, while the Assyrian recipes for laying ghosts and those in the
Cyrenean decree appear to have been used only in times of sickness or some
other serious problem. The bound statues of Ares at Syedra and in Thrace, and
of Actaeon at Orchomenus, were similarly a reaction to a specific crisis, although
in the first and last instances it is probable that these statues were regularly
treated to sacrifice thereafter, as was the bound Ares in Boeotia (as the quip of
Herodes about the Aloadae suggests).
108. For Sileni and goat demons, see M. Nilsson, Greek Popular Religion (New York, 1940) [=
Greek Folk Religion (Philadelphia, 1961)] 10-15 and fig. 5. For the attacks of Pan, see P. Borgeaud,
The Cult of Pan in Ancient Greece, trans. K. Atlass and J. Redfield (Chicago, 1988), 88-116; W. K.
Pritchett, The Greek State at War 3 (Berkeley and Los Angeles, 1979) 45, 162-63.
109. Ritner (supra, n. 21).
To some extent, the presence of these public "voodoo dolls" acts lik
active traces on a Geiger counter, indicating an elevated level of anxie
particularly dangerous neighbors. In Egypt, for example, although the
rituals at Abydus and elsewhere carefully named all of the possible enem
rivals of the state, the high incidence of bound figurines of Nubians and sim
inscribed plaques along the Nubian border (many of which were of hur
inferior execution) may indicate a special concern about these people a
points in time. The public binding ceremonies of the Greeks, Etruscan
Romans reveal a similar anxiety about invaders from the north. To be s
old Etrusco-Roman ritual of burying Gauls and Greeks alive symbolicall
at inhibiting all enemies, but it was only performed in historical times to w
danger from the northern border. For the Greeks of late antiquity, the
ans to the north were the Goths, against whom were directed the thr
silver statues described by Olympiodorus and the bound statue of Are
tioned in the Palatine epigram. There are, however, indications that suc
of "Ares' brood" in the north were commonplace as early as the Classica
see, for example, the bizarre oracle at the end of Euripides' Bacch
Cadmus and Harmonia (the daughter of Ares) will someday return f
north leading a barbarian horde to sack many Greek cites, only to fall s
taking Delphi.110
In Greece there is some evidence for the annual performance of m
(re)binding rituals, but it is rarely explicit (e.g., the Boeotian treatment
implicit in the myth and anecdote discussed above). There are sever
notices in Pausanias and other, later authors of the existence of bound st
Artemis, Dionysus, Hera, and Athena that were apparently released on
from bondage and may have received special sacrifice.'1l In some ca
bound or incarcerated statues were thought to be important to the he
continued safety of the city.112 It also appears that lore about binding st
110. Ba. 1330-39 (cf. 1355-60), with E. R. Dodds, Euripides: Bacchae2 (Oxford, 1960
for a full discussion of this oracle and similar ones, which date as early as the time of the Per
111. A wooden image of Artemis Eurynome at Phigalia was bound in chains, and its
was only opened once each year (Paus. 8.41.6); a small helmeted image of Artemis Soteira
was kept out of sight (Paus. 7.27.3, Plut. Arat. 32.2); a small wooden image of Artemis O
"Lygodesma" was bound by willows (Paus. 3.16.7-11); a small wooden image of D
Asymnetes at Sparta was shut up in a box and only brought out once annually (Paus. 7.1
armed Aphrodite Morpho, also at Sparta, had its feet bound by golden chains (Paus. 3
and a small statue of Athene, the so-called "Palladium," was hidden in the temple at Troy
Mor. 309-10, Dercylus FHG 4.377.5; cf. the Palladium at Rome: Seneca, Contr. 4.2, Di
Pliny, NH 7.141; Juv. 6.265). The Gorgon's Lock at Tegea was sealed in a brazen pitcher; t
in which it was kept was only opened once a year (Paus. 8.47.5, Apollod. 2.7.3). Polemon
in passing that among the Chians a statue of Dionysus and among the Erythrians a statue of
is bound (frag. 90 Preller = FHG 3.146.90 = schol. Pind. 01. 7.95a). For a wide-ranging
see Meuli and Merkelbach (supra, n. 3) and Graf (supra, n. 3).
112. The statue of Artemis Soteira at Pellene saved the city from the Aetolians (Plut.
the Tegeans had a similar legend about the Gorgon's Lock (supra, n. 111). The protective pow
Trojan Palladium is well attested; see Verg. Aen. 2 passim, Konon FGrH 26 F 34, Apoll
their subsequent calendrical ceremonies was disseminated throughout Greece by

the Panhellenic oracles, those renowned storehouses of traditional knowledge to
which both cities and individuals turned for advice in times of illness or catastro
phe. Thus it is at the command of the Delphic Oracle that the Orchomenians
bind a statue of Actaeon with iron chains, while the Syedrans follow the com
mands of the Clarian Apollo when they similarly treat an effigy of Ares. The
statues of the Spartan Pausanias and the fifty youths of Siris are set up upon the
advice of Delphi. The burial of the Gauls and Greeks at Rome as "living effigies"
is likewise performed in times of crisis and at the behest of the Sibylline Books.
The stipulations for the use of xoXoooot dolls in the Cyrenean decree also
purport to be the words of the Delphian Apollo, although the slapdash assembly
of that fourth-century document from obviously disparate sources greatly vitiates
the credibility of this claim. The Cyrenean code is most interesting, because it is
the sole testimony to the early existence of ritual handbooks to which one could
turn when a house was haunted or when any other form of pollution or supernatu
ral discomfort existed. We know that such documents existed (unfortunately not
in durable media) at Athens and elsewhere, but the great bulk of them and even
the nature of their contents are unknown to us.113 In Assyria, a number of such
handbooks have survived, primarily because of the antiquarian interests of the
seventh-century-B.c. monarch Assurbanipal, who sent his secretaries far and
wide to collect, copy, and study all the ritual handbooks that they could find.
One can only wonder what our picture of Greek apotropaic ritual would look
like if some similarly motivated individual had performed such a favor for the
Greek states and inscribed them on more durable materials.
Despite the many broad correspondences between the Greek, Near Eastern,
and Egyptian rites discussed in this survey, the scantiness of the evidence pre
vents us from knowing whether such parallels arise independently in different
cultures as part of a universal human response to danger or impending disaster
(e.g., the tying of knots and the manipulation of effigies are forms of magic
known to many cultures), or whether these similarities result from a process of
cultural borrowing from the society that first "invented" the rituals (e.g., as
seems the case in several forms of Hittite rituals, which were probably modeled
on Babylonian or Hurrian antecedents). It is, nevertheless, a priori more likely
that the Greeks would imitate in matters of ritual the more advanced and cultur
ally sophisticated societies of the East, from which they indisputably borrowed
their alphabet and so much of their artistic and literary esthetic.114 As I have
argued elsewhere, however, it is impossible in most cases to know exactly when
113. Tresp (supra, n. 6) collects all the fragments of what was probably (at least at Athens) a
considerable corpus of materials.
114. For the "Orientalizing period" of Greek art, see the general summary in J. Boardman,
The Greeks Overseas2 (London, 1980) 54-85. For similar influence on Greek philosophy and litera
ture, see M. L. West, Early Greek Philosophy and the Orient (Oxford, 1971); idem, "The Rise of the
Greek Epic," JHS 108 (1988) 169-71. Burkert (supra, n. 57) demonstrates that a similar and perva
sive influence was being exercised in the area of magical/religious ritual.
such influence occurred,15 and the best one can do is to assume some
influence at an early date, without pressing the unwilling evidence furth
I would argue that the Assyrian ghost-banning rites in particular we
bly absorbed into some areas of Greece rather early on, as is manifested
apparently traditional use of xokoooot in the Argolid, Thera, and Cyr
passage discussed above from the Cyrenean lex sacra just may, in fact, h
an Assyrian ritual text as its source."6 The process of borrowing was
edly aided, as Burkert suggests, by the existence of wandering professio
the ivXacywyot whom the Spartans summoned to ban the ghost of Paus
well as by the existence of Panhellenic oracles, which seem to play a co
role in ordering the implementation of these types of rituals and may have
as a clearinghouse of sorts for "religious technology,""' just as they wer
tories of geographical and ethnological information useful to petition
ested in founding colonies.18 The origin, however, of the Greek tradi
binding Ares and foreign armies in effigy is much more difficult to ass
seems to be related in a much more general and diffuse way to myths an
found in Egyptian, Assyrian, and Hittite culture. A basic stumbling b
mains: the relatively small distances that separate these ancient cultur
near the ancient Mediterranean Basin allow for the possibility of an
number of cross-borrowings on all sides-a situation very similar, in fact
fearful and treacherous state in some Greek manuscript traditions that
gists call contamination. It is in just such a light that we should also exa
corpus of "voodoo dolls" collected in the Appendix; although I hav
labeled them Greek, Etruscan, and Roman according to the immediate
contexts of their discovery, their long and complex cultural heritage per
for a more general appellation: perhaps "Mediterranean," as it is th
inland sea that facilitated the spread, growth, and continued "cross-fert
of these curious and widespread artifacts.
The Center for Hellenic Studies

and The University of Chicago
115. Faraone (supra, n. 35: 275-80) and "Aphrodite's KEYTOX and Apples for
Aphrodisiacs in Early Greek Myth and Ritual," Phoenix 44 (1990) 239-43.
116. Aside from the close similarities with regard to the subject matter of the Cyre
Assyrian rituals (supra, nn. 57, 58, 63, 64, 67), the actual literary form of the ghost-banning
of the lex sacra may depend ultimately on some Assyrian text, especially the peculiar combi
conditional statement indicating the distress of the victim, who is described in the thi
singular (e.g., "if a visitant is sent against the house" or "if he knows his name"; cf. M
dead persons meet with a man"), followed by directions for the appropriate ritual in th
person singular, which in Assyrian texts are addressed to the professional magician, and
account for the hitherto unexplained second-person-singular verb in the Cyrenaean text:
have done the customary things" (supra, n. 59).
117. Burkert (supra, n. 57).
118. See W. G. Forrest, "Colonisation and the Rise of Delphi," Historia 6 (1957
Graham (supra, n. 54); R. Drews, "The Earliest Greek Settlements on the Black Se
(1976) 18-31.
APPENDIX: SURVEY OF GREEK, ETRUSCAN AND ROMAN "VOODOO

DOLLS"
The following survey is greatly indebted to the earlier works of C. Dugas, BCH 39 (1915)
413-23; L. Mariani, Ausonia (1910) 39-47; K. Preisendanz, APF 11 (1935) 163-64; and
D. R. Jordan in Praktika of the 12th International Congress of Classical Archaeology,
Athens, Sept. 4-10, 1983 (Athens, 1988) 273-77. I have also benefited from my private
conversations and correspondence with J. Winkler, W. Brashear, D. R. Jordan, R.
Kotansky, F. Maltomini, and H. S. Versnel.
With regard to these artifacts, I limit my definition of "voodoo doll" to figurines that
display two or more of the following characteristics:
1. The arms (and occasionally the legs) have been bound, or twisted behind the
back as if they were bound;
2. Nails have been hammered into the doll;
3. The head, the feet, and occasionally the entire upper torso have been twisted
unnaturally around, so that the chin and the toes point in the same direction
as the buttocks;
4. The doll has been deposited in a container with a tight-fitting cover;
5. Some part of the doll has been inscribed with the name of the victim;
6. It was discovered in a grave, a sanctuary, or a body of water.
I have also included here some drawings of bound figures on defixiones (e.g.,
nos. 23, 24), which, although they are not figurines, nonetheless serve the same
function.
ATTICA
1. Lead male figurine (18 cm tall) from a grave on Odos Panepistimiou in Athens in
disturbed fifth-century-B.c. context. Its palms are pressed together behind its back, an
there seems to be some indications of binding either by a wire or (less likely) by a nail. Its
head, with its carefully modeled hair and features, faces the same direction as the but
tocks. (0. Alexandri, AD 27 B.1 Chronika [1972] 75, grave no. 32, pl. 57.)
2. Lead male figurine (11.5 cm tall) of unknown date found in the bed of the Ilissu
River. Its hands are twisted behind its back and seem to be deliberately covering its anus
Its (bearded?) head faces the same direction as the buttocks. (Fig. 6; Ath. Nat. Mus. inv.
no. 7877; P.-F. Perdrizet, "Les pieds ou les genoux au rebours," Melusine 9 [1898-99]
194-96.)
3. Lead figurine (8 cm tall) of a (bearded?) male; exact provenance and date un
known. Its mittenlike hands are crossed behind its back as if bound, and its head faces the
same direction as the buttocks. Its feet appear to have been modeled facing forward in a
normal fashion and then roughly pinched and twisted about (P1. 1; Ath. Nat. Mus. inv.
no. 10807).
4. Lead male figurine (11 cm tall) of "high artistic quality" found in Athens along
Odos Eolou. Its hands are bound behind its back, and it was discovered in a covered box
formed from two folded sheets of lead. Dates "on stylistic grounds" to the fourth century
B.C. or later. (F. Cumont, "Une figurine grecque d'envofitement," CRAI [1913] 412-21.)
5. Lead male figurine (6 cm tall) from a grave in the Ceramicus. It has exag
genitals, and its hands are twisted behind its back as if bound. A single name is in
its right leg. Found in an oval-shaped covered box formed from two sheets of
lead; the interior of the lid is inscribed with nine names; ca. 400 B.C. (J. Trumpf, "F
und Rachepuppe," MDAI[A] 73 [1958] 94-102, pls. 71.3-4, 72.1-2; D. R. Jord
Archaeological Evidence for the Practice of Magic in Classical Athens," in Prakti
12th International Congress of Classical Archaeology [Athens, 1988] 273-77.)
6. Three lead male figurines (6 cm tall) found in a grave in the Ceramicus
meters from the grave that contained no. 5, above. Each had apparently been
inside one of the three inscribed covered lead boxes that were discovered in
grave. One has exaggerated genitals, and two have their hands bound behind the
All of the boxes and two of the dolls are inscribed with names. They have been
about 400 B.C. (PI. 5; Keiramikos Museum inv. no. KER 8805; B. Schlorb-V
"Eridanos: Necropole," MDAI[A] 81 [1966] 38, pl. 51.1; J. Boardman and D
Greek Burial Customs [Ithaca, 1971] 217, pl. 46; D. R. Jordan, loc. cit. [supra,
7. Lead male figurine (6 cm tall) from a grave in Attica. Its head has been
hacked off, and its feet and hands are violently twisted and bound behind its
lead bands. Its stomach is pierced with two iron nails. It has been dated to th
century B.C. (Fig. 7; R. Wiinsch, "Eine antike Rachepuppe," Philologus 61 [190
[In 1857 S. A. Koumanoudis reported seeing a terra-cotta figurine in Athe
scribed with the name "Noema": see Pape-Benseler, s.v. Noema; and 0. Masson
Benseleriana VIII," ZPE 42 (1981) 199. I thank Professor Jordan for this refer
ARCADIA
8. Cast-bronze male figurine (16.2 cm tall) from Tegea, now in Berlin, dated by
Lamb and Neugebauer to the early seventh century B.C. Its head and feet are twiste
completely around, and its hands are raised like a ballroom dancer. (Staatliche Museen
Berlin, Antiquarium inv. no. 6405; W. S. Lamb, "Arcadian Bronze Statuettes," BSA 27
[1925-26] 133-34; K. A. Neugebauer, Katalog der statuarischen Bronzen im Antiquari
1 [Berlin, 1931] 66-67, no. 163.)
9. Five bronze figurines (6-8 cm tall) from Alonistena dated by Neugebauer to t
Archaic period. All were probably meant to portray males, but male genitalia are clear
portrayed on only one. In each case the hands are bound behind the back and the feet
bound together. The shortest has its head facing the direction of its buttocks, while
heads of the other four are twisted unnaturally ninety degrees to the side. (Pls. 2-4a; Ath
Nat. Mus. inv. no. 15130; Neugebauer [supra, no. 8] 67 n. 2.)
CEPHALONIA
10. Bronze male figurine (2.9 cm tall) from Sime, now in the British Museum,
described by Neugebauer as "nach-archaisch." Its feet and one of its arms are twiste
completely around, and its head is wrenched unnaturally to the left, so that the chin l
directly over the shoulder (Fig. 3; H. B. Walters, BMCBronzes [London, 1899] no. 216
C. Q. Giglioli, "Curiosita archeologiche," SE 3 [1929] 529-30, fig. 1; Neugebauer [supr
no. 8] 66.)
DELOS
11. Four cast-bronze male figurines (6-9 cm tall) found together in a house
Hellenistic date near the agora. Three have their hands bound behind their backs, a
two are ithyphallic (Pls. 4b, 6-9; Musee du Delos inv. nos. A 2157-60; C. Duga
"Figurines d'envoftement trouvees a Delos," BCH 39 [1915] 413-23.)
12. Four figurines (6-7 cm tall), two male and two female, each roughly carved from
rectangular slab of lead; the males are each pierced through the eyes, ears, and mouth w
iron nails; their right hands seem to cover or hold the head of the nail that penetrates t
mouth. The females are simply bound around the neck with thick collars. They were
covered buried together in the retaining wall of the sanctuary of Zeus Hypsistos and dat
the first century B.C. (P1. 10; Musee du Delos inv. nos. 3787-90; A. Plassart, Les sanctuair
et les cultes du mont Cynthe, Exploration Archeologique de Delos 11 [Paris 1928] 292-
CRETE
13. Cast-bronze male figurine (14 cm tall) with its hands twisted behind the back as if
bound. Date and exact provenance unknown. (L. Mariani, "Osservazioni intorno all
statuette plumbee sovanesi," Ausonia [1910] 44 n. 5, fig. 12.)
14. Bronze male figurine (height not given) apparently cast without shoulders and
head, with its hands twisted behind the back as if bound. Date and exact provenance
unknown. (Mariani [supra, no. 13] 44, fig. 13.)
EUBOEA
15. Flattened lead "gingerbread man" (9 cm tall, 5 cm wide) from Carystus, whose
sex is not indicated but which is inscribed on the chest with two fourth-century-B.c. texts
that aim at binding a woman. There appears to be some attempt to depict a rope or
chain(?) attaching the hands to the top of the head. (P1. 11; L. Robert, Collectio
Froehner 1 [Paris, 1936] 17-18 no. 130, pl. viii.13; SGD 64.)
SICILY AND ITALY

16. Crude lead figurine (height not given) from Sicily of a demonic-looking man with
arms upraised like a ballroom dancer (cf. nos. 8, 17) and with head and feet twisted
completely around. There are ten names inscribed upon his chest; the letter-forms date t
the late fifth or early fourth century B.C. (D. Comparetti, "Figurina plumbea iscritta," i
Miscellanea di studi sicelioti ed italioti in onore di Paolo Orsi, ASSO 28 [Catania, 1921]
194-97; SGD 122.)
17. Bronze figurine (height not given) from Sicilian Naxos "close in style to" the
bronze figurine from Tegea, no. 8 above. (D. E. L. Haynes, "A Group of East Gree
Bronzes," JHS 72 [1952] 75; no photograph or drawing.)
18. Two lead figurines, one male (18 cm tall) and one female (16 cm tall), from a
sixth-century-B.c. Etruscan tomb in Sovana. They are of high artistic quality and are
depicted with arms twisted behind their backs as if bound. A masculine name is inscribe
on the left thigh of the male and a feminine one on the left hip of the female in an
Etruscan alphabet that dates to the late fourth or early third century B.C. (Pls. 12, 13; B
Nogara, "Due statuette etrusche di piombo trovate recentemente a Sovana," Auson
[1910] 31-39; Mariani [supra, no. 13] 39-47.)
19. Bronze male statuette (9 cm tall) from Citta di Castello; it seems to depi
Etruscan Heracles wearing a wolf- or dogskin hat and leaning on a knotty club. Its
completely twisted about, and its legs are broken off below the knees (Fig. 5; A.
"Curiosita Archeologiche," SE 1 [1927] 475-76, pl. 72a).
20. Eight terra-cotta figurines (7-9 cm tall), four males and four females, disc
in a pile of cremation ashes and bones in a grave in Puteoli; each is inscribed in
with a different name. They date no later than the first century B.C. (Fig. 4; A. S
and D. Vaglieri, "Di alcune figurine di terra cruda, sulle quali si leggono nomi g
NSA [1897] 529-34, figs. 1-8; DT 200-207.)
21. Torso of a lead male figure in relief (6 cm tall) of unknown provenance
Museo Nazionale Romano nelle Terme Diocleziane. It has exaggerated nipples,
and genitals (perhaps at one time it was ithyphallic), and it is inscribed in Latin w
masculine name. The inscription dates to the first century A.D. The arms, now bot
seem to have been twisted behind the back. There are the remains of an iron nai
right shoulder near the neck. (Mariani [supra, no. 13] 42-44, fig. 11; K. Preisend
"Die griechischen und lateinischen Zaubertafeln," APF 11 [1935] 164.)
22. Terra-cotta bust of a couple and young child (11.6 cm tall, 10 cm wide) f
second-century-A.D. cemetery beneath the Basilica of San Sebastiano in Rome. A
ently a common sort of ex voto, it was inscribed in Latin prior to firing; each fig
been labeled with a name, and the back is inscribed with what seems to be a bindin
(Museo Nat. Rom. inv. no. 82684; M. Della Corte, "Sigillum-Devotio," RAAN 18 [1
1-13; H. Solin, Eine neue Fluchtafel aus Ostia, Commentationes Humanarum Litt
42.3 [Helsinki, 1968] 29 no. 33.)
23. An elaborate late-antique lead defixio (text in Greek) from Rome inscribed
four bound figures on one side and three on the other. In each case the legs are
beneath the body and the arms are crossed over the belly. In three of the drawings,
seem to be indicated by horizontal lines drawn through the point at which the arms
are crossed. (R. Wiinsch, Sethianische Verfluchungstafeln aus Rom [Leipzig, 1900]
no. 20; DT 159.)
24. Another Greek lead defixio from the same cache as no. 23, above. It depic
demons binding a headless victim in chains. (Wiinsch [supra, no. 23] 40-41 no. 2
167.)
NORTH AFRICA
25. Two elongated lead figurines, one male (14 cm tall) and one female (11.7 cm tall),
from a sewer tunnel in Volubilis (Morocco) dating to the Roman period. Their legs are
slightly twisted, and the hands of the male are crossed over its belly; the hands of the
female are missing. (G. Souville, "Volubilis: Le collecteur principal du decumanus maxi
mus," BAM 2 [1957] 175-83 nos. 36, 37, pl. ii.3, 4.)
26. Lead figurine (13 cm tall) similar to the preceding pair but of slightly better
execution. It was found in Valentia Banasa, a Roman colony in Morocco approximately
75 km from Volubilis, and dates to the Roman period. It is now in the museum at Rabat.
(R. Thouvenot, Une colonie romaine de la Mauretanie Tingitane: Valentia Banasa, Publi
cations de l'Institut des Hautes Etudes Marocaines 36 [Paris 1941] 91.)
26a. A late-antique lead defixio from Carthage inscribed (like nos. 23, 24) with draw
ings of bound men, presumably depicting the charioteers who are cursed in the accompany
ing text. (D. R. Jordan, "New Defixiones from Carthage," in J. H. Humphrey, ed., The
Circus and a Byzantine Cemetery at Carthage 1 [Ann Arbor, 1988] 117-34)
Egypt
27. Unbaked clay female figurine (9 cm tall) from "Moyenne Egypte" and probably
dating to the fourth century A.D. It is in a kneeling position, has its hands bound behind its
back, and is pierced with thirteen bronze nails through the eyes, ears, mouth, top of the
head, palms of each hand, soles of each foot, solar plexus, anus, and vagina. It had been
wrapped in a lead tablet inscribed with an erotic defixio text (in Greek) and then set
within a crudely fashioned clay pot. (P. du Bourguet, "Ensemble magique de la periode
romaine en Egypte," RLouvre [1975] 255-57; idem, "Une ancetre des figurines d'envofte
ment perches d'aiguilles, avec ses complements magiques au Musee du Louvre," in Livre
du Centenaire de l'Institutfrangais d'Archeologie orientate, Melanges de 1'Institut franqais
d'Arch6ologie orientale 104 [1980] 225-38; SGD 152.)
[A recent catalogue for a public auction sponsored by Laurin, Guilloux, Buffetaud,
and Tailleur on October 27-28, 1987, advertised the sale of two figurines, one terra-cotta
(cat. no. 259; height 12.5 cm) and one wax (cat. no. 260; height 12.7 cm), that are nearly
identical to the Louvre figurine, no. 27, above. Both are of Egyptian provenance, de
picted kneeling with hands bound behind the back, and pierced with thirteen needles. The
wax figure is accompanied by a standing wax male figurine, which is unbound. I thank Dr.
W. Brashear for bringing this catalogue to my attention.]
28. Two wax figurines (height not given, but Prof. F. Maltomini informs me that they
are 7 cm tall), a male and a female entwined in the act of lovemaking, found in Upper
Egypt north of Assiut wrapped in an inscribed fifth-century-AD papyrus and wedged
inside an uninscribed clay pot. (D. Wortmann, "Neue magische Texte," BJ 168 [1968] 85
102 no. 4, pls. 8, 9.)
28a. Two wax figurines (10.8 cm tall), a male and a female entwined in lovemaking.
Origin unknown, but given their apparent similarities with no. 28 I would be surprised if
they were not from Egypt and did not date to the fourth or fifth century A.D. They are now
in Munich, purchased with a covered clay jug and a much-faded papyrus said to be
inscribed in Coptic, but Dr. Brashear informs me that it is indeed Greek, and that he will
publish the text in Studien zur altagyptischen Kultur. (Staatliche Sammlung Agyptischer
Kunst inv. nos. AS G791-93; D. Wildung, "Requisiten eines Liebeszauber" MJb [1988]
210-11.)
29. Two wax figurines (height not given), a male and a female, found in Hawara near
a Greek erotic defixio. The male has an ass's head (= Seth?) and exaggerated genitals; the
female has her hands bound behind her back and a hairstyle very similar to that of figurine
no. 27, above (C. C. Edgar, "A Love Charm from the Fayoum," BSRAA 21 [1925] 42
47; D. R. Jordan, "A Love Charm with Verses," ZPE 72 [1988] 247 n. 5; SGD 153.)
30. Mud figurine (height not given) tied to a rolled-up fourth-century-A.D. Greek
erotic charm, PGM LXVIII. (A. S. Hunt, "An Incantation in the Ashmolean Museum,"
JEA 15 [1929] 155-57, pl. xxxi.1.)
31. Broken wax figurine (10 cm tall) of unknown provenance molded around a small,
rolled papyrus, which is undoubtedly inscribed with a victim's name and a spell. Its navel
has some human hair embedded in the wax. (BM inv. no. 37918; E. A. W. Budge, British
Museum: A Guide to the ... Egyptian Rooms [London, 1922] 51 no. 77; M. J
"Wax in Egyptian Magic and Symbolism," OMRL 63 [1983] 12).
PALESTINE
[Jordan informs me per litteras that the loose clayey dirt found within on
eighty unpublished defixiones from an early Christian necropolis in Tyre (SGD
probably the remains of a clay figure.]
32. Sixteen lead slab figurines (of various heights) found in Tell Sandahann
the sanctuary of an unidentified deity, together with Greek and Hebrew inscrip
limestone. Their hands and feet are bound together with bronze wire. They dat
ing to archaeological context no later than the first century B.c. (Wiinsch [supra
31; DT 17.)
ASIA MINOR
33. Bronze female figurine (7.3 cm tall) with head completely twisted about found
near Smyrna(?) amongst a large number of small bronze human and animal figurines.
(BM inv. no. 1951.3-29.1; D. E. L. Haynes, [supra, no. 17] 75 no. 9, pl. IIc).
BLACK SEA AREA

34. Lead box (11 cm x 5 cm) with cover from a grave in Panticapaeum. The floor of
the box was once inscribed with a drawing (now gone) of a person bound with ropes. Two
narrow lead scrolls were discovered within the box, which name at least six individuals.
Jordan (per litteras) dates the letter-forms to the fourth century B.C. or later. On the
exterior of the box and its lid four bands were inscribed as if depicting ropes used to bind
the box shut. (K. Preisendanz, "Die stettiner Sargzauber," FuF 6 [1930] 149, fig. 2).
FARAONE FIGURES 1 & 2
1G29
Derchain, Le papyrus Salt 82
Figure 2. Egyptian clay disks

und
und Nubien
Nubien [Lund,
[Lund, 1941]
1941] 104
104
FIGURES 3 & 4 FARAONE
Figure 3. Bronze figurine from Cephalonia. (After

Mdlusine 9 [1898-99] 104, fig. 1)
Figure 4. Clay figurines from Puteoli. (After NS

[1897] 529-30, figs. 1,2)
FARAONE FIGURE 5
Figure 5. Bronze figurin
FIGURES 6 & 7 FARAONE
4!y/
Figure 6. Lead figurine from Athens. (After

Melusine 9 [1898-99] 104, fig. 2)
Figure 7. Decapitated lead figurine from Athens. (After Philologus

61 [1902] 37)
FARAONE PLATE 1
: .~ ~ ..
./
L~~~~~~~~~~~
Museum~~, AthensC
PLATES 2 & 3 FARAONE
Plate 2. Bron
Alonistena (A
(Photo: Natio
Museum, At
Alonistena (Athens NM 15
(Photo: National Archaeolo
Museum, Athens)
FARAONE PLATE 4
Plate 4 a, Bro
(Athens NM 1
Archaeologic
..
PLATE 5 FARAONE
.....*' ..... . .
? . r S . '~' fi .Z; g . . . :
Plate 5. Lead figurines in inscribed lead c

(ca 400:::
(ca. B.C.).
400 B.c.).
FARAONE PLATE 6
Plate 6a,b. Bronze fig

Delos inv. no. A 2159)
PLATE 7 FARAONE
*" :"~'z:~ .
Plt ab Boz iurn on n eo na os fHllnsiatMsed

Delos inv. no. A 2157)(front and side'views)
Plate 7a,b. Bronze f

Delos inv. no. A 2157
FARAONE PLATE 8
Plate 8a,b. Bronze f

Delos inv. no. A 2158
PLATE 9 FARAONE
* ~ ~ ~~~~~
?._
?~ ~ ~ ~ ~ ~ ~.~ ~ ~ ~ ~ ~ ~ ~. i,
ll.
.. ..
. _L~~~~~~~~~
_!~~~~~~~~~~~~~~~~~~~
_lt ab rnefgrn on nDlsi os fHleitcdt Msed

D6 o n . - ,no. A;10 fotadsd iw)
FARAONE PLATES 10 & 11
Plate 10. Four lead figurines dated

hidden in the outside of the retain
Hypsistos on Delos (Musee du De
Plate 11. Flattened lead "gingerbr
PLATES 12 & 13 FARAONE
Plate 12. Two ea

figurines from an
Sovana (front vi
Plate 13. Two early Hellenisti

figurines from an Etruscan tom
Sovana (rear view).

1991 Binding and Burying With Cover Page v2

Uploaded by

Copyright:

Available Formats

1991 Binding and Burying With Cover Page v2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1991 Binding and Burying With Cover Page v2

Uploaded by

Copyright:

Available Formats

Accelerat ing t he world's research.

Binding and Burying the Forces of

Cite this paper Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers Download a PDF Pack of t he best relat ed papers 

Binding and Burying the Forces of Evil:

? 1991 BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIA

source of information-the literary, epigraphical, and archaeological evidence

BINDING ARES IN GREEK MYTH AND RITUAL

Ares had to endure it when strong Ephialtes and Otos,

Pamphylians of Syedra inhabiting a land in the bordering territory of the

Eiloxe OVQ6Lo g otog ETl T r XOovL xeXXLTaL "AQSg,

[Inscribed] on the base of an Ares [statue] that lies buried in Thrace:

The historian Photius refers to is Olympiodorus, an occasionally tendentious

also be suspected as fanciful embroideries drawn from the author's

Opposite this temple [sc. of Poseidon] is an ancient image of Enyal

BOUND STATUES IN EGYPT AND THE NEAR EAST: SOME COMPARATIVE

In Egypt "voodoo dolls" were deployed at the highest levels of stat

empreintes magiques de Gizeh et les Morts Dangereux," MDAI(Kairo) 16 (1958) 252-7

inscribed with the names of individual Egyptians who are described

31. Posener, Princes etpays (supra, n. 29) 15-20.

One rather tempting explanation is the "trickle-down effect": Junker, struck by

[You fashion] seven figurines of asphalt, seven of sheep tallow, and

31 That sorcery, witchcraft, magic, [and] evil spells, whether of a man

39. G. Meier, Die assyrische Beschworungssammlung Maqlu, Archiv fur Orientforschung 2

13 If a [dead] ghost appears to a living person [ . . . a figurine] of clay

. . .knead clay, make a figurine

In contrast to the treatment of images of demons or sorcerers in the

46. Of the 88 Assyrian ghost-banning recipes collected by Scurlock, only 17 (MMG

CYRENEAN KOAOXZOI AND EARLY GREEK GHOST-BANNING RITUALS

Two fourth-century-B.c. inscriptions from Cyrene attest to a similar use of

images, himself, his seed, and his property" (trans. A. J. Graham).

Visitant sent [by spells] from afar.

If a visitant is sent against the house [and] if he [sc. the househo

This ritual is designed to prevent the attack of a LXEoLOg eraxto;g, a hostile

female) of wood or mud statuettes (xoXo0aoi) ;65 after providing a ritual

of seventh-century-B.c. date discovered on the island of Thera (the so-called

Pausanias within Athena's temple. A similar situation seems to have occur

qvXXaycyo yoi tVE y6 e qEg e tv OexTaXik ovTc xxako ievol, oittveg

Some sorcerers in Thessaly are in this way called vuXayoyoi, who w

The comment is doubly instructive. First of all, it explains why Heracles

75. Justin, Pompei Trogi Epit. 20.2.

us with a Plutarchan fragment that mentions a fact Thucydides apparently left

tional split in the ancient Greek understanding of the exact operations

xoXoooot as substitutes for reburial.86 It is in fact interesting to note th

PRIVATE "VOODOO DOLLS" IN THE GREEK AND ROMAN WOR

There is widespread evidence that throughout the Greek and Roman

personal Greek or Etruscan names, we might understand in the pairing

A woman is represented as carrying a white boy asleep on her right arm;

Etruscans and Romans as well) were accustomed to bind or bury effigies

their subsequent calendrical ceremonies was disseminated throughout Greece by

The Center for Hellenic Studies

APPENDIX: SURVEY OF GREEK, ETRUSCAN AND ROMAN "VOODOO

SICILY AND ITALY

BLACK SEA AREA

Derchain, Le papyrus Salt 82

Figure 2. Egyptian clay disks

Figure 3. Bronze figurine from Cephalonia. (After

Figure 4. Clay figurines from Puteoli. (After NS

Figure 5. Bronze figurin