1 Introduction

In this paper, we publish a lead circus curse tablet written in Jewish Palestinian Aramaic. The tablet should be viewed in light of the widespread use of so-called defixiones agonisticae (agonistic binding spells) in chariot races during the Roman Byzantine Period. Such curse tablets were prepared against charioteers, horses, and circus factions.¹ They consist of thin lead sheets inscribed with aggressive incantations that include the names of the victims and which seek their failure in the races. The tablets were folded and buried or deposited at various points around the circus or in nearby cemeteries or wells.

Most of the documented circus curse tablets that survive are inscribed in Greek or Latin. The tablet under discussion, however, was written in Jewish Palestinian Aramaic using the Aramaic square script. This strongly suggests that the person who produced/inscribed the tablet was Jewish.² This is the first known example of a Jewish Aramaic circus curse and one of a very small number of Jewish curse tablets (see below 3. Circus Magic in Jewish Magical Literature).

2 Archaeological Context and Description of the Tablet

Our tablet (Princeton Art Museum excavation no. 3608-I57; Princeton University Art Museum accession no. 2011–143) was found in 1935 during one of the excavations conducted by W.A. Campbell in Antioch on the Orontes (Syria) and is datable to the fifth or sixth century CE.³ It was deposited in a drain near the meta prima (first turning-post).⁴ Its placement at this point is significant not just because the turning-posts represent the most dangerous part of the hippodrome but also because the drains were believed to provide a connection with the netherworld.⁵

The lead tablet itself measures 20.0 cm wide by 9.8 cm high. It was found rolled up seven times and pierced with a nail.⁶ The nail damaged the upper part of the tablet, in particular the left edge, from which a fragment is missing. The left bottom part of the tablet is also missing.

The upper part of the tablet shows three heads in frontal view (marked as HEADS 1–3). These consist of two small heads with long hair on either side of a larger oblong head with curls or a head cover. Heads depicted on circus tablets usually represent the charioteers who are the targets of the curse.⁷ The closest parallel to our drawing is PGM XXXV, a papyrus amulet for victory that includes a similar composition featuring three heads. This papyrus also includes two proper names; these are written from left to right between the first and the second head and have been identified as the names of the amulet’s beneficiary.⁸ In our tablet, the nail was driven into the upper half of the rolled tablet. If the three faces represent the intended victims of the curse, then it is possible that this was done purposefully in order to pierce their faces with the nail and thereby achieve the total destruction of these three individuals. That said, it seems more likely that the faces represent angels or other heavenly forces. This interpretation is based on the word bšmh (‘in the name of’) which was inscribed next to HEAD 1 and seems to relate to the drawings.

3 Circus Magic in Jewish Magical Literature

Jewish magical literature preserves two recipes for chariot races involving the use of lamellae (thin metal sheets).⁹ The most prominent example is to be found in Sefer ha-Razim (The Book of Mysteries), a Jewish magical book written in Hebrew in Palestine or Egypt sometime in Late Antiquity. The recipe in question reads as follows:

If you wish to race horses (even) when they are exhausted, so that they will not stumble in their running, that they will be swift as the wind, and the foot of no living thing will pass them, and they will win popularity in their running, take a silver lamella and write upon it the names of the horses and the names of the angels and the name of the prince who is over them and say: I adjure you, angels of running who run amid the stars, that you gird with strength and courage the horses that N is racing and his charioteer (hnywkws) who is racing them. Let them run and not become weary nor stumble. Let them run and be swift as an eagle. Let no animals stand before them, and let no other magic or witchcraft affect them. Take the lamella and conceal it in the racing lane (of the one) you wish to win.¹⁰

This recipe reflects a close acquaintance with the practice of producing circus curse tablets. Certain details in the magical instructions, such as inscribing the names of the charioteers and hiding the tablet beneath the racing lane, suggest that the author of this recipe in Sefer ha-Razim was familiar with different elements of this specific practice. The Greek term for charioteer, ἡνίοχος, which appears in transliterated form (hnywkws) in this text, also shows that the author was aware of the professional jargon of the circus.¹¹

Another Jewish recipe for chariot races is documented in a text related to the magical sword literature.¹² Although this text has been translated and discussed by a few scholars in the past, we propose a new translation of it here that is based on a new reading as well as findings from our Aramaic circus curse tablet:

And if you wish to race (horses) well, write the sword (i.e., certain magical words) on a sheet of papyrus,¹³ and put wine in a new clay (vessel), and they (the charioteers) shall drink and wash their faces, and they will win. If you wish to break him (the horse/charioteer), write the sword on a copper¹⁴ lamella and bury¹⁵ it in qmṭwn, and they will be broken.¹⁶

This recipe is comprised of two potential magical manipulations of the race. The first, which involves drinking and washing the charioteer’s face with wine, is meant for gaining victory. The second is designed to destroy (literally ‘to break’) the rival in the race, that is, to result in his defeat. In agreement with the recipe from Sefer ha-Razim cited above, it involves the use of an inscribed lamella that should be buried in a certain spot: qmṭwn. This word has intrigued scholars, as it is unparalleled in Rabbinic literature and in other contemporary Hebrew texts. Margalioth suggested correcting the reading to qmpwn, a word documented in Rabbinic literature with the meaning of ‘field’, ‘open area’, and usually explained as deriving from Latin campus.¹⁷ Weiss has shown that, in some cases, this word was specifically connected to chariot races and further suggested a different etymology for qmpwn, namely the Greek word καμπή—the turning-post at the hippodrome.¹⁸ Below, we will show that the word qmṭh, which occurs twice in our curse tablet, reflects the Greek word καμπτός. In light of this, we suggest that the word qmṭwn in the sword fragment is not corrupt, as previously claimed, but is an accusative form of the Greek word καμπτός (see further the commentary to block D). This fragment from the sword literature thus reflects professional knowledge of circus magic and preserves the tradition—popular among practitioners—of burying curse tablets near the turning-post of the hippodrome.

Another echo of circus magic is found in the magical composition Havdala de-Rabbi Aqiva. In this text, a spell for overturning curses and evil magic is accompanied by a poetic set of allegories for overturning the regular order of things. The relevant section starts as follows: ‘Their horse will be killed and break, the charioteer will fall, and the chariot will be overturned’ (12:13).¹⁹ As Harari has suggested, though this phrase is not explicitly designed to defeat one’s rivals in a chariot race, it might reflect a quotation from a lost Hebrew magical incantation for the purpose of victory in the hippodrome.²⁰

Archaeological evidence reflects the fact that the practice of writing amulets on lamellae was well-developed within Jewish tradition from the Early Byzantine period. More than one hundred Jewish Aramaic amulets datable from the fourth century to the beginning of the seventh century CE are known to date.²¹ Nearly all of these lamellae were written for apotropaic purposes, and they were commonly worn on their owner’s person in order to extend their protective function. An exception is a bronze lamella that was excavated from beneath the threshold of the ancient synagogue at Ḥorvat Marish in northern Israel. This lamella, datable to the beginning of the seventh century CE, was written with the aim of subjugating the townspeople to the amulet’s beneficiary. Naveh, who published this text, pointed out that, unlike most of the other bronze and copper lamellae, this artifact consisted of a relatively thick metal leaf, which may reflect an attempt to imitate a lead tablet.²² Another aggressive magical object known to scholars is a love spell written on a potsherd that was found in Ḥorvat Rimmon in southern Israel, datable to the fifth or sixth century CE.²³ However, both of these objects clearly do not form part of the group of lead curse tablets that were widespread in the Roman-Byzantine world.²⁴ The Aramaic circus text discussed here is thus the only documented example of a curse tablet in the Jewish tradition.

4 Antiochene Jews and Chariot Races

The Jewish community in Antioch has a long and well-documented history.²⁵ The Jews of Antioch attended chariot races and were dedicated supporters of the different factions. In his polemical anti-Jewish work Adversus Judaeos, John Chrysostom (d. 407) describes the Jews as active partisans (PG 48.848). Several other Late Antique sources suggest that the Jews in the Byzantine Empire were generally identified with the Blue faction,²⁶ while John Malalas, an Antiochian chronicler of the sixth century, gives a report about fighting that took place between the Green and the Blue factions at the end of the fifth century and which resulted in rioting, the desecration of Jewish graves, and the burning of synagogues.²⁷ The Aramaic tablet under discussion mentions the Blue faction several times, but the context is unclear, and it is not possible to state whether the tablet was written for or against the Blue faction.

In a few places in his writings, John Chrysostom also refers to the Jews as powerful exorcists and complains that their charms and amulets attracted Antiochian Christians to their synagogues for healing.²⁸ Although Chrysostom’s accusations should be approached with an appropriate degree of caution, they nonetheless suggest that some Antiochian Jews were involved in magical practices. They may also shed further light on the circumstances of the production of our Aramaic circus tablet, even suggesting the possibility that the tablet was ordered from a Jewish practitioner by non-Jewish clients in the very fashion criticised by Chrysostom.

5 Balaam and His She-Ass

Balaam and his she-ass, known from the Balaam narrative in Numbers 22–24, are referred to in line 14 of the tablet. This line invokes the angel that positioned itself before Balaam’s she-ass as Balaam was on his way to Balak the king of Moab. According to the biblical text, Balaam was a non-Israelite seer from Petor, located on the River in the land of the Ammonites.²⁹ He is said to have had knowledge of the Most High (Num. 24.16), and when asked by Balak to curse the invading Israelites, he blesses them instead (Num. 23.7).

Balaam’s she-ass plays an important role in this biblical text. As Balaam journeys to Moab on her back, it is she and not Balaam who sees the angel of God barring their way. Three times she shies away from the angel of God, and three times Balaam strikes her in order to spur her to continue on their way. After the third beating, the she-ass opens her mouth and speaks to Balaam (‘What have I done to you that you have beaten me these three times’, Num. 22.28 [JPS]). Following her speech, God opens Balaam’s eyes whereupon Balaam also sees the angel and acknowledges his sin. He is commanded by the angel to continue on his way to Moab and to tell Balak the words that God will put in his mouth. In three subsequent oracles (Numbers 23–24), Balaam blesses the Israelites three times.

Balaam is a complex and ambiguous figure in Jewish tradition. Already within the Hebrew Bible itself there are clear instances of the negative reception of his figure.³⁰ Both Num. 31.8 and Josh. 13.22 mention that he was killed by the Israelites, apparently because he led them astray (see Num. 31.16). According to Deut. 23.3–5 and Josh. 24.9–10, Balaam had the intention to curse the Israelites, but God turned the curse into a blessing. In Josh. 13.22, Balaam is referred to as a qōsem ‘diviner’, one of the professions in the list in Deut. 18.10–11 whose practitioners the Israelites should neither appoint nor consult.³¹

Interpretational problems and ambiguities within the Hebrew text of Numbers 22–24 as well as Balaam’s negative reception in the Hebrew Bible itself (outside of the Balaam cycle) furnish the basis for his predominantly negative reception in post-biblical Jewish literature.³² In this literature he is also associated with the practice of divination and magic, while a good example of his negative reception is to be found in Targum Pseudo-Jonathan. According to Pseudo-Jonathan, Balaam travelled to Balak with the intention of cursing the Israelites. In this way, the translator of the Aramaic Targum seeks to explain why, according to the Biblical Hebrew text, God was angry about Balaam’s leaving (Num. 22.5). Pseudo-Jonathan also states that Balaam was accompanied on his journey by Jannes and Jambres (TgPsJ Num. 22.22), who are described in the Targum as the chiefs of the magicians in Egypt (TgPsJ Exod. 1.15; 7.11; Balaam himself is also linked to Egypt in TgPsJ Exod. 9.21). In its description of Balaam’s death (Num. 31.8), Targum Pseudo-Jonathan describes how Balaam tried to escape death by using a magical word to enable him to fly away (thereby depicting him as a demon).³³ In this Targum, Balaam is also referred to as blʿm ršyʿʾ ‘Balaam the villain’ (TgPsJ Num. 23.9, 21; 24.25) and blʿm ḥybʾ ‘Balaam the sinner’ (TgPsJ Num. 23.10; Num. 31.8). This Targum also makes a connection between the name Balaam (blʿm) and the verb blʿ ‘to swallow’, as Balaam’s intention was to swallow the House of Israel according to Targum Pseudo-Jonathan (Num. 22.5).³⁴ In addition, Balaam is identified with Laban (lbn) the Aramean (Num. 22.5; 31.8), who is described as a deceiver in the biblical story about Jacob and Laban’s daughters Rachel and Leah (Genesis 29–31); according to later Jewish tradition, Laban was also a ‘villain’ (lbn ršyʿʾ). The traditions connecting the name Balaam to the verb blʿ and identifying Balaam with Laban are also found in TgPsJ Num. 22.5:

He (= Balak) sent messengers to Laban the Aramean, that is Balaam (for he sought to swallow the people of the House of Israel), the son of Beor, who acted foolishly from the greatness of his wisdom. He did not spare Israel, the descendants of the sons of his daughters; and his dwelling house was in Padan, that is Pethor, according to its name, “Interpreter of Dreams.” And it was built in Aram on the Euphrates, a land in which the population served and worshiped him—(and Balak sent) to call him, saying: Behold, a people has come out from Egypt, and behold, it obscures the view of this land and is encamped before me.³⁵

When we compare this translation with the biblical text of Num. 22.5, the differences become apparent:

And he (= Balak) sent messengers to Balaam son of Beor in Pethor, which is by the Euphrates (lit. ‘the River’), in the land of his kinsfolk, to invite him, saying, “There is a people that came out of Egypt; it hides the earth from view, and it is settled next to me.”

JPS

In contrast to the negative receptions of Balaam’s figure outlined above, there also exists an altogether positive depiction of the figure of Balaam in post-biblical literature (both in Jewish and in Christian sources). It is based on a Messianic interpretation of Num 24.17–19, and on vs. 17 in particular (Balaam’s last oracle).³⁶ In this passage, Balaam foresees the coming of a great leader: ‘A star shall proceed from Jacob, and a scepter shall rise from Israel’ (cf. also Gen. 49.10: ‘The scepter shall not depart from Judah, nor the ruler’s staff from between his feet’). In later times, these verses in Numbers 24 were interpreted as Balaam’s prediction of the arrival of the Messiah.

The figure of Balaam is clearly linked with divination and magic (see above) in the Hebrew Bible and in post-biblical Jewish literature, though Balaam does not appear very frequently in the Jewish magical tradition. The following instances are known to us.

First, the name Balaam is mentioned in one of the Greek magical papyri (PGM XXXV, l. 36). In line 36 of this text, the name Balaam (Βιλιάμ) forms part of an adjuration.³⁷ Remarkably, just as in our tablet, this text contains drawings of three heads (see also above 2. Archaeological Context and Description of the Tablet). The function of Balaam’s name in this particular context, however, as well as the connection between Balaam and the three heads is unclear.

Second, the she-ass of Balaam the diviner appears in a Syriac remedy for colic.³⁸ This Christian remedy includes a historiola that narrates an encounter between Solomon and a talking she-ass, a descendant of the she-ass of Balaam the diviner. Meetings between a human and a demon are often narrated in historiolae in Jewish magical texts.³⁹ Examples include the meetings between the prophet Elijah and Lilith as well as Rabbi Ḥanina ben Dosa and Agrat daughter of Maḥlat.⁴⁰ These parallels suggest that the she-ass in the Syriac text is a demon as well.⁴¹ According to Zellmann-Rohrer, there may have been an analogy at work between the stopping of the movement of the she-ass by the angel and the stopping of the motion of the bowels.⁴²

Third, as mentioned previously, the Jewish tradition (see above on Targum Pseudo-Jonathan) connects Balaam with the Egyptian magicians Jannes and Jambres.⁴³ This connection is also made in the magical treatise Havdala de-Rabbi Aqiva, in an incantation against sorcery (referred to above in 3. Circus Magic in Jewish Magical Literature):

View Full Size

Figure 1

Princeton Art Museum excavation no. 3608-I57

Citation: Aramaic Studies 21, 1 (2023) ; 10.1163/17455227-bja10041

• Download Figure
• Download figure as PowerPoint slide

Just as the (magical) acts of Joannis and Joanbris before Moses in Egypt did not come out so will also (the works) of people that practice sorcery, whisper (incantations), foretell the future, and give evil advice against us not come out. Their horse will be killed and crushed, the charioteer will fall, and the chariot will be overturned.⁴⁴

With respect to the interpretation of our tablet, it is significant that in later Jewish literature the figure of Balaam is linked with the story in Exodus 15 not just through the magicians Jannes and Jambres, but through the verb blʿ and the notion ‘to sink’ (on the verb ṭmʿ, which appears in our tablet and in several Aramaic translations of the Hebrew Bible, see the commentary to l. 4). This link between the two passages is also made in our tablet, which clearly evokes the story of Balaam, though the tablet places more emphasis on the angel that stood in the way of Balaam and his she-ass than it does on the figure of Balaam or his she-ass. There are several points to be made concerning the role of the angel, the she-ass, and Balaam in our text. First, the angel is in all likelihood invoked in the tablet in order to impede certain horses in the circus. Second, the text appears to draw an analogical connection between the circus horses, the horses of Pharaoh at the Reed Sea, and Balaam’s she-ass in that they all need to be stopped.⁴⁵

View Full Size

Figure 2

Reflectance Transformation Imaging (RTI) of Princeton Art Museum excavation no. 3608-I57

Citation: Aramaic Studies 21, 1 (2023) ; 10.1163/17455227-bja10041

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Figure 3

Hand-copy of Princeton Art Museum excavation no. 3608-I57

Citation: Aramaic Studies 21, 1 (2023) ; 10.1163/17455227-bja10041

• Download Figure
• Download figure as PowerPoint slide

6 The Inscription on the Tablet

The inscription was engraved superficially onto the surface of our lead tablet and is almost invisible to the naked eye. We have been able to decipher it thanks to photos taken with Reflectance Transformation Imaging (RTI).⁴⁶ The two upper lines of the inscription and the left fragment are heavily damaged, and the reading of these sections is highly uncertain. In addition, the lower left part is completely missing. The lines of the text curve downwards, especially towards the left edge of the tablet, while a straight margin characterises the right-hand side of the text. In line 7, it seems that the person who inscribed the tablet began incising before stopping after two words and tried to erase the letters he had written by abrasing the surface; he then inscribed the next line (l. 8) over line 7. Line 8 overlaps partially with line 9, and line 10 overlaps partially with line 11.

View Full Size

Figure 4

Script chart

Citation: Aramaic Studies 21, 1 (2023) ; 10.1163/17455227-bja10041

• Download Figure
• Download figure as PowerPoint slide

The reconstruction of the layout of the upper part of the tablet is complicated. A study of the strokes shows that HEADS 1–3 were drawn first; only then did the person who inscribed the tablet add the text. The heads separate the area that surrounds them into four small blocks of text that should be read from top to bottom. We have marked these segments as blocks A–D. Blocks B, C, and D appear to relate to the drawings to their right. In what follows, we ‘read’ the drawings together with the blocks of text: As in a picture book, we view the heads as ‘words’ in a continuous text.⁴⁷ The word bšmh, which appears in the area of block A to the right of HEAD 1, should be read as part of block B. This is clear from the size and direction of the letters, which does not correspond with other words in block A. The leftmost part of the tablet is poorly preserved which makes it hard to determine where block D begins. We interpret it as consisting of eight short lines beginning right under line 3. That said, it is equally possible that this block begins nearer to the bottom of the tablet, and that line 3 accordingly extends into another small ‘block’ of a few lines (see also the commentary to block D).

7 The Script

The tablet is written in the Aramaic square script (see Fig. 4). The letters’ height is 3–6 mm. The hand is cursive and somewhat sloppy. Some letters have the same form, such as b and k; d and r; h, ḥ, and q; w and y; s, final m, and sometimes q; and n and l in a few cases. At some points, it is virtually impossible to differentiate between these letters. In one case at least, the letter h was written in cursive fashion and looks like s. In this inscription, final letter forms of m and n are attested. In general, the words are separated by small blank spaces, but sometimes the letters are squeezed together without apparent blanks. Despite all of this, it seems that the inscription was written by an experienced person. Certain habits can be identified, such as the ligature of ʾ and l (gbrʾl, l. 5; mhddʾl, l. 6).

8 Linguistic and Orthographic Characteristics of the Text

The Aramaic text displays several characteristics of Jewish Palestinian Aramaic.

9 Transliteration of the Text⁶¹

1 ḣrṡṗẏ hṡṅ.ṁ ḣẇ.ṁ smk̇. .ṫṅṡḣ…[ ]nyṡ.[

2 q̇lrwk̇s hy..q̇ḥ ṗḣẏnyqs ḥ. .ṡplyh ḋ.sl[ ]ḣylṡ ..

3 [ ] ṭ̇ṁṡ prsy lṭ[ ]. ʾwlmps

Block A

[ ] k̇ṙ..[ ]

[ ].k̇ẏ.h pṁh yhwh

Block B

bšṁh (HEAD 1) dswsṫ[h]

dbnṭ ẏḣẇḣ lh

ṫṧbq̇ lwn ʿl̇

Block C

(HEAD 2) dswsth

hdn sr

bhdn

bṭlh

Block D

[ ].. q̇ṙṙ

[ ].m….

…..h

k…..r

(HEAD 3) sws[w]th ḋbn[ṭh]

.. [ ] .. qmṭh

[ ] ..lʾyh

w.[ ]kyh ṡ

4 l̇qwrṙ kdw ṭmʿwn ʾlẏn swsth ḋktbyn

5 ʾ{b}šbʾt bš mkʾʾl wbšm gbrʾl wbš rpʾl ḃš [ ]yh ṡ

6 wbšmh dhw hwh yhwq mhddʾl ṙmnʾl wswsth hm[ ].šmḣ

7 erased: {hln ṁmrtṅ ḋ|||}

8 ʾln šhmṙ.. [ʾ]ln swswth dbnṭh k̇lḥd w{h}ḃšṁḣ ḋ[

9 ʾšbʾ wmšbʿ lšmh dmkʾl wbš gḃry.ʾl [

10 wbšmh ḋrpʾl wḃ..ʾl̇

11 bšmh dẏhwh mlkh dyšḋh ḋ…yṅ ṩḋw bm [

12 dʾln swst klḥd wbšmh dʾl̇ṅ dṣ.. ʾl lhdḋ[

13 lšmh dyhwh bhdn bṭlh ʿl qmṭh ʿl̇ẏh wmk̇šp.[

14 żẏl̇ṡ ʾt mlʾkh dqm qdm ʾtnh dblʿm qwm qdm ḋḃ[

15 swsth dṩ.yh wk̇ḣ dbnyṭh ktbh mṅ ddẏẏ ʿbdy[t.

10 Translation of the Text⁶²

1 … Symmach[os] (?) …

2 Klearchos (?) … Hipponikos (?) … [A]sphaleia (?) … Hel(i)os (?) …

3 Parsāy (?) … Olymp(i)os

Block A

…

… the mouth (of) yhwh (?)

Block B

In the name (of) (head 1) of [the] horses

of (the) Blue (faction). Yhwh, do not

let them …

Block C

(in the name of head 2) of the horses

this angel

on this

lamella

Block D⁶³

… charioteer …

…

…

…

(in the name of head 3 of) the horses of [the] Blue (faction)

… the turning-post

…

…

…

4 for a charioteer (?). Now, sink these horses that are written (= whose names are written).

5 I adjure in the name of Michael and in the name of Gabriel and in the name of Raphael, (and) in the na[me of …

6 and in the name of hw hwh yhwq mhddʾl (?) rmnʾl (?) and the horses … the (or: his) name

7 erased: {these …}

8 these … [th]ese horses of the Blue (faction) each one (?) and in the name of …

9 I will adjure and I adjure in the name of Michael and in the name of Gabriel (?) …

10 and in the name of Raphael and in …

11 in the name of yhwh the king, who cast (?) … desolation (?) …

12 of these horse(s) each one (?) and in the name of these …

13 in the name of yhwh, on this lamella, at the upper turning-post …

14 … You, the angel that stood in front of the she-ass of Balaam, stand in front of …

15 the horse(s) which … and here/so of the Blue (faction) vacat [I] made the writing from my own (possessions) (?).

11 Commentary to the Text

Lines 1–3

The first three lines probably contain hipponyms rather than the names of charioteers.⁶⁴ Among these in all likelihood are many Greek hipponyms; this is suggested by the high number of words containing an emphatic consonant q or ṭ, and/or an -s at the end of the word, which may reflect the Greek nominative ending -ος in at least some of the cases.⁶⁵ In what follows, those readings that defy explanation will not be commented on.

Line 1⁶⁶

smk̇.—This may reflect the beginning of a hipponym that commences with the Greek name element σύν ‘with’. Such names are attested in Greek tablets from Antioch.⁶⁷ A good candidate here would be Σύμμαχος ‘allied fighter’,⁶⁸ but this well-known name is not to our knowledge attested as a hipponym.

]nyṡ.[—Probably part of another hipponym. The final -s may indicate a Greek hipponym that we are unable to identify. It is also possible to read ]nyq.[ here (on s and q, see 7. The Script) and to argue that it forms the last part of the names Ἱππονίκος or Ἑλληνίκος, either in a m. or a f. form (see also the commentary on the hipponyms in l. 2). The Greek word νίκη ‘victory’ (and derived forms) is a frequent element in hipponyms.⁶⁹

Line 2

q̇lrwk̇s—Note that the reading of the first letter is uncertain. The final -s suggests that this is a Greek hipponym. If this is the case, it may represent a scribal error for qlrkws and reflect the Greek name Κλέαρχος, which occurs in a list of hipponyms in a Greek curse tablet from Antioch. The horses mentioned in the list on this Greek tablet belong to the Blue faction of the city.⁷⁰

hy..q̇h—This possibly reflects the name Ἱππονίκη or Ἑλληνίκη, f. forms of Ἱππονίκος and Ἑλληνίκος, respectively,⁷¹ both of which are known as hipponyms.⁷² Female hipponyms have also been raised as a possibility in one of the Greek curse tablets from Antioch.⁷³

ṗḣẏnyqs—The reading is highly uncertain, though this word seems to reflect a Greek hipponym. One wonders if phynyqs is a scribal error for hypnyqs, Greek Ἱππονίκος.⁷⁴ A related form of this name is found in a Greek curse tablet from Antioch.⁷⁵ Alternatively, one might think of the well-known hipponym Φοῖνιξ / Phoenix.⁷⁶ It must be stressed, however, that the Aramaic letter h in general is not used to render the spirantisation of the Greek consonant phi.⁷⁷ Either way, the spelling is not what one would expect.

.ṡplyh—Possibly another hipponym. It may reflect the Greek noun ἀσφάλεια, ‘security against stumbling or falling’ and may be compared with the hipponym Securus which is found in a Latin curse tablet from Hadrumetum (second century CE).⁷⁸ As a Greek loanword, it is found in an adverbial phrase in the Aramaic Bar Kosiba letters from Naḥal Ḥever, datable to the first half of the second century CE (no. 54: 5, 9, 15 bʾsplyh and bʾsplyʾ ‘in security’); the noun is also known from Syriac (ʾsplyʾ).⁷⁹ As a result, the possibility that the word functions as a noun in our tablet (and does not represent a hipponym) cannot be excluded.

ḣylṡ—We assume that this word reflects the name Ἥλιος, even though the expected spelling would be h(y)lys or h(y)lyws. The name Helios is well known as a hipponym and as the name of charioteers.⁸⁰ The spelling hylws is to be found in a new Genizah manuscript of Sefer ha-Razim.⁸¹

Line 3

prsy ‘Persian’—Probably a hipponym (possibly with the Aramaic gentilic ending sg. m. -āy of the absolute state).⁸² One of the semantic fields represented in hipponyms is the origin of the horse (divine descent, actual sire, ethnic provenance, or some other source).⁸³ Other interpretations are less likely.⁸⁴

lṭ[—In light of the hipponyms in this section of the tablet we should probably assume a hipponym here as well, but the damage to the tablet at this point is too extensive to make any convincing suggestions in this regard.⁸⁵

We also note that this is the only point in the tablet where a form of the Common Aramaic root lwṭ might appear—either a verbal form (lwṭ ‘to curse’) or a form of the related f. noun lūṭā ‘curse’. The form could reflect a defective spelling of the passive participle (līṭ ‘cursed’) or a defective spelling of the noun; any inflectional endings are not legible. For obvious reasons, this suggestion is highly tentative.⁸⁶

ʾwlmps—This is probably the Greek hipponym Ὄλυμπιος or Ὀλύμπιος. To date, this name is known only as that of a charioteer and not as a hipponym.⁸⁷ The hipponym Olypius (error for Olympius) ‘originating from the Olympus’, however, appears in a Latin mosaic inscription.⁸⁸ The hipponym Ὀλυμπιονίκη ‘victory at Olympia’ is more common and appears in circus curse tablets.⁸⁹

Block A

bšmh—We assume that this word should be read with the text of block B; see below.⁹⁰

pṁh yhwh ‘the mouth ⟨of⟩ yhwh’ (?)⁹¹—The word pmh is either an emphatic sg. m. form (‘the mouth’) or a pron. sf. sg. 3m. (‘his mouth’). It is not immediately clear to whose mouth the text refers. Neither is it clear whether the form represents an isolated word or not. As an isolated word, pmh might refer to a human mouth, such as the mouth of the charioteer in the next line, or perhaps even to the mouth of Balaam’s she-ass, as she is probably mentioned in line 14 (qdm ʾtnh dblʿm ‘before the she-ass of Balaam’; see also 5. Balaam and His She-Ass).⁹² The word pmh might equally have been intended to evoke the biblical image of ‘the (mouth of the) earth’, which takes/swallows human beings (see Gen. 4.11, Exod. 15.12, and other texts). It should be noted that the mouth of yhwh, the mouth of Balaam’s she-ass, and the mouth of the earth all play a role in the Balaam narrative and its reception; even Balaam’s own name, according to tradition, is connected with an action of the mouth (‘to swallow’; see 5. Balaam and His She-Ass).

If pmh is not an isolated word, then it must be connected with the divine name written on the same line. It is possible that we are dealing here with a lapsus on the part of the person who inscribed the tablet, who may have intended to write pmh dyhwh ‘the mouth of yhwh’ but forgot the genitive particle d. This phrase would echo the biblical phrase ky py yhwh dbr ‘because the mouth of yhwh has spoken’ (Isa. 40.5). The phrase ‘the mouth of yhwh’ is also found in the Balaam narrative in Num. 22.18 (see 5. Balaam and His She-Ass).

Block B

bšṁh (HEAD 1) dswst[h] dbnṭ ‘In the name of HEAD 1 of the horses of the Blue faction’—On this reading of the different segments of the text that surround the three heads, see 6. The Inscription on the Tablet, and the commentary to block A. Our interpretation is complicated by the fact that we remain unsure about the referents of the heads and the relation between the text and the drawings. The heads may represent the three archangels, but we cannot exclude the possibility that they represent three charioteers (regardless of whether they are referred to in the text or not). An argument in favor of the latter interpretation is the fact that the heads were pierced with a nail after the tablet had been rolled up. An argument in favor of the former interpretation is the fact that the name of an archangel appears to have been written below each of the heads. If the angelic interpretation is correct, the piercing of the heads—though unfortunate—was surely unintentional.⁹³ We suggest that the text in this portion of the tablet can be read profitably as a kind of picture book, and that the word bšmh, which is written to the right of HEAD 1, represents the beginning of a spell that is followed by HEAD 1 dswst[h] dbnṭ, continued by HEAD 2 dswsth, and followed by HEAD 3 dswsth dbn[ṭh]. The phrase bšmh ‘in the name of’ is familiar as part of the opening formula in many spells, and it would represent the first appearance in our text of such an introductory formula.⁹⁴

Note, however, that there is no trace of the particle d- (‘of’) following bšmh. Nonetheless, if we render the picture-book format of this section of the tablet in the form of a running text, we arrive at the following fluent reading: ‘In the name of Michael (HEAD 1) of the horses of the Blue faction, (in the name of) Gabriel (HEAD 2) of the horses, (in the name of Raphael) (HEAD 3) of the horses of the Blue (faction)’. If, in contrast, the heads represent a trio of charioteers targeted by our tablet, the text may still be read as a picture book, though the word bšmh would then form part of the text in block A.

dswsṫ[h] dbnṭ ‘of the horses of the Blue (faction)’—As would be expected of a lead tablet found in a circus, there is frequent reference to horses in our text. The form swsth occurs several times in this tablet (ll. 4, 6, 15; block C); in addition, the forms swswth (in l. 8; perhaps also in block D: sws[w]th) and swst (l. 12) are attested.

The vocal ending of the sg. m. absolute form of the word for horse (written with -y or -h and representing the vowel -ē) is characteristic of Aramaic. The emphatic form of this word is usually sws(y)yʾ (in Jewish Palestinian Aramaic sws[y]yh). The plural of the m. noun is largely formed with the ending -wān (spelled -wn) in the absolute form, -wāt (spelled -wt) in the construct form, and -wātā (spelled -wtʾ or -wth) in the emphatic form.⁹⁵ There is a considerable amount of variation in the spelling of this noun across the different dialects of Aramaic.

The f. form of the noun (swsh, with a f. ending -ā) is not widely attested in Aramaic but examples are known of the sg. f. emphatic form swstʾ and the pl. emphatic form swsytʾ.⁹⁶

All the attestations of this noun in our text are probably based on the pl. m. formation of the noun, with nominal endings that are normally used for f. nouns (absolute: -n; construct state: -t; emphatic state: -th). From a formal point of view, only the form swswth in line 8 can be identified as a pl. m. (due to the medial consonant -w-), though it is possible and even probable that the plural is implied in each occurrence of the noun in this text. This assumption is based on the use of the demonstrative pronoun ʾl(y)n ‘these’ (pl. c.), which modifies both swsth and swst in the tablet (l. 4 ʾlẏn swsth; l. 12 dʾln swst). However, the possibility that several of these forms are in fact f. ‘mare(s)’ (especially in l. 15; see the commentary to this line) cannot be excluded.

In three cases, the word for horse is found in a d-phrase (a genitive construction). In each case, the second term in the construction refers to the Blue faction (block B dswst[h] dbnṭ; block D sws[w]th dbn[ṭh]; l. 8 swswth dbnṭh). The word dbnṭ also appears in line 15 where it is preceded by wkh ‘and here’/‘and so’ and followed by dṩ.yh (meaning uncertain; see the commentary to ll. 12 and 15). In lines 8 and 15 the word bnṭ probably occurs in the sg. emphatic form (with the ending -h; l. 8 swswth dbnṭh; l. 15 swsth dṩ.yh wkh dbnyṭh), but in Syriac the word bnṭʾ is a pl. m. (marked with the syāmē) and reflects the Greek plural form Βένετοι ‘Veneti, members of the Blue faction in the Byzantine circus’.⁹⁷

ẏḣẇḣ—It is possible that yhwh is invoked here and that the divine name should be connected with the following words.

lh ṫṧbq̇ lwn ʿl̇—The reading is tentative. The second word is a verbal form and should perhaps be connected with the word lh ‘not’,⁹⁸ which is written above the last part of this line and which slants downwards (see drawing). It should be noted, however, that the letters of lh tšbq lwn are smaller in size than the rest of text. The l which follows the verbal form could be the preposition l-, in which case we might be looking at the preposition l- combined with a pron. sf. pl. 3m.⁹⁹ The phrase could thus be read as ‘do not let/allow them (to enter/to win)’.¹⁰⁰ Similar short commands to a divinity to impede the horses in a race are also found in a Greek curse tablet discovered in the circus at Antioch.¹⁰¹ It should be noted that the line tšbq lwn ʿl̇ was added after the following line was written. The interpretation of the last word of the line is equally slippery. A connection with the verb ʿll ‘to enter’ is attractive due to the context, but this is grammatically impossible, as one would expect lmʿl ‘to enter’, or the like.

Block C

dswsth ‘of the horses’—The word is located at the top of block C, just as in block B and probably in block D as well (where it appears in a damaged form), and is read as part of a spell. Note that in block B and D this noun phrase is followed by dbnṭ(h) ‘of (the) Blue faction’ (dbnṭ in block B; dbn[ṭh] in block D; see also dbnṭh in l. 8). On the interpretation of this part of the text in block C, see the commentary to blocks A and B.

hdn sr ‘this angel’—The demonstrative pronoun hdn occurs in Jewish Palestinian Aramaic with an attributive function as well as a substantive function (see also 8. Linguistic and Orthographic Characteristics of the Text).¹⁰² Attributive demonstrative pronouns are usually found in an initial position in Jewish Palestinian Aramaic.¹⁰³ The meaning ‘angel’ is not absolutely certain here, and the interpretation is somewhat problematic due to the fact that the three heads drawn on the tablet were pierced by the nail, though the possibility that this happened accidentally cannot be excluded, regardless of the negligence this might suggest (see also the commentary to block B above). If correct, however, this noun phrase may point to the adjacent head (either HEAD 2 or HEAD 3). The word is found here in the spelling with s, though the older spelling with ś is the regular spelling in Jewish Palestinian Aramaic.¹⁰⁴ It has the meaning ‘lord’, ‘prince’¹⁰⁵ and ‘angel, guardian angel’¹⁰⁶ and is considered to be a loanword from Biblical Hebrew due to the frequent spelling of the word with ś in several dialects of Aramaic. The word is known in other magical contexts, for instance, in the following text from the Palestinian Talmud (p. San. 7:11 [25d]): ‘Rabbi Yehoshua commanded the angel of the sea and he swallowed him (the magician)’ (gzr rʺ yhwšʿ ʿl śrh dymʾ wblʿyh).¹⁰⁷ There are no visible traces of an emphatic state ending attached to sr, and such an ending may have been omitted by the person who drafted the text of the tablet (see also dʾln swst in l. 12).¹⁰⁸ It is also possible to translate this phrase as a nominal clause (‘this is an angel’), though this reading is less likely.

bhdn bṭlh ‘on this lamella’—The word bṭlh is probably related to Greek πέταλον ‘lamella’, a word that is often applied to the metal sheets used for the inscription of magical texts.¹⁰⁹ The word appears twice in the self-designation bhdn bṭlh ‘on this lamella’ (block C; l. 13). The related word pṭlwn (pslyn) is documented already in the Tosefta, with the same technical meaning as the Greek: ‘pslyn on which one writes the amulet’ (pslyn šhwʾ kwtb ʿlyw ʾt hqmyʿ). It also appears in later Jewish magical material.¹¹⁰ The word pṭlʾ / pʾṭlʾ is also known from Syriac (pl. m. < πέταλα; marked with a syāmē). It is modified by a noun designating copper (pṭlʾ dnḥšʾ), tin (pṭlʾ dʾnkʾ) and lead (pṭlʾ dʿprʾ).¹¹¹ Noteworthy is the fact that, in all the Aramaic sources, the Greek π in this word is regularly rendered with p,¹¹² as is the normal practice in Aramaic for the rendering of Greek loanwords with the voiceless labial stop π. There are, however, several instances in Rabbinic literature where Greek π is rendered with the voiced labial stop b in loanwords.¹¹³ We might add that the change p > b is in general a well-known feature in many Semitic languages, including Hebrew and Aramaic.

Block D

q̇ṙṙ—Possibly ‘charioteer’, a word which occurs in Syriac and which is derived from Latin carrarius.¹¹⁴ It appears, for instance, in the Peshitta of 2 Macc. 9.4, where it renders Greek ἀρματηλάτης.¹¹⁵

sws[w]th ḋbn[ṭh]—See blocks A, B, and C for the interpretation of this phrase as part of a spell beginning with bšmh. This accords with the location of the word sws[w]th to the left of HEAD 3. In fact, block D may start at this point (see also 6. The Inscription on the Tablet).

qmṭh—The same form of the word also occurs in line 13. The word is probably based on the Greek καμπτός ‘flexible’ and may be interpreted in this context as a substantivised form of this Greek adjective.¹¹⁶ The Greek ending has been removed and the final -h reflects the emphatic state ending -ā.¹¹⁷ The spelling also reflects the deletion of π in the input form (see n. 118). Both qmpṭws and qmṭws (with deletion of π) are attested in Syriac.¹¹⁸ As our tablet was excavated near the first turning-post (see 2. Archaeological Context and Description of the Tablet), this reading would fit well with the context of the tablet.¹¹⁹

Line 4

l̇qwrṙ–qwrr may be a variant spelling of qrr ‘charioteer’ (see above block D). The presence of the vowel letter w for u or o in our tablet can be explained by the close proximity of q and a liquid.¹²⁰ The phenomenon as such is not limited to loanwords. There may be another example of the word in block D (q̇ṙṙ), but note that the reading of the word is uncertain (on the similarity of d and r, see 7. The Script).

Another possibility is to derive l̇qwrṙ from the well-known verb qrr ‘to freeze’, but the form defies further interpretation, and it is therefore unclear how the word would exactly fit in the context.¹²¹ For a possible parallel in a Greek curse tablet, see, e.g., the curse directed at Babylas the Greengrocer from Antioch. In that tablet, ‘to drown’ and ‘to chill’ occur in close proximity: ‘Just as you are sinking and are chilled in the disused cistern, so too sink and chill the soul of Babylas’.¹²²

kdw ṭmʿwn ʾlẏn swsth—Because the forms of w and y are so similar in our text, there are several ways to read this line. Alternatively, the first word may be read as kd ‘when’ (instead of kdw ‘now’¹²³), in which case the third letter would be a y representing the prefix of a pref. form (kd yṭmʿwn). The verb form derives from the root ṭmʿ. This verb is well known from the western dialects of Late Aramaic (Jewish Palestinian Aramaic, Christian Palestinian Aramaic, and Samaritan Aramaic).¹²⁴ In Jewish Palestinian Aramaic, the peʿal of the verb has only intransitive meanings, such as ‘to be swallowed up’, ‘to drown’, ‘to sink’, and ‘to set’ (said of the sun); in the paʿel, it has transitive meanings, such as ‘to bury’, ‘to cause to drown’, and ‘to sink’.¹²⁵ We have interpreted the form here as a paʿel imperative pl. 2m.¹²⁶ Imperative forms of Greek verbs that denote ‘to sink’ (βολίζω) and ‘to submerge’ (καταβολοθρίζω) are frequently attested in Greek circus curses from Antioch and Caesarea.¹²⁷ All these tablets were found in drains, cisterns, or wells, and there must presumably have been an analogy at work between the sinking of the tablets and the sinking of the tablets’ victims.

At the same time, the verb ṭmʿ in our tablet also evokes the Hebrew text of Exodus 15 (see also 5. Balaam and His She-Ass), a song about the destruction of Pharaoh’s cavalry, which ‘sank as lead into the mighty waters’ (JPS) (Exod. 15.10: ṣllw kʿprt bmym ʾdyrym). The correspondence between the sinking of the cavalry and of lead is an excellent example of analogical magic, especially in our equine context. Presumably the exposure of the lead tablet to the water in the drainage channel where it was deposited was seen as sufficient for the purposes of the effective analogy.¹²⁸ According to Exod. 15.12, and contra Exod. 15.10, it was the earth and not the sea that swallowed the Egyptians (‘you stretched your right hand and the earth swallowed them’). One of the Rabbinic traditions that tries to explain this discrepancy recounts a dispute between the earth and the sea about who will receive the slain Egyptians. This dispute is found in Targum Pseudo-Jonathan and all the extant Palestinian Targums to Exod. 15.12 (Fragmentary Targum [ms. P and ms. V]; Targum Cairo ms. W; Targum Neofiti I). The rabbis use this dispute in order to try and explain that, according to the biblical text, it was the earth and not the sea that swallowed the Egyptians:

The land and the sea were arguing together, one with the other. The sea said to the land, ‘Receive your children’. And the land was saying to the sea, ‘Receive your slain’. The sea did not want to sink them (ṭmʿ)¹²⁹ and the earth did not want to swallow (blʿ) them. Immediately, O Lord, you inclined your right hand over the earth with an oath that they would not be demanded of her in the world to come. The earth then opened its mouth and swallowed them.¹³⁰

Targum Pseudo-Jonathan

The Palestinian Targums and Targum Pseudo-Jonathan use the verb ṭmʿ to translate the Hebrew verb ṭbʿ in Exod. 15.4 (‘they were sunk [puʿal perfect] in the Sea of Reeds’).¹³¹ Most of the Palestinian Targums and Targum Pseudo-Jonathan to Exod. 15.1 and 15.21 also use the verb ṭmʿ in an expansion of the Hebrew text,¹³² apparently in order to create a parallel with the biblical text of Exod. 15.4 which reads: ‘Pharaoh’s chariots and his army he has thrown into the sea and the best of his officers were sunk in the Sea of Reeds’ (translation Folmer). The biblical text of verses 1 and 21 has only ‘the horse and its rider(s) he has thrown into the sea’.

There is a reference to Pharaoh’s chariots in one of the Greek curse tablets from Antioch: ὡς ἔβαλες τὼ ἅ⟨ρ⟩μα τοῦ Φαραῶνος οὕτος βάλη τὶν δύσ[ω]ληψιν αὐτοῦ ‘As you smote the chariot of Pharaoh, so smite his (Babylas’) obstinacy (or trickery)’.¹³³ ‘The chariot of Pharaoh’ in the Greek tablet is an explicit reference to the story in Exodus 15.¹³⁴ The image of God as a charioteer pushing his chariot over the Sea of Reeds is also found in Jewish Babylonian Aramaic magic bowls, e.g., in the following adjuration: bšmyh dyhwq yhwq dydḥq mrkbtyh ʿl ymʾ dswp ‘in the name of God who pushes his chariot over the Sea of Reeds’.¹³⁵ The Sea of Reeds also appears in a Syriac bowl: ʾyk dʾmr mwšʾ lymʾ dswp wqmw ʾyk šwrʾ dmn trwyhwn gysh ‘just as Moses said to the Sea of Reeds and walls (of water) stood from two sides’.¹³⁶ In the following amulet from Ḥorvath Marish, written in Jewish Palestinian Aramaic and intended for the subjugation of the fellow townspeople of Yose son of Zenobia, God is depicted as the God who subdued the sea with his horses: hyk dkbšt ymh bswsyk ‘just as you have subdued the sea with your horses’.¹³⁷

Line 5

ʾ{b}šbʾt—The person who inscribed the tablet erred after ʾ and wrote b. He then wrote the š over the b. This verbal form (afʿel sf. conj. sg. 1c. of the verb šbʿ ‘to adjure’) is not spelled with ʿ but with ʾ. The same spelling is used in ʾšbʾ (l. 9), which is a pref. conj. form of the same verb (afʿel pref. conj. sg. 1c. of šbʿ ‘to adjure’) (see the discussion in 8. Linguistic and Orthographic Characteristics of the Text). In this verb’s third appearance, it is spelled with ʿ (l. 9 mšbʿ, afʿel part. sg. m.). This form misses an overt expression of the subject, but note that the subject can be inferred from ʾšbʾ, to which it is juxtaposed.

bš mkʾʾl—This is an abbreviation of bšm mkʾl rather than a case of haplography (see also bš rpʾl in l. 5 and bš gḃry.ʾl in l. 9).¹³⁸ Note that the angelic name Michael is written without the vowel letter y, and that the ʾ is written twice.¹³⁹

gbrʾl—The shape of the ʾ with two additional strokes in the shape of a Greek lambda is unusual. We have interpreted this as a ligature of ʾ and l. Another example of this ligature appears in line 6, in the word mhddʾl.

bš rpʾl ‘in the name of Raphael’—This is another instance of the abbreviation of bšm. Michael, Gabriel, and Raphael are often grouped together in Jewish literature from the Second Temple period onwards—typically in groups of four or seven special angels.¹⁴⁰ The names Michael, Sariel,¹⁴¹ Raphael, and Gabriel appear in Aramaic fragments of 1 En. 9:1 from Qumran (4Q202 col. iii:7) which probably belong to the oldest strand of this tradition (cf. 1QM ix.15–16).¹⁴² Qumran also witnesses the tradition of seven archangels.¹⁴³ Even though the ordering of the archangels is not always the same, there is a clear tendency to name Michael first.¹⁴⁴ Parallels to the order Michael, Gabriel, Raphael—as in our tablet—can be found in a Greek curse tablet from Antioch.¹⁴⁵ This order is also found in Jewish magical texts from Late Antiquity and the Cairo Genizah,¹⁴⁶ alongside the order Michael, Raphael, Gabriel.¹⁴⁷ The three names sometimes appear in a sequence consisting of only these three angels. There are also examples of this in the magic bowls.¹⁴⁸

Line 6

wbšmh dhw hwh yhwq ‘and in the name of hw hwh yhwq’—The name yhwq is known from a magical historiola that appears in some magic bowls and which recalls the crossing of the Sea of Reeds in Exodus 15: bšmyh dyhwq yhwq ddḥyq mrkbtyh ʿl ymʾ dswpʾ ‘in the name of yhwq yhwq who pushed his chariot over the Sea of Reeds’.¹⁴⁹

mhddʾl rmnʾl—The final element -ʾl attached to both of these words suggests that they are possibly angelic names. We have no direct parallels for such names of angels, but there are several remarks that can be made.

The interpretation of mhddʾl or mḥddʾl is problematic.¹⁵⁰ There are many angelic names that somehow echo in this name, but it is difficult to draw any etymological connection with these other names. From an etymological point of view, it makes sense to connect the name with the root ḥdd (paʿel ‘to sharpen’).¹⁵¹ The name mhddʾl / mḥddʾl may have echos in hdd[ʾl or ḥdd[ʾl (l. 12), but this is very speculative.

From a semantic point of view, it is attractive to attempt a connection with the angelic name yhdʾl, attested in an Aramaic fragment from Qumran, but this is not warranted from an etymological point of view, as yhdʾl seems to derive from hdy ‘to lead’ (paʿel).¹⁵²

There may be a connection between rmn’l and the Aramean divinity Ramman.¹⁵³ A connection with the name rmyʾl is not very likely from an etymological point of view but should be considered nonetheless.¹⁵⁴

Lines 7–8

Line 7 and the beginning of line 8 are almost incomprehensible. Line 7 seems to be a false start. The nine or ten letters in this line look faded and were apparently erased. It is possible that line 8 is a new attempt to write the same text, as the initial words in both lines are almost identical (hln and ʾln, respectively, possibly the plural demonstrative pronoun).

Line 8

.ln swswth dbnyṭh—The reading of the first word is uncertain, but it probably reflects the demonstrative pronoun ʾln, attested elsewhere in this inscription (for ʾlẏn, see l. 4; for the defective spelling ʾln, see l. 12).¹⁵⁵

k̇lḥd—The reading of this word is uncertain. The same sequence of letters also appears in line 12 (klḥd). At present, the most satisfactory solution seems to be to read klḥd or kl ḥd and to interpret this as ‘every one’ or ‘each one’.¹⁵⁶

Line 9

ʾšbʾ wmšbʿ—On the weakening of the gutturals, see 8. Linguistic and Orthographic Characteristics of the Text and the commentary on line 5. The afʿel pref. conj. sg. 1 is followed by an afʿel participle (sg. m.). Both participles and sf. conj. forms sg. 1 (the latter with performative function; also in l. 5 of this tablet) of this verb are common in adjurations. To our knowledge, there are no parallels for pref. conj. forms of this verb in incantations.¹⁵⁷

lšmh—The preposition l in this formula is unusual; the expected formula is bšmh ‘in the name of’. The combination with l also appears in line 13,¹⁵⁸ though it is also possible to read a corrupted b in line 13 instead.

wbš gḃryʾ.l—This is presumably another instance of the abbreviation of bšm to bš, which also occurs in line 5 in combination with the angelic names mkʾʾl and rpʾl. The name that follows bš in line 9 is probably that of an angel: It ends with the typical element -ʾl and could be Gabriel, but the reading is uncertain. The lower stroke of the b is omitted, and a letter or a sign appears between ʾ and l that we are unable to identify. If the name gbryʾl does indeed appear here, line 9 would contain a further instance of the trio Michael, Gabriel, and Raphael in addition to their appearance in line 5. It is also possible to read gdryʾ.l here; the angelic name gdryʾl is known from several mss. of Sefer ha-Razim.¹⁵⁹

Line 10

wbšmh drpʾl—‘and in the name of Raphael’

What follows after this phrase is difficult to decipher because either line 10 or line 11 was written over the other, and we cannot tell which of the two lines the letters -ʾl belong to (see l. 11 below). It is possible that another angelic name follows rpʾl as the ending -ʾl can be identified here.

Line 11

bšmh dẏhwh mlkh dyšḋh—‘in the name of yhwh the king,¹⁶⁰ who cast (?) …’

The reading of the last word is uncertain. It seems most plausible to interpret the word dyšdh as a form of the verb šdy (peʿal) ‘to throw’, ‘to cast down’, represented by a pref. conj. preceded by d- or a sf. conj. preceded by a plene spelling of d-.¹⁶¹ A sf. conj. would suit the context best as it would link this sentence to Exodus 15. The verb šdy is a Common Aramaic verb; it occurs in Official Aramaic, Qumran Aramaic, and in several dialects of Late Aramaic.¹⁶² It also appears in Targum Onqelos to Exod. 15.4:¹⁶³ rtyky prʿh wmšrytyh šdy bymʾ ‘The chariots of Pharaoh and his armies he has thrown into the sea’. For a similar usage of the verb šdy in the peʿal, see the following magic bowl: šdynʾ lkyn lʿwmqy ʾrʿh tḥtytʾ dlyt ʾypšr lmysq mynh ‘and I shall throw you to the depths of the netherworld, from which it is not possible to ascend’.¹⁶⁴ Exodus 15 is often referred to in Jewish magical texts, and a reference to Exodus 15 is also to be found in a Greek tablet from Antioch (see the commentary to l. 4 above).¹⁶⁵

Another possibility is to read the sequence dyšdh as ‘of the demon’ (featuring either a plene spelling of the genitive particle də- or, less likely, the older form of this particle: dy). This would give ‘the king of the demon’ (mlkh dyšdh), a title which is more often found in Jewish magical literature, but as far as we know it is never given to God.¹⁶⁶

As mentioned in the commentary to line 10, the reading of lines 10 and 11 is complicated by the fact that they overlap partially; the reading of the word following mlkh in line 11 is complicated by the problematic positioning of the letters ʾ and l. It is exactly at the point of these two letters that line 10 and the last part of the word dyšdh in line 11 overlap, and it is possible that these two letters belong not to line 10 but to a word in line 11. If the two letters belong to line 11, then the word following mlkh should be read dyśrʾl: bšmh dyhwh mlkh dyśrʾl ‘in the name of yhwh the king of Israel …’. This interpretation suits the context best, but because of the difficulties mentioned we do not present it as our primary reading.

The last part of line 11 defies interpretation. Only the word ṣdw, ‘desolation’ or ‘affliction’ can be read with some confidence.¹⁶⁷ The related adjective ṣdy is found in the phrase ʾtr ṣdy ‘a desolate place’ in a magic bowl.¹⁶⁸

Line 12

dʾln swst—Probably a scribal error for sws(w)th (l. 8 swswth) as one would expect the emphatic form in an attributive noun phrase with a deictic pronoun as the modifying element.

dṣ..—We also encounter a word that begins with dṣ[ in line 15. It is possible that these two forms derive from the same root, possibly a peʿal passive participle of a form of a root with initial ṣ and a final y (for the y, see l. 15), preceded by the particle d. Unfortunately, the second root consonant is unclear (both here and in l. 15).

lhdḋ[—For this word, see the comments on mhddʾl in line 6. We might hypothesise another angelic name here, but this is highly tentative for obvious reasons. We do not have a direct parallel for this name.

Given the regional context of our tablet, a connection with the foremost Aramean weather-god Hadad—even at this late period—should be considered. There is evidence in a letter written to the Antiochene rhetorician Libanios (d. 363 CE) for the continuity of the cult of Hadad at Aleppo at least until the middle of the fourth century CE.¹⁶⁹ A connection with Hadad is possible, but it must be stressed that this is highly speculative. See also the comments on Ramman in line 6.

It is also possible, instead of hdd[ʾl, to interpret the second d of the word as a r and to read hdr[ʾl or ḥdr[ʾl. The angelic name ḥdryʾl in a magic bowl text might be related.¹⁷⁰

An alternative possibility is to set aside angelic names altogether and to interpret ḥdd as a form of the verb ḥdd. The root ḥdd is known in Aramaic from the verb ‘to sharpen’ (paʿel) and from several nominal formations,¹⁷¹ but it is not clear which form and function would fit in this context (see also the commentary on mhddʾl in l. 6).

Finally, it is also possible to interpret lḥd as ḥd ‘one’¹⁷² preceded by the preposition l. In this case, the second d should be connected with the following, damaged part of the line.

Line 13

lšmh dyhwh ‘in the name of yhwh’—On the preposition l in this formula, see the commentary to line 9.

bhdn bṭlh ‘on this lamella’—On this noun phrase, see block C.

ʿl qmṭh ʿl̇ẏh ‘at the upper turning-post’—On the word qmṭh, see block D.

ʿl̇ẏh—The reading of this word is uncertain. The word is possibly an emphatic form of the word ʿlyh ‘upper’ modifying the preceding word qmṭh (emphatic sg. m.).¹⁷³ In that case, ‘the upper turning-post’ would represent a technical term referring to the first turning-post (where the tablet was found) or to the second turning-post.¹⁷⁴ The word ʿl̇ẏh possibly reflects the word ἄνω in the noun phrase ὁ ἄνω καμπτός ‘the upper turning-post’ which is found in the work of the Antiochene writer Malalas.¹⁷⁵

wmk̇šp[—The reading of this word is also uncertain. If this reading were correct, then the form could be a participle of kšp ‘to practice sorcery’. This root is more common in Hebrew than in Aramaic (where in all likelihood it is a loan from Hebrew).¹⁷⁶

Line 14

żẏl̇ṡ—Unfortunately, the reading of this word, especially its first letter, is highly tentative. The final -s suggests that a Greek loanword might provide a possible explanation. We have considered ζῆλος ‘jealousy’, ‘rivalry’,¹⁷⁷ but not enough of the context is preserved to comment any further on the plausibility of this suggestion.

ʾt mlʾkh dqm qdm blʿm qwm qdm ḋḃ[… ‘you the angel that stood before Balaam stand before ḋḃ[…’

ʾt—This probably reflects the personal pronoun sg. 2m. ‘you’. This is the regular spelling of this pronoun in Jewish Palestinian Aramaic.¹⁷⁸ For the direct address ‘You, the angel that stood before the she-ass of Balaam, now stand before ḋḃ[…’ see, for instance, the direct address in the following magical text: ʾtwn mlʾkyh qdšyh dqymy[n] qdm krsyh dʾlh rbh ‘you, the holy angels who stand in front of the throne of the Great God’.¹⁷⁹

The invocation of the angel that stood before the she-ass of Balaam is a reference to the biblical story of Balaam son of Beor and his she-ass in Num. 22.1–35.¹⁸⁰ The name Balaam might be linked with the events recounted in Exodus 15 (see 5. Balaam and His She-Ass) through the association of the name Balaam with the verb blʿ ‘to swallow’. This verb is also found in Exod. 15.12, where the earth is said to have swallowed up the Egyptians, while the verb ṭmʿ is used to translate blʿ in several Aramaic versions of Exod. 15.12. The same verb is also found in line 4 of our tablet. As such, both line 4 and line 14 may evoke the narrative of Balaam and the events related in Exodus 15. For further details on the figure of Balaam in the Hebrew Bible and in later Jewish tradition, see above 5. Balaam and His She-Ass.

What follows after qdm is unclear as the reading of both the d and the b are uncertain. The fact that the angel in Num. 22.1–35 blocked the path of Balaam’s she-ass and its rider leads us to believe that what follows after qdm is the name of the horse and/or charioteer whose course was to be blocked. For that reason, the sequence ḋḃ[ should perhaps be restored as ḋḃ[nṭh ‘(those) of the Veneta’, but this is very tentative. Alternatively, the text can be restored as ḋḃ[b ‘adversary’.¹⁸¹

Line 15

swsth dṩ.yh—The word dṩ.yh is possibly connected to the word dṣ.. in line 12 (see the commentary to l. 12). The ending -h is suggestive of a sg. f. absolute form, but this does not sit well with the preceding m. noun swsth, unless this instance represents a sg. f. form of the noun (‘mare’; see also the commentary to l. 2 on the possible appearance of the names of mares in curse tablets from Antioch).

wk̇ḣ ‘and here’—A word pointing to the here and now would fit well with the context of a magical text, but its precise relation with what follows is unclear. An alternative interpretation is not self-evident.

ktbh mṅ ddẏẏ ʿbdy[t.—‘The (this) inscription I made from my own [funds/possessions]’

The phrase is preceded by a small vacat, and it is attractive to read this final part of the text as a kind of colophon, though its reading and interpretation are far from certain.

The word ktbh is probably a noun in the singular with the emphatic ending (‘the writing’, ‘the document’). If this interpretation is correct, then it is similar in function to the self-designation hdn bṭlh that appears elsewhere in our text (block C and l. 13), albeit without a demonstrative pronoun. See also the phrase hdn ktbh in the magic bowls.¹⁸²

The presence of the verb ʿbd ‘to make’ seems certain here.¹⁸³ The y following ʿbd may be interpreted either as the beginning of a new word, perhaps a short personal name, or as part of the preceding word. The remains of another letter are visible following the y, possibly a t. If so, it is possible that the text here reads ʿbdy[t], a sf. conj. 1c. with the plene spelling of the afformative (ʿbdy[t] ‘I made’).

At first glance, the reading ddwy seems more likely than ddyy due to the different lengths of the last two letters. The latter reading—ddyy, with a shorter y following the first—should nevertheless be given serious consideration. The word ddyy / ddwy should probably be understood as a word derived from yd ‘hand’, most likely the possessive particle dyd (< d- + yd) with a pron. sf. sg. 1c. or perhaps sg. 3m.¹⁸⁴ Each of these two interpretations is problematic in one way or another. First, the possessive particle is usually spelled in plene fashion (dyd-). Second, the possessive particle is normally combined with pronouns that in turn are combined with singular masculine nouns (i.e, -h instead of -wy; -y instead of -yy; see the text quoted below). Our tablet, however, presents us with either -wy or with -yy, both of which are normally attached to plural masculine nouns (on the typical form of the pron. sf. -wy in Jewish Palestinian Aramaic, see n. 56). Third, one would not expect the vowel-letter y (representing the original consonant of yd ‘hand’) to be dropped in either of these cases. Despite these problems, the interpretation of ddyy / ddwy as a noun which developed from a possessive particle seems to be the most likely option. It is possible that the double y in ddyy is an instance of dittography, or that the person who inscribed the tablet perhaps intended to write dydy but made a mistake.

The form ddwy cannot be explained in a similar manner (as a scribal error), and it is for this reason that the interpretation ‘The (this) inscription I made from my own [funds/possessions]’ is to be preferred over the interpretation ‘The (this) inscription PN (y[…) made (ʿbd) from his own [funds/possessions] (mn ddwy)’.¹⁸⁵ Whether or not the sentence is phrased in the first person remains somewhat uncertain, though this question is obviously of importance to the interpretation of the tablet.

The phrase calls to mind the following line from an ancient mosaic inscription in the synagogue at Noʿara (ʿAyn Dūk, sixth century CE), where mn precedes dydh: dyhb ṭymy psypsh mn dydh ‘who gave the price of the mosaic from his own (possessions)’.¹⁸⁶

Acknowledgements

We are grateful to Robert Daniel for reading and commenting on this paper and to Julia Dweck, chief curator of the Princeton University Art Museum, for her permission to publish the curse tablet in Aramaic Studies. The tablet is a gift of the Committee for the Excavation of Antioch to the Princeton University Museum. We would like to thank Gideon Bohak, Yoel Elitzur, Christopher Faraone, Raquel Martín-Hernández, Ortal-Paz Saar, Guy Stiebel, Aaron Butts, and the anonymous reviewers for their helpful suggestions, as well as Daniel Waller for checking the English.

Abbreviations