Notes on a Recently Published Magic Bowl

[AS 2.2 (2004) 207-22] ISSN 1477-8351

NOTES ON A RECENTLY PUBLISHED MAGIC BOWL*

Matthew Morgenstern

University of Haifa

Dedicated to Prof. Joseph Naveh on the occasion of his 75th Birthday

1. Introduction

Scholars of Jewish Babylonian Aramaic have recently been presented with two editions of an interesting magic bowl from the British Museum Collection (BM 135563). The first edition to appear was published by C. Miiller-Kessler and Th. Kwasman (from hereon MKK),l anticipating by several months the appearance of the 'official' edition by J.B. Segal (from hereon Segal).2 In the opinion of Miiller-Kessler and K wasman, the bowl is unique in refiecting an Aramaic idiom close to that found in the majority of the Talmudic tractates. A detailed discussion of linguistic variation in Jewish Babylonian Aramaic te~s will be provided in a future article. However, it seems that both available editions of this bowl may be improved upon, and accordingly 1

·The body of this article was written in the summer of 2002. The appearance of M. Sokoloff's A Dictionarg 0/ Jewish Babylonian Aramaic 0/ the Talmudic and Geonic Periods (RAmat Gan: Bar-Ilan University Press, 2002; from hereon DJBA) has enabled me to provide further examples for some of the phenomena discussed, and to improve my interpretation of the text at several points. It will be noted, however, that my interpretation does not always agree with that presented in the dictionary. The text of the bowl has been collated in the British Museum with kind permission of the trustees.

1C. Miiller-Kessler and Th. Kwasman, 'A Unique Talmudic Aramaic Incantation Bowl', JAOS 120 (2000), pp. 159-65. The same interpretation is refiected in C. Miiller-Kessler's lengthy review of Segal's volume, 'Die Zauberschalensarumlung des British Museum', A/O 48/49 (2001/2002), pp. 115-43. Our bowl is discussed therein on page 129.

2J.B. Segal, Catalogue 0/ the Aramaic and Mandaic Incantation Bowls in the British Museum (London: British Museum Press, 20(0), pp. 92-93, pl. 53.

,Rםacl The Continuum Pub1ishing Group Ltd 2004, The Tower Building, 11 York @ . London SE1 7NX and 15 East 26th Street, 8uite 1703, New York, NY 10010, USA

Aramaic Studies 2.2 (2004) 208

present here my own reading and translation. Overall, the readings of MKK have been preferred, but it seems to me that their many textual emendations require reconsideration. Needless to say, 1 have benefited greatly from both the available editions.

For the sake of comparison, 1 shall begin by presenting my collation of the text accompanied by an apparatus listing MKK's and Segal's readings. Detailed discussion of their interpretations will follow in my commentary. My own edition is presented schematically, to highlight the repetitive structure of the text which often enables us to improve the readings. 1 hope to show that while not free of difficulties, the bowl presents a text that is more cogent and better structured than either of the existing editions suggest.

Transcription and translation . 2

A ) 1 ( אבהיבתאנאכני

אחחnבגונשזדוכח

דמיבאלבבליאה ) 2 (

אבהיבתאנבאסופי

אחח ) n) 3בגונשזדוכח

דמיבאלבודםסיחא

B אבהפחחיאהבארעא

כיפליאלדאביש

נאהדרמאמשיהי ) 4 (

לי t3פםיאלדאביש

אבהמרידאההדזיפא

מיבי ) 5 (אכיליאלדאביש

נאהמרריבהר

מיביחשילאדאביש

C רחיץביחי

מרימאאיpסןפחי

D בישיחדשי ) 6 (אליארחומללחאפקיפגעי

E בגושבדוכחאבהnאחח

מלנלאבפקבאאלפיהו

2# . overcome' (see Sokoloff, DJBA, p. 700 s.v ' 3םציThe scribe originally wrote . 9 reach'. Compare line ' סמיbut altered the text to םצי),

@ 2004.

MORGENSTERN A Recently Published Magic Bowl 209

כישיחלרשי ) 7 (הלוומארנאוס

ולמלחאקפיפגעיפקי

F 4מדיכאנלאאכולדוח

5מידש;גחאאירחשווח

מידשיפנארשףוחו

" 8) D ( בישיחרשימלילהאכןולמלחאפקיפגעיפגע

מידאכלחניכוהיכי Fו

מדחשיוניחשי

מדשיפחונישוף

אחפחיחא ) 9 ( 6דארעא Bו

לךכיףאלדנאישאחדרמאמשי

מIבילךלאדאניש

אחמרירחאהרזיפא

אכימינךלאדאניש

חאמרארי ) 10 (מדנהרמיכןחשילאדאניש

רחיץביחיך Cו

מרימאאיסכופחיך

G אבאדניכו 8היזךדיואילא

מאנשחרניכו

קמחיכוIבחיעל

להופילוזילו ) 11 (

" F ' דנהמאבלסיה

וניחבןמיניהדניכו

דמיאבחצביה

וניחבןמיבהודבישחי

דמישחא lבאצוהיהס

. 4TIle kaph is malformed, and the reading is according to the context 5The reading should be.מיתשדיגא

. 8 at the start of line למילof מof this word has intruded on the 6הThe which is written at the start of the next 7 ,,מרראי A false start for the next word

. line The extended head of the third letter lends it the appearance 8 :.מיגדCorrect to

of a kaph rather than a nun, which lacks a left-protruding upper stroke in script . on this bowl

. 9For the interpretation of these letters, see commentary

@ 2004.


וניחבןמיניהדנירשף

הלסמאןאמןצבאוחיהוהחיוpסברשם ) 12 (11 : Previotls readings

) 2 ([דמינא Segal :.רמינא 11 ) 3 ([דמינא Segal :ר~ינא 11[כילפי MKK :לי.כיף 11 ) 4 ([הי Segal :.הו 11לי[tפמי MKK :לי.מצי Segal :.מצילי 11מרירחא[החיפא

: Segal כאי 11 ) 5 ([מררי.MKK emend to מרידחאזיפא.חד 11[אכילי: Segal .מרדי 11 ) 6 ([פגעי Segal :.פקעו 11[קפי MKK emend to.פקדי Segal :·קפו

11[מללנא MKK emend to.מלמנלא 11 ) 7 (קפי l] MKK erase. Segal :.קפו 11

: Segal דחאיכול.: MKK קפו 11אכול[דוח.: Segal 2.פקדי] MKK emend to פקיאכול.אחו 11[מידאכלנא MKK :דאילנא,מי emended toארכינלא.מי Segal :מיד

: MKK ואתיתשי 11[מידשינחא.emended to וחיאיתשי,: MKK זאלנא 11איתשו[ורח.דשיחנא,מי emended toדתשינא.מי Segal :דישנחא.מי 11רשף[וחו MKK :.וחישוף

: MKK שיפנא.מיד 11 ) 8 (בישיחרשי[למיל: Segal דשיפנא.מי: MKK 11[מידשיפנא

בישילחרשי,מלי emended toבישילחרשי.מסלילאנ Segal :בישיחרשי.ומליל 11

[פגע MKK erase .[קפי MKK emend to.פקדי 11[האכן MKK :.האכי 11[היכיMKK אכלח.מיד 11[מדתשי: Segal דאכלח.מי: MKK חיכי 11[מידאכלח.: MKK

emend toדשחיח.מי 11[מדשיפח MKK emend toדשיפח.מי 11 ) 9 (לך[כיף Segal :.כיפלך 11[מטילך MKK :לך.tפמי 11מרירחא[החיפא Segal :זיפאחד

.מרידחא 11[אכימינך MKK :מינך.אכי 11 ) 10 ([מרארי Segal :.מראדי 11[מיכך MKK, Segal :.מינך 11אאבדניכוחיזךידו[איאל MKK :אבדניכוזידנא.אילחאיך

emended to אמשרניכו,: MKK ניבואבדזידנא.יאאלחיך 11[מאש[ד]רניכו: Segal .מדשרנכיו Segal :.מאדשניבו 11 ) 11 ([ופילו Segal :.וכולו 11[וניחבן Segal :ניחבל

) all three occurrences (. 11[באצוחיה MKK :.באובחיה

3. Translation

, 1 I sit at my gate, 1, Gusnazdukht daughter of Ahat 2 I look like the Babylonian. I sit at my vestibule, 1, Gusnazdukht

daughter 3 of Ahat, 1 look like the Borsippean. 1 am on the wide Earth, that

, no man can overpower me 4 I am the high heavens, that no man can reach me. 1 am the bitter

, harzipa, that no man can eat me . 5 from me. 1 am the Merari River, that no man can drink from me

My house is secure, my threshold is raised. The evil sorcerers came , to me

. See commentary אנוחיה.lOPerhaps read 11 Bot!l MKK and Segal present their transcriptions in transliteration. 1 have

-rendered their versions into the Hebrew script, which naturally involves the em . ployment of finalletters which their transcriptions do not distinguish

@ 2004.


6 Pag'i, Paqi and Mallalta. 1, Gusnazdukht daughter of Ahat go out to them, I speak and say to them,

7 to the evil sorcerers, Paqi Pag'i Paqi and Mallalta: 'Go eat from what 1 eat, and go drink from what 1 drink, and go anoint from what 1 anoint.

8 The evil sorcerers, Paga', Pag'i, Paqi and Mallalta, said thus: 'How can we from what you eat and drink from what you drink and anoint from what you anoint? For

9 you are the wide Earth that no man can compel, and you are the high heavens, that no man can reach. You are the bitter Harzifaherb from which no man can eat. You are the

10 Merari River that no man can drink. Your house is secure, your threshold raised!' 'If not, ... to your practitioner, to your sender, to your trays (?) of flour.

11 Go and put it in his breadbasket that he may eat from it and be sickened, in his water barrel that he may drink from it and be sickened, in his troughjflask of oil that he may anoint with it and be sickened'.

12 In the name of Tiqos YHWH Sebaoth Amen Amen Selah.

4. Commentary

to sit at the gate' is found in the ' אבבאית"בThe idiom 1 .יתבנא:אבבי . b. B ( דינאולידיןאבבאליתיבהאיוכוליוחכיםהואיל. Babylonian Talmud, e.g

since he is so clever, let him sit at the gate ' נ)Bat. 58a [MS Hamburg 12 . and judge'. Compare b. l:Iul. 99b

The interpretation 'portico', suggested by MKK, is clearly באסופי:

superior to SegaI's 'rubbish' (apparently based upon Hebrew,(אפשה and provides a parallel with.בבא

is the פm יתאMKK and Segal agree that the reading פ:m 3 .יתא is found in line 9. It חפיתא,, result of dittography. The correct reading

seems likely that the error occurred as a result of the scribe's breaking ,' to refill his pen. SegaI translates 'crumbling פm יoff after the letters

but it seems that MKK 'wide' is to be preferred. MKK overlook the . 9 presumably in the light of the parallel in line בארעא,preposition of . Certainly, assuming a scribal error here produces a smoother reading

See aIso our comments onדרמאמשי.הי/הי

12Soo further Sokoloff, DJBA, p. 183, s.v ..בבא

@ The Continuum Publishing Group Ltd 2004.


-Segal translates 'restrains', while MKK prefer 'bend'. How כיפלי:-ofien has the meaning in Babylonian Aramaic of 'over כרף/כפף, ever

13 . ' come, force and translates 'It is 1 myself דרמאמשיהרSegal reads 4 .דרמאשמי:הי

that cast', but this interpretation cannot be supported in the light of line 9, nor does it fit the context. MKK do not relate to the first two letters in their translation or discussion. Prof. J. Naveh has suggested oral communication) that it may be a presentative particle. We may (

-14 Al היד.or האיbehold' or Mandaic ' האthink of a form such as JBA though MKK's translation of 'the high heaven' is adopted here, it is

for regular משיnot without problems, most notably the unusual form In favour of the interpretation 'heavens' stands the statement שמיא.

whom no man can reach'. One possible explanation for ' לימטיאלראבישthis phenomenon is that this form represents metanalysis of compound

was pronounced / ad-/ as it דS;Jmayyadramii/, 15 and that the particle / is in the Mandaic reading tradition. 16 The rarity of the / -ayya / ending

, in this dialect might have encouraged such a metanalysis. Alternatively the spelling may reflect an early attestation of the neo-Aramaic plural form /simme/, found in the Jewish Aramaic dialects of Northwestern

17 . Iraq MKK's reading, based on the Talmudic parallel from מרירהא:דחיפא

an ' מרידתאזיפאחדb. Pes 39a, is clearly superior to Segal's bizarre offensive eyebrow'. A form identical to that in the magic bowl is found

18 . in the Columbia manuscript of Pesahim MKK assume here dittography, and correct the text 4-5 .מיבי:אכילי

-with the apocopation of the lamed of the partici מיבי,אכיto read ple / iikil/. The apocopation of a final consonant is a marked feature of Jewish Babylonian Aramaic, and is widely attested in manuscripts

. 597 . 13SokoJoff, DJBA, p , 14T11. N6Jdeke, Mandiiische Gmmmatik (HalIe: BuchhandJung des Waisenhauses

. 72 § ,) 1875 15My reconstruction of Jewish BabyJonian Aramaic phonology is based upon

tlle detaiJed study of the earliest manuscripts. A full description will be published . shortly

lfiR. Macucll, Handbook 0/ Classical and Modern Mandaic (Berlin: Walter de . 10 . Gruyter, 1965), p

, 17y. Sabar, A Jewish Neo-Ammaic Dictionary: Dialects 0/ Amidya, Dihok , 300 . Ne1Wa and Zakho, Northwestern Jraqi (Wiesbaden: Harrassowitz, 2002), p

.) simme ( משה. v מ.. with literature 18 .,הרזיפאNoted in Sokoloff, DJBA, p. 390 s.v



9 demonstrating phonetic orthography.19 The comparison with lines -and 11 supports this correction. It would seem likely that the extra

line 4). An ( ליטמיline 3) and ( כילפיresulted from the forms ליneous for ליהhere appears here instead of ליalternative explanation is that

20 .' the reading jlej, and is the so-called 'ethical dative 5 .אנהמררי:נהר The readingמררי ) MKK) is superior to Segal's,מרךי

both in terms of the material reading and the context (see the following 1 ' from whom no man can drink'). MKK translate ' מיניחשיאלךנאיש

am the bitter river'. However, the emphatic form implies that this is . b בורןנהר, b. M. Qat. 4b בורניץנהרperhaps a proper noun. Compare

b. 'Abod. Zar. 39a. It is feasible that t,his 24 . ,נומךאונהרנחןנהרb ~ B. Me was the name of a river that would have been familiar to the readers

. and writers Segal's translation, 'my house is secure', is to be preferred רחיץ:ביחי

;' over MKK's assumption of an Akkadian loanword 'to fiood, rinse there is no reason to follow them in rejecting the Aramaic root in this context. Up to this point, the speaker has declared her strength and might: she sits firm in her home, she is like the unconquerable earth and heavens, the inedible bitter herb, and the undrinkable bitter

,' river. Finally, she declares 'my house is secure, my threshold high emphasizing her security and impenetrability. This line, breaking the sequence of the previous four statements that are all formulated in the

concludes section B כי,... identical pattern ('1 am ... whom no man can in my schematic presentation of the text. It is refiected again in section

.' B MKK are correct in assuming that these are 6 .ולםלחאפקי:פנעי

the names of evil forces, rather than imperative verbal forms as Segal suggests. It is possible that these words are in apposition withבישי,חרשי

and should be translated as though they are personal names. MKK Since the קפנך]י.to פקיconsistently (lines 6, 7, 8) 'correct' the reading

19 A1though phonetic orthography is found to a certain degree in al1 Jewish Baby10nian Aramaic texts, its widespread use is charactetistic of certain groups of manuscripts, particular1y ear1y Eastern manuscripts and Yemenite texts. For an important preliminary description of manuscript groups, see S.Y. Friedman, 'The Manuscripts of the Babylonian Talmud: a Typology Based upon Orthographic and

-LiIlguistic Features', in M. Bar-Asher (ed.), Studies in the Hebrew and Jewish Lan . 163-90 , g'uages Presented to Shelomo Morog (Jerusalem: Mossad Bialik, 1996), pp

20The e1ision of the final he of the 3rd masculine singular pronoun is quite widespread in Jewish Baby10nian Aramaic texts, for examp1e in the conImon

See y, Kara, Babylonian Aromaic in the Yemenite עלמאכולי =עלמא.כוליהform , 75-76 , Manv.scripts 01 the Talmud (Jerusa1em: Magnes Press, 1983) pp



, is never attested פךיpis repeated three times in this text, while פיpform the suggested correction would not appear to be justified. The names

may have sounded almost identical to the scribe. The פיpand פגעי

Babylonian Geonim attest to the dropping of the 'ayin in the form :פיקעאnבחחוניוחשניבקעקמוםהואלקלושןפיקאלמימררנבןרגיליןבפיקעא

Harkavy, Responsa, p. 22 lines 29-30). Interchanges of qof ( לשאךם... and gimel are less common in Jewish Babylonian Aramaic, though not unknown, particularly in loan-words. Compare, for example, the

21 .' club ' גולפאand רקלפאgum Arabic', or ' קומאand גומאspellings -Segal and MKK both read as a participle, with MKK 'cor מללנא:

The form in the bowl would appear 1נלםלנא[].recting' the reading to to refiect haplology or the assimilation of the participle morpheme to the first root letter. A similar phenomenon is attested in the Yemenite

= ) b. Rosh Hash. 15a ( ממששי: manuscripts of the Babylonian Talmud . in Spanish MS Enelow and MS St. Petersburg Antonin 502 fol ממשמשי

. 5 and אכול(א)רחThis reading is derived from Segal's 7 .אכול:ךחו

While the first letter is malformed, MKK's reading of ךחיאכול.MKK's , a dalet is clearly superior to Segal's suggestion of an aleph. However

Segal's reading of two words is supported by the subsequentאיהשו,וחוto -ךis hard to justify. The use of the חיאיהשוfor which the reading

22 . introduce direct speech is widespread in the Aramaic dialects with כאולו,represents an apocopated form of the imperative אכול

the loss of the final unstressed vowel. Such forms are attested in the Yemenite manuscripts of the Babylonian Talmud, and are the standard in Classical Syriac and Mandaic.23 The apocopation does not affect the

in which the ending was derived from an original diphthong ךחן,form , and was certainly stressed. For the form in the Babylonian Talmud

b. B. Bat. 22a [MS ( רבאבימשינאבשראאכלןחןאבייביגרמיואבםרשיחןsee b. Ber. 25a [MS Oxford ( רחm גרסיוהניאבונהנירב.ךביציפיהניו.)] Hamburg

bears the prosthetic aleph, which is attested 24.([366איהשו The form . b ( איהשיחוטףכאולחטוף. on both perfect and imperative forms, e.g

is >כIn Mandaic the shift from.p לפא.כןלפאp21 .,וSee Sokoloff, DJBA, p. 991 S.v . 41 § common in dissimilated fornlS. See Noldeke, Mandiiische Gmmmatik,

12 22See J.N. Epstein, 'Notes on Post-Talmudic-Aramaic Lexicography', JQR . 1.4 meaning -1922 .ד), pp. 305-307,382-83, and Sokoloff, DJBA, p. 307, s.v (

-23Kara, Yemenite Manuscripts, p. 152. Th. Noldeke, Compendious Syriac Gmm -fnar (LondoIl: Williams and Norgate, 1904), 158C. Noldeke, Mandaische Gmm '

. 173 § , 1natik 24Compare Sokoloff, DJBA, p. 176 s.v ..אחי



Erub. 54a [MS Vatican 109]) 'eat quickly (and) drink quickly,.25 An ' : imperative form with the prosthetic aleph appears in a magic bowl

Naveh-Shaked 7:8).26 Sokoloff ( דמיהמןאריתשאיבישריהמןאיכוליעלהויפולי

as in the citation אכל,notes that this verb is often found together with for they eat, drink ' ורווחשוודאכלןfrom b. 'Erub. brought above, and

27 .)] and get drunk' (b. Meg. 13b [MS Columbia It seems preferable to read jmidda'ii};liiniij, given the deמידאכנלא:

fective spelling ofמדשיפח in line 8 and the parallelsמיניה,דניכודביתשיin line 11. MKK's suggestion to read the prefix מיניהדנירשףand מיניה

, is not impossible. However -מאידas equivalent to the Talmudic מיד

is either רמשיפחin that case, we would have to assume that the form a scribal error or refiects the contraction of jmayda-j > jme:da/. The

-is found in Babylonian Aramaic in the in מיand מאיinterchange of terrogative particle.28 However, given the parallels of line 11, it would

/-in this text as refiecting /midda -מידseem preferable to view the form . and 1 have translated accordingly here

-This is the reading of the bowl. MKK suggest cor 8 .בישיחרשי:מליל but such a drastic emendation נמ]למי,j [בישיחלרשינלאrecting the text to

is unnecessary, and significantly alters the sense of the narrative. Seas verbal predicate of the subject מלילgal correctly identifies the form

but assumes that the content of the evil sorcerers' speech is בישי,חרשי

while 1 assume, as mentioned in פי,פגעיpלאnולמcontained in the words my commentary upon line 6, that these are evil spirits themselves. The

for the plural refiects the apocopation of the final unstressed מלילform 29 . vowel

25For the imperative, see J.N. Epstein, A Grnmmar oJ Baby/onian Arnmaic . Jerusalem: Magnes Press, 1960), p. 96, and N6ldeke, Syriac Grnmmar, §51, 176D (

The perfect forms are well attested in the Aramaic dialects, e.g. Biblical Aramaic j see A. Dodi, 'The Grammar of Targunl איתשיתי: I:t, Dan 5.3,4, and TgOnq ין~

Onkelos According to Geniza Fragments', PhD Thesis Submitted to the Senate of Bar-Ilan University, Ramat Gan, Israel, 1981, p. 69. For other forms in Jewish

. 281-83 . Babylonian Aramaic see Kara, Yemenite Manuscripts, pp 26Cited by H. Juusola, Linguistic Pecu/iarities in the Arnmaic Magic Bow/ Texts

. 191 . Studia Orientalia, 86j Helsinki: The Finnish Oriental Society, 1999), p ( . 1c 27 .תשיSokoloff, DJBA, p. 1184, s.v

. and endnotes 28 .ז;ניSokoloff, DJBA, p. 661 s.v , 29See E.Y. Kutscher, Hebrew and Arnmaic Studies (Jerusalem: Magnes Press

1977), pp. 243-45 (Hebrew Section), Kara, Yemenite Manuscripts, pp. 144-45. For , possible attestations of such forms in the magic bowls, see Juusola, Magic Bowls

. 164-69 . pp



and translates '[and the (ן)ני(כו)היכיאהכSegal reads ניכוהיכי:האכןspell was] thus. How? 'We shall eat". However, as noted, he did not identify the names of the demons and his translation is fragmented and

Thus my palate will eat', and translate ' ניכוחיכיהאכיforced. MKK read all of the nun-prefixed forms in the third person singular. But this

as the palate does not ומשיפת,ונישוףreading does not suit the statement anoint. Moreover, the 'justification' section, which begins withראערה

,) and continues to line 10 (B' in our schematic presentation אתפתיתא

is clearly addressed in the second person to the speaker in the opening of the text, Gusnazdukht daughter of Ahat. MKK's ascription of these words to Gusnazdukht herself depends upon their emendation of the

. text in line 8 and cannot be justified from the text itself The simplest reading may be gained by assuming that the direct

-how can we eat?'.30 The di ' ניכו,היכיspeech begins with the words , rect speech represents the demons' response to the invitation to eat

to drink and to anoint themselves. The demons complain that they cannot partake of the same pleasures as the mighty Gusnazdukht. The

and ' מליל,is taken as an adjectival complement of the verb האכןword is marginally better than האכןthey said thus:'. Graphically reading

though it assumes the use of a medial nun in a final האכי,the reading . position

8-9: Here we find the demons' explanation for their inability to partake of the delights that Gusnazdukht enjoys, repeating her words

. from the opening section -The material reading here is dif 10 .אמשרניכואאכרניכוזירותיר:אילא

and translates 'your מאשרניכוניכואכרזיראנאיאלתיךficult. Segal reads -as proper names. This sug מאשרניכוand אכרניכוentry haughty', taking

gestion has little to recommend it. MKK readבארניכוזירנאאילתאיך-and translate 'your upper room is inundated; your prac םi{א}מר,j ר]רניכו

titioners of (a ritual), your scatterers', but this interpretation also de.' in line 5 means 'flooded, rinsed רחיץpends upon the assumption that

The key to the passage seems to lie, as they suggest, in the parallels : from similar magic bowls. MKK cite three parallels

-1 British Museum 91776, first published by Gordon in 1941 and pre sented according to Miiller-Kessler's collation :עלחזרוזאילואיאלדלחמאכסליהליהופילוזילורמשררינכומרiםרנ]אועלרכערינכואנ jעכר

ולישוףולזיקסיאוליתשי ] ...וללחאמrונליכולרסי[iםחאנתי]ה Jכארמיאכחצכיה

rather, go and return to the practitioner who produced ' וליכוכסישחא

30S0 Sokoloff, DJBA, 377b, S.V ..היכי



you and to the sender who sent you; go and fall into his bread basket, into his cask of water, and into his fl[as]k of oi[l] , and let him eat bread and], drink water and be harmed, and anoint oil and be [

.' pained , 2 A bowl published by Gordon in 197831 (Zion Research Library

Boston, no. 48) lines 8-9 : 32דפנכותקןלביאתעוולובהדתאיחוןבאורחאחדרודמיאומבאומאנישבעאלדיכאלבממביכוןדחלמאבאלסלהופילומיביה )!(

Return by the path on ' ויכעובויכעבדישוףדמשאחוכמבאויקפדיחשי

which you came, and enter the house from which you left, and fall into the bread basket that he may eat and not be satiated, and the water vessel that he may drink and ... 33 and into the oil vessel that

'. he may anoint and be pained A text published by Gordon 3 :עלויאזלוןויחדרוןלאחוריהוןיחהפכוןוכוהלו

משדרביהוןעול.עבדביהןן 34

The parallels would suggest that the reading should be :אאבדביכן ...איאל, If not, ... to your practitioners, your senders ' מקחיכןחטיעלמאש[ד]רביכן

paralleled in BM 91776 cited איאל,to your plates (?) of flour'. The word above, refers to the aforementioned invitation. The protagonist says- to the demons, if you do not or cannot partake of what 1 partake, then

. return to your sender is very אאבדביכןand אילאThe interpretation of the letters between

uncertain, and the text may be corrupt. Repeated inspections of the bowl in the facsimile and in the original have not provided a wholly

31C. Gordon, 'Two Aramaic Incantations', in G.A. Thttle (ed.), Biblical and Near , W.S. LaSor (Grand Rapids: W.B. Eerdmans ןסEastern Studies: Essays in Honor

1978), pp. 231-43. The text was published with an accompanying hand-copy, but without a photograph. Some of the readings could perhaps be improved by viewing

. tlle original 32The scribe initiaily wrote a kaph, but erased it and wrote in wrote a qoph. This

which, while rare, is not unknown in Jewish Babyloכ<ק,may refiect a shift of , Sokoloff ( קפיזא/בכיזאnian Aramaic texts. Note, for example, the Persian load-word

,)] San. 29b [MS Yad Harav Herzog ( עככראDJBA, p. 1032), or the personal name er Halachot Pesuqot, Codex Sassoon ןSe ( ראpעכthat appears in Geonic literature as

263 [Facsimile Edition, Jerusalem: Makor Publishing, 1971], p. 134line 15, parailels Hilchot Reu, ed. Schlossberg [Versailles: Cerf et fils, 1886], p. 61 line 28). For the

.' shift q>fricativek in modern Arabic, compare Iraqi Arabic waxt « waqt), 'time is probably incorrect. Perhaps in the light of BM 91776, cited 33ןיפקThe reading

above, we should read.ןיזיק,' 34C.H. Gordon, 'Aramaic Magical Bowls in the Istanbul and Baghdad Museums

Archiv Orientalia 6 (1934), pp. 319-34. 1 have slightly extended the citation here 1 read ויתהפכוזכולהand made a minor correction to Gordon's reading (in place of

jיתהפכן.(כןהלן



satisfactory material reading. From the parallels we would expect to find here imperative forms meaning 'to back' or 'return'. It is tempting

refiect a scribal error תיףירוto assume that the letters that appear as a minor emendation, particularly in the light of the resh תיהררו,for

written as a dalet in the wordארעה ) line 3). Instead of MKK'sאיתאליךyour upper room is inundatedj your practitioners' we get ' משרניכוזירנא

if not, go back to your practitioner'. This reading ' אמשרניכותיהררו,איאל

and interpretation both suits the context and is paralleled in meaning though not in precise formulation) by the other magic texts, but is (

. based on emendation Following MKK's suggestion, on the basis of Gordonועל)ערבניכוןעל

, However מארשרניכו.to 35מאשרניכו we have emended the form 1,(משררניהו the interpretation suggested here does not require the deletion of the initial aJeph. Rather, it is to be taken as the phonetic equivalent of the

. found in Gordon's text עלStandard Aramaic MKK translate the whoJe expression, according to their reading

your practitioners of (a rituaJ), your ' קמחיכןטחיעל}מש[]דרניכו{אאברניכו

scatterers concerning the spreading of your fiour', and convincingly , with the use of Akkadian epesu 'to practice witchcraft עברניכןlink . perform a divination, ritual' (CAD E 191 b s.v. epesu and 207b s.v -to an Akka משררנאdullu). However, their attempt to link the word . dian apotropaic ritual mentioned in the Maqll1 texts is less convincing

is not רש"רin Aramaic means 'one who sends', and the root משררנא

. llsed for scattering -used with this mean מקIn Jewish Babylonian Aramaic, we find

ing,36 including in a magic context :רביתיהאקסופאnעלאובפאיהומוקand cast [the earth] before him or on the threshold of his house' (Harba '

de-Moshe 42.10).37 The verbרשי is also employed in this meaning :רהוהwho would cast dust from it and ' ניריוהוןניליחרביךוהוועפראמניהשארי

. it would become swords,38 straw and it would become arrows' (Ta'an 24a [MS Yad Harav Herzog]), and in a magic context :למצמרבעיתאם

. 35See the previous footnote ,) 36D. Boyarin, 'Towards the Talmudic Lexicon 11', in M. A. Friedman et al. (eds

Teuda 111: Studies in Talmudic Liternture, in Post Biblical Hebrew and in Biblical .) Exegesis (Tel Aviv: Tel Aviv University, 1983), pp. 113-18 (Hebrew

.) 1997 , 37The reference is to Y. Harari, Harba de-Moshe (Jerusalem: Akademon This example is drawn from Sokoloff, DJBA, p. 195 s.v ..בזק

lt is either a graphic error עפרא.f סis to be taken as the verbal predicate 38ןהןן

or the result of attraction to ןהןן),influenced by the following correct form ( ןהןיfor tlle plural conlplement.חרבי

. 2004 Publishing Group Ltd nשThe Continu @


אלפיהוןוךשיייהיטנאי"וערייגבניה"מןרמשאלרנלךרחתמותעפראעלאמ'יתהון ' and if you want to bind them [wild animals] recite fromגבניה toהינתאי

magic words in the Harba text] over the earth under your left foot [ and cast it before them'. Compare also the less specific useס f:הר"רהאירתרננולאככרעיחזיולמחראפוריהולהיררהנהילהקיאמטלייתילמידערבעימאן

) Ber. [MS Oxford 366] 6a) 'Whoever wants to know (about the demons ( should bring sifted ash and put it around his bed, and the next day it will look like chicken feet'. These usages would appear to refiect closer

. here מךשרנאparallels to Babylonian magic practices than the use of Our text is better translated 'Go back to your practitioner, to your

sender'. Moreover, as 1 have tried to demonstrate above, the aleph should not be deleted, but is paralleled by the aleph of מאש[ר]רניכוon

.אאברניכו

As noted above, the motif of sending spirits back to the sender is quite common in the magic texts. Indeed, the frequent openingהפיכי

which is often translated 'overturned', may also be translated הפיכי,

. has both meanings הפךturned back', as in Aramaic the root ' The meaning of this verb has not been clearly 11 .וניחבןמיניה:רניכו

-established. Segal translates it 'be spoiled', based on his incorrect read MKK leave it untranslated. Sokoloff has tried to connect it to וניחבל.ing

, to cease', and suggested reading the form here as an ithpe'il ' חבן,Syriac -meaning 'and it will be left over'. The interpretation suggested here re

39 .' which can mean 'to become sick חב"ל,to JBA חב"ןlates the root , The shift of l>n is quite well attested in Jewish Babylonian Aramaic

of bread' (line ' רנהמאand is attested elsewhere in this bowl in the form 40 It is also found 11.חלמא), which is derived from Standard Aramaic

Come let us all be one ' חרעמאכולןליהןיתאin verbal prefixes, such as -people' (b. San. 39a [MS Yad Harav Herzog]).41 In verbal roots, it oc

curs in root-initial position in the common Talmudic formנקט >.לקט 42. b ( מירוין, In root-final position there is apparently one other example

B. Bat. 12a [MS Hamburg, Harkavy, Responsa, p. 26 line 11]), which 2.25.43 in the Geniza fragment TS G מירוילappears also as

, 391 owe this suggestion to Prof. J. Naveh, who also drew my attention to Sokoloff . is well recorded חבלwhere this Ineaning of חבל,. DJBA, p. 426 s.v

. 19 . 40See Epstein, Grnmmar, p . 167 , 151-52 . 41See further Kara, Yemenite Manu.scripts, pp

. 366-67 . 42See Epstein, 'Notes', pp ,) 1942 , 43published by S. Assaf, Responsa Geonica (Jerusalem: Mekize Nirdamim

. 143 . p


Aramaic Stv.dies 2.2 (2004) 220

is found in line 8 of חב"ןFurthermore, is possible that the root , bowl 11 from the Mousaieff collection. Shaked, who published the bowl

transcribed this lineביחיהמןיחכוןינפיקיחכוןיחב[]לוןחבאלעליכוןוהיחיחand translated 'and 1 brought upon you an unjurer חחויבראבונא,ךהךין

that may injure you, and may cause you to leave the house of this relies on יחב[]לוןf סAbuna son of Hatpoi,.44 However, Shaked's reading

an emendation, and also produces a plural verb in a context where a singular verb is required. Accordingly, it seems preferable to read,יחבין

involving only the minor change of Shaked's material reading from a waw to a yod and requiring no change in translation. The form is a

appears in חב"ןpa "il imperfect 3rd masculine singular. The verbal root , is employed in a parallel text חב"לthe Mousaieff bowl where the root

Montgomery 9 lines 7-8 : 1]יחכ[וואלקפאיחהוןלחבאלהמבלאעליכוןאריחיחי45 If the verb in the Mousaieff bowl is indeed in the pa "il, it בחיהו.מןj

seems likely that the verb in our bowl is in the itpa"al, an assumption that would suit the defective spelling of.ניחבן

thus refiects the other similar curses cited from the ניחבןThe word magic bowls mentioned above, many of which relate to bodily pains caused by bewitched food. We may note that the idea that demons dwell within food is also found in the homilies of Aphrahat, the scholar

: of the early fourth century, wherein we read

r<'a בנטנ.aז:r6ני~ cו..מ :ta r<.ז. ~ r<' ז:a r<'aUr<' _rנ CL1 ~ r<' , :tב r<' .x.נ~ ~ 't.ככ.פ~ m ז:r<'aUr<' ~ :ta י~ז::םt.:נ.a ~זנ..בבונC

cז:ו..מ a u .נב:עr..::ני a בכיב:ta U\.., :זם::יינ.נו:r6נניr. :ז:~ז m ~ :ז'iכ ~ r<'

a ~r<ב ma :ו..~ cו..מ :ta ~ .ou.a ~זנ..בבונCr<l...... :זr<z....:כ ' And he says: 'AII foods are sanctified by the word of God and by prayer

1 Tim 4:4). So when somebody has need of one of God's creatures, be( ' fore he eats he invokes and praises and blesses the name ofthe creatures

creator, and prays, and then they are sanctified, and the evil spirit flees 46 . from within them

-44S. Shaked, 'The Poetics of Spells, Language and Structure in Aramaic Incan . tations of Late Antiquity. 1: The Divorce Formula and its RAmifications', in T

Abusch and K. van der Toorn (eds.), Mesopotamian Magic, Textual, Historical and Interpretive Perspectives (Groningen, 1999: Styx Publications), pp. 173-95. The

. 188-89 . transcription of this bowl and its translation appear on pp , m Nippur (Philadelphia ןro 45 J.A. Montgomery, Ammaic Incantation Texts

. 161 . 1913), p : Aphmates the Persian Sage (London 46ןסW. Wright (ed.), The Homilies

. 3-7 . 30911 . Williams and Norgate, 1869), p



, was suggested in MKK אכובחיהThe reading 11 .ךסיהשא:אכובחיה explained in a separate article,47 and preferred over Segal's reading

-in Miiller-Kessler's review.48 In the light of the parallel materi אכצךחיהis indeed באובחיהals that Miiller-Kessler has cited, it seems likely that

-the correct reading. However, the right-hand stroke that lends the let ter the appearance of a sade may be clearly discerned on both the original bowl and in the photographs, and is written in the same ink as the remainder of the text. The scribe may have begun to write a different letter before breaking off to write the nun, without erasing

cannot אצןחיהthe previous pen-stroke. As the text stands, the reading . be definitively excluded

A common closing formula, based on the doubled use of לסהמאן:אסןalready found at the end of the books of Psalms and used further מאן

in Second Temple liturgical settings.49 Its use in the magic texts is 50 . widespread

5. Conclusion

The interpretation of this magic bowl presented here requires far fewer textual emendations than found in MKK's edition, and proves a far more coherent text than provided by Segal.

The narrative is based around a simple structure. The opening section (A) sets the scene, and introduces the 'heroine' of the bowl, who is presumably the client. In (B), she declares her invincibility, culininating in her proclamation (C) that her house is secure and threshold raised. (D) reports the arrival of the demons, while (E) reports the heroine's approach. The demons' names are listed again in (D'). The heroine challenges the demons in (F), and they respond (D") saying

36 in the Babylonian Talmud', ANES תנאאf 47סC. Miiller-Kessler, 'The Meaning . 169-71 . 1999), pp (

. 48See note 1 above Ps. 41.14, 72.19, 99.53. For the Second Temple 49מאוןמאןFor the Psalms, see

: liturgy, see E. Chazon, 'A Liturgical Document From Qumran and its Implications 102 . Words of the Luminaries" , (PhD Dissertation, Hebrew University, 1991), pp "

.) Hebrew ( 84a, as noted aשappears in the Babylonian Talmud, YO 50לסהאמןמאוThe formula

by Montgomery, Ammaic Incantation Texts, p. 63. It has now also been found ill : Palestinian amulets, such as J. Naveh and S. Shaked, Amulets and Magic Bowls

Late Antiquity (Jerusalem: Magnes Press, 1987), amulets ןסAmmaic Incantations Late Antiquity 1.12,2.11,4.23ןס, Magic Spells and Fonnulae: Ammaic Incantations

. 29.5,9 , 19.33 , 17.32-33 , 16.24-25 Jerusalem: Magnes Press, 1993), amulets (



-cibil iזthat they cannot take up the challenge (F') on account of her invi .)' ity (B'), which they recount, culminating in the proclamation in (C

Since the demons have failed to meet the challenge, they are sent back against their sender (G), who must now face the dangers he sought to

. impose upon the heroine; in effect, meet her challenge The text has been carefully crafted. It contains an opening in which

-the heroine is presented and her identity clearly established by repeti , tion. The main part of the narrative is told through repeated stanzas

which themselves are structured around a series of parallelisms. The indicates that the text is drawing to a close. The denoueאילאword

ment of the final lines, in which the three verbs of eating, drinking ' and anointing are repeated once again in curses against the demons

sender, provides the dramatic climax, while the post-scriptלסהמאןמאן51 . signals that the text is complete

51The formula also appears in magic texts signaling the end of a section. 1 shall return to the literary aspects of the magic bowls in a future article.

. 2004 Publishing Group Ltd nשThe Continu @

[AS 3.2 (2005) 203-204]DOI: 10.1177/1477835105059093

Additional Notes to the Aramaic Magic Bowl

BM 135563

Matthew Morgenstern

University of Haifa

In a recent issue of this journal I republished an Aramaic magic bowlBM 135563.1 In this brief note, I wish to present an alternative inter-pretation of one difficult word, and to provide further support for theinterpretation of another.

10. ykyjmq yjf—In my article (p. 211) I hesitatingly followed Muller-Kessler and Kwasman’s interpretation and translated ‘your trays (?)of flour’. However, as is clear from the entry in M. Sokoloff’s dictio-nary, ayjf is really a greased dish for an oven, not a tray for flour.2

Prof. Moshe Assis of Tel Aviv University has suggested to me that theform yjf may be better understood as an apocopated form of the Qalmasculine singular participle from the root ÷˜˜jf ‘grind’. Accordingly,we would understand ykyjmq yjf to mean ‘the grinder of your flour’.This interpretation better suits the context, and is also supported bythe many apocopated forms found in the language of the bowl.

11. ÷bjyn—In my article (pp. 219-20) I suggested that this word is de-rived from a phonetic variant of the root l˜˜bj, and that the same root isattested in a bowl from the Mousaeiff Collection published by Shaked. Ihave now found further evidence for the root from an important Rab-binic manuscript, MS Paris 1402 of Halachot Gedolot.3 In a citation

1M. Morgenstern, ‘Notes on a Recently Published Aramaic Magic Bowl’, AS 2(2004), pp. 207-22.

2M. Sokoloff, A Dictionary of Jewish Babylonian Aramaic of the Talmudic andGeonic Periods (Ramat Gan: Bar Ilan University Press; Baltimore: The John Hop-kins University Press, 2002), p. 499 s.v. ayjf.

3A linguistic assessment of this manuscript will appear in my forthcoming mono-graph, Studies in Babylonian Aramaic.

c© SAGE Publications (London, Thousand Oaks CA, and New Delhi), 2005

204 Aramaic Studies 3.2 (2005)

from b. BM 97a that appears on folio 30a lines 6-7 we read yrmad akyawtymw ÷bhya yrbk[ lka ‘some say that it ate mice, fell ill an died’. The sametext appears in the Spanish manuscript of the tractate, MS Hamburg165, as tymw lbjya yrbk[ lka.

The shift from het to he characterizes the early eastern manuscripts,4

while the interchange of nun and lamed attests to the existence of thesecondary root already found in the magic bowl. This now provides an-other example of phonetic features of Babylonian Aramaic attested inthe magic bowls and preserved only in the best Rabbinic manuscripts.5

4I deal with this phenomenon in detail in my forthcoming work mentioned inn. 3. Meanwhile, see Y. Kara, Babylonian Aramaic in the Yemenite Manuscriptsof the Talmud (Jerusalem: Magnes Press, 1983, Hebrew), pp. 58-62.

5See my article, ‘Linguistic Notes on Magic Bowls in the Moussaieff Collection’,BSOAS (forthcoming).

Notes on a Recently Published Magic Bowl

Documents

Transcript of Notes on a Recently Published Magic Bowl