The challenges presented by the Aramaic incantation bowl texts are well known to all those who work on them, and there is no doubt that the readings and interpretations presented in the existing editions may often be improved through collation and in the light of new sources. Many formulae that are difficult to read in one copy, due to the shape of the script or damage, are often more easily deciphered in parallel copies. The amount of material that remains unpublished is significantly greater than what has already been published, and this large body of new material provides a great advantage to the present generation of scholars over its predecessors.

In an article published in this journal (BSOAS 75/1, 2012, 1–31), Christa Müller-Kessler (henceforth MK) has offered readings and explanations for 74 ‘puzzling’ words in the Aramaic incantation bowls and related texts.Footnote ¹ Some of these proposals are new, while others had previously been published and are conveniently gathered and arranged in alphabetical order. Since MK's study generally deals with passages that were not satisfactorily explained by their original editors, it naturally concentrates on some of the more difficult and challenging parts of the magic corpus. Some of the explanations proposed, such as אלימני ‘young ones’ (no. 7) or אספריס ‘sphere’ (no. 10), represent a significant advance in the reading and understanding of these texts.Footnote ² Others appear less convincing to the present authors. In this article, alternative readings and interpretations are presented for a number of the passages discussed in the aforementioned study, often in light of unpublished parallels. Wherever possible, the readings proposed here are supported with photographs and representative examples. The numbering of the entries here follows that of MK. However, since not all the entries in MK's article are discussed, the numbering below is not complete but rather jumps between entries in the original article.

1. אזה. On the basis of a parallel offered for sale at Christie's in 2000, MK has proposed correcting Naveh and Shaked's reading ובטילו אזה ‘they annihilated ʾzh’ (AMB B13: 9) to read ובטיל ראזה ‘and annul the mystery’. MK writes (p. 3): “The space between ראזה and ובטיל is quite large, therefore the badly executed resh is not part of ובטיל”. However, Naveh and Shaked's material reading is closer to the correct interpretation. From the unpublished parallels it is clear that the formula is ובטיל ואזח ‘and annul and remove’.Footnote ³ This is also the correct reading in the Christie's bowl.Footnote ⁴ (See Figures 1–5.)

Figure 1. ובטיל ואזח (Christie's bowl)

Figure 2. ובטיל ואזח (JNF 29: 5)

Figure 3. ובטיל ואזח (JNF 30: 7)

Figure 4. ובטיל ואזח (JNF 43: 7)

Figure 5. ובטיל ואזח (AMB B13: 9)

MK further notes: “The plene spelling of ראזה is obviously induced by a Mandaic ‘Vorlage’, since only this Aramaic dialect tends here to spellings with aleph”. This assumption is not convincing. First, the spelling ܪܐܙܐ is amply attested in Syriac, both in the magic bowlsFootnote ⁵ and in Christian Syriac literature.Footnote ⁶ Furthermore, we find the form ראזי, for example, in a typically Jewish context in HS 3047 (= TMHC 7, B17; reading from unpublished photographs):Footnote ⁷

1. 1 אסותא דישלמא על ביתה דאדורב֯יט […] ש̊יר תש̊באחות למאלך יה

2. 2 יה יה יהו מ̊ל̊ך יהוה בצור עולםם אהיא אשר אה̊יא [מלך די ממלל] ב̊ראזי פרי{x}שתאFootnote ⁸ כי אל חז̊ק

3. 3 וגיבורFootnote ⁹ שלח ביד [שלמי]א̊ל רפאלFootnote ¹⁰ קנאלFootnote ¹¹ למ(י)כאל מי..[..Footnote ¹² למיסר (?) כל עין בי]ש̊ה כל מאמר כול לוטה א…

4. 4 שלמ[…]למא

1. 1 (May there be) complete healing upon the house of Ādur-beṭ son of [PN]. A song of praise to the King, Yah

2. 2 Yah Yah Yahu, the King, YHWH, by the Rock of Ages, “I am who I am”, [a King who speaks] with wondrous mysteries, for a strong

3. 3 and mighty God has sent by means of [Šalmi]el, Raphael, Qanael, {to} Michael, ..[.. in order to bind (?) every e]vil [eye], every utterance (and) every curse. …

4. 4 …

Note also the plene spellings with ʾaleph in תשבאחות and מאלך in the same context.Footnote ¹³

The use of ʾaleph as a vowel marker for ā has a long history in the Jewish Aramaic dialects, including linguistically reliable early Eastern Talmudic manuscripts, and cannot be adduced as evidence of a Mandaic Vorlage.Footnote ¹⁴

3. איכא. The attestation of this word in MSF B23 is based on an incorrect reading.Footnote ¹⁵ See below, no. 23.

MK presents a new interpretation of BM 91767: 11 (= CAMIB 040A): דאכא כ/בינא בחילן דילן, which she interprets as ‘where exists understanding/nature in our own power’. However, the proposed translation of this difficult passage is not suited to the wider context. A different translation is presented on p. 22 (entry no. 58, עד אמא ד-), where the passage is cited in greater detail, עדאמי דנמטי זבן ואידן דאכא בינא, translated ‘until comes time and season so that there is understanding’,Footnote ¹⁶ but it is hard to justify בינא as an absolute noun in JBA with this meaning. We may hesitantly propose that the correct reading is עדאמי דנמטי זבן ואידן דאנא בינא ‘until the time and season that I desire shall come’, in accordance with Segal's basic material reading (Segal deletes בינא and puts דאנא in the following sentence). The letter in question looks more like kaf (as read by MK, and by Morgenstern Reference Morgenstern2007a), but the reading דאנא is supported by the following contexts:

1. 1. JNF 322: 5–6:

   והוא ישתמע ל…אגושנסף בר אימי בכל עידן דאנאיבעיה לעלם

   And he shall heed … a-gušnasp son of Immay whenever I desire it foreverFootnote ¹⁷

2. 2. JNF 90: 11–12 (Ford, Reference FordForthcoming):

   תלתין יומי ירחא תרי עשר ירחא שתא שבתא כסא ורישרחא כמא דאנא תליפא בר אימי צבינא מן שעתא ה[דא ולעלם]

   thirty days of the month, twelve months of the year, and on the sabbath, full moon and first day of the month, just as I, Talifa son of Immay, wish, from [this] moment [and forever]

The following example lacks the independent pronoun אנא, but further demonstrates the use of a 1st person singular active participle in a subordinate clause corresponding to ד … בינא in a similar context.

1. 3. MS Berlin Or. Sim. 6, 4a: 7–9 (Bohak and Morgenstern Reference Bohak and Morgenstern2014: 29*):

   תריסר ירחי שתא לא בכצומי ולא בכלדאיי אסיא לא נשתכח ליה א… דשרינא ליה ומית

   twelve months of the year, neither amongst magicians nor amongst Chaldeans may a healer be found for him, [until] I release him and he dies.

4. MK has interpreted כשורי אילן in Moussaieff 145: 11 as ‘beams of wood’. However, אילן means ‘tree’, not ‘wood’.Footnote ¹⁸ Furthermore, Levene's interpretation of אילן as a demonstrative pronoun is supported by the published parallel from MS 2053/159: 12, which reads כשוריה אילין (see Figure 6).Footnote ¹⁹

Figure 6. כשוריה אילין (MS 2053/159: 12)

11. אצמומי. The relationship between this form and the Mandaic forms presented by Drower and Macuch (Reference Drower and Macuch1963: 345) has already been posited by Sokoloff (Reference Sokoloff2002: 160a). While it is generally agreed that these words historically derive from Greek στόμωμα, the question remains whether the same Greek etymon has produced synchronically semantically different words in Aramaic. MK states that אסטמא (Sokoloff Reference Sokoloff2002: 147) and סטומא (Sokoloff Reference Sokoloff2002: 798) should be subsumed under a single lexical entry along with איסטמומי פרזלא ‘spears of iron’, and writes: “A similar explanation was given by the Geonim, respectively”. In fact, the Geonim did not regard the form סטומא in b. Ber. 62b as referring to a weapon, but rather to spikes of iron that are attached to a hoe to make it sharp. The ל- in the Talmudic text is not necessarily a genitive particle, as suggested by MK, but rather a dative one: ‘is like a steel spike to an iron’, i.e. something that is added to strengthen it.

14. MK defends Naveh and Shaked's interpretation (Reference Greenfield and Naveh1985: 200–01) of ביבי ‘canals’ in AMB B13: 6 against the doubts expressed in Sokoloff (Reference Sokoloff2002: 199), but all of the parallel texts read here בירי ‘wells’, including the Christie's bowl, and this is also the correct reading in AMB B13. What appears to be the lower stroke of a second ב in the published photograph of AMB B13: 6 is in fact part of a crack in the bowl (see Figures 7 and 8).Footnote ²⁰

Figure 7. ימא דלא בירי (JNF 30: 4)

Figure 8. דלא בירי (AMB B13: 6)

For the relationship between the sea and wells, which presumably supply it with water, compare Gs 16: 23–24 (reading with MSS Paris A and B): wkwlhyn byryʾ wyʾmʾmyʾ yʾbšyʾ ‘and all the wells and seas will dry up’.

19. ניזי נורי. MK is correct in reading the resh of Naveh and Shaked's ניזי נורי (AMB B13: 15) as a kaph and translating ‘let us go and devour’. However, in accordance with Aramaic grammar, the reading must be ניכו with yod in the second place, rather than MK's נוכו. This reading is supported by the parallels.

23. Naveh and Shaked marked part of their reading of MSF B23: 4 as uncertain: ונידרא וקיריתא ולוטתא ושיקופי(תא) (דאיכא ב)גיתא ומללתא ‘vow, calamity, curse, affliction that is (?) in the world, (magical) words’. They suggested that the hapax legomenon גיתא may be a loanword from Middle Persian gētig ‘the material world, the world’.Footnote ²¹ The existence of this lexeme was subsequently accepted by Sokoloff Reference Sokoloff2002: 284, s.v. 1גֵּיתָא # ‘inhabited world’. MK has proposed reading the doubtful words with greater certainty and with a minor emendation as ושיקופת<י>א דאיכא בגותא ‘and plague that exists in the inside (of PN bar PN)’.Footnote ²² She interprets the hapax legomenon גותא as a native Aramaic term meaning ‘inside’.Footnote ²³ However, in spite of the existence of etymologically related terms in JBA and cognate dialects,Footnote ²⁴ both גיתא ‘world’ and גותא ‘inside’ presently appear to be “ghost words” in JBA, as the correct reading of the text is ונידרא וקיריתא ולוטתא ושיקופ^יתא גיסא פגיתא ומללתא ‘and vow, and imprecation, and curse, and smitings, band (of demons), attack (by a demon or sorcerer), and (magical) word’. This reading, which may be discerned from the published photograph, is confirmed by similar sequences in two unpublished parallel bowls, one written in JBA and the other in Syriac:

1. 1. Wolfe 55: 3: …] ושיקופתא גיסא פגעא ומללתא קיריא קרותא […;

2. 2. MS 2055/11: 3–4: Footnote ²⁵

The demon פגיתא (and its by-form פגעיתא) is well attested in Babylonian Aramaic magic texts,Footnote ²⁶ and it appears in collocation with מללתא in several bowls, e.g מן מבכלתא ומן ליליתא ומן פגיתא ומן מללתא ומן קרותא (JNF 255: 13); ורוח פתכרי ורוחא פורחתא פגיתא ומללתא ונוסיא ונאלה ובני איגרי ובני חיצבי (MS 1927/7: 2); מן נידרא דכלתא ודחמתה ומן לוטתא דימא ודיברתה ומן פגיתא ומן מללתא (MS 2053/189: 22–3, see Figure 9).

Figure 9. ומן פגיתא ומן מללתא (MS 2053/189: 23)

For דאיכא of the previous studies we read גיסא. The latter term is also well attested in lists of demons in Babylonian Aramaic magic texts. In addition to the above-cited parallels, compare the collocation of גיסי with ולוטתא וקריתא ושיקופתא … ומללתא in an enumeration of malevolent forces in CBS 9009 (AIT 12): 9–10: אינון יבטלון וישמתון כל גיסי וקיבלי ואנקי ולוטתא וקריתא ושיקופתא ואשלמתא ומללתא ושידי ודיוי ושובטי וליליתא {ו.}Footnote ²⁷ ופתכרי ומבכלתא וכל מידעם ביש ‘they shall annul and ban all bands (of demons), and counter-charms, and necklace-charms, and curses, and imprecations, and smiting, and spells, and (magic) words, and demons, and dēvs, and plagues, and liliths, and idol-spirits, and mevakkaltas, and everything evil’.Footnote ²⁸ A similar collocation of gysʾ with šyqwptʾ and qʾryʾ appears in a list of malevolent forces in a late Mandaic formulary: mn ʿšʾtʾ gysʾ qʾryʾ wšyqwptʾ wšyqwptʾ (DC 46. 180: 12–3).Footnote ²⁹

25. MK rejects Naveh and Shaked's reading ככיה ככי דדיבתה ‘his molar teeth are the molar teeth of a she-wolf’ (AMB B13: 4) in favour of ככיה ככי {ב}דיבוזה, based on her reading דבוזא ‘of a goat’ in l. 3 of the Christie's bowl. She derives בוזא from Middle Persian buz ‘goat’. Although MK's proposed emendation seems necessary and is supported by the unpublished parallels, the actual reading of the Christie's bowl would appear to be דברזא ‘of a wild boar’ (see Figure 10). The interpretation of ברזא was suggested to us by Shaul Shaked, who immediately identified the term with Middle Persian warāz ‘boar’ based on our new reading. The same word can be identified in the PN בראזדוך (VA 2180: 8 and 10; unpublished), which is equivalent to Warazduχt.Footnote ³⁰ In the PNN warāz / בראז ‘wild boar’ undoubtedly refers to the Zoroastrian god Bahrām.

Figure 10. דברזא (Christie's bowl)

The same reading occurs in several unpublished parallels written by the same hand as AMB B13 and the Christie's bowl, for example, JNF 29: 2 (see Figure 11).

Figure 11. ככיה ככי דברזא (JNF 29: 2)

In AMB B13: 4 one may thus read ככיה ככי {ב}דיברזה ‘his tusks are the tusks of a wild boar’.Footnote ³¹ The term is equivalent to Classical Syriac ܘܪܙܐ, which is defined by Bar Bahlul (following Bar Serošway) as ‘wild boar’.Footnote ³² The spelling with beth also occurs in Syriac in the title of the Sassanian general Shahrbarāz/Shahrwarāz ( = شهروراز).Footnote ³³ The variant spellings reflect the shift w > b already in Middle Persian.Footnote ³⁴

The word ככא was translated ‘molar tooth’ in AMB and that meaning was accepted by Sokoloff Reference Sokoloff2002: 580 (in accordance with the usual meaning of the word in JBA), but in the present context it can only be translated ‘tusk’. The same usage is attested in MS 2053/261: 6–7 (Ford and Ten-Ami Reference Ford and Ten-Ami2012: 223): בעקוסא דעקרבא ירוקתי בככיה דסימור חזורא ‘with the stinger of a yellow scorpion, with the tusk(s) of Simur the pig’.Footnote ³⁵ Compare Syriac ‘tusk of an elephant’.Footnote ³⁶

31. MK reads נטרי אירי^! מגיני in the Borsippa bowl, line 9 for Harviainen's (Reference Harviainen1981: 5) נטריא והרמגוני and Greenfield and Naveh's (Reference Greenfield and Naveh1985: 103) נטרי אגר דמגיני. She understands אירי as ‘a dissimilated variant (/ʿ < ʾ/) [sic] and corrupted spelling of the Syriac form ʿyywrʾ ‘watchers’ (√ʿwr) and its Mandaic counterpart ʿyryʾ in the parallel Syriac bowl Louvre AO 17.284: 8 and the parallel Mandaic lead roll Macuch Ia: 48–9, respectively.Footnote ³⁷ One should first note that the final letter of the word in question is clearly written as daleth or resh, rather than a misshaped yodh, and the reading אירי, whatever its merits, must be considered an emendation.Footnote ³⁸ MK's reading ʿyryʾ is clearly legible in the published photograph of the Mandaic lead roll and is independently corroborated by the equivalent reading ʿydyʾ in the parallel Mandaic magic bowl MS 2054/102: 15–6: hynwn nhwylwn nʾṭ^w^rʾ ʿydyʾ wmgnyʾ hynwn nynṭrwnwn ‘they shall be for them a guard … and shields. They shall guard them’. The equivalent reading is also found in a new parallel Syriac bowl T 027996:10 (Ford and Abudraham, Reference Ford, Abudraham and Regevin press): ‘and they shall be for Zuṭay, who is [called Babay], son of Imma-d-immah, and for his house, an alert guard and a sheltering shield’. The Louvre bowl, however, in fact reads ‘helpers’. Louvre AO 17.284 was first published by de la Fuÿe with a hand copy in 1924 and was subsequently independently re-edited by Müller-Kessler in 1998 with a photograph. De la Fuÿe's hand copy, apparently executed before the bowl was damaged, clearly reads , which he rendered as ‘les aides’. Marco Moriggi, who is preparing a new edition of the Louvre bowl, kindly collated the text for us from a new high-resolution photograph and confirmed de la Fuÿe's reading. In an email (15.11.2012) he writes as follows: “The dalath is there, just a little portion of the left stroke is missing at the bottom of the sign”.Footnote ³⁹ The reading is corroborated by an unpublished parallel magic bowl (PC 1:6) which reads at this point ‘[may they (the angels) be] helpers, and protectors, and sheltering shields’ (see Figure 12).Footnote ⁴⁰

Figure 12. (PC 1)Footnote ⁴¹

Syriac ‘helper’ appears to be cognate with Mandaic ʾdyʾwrʾ ‘helper’, a loanword from Parthian adyāvar.Footnote ⁴² The Mandaic term, in fact, appears in the corresponding context in an unpublished parallel Mandaic magic bowl (MS 2054/100: 41–2): hlyn mlʾky [ ʾ ] nʿhwylẖ ʾdyʾwrʾ … wm.. wmgʾnyʾ … ‘these angels shall be for him a helper … and shields …’. The writing of the Syriac term with a non-etymological ʿayin reflects the weakening of the pharyngeal ʿ > ʾ.Footnote ⁴³ The same word occurs in the JBA incantation bowl MS 2053/159: 14, where it is written with ʾaleph: אידיורא הוי לן בקרבה רבה וקול[ן ביומא] רבא דקירסא ‘be a helper to us in the great battle, and stand by [us on the] great [day] of misfortune!’Footnote ⁴⁴ The collocation of ‘helpers’ and ‘shields’ in Louvre AO 17.284 corresponds to the collocation of אידיוריה ‘his auxiliaries (lit. helpers)’ and מגינא ‘shield’ in Moussaieff 4: 4–6 (Shaked Reference Shaked2006: 373), where one may read as follows:Footnote ⁴⁵

The cognate abstract noun אידיורותא ‘help, assistance’ is likewise attested with respect to angels in Wolfe 41: 10–11: אילין מלאכי[ן] קרובו ואיתו ואידיורותא הוו ליה לרב משרשיה בר נינסי ‘These angels – draw near and come and be of assistance to Rav Mešaršia son of Ninsay’.Footnote ⁵⁰

In light of the correspondence with / ʾdyʾwrʾ ‘helper(s)’ in some of the Syriac and Mandaic parallels and the use of JBA אידיורא ‘helper’ and אידיורותא ‘help’ with reference to angels in other incantation bowls in similar contexts, we would read אידר in the Borsippa bowl and would suggest that the word may be a corruption of אידיורי ‘helpers’. A derivation from Syriac ‘aid, help; pl. auxiliary troops’ is also conceivable. Hopefully additional parallels will come to light and provide a definitive solution for this word. The variant readings ʿyryʾ / ʿydyʾ in Mandaic and in Syriac appear in three discrete textual witnesses and represent a distinct textual tradition. The writer of Macuch Ia could conceivably have understood ʿyryʾ as ‘watchers’, as proposed by MK (we would translate nʾṭryʾ ʿyryʾ as ‘alert guards’). The precise relationship between the two textual traditions, however, remains to be determined.

34. MK's material reading וזיגודיתא is correct, but the possibility of emending the text to <מ>זיגודיתא should not be considered,Footnote ⁵¹ since the word is equivalent to Mandaic sygwdtʾ, pl. sygwdyʾtʾ, Footnote ⁵² and is otherwise attested in the form זיגודתא in VA 2430: 7–8 (unpublished): ושלניתא וחטטיתא וזיגודתא ודיוי בישי ורוחי בישתא. The Mandaic plural form sygwdyʾtʾ suggests that זיגודיתא is likewise the plural form of זיגודתא.

36. זלעיקא. See the comments on no. 57.

37. חי חי מץ. The basic reading in Moussaieff 101: 11 is טיטינוס חיחימץ, as read by LeveneFootnote ⁵³ (see Figure 13). If we understand Müller-Kessler correctly, she recognizes that חיחימץ is indeed the material reading in Moussaieff 101, but believes that the formula “can be understood through its text parallels which constantly [sic] show בשים טיטוניס חיי חמץ פגרי רעש ופגרי רגש ‘In the name of Ṭiṭinus my life turned sour, my body reacted, and my body trembled’”.Footnote ⁵⁴ She cites TMHC 7, Bowls 5–7; Istanbul 1167 (= Gordon Reference Gordon1934, Bowl B); and BM 117824 (= CAMIB 027A) as published examples of this reading. It should first be noted that the basic reading in both Istanbul 1167: 8 and BM 117824: 18 is, in fact, identical to that of Moussaieff 101 (see Figures 14 and 15).

Figure 13. חיחימץ (Moussaieff 101: 11)

Figure 14. טוטינוס חיחימץ (Istanbul 1167: 8)Footnote ⁵⁵

Figure 15. חיחי מץ (BM 117824: 18)Footnote ⁵⁶

The same basic reading is shared by additional textual witnesses, likewise in collocation with טיטינוס or equivalent names, as in JNF 123: 8 (see Figure 16).

Figure 16. טוטינוס חי חי חמץ (JNF 123: 8)

JNF 210: 12 reads טיטיוניס but nonetheless like the Moussaieff bowl reads חי חי מץ (see Figure 17).

Figure 17. טיטיוניס חי חי מץ (JNF 210: 12)

The precise reading of חיחימץ has been debated. Levene read חי חי מץ, while one of the present writers proposed reading הוהי מץ and regarding מץ as Atbash (a well-known alphabetic code) for יה.Footnote ⁵⁷ The interpretation of מץ as Atbash was first proposed by Montgomery in relation to a different formula, CBS 16917 (AIT 14): 2: בשום אגרבוס קדישא בשום מץ מץ בשום סף סף יהוק יהוק דדחיק ית מרכבתיה על ימא דסוף ‘in the name of the Holy ʾgrbws, in the name of mṣ mṣ, in the name of sp sp yhwq yhwq who pushed his chariot upon the Sea of Reeds’.Footnote ⁵⁸

Similarly, in MS 2053/278: 9–10 we read: ובשום מץ מץ מץ מץ מץ מץ (see Figure 18). Since in these cases the letter sequence מץ is repeated, it cannot be a part of another word, and must be regarded as a holy epithet.

Figure 18. ובשום מץ מץ מץ מץ מץ מץ (MS 2053/278: 9–10)

מץ is also found as a divine name in several magic formulae preserved in the Cairo Geniza, e.g. בשמיה דמץ מץ רבה גיברא ודחילא דאיתגלי למשה בסנה ‘in the name of mṣ mṣ, the great mighty and awesome,Footnote ⁵⁹ who appeared to Moses in the bush’ (T-S K 1.162 1a, 6–7).Footnote ⁶⁰ As discussed by Herrmann, the name also appears in the Hekhalot literature,Footnote ⁶¹ including several texts that bear a close resemblance to the formula at hand, e.g. והויאל השר ואחיאל הקדוש וטיטינוס חי חי מץ ואטטינו (T-S K 21.95.A 2b: 26).Footnote ⁶² One may also note that the name מץ occurs in collocation with a phrase resembling פגרי רעש פגרי רגש in T-S K 21.95.P, 1b: 11–14:Footnote ⁶³

Accordingly, the interpretation חי חימץ ‘my life soured’ is unlikely. Furthermore, the above-quoted Genizah parallel from T-S K 21.95 clearly reads חי חי מץ (see the published photograph and Schäfer's transliteration), and this is to be regarded as the correct reading in the present formula, in accordance with Levene. TMHC 7, Bowls 5–7, all written by the same hand, would appear to preserve a variant reading that would best be transliterated חי יה מץ (see Figure 19).

Figure 19. חי יה מץ (TMHC 7, B6: 4)

There, too, מץ is undoubtedly a divine name (Atbash for the preceding יה – see above) and, accordingly, the interpretation ‘my life soured’ must be rejected.

42. For BM 117880: 10 (= CAMIB 081M), MK proposes that the verbal form lʾpwlyʾ bnyʾ ʾnʾšʾ ‘to prostrate humankind’Footnote ⁶⁴ is “a scribal error for lʾp〈d〉wlyʾ” and states: “This emendation is possible on account of similar usage in wbʾyʾ lpdwlyʾ ʿtʾtʾ mn gbrʾ ‘and she tries to separate wife from man’ instead of emended lprw{l}〈ṭ〉yʾ (YBC 2364: 23–4)”.Footnote ⁶⁵ The question here is not whether the emendation is possible, but whether it is necessary. In fact, the verb as it stands is both lexically and grammatically suited to its context. Moreover, as MK herself notes, the textual witnesses to this formula differ greatly at this point. While the BM version reads lʾpwlyʾ, a parallel she herself published reads lʾnpwqyʾ ‘to make leave’. An unpublished parallel in the Moussaieff Collection reads lʾnʾqwpyʾ ‘to strike’ (see Figure 20).

Figure 20. wlʾnʾqwpyʾ bnyʾ ʾnʾšʾ (Moussaieff 25: 11)

Given this wide variety of readings, all of which are grammatically and contextually acceptable, it is not possible to establish an ‘original’ reading for this formula, and there is certainly no justification for emending it to an unattested reading on the basis of speculation alone.

43. In Moussaieff 145: 4, first published in Levene Reference Levene2003a: 100, MK proposes reading לימעביא, which she interprets as a pəʿal infinitive and translates ‘to accumulate’. This proposal may be questioned for several reasons: (1) the infinitive pattern מקטיא is not normal in Babylonian Aramaic; (2) if the pattern were employed, we would expect to find a shewa after the l- preposition, i.e. ləmiqṭəyā (or perhaps ləmiqiṭyā), which would not normally be represented with a yod; (3) the pəʿal stem of עב”י has the stative meaning ‘to be thick’, and is never transitive; in the paʿil it has the meaning of ‘to thicken’ and not ‘to accumulate’Footnote ⁶⁶ ; (4) the word appears in the context כד דיוי נפק ליקרבא וליליתא נפק לימ()א ‘when the devs went out to battle, and the liliths went out to …', and hence we would expect a noun.

It appears that the correct reading here is מנוביא ‘wailing’. The reading is somewhat difficult in the Moussaieff bowl, where part of the ink of the beth appears to have flaked off, but is much clearer in the parallel from the Schøyen Collection, where Shaked correctly read למנוביא.Footnote ⁶⁷ The letter that MK reads as an ʿayin is clearly written as two letters in the Moussaieff bowl (see Figures 21 and 22).Footnote ⁶⁸

Figure 21. לימנוביא (Moussaieff 145: 4)

Figure 22. למנוביא (MS 2053/159: 4)

The reading נפק לימנוביא can also be discerned in the unpublished parallel bowl MS 2053/17: 4, written by the same hand as Moussaieff 145 (see Figure 23).

Figure 23. לימנוביא (MS 2053/17: 4)

מנוביא is a previously unattested Jewish Babylonian Aramaic form of Mandaic mnwmbyʾ ‘mourning’, derived from the Akkadian verb nubbû ‘to mourn’.Footnote ⁶⁹ While the male devs go out to fight, it is the female liliths’ role to mourn.Footnote ⁷⁰

46. In his original publication of Moussaieff 1: 5, Shaked marked several letters as uncertain: ומחתא מוח ו(מו)ל(ה) מן אודנה.Footnote ⁷¹ MK reproduces the same reading without brackets and presents a different interpretation. However, examination of both this text and the parallel in MS 2046: 4 indicates that Shaked's hesitation was well placed (see Figure 24). Unfortunately, both copies of the text are damaged at this point, but for the word that was read ומולה, MS 2046 most likely reads ונטלא from the root נט”ל ‘to pour’.Footnote ⁷² The third letter is poorly preserved, but the nun is clear.Footnote ⁷³

Figure 24. ונ(ט)לא (MS 2046: 4)

We may tentatively transcribe and translate דמחתא מיה ונ(ט)לא מן אודנה ‘who brings down water and pours (it) from her ear’.

51. Shaked read Moussaieff 1: 6 as follows: מן רוח (ח)יצ(ב)אי מן רוח מזריבי מן רוח בי קברי ‘from the spirit of jugs; from the spirit of drain-pipes; from the spirit of the cemetery’. He interpreted מזריבי as a metathesized form of JBA מרזיבי ‘drain-pipes’ and adduced convincing evidence from the Babylonian Talmud for the belief in demons infesting drain-pipes.Footnote ⁷⁴ MK has accepted Shaked's basic identification of the term, but has emended מזריבי to מרזיבי and translated ‘spirit of gutter demons’, understanding מרזיבי itself as a type of demon. She finds support for this interpretation both in Mandaic, where nʾrzwbyʾ = mʾrzwbyʾ signifies both ‘gutters’ and ‘gutter-demons’, and in several JBA bowls where she reads מרזבין or מרזביא, both referring to a type of demon. The same spelling רוח [מ]זריבי, however, occurs in the parallel MS 1927/63: 6, and another parallel, MS 2046: 6, reads רוח ניזריבי (or רוח ניזרובי), again with the zayin preceding the resh.Footnote ⁷⁵ The fact that all three texts show metathesized forms suggests that מזריבי is what the scribe intended to write and that, accordingly, the text should not be emended. Furthermore, the parallel terms רוח חיצבאי ‘spirit of jugs’ and רוח בי קברי ‘spirit of the cemetery’ suggest that מזריבי denotes the place that the demon haunts and not the demon itself.Footnote ⁷⁶

In fact, the existence of a category of demons called מרזבין ‘gutter-demons’ or מרזביא ‘idem’ in the extant JBA bowls in general seems unlikely to the present authors. Not all of the bowls that MK cites have been published with photographs that allow independent verification of the reading, but when we have been able to verify the reading, what MK reads as zayin appears in fact to be waw, usually in accordance with the reading by the original editor of the text. The reading in MSF B15: 6 (Naveh and Shaked: וימרובין) is difficult, but compare the letter in question with the waw in ומן directly below in line 7, and contrast the zayin in מזיקין (line 6) and in ראזי (line 8). The tip of the vertical stroke consistently extends above the horizontal stroke of the zayin, but one sees this neither in the waw nor in the letter in question. Geller (Reference Geller and Rendsburg1980) read מריבין in Bowl A:18 (MK: מרזבין). The word is clearly visible in the published photograph. The letters waw and zayin are consistently distinguished in this text. The head of the waw usually extends slightly to the left or is occasionally without a prominent point, whereas the head of the zayin extends to the right. Compare, for example, the waw and zayin in אזרמידוך (lines 8 and 15), וחיזוני (line 5), and וגזרית (line 6) and cf. also the zayin in דזעקי (line 12). The letter in question resembles a typical waw with the head extending to the left and is distinct from all clearly preserved examples of zayin. In Moriah I: 25, Gordon indeed read ומרזבין, but marked the zayin as only partially preserved.Footnote ⁷⁷ No photograph of the bowl has been published, but Gordon's own handcopy suggests the reading מרובין. Compare the traces of the letter in question with the waw at the beginning of the word and with the zayin in אזרמידוך (line 26), see Figures 25 and 26.Footnote ⁷⁸

Figure 25. ומר.בין (Moriah I: 25)

Figure 26. אזרמידוך (Moriah I: 26)

MK reads מרזביא in Nippur 12 N 387: 4 (MK: line 3),Footnote ⁷⁹ but the letter in question lacks the short upper horizontal stroke of a zayin.Footnote ⁸⁰ Contrast the zayin in זויתא (line 2), see Figures 27 and 28.Footnote ⁸¹

Figure 27. מרוביא (Nippur 12 N 387:4; Gibson 1978, fig. 80)

Figure 28. זויתא (Nippur 12 N 387:2; Gibson 1978, fig. 80)

The same demon is named in VA 3854: 5, where Levene correctly read מרוביא (MK: מרזביא).Footnote ⁸² See Figure 29.

Figure 29. ומן מרוביא (VA 3854: 5)

Although it is sometimes difficult to distinguish between waw and zayin in the script of the JBA bowls, in VA 3854 the two letters are quite distinct. Contrast, for example, the zayin as shown in Figures 30–32.

Figure 30. וזיקין (VA 3854: 9)

Figure 31. בעיזקת_יה (VA 3854: 13)

Figure 32. מזוזות (VA 3854: 45)

The demon מרוביא is otherwise attested in DCG 3: 6: ובר איגרי ומרוביה וכל סטני ‘and the roof spirit, and MRWBYH, and all satans’ (see Figure 33).

Figure 33. ומרוביה (DCG 3: 6)

Contrast the zayin in the words shown in Figures 34 and 35.

Figure 34. ייזלון (DCG 3: 2)

Figure 35. יזלון (DCG 3: 4)

Similarly, MS 2053/2: 12 reads: ומשמתתא ורונאי ומרוביא ודנחיש ובר איגרי ‘and the banned demoness, and Ronay, and MRWBYʾ, and Danaḥiš, and the roof spirit’. What appears to be a defective spelling of the same term occurs in JNF 271: 8–9: ותיקלי ית שידא ודיוא ודנחיש וששנצר ומרבה ורוחא בישתא ‘and may you roast the demon, and the dēv, and Danaḥiš, and ŠŠNṢR, and MRBH, and the evil spirit’. The collocation with the demon ששנצר is not accidental, for מרוביא is amply attested specifically as the epithet of this demon (also written שנאצר and שלאצר). For example, אנתה שלאצר מרוביה שידה בישה וכל שום דאית לך ‘you, ŠLʾṢR MRWBYH, the evil demon, and any (other) name that you have’ (JNF 60: 5); ויתסי מן שנאצר מרוביה שידא בישה ‘and may he be healed from ŠNʾṢR MRWBYH, the evil demon’ (JNF 141: 2). The reading מרוביה in these contexts cannot be doubted, since the same demon occurs in a Syriac amulet (Naveh Reference Naveh1997, line 12), where there is no ambiguity between waw and zayin in the script (see Figure 36).

Figure 36. (Naveh 1997: 35)

Naveh translated “Sheshnaṣar the educator”, but stated that the expression “seems rather to belong to the list of magic words of lines 9–11”, rather than to the list of malevolent elements in line 12. The evidence from the incantation bowls, however, confirms the existence of as an independent demon.Footnote ⁸³

52. For Segal's מרזקופתא ‘hanging’ in BM 91771: 7 (= CAMIB 039A), MK proposes reading מרזהיפתא,Footnote ⁸⁴ which she explains as “a nominal form of the saf‘el סרהב ‘to hasten, to be angry’ based on the sound shifts /z/ < /s/ and /p/ < /b/”. MK is correct that Segal's reading is unsatisfactory, but rather than posit a series of sound changes and metathesis, one may simply read מסחיפתא ‘overthrower’ from the attested root סח"פ ‘to throw down, overturn’ (see Figure 37).Footnote ⁸⁵ A similarly written samekh is found in line 17 of the same text (see Figure 38).

Figure 37. מסחיפתא (BM 91771: 7)

Figure 38. ואיסרי (BM 91771: 17)

Compare the demonic epithet mʾsyhpʾn ‘overthrower’ in DC 37: 64–6 // BL MS Add. 23602B fol. 26 (unpublished):Footnote ⁸⁶ mšʾrhybʾnʾ lʾk mʾsyhpʾn dʾywʾ tbʾr hʾylʾk mn pʾgrʾ ḏ-yʾhyʾ byhrʾm br hʾwʾ symʾt ‘I terrify you, Masihpan the dēv. Break your power from the body of Y. son of H.!’Footnote ⁸⁷

54. MK presents a revised reading ניפ<ר>הזוניה in the Borsippa bowl: 10, which she interprets as a miscopying for the quadraliteral root prhz.Footnote ⁸⁸ Although this emendation is plausible, she further states (p. 20): “Obviously Mandaic ʾprwz ʾlʾhyʾ (DC 40: 491; unpubl.) is a short form of prhz as well, and not a loan from Hebrew”. Here MK has been misled by Drower and Macuch Reference Drower and Macuch1963: 379, s.v. PRZ, wherein ʾprwz is mistakenly interpreted as a verb. The full context of this text is bšwm ʾ ḏ-zʾn ʾprwz ʾlʾhyʾ, in which ʾprwz is certainly an epithet of the divine Zan, probably derived from Persian afroz ‘dazzling, illuminating’. In any case, it has nothing to do with the verb prhz.

55. MK writes: “עברא, in ותובי כי עברא על ליביה כ[…]תא על מחו ‘and sit like a slave on his heart, like a … on his brain’ (BM 91767: 4–5 = CAMIB 040A) is clearly to be read נירבא, and not with ‘ayin and resh עברא ‘bolt’ as suggested by Ford in Morgenstern Reference Morgenstern2007a: 13”. MK's material reading is not possible, since in the text itself the beth precedes the resh (see Figure 39).Footnote ⁸⁹

Figure 39. עברא (BM 91767: 4)

With respect to the reading of the first letter as ʿayin, rather than MK's nun + yodh, compare the ʿayin in ועמורה (line 15, see Figure 40).

Figure 40. ועמורה (BM 91767: 15)

Cf. also the form of the ʿayin especially in רוחא על (line 1), אשבעית (line 2), מעוהי (line 4), וארבעין (line 9) and כל עשב (line 15).

MK further states: “It is an obvious misspelling of צירבא ‘lead or purified silver’ (AO 1177: 4)”, though according to her own testimony presented elsewhere,Footnote ⁹⁰ AO 1177: 4 reads נירבא not צירבא.

In that previous study, MK cast doubt upon the existence of the lexeme nyrb ʾ [/nirbā/?], an unidentified hard metal, and suggested that it resulted from a scribal error for ṣyrbʾ [/ṣirbā/?] that arose due to the graphic similarity of ligatured wn and ṣ in Mandaic.Footnote ⁹¹ Recent findings provide evidence that seems to point in the opposite direction. The form nyrbʾ is attested in several Eastern Aramaic magic texts from late antiquity in different scripts (see Figures 41–43).

Figure 41. דנירבא (MS 1927/35: 11)

Figure 42. (Wolfe 24: 13)

Figure 43. wnyrbʾ (BM 91715: 14)Footnote ⁹²

By contrast, ṣyrbʾ is found only at a single point in parallel copies of the šafta ḏ-qaština, all of which are very late and unreliable manuscripts.Footnote ⁹³ It is therefore likely that ṣyrbʾ is to be regarded as a scribal error that corrupted the textual tradition of this word in this specific context.Footnote ⁹⁴ On the basis of these late attestations of a single example, there is certainly no reason to posit that all the evidence for nyrbʾ found in epigraphic texts from late antiquity written in several Aramaic dialects is to be ascribed to a scribal error in a posited but unattested Mandaic Vorlage.

56. MK has suggested that the root s-g-m ‘shut up’ is a ghost in Aramaic that arises from the graphic similarity of g and ṭ in the Mandaic script, even though it is attested in several independent formulae in both the Jewish and Mandaic scripts. The reading סגמה פומה in Moussaieff 1: 11 also occurs in the unpublished parallel MS 2046: 11. The explanation that all these attestations stem from a scribal error or a misreading of Mandaic is not convincing, and so an etymology must be sought elsewhere. We would cautiously propose that it is a denominative verb derived from the noun ‘bolt’.Footnote ⁹⁵

57. In CBS 16041: 15 = AIT 27 (unpublished section), MK has read באדא אכלא רבא דזיוא {סורא} ספסי{פ}〈ר〉א רבא דזלעיקא, which she translates ‘with a great mace of splendour, a great sword of ray’.Footnote ⁹⁶ MK writes that ספסיפא “is an obvious scribal error for Iranian ספסירא ‘sword’”. However, the emendation appears to be unnecessary and most likely based upon a misreading of the text. An unpublished parallel to AIT 27 is found in the Moussaieff Collection, wherein the phrase אכלא רבא דזיוא וסיפא ספסיפא רבא דילעולם ‘a great mace of light and a great eternal burning sword’ may be clearly read in two places (see Figures 44 and 45). For the previously unattested ספסיפא ‘burning’, compare Syriac ‘blazing, burning’;Footnote ⁹⁷ it is presumably employed here in alliteration with the common Aramaic סיפא ‘sword’.Footnote ⁹⁸

Figure 44. סיפא ספסיפא רבא דילע[ו]לם (Moussaieff unnumbered: 13–4)

Figure 45. בההוא סיפא ספסיפא רבא דילעולם (Moussaieff unnumbered: 15)

The hapax legomenon דזלעיקא in MK's reading of AIT 27 (see no. 36) corresponds to the common דילעולם ‘eternal’ in the Moussaieff bowl, the latter reading occurring in both contexts. Although variant readings are always a possibility (cf. the note to no. 42, above), in light of the apparent poor state of preservation of AIT 27 and the new evidence from the parallel, a collation would be in order before accepting זלעיקא ‘ray’ into the JBA lexicon. It is not clear to the present authors how MK analyses באדא. The word most likely corresponds to בההוא in the second occurrence of the phrase in the Moussaieff bowl. Here, too, a collation would be a desideratum in light of the new parallel. In any case, the major emendation proposed by MK in this context in AIT 27 does not appear justified.

58. See above, no. 3 (n. 16).

62. The original editor of the Borsippa bowl, Tapani Harviainen (Reference Harviainen1981), read in line 9 פרחוני. On the basis of a Mandaic parallel, pw{r}ršʾnʾ, MK has proposed emending the word in the Borsippa bowl to read פרשוני, which she translates ‘divisions’. But the correct reading is פרהזני ‘protectors’, derived from Middle Persian parhēz (see no. 54).Footnote ⁹⁹ The variant reading in the Mandaic version should not be employed to emend the comprehensible Jewish Aramaic text.Footnote ¹⁰⁰

66. MK rejects Levene's interpretation of the expression אכיף ימא קינא (Moussaieff 145: 9) and its variant אכיף ימא {ק^אי^נא} קאינה (MS 2053/159: 9) as ‘I am standing upon the shore of the sea’Footnote ¹⁰¹ in favour of the translation ‘he bend down the reed sea’. According to MK, this clause would belong to the preceding narrative in the third person, rather than the following section (beginning at the end of line 9) where the first person is used. She states that “קאינא was obviously borrowed in this spelling from Mandaic qʾynʾ [‘reed’]” and that “ימא קינא corresponds to yʾmʾ ḏ-swp in Mandaic”. According to MK, the spelling קינא (along with other features) would speak for a Mandaic forerunner to the text. It is nevertheless the opinion of the present authors that Levene's interpretation here is correct, for the reasons outlined below.

Without entering into the theoretical question of why Mandaic yʾmʾ ḏ-swp ‘reed sea’ would not have been rendered here with its common JBA etymological correspondent ימא דסוף,Footnote ¹⁰² one may first note that the interpretation of the present context as a reference to a ‘reed sea’ is syntactically unlikely, as such a meaning would normally require the reading ימא דקינא* or ים קינא*.Footnote ¹⁰³ The structure of the text also speaks against the proposed analysis. The following discussion is based on the version in MS 2053/159.Footnote ¹⁰⁴ Lines 1–2a read as follows:

In light of parallel material, the forms קאינה and קינא were explained in Morgenstern Reference Morgenstern2007b: 265 as 1 c.s. participles showing assimilation of the 3^rd radical to the n of the appended personal pronoun.Footnote ¹⁰⁶ Further evidence for the derivation of these forms from קו”ם is now forthcoming in a new bowl formula, wherein the standard and phonetic orthographies appear side by side: אנא דוכתיש בת בהרוי בבאבי קימנא לבאביל דמינא בסופי קינא {ד}לבורסיף דמינא ‘I, Dukhtīč daughter of Bahārōy, stand at my doorway (and) I resemble Babylon, I stand in my vestibule (and) I resemble Borsippa’ (Davidovitz 2: 1–2a).Footnote ¹⁰⁷

This use of a 1 sing. participle of קו"ם at the beginning of an incantation (or section thereof) is quite common in Aramaic historiolae. In addition to the preceding texts one may adduce the following representative examples: בביל קמינא לבבלהא דמינא בברסיף קימנא לברספיא דמינא ‘I stand (in) Babylon, I resemble a Babylonian; I stand in Borsippa, I resemble a Borsippean’ (JNF 90: 2–4); ʿl klyl nhwr ʾyʾr qʾymnʾ ‘I stand upon the wreath of light of ether’ (BM 117880: 5 and parallels); ʿl bʾbʾ ḏ-byt hyyʾ qʾyymnʾ ʾnʾ hw qʾštynʾ qʾšyšʾ ḏ-mn byt hyyʾ qʾl ṭwryʾ šʾmʾnʾ wqʾl pʾqʾtʾ ḏ-ʾpqʾ ‘I stand at the door of the House of Life, I am the Elder Archer from the House of Life; I hear the sound of the mountains, and the sound of the valleys that were split’ (DC 43 J 3–5 // DC 39 6–9 // Oxf. Bod. MS Syr.g.2(r) 11–13, unpublished); bmyṣʾt ʾlmyʾ wdʾryʾ Footnote ¹⁰⁸ qʾyymnʾ gymrʾ ʾnʾ gmyrʾ ‘I stand in the midst of the eternities and generations, I am the perfect gem (?)’ (DC 12: 211 // BM Or 6593: 447–9, unpublished); ʿl ʾmynṭwl ḏ-ʿl bʾbʾ ḏ-byt hyyʾ qʾyymnʾ wqʾrynʾlwn lʿwtryʾ ʾdyʾwrʾy Footnote ¹⁰⁹ ‘for I stand at the gate of the House of Life, and I call the Uthras, my helpers’ (DC 26: 307–8 // DC 28: 406–8);Footnote ¹¹⁰ lryš tws tʾnynʾ qʾyymnʾ ʾnʾ mʾlkʾ ḏ-ʾlʾhyʾ wdʾyʾnʾ rbʾ ḏ-ʿstyrʾtʾ ‘I stand upon the head of Tus, the dragon, I am the king of the gods and the great judge of the goddesses’ (DC 26: 542–4 // DC 40: 556–8); ʿl ʾrqʾ ḏ-ʾnhʾšʾ qʾyymnʾ wlbʾbʾ ḏ-byt ʾlʾhyʾ ‘I stand upon the earth of copper, and at the gate of the house of the gods’ (DC 26: 550–1 // DC 40: 565–7; unpublished); lṭwrʾ ḏ–rʾzyʾ qʾy^y^mnʾ wlṭwryʾ ḏ-rʾzyʾ mʾsgynʾ ‘I stand upon the mountain of mysteries, and I walk upon the mountains of mysteries’ (DC 40: 79–80, unpublished); lʾrqʾ ḏ-nhʾšʾ qʾyymnʾ wlbʾbʾ rbʾ ḏ-bythyyʾ ‘I stand upon the earth of copper, and at the great gate of the House of Life’ (DC 40: 681–2; unpublished).

MK accepts this general interpretation of קאינה in MS 2053/159: 1 // M145: 1,Footnote ¹¹¹ but, as noted above, rejects it in line 9 of both texts, appealing to the structure of the text. The structure of lines 9–10, however, strikingly parallels that of line 1:

One may note the similarity between the first bicolon in each context. Both lines in each case begin with the preposition א- followed by the same word, the first time without final ʾaleph and the second time with final ʾaleph and followed by the word רבה. The parallelism A // B rabbā is amply attested in the incantation bowls and related magical literature. In addition to mʾlkʾ ḏ-ʾlʾhyʾ ‘the king of the gods’ // dʾyʾnʾ rbʾ ḏ-ʿstyrʾtʾ ‘the great judge of the goddesses’ (DC 26/40) and ʾrqʾ ḏ-nhʾšʾ ‘earth of copper’ // bʾbʾ rbʾ ḏ-bythyyʾ ‘the great gate of the House of Life’ (DC 40), quoted above, see, for example: בצורת עיזקתא דישלומו מלכא בר דויד ובחתמא רבה דיחתימין ביה שמיא וארעה ‘by the image of the signet-ring of King Solomon son of David, and by the great seal by which heaven and earth are sealed’ (JNF 245: 10–11); ובציצי דצבאות ובזיוא רבא דקדוש ‘and by the brightness of Sebaoth and by the great radiance of the Holy One’ (AIT 7: 5); אליסור בגדנא מלכהון דשידי שליטא רבא דיליליתא ‘Elisur Bagdana, the king of the demons, the great ruler of the liliths’ (IM 141802: 1);Footnote ¹¹² דהוא [רמ]א אידיא בימא ותיוהא רבה ברביתא ‘who [pu]t high water in the sea and a great agitation in the ocean’ (Tarshish Bowl: 14–15; Ford and Ten-Ami Reference Ford and Ten-Ami2012).

Next comes a parallel pair, which in the second line in each context is preceded by the relative pronoun ד. Line 1 shows the repetitive parallelism גלל // גללא, whereas line 9 has the synonymous parallelism ימא // רביתא. The latter parallel pair is otherwise attested in both JBA (see the Tarshish Bowl, quoted above) and Mandaic: kmʾ šʾpyryʾ yʾmʾ wšʾpyryʾ ḏ-mʾsgyn bgʾwẖ Footnote ¹¹³ ḏ-rbytʾ ‘how beautiful is the sea, and how beautiful are those that go about within the ocean!’ (DC 21: 34–5 // DC 29: 40–1 // MS Berlin 22–3).Footnote ¹¹⁴ For the construct chain כיף ימא ‘the shore of the sea’, compare לכיף ימא סליקנא ולבני ימא משילנא ‘I go up to the shore of the sea, and ask the inhabitants of the sea’ (BM 91767: 13).Footnote ¹¹⁵ The corresponding construct chain טור גלל ‘a mountain of rock’ in context 1 is otherwise attested in Wolfe 10: 9: ואמטיאו יתיכי לטורא טור גללא מן יומה דין ולעלם ‘and they brought you to a mountain, a mountain of rock, from this day and forever’.

Finally, the first line of each bicolon ends with קאינה. Given the formal similarity between the two contexts, קאינה must have the same meaning in each case, namely, it must be the 1 sing. participle of קו”ם. As shown above, this use of qāyimnā is quite frequent at the beginning of an incantation or section thereof. Accordingly, in line 9 it must signal the beginning of a new section. In both contexts the lines following the initial bicolon are likewise formulated in the first person singular and contain verbs of communication (verbs of hearing in context 1 and verbs of speech in context 2). One may thus propose the following translation (in general accordance with Levene):

69. MK is correct in removing טלניתא from the JBA lexicon, the reading originating under the influence of an infelicitous proposal by Scholem to emend שלניתה in the original editions of two bowls to טולניתה.Footnote ¹¹⁶ MK's comments have been appropriately accepted by Sokoloff in his corrections to his dictionary. The same reading שלניתא (with its orthographic variant שלוניתא and phonetic variants שלנניתא, שנניתא, and שניתא) appears in the published and unpublished parallels known to the present authors.Footnote ¹¹⁷ It should nevertheless be noted that the demon טולניתא (essentially the JBA form corresponding to JPA טלניתא) is amply attested in the JBA incantation bowls. In addition to the sole (remaining) reference cited by Sokoloff Reference Sokoloff2002, s.v., see the example in Figure 46.

Figure 46. כל טולניתא (MS 2053/261: 4)Footnote ¹¹⁸

In Moussaieff 164: 11, however, the form שולניתהין is not merely a plene spelling of שלניתא ‘robbing one’, as proposed by MK, but most likely the plural of a different lexeme. In several Mandaic magic texts we find an Akkadian loanword שוליתא (šəwālīṯā) ‘(female) apprentice’ or ‘maidservant’.Footnote ¹¹⁹ However, in two copies of this formula, one in JBA and the other in Mandaic, we find instead the form שולניתא/šwlnytʾ.

Figure 47. זניתא ושולניתה (JNF 247: 10).

מן לוטת אבא ואימא ומן לוטת אחא ואחתא ומן לוטת יחמתא וכלתא ומן לוטת זניתא ושולניתה

Figure 48. drptʾyʾ wšwlnytʾ (MS 2054/50: 26)

Compare also:

Figure 49. ויתמני שולתהון (MS 2053/29: 7).

With respect to the general context, Levene's reading of ר(כ)בין ‘are riding’ in Moussaieff 164: 11 is correct.Footnote ¹²¹ The samekh is always written in this text with a rounded head, e.g. in the name of the client סמי בת פרסיתא (see Figure 50).

Figure 50. דסמי בת פרסיתא (Moussaieff 164: 12)

What appears to be an extra stroke in the kaph (giving the impression of samekh) is most likely the result of a small crack in the surface of the bowl into which some ink spilled from the base of the letter (see Figure 51).

Figure 51. רכבין (Moussaieff 164: 11)

MK's translation of אם לא איתי עליכון as ‘and if he does not bring upon you …’ does not fit the context. The correct translation is that provided by Levene: ‘And if not, I shall bring against you …’.Footnote ¹²² The entire passage may be translated: ‘And if not, I shall bring against you a reed of seven pieces that seven sorcerous women are riding, (they) and their maidservants’.Footnote ¹²³

70. This paragraph contains three misreadings of the sources. BM 91771: 2 does not read ועמומתית but rather שמומתית or שמומתות, most likely with the meaning ‘bans’(see Figure 52). The complete context is ושלאמתא שופהרי שמומתות בית כנישתא ‘and spells, šwphry, bans of [i.e. pronounced in] the synagogue’.Footnote ¹²⁴

Figure 52. שמומתות (BM 91771: 2)

VA 2416: 10 (see Figure 53) does not read ושפורי ושמהתא but rather ושיפורי ושמתתא, as read by Wohlstein.Footnote ¹²⁵

Figure 53. ושיפורי ושמתתא (VA 2416: 10)

DC 47 (222) does not read br špwhr ʾ as MK reports, but rather brspwhrʾ as a single word and with an s as Drower (Reference Drower1946: 331) recorded (collated from the original manuscript).Footnote ¹²⁶

Wohlstein translated שיפורי as ‘Aechtungen’ and, in light of the collocation with שמתתא ‘bans’, compared the reference to שיפורי in b.MQ 16a: בארבע מאה שיפורי שמתיה ברק למרוז ‘Baraq banned Meroz with four hundred shofars’.Footnote ¹²⁷ Sokoloff Reference Sokoloff2002: 1139 accordingly classifies שיפורי in the above-cited bowls s.v. שיפורא ‘shofar’, meaning c: ‘used for proclaiming a ban’, noting that in the examples from the incantation bowls the term refers to a type of demon. MK rejects this derivation of שיפורי, stating that in the above-cited texts the word “has no connection with the Hebrew [sic] word שיפורא ‘shofar, trumpet’”, and that both occurrences are “shortened variants of שופהרי, meaning something like ‘exorcism’ or ‘slander’”. Although the collocations שיפורי ושמתתא and שופהרי שמומתות בית כנישתא appear to support MK's basic identification of שיפורי with שופהרי, they also support the interpretation of שיפורי as referring to the use of the shofar in excommunicating. This interpretation is confirmed by the occurrence of שיפורי in an enumeration of maleficent forces in collocation with תברי “‘broken’ sounds of the shofar” Footnote ¹²⁸ in DS 9: 6 (= JNF 317): נידרי ושבועתאתה תברי ושיפורי שיקופתא ענקא לוטתא וקרותא “vows, oaths, ‘broken’ sounds of the shofar, shofars, smiting, ʿnqʾ-demon, curse, and imprecation” (see Figure 54).

Figure 54. תברי ושיפורי (DS 9: 6)

An even more explicit collocation appears in VA 3381: 10–11: וכל לוטתא ושיפורי ושמתתא ותיברי וגזירתא ואחרמתא ‘all curses, shofars, bans, “broken” sounds of the shofar, court oaths, anathemas’.Footnote ¹²⁹

The context of bans (as maleficent forces) in these texts is unmistakeable, as the word תברי occurs in the Talmud with the meaning ‘“broken” sounds of the shofar’ in b. MQ 17a–17b,Footnote ¹³⁰ precisely in a context of banning with a shofar(-ban):

It would thus appear that Wohlstein's identification of שיפורי in the magic bowls with Hebrew שופר ‘shofar (used for proclaiming bans)’ and Aramaic שיפורא ‘idem’ was correct, and that שופהרי is most probably a variant of שיפורי, rather than being the basic form as posited by MK.Footnote ¹³¹

73. MK cites the parallel to AMB B13: 6 from the Christie's bowl, and correctly notes that the emendation proposed by Naveh and Shaked for {ת}יתי בידיה חרבא דקטלא ‘he(?) comes and in his hand there is a sword of slaying’ is not supported by the other textual witnesses, which all read תיתי.Footnote ¹³² However, her own translation, ‘you shall come with a sword that kills’, cannot be accepted as it does not take account of the 3 s. possessive pronoun on בידיה ‘in his hand(s)’. Since חרבא is feminine, it may be taken as the subject of the verb (in the G-stem), providing an alternative translation of ‘let a sword of killing come into his hand’. The direct invocation of lord Bagdana would then begin with the following word את ‘You!’, which would also suit the use of the personal pronoun to change the discourse to direct address.