Aramaic Incantation Texts between Orality and Textuality

Orality and Textuality in the Iranian World

Patterns of Interaction across the Centuries

Edited by

Julia Rubanovich

LEIDEN | BOSTON

For use by the Author only | © 2015 Koninklijke Brill NV

Charles

Rectangle

Contents

Acknowledgments ix List of Illustrations x Notes on Transliteration and Abbreviations xii Notes on Contributors xiii

Introduction: New Perspectives on Orality in Iranian Studies 1Julia Rubanovich

Part 1Approaching Orality

1 Memory and Textuality in the Orality-Literacy Continuum 19Karl Reichl

2 Orality and EsotericismReflections on Modes of Transmission in Late Antiquity 43

Shaul Shaked

Part 2Sacred Traditions and Oral History

3 Irano-Talmudica IIIGiant Mythological Creatures in Transition from the Avesta to the Babylonian Talmud 65

Reuven Kiperwasser and Dan D.Y. Shapira

4 The Islamic Ascension Narrative in the Context of Conversion in Medieval IranAn Apocalypse at the Intersection of Orality and Textuality 93

Maria E. Subtelny

5 The Motif of the Cave and the Funerary Narratives of Nāṣir-i Khusrau 130

Jo-Ann Gross


Charles

Rectangle

contentsvi

Part 3Iranian Epic Tradition

6 ʻThe Ground Well Trodden But the Shah Not Found . . .ʼOrality and Textuality in the ʻBook of Kingsʼ and the Zoroastrian Mythoepic Tradition 169

Yuhan Sohrab-Dinshaw Vevaina

7 ʻThe Book of the Black Demon,ʼ or Shabrang-nāma, and the Black Demon in Oral Tradition 191

Gabrielle R. van den Berg

8 Why So Many Stories? Untangling the Versions of Iskandar’s Birth and Upbringing 202

Julia Rubanovich

9 Some Comments on the Probable Sources of Ibn Ḥusām’s Khāvarān-nāma and the Oral Transmission of Epic Materials 241

Raya Shani

10 Professional Storytelling (naqqālī) in Qājār Iran 271Ulrich Marzolph

Part 4Oral and Literary Traditions as Channels of Cultural Transformation

11 The Literary Use of Proverbs and Myths in Nāṣir-i Khusrau’s Dīvān 289

Mohsen Zakeri

12 Classical Poetry as Cultural Capital in the Proverbs of Jews from IranTransformations of Intertextuality 307

Galit Hasan-Rokem

13 Gashtak: Oral/Literary Intertextuality, Performance and Identity in Contemporary Tajikistan 316

Margaret Mills and Ravshan Rahmoni


Charles

Rectangle

contents vii

14 The Tale of ʻThe Old Woman on the MountainʼA Jewish Folktale from Afghanistan 342

Tsila Zan-Bar Tsur

Part 5Performative Aspects of Orality in Visual Artefacts

15 Aramaic Incantation Texts between Orality and Textuality 365Charles G. Häberl

16 Between Demons and KingsThe Art of Babylonian Incantation Bowls 400

Naama Vilozny

17 Between Written Texts, Oral Performances and Mural PaintingsIllustrated Scrolls in Pre-Islamic Central Asia 422

Frantz Grenet

Index 447


Charles

Rectangle

Part 5 Performative Aspects of Orality in Visual Artefacts

⸪


Charles

Rectangle

© koninklijke brill nv, leiden, 2015 | doi 10.1163/9789004291973_017

CHAPTER 15

Aramaic Incantation Texts between Orality and Textuality

Charles G. Häberl

The subject of this paper is a corpus of texts composed within a relatively nar-row timeframe, during the late Sasanian and early Islamic period. These texts were inscribed upon terracotta bowls and buried in courtyards and under thresholds. Thus far, most attempts to interpret these texts have been narrowly philological, and the primary focus of these attempts has been their relation-ship to other written texts, particularly those from the canons of religious lit-erature. The philological approach has yielded impressive results towards the interpretation of these texts, but much about them still remains a mystery.

In this paper, I discuss a common feature of this genre – namely, the illocu-tionary act – in order to illustrate its fundamentally performative nature. I sub-mit that our interpretation of these texts can be much refĳined by approaching them not only as written texts, but also as transcriptions of ritual utterances, and therefore the actual speech of the magician. While the inscriptions on the bowls were obviously the fĳinal product of the ritual, and the texts are clearly products of the literate cultures of late antique Mesopotamia, they nonethe-less bear all of the hallmarks of oral composition. As such, no treatment of these texts can be considered comprehensive unless it attempts to address issues of their composition, transmission, and reception.

Introduction

For over sixteen decades, ever since Austen Henry Layard began discover-ing Aramaic incantation texts inscribed upon terracotta bowls during the course of his excavations1 and convinced the British Museum to acquire a col-lection of them, scholars have examined these texts in the hope of answering the age-old question: ‘What possessed these people to do something like this?’

1 Layard 1853. The bowls (sometimes designated ‘Ellis 1–7’ after Layard’s consultant Thomas Ellis) are discussed on pp. 509–26.


Charles

Rectangle

366 Häberl

Written upon unglazed plain-ware bowls drawn from the repertoire of ordi-nary Sasanian household pottery, the texts seem to betray our high expecta-tions of the ancient civilizations of Mesopotamia, the birthplace of literacy. The handwriting of their authors is almost always sloppy to the point of illeg-ibility, their language variously characterized as ‘corrupt,’ ‘debased,’ or ‘full of mistakes,’ and their content often described as ‘formulaic,’ ‘repetitive and stilted.’2 Henri Pognon, who published the fĳirst corpus of Mandaic incantation texts in 1898, warned his readers:

Not all of the Mandaic inscriptions of Khouabir are interesting: they are fĳilled with mistakes and inaccuracies, and some, composed of scraps of sentences borrowed from diffferent formulas and written by ignorant scribes, do not, so to speak, make any sense. Others contain so many mis-takes that they would be nearly untranslatable if the formulae which are read in them were not found written more correctly in other inscriptions (Pognon 1898: 15).3

Furthermore, many of them (perhaps as many as one in fĳive) appear at fĳirst glance to have been ‘written’ by complete illiterates, consisting of a series of repeated squiggles, dots, and lines not resembling any known alphabet – not even Pahlavi.4 Pognon was of the opinion that these were created by ‘charla-tans’ in imitation of legitimate scribes in the hopes of duping the predomi-nantly unlettered and therefore presumably gullible public.5

It was for these reasons (among others) that Rudolf Macuch and Ethel Drower refused to dignify the Mandaic texts among them by incorporating their contents into their dictionary of Mandaic, except where they ‘grudgingly

2 Segal 2000: 29. Bowls from this volume are henceforth designated as ‘Segal 001A–075A’ and ‘Segal 076M–116M.’

3 The translations, if not stated otherwise, are mine. Bowls from this volume are henceforth designated as ‘Pognon 1898 1–31.’

4 Twenty-two out of the 120 in the collections of the British Museum. Erica C.D. Hunter esti-mates that these ‘pseudoscript’ bowls comprise some 20% of the known corpus of incanta-tion bowls (personal communication, December 2, 2008).

5 Pognon 1898: 15. Pognon’s explanation for the pseudo-script bowls was adopted by those who followed him; see, for example, Montgomery 1913: 27–28: ‘. . . but many were written by care-less scribes, and many by illiterate ones, probably often by laymen, who afffected to write their own prescriptions,’ and Rossell 1953: 13, n. 1: ‘Few of the clients for which the bowls were inscribed could read. The many “fakes” unearthed show that this fact was made use of by some unscrupulous “scribes”.’ To my knowledge, this explanation has never been seriously challenged.


Charles

Rectangle

367Aramaic Incantation Texts Between Orality And Textuality

quoted examples worth an entry or a reference and have omitted much that is doubtful or obviously corrupt’ (Drower and Macuch 1963: vi). In the introduc-tion to his Handbook of Classical and Modern Mandaic, Macuch expressed his opinion as follows:

Magic bowls and rolls usually contain a mass of hardly decipherable or completely incomprehensible nonsense. They were written against the demons who were supposed to understand their magic language. Their defective and often careless writing makes their reading difffĳicult and their interpretation doubtful. The picture of the language they give us is very incomplete (Macuch 1965: lix).

Macuch’s explanation, that language of the bowls was a kind of glossolalia intended for demons and not humans, is certainly original, and there is some evidence that incomprehensible phrases (voces magicae) were a component of the incantation ritual. This explanation has not been followed by any other scholars for the obvious reason that most of the bowl texts are composed in comprehensible, if non-standard, forms of Aramaic, and that it is not neces-sarily the language in which they were composed but rather the circumstances of their composition and preservation which have rendered them difffĳicult to read. Most scholars who have examined the language of these texts have done so explicitly in comparison with the standardized, written forms of Aramaic used by the communities who created them, occasionally allowing some degree of latitude for the ‘vernacular’ nature of the texts.6 The comparison is almost always to the detriment of the incantation texts, as indicated by words such as ‘mistaken,’ ‘inaccurate,’ ‘corrupt’ and ‘defective,’ which were frequently deployed by these scholars.7

Pognon and Macuch, among other scholars of the bowl texts, have also oper-ated under the assumption that the primary focus of the incantation ritual was the writing of the text, which has often given them occasion to deplore the careless, sloppy, inconsistent, and otherwise lamentable state of this writing. It cannot be said that most scholars of the incantation texts have attempted to examine the texts and the language in which they are composed in their own

6 See, e.g., Segal 2000. J.B. Segal remarks: ‘The incantation texts are written in popular language. We should not, then, expect them to conform to the standard norms of grammar’ (ibid.: 30).

7 In addition to the scholars and works mentioned above, see also Yamauchi 2005: 72: ‘Few of the magician’s clients are literate enough to criticize what the magician has written for them [. . .]. Therefore, it should not surprise us to meet numerous errors in our texts.’ Yamauchi does acknowledge that ‘some of the variations are probably dialectal forms’ (ibid.: 73).


Charles

Rectangle

368 Häberl

right and according to their own merits. It is undeniable that these texts are formulaic and repetitive; far from reducing their value as sources for study of the region, its languages and literatures, however, this observation provides the key to understanding these texts and refĳining their interpretation.

The Praxis of the Incantation Ritual

To what extent is the ritual that produced these texts reflected within the texts themselves? We know virtually nothing about the praxis, and the texts only rarely offfer clues.8 These clues do suggest that the act of writing the text, described as taking place while seated upon a stone, was an integral part of the ritual; the action is usually described with the imperfect (Example 1 below), suggesting that the writing occurred during the course of the ritual, but the action is occasionally rendered by the perfect as well (Example 2 below):

1. eʾtteḇ w- eʾḵtoḇennen eʾl-kāsā haḏṯā

I sit and write them [i.e., all the curses] upon a new bowl . . .92. yaṯḇiṯ u-ḵṯaḇtennen [. . .] ʾel-kāsā haḏṯā

I sat and wrote them [. . .] upon a new bowl . . .10

‘By the ban of Buḡdānā,’11 an incantation which is uniquely attested both in the bowls written in the square Aramaic script and those written in the Mandaic script, includes the expression ‘I have written (or perhaps prescribed) this against you . . .’ (Aramaic: hā kəṯaḇiṯ liḵ; Mandaic: hāzen kəṯaḇteleḵ),12 referring explicitly to the text of the incantation that was still in the process of being written. Furthermore, in addition to this quite literal reference to writing, the text also makes a fĳigurative reference to the writing of a deed of divorce (Aramaic geṭ) as a metaphor for banishing the liliths that are plaguing the cli-ent. This particular incantation text demonstrates the centrality of writing to

8 These clues are confĳined largely to the corpus of bowls in the Mandaic script, as the bowls in other scripts are generally mute about praxis (Erica C.D. Hunter, personal communica-tion, 2 December 2008).

9 Found in eight bowls: Pognon 1898 18, Texts I–III in Lidzbarski 1902, and Segal 094–097M.10 Found in three bowls: Pognon 1898 15 and 24, and Text M in Gordon 1937: 95–100.11 ‘By the ban of Buḡdānā’ (ʿal-ʾissur Buḡdānā) is the fĳirst line of the incantation proper,

appearing either at the very beginning of the text or following a short introductory formula.

12 These two phrases are taken from Montgomery 1913, 11, and Louvre AO 2629, a.k.a. Lidzbarski 1902 V, respectively.


Charles

Rectangle


the incantation ritual to a greater degree than any other attested in the corpus, identifying it as the product of an environment steeped in literacy.

Intriguingly, two texts indicate that the object of the incantation was ‘thor-oughly bound, sealed, tied, and charmed with whispers’ (ʾăsure ʾăsirin wa-ḥṯume

ḥăṯimin wi-qṭure qiṭrin u-lḥuše ləḥišin),13 and a single text appears to include the instructions ‘you murmur and you whisper’ (tirtəʿim wə-lāḥəšittum).14 Thus, in addition to the written component of the ritual, there was also likely an oral component. This is not at all unexpected; in languages such as Arabic, Hebrew, and Aramaic, the verb describing the act of reading is not distinguished from that of reciting a text aloud, suggesting that those who composed these texts, like others the world around, spoke their compositions aloud while reading them. How, then, were these texts transmitted from magician to magician?15 Were they primarily compositions in writing that happened to be recited, or primarily oral compositions that happened to be written down? The evidence, I contend, overwhelmingly suggests the latter.

The Transmission of the Incantation Text

It is worth noting that the belief in the magical power of words, and particu-larly names, is near universal across preliterate cultures. But what, exactly, is intended by the term ‘magic’? Since Frazer, it has been customary to speak of ‘magic’ in opposition to other abstract concepts such as ‘religion’ and ‘sci-ence’. As the classical and ancient Near Eastern sources attest to a distinction between what we might deem magical practices and religious ones, this approach has enjoyed a certain degree of legitimacy, even if it has been sub-ject to much criticism.16 Unfortunately, by defĳining ‘magic’ not independently

13 Naveh and Shaked 1993, Bowl 14, which is parallel to Montgomery 1913 5. Note that the Aramaic root l-ḥ-š can refer both to the act of whispering and charming; cf. the Qurʾanic ʾaʿūd̲h̲u [. . .] min sh̲̲arri ʾl-waswāsi ʾl-khannāsi (‘I seek refuge [. . .] from the evil of the retreating whisperer’; Q 114: 1–4).

14 Segal 033A: 10–11. See also Segal 2000: 27 for a fuller discussion on whispering in the incan-tation texts. Yamauchi also addresses the oral component of the ritual in Yamauchi 2005: 54–55, and explicitly compares it to magical rituals as performed in oral cultures today.

15 The question of the transmission of these texts was fĳirst addressed in the literature by Naveh and Shaked, who submit that ‘. . . the mechanism of the transmission of magical formulae [. . .] is not entirely that of a scribal tradition, but is akin in some respects to oral transmission’ (Naveh and Shaked 1998: 27).

16 For a history of the discussion, see Graf 1997: 8–19. In the context of the ancient Near East, note the numerous prohibitions contained within the text of the Hebrew Bible against ritual practices that we would consider ‘magical,’ such as Deut. 18:10–12 and Is. 8:19–22.


Charles

Rectangle

370 Häberl

according to its own distinctive features but rather in opposition to these other categories, we neglect to explain precisely what it entails. Furthermore, the contrived nature of these categories is apparent from the lack of precise boundaries between them, which have never been fĳixed and probably never will be. Any attempt to defĳine the term ‘magic’ either within the context of late antique Mesopotamia or even according to its use in recent scholarship would fall well beyond the ambit of this paper, so for the sake of brevity I shall restrict the discussion to the bowl incantation rituals, which may be considered as ‘magical’ insofar as they do not bear the hallmarks of contemporary religious rituals,17 but nonetheless represent an important element of the ancient Near Eastern relationship between the human and the supernatural.18

While rituals are a universal feature of human societies, it has long been acknowledged that magical rituals (strictly defĳined) are more typical of prelit-erate cultures than literate ones. The incantation texts in particular bear many of the hallmarks of oral compositions rather than of literary compositions: they were composed in the popular language rather than any literary standard, and most are attested in multiple variants, similar to one another in their con-tent and formulaic structure, but nevertheless difffering in some way from one another as well.

Additionally, although these texts are obviously only attested in written form, they are wholly absent from the canons of the contemporary religions that have survived to the present date, perhaps because magical rituals such as these were generally anathematized by the normative religious traditions of the region. As a result, it is likely that they were transmitted from person to person, and the evidence of the bowls, which preserve multiple attestations of many incantations, suggests that the same incantations were performed repeatedly over time. For these reasons, they are best classifĳied with other artistic forms of communication conventionally described as folklore.19

17 Note, for example, that the bowl texts are, as a rule, composed in the profane language of the people rather than the sacred language of scripture; for more on the distinction between sacred and profane language in the context of magical and religious rituals, see Tambiah 1968: 179–82.

18 Following the model established by Graf: ‘. . . it will be necessary for us to consider and analyze the ancient use of the term magic as it constitutes an element of the indigenous discourse on the relationship between the human and the supernatural’ (Graf 1997: 19). [For an attempt to interpret Mesopotamian incantation bowl paintings in the context of Jewish magic, see N. Vilozny’s article in this volume.]

19 Cf. Dundes 1999: 2: ‘. . . the basic distinctive criteria of folklore: namely, multiple existence and variation.’ For a more recent defĳinition of folklore, with specifĳic reference to folklore from the region of the Middle East, see Reynolds 2007: 25–28.


Charles

Rectangle


Much like folklore, many of the attested texts can be found written in the scripts of diffferent religious communities, indicating that these texts traversed sectarian boundaries. ‘By the ban of Buḡdānā’ is usually described as a Jewish incantation on the basis of its content, which employs the common Jewish trope of the divorce document as a model to dismiss demons, and makes ref-erence to the ‘great inefffable name’. This same device also appears in several bowls20 that specifĳically attribute it to R. Joshua b. Peraḥya, whom Jewish tradi-tion dates to the reign of Alexander Jannaeus (103–76 BCE) (Montgomery 1913: 226). Apart from the ‘smoking gun’ provided by this explicit attribution, the tradition of issuing divorce documents does not exist among the Mandaeans and the faith makes no accommodation for divorce in the present date (Drower 2000: 62). Consequently, the hypothesis of a Jewish origin for these particular formulae would appear to be well-grounded.

The fact that the texts transcend sectarian boundaries is of no small rel-evance to the question of their transmission. It suggests that these formu-las were transmitted in the manner that folklore the world around has been transmitted, from person to person and by word of mouth rather than written sources. This is further supported by the orthography of the texts, which devi-ates from the standard orthography in a way that suggests vernacular influ-ence21 or even unfamiliarity with the standard. A particularly glaring example of this phenomenon, a Hebrew formula (Num. 10:35) embedded within an incantation text, was published by Naveh and Shaked, as in Example 3 below:

3. AMB Bowl 3: <WḤYḤY BYN NWSʿ HʾRWN WYMR MWŠH QWMʾ YHWH WYPWṢW ʾYBʾK WYNSW M[. . .] MYPNK>

Masoretic Text: Wa-yhi bi-nsoaʿ hā-ʾāron, way-yomer Moše: ‘Qumā YHWH,

wə-yāp̄uṣu ʾoyḇeḵā, wə-yānusu məśan eʾḵā mip-pāneḵā’

And it came to pass, when the ark set forward, that Moses said: ‘Rise up, O LORD, and let Thine enemies be scattered; and let them that hate Thee flee before Thee’.22

Scarcely a single word within this formula has been spelled according to the standard orthography, and the reinterpretation of the preposition bə- with the infĳinitive construct nəsoaʿ (indicating an adverbial clause ‘when it set out’) as

20 Montgomery 1913 8, 9, 17 [in Mandaic], as well as 32 and 33 [in Syriac]; AMB Bowl 5.21 Rossell argues that the ‘. . . the unlearned style with its many variations of spelling fre-

quently reflects actual speech, thus throwing new light on the phonetics and other lin-guistic features [of the language]’ (Rossell 1953: 13).

22 Translation from the 1917 Jewish Publication Society edition.


Charles

Rectangle

372 Häberl

the preposition ben ‘between’ and the active participle noseaʿ ‘setting out’ sug-gests rote learning rather than any in-depth familiarity with Hebrew grammar. If we then assume that these texts were transmitted by word of mouth rather than through written sources, what relevance does this bear upon our under-standing and interpretation of these texts?

In his work on the psychodynamics of orality, Walter Ong uses the term ‘oral universe of communication’ when discussing the diffference between the ‘mentalities’ of primarily oral and primarily literate cultures (Ong 2002: 2). One of the fundamental theses of Ong’s work is that it is difffĳicult for people who have been steeped in a literary tradition to comprehend the context within which oral compositions are created and the roles that they play in a primarily oral society. With reference to the bowl incantations, we may naturally enough interpret these texts purely as epigraphic data, divorced from this context, but for the composers of these texts, it would be just as natural to feel there is little if any existence to them apart from it.23 The key to understanding the nature of these texts, the contexts in which they are created and the roles that they play in late antique Mesopotamian society, lies within one of the most common features of the text themselves: the illocutionary act.

Illocutionary Acts within the Incantation Texts

Bearing in mind that sound is dynamic in comparison to the static nature of printed material – as it is being produced, an utterance is immediately gone, and cannot be captured or frozen (although it can obviously be recorded with the proper technology). While speech, like writing, is a vehicle for the commu-nication of thoughts, it difffers in that it consists of a series of utterances, which are transient actions, as opposed to the static and enduring physical signs that are the components of writing. Speech serves as an action most transparently in the form of illocutionary acts, according to which real actions are accom-plished by certain performative utterances (such as ‘I hereby pronounce you man and wife,’ ‘I baptize you in the name of the Father, the Son, and the Holy Ghost,’ or ‘I name this ship Generalissimo Stalin’).24 In English, these utterances are grammatically expressed in the fĳirst person (singular or plural) and in the

23 For a possible parallel in the composition of folk songs, see Herzog 1965: 174.24 The term ‘performative utterance’ was fĳirst applied to this phenomenon by John L. Austin

in his 1955 William James lecture series delivered at Harvard University; see Austin 1962; idem 1979: 233–52.


Charles

Rectangle


present tense. In Aramaic as well as in other related Semitic languages such as Arabic and Hebrew, they are regularly expressed in the perfect instead, as in Examples 4–5 below:

4. ṭallaqtuki (Arabic) I (hereby) divorce you.5. b-i nišbaʿ-ti [. . .] ki-l-ḥorbā yi-hye ha-bbayiṯ ha-zze (Hebrew) I (hereby) swear by myself [. . .] that this house shall become a ruin (Jer.

22:5).25

Returning now to the act of writing as described in the incantations them-selves, recall that it appears both frequently in the imperfect (‘I write’) and also occasionally in the perfect, which might seem anomalous with reference to an action which has not yet been completed were not for the fact that the Semitic perfect also designates illocutionary acts. Additional evidence supporting the interpretation of the phrase ‘I write’ as an assertion and therefore an implicit performative (i.e., ‘[I assert that] I write’) is provided by other explicit perfor-mative utterances that occur in the course of the incantation texts, such as ‘I adjure you,’ ‘I bid you swear,’ and ‘I invoke’. In the British Museum collection of incantation bowls, the phrase ‘I adjure you’ occurs 17 times in the particip-ial present tense, 19 times in the perfect, and once in the imperfect, often in the same exact context, such as in the formula from ‘By the ban of Buḡdānā,’ ‘I adjure you that you be smitten in the membrane of your heart . . .’.26 The occurrences of these performative utterances in the British Museum collec-tion are documented in Table 1 below.

25 The above examples are taken from W. Wright, A Grammar of the Arabic Language II §1(d), and W. Gesenius, E. Kautzsch, and A.E. Cowley, Gesenius’ Hebrew Grammar §106 2(b), which deal with the use of the perfect in performative speech acts.

26 This exact formula is found in at least twelve bowls: three in the participial present (Segal 013A, Montgomery 1913 11, and Text A in Geller 1997), and nine in the perfect (Montgomery 1913 18, Iraq Museum 5497 published as Text G in Gordon 1934a, Iraq Museum 9737 published as Gordon 1934b, Iraq Museum 11113 published in Gordon 1941: 350–52, Iraq Museum 18N18 published in Hunter 1995, Schøyen MS 1928/47 published in Shaked 1999: 193–94, Louvre AO 2629, Segal 098M, and Segal 099M).


Charles

Rectangle

374 Häberl

table 1 Performative utterances in the incantation texts

I bid (you) swearomiṯā perfect (x1: 079M: 3)momenā participial present (x5: 027A: 6, 031A: 5, 039A: 9, 056A: 4, 057A:

13)mominā participial present (x12: 078M: 6,12; 094M: 14; 095M: 16, 096M:

14, 097M: 8, 098M: 3,6, 099M: 7, 100M: 7, 103M: 20, 105M: 9)I adjure (you)ʾašbaʿiṯ perfect (x10: 019A: 9, 027A: 4, 028A: 15, 040A: 2, 041A: 3, 045A:

1, 048A: 1,12, 059A: 6, 069A: 4)ašbiṯā perfect (x9: 078M: 5,12, 094M: 13, 096M: 13, 098M: 6, 099M:

4,7, 105M: 5,9)ʾašbaʿ imperfect (x1: 048A: 26)mašbaʿnā participial present (x15: 013A: 4,5, 019A: 8,9, 027A: 6, 030A:

3,5, 031A: 5, 042A: 1, 044A: 5, 045A: 3,4,5, 046A: 1, 055A: 4)mašbānā participial present (x2: 098M: 4, 100M: 8)I invokeqəreṯ perfect (x1: 034A: 4)qəriṯ perfect (x4: 082M: 7,8,9,10)qārenā participial present (x3: 019A: 3, 029A: 5, 043A: 1)

Each phrase appears either equally in the perfect and the participial present (denoting either the present progressive or immediate future) or predomi-nantly in the participial present. One might also add other inexplicit illocu-tionary acts, especially the passive participial forms məzammənā ‘designated [is . . .],’ hăp̄iḵā ‘overturned [is . . .],’ kəḇišā ‘subdued,’ ʾăsirā wa-ḥṯimā ‘bound and sealed,’ and šərir wə-qim ‘confĳirmed and established,’ which are among the most characteristic phrases of these texts and belong to the same category of performative utterances as passives in the English language, such as ‘this meeting is now adjourned,’ ‘you are hereby authorized,’ ‘war is declared,’ and so on. The frequent occurrence of these explicit and/or inexplicit illocutionary acts throughout the corpus indicates the fundamentally performative nature of the incantation texts, which has thus far escaped the attention of all schol-ars of these texts save Shaul Shaked.27 Shaked explicitly compares the incan-tations with the formulaic language of legal documents which creates, by its mere utterance, a new legal situation, and cites J.L. Austin on performativity

27 Shaul Shaked explicitly identifĳied the performative nature of these texts in Shaked 1999.


Charles

Rectangle


in this regard. Although Shaked does not feel that much more could be gained by applying Austin’s work to the incantation texts, he does acknowledge that ‘. . . the legal framework in these spells is a make-believe, it is based on a literary (as well as on a magical) convention, and needs to be supplemented by a whole range of other devices to make it efffective’ (Shaked 1999: 174) or, to use Austin’s terminology, felicitous.

Readers familiar with Austin’s work will undoubtedly recognize ‘I name this ship Generalissimo Stalin’ as a banner example of an ‘infelicitous’ act, one which is unsuccessful due to the circumstances under which it was performed and its reception by its audience. Austin writes:

Suppose that you are just about to name the ship, you have been appointed to name it, and you are just about to bang the bottle against the stem; but at that very moment some low type comes up, snatches the bottle out of your hand, breaks it on the stem, shouts out, “I name this ship Generalissimo Stalin,” and then for good measure kicks away the chocks. We agree that the ship certainly isn’t now named the Generalissimo

Stalin, and we agree that it’s an infernal shame (Austin 1979: 239).

Just as the success of a given illocutionary act depends upon the details of its pro-duction and reception, so too must it be noted that courses of action and even basic attitudes towards a given issue depend upon the efffective use of words, or rhetoric, particularly in pre-literate cultures. For most literate Westerners, such rhetoric is reserved exclusively for certain rare ritual occasions such as a legal case, a political debate, or the oral defense of a doctoral dissertation, but something of the role that rhetoric can play in even the most mundane interac-tions can be ascertained by comparing business in an American supermarket and in a Middle Eastern souq.28 In the former, the prices of the products are fĳixed for all customers, and the process has been entirely automated; indeed, often no direct interaction between the customer and any other human being is necessary. In the latter case, ‘business’ is not simply business; it is a ritual in which the efffective use of rhetoric can make the diffference between making a great deal or losing the shirt offf one’s back. Likewise, all parties to such a ritual perceive the circumstances under which it is performed, the efffective use of rhetoric in its articulation, and its reception at the moment of utterance as integral to its success. For this reason, the potential for shenanigans of the sort suggested by Pognon and Rossell is extremely limited if not completely absent; anyone attempting to reproduce the incantation ritual would need to

28 I have borrowed this metaphor from Ong 2002: 67–68.


Charles

Rectangle

376 Häberl

be well-versed in all of its performative components in order for the potential clients to be convinced of its felicity.

The Production of the Incantation

Let us now return to the production of each text. Ong notes that the act of writing remained entirely subordinate to speech throughout much of recorded history. It was only through gradual stages that written compositions become compositions in writing, rather than a verbatim mimicry of speech. In fact, even today many cultures have adopted writing without having internalized it, with the result that their written literature mirrors their speech, following the same principles of composition (Ong 2002: 36–57). Of relevance to the incan-tation texts, Ong notes that certain literary compositions, particularly those from cultures that have not yet fully internalized literacy, betray certain hall-marks of oral composition, among which the most salient is an extensive use of formulaic expressions. He defĳines the formula in a more general context as phrases or set expressions (such as proverbs) that are more or less exactly repeated, and that have a function in oral culture more crucial and pervasive than any they may have in literate societies (ibid.: 25–26). Ritual and magical texts difffer from other oral compositions in that their recitation tends to be more fĳixed and that greater effforts are made at reproducing the formulae ver-batim, although the rate of success varies, as is obvious from the study of the incantation texts, particularly those that reproduce the same material.

While scholars have been aware of the formulaic nature of the incantation texts from the very inception of their study, few went beyond simply acknowl-edging the existence of ‘duplicate’ inscriptions, and marshaling them, where needed, to reconstruct any text that might be missing or ‘corrupt’. These recon-structions were generally motivated by a belief in an Urtext that might be recovered by comparison of the imperfect copies that had survived, a model imported from the fĳield of manuscript studies. Montgomery established a schema for the comparison of several texts, including all the versions of ‘By the ban of Buḡdānā’ that were then known (Montgomery 1913: 167–73), by placing each text in a parallel column to economize space and highlight the corre-spondences between them. Naveh and Shaked employed a similar schema for their Bowl 12 in Amulets and Magic Bowls, comparing its text with an unpub-lished bowl in the Metropolitan Museum in New York and an amulet in the Israel Museum (Naveh and Shaked 1998, Amulet 15; ibid.: 188–97). Shaked also employed such an analysis in developing his typology of the divorce for-mula in his article on the ‘Poetics of Spells’ (Shaked 1999), which includes a previously published version of the ban of Buḡdānā incantation (from the


Charles

Rectangle


Schøyen Collection, MS 1928/47). Likewise, Müller-Kessler has frequently applied a structural analysis to incantations across a variety of corpora,29 but the most comprehensive attempt to analyse the bowl incantations structurally has been Segal’s Aramaic and Mandaic Incantation Bowls in the British Museum. In this work, Segal identifĳies recurring formulae and four distinct incanta-tion types, which he labels ‘Refrains’. Building upon this typology, Hunter has collected and discussed all of the attested variants of the fĳirst of Segal’s four Refrains (Hunter 2002).

While the structural analysis developed by these scholars and others work-ing within the fĳield has served extremely well for the purpose of reconstructing these texts, particularly in the areas where they are most unclear, it does not address the signifĳicance of the variation within the text or attempt to iden-tify the principles motivating this variation, beyond ascribing any variation to ‘corruption’ from a putative pristine original. Considering that each ver-sion represents the product of a performative ritual rather than the scholastic reproduction of a manuscript, and is the instantiation not of a written Urtext but of a fluid oral tradition, this approach neglects a potentially important source of signifĳicant data for the interpretation of these texts.

As these texts are clearly vernacular compositions rather than formal works of literature, the modes of structural analysis pioneered by folklorists through-out the last century offfer more appropriate models for their analysis.30 Such an analysis requires that all the attested variants of a given type of incantation be gathered and the nodes of variation within that type be identifĳied. Once these nodes have been identifĳied, the principles governing the selection of each vari-able can then be considered (Propp 1968: 9). Ultimately, this analysis will arrive at a description of the incantation type according to its component parts and the relationship of these components to each other and to the whole (ibid.: 19).

As in other frequently repeated compositions, the most distinct nodes of variation can often be found at the margins of the incantation, which are struc-tured not unlike a letter with introductory (‘To whom it may concern . . .’) and concluding formulae (‘Sincerely . . .’). These marginal elements are sometimes optional, and vary considerably, as represented in Tables 2 and 3.31

29 See, e.g., Müller-Kessler 1996; 1998; 1999a; 1999b; Müller-Kessler and Kwasman 2000.30 The primary reference for this sort of analysis remains Thompson 1955–58, although the

original Aarne-Thompson classifĳicatory system has subsequently been expanded by other scholars to accommodate the exigencies of the folkloric material in which they work.

31 Cf. Herzog 1965: 173. In his article, Herzog refers specifĳically to the margins of folksongs that are repeated multiple times in the course of a performance. The initial rendition and the fĳinal rendition often have special features that mark these elements offf as marginal. See Herzog 1965: 173–74.


Charles

Rectangle

378 Häberl

table 2 Introductory Formulae

Health [and guarding and sealing] from heaven for . . . ʾasuṯā [u-nṭurṯā wa-ḥṯimṯā] min šəmayyā lə

Sealed and doubly sealed be the body and bones/house and threshold of-

ḥăṯimā wi-mḥattam pagreh wə-ḡarmohi / bayṯeh wi-squbteh [sic]This deed of divorce / counter-charm / amulet is [designated] for . . .[məzamman / məzammənā] hāḏen giṭṭā / qiḇlā / qəmeʿā ḏənā lə-32

table 3 Concluding Formulae

Amen, Amen [Selah, Hallelujah].āmen āmen [selā halləluyāh]Confĳirmed and established.šərir wə-qim

Peace be upon you, may you have peace.šəlāmā ʿəlāḵ šəlāmā nihyelāḵ

Apart from the introductory and concluding formulae at the margins of the texts, the texts are invariably composed in formulaic language as well, and the internal formulae themselves are often the nodes of considerable variation. While these phrases generally appear in binary pairs, they can be reduced to simply one term or extended to multiple terms; for example, the common for-mula ʾăsirā wa-ḥṯimā (‘bound and sealed’) is extended to ʾăsure ʾăsirin wa-ḥṯume

ḥăṯimin wi-qṭure qiṭrin u-lḥuše ləḥišin (‘thoroughly bound and sealed and tied and charmed with whispers’) in a variant attested in two bowls (Naveh and Shaked 1993, Bowl 14; Montgomery 1913 5).

This variation, which is the signature of oral composition, is amply dem-onstrated by the attested versions of the aforementioned ‘By the ban of Buḡdānā,’ which is one of the more frequently reproduced incantation texts. Of the bowls currently in the possession of the British Museum, fully three (two in the Mandaic script and one in the square script) contain this

32 Harviainen (1995: 56) notes that this is the sole introductory formula which appears in bowls of more than a single script, but it had not yet been attested in the Mandaic-script bowls. An incantation beginning with this same introductory formula was subsequently published by Segal 2000 as Bowl 091M.


Charles

Rectangle


incantation, and it is attested elsewhere in at least nine other published variants.33 Of the twelve attested variants of this incantation, Segal 013A has the distinction of being the very fĳirst incantation bowl to receive scholarly treatment, in Austen Henry Layard’s Discoveries in the Ruins of Nineveh and

Babylon (Layard 1853: 512–14).After the optional introductory formula, the text of ‘By the ban of Buḡdānā’

is composed of as many as four discrete sections, all of which offfer opportuni-ties for improvisation (and therefore variation). These sections can be fruit-fully compared to the parts of a contract:

I. Introductory Formula (Example 6 below) II. Preamble/Recitals (Examples 7–9 below) III. Body (Examples 10–17 below) a. Divorce/Expulsion Formula (Examples 10–15 below) b. Injunctions (Example 16 and Table 5 below) c. Restraining Orders (Examples 16–17 below) IV. Voces Magicae (Table 6 and Examples 18–20 below) V. Ending (Examples 21–25 below) VI. Concluding Formula (Example 26 below)

The Preamble identifĳies the contract by name and indicates the primary par-ties to it, typically the enchanter’s client or clients and the lilith. Recitals pro-vide some background to the contract (in this case, the activities of the lilith that have necessitated it) and any third parties, such as those beings to whose authority the enchanter appeals. The Body contains the heart of the contract, including the formula which divorces or expels the lilith, injunctions demand-ing that the lilith leave, and orders restraining her from troubling the clients and their property in the future. The Body is typically followed by some voces

magicae, mysterious words or phrases that have no apparent meaning. The contract concludes with an Ending which formally seals the document, and optionally a concluding formula.

These parts of the contract always appear in the same order, although they are not always reproduced in full, and one of the Mandaic versions interpolates some additional material into the body of the incantation that is not found in any of the other variants. Likewise, of those versions which are framed with introductory and/or concluding formulae, no two agree. An analysis of version Segal 013A follows:

33 In order of publication: Segal 013A, Louvre AO 2629, Montgomery 1913 11 and 18, Iraq Museum 5479, 9737, 11113, Iraq 18N18, Geller 1997 A, Schøyen MS 1928/47, Segal 098M and 099M.


Charles

Rectangle

380 Häberl

I. Introductory Formula

6. Haḏen giṭṭā lə-šeḏā u-l-[. . .] u-l-Sāṭānā u-l-Neriḵ u-l-Ḏaḵyā u-l-ʾAḇiṭur

This deed of divorce is for the demon and for [. . .], Satan, Nerig, Dakyā, Abiṭur

wə-li-ḏnā liliṯā ḏi-yiḇṭəlun min Bahrānduḵ baṯ Newānduḵ

and for this lilith so that they may withdraw from Wahrāmduxt, daughter of Nēwānduxt,

wə-min Māhdāḏ bar Iṣpandārmuḏ wə-min-bayṯah kulleh

and from Māhdād, son of Spandārmad (Esfandārmoz), and from her entire household.

Segal 013A opens with a fairly conventional introductory formula of the type that is also found elsewhere in the British Museum collection in fĳive other bowls in the square script (Segal 014A, 039A, 040A, 041A, and 066A), one incan-tation in the Mandaic script (091M), and two bowls in the Syriac script (119ES and 120SY), without any apparent connection to the formulae that follow.

The various malefĳicent influences against which the geṭ is intended to protect are drawn from a variety of backgrounds. The šeḏā is ultimately of Mesopotamian origin, but also known from the Bible (Deut. 32:17; Psalm 106:37). Sāṭānā is, of course, the infamous Adversary. Neriḵ resembles the Mandaic Neriḡ, whose name is ultimately derived from Nirgallu, the Babylonian god of war and pestilence; in the Mandaean tradition, this being is usually iden-tifĳied with the planet Mars. Daḵyā is an otherwise unattested being whose name would appear to mean ‘the pure,’34 and ʾAḇiṭur superfĳicially resembles the Mandaean Aḇaṯur, guardian of the fĳinal purgatory through which the souls of Mandaeans must pass and be weighed before they may enter the world of light. Liliths, who are the focus of this particular incantation, are a special class of female demons who prey upon women and small children.

Note that the names of the clients are typically followed by the names of their mothers rather than their fathers, possibly on the grounds that mater

certa, pater incertus. All four names are unmistakably Iranian. This suggests that the clients were Zoroastrian, rather than Jewish or members of any of the other ethno-religious communities of late antique Mesopotamia, but in the absence of further evidence one cannot necessarily assume this to be the case (Segal 2000: 24).

34 Shaked amends <DKYH> to <DNḤYŠ> Danaḥiš, a name of Iranian origin attested in other incantation texts (Shaked 1999: 184 and n. 45; for an etymology of this name and its other attestations, see idem 1985: 520–22).


Charles

Rectangle


II. Preamble/Recitals

7. ʾal-ʾissur Buḡdānā malḵəhon d-šeḏe u-d-diwe wə-šalliṭā rabbā ḏ-lilyāṯā

By the ban of Buḡdānā, king of demons and devils, and great ruler of liliths,

mašbaʿnā ʿălaḵi Ḥəḇalus liliṯā baṯ bəraṯah də-Zarnay liliṯā

I adjure you, Ḥbsls the lilith, daughter of Zrny the lilith’s daughter . . . ʾim dəḵar ʾim nəqeḇā

. . . whether male or female.

Since Lidzbarski’s publication of his variant of this text,35 it has been custom-ary to interpret the fĳirst word of this formula as a non-standard form of the preposition ʿal ‘on, upon’ but also ‘by; on the grounds of,’ making the fĳirst of the two lines above an adverbial clause modifying the second.36 This is the form found in one of the remaining variants in the square script (Iraq 9737), and its cognate (ʾəl) is found in two of the Mandaic variants.37 The term ʾissur, much like the term geṭ, is borrowed from the vocabulary of Halakhic jurisprudence,38 and means a ban or interdict, although the construct phrase of which it con-stitutes the nomen regens is ambivalent as to whether Buḡdānā is the intended recipient of the ban or its author. In the context of the incantation, the for-mer seems much more likely. The rectum of this construct phrase, Buḡdānā, is reasonably well known from many other inscriptions, including an especially detailed text published by Naveh and Shaked.39

In most versions, the title malkā ‘king’ appears in apposition to Buḡdānā. Further evidence for this supposition is provided by Naveh and Shaked (1998, Bowl 13), which identifĳies Buḡdānā as malkā rišā ḏ-šitte malḵəwāṯā (‘the king,

35 Lidzbarski 1902 (Louvre AO 2629).36 Shaked (1999: 185) dissents, suggesting instead that the enchanters interpreted both this

word and the following one as names. It certainly does appear to have been treated as such in some versions, but this does not necessarily rule out an alternate interpretation for other versions.

37 Louvre AO 2629 and Segal 098M. It must be noted that the voiced pharyngeal fricative ʿ has either merged with the glottal stop ʾ in Mandaic or has been deleted entirely.

38 Perhaps because the Mandaeans have no precise parallel to the ʾissur, the Mandaic ver-sions begin with the passive participle ʾəsirā ‘bound’.

39 Naveh and Shaked 1998 (Bowl 13). Shaked (1985: 516–19) suggests that this name derives from a reconstructed Iranian *bagadāna- ‘temple,’ which came to refer to the god therein worshipped by synecdoche, much as the term ‘the White House’ might refer to the President of the United States. The referent was then demonized by those who inscribed the bowls.


Charles

Rectangle

382 Häberl

head of sixty kingdoms’). What would at fĳirst glance appear to be the most logical interpretation is complicated by some of the other attested versions. Iraq 9737 offfers us ‘Buḡdānā and the kings of the devils’; Iraq 5497 offfers us ‘Buḡbānā of the king of kings of demons and devils’; Louvre AO 2629 offfers ‘Aḇuḡdānā to the kings of the devils,’ and Segal 098M inexplicably suggests ‘Buḡdānā, I swear I am the angels (malāḵawn, presumably a syncopated form of malāḵay-hon ‘their angels’ with a proleptic possessive sufffĳix) of the devils!’ The last variant is suggestive of oral transmission, given the similarity between the words malḵay-hon ‘their kings’ and malāḵay-hon ‘their angels’.

In most versions of the text, the name of the lilith is followed by a detailed description of her location and activities, although there is a considerable degree of latitude in the way that these things are expressed. In Segal 013A, this entire description is reduced to the four words ‘either male or female,’ pos-sibly indicating that the enchanter has skipped over part of the formula in the course of committing it to writing, leaving a lacuna.

8. mašbaʿnā ʿălaḵi ḏə-ṯimməḥan bə-ṭerpəsā lə-libbəḵon

I adjure you that you (f.sg.) be struck in the pericardium of your (m.pl.) heart

u-ḇ-morāniṯeh də-Tyqs gibbārā ḏ-hu šalliṭ ʿal-šede wə-ʿal-lilyāṯā

and with the lance of the hero Tyqs, who has control over demons and liliths.

Where Montgomery 1913 11, Geller 1997 A, and Louvre AO 2629 have tiṯməḥan

‘you are struck,’ Segal 013A has a more colloquial form timməḥan with assimi-lation of the ‘reflexive’ t- afffĳix. Strangely, the form of the verb (f.sg.) does not agree with the possessive pronoun on the phrase which follows (‘your [m.pl.] heart’). This confusion in agreement can be found throughout most versions of the incantation. Montgomery noted that many of the enchanters who used this text must also have had difffĳiculties with the technical term ṭarpəšā or ṭarpəsā ‘pericardium,’ as evidenced by the variation in the rendering of this word.40 Technical terms like ʾissur and ṭarpəsā appear to have been imper-fectly transmitted in most versions of this text, as one would expect if they were transmitted orally.

40 Montgomery 1913 11 ṭwrps, 18 ṭrps; Iraq 5497 ṭwpry, 9737 ṭprs, 11113 ṭpsy; Geller 1997 A ṭprs; Schøyen MS 1928/47 ṭrpsy; Segal 013A ṭyrpsʾ; Louvre AO 2629 ṭarpus; Segal 098M ṭarabus, and 099M tras. The most extreme example of this confusion is represented by Segal 099M: də- eʾṯməhi ḇ-eh bə-ṯəras gibbārā ‘that he was struck by it in the peri . . . pera . . . peri-whatsis of a hero’.


Charles

Rectangle


For the most part, the names of the major dramatis personae are fairly stable throughout the variants of the incantation. The name Buḡdānā in particular does not vary signifĳicantly from version to version, apart from two of the three versions in the Iraq Museum that were published by Gordon. The names of the liliths are likewise fairly consistent, despite some minor confusion over the name of the object of the incantation.41 There is, however, a surprising lack of agreement among the names given for the protective genius, as demonstrated by Table 4 below:

table 4 Variation in Names

Text Bugdānā Lilith(s) Grandmother Hero

Montgomery 1913 11 bgdnʾ ḥlbs zrny qyl. .s

Montgomery 1913 18 bgdnʾ bḥlbs zrny sq . . .

Iraq 5497 bwgbnʾ ḥbls zrnʾy qtrws

Iraq 9737 bgdynʾ ḥbls zrny qtrws

Iraq 11113 bgdnʾ ḥbls zrnʾy —Iraq 18N18 — — — sqrwṭ

Geller 1997 A — — — ʾytqdš

Schøyen MS 1928/47 bgdnʾ ḥbls zrny ḥwnrws

Segal 013A bgdnʾ ḥbsls zrny tyqs

Louvre AO 2629 abugdana hldas and taklat zarnia qaṭriauisSegal 098M bugdana hlabus zarnik ʿqarusSegal 099M bugdana hlbus — —

The only two versions which agree are Iraq 5497 and 9737, which both contain the name qtrws. Segal (2000: 56) interprets this name as a loan from Greek Καθαρός ‘pure,’ citing the name Daḵyā ‘pure’ which appears above in the intro-ductory formula to Segal 013A. A fourth version from the Iraqi Museum, Iraq 18N18, contains what appears to be a near anagram of the same name, sqrwṭ, apart from the fĳinal voiceless velar plosive. While such direct influence from

41 The form cited by Lidzbarski, hldas, may well be the result of the similarity between the two characters b and d in the Mandaic script. After controlling for other common lapsi

calumi, the attested forms appear to reduce to either Ḥbls (x5) or Ḥlbs (x5), but the name cannot be further reconstructed with any certainty.


Charles

Rectangle

384 Häberl

Greek would not normally be expected in a Mesopotamian text of this period, it is perhaps signifĳicant that all of the analogous names from the remaining attested versions end in a sibilant.

Given the fairly consistent rendering of Buḡdānā’s name and the names of the lilith and her grandmother, it seems unlikely that the extreme variation in the name of the protective genius is due to corruption. What could possibly condition the enchanters to substitute a diffferent name for each version of the bowl? One possible explanation for this variation is avoidance of a taboo name, of which the most obvious example is the Jewish taboo against pronounc-ing the name of God (illustrated below in section V. Ending). Crossculturally, avoidance of this sort is generally restricted to the names of evil spirits, certain animals, certain body parts, and deceased ancestors, but it is not unknown for the names of protective beings to be avoided as taboo, as the example of the inefffable name of God demonstrates.

III. Body

Segal 099M interpolates a few lines at this point not found in any other version, including the other Mandaic versions:

9. u-np̄aq kulhon qəryāṯā min bayṯah də-māraṯ Miryām paṯ Mādanoš

and all mishaps hereby leave from the house of Lady Miryām daughter of Mādanōš.

Həze ḏə-ḇaṭṭelteḵ min bayṯeh də-Yohānān hā həze

See that I hereby suspend you from the house of Yoḥānān. Look, see ḏi-ḵraḵteḵ min zireh wə-min bazireh wə-min bəneh

that I hereby circumvent you from the semen and seed,42 and from the sons

wə-min bənāṯeh də-Yohānān wə-min māraṯ Miryām paṯ Mādanoš

and from the daughters of Yoḥānān and from Lady Miryam daughter of Mādanōš.

In all versions, the following lines refer explicitly to the writing of the incanta-tion, which was then taking place; some of the versions even specify that the enchanter is in the process of writing a geṭ (Montgomery 1913 18, Iraq 5497, Iraq 11113, Iraq 18N18, Geller 1997 A, and Schøyen MS 1928/47).

42 Cf. JB 202: from his ‘night seed’ (=semen) and from his ‘day seed’ (=seeds of grain).


Charles

Rectangle


10. hā ḵəṯaḇiṯ bə-ḵon hā hā ḇaṭṭəliṯ yāṯəḵon

Look, I hereby write concerning you.43 Look, look, I hereby suspend you minnah wə-min bayṯah də-Ḇahrānduḵ baṯ Newānduḵ wə-min bərah

from her and from the house of Wahrāmduxt, daughter of Nēwānduxt, and from her son.

The presentative particle hā (here translated ‘look’) serves to the attention of the addressee (in this case, the evil spirits) to new information, further dem-onstrating the performativity of the text. The new information consists of the actions which are accomplished through the utterance of the verbs that fol-low it. The second of the two verbs is another minor locus of variation. Only the present version and Segal 099M have baṭṭəliṯ or its Mandaic equivalent baṭṭelteḵ ‘I suspend you’. While the characters b and p or l and r are not gener-ally confused with one another in either script, baṭṭəlit and paṭṭərit are phonet-ically similar, suggesting that these forms were transmitted orally rather than through the medium of text. Four of the attested versions have paṭṭəriṯ44 and two have tarrəḵiṯ (Iraq 9737 and Schøyen MS 1928/47), both of which mean ‘I (hereby) divorce’ in this context.

Two of the bowls from the Iraq Museum, Iraq 5497 and 11113, have a tripar-tite divorce formula incorporating both those synonyms and a third synonym as well, reminiscent of the irrevocable ṭalāqu bi-θ-θalāθati, or so-called ‘triple divorce’ of Islamic tradition:

11. hā paṭṭəriṯ yāṯəḵi wə-hā šabbəqiṯ yāṯəḵi wə-hā ṯarrəḵiṯ

Look, I hereby divorce you! Look, I hereby divorce you! Look, I hereby divorce

yāṯəḵi bə-ḡeṭ piṭṭurin

you with a deed of divorce! (Iraq 5497)12. paṭṭəriṯ wə-šabbəqiṯ wə-ṯarrəḵiṯ yāṯəḵi

I hereby divorce, divorce, and divorce you! (Iraq 11113)

Most versions in the square script continue with a simile, comparing the cli-ent’s act of divorcing the lilith to the parallel institution of divorces between demons:

43 For the use of bə- here in place of the expected ʿ al, cf. Daniel 6:18: ִּדי ָלא-ִתְׁשֵנא ְצבּו ,ְּבָדִנֵּיאל di lā-ṯišne ṣəḇu bə-ḏāniyyel (‘that nothing concerning Daniel might be changed’).

44 Montgomery 1913 11, 18; Iraq 18N18, and Geller 1997 A. Additionally, Louvre AO 2629 and Segal 098M have the form ʾap̄ṭarteḵ ‘I (hereby) cause you to leave; discharge you,’ which derives from the same root as paṭṭərit.


Charles

Rectangle

386 Häberl

13. kəmā ḏə-ḵāṯḇin šedin giṭṭin wə-yāhḇin li-nšayhon wə-ṯuḇ

Just as demons write deeds of divorce, give them to their wives,45 and never

lā-ḥāḏrin ʿălayhon šəqulu giṭṭayḵon wə-qabbəlu mawmāṯḵon

swarm over them again,46 take your deeds of divorce and accept your oaths.47 (Segal 013A)

Two of the Mandaic versions, Segal 098M and 099M, dispense with this simile, at least explicitly:

14. hā eʾkkəṯiḇyān wə-həzen ʾap̄ṭartəḵen min Bā-ḏə-ros paṯ Qāqāy

Look, they are written,48 and see, I hereby make you leave Bā ḏə-Ros, daughter of Qāqāy,

kəmā ḏi-ḵṯiḇ šəmeḵ šəqul giṭṭeḵ wə-gabbil ʾumāmāṯḵen

[. . .] just as your name is written. Take your deed of divorce and receive your oaths! (Segal 098M)

15. həze ḏə-ḵṯaḇleḵ giṭṭā kəmā hāzen giṭṭā

See, he (or: they) hereby writes a deed of divorce for you like this deed [that]

45 Montgomery 1913 11 and 18, Iraq 9737, Geller 1997 A, Schøyen MS 1928/47, and Louvre AO 2629 simply have ‘just as demons write deeds of divorce for their wives’; Louvre AO 2629 also adds a typically Mandaean touch, bə-ḵušṭā (‘in truth’). Iraq 5497 has ‘just as demons (šede) and devils (diwe) write and give deeds of divorce to their wives’. This portion is missing from Iraq 11113 and Segal 099M.

46 Most versions have the expected hāḏrin ‘they return’ here, but Montgomery 1913 11, Iraq 9737, Iraq 11113, Geller 1997 A, and Segal 013A have ḥāḏrin ‘they swarm around’ instead. Iraq 5497 interpolates ʾap̄ ʾanti liliṯā bištā lili ḏiḵrā lili nəqeḇṯā wə-ḥānqəṯā u-ḇraṯā wə-š . . . (‘you are also an evil lilith, a male lilith and a female lilith, a strangler, a daughter, and . . .’).

47 So Montgomery 1913 18, Iraq 9737, Iraq 18N18, Iraq 11113, Geller 1997 A, Schøyen MS 1928/47, Segal 013A, Louvre AO 2629, and Segal 098M. Montgomery 1913 11 simply summarizes this part and the following with ‘take your deed of divorce from Nēwānduxt daughter of Kafni,’ and Iraq 5497 extends ‘deed of divorce’ to giṭṭəḵi u-sp̄ar tiruḵayḵi wə-ʾiggəraṯ

šibbuqayḵi (‘your deed of divorce, your document of divorce, and your letter of divorce’).48 Against Segal 2000: 129 I interpret this form as a Gt perfect 3rd fem. pl. form. The initial

glottal stop has become elided after hā, and the t- prefĳix has assimilated to the fĳirst root letter, as in the Gt participle mekkəṯiḇ ‘written’ and the Dt participle meqqəbar ‘buried’ from Example 15. In the text, the stem is followed by the 3rd fem. sing. sufffĳix -aṯ, and then subsequently by the (conspicuously Neo-Mandaic) 3rd fem. pl. sufffĳix -yān, as if the enchanter had originally written the singular form and then changed his or her mind. After using the 2nd fem. pl. sufffĳix for the next few verbs, s/he then returns back to the singular in the following line.


Charles

Rectangle


kəṯaḇiṯ ḏə-həḇile lə-Yoḇhānān u-l-ḵol mān də-hāzen puḡdāmā

I hereby write to destroy her49 for Yoḥānān and for all for whom this command

mekkəṯiḇleh wə-meqqəbarleh bə-ḇāḇeh

is written and buried at his gate. (Segal 099M)

Once the deed of divorce has been written and served, the enchanter follows with a series of injunctions to leave, often specifĳically enumerating the people, places, and things from which the lilith is barred:

16. wə-p̄uqu u-qḏuḥu wə-ʿiruqu wə-ʾizelu

And go out, flee, scram, and depart min bayṯah də-Ḇahrānduḵ baṯ Newānduḵ

from the house of Wahrāmduxt, daughter of Nēwānduxt. (Segal 013A)

Some versions50 also include an admonition against appearing to the client ever again in any form:

17. wə-lā-ṯiṯḥăzin ləhon lā-ḇ-ḥezwe ḏ-imāmā wə-lā-ḇ-harhure lelyā

And do not appear to them either in daytime visions or in nighttime fancies

wə-lā-ḇi-ḏmuṯ gaḇrā wə-ʾittəṯā

and not in the form of a man or woman. (Iraq 5497)

The extreme degree of variation within the sequence of the diffferent injunc-tions is strongly suggestive of oral transmission, as Table 5 below demon-strates. Note that no two of the texts preserve the same commands in the same order, and that in place of the expected ʿiruqi ‘flee,’ which is synonymous with the other commands, several versions offfer ʿiqiri ‘uproot’. Another interesting feature of this sequence of commands is the fact that the majority of versions use the feminine singular form of the imperative (e.g., qəḏuḥi ‘flee’), whereas Montgomery 1913 18 and Segal 013A use the plural form (e.g., qəḏuḥu ‘flee’).

49 Lit., ‘that she be destroyed’ həbil-e < *həḇil-ay <3rd f.sg. pass. ptc. həḇil-a +3rd f.sg. pron. hi. Cf. JB 46 li-hḇilennon ‘to destroy them’.

50 In addition to Iraq 5497, Montgomery 1913 11 has ‘do not appear to her either in the night or the daytime, do not sleep with her (lā ṯiškəḇin ʿimmah) and do not kill her sons or daughters (lā ṯiqṭəlin yāṯ bənah u-ḇnāṯah)’. Montgomery 1913 18 simply has ‘do not appear to him any more (ʿoḏ)’ and Louvre AO 2629 has ‘do not appear to them either in their nighttime dreams (bə-helmayyon d-lelyā) or their daytime visions’.


Charles

Rectangle

388 Häberl

The one text that addresses two liliths, Louvre AO 2629, uses the epicene singu-lar form of the imperative (e.g., qəḏā ‘flee’).

Just as the location and activities of the lilith represent a signifĳicant locus for variation, so too does the extent of the incantation’s coverage. If we consider the incantation to be something like an ‘insurance policy,’51 and the named clients to be the ‘policyholders,’ then the extent of the coverage beyond the initial policyholders, such as property and dependents, difffers from version to version.

table 5 Variation in sequence of injunctions

Text 1st injunction 2nd injunction 3rd injunction 4th injunction

Montgomery 1913 11 — — — —Montgomery 1913 18 qəḏuḥu ‘flee’ ʿiriqu ‘flee’ — —Iraq 5497 ʿiqiri ‘uproot’ qəḏuḥi ‘flee’ puqi ‘go out’ ʾiṯʿaqqəri ‘be

uprooted’Iraq 9737 qəḏuḥi ‘flee’ ʿiqiri ‘uproot’ puqi ‘go out’ —Iraq 11113 — — — —Iraq 18N1852 ʿiqiri ‘uproot’ qəḏuḥi ‘flee’ puqi ‘go out’ ʾiṯraḥḥəqi

‘depart’Geller 1997 A qəḏuḥi ‘flee’ zuʿi ‘move’ puqi ‘go out’ ʿiruqi ‘flee’Schøyen MS 1928/47 puqi ‘go out’ qəḏuḥi ‘flee’ pəṭuri ‘abandon’ ʿiruqi ‘flee’Segal 013A puqu ‘go out’ qəḏuḥu ‘flee’ ʿiruqu ‘flee’ ʾizilu ‘go (pl.)’Louvre AO 2629 qəḏā ‘flee’ puq ‘go out’ ʾəruq ‘flee’ eʾṯrahhaq

‘depart’Segal 098M qəḏā ‘flee’ ʾəqir ‘uproot’ puq ‘go out’ eʾṯrahhaq

‘depart’Segal 099M ʾəqar ‘uproot’ puq ‘go out’ zəhā ‘flee’ eʾṯrahhaq

‘depart’

51 A metaphor fĳirst suggested to me by Erica C.D. Hunter (personal communication).52 Iraq 18N18 has ʿiqiri mādorəḵi wə-šannu malḵuṯəḵi qəḏuḥi ʾiṯbəʾiri p̄uqi wə-ʾiṯraḥḥəqi

(‘uproot your dwelling, change your kingdom, flee, dispose of yourself (Gt *b-ʿ-r), go out, and depart’).


Charles

Rectangle


IV. Voces Magicae

Several variants of this text incorporate some unintelligible voces magicae53 introduced by the phrase bə šum in the penultimate formulae of six variants of the incantation:

table 6 Variation in Voces Magicae

Text Vox Magica

Montgomery 1913 11 MMYNTŠ ŠMR DY HBNZYG YW YDYD YṬ YṬ YṬ

Montgomery 1913 18 MTHMN YʾBNBYG DYD ṬDṬ

Iraq 5497 —Iraq 9737 RT MHṢ MHṢ MHṢ

Iraq 11113 —Iraq 18N18 —Geller 1997 A —Schøyen MS 1928/47 —Segal 013A HYGYH DYH TṬM

Louvre AO 2629 —Segal 098M TMAHṢ ABD HYGUT AID AUD ṬṬUZ

Segal 099M MATHIMIṢ

The phrase bə-šum is usually translated literally as ‘in the name of . . .’, parallel in some respects to the legal formula ‘by the authority granted to/vested in me by . . .’. This would logically require that the following refers to some being or beings, even though this interpretation is not always tenable, as the examples above indicate. Segal notes that this same phrase is used to introduce entire formulae in two other bowls from the British Museum:

18. bə-šum kəḇišin dišin dariḵin

In the name of: ‘Suppressed, trampled, and trodden . . .’. (Segal 043A: 3–4)

53 Voces magicae are enigmatic phrases lacking recognizable semantic content, at least to the enchanter and his or her intended audience; see Bohak 2008: 258–64.


Charles

Rectangle

390 Häberl

19. bə-šum ʾiṭṭer yamin ḥayyāḇā bayṯā min šəmayyā

In the name of: ‘The left-handed is condemned, the house from heaven [is] . . .’. (Segal 044A: 10)

20. bə-šum haškəḥit məšoḇədāyṯ (<*məšaʿbədāʾiṯ) In the name of: ‘Submissively, I have found/attained . . .’. (Segal 044A: 11)

These strangely truncated examples resemble the initial lines of incantation formulae, which (as noted above), can stand in place of the entire composition. This suggests that bə-šum might be employed to introduce quotations in some contexts, and perhaps even direct speech, as in the case of the voces magicae. It is entirely possible that the text which follows might be meaningful in some language other than Aramaic, though not necessarily to the magician or his intended audience,54 thus to some extent vindicating Macuch’s observation.

V. Ending

The Ending ‘seals the deal,’ so to speak. Seven of the versions (including two of the Mandaic ones) make this metaphor explicit by employing the rhetori-cal device of a ‘signet ring upon which is engraved the great inefffable name’ of God, a device clearly borrowed from the lore of Judaism. This section contains more explicitly Jewish content than any of the other sections, and perhaps not surprisingly this content does not always survive intact in every attested instantiation of the text.

21. ḥăṯimṯā bə-ʿizqəṯeh də-Tyqs gibbārā u-ḇ-ʿizqəṯeh

Sealed with the signet ring of Tyqs the Hero and with the signet ring of də-ṣurgəli ḏə-ʿălohi šem məp̄oraš rabbā

Ṣurgəli (“Drawngrave”) upon which is the great inefffable name. (Segal 013A)55

54 Versnel (2002: 137) gives an example of a very common medieval charm against worms, sisim hemma mulahos usmonim pilagrim velamos einmisspar, which was revealed in the 19th century to be a Latinized version of Song of Solomon 6:8 (‘There are threescore queens, and fourscore concubines, and virgins without number’).

55 This section is entirely missing from Iraq 11113, Iraq 18N18, and Segal 099M. Montgomery 1913 11 simply has ‘with the signet ring upon which is drawn (ṣir) and engraved (u-ḡlip̄) the inefffable name’; Montgomery 1913 18 preserves only the last four letters of ‘inefffable’; Iraq 5497 has ‘sealed upon your deed of divorce’; Iraq 9737 includes only the phrase ‘inefffable name’ immediately after the voces magicae; Geller 1997 A has: ‘this . . . that was sealed [ʾiṯḥattəmaṯ] in the name of Ṣurgəlip̄ upon which’, whereupon the enchanter switches


Charles

Rectangle


The syntax and vocalization of the last phrase (bounded by the letter H in superscript) indicates that it is an obvious loan from Hebrew. The Aramaic cognate of məp̄oraš is məp̄āršā ‘separated,’ in which the vowel ā of the second syllable has not undergone the shift to o that is typical of ‘Canaanite’ languages such as Hebrew and Phoenician, an important isogloss distinguishing these languages from Aramaic. The presence of this phrase in most versions of the text, including two of the Mandaic ones, lends credence to the thesis that it originated in Jewish circles. Five of the attested versions continue where Segal 013A leaves offf, adding the following phrase, presumably in reference to the inefffable name:

22. siddur ʿālmā mi-ššešeṯ yəme ḇəreʾšiṯ

the natural order since the six days of creation. (Schøyen MS 1928/47)56

Bəreʾšiṯ (lit., ‘in the beginning’) is, of course, the phrase which begins the fĳirst of the fĳive books of the Pentateuch. The title of this work has become a byword for the creation in the Jewish tradition.

The content of this section, which is rather specifĳic to Judaism and even partially composed in a separate language, obviously did not lend itself well to transmission. One of the Mandaic versions, Segal 098M, illustrates some of the difffĳiculties the enchanter faced when confronted with phrases like gəlip̄

ʿălohi ‘engraved upon it,’ which has become Gəlip̄ ʾalāhā ‘the god Gəlip̄,’ and šem məp̄oraš ‘the inefffable name,’ which has here become šəmā ḏə-p̄ārušā ‘name of the discerner’.

23. hā ʾəsirā u-hṯimā ḏə-Ḇā-ḏə-ros paṯ Qāqāy

Look, the (geṭ?) of Bā ḏə-Ros, daughter of Qāqāy, is bound and sealed bə- eʾzqəṯon ḏi-ḡlip̄ ʾalāhā bə-šəmā ḏə-p̄ārušā

with the signet ring of the god Gəlip̄ (‘Engraved’) in the name of the dis-cerner

ḏə-ḇeh ʾəmareh u-ḇ- eʾšmeh kulləhon ʾalāme

with which he said it and in whose name are all things. (Segal 098M)

Louvre AO 2629 introduces another element of Jewish origin, the Ring of Solomon, which is surprisingly not found in any other version, Mandaic or otherwise:

to ‘with the signet-ring upon which is drawn the name . . .’, and Schøyen MS 1928/47 has ‘bound and sealed by the signet-ring on which the inefffable name is engraved and incised’.

56 Montgomery 1913 11, Montgomery 1913 18, Iraq 9737, Schøyen MS 1928/47, and Louvre AO 2629 contain variations on the phrase ‘since the six days of creation’.


Charles

Rectangle

392 Häberl

24. ʾəsirāṯe u-həṯimāṯe lilyāṯā ziḵre wə-nuqbāṯā bə- eʾzqəṯeh

Liliths both male and female are bound and sealed57 with the signet ring of

di-Šlemon malkā bar Dāwiḏ də-ṣirgəlip̄ ʾalāhā šəmeh rabbā wə-yaqqirā

King Solomon son of David, of the god Ṣirgəlip̄, his name is great and honorable. (Louvre AO 2629)

Serendipitously, this very same ring makes an appearance in that section of the Babylonian Talmud dedicated explicitly to divorce documents, Tractate Giṭṭin. The seventh chapter of this tractate contains an aggadah in which Solomon sends his trusted offfĳicial Benayahu bar Yehohada to entrap the demon Asmodeus with a ring on which was engraved the name of God (da-ḥqiq

ʿălah šem), and with a chain that was likewise engraved (BT Gittin 68a–b). It is interesting to note that the somewhat enigmatic phrase which appears in Montgomery 1913 11 as ‘drawn and engraved upon it’ (ṣir u-ḡlip̄ ʿălohi) has clearly been reanalysed as a name in several versions of this text, and in the two Mandaic versions cited above has even been deifĳied, thus making a mock-ery of the inefffable name (Montgomery 1913: 179).58

Perhaps not unexpectedly, the Hebrew phrases at the very end have become garbled in the version published by Lidzbarski:

25. gəlip̄ ʾalāhā šəmeh məp̄āršā

The god Gəlip̄, his name is held separate mirešešaṯ šeṯ yawme bərešiṯ

from the beginning, the six days of creation. (Louvre AO 2629)

Mirešešaṯ appears to be a blend of the Hebrew words mi- ‘from’, reʾšiṯ ‘begin-ning’ and šeš ‘six’ or possibly the Jewish Babylonian Aramaic middə- ‘from (the fact that)’ and Hebrew šḗšeṯ ‘six’.59 Having made a hash of the phrase, the enchanter then attempted to convey its sense in Mandaic.

57 Both passive participles are marked with both masculine and feminine plural sufffĳixes, which is ungrammatical but certainly in keeping with the content of the text.

58 Note that the initial words of our incantation ʾal-ʾissur may also have been interpreted as a proper name by some enchanters.

59 I owe the latter suggestion to an anonymous reviewer, who posited a scenario in which a Jewish mediator misread the dalet for a resh and transmitted it thus to the Mandaean practitioner. As these two characters are not infrequently conflated in writing, this would appear to be a point in favour of a textual rather than oral transmission. Although I


Charles

Rectangle


VI. Concluding Formula

26. ʾāmen ʾāmen ʾāmen selā

Amen, amen, amen, selah. (Segal 013A)

This is one of the more popular concluding formulae to be found in the incan-tations. ʾĀmen ‘truly’ typically concludes doxologies and other hymns. The word selā ‘selah’ is particularly associated with the Psalms, where it concludes certain psalms or divides one strophe from the next, and appears to serve as a choral direction to the worshippers (Snaith 1952). It is unlikely that it contin-ued to serve that purpose at the time this text was composed, but it nonethe-less served to indicate the end of the composition in this and other bowls.

Having compared all of the versions of this incantation currently available to scholarship, isolated the variants, and identifĳied a few possible factors con-ditioning the selection of these variants, I can now tender a few tentative con-clusions about the production of each version of this incantation. While it is clear that each follows an established model and maintains a certain unity, it is equally clear that each performance of the text also offfered ample opportuni-ties for improvisation. The most obvious opportunities are offfered by the selec-tion of the framing formulae, the name of the protective genius, and the voces

magicae introduced by bə-šum, which clearly difffer from version to version. Less immediately obvious are the multiple nodes of variation contained within the text, such as the description of the lilith in the Recitals, the sequence and number of injunctions against her, and the sequence and number of people, places, and objects from which the lilith is restrained. These should be dis-tinguished from the variation caused by miscommunication in the transmis-sion of the incantation, such as the way in which various technical terms are rendered, or the various ways in which the Hebrew text and other specifĳically Jewish content in the Ending have been garbled. The latter are clearly mistakes of the sort noted by Pognon and his successors when they discuss how one ver-sion might be written ‘more correctly’ than another, whereas the former repre-sent the individual initiative of each enchanter in performing the incantation. Both nonetheless represent the unmistakable signatures of oral transmission.

have my doubts that the conjunction middə- is appropriate for this context, I would like to suggest that the intervocalic environment of the d could result in it surfacing as an alveolar flap /ɾ/, a common enough phenomenon across many languages, e.g. Neapolitan madonna [maɾonnə] ‘the Virgin Mary,’ or (North American) English riding [ɹaɪɾɪŋ], with-out assuming a textual intermediary.


Charles

Rectangle

394 Häberl

The Reception of the Incantation

Our analysis should not conclude with the production of the text. One of Ong’s personal contributions to the study of oral compositions was a major shift of focus from their production, which had been the primary concern of previous scholars, to their reception, for the simple reason that oral compositions are necessarily addressed directly to a specifĳic audience, which explains many of the distinctive characteristics that separate them from written compositions. For example, Ong notes that oral compositions more commonly use coordi-nating conjunctions than subordinating conjunctions. He explains this phe-nomenon by noting that written compositions lack the full context of the oral utterance, including non-verbal ‘paralinguistic’ cues such as intonation and body language, and are therefore much more dependent upon linguistic struc-ture and a much more elaborate syntax to impart meaning (Ong 2002: 37–38). To eyes accustomed to this elaborate syntax, such as ours, the repeated use of the coordinating conjunction wə- in these texts must seem excessive and simplistic, but in the context of an oral recitation, elaborate syntax is not only unnecessary but might even hamper comprehension. Within the context of medieval Persian epic prose narratives, Rubanovich (2012: 668) notes coordi-nating conjunctions such as these perform an important pragmatic function in structuring the content of the discourse, and facilitate the audience’s aural reception of the text by punctuating shifts in the story line.

Additionally, oral expressions tend to rely more heavily upon parallel terms, phrases, or even entire clauses, as well as epithets. Ong (2002: 38–39) relates this tendency to the oral reliance upon formulae to implement memory. To the literate eye, these appear to be cumbersome, tiresomely redundant, and in a word ‘clichéd’. Oral compositions prefer not merely the son, but ‘the mascu-line son,’ not merely the daughter, but ‘the feminine daughter,’ and not merely the animals, but ‘the animals with the cloven hoof and the animals without cloven feet’. The angels and demons in these texts are generally identifĳied as such, given names like Buḡdānā the king, Ḥabsalas the lilith, Tiqos the hero, and Gəlip̄ the god.

Oral compositions tend to be redundant. Ong explains this redundancy by noting that it serves several purposes – it helps to preserve continuity when there is no model to follow such as a written text, by forcing the speaker to consider simultaneously what he is saying, what he has already said, and what follows in the series, keeping him on track, and the repetition of information allows larger audiences to keep track of what is being said even when they have difffĳiculty hearing all of the information (Ong 2002: 39–41).

Oral compositions also tend to be somewhat conservative vis-à-vis writ-ten compositions. As knowledge in an oral culture tends to disappear unless


Charles

Rectangle


it is frequently repeated, oral cultures must invest great energy in repeating knowledge that has been acquired over and over again. This need encour-ages and even demands a highly traditionalist or conservative mindset that shuns intellectual experimentation, and for good reason (Ong 2002: 41–42). I do not intend to suggest that oral cultures are bereft of creativity or originality, but merely that this creativity consists chiefly of variations upon established themes, as in the example of the incantation above.60

Finally, knowledge is rarely presented in the abstract, but almost always in reference to situations derived from the human experience. As a corollary, oral compositions are always directed towards a particular audience, as it would be pointless to deliver them while alone. In fact, one of the most common situ-ations that occasion an oral composition is that of a competition. Poems and proverbs are used not merely for storing knowledge, but also to engage others in verbal and intellectual combat. Likewise, conflict is one of the most com-mon topics of oral compositions – including, for obvious reasons, the incanta-tion texts.

Conclusion

In the article that opens this volume, Karl Reichl observed that there is a kind of continuum between orality and textuality, particularly with reference to oral compositions for which we possess only written attestations. The bowl incantations are comfortably situated within this continuum. They are obvi-ously not entirely oral, given the fact that they survive only in written form, and furthermore contain rather explicit references to the act of writing and other characteristics of a literate culture. Nevertheless, they are not entirely textual either, given the degree of variation between each attested version and the explicit opportunities for improvisation within the text.

The surviving versions of our incantation and the conditioned variation between each suggest that they were composed and transmitted orally, and only secondarily written down. Internal evidence, such as the use of illocution-ary acts within the text, indicate that they were the products of a performative ritual rather than the scholastic and mechanical reproduction of a manuscript, and that they were the instantiations not of a written Urtext but of a fluid oral tradition. Therein lies the answer to the puzzle posed by the poor quality of

60 Hunter (2002: 273) notes: ‘[Refrain A] shows the degree of improvisation that took place during the transcription of incantations. Within the conventions of the genre, the writ-ing process was flexible and innovative, and the copyists expressed their individuality accordingly’.


Charles

Rectangle

396 Häberl

the writing (not to mention the large number of bowls covered not with actual writing but with a crude imitation or ‘pseudo-script’).

In a primarily oral milieu such as that of the Middle East in antiquity, compositions of this kind were likely composed not to be recited alone but rather in the presence of others, as a variety of performance. Returning to the Generalissimo Stalin, it is clear that the other components of this perfor-mance – the whispered incantations, the construction of a magical circle, the inscribing of a new bowl, and fĳinally the burial of that bowl – would have to be done properly in order to satisfy this audience of the felicity and efffĳicacy of the ritual. The writing which was a reflection of the oral composition, ‘trapped’ as it were within the bounds of the circle formed by the rim of the bowl, was quite possibly of secondary importance, and the circumstances of its execu-tion (on an unglazed plain-ware bowl during the course of a ritual) certainly did not lend themselves to great feats of calligraphy, in any case. Given these circumstances, and the fact that writing is typically much slower than speech (being typically about one tenth its speed), it is not surprising that some magi-cians may have merely mimicked the act of writing during the course of the ritual, thereby producing the so-called ‘pseudo-script texts’. We therefore need not dismiss their creators as charlatans, as they would also need to be completely versed in the other components of the ritual in order to satisfy their clients.

In addition to illuminating much about the thought world of the people of late antique Mesopotamia, these bowl texts also have great potential to teach us much about the language they spoke, provided that they are not confused or conflated with the standard versions of this language represented by other lit-erary traditions. In fact, as hastily scrawled transcriptions of the actual speech of the magician, presumably recorded more or less verbatim, they offfer us insight into aspects of the spoken language that are not provided by the more prestigious literary texts.

Bibliography

Austin, John L. 1962. How to Do Things with Words. The William James Lectures delivered

at Harvard University in 1955. Cambridge, Masssachusetts: Harvard UP.———. 1979. Philosophical Papers. Ed. James O. Urmson and Geofffrey J. Warnock.

Oxford: Oxford UP.Bohak, Gideon. 2008. Ancient Jewish Magic: A History. Cambridge: Cambridge UP.Drower, Ethel S. 2000. The Mandaeans of Iraq and Iran. 2nd rpt. ed. Piscataway, New

Jersey: Gorgias.


Charles

Rectangle


———, and Rudolf Macuch. 1963. A Mandaic Dictionary. London: Oxford UP.Dundes, Alan. 1999. Holy Writ as Oral Lit: The Bible as Folklore. Lanham, MD: Rowman

and Littlefĳield.Geller, M.J. 1997. ‘More magic spells and formulae.’ BSOAS 60, pp. 327–35.Gordon, Cyrus H. 1934a. ‘Aramaic Magical Bowls in the Istanbul and Baghdad Museums.’

Archiv Orientální 6/2, pp. 319–34.———. 1934b. ‘An Aramaic Incantation.’ The Annual of the American Schools of Oriental

Research 14, pp. 141–44.———. 1937. ‘Aramaic and Mandaic Magical Bowls.’ Archiv Orientální 9, pp. 84–106.———. 1941. ‘Aramaic Incantation Bowls.’ Orientalia 10, pp. 116–41, 272–84, 339–60.Graf, Fritz. 1997. Magic in the Ancient World. Trans. Franklin Philip. Cambridge,

Massachusetts: Harvard UP.Greenfĳield, Jonas C., and Joseph Naveh. 1985. ‘A Mandaic Lead Amulet with Four

Incantations.’ Eretz Israel 17, pp. 97–107. [in Hebrew]Harviainen, Tapani. 1995. ‘Pagan Incantations in Aramaic Magic Bowls.’ In Studia

Aramaica: New Sources and New Approaches. Papers Delivered at the London

Conference of the Institute of Jewish Studies, University College London, 26th–28th

June 1991. Ed. Markham J. Geller, Jonas C. Greenfĳield and Michael P. Weitzman. Journal of Semitic Studies Supplement 4. Oxford: Oxford UP, pp. 53–60.

Herzog, George. 1965. ‘Stability of Form in Traditional and Cultivated Music.’ In The

Study of Folklore. Ed. Alan Dundes. Englewood Clifffs, New Jersey: Prentice-Hall, pp. 169–74.

Hunter, Erica C.D. 1995. ‘Combat and Conflict in Incantation Bowls: Studies On Two Aramaic Specimens from Nippur.’ In Studia Aramaica: New Sources and New

Approaches. Papers Delivered at the London Conference of the Institute of Jewish

Studies, University College London, 26th–28th June 1991. Ed. Markham J. Geller, Jonas C. Greenfĳield and Michael P. Weitzman. Journal of Semitic Studies Supplement 4. Oxford: Oxford UP, pp. 61–75.

———. 2002. ‘Manipulating Incantation Texts: Excursions in Refrain A.’ Iraq 64, pp. 259–73.

Layard, Austen Henry. 1853. Discoveries in the Ruins of Nineveh and Babylon. New York.Lidzbarski, Mark. 1902. Ephemeris für semitische Epigraphik. Giessen. Vol. 1.Macuch, Rudolf. 1965. Handbook of Classical and Modern Mandaic. Berlin: Walter de

Gruyter.Montgomery, James A. 1913. Aramaic Incantation Texts from Nippur. Philadelphia: U of

Pennsylvania, University Museum.Müller-Kessler, Christa. 1996. ‘The Story of Bguzan-Lilit, Daughter of Zanay-Lilit.’ JAOS

116/2, pp. 185–95.———. 1998. ‘Aramäische Koine – Ein Beschwörungsformular aus Mesopotamien.’

Baghdader Mitteilungen 29, pp. 331–48, pls. 1–2.


Charles

Rectangle

398 Häberl

———. 1999a. ‘Puzzling words and spellings in Babylonian Aramaic magic bowls.’ BSOAS 62/1, pp. 111–14.

———. 1999b. ‘Interrelations between Mandaic Lead Rolls and Incantation Bowls.’ In Mesopotamian Magic: textual, historical, and interpretative perspectives. Ed. Tzvi Abusch and Karel van der Toorn. Ancient Magic and Divination 1. Groningen: Styx Publications, pp. 197–209.

———, and Theodore Kwasman. 2000. ‘A Unique Talmudic Aramaic Incantation Bowl.’ JAOS 120/2, pp. 159–65.

Naveh, Joseph, and Shaul Shaked. 1993. Magic Spells and Formulae: Aramaic Incantations

of Late Antiquity. Jerusalem: The Magnes Press, The Hebrew University.———. 1998. Amulets and Magic Bowls: Aramaic Incantations of Late Antiquity. 3d ed.

Jerusalem: The Magnes Press, The Hebrew University.Ong, Walter J. 2002. Orality and Literacy: The Technologizing of the Word. 2nd ed.

London and New York: Routledge.Pognon, Henri. 1898. Inscriptions mandaïtes des coupes de Khouabir. Texte, traduction et

commentaire philologique. Paris. 3 parts in one volume.Propp, Vladimir. 1968. Morphology of the Folktale. Trans. Laurence Scott. Austin: U of

Texas Press.Reynolds, Dwight F. 2007. Arab Folklore: a handbook. Greenwood Folklore Handbooks.

Westport, Connecticut and London: Greenwood Press.Rossell, William H. 1953. A Handbook of Aramaic Magical Texts. Shelton Semitic Series

2. Skylands, New Jersey: Department of Semitics of Shelton College.Rubanovich, Julia. 2012. ‘Orality in Medieval Persian Literature.’ In Medieval Oral

Literature. Ed. Karl Reichl. De Gruyter Lexikon. Berlin and New York: de Gruyter, pp. 653–79.

Segal, Judah B. 2000. Catalogue of the Aramaic and Mandaic Incantation Bowls in the

British Museum. With a contribution by Erica C.D. Hunter. London: British Museum Press.

Shaked, Shaul. 1985. ‘Bagdāna, King of the Demons, and other Iranian terms in Babylonian Aramaic magic.’ In Acta Iranica 25. Papers in Honour of Professor Mary

Boyce. Ed. H.W. Bailey, A.D.H. Bivar, Jacques Duchesne-Guillemin and John R. Hinnells. Leiden: E.J. Brill, pp. 511–25.

———. 1999. ‘The Poetics of Spells. Language and Structure in Aramaic Incantations of Late Antiquity. 1: The Divorce Formula and its Ramifĳications.’ In Mesopotamian

Magic: textual, historical, and interpretative perspectives. Ancient Magic and Divination 1. Ed. Tzvi Abusch and Karel van der Toorn. Groningen: Styx Publications, pp. 173–95.

Snaith, Norman H. 1952. ‘Selah.’ Vetus Testamentum 2, pp. 43–56.Tambiah, Stanley J. 1968. ‘The Magical Power of Words.’ MAN: The Journal of the Royal

Anthropological Institute 3/2, pp. 175–208.


Charles

Rectangle


Thompson, Stith. 1955–58. Motif-Index of Folk Literature. A Classifĳication of Narrative

Elements in Folktales, Ballads, Myths, Fables, Mediaeval Romances, Exempla,

Fabliaux, Jest – Books and Local Legends. Bloomington: Indiana UP. 6 vols.Versnel, Henk S. 2002. ‘The Poetics of the Magical Charm: An Essay on the Power

of Words.’ In Magic and Ritual in the Ancient World. Ed. Paul Mirecki and Marvin Meyer. Religions in the Graeco-Roman World 141. Leiden: E.J. Brill, pp. 105–58.

Yamauchi, Edwin M. 2005. Mandaic Incantation Texts. Rev. ed. Piscataway, New Jersey: Gorgias Press.


Charles

Rectangle

Documents

Aramaic Incantation Texts between Orality and Textuality