Neo-Mandaic in Mandaean Manuscript Sources

NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 1 MATTHEW MORGENSTERN 1. INTRODUCTION Mandaic is a south-eastern variety of Aramaic that is closely related to the Jewish Babylonian Aramaic of the Talmud and post-Talmudic rabbinic literature. It is essentially a communal dialect, in that all attested sources demonstrating uniquely Mandaic features are written in the script that is employed for Mandaean religious documents.2 Furthermore, the overwhelming majority of the texts written in this lan1 This article is based upon a lecture presented at the conference Neo-Aramaic Dialectology: Jews, Christians, and Mandaeans, held at the Institute of Advanced Studies, Jerusalem on 26–27 June 2013. I wish to thank the organizers of the conference, Professors Simon Hopkins, Steven Fassberg and Hezy Mutzafi for inviting me to speak, and Professor Geoffrey Khan for accepting the paper in its present form for publication. I also wish to thank Dr. Tania Notarius, Maleen Schlüter, Elisheva Bard, Tom Alfia and Livnat Barkan for their assistance in preparing the materials discussed herein. Professors Shaul Shaked, Hezy Mutzafi, Charles Häberl and James Nathan Ford kindly shared with me their unpublished works. Citations from the Rbai Rafid Collection are reproduced by kind permission of the custodian of the collection. Mandaic written sources are provided in Mandaic script followed by a Latin transliteration in round brackets, e.g. #‫ר‬#‫( חיו‬hiuara) ‘white (m.sg.)’. The transliteration system follows that employed Rudolf Macuch, except for ‫ ע‬and ! (the Arabic pharyngeal ‫ ﻉ‬employed exclusively in loanwords and proper nouns), which are represented respectively by (ʿ) and (ʕ). Neo-Mandaic words from spoken sources are presented in italics, e.g. həwɔrɔ ‘white’ according to a standardized transcription, while for the sake of simplicity, in the references preference has been given to the glossaries that accompany these texts. This research was supported by the Israel Science Foundation grant no. 419/13. 2 Exceptions are citations of Mandaean literature found in Syriac, and some later manuscripts with accompanying transcriptions into Arabic letters to aid the study of the texts or to 368 MATTHEW MORGENSTERN guage contain elements that reflect the Mandaean religion. This includes the evergrowing corpus of spells written on clay bowls and lead scrolls, which according to most estimations were written between the fourth to seventh centuries CE3 and provide the earliest material evidence for the language, even though the original date of the compositions contained in both this epigraphic corpus and the later manuscripts may be earlier and extend as far back as the second century CE. Most Mandaean texts are not preserved in early epigraphic sources but rather in much later manuscripts. The earliest Mandaean manuscript known to scholarship remains the Bodleian Library’s codex Marsh. 691, a small selection of prayers copied in Ḥuweiza in 936 AH (1529–1530 CE). Very few manuscripts survive from the 16 th century, and several works survive only in manuscripts that were copied as late as the 20th century. The lack of primary textual sources over a period of some 900 years presents a considerable impediment to the diachronic study of Mandaic. Yet while the fact that the available textual witnesses were copied so long after the redaction of the works that they contain is very frustrating for the study of pre-modern Mandaic, thanks to the researches of Jorunn Buckley, it has become clear that even late manuscripts may teach us much about the more recent history of the Mandaean community.4 During the course of our research projects related to the preparation of a new dictionary of the Mandaic language,5 considerable efforts have been made to expand the corpus. Thanks to the collaboration of several scholars and members of the Mandaean community around the world, these labours have proven successful, and we are now in possession of twice as many written sources than were previously available.6 These cast significant light not only on the textual transmission and the contents of Mandaean literature, enabling us to make many improvements to the interpretation of the Mandaean lexicon (the primary aim of the research projects), but also cast considerable light on the pre-history of contemporary Neo-Mandaic (NM). render them readable to lay Mandaeans who need to recite them as part of the ritual. 3 On the archaeological evidence see, e.g., Hunter (1995). 4 See in particular Buckley (2010). 5 Two research projects have been funded by the Israel Science Foundation: ‘Materials for a New Dictionary of Mandaic’ (2010–2014) and ‘The Lexical Analysis of Mandaic’ (2013– 2016). 6 The largest single contribution has been the acquisition of digital images of the Rbai Rafid Collection. A full account of the current state of evidence for Mandaean literary works is presented in Morgenstern (in preparation). NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 369 As we shall see in what follows, the discovery of NM as a linguistic entity worthy of philological study was a gradual process. Macuch’s claims to have been the first scholar to discover that Mandaic was still spoken have now been conclusively disproven by Häberl,7 but his contributions to the study of NM were for many years the only significant source of information about the idiom. The recent publications by Häberl and Mutzafi represent valiant efforts to salvage from aging informants what remains of this critically endangered language, as well as to submit Macuch’s findings and conclusions to closer scrutiny.8 The current paper aims to provide an outline of the sources available to us for understanding the development of NM prior to the pioneering fieldwork of Rudolf Macuch in the 1950s. While some of these have been mentioned previously in the literature,9 others have not been given sufficient attention. Furthermore, our recent manuscript discoveries provide valuable new evidence for the earlier stages of NM. Within the scope of this article, it is not possible to discuss all these new findings in detail, and these will be presented in separate studies that consider the different types of sources in greater depth. 2. EINE SPRACHE OHNE FORSCHUNGSGESCHICHTE In describing the state of scholarship in NM in his groundbreaking (if flawed) chrestomathy of NM texts, Rudolf Macuch observed: Obwohl man schon seit den ersten europäischen Kontaken mit den Mandäern durch katholische Missionare des 17. Jh. und später durch Reisende und Orientalisten, die zu diesem Völkchen durchgedrungen waren, wußte, daß die Mandäer ein der Sprache ihrer heiligen Bücher verwandtes Idiom sprachen, hat das gesprochene Mandäisch im Gegensatz zu den neusyrischen und westaramäischen Dialekten keine Forschungsgeschichte.10 Macuch ascribed this lack of recognition to the Mandaeans’ own suspicion of outsiders. However, it seems that Macuch’s statement is both an exaggeration and, at the same time, overlooks an important source for scholarly scepticism regarding NM.11 7 Häberl (2009: 25–26), and see further below. Häberl (2009, 2010; 2013); Mutzafi (2014) and Mutzafi (forthcoming). 9 See in particular Macuch (1989: 4–5): ‘Ältere und neuere Quellen der mandäischen Volkssprache’. 10 Macuch (1989: 5). 11 For an account of the exposure of western scholars to Mandaic prior to Macuch’s studies 8 370 MATTHEW MORGENSTERN Theodore Nöldeke’s seminal Mandäische Grammatik of 1875 provided what remains the most important description of Classical Mandaic (CM) as it is found in manuscripts. (The epigraphic corpus was unknown at that time.) In summarizing the nature of his sources, Nöldeke ascribed a particularly late date to two of them: Die letzte Form der Sprache zeigen endlich die jüngsten Theile des Asfar Malwâše und die Berichte der Abschreiber über ihre Zeit (16. Jahrh. bis zur Gegenwart). Hätten wir in diesen Stücken wirklich einen modernen lebenden Dialect, so wären sie von grosser Wichtigkeit; aber sie bieten uns nur ein unerquickliches Gemisch von Formen der alten Sprache, welche man noch immer zu schreiben meint, und ganz jungen. Nicht bloss der Wortschatz, sondern auch die Grammatik ist von arabischen und persischen Elementen durchdrungen. Man sagt z. B. ‫„ ראבתאר‬grösser“ mit dem persischen Suffix tar, und gebraucht im aramäischen Text arabische Formen wie ‫يظ ر = יידהאר‬. Ein Studium der lebenden Sprache, welche den Texten zu Grunde liegt, wäre natürlich von Interesse, aber diese wird hier eben wegen der Rücksicht, die man auf die alte Sprache und vielleicht auch auf fremde Schriftsprachen nimmt, durchaus nicht treu ausgedrückt.12 Nöldeke’s opinion was similarly negative regarding the 17th century Glossarium, now known as the Leiden Glossarium but at that time in Amsterdam: In diesen letzten Zeitraum fällt auch das von einem katholischen Missionär mit Hülfe eines Mandäers verfasste arabisch-mandäisch-lateinisch-persischtürkische Glossar, welches in einem Amsterdamer Codex enthalten ist. Dasselbe erweist sich bei äusserst behutsamem Gebrauch nützlich, kann aber den Unkundigen leicht sehr stark irre führen. Der Verfasser selbst hat sich oft genug geirrt und hatte keine Kenntniss von der Literatur.13 This is not to say that Nöldeke regarded Mandaic as a dead language, as he clarified shortly thereafter: Gern hätte ich die Entwicklung des Mandäischen bis auf unsere Zeit dargestellt, aber wir haben eben nur für die ältere Periode zuverlässige Quellen, da ja, wie gesagt, die jüngeren Schriften keineswegs die Sprache ihrer Gegenwart rein darstellen.14 see Häberl (2009: 16–26) and Mutzafi (2014: 5–9). 12 Nöldeke (1875: XXIV–XXV). 13 Nöldeke (1875: XXV). 14 Nöldeke (1875: XXV). NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 371 We must recall that Nöldeke had no direct contact with the Mandaeans, and was unable to determine the true nature of the spoken Mandaean language. By contrast, Ethel Stefana Drower, the most important western student of Mandaean culture,15 encountered the spoken Mandaic of Iran in the 1930s when she visited Khorramshahr with Sheikh Abdallah Khaffagi.16 Drower also recorded a version of the popular Mandaean legend The Bridge of Šuštar in a Latin transcription and partial translation. 17 This experience with spoken Mandaic seems to have influenced Drower’s view of the language of the Book of the Zodiac, as she writes: I venture to think that Noldeke (sic) is mistaken, and that the language is not artificially archaic, but represents a transitional period. In the later fragments, in which Arabic and Persian elements are, as he says, very evident, we get something very near the spoken Mandaean of today, hence, philologically, it is of importance.18 As we shall see below, both Drower and Nöldeke were correct to a degree. While the literary idiom of almost all late Mandaic texts is to a degree archaizing 15 It is interesting to note that the most accomplished translator of Mandaean texts, Mark Lidzbarski, whose publications provided the foundation for all subsequent research of Mandaean literature, had almost nothing to say about late Mandaic, beyond noting that Arabic loanwords indicate that some of the Mandaean wedding poems, drawn from folk songs are late. See Lidzbarski (1920: XI ft. 2), and the discussion of NM poetry below. 16 Buckley (2010: 115). An account of Drower’s visit, and her impressions upon seeing Mandaean children conversing in the Mandaic language, is provided by Nasser Sobbi in Häberl (2009: 276–277). 17 The text has been published with a translation and notes in Häberl (2013). Examination of Drower’s notes, kindly made available to me by Häberl, reveals a fair level of competence on Drower’s behalf, and a good ability to relate the words in her phonetic transcription to the related lexemes in literary Mandaic, Arabic and Persian. All of this belies Macuch’s claim (Macuch 1965: xlvi) that “The existence of a vernacular dialect spoken by Mandaean laymen in Khuzistan, which as a living language deserves, at least, as much attention as the traditional pronunciation of the literary tongue, remained completely unknown until my personal discovery at the occasion of my visit to the Mandaean community of Ahw̄z in 1953”, as Häberl (2009: 25–26) justifiably noted. 18 Drower (1949: 2). Drower’s recognition of Neo-Mandaic as a language was tempered by a distinct distaste for it. Even in her last published statement on the matter, she described it as “a modernized and debased form of the tongue, incorporating a number of Persian and Arabic words” (Drower 1960: 1). 372 MATTHEW MORGENSTERN and influenced by the literary language of earlier Mandaean works, the late texts are replete with neologisms and expressions drawn from the vernacular. 3. WHAT IS NEO-MANDAIC? As we mentioned, there are no primary textual witnesses for the Mandaic language between the inscribed bowls and lead scrolls of the late Sassanid and early Islamic period and the manuscripts of the 16 th century. While it is certain that texts were being copied and composed during this period,19 they are only known today from later exemplars. The precise dating of almost all works of Mandaic literature is impossible, but it seems likely that the Ginza Rba was redacted in something close to its present form in the early Islamic period. However, Ginza Rba as it survives is unlikely to be linguistically identical to the work as it was redacted. Even a superficial comparison of the epigraphic texts with the earliest surviving manuscripts reveals differences, most notably in the orthography; while in the epigraphic texts, the use of matres lectionis to mark medial a/ā vowels (and to a lesser extent the other full vowels) is optional, in the Mandaic manuscripts it is obligatory. This implies that the copies of early works now in our hands have undergone some linguistic change.20 Owing to the large gap in primary sources, it is hard if not impossible to isolate when specific grammatical or lexical neologisms entered the Mandaic language. This is particularly true in the fields of phonology and morphology.21 Ostensibly late elements found in Mandaean works may result from the later copyist who consciously or unconsciously introduced vernacular elements into the written text, and not stem from the time that the composition was originally put down in writing. For example, #‫ר‬#‫( חיו‬hiuara) ‘white (m.sg.)’ and #‫רתי‬#‫( חיו‬hiuartia) ‘white (f.sg.)’ appear in Gy 9: 8 and 15 according to CS 1, the oldest manuscript of the Ginza Rba which was copied in 968 AH (c. 1560 CE). By contrast DC 22, which was copied in 1247 AH (c. 1831 CE) reads #‫ר‬#‫ו‬#‫( ח‬hauara) and #‫רתי‬#‫ו‬#‫( ח‬hauartia). #‫ר‬#‫( חיו‬hiuara) is the older form, and is found in the epigraphic corpus (e.g. BM 117880 [Segal 19 Buckley (2010: 231–273). The language of even the earliest manuscripts differs in many aspects from that of the epigraphic texts, most of which have not been published. A comprehensive study of the orthography, phonology and morphology of the epigraphic texts is now being prepared by Mr. Ohad Abudraham at Ben Gurion University, Be’er Sheva, Israel. See meanwhile Morgenstern (forthcoming) for a discussion. 21 Syntactic differences, such as verbal use, are probably less subject to scribal alteration. For an attempt to trace the development of syntactic patterns in Mandaic texts, see the study by Häberl in this volume. 20 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 373 081M]:16),22 while #‫ר‬#‫ו‬#‫( ח‬hauara) represents a spelling that reflects NM həwɔrɔ, wherein the shewa may be realized as a rounded vowel [ʊ].23 The digraph ‫ו‬# (au) represents this rounded vowel. The classical spelling, which is retained in early copies of the Ginza Rba, has been substituted in the later manuscript by the more phonetic spelling, which reflects the later developments of spoken language. The distribution of these apparent NM forms is not always according to earlier or later manuscripts. For example, one of the most salient features of NM is the shift of īṯ, ūṯ to ext, oxt.24 This process is already attested in the colophon and ritual instructions in the oldest known manuscript, Marsh. 691,25 and in the Leiden Glossarium in the form #‫( מכת‬m̱ṯa) ‘death’, Lat.: mors, Ar.: ‫موت‬ َ (154:1). More examples are now forthcoming from early colophons, e.g. #‫( שופרוכת‬šuprukta) ‘kindness, favour’ (NM ešbəroxtɔ)26 in the colophon of RRC 1C from 1074 AH (1663–1664 CE). This phenomenon is not attested in the epigraphic sources discovered to date, but is attested in the form #‫( אליכת‬alikta) ‘fatty tail’ (NM əlexta)27 in CS 1 (Gy 234:2), which as we have seen is the earliest surviving manuscript of the Ginza. However, in CS 2, which was copied in 1042 AH (1632–1633 CE), i.e. over eighty years later, we find the more conservative form #‫( אלית‬alita). Given what we know about Mandaic in its earlier period, it is likely that #‫( אלית‬alita) is the original reading, since the Ginza is an early text. In DC 12:190, a copy of Pašar Haršia, a Mandaean amulet formula of unknown date copied in 1196 AH (1781–1782 CE), we find #‫( גוריכת‬gurikta) ‘bitch’, and the same reading is shared by RRC 1X which was copied in 1248 AH (1832-1833 CE); but in Ms Berlin or. 8° 3634 d, which was copied in 1231 AH (1816 CE), the reading for the same text is #‫( גורית‬gurita), representing the earlier form 22 242). The correct reading was presented in Müller-Kessler (2001–2002: 131) and Ford (2002a: See Häberl (2009: 85), where the forms həwārā ‘white’ and həwarānā are transcribed phonetically as [ˈhwɔ .rɔ] and [hʊ.wɛ.ˈrɔ .nɔ]. Note that Macuch (1993: 383) heard an o vowel quite distinctly. 24 Kim (2011: 325); Mutzafi (2014: 172 ft. 122). 25 Nöldeke (1875: 78) recognized it as a late form. 26 Macuch (1989: 245); Macuch (1993: 402). 27 Mutzafi (2014: 92–93). 23 374 MATTHEW MORGENSTERN without the NM sound shift.28 Once again, we find the typologically earlier form in a later manuscript; but since the date of this composition is unclear, it is not possible to determine if the typologically earlier #‫( גורית‬gurita) is the original reading or a classicizing correction. Moreover, we must also consider the possibility that some of these “late” elements may have much earlier roots, and represent a hidden non-literary stratum that only rarely found expression in writing.29 For example, the “Neo-Mandaic” form for water, menā/mienā,30 already known from late manuscript sources,31 has now been identified in the Mandaic magic bowls from the late Sassanid or early Islamic period.32 We may also cite a grammatical example: while in CM and the majority of the pre-classical texts the 3m.pl. possessive morphemes (also employed with prepositions) are ‫( –חון‬-hun), ‫( –איחון‬-aihun), ‫( –ון‬-un) or ‫( –איחון‬-aiun) (in pre-classical texts written also with the defective spellings of ‫( –יחון‬-ihun) and ‫( –יון‬-iun)), in NM these forms have been replaced by -u, which is employed also as an object pronoun. Nöldeke already noted that sporadic examples of ‫( ו‬-u) are found attached to 28 Only a few lines later, all of these manuscripts read #‫( חוכצ‬hukṣa) ‘palm-frond’. Drower and Macuch (1963: 135) derived this word from BA ‫( הוצא‬Sokoloff 2002: 373) and Syriac ‫ܐ‬ (Sokoloff 2009: 430), and indeed the direct cognate of these forms is now found in at least two pre-classical Mandaic sources as #‫( חוצ‬huṣa): MS 2087/9:22, a lead scroll which I am currently preparing for publication, and JNF 40:19, a magic bowl which will be published in Ford and Morgenstern (forthcoming). If #‫( חוכצ‬hukṣa) and #‫( חוצ‬huṣa) are related, then we have an exceptional example of the shift of ū to ox in different contexts, though see also NM həwexṣɔ, həwexṣ ‘a dish made of rice flour, dates and sesame’, derived from həḇiṣā (Mutzafi 2014: 14.). It also remains unclear if Mandaic #‫( עוצ‬ʿuṣa, Drower and Macuch 1963: 334-335) is related to this word, as Nöldeke (1875: 61) proposed. Mandaic #‫( חוצ‬huṣa)/#‫( חוכצ‬hukṣa) and its cognates always refer to palm branches, which is not the case with #‫( עוצ‬ʿuṣa). 29 Several characteristic NM forms are already attested in early Jewish Babylonian Aramaic sources; see Morgenstern (2010). 30 Macuch (1989: 236); Macuch (1993: 414–415); Häberl (2009: 338). 31 Drower and Macuch (1963: 242) s.v. maina; 267, s.v. mina. 32 See the correct reading of BM 103365 [Segal 102M]:12, ‫ינה‬#‫( מ‬mainẖ) ‘his water’, pre- sented in Ford 2002a: 259. The same form is now attested in defective spelling, ‫( מינה‬minẖ), in an unpublished bowl from the Martin Schøyen Collection, MS 2054/72:3. NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 375 prepositions in the corpus that he examined, and we now have many more examples of ‫( ו‬-u) and ‫( וה‬-uẖ) from the magic texts.33 A morpheme which might previously have been considered to represent NM influences on manuscript traditions has now been shown to have its roots in the earliest attested levels of the language.34 4. THE LEIDEN GLOSSARIUM The provenance of the Leiden Glossarium has been convincingly established thanks to the work of Borghero.35 The evidence indicates that it was composed in Basra in 1651 by the Carmeline missionary Matteo de San Giuseppe. As we saw, Nöldeke’s opinion of this text was not favourable. Indeed it is true that the Glossarium contains many errors, and the author appears to have had difficulty in distinguishing between the phonemes s and ṣ (though some of the errors might reflect genuine interchanges between these phonemes).36 Nonetheless, precisely on account of the author’s lack of familiarity with the written language, which Nöldeke regarded as a shortcoming, the Glossarium is an invaluable source of contemporary Iraqi Mandaic. In addition, the author’s unique method of marking Mandaic vowels and plosive and fricative pronunciations of the consonants provides unparalleled information about the phonology of NM in the 17th century.37 Borghero has drawn attention to the fact that while in CM the form of the numeral ‘three’ is #‫ת‬#‫( תל‬tlata), the Glossarium employs #‫ת‬#ֺ ‫( כל‬klaṯa), a NM isogloss that is found in all of its forms and clearly distinguishes NM from Classical Mandaic, and has recalled Macuch’s reaction to this form:38 The author of the “Glossarium Sabico-Arabicum-Latinum-Turcum Persicum” gives some words as pronounced only as a proof of his ignorance of the regular spelling. Reading his work before my visit to the Mandaeans, I was especially struck by the word klata “three” written with an initial k as klata. I imagined that the author was partially deaf. My interviews with the Mandaeans, 33 For a detailed discussion see Morgenstern (forthcoming §4). For additional examples drawn from the field of verbal morphology, see Abudraham and Morgenstern (forthcoming). 35 Borghero (1999–2000). 36 Borghero (2004). 37 Ibid. 38 Ibid. (74–75). 34 376 MATTHEW MORGENSTERN however, proved to me that this dissimilated form represents the real pronunciation.39 However, much evidence previously existed for the authenticity of the Glossarium’s form. Already in 1900, Zwemer had recorded examples of the Mandaic language including the days of the week, amongst which was #‫ חובשאב‬#‫ת‬#‫( אכל‬aklata hubšaba) aklatha hofshaba ‘Tuesday’.40 Moreover, evidence was forthcoming from the colophons of manuscripts already in western collections for the existence of such a form. The colophon of CS 23 (fol. 96b) attests that the manuscript was copied #‫ר אלפ‬#‫ת‬#‫י אב‬#‫ אמ‬#‫ת‬#‫ ‏כל‬... ‫ת‬#‫) לשנ‬lšnat…41 klata amai abatar alpa( ‘in the year of three hundred after a thousand’ i.e. 1300 AH (= 1882–1882 CE). Yet another example is found in DC 36 in an historical account added after the cholera epidemic of 1247 (1831 CE). In line 2725, the scribe corrects #‫ת‬#‫( כל‬klata) to #‫ת‬#‫( תל‬tlata), demonstrating his awareness of both the colloquial form and the literary standard. Borghero has also shown that although much material from the Glossarium was incorporated into Drower and Macuch’s Mandaic Dictionary (henceforth MD), many entries have been omitted.42 There is no apparent justification for the arbitrary selection, and it has led to useful information being omitted. For example, the Glossarium (31:1) records a form #‫( אנונ‬anuna) Lat.: aures, Ar.: ‫ا ا‬, i.e. ‘ears’, which must be related to the plural form onɔnɔ ‘ears’ in the dialect of Ahvaz.43 This valuable evidence of an etymologically Aramaic lexeme is not recorded in MD, while numerous loanwords from Arabic found in the Glossarium are presented. Even when the material is recorded, the irregular spellings of the Glossarium’s author and the dictionary’s problematic method of recording entries has on occasion led to the Glossarium’s evidence being separated from the main lexeme it represents. For example, in two places the Glossarium presents a reduplicative plural for the noun sheep: Mand.: ‫י‬#‫ר‬#‫ אנבר‬. #‫( אנברונ‬anbruna, anbrarai), Ar.:‫‏ح ا ‏‬,‫‏ح ل‬. 39 Macuch (1965: 1–2). Zwemer (1900: 287). I owe this reference to Charles Häberl. 41 Here an erroneous date has been corrected. 42 Borghero (2004: 64). 43 Macuch (1989: 246); (1993: 392). 40 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 377 Lat.: ovis, Turk.: 44 ‫ قويين‬Pers.: 45 ‫( كوسف‬Gl. 68:13); Mand.: ‫רי‬#‫( אנבר‬anbrari), Ar.: ‫رعية‬, Lat.: grex, Turk.: ‫رعية‬, Pers.: ‫رعيت‬ (Gl. 85:9). In addition to these, we find once a non-reduplicative plural: Mand.: ‫י‬#‫ אנבר‬.#‫( אנבר‬anbra anbrai), Ar.: ‫خر ف‬, Lat.: ovis, Turk.: ‫قويين‬, Pers.: ‫( كوسف‬Gl. 85:9). This reduplicative plural is not mentioned in the MD’s entry ʿmbra ‘sheep, lamb’46 but instead is listed as a separate entry, ambra,47 a singular form that is not attested in the Glossarium at all. Furthermore, the reduplicative plural is also attested in the Mandaic corpus in the form ‫ריא‬#‫( עומבר‬ʿumbraria) in DC 46. 15:13 but is again listed separately in the MD under the lemma umbra.48 It is not surprising that the single attestation of the reduplicated form in the Mandaic literary corpus occurs in the instructions for preparation of an amulet in DC 46 since, as we shall see below, these instructions are replete with NM lexemes and forms. However, a reader of the MD who wished to examine the distribution of the plural forms of this noun in the various historical stages of Mandaic would be hard pressed to locate them in the dictionary. More work remains to be done on comparing the vocabulary and morphology of the Mandaic in the Glossarium to those of the surviving NM dialects.49 The Glossarium has been employed to excellent effect to establish the earlier history of several NM lexemes still used today,50 but further attention needs to be given to the lexemes that are no longer in use. A preliminary survey reveals that many of them are Arabic loanwords, and it is possible that some were local loanwords that were always specific to Basra.51 Some of these loanwords were so deeply entrenched in the For standard Ottoman Turkish ‫قويو‬. For standard Persian ‫گوسپ‬. 46 Drower and Macuch (1963: 352). 47 Ibid. (22). 48 Ibid. (344). The initial ʿ is accidentally omitted in the dictionary, though the form is listed under ʿ. 49 Ms. Tom Alfia of the University of Haifa is currently working on an MA dissertation that will cover aspects of these topics. 50 Mutzafi (2014, passim). 51 The lack of evidence for the dialects of this period makes it difficult to establish the ex44 45 378 MATTHEW MORGENSTERN Mandaic of the period that they were used to translate a different Arabic word, e.g. ‫‘ ات ف‬destroy’ (Lat.: perdere) in Gl. 21:15 is translated ‫( ֺאסיא‬ȧia),52 from the Arabic root ḍ-y-ʿ. Similarly, the extremely common root n-z-q ‘be close’ apparently a denominative derivative of Persian nazdiq, has almost entirely disappeared from the modern spoken languages, though nazdiq itself ostensibly survives in the dialect of Khorramshahr53 and naziq is employed in the colophon of DC 35.54 The Glossarium also indicates that many Aramaic roots and lexemes which have now been lost continued to function in 17 th century Iraq, e.g. #‫( חבר‬hbra) ‘friend’, Lat.: amicus, Ar.: ‫( ص حب‬Gl. 156:9) which has now been superseded by Persian dūs,55 or Arabic rafīq.56 The rootׁ‫אפך‬ ֺ (ap̱) and its secondary development ‫תפך‬ ֹ (tp̱)57 are attested in a very wide range of meanings in the Glossarium: ׁ‫תפך‬ ֺ (tp̱), ׁ‫גאמתפך‬ ֺ (gamtp̱), Lat.: turbari, Ar.: ‫اضطر‬, ‫( يضطر‬Gl. 9:8–10) ׁ‫תפך‬ ֺ (mtp̱),ׁ‫גאמתפך‬ ֺ (gamtp̱), Lat.: venire in dissidium, Ar.: ‫اخت ف‬, ‫( ي ت ف‬Gl. 14:7–8) ׁ‫אתפך‬ ֺ (atp̱), Lat.: motus, commotio, Ar.: ‫( اضطرا‬Gl. 30:2) ‫אפכאיי‬ ֺ (ap̱aii), Lat.: fluctus, Ar.: ‫( امواج‬Gl. 30:3) ‫אפכא‬ ֺ (ap̱a), ‫גאפך‬ ֺ (gap̱) Lat.: seduci, Ar.: ّ‫تضل‬, ّ‫( يتضل‬Gl. 52:3–4) ׁ‫אפך‬ ֺ (ap̱),ׁ‫גאפך‬ ֺ (gap̱) Lat.: advolvere, revolvere, Ar.: ‫حرج‬, ‫( يحرج‬Gl. 62:1–2) tent to which Iranian spoken Mandaic differed from that of Iraq during the 17 th century. 52 The strange use of Mandaic ‫( ס‬s) with the diacritical point seems to have arisen from the author’s confusion of this letter with the graphically similar Arabic ‫ص‬. See Borghero (2004: 69). 53 Häberl (2009: 341). 54 Macuch (1955: 362). It is striking, however, that neither nazdiq nor naziq are actually used in the published corpora of spoken texts. 55 Macuch (1993: 381). 56 Macuch (1989: 254). Hezy Muzafi informs me that rāhem (Macuch 1989: 254) is a classicism. 57 Compare Sokoloff (2002: 1225), s.v ‫תפך‬. NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 379 ׁ‫אפך‬ ֺ (ap̱),ׁ‫תפך‬ ֺ (tp̱),ׁ‫גאמתאפך‬ ֺ (gamtap̱) Lat.: evertere, subvertere, revolvere, Ar.: ‫ق ب‬, ‫( يق ب‬Gl. 133:5–6) ׁ‫תפאך‬ ֺ (tpa̱),ׁ‫גאמתאפך‬ ֺ (gamtap̱) Lat.: adversari, Ar.: ‫ق‬, ‫( يق‬Gl. 134:5–6) ׁ‫אתפך‬ ֺ (atp̱),ׁ‫גאמתאפך‬ ֺ (gamtap̱) Lat.: revolvere, Ar.: ‫ خرج‬, ‫( ي خرج‬Gl. 181:14) In spite of this widespread usage in the Glossarium, the root has been entirely superseded in the living Mandaic dialects by a different Mandaic root, k-m-r.58 The verbal root ‫בל‬#‫( ק‬qabl), imperfective ‫בל‬#‫מק‬#‫( ג‬gamqabl) ‘receive’ (Lat.: recipere, Ar.: ‫اقبل‬, ‫يقبل‬, Gl. 7:7–8, Lat.: recipere, Ar.: ‫قبل‬, ‫يقبل‬, Gl. 132:7–8 and Lat.: excipere, Ar.: ‫استقبل‬, ‫يستقبل‬, Gl. 19:11–12) is no longer employed, and survives in NM only in the loan-cognate qabul (< Persian < Arabic) ‘acceptance’.59 Semantically, its place is taken by the roots l-x-̣60 and d-r-y.61 #‫פ‬#‫( קל‬qlapa) ‘shell’ (Lat.: cortex, Ar.: ‫قشور‬, Gl. 138:6) has been superseded in NM by miškɔ,62 a lexical impoverishment that loses the original distinction between the softer skin and harder shell. #‫נ‬#‫ת‬#‫חור‬ (huratana) ‘elders’ (Lat.: senes, Ar.: ‫مشي ة‬, Gl. 155:12), and ‫חוראן‬, ‫( חוראתאני‬huran, huratani) ‘elder, elders’ (Lat.: senis, Ar.: ‫شيخ‬, ‫ مشي ة‬Gl. 96:15) are forms that are attested in neither literary nor in spoken Mandaic sources. In contemporary NM, deqen həvar or rīš həvar ‘elder’ is employed in the dialect of Ahvaz (literally ‘white beard’ and ‘white head’),63 attested in the plural as daqqen həvarānā ‘elders’ in the dialect of Khorramshahr.64 It is possible that the increased urbanization and secularization amongst the Iraqi Mandaeans, which brought them into greater contact with Arabic, may have brought about the decline of this type of spoken Mandaic in Iraq.65 It is also feasible that already by the mid-19th century, the Iraqi dialect had been supplanted by the 58 Macuch (1989: 229); Macuch (1993: 406); Häberl (2009: 331). Macuch (1993: 340); Häberl (2009: 345). 60 Macuch (1989: 232); Macuch (1993: 330); Häberl (2009: 334). 61 Macuch (1989: 313–314); Macuch (1993: 382–383); Häberl (2009: 310). 62 Mutzafi (2014: 129–130). 63 Macuch (1993: 385). 64 Häberl (2009: 311). 65 On Iraqi NM and its demise, see Mutzafi and Morgenstern (2012). As Buckley (2010: 6) notes, Petermann did not find Mandaeans in Baghdad in 1854, and M. N. Siouffi, who also came to Baghdad in 1873, did not mention it as a city where Mandaeans lived. By Drower’s time, it was one of the important Mandaean population centres. 59 380 MATTHEW MORGENSTERN dialects from Iran. Certainly the colophons of Mandaic manuscripts attest to great movement over the whole southern Mesopotamian area during this period. 5. FROM BAGHDAD TO PARIS Sustained texts written by Mandaeans in vernacular Mandaic—in contrast to late Mandaic literary texts containing strong influences of spoken Mandaic—are extremely rare. A selection of such texts was published in 1904 by J. de Morgan in his volume of Mandaic manuscript finds under the title “Histoires en mandaïte vulgaire”, but was almost entirely ignored by scholars until Macuch presented a study of them in 1989 as part of his Chrestomathie.66 The origin of de Morgan’s texts remained uncertain, and de Morgan did everything that he could to conceal their source, writing dramatically in the introduction to the volume: Je ne dirai pas comment je me suis procuré ces manuscrits, ni quels ont été mes intermédiaires: ce serait mettre en danger l’existence de ceux qui vivent parmi les Sabéens ou qui doivent encore traverser leur pays. Divulguer un livre est considéré chez ces gens comme une trahison méritant la mort, et chaque Sabéen est chargé de l’exécution de cet arrêt.67 As Macuch observed, the vernacular texts appeared to have been composed a short time prior the acquisition of the manuscripts, “etwa am Ende der siebziger oder Anfang der achtziger Jahre des vorigen Jahrhunderts”.68 In particular, Macuch noted that the text mistitled “Historiette racontщe en langue mandaïte vulgaire”,69 in reality a letter, mentions ‫ס‬#‫סת‬#‫תריא אנ‬#‫( פ‬patria anastas), adding: “Sollte die Identifizierung eines der im mandäischen Brief … angesprochenen christlichen Priesters ... mit Père Anastase Marie ... stimmen, hätte dieser damals noch ziemlich jung (höchstens 30 Jahre alt) gewesen sein müssen. In diesem Fall wäre aber kaum eine frühere Datierung möglich.”70 Given that Père Anastase Marie de St. Elie (1866– 1947) of the Carmelite Mission in Baghdad was known for his contacts with Man- 66 Macuch (1989: 168–191). De Morgan (1904: VIII). 68 Macuch (1989: 12). 69 De Morgan (1904: 282). 70 Macuch (1989: 12). 67 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 381 daeans and his expertise in the Mandaic language,71 Macuch’s identification was quite justified, and was subsequently adopted by Häberl in his own study of the text at hand; as Häberl succinctly stated: “It is unlikely that there were multiple Mandaeophone Père Anastases running around Baghdad at the turn of the century”.72 A recent discovery now removes all doubt regarding the identification. Amongst the Mark Lidzbarski papers held by the Bibliothek der Deutschen Morgenländischen Gesellschaft is a letter from Père Anastase to Lidzbarski written on 20 November 1913. Although the letter deals primarily with questions relating to fishing terminology, which occupied Lidzbarski’s interest during his study of Draša ḏ-Yahia (Johannesbuch),73 it appears from the response that Lidzbarski asked Anastase other questions, including several relating to issues of pronunciation and the existence of other manuscripts. Anastase responded: J’ai reçu des leçons, il y a près de 10 ans d’un professeur mandéen qui me faisait prononcer comme je vous le dis. Ce professeur comprenait un peu sa langue, mais il savait bien le mandaïte vulgaire actuel. Je possède des lettres et des historiettes en ce dialecte que je payais par page (5 francs chacune). Leur langue a fameusement baissé. Tous les manuscrits que possèdent depuis 15 ans les bibliothèques de Londres, Paris, Berlin et autres ont щtщ acquis par moi. Ceux qu’a publiщs M. de Morgan ne font pas exception.74 71 72 Macuch (1989: 184–185) noted these contacts to lend support to the identification. Häberl (2010: 551–552), who presents additional information about this interesting fig- ure. 73 The letter is the first of the two mentioned explicitly in Lidzbarski (1915: 141–142), and cited sporadically through the work. 74 The reference to Berlin is interesting, since the manuscript of “Iniani” published by de Morgan—now Codex Sab. 28 in the Bibliothèque nationale de France—is the twin of another copy found in the Berlin Staatsbibliotek codex Mq 1032. Codex Sab. contains a French description of the manuscript’s contents that seems to have served as a basis for the description of the “Iniani” published by de Morgan, but the writer’s identity and place of origin have been erased. The author might be Anastase, who appended similar descriptions to other manuscripts he supplied to western libraries (e.g. Bodleian Syr. f. 2 and Syr. g. 2.) but comparison of the handwriting with known samples of Anastase’s writing has not led me to any firm conclusion. 382 MATTHEW MORGENSTERN Again, these manuscripts are exceptional, in that they are written entirely in the vernacular and represent genres—personal letters and folk literature—that are not commonly found in Mandaean libraries. I have found nothing similar in nature in any of the Mandaean collections that I have examined. From Anastase’s response it seems likely that they were written specifically for him. The teacher’s lack of familiarity with the literary idiom would also explain why these texts are so free of classicisms which characterize even the more colloquial writings and speech of learned Mandaeans, and why the spellings are so often corrupt. Both Macuch (1989: 11-12) and Häberl (2011: 560) have emphasized the great similarity between the language of the de Morgan (Anastase) texts and that of the contemporary NM dialects. This resemblance stands in contrast to other sources of NM, which show some variance from the currently spoken dialects. 6. MANDAEAN COLOPHONS Mandaic literary texts are generally accompanied by a two-part colophon. The first part provides the name of the work, the name of the scribe, and recounts the textual history of the work, i.e. “I, PN son of PN, copied from a manuscript that was copied by PN son of PN, who copied from a manuscript that was copied by etc.”. The personal names are often accompanied by honorific titles. Following the textual history, the second part lists the date and circumstances in which the manuscript was copied. In several manuscripts, this is in turn followed by historical accounts of important events, e.g. persecutions or disasters, and the reconstruction of Mandaean communities following these calamities. Although the copying histories and dating formulae follow a set literary model, they occasionally integrate elements from the contemporary language of the copyists. This is all the more true of the historical accounts, wherein the copyists had to describe current events and society. It is in the colophons that we find some of the strongest influences of contemporary spoken Mandaic. Jorunn Buckley has used the material in the colophons to great effect to try to reconstruct the lost history of the Mandaeans.75 They are equally valuable as a source of NM. In discussing the orthography of Mandaic, Nöldeke already observed the colophons would frequently employ diacritical symbols to represent sounds in words borrowed from Arabic and Persian that did not exist in CM, namely ʿ, ḥ, j and č.76 However, the colophons available to Nöldeke rarely contained sustained NM passages, and the influences such as they are in those texts are limited primarily (though 75 76 Buckley (2010). Nöldeke (1875: 1–3). NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 383 not exclusively) to loanwords. It is therefore not entirely surprising that Nöldeke would have regarded these colophons as reflecting a literary idiom written in a classicising Mandaic and incorporating elements of the local spoken languages rather than representatives of a living language. Again, Macuch was the first scholar to explicitly declare materials from the colophons to be written in NM rather than simply containing late elements. In his review of Drower’s edition of DC 35, The Baptism of Hibil Ziwa, Macuch remarked “Der Schluß der Handschrift ist höchst interessant. Er wurde im Neumandäischen der zweiten Hälfte des vorigen Jahrhunderts vom Kopisten, welcher Petermanns mandäischer Lehrer war... zugefügt”.77 On the basis of his first-hand knowledge of NM, Macuch was able to improve the readings and interpretations of several passages, though it must be stated that Drower had done remarkably well in translating the colophon.78 As we have seen, there is now clear evidence that Drower had first-hand acquaintance with the rạna, both the Mandaising Arabic speech of the Iraqi Mandaeans and the NM speech of Mandaeans of Iranian origins.79 Nevertheless, certain structures escaped her. For example, in two places I have found that Drower did not correctly interpret the possessive structure noun+d+pronoun, which is one of the salient features that distinguish NM from CM. In the same late historical colophon from DC 35 that Macuch mentioned, which relates an event following the cholera epidemic of 1247 AH (1831–1832 CE) up to 1265 AH (1848–1849 CE), we read #‫ב‬#‫דיון עור‬#‫ן עב‬#‫תד‬#‫ר‬#‫ר װ‬#‫שושת‬-$ #‫בעוחר‬ ‫י‬#‫ת‬#‫כס‬#‫י ות‬#‫( לױימיד מינד‬bʿuhra ḏ-šuštar g̤aratdan ʿbadiun ʿuraba lh̤ imid mindai utaksatai). Drower translated: “On the Šuštar road, Arabs, for greed, surrounded us and were covetous of my things and my clothes”, and wrote of the word ‫ן‬#‫תד‬#‫ר‬#‫װ‬ (g̤aratdan) “P[ersian] ‫‘ ڭردان‬a circuit’, ‘a turn’; hence here ‘made a circuit about us’, Macuch (1955: 362). It was this review that drew Drower’s attention to Macuch and his work and ultimately led to their collaboration on the Mandaic Dictionary. See Macuch (2008: 13). Macuch (1989: 4–5) also called attention to the significance of the colophons as a source of Neo-Mandaic. 78 Contra Macuch (1989: 5), who wrote of Drower’s attempts “daß sie trotz ihres langjährigen Verkehrs mit den Mandäern im Irak, bei denen sich schon seit längerem das bei den Mandäern der südiranischen Provinz Chūzist̄n noch am Leben erhaltene neumandäische Idiom im Absterben befand, die größten Schwierigkeiten hatte und selten den richtigen Sinn traf”. 79 On the distinction between these two see Mutzafi and Morgenstern (2012: 160–162). 77 384 MATTHEW MORGENSTERN ‘surrounded us’. The sentence is obscurely phrased, but the meaning is obvious from the context”.80 However, the correct interpretation of the phrase would appear to be ġārat-dan əwadyon ‘(they) raided us’ wherein ġārat-dan is derived from Arabic/Persian ‫‘ غارة‬making a hostile incursion into an enemy’s country; a raid; plunder, pillage; havoc, devastation’81 construed as a phrasal verb on the model of Persian ġārat kardan. Similarly, in the colophon of BL. Or 6592, copied in Muḥammara in 1289 AH (1872–1873 CE) we read: #‫ כדיב חו‬#‫דת‬#‫כירדה ח‬#‫לדה ו‬#‫ שארױ או‬#‫אנסית מן חד‬$ #‫חו כתיב חו‬#‫וה במ‬#‫ כידבית על‬#‫ אנ‬#‫חו כדיב חו‬#‫ מ‬#‫ת‬#‫לה אס‬#‫ת‬#‫חו‬#‫( ל‬ḏ-ansit mn hda šarh̤ aualdẖ uakirdẖ hadta kdib hua lahuatalẖ asata mahu kdib hua ana kidbit ʿlauẖ bmahu ktib hua). Drower translated “I copied from a šaḥr (!) belonging to his father (?) (walidẖ) and its latter part was newly written. He had no copies. Whatever was written on it (in it) I wrote down, I wrote whatsoever was written”.82 It seems better to translate: ‘that I copied from a šarḥ, the beginning of which and end of which had been written anew. It didn’t have copying traditions. 83 Whatever was written I wrote as it was written’.‫כירדה‬#‫לדה ו‬#‫( או‬aualdẖ uakirdẖ) are NM forms awwaldi ‘its beginning’ and āxirdi ‘its end’. As Häberl has pointed out regarding the NM of Khorramshahr, “foreign nouns will only take possessive suffixes via an intermediary morpheme, -d-”.84 A full study of the language of these colophons has yet to be carried out.85 Nonetheless, even a cursory glance at MD reveals that several lexemes that are characteristic of NM are found only in the colophons. While some have survived into contemporary NM, e.g. the verbal root ‫( בגץ‬b-g-ṣ) ‘stay, wait’,86 others are appar- 80 Drower (1953: 88). Steingass (1892: 877). 82 Drower (1962: 36). 83 For this use of #‫ת‬#‫( אס‬asata) compare ‫ לה‬#‫חו‬#‫ יאתיר ל‬#‫ת‬#‫ס‬#‫( ו‬uasata iatir lahualẖ) ‘and it had no more copying traditions’ (DC 36:825). 84 Häberl (2009: 33). An exception to this rule is found twice in Macuch (1993: 212): merīšdẖ “von seinem Anfang” (l. 1181); ʾāx rīšdẖ “sein anderes Ende” (l. 1182). 85 I am currently preparing such a study for publication. 86 Drower and Macuch (1963: 52); it is already found in the Glossarium. 81 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 385 ently not employed in the surviving dialects, e.g. ‫( מוד‬mud)87 with its variant ‫אמוד‬ (amud) ‘according to, whatever’. 88 A similar word that escaped the attention of Drower and Macuch is ‫( מול‬mul), which from the context means ‘that (which)’. I have found it in two contexts, one of which (DC 27) was known to the authors of MD: ‫י חזאת מן גינזא אנסית‬#‫( וכול מול אינ‬ukul mul ainai hzat mn ginza ansit) ‘and all that my eye saw from the library89 I copied’ (DC 27:558, copied 1088 AH [16771678 CE]);90 and ‫ כול מול‬#‫שאמ‬#‫יכון ל‬#‫ן חזאת עודונ‬#‫זיא מן כול אינ‬#‫ ח‬#‫יכון ל‬#‫אינ‬ #‫ת מן בניא חשוכ‬#‫ן שימ‬#‫( עודנ‬ainaikun la hazia mn kul ainan hzat ʿudunaikun lašama kul mul ʿudnan šimat mn bnia hšuka) ‘may your eye not see any of that which our eye has seen and may your ear not hear all that which our ear has heard from the sons of darkness’ (RRC 2M, also copied 1086 AH [1675–1676 CE]). ‫ן‬#‫( אכתינ‬aktinan) ‘we are’ (DC 51:834), found in the colophon of an amulet formulary written in 1277 AH (1860-1861 AH), was correctly identified in MD as a NM form.91 Another example occurs in the same colophon: ‫( עכתינו‬ʿktinu) ‘they are’ (DC 51:816).92 It is worth noting that in both cases, the forms of the copula in DC 51 differ from those of the contemporary NM dialects, which are respectively extan ‘we are’ and extu ‘they are’.93 But while ‫ן‬#‫( אכתינ‬aktinan) ‘we are’ was accurately identified in MD, the wider context has not been correctly analysed. The full context (DC 51:835–836) reads: #‫ וגוצ‬#‫מ‬#‫ וג‬#‫מ‬#‫ן גוד ח‬#‫ אכתינ‬#‫מ‬#‫( וליליא ועומ‬ulilia uʿumama aktinan gud hama ugama uguṣa), and should be translated ‘and day and night we were in trouble and distress and worry’. MD emends ‫( גוד‬gud) to the graphically similar ‫( גוב‬gub) and interprets this as a phonetic spelling for guw ‘in’;94 however, since ‫( גוד‬gud) is now attested numerous times in several colophons in the Rbai 87 Ibid. (260). Ibid. (22). Perhaps a reflex of this lexeme is found in NM hemmed ‘whatever’. See Macuch (1989: 218; 1993: 387). 88 89 For this meaning of ‫( גינזא‬ginza) see Drower and Macuch (1963: 90). For a different interpretation see Burtea (2008: 118–189). 91 Drower and Macuch (1963: 18). 92 Drower and Macuch (1963: 349) s.v. ʿkt-. 93 Macuch (1965: 380); Macuch (1993: 92); Häberl (2009: 231). 94 Drower and Macuch (1963: 82). 90 386 MATTHEW MORGENSTERN Rafid Collection95 there is clearly no need for emendation. Like ‫( מוד‬mud), ‫גוד‬ (gud) may be explained as a compound of gu and d-. #‫מ‬#‫( ח‬hama) is erroneously derived in MD from Arabic ‫ َحم‬,96 but ‫ هم ‏ غم‬is a fairly common word-pair in Arabic, and the NM Mandaic transcription should be interpreted accordingly. As with the Glossarium, the colophons employ loanwords which are now no longer in use, e.g. #‫( תוף וגוליל‬tup ugulila), apparently ‘cannon(s) and ball(s)’, from Turkish top ve gülle. The words are used to describe the weapons employed by the Turkish army against the Arab militias of the Muntafiq tribe (DC 43I:181), and thus represent the borrowing of terms for contemporary military technology which are now obsolete. 7. SFAR MALWAŠIA (THE BOOK OF THE ZODIAC) Above we saw that this work, which comprises of assorted prognostications, was already known to Nöldeke,97 who expressed a negative opinion regarding its language and style, and was published by Drower, who regarded it as reflecting a transitional period between CM and NM.98 All copies of this book known today may be traced back to the work of a single scribe, Yahia Ram Zihrun br Mhatam,99 from whose work Drower’s manuscript was directly copied in Qurna in 1247 AH (18311832 CE). The work is clearly divided into several original collections, the first two of which have their own colophons (DC 31. 1–106; 106–256). The second colophon ultimately extends back to one Muʿalia br Anuš Bihdad, whom Buckley has tentatively placed around 1100 CE.100 The first colophon extends back a further two generations, but neither of the scribes mentioned is known from other texts. The contents of these sources may be in parts much older, and evidence has been adduced for their reliance upon Babylonian (Akkadian) models101 and for their similarity to 95 So far I have identified the form in RRC 1C, 2M, 2O and 4G. Drower and Macuch (1963: 122). 97 Nöldeke made use of MS Berlin Cod. Petermann 155, copied in Suq eš-Šuiū̱ in 1270 AH (1853–1854 CE). 98 Drower (1949). 99 This scribe was also active in Qurna, where he copied Oxf. Ms. Syr. G 2 (R) in 1231 AH (1816 CE) and Oxf. Ms. Asiat. Syr. C 13 in 1233 AH (1817–1818 CE); see Buckley (2010: 271). He later copied RRC 3F in 1238 AH (1822–1823 CE). 100 Buckley (2010: 241). 101 Rochberg (1999–2000). 96 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 387 Jewish materials found in the Cairo Geniza.102 The third collection (DC 31. 257– 289), which lacks a textual history, is quite clearly of much later origin, and is replete with Arabic loanwords and influences of NM. Let us consider a single passage: ‫לין‬#‫ מן ח‬#‫ניא ודר‬#‫כירמ‬$ ‫יניא איתיא תותיא‬#$ ‫בדית תותיא‬#$ ‫ישֵיא‬ ֵ #‫ו‬#ֵ‫לין ֵח‬#‫ח‬ ‫ך‬#‫רס‬#‫ ם ב‬#‫ריר ולילי‬#ֵֵ‫!ימ ונוחלה בח‬#‫ לתותיא ושוגה נ‬#‫תנ‬#‫ניא ו‬#‫רמ‬#‫ד‬ ‫יניא‬#-$ ‫ויא זיחוא‬#‫ר ויאק‬#‫ך ומינח‬#‫ינ‬#‫שאכבית מליא ב‬ (halin hauaiš̤ia ḏ-abdit tutia ḏ-ainia aitia tutia ḏ-kirmania udra mn halin darmania uatna ltutia ušugẖ naʕim unuhlẖ bh̤arir ulilia kḏ barsak šakbit mlia bainak uminhar uiaqauia zihua ḏ-ainia) These are the medicines that you may make a collyrium for the eyes; bring the collyrium (stone) of Kirman, and take of these medicines and set down the collyrium (stone) and mill it fine, and sift it with silk, and at night, when you lie in your bed, fill it in your eye and your eyesight will be enlightened and grow strong (DC 31. 286:38–41). Most of the language here is late: ‫יײיא‬#‫ו‬#‫( ױ‬h̤ auaiš̤ia) is an Arabic loanword that appears here in its NM meaning of ‘medicine’;103 ‫איניא‬$ ‫( תותיא‬tutia ḏ-ainia) is a calque of Persian tūtiyāʾi dīda ‘A collyrium or medicine for the eyes’;104 ‫ושוגה‬ (ušugẖ) appears in the NM meaning of ‘to rub a special type of stone against a millstone so as to extract a substance used as eye medicine’;105‫!ימ‬#‫( נ‬naʕim) is an Arabic loanword as the pharyngeal ʕ indicates; similarly ‫ריר‬#‫( ױ‬h̤ arir), as indicated by the diacritical pointing that marks the pharyngeal ḥ; ‫ויא‬#‫( ויאק‬uiaqauia) appears to be based upon Arabic root q-w-y ‘be strong’, and employs the y- verbal prefix of Arabic; and ‫( זיחוא‬zihua)106 corresponds to NM zehwɔ ‘light’; while NM zehəw al-inɔ means ‘eyesight’. 107 ‫ניא‬#‫רמ‬#‫( ד‬darmania), though a loanword from Persian, is already found in more ancient Mandaic sources.108 102 Bohak and Geller (2013). Mutzafi (2014: 8). 104 Steingass (1892: 333). 105 Mutzafi (2014: 97–98). 106 In DC 31 the ‫( ח‬h) is written above the line as scribal correction, but the reading is supported by the other textual witnesses. 107 Mutzafi (2014: 174–176). 108 Pognon (1898: 123, text 4:1). 103 388 MATTHEW MORGENSTERN What is significant about this passage is that within the space of a few lines, we find such a high concentration of modern vocables. True, we still find classicisms here, e.g. ‫לין‬#‫( ‏ח‬halin) ‘these’ for NM ahni(n) or hanni,109 and ‫( ם‬kḏ) ‘when’ for NM ke, or waxti ke,110 but that fact that so many of the late usages are shared with contemporary NM shows that the thin façade of the literary conservatism cannot conceal the spoken NM of the scribe who formulated this passage. 8. GRIMOIRES AND AMULETS Beyond the religious works, a large proportion of Mandaic literature comprises amuletic materials, mostly for healing and protection. The style and content of many of the amulet formulae preserved in late manuscripts accord with those found in the much earlier epigraphic materials, demonstrating the great antiquity of the manuscripts’ formulae.111 In recent times, many of the shorter formulae were collected into grimoires (magical ‘recipe books’) which provided professional amulet writers with their raw materials.112 Numerous Mandaic amulets written in modern times are found in western libraries and private collections, of which some are actual objects written for real clients, while others are formularies which could serve as a model for the copying of amulets. Four substantial grimoires are known to me, all from western collections: DC 45, DC 46, CS 24 and CS 27. None of these are dated, but all appear to have been copied in the late 19th or early 20th centuries. Some additional fragments are found in libraries and private collections.113 Several lexemes from CS 27 entered MD from Lidzbarski’s card index of words, e.g. ‫יזא‬#‫( נ‬naiza) ‘[term] used of a reed’,114 while selections from DC 45 and DC 46 were published by Drower115 and most of their vocabulary was also recorded in MD. 109 Macuch (1989: 52; 1993: 56); Häberl (2009: 162); and see Morgenstern (2010: 512– 516). 110 Macuch (1989: 95); Häberl (2009: 332; 361). For previous discussions and literature, see Ford (2002b: 44–45) and Müller-Kessler (2010: 453–454). 112 A description of the structure of the grimoires appears in Morgenstern and Alfia (2013), but at the time of writing, only some of the manuscripts were known to us. 113 A full list of these manuscripts is found in Morgenstern (in preparation). 114 Drower and Macuch (1963: 283). Lidzbarski’s acquaintance with CD 27 is evident from his card index which is now held in the library of the Deutsche Morgenländische Gesellschaft. I have found no evidence that suggests that Drower herself saw either of the manuscripts from Paris. 111 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 389 In many modern manuscript amulets or formularies, and to an even greater extent in these grimoires, the protective formulae were furnished with instructions for the preparation of the amulet.116 Since these instructions had to be comprehensible to the modern copyist, and related to the realia of contemporary life (e.g. materials on which the amulet is to be written, the type of ink to be used, the location in which it is to be placed etc.), these instructions tend to be extremely rich in NM lexemes.117 In addition to the Mandaic formulae, a large number of formulae in the grimoires and occasionally in the amulet scrolls are in Arabic, usually in Mandaic transcription and often corrupt. The transcription shows clear signs of the Muslim gelet type dialects.118 Nonetheless, even when the formulae are in Arabic, the instructions for their use are generally presented in the same type of late literary Mandaic replete with NM influences that characterizes the other rubrics. Furthermore, not all of the Mandaic formulae are of the same ancient date. For example, a lengthy series of amulet formulae presented in narrative form, each of which begins #‫יו‬#‫ד‬$ #‫ב‬#‫( ב‬baba ḏ-daiua) ‘a spell for (lit. of) a dev’, is unmistakably of late origin and contains many NM lexemes.119 It seems that Drower was correct that the series, which is preserved in DC 46 and CS 24, was based upon an Arabic model.120 NM evidence was applied to good effect to interpret the late language of the baba ḏ-daiua texts, and to the examples already identified in MD we may now add one more. Mutzafi has recently drawn attention to the NM use of the root r-d-f in the meaning of ‘to shiver, tremble’.121 This usage of r-d-f may be used to interpret one of the baba ḏ-daiua texts, in which the demon states #‫ת‬#‫עוחר‬$ ‫י לרישא‬#‫דוכת‬ #‫דיפנ‬#‫שא מר‬#‫בר אנ‬$ ‫ויא ועדיא וליגריא‬#‫( ח‬duktai lriša d-ʿuhrata hauia uʿdia uligria 115 Drower (1943). Morgenstern and Alfia (2013: 156–157). 117 For examples see Mutzafi and Morgenstern (2012: 162–163). 118 Ibid. (163–164). 119 Morgenstern and Alfia (2013: 156). 120 Drower (1943: 155–1556); she was apparently unaware that a second copy existed in Paris. 121 Mutzafi (2014: 24–25). 116 390 MATTHEW MORGENSTERN ḏ-br anaša mradipna) ‘my haunt is at the junction (?) and I cause the hands and legs of a man to tremble’ (DC 46. 63: 11–12).122 Not all of the late linguistic material has been presented accurately in MD or explicitly marked as post-classical. For example, Mutzafi has noted the use of derkɔ, derk in meaning ‘bank (of river, stream, canal)’ as characteristic of NM.123 The same definition is recorded in MD with a citation but no reference.124 Examination of the sources reveals that the citation, #‫חר‬#‫נ‬$ ‫( תרין דירכיא‬trin dirkia ḏ-nahra), is drawn from the instructions for writing on an amulet found in two parallel copies in DC 45. 70:16 and DC 46. 210:1. The phrase #‫חר‬#‫נ‬$ #‫( דירכ‬dirka ḏ-nahra) appears elsewhere in these manuscripts meaning ‘bank of the river’, but always within the context of such instructions. I have not found it elsewhere in Mandaic literature. Here we have a clear example wherein these instructions employ a NM usage which is not otherwise attested in Mandaic literature, but which is not unambiguously marked as a late usage in MD. MD’s somewhat haphazard system of marking post-classical lexemes and definitions has proven an additional hindrance to the diachronic study of Mandaic.125 Regarding loanwords in these texts, there are several indications these are drawn from the spoken language. For example, ‫ר‬#‫( ביסמ‬bismar) ‘nail’ is not simply a ‘coll[oquial] corruption of Ar[abic] ‫ ’ ِم ْس ر‬as MD suggests,126 but rather reflects the spoken Arabic of Baghdad.127 The number of loanwords in the baba ḏ-daiua texts and the rubrics is considerable. 9. LATE MANDAIC POETRY In his discussion of Mandaic poetry, Lidzbarski noted that while most of the wedding ritual is of a religious nature, some “volkstümliche Liedchen” are added, but notes “Bei ihrem volkstümlichen Charakter, der vom Inhalt des sonstigen man- 122 Reading and interpretation contra Drower and Macuch (1963: 425) s.v. RDP Part. pres. Mutzafi (2014: 101–102). 124 Drower and Macuch (1961: 109) s.v. dirka. 125 Morgenstern and Mutzafi (2012: 162–164); Morgenstern (2012: 184). 126 Drower and Macuch (1963: 62). 127 Woodhead and Beene (1967: 35). 123 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 391 däischen Schrifttums stark abweicht, ist das Verständnis sehr schwierig”.128 Although Lidzbarski seems to have regarded these poems of being early, he also remarked as an aside “Natürlich können im Einzelnen die Stücke jung sein und sind es auch, wie die arabischen Wörter zeigen”.129 I believe that Macuch was correct to regard that poems as evidence for early NM, and regard his assessment that they “bilden eine Zwischenstufe zwischen einer älteren und der heutigen Volkssprache” to be more accurate than Lidzbarski’s assessment.130 The earliest evidence for these wedding songs is found in CS 15, which was copied in Basra in 1086 CE (=1675–1676 CE).131 The texts were published by Drower, without reference to the older Paris manuscript, primarily on the basis of DC 38, which was copied in 1216 AH (1801–1802 CE).132 It is clear from the translation that Drower did not always understand their late language. Some of these errors were corrected in MD, though the correct interpretations are not always immediately apparent since they were not recorded under their lexical entry. For example, in one song we read #‫אנ‬$ ‫ן‬#‫ בורכ‬#‫ח‬#‫ת אל‬#‫ אנ‬#‫שט‬#‫י פ‬#‫ר‬#‫ למ‬#‫ ועד‬#‫ימ‬#‫ ק‬#‫אתותיא דיקל‬ ‫ לביחומ‬#‫נ‬#‫ל‬#‫( ק‬atutia diqla qaima uʿda lmarai pašṭa anat alaha burkan ḏ-ana qalana lbihum),133 which Drower translated: ‘Standing beneath the date-palm, And outstretching (my) hand to my lord, (I pray), “Thou, God bless me, So that I may speak with the stranger!”’.134 In MD, s.v. bihum, the latter part has been corrected to ‘thou, God, bless me that I may go to the stranger (?)’,135 wherein #‫נ‬#‫ל‬#‫( ק‬qalana) is cor- 128 Lidzbarski (1920: IX). Ibid.: XI. 130 Macuch (1989: 16). 131 Macuch (1989: 4) erroneously reported that the manuscript is CS 11 (this was the old numbering of the Paris manuscripts, still employed by Nöldeke) and that it was copied in 1529/30 (this is the date of Oxf. Marsh. 691, which does not contain these songs). 132 Drower (1950). 133 Published in Drower (1950: 19). Drower’s edition mistakenly omits part of the text, and reads: Atutia diqla anat alaha burkan ḏ ana qalana lbihum, thought the missing words are represented into the translation. 134 Ibid.: 69. 135 Drower and Macuch (1963: 59). In light of Mutzafi’s discovery of 1 f.s. participles in the pattern of našqɔ́ as well as more conservative našqɔnɔ we may translate the first part of this verse ‘I stand beneath the date-palm, and stretch out (my) hand to my lord’. The personal pronoun of ‫י‬#‫ר‬#‫( מ‬marai) ‘my lord’ requires such an interpretation. See Mutzafi (forthcoming). 129 392 MATTHEW MORGENSTERN rectly identified as NM qalɔnɔ ‘I go’.136 Similarly, ‫חיך‬#‫זיא בעוחרה וג‬#‫( ט‬ṭazia137 bʿuhrẖ ugahik),138 which Drower translated ‘That those who mourn in the roads will laugh’ has been corrected in MD s.v. ṭ- to ‘who goes’.139 Nevertheless, some modern forms have been overlooked. In the continuation of the previous text, we read in DC 38: #‫ר‬#‫גול חיו‬-$ #‫ונ‬#‫ מיל‬#‫ שאדויא סקיל‬#‫נ‬#‫ב‬#‫( ק‬qabana šaduia sqila milauna ḏ-gul hiuara), which Drower translated ‘They clad him in a robe, an elegant one, Coloured rose and white’,140 deriving #‫נ‬#‫ב‬#‫( ק‬qabana) from Arabic ‫قب ء‬.141 It seems, however, that #‫נ‬#‫ב‬#‫( ק‬qabana) is to be interpreted as ‘I (f.sg.) desire’, and that #‫נ‬#‫ב‬#‫( ל‬labana) in the same text means ‘I (f.sg.) don’t desire’.142 The remainder of the poem remains difficult to interpret, particularly since the manuscripts contain differences at several crucial points. Aside from the wedding songs, I know of only one other example of late Mandaic poetry. It is a polemical poem written against the consumption of meat that has not been ritually slaughtered. The poem appears at the end of RRC 4G, a copy of Draša ḏ-Yahia (the Mandaean Book of John) copied in Qurna in 1248 AH (1832– 1833 CE) by Yahia Bihram br Adam Yuhana, and appears to be his own composition.143 Although it is mostly formulated in a classicizing style, the author employs post-classical forms such as the demonstrative pronoun ‫ך‬#‫( ת‬tak) ‘that’144 for Classi- 136 32). For the etymology of this form see Morgenstern (2010: 519–523) and Mutzafi (2014: CS 15. 20b:2 reads ‫אזיא‬$ (ḏ-azia). On interchanges of ḏ and ṭ see now Ford (2012). Drower (1950: 18). 139 Drower and Macuch (1963: 171). See also Macuch (1965: 11) (where, however, ṭazia is erroneously identified with ḏ-masgia) and Macuch (1989: 1). 140 Drower (1950: 66). 141 Drower and Macuch (1963: 398 s.v.). 142 Hezy Mutzafi (personal communication) confirms that in the Neo-Aramaic dialect of Ahvaz, the form qəbɔnɔ is employed in this category. 143 On Yahia Bihram’s life and works see Buckley (2010: 133–147). 144 This form of the demonstrative pronoun is not recorded in MD or Macuch’s supplements (Macuch 1965: 527–543; Macuch 1976: 1–146) but is the far-deixis form of #‫( ת‬ta), for which see Drower and Macuch 1963: 477 s.v. It is already found in Drower’s collection in the expression #‫נ‬#‫ך עד‬#‫( בת‬btak ʿdana) ‘at that time’ DC 42: 828. The earliest example of ‫ך‬#‫ת‬ (tak) ‘that’ currently known to me is in the NM colophon of RRC 1C, copied in 1074 AH (1663–1664 CE). 137 138 NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 393 cal Mandaic ‫תה‬#‫נ‬#‫( ח‬hanatẖ) or NM ax or axu.145 The text also employs occasional loanwords, e.g. #‫( ײאר‬š̤ara) ‘cure’ from Persian čāre.146 Though not a rich source of NM, the poem nonetheless demonstrates that even classicizing writings were not free of NM influences. As with the grimoires and magic texts, it seems that during the later stages of editorial redaction, some parts of the classical liturgy were provided with rubrics that show influences of post-classical Mandaic, e.g. ‫( חידוכתיא‬hiduktia) ‘wedding days’.147 These late usages in the rubrics do not necessarily indicate that the liturgical poems that they accompany are late. 10. SHEIKH NEJM’S GLOSSARY This text has been discussed in detail in a recent article, wherein it was shown to contain several NM lexemes.148 Although not a rich source of material, it indicates the degree to which NM remained familiar to Iraqi Mandaeans in the 20 th century and was prone to affect their writings. 11. CONCLUSION All surviving Mandaic manuscripts were copied from the 16th century onwards, i.e. some 900 years later than the epigraphic texts (the lead scrolls and the inscribed clay bowls). During this period, the earlier parts of the Classical Mandaic corpus were redacted, the orthography was to an extent standardized, and the later texts were probably composed. By the time the earliest surviving manuscripts were copied, a very different type of Mandaic was apparently being spoken, and this is reflected in the manuscripts in the ways outlined above. Evidence for this spoken language is unambiguously forthcoming from the 17th century onwards, particularly in the Leiden Glossarium and in some colophons. Fewer sources may be dated with certainty to the 18th century, but a considerable body of material is available from the 19th century. All of these sources demonstrate that the NM dialects spoken today 145 Macuch (1989: 52); Macuch (1993: 56); compare Morgenstern (2010: 515). Compare Häberl (2009: 309). In light of the diacritical marks in RRC 4G and the pronunciation in NM, Häberl (2010: 556-557) was correct to read ‫י‬#‫( שאר‬šarai) in the NM letter published by De Morgan as čāre (contra Macuch 1989: 184, 257, who read šāre). 147 Lidzbarski (1920: 239: 7, 245: 1; 250: 2). See already Nöldeke (1875: 78). 148 Mutzafi and Morgenstern (2012). 146 394 MATTHEW MORGENSTERN represent only a small amount of the Mandaic linguistic riches and varieties that existed when NM was spoken by much larger circles of Mandaeans over a wider geographical area. The devastating effects of the cholera epidemic, persecutions and political upheavals that befell the Mandaean community in the 19 th century, and which are described so vividly (in NM) in the historical accounts appended to the colophons, greatly reduced the scope of NM use. Today, the last remaining vestiges of this language survive on a communal level only in Ahvaz. A preliminary survey of written evidence for NM prior to Macuch’s fieldwork indicates that they may serve as a valuable source for uncovering these last chapters in the lost history of the Mandaic language. It also makes clear that, following the title of Macuch’s seminal work of 1965, MD must be understood as a Dictionary of Classical and Modern Mandaic. REFERENCES Manuscript abbreviations CS—Codex Sabéen, Bibliothèque nationale de France, Paris. DC—Drower Collection, Bodleian Library, Oxford. RRC—Rbai Rafid Collection. Bibliography Bohak, Gideon, and Mark Geller. 2013. “Babylonian Astrology in the Cairo Genizah.” In Envisioning Judaism: Studies in Honor of Peter Schafer on the Occasion of his Seventieth Birthday, edited by Raʿanan S. Boustan, Klaus Herrmann, Reimund Leicht, Annette Y. Reed and Giuseppe Veltri, 607–622. Tübingen: Mohr Siebeck. Borghero, Roberta. 1999–2000. “A 17th Century Glossary of Mandaic.” ARAM 11: 311–319. —. 2004. “Some Linguistic Features of a Mandaean Manuscript from the Seventeenth Century.” ARAM 16: 61–83. Buckley, Jorunn J. 2010. The Great Stem of Souls. Piscataway, NJ: Gorgias Press. Burtea, Bogdan. 2008. „Zihrun, das verborgene Geheimnis“: Eine mandäische priesterliche Rolle. Edition, Übersetzung und Kommentierung der Handschrift DC 27 Zihrun Raza Kasia. Wiesbaden: Harrassowitz. Morgan, Jacques de. 1904. Mission scientifique en Perse, tome V (études linguistiques), deuxième partie: Textes mandaïtes, histoires en Mandaïte vulgaire. Paris: Imprimerie Nationale. Drower, Ethel S. 1943. “A Mandæan Book of Black Magic.” Journal of the Royal Asiatic Society 1943: 149–181. NEO-MANDAIC IN MANDAEAN MANUSCRIPT SOURCES 395 —. 1949. The Book of the Zodiac = Sfar malwašia: D. C. 31. London: Royal Asiatic Society. Drower, Ethel S. 1950. Šarh d-qabin d-Šišlam Rba (D.C. 38): Explanatory Commentary on the Marriage-ceremony of the Great Šišlam. Roma: Pontificio Istituto Biblico. —. 1953. The Haran Gawaita and the Baptism of Hibil-Ziwa. Vatican City: Biblioteca Apostolica Vaticana. —. 1960. Alf Trisar Šuialia, The Thousand and Twelve Questions. Berlin: AkademieVerlag. —. 1962. The Coronation of the Great Šišlam, being a Description of the Rite of the Coronation of a Mandaean Priest according to the Ancient Canon. Translation from Two Tanuscripts Entitled “The Coronation of Šišlam-rba”, DC 54 Bodleian Library, Oxford (1008 A.H.) and Or. 6592, British Museum (1298 A.H.) with Discussion of the “Words Written in the Dust.” Leiden: Brill. Drower Ethel S., and Rudolf Macuch. 1963. A Mandaic Dictionary. Oxford: The Clarendon Press. Ford, James N., and Matthew Morgenstern. 2002a. “Review of J.B. Segal, Catalogue of the Aramaic and Mandaic Incantation Bowls in the British Museum.” Jerusalem Studies in Arabic and Islam 26: 237–272. —. 2002b. “Another Look at the Mandaic Incantation Bowl BM 91715.” JANES 29: 31–47. —. 2012. “Phonetic Spelling of the Subordinating Particle d(y) in the Jewish Babylonian Aramaic Bowls.” Aramaic Studies 10: 215–247. —. in preparation. New Mandaic Incantation Bowls. Häberl, Charles. 2009. The Neo-Mandaic Dialect of Khorramshahr. Wiesbaden: Harrassowitz. —. 2010. “Neo-Mandaic in Fin de sìcle Baghdad.” Journal of the American Oriental Society 130 (4): 551–560. Hunter, Erica. 1995. “Combat and Conflict in Incantation Bowls: Studies on Two Aramaic Specimens from Nippur.” In Studia Aramaica: New Sources and New Approaches. Journal of Semitic Studies Supplement Series 4, edited by Markham J. Geller, Jonas C. Greenfield and Michael P. Weitzman, 61–76. Oxford: University Press. Kim, Ronald. 2011. “Review of Häberl 2009.” Journal of the American Oriental Society 131: 323–327. Lidzbarski, Mark. 1915. Das Johannesbuch der Mandäer. Giessen: Töpelmann. —. 1920. Mandäische Liturgien, mitgeteilt, übersetzt und erklärt. Berlin: Weidmann. Macuch, Maria. 2008. “‘And Life is Victorious!’ Mandaean and Samaritan Literature—In Memory of Rudolf Macuch (1919–1993).” In Und das Leben ist siegreich, And Life is Victorious, edited by Rainer Voigt, 9–16. Wiesbaden: Harrasowitz. 396 MATTHEW MORGENSTERN —. “Review of Drower 1953.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 105: 357–363. —. 1965. Handbook of Classical and Modern Mandaic. Berlin: De Gruyter. —. 1976. Zur Sprache und Literatur der Mandäer. Berlin: De Gruyter. —. 1989. with Klaus Boekels. Neumandäische Chrestomathie mit grammatischer Skizze, kommentierter Übersetzung und Glossar. Wiesbaden: Harrassowitz. —. 1993. Neumandäische Texte im Dialekt von Ahwaz. Wiesbaden: Harrassowitz. Morgenstern, Matthew. 2010. “Diachronic Studies in Mandaic.” Orientalia 79: 505– 525. —. 2012. “Review of Häberl 2009.” Journal of Semitic Studies 57: 182–184. —. forthcoming. “Rare Forms in Eastern Aramaic.” Language Studies 16. [in Hebrew] —. in preparation. “Mandaic Literature: A Guide to Manuscript Sources.” Morgenstern, Matthew, and Tom Alfia. 2013. “Arabic Magic Texts in Mandaic Script: A Forgotten Chapter in Near-Eastern Magic.” In Durch Dein Wort ward jegliches Ding! / Through Thy Word All Things Were Made! – II Mandäistische und Samaritanistische Tagung , 163–179. Wiesbaden: Harrassowitz. Müller-Kessler, Christa. 2001–2002. “Die Zauberschalensammlung des British Museum.” Archiv für Orientforschung 48/49: 115–145. —. 2010. “A Mandaic Incantation against an Anonymous Dew Causing Fright: Drower Collection 20 and its Variant DC 43 E.” ARAM 22: 453–476. Mutzafi, Hezy. 2014. Comparative Lexical Studies in Neo-Mandaic. Leiden: Brill. —. forthcoming. “Verbal Conjugations in the Neo-Mandaic Dialect of Ahvaz.” Nöldeke, Theodor. 1875. Mandäische Grammatik. Halle an der Salle: Waisenhaus. Pognon, H. 1898. Inscriptions mandaïtes des coupes de Khouabir. Paris: Imprimerie Nationale. Rochberg, Francesca. 1999–2000. “The Babylonian Origins of the Mandaean Book of the Zodiac.” ARAM Periodical 11–12: 237–247. Reprinted in Rochberg, Francesca. In the Path of the Moon: Babylonian Celestial Divination and its Legacy. Ancient Magic and Divination 6, 223–235. Leiden: Brill, 2010. Sokoloff, Michael. 2002. A Dictionary of Jewish Babylonian Aramaic of the Talmudic and Geonic Periods. Ramat Gan: Bar Ilan University Press, and Baltimore and London: The John Hopkins University Press. Steingass, Francis J. 1892. A Comprehensive Persian-English Dictionary, 2 vols. London: Routledge & K. Paul. Woodhead, Daniel R., and Wayne Beene. 1967. A Dictionary of Iraqi Arabic: ArabicEnglish. Washington, D.C.: Georgetown University Press. Zwemer, Samuel M. 1900. Arabia: The Cradle of Islam. New York: Fleming H. Revell Company.