Studia Semitica Upsaliensia 30 19. Torkel Lindquist, A War of Words: From Lod to Twin Towers. Defining Terrorism in Arab and Israeli Newspapers 1972–1996 (2001) – A Study in Propaganda, Semantics and Pragmatics. 2003. 20. Assad Sauma, Gregory Bar-Hebraeus’s Commentary on the Book of Kings from His Storehouse of Mysteries. A Critical Edition with an English Translation, Introduction and Notes. 2003. 21. Astrid Ottosson Bitar, “I Can Do Nothing against the Wish of the Pen”. Studies in the Short Stories of Widād Sakākīnī. 2005. 22 Hallvard Hagelia, The Tel Dan inscription. A critical investigation of recent research on its palaeography and philology. 2006. 23 Tal Davidovich, The Mystery of the House of Royal Women. Royal Pīlagšīm as Secondary Wives in the Old Testament. 2007. 24 The Professorship of Semitic Languages at Uppsala University 400 years. Jubilee Volume from a Symposium held at the University Hall, 21–23 September 2005. Edited by Bo Isaksson, Mats Eskhult & Gail Ramsay. 2007. 25. Søren Holst, Verbs and War Scroll. Studies in the Hebrew Verbal System and the Qumran War Scroll. 2008. 26. Ablahad Lahdo, The Arabic Dialect of Tillo in the Region of Siirt (South-eastern Turkey). 2009. 27. Sina Tezel, Arabic Borrowings in Ṣūrayt/Ṭūrōyo within the Framework of Phonological Correspondences. In Comparison with other Semitic Languages. 2011. 28. Jakob Andersson, Kingship in the Early Mesopotamian Onomasticon 2800–2200 BCE. 2012. 29. Ablahad Lahdo, The Arabic Dialect of Tillo in the Region of Siirt (South-eastern Turkey). Textbook. 2016. 30. Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan. Edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner and Judith Olszowy-Schlanger. 2018. Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan Edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner and Judith Olszowy-Schlanger ACTA UNIVERSITATIS UPSALIENSIS Studia Semitica Upsaliensia 30 Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan Edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner and Judith Olszowy-Schlanger © AUU and the authors 2018 ISSN 0585-5535 ISBN 978-91-513-0290-4 Distributor: Uppsala University Library, Box 510, SE-751 20 Uppsala, Sweden www.uu.se, acta@ub.uu.se Printed in Sweden by DanagårdLiTHO AB, 2018 Contents Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan ............................................................... 7 THE EDITORS Part 1: Linguistics, Grammar and Exegesis Semitic Long /i/ Vowels in the Greek of Codex Vaticanus of the New Testament ...................................................................................................... 15 PETER J. WILLIAMS Biblical Hebrew Tense–Aspect–Mood, Word Order and Pragmatics: Some Observations on Recent Approaches .................................................. 27 AARON D. HORNKOHL Long or Short? The Use of Long and Short Wayyiqṭols in Biblical, Parabiblical and Commentary Scrolls from Qumran .................................... 57 JOHAN M. V. LUNDBERG Unmarked Modality and Rhetorical Questions in Biblical Hebrew ............. 75 ELIZABETH ROBAR The Shewa in the First of Two Identical Letters and the Compound Babylonian Vocalisation ............................................................................... 98 SHAI HEIJMANS ‫הֶ חָ כָם‬, but ‫הַ חָ כְ מָ ה‬: Some Notes on the Vocalisation of the Definite Article in Tiberian Hebrew ..................................................................................... 111 DANIEL BIRNSTIEL The Use of Dageš in the Non-Standard Tiberian Manuscripts of the Hebrew Bible from the Cairo Genizah ....................................................... 132 SAMUEL BLAPP The Ashkenazic Hebrew of Nathan Nata Hannover’s Yeven Meṣula (1653) .......................................................................................................... 151 LILY KAHN Medieval Jewish Exegetical Insights into the Use of Infinitive Absolute as the Equivalent of a Preceding Finite Form ................................................. 181 FIONA BLUMFIELD Implementation as Innovation: The Arabic Terms Qiṣṣa and Ḵabar in Medieval Karaite Interpretation of Biblical Narrative and its Redaction History ........................................................................................................ 200 MEIRA POLLIACK Patterns of Diffusion of Phonological Change in the North-Eastern NeoAramaic Dialect of Azran ........................................................................... 217 LIDIA NAPIORKOWSKA The Neo-Aramaic Dialect of Telkepe ......................................................... 234 ELEANOR COGHILL ‘The King and the Wazir’: A Folk-Tale in the Jewish North-Eastern NeoAramaic Dialect of Zakho ........................................................................... 272 OZ ALONI Part 2: Texts, Scribes and the Making of Books and Documents Crossing Palaeographical Borders: Bi-Alphabetical Scribes and the Development of Hebrew Script – The Case of the Maghrebi Cursive ....... 299 JUDITH OLSZOWY-SCHLANGER Beyond the Leningrad Codex: Samuel b. Jacob in the Cairo Genizah ....... 320 BENJAMIN M. OUTHWAITE Arabic Vocalisation in Judaeo-Arabic Grammars of Classical Arabic ....... 341 NADIA VIDRO The Structural and Linguistic Features of Three Hebrew Begging Letters from the Cairo Genizah ............................................................................... 352 ESTARA J ARRANT Birds of a Feather? Arabic Scribal Conventions in Christian and Jewish Arabic.......................................................................................................... 376 ESTHER-MIRIAM WAGNER A 19th Century CE Egyptian Judaeo-Arabic Folk Narrative: Text, Translation and Grammatical Notes ........................................................... 392 MAGDALEN M. CONNOLLY Popular Renditions of Hebrew Hymns in 19th Century Yemen: How a Crudely Formed, Vocalised Manuscript Codex Can Provide Insights into the Local Pronunciation and Practice of Prayer .......................................... 421 REBECCA J. W. JEFFERSON The Status Quaestionis of Research on the Arabic Bible ........................... 442 RONNY VOLLANDT Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan THE EDITORS The work of Geoffrey Khan has had a tremendous impact on a vast array of domains of study, including Hebrew, Arabic, Aramaic, Semitic grammar and linguistics, Bible vocalisation tradition, Cairo Genizah studies, palaeography, codicology and Arabic papyrology. His fresh insights into Semitic syntax and into the pronunciation traditions of the Hebrew Bible, supported by the study of Arabic transcriptions, his pioneering work on Hebrew grammatisation and the history of Hebrew linguistic ideas, his daring fieldwork trips to rescue endangered Neo-Aramaic dialects as well as his precise philological and editorial work on medieval manuscripts and other documents have all transformed our perceptions of these fields. The diversity of fields covered by the research interests of the contributors to this volume reflects the richness of the themes investigated by Geoffrey Khan. This Festschrift is a collection of twenty-one papers written by the disciples of Geoffrey Khan, the Regius Professor of Hebrew at the University of Cambridge. His doctoral students and the post-doctoral researchers whom he has been guiding since 1989 gather here to celebrate his sixtieth birthday. Some of us have since gone on to become university professors and senior researchers, while others are still working on their doctoral and post-doctoral research projects. We all feel and proclaim a deep gratitude to Geoffrey Khan, who has led us with passion through the intricacies of philological and linguistic methods. We have all benefited from his immense erudition and amazing kindness, and he has shared with us his rigorous scientific approach while ensuring that everyone has had the freedom of thought and initiative to follow his or her own path. A typical Geoffrey supervision experience consists of presenting to him the fruits of the labours of past weeks in a somewhat garbled fashion, which then prompts him to say “So do you mean ‘…’?”, effortlessly transforming all you have just said into a fully formed and reasonable theory, while all the time giving you the impression that this is indeed what you had been thinking. A good teacher explains. An exceptional teacher inspires. Geoffrey 7 Khan is just such an extraordinary teacher. His encouragement, care and intellectual generosity turned many fledgling students into scholars, forming a Khan school. We are grateful to be counted among the students of this outstanding scholar. Eager to show our affection for him as mentor and friend, we are presenting Geoffrey with this ‘liber discipulorum’ to mark his sixtieth birthday. The twenty-one papers contribute to different fields ranging from Semitic linguistics and biblical grammar to codicology and the textual study of 19th century books in Hebrew or Arabic. This rich array of objects of study and of disciplinary methodologies is, however, united by a common approach. This focuses on a careful description of the ‘document’ – whether it is written (a manuscript) or oral (a recording of a speaker of a contemporary endangered language) – prior to its interpretation in the light of the most recent ideas of the relevant disciplines. In this manner, our contributors re-examine wellworn assumptions or shed light on old questions. The interrelated questions of language, texts and their material vehicle – the manuscripts – are the leading thread of the volume. The first group of contributions, in Part 1, ‘Linguistics, Grammar and Exegesis’, deal with various aspects of Semitic grammar, linguistics and the history of grammatical and exegetical thought. Several deal with the analysis of the language of the Hebrew Bible. They propose new interpretations of biblical language phenomena such as variations in word order or modality, explore new sources for reconstructing phonetics, and provide fresh research into specific questions of different traditions of Bible vocalisation. Peter J. Williams illustrates one of the under-appreciated features of the famous Codex Vaticanus of the New Testament: its consistent distinction between short and long /i/ vowels in spelling. This distinction is applied to long and short vowels in Greek as well as in Latin loan-words, but also applies to words of Semitic origin. Williams reassesses, moreover, the implications of this for our understanding of early Christian knowledge of Hebrew and Aramaic. Aaron D. Hornkohl’s study examines a series of weaknesses in current approaches to Biblical Hebrew grammar which, it is argued, hinder progress to a fuller understanding of the factors driving word order variation in Biblical Hebrew and to an appropriate interpretation of their intended effects. The use of long and short wayyiqṭols in biblical, parabiblical and commentary scrolls from Qumran takes centre stage in Johan M. V. Lundberg’s article. In the following contribution, Elizabeth Robar revisits the assumption traditionally held by scholars that modality is a part of the verbal system in Biblical Hebrew, expressed through a paradigm often called ‘volitional’, with clearly modal functions. She suggests that modality cannot be confined to the verbal system, but belongs also in the lexicon and in syntax. Shai Heijmans sheds new light on the shewa medium, looking at a shewa that is considered to be silent, but appears in a position where a vocal shewa 8 is expected. While the Tiberian vocalisation system does not distinguish graphically between a silent and a vocal shewa, the Compound Babylonian vocalisation system does. In his contribution, he explores the pronunciation of the so-called shewa medium – when it is silent and when it is vocal – in the Compound Babylonian vocalisation according to Codex Babylonicus Petropolitanus. Daniel Birnstiel focuses on the difficult issue of the vowel of the definite article placed before a guttural consonant in pre-Masoretic and Tiberian Hebrew. Quoting rich evidence from the Bible itself, from transcriptions in Greek and from comparisons with other Semitic languages, the paper explains various phonetic changes in the realisation of consonants and vowels and the way they affected the vowel of the article. Samuel Blapp deals with the use of dageš in the Non-Standard Tiberian manuscripts of the Hebrew Bible. The data presented in this article comes from his recently submitted PhD thesis, and is based on six manuscripts from the Taylor-Schechter Genizah collection at Cambridge University Library. Later periods of the development of the Hebrew language are also represented. For example, Lily Kahn’s paper investigates the 17th century variety reflected in Nathan Nata Hannover’s chronicle of Chmielnicki’s pogroms written in Eastern Europe. Through a detailed grammatical and philological analysis of the chronicle’s language, the paper proposes a much-needed definition of some pertinent features characteristic of ‘Ashkenazi Hebrew’ – the main literary language of north-central European Jews from the Middle Ages to the 20th century. Two papers discuss medieval approaches to grammatical phenomena and the narrative structure of the Bible. In her contribution, Fiona Blumfield sets out to examine the comments of the medieval Jewish exegetes on the use of infinitive absolute as the equivalent of a preceding form. Its origin, use and semantics in Biblical Hebrew have occupied modern grammatical researchers, but none of their interpretations seem entirely satisfactory. Blumfield asks whether it is possible to apply medieval insights to the analysis of this category. Meira Polliack’s study examines the use of the Arabic term qiṣṣa as a technical term for biblical ‘story/narrative’ in Judaeo-Arabic exegesis. It aims at a more detailed understanding of the medieval exegetes’ conceptualisation of the biblical story as a structural-thematic unit that forms part of the final form of the biblical text. The final three contributions in this first part are dedicated to Neo-Aramaic linguistics. They present original research on three dialects of North-Eastern Neo-Aramaic (NENA). Strongly endangered by tragic political events affecting the Iraqi and Kurdish population, these dialects are in need of linguistic recording and description, as precious elements of intangible cultural heritage. Lidia Napiorkowska deals with variation encountered in a language during the process of documentation and the challenges this poses to descriptive linguists. In the NENA dialect of Azran, the phonological processes of consonant and vowel fronting appear to be far from regular, and Lidia demonstrates that 9 these changes are conditioned by a number of factors, including word frequency and the coherence of morphological patterns, as well as language contact. Eleanor Coghill gives a grammatical sketch of a NENA dialect spoken by the Chaldean Catholic Christians of the town of Telkepe, north of Mosul. Since the capture of the town by ISIS in 2014, the dialect has become severely endangered. This contribution outlines the main features of the dialect, comparing and placing it within the NENA dialect context, and noting unusual features in the phonology, morphology and syntax, as well as contact influences from Arabic. The final contribution of this section, by Oz Aloni, provides the transcription and translation of a folk-tale told in the Jewish NENA dialect of Zakho. This folk-tale is rather unusual, since it is built around a relatively uncommon motif in folk-literature, that of gender transformation. Manuscripts and their production, and in particular the work of scribes, are the subject of Part 2 of this volume, ‘Texts, Scribes and the Making of Books and Documents’. The papers all centre around the manuscripts discovered in the Cairo Genizah and other similar collections. This part opens with Judith Olszowy-Schlanger tackling the question of the acquisition and use of writing skills in both Hebrew and Arabic scripts by the scribes of the Genizah world, and the impact that this bi-alphabetism had on the palaeographical changes of the Hebrew script itself. Benjamin M. Outhwaite presents fresh Genizah evidence of the life and work of the master calligrapher Samuel b. Jacob, the scribe of the famous Leningrad Codex. Despite the prominence of his work, little is known of Samuel b. Jacob himself. A colophon of the codex places him in Fusṭāṭ in the first decade of the 11th century, a time and place richly documented in the Cairo Genizah, and now from a few further manuscript discoveries in the Genizah we are able to better trace some new facts about his work, life and possible journey from penury to scribal perfection. Nadia Vidro’s contribution looks at the Arabic vocalisation found in grammars of Classical Arabic copied in Hebrew characters. She provides an edition and analysis of a Genizah fragment consistently vocalised with Arabic signs, suggesting that the fragment is a vocalisation exercise performed by a learner of the Classical Arabic language and its grammar. The cultural and scribal contacts between the Jewish and Muslim epistolary traditions are discussed by Estara J Arrant, who edits and examines three Cairo Genizah documents from the Fatimid period, which are petitions written by Jewish individuals to prominent and influential members of the Jewish community. Esther-Miriam Wagner explores commonalities and differences between Judaeo-Arabic and Christian Arabic by comparing mercantile correspondence of the 18th and 19th centuries from the Cairo Genizah and from the Prize Papers collection, placing linguistic phenomena into the context of literacy, the use of script, and general scribal norms. Magdalen M. Connolly investigates a 19th century Egyptian Judaeo-Arabic folk narrative. Narrated from the perspective of a rural Jewish community oppressed by their ‘uncircumcised’ rulers, this folk-tale – found in the manuscript BNF Hébreu 583 10 and dated to 1839 – depicts a mythologised episode from the life of Abraham ibn ʿEzra, who liberates the community from its tyrannical rulers. Rebecca J. W. Jefferson’s contribution is on popular renditions of Hebrew hymns in 19th century Yemen. It provides a detailed description of a previously unknown Yemenite manuscript codex from the early 19th century. The content consists primarily of Hebrew lamentations and penitential hymns for the Ninth of Av and for personal mourning, as well as texts regarding funerary practices and burial procedures. Rebecca additionally analyses a sample text from the codex. In the last contribution, Ronny Vollandt discusses the status quaestionis of research on the Arabic Bible, surveying the present state of research with a programmatic outlook on what is still to be achieved. It brings together different strands of a dynamic field, which has gained considerable momentum since the turn of the new millennium. This collective volume has benefitted from the help and support of several colleagues. We warmly thank Timothy Jowan Curnow for copy-editing and proofreading the volume. His exceptional skills turned the final stages of preparation into something pleasant. We are grateful to the EPHE-PSL Paris, the LMU Munich and the Carlo Landberg Foundation, administered by Uppsala University, for their support. Thanks are also due to George Kiraz and Melonie Schmierer-Lee, who kindly arranged a first hard copy of this book for the occasion of Geoffrey Khan’s birthday gathering. Last but not least, we are indebted to Colette Khan, who has been so wonderful at keeping a secret. 11 PART 1 Linguistics, Grammar and Exegesis Semitic Long /i/ Vowels in the Greek of Codex Vaticanus of the New Testament PETER J. WILLIAMS Tyndale House, Cambridge It is commonly believed that by the time of the New Testament, and certainly by the time of the earliest manuscripts of the New Testament, there was no distinction between the long and short /i/ vowel in Greek.1,2 According to this view, whereas in earlier forms of Greek, iota could represent a long or short /i/, this distinction no longer held by the time of the New Testament. This article argues that the 4th century Codex Vaticanus (hereafter B), arguably the most prestigious witness to the New Testament, provides evidence against this view.3 The distinction between long and short /i/ can be seen in B in the spelling of many words. One only has to read a single page of the manuscript with this question in mind to be overwhelmed by the mass of supporting data. Here we can establish the distinction by simply appealing to a single word. The verb which in Classical Greek is spelled κρίνω ‘I judge’ (present), κρινῶ (future) has a long stem for the present tense and a short stem for the future. Some time before our earliest New Testament manuscripts, the long /i/ could be marked by the digraph epsilon-iota. This can be seen in Matthew 7:1–2 which, according to the spelling of the first hand of B, reads: μη κρεῑνετε ινα μη κρῐθητε εν ω γαρ κρῐματι κρεῑνετε κρῐθησεσθε. Here the macron and breve have been added to show the distinction in Classical vowel length, 1 My first introduction to my future doctoral supervisor, Geoffrey Khan, was hearing him lecture on vowel length in Hebrew as reflected in Arabic script. It is a pleasure therefore to dedicate to him an essay on vowel length in Hebrew as reflected in Greek script. 2 See Gignac 1981, p. 191, who states that “The confusion of ει and ι, found already in some classical dialects, is paralleled throughout Koine Greek … With the loss of quantitative distinction, there was no longer any question of short or long /i/ in pronunciation, but only of an /i/ sound indifferent in length.” This statement is slightly nuanced by n. 2, which allows that the spelling ει for short [i] makes a later appearance in Asia Minor. Buth, 2012, p. 219 says: “It is certain that ει and ι were both pronounced [i] for the Roman period Koiné. Likewise, there was no distinction between long and short time. There was no ‘short ιωτα’ and ‘long ιωτα’.” 3 The Old and New Testament sections of B are thought to be by two different scribes. What is said in this article applies to the New Testament, but many of the same features also appear in the Old Testament. 15 which corresponds in turn with the alternation in B between epsilon-iota and iota. An even more striking example of this in B is in 1 Corinthians 5:12–6:7:4 5:12 τι γαρ μοι τους εξω κρεινειν· ουχι τους εσω υμεις κρεινετε. 13 τους δε εξω ο θ̅ς κρινει εξαρατε τον πονηρον εξ υμων αυτων· 6:1 τολμα τις υμων πραγμα εχων προς ετερον κρεινεσθαι επι των αδικων και ουχι επι των αγιων· 2 η ουκ οιδατε οτι οι αγιοι τον κοσμον κρινουσιν· και ει εν υμιν κρεινεται ο κοσμος· αναξιοι εστε κριτηριων ελαχιστων· 3 ουκ οιδατε οτι αγγελους κρινουμεν. μητι γε βιωτικα· 4 βιωτικα μεν ουν κριτηρια εαν εχητε τους εξουθενημενους εν τη εκκλησια τουτους καθιζετε· 5 προς εντροπην υμιν λαλω· ουτως ουκ ενι εν υμιν ουδεις σοφος ος δυνησεται διακρειναι ανα μεσον του αδελφου αυτου· 6 αλλα αδελφος μετα αδελφου κρεινεται και τουτο επι απιστων· 7 ηδη μεν ουν ολως ηττημα υμιν εστιν οτι κριματα εχετε μεθ εαυτων In this passage the six present and aorist verbs have epsilon-iota.5 However, the three nouns and the three verbs plausibly interpreted as future have iota. Again, this is absolutely in line with Classical distinctions in vowel length. It also means that we do not have to wait until the widespread use of accents from the 9th century onwards to tell the difference between the present and future of κρ(ε)ινω. Not only do we find that so extensive a passage consistently marks the distinction, but we also find an absence of hypercorrection in B in the entire New Testament. According to unpublished data gathered by Patrick James in preparation for the Tyndale House Greek New Testament,6 B avoids all examples of representing etymological short /i/ by epsilon-iota in verbs or nouns relating to κρ(ε)ινω, totalling over 309 cases. We find the same tendency for B to distinguish long and short vowels in many lexemes, and also find that in some patterns B is joined by other witnesses. For instance, if we just consider verbs in Luke’s Gospel, long /i/ is represented in a wide range of witnesses by epsilon-iota for common verbs including γ(ε)ινομαι, γ(ε)ινωσκω, κλ(ε)ινω, κρ(ε)ινω and μ(ε)ισεω. In fact in all these cases, epsilon-iota is the strongest attested spelling in early witnesses. A natural question to arise in relation to B is whether these spellings are inheritances from the earliest forms of the text, or alternatively are innovations by the scribe of this manuscript or one of its forebears. The answer to this is not simple: in this matter, as in many others, B appears sometimes as completely isolated and at other times as well supported by other manuscripts. The isolation may be so total that it might suggest that B contains 4 I am grateful to Patrick James for pointing this passage out. The aorist active is also long by compensatory lengthening for the loss of original sigma after the nu. 6 Jongkind, 2017. 5 16 an innovation, or the reading may be so widely shared that we are more likely to conclude that it was inherited. This pattern of readings is consistent with the view that the scribe of B was rather knowledgeable. He not only distinguishes historical vowel length in Greek with almost complete consistency, but may also have been aware of Latin vowel length.7 Here, however, we consider what is potentially an even more impressive feat in representing vowel length, namely the representation of Hebrew (and occasionally Aramaic) vowel length. 1 The genealogies of Jesus We consider this question first in relation to the genealogies of Jesus in the first chapter of Matthew and the third chapter of Luke. An advantage of considering the genealogies is that they contain many names which show little or no sign of nativisation in Greek. In the genealogy of Matthew 1:1–16, the /i/ sounds occur in a variety of phonetic contexts.8 Here are a few broad classes: • • • • Names beginning with a pre-vocalic consonant where iota represents yodh: ιακωβ, ιεχονιας, ιουδας, ιωαθαμ, ιωραμ, ιωσαφατ and ιωσηφ. Presumably by assimilation to the prevalence of the beginning ιω(occurring in four other names in this list and representing the most common initial pattern), we also have ιωβηδ as the name of David’s father Obed (‫)עֹובֵ ד‬. Names containing iota in a post-consonantal pre-vocalic position: αβιουδ, ελιακειμ, ελιουδ and σελαθιηλ. Names containing stressed /i/ and arguably a long /i/ represented by epsilon-iota: αμειναδαβ, αχειμ, δαυειδ and ελιακειμ. In each case the 9th century accentor of B placed the accent on the final syllable, though the Hebrew equivalent of αμειναδαβ also has secondary stress on the /i/ vowel (e.g. Ruth 4:19 and 20 in the Masoretic text). Names with a final Yahwistic element, where the data are less clear. Some have epsilon-iota, while others have just iota: οζειας, ουρειου and ιωσειας vs εζεκιας, ιεχονιας and αβια. ουρειου is the only genitive here and αβια is unique for being treated as indeclinable.9 Looked at Pontius Pilate, for instance, is called πειλατος (e.g. Matthew 27:13), consistent with the long initial vowel in Latin. 8 Where possible the names are given in their nominative forms as they occur in the genealogy, which typically has both accusative and nominative. The exception to this is the genitive ουρειου. In instances where the nominative and accusative differ, the final consonant for the nominative is sigma and for the accusative is nu. The vowels do not change. 9 It lacks sigma and nu in the nominative and accusative respectively, and receives an accent on the final syllable from the accentor of B, when all other forms have accent on the preceding /i/ vowel. 7 17 • from a purely Greek angle we thus have the contrasting pairs οζειας and ιωσειας vs εζεκιας and ιεχονιας, which do not provide enough data to draw any firm conclusions. In a category of its own is ισαακ, the only case where an /i/ vowel precedes a consonant at the beginning of a name, whether or not this name was pronounced in Hebrew with initial /y/. We see broadly similar patterns in B for Luke 3:23–38: • • • • Thirteen names begin with a pre-vocalic consonant where iota represents yodh: ιακωβ, ιανναι, ιαρετ, ιεσσαι, ιησου, ιουδα, ιωαναν, ιωβηλ,10 ιωδα, ιωναμ, ιωρειμ, ιωσηφ and ιωσηχ. This is a large class, as it was for Matthew 1:1–16. An equally large class consists of names containing stressed /i/ and arguably a long /i/ represented by epsilon-iota. The digraph occurs in these final syllables whether it ends in a consonant or not: αδδει, αρνει, εσλει, ηλει, ηλειει,11 λευει, μελχει, νηρει, αδμειν, δαυειδ, ελιακειμ, ιωρειμ and σεμεειν. Names containing iota in post-consonantal pre-vocalic position: ελιεζερ, μαθθαθιου, ματταθιου and σαλαθιηλ. Again, ισαακ in a class of its own. However one classifies the examples, we see two clear patterns in both genealogies: • • Initial consonantal and therefore pre-vocalic /y/ is always a single iota. Stressed final /i/ is always spelled epsilon-iota. Thus the distribution of epsilon-iota vs iota in Hebrew names is not random. More difficult is the question of vowel length, where our method with ancient sources is necessarily inferential. We may at least say that all of the above cases of epsilon-iota might be long. Because epsilon-iota always occurs in what in Hebrew would be the stressed position (or, in the case of αμειναδαβ, where there was secondary stress) and because these vowels are all probably etymologically long, it becomes plausible, in light of other tendencies in B, that these vowels were considered long by the scribe of B or whoever formed his tradition. 10 Probably a corruption of ιωβηδ, itself an assimilation to the prevailing ιω- prefix rather than a representation of yodh. 11 This form in Luke 3:24 appears to be a mistake corrected by the first corrector to λευει. 18 2 Three common words In the rest of this article we will be considering individual words which occur in speech or within a wider discourse. The method we will use is to put B alongside other early witnesses, including Codex Sinaiticus (‫ = א‬Codex Sinaiticus; ‫א‬a = Codex Sinaiticus Scribe A; ‫א‬d = Codex Sinaiticus Scribe D), which is also from the 4th century, and various papyri. All the papyri referenced in this article are plausibly dated to the 3rd century.12 We begin by considering in table 1 the spelling of Elijah, which occurs 29 times in the New Testament. B spells every single occurrence with epsiloniota, offering the forms ηλειας, ηλειαν, ηλειου and ηλεια. Table 1. Epsilon-iota vs iota in the name Elijah Elijah Epsilon-iota Iota Matthew 11:14 Matthew 16:14 Matthew 17:3 Matthew 17:4 Matthew 17:10 Matthew 17:11 Matthew 17:12 Matthew 27:47 Matthew 27:49 Mark 6:15 Mark 8:28 Mark 9:4 Mark 9:5 Mark 9:11 Mark 9:12 Mark 9:13 Mark 15:35 Mark 15:36 Luke 1:17 Luke 4:25 Luke 4:26 Luke 9:8 Luke 9:19 Luke 9:30 Luke 9:33 John 1:21 John 1:2513 Romans 11:2 James 5:17 Total Total excluding B B B ‫א‬d B ‫א‬d B ‫א‬d B ‫א‬d B ‫א‬d B B B B ‫א‬B B B B B B ‫א‬d B ‫א‬d B ‫א‬d B B ‫א‬a B P75 ‫א‬a B P75 B P45 P75 ‫א‬a B P45 P75 B P66 P75 B P66corrector P75 B B B 51 22 ‫א‬a ‫א‬d ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a ‫א‬a P66 ‫א‬a ‫א‬a ‫א‬a 18 18 12 This dating is for the practical purposes of this article of establishing early spelling. The argument is not affected if some of them are from the 2nd or 4th centuries. 13 At John 1:25, P119 has ]ιας ουδ[ε. Unfortunately it is not possible to know whether it spelled ηλειας or ηλιας. 19 What makes the spelling with epsilon-iota so impressive is that it is the majority spelling of manuscripts from the 4th century or earlier, occurring in 51 of 69 occurrences. A problem with this, of course, is that B’s own 29 uses of epsilon-iota have been counted in the total. If we remove these we still have 22 occurrences of epsilon-iota in other witnesses against 18 of simple iota. Epsilon-iota is still the majority spelling at 55 per cent of occurrences. We must further note that the occurrence of epsilon-iota in ‫ א‬on 12 out of 29 instances is striking given ‫’א‬s strong scribal tendency to use simple iota, even where forms are traditionally spelled with epsilon-iota. Moreover, we have a further consideration: The New Testament of ‫ א‬was written by two scribes, known as Scribe A and Scribe D. Scribe A penned the majority of the New Testament, but Scribe D replaced six of the leaves in the New Testament with his own work.14 Scribe D is thought to be much better at spelling than Scribe A.15 Thus 8 of the 12 occurrences of epsilon-iota in ‫ א‬are provided by the better speller – a scribe who only once allows plain iota, namely in Matthew 16:14. This occurrence is easily explained by the phenomenon in Greek manuscripts whereby a scribe corrects the first occurrence of an unexpected spelling because he considers it a mistake. However, when he sees it for a second time he accepts that it is not a mistake and copies it accurately. Thus the testimony of ‫ א‬strongly points towards a plentiful representation of epsilon-iota in its Vorlage. In the 3rd century, though the papyri overwhelmingly attest epsilon-iota (in 9 of 10 occurrences) the instance of ηλιας in John 1:25 in P66 indicates that the pattern is not uniform.16 Next we consider two contrasting cases: the word ‘rabbi’, in table 2, and the word ‘Pharisee’, in table 3. B spells the words as ραββει and φαρεισαιος respectively. Because of the large number of occurrences, and in order to maximise the number of witnesses from the 4th century or earlier, we will only consider these in the Gospel of John. Table 2. Epsilon-iota vs iota in the word ‘rabbi’ Rabbi Epsilon-iota Iota John 1:38 John 1:49 John 3:2 John 3:26 John 4:31 John 6:25 John 9:2 John 11:8 Total Total excluding B P66 P75 ‫ א‬B P66 P75 ‫ א‬B P66 P75 ‫ א‬B P66 P75 ‫ א‬B P66 P75 ‫ א‬B P75 ‫ א‬B P66 P75 ‫ א‬B P45 P66 P75 ‫ א‬B 32 24 0 0 14 Jongkind, 2007, p. 40. Jongkind, 2007, pp. 90–94. 16 The fact that the contemporaneous corrector of P66 inserted epsilon in John 1:25 may give us greater confidence that P66’s Vorlage had the longer spelling. 15 20 Table 3. Epsilon-iota vs iota in the word ‘Pharisee’ Pharisee Epsilon-iota Iota John 1:24 John 3:1 John 4:1 John 7:32a John 7:32b John 7:45 John 7:47 John 7:48 John 8:13 John 9:13 John 9:15 John 9:16 John 9:40 John 11:46 John 11:47 John 11:57 John 12:19 John 12:42 John 18:3 Total Total excluding B B B B B B B B B B B B B B P45 B P45 B B B B B 21 2 P66 P75 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P66 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P66 ‫א‬ P66 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ P6 P66 ‫א‬ P66 ‫א‬ P66 ‫א‬ P66 P75 ‫א‬ P66 P75 ‫א‬ ‫א‬ 50 50 The contrast could hardly be starker. In the case of the first word, ραββι receives no support while ραββει receives complete support. In the case of the second word, φαρισαιος receives almost universal support while φαρεισαιος has support only from P45 and B. 3 Three similar words Next we consider three words which, whatever their origin, come from Semitic texts, and appear similar in Greek: Sion (in table 4), Sidon (in table 5),17 and Sinai (in table 6). Table 4. Epsilon-iota vs iota in the place-name Sion Sion Epsilon-iota Iota Matthew 21:5 John 12:15 Romans 9:33 Romans 11:26 Hebrews 12:22 1 Peter 2:6 Revelation 14:1 Total Total excluding B B P66 P75 B B P46 B P46 (B not extant) B P47 (B not extant) 10 5 ‫א‬ ‫א‬ ‫א‬ ‫א‬ ‫א‬ ‫א‬ ‫א‬ 17 7 7 The words for ‘Sidonia’ and ‘Sidonian’ are also relevant here, but left out for simplicity. 21 Table 5. Epsilon-iota vs iota in the place-name Sidon Sidon Epsilon-iota Iota Matthew 11:21 Matthew 11:22 Matthew 15:21 Mark 3:8 Mark 7:31 Luke 6:17 Luke 10:13 Luke 10:14 Acts 27:3 Total Total excluding B B B B B P45 B ‫א‬B P45 B P45? P75 B B 14 5 ‫א‬ ‫א‬ ‫א‬ ‫א‬ ‫א‬ P75 ‫א‬ ‫א‬ ‫א‬ 9 9 Table 6. Epsilon-iota vs iota in the place-name Sinai Sinai Epsilon-iota Iota Acts 7:30 Acts 7:38 Galatians 4:24 Galatians 4:25 Total Total excluding B B ‫א‬ ‫א‬B ‫א‬ ‫א‬ P46 B P46 B 5 2 5 4 In all three words, B prefers epsilon-iota and is also on the side which is overall better attested. 4 Less common words In less frequent words, we still see a strong preference for epsilon-iota for Semitic long /i/ in Greek witnesses up to the 4th century, as we can see from the forms for the place-name Chorazin (in table 7) and the word ‘cherubim’ (in table 8). Table 7. Epsilon-iota vs iota in the place-name Chorazin Chorazin (χοραζειν) Epsilon-iota Matthew 11:21 Luke 10:13 Total Total excluding B ‫א‬B P45 P75 ‫ א‬B 6 4 Iota 0 0 Table 8. Epsilon-iota vs iota in the word ‘cherubim’ Cherubim (χερουβειν) Epsilon-iota Iota Hebrews 9:5 Total Total excluding B P46 B 2 1 ‫א‬ 22 1 1 We also get an /i/ vowel in Semitic speech in the representations of talitha and the rather similar name Tabitha (see table 9). Table 9. Epsilon-iota vs iota in the word talitha and the name Tabitha Talitha and Tabitha Epsilon-iota Iota Mark 5:41 (talitha) Acts 9:36 (Tabitha) Acts 9:40 (Tabitha) Total Total excluding B B P45 P53 B P53 B 6 3 ‫א‬ ‫א‬ ‫א‬ 3 3 The different forms of Jesus’s dereliction cry in Matthew and Mark contain an /i/ vowel and are spelled thus in B: ελωει ελωει λεμα σαβακτανει (Matthew 27:46) and ελωι ελωι λαμα ζαβαφθανει (Mark 15:34). It is interesting to note that B is isolated as the only witness to have the phrase in these two ways. Arguably the best early attested form as represented in the Tyndale House Greek New Testament is ηλει ηλει λεμα σαβαχθανει (Matthew 27:46) and ελωι ελωι λαμα σαβαχθανει (Mark 15:34). In particular, the form ελωει for the Matthaean occurrence is unique to B, as is the form σαβακτανει, perhaps because some scribe felt that it better represented ‫ שבקתני‬than σαβαχθανει, which follows Greek rules of assimilating the aspirate χ to the following θ but in doing so creates the counterintuitive equation of χ and ‫ק‬. The form ζαβαφθανει is also unique to B in Mark, but may be a corruption of some representation of the Hebrew ‫( עזבתני‬cf. Codex D’s unique ζαφθανει). 5 Interpretation of the data When we consider B’s treatment of etymological long /i/ Greek words, and then its treatment of Hebrew names in the genealogy of Jesus, and then its treatment of the various individual words above, we see a clear pattern: B with striking consistency uses epsilon-iota for what looks etymologically like a long /i/, either in Hebrew/Aramaic or in Greek. This is the case with common words, mid-frequency words and low frequency words – a pattern which must either point to the extremely careful work in transmission or in ensuring editorial consistency. The question we now pose is how we should interpret these data. We consider first two contrasting hypotheses: a. The Preservation Hypothesis. One possibility is that B is overwhelmingly preserving the original or earliest form of the text. In that case we would have to posit that B represented a stream of copying which was exceedingly faithful, even in the minutiae of spelling; there was no link in the chain of transmission which was weak in this issue. 23 Moreover, the authors of the New Testament or their amanuenses consistently distinguished short and long vowels in their writing. This is a high demand, which requires learning and care in both the authors and in the entire line of scribes leading to this 4th century manuscript. There is a further downside to this hypothesis: it requires us to posit selective carelessness on the question of spelling in many other witnesses. According to this hypothesis, these witnesses consistently preserved the long spelling of ραββει, while regularly adapting an original spelling of φαρεισαιος to the shorter φαρισαιος. b. The Innovation Hypothesis. A rather different view is that the scribe of B (or, in a slightly more complex hypothesis, a previous scribe whose work was faithfully copied into B) actually knew the distinction between long and short vowels and imposed this upon the text he received even though the text he received did not actually show that distinction at all. The problem with this view is that it becomes hard to imagine how someone could have such extensive knowledge of both Greek and Hebrew vowel length in a context where no one else preserved vowel length. From considering the weaknesses of these two extreme views, we turn to a middle way which may actually have more explanatory power. This position is that the correlation between epsilon-iota and historically long /i/ in B is mostly preservation and partly innovation. The principal disadvantage with this view is that it takes a unitary phenomenon (the correlation between epsilon-iota and long /i/) and splits it in two. The advantages, however, are several: a. we do not have to suppose that all other manuscripts strayed from the earliest forms with such frequency as in the Preservation Hypothesis; b. we allow for a range of care in transmission such that some scribes do not preserve spelling, while many do; c. we explain how traditions about short and long vowels could be available to the scribe of B; d. we do not require the scribe to be introducing the epsilon-iota grapheme for a long /i/ de novo; e. we explain a significant correlation between the spelling of B and many of the earliest papyri and the way epsilon-iota is sometimes used in the earliest majuscules; f. we explain B in a way consistent with its scribal tendencies elsewhere: the scribe is clearly sufficiently learned to seek to reinforce the tendencies of the authors themselves; 24 g. B may be regarded as carefully put together and as representing a high quality tradition, without being accorded an almost unchallengeable authority at the expense of other manuscripts.18 6 Conclusions We have seen that B distinguishes long and short /i/ for Greek words and for Semitic ones too. The almost complete consistency with which it does this is impressive. For many or even most of these forms it receives considerable support from other Greek witnesses of the 3rd and 4th centuries, but at times it is much more isolated, especially in the case of φαρεισαιος for the more common φαρισαιος. This leads to the conclusion that whereas many or perhaps even a majority of B’s uses of epsilon-iota for originally long Semitic /i/ are preservations from the beginnings of the tradition, not a few cases are innovations. From the fact that long Semitic /i/ is differentially preserved in writing in some parts of the Greek manuscript tradition until the 4th century it is not a very big step to conclude that knowledge of its pronunciation may also have been preserved this late. A simple objection to this inference is the observation that traditional spelling can lag long behind pronunciation. Scribes may learn received spelling and use it long after the spelling no longer represents pronunciation. This is true and is abundantly the case for Greek. Yet this insight cannot be used to explain all the New Testament data precisely because the use of epsilon-iota to represent the long /i/ is not a traditional spelling but an innovation. While traditional spellings may not represent pronunciation, innovative spellings typically do. The early papyri of the New Testament give testimony to a prior spelling reform away from the use of iota for the long /i/ towards epsiloniota, and this can only be interpreted as an attempt to codify a pronunciation difference between short and long /i/. This spelling reform, of course, did not begin with the scribes of Christian literature, but given the significant correlation between epsilon-iota and Semitic long /i/ in early witnesses of the New Testament, we may conclude that it probably did not post-date the composition of the New Testament books. This includes the tendency to omit the definite article with the name Ἰησοῦς in John’s Gospel, and to increase the frequency of the name ‘Christ Jesus’ over the name ‘Jesus Christ’ in the Pauline Corpus. See Jongkind, 2017, p. 507. 18 25 References Buth, R., 2012, “Koiné Greek pronunciation”, Biblical Language Center, https:// www.biblicallanguagecenter.com/koine-greek-pronunciation/, accessed 11 July 2017. Gignac, F. T., 1981, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, vol. I: Phonology, Milan: Istituto Editoriale Cisalpino. Jongkind, D., 2007, Scribal Habits of Codex Sinaiticus, Piscataway: Gorgias Press. ——— (ed.), 2017, The Greek New Testament, Produced at Tyndale House, Cambridge, Wheaton: Crossway and Cambridge: CUP. 26 Biblical Hebrew Tense–Aspect–Mood, Word Order and Pragmatics: Some Observations on Recent Approaches AARON D. HORNKOHL University of Cambridge 1 Introduction1 As has been observed more than once in recent years, consensus remains elusive on a ‘grand unified theory’ of constituent order variation in the Biblical Hebrew (henceforth BH) verbal clause.2 Useful concepts have been proposed and promising theories have been applied to – and, indeed, at times forced upon – the Hebrew data, but even the most successful, highly explanatory approaches leave large numbers of cases unexplained.3 This study aims at nothing so ambitious as the proposal of a comprehensive system, but is merely one in a series of recent treatments offering critical observations on traditional and current approaches, suggesting elements that, it is here argued, should be considered integral within such an all-encompassing explanation. Due to concerns of space, no attempt at bibliographic exhaustiveness is made; rather, references are generally limited to especially influential, recent and/or relevant research. It is hoped that the particular critique and synthesis offered in this modest article might pave the way for further scholarly development, or, at the very least, may prove of clarificatory benefit to interested parties. 2 Verbal semantics: TAM There is no sound way of treating word order variation in BH that does not take into account the ‘in-built’ semantics of the language’s individual verbal 1 For as long as I’ve known him, my friend, colleague and mentor, Prof. Geoffrey Khan, has been an inexhaustible source of help and support as well as a worthy example of scholarship and humanity. As his first published monograph, Studies in Semitic Syntax (based on his 1984 PhD dissertation), dealt with the pragmatics of extraposition in the Semitic languages, it is a pleasure to contribute this study on the related matter of constituent order in Biblical Hebrew to a Festschrift in honour of his 60th birthday. 2 Moshavi, 2010, pp. 18–47, 119–120; Holmstedt, 2011, p. 2; Shimasaki, 2013, pp. 763–766. The lack of consensus touches on verbless/nominal clauses as well; see e.g. Miller, 1999. 3 Moshavi, 2010, p. 119. 27 forms, especially – but not exclusively – the respective meanings of the four principal finite forms, qatal, yiqtol, wayyiqtol and weqatal, in terms of tense, aspect and mood (henceforward TAM). This is because the choice between these, while to some extent pragmatically conditioned, depends, at least partially, on fixed semantic values.4 As such, before coming to the issue of constituent order, it is necessary to dedicate some space to BH verbal TAM. The approach adopted here is principally synchronic, but informed by awareness of historical developments in Hebrew and the Semitic languages more broadly, and by sensitivity to cross-linguistic evidence from the related domains of diachronic typology and grammaticalisation. In other words, for purposes of the present discussion it is deemed more important to explain how the BH verbal system works than the exact routes and processes required for development from a preceding system,5 although it is readily admitted that certain assumptions about its ancestor(s) help to explain features unexpected within a purely synchronic perspective.6 2.1 Tense-prominence In Bhatian typological terms,7 it is claimed here that BH is best seen as a tenseprominent language. Whatever the nature of the verbal system in the linguistic ancestor(s) of BH, BH verbs seem to encode temporal values – whether absolute or relative – more basically and consistently than either aspectual or modal values. Qatal and wayyiqtol correlate strongly with past time, yiqtol and weqatal with future orientation, while also encoding limited non-future values. The active participle, a morphological substantive with both nominal and verbal functions, must also be accorded full membership in the verbal system, since it defaults for both actual and relative present, and can also bear generic present and imminent-future force.8 4 In agreement with Cook, 2012, pp. 272–275. Thus the present approach differs from those in which great emphasis is placed on the precise character of the proto-system from which the BH system could conceivably have developed. 6 For example, while eventualities (actions, events, states) depicted using qatal forms are most often in the past, regular exceptions to this norm are present tense uses of stative verbs, verbs denoting feeling or thought, and performatives. These apparently irregular qatal uses are illuminated by comparative evidence, e.g., the old use of qatal as a ‘verbal adjective’ in Akkadian. From this perspective, present tense uses of qatal in BH may be considered remnants of a more ancient verbal system, which were eventually superseded. A helpful discussion of the development is provided by Cook (2012, pp. 201–211), who, however, in light of the qatal form’s nonpast meanings in BH, sees the form as basically aspectual, labelling it ‘perfective’. Beyond the aforementioned present tense usages (which, in terms of aspect, are actually imperfective), the use of qatal to signal the future perfect (‘he will have done’) is adequately explained on the assumption that BH verb forms encode relative tense (see Cohen, 2013, pp. 19–20, 57–58; cf. Cook, 2012, p. 202). Note that unlike in Cook’s (2012, pp. 208–210) treatment, weqatal is here considered a separate form from qatal. 7 Bhat, 1999. 8 In agreement with Joosten, 1989; Joosten, 2002; Hatav, 1997, pp. 89–116; Cook, 2012, pp. 223–225; Cohen, 2013, pp. 125–149. 5 28 2.2 Aspect Grammatical aspect broadly refers to the portrayal of an eventuality as either an entirety with endpoints, that is, perfective, or as a less defined process, duration or iteration without reference to endpoints, that is, imperfective.9 Crucially, only in the sphere of the past did BH users routinely select between perfective and imperfective presentations: qatal and wayyiqtol default as perfective; yiqtol, weqatal and the compound haya qotel encode iterative/habitual imperfectivity; and the participle serves to portray durativity/on-goingness. Unsurprisingly, in the present, the choice was merely between shades of imperfectivity: the actual present is generally conveyed using the participle, the generic present with yiqtol and weqatal.10 In the future, though duration and/or repetition characterise many eventualities, the explicit portrayal of future imperfectivity through devices internal to the BH verb system is extremely uncommon. Rather, imperfective interpretation depends on context, adverbials, lexical aspect or some combination thereof. Examples (1)–(4) illustrate the default perfective or undefined aspect of future verbs, as well as the use of adverbs and rare compound structures for the explicit signalling of future imperfectivity. Examples (1)–(4): Yiqtol/weqatal aspect (1) ‫ַת־ספֶר ּולְ כ ָ ָָ֑דּה וְ נ ַ ֶָּ֥ת ִּתי ל֛ ֹו אֶ ת־עַכְ ָ ָ֥סה בִ ִ ֵ֖תי לְ ִא ָ ָּֽׁשה׃‬ ֵ֖ ֵ ‫אֲשֶ ר־יַכֶּ֥ה אֶ ת־קִ ְרי‬ ‘Whoever strikes Kiryath-Sepher and captures it – I will give him Achsah my daughter as a wife’ (Josh. 15:16) (the verbs default as perfective) (2) ‫יֹומם הַ ֶ ָ֥שמֶ ש ָּֽׁלא־י ַָ֗ככָה וְ י ֵ ָָ֥רחַ בַ ָ ָּֽליְ לָה׃‬ ָ ָ֗ ‘By day the sun will not strike you, and the moon at night’ (Ps. 121:6) (context and adverbials lead to imperfective interpretation) (3) ... ‫שׁש הָ עִ וֵרֶׁ֙ בָ ֲא ֵפ ָָ֔לה‬ ֵּׁ֤ ַ‫שׁש ַ ָּֽׁבצָ ה ַ ֳַ֗ריִ ם ַכ ֲא ֶ ֶׁ֙שר יְמ‬ ֵּׁ֣ ַ‫וְ הָ ִ֜ ִּייתָ ְממ‬ ‘and you will (continually; repeatedly) grope at noon as a blind person gropes in the dark …’ (Deut. 28:29) (the compound verbal structure and the generic present context signal imperfectivity) 9 Grammatical aspect is also called viewpoint aspect (Cook 2012, pp. 26–27, 199–201). It contrasts with situation aspect (including lexical aspect or Aktionsart), which is a semantic property of individual verbs (Cook, 2012, pp. 19–25, 194–199), and phasal aspect, which refers to different stages within an eventuality (Cook, 2012, pp. 25–26, 191–194). 10 Though there also exists a gnomic or generic use of qatal (see Cook, 2012, pp. 214–216), presumably explicable as a holdover from a previous TAM system (see above, n. 6). 29 (4) ... ‫מ ֵּׁ֣רים הַ ְשע ִ ָָ֔רים‬ ִּ ְ ֹׁ‫ֲשר ִּ ָּֽי ְהיֵּׁ֤ ּו ִּ ָּֽמטַ ה ֲִּרים֙ ּובָ ִּאים֙ ׁש‬ ֶׁ֙ ֶ ‫וָא ְמ ָ ָ֣רה לַלְ וִ ַ֗ ִים א‬ ‘and I said to the Levites that they should (regularly) purify and come, guarding the gates …’ (Neh. 13:22) (compound verbal structure signals imperfectivity) Crucially, where the inherent meaning of a verb and/or the context do not invite an imperfective reading, mere use of the yiqtol or weqatal forms, that is, those forms traditionally labelled ‘imperfect’, is insufficient for conveying future iterativity or duration. From this perspective, whatever the diachronic course of grammaticalisation or typology paths, the BH verbal system appears to be more tense- than aspect-prominent, though it must be emphasised that the language possesses clear means of marking both tense and aspect. Another argument against a primarily aspectual view of BH has been advanced by Joosten. He notes that The most prominent features attached to the imperfective in recognized aspect languages are the expression of real present and of attendant circumstances in the past. Since neither of these functions is regularly expressed by yiqtol in BH there is no point in classifying yiqtol as imperfective.11 Presumably, this would also apply to Joosten’s view of weqatal, since he views it as a semantically identical syntactic alternative to yiqtol.12 As noted above, the depiction of the actual present and of attendant circumstances in the past is generally achieved through employment of the active participle. These uses of the participle can be explained as functions of either relative present tense or imperfective aspect. 2.3 Mood/modality The realm of BH modality is complicated.13 On the one hand, if the discussion concerns the distinction between indicatives and explicitly marked directivevolitives, such as the jussive, imperative and cohortative, then the BH verbal system clearly cannot be considered mood-prominent, for the simple reason that many of its dedicated means for marking volition – whether morphological, such as the imperative, the jussive (short yiqtol) or the cohortative (lengthened 1st-person yiqtol); lexical, such as the negative ‫ אַ ל‬and the particle ‫ ;נָא‬or syntactic, like yiqtol plus initial word order – are not consistently employed in apparently volitive contexts and also not infrequently crop up where the context seems to call for indicatives. Many of the more important means of directive-volitive marking seem to have been in flux. From the limited perspec11 Joosten, 2002, p. 53. Joosten, 1992, p. 13. 13 See Robar’s contribution to the present volume. 12 30 tive of the directive-volitional dimension of deontic modality, such inconsistency and irregularity militates against viewing BH as a mood-prominent language, at least in so far as it has been preserved.14 Where things are less clear is in the broader domain of what might be termed ‘unmarked deontic modality’. It is well known that in BH the futureoriented forms yiqtol and weqatal frequently convey various nuances related to the freedom to act and the imposition of one’s will, ranging from permission through obligation to declarative future and variously corresponding to English may, should, must and will. Indeed, it is sometimes difficult to distinguish between indicative future and such modal nuances. While there are theoretical/philosophical views, supported by both cross-linguistic evidence and diachronic typology paths, according to which it is argued that future-oriented utterances should by their nature be deemed modal,15 there are arguably valid reasons for doubting whether this should necessarily be seen as a linguistic universal.16 While an important issue, however, it is beyond the scope of the present study to pursue it any further. Even so, it seems of critical import that it is generally in the domain of future-oriented eventualities that modality comes into play in BH. This is significant because in a mood-prominent language one would expect the explicit marking of modality across the full range 14 Cf., for example, Dallaire (2014, pp. 3–4, 122, 125–141), who, on the basis of comparative Semitic evidence, concludes that full (i.e. non-short) yiqtol forms with volitive meanings represent not the encroachment of indicative yaqtulu into the modal domain otherwise reserved for jussive yaqtul, but the existence in BH of yaqtula, an ostensible volitive form whose Tiberian reflex, yiqtol, can no longer be distinguished from that of indicative yaqtulu. One of Dallaire’s concerns is gratuitous textual emendation, in which full yiqtol forms in volitive contexts are ‘corrected’ to short forms. However, as Hornkohl (in press) contends, scribal mistakes are not the only alternative to an explanation of the volitive use of full yiqtol assuming derivation from yaqtula. Genuinely ancient expansion of full yiqtol at the expense of short yiqtol is also a possibility. 15 See, e.g., Hatav, 1997, p. 29; Joosten, 1997, p. 58; Joosten, 2012, p. 33; Penner 2015; cf. Hornkohl, 2016, pp. 306–307. 16 Just as some of the parameters of tense and aspect are often conflated (e.g. the notions of the future tense’s ‘yet to happen-ness’ and the imperfective aspect’s ‘non-completive-ness’), so some conceptions of modality appear to confuse dimensions that it seems proper to keep separate, e.g. English would to denote past habituals versus English modal would to denote future in the past (see Cook, 2012, pp. 142–143). Despite the notional attraction of theories that subsume certain temporal and/or aspectual dimensions under modality, it is problematic to prejudge indicative declarations about the future or about repeated actions in the present or past to be logical and linguistic impossibilities (cf. Joosten, 1997, p. 58 on future and Joosten, 2012, p. 26 on imperfective past). It is instructive that though explaining the habitual/iterative force of yiqtol and weqatal (and presumably haya qotel) as a function of mood rather than aspect, Joosten (2002, p. 62) recognises that “iteration is not itself modal”. Cf. also Bhat (1999, pp. 175–178), who contends that the TAM classification of, inter alia, future and imperfective past should be done on a language-by-language basis with reference to the way in which each language patterns in terms of TAM-prominence. In other words, a function rightly classified as modal in a mood-prominent language may legitimately be viewed as tense- or aspect-based in a language in which mood is less prominent. 31 of tenses and/or aspects, not in one alone.17 As a response to the claim that BH yiqtol and weqatal are fundamentally modal, it seems fair to ask why their unambiguous modal uses are restricted to the realm of future-oriented eventualities.18 Put somewhat simplistically, in a mood-prominent version of BH one would expect the indicative forms, that is, qatal and wayyiqtol, to function regularly for declarative certainties regardless of an eventuality’s temporal relation to speech time, and the modal forms, that is, yiqtol and weqatal, to encode volition, uncertainty and/or irrealis/conditional/hypothetical/counterfactual status across all tenses. But this situation does not obtain in BH. Shades of directive-volitive modality are (inconsistently) discernible only in the realm of eventualities temporally posterior to speech or reference time; nuances of deontic modality cannot generally be perceived on the basis of details internal to the verb system; and non–future-oriented conditionals and the like are typically not encoded with yiqtol or weqatal. 2.4 The importance of a balanced, data-centric approach to TAM Whatever the best explanation for how exactly BH manages to cover the range and combinations of TAM values given the relatively few forms and structures at its disposal, the temptation to oversimplify must be resisted. Valid, sufficiently comprehensive accounts are unlikely to be of the sort that reduce the BH verb system to a unidimensional dichotomy or that exaggerate the prominence of a single parameter. One of the problems with accounts that focus on the system’s historical development, attempting to arrive at its basic or original TAM essence, is that students are often left with an under-appreciation of the ways in which the system developed to handle multiple TAM values, and combinations thereof, using relatively limited grammatical resources.19 17 See Bhat (1999, pp. 130–140), according to whom the degree to which languages are classified as mood-prominent should be measured in terms of such criteria as grammaticalisation, obligatoriness, systematicity and pervasiveness of mood. Not all of these dimensions may be reliably traced in the history of ancient Hebrew. But, as in the case of aspectual marking, it would seem that modal marking in BH – directive-volitional as well as the ‘unmarked deontic’ variety mentioned above – lacks the obligatoriness, systematicity and pervasiveness of tense marking. Beyond the restriction of clear modality to temporally posterior eventualities and the use of the ostensibly modal forms for TAM values that may be otherwise explained (e.g. future, imperfective past), the split between first conditionals, on the one hand, and second and third conditionals, on the other, is to be observed. Crucially, conditionals with protases referring to realisable eventualities employ yiqtol after ‫( ִאם‬Gen. 18:26, 28:20, etc.), whereas irrealis hypotheticals use qatal after ‫( לו‬Num. 14:2 (2x), 20:3, Deut. 32:29, Josh. 7:7, Judg. 8:19, 13:23; significantly, yiqtol after ‫ לו‬serves to mark non-past-oriented wishes and doubts: Gen. 17:18, 30:34 (jussive), 50:15). In other words, the more realis eventuality is expressed by means of the allegedly modal form, while the clearly irrealis eventuality is expressed with what is otherwise considered the indicative form. 18 As argued above, the past imperfective use of yiqtol and weqatal is not unambiguously modal. 19 See Buth, 1992. For the sake of comparison, it is worth pointing out that the ‘primitive’ BH verbal system, with its relatively meagre number of forms for indicating TAM, possesses a 32 Another problem is the imposition of perspectives arguably foreign to the system. While cross-linguistic probabilities, comparisons with other languages (especially Semitic cognates) and typological development paths can provide helpful evidence, it is necessarily circumstantial. Analysts must rigorously and transparently chart the verbal semantic values of large numbers of unambiguous examples along all three TAM axes without undue influence from reductionist approaches, no matter their philosophical attraction, crosslinguistic support or basis in grammaticalisation theory and/or diachronic typology. Though subject to criticism on certain points, Penner’s recent statistical analysis of TAM in the Dead Sea Scrolls is a welcome contribution with promising potential as a model.20 Only by means of such studies are we likely to minimise subjectivity and theoretical bias in our characterisation of the BH verbal system, and, in so doing, to carve out a place in which BH linguistic investigation not only benefits from the wider world of linguistics, but contributes meaningfully thereto. 3 Verbal semantics, syntax and word order Before moving on, it is worth pausing to clarify why understanding of BH verbal TAM semantics is so central to a discussion of constituent order. Two points are crucial. First, it emerges that the four principal finite verbal forms have fixed semantic values, and also that they occur in sets of semantically equivalent alternative pairs: wayyiqtol and qatal are past, perfective, indicative, whereas weqatal and yiqtol are either future/modal, with generally perfective/undefined aspect, or have habitual/iterative aspect in past or present.21 Second, from a syntactic perspective, the members of each pair occur in complementary distribution, the waw-initial forms coming clause-initially, those without waw – with the exception of directive-volitive yiqtol22 – being reserved for syntactic environments in which a clause-initial element precludes use of the waw-initial form.23 But we must go further than these observations, which, though accurate, have little explanatory value. The apparent systemic redundancy of semantically equivalent forms that alternate depending on word order demands an explanation. One thing is certain: no analysis confined to greater variety of morphological resources than Modern Israeli Hebrew, speakers of which, nevertheless, manage passably to communicate TAM distinctions. 20 Penner, 2015; cf. Hornkohl, 2016. 21 Buth, 1992, p. 104. 22 Whether or not they exhibit distinct volitive morphology, it is notable that clause-initial (we)yiqtol forms usually have explicit volitional, purposive or final force. 23 Notably absent from this description is we+unconverted qatal, which is best seen as marginal within the BH verbal system, principally restricted to hendiadys, poetry and post-exilic BH (especially Qohelet, whereas most LBH books chiefly preserve the classical system). See Hornkohl, 2014, pp. 254–266, 287–293; cf. Longacre, 1994, pp. 83–84; Robar, 2014, pp. 152–159. I am grateful to my friend and former teacher, Randall Buth, for enlightening discussions on this topic. 33 the traditional domains of syntax and semantics promises to furnish a satisfactory account of the BH verbal system. This is where pragmatics has much to contribute. 4 Pragmatics and word order Over the years, scholars have discerned consistent, communicatively significant meanings and effects beyond the basic semantics of individual verb forms and clause-level syntax, meanings and effects that operate at a higher, discourse dimension. 4.1 Information structure Thus, long ago scholars noted that the qatal and yiqtol forms not preceded by waw often followed constituents that bore special ‘emphasis’.24 4.1.1 Focus Consider examples (5)–(10), in which the pre-verbal element in the second clause supplies the answer to a content question posed in the first. Examples (5)–(10): Fill-in focus with wh-questions (5) ‫ׁשר ל ֹׁא־יָּֽעָ ׂ֔שּו ע ִ ֵָ֖שיתָ עִ מָ ִ ָּֽׁדי׃‬ ֵּׁ֣ ֲ‫ מַ ע ֲִּשים֙ א‬... ֶׁ֙‫ָּֽמה־ע ִ ִָׂ֤שיתָ ָָּ֙לנו‬ ‘What have you done to us? … Deeds that are not done you have done with me!’ (Gen. 20:9) (6) ... ‫חֹוקה ְמאֹׁ ד֙ ָבָ֣או עֲבָ ָ֔ ֶדיָך‬ ֵּׁ֤ ָ ‫ מ ֙ארץ ְר‬... ‫ִ ָ֥מי אַ ֶ ֵ֖תם ּומ ַ ֶּ֥איִּ ן תָ ָּֽׁבאו׃‬ ‘Who are you and where are you from? … From a very distant land your servants have come.’ (Josh. 9:8b–9a) (7) ... ‫הּודה ַיעֲלֶ ֶ֑ה‬ ֵּׁ֣ ָ ְ‫ י‬... ‫ל־הכְ ַנע ֲִנ֛י בַ ְתחִ לָ ֵ֖ה לְ הִ לָ ָ֥חֶ ם ָּֽׁבֹו׃‬ ָּֽׁ ַ ֶ‫ִּ ֵּׁ֣מי ַי ֲעלֶה־לָ ָּ֧נו א‬ ‘Who will go up for us to the Canaanites first to fight against them? … Judah will go up.’ (Judg. 1:1–2) (8) ‫ָשה הַ דָ ָבָ֥ר הַ ֶזָּֽׁה׃‬ ֵ֖ ָ ‫ן־יֹואׁש ע‬ ָ ׂ֔ ‫ גִּ ְדעֹון֙ ב‬... ‫ָשה הַ דָ ָבָ֣ר הַ זֶ ֶ֑ה‬ ֵ֖ ָ ‫ִּ ֶּ֥מי ע‬ ‘Who did this thing? … Gideon son of Joash did this thing.’ (Judg. 6:29) 24 See, e.g., Gesenius, 1910, §142a; Muraoka, 1985, pp. 1–46; Joüon and Muraoka, 2006, §§146a, 155na–nb, nh. 34 (9) ... ‫ת־ׁש ְמׁשֹון֙ ָעלִָ֔ ינו‬ ִּ ‫ לאֱ ֵּׁ֤סֹור א‬... ‫יתם עָלֵ ֶ֑ינו‬ ָ֣ ֶ ִ‫ל ָ ָָ֖מה עֲל‬ ‘Why have you come up against us? … To capture Samson we have come up.’ (Judg. 15:10) (10) ‫יתָך‬ ֶ֑ ֶ ֵ‫ ָ ֶּ֥מה ָר ֵ֖או בְ ב‬... ‫חֹוקה ָבֵ֖או‬ ָ֛ ָ ‫ מ ֶ֧ארץ ְר‬... ‫ָ ָ֥מה אָ ְמ ָ֣רו׀ הָ ֲאנ ִ ָָ֣שים הָ ַ֗ ֵאלֶה ּומאַ ִּ֙ין֙ י ָָ֣באו אֵ ֶָ֔ליָך‬ ‫יתי֙ ָר ָ֔או‬ ִּ ‫ׁשר ְבב‬ ֵּׁ֤ ֲ‫ ֵּׁ֣את כָל־א‬... ‘What did these men say to you and where have the come from? … From a distant land they have come … What have they seen in your house? … Everything in my house they have seen.’ (2 Kings 20:14– 15a) Things have come a long way since scholars were satisfied with a term and explanation as vague as ‘emphasis’. Nowadays, such cases are usefully explained with reference to the notion of information structure, according to which the pre-verbal constituents in the foregoing examples are considered to be marked for focus. While variously defined, for purposes of the present study focus is understood as marked rhematic (or comment) material or as that element designated by explicit marking as the most salient, or newsworthy, piece of information in the clause, often for purposes of contrast, fill-in, identification, contra-expectation or reinforcement.25 4.1.2 Topic However not all pre-verbal, apparently emphasis-bearing constituents can properly be considered focal in this way. In many cases, such an element is highlighted as a topic, a marked point of reference for ensuing information.26 Consider examples (11) and (12). Examples (11)–(12): Topic (11) ‫ת־המַ חֲנֶ ֶ֑ה ּובַ בָֹׁ֗ קר ָ ָּֽׁהיְ תָ הֶׁ֙ ִשכְ ַבָ֣ת הַ ָ֔ ַטל סָ ִ ֵ֖ביב ַ ָּֽׁלמַ ח ֲֶנָּֽׁה׃‬ ָּֽׁ ַ ֶ‫וַיְ ִ ָ֣הי בָ ֶָ֔ע ֶרב ו ַ ַָ֣תעַל הַ ְש ָָ֔לו ו ְַתכַ ֵ֖ס א‬ ‘And it was in the evening and quail came up and covered the camp. And in the morning a layer of dew was around the camp.’ (Ex. 16:13) 25 For various formulations of this definition see Buth, 1995, p. 84; Buth, 1999, p. 81; Shimasaki, 2002, p. 42; Holmstedt, 2009, pp. 126–129. See Hornkohl, 2005, pp. 31–36 for examples, and also the careful discussion in Moshavi, 2010, pp. 90–97, 121–143. Alternatively, some use the term to refer to the new information provided by a sentence, regardless of markedness (Lambrecht, 1994, p. 206). Traditionally, this was called the predicate, but since today that term is generally considered to have a purely syntactic definition (in contrast to ‘subject’), some now refer to the ‘psychological predicate’; other terms include ‘comment’, ‘rheme’ and ‘focus’ (as opposed to capitalised ‘Focus’ in Functional Grammar, which is used in reference to marked rhematic information). 26 Buth, 1995, p. 84. Moshavi (2010, pp. 97–103, esp. 101–102) provides an instructive discussion, showing that marked topics in BH can relate to the immediately preceding context or the immediately following context or can signal relations between text segments, like discourse connectives. See also Khan, 1988, p. 86–87. 35 (12) ... ‫ ּולְ אָ ָ ֵּׁ֣דם אָ ַ֗ ַמר‬... ‫שה אָ ַ֗ ַמר‬ ֵּׁ֣ ָ ‫ ָּֽאל־הָ ִּא‬... ‫ֹלהים ׀ ֶ ָּֽׁאל־הַ נָחָ ש‬ ָ֥ ִ ‫וַיאמֶ ֩ר יְ ה ֶׁ֙ ָוה ֱא‬ ‘And Yhwh God said to the serpent … To the woman he said … And to the man he said …’ (Gen. 3:14, 16–17) This is common with scene-setting prepositional phrases, as in example (11), where ‫ ובַ ַ֗בקֶ ר‬unambiguously orients the reader to the temporal setting of what is about to be related. But it also occurs with other types of constituents, as in example (12), where the salient information in each clause is communicated not by means of the pre-verbal prepositional phrases used in reference to two of the addresses, but in the ensuing curses. The fronted prepositional clauses help merely to organise the information relevant to each addressee, serving to juxtapose the members of the set. 4.1.3 Contrastive focus vs contrastive topic The difference between focus and topic deserves further elaboration, because the two phenomena are sometimes confused, particularly when they are used for purposes of contrast. Consider the cases of contrastive focus in examples (13)–(15). Examples (13)–(15): Contrastive focus (13) ... ‫ָּֽׁלא־ ִיקָ ֵר ֩א ִש ְמָךֶׁ֙ ֜עֹוד ַיעֲקָֹׁ֗ ב ִ ִׂ֤כי ִאם־יִּ ְש ָרא ֙ל יִ הְ יֶ ָ֣ה ְש ָ֔ ֶמָך‬ ‘Your name will no longer be called Jacob, but Israel will be your name’ (Gen. 35:10b) (the first clause exhibits standard predicate focus, while in the second the verbal complement is marked via fronting for argument focus as the salient piece of – in this case, contrastive – information) (14) ‫ֵיהם‬ ָּֽׁ ֶ ‫ִ ָ֣כי ִׂ֤לא אָֹּֽׁ ְתך֙ מָ ָ֔ ָאסו ִ ָּֽׁכי־אֹׁ ִּ ֶּ֥תי מָ ֲא ֵ֖סו ִמ ְמֹלָ֥ ְך ֲעל‬ ‘For not you have they rejected, but me they have rejected from ruling over them’ (1 Sam. 8:7b) (in both clauses focus via fronting of the direct objects conveys counterexpectation) (15) ‫ ֶּ֥הּוא‬... ‫ֲשר יִ הְ יֶ ֵ֖ה ִמבָ נֶ ֶ֑יָך‬ ָ֥ ֶ ‫ימֹותי ֶ ָּֽׁאת־ז ְַרעֲָךֶׁ֙ אַ ח ֶ ֲָ֔ריָך א‬ ִׂ֤ ִ ‫ ַוה ִ ֲָּֽׁק‬... ‫ֶה־לי הַ ַבֵ֖יִ ת ל ָ ָָּֽׁשבֶ ת‬ ָ֥ ִ ‫ָ֥לא אַ ָ ָ֛תה ִתבְ נ‬ ... ‫ֶה־לי ָבֶ֑יִ ת‬ ֵ֖ ִ ‫יִ בְ נ‬ ‘You will not build me the house for dwelling … but I will establish your seed after you who will be from your sons … He will build me a house.’ (1 Chr. 17:4, 11–12) (independent subject pronouns are unnecessary in BH; when used they almost obligatorily precede the verb for which they serve as subject, often, as here, marking contrastive focus) 36 In the foregoing examples, each pre-verbal constituent is marked as the salient point of contrast between two eventualities. This is not be confused with contrastive topic, such as in examples (16)–(20), where the pre-verbal constituent in each case does not itself constitute the salient point of contrast, but is one of a set of entities held up for comparison with respect to the salient point of contrast (that is, the newsworthy or previously unknown information), which is conveyed by the verbal predicate or some element therein.27 Examples (16)–(20): Contrastive topic (16) ‫ֹלהים׀ לָאֹור֙ יׂ֔ ֹום וְ ל ַֹ֖חשֶׁ ְך ָ ָ֣ק ָרא ָ ָ֑ליְ לָה‬ ִׂ֤ ִ ‫ַו ִיקְ ָ ֶׁ֙רא ֱא‬ ‘God called the light day, whereas the dark he called night’ (Gen. 1:5) (the ‘new’ information (in bold) consists of the respective names bestowed upon the light and darkness; ‘and the dark’ is fronted as a marked topic to provide a reference point for the comparison (in italics)) (17) ‫א֙ ִ֛תי ה ִּ ֶּׁ֥שיב עַ ל־כ ִּ ַָ֖ני וְ א ֥תֹו תָ ָ ָּֽלה׃‬ ‘Me he restored to my office, whereas him he hanged’ (Gen. 41:13) (the fronted direct objects (in italics) are topics in reference to which the main points of contrast, the difference in treatment in standard predicate focus (in bold), are related) (18) ‫יֹוסף אֶ ת־אֶ ָחֶ֑יו וְ הֹ֖ם ֶּ֥ל ֹׁא הִ כִ ָּֽׁרהו׃‬ ֹ֖ ‫ַויַכֶּ֥ר‬ ‘And Joseph recognised his brothers, but they did not recognise him’ (Gen. 42:8) (the salient point of contrast is the brothers’ non-recognition of Joseph; the fronted subject pronoun is not obligatory, but here serves as a marked topic to heighten the contrast) (19) ‫׃‬... ‫וְ נִּ ֙ ַגׁש מ ֶׁ ֶׁ֤שה לְ בַ ּדֹו֙ אֶ ל־יְ ה ָ֔ ָוה וְ הֹ֖ם ֵּׁ֣ל ֹׁא יִ גָ ֶ֑שו וְ הָ ָ֕ ָעם ֶּ֥ל ֹׁא ַיעֲלָ֖ ּו‬ ‘And Moses alone will approach Yhwh, but they will not approach, and the people will not ascend …’ (Ex. 24:2) (the principal difference is in the actions, the fronted subjects serving to bring the differences into clearer relief) (20) ‫ֲמֹותּה וְ ֹ֖רּות ָ ֶּּ֥ד ְבקָ ה ָ ָּֽׁבּה׃‬ ָ ָ֔ ‫שק עָ ְרפָ ה֙ ַלח‬ ֵּׁ֤ ַ ‫ו ִַּת‬ ‘And Orpah kissed her mother-in-law, but Ruth clung to her’ (Ruth 1:14) (the women are compared with respect to their actions toward their mother-in-law) 27 The identification of such pre-verbal constituents as topics is in agreement with van der Merwe, Naudé and Kroeze, 2000, p. 348 and Moshavi, 2010, pp. 98–100, 144–146; cf. Holmstedt, 2009, pp. 127–129. Buth (1994a, p. 223) views such cases of contrastive topics simultaneously as instances of marked topic and marked focus. 37 4.2 The inadequacy of information structure There is undeniable value in recognising within BH word order variation functions known from information structure. However, in terms of explanatory power, information structure alone proves inadequate, as there are a great many cases of XV order (where X stands for any clausal argument or adjunct, and V for the verb) to which the functions of focus and topic do not obviously apply. Indeed, according to Moshavi’s statistics for the prose sections of Genesis, in just 56.6 per cent of the cases can a pre-verbal constituent be so explained.28 4.3 Basic subject-verb word order? 4.3.1 Arguments for basic subject-verb word order in BH The pre-verbal placement of one category of constituents, namely subjects, might conceivably be accounted for on the assumption that the BH verbal clause has a basic subject-verb (SV) word order, an opinion that has been held by a minority of scholars past and present.29 On this view, SV linearisations, by dint of representing the default, neutral order, require no explanation. Holmstedt has put forth the most lucid and comprehensive case for just such a view.30 Focusing on a corpus consisting of main clauses in Genesis (though he has also applied his method to Jonah, Ruth and Proverbs), his approach is fundamentally frequency-based, but refined according to the criteria of syntactic distribution, clause type and pragmatic marking. This filtering is indispensable, since in any given language the numerically dominant – and thus apparently basic – order might have a limited syntactic distribution,31 obtain only in a specific clause type or types32 and/or exhibit pragmatic markedness.33 On the basis of syntactic distribution and clause type, Holmstedt excludes 28 Moshavi, 2010, p. 119. And this, despite the fact that Moshavi’s definition of topic extends beyond many traditional definitions (see n. 26 above). 29 Joüon, 1923, §155k; Greenberg, 1965, §27.4; DeCaen, 1999, p. 118 n. 22; Gross, 1999, p. 30 n. 46; Holmstedt, 2003; Holmstedt, 2005; Holmstedt, 2009; Holmstedt, 2011; Holmstedt, 2016, pp. 46–49; Cook, 2012, pp. 235–236. 30 Holmstedt, 2011. 31 Holmstedt, 2011, pp. 7–13 (see also Holmstedt, 2009, pp. 117–119). To be precise, Holmstedt uses the criterion of distribution not to establish basic word order in BH, but to justify his excluding from consideration clauses with wayyiqtol, with its virtually obligatory clause-initial position, which, in his view, “distorts” (2009, p. 120) or “skews” (2011, p. 17 n. 35) the word order profile of much of the Hebrew Bible. Of course, this is true of wayyiqtol only if SV order really is basic in BH. If, on the other hand, basic VS order can be established (without consideration of wayyiqtol clauses), it is reasonable to conclude that the VS order of wayyiqtol clauses is simply one among several manifestations of the basic, unmarked order, rather than a distortion thereof. 32 Holmstedt, 2011, pp. 13–20 (see also Holmstedt, 2009, pp. 116–117). Again, Holmstedt uses this criterion not to establish basic word order, but to argue that wayyiqtol clauses should be filtered out of the data to be considered when measuring frequency. 33 Holmstedt, 2011, pp. 20–25 (see also Holmstedt, 2009, pp. 119–120). 38 wayyiqtol clauses from consideration, as the form is essentially confined to a single syntactic environment – the verb-initial clause – and is used almost exclusively in one clause type – indicative, past temporal, narrative clauses. Further, since BH, like other languages, frequently displays word order variation motivated by pragmatic factors, clauses pragmatically marked for topic and focus are winnowed, resulting in sets of what Holmstedt considers pragmatically neutral SV and VS clauses. According to Holmstedt’s application of the methodology to eligible clauses in Genesis, pragmatically neutral SV clauses (of which he counts 47) outnumber their VS counterparts (of which he counts 2634) by a ratio of nearly 2:1.35 Holmstedt stops short of citing such results as proof of basic SV order in the BH verbal clause, but he understandably sees them – along with the acknowledged basic subject-predicate order of the BH verbless clause – as fatally problematic for the basic VS view. 4.3.2 Problems with the view that basic word order in BH is SV 4.3.2.1 SV order and pragmatic markedness The methodology just summarised is arguably sound. Its insistence on linking basic word order to pragmatic neutrality is especially laudable. Holmstedt’s study is reasonable, well-documented and compellingly argued. Its conclusions are so fundamental and far-reaching that ignoring them is out of the question. And there is much of value in the study, both in terms of the questions it addresses, the weaknesses it cites in competing arguments, the supporting arguments that it furnishes for its central claim and, most importantly, the methodology it proposes. Be that as it may, at certain points Holmstedt’s application leads one to doubt the reliability of the results obtained and, consequently, the conclusions drawn therefrom. Most crucially, its central conclusion, that a basic SV word order for the BH verbal clause better accounts for the data than a basic VS order, proves unconvincing for the simple reason that, if rerun in strict accordance with the methodology that Holmstedt rightly advocates, the experiment produces very different results. For a provisional presentation of such results, see table 1. The remainder of this paper consists of examples and discussion, including the data behind the table. 34 The figure does not include wayyiqtol, negative or volitive clauses, which Holmstedt excludes on various grounds. 35 On three occasions Holmstedt counts a single example of SV word order in a verse arguably containing two cases: Gen. 15:17 ‫ הַ שֶ ֶֶׁ֙משֶׁ֙ ָ֔ ָבאָ ה‬and ‫ ַו ֲעל ָ ֵָ֖טה הָ יָ ֶ֑ה‬, 43:23 ‫ָּֽׁאֹלהִׂ֤י א ֲִביכֶםֶׁ֙ נ ֶַָׁ֙תן‬ ֵ ‫ א ֱָֹּ֙להֵ י ֶ֜כם ֵו‬and ‫כ ְַספְ כֶ ֵ֖ם ָבָ֣א‬, 44:20 ‫ וְ אָ ִ ֶׁ֙חיו ֜ ֵמת‬and ‫וְ אָ ִ ָ֥ביו אֲהֵ ָּֽׁבֹו‬. Presumably, the ‫ ע״ו‬forms were interpreted as participles (though ‫ ָ֔ ָבאָ ה‬in Gen. 15:17 is accented as a qatal form). Gen. 19:23 has two cases of SV word order and is included in both of Holmstedt’s lists, but it is not clear which case is intended in each list. Also, Gen. 50:23, with just one SV clause, is included in both the marked topic list (Holmstedt, 2011, pp. 23–24 n. 53) and the unmarked SV list (Holmstedt, 2011, p. 24 n. 54). 39 Table 1. Holmstedt’s results and the results from re-running the test Holmstedt’s statistics Statistics from re-run test marked SV (topic): 11236 unmarked SV (no topic or focus): 4738 unmarked VS: 2640 marked SV (topic / focus): 62 / 937 SV unmarked for topic or focus: 1139 unmarked VS: 23 From a theoretical standpoint, under the assumption of a pragmatically neutral SV order, it is natural to suppose that pre-verbal positioning, that is, fronting, would entail pragmatic marking only in the case of non-subject constituents, the pre-verbal slot being default for subjects. The pragmatic marking of subjects would presumably necessitate alternative strategies, such as intonation, focalising particles and special structures. As Holmstedt shows, however, there is strong correlation between pre-verbal subjects and topic marking. In- 36 Holmstedt, 2009, p. 23 n. 53. The discrepancy between the figures in this line is due primarily to the fact that many of the cases of SV order in which Holmstedt identifies a topic are classified here as marked for some other pragmatic value. In other words, it is maintained here that the element fronted in each of the following does not convey the clause- or discourse-level ‘about-ness’ of a genuine topic, but some other argument-centred value (i.e. focus), or, more frequently, discourse-level marking of the entire clause for purposes of pragmatic discontinuity (on which concept see below, section 4.4.3 and n. 64): Gen. 2:6 (‫ ;וְ אֵ ד ַי ֲעלֶה‬sentence focus), 4:1 (‫)וְ הָ אָ דָ ם יָדַ ע‬, 7:19 (‫)ג ְָברו וְ הַ מַ יִ ם‬, 8:5 (‫)וְ הַ מַ יִ ם הָ יו‬, 9:2 (‫ומֹור ֲאכֶם וְ ִח ְתכֶם יִ ְהיֶה‬ ַ ), 13:14 (‫)וַיהוָה אָ מַ ר‬, 14:3 (‫)כָל־אֵ לֶה חָ ְברו‬, 17:16 ( ‫מלְ כֵי ע ִַמים‬ ‫ ; ִממֶ נָה יִ הְ יו‬focus), 18:17 (‫)וַיהוָה אָ מָ ר‬, 19:15 (‫)וכְ מֹו הַ שַ חַ ר ָעלָה‬, 19:23 (whether ‫ הַ שֶ מֶ ש יָצָ א‬or ‫ וְ לֹוט בָ א‬is intended), 19:24 (‫)וַיהוָה ִה ְמ ִטיר‬, 21:1 (‫)וַיהוָה ִה ְמ ִטיר‬, 21:7 (sic; read 21:6, ‫)כל־הַ שמֵ ַע יִ צְ חַ ק‬, 22:1 (‫)וְ הָ אֱֹלהִ ים נִ סָ ה‬, 24:1 (two cases, ‫ וְ אַ ְב ָרהָ ם זָקֵ ן‬and‫)וַיהוָה בֵ ַרְך‬, 24:35 (‫)וַיהוָה בֵ ַרְך‬, 24:40 (‫יְ הוָה אֲשֶ ר־‬ ‫)הִ ְתהַ לַכְ ִתי לְ ָפנָיו יִ ְשלַח‬, 24:56 ( ַ‫)וַיהוָה ִהצְ לִ יח‬, 25:34 (‫)וְ ַיעֲקב נָתַ ן‬, 26:26 (‫) ַואֲבִ ימֶ לְֶך הָ לְַך‬, 27:6 (‫)וְ ִר ְבקָ ה אָ ְמ ָרה‬ 27:30 (‫ ;וְ עֵשָ ו אָ חִ יו בָ א‬sentence focus), 28:3 (‫)וְ אֵ ל שַ דַ י יְ בָ ֵרְך‬, 29:9 (‫( ;וְ ָרחֵ ל בָ אָ ה‬sentence) focus), 31:5 (‫)וֵאֹלהֵ י אָ בִ י הָ יָה‬, 31:19 (‫)הָ לְַך וְ לָבָ ן‬, 31:29 (‫ ;וֵאֹלהֵ י א ֲִביכֶם אֶ מֶ ש אָ מַ ר‬of questionable relevance because of the double fronting), 31:34 (‫)וְ ָרחֵל לָקְ חָ ה אֶ ת־הַ ְת ָרפִ ים‬, 34:5 (‫)וְ ַיעֲקב שָ מַ ע‬, 34:7 (‫)ובנֵי ַיעֲקב בָ או‬, ְ 35:10 (‫ ;יִ הְ יֶה ְשמֶ ָך יִ ְש ָראֵ ל‬focus), 35:11 (‫ ;גֹוי וקְ הַ ל גֹויִ ם יִ ְהיֶה ִממֶ ָך‬focus), 37:3 (‫)וְ יִ ְש ָראֵ ל אָ הַ ב‬, 37:33 (‫ָרעָה חַ יָה‬ ‫( ; ֲא ָכל ְָתהו‬sentence) focus), 37:36 (‫)וְ הַ ְמדָ נִ ים מָ כְ רו‬, 41:10 (‫)פ ְַרעה קָ צַ ף‬, 41:16 (‫ֱֹלהים ַי ֲענֶה‬ ִ ‫ ;א‬focus), 41:56 (‫)וְ הָ ָרעָב הָ יָה‬, 41:57 (‫)וְ כָל־הָ אָ ֶרץ בָ או‬, 44:19 (‫)אֲדנִ י שָ אַ ל‬, 46:31 (‫ ; ְבאֶ ֶרץ־כְ ַנעַן בָ או אחַ י ובֵ ית־אָ ִבי אֲשֶ ר‬sentence focus?), 47:5 (‫ ;אָ ִביָך וְ אַ חֶ יָך בָ או‬sentence focus?), 50:23 (‫ן־מנַשֶ ה ילְ דו גַם ְבנֵי מָ כִ יר‬ ְ ֶ‫ ;ב‬focus). The instance Gen. 5:29, with a single SV clause, is included by Holmstedt in his list of SV clauses with marked topic as well as in that containing unmarked SV clauses. Having a pronominal subject, its eligibility for consideration according to Holmstedt’s criteria is questionable; see below, n. 44. It is also arguably explicable as an instance of focus; see below, n. 47. As such, it has been excluded from the count. Gen. 3:11 appears to be a mistake for Gen. 3:1, while Gen. 21:26, 44:8 and 49:9 seem to have been listed by mistake. It is to be emphasised that the classification of pragmatic effects involves the subjective judgement of individual readers, so that differences of opinion are not unexpected. The important point is that scholars be furnished with notional frameworks in which to compare approaches to individual clauses. 38 Holmstedt, 2009, p. 24 n. 54. 39 The discrepancy between the totals in this line is due to two factors. First, adhering strictly to Holmstedt’s criteria, 32 of the SV clauses he considers unmarked for information structure ought to be considered ineligible for consideration due to other factors; see the discussion below and n. 44. Additionally, in some nine cases (four of which are not excluded on other grounds), the SV order appears to involve either focus or topic; see below, examples (21)–(29). 40 Holmstedt, 2009, p. 25 n. 55. 37 40 deed, he finds marked SV cases (112) more than twice as common as unmarked SV cases (47).41 Intuitively, the fact that SV order, like XV more generally, correlates so highly with pragmatic markedness would seem to militate against seeing SV order as basic in BH. 4.3.2.2 Alleged unmarked cases of SV word order But a more serious problem arises when one rigorously applies the proposed criteria for establishing basic word order to Holmstedt’s ostensibly neutral clauses. His discussion of clause type42 explicitly cites Siewierska’s oftquoted dictum: Within the context of typological studies the term ‘basic order’ is typically identified with the order that occurs in stylistically neutral, independent, indicative clauses with full noun phrase (NP) participants, where the subject is definite, agentive and human, the object is a definite semantic patient, and the verb represents an action, not a state or an event.43 She then goes on to note that this basic order “need not be the dominant linearization pattern in a given language”. The problem is that, according to these guidelines, the majority of Holmstedt’s 47 neutral SV cases must be disqualified. Some 32 involve a stative, passive or middle verb; a subject that is nonhuman, non-agentive or both; or some combination of the above;44 by comparison, just three of his 26 pragmatically neutral VS clauses are excluded on 41 Holmstedt, 2011, pp. 23–24 n. 53. This figure includes neither clauses with pre-verbal pronominal subjects, which Holmstedt justifiably sees as pragmatically marked (though an argument could be made that these should be counted, since it might well be markedness by means of fronting that leads to the use of the independent subject pronouns), nor modal clauses with pre-verbal subjects, in which category Holmstedt (following others) includes negative clauses (2011, p. 20 n. 44; cf. Bhat, 1999, pp. 178–179). Interestingly, Holmstedt identifies no cases of marked SV word order with the subject marked for focus. In n. 37 above, at least five of Holmstedt’s cases have been classified as marked for argument focus, but some of these are debatable. Consider, by way of example: ‫‘ ָּֽׁלא־יִ קָ ֵר ֩א ִש ְמָךֶׁ֙ ֜עֹוד ַיע ֲַ֗קב ִ ִׂ֤כי ִאם־יִּ ְש ָראל֙ יִ ְהיֶ ָ֣ה ְש ָ֔ ֶמָך‬no longer will your name be called Jacob, but Israel will be your name’ (Gen. 35:10). Here the fronted proper noun is arguably construable as a focalised subject, but is probably better understood as the complement of the verb ‫‘ יִ ְהיֶה‬will be’, the subject being ‫‘ ְשמֶ ָך‬your name’, as in the preceding clause. As Buth (1999, pp. 100–101) opines, the subject of a clause is normally the more definite and the more presupposed. While ‘Israel’ here is more definite than ‘your name’, in the immediate context ‘your name’ is far more presupposed as the issue at hand, and is thus the better candidate for clausal subject. 42 Holmstedt, 2011, p. 13. 43 Siewierska, 1988, p. 8. 44 Gen. 1:2 (‫)וְ הָ אָ ֶרץ הָ יְ תָ ה‬, 2:5a (two cases, ‫שיחַ הַ שָ דֶ ה טֶ ֶרם יִ ְהיֶה‬ ִ ‫ וְ כל‬and ‫)וכָל־עֵשֶ ב הַ שָ דֶ ה טֶ ֶרם יִ צְ מָ ח‬, 2:6 (‫)וְ אֵ ד ַי ֲעלֶה‬, 3:1 (‫)וְ הַ נָחָ ש הָ יָה‬, 3:22 (‫)הָ אָ דָ ם הָ יָה‬, 6:1 (‫)ובָ נֹות ילְ דו‬, 6:4 (‫)הַ נְ פִ לִ ים הָ יו‬, 7:6 (‫)וְ הַ מַ בול הָ יָה‬, 7:10 (‫)ומֵ י הַ מַ בול הָ יו‬, 7:11 (‫) ַואֲרבת הַ שָ מַ יִ ם נִ פְ תָ חו‬, 9:2 (‫)ומֹור ֲאכֶם וְ ִח ְתכֶם יִ ְהיֶה‬, ַ 11:3 (‫)וְ הַ חֵ מָ ר הָ יָה‬, 15:12 ( ‫וְ תַ ְרדֵ מָ ה‬ ‫)נָפְ לָה‬, 15:17 (two cases, ‫ הַ שֶ מֶ ש בָ אָ ה‬and ‫) ַו ֲעלָטָ ה הָ יָה‬, 19:23 (two cases, ‫ יָצָ א הַ שֶ מֶ ש‬and ‫)וְ לֹוט בָ א‬, 19:31 (‫)אָ בִ ינו זָקֵ ן‬, 34:10 (‫)וְ הָ אָ ֶרץ ִת ְהיֶה‬, 35:11 (‫)גֹוי וקְ הַ ל גֹויִ ם יִ ְהיֶה‬, 36:12 (‫)וְ ִת ְמנַע הָ יְ תָ ה‬, 36:13 (‫)אֵ לֶה הָ יו‬, 36:14 (‫)וְ אֵ לֶּה הָ יּו‬, 42:19 (‫)א ֲִחיכֶּם אֶּ חָ ד יֵאָ סֵ ר‬, 43:23b (‫)בָ א כ ְַספְ כֶם‬, 44:20b (two cases, ‫ וְ אָ ִחיו מֵ ת‬and ‫)וְ אָ ִביו אֲהֵ בֹו‬, 45:16 (‫)וְ הַ קל נִ ְשמַ ע‬, 48:10 (‫)וְ עֵינֵי יִ ְש ָראֵ ל ָכ ְבדו‬, 49:22 (‫)בָ נֹות צָ עֲדָ ה‬, 49:26 (‫)ב ְרכת אָ ִביָך ג ְָברו‬, ִ 50:23 ( ‫ְבנֵי‬ ‫ן־מנַשֶ ה ילְ דו‬ ְ ֶ‫)מָ כִ יר ב‬. Holmstedt’s list of topic SV cases and his list of unmarked SV cases each 41 the same grounds.45 In fifteen of the 32 allegedly neutral SV cases with stative, passive or middle semantics, the verb is the tensed copula ‫הָ יָה‬.46 The dubious relevance of these clauses is seen not only in their lack of agentive subjects, action and transitivity, but in their striking similarity to verbless clauses, with the expected neutral subject-predicate order. If wayyiqtol clauses are to be excluded because of their restriction to a specific clause type, so, arguably, should these. Finally, in at least nine cases (four of which are not excluded on other grounds), the pre-verbal subject in an allegedly neutral SV clause looks to be marked for one of the information structure functions; these are given in examples (21)–(29).47 Examples (21)–(25): Arguable cases of focal fronting among Holmstedt’s unmarked SV clauses (21) ... ‫ֹלהים יִ ְראֶ ה־לָ֥ ֹו הַ ֶ ֛שה לְ עלָ ֵ֖ה בְ ִנֶ֑י‬ ִִּ֞ ֱ‫ א‬... ‫הִ נֵ ִׂ֤ה הָ אֵ שֶׁ֙ וְ ָ ָ֣הע ָ֔ ִֵצים וְ אַ יֵ ָ֥ה הַ ֶ ֵ֖שה לְ ע ָ ָּֽׁלה׃‬ ‘Here’s the fire and the wood, but where is the sheep for the offering? … God will provide for himself the sheep …’ (Gen. 22:7b–8a) (fill-in: that someone must provide the offering is assumed, but it is not known who; the fronted element supplies the missing information) (22) ... ‫גֹויָ֖ם יִ הְ יֶ ָ֣ה ִמ ֶ ֶ֑מ ָך‬ ִּ ‫פְ ֵ ָ֣רה ְור ָ֔ ֵבה ָ֛גֹוי ּוקְ ַ ֶּ֥הל‬ ‘Be fruitful and multiply. A nation, even an assembly of nations will be from you.’ (Gen. 35:11a) (contra-expectation: notwithstanding the commands to be fruitful and multiply, that Jacob’s progeny should become a nation, much less an assembly thereof, is unexpected; excluded from previous counts due to use of ‫)הָ ָיה‬ include two examples from Gen. 49; one wonders if the poetry here qualifies as “stylistically neutral”. The relevance of four verses, two of which are listed above, is questionable on the grounds that the subject is encoded using a pronoun: Gen. 5:29, 36:13, 36:14, 38:28. 45 Gen. 19:28 (‫עלָה קִ יטר הָ אָ ֶרץ‬ ָ ), 27:41 (‫ ;יִ קְ ְרבו יְ מֵ י אֵ בֶ ל אָ ִבי‬probably jussive), 42:28 (‫)כ ְַספִ י הושַ ב‬. 46 See above, n. 44. 47 There are other cases in which a pre-verbal subject can be read as marked in terms of information structure. Focus: ‫‘ ִ֞זה יְ ַנח ֵ ֲִׂ֤מנו ִ ָּֽׁממַ עֲשֵ נֶׁ֙ וֶׁ֙ ומֵ עִ צְ ָ֣בֹון י ָ֔ ֵָדינו‬This one [as opposed to previous children] will comfort us from our deeds and from the toil of our hands’ (Gen. 5:29). In two further cases, the fronting is arguably intended to mark the entire clause for what Lambrecht (1994, pp. 222– 224, 233–235) calls sentence focus: ‫‘ וְ אָ ַ֕ ַמ ְרנו חַ יָ ֶּ֥ה ָר ָעָ֖ה אֲ כ ָ ָָ֑ל ְתהּו‬and we will say: “A wild animal has eaten him!” ’ (Gen. 37:20); ‫ֹלהים מָ צָ א֙ את־ע ֲֵֹּׁ֣ון עֲבָ ׂ֔דיך‬ ִָּ֗ ֱ‫‘ הָ א‬God has found out your servants’ guilt’ (Gen. 44:16). Constituent marking in such cases (fronting in BH) signals that the entire clause, not just the fronted element, constitutes marked rhematic information, usually as an answer to the (implied) question ‘What happened?’. Alternative pragmatic readings are also possible. 42 (23) ‫ׁשי הַ מָ קֹום֙ ָ ָּֽׁא ְמ ָ֔רו לא־הָ יְ ָ ָ֥תה בָ זֶ ֵ֖ה קְ דֵ ָ ָּֽׁשה׃‬ ֵּׁ֤ ְ‫אתיהָ וְ ֶׁ֙ ַגם אַ נ‬ ֶ֑ ִ ָ‫ָ֣לא ְמצ‬ ‘I haven’t found her. Even the locals said “There was no prostitute here.” ’ (Gen. 38:22) (focalising particle, contra-expectation: even those expected to know do not know) (24) ‫ֵאמר זָ֖ה י ָָצָ֥א ִראש ָנָּֽׁה׃‬ ָ֔ ‫ו ִַת ַ ָ֣קח הַ ְמ ַי ֶַ֗לדֶ ת ו ִַתקְ ָּ֙שר עַל־י ִָׂ֤דֹו שָ נִיֶׁ֙ ל‬ ‘And the midwife took and tied a scarlet thread on his hand, saying: “This one came out first.” ’ (Gen. 38:28) (identification, disambiguation: given the cultural importance attached to primogeniture, and the ambiguity involved in the birth of twins, identification of the firstborn is fitting information to be focalised) (25) ‫אֹלהי‬ ֵּׁ֤ ‫יכם ָּֽו‬ ִ֜ ‫ל־תי ָ ַ֗ראו אֱ ֹֹ֙לה‬ ִ ַ‫י־שם כַסְ פֵ ֵ֖נו בְ אַ ְמ ְתח ֵ ָּֽׁתינו׃ וַיאמֶ ֩ר שָ לֶֹׁ֙ום ָל ֶ֜כם א‬ ָ֥ ָ ‫ָ֣לא י ָ֔ ַָדעְ נו ִמ‬ ... ‫אֲ ִּביכם֙ נ ֶַָׁ֙תן לָכֶ ִׂ֤ם מַ ְטמֹוןֶׁ֙ בְ אַ ְמ ְת ָ֣חתֵ י ֶָ֔כם‬ ‘ “We don’t know who put our silver in our sacks.” And he said “Peace be with you. Don’t be afraid. Your God and the God of your fathers put treasure in your sacks.” ’ (Gen. 43:22b–23a) (fill-in: the fronted element answers the indirect question in the preceding verse) Examples (26)–(29): Arguable cases of topic fronting among Holmstedt’s unmarked SV clauses (26) ‫בַ יָ֣ ֹום הַ ֶַ֗זה נִ בְ קְ עוֶׁ֙ ָ ָּֽׁכל־מַ עְ יְ נתֶׁ֙ ְת ָ֣הֹום ַר ָ֔ ָבה וַאֲ רֻ ֥בת הַ שָ ַ ֹ֖מים נִ פְ ָ ָּֽׁתחו׃‬ ‘on that day burst forth all the springs of the great deep, while the floodgates of the sky opened’ (Gen. 7:11) (excluded from previous counts due to passive/intransitive verb and non-agentive/non-human subject48) (27) ‫ו ְַת ִ ֶׁ֙הי ל ֶָהִׂ֤ם הַ לְ בֵ נָהֶׁ֙ לְ ָ֔ ָאבֶ ן וְ ַהַ֣ח ָ֔ ָמר הָ יָ ָ֥ה ל ֶָהֵ֖ם ל ַָּֽׁחמֶ ר׃‬ ‘And brick served them as stone, while bitumen served them as mortar’ (Gen. 11:3) (excluded from previous counts due to use of ‫ הָ ָיה‬and non-agentive/nonhuman subject) (28) ... ‫ִאם־כֵנִ ָ֣ים אַ ָ֔ ֶתם אֲ חיכֶׁ ַ֣ם֙אֶׁ ָ֔ ָחד יֵאָ ֵ ֵ֖סר בְ ֵ ָ֣בית ִמ ְשמַ ְרכֶ ֶ֑ם וְ אַ תֶׁ ם֙ לְ כָ֣ ו‬ ‘If you’re honest, let one of your brothers remain in the place of detention, whereas you, go …’ (Gen. 42:19) (excluded from previous counts due to use of passive verb and nonagentive/non-human subject) The prepositional time adverbial ‘on this day’ is also topicalised as a temporal, scene-setting device (see above, section 4.1.2). 48 43 (29) ‫ָּֽׁאֹלהי אֲבִ יכֶםֶׁ֙ נ ֶַָׁ֙תן לָכֶ ִׂ֤ם מַ טְ מֹוןֶׁ֙ בְ אַ ְמ ְת ָ֣חתֵ י ֶָ֔כם כ ְַספְ כֶׁ ֹ֖ם‬ ִׂ֤ ֵ ‫יראו א ֱָֹּ֙להֵ י ֶ֜כם ֵו‬ ַ֗ ָ ‫ל־ת‬ ִ ַ‫וַיאמֶ ֩ר שָ לֶֹׁ֙ום ָל ֶ֜כם א‬ ... ‫ָבָ֣א אֵ לָ ֶ֑י‬ ‘And he said “Peace be with you. Don’t be afraid. Your God and the God of your fathers put treasure in your sacks. Your money came to me.” ’ (Gen. 43:23) (excluded from previous counts due to use of intransitive verb and nonagentive/non-human subject) 4.3.2.3 Ramifications: Is VS basic after all? In total, at least 36 of the 47 cases of alleged unmarked SV order are excluded due to clause type, pragmatic marking or both, leaving just 11 cases of apparently unmarked SV order against 23 unmarked VS clauses.49 Applying Holmstedt’s criteria strictly, VS emerges as the numerically dominant order for pragmatically neutral clauses. Though postulating a basic VS word order admittedly leaves a situation of asymmetry between the respective default orders of verbal and verbless clauses, basic order is unified in main and subordinate verbal clauses.50 Returning to the question with which this section began, the assumption of basic SV word order in the BH verbal clause is of little help in accounting for the many XV clauses that cannot be explained in terms of information structure. Topic and focus get one only so far; the need for a complementary dimension or dimensions is evident. 4.4 Alternative pragmatic values and word order variation The greatest problem with positing the all-sufficiency of information structure as an explanation for fronting in the BH verbal clause is that the relevant marking is generally conceived of as applying narrowly to fronted arguments in agreement with their informational status. Yet frequently in BH, elements are fronted for purposes of marking something special about the entire clause. Though some formulations of information structure include the possibility of whole-clause marking by means of a fronted constituent – for example, Lambrecht’s sentence focus,51 which may usefully explain certain cases of preverbal positioning in BH – this is inappropriate for the vast majority of XV instances in BH in which X is neither topic nor focus. 49 This figure should be considered provisional, since it is possible that the SV order in one or more cases may signal pragmatic marking for purposes other than topic or focus; see below. 50 Holmstedt himself (2011, pp. 16–18, 28; see also 2016, p. 146 and n. 17) recognises neutral VS word order in subordinate clauses. Indeed, he argues compellingly (2011, p. 16) that this order in subordinate clauses reflects an earlier unmarked order. See also Buth, 1995, pp. 82– 83. While the concept of ‘basic’ or pragmatically neutral constituent order is useful for explaining marked patterns, it may be advisable to resist the dogmatism involved in assigning a basic word order. It may be more profitable simply to recognise functional correlations between various orders and their semantic/pragmatic values and effects. I thank Profs. Geoffrey Khan and Christo van der Merwe for discussing this perspective with me. 51 Lambrecht, 1994, pp. 222–224, 233–235. 44 4.4.1 Temporal succession (sequentiality) An obvious candidate for a parameter to complement information structure is temporal succession. Wayyiqtol has long been associated with temporal sequence, X qatal with departures therefrom, especially anteriority.52 See examples (30)–(35). Examples (30)–(35): X qatal signalling anteriority (30) ‫ימלְֶך בַ חֲלָ֣ ֹום הַ לָ ֶ֑יְ לָה ו ַָ֣יאמֶ ר ַ֗לֹו הִ נְ ָךָ֥ מֵ תֶׁ֙ עַל־הָ ִא ָ ָ֣שה אֲשֶ ר־ל ָ֔ ַָקחְ תָ וְ ִ ֵ֖הוא‬ ֵ֖ ֶ ִ‫ֹלהים אֶ ל־אֲב‬ ֛ ִ ‫ַוי ָָּ֧בא ֱא‬ ָ‫ימלְך ָ֥לא קָ ַ ָ֖רב אֵ לֶ ֶ֑יה‬ ֶ֕ ‫בְ עָ֥לַת ָ ָּֽׁבעַל׃ וַאֲ ִּב‬ ‘And God came to Abimelech in a dream. And he said to him, “You’re dead on account of the woman you have taken, since she’s married.” But Abimelech had not approached her …’ (Gen. 20:3–4) (note that relative VS order was possible here, e.g., ‫*וְ לא קָ ַרב אֵ לֶיהָ אֲבִ ימֶ לְֶך‬, but this would be given to a sequential reading, i.e., ‘so Abimelech did not approach her’) (31) ‫ ל ָ֛בֹוא אֶ ל־יִ צְ ָחָ֥ק אָ ִ ֵ֖ביו ַ ָ֥א ְרצָ ה כְ ָ ָּֽׁנעַן׃ וְ ל ָ ֵָּׁ֣בן הָ ַׂ֔לְך לִ גְ ֵ֖זז‬... ֶׁ֙‫ָל־רכשֹו‬ ְ ‫ָל־מקְ ַ֗ ֵנהו וְ אֶ ת־כ‬ ִ ‫וַיִ נְ ַהָ֣ג אֶ ת־כ‬ ‫אֶ ת־צאנֶ֑ ֹו ו ִַתגְ ָ֣נב ָר ָ֔ ֵחל אֶ ת־הַ ְת ָר ִ ֵ֖פים ֲא ֶ ָ֥שר לְ אָ ִ ָּֽׁביהָ ׃‬ ‘And he drove all his livestock and all his possessions … to come to his father Isaac in the land of Canaan. Meanwhile, Laban had gone to shear his sheep, and Rachel stole her father’s household gods.’ (Gen. 31:18–19) (cf. the alternative, default ‫‘ * ַו ֵילְֶך לָבָ ן‬and Laban went’, likely to give rise to a temporally successive reading) (32) ... ‫ַוי ֵ ַָ֥שג ל ָָבֵ֖ן ֶ ָּֽׁאת־ ַיע ֲֶ֑קב וְ ַיעֲקָֹׁ֗ ב תָ ַ ֵּׁ֤קע ֶ ָּֽׁאת־אָ הֳלֹוֶׁ֙ בָ ָ֔ ָהר‬ ‘And Laban caught up with Jacob. Now, Jacob had pitched his tent in the hill country …’ (Gen. 31:25) (33) ‫ַוי ֶָׁ֙בא ל ֜ ָָבן בְ ָ֥אהֶ ל ַיע ֲָ֣קב׀ ובְ ָ֣אהֶ ל ל ַ֗ ֵָאה ובְ ֛אהֶ ל ְש ֵ ָ֥תי הָ ֲאמָ ֵ֖הת וְ ָ֣לא מָ ָצֶ֑א ַויֵצֵ אֶׁ֙ מֵ ָ֣אהֶ ל ל ָ֔ ֵָאה ַוי ֵָ֖בא‬ ... ‫ֵיהֶ֑ם‬ ֶ ‫בְ ָ֥אהֶ ל ָר ֵ ָּֽׁחל׃ וְ ָר ִ֞חל לָקְ ָ ֵּׁ֣חה אֶ ת־הַ ְת ָר ַ֗ ִפים ו ְַת ִש ֵ ֛מם בְ כַ ָ֥ר הַ ג ָ ֵָ֖מל ו ֵ ַָ֣תשֶ ב ֲעל‬ ‘And Laban entered Jacob’s tent and Leah’s tent and the tents of the two maidservants, but he did not find (them). And he left Leah’s tent and entered Rachel’s tent. But Rachel had taken the household gods and put them in her saddlebag and sat upon them …’ (Gen. 31:33–34) Note, for example, Rashi on ‫‘ וְ ָהָ֣אָ ָ֔ ָדם י ַ ֵָ֖דע אֶ ת־חַ וָ ָ֣ה ִא ְש ֶ֑תֹו‬and the man knew Eve, his wife’ (Gen. 4:1): ,‫ וכן ההריון והלידה; שאם כתב וידע אדם‬,‫ קודם שחטא ושנטרד מגן עדן‬,‫כבר קודם עניין שלמעלה‬ ‫‘ נשמע שאחר שנטרד היו לו בנים‬Already before the above matter, before he sinned and was banished from the garden of Eden, and so too the pregnancy and birth. For if it had written ‫וידע אדם‬, it would have meant that after he was banished he had children’ (Moshavi, 2010, p. 23). 52 45 (34) ... ‫הּורד ִמצְ ָ ֶ֑ריְ מָ ה‬ ֵּׁ֣ ַ ‫יֹוסף‬ ָ֖ ְ‫ו‬ ‘And Joseph had been taken down to Egypt …’ (Gen. 39:1) (chapter 39 continues the narrative of chapter 37; since the events of the intervening chapter extend generations beyond the narrative of chapters 37 and 39, 39:1 marks a retreat in relation to the end of chapter 38) (35) ‫ת־שנֵ ָ֥י הָ ֲאנ ִ ֵָ֖שים ַו ִָּֽׁתצְ פְ נֶ֑ ֹו ו ַָ֣תאמֶ ר׀ ֵַ֗כן ָבִׂ֤או אֵ לַיֶׁ֙ ָ ָּֽׁה ֲאנ ָ֔ ִָשים וְ ָ֥לא י ַ ֵָ֖דעְ ִתי מֵ ַ ָ֥איִ ן‬ ְ ֶ‫ו ִַת ַ ָּ֧קח ָ ָּֽׁה ִא ָ ֛שה א‬ ‫ַיְהי הַ שַ֜ עַר לִ ְסגַ֗ ֹור בַ ָּ֙חשֶ ְךֶׁ֙ וְ הָ ֲאנ ִ ָָ֣שים י ָ֔ ָָצאו ָ֣לא י ָ֔ ַָדעְ ִתי ָ ָ֥אנָה הָ לְ כֵ֖ ו ָ ָּֽׁה ֲאנ ִ ֶָ֑שים ִרדְ ָ֥פו מַ ֵה֛ר‬ ֶׁ֙ ִ ‫ֵ ָּֽׁהמָ ה׃ ו‬ ‫יהֵ֖ם ִ ָ֥כי תַ ִשיגָּֽׁ ום׃ וְ ִּ ָ֖היא הע ֱָלֵּׁ֣תַ ם הַ גָ ֶ֑ ָּ֑גָה ַו ִָּֽׁת ְט ְמנֵםֶׁ֙ בְ פִ ְש ֵ ָ֣תי הָ ֵָ֔עץ הָ עֲרכָ֥ ֹות לָ ֵּ֖ה עַל־הַ ָגָּֽׁג׃‬ ֶ ‫אַ ח ֲֵר‬ ‘And the woman took the two men and she hid them. And she said “So it is that the men came to me, but I didn’t know where they were from. And the gate was about to close in the dark when the men left. I don’t know where the men went. Hurry, chase after them, so that you catch them!” But she had taken them up to the roof and hidden them in the flax-stalks laid out for her on the roof.’ (Josh. 2:4–6) 4.4.1.1 Non-sequential wayyiqtol There is an obvious correlation between wayyiqtol and event order, 53 but as a semantic property, temporal succession is too narrow and rigid to be comprehensively applicable. Gauged against a strict conception of event sequence, many wayyiqtols prove to flout a precise notion of sequentiality, as shown in examples (36)–(38), which could be multiplied.54 Examples (36)–(38): Temporally non-successive wayyiqtol (36) ... ֶׁ֙‫ברוְך אַ בְ ָרם‬ ִׂ֤ ‫ֹׁאמר‬ ָ֑ ַ ‫ַוָָּּֽׁ֑יְבָ ְרכֵ ֵ֖הו ַוי‬ ‘And he blessed him and said “Blessed be Abram …” ’ (Gen. 14:19) (the speaking and blessing are the same act) (37) ‫וַיִ ְשאוֶׁ֙ אֶ ת־יֹו ָ֔ ָנה ַו ְי ִטלֵ֖הו אֶ ל־הַ יָ ֶ֑ם ַו ַיע ֲָ֥מד הַ יָ ֵ֖ם ִמזַעְ פָּֽׁ ֹו׃ ו ִ ַָּֽׁי ְיר ָּ֧או הָ ֲאנ ִ ָ֛שים י ְִר ָ ָ֥אה גְ דֹולָ ֵ֖ה אֶ ת־יְ הוָ ֶ֑ה‬ ... ‫חּו־זבַ ח֙ ַ ָּֽׁליה ָ֔ ָוה ַוָָּּֽׁ֑ ִי ְד ֵ֖רו ְנדָ ִ ָּֽׁרים׃ וַיְ ַ ֵּׁ֤מן יְ הוָה֙ ָ ָ֣דג גָדָ֔ ֹול לִ בְ ֹלֵ֖ ַע אֶ ת־יֹונָ ֶ֑ה‬ ֙ ‫ו ִּ ַָּֽיז ְְב‬ ‘And they picked up Jonah and threw him into the sea. And the sea ceased from its raging. And the men greatly feared Yhwh. And they made sacrifices and took oaths. And Yhwh appointed a great fish to swallow Jonah.’ (Jon. 1:16–2:1) (it is unlikely that the sailors are to be envisioned as having made their sacrifices while on the high seas or, arguably, that God waited till they had got back to shore to sacrifice before appointing the fish to swallow Jonah55) 53 Hatav (1997, p. 57) observes that just six per cent of the 2,445 wayyiqtol clauses in her corpus of classical Hebrew prose are clearly non-sequential. 54 Buth, 1994b; Buth, 1995, pp. 86–87; Cook, 2004, pp. 257–261; Cook, 2012, pp. 289–298; Moshavi, 2010, pp. 29–30. 55 Buth, 1994b, pp. 143–144; Buth, 1995, p. 87. 46 (38) ‫ן־א ָ ָ֣שה זֹונָ ֶ֑ה וַיֶּ֥ ֹולד גִּ לְ ָעָ֖ד אֶ ת־יִפְ ָ ָּֽׁתח׃‬ ִ ֶ‫וְ ִיפְ ָ ָ֣תח הַ גִ לְ ע ִַָ֗די הָ יָהֶׁ֙ גִ ָ֣בֹור ָ֔ ַח ִיל וְ ֵ֖הוא ב‬ ‘And Jephthah the Gileadite was a great warrior. And he was the son of a prostitute. And Gilead fathered Jephthah.’ (Judg. 11:1) (Gilead obviously bore Jephthah before the latter became a mighty warrior56) These examples show that wayyiqtol is best viewed as the default – that is, unmarked – past narrative form, and that its association with temporal succession, while real, is indirect. In other words, BH writers needed no special reason to utilise wayyiqtol in past tense narrative accounts, but did need a special reason to use an alternative. 4.4.1.2 Sequential X qatal Just as wayyiqtol does not automatically entail temporal succession, strictly speaking, only a minority of X qatal cases call for translation using the pluperfect.57 As long as temporal succession in discourse is linked on a one-toone basis with real-world temporal sequence, it will prove too rigid a parameter to account for XV instances left unexplained by recourse to information structure. However, even if it is considered a purely pragmatic, presentationcentred property, that is, as being more dependent on a writer’s discretion for purposes of text structure than on actual real-life event sequence, it seems doubtful that the concept of temporal succession is sufficiently broad to encompass the range of meanings and effects associated with wayyiqtol and X qatal not covered by information structure, some of which, significantly, seem to have nothing to do with sequence or ordering. 4.4.2 Grounding We turn now to another potential complementary parameter. With specific regard to narrative texts, the notion of grounding – foreground vs background – has been applied to biblical texts. Wayyiqtol is seen as carrying forward the sequence of mainline events and actions, with alternative forms and structures used to furnish ancillary information. But there are instances of non-wayyiqtol forms and structures – including X qatal – ill-suited by the label ‘background’. Examples (39)–(40) present cases of what are arguably mainline sequential events in which fronting does not obviously mark topic or focus. 56 Buth, 1995, p. 86. According to Moshavi (2010, p. 119), in only 11.2 per cent of clauses with fronting in the prose of Genesis does the fronting mark the entire clause, and in only a portion of these is an anterior reading called for. 57 47 Examples (39)–(40): X qatal for foreground eventualities (39) ‫וַיָ ִָּׂ֤֑קָ ם יֹונָהֶׁ֙ לִ בְ ָ֣רחַ תַ ְר ָ֔ ִשישָ ה ִמלִ פְ נֵ ֵ֖י יְ הוָ ֶ֑ה ַו ֵֶׁ֙י ֶרד י ָ֜פֹו וַיִ ְמ ָצָ֥א אֳנִ יָ ָ֣ה׀ בָ ָ ָ֣אה תַ ְר ַ֗ ִשיש וַיִ ֵֶׁ֙תן ְשכ ָ ָ֜רּה‬ ‫יְהוָּֽׁה׃ ַ ָּֽויה ָ֗ ָוה ה ִּ ֵּׁ֤טיל ָּֽׁרוחַ ־גְ דֹולָהֶׁ֙ אֶ ל־הַ ָָ֔ים וַיְ ִ ָ֥הי ַ ָּֽׁסעַר־‬ ָ ‫וַיֵ ִׂ֤ ֶָּ֑רד בָ ּהֶׁ֙ ל ִָׂ֤בֹוא עִ מָ הֶ םֶׁ֙ תַ ְר ָ֔ ִשישָ ה ִמלִ פְ נֵ ֵ֖י‬ ‫ג ֵָ֖דֹול בַ יָ ֶ֑ם וְ ָ ֵּׁ֣האֳ נִּ ָׂ֔יה ִּח ְש ָ ָ֖בה לְ הִ שָ ֵ ָּֽׁבר׃‬ ‘And Jonah arose to flee to Tarshish from before Yhwh. And he went down to Joppa, where he found a boat headed for Tarshish, gave payment, and set off to go with them to Tarshish from before Yhwh. And Yhwh cast down to the sea a mighty wind. And there was a great storm in the sea. And the boat threatened to break up.’ (Jon. 1:3–4)58 (40) ‫שמׁש י ָ ֵָּׁ֣צא ַעל־הָ ָ ֶ֑א ֶרץ וְ לָ֖ ֹוט ָ ֶּ֥בא ָּֽׁצע ֲָרה׃ ַ ָּֽויה ָ֗ ָוה ִּה ְמ ִּ ֶ֧טיר עַל־סְ ֛דם וְ עַל־עֲמ ָ ֵ֖רה גָפְ ִ ָ֣רית ו ֵ ֶָ֑אש‬ ָ֖ ַ‫ה‬ ‘The sun came up over the land. And Lot came (or was coming) to Zoar. And Yhwh rained down upon Sodom and upon Gomorrah brimstone and fire.’ (Gen. 19:24)59 They may be wrapped in the packaging typically associated with supporting information, but, far from being backgrounded, the events they relate are highlighted foreground.60 That scholars often differ on distinguishing foreground from background demonstrates the subjectivity of the distinction.61 4.4.3 Discourse continuity and discontinuity As a concept that more satisfactorily complements information structure in accounting for BH constituent order variation, the notion of discourse continuity seems useful. So far as I can discern, the first to elaborate on discourse continuity was Givón.62 Since then ‘continuity’ as a pragmatic dimension within Hebrew discourse has been variously cited and defined.63 Givón lists three aspects of discourse continuity: thematic, action and topic. Of particular 58 Buth, 2005, p. 77 n. 17. Buth, 1994a, pp. 226–227. 60 Bailey and Levinsohn, 1992, pp. 200–204; Buth, 1995, pp. 85–88; Moshavi, 2010, pp. 27– 31. 61 Heimerdinger (1999, pp. 75–76) is justifiably critical of Longacre’s (1989, pp. 76–78) rather circular identification of foreground with wayyiqtol. Bailey and Levinsohn (1992, p. 204) not unreasonably conjecture that Longacre may have confused the notions of ‘backbone’ and ‘foreground’. See also Hornkohl, 2005, pp. 56–60. If the concept of grounding is to be used at all in discourse studies, it may be useful to think in terms of three grounds: background, (neutral) mid-ground and (highlighted) foreground. Conceivably, identical marking could be used to signal either backgrounded or foregrounded entities/eventualities; cf. the helpful, if tentative, discussion in Robar, 2014, pp. 65–67. 62 Givón, 1983, pp. 7–8. 63 Bandstra, 1992, pp. 116ff; Buth, 1995, p. 97; Bailey and Levinsohn, 1992; Dooley and Levinsohn, 2001, pp. 37–42; Hornkohl, 2005, pp. 53–78; Robar, 2014, pp. 148–188. It gives me great pleasure to note that the honouree of the present volume himself noted some of the discontinuity functions of XV word order. Already in his 1988 monograph on extraposition in Semitic languages Geoffrey Khan included a section (pp. 86–88) on ‘Other devices for marking span boundaries’, among them SV clauses. 59 48 relevance to the present discussion is action continuity, which Givón explicitly relates to temporal sequentiality and adjacency. Since the notion of pragmatic continuity is not directly linked to real-world temporal succession, but constitutes a higher-level discourse property, the application of which is at least partially subject to a language user’s discretion, it promises to provide the practical and theoretical flexibility precluded by approaches based on semantic parameters, such as temporal succession, which often fail to comprehend the optionality of pragmatic marking. It is also, arguably, less subjective than the notion of grounding. It seems obvious that the waw-consecutive forms encode action continuity in the sorts of discourse in which they feature and, at the same time, that action continuity is a default quality of those types of discourse. For example, on the assumption that wayyiqtol encodes action continuity in narrative and report, it is not surprising that it should correlate highly with real-life sequence, but also not infrequently serve to depict eventualities that are not strictly or even remotely sequential. As long as readers can successfully sequence story events based on content, context and logic, explicit grammatical marking of backtracking is unnecessary. Similarly, the notion of pragmatic discontinuity fits well with a variety of effects associated with X qatal that are not covered by information structure, specifically those aforementioned cases of pre-verbal constituents in which the fronting marks the entire clause rather than the fronted element itself: departure from temporal succession; departures from the mainline story for purposes of providing ancillary information; signalling scene and unit boundaries (including the onset of direct speech), where, to be sure, the same structure also frequently indicates an actual change of clause-level or discourse topic and/or a temporal retreat (see below); even, possibly, the use of X qatal for such effects as sentence focus, highlighting and dramatic pause.64 While some of these are explicable according to the concepts of temporal succession and/or grounding, discontinuity seems a particularly apt concept for uniting them under a single, explanatory heading.65 64 Returning to the marked SV cases in Genesis in which, against Holmstedt (2011, pp. 23–24 n. 53), the fronted element is here understood as marked for discontinuity rather than topic (n. 37 above), the following classification of effects is suggested (note that a few cases have multiple effects): new unit/scene/direct speech or new tack in direct speech – 4:1, 9:2, 13:14, 18:17, 19:15, 21:1, 22:1, 24:1a, 24:35, 24:40, 26:26, 27:6, 28:3, 34:5, 37:36 (end of scene), 41:10, 44:19, 46:31 (sentence focus?), 47:5 (sentence focus?); off-line information (parenthesis, description, summary, restatement, commentary, interruption) – 8:5, 14:3, 21:6, 24:56, 25:34, 31:5, 37:3, 41:56, 41:57 (end of scene); in a subcategory of these are background clauses in which the action should be understood as temporally anterior: 24:1b, 31:19, 31:34, 34:5, 34:7; dramatic pause/highlighting – 7:19, 19:23, 19:24. As noted above, the identification of these effects is subjective. 65 The fullest explanation is that of Buth, 1995, pp. 84–93, 95–99; see also Khan, 1988, pp. 86– 88; Bailey and Levinsohn, 1992, pp. 188–200; Hornkohl, 2005, pp. 41–78. 49 Positing continuity as an inherent property of discourse brings with it another benefit. It allows us to slice through the Gordion knot inherent in a system in which all verbal forms have been described as marked. If action continuity represents a default property of certain types of discourse, and if the waw-consecutive forms are recognised as encoding continuity, then wayyiqtol and weqatal emerge as default, unmarked forms within the respective types of discourse in which each figures prominently. Since both information structure and discourse continuity have been recognised by scholars as pragmatic dimensions signalled by means of BH word order variation, it is appropriate at this juncture to explain how what has been proposed here tallies with other recent proposals. Several approaches to BH constituent order admit only one dimension, most often information structure, which, as we have seen, fails to account for a significant proportion of XV structures.66 Other approaches include two (or more) dimensions, usually information structure and either temporal succession or grounding (or a combination of the two),67 but the explanation is characterised by the weaknesses highlighted above, such as lack of comprehensiveness, conceptual rigidity (in terms of temporal succession) and/or subjectivity (in terms of grounding). Happily, several scholars promote approaches that combine information structure and continuity/discontinuity. Yet even among these there is variety, especially concerning the degree to which scholars are willing to group sometimes apparently disparate effects within broader pragmatic categories. One sort of approach, exemplified most comprehensively by Moshavi, presents a taxonomy of constituent order effects representing a spectrum of parameters for which she seeks no common heading. Focusing on clauses with fronted subjects, she boils factors down to those which mark the preposed element itself for the information structure notions of topic or focus68 and an assortment of effects having to do with the special marking of the entire clause: anteriority,69 simultaneity,70 background information,71 new narrative unit or scene within a narrative.72 Moshavi subsequently attempts to deal with a sizable ‘residue’ of clauses, which are especially typical of direct speech, offering such categories as justification, affirmation and boasting.73 Moshavi’s approach is somewhat reminiscent of my own, according to which focal frontings are distinguished from non-focal frontings, the latter being further subdivided into those in which the preposed constituent is itself marked as a topic and those in which the entire clause is set off.74 Significantly, 66 Heimerdinger, 1999; Holmstedt, 2011. Longacre, 1989. 68 Moshavi, 2010, pp. 104–112. 69 Moshavi, 2010, pp. 112–113. 70 Moshavi, 2010, pp. 113–114. 71 Moshavi, 2010, p. 114. 72 Moshavi, 2010, p. 115. 73 Moshavi, 2010, pp. 115–120. 74 Hornkohl, 2005, pp. 35–78. 67 50 however, I attempted to explain all frontings that mark the entire clause in terms of discourse discontinuity.75 Bailey and Levinsohn, joined by Buth, go one step further.76 They deal with focal fronting as a function of information structure, but group all other frontings, both those which mark the preposed element as a topic and those in which the preposed element marks off the entire clause, under the umbrella of discontinuities. Bailey and Levinsohn use the terminology associated with information structure, that is, focus versus topicalisation, but it is clear from their discussion that they view topicalisation in terms of discontinuity. For his part, Buth coins the term ‘contextualising constituent’ (to replace his earlier ‘pseudo-topic’) as a broad category that subsumes constituents fronted for their own informational status as well as those preposed to signal something special about the entire clause in which they occur. Notwithstanding the differences among the approaches just summarised, it is of crucial import for the present study to note that all (a) recognise the explanatory inadequacy of systems that attempt to account for BH constituent order variation on the basis of a single parameter, (b) posit the utility of the information structure values of focus and topic, (c) acknowledge the reality of whole-clause marking by means of fronted elements, and (d) adumbrate similar lists of (especially narrative) effects associated with discourse discontinuity. The studies of these scholars represent a consensus regarding the need for a multidimensional account of BH word order variation encompassing both information structure and continuity/discontinuity. Even so, the differences between approaches merit consideration. Most significantly, Moshavi wonders whether Buth is successful in unifying all the different types of discontinuities. By applying the term Contextualising Constituent to cases in which the preposed item does not actually function as Topic or Setting, Buth is calling very different things by the same name. 77 Language users evidently made use of a single multi-functional structure, the XV order, effectively to halt forward progress of the default discourse continuity iconically communicated by the waw-consecutive forms for purposes of specially marking both genuine topics and whole clauses. Further, the intersection between information structure and discontinuity is abundantly evident 75 The discontinuity effects that I enumerated are similar to some of those mentioned by Moshavi: non-sequentiality (Hornkohl, 2005, pp. 62–66), non-storyline information (circumstantial, parenthetical, explanatory, background, summary; pp. 67–68), start of new literary unit (episode or paragraph; pp. 68–70), intra-episode scene switching (pp. 70–71), dramatic slowing or pause (pp. 71–74), change of tack/start of new theme in direct speech, including restatement, explanation, detailing, and redirection (pp. 74–78). 76 Bailey and Levinsohn, 1992; Buth, 1995. 77 Moshavi, 2010, p. 40. 51 in cases in which the marking of a genuine topic (whether sentence topic, discourse topic or both) coincides with a discontinuity effect. The preposing of actual clause- and discourse-level topics also frequently achieves other effects, as can be seen in examples (41)–(44). (41) ... ‫וְ הַ נָחָ ׁש הָ יָה עָרום ִמכל חַ יַת הַ שָ דֶ ה אֲשֶ ר עָשָ ה יְ הוָה ֱאֹלהִ ים‬ ‘Now, the serpent was more cunning than any other beast of the field that Yhwh God had made …’ (Gen. 3:1) (the fronting serves to mark both the serpent as a topic of the ensuring discourse and the start of an episode78) (42) ... ‫ ּולְ אָ ָ ֵּׁ֣דם אָ ַ֗ ַמר‬... ‫שה אָ ַ֗ ַמר‬ ֵּׁ֣ ָ ‫ ָּֽאל־הָ ִּא‬... ‫ֹלהים ׀ ֶ ָּֽׁאל־הַ נָחָ ש‬ ָ֥ ִ ‫וַיאמֶ ֩ר יְ ה ֶׁ֙ ָוה ֱא‬ ‘14 And Yhwh God said to the serpent … 16 To the woman he said … 17 And to the man he said …’ (Gen. 3:14, 16–17) (similarly, the non-subjectival frontings here serve as genuine topics highlighting the addressee of each curse, put also serve to segment the text into paragraphs) (43) ... ‫הורד ִמצְ ָ ֶ֑ריְ מָ ה‬ ָ֣ ַ ‫יֹוסף‬ ָ֖ ְ‫ו‬ ‘And Joseph had been taken down to Egypt …’ (Gen. 39:1) (here the fronted element serves as discourse topic, reactivated after the Judah-excursus of chapter 38; it signals a new literary unit; and it marks a temporal retreat since, as mentioned in example (34), the Judahoriented events of chapter 38 extend generations beyond the Josephoriented events recounted in chapters 37 and 39) (44) ... ‫ת־מקְ נֵהו בַ שָ דֶ ה‬ ִ ֶ‫וְ ַיעֲקֹׁ ב שָ מַ ע כִ י ִטמֵ א אֶ ת־דִ ינָה בִ תֹו ּובָ נָיו הָ יו א‬ ‘Now Jacob heard that he had defiled his daughter Dinah. But his sons were with his livestock in the field …’ (Gen. 34:5) (the first fronting may be genuinely topical as well as mark a scene change, while the second arguably serves several purposes at once: to raise the sons to a discourse topic; to heighten the contrast between Jacob’s hearing and the sons’ ignorance; and to avoid the wayyiqtol structure ‫*וַיִ הְ יו בָ נָיו‬, which would have sounded like a temporally subsequent situation) 78 Cf. ‫שתֹו‬ ְ ‫‘ וְ הָ אָ דָ ם יָדַ ע אֶ ת־חַ וָה ִא‬Now, the man had marital relations with his wife, Eve …’ (Gen. 4:1), in which the XV structure would seem to mark a new literary unit, but in which the preposed argument ‘the man’ seems in no way fronted on account of its own special informational status, since he is not a topic of the ensuing discourse. Indeed, the only thing that would be different given the alternative VS order ‫ * ַו ֵידַ ע הָ אָ דָ ם אֶ ת־חַ וָה ִא ְשתֹו‬would arguably be that continuity would be maintained with the previous episode. Pace Rashi (see above, n. 52), it seems more likely that the discontinuity here signals span onset than anteriority. 52 Despite the polyvalence of XV constituent order, it seems notionally valid, and indeed preferable, to distinguish between actual discourse topics and elements brought forward for clause marking. The two phenomena exploit the same syntactic structures, but involve different cognitive processes, since an actual topic, like a setting, orients understanding of an utterance around itself in a way that clause-marking frontings do not.79 The explanatory power of a continuity/discontinuity approach is largely untested. Moshavi found that by taking into account both information structure and such effects as anteriority, simultaneity, background, scene/unit demarcation, justification, affirmation and boasting, she could account for about 70 per cent of the approximately 400 XV structures in Genesis, with the lion’s share (nearly 84 per cent) of the unexplained instances coming in direct speech.80 Whether most or many of these can be compellingly explained in terms of discourse discontinuity or whether there is need for another complementary explanation or explanations remains to be seen.81 However, given the utility of an approach to explain BH word order incorporating an intersection of information structure and continuity/discontinuity, no future study should ignore these parameters. 5 Conclusion To summarise: Any legitimate treatment of BH word order must include a view of the language’s verbal system that incorporates not just semantics (tense, aspect and mood) and syntax, but pragmatics as well. In view of the semantic-functional polyvalence of XV structures, in some of which X is itself marked for topic or focus, and in others of which the fronted X marks the 79 Moshavi (2010, p. 39) understandably questions Buth’s (1994a, p. 223) claim that a contrastive topic is simultaneously a marked topic and a marked focus. However, on a few other points she seems to have misunderstood Buth. For example, she writes that “[o]ne type of discontinuity concerns a particular constituent in the clause, which is preposed to mark it as Focus or Contextualizing Constituent” (Moshavi, 2010, p. 39). However, Buth clearly distinguishes between focus, on the one hand, and topic/discontinuity/contextualising constituent, on the other (1995, pp. 84–85, 100). Similarly, Moshavi’s criticisms (2010, p. 40) regarding the marking of continuous discourse topics by means of discontinuity structures and of foregrounded dramatic pauses by means of constructions typically associated with background arguably stem from confusion born of conceptual complexity and terminological opacity. I can see no logical contradiction in the use of a decelerating/interrupting discourse tool variously to highlight a continuous or resumed topic, to parenthesise ancillary information and to spotlight particularly significant events. For all such effects the author exploits the non-routine XV order, which disrupts normal progress, but leaves the reader to decode the intended effect(s). 80 Moshavi, 2010, p. 119. 81 Since, among other things, languages tend to preserve examples of constituent orders inherited from obsolete stages or borrowed from foreign linguistic systems, orders that are exceptional within contemporary usage rather than representative thereof, the likelihood that any purely pragmatic system can fully account for all word order variation seems small. 53 entire clause, at least two complementary pragmatic dimensions must be posited. On the relevance of one of these – information structure – to BH, there is broad consensus. The exact nature of the other dimension is disputed. Several potential candidates have been proposed. Here it is argued that the notion of ‘discourse continuity’ is sufficiently flexible, because of its indirect link to real-world semantics, and at the same time broad enough to cover a range of effects associated with fronting employed to mark the entire clause in BH. Finally, the apparent problem of positing a system in which all alternatives are pragmatically marked is resolved through recognising action continuity as an inherent property of various types of BH texts and the encoding of continuity in the so-called waw-consecutive forms, whereby they emerge as default, their waw-less counterparts as marked. References Bailey, N. A. and Levinsohn, S. H., 1992, “The function of preverbal elements in independent clauses in the Hebrew narrative of genesis”, Journal of Translation and Textlinguistics 5/3, pp. 179–207. Bandstra, B. L., 1992, “Word order and emphasis in Biblical Hebrew narrative: syntactic observations on Genesis 22 from a discourse perspective”, in W. R. Bodine (ed.), Linguistics and Biblical Hebrew, Winona Lake, IN: Eisenbrauns, pp. 109– 123. Bhat, D. N. S. 1999. The Prominence of Tense, Aspect, and Mood, Amsterdam: John Benjamins. Buth, R., 1992, “The Hebrew verb in current discussions”, Journal of Translation and Textlinguistics 5/2, pp. 91–105. ———, 1994a, “Contextualizing constituent as topic, non-sequential background and dramatic pause: Hebrew and Aramaic evidence”, in E. Engberg-Pedersen, L. F. Jakobsen and L. S. Rasmussen (eds), Function and Expression in Functional Grammar, Berlin: Mouton de Gruyter, pp. 215–231. ———, 1994b, “Methodological collision between source criticism and discourse analysis: the problem of ‘unmarked temporal overlay’ and the pluperfect/nonsequential wayyiqtol”, in R. D. Bergen (ed.), Biblical Hebrew and Discourse Linguistics, Dallas, TX: Summer Institute of Linguistics, pp. 138–154. ———, 1995, “Functional Grammar, Hebrew and Aramaic: an integrated, textlinguistic approach to syntax”, in W. R. Bodine (ed.), Discourse Analysis of Biblical Literature: What It Is and What It Offers, Atlanta, GA: Scholars Press, pp. 77– 102. ———, 1999, “Word order in the verbless clause: a generative-functional approach”, in C. Miller (ed.), The Verbless Clause in Biblical Hebrew: Linguistic Approaches, Winona Lake, IN: Eisenbrauns, pp. 79–108. ———, 2005, Living Biblical Hebrew: Introduction Part Two – ʾulpan le-ʿivrit miqraʾit bet, Jerusalem: Biblical Language Center. Cohen, O., 2013, The Verbal Tense System in Late Biblical Hebrew Prose, Winona Lake, IN: Eisenbrauns. Cook, J. A., 2004, “The semantics of verbal pragmatics: clarifying the roles of wayyiqtol and weqatal in Biblical Hebrew prose”, JSS 49, pp. 247–273. 54 ———, 2012, Time and the Biblical Hebrew Verb: The Expression of Tense, Aspect, and Modality in Biblical Hebrew, Winona Lake, IN: Eisenbrauns. Dallaire, H., 2014, The Syntax of Volitives in Biblical Hebrew and Amarna Canaanite Prose, Winona Lake, IN: Eisenbrauns. DeCaen, V., 1999, “Verbal and verbless clauses within Government Binding Theory”, in C. Miller (ed.), The Verbless Clause in Biblical Hebrew: Linguistic Approaches, Winona Lake, IN: Eisenbrauns, pp. 109–131. Dooley, R. A. and Levinsohn, S. H., 2001, Analysing Discourse: A Manual of Basic Concepts, Dallas, TX: SIL International. Gesenius, W., 1910, Gesenius’ Hebrew Grammar, E. Kautzsch (ed.), A. E. Cowley (trans.), Oxford: OUP. Givón, T., 1983, “Topic continuity in discourse: an introduction”, in T. Givón (ed.), Topic Continuity in Discourse: A Quantitative Cross-Language Study, Amsterdam: John Benjamins, pp. 1–41. Greenberg, M., 1965, Introduction to Hebrew, Englewood Cliffs, NJ: Prentice-Hall. Gross, W., 1999, “Is there really a compound nominal clause in Biblical Hebrew?”, in C. Miller (ed.), The Verbless Clause in Biblical Hebrew: Linguistic Approaches, Winona Lake, IN: Eisenbrauns, pp. 19–49. Hatav, G., 1997, The Semantics of Aspect and Modality: Evidence from English and Biblical Hebrew, Amsterdam: John Benjamins. Heimerdinger, J. M., 1999, Topic, Focus and Foreground in Ancient Hebrew Narratives, Sheffield: Sheffield Academic. Holmstedt, R. D., 2003, “Adjusting our focus: a review of Focus Structure in Biblical Hebrew: A Study of Word Order and Information Structure, by Katsuomi Shimasaki”, HS 44, pp. 101–113. ———, 2005, “Word order in the book of Proverbs”, in R. L. Troxel, K. G. Friebel and D. R. Magary (eds), Seeking Out the Wisdom of the Ancients: Essays Offered to Honor Michael V. Fox on the Occasion of His Sixty-Fifth Birthday, Winona Lake, IN: Eisenbrauns, pp. 135–154. ———, 2009, “Word order and information structure in Ruth and Jonah: a generative-typological analysis”, JSS 54/1, pp. 111–139. ———, 2011, “The typological classification of the Hebrew of Genesis: subject-verb or verb-subject?”, JHS 11/14, http://www.jhsonline.org/Articles/article_161.pdf. ———, 2016, The Relative Clause in Biblical Hebrew, Winona Lake, IN: Eisenbrauns. Hornkohl, A. D., 2005, “The pragmatics of the x+verb structure in the Hebrew of Genesis: the linguistic functions and associated effects and meanings of intraclausal fronted constituents”, Ethnorêma 1, pp. 36–122. ———, 2014, Ancient Hebrew Periodization and the Language of the Book of Jeremiah: The Case for a Sixth-Century Date of Composition, Leiden: Brill. ———, 2016, “Review of K. M. Penner, The Verbal System of the Dead Sea Scrolls”, Revue de Qumrân 28/2, pp. 304–309. ———, in press, “Review of H. Dallaire, The Syntax of Volitives in Biblical Hebrew and Amarna Canaanite Prose”, JSS. Joosten, J., 1989, “The predicative participle in Biblical Hebrew”, ZAH 2, pp. 128– 159. ———, 1992, “Biblical weqāṭal and Syriac hwā qāṭel expressing repetition in the past”, ZAH 5, pp. 1–14. ———, 1997, “The indicative system of the Biblical Hebrew verb in its literary exploitation”, in E. van Wolde (ed.), Narrative Syntax and the Hebrew Bible, Leiden: Brill, pp. 51–71. 55 ———, 2002, “Do the finite verbal forms in Biblical Hebrew express aspect?”, JANES 29, pp. 49–70. ———, 2012, The Verbal System of Biblical Hebrew: A New Synthesis Elaborated on the Basis of Classical Prose, Jerusalem: Simor. Joüon, P., 1923, Grammaire de l’hébreu biblique, Rome: Pontifical Biblical Institute Press. ——— and Muraoka, T., 2006, A Grammar of Biblical Hebrew, Rome: Pontifical Biblical Institute Press. Khan, G. A., 1988, Studies in Semitic Syntax, Oxford: OUP. Lambrecht, K., 1994, Information Structure and Sentence Form: Topic, Focus, and the Mental Representation of Discourse Referents, Cambridge: CUP. Longacre, R. E., 1989, Joseph: A Story of Divine Providence – A Text-Theoretical and Textlinguistic Analysis of Genesis 37 and 39–48, Winona Lake, IN: Eisenbrauns. ———, 1994, “Weqatal forms in Biblical Hebrew prose: a discourse-modular approach”, in R. D. Bergen (ed.), Biblical Hebrew and Discourse Linguistics, Dallas, TX: Summer Institute of Linguistics, pp. 50–98. van der Merwe, C. H. J., Naudé, J. A. and Kroeze, J. H., 2000, A Biblical Hebrew Reference Grammar, Sheffield: Sheffield Academic. Miller, C. L. (ed.), 1999. The Verbless Clause in Biblical Hebrew, Winona Lake, IN: Eisenbrauns. Moshavi, A., 2010, Word Order in the Biblical Hebrew Finite Clause, Winona Lake, IN: Eisenbrauns. Muraoka, T., 1985, Emphatic Words and Structures in Biblical Hebrew, Jerusalem: Magnes and Leiden: Brill. Penner, K. M., 2015, The Verbal System of the Dead Sea Scrolls: Tense, Aspect, and Modality in Qumran Hebrew Texts, Leiden: Brill. Robar, E., 2014, The Verb and the Paragraph in Biblical Hebrew: A Cognitive-Linguistic Approach, Leiden: Brill. Shimasaki, K., 2002, Focus Structure in Biblical Hebrew: A Study of Word Order and Information Structure, Bethesda, MD: CDL. ———, 2013, “Disjoining in discourse”, in G. Khan (ed.), Encylopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. I, pp. 763–766. Siewierska, A., 1988, Word Order Rules, New York, NY: Croon Helm/Methuen. 56 Long or Short? The Use of Long and Short Wayyiqṭols in Biblical, Parabiblical and Commentary Scrolls from Qumran JOHAN M. V. LUNDBERG University of Cambridge 1 The diachrony of Biblical Hebrew verbal morphology and copying practices1 The Biblical Hebrew verbal system has received much attention in the last century, especially the functions and meanings of the prefix conjugations (i.e. short, long and lengthened yiqṭol)2 vis-à-vis those of the suffix conjugation. The short yiqṭol was originally perfective and consequently functioned as a preterite in narrative texts.3 This changed, however, when the suffix conjugation was grammaticalised as a perfective or perfect. As a result, the short form was replaced and only remained in volitive forms (i.e. jussives) and in wayyiqṭols (i.e. consecutive imperfects). The volitive function is mostly preserved in third person forms and second person negative imperatives. Wayyiqṭol forms, on the other hand, exist in a complementary system together with the suffix conjugation and dominate the narrative sections of the Hebrew Bible.4 This role in narratives was taken over by the suffix conjugation in Rabbinic Hebrew.5 In Late Biblical Hebrew and the non-biblical Dead Sea Scrolls, the use of wayyiqṭol reflects its classical usage even though much of the modal system had collapsed.6 1 I would like to thank Holger Gzella, Aaron Hornkohl and Martin Heide for their insightful comments and suggestions during the process of writing this article. All errors remain my own. 2 The lengthened yiqṭol is used in first person forms, primarily for marking volition (i.e. the cohortative) but it is also found in wayyiqṭol forms (Hornkohl, 2014, p. 159). Discussion of the lengthened form is mostly left out of this article because it is not attested with III-y verbs (cf. Hornkohl, 2014, p. 161 n. 6, 163 n. 15). 3 See Cook, 2012, pp. 256–265 for this function; cf. Gzella, forthcoming for a summary of the North-West Semitic material. 4 Cf. Gzella, forthcoming. 5 See Geiger, 2013, pp. 740–741 for a short survey of tense in relation to the suffix and prefix conjugations in Rabbinic Hebrew. 6 Fassberg, 2013, pp. 666–667; Hornkohl, 2014, pp. 256–257. 57 Long and short yiqṭols were originally distinguished by word-final unstressed short vowels in all singular persons except the feminine. When these vowels disappeared at the beginning of the first millennium BCE, most long and short forms merged.7 The distinction was only retained in some weak roots and in the causative stem. This leads to the topic of this article. Evidence from late biblical compositions indicates that authors or scribes sometimes used long wayyiqṭols instead of their short counterparts. For example, the long form wayyaʿănε is used in 2 Kings 1:10 while wayyaʿan is used in an identical construction two verses later.8 Moreover, these long forms are used with different types of verbs and have the same semantic function as distinctly short wayyiqṭols.9 This pattern points to a process of morphological erosion suggesting that long and short wayyiqṭols were becoming interchangeable. At the same time, the text in Chronicles contain several passages where a short form corresponds to the long form used in Kings.10 The aim of this article is to investigate how this process of morphological erosion influenced the copying practices of the scribes who made the Dead Sea Scrolls. To achieve this goal the following will be addressed: Was the use of long and short wayyiqṭols systematic or incidental? a. Were scribes consistent in their use of long and/or short wayyiqṭols? b. Were long forms used more often with specific verbs or verb types? c. Is it possible to determine whether scribes ‘classicised’ long forms (i.e. exchanged long wayyiqṭols for short ones)? d. Are there observable differences between manuscripts, genres and different kinds of documents (e.g. between biblical texts, rewritten biblical texts or pesharim)? In short, this comparison of long and short wayyiqṭols serves as a lens elucidating some scribal habits (especially copying practices). 1.1 Verb types with distinctive short forms As was noted above, vestiges of the previous distinction between long and short forms only remained in three contexts after the loss of the final short vowels: (1) in all verbal stems of III-y verbs except pual and hophal; (2) in 7 Cf. Gzella, forthcoming. For further discussion of these forms see Gzella, forthcoming; cf. Hornkohl, 2014, pp. 171– 172. 9 Gzella, forthcoming. 10 2 Chr. 18:23, 33, 34; 2 Chr. 21:9; and 2 Chr. 34:27 can be compared with 1 Kings 22:24, 34, 35; 2 Kings 8:21; and 2 Kings 22:18. Gzella, forthcoming notes that there is a tendency in Chronicles to classicise the language of Kings; Bloch, 2007, p. 156 notes that if the first of two parallel clauses contains a long wayyiqṭol from a III-y root, the second clause invariably has a short form; cf. Hornkohl, 2014, p. 177 n. 65. 8 58 hollow roots; and (3) in the causative stem of triconsonantal roots.11 Long forms of III-y verbs end in a final short vowel while short forms end with a closed syllable.12 The presence of the final vowel is indicated by the vowel letter ‫ ;ה‬e.g. wayyaʿaś (‫ ; ַו ַיעַש‬Gen. 19:3) and yaʿăśε (‫ ; ַיעֲשֶּ ה‬Gen. 18:25). A ‫ ו‬or ‫ י‬was often used in long forms of hollow roots to mark the presence of a long vowel (/ū/ or /ī/); e.g. wayyå̄måṯ (‫ ; ַויָמָ ת‬Gen. 25:8) and yå̄mūṯ (‫ ;יָמות‬Gen. 38:11) or wayyå̄śεm (‫ ; ַויָשֶ ם‬Gen. 4:15) and yå̄śīm (‫ ;י ִָשים‬Gen. 30:42). Lastly, long and short forms in the causative stem can be distinguished through the length and quality of their theme vowel. Short forms have a short i-vowel or an e-vowel, the latter being represented by ṣere in the Masoretic text (MT). By contrast, long forms are characterised by a long i-vowel.13 In the MT, a ‫ י‬is usually used to mark this /ī/, distinguishing long forms from short ones; e.g. short wayyaḇdēl (‫ ; ַויַבְ דֵ ל‬Gen. 1:7) and long yaḇdīl (‫ ;יַבְ דִ יל‬Lev. 1:7). Because the difference between these forms is a matter of vocalisation, it is in many cases impossible to distinguish between long and short forms without the help of vowel points. For a study of the Dead Sea Scrolls, which only includes the consonantal text, it is therefore necessary to narrow down the corpus. The focus of this article will be on III-y verbs, as they are the least ambiguous. In the other cases, long and short forms are often indistinguishable. In hollow roots a ‫ ו‬or ‫ י‬could be used to mark short vowels. Similarly, a ‫י‬ in the hiphil could mark an e-vowel rather than /ī/. Additionally, it is important to note that first person wayyiqṭols outside the Torah tend to be long even in the MT.14 It is, therefore, relevant to investigate whether scribes were more prone to add a ‫ ה‬to first person forms than to other forms; or whether they classicised long first person wayyiqṭols. 1.2 The corpus This article covers relevant wayyiqṭol forms in three types of Qumran manuscripts: 1. Biblical texts. This category consists of the fragments from various books of the Hebrew Bible. It is by far the largest category and contains the bulk of the material.15 11 See Gzella, forthcoming for further bibliographic references; cf. Waltke and O’Connor, 1990, pp. 543–544 for additional forms with a different vocalisation. 12 Joüon and Muraoka, 2006, pp. 190–191. 13 Joüon and Muraoka, 2006, p. 148. 14 Cf. Hornkohl, 2014, p. 160–161. The use of short first person forms in the Torah and the Mesha stele show that short first person forms are more ancient than long first person forms. 15 The Biblical Qumran Scrolls: Transcriptions and Textual Variants (Ulrich, 2010) has greatly facilitated the collection of the data, serving as the point of departure. References to biblical manuscripts follow the same system as this edition. 59 2. Parabiblical texts. These texts are sometimes termed ‘rewritten bible’, and manuscripts within this category often contain clauses, sentences or paragraphs from biblical texts. Significant parts of these texts, however, do not occur in the Hebrew Bible. Instead, the text has been rephrased and expanded. For this article, the focus will be on quotes in comparison with the MT. 3. Commentaries and exegetical texts. The commentaries known as pesharim have a basic structure. First a biblical prophecy is quoted and then it is applied to the community at Qumran and/or their adversaries. Apart from these texts, there are also several other types of exegetical works referencing biblical passages and explaining their content. As with the parabiblical texts, this study will focus on quotations from the Hebrew Bible. 2 2.1 Long and short wayyiqṭols III-y Long wayyiqṭols are easily identifiable in the consonantal text because the presence of the final vowel is marked by the vowel letter ‫ה‬. These forms are used more frequently in Kings, Ezekiel and Jeremiah than in any other part of the Hebrew Bible; especially compared to the Torah, which only contains a few examples (Gen. 25:48; Deut. 1:16, 18).16 It would therefore be interesting to compare the use of long forms in Qumran manuscripts of these three books with the MT. However, the fragmentary nature of the material preserved at Qumran makes a reliable comparison impossible. Other manuscripts will provide the bulk of the data presented here, most prominently 1QIsaa. 2.1.1 Biblical texts ‫היה‬ Short wayyiqṭols of ‫ היה‬are used very frequently in the MT. These forms can therefore be easily found in the biblical Dead Sea Scrolls. In total, there are seventy-one attestations of the third person masculine singular ‫ויהי‬.17 The high 16 See Gzella, forthcoming. He also notes that the concentration of long forms in Ezekiel is higher than in Kings, and that the picture could be skewed by the general lack of wayyiqṭols in the former. He writes: “When one compares the situation in 1–2 Kings with other narrative compositions like 1–2 Samuel, it becomes evident that the former is exceptional.” While it is possible that the picture is skewed, it should be noted that the non-narrative style could have contributed to the use of long forms. Cf. Hornkohl, 2014, p. 161 n. 6. 17 Attestations occur in the following manuscripts: 4QGen b (1:5, frg. 1i, 5; 1:7, frg. 1i, 8; 1:8, frg. 1i, 9; 1:9, frg. 1i, 11; 1:11, frg. 1i, 13; 1:13, frg. 1i, 15; 1:19x2, frg. 1i, 22; 4:3 frg. 3i, 2); 4QGenc (40:8, frg. 1ii, 15); 4QGend (1:19, frg. 1, 1; 1:23, frg. 1, 6); 4QGen f (48:1, frg. 1, 1); 60 frequency of this form is a result of its use as a discourse marker. The third person feminine form ‫ ותהי‬is attested seven times in the scrolls, one of which is used instead of the first person form ‫ואהי‬.18 Since ‫ ויהי‬was used with such frequency in the biblical corpus, it is not surprising that some manuscripts contain alternative forms. In one instance the plural ‫ ויהיו‬is used instead of ‫ותהי‬: 1QIsaa (9:18, col. IX, 9).19 More interesting, though, is the occasional use of ‫ ויהי‬instead of ‫( והיה‬4QSama: 2 Sam. 6:16, frgs. 68–76, 21) and the reverse (1QIsaa: 22:7, col. XVII, 10; 48:18, col. XL, 23;20 4QPsq: 33:9, col. I, 7).21 These switches suggest that the classical system of wəqaṭal and wayyiqṭol had collapsed or was in the process of collapsing.22 Consequently, scribes sometimes failed to use ‫ ויהי‬and ‫ והיה‬with the correct function as discourse markers (perhaps conditioned by vernacular Hebrew or Aramaic).23 Furthermore, it should be noted that the replacement of ‫ ויהי‬with 4QGeng (1:6, frg. 1, 6; 1:7, frg. 1, 8; 1:9, frg. 1, 11; 1:13, frg. 2, 1; 1:15, frg. 2, 4); 4QGen k (1:15, frg. 2, 4); 4QGen-Exoda (Gen. 37:23, frg. 8, 2; 39:18, frg. 9, 7; 39:19, frg. 9, 8; 39:20, frg. 9, 10; Ex. 7:10, frg. 33, 7); 4QpaleoExodm (6:28, col. I, 5; 8:14, col. III, 26; 32:30, col. XXXVIII, 32); 4QExodc (7:21, col. I, frgs. 1–4, 17; 9:22, col. II, frgs. 11–19, 31; 12:41x2, col. V, frg. 32, 10, 11; 17:15, col. VIII, frgs. 35–36, 23); 4Exod-Levf (Ex. 40:17, col. II, frg. 2ii, 12); 4QNumb (31:32, col. XXVI, frgs. 55ii, 57–59, 20; 31:32, col. XXVII, frgs. 60–64, 7); 4QDeuth (1:3, frg. 1, 3; 4:33, frg. 8, 5); 4QDeutn (5:23, col. V, 3; 5:26, col. V, 8); 4QDeutj (5:26, col. III, 9); 4QpaleoDeutr (33:5, frgs. 42–43, 3); 4QJoshb (17:13, frg. 5, 5); 1QJudg (9:42, frgs. 7–8, 3); 4QSama (1 Sam. 5:9, col. V, frgs. b–c, 5; 2 Sam. 3:6, frgs. 55–57a–b, 58, 12; 4:4, frgs. 61i–62, 33; 11:16, frgs. 93–94, 1); 4QSamb (1 Sam. 20:27, frgs. 6–7, 1); 4QSamc (2 Sam. 15:1, col. II, frgs. 5ii–7i, 21); 1QSam (2 Sam. 21:18, frgs. 4–7, 4); 1QIsaa (7:1, col. VI, 11; 36:1, col. XXVIII, 29; 37:1, col. XXX, 4; 37:38, col. XXXI, 18; 38:4, col. XXXI, 24; 38:9, col. XXXII, 1; 48:19, col. XL, 24; 63:8, col. LI, 4); 1QIsab (48:19, col. XXI, 3; 63:8, col. XXVII, 17); 4QIsad (48:19, col. IV, frgs. 6–10, 3); 4QIsae (12:2, frgs. 17, 18i, 19, 2); 4QJera (13:6, col. VIII, part 1, 5; 33:19, col. XXV, 4); 2QJer (42:7, frg. 1, 1); 4QXIIa (Jon. 1:4, col. V, 9); 4QXIIg (Jon. 1:1, frgs. 76–78i, 79–81, 4; 3:1 frgs. 78ii, 82–87, 10); 4QRutha (1:1x2, 1); 2QRutha (2:17, col. II, 5); 1QDana (1:16, col. I, 8). The first part of the verse in 4QSamc is inserted above the line as a later addition. 18 4QSama (1 Sam. 2:17, col. III, frgs. a–e, 5; 5:9, col. V, frgs. b–c, 5); 1QSam (2 Sam. 23:11, frg. 8, 4); 1QIsaa (23:3, col. XVIII, 7; 59:15, col. XLVIII, 27); 4QIsaa (23:3, frgs. 11ii, 15, 28). The attestation which corresponds to ‫ ואהי‬occurs in 4QPsa (69:12, col. III, frgs. 19ii–20, 31). 19 Similarly, the form ‫ ותהי‬has been replaced by ]...[‫ ויה‬in 4QSama (2 Sam. 8:2, frgs. 80–83, 3). The lone example of the verb ‫ זנה‬might also illustrate this phenomenon. The MT has ‫ ותזן‬but in 1QIsaa (57:3, col. XLVII, 3) it was written ‫ ותזנו‬or ‫ותזני‬. Parry and Qimron, 1998, p. 95 read it as the former. Formally it could also be ‫ ותזני‬since the last two letters are similar to those of the verb ‫ בנה‬in the same line. 20 In 1QIsaa (12:2, col. XI, 8), ‫ היהא‬is used instead of ‫ויהי‬. The same switch from the prefix conjugation to the suffix conjugation also occurs with plural forms. In 1QIsa a (66:2, col. LIII, 3), ‫ והיו‬is used instead of ‫ויהיו‬, which is used in the MT and in 1QIsab (66:2, col. XXVII, 10). 21 Another example of a switch from ‫ ויהי‬to ‫ והיה‬could occur in 4QSama (2 Sam. 6:13, frgs. 68– 76, 18), but several of the letters are damaged and only the downstrokes are visible on the photograph (frg. 75, pl. XVI in DJD XVII, Cross et al., 2005). From the photograph, it would perhaps be possible to reconstruct ‫ויהי‬. 22 Fassberg, 2013, pp. 666–667 uses Isa. 22:7 as an example of how the “classical system of consecutive tenses had collapsed”. 23 Cook, 2012, pp. 309–312 notes that these forms were used both as copula predicates as well as discourse markers in narratives; cf. Kutscher, 1974, pp. 350–358, which contains a discussion of the collapse of “consecutive tenses” in 1QIsaa. 61 ‫ והיה‬in Isaiah is more common than the reverse, while the opposite is true in 2 Samuel. It is therefore possible that scribes (consciously or unconsciously) switched to the discourse marker that was most common in the book they copied, or failed in their attempt to imitate the classical style (i.e. hypercorrection).24 The Qumran material contains five examples that look like typical long forms. The form ‫ ותהי‬is replaced by ‫ ותהיה‬in both 1QIsaa (5:25, col. V, 13; 29:11, col. XXIII, 19; 29:13, col. XXIII, 23) and 4QIsaf (5:25, frg. 8, 1). It is noteworthy that two of these attestations occur in the same verse. However, their use in 1QIsaa, which also contains several short forms (both ‫ ויהי‬and ‫)ותהי‬, is more important. In addition to the aforementioned long forms, this manuscript also contains ‫( ונהיה‬64:5, col. LI, 19) and ‫( ותהיי‬9:5, col. VIII, 23).25 It is important to note that these long forms are third person feminine singular or first person plural. The form ‫ ויהי‬is sometimes changed, but never through the addition of the vowel letter ‫)ויהיה( ה‬. ‫עשה‬ Wayyiqṭols of the verb ‫ עשה‬are used frequently in the MT. The biblical scrolls contain seventeen attestations of distinct short forms with a short counterpart in the MT.26 Two manuscripts exhibit three long forms instead of the short ones used in the MT: 1QSam (2 Sam. 23:10, frg. 8, 3) and 1QIsa a (5:2, col. IV, 14; 5:4, col. IV, 16).27 Three observations are relevant. First, there are no distinct short forms of ‫ עשה‬in these manuscripts. Secondly, the fragment from 1QSam includes four other wayyiqṭols, two long (‫ ;ויכה‬see ‫ נכה‬below) and two short (‫ ויהי‬and ‫)ותהי‬. Thirdly, the two attestations in 1QIsaa occur in the song of the vineyard, which is a short narrative within the prophetic discourse. Two explanations should be considered: the non-narrative character of Isaiah could have caused the scribe to use long forms; or it is possible that the scribe(s) consistently used the long form ‫ ויעשה‬of this verb.28 24 Cf. Kutscher, 1974, p. 329; Elwolde, 2015, p. 100. Parry and Qimron, 1998, p. 17 note that there is no space between the verb and the following word in 9:5, and they also suggest that someone might have tried to erase the additional ‫י‬. 26 4QGenb (1:16, frg. 1i, 18); 4QpaleoGen-Exod l (Ex. 18:25, frg. 20, 13); 4QpaleoExodm (7:6, col. I, 17; 9:6, col. V, 3; 37:10, col. XLV, 3; 37:11, col. XLV, 5; 37:12, col. XLV, 5; 37:12, col. XLV, 6; 37:15, col. XLV, 9); 4QExod-Levf (Ex. 39:22, col. II, frg. 1ii, 6; 40:16, col. II, frg. 2ii, 10); 1QJudg (6:20, frg. 1, 2); 4QSama (2 Sam. 15:1; frgs. 112–114a–b, 1); 1QIsaa (20:2, col. XVI, 6); 4QIsaa (20:2, frgs. 10, 11i, 12–14, 18); 4QIsab (20:2, frgs. 10–13, 40); 4QChr (2 Chr. 29:2, frg. 1ii, 4). Occasionally, short forms are changed. Verb forms that were changed include 4QSamc (2 Sam. 15:1, col. II, frgs. 5ii–7i, 21), where the text reads ]‫ ואבשלום יע[שה‬instead of ‫ויעש‬, which is used in the MT. The text in 4QSama is broken after the last letter, so the form could theoretically be long. 27 It is also noteworthy that the verb in Isa. 5:4 is written ‫ וישה‬without the initial root letter ‫ע‬. 28 Elwolde, 2015, p. 100, following Kutscher, 1974, p. 328, notes the relative frequency of these forms in 1QIsaa. 25 62 ‫ראה‬ The verb ‫ ראה‬is used frequently in the Hebrew Bible. The Qumran corpus contains thirty wayyiqṭols, the clear majority of which (twenty-five of them) have a distinct short form. These short forms are primarily distributed throughout narrative sections of the Hebrew Bible.29 In 4QXIIe (Zech. 5:9, frgs. 14–15, 2) there is one attestation of a long wayyiqṭol with a short counterpart in the MT. In this instance, it is also noteworthy that this verbal form is added in a smaller script above the line. If this was a later addition, it is possible that the author wrote the verb from memory and therefore used a long form rather than a short one. Moreover, there are four attestations where long forms are used in both the MT and Qumran.30 1QIsaa contains both short and long wayyiqṭols of ‫ ראה‬mirroring the MT. One of the short forms also agrees with another example in 4QIsae. The long form in 1QIsaa is a first person wayyiqṭol, just like the three long forms in 4QEzeka, 4QDana and 4QDanc. This indicates that these manuscripts exhibit the same tendency as the MT in the use of long first person wayyiqṭols. ‫עלה‬ The verb ‫ עלה‬is attested fourteen times in the biblical manuscripts, twelve of which are short.31 Long wayyiqṭols of ‫ עלה‬are rare in the MT. It is therefore not surprising that only one manuscript contains an example of a long form that also corresponds to a long form in the MT: 4QJerc (10:13, col. V, 2). Additionally, 1QIsaa (37:14, col. XXX, 19) contains a long form that corresponds 29 4QGenb (1:10, frg. 1i, 11; 1:12, frg. 1i, 14; 1:18, frg. 1i, 21; 1:21, frg. 1i, 26); 4QGen d (1:18, frg. 1, 1); 4QGenf (48:8, frg. 1, 13); 4QGeng (1:4, frg. 1, 3; 1:18, frg. 2, 8); 4QGenk (1:9, frg. 1, 1); 4QExodb (2:5, frgs. 3i–4, 4; 2:12, frgs. 3i–4, 13); 4QpaleoGen-Exodl (Ex. 3:4, frgs. 3–4, 12); 4QLev-Numa (Num. 22:23, frg. 64, 2); 4QNumb (24:1, col. XVII, 24 ii, 27–30, 12); 4QDeuth (33:21, frgs. 11–15, 11); 1QJudg (6:22, frg. 1, 4); 4QJudga (6:12, frg. 1, 7); 4QSama (2 Sam. 6:16, frgs. 68–76, 22; 18:10, frgs. 128–132, 19; 20:12, frgs. 144–145, 6); 4QSamb (1 Sam. 16:6, frg. 4, 5); 1QIsaa (59:15, col. XLVIII, 28; 59:16, col. XLVIII, 28); 4QIsa e (59:16, frg. 25, 1). The last letter and the beginning of the verbal form in 4QJudga is damaged. 30 1QIsaa (6:1, col. V, 21); 4QEzeka (11:1, frg. 2, 10); 4QDan a (8:2, frg. 14, 12) 4QDan c (10:8, col. I, 7). The initial ‫ ו‬in 1QIsaa was omitted but the function still indicates that this is a wayyiqṭol. The beginning of the verb in 4QEzeka is within brackets in the edition and the last two letters are marked as uncertain. The photograph contains only traces of what could be the combination ‫אה‬. It is also worth noting that an additional ‫ ה‬is added after verbs ending in ‫א‬: cf. ‫ואשאה‬ instead of ‫ ואשא‬in Dan 8:3 (4QDanb, frgs. 16–18i, 19, 5). 31 4QpaleoGen m (26:23, 3); 4QpaleoExodm (24:9, col. XXVI, 31); 4QNumb (23:30, col. XVII, 24 ii, 27–30, 12); 1QJudg (6:21, frg. 1, 3); 4QSama (1 Sam. 11:1, col. X, frg. a, 9; 2 Sam. 6:17, frgs. 68–76, 24); 6QpapKgs (1 Kings 22:29, frg. 5, 2); 1QIsaa (53:2, col. XLIV, 5); 1QIsab (53:2, col. XXIII, 10); 4QXIIa (Jon. 2:7, frg. 21, 2); 4QXIIg (Jon. 2:7, frgs. 78ii, 82–87, 6; 4:6, frgs. 88–91i, 3). The example from 4QpaleoExodm is probably short, but there is a break in the text partly damaging the ‫ל‬, so that it is impossible to discern whether the form was long or short. Part of the example in 1 Sam. 11:1 in 4QSama was added in smaller letters above the line. The first two letters of the verb in 6QpapKgs are damaged but the last two are intact showing that the form is short. After the ‫ ל‬in 1QIsaa there is an erased letter that could be a ‫ ו‬or the right side of a ‫ה‬. 63 to a short one in the MT. The presence of this form and the short one in Isa. 53:2 points to an inconsistency in the use of long and short forms of the same verb in 1QIsaa. Perhaps this inconsistency was caused by the small number of wayyiqṭol forms being used in some parts of the scroll. ‫נטה‬ The verb ‫ נטה‬is attested seven times, five of which have distinct short forms.32 It is noteworthy that one example is from Isa. 5, which contains several long wayyiqṭols.33 The two long forms are attested in 4QExodc (10:13, col. III, frgs. 20–29, 15) and 4QpaleoGenm (26:25, 5).34 These two forms are interesting because they occur in passages of Genesis and Exodus and are the only long forms from Torah manuscripts. These attestations are important because they show that there are occasional long forms even in Torah manuscripts (with which scribes are known to have exercised great caution in terms of preservation). ‫נכה‬ There are five wayyiqṭols of ‫ נכה‬in the corpus. Three short forms are attested in 4QSama (1 Sam. 5:9, col. V, frgs. b–c, 6), 1QIsaa (37:36, col. XXXI, 16) and 11QPsa (105:33, col. I, 735). It is noteworthy that the only wayyiqṭol of this verb in 1QIsaa is short, since the MT has the long form ‫ויכה‬.36 Furthermore, two long forms corresponding to short forms in the MT are attested in 1QSam (2 Sam. 23:10, frg. 8, 2; 23:12, frg. 8, 5).37 ‫ענה‬ There are ten wayyiqṭols of ‫ ענה‬in the corpus, nine of which are short.38 Not surprisingly, the only long form occurs in 1QIsaa (21:9, col. XVI, 24). The odd spelling of this form is noteworthy (‫)ויעני‬. 32 4QGen-Exoda (Gen. 35:21, frg. 5, 4; 39:21, frg. 9, 11); 4QpaleoExod m (10:22, col. VII, 29); 4QNumb (22:33, col. XV, frgs. 20–22, 29); 1QIsaa (5:25, col. V, 12). 33 Note, however, that the long forms occur in the song of the vineyard, unlike the short form of ‫נטה‬. 34 The long form in 4QpaleoGen m is written ‫ ויטי‬with the second ‫ י‬as a vowel letter. Note, however, that this letter is damaged. 35 The first letters of this attestation are gone, but the final ‫ ך‬is visible indicating that it was a short form. 36 Other third person long forms of this verb occur in Josh. 10:40; 1 Kings 22:24, 34; 2 Kings 2:8, 14, 14; 2 Kings 8:21; Jer. 20:2; Jer. 52:27. 37 The second form is written ‫ ויה‬instead of ‫ויכה‬, with a ‫ כ‬inserted above the line. 38 4QGen-Exoda (Gen. 27:39, frg. 2, 1); 4QExodc (15:21, col. VI, frg. 33ii, 43); 4QNumb (22:18, col. XV, frgs. 20–22, 11); 4QSama (2 Sam. 13:32, frgs. 102ii, 103–106i, 107–109a–b, 29); 4QSamb (1 Sam. 20:28, frg. 6–7, 2; 21:5, frg. 6–7, 16); 4QIsaa (21:9, frgs. 10, 11i, 12–14, 29); 4QXIIc (Am. 7:14, frgs. 30–33, 15); 4QXIIe (Zech. 6:5, frgs. 14–15, 11). There is a break after the ‫ נ‬in 4QSama. The last letter could be a final ‫ ן‬since there is no trace of a ‫ה‬. The first two letters in 4QSamb are slightly damaged, but the end of the verb is preserved, showing the distinct short form. 64 ‫בכה‬ There are five wayyiqṭols of ‫ בכה‬in the corpus, four of which are short.39 One of these short forms is noteworthy for two reasons. First, it is written ‫ ואך‬without the initial root consonant ‫ב‬. Moreover, the MT contains the long form ‫ואבכה‬. This could perhaps be an attempt to classicise the psalm. However, if the omission of ‫ ב‬was a scribal mistake, it is less likely that the short form represents a conscious attempt to classicise the text. Lastly, there is one attestation of a long form in 1QIsaa (38:3, col. XXXI, 23), written ‫ ויבכא‬rather than ‫ויבכה‬. ‫בנה‬ The biblical Qumran manuscripts contain four wayyiqṭols of the verb ‫בנה‬, two of which are probably short: 4QGen-Exoda (Ex. 1:11, frgs. 17–18, 8) and 4QExodc (17:15, col. VIII, frgs. 35–36).40 The other two attestations are distinctively long while the MT has short forms: 4QSama (2 Sam. 5:9, frgs. 61 ii, 63–64 a–b, 65–67, 20) and 1QIsaa (5:2, col. IV, 13). Some observations are in order. First, it is noteworthy that the form in 1QIsaa is spelled with an ‫ א‬instead of a ‫)ויבנא( ה‬. Moreover, it is interesting that this long wayyiqṭol occurs in the song of the vineyard together with several other long forms. Lastly, it is noteworthy that the long form in 4QSama is the only one in a manuscript which includes short forms of most verbs in this section (including ‫היה‬, ‫ראה‬, ‫עשה‬, ‫עלה‬, ‫נכה‬, ‫ענה‬, ‫בכה‬, ‫חרה‬, ‫ צוה‬and ‫)חנה‬. This suggests that even scribes who consistently used short forms occasionally made use of long ones. ‫קנה‬ Only one wayyiqṭol of ‫ קנה‬is attested in the corpus and occurs in 4QJera (13:2, col. VIII, part 1, 3). The form is long just like its counterpart in the MT. Other III-y verbs The verbs in this section have two things in common: there are relatively few attestations, and all attestations are short. The verb ‫ חרה‬is attested nine times in the corpus.41 Furthermore, nine attestations of ‫ צוה‬occur in the biblical scrolls, all of which are in the piel.42 The verb ‫ כסה‬is attested three times, in 39 4QGen-Exoda (Gen. 27:38, frg. 2, 1); 4QSama (1 Sam. 24:17, frgs. 26–27, 6; 2 Sam. 3:32, frgs. 61i–62, 15); 4QPsa (69:11, col. III, frgs. 19ii–20, 30). The beginning of this verb in 4QGenExoda is missing, but the last two root consonants are visible, indicating that it is a short form. 40 It should be noted that the text is broken after the ‫ נ‬of ‫ויבן‬. 41 4QGenb (4:5, frg. 3i, 5); 4QGen-Exoda (Gen. 39:19, frg. 9, 9); 4QNumb (12:9, col. I, frgs. 1– 4, 15); 4QLev-Numa (Num. 12:9, frgs. 60–61, 6; 32:13, frg. 68i, 5); 4QSama (2 Sam. 6:7, frgs. 68–76, 8; 6:8, frgs. 68–76, 9); 4QSamb (1 Sam. 20:30, frgs. 6–7, 4). The first example from 4QLev-Numa is relatively damaged. 42 4QGen b (2:16, frg. 1ii, 3); 4QExoda (5:6, frgs. 22ii, 26, 5); 4QExodb (5:6, frg. 6ii, 4); 4QDeutc (27:1, frg. 42, 2); 4QDeutb (31:25, frgs. 5–8, 1); XJosh (1:10, col. I, 22); 4QSama (2 Sam. 11:19, frgs. 93–94, 4; 13:28, frgs. 102ii, 103–106i, 107–109a–b, 22); 2QRutha (2:15, col. II, 2). The 65 4QNumb (22:11, col. XV, frgs. 20–22, 4), 4QSamb (1 Sam. 19:13, frg. 5, 4) and 1QIsaa (37:1, col. XXX, 5). The verb ‫ מנה‬is, likewise, attested three times: once in 4QXIIa (Jon. 2:1, col. VI, 18) and twice in 4QXIIg (Jon. 4:6, frgs. 88– 91i, 2; 4:8, frgs. 88–91i, 6). The verb ‫ קוה‬is attested twice in 4QIsaa (5:2, col. IV, 14; 5:7, col. IV, 20). These examples are especially interesting because they occur in the song of the vineyard, which includes long forms of both ‫ ראה‬and ‫עשה‬. There is no obvious explanation as to why the scribe used both short and long forms in this passage. Other verbs that are attested twice include: ‫ גלה‬in 1QIsaa (22:8, col. XVII, 11) and 2QRutha (3:7, col. V, 6); ‫ הרה‬in 4QIsaa (8:3, col. VII, 21) and 4QIsae (8:3, frgs. 4–10, 1); ‫ חדה‬in 4QExodc (18:9, col. VIII, frgs. 35–36, 34) and 4QpaleoExodm (18:9, col. XVIII, 13); and ‫ פנה‬in 4QExodb (2:12, frgs. 3i–4, 12) and 4QNumb (12:10, col. I, frgs. 1–4, 16). Lastly, ten additional verbs are attested once in the biblical scrolls.43 Summary The pattern that emerges from this survey indicates that short forms were used with relative consistency. Distinctive long forms are found in less than a fifth of the manuscripts that contain wayyiqṭols (4QpaleoGenm, 4QExodc, 1QSam, 4QSama, 1QIsaa, 4QIsaf, 4QJera, 4QJerc, 4QEzeka, 4QXIIe, 4QDana and 4QDanc). Moreover, when long forms are used, short forms of the same or another verb are almost invariably attested in the same manuscript. The main manuscripts that depart from this general pattern are 1QSam, 1QIsa a, 4QIsaf and 4QJerc.44 Unfortunately, only a few fragments have been preserved of 1QSam, and all three long forms are attested in the same fragment (frg. 8; 2 Sam. 23:9–12). 4QIsaf and 4QJerc are likewise very fragmentary. Consequently, it is impossible to know whether the use of long forms reflects the compositional style of these scrolls or if these forms were restricted to a few passages. On the other hand, the fact that three long forms are attested in the same passage of 1QSam, and that the two short forms are ‫ ויהי‬and ‫ותהי‬, could indicate that the scribe generally used long forms with other verbs than ‫היה‬. Alternatively, the manuscript could have contained an equal number of long and short forms (used interchangeably); or long forms could have been used with specific verbs. example in 4QExoda is relatively damaged and it is not clear whether a final ‫ ה‬could have been there in the original. 43 ‫ שתה‬in 2QRutha (3:7, col. V, 6); ‫ שחה‬in 4QSamb (1 Sam. 20:41, frg. 6–7, 12); ‫ שעה‬in 4QGenb (4:4, frg. 3i, 4); ‫ שקה‬in 4QExodb (2:17, frgs. 3i–4, 20); ‫ צפה‬in 4QpaleoExodm (37:15, col. XLV, 10); ‫ אדה‬in 8QPs (18:11, frgs.11–13, 2); ‫ רצה‬in 4QPsc (50:18, col. II, frgs. 15ii–16, 23); ‫ תכה‬in 4QXIIg (Jon. 4:7, frgs. 88–91i, 5); ‫ עטה‬in 1QIsaa (59:17, col. XLIX, 1); ‫ חנה‬in 4QSama (1 Sam. 11:1, col. X, frg. a, 9). There is a break after the ‫ פ‬of ‫ צפה‬in 4QpaleoExodm. The same verb also occurs a few lines before in Ex. 37:11, but the break is after the ‫צ‬. Part of the text in 4QSama was first omitted and then added in smaller letters above the line. 44 Cf. Qimron, 1986, p. 45 n. 8 who only lists the first two, presumably because he refers to Kutscher’s previous study of 1QIsaa. 66 While the three manuscripts discussed above are fragmentary, 1QIsaa is almost complete. The thirty wayyiqṭols that are attested can be divided into three categories: ‫( ויהי‬eight cases), other short forms (eleven) and long forms (ten). The reason for distinguishing between ‫ ויהי‬and other short forms is simple. It is very common and has a distinct function in the discourse, making it less susceptible to change. Moreover, there are no attestations of a form ‫ ויהיה‬in any manuscript from Qumran. The remaining long and short forms fall into two categories of equal size. This use of long wayyiqṭols was already noted by Kutscher. He also observed the occasional odd spelling of these forms and compared them with long wayyiqṭols in the Samaritan Pentateuch. Of particular interest are forms spelled with ‫ י‬or ‫ א‬instead of ‫( ה‬Isa. 21:9, ‫ ;ויעני‬Isa. 5:2, ‫ ;ויבנא‬Isa. 38:3, ‫)ויבכא‬. These spellings are not attested in other Qumran manuscripts, but similar examples exist in other corpora.45 As for the use of long forms in manuscripts of the Torah, it is noteworthy that the biblical Qumran scrolls exhibit the same pattern as the MT. There are a total of fifty-eight wayyiqṭols in twenty-two manuscripts (not counting attestations of ‫)ויהי‬. Only two of these are long, indicating that scribes were consistent in their use of short forms (although some of the manuscripts are very fragmentary).46 As table 1 indicates, most long forms occur in prophetic books. There are two attestations of long forms in manuscripts of the book of Jeremiah, and both correspond to long forms in the MT. Similarly, 4QEzeka contains an attestation of a long wayyiqṭol with a long counterpart in the MT. The data from 4QDana and 4QDanc follow a similar pattern. It is important to note that these verbs are first person wayyiqṭols. It was noted above that the MT tend to use long first person wayyiqṭols outside the Torah.47 Long first person forms in Qumran correspond to long forms in the MT, while forms with an additional vowel letter tend to be third person (cf. 1QSam and 1QIsaa). This pattern underscores the accuracy of the scribes’ copying practices. In short, the Qumran manuscripts contain a larger number of long wayyiqṭols than the MT, but the great majority are still short.48 In other words, short wayyiqṭols were the norm and there is little evidence for the kind of oscillation between long and short forms that can be observed in Kings (see section 1).49 It should be stressed, however, that long and short forms of the same verb seldom occur in one and the same manuscript. Moreover, when there is 45 Kutscher, 1974, pp. 158–159, 163, 328; cf. Qimron, 1986, p. 20; see also Reymond, 2014, p. 189 n. 130. 46 Note also that both are long forms of the doubly weak verb ‫נטה‬. 47 Cf. Hornkohl, 2014, p.160–161. 48 Note, however, 11Q19 (Temple Scroll) contains two attestations of ‫( ויעש‬56:8, 59:16). What makes these interesting is that Smith, 1991, p. 49 correctly classifies both as “unconverted imperfect” forms. A survey of 11Q19 reveals that the short form only occurs when the verb is preceded by a ‫ו‬. All other yiqṭols of the root ‫ עשה‬have the long form ‫יעשה‬. Note however the ‫ה‬ in the form ‫ ויטה‬from the root ‫ נטה‬in 11Q19 51:17 (cf. Smith, 1991, p. 49). 49 Cf. Gzella, forthcoming. 67 a long form in the MT there is generally one in the current corpus. The only possible example of a long form being shortened occurs in 1QIsaa (Isa. 37:36, col. XXXI, 16). The abundance of long forms in this manuscript makes it implausible to interpret this as a classicism. Consequently, it is unlikely that these scribes consciously classicised long forms.50 This lack of classicisms is not surprising since they tend to appear more often in original or new compositions.51 Table 1. Long and short forms in Qumran manuscripts 52 Manuscript Short forms / long forms (1st person long forms) Manuscript Short forms / long forms (1st person long forms) 4Genb 4QGenc 4QGend 4QGenf 4QGeng 4QGenk 4QGen-Exoda 4QpaleoGenm 4QpaleoGen-Exodl 4QpaleoExodm 4QExoda 4QExodb 4QExodc 4QExod-Levf 4QLev-Numa 4QNumb 4QDeutb 4QDeutc 4QDeuth 4QDeutn 4QDeutj 4QpaleoDeutr 4QJoshb XJosh 1QJudg 4QJudga 1QSam 4QSama 17 / 0 1/0 3/0 2/0 7/0 2/0 11 / 0 1/1 2/0 14 / 0 1/0 5/0 3/1 2/0 3/0 9/0 1/0 1/0 3/0 2/0 1/0 1/0 1/0 1/0 4/0 1/0 2/3 22 / 1 4QSamb 4QSamc 6QpapKgs 1QIsaa 1QIsab 4QIsaa 4QIsab 4QIsad 4QIsae 4QIsaf 4QJera 4QJerc 4QEzeka 4QXIIa 4QXIIc 4QXIIe 4QXIIg 4QPsa 4QPsc 8QPs 11QPsa 2QRutha 4QRutha 1QDana 4QDana 4QDanc 4QChr 50 7/0 3/0 1/0 19 / 10 (2) 3/0 5/0 1/0 1/0 3/0 0/1 2 / 1 (1) 0/1 0 / 1 (1) 3/0 1/0 1 / 1 (1) 7/0 2/0 1/0 1/0 1/0 4/0 2/0 1/0 0 / 1 (1) 0 / 1 (1) 1/0 Cf. Gzella, forthcoming. Cf. Joosten, 2013, p. 454 who notes that there is an abundance of classicisms in some Qumran scrolls, particularly the Thanksgiving Scroll, 1QH. In an earlier article, Joosten, 1999 restricts his discussion of classicisms to texts composed in the post-biblical era. 52 The second and fourth columns contain the total number of short and long wayyiqṭol forms in each manuscript. Where there is a number in parentheses following this, it refers to the number of long first person forms. For example, in 1QIsaa there are nineteen short forms and ten long forms, two of which are first person forms. 51 68 Moreover, it seems wise to distinguish between classicisms introduced by scribes copying biblical texts and those made by authors of new documents. To correctly copy a short form is very different from removing a final vowel letter, especially if the short form was no longer used at the time. It requires a different kind of awareness on the part of the scribe to “turn back the clock”. A note of caution: The fragmentary nature of the manuscripts (especially of relevant passages in Kings and Ezekiel) makes it difficult to assert that scribes never classicised texts. Lastly, a study of the use of wayyiqṭols in comparison with other features that are typical of manuscripts from Qumran lies beyond the scope of this article. However, a brief comparison should still be made between the use of long wayyiqṭols and the distribution of yiqṭols and cohortatives. Qimron has argued that the normal yiqṭol and the cohortative appear in complementary distribution. The latter was used clause-initially and the former was used in all other positions.53 No similar pattern can be observed in the case of long wayyiqṭols. But that is expected, since the latter are always clause-initial. Though it is hard to prove, it cannot be ruled out that some long forms were used because of the position they shared with cohortatives. 2.1.2 Parabiblical texts Attestations in parabiblical texts are restricted to manuscripts containing a rewritten form of the Torah. The form ‫ ויהי‬is attested once in 4Q365 (Ex. 14:20, frg. 6a, col. i, 11) and ‫ ו[יע]ש‬is attested in 4Q365 (Ex. 36:33, frg. 12a, col. i, 1). The form ‫ וירא‬is attested twice in 4Q364 (Gen. 38:16, frg. 9a–b, 3; Ex. 14:21, frg. 6a, col. i, 12).54 Moreover, the form ‫ ויט‬is attested twice in 4Q364 (Gen. 38:16, frg. 9a–b, 3; Ex. 14:21, frg. 6a, col. i, 12). The verb ‫ עלה‬is also attested twice. The first person long form ‫ ונעלה‬is used in 4Q364 (Deut. 3:1, frg. 24a–c, 15).55 This form is interesting because it occurs in a quote from the Torah. A short wayyiqṭol occurs in 4Q158 (frg. 4, col. II, 4). This is the only short form in either pesharim or parabiblical texts that does not occur in a quote. 2.1.3 Pesharim and exegetical texts The texts in this category provide only a small number of examples relevant for this study. There are two main reasons for this lack of evidence. The first concerns the genres commented on, namely that most pesharim are commentaries on prophetic books that contain few wayyiqṭols.56 Moreover, the quotes 53 Qimron, 1997, pp. 174–181. Only the last letter of the second example is preserved: ‫[ואר]א‬. 55 Apart from the long form, it is also noteworthy that the text omits the preceding wayyiqṭol in the MT ‫ונפן‬. 56 A good example is the relatively complete commentary on the book of Habakkuk. Even though most of the book is quoted in 1QpHab, there are only two relevant forms in the book of Habakkuk. Similarly, there are no relevant wayyiqṭols in the books of Micah and Nahum. 54 69 from these books, as well as the comments, are often concerned with future events. These factors are sufficient to explain the lack of attestations in these documents. There are, however, ten possible attestations in the corpus, all of which occur in quotes from the Hebrew Bible. Four attestations have the form ‫ויהי‬.57 The form ‫ ותהי‬is attested once in 4Q162 (Isa. 5:25, col. II, 9). Two short forms of ‫ )ויט( נטה‬are attested in commentaries on Isaiah: 4Q162 (Isa. 5:25, col. II, 8) and 4Q163 (Isa. 29:11, frgs. 15–16, 2). It is possible that another attestation of this form occurs in 4Q254 (Gen. 49:15, frgs. 5–6, 1). Unfortunately, very little is visible on the plates in DJD XXII.58 Lastly, the form ‫ וירא‬is attested in 4Q252 (Gen. 8:13, col. I, 22) and ‫ ותעש‬in 1QpHab (Hab. 1:14, col. V, 1). It is noteworthy that the MT has the long form ‫ ותעשה‬in Hab. 1:14. However, it is difficult to use this as evidence that the scribe classicised the text because it is the only relevant form in Habakkuk. The small number of attestations make any conclusions tentative. Yet the available evidence suggests that the scribes who wrote or copied these manuscripts were either familiar with the short forms (as a written convention) or made sure that they carefully copied the quotes containing these wayyiqṭols. 2.2 Hollow roots 2.2.1 Verbs with a medial ‫ו‬ The data discussed in this section concerns three weak verbs with a medial ‫ו‬: ‫מות‬, ‫ קום‬and ‫בוא‬. The biblical manuscripts contain eight clear wayyiqṭols of the root ‫מות‬. Two of these have the distinct short form ‫וימת‬, and six have the form ‫וימות‬.59 Two observations are noteworthy: different forms are not attested in the same manuscript;60 and forms with ‫ ו‬are more common. There are thirteen clear wayyiqṭols of the verb ‫קום‬: six are distinctly short, and seven contain a ‫ו‬.61 Again, the sample is relatively small due to the fragmentary nature of the scrolls. It is still noteworthy, though, that there are more 57 4Q252 (Gen. 7:12, col. I, 5; 8:6, col. I, 12), 4QFlor (Isa. 8:11, frg. 1, pl. 286, col. I, 15), and 4Q254 (Gen. 19:17, frgs. 5–6, 4). 58 On the top right of frgs. 5–6 on plate XV in DJD XXII (Brooke et al., 1996). 59 The form ‫ וימת‬occurs in 4QExodb (Ex. 1:6, frg.1) and 4QRuth a (Ruth 1:3, 3). The form ‫וימות‬ occurs in 4QNumb (Num. 35:20, col. XXXI, frgs. 80–84, 11; 35:23, col. XXXI, frgs. 80–84, 15) and 4QSama (2 Sam. 3:27, frg. 61i–62, 7; 2 Sam. 6:7, frgs. 68–76, 9; 2 Sam. 10:18, frg. 88, 2; 2 Sam. 11:17, frgs. 93–94, 3). 60 There are several occasions where the editors of individual manuscripts reconstructed part of the text. One interesting example is ‫ וימת‬in 2 Sam. 17:23 (4QSama, frgs 126–27, 2). Since this form is not attested elsewhere in 4QSama, it seems reasonable to assume that this reconstruction is a mistake on the part of the editor. Reconstructions are always in the realm of speculation, but all other relevant wayyiqṭols of ‫מות‬, ‫ קום‬and ‫ בוא‬in 4QSama have ‫ו‬, so it stands to reason that any reconstructed form should be spelled with ‫ו‬. 61 The short forms occur in 4QGenb (4:8, frg. 3i, 8); 4QExod-Levf (Ex. 40:18, col. II, frg. 2ii, 14); 4QJudgb (19:7, frg. 1, 4); 4QSamb (1 Sam. 21:1, frg. 6–7, 14; 23:13, frg. 10–23, 4); and 2QRutha (2:15, col. II, 1). The forms with ‫ ו‬occur in 4QNumb (11:32, col. I, frgs. 1–4, 1; 16:2, 70 forms with ‫ו‬.62 Moreover, several of these occur in 4QNumb and 4QSama, where similar forms of ‫ מות‬are attested. The biblical manuscripts contain twenty-four wayyiqṭols of the verb ‫בוא‬, seven of which are distinctly short.63 The other seventeen attestations of this verb have ‫ו‬.64 To summarise, scribes were consistent, in that they either wrote these wayyiqṭols with a ‫ ו‬or without it (the only exception is 2QRutha, which contains both ‫ ויבא‬and ‫)ויבוא‬.65 At first glance the extensive use of ‫ ו‬could be interpreted as an indication that scribes used long wayyiqṭols. However, the use of ‫ ו‬as a mater lectionis for both long and short vowels in Qumran suggests that these forms remained short while the vowel was written plene (especially in the case of ‫בוא‬, where the vowel had been lengthened due to the quiescent ‫)א‬.66 Moreover, 4QNumb, 4QSamc and 4QXIIg were written in the Qumran scribal practice, which is characterised by an extensive use of vowel letters.67 The consistent use of ‫ ו‬within individual manuscripts also indicates that this is a matter of plene spelling. If some scribes were using occasional long forms, a less uniform pattern would be expected (as can be seen among III-y verbs). 2.2.2 Verbs with a medial ‫י‬ Wayyiqṭols of verbs with a medial ‫ י‬are rare in both the Hebrew Bible and in Qumran manuscripts. Only two different verbs are attested. The form ‫ויקא‬ from ‫ קיא‬is attested in 4QXIIg (Jon. 2:11, frgs. 78ii, 82–87, 10). Furthermore, there are eleven relevant wayyiqṭols of the verb ‫שים‬. Ten of these have forms without ‫ י‬and occur in four manuscripts.68 The only attestation with a ‫ י‬occurs col. VI, frgs. 6–10, 13); 4QSama (1 Sam. 28:23, frg. 44, 3; 2 Sam. 12:17, frgs. 100–101, 3; 13:31, frgs. 102ii, 103–106i, 107–109 a–b, 28; 14:31, col. II, frgs. 5ii–7i, 15); and 4QXIIg (Jon. 3:3, frgs. 78ii, 82–87, 12). 62 By way of comparison, it could be noted that ‫ ויקם‬is attested in 1QS 5:8 as an “unconverted imperfect” along with two other imperfects of hollow roots: ‫ וישם‬in 1QS 3:18 and ‫ ויכן‬in 1QS 11:13 (Smith, 1991, pp. 41–42). 63 4QGen-Exoda (Gen. 35:27, frg. 5, 8); 4QpaleoGen-Exodl (Ex. 3:1, frgs. 3–4, 9); 4QExodc )17:8, col. VIII, frgs. 35–36, 15); 4QSamb (1 Sam. 20:37, frgs. 6–7, 9); 5QKgs (1 Kings 1:28, frg. 1c, 2); 4QIsab (39:3, frgs. 24–25, 6); 2QRutha (3:7, col. V, 6). 64 4QpaleoExodm (18:12, col. XVIII, 18); 4QNumb (13:22, col. II, frgs. 3ii, 5, 13; 22:9, col. XV, frgs. 20–22, 2); 4QJosha (10:9, frgs. 19–22, 2); 4QSama (2 Sam. 3:35, frgs. 61i–62, 19; 11:4, frgs. 89–92, 5; 12:16, frgs. 100–101, 3; 13:24, frgs. 102ii, 103–106i, 107–109 a–b, 17); 4QSamc (2 Sam. 14:33, col. II, frgs. 5ii–7i, 20; 15:13, col. III, frgs. 7ii–11, 13); 1QIsaa (36:22, col. XXX, 3; 37:1, col. XXX, 5; 38:1, col. XXXI, 20; 39:3, col. XXXII, 20); 1QIsa b (39:3, col. XVI, 15); 4QXIIg (Jon. 2:8, frgs. 78ii, 82–87, 7); 2QRutha (2:18, col. II, 6). The form in 2 Sam. 12:16 of 4QSama is partially damaged. 65 See also Smith, 1991, p. 50 for the forms ‫ וימות‬and ‫ וימת‬in 11Q19. 66 For the use of ‫ ו‬for long and short o/u, see Qimron, 1986, pp. 17–18; Tov, 2013, p. 670. See Kutscher, 1974, pp. 126–148 for a discussion of ‫ ו‬as a vowel letter in 1QIsaa. 67 Tov, 2004, p. 339. Tov’s table indicates that vowel letters were used consistently in 4QNumb and 4QSamc. 68 4QGen-Exoda (Ex. 2:3, frg. 19i, 7); 4QExod-Levf (Ex. 40:18, col. II, frg. 2ii, 13; 40:19, col. II, frg. 2ii, 15; 40:21, col. II, frg. 2ii, 18; 40:24, col. II, frg. 2ii, 21); 1QIsa a (49:2, col. XL, 29; 51:3, col. XLII, 16); 1QIsab (49:2, col. XXI, 9; 51:3, col. XXII, 11; 51:16, frg. 36, 2). The vellum of 4QExod-Levf is relatively well preserved but the ink is fading so the manuscript is very difficult to read from the photographs and the plates. Furthermore, there could be another 71 in 4QExodb (2:3, frgs. 3i–4, 2). The small sample makes it difficult to draw reliable conclusions. The available evidence suggests that scribes typically distinguished between long and short forms of ‫שים‬. The form with a ‫ י‬in 4QExodb may be a scribal mistake. But it is also possible that the orthography was influenced by that of long yiqṭols, especially since ‫ י‬does not occur as a vowel letter for /ɛ/.69 2.3 Causative forms of strong roots A brief note should be added concerning wayyiqṭols of strong roots in the hiphil stem. Long yiqṭol forms of these roots typically contain an additional ‫י‬ in both the MT and the biblical scrolls from Qumran. Wayyiqṭols in these corpora lack this distinguishing feature. Here, it is sufficient to note that these short forms share this feature with those of the verb ‫שים‬. 3 Conclusion What can be said about copying practices based on this survey? How did the morphological erosion of wayyiqṭols influence the scribes who made the Dead Sea Scrolls? The documents in the corpus provide a relatively homogenous picture. Most manuscripts consistently use short wayyiqṭols in places where the MT has a corresponding short form. Only 1QSam and 1QIsaa depart from this pattern, and the latter contains as many short wayyiqṭols as long ones. What is the best explanation for this pattern? The use of non-conversive alternatives to wayyiqṭol in non-biblical compositions suggests that the use of wayyiqṭol was in decline. Moreover, the occasional long wayyiqṭol forms in biblical texts and quotes point to a process of morphological erosion. Consequently, it would have been very easy for scribes to add an additional ‫ה‬. However, the scribes who copied the manuscripts in the corpus were very consistent in their use of distinct short forms. The same consistency is attested in the accurate copying of long first person forms. Two explanations can account for this pattern: it could point to a carefully executed copying process (particularly in the case of the Torah); or, if the scribes were familiar with the style of the texts they copied, it is also possible that the short form was a part of their literary toolbox (even if it was not used creatively). A combination of these two factors is the best explanation for the consistent copying of short forms. As for 1QIsaa, it is plausible that the non-narrative style of the book contributed to the higher frequency of long forms. example in 4QExod-Levf (Ex. 40:26, col. II, frg. 2ii, 24). A normal yiqṭol (‫ )אשים‬is used in Isa. 51:16 of 1QIsaa instead of the MT form ‫ואשים‬. 69 Cf. Qimron, 1986, p. 19; Tov, 2013, pp. 670–671; and Reymond, 2014, pp. 39–43, who note that ‫ י‬primarily represents /ī/ and sometimes /ē/ but not /ε/. 72 Lastly, a cautionary note. Many manuscripts are fragmentary, and the great Isaiah scroll is the only one that can be studied systematically. It is therefore possible that more long forms were present in passages that have not been preserved. References Bloch, Y., 2007, “From linguistics to text-criticism and back: wayyiqṭōl constructions with long prefixed verbal forms in Biblical Hebrew”, Hebrew Studies 48, pp. 141–170. Brooke, G.J., et al., 1996, Discoveries in the Judean Desert, vol. XXII: Qumran Cave 4. XVII: Parabiblical Texts, Part 3, Oxford: Clarendon Press. Cook, J., 2012, Time and the Biblical Hebrew Verb: The Expression of Tense, Aspect, and Modality in Biblical Hebrew, Winona Lake: Eisenbrauns. Cross, F. M., Parry, D. W., Saley, R. J. and Ulrich, E. C., 2005, Discoveries in the Judean Desert, vol. XVII: Qumran Cave 4. XII: 1–2 Samuel, Oxford: Clarendon Press. Elwolde, J. F., 2015, “Isaiah 5:1–7 in the Dead Sea Scrolls, MT, and Versions: basic data and preliminary observations”, in A. Puig i Tàrrech (ed.), Relectures de l’escriptura a la llum del Concili Vaticà II (1): “La vinya”, Barcelona: Publicacions de l’Abadia de Montserrat and Associació Bíblica de Catalunya, pp. 89– 132. Fassberg, S. E., 2013, “Dead Sea Scrolls: linguistic features”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. I, pp. 663–669. Geiger, G., 2013, “Tense: Rabbinic Hebrew”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. III, pp. 740–742. Gzella, H., forthcoming, “Untypical wayyiqṭol forms in Hebrew and early linguistic diversity”. Hornkohl, A. D., 2014, Ancient Hebrew Periodization and the Language of the Book of Jeremiah: The Case for a Sixth-Century Date of Composition, Leiden: Brill. Joosten, J., 1999, “Pseudo-classicisms in Late Biblical Hebrew, in Ben Sira, and in Qumran Hebrew”, in T. Muraoka and J. F. Elwolde (eds), Sirach, Scrolls, and Sages: Proceedings of a Second International Symposium on the Hebrew of the Dead Sea Scrolls, Ben Sira, and the Mishnah, Held at Leiden University, 15–17 December 1997, Leiden: Brill, pp. 146–159. ———, 2013, “Classicism: Biblical Hebrew”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. I, p. 454. Joüon, P. and Muraoka, T., 2006, A Grammar of Biblical Hebrew, 2nd revised edn, Rome: Editrice Pontificio Istituto Biblico. Kutscher, E. Y., 1974, The Language and Linguistic Background of the Isaiah Scroll (IQIsaa), Leiden: Brill. Parry, D. W. and Qimron, E. (eds), 1998, The Great Isaiah Scroll (1QIsaa): A New Edition, Leiden: Brill. Qimron, E., 1986, The Hebrew of the Dead Sea Scrolls, Atlanta: Scholars Press. ———, 1997, “A new approach to the use of forms of the imperfect without personal endings”, in T. Muraoka and J. F. Elwolde (eds), The Hebrew of the Dead Sea Scrolls and Ben Sira: Proceedings of a Symposium Held at Leiden University, 11–14 December 1995, Leiden: Brill, pp. 174–181. 73 Reymond, E. D., 2014, Qumran Hebrew: An Overview of Orthography, Phonology, and Morphology, Atlanta: SBL. Smith, M. S., 1991, The Origins and Development of the Waw-Consecutiv, Atlanta: Scholars Press. Tov, E., 2004, Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert, Leiden: Brill. ———, 2013, “Dead Sea Scrolls: orthography and scribal practices”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. I, pp. 669–673. Ulrich, E. C. (ed.), 2010, The Biblical Qumran Scrolls: Transcriptions and Textual Variants, Leiden: Brill. Waltke, B. K. and O’Connor, M. P., 1990, An Introduction to Biblical Hebrew Syntax, Winona Lake: Eisenbrauns. 74 Unmarked Modality and Rhetorical Questions in Biblical Hebrew ELIZABETH ROBAR Tyndale House, Cambridge 1 Modality and systems of modality1 Modality refers to the mood, opinion or attitude a speaker expresses toward an idea. It may be expressed systematically through the verbal system by modal inflection of verbs, as in Latin and (Classical) Greek: dat; det; da! dídosin; dói; dós! ‘he gives; he should give; give!’ Alternatively, but still within the verbal system, modality may be expressed through auxiliary verbs, as in German and English: ich darf gehen; ich muss gehen; ich kann gehen; ich soll gehen I may go; I must go; I can go; I should go Modality may equally be conveyed outside of the verbal system. In Latin, interrogative modality may be expressed by tone of voice, an interrogative pronoun or the particle -ne appended to the first word of the sentence. mihi dat? quis mihi dat? mihine dat? 1.1 Biblical Hebrew: Foretaste Biblical Hebrew, like any language, expresses modality, but its full grammatical arsenal has not yet been fully appreciated by scholars. Grammars often 1 An earlier version was first presented at the conference of the Society of Old Testament Studies in Nottingham on 4 January 2017 as ‘Unmarked modality: Rhetorical questions and theological presuppositions’. 75 mention the inflectional varieties of yiqtol, including the ‘volitional’ paradigm, but Biblical Hebrew has more ways of expressing modality. There are lexically modal verbs such as ‫ יוכַל‬and ‫אָ בָ ה‬: ‫ָׁשבת י ְַח ָ ָּּֽדו׃‬ ֶּ֥ ‫ָשבֶ ת יַחְ ָ ֶ֑דו ִ ָּֽׁכי־הָ יָ ִׂ֤ה ְרכושָ םֶׁ֙ ָ ָ֔רב וְ ֶּ֥ל ֹׁא ָיָּֽכְ לָ֖ ּו ל‬ ָ֣ ֶ ‫ָשא א ָ ֛תם הָ ָ ֵ֖א ֶרץ ל‬ ָ֥ ָ ‫וְ לא־נ‬ ‘so that the land could not support both of them dwelling together; for their possessions were so great that they could not dwell together.’ (Gen. 13:16, ESV)2 ‫וַיְ חַ זֵ ָ֥ק יְ הוָ ֵ֖ה אֶ ת־לֵ ָ֣ב פ ְַר ֶ֑עה וְ ֶּ֥ל ֹׁא אָ ָ ָ֖בה לְ ׁשַ לְ ָ ָּֽחם׃‬ ‘But the LORD hardened Pharaoh’s heart, and he was not willing to let them go.’ (Ex. 10:27, NIV) There are some idiomatic uses, such as ‫ אין‬with a personal pronoun for the person unwilling: ‫א־ת ְר ָ֣או ָפ ָ֔ ַני בִ לְ ִ ֵ֖תי אֲחִ יכֶ ָ֥ם ִא ְת ֶ ָּֽׁכם׃‬ ִ ‫וְ ִּאם־אינְ ךֶּ֥ ְמׁשַ ָ֖לחַ ָ֣לא נ ֵ ֵֶ֑רד ִ ָּֽׁכי־הָ ִִ֞איש אָ ַ ִׂ֤מר אֵ ֵָּ֙לינוֶׁ֙ ָּֽׁל‬ ‘But if you will not send him, we will not go down, for the man said to us, “You shall not see my face, unless your brother is with you.” ’ (Gen. 43:5, ESV) ‫יס ָ֣רו א ָ֔תֹו וְ ָ֥לא יִ ְש ַמֵ֖ע‬ ְ ְ‫ומֹורה אינֵּׁ֣נּו ׁשֹׁ ׂ֔מעַ בְ ָ֥קֹול אָ ִ ֵ֖ביו ובְ ָ֣קֹול ִא ֶ֑מֹו ו‬ ָ֔ ֶ ‫סֹורר‬ ָ֣ ֵ ‫ִ ָּֽׁכי־יִ הְ יֶ ָ֣ה לְ ִַ֗איש בֵֵּ֚ ן‬ ‫ֵיהם׃‬ ָּֽׁ ֶ ‫ֲאל‬ ‘If a man has a stubborn and rebellious son who will not obey the voice of his father or the voice of his mother, and, though they discipline him, will not listen to them …’ (Deut. 21:18, ESV) And particles may also indicate a modal meaning: ‫ֹלהים ָ֣לא ָּֽׁתאכְ ָ֔לו ִמ ֵ֖כל עֵ ָ֥ץ הַ ָ ָּֽׁגָּ֑ן׃‬ ִָ֔ ‫אַַ֚ ף ִּ ָּֽכי־אָ ַ ָ֣מר ֱא‬ ‘Did God actually say, “You shall not eat of any tree in the garden”?’ (Gen. 3:1b, ESV) And, of particular interest, word order may be a linguistic signal that the verb is to be understood as modal. 2 In examples throughout this paper, English translations are taken from: ESV Study Bible, English Standard Version, Crossway, 2016 (ESV); New International Version, Biblica, 2011 (NIV); A New Translation of the Holy Scriptures according to the Traditional Hebrew Text, The Jewish Publication Society, 1985 (JPS); New English Translation, Biblical Studies Press, 1996–2005 (NET). 76 ‫אֹותֹו‬ ֶ֑ ‫יְהֹוש ַע אֶ ל־הָ ַָ֗עם ע ִּ ֵּׁ֤דים אַ תם֙ בָ ֶָ֔כם ִ ָּֽׁכי־אַ ִ֞ ֶתם בְ חַ ְר ֶ ָ֥תם לָכֶ ֛ם אֶ ת־יְ הוָ ֵ֖ה ַלע ֲָ֣בד‬ ֜ ‫ו ַָּ֙יאמֶ ר‬ ‫אמ ֵ֖רו ע ִ ֵָּֽׁדים׃‬ ְ ‫וַי‬ ‘Then Joshua said to the people, “You are witnesses against yourselves that you have chosen the LORD, to serve him.” And they said, “We are witnesses.” ’ (Josh. 24:22, ESV) ‘Thereupon Joshua said to the people, “You are witnesses against yourselves that you have by your own act chosen to serve the LORD.” “Yes, we are!” they responded.’ (Josh. 24:22, JPS) ‘Joshua said to the people, “Do you agree to be witnesses against yourselves that you have chosen to worship the LORD?” They replied, “We are witnesses!” ’ (Josh. 24:22, NET) The NET translation of this as a question is well justified by the repetition of the ‫ע ִֵדים‬, which can be translated as ‘Yes, we will be witnesses’. Biblical Hebrew does not answer yes/no questions in the affirmative with a word for ‘yes’: instead, the affirmative is indicated by repeating the key word or phrase.3 In this case, it also means that Joshua’s initial statement could be understood as a question: ‘Will you be witnesses against yourselves?’ Or, making the modality explicit, ‘Are you willing to be witnesses?’ They answer, ‘Yes, we are willing to be witnesses’. But, as the ESV and other translations demonstrate, Joshua’s challenge to the people does not have to be understood as a question. A sternly stated ‘You are witnesses against yourselves’ can equally convey the meaning. Joshua is imposing his will on the people and challenging them to accept it. This imposition of will is the definition of deontic modality, in which what is at stake is not a proposition itself, but one party imposing its will on another with regard to that proposition. Here is the first of many cases in which there is an overlap between kinds of modality, such as the imposition of will, deontic modality, with the requirement for a response, interrogative modality. 1.2 When modality is significant Biblical Hebrew modality is not restricted to the volitional yiqtol paradigm, which explains why modality presents such a challenge for non-native speakers and grammarians. Comparing Bible translations, let alone interpretations, is sobering, as they can vary so widely in what modality they detect. Deciding on the modality can be an exercise in frustration. Consider Isaiah 1:18: 3 I was vaguely aware of this, but I thank Aaron Hornkohl for making it explicit for me. 77 ‫ַשלֶג יַלְ ִָ֔בינו ִאם־י ְַא ִ ָ֥דימו כַתֹולָ ֵ֖ע‬ ָ֣ ֶ ‫ם־יהְ יָּ֙ ו חֲטָ אֵ י כֶ ִׂ֤ם כַשָ נִ יםֶׁ֙ כ‬ ָּֽׁ ִ ‫אמָ֣ר יְ הוָ ֶ֑ה ִא‬ ַ ‫לְ כו־נָ ֛א וְ נ ִָוָּֽׁכְ ָחֵ֖ה י‬ ‫כ ֶַצָ֥מֶ ר יִהְ יָּֽׁ ו׃‬ ‘Come now, let us reason together, says the LORD: though your sins are like scarlet, they shall be as white as snow; though they are red like crimson, they shall become like wool.’ (Isa. 1:18, ESV) In this verse, does God promise that crimson sins can be made white as snow? Or does he ridicule that same thought? It all hangs on the modality intended. If assertive modality, ‘though your sins are like scarlet, they shall be as white as snow’, then God is the saviour promising forgiveness. If interrogative modality, ‘though your sins are like scarlet, shall they be as white as snow?’, then God is the judge, dismissing nonsense and anticipating judgment. The modality makes all the difference.4 Since our grammars tell us that interrogative modality in Biblical Hebrew is marked by the interrogative particle ‫ ֲה‬or by a question word such as ‫ ִמי‬or ‫מָ ה‬, questions lacking these particles are given the label ‘unmarked’. In other words, these are possibly questions, if a grammarian can justify why we ought to analyse them as questions even though they lack the expected marking. This will be the topic of this paper, with a particular focus, eventually, on unmarked rhetorical questions. 2 Modality in BH: Function to form Discussions of modality in Biblical Hebrew usually work from form to function: beginning with the cohortative, imperative and jussive forms, and observing that weqatal and yiqtol are compatible with many kinds of modality. Hatav’s conclusion is representative: “all the modals (except for the counterfactuals) can be, and usually are, denoted by the two forms yiqtol and wqatal”,5 alongside the volitional paradigm. Without necessarily disputing this, it may be worthwhile to approach the topic from the other direction: beginning with modal function, as best we can determine from the semantics evident in English translation, and then observing what forms are involved. What follows is a brief sampling of each of the three main areas of modality: epistemic, deontic and evidential. To protect against special pleading, I employ a few standard translations (ESV, NIV, JPS, NET)6 as evidence for the modal semantics claimed. 4 There is clearly a nuanced range of possibilities, but these two demonstrate the critical role of modality within the discussion. For a review of proposals, see Williamson, 2006, pp. 103–118. 5 Hatav, 1997, p. 150. 6 For details of these translations, see n. 2. 78 2.1 Possibility (epistemic) Possibility, within what is called epistemic modality, or modality referring to the speaker’s knowledge or certainty regarding a proposition, provides one view of the various strategies within Biblical Hebrew for conveying modality. The verb ‫‘ יוכַל‬to be able’ is the first contender, and then the verb stem, or binyan, of the niphal. In Genesis 13:16, possibility is introduced by ‫ִאם־יוכַ ָ֣ל‬ ‫‘ ִַ֗איש‬if anyone can’, and then it is continued with the niphal stem, in this case passive voice, ‫‘ ַגָּֽׁם־ז ְַר ֲעָךֵ֖ ִימָ ֶנָּֽׁה‬so your seed will be able to be counted’.7 ָ֖‫ם־יּוכֵּׁ֣ל ִָּ֗איׁש לִ ְמנֹותֶׁ֙ אֶ ת־עֲפַ ָ֣ר הָ ָ֔ ָא ֶרץ ַ ָּֽגם־ז ְַרעֲך‬ ַ ‫ֲשר ׀ ִּא‬ ָ֣ ֶ ‫וְ שַ ְמ ִ ָ֥תי ֶ ָּֽׁאת־ז ְַרעֲָךֵ֖ ַכעֲפַ ָ֣ר הָ ָ ֶ֑א ֶרץ א‬ ‫יִּ מָ נָּֽה׃‬ ‘I will make your offspring as the dust of the earth, so that if one can count the dust of the earth, your offspring also can be counted.’ (Gen. 13:16, ESV) In Genesis 16:10, the seed ‘cannot be counted’ for the abundance, ‫וְ ָ֥לא ִיסָ פֵ ֵ֖ר‬ ‫מֵ ָּֽׁרב‬. ‫ו ִַׂ֤יאמֶ ר לָּהֶׁ֙ מַ לְ ַ ָ֣אְך יְ ה ָ֔ ָוה הַ ְר ָבָ֥ה אַ ְר ֶבֵ֖ה אֶ ת־ז ְַרעֵ ְֶ֑ך וְ ֶּ֥ל ֹׁא יִּ סָ פָ֖ר מ ָּֽרֹׁ ב׃‬ ‘The angel of the LORD also said to her, “I will surely multiply your offspring so that they cannot be numbered for multitude.” ’ (Gen. 16:10, ESV) A conditional context, ‫ ִאם‬followed by ‫וְ הָ יָה‬, can convey possibility: ‘if’ Esau attacks the first camp, the second camp ‘might be able to escape’, in Genesis 32:9. ‫יטה׃‬ ָּֽ ָ ‫ֵשו אֶ ל־הַ מַ חֲנֶ ָ֥ה הָ אַ ַחֵ֖ת וְ הִ כָ ֶ֑הו וְ הָ יָ ָ֛ה הַ מַ חֲנֶּ֥ה הַ נִּ ְׁש ָ ָ֖אר לִּ פְ ל‬ ֛ ָ ‫ו ַַ֕יאמֶ ר ִּאם־י ָָ֥בֹוא ע‬ ‘… thinking, “If Esau comes to the one camp and attacks it, the other camp may yet escape.” ’ (Gen. 32:9, JPS) The adverb ‫ אולַי‬can make possibility explicit: ‘maybe I will have children through her’, ‫ אולַ ָ֥י ִאבָ נֶ ֵ֖ה ִמ ֶ ֶ֑מנָה‬in Genesis 16:2. ‫אּולי ִּאבָ נָ֖ה ִּמ ָ֑מנָה‬ ֶּ֥ ַ ‫ל־שפְ חָ ָ֔ ִתי‬ ִ ֶ‫ו ַָּ֙תאמֶ ר שָ ַ ֜רי אֶ ל־ אַ בְ ָ ַ֗רם הִ נֵה־ ִ֞ ָנא עֲצָ ַ ִׂ֤רנִ י יְ הוָהֶׁ֙ ִמ ֶָ֔לדֶ ת בא־נָאֶׁ֙ א‬ ‫ַיִש ַ ָ֥מע אַ בְ ָ ֵ֖רם לְ ָ֥קֹול שָ ָ ָּֽׁרי׃‬ ְ ‫ו‬ ‘And Sarai said to Abram, “Behold now, the LORD has prevented me from bearing children. Go in to my servant; it may be that I shall obtain children by her.” And Abram listened to the voice of Sarai.’ (Gen. 16:2, ESV) 7 Cf. the Latin future participle. 79 The conjunction ‫ פֶן‬can convey an undesired possibility: ‘he might also die’, ‫ַם־הוא‬ ֵ֖ ‫ פֶן־י ָָ֥מות ג‬in Genesis 38:11. ‫ו ַָ֣יאמֶ ר יְ הודָ ֩ה לְ תָ ָּ֙ ָמר ַכל ָ֜תֹו ְש ִ ָּ֧בי אַ לְ מָ נָ ָ֣ה בֵ ית־אָ ִַ֗ביְך עַד־יִגְ דַ ֶׁ֙ל שֵ לָ ָ֣ה בְ ָ֔ ִני ִ ָ֣כי אָ ָ֔ ַמר פן־י ֶָּ֥מּות‬ ‫ַם־הּוא כְ אֶ ָחֶ֑יו ו ֵ ַָ֣תלְֶך תָ ָ֔ ָמר ו ֵ ֵַ֖תשֶ ב ֵבָ֥ית אָ ִ ָּֽׁביהָ ׃‬ ָ֖ ‫ג‬ ‘Then Judah said to his daughter-in-law Tamar, “Stay as a widow in your father’s house until my son Shelah grows up” – for he thought, “He too might die like his brothers.” So Tamar went to live in her father’s house.’ (Gen. 38:11, JPS) It is true that, in each case, either a yiqtol or weqatal form is involved, but it is often not the verbal conjugation that conveys the meaning of possibility. The stem, or binyan, or a particle, or a conjunction, or a conditional structure, are all involved as well, and in fact, were it not for these, the yiqtol or weqatal would likely not be considered modal. 2.2 Permission (deontic) Permission comes from another side of modality, deontic modality, concerned with obligation and permission, imposing one’s will on another. This is familiarly expressed with the jussive, both in verb-initial and verb-medial clauses, as in Genesis 24:55: ‫ תֵ שֵָּ֙ ב הַ ַנע ָ ֲָ֥ר ִא ָ ֛תנו י ִ ֵָ֖מים ָ֣אֹו ע ֶָ֑שֹור‬followed by ‫אַ ַחֵ֖ר תֵ ֵ ָּֽׁלְך‬. ‫יה וְ ִא ָ֔ ָמּה ת ֹׁ֙שב הַ ַנע ָ ֲֶּ֥ר ִּא ָ ָ֛תנּו י ִ ֵָ֖מים ָ֣אֹו ע ֶָ֑שֹור אַ ַ ָ֖חר ת ָּֽלְך׃‬ ֶָׁ֙ ‫ו ִַׂ֤יאמֶ ר אָ ִ ֶׁ֙ח‬ ‘Her brother and her mother said, “Let the young woman remain with us a while, at least ten days; after that she may go.” ’ (Gen. 24:55, ESV) Permission may also be expressed with a non-jussive form, but with a fronted object, as in Genesis 42:37, with Reuben’s desperate plea to his father, and Genesis 47:24, in which the Egyptians are permitted four-fifths of their yield. ‫יאנו אֵ לֶ ֶ֑יָך ְתנָ ִׂ֤ה אתֹוֶׁ֙ עַל־י ִָָ֔די‬ ֵ֖ ֶ ִ‫ם־לא אֲב‬ ְ ‫ֵאמר א‬ ָ֔ ‫ו ִַׂ֤יאמֶ ר ְראובֵ ןֶׁ֙ אֶ ל־אָ ִ ָ֣ביו ל‬ ָ֥ ‫ת־ׁשנֵּׁ֤י בָ נַי֙ תָ ִּׂ֔מית ִא‬ ‫יבָ֥נו אֵ ֶ ָּֽׁליָך׃‬ ֶ ‫ַוא ֲִנֵ֖י א ֲִש‬ ‘Then Reuben said to his father, “You may kill my two sons if I do not bring him back to you. Put him in my care, and I will return him to you.” ’ (Gen. 42:37, JPS) ‫ישית לְ פ ְַר ֶ֑עה וְ אַ ְר ַ ֵּׁ֣בע הַ יָדֹֹׁ֡ ת יִּ ְהיֵּׁ֣ה לָכם לְ ֶָּ֙ז ַרע הַ שָ ֶ ָּ֧דה וָּֽׁ לְ אָ כְ לְ כֶ ֛ם‬ ֵ֖ ִ ‫בואת ונְ תַ ֶ ָ֥תם ח ֲִמ‬ ָ֔ ‫וְ הָ יָהֶׁ֙ בַ ְת‬ ‫ֲשר בְ בָ תֵ יכֶ ֵ֖ם וְ ֶלא ֱָ֥כל לְ טַ פְ ֶ ָּֽׁכם׃‬ ָ֥ ֶ ‫וְ ַלא‬ ‘But when the crop comes in, give a fifth of it to Pharaoh. The other four-fifths you may keep as seed for the fields and as food for yourselves and your households and your children.’ (Gen. 47:24, NIV) Permission may also be conveyed with a cohortative form, as in Amos 8:5. 80 ‫ה־בר‬ ָ֑ ָ ָ‫ֵאמר מָ ִ֞ ַתי ַיע ֲִׂ֤בר הַ ָּ֙חדֶ שֶׁ֙ וְ נ ְַׁש ִּ ֵּׁ֣ב ָירה ׂ֔שבר וְ הַ שַ ָבֵ֖ת וְ נִּ פְ ְתח‬ ַ֗ ‫ל‬ ‘saying, “When will the new moon be over, that we may sell grain? And the Sabbath, that we may offer wheat for sale …” ’ (Am. 8:5, ESV) Every variation of yiqtol is compatible with permission, making it clear that there must be more to deontic modality than just volitional forms. 2.3 Certainty and evidential modality Certainty sits astride epistemic and evidential modality, with evidential modality focused on the grounds for asserting knowledge. ‘Deductive modality’ is Palmer’s8 technical term for the certainty that does not identify the evidence, but insists that any and all evidence available leaves as the only possible conclusion the statement at hand. In English, we often prefer intonation, adverbs and punctuation to express the strongest certainty: I am absolutely dead positive about this!! Hebrew has several strategies for conveying evidential modality, among them ‫הֲלא‬, ‫ הִ נֵה‬and the paronomastic infinitive absolute.9 In Judges 5:30, Sisera’s mother insists that the only possible explanation for his delay is the abundance of plunder that takes time to gather and divide. ‫יס ָ ָ֔רא ְשלַ ָ֥ל צְ בָ ִ ֵ֖עים ִרקְ ָ ֶ֑מה ֶצָ֥בַ ע‬ ְ ‫ֲה ֹ֙ל ֹׁא יִּ ְמ ְצ ִ֜אּו יְ חַ לְ ֵּׁ֣קּו ׁשָ ָָ֗לל ַ ִׂ֤רחַ ם ַרחֲמָ ֶַׁ֙ת ִיםֶׁ֙ לְ ָ֣ראש ָ֔ ֶגבֶ ר ְשלַ ִׂ֤ל צְ בָ עִ יםֶׁ֙ לְ ִ ָ֣ס‬ ‫ארי שָ ָ ָּֽׁלל׃‬ ָ֥ ֵ ְ‫ִרקְ מָ ַ ֵ֖תיִ ם לְ צַ ו‬ ‘They must be dividing the spoil they have found: A damsel or two for each man, Spoil of dyed cloths for Sisera, Spoil of embroidered cloths, A couple of embroidered cloths Round every neck as spoil.’ (Judg. 5:30, JPS) In 1 Samuel 21:12, the servants of Achish have recognised David and are exclaiming regarding the mismatch between his current, refugee situation and what everybody knew to be true of his reputation. ‫ֵאמר הִ כָ ִׂ֤ה‬ ָ֔ ‫אמ ֜רו עַבְ ֵ ִׂ֤די אָ כִ ישֶׁ֙ אֵ ָָ֔ליו הֲלֹוא־זֶּ֥ה דָ ִּ ָ֖וד ֶ ָ֣מלְֶך הָ ָ ֶ֑א ֶרץ הֲלָ֣ ֹוא ָל ֶַ֗זה ַיעֲנִׂ֤ ו בַ ְמחלֹותֶׁ֙ ל‬ ְ ‫ו ַָּ֙י‬ ‫שָ או ֶׁ֙ל בְ ֲא ָלפָו וְ דָ ִ ֵ֖וד בְ ִרבְ בתָ ו‬ ‘The courtiers of Achish said to him, “Why, that’s David, king of the land! That’s the one of whom they sing as they dance: Saul has slain his thousands; David, his tens of thousands.” ’ (1 Sam. 21:12, JPS) 8 9 Palmer, 2001. Callaham, 2010. 81 The word ‫ הִ נֵה‬can function as a sensory evidential, as in Genesis 1:31, where God looked, and, based on the evidence from his senses, he concluded that it was very good. ‫־טֹוב ְמ ֶ֑אד‬ ֵ֖ ‫ֲשר עָשָָ֔ ה וְ ִּהנה‬ ָ֣ ֶ ‫וַיַ ִׂ֤ ְָּ֑רא אֱֹלהִ יםֶׁ֙ אֶ ת־כָל־א‬ ‘And God saw everything that he had made, and behold, it was very good. And there was evening and there was morning, the sixth day.’ (Gen. 1:31, ESV) The paronomastic infinitive, though still disputed with regard to its details, seems to express a kind of certainty: ‫ֵה־בֵ֖ן לְ שָ ָ ָ֣רה ִא ְש ֶ ֶ֑תָך וְ שָ ָ ָ֥רה ש ַ ֛מעַת פֶ ָ֥תַ ח הָ ֵ֖אהֶ ל וְ ָ֥הוא‬ ֵ ‫ו ַַ֗יאמֶ ר ֵּׁׁ֣שֹוב אָ ֵּׁׁ֤שּוב א ֹ֙ליך֙ כָעֵ ָ֣ת חַ ָָ֔יה וְ הִ נ‬ ‫אַ ח ָ ֲָּֽׁריו׃‬ ‘The LORD said, “I will surely return to you about this time next year, and Sarah your wife shall have a son.” And Sarah was listening at the tent door behind him.’ (Gen. 18:10, ESV) 3 Modality in BH: Form inventory The above overview provides us with a strikingly different inventory of means of indicating modality in Biblical Hebrew than generally assumed. Two phenomena will receive further analysis here: word order and particles. 3.1 Word order We may begin with word order and recall the usual conclusion that modality is determined by clause-initial, non-consecutive yiqtol forms. The difference between ‫יִש ָ֥פט יְ הוָ ֵ֖ה‬ ְ in Genesis 16:5 and ‫יִש ָּֽׁפט‬ ְ ‫ֹלהים‬ ָ֣ ִ ‫ ֱא‬in Psalms 82:1 is ‘let the Lord judge’ versus ‘God judges’ or ‘God will judge’. Of great curiosity, then, is the observation that in the examples below it is the fronting of a constituent before the verb, in verbal clauses, or the fronting of a non-verbal predicate before a pronoun, in verbless clauses, that seems to indicate modality. This may be what Gesenius meant when he wrote, regarding unmarked questions: A question need not necessarily be introduced by a special interrogative pronoun or adverb. Frequently the natural emphasis upon the words is of itself sufficient to indicate an interrogative sentence as such.10 10 Gesenius, 1910, p. 473, §150a. Gesenius notes, without evaluation, that H. G. Mitchell considered that there are only 39 such examples, however, of which 12 to 17 must be considered corruptions. We can see that Mitchell was not a fan of unmarked questions. 82 In our first three examples, a redundant pronoun is fronted before the verb, whether inflected or participial. In Joshua 22:18, the ESV uses the earlier question in verse 17 (‘Have we not had enough of the sin at Peor …’) as justification for continuing interrogative modality in verse 18: ‫‘ וְ אַ תֶ םֶׁ֙ תָ שָ֣בו‬that you too must turn away?’ As with Joshua 24:22 in section 1.1 above, there is clear deontic modality, ‘you must turn away’, that may also be considered interrogative, as in the ESV and NIV. The JPS and NET put their own modal twist in their translations, opting for an exclamation point instead of a question mark. ‫הוָּֽׁה׃‬ ָ ְ‫הַ ְמעַט־ ָָּ֙לנוֶׁ֙ אֶ ת־עֲֹוָ֣ ן פְ ָ֔עֹור ֲא ֶ ִׂ֤שר ָּֽׁלא־הִ טַ ֶַׁ֙ה ְרנוֶׁ֙ ִמ ָ֔ ֶמנו עַ ֵ֖ד הַ יָ֣ ֹום הַ זֶ ֶ֑ה וַיְ ִ ָ֥הי הַ נֶ ֵָּ֖֑גֶף בַ ע ֲַדָ֥ת י‬ ‫אַתם ִת ְמ ְר ִׂ֤דו הַ יֹוםֶׁ֙ ַ ָּֽׁביה ָ֔ ָוה ומָ ַ֕ ָחר ֶ ָּֽׁאל־ ָכל־ע ַ ֲָ֥דת יִ ְש ָר ֵ ֵ֖אל‬ ֶ ִ֞ ‫ׁשבּו הַ יָ֔ ֹום מֵ אַח ֵ ֲֵ֖רי יְ הוָ ֶ֑ה וְ הָ ַָ֗יה‬ ֵּׁ֣ ָ‫וְ אַתם֙ ת‬ ‫יִ קְ ָּֽׁצף׃‬ ‘17 Have we not had enough of the sin at Peor from which even yet we have not cleansed ourselves, and for which there came a plague upon the congregation of the LORD, 18 that you too must turn away this day from following the LORD? And if you too rebel against the LORD today then tomorrow he will be angry with the whole congregation of Israel.’ (Josh. 22:17–18, ESV) ‘17 Was not the sin of Peor enough for us? Up to this very day we have not cleansed ourselves from that sin, even though a plague fell on the community of the LORD! 18 And are you now turning away from the LORD? If you rebel against the LORD today, tomorrow he will be angry with the whole community of Israel.’ (Josh. 22:17–18, NIV) ‘17 Is the sin of Peor, which brought a plague upon the community of the LORD, such a small thing to us? We have not cleansed ourselves from it to this very day; 18 and now you would turn away from the LORD! If you rebel against the LORD today, tomorrow He will be angry with the whole community of Israel.’ (Josh. 22:17–18, JPS) ‘17 The sin we committed at Peor was bad enough. To this very day we have not purified ourselves; it even brought a plague on the community of the LORD. 18 Now today you dare to turn back from following the LORD! You are rebelling today against the LORD; tomorrow he may break out in anger against the entire community of Israel.’ (Josh. 22:17–18, NET) In Judges 11, the Amorites demand that the Israelites return land that allegedly rightfully belongs to the Amorites. In the recital of how the land came to be in Israelite hands comes verse 23, where all four sample translations recognise interrogative modality in ‫ ;וְ אַ ָ ֵ֖תה ִת ָיר ֶ ָּֽׁשנו‬this can be seen in the representative translation from NIV. 83 ‫ת־ה ֱאמ ִ ָ֔רי ִמפְ נֵ ֵ֖י ע ַָ֣מֹו יִ ְש ָר ֵ ֶ֑אל וְ אַ ָ ָ֖תה ִּת ָיר ָּֽׁשנּו׃‬ ָ֣ ָ ֶ‫הֹורישֶׁ֙ א‬ ִ ‫ֹלהָ֣י יִ ְש ָר ַ֗ ֵאל‬ ֵ ‫וְ ע ִ֞ ַָתה יְהוָ ָ֣ה ׀ ֱא‬ ‘Now since the LORD, the God of Israel, has driven the Amorites out before his people Israel, what right have you to take it over?’ (Judg. 11:23, NIV) The reason is clear: the context makes nonsense of assertive modality: it cannot be that the text is asserting that the Amorites will or ought to take possession of the land God has given the Israelites. The context unequivocally denies that the Amorites have any right to the land. The only valid translation, therefore, requires an interrogative, challenging modality: ‘what right have you to take the land?’ An example with such a clear context, that forces all translations to identify interrogative modality, is ideal for analysis: can we discover anything, hitherto unnoticed, that might mark the interrogative modality, syntactically or otherwise? The obvious options are the redundant pronoun and its position in front of the verb. The founding of Dan (formerly Laish) is recounted in Judges 18. In Judges 18:9, the versions differ on whether to express incredulity or disgust in translating ‫וְ אַ ֶ ָ֣תם מַ חְ ָ֔ ִשים‬: if incredulity, a pregnant question is appropriate; if disgust, then negative words in an exclamation are suitable. There is unanimity in recognising that this is not a straightforward assertion. ‫טֹובֵ֖ה ְמ ֶ֑אד וְ אַ ֵּׁ֣תם מַ ְח ׂ֔ ִּׁשים אַ ל־‬ ָ ‫ֵיהם ִ ִׂ֤כי ָר ִ ֶׁ֙אינוֶׁ֙ אֶ ת־הָ ָ֔ ָא ֶרץ וְ הִ נֵ ָ֥ה‬ ֶ ָ֔ ‫אמ ַ֗רו קוֵּ֚ מָ ה וְ ַנעֲלֶ ָ֣ה ֲעל‬ ְ ‫וַי‬ ‫תֵ עָ ָ֣צְ ָ֔לו לָלֶ ָ֥ כֶת ל ֵָ֖בא ל ֶ ָָ֥רשֶ ת אֶ ת־הָ ָ ָּֽׁא ֶרץ׃‬ ‘They said, “Arise, and let us go up against them, for we have seen the land, and behold, it is very good. And will you do nothing? Do not be slow to go, to enter in and possess the land.” ’ (Judg. 18:9, ESV) ‘They answered, “Come on, let’s attack them! We have seen that the land is very good. Aren’t you going to do something? Don’t hesitate to go there and take it over.” ’ (Judg. 18:9, NIV) ‘They replied, “Let us go at once and attack them! For we found that the land was very good, and you are sitting idle! Don’t delay; go and invade the land and take possession of it.” ’ (Judg. 18:9, JPS) ‘They said, “Come on, let’s attack them, for we saw their land and it is very good. You seem lethargic, but don’t hesitate to invade and conquer the land.” ’ (Judg. 18:9, NET) As mentioned mentioned in section 1.1 above, Joshua 24:22 demonstrates the verbless clause in which the predicate is before the pronoun, ֶׁ֙‫ע ִ ִֵׂ֤דים אַ תֶ ם‬. Four more examples have different constituents fronted. In Genesis 17:12, the subject is fronted: the eight-day-old male must be circumcised. 84 ‫ֲשר‬ ֛ ֶ ‫ומקְ נַת־ ֶׁ֙ ֶכסֶ ףֶׁ֙ ִמ ָ֣כל בֶ ן־ ֵנ ָָ֔כר א‬ ִ ‫ן־ׁשמֹׁ נַ ֵּׁ֣ת י ִָָּ֗מים יִּ ֶּ֥מֹול לָכֶ ֛ם כָל־זָכָ ֵ֖ר לְ דרתֵ יכֶ ֶ֑ם יְ ִ ָ֣ליד ָ֔ ָביִ ת‬ ְ ‫ּוב‬ ‫ָ֥לא ִ ָּֽׁמז ְַרעֲָךֵ֖ ָּֽׁהוא׃‬ ‘He who is eight days old among you shall be circumcised. Every male throughout your generations, whether born in your house or bought with your money from any foreigner who is not of your offspring …’ (Gen. 17:12, ESV) ‘For the generations to come every male among you who is eight days old must be circumcised, including those born in your household or bought with money from a foreigner – those who are not your offspring.’ (Gen. 17:12, NIV) In Judges 15:18, the adverb ֶׁ֙‫ וְ עַתָ ה‬is fronted, when Samson complains to God, ‘does this have to happen, that I now die of thirst?’ ֙‫אמר אַ תָ הֶׁ֙ ָנ ַ ָ֣תתָ בְ ַיָּֽׁד־עַבְ דְ ָָ֔ך אֶ ת־הַ ְתשועָ ָ֥ה הַ גְ דלָ ֵ֖ה הַ ֶ֑זאת וְ עַ תָ ה‬ ַ ָ֔ ‫וַיִ צְ מָ א֮ ְמא ֒ד ַו ִיקְ ָ ִׂ֤רא אֶ ל־יְ הוָהֶׁ֙ וַי‬ ‫אָ ֵּׁ֣מּות בַ צָ ׂ֔ ָמא וְ ָנ ַפלְ ִ ֵ֖תי בְ יַ ָ֥ד הָ ע ֲֵר ִ ָּֽׁלים׃‬ ‘And he was very thirsty, and he called upon the LORD and said, “You have granted this great salvation by the hand of your servant, and shall I now die of thirst and fall into the hands of the uncircumcised?” ’ (Judg. 15:18, ESV) ‘He was very thirsty and he called to the LORD, “You Yourself have granted this great victory through Your servant; and must I now die of thirst and fall into the hands of the uncircumcised?” ’ (Judg. 15:18, JPS) In 1 Samuel 22:7, with a fronted indirect object, ‫ גַם־לְ כלְ ֶַ֗כם‬, Saul challenges his servants regarding how David would treat them: ‘would he [really] give you fields and vineyards?’ ‫ו ַָ֣י אמֶ ר שָ ַ֗אול ַ ָּֽׁלעֲבָ דָ יוֶׁ֙ הַ נִ צָ ִ ָ֣בים ָע ָָ֔ליו ִש ְמעו־נָ ֵ֖א בְ נֵ ָ֣י ְי ִמינִ ֶ֑י גַם־לְ כלְ ָ֗כם יִּ ֵּׁ֤תן בֶ ן־יִשַ יֶׁ֙ שָ ָ֣דֹות‬ ‫וכְ ָר ִָ֔מים לְ כלְ כֶ ָ֣ם י ָ֔ ִָשים שָ ֵ ָ֥רי ֲאל ִ ֵָ֖פים וְ שָ ֵ ָ֥רי מֵ ָּֽׁאֹות׃‬ ‘And Saul said to his servants who stood about him, “Hear now, people of Benjamin; will the son of Jesse give every one of you fields and vineyards, will he make you all commanders of thousands and commanders of hundreds …” ’ (1 Sam. 22:7, ESV) ‘Saul said to them, “Listen, men of Benjamin! Will the son of Jesse give all of you fields and vineyards? Will he make all of you commanders of thousands and commanders of hundreds?” ’ (1 Sam. 22:7, NIV) ‘Saul said to the courtiers standing about him, “Listen, men of Benjamin! Will the son of Jesse give fields and vineyards to every one of you? And will he make all of you captains of thousands or captains of hundreds?” ’ (1 Sam. 22:7, JPS) 85 ‘Saul said to his servants who were stationed around him, “Listen up, you Benjaminites! Is Jesse’s son giving fields and vineyards to all of you? Or is he making all of you commanders and officers?” ’ (1 Sam. 22:7, NET) And as seen in section 2.2, a fronted object is used in Genesis 42:37 when Reuben grants permission for his own two sons to be put to death, as compensation for Benjamin, if he were not to return safely from Egypt. 3.1.1 Contrast Our last example in this section may be illuminating. In 2 Samuel 15:20, an adverbial temporal is fronted, but it is the contrast between two different temporals, ‫‘ ְת ָ֣מֹול‬yesterday’ and ‫‘ הַ יִ֞ ֹום‬today’, that induces the modality: ‘You came only yesterday; how can I make you wander today?’ ‫בֹואָך וְ הַ יִ֞ ֹום אֲנֹועֲָך עִ ֶָׁ֙מנוֶׁ֙ ָל ֶָ֔לכֶת ַוא ֲִנָ֣י הֹו ֵָ֔לְך עַ ָ֥ל אֲשֶ ר־אֲנִ ֵ֖י הֹולֵ ְֶ֑ך ָ֣שוב וְ הָ ֵ ָּ֧שב אֶ ת־‬ ֶ ַ֗ ‫ְת ֵּׁ֣מֹול ׀‬ ‫אַ ֶח֛יָך עִ ָ ֵ֖מְך ֶחָ֥סֶ ד ֶו ֱא ֶ ָּֽׁמת׃‬ ‘You came only yesterday, and shall I today make you wander about with us, since I go I know not where? Go back and take your brothers with you, and may the LORD show steadfast love and faithfulness to you.’ (2 Sam. 15:20, ESV) Contrast has been present in many of the examples given so far, if most obvious here. Hebrew linguists are most familiar with word order, that is, constituents fronted before the verb, to mark topic and focus. Default syntax can provide an unmarked topic and focus, but when a constituent is fronted to mark the topic or focus, that is nearly always specifically a contrastive topic or focus. Word order with fronted constituents, and contrast, are highly correlated. And yet it has been observed that topic and focus account for, at best, half of the instances of marked, or non-default, word order.11 It may be more helpful to see the marked word order as often signalling contrast, namely, discontinuity with the previous discourse. Marked topic and focus are one instantiation of this contrast, but are not to be identified with it. Here in 2 Samuel, the semantics of the contrast also require a modal interpretation of ִׂ֤‫‘ א ִ ֲָּֽׁנועֲָך‬should I make you wander’. The English translation with a modal verb is a recognition of a contrast present already in the Biblical Hebrew. Biblical Hebrew has no need to doubly mark the modality, as the context is sufficient; it is only English that expresses both the contrast (‘yesterday’ vs ‘today’) and also the modality on the verb (‘should I make you wander’). But the modality is clearly present in BH, even if not marked by a particular verbal form. 11 Referring specifically to Genesis, for example, see Moshavi, 2010, p. 119. See also Hornkohl’s paper in this volume. 86 If it is contrast, or a similar notion, that motivates the marked word order, then the modality perceived and expressed in English translations may be genuinely grammatical modality in English but not in Biblical Hebrew: that is, semantically present in the Biblical Hebrew (and thus justifiably translated) but not explicitly marked in the grammar. 3.2 The particle ‫אַ ף כִ י‬ The compound particle ‫ אַ ף כִ י‬is often recognised as its own unit, meaning ‘furthermore; yea, that!’12 or ‘how much more/how much less when’.13 This latter definition from HALOT is arrived at by combining the ‘also, even’ meaning of ‫ אַ ף‬with the ‘when’ meaning of ‫כִ י‬. ‫מֹותי׃‬ ָּֽׁ ִ ‫ֵהָ֣ן בְ עֹודֶ ִנ ֩י ָּ֙ ַחי עִ מָ ֶ֜כם הַ יַ֗ ֹום מַ ְמ ִ ִׂ֤רים הֱיִ תֶ םֶׁ֙ עִ ם־יְ ה ָֹוָ֔ה וְ ַ ָ֖אף כִּ י־אַ ח ֵ ֲָ֥רי‬ ‘Behold, even today while I am yet alive with you, you have been rebellious against the LORD. How much more after my death!’ (Deut. 31:27, ESV) ‫א־רבְ ָ ָ֥תה מַ כָ ֵ֖ה בַ פְ לִ ְש ִ ָּֽׁתים׃‬ ָ ‫ֲשר מָ ָצֶ֑א ִ ָ֥כי ע ָ ַ֛תה ָּֽׁל‬ ָ֣ ֶ ‫יְבֵ֖יו א‬ ָ ‫ָ֗ ַאף ֹ֡ ִּכי לו ֩א אָ ָּ֙כל אָ כַ ִׂ֤ל הַ יֹוםֶׁ֙ הָ ָָ֔עם ִמ ְשלַ ָ֥ל א‬ ‘How much better if the people had eaten freely today of the spoil of their enemies that they found. For now the defeat among the Philistines has not been great.’ (1 Sam. 14:30, ESV) ‫וש ֵ ִׂ֤מי הַ שָ ֶַׁ֙מ ִיםֶׁ֙ ָ֣לא יְ כַלְ כְ ָ֔לוָך ֶ֕ ַאף ִּ ָּֽכי־הַ ַבָ֥יִת‬ ְ ‫ֹלהים עַל־הָ ָ ֶ֑א ֶרץ הִִ֠ נֵה הַ שָ ֜ ַמ ִים‬ ֵ֖ ִ ‫ֵּ֚ ִכי ַ ָּֽׁהא ְמ ָ֔ ָנם י ֵ ֵָ֥שב ֱא‬ ‫יתי׃‬ ִ ‫ֲשר בָ ִ ָּֽׁנ‬ ָ֥ ֶ ‫הַ זֶ ֵ֖ה א‬ ‘But will God indeed dwell on the earth? Behold, heaven and the highest heaven cannot contain you; how much less this house that I have built!’ (1 Kings 8:27, ESV) This selection of examples reveals how contextually determined each translation is: ‘how much more’, ‘how much better’, ‘how much less’, all from the same compound particle. The linguist sees this phenomenon and hopes to find an underlying category that holds true in all instances, even if expressed differently according to context. From the lexica, it is clear a comparison is made, in which the first situation makes the evaluation of the second situation selfevident. That is, this is a kind of deductive modality, in which the proposition is presented as the only possible conclusion from the available evidence.14 If rebellion happened during a leader’s life, that is cause enough to deduce it would continue after his death. If the people had eaten, that is reason enough to have had strength to fight well. If heaven cannot contain God, there is no cause to believe a house built by man could. 12 Brown, Driver and Briggs (BDB), 1907, p. 65. Koehler and Baumgartner (HALOT), 2000, vol. I, p. 76. 14 Palmer, 2001, p. 25. 13 87 A gloss in casual English that might catch both the comparison and the exclamation might be: ‘[if x,] how could [y]?’ The above translations might then be modified to: ‘if you rebelled during my life, how could you do otherwise after my death?’, ‘if the people had eaten, how could the slaughter of the Philistines not have been great!’ and, lastly, ‘if heaven cannot contain God, how could this house?’ This approach, seeking a valid underlying linguistic meaning, and even, where possible, a translation gloss that captures this, helps makes sense of van der Merwe, Naudé and Kroeze’s comment in their reference grammar, that A speaker/narrator uses ‫ אף‬to introduce a rhetorical question that must be joined to a preceding statement … By using ‫ אַ ף כִ י‬the speaker indicates that what has been suggested in the rhetorical question can only be confirmed in the light of a preceding situation. As with ‫ גַם‬an argument that has been added to another is involved. The second argument is then the one bearing persuasive power. 15 This all comes together in our final example. The lexica and grammars struggle with the use of ‫ אַ ף כִ י‬in Genesis 3:1. ‫ל־ה ִָ֣אשָָ֔ ה אַַ֚ ף ִּ ָּֽכי־אָ ַ ֵּׁ֣מר‬ ָ ֶ‫ֹלהים ו ֶַׁ֙יאמֶ רֶׁ֙ א‬ ֶ֑ ִ ‫ָשה יְ הוָ ָ֣ה ֱא‬ ֵ֖ ָ ‫ֲשר ע‬ ָ֥ ֶ ‫וְ הַ נָחָ שֶׁ֙ הָ יָ ָ֣ה ע ָָ֔רום ִמכ ֶׁ֙ל חַ יַ ָ֣ת הַ שָ ָ֔ ֶדה א‬ ‫ֹלהים ָ֣לא ָּֽׁתאכְ ָ֔לו ִמ ֵ֖כל עֵ ָ֥ץ הַ ָ ָּֽׁגָּ֑ן׃‬ ִּׂ֔ ֱ‫א‬ ‘Now the serpent was more crafty than any other beast of the field that the LORD God had made. He said to the woman, “Did God actually say, ‘You shall not eat of any tree in the garden’?” ’ (Gen. 3:1, ESV) ‘Now the serpent was the shrewdest of all the wild beasts that the LORD God had made. He said to the woman, “Did God really say: You shall not eat of any tree of the garden?” ’ (Gen. 3:1, JPS) ‘Now the serpent was more shrewd than any of the wild animals that the LORD God had made. He said to the woman, “Is it really true that God said, ‘You must not eat from any tree of the orchard’?” ’ (Gen. 3:1, NET) One can hardly gloss it as ‘how much more’, ‘how much less’ or something similar. What preceding situation is there for comparison to what is being said? And so, BDB provides an additional meaning for ‫אַ ף כִ י‬, exclusively for Genesis 3:1. It refers to the entry for ‫‘ הַ אַ ף‬indeed? really?’ By treating the ‫אַ ף‬ ‫ כִ י‬as if it were ‫הַ אַ ף‬, BDB is able to create a new definition for ‫אַ ף כִ י‬.16 I am not sure lexical semanticists would approve. Joüon and Muraoka, on the other hand, consider this particular example not as a case of the compound particle, as that clearly does not fit, but as a simple ‫ אַף‬followed by ‫ כִ י‬introducing a 15 16 Van der Merwe, Naudé and Kroeze, 1999, p. 313. Possibly influenced by Rashi’s interpretation of ‫ אַ ף כִ י‬as ‫‘ שמא‬perhaps’ in Miqraʾot Gedolot? 88 subject clause, and they insert an implied ‘(is it)’ to make the meaning ‘(is it) also that he said’, which they equate to ‘is it then true that he said?’17 From the context of Genesis 3, we can intuit that these scholars are close, if not exactly right, in their conclusions. But if we recall the other uses of ‫אַ ף‬ ‫כִ י‬, we might have more to offer. With the gloss of ‘how could’ proposed for the other examples, a possible translation for the second half of Genesis 3:1 might be ‘How could God say that you must not eat from any tree in the garden?’ This translation makes clear the deductive modality involved. That is, the serpent is not merely casting doubt on the situation at hand, he is declaring it to be impossible. The shift is from a simple incredulous question, which is the usual interpretation, to a rhetorical question. As a rhetorical question, it does not permit a response, but instead asserts the impossibility of God saying any such thing. The implication must be that, based on what we know of God already, it is impossible that he could say something like this. The woman’s response shows her struggle with this, because she knows that, actually, it does very well permit a response. The serpent is wrong! But he is also subtle and devious, and if he does not persuade her to his view, he succeeds in dragging her into doubt. Here is the beauty of literary artistry. And an example of particles requiring that we translate a verb modally, although in BH the particle was sufficient marking. 3.3 The particle ‫נָא‬ Biblical Hebrew has another, largely unrecognised, category of overlapping modality. What we call an imperative by form may in fact be used as a polite interrogative. Note the role of ‫ נָא‬in every example below. In Judges 19, the story of the Levite and his concubine starts with politeness and hospitality, even if it ends otherwise. In verse 6, the father requests that the Levite stay, spend the night and enjoy himself. This is expressed with two imperatives and the functional equivalent, the jussive, for ‘let your heart be merry’. But the function is that of a question, which JPS captures well: ‘Won’t you stay?’ The meaning is not much different from NIV’s ‘Please stay tonight’, but it is more transparent: there is no directive, or command, but instead a request. This becomes clear in the remaining examples. ‫ֵיה֛ם יַחְ ָ ֵ֖דו וַיִ ְש ֶ֑תו ו ַ֜יאמֶ ר א ִ ֲִׂ֤בי ַ ָּֽׁה ַנע ֲָרהֶׁ֙ אֶ ל־הָ ִָ֔איש ָּֽהֹואל־נָ ֶּ֥א וְ ִּ ָ֖לין וְ י ַ ִָ֥טב לִ ֶ ָּֽׁבָך׃‬ ֶ ‫ַוי ְֵשבַ֗ ו וַיאכְ לָּ֧ ו ְשנ‬ ‘So the two of them sat down and they feasted together. Then the girl’s father said to the man, “Won’t you stay overnight and enjoy yourself?” ’ (Judg. 19:6, JPS) 17 Joüon and Muraoka, 2006, p. 557 §157n2. 89 In Judges 19:11, the servant must be asking, ‘Shall we stop at this city of the Jebusites, and spend the night?’ because his master replies, ‘No’. ‫בוסי‬ ָ֥ ִ ְ‫ל־עיר־הַ י‬ ָּֽׁ ִ ֶ‫ֵ ָ֣הם עִ ם־יְבָ֔ ו ס וְ הַ יֵ֖ ֹום ַ ָ֣רד ְמ ֶ֑אד ו ַָּ֙יאמֶ ר הַ ֜ ַנעַר אֶ ל־ ֲאד ַ֗ ָניו לְ כָה־נָ ָ֛א וְ נ ָָ֛ס ָּורה א‬ ‫הַ ֵ֖זאת וְ נ ִ ָָ֥לין ָ ָּֽׁבּה׃‬ ‫א־מבְ נֵ ָ֥י יִ ְש ָר ֵ ֵ֖אל ֵ ֶ֑הנָה וְ ע ַָב ְֵ֖רנו עַד־גִ בְ ָ ָּֽׁעה׃‬ ִ ‫ֲשר ָּֽׁל‬ ֛ ֶ ‫ל־עיר נָכְ ִ ָ֔רי א‬ ָ֣ ִ ֶ‫ו ִַׂ֤יאמֶ ר אֵ לָיוֶׁ֙ ֲאד ָ֔ ָניו ֵּׁ֤ל ֹׁא נָסורֶׁ֙ א‬ ‘When they were near Jebus and the day was almost gone, the servant said to his master, “Come, let’s stop at this city of the Jebusites and spend the night.” His master replied, “No. We won’t go into an alien city, whose people are not Israelites. We will go on to Gibeah.” ’ (Judg. 19:11–12, NIV) Similarly, in the next three examples, the question is proven to be a question by the affirmative replies, in Biblical Hebrew style, repeating the salient part of the question. In 2 Samuel 14:12, after ‘May your servant speak?’, the king answers, ‘Yes’ or ‘Speak’. ‫ו ֶַׁ֙תאמֶ רֶׁ֙ ָ ָּֽׁה ִאשָָ֔ ה ְתדַ בֶ ר־נָ ָּ֧א ִשפְ חָ ְתָך֛ אֶ ל־ ֲאדנִ ָ֥י הַ ֶ ֵ֖מלְֶך דָ ָבֶ֑ר ו ֵַ֖יאמֶ ר ּדַ ָּֽב ִּרי׃‬ ‘Then the woman said, “Please let your servant speak a word to my lord the king.” He said, “Speak.” ’ (2 Sam. 14:12, ESV) In 2 Kings 6:3, Elisha is asked, ‘Will you come?’ and he replies, ‘Yes’ or ‘I will come’. ‫ו ֶַׁ֙יאמֶ רֶׁ֙ ָ ָּֽׁהאֶ ָ֔ ָחד ֶּ֥הֹואל נָ ָ֖א וְ ֵּׁ֣לְך אֶ ת־עֲבָ ֶ ֶ֑דיָך ו ֵַ֖יאמֶ ר אֲנִ ָ֥י א ָּֽלְך׃‬ ‘Then one of them said, “Will you please come along with your servants?” “Yes, I will come”, he said.’ (2 Kings 6:3, JPS) In Ruth 2:2, Ruth asks permission to go into the fields to gather grain, and Naomi grants her permission, ‘Yes, you may go’. ‫א־חֵ֖ן‬ ֵ ָ‫ֲשר אֶ ְמצ‬ ָ֥ ֶ ‫וַתאמֶ ֩ר ָּ֙רות הַ מֹואֲבִ ָ֜יה ֶ ָּֽׁאל־ ָנע ֳִַ֗מי ָּֽאלְ כָה־נָ ֵּׁ֤א הַ שָ דֶ הֶׁ֙ ַו ֲאלַקֳ ָ ָ֣טה בַ ִשבֳלִָ֔ ים אַ ַ֕ ַחר א‬ ‫בְ ֵעינָ ֶ֑יו ו ַָ֥תאמֶ ר לָ ֵּ֖ה לְ ִּ ֶּ֥כי בִ ִ ָּֽׁתי׃‬ ‘Ruth the Moabite said to Naomi, “I would like to go to the fields and glean among the ears of grain, behind someone who may show me kindness.” “Yes, daughter, go”, she replied.’ (Ruth 2:2, JPS) In all of these, the imperative with ‘please’ is also acceptable in English, because we can also ask a question with a polite imperative. But we must also recognise the interrogative at play that justifies the answer ‘yes’ or ‘no’. What we have is a weak deontic modality, in which the request is a weak imposition of will. The request admits of a refusal as well as an acquiescence, which accounts for the interrogative function and the ease of an interrogative interpretation. Once again, deontic modality and interrogative modality overlap each other. 90 4 Rhetorical questions 4.1 Literature review (Mitchell, Brongers) We may now fit this sketch of modality and its signals in with other scholarship on Biblical Hebrew. Mitchell argued strongly, back in 1907, that the unmarked question, rhetorical or not – what he called omission of the interrogative – is mostly a misidentified creature. Instead, what we mostly have are cases of textual corruption or mistaken exegesis. He concludes: If, therefore, one were required to make a statement on the subject, one would have to say that in direct single or initial questions ‫ ה‬is omitted before the article, and sometimes in exclamatory questions for the purpose of indicating more clearly the incredulity, irony, or sarcasm which prompted them, but which can be adequately expressed only by the human voice. 18 In describing these unmarked questions, he considers these three main categories (incredulity, irony and sarcasm) as “so many varieties of what might be called exclamatory questions, and appropriately marked by a double punctuation (! ?)”.19 What he called an exclamatory question, we now call a rhetorical question, having the form of a question but the function of an exclamation or strong assertion. This is Mitchell’s summation of the twenty-seven cases of genuine unmarked questions, according to his assessment. But he does not include examples which the versions nearly unanimously translate as questions, though there is no interrogative particle in sight. In Exodus 8:22, the apodosis of a conditional sentence, ‫וְ ָ֥לא יִ סְ קְ לָּֽׁנו‬, is translated by most as a question, even with the interrogative particle ‫ הְ לָא‬in Targum Onqelos. Targums Neofiti and Pseudo-Jonathan add their own expressions to convey the modal meaning they sense in these two final Hebrew words. ‫ֵיהֵ֖ם וְ ֶּ֥ל ֹׁא יִּ ְסקְ ָּֽלנּו׃‬ ֶ ‫ֵהָ֣ן נִ ְז ִ֞ ַבח אֶ ת־תֹוע ֲַבָ֥ת ִמצְ ַ ֛ר ִים לְ עֵינ‬ ‘If we sacrifice offerings abominable to the Egyptians before their eyes, will they not stone us?’ (Ex. 8:22b, ESV) ‘If we sacrifice that which is untouchable to the Egyptians before their very eyes, will they not stone us!’ (Ex. 8:22b, JPS) ‘If we make sacrifices that are an abomination to the Egyptians right before their eyes, will they not stone us?’ (Ex. 8:22b, NET) 18 19 Mitchell, 1907, p. 129. Mitchell, 1907, p. 129. 91 ‫צראֵ י דָ חְ לִ ין לֵיה ִמנֵיה ְאנַחנָא נ ְָסבִ ין לְ דַ בָ חָ א‬ ָ ‫ו ְַאמַ ר משַ ה לָא תָ קֵ ין לְ מַ עְ בַ ד כֵין ְא ֵרי בְ עִ ָירא דְ ִמ‬ ‫ימרּון‬ ְ ‫צראֵ י דָ חְ לִ ין לֵיה וְ ִאנון יְ הֹון חָ זַן ְהלָא י‬ ָ ‫קדם יוי ְאלָהַ נָא הָ א נְ דַ בַ ח יָת בְ עִ ָירא ְד ִמ‬ ‫לְ ִּמרגְ מַ נָא׃‬ ‘… will they not command to stone us?’ (TG Onqelos) ‫לית אפׁשר דלא ירגמון יתן׃‬ ‘… it is not possible that they not stone us.’ (TG Neofiti, Esther Sheni) ‫מן דינא הוא לאטלא יתן באבנין‬ ‘… by right they could stone us with stones.’ (TG Ps-Jonathan) We might even broaden the field of rhetorical exclamations if we listen to Brongers20 argue about those in which ‫ הֲלא‬and ‫ הִ נֵה‬are actually interchangeable, both functioning to call attention to the obvious: compare 1 Samuel 9:21 and Judges 6:15, for example. ‫ומ ְשפַחְ ִתיֶׁ֙ הַ צְ עִ ָ ָ֔רה ִמ ָ ָּֽׁכל־‬ ִ ‫ַו ַָּ֙יעַן שָ ֜אול ו ַַ֗יאמֶ ר ה ֲֹ֙לֹוא בֶ ן־יְ ִמינִ ִׂ֤י ֶָׁ֙אנכִ יֶׁ֙ ִמקַ טַ נֵיֶׁ֙ ִשבְ ֵ ָ֣טי יִ ְש ָר ָ֔ ֵאל‬ ‫ִמ ְשפְ ֵ֖חֹות ִשבְ ֵ ָ֣טי בִ נְ י ִ ֶָ֑מן וְ ָָּ֙למָ הֶׁ֙ ִד ַב ְָ֣רתָ אֵ ַָ֔לי כַדָ ָבֵ֖ר הַ ֶזָּֽׁה׃‬ ‘Saul answered, “Am I not a Benjaminite, from the least of the tribes of Israel? And is not my clan the humblest of all the clans of the tribe of Benjamin? Why then have you spoken to me in this way?” ’ (1 Sam. 9:21, ESV) ‫אֹושי ַע אֶ ת־יִ ְש ָר ֵ ֶ֑אל ִּהנֵּׁ֤ה אַ לְ פִ יֶׁ֙ הַ ַ ָ֣דל בִ ְמנַשֶָ֔ ה וְ אָ נ ִ ָ֥כי הַ צָ ִ ֵ֖עיר בְ ֵבָ֥ית‬ ֵ֖ ִ ‫ו ִַׂ֤יאמֶ ר אֵ לָיוֶׁ֙ ִ ָ֣בי ֲאד ָ֔ ָני בַ ָ ָ֥מה‬ ‫אָ ִ ָּֽׁבי׃‬ ‘And he said to him, “Please, Lord, how can I save Israel? Behold, my clan is the weakest in Manasseh, and I am the least in my father’s house.” ’ (Judg. 6:15, ESV) Brongers concludes that, in spite of the interrogative particle in ‫הֲלא‬, there can be no interrogative modality involved.21 I think we could resolve Brongers’ dilemma, however, by recognising that form and function do not always map consistently between languages. In BH, these exclamations, as Brongers calls them, were appropriately considered rhetorical questions (with interrogative marking) or as exclamations (with ‫)הִ נֵה‬. In English, the exclamation sounds far more natural than the question. The dilemma, however, is purely one of the target language, English; it is not a problem in the Biblical Hebrew. Consider how many examples there have been so far in which the versions have waffled between question marks and exclamation points. Instead of distinguishing between discrete categories, we have overlapping categories, here of surprise modality and interrogative modality. 20 Brongers, 1981. He finds confirmation in the verb ‫ שָ אַ ל‬never being used in conjunction with ‫הֲלא‬, but instead ‫ אָ מַ ר‬or ‫קָ ָרא‬. 21 92 4.2 Summary What we find, then, is a more complex situation than generally assumed. Modality in BH is far more than the volitional paradigm, and the different kinds of modality can overlap, to the consternation of translators. Also, the reality is that mismatches of form and function are to be expected in human language. Where the interrogative form is not accompanied by the expectation of an answer, either in Biblical Hebrew or in English, we call it a rhetorical question. If either interrogative or exclamatory particles, in BH, or punctuation, in English, seem interchangeable, then we may even call them rhetorical exclamations. 4.3 Identifying questions How, then, to identify such mismatches? Are there rhetorical questions heretofore assumed to be normal questions? Are there unmarked questions, heretofore assumed to be assertions, that are actually unmarked rhetorical questions? This paper cannot answer any of these questions comprehensively, but some helpful observations have already been made. First, it must be affirmed that there are many cases of interrogative modality in BH that are not marked by an interrogative particle. In spite of Mitchell’s protestations, the intuition that has led so many versions to nonetheless translate questions, even in the absence of a BH interrogative, is not always misguided. This paper attempts to demonstrate that, although lacking the interrogative particle, there may be other forms of marking, that show interrogative modality was indeed intended. We have seen three types of marking: 1. The most obvious evidence for an intended question comes when an answer is supplied. Joshua 24:22 (discussed in section 1.1 above) was one example of this. These are obviously not rhetorical questions, and may bear additional modality, such as deontic modality. Translations may vary in which modality they express in the target language. 2. Some particles, not normally considered interrogative, may in fact indicate interrogative modality, whether rhetorical or not. The particles ‫ אַ ף כִ י‬and ‫ נָא‬were the two discussed here (in sections 3.2 and 3.3, respectively). The variation in translation between exclamation points and question marks points to the dual nature of this category, as well: both surprise modality (exclamation) and interrogative modality (question). 3. Word order, in particular non-verb-first clauses, are commonly explained pragmatically, as expressing topic or focus. But, as already mentioned in section 3.1.1, marked topic and focus by no means exhaust the cases of non-verb-first clauses. The examples of unmarked 93 questions above may suggest an explanation for some of these unresolved cases: non-verb-first word order may mark a contrast that induces some kind of modality. Most examples of modality shown through word order observed above actually fall neatly into Mitchell’s categories of incredulity, sarcasm and irony, suggesting that non-verb-first word order may mark interrogative modality for rhetorical questions, or, at times, rhetorical exclamations. Many of these examples include a personal pronoun before the verb, such as: ‫וַיאמֶ ר֮ נָתָ ֒ן ֲאדנִ ָ֣י הַ ָ֔ ֶמלְֶך אַ ָ ֵּׁ֣תה אָ ׂ֔ ַמ ְרתָ ֲאד ִניָ ֵ֖הו יִ ְמֹלָ֣ ְך אַ ח ָ ֲֶ֑רי וְ ֵ֖הוא י ֵ ֵָ֥שב עַל־כִ ְס ִ ָּֽׁאי׃‬ ‘Have you [really] said, “Adonijah shall reign after me?” ’ (1 Kings 1:24, ESV) ‘… you must have said, “Adonijah shall succeed me as king…” ’ (1 Kings 1:24, JPS) ‫לּוכָ֖ה עַל־יִ ְש ָר ֵ ֶ֑אל ִׂ֤קום ֱאכָל־ ֶָּ֙לחֶ םֶׁ֙ וְ י ִַטָ֣ב‬ ָ ‫ֲשה ְמ‬ ֶּ֥ ‫ו ִַׂ֤תאמֶ ר אֵ לָיוֶׁ֙ ִאיזֶ ָ֣בֶ ל ִא ְש ָ֔תֹו אַ תֶָ֕ ה ע ָ ַ֛תה תַ ע‬ ‫ֵאלי׃‬ ָּֽׁ ִ ‫לִ ָ֔ ֶבָך ֲא ִניֶׁ֙ אֶ ֵ ָ֣תן לְ ׂ֔ך אֶ ת־כֶ ֵֶ֖רם נ ָָ֥בֹות הַ ִיז ְְרע‬ ‘Do you now govern Israel? Arise … [!]’ (1 Kings 21:7, ESV) ‘You are the king of Israel! Get up …’ (1 Kings 21:7, NET) ‫ָכלָה֙ אַ ָ ֵּׁ֣תה עֹׁ ׂ֔שה ֵ ֵ֖את ְשאֵ ִ ָ֥רית יִ ְש ָר ֵ ָּֽׁאל׃‬ ‘Will you make a full end of the remnant of Israel?!’ (Ezek. 11:13, ESV) ‘You are wiping out the remnant of Israel?!’ (Ezek. 11:13, JPS) 5 Rhetorical questions and theological presuppositions This brings us, finally, to rhetorical questions. The above observations make some questions more identifiable, but there still remain questions with context as the only clue. For example, 1 Samuel 24:20, with no cultural or literary context, would be read as, ‘When a man finds his enemy, and he sends him away safely …’ with no modality at all, but simply a compound temporal clause. But within its context here, it becomes a rhetorical question. David had the opportunity to harm Saul, who had been seeking his life, but David refused. Saul now acknowledges the unexpected nature of David’s kindness to him, for to let an enemy go, safe, was contrary to all norms in the ancient Near East. ‫טֹובֶ֑ה‬ ָ ‫ת־איְבָ֔ ֹו וְ ִשלְ ֵ֖חֹו בְ ֶד ֶָ֣רְך‬ ָ֣ ֶ‫וְ ִ ָּֽׁכי־ ִי ְמ ָצָ֥א ִאישֶׁ֙ א‬ ‘For if a man finds his enemy, will he let him go away safe?’ (1 Sam. 24:20, ESV) 94 Even where there are additional signals, such as word order, we observed above that in Judges 11:23 it was the context that made the translations unanimous in detecting a question. In Isaiah 1:18 (see section 1.2), it is the context and meaning of the verse that make Duhm and followers reject the standard translation, in which God promises that sins can become white as snow.22 For forgiveness is the removal of sin, which is forever stained. To change the colour of sin, to make sin no longer sinful, is antithetical to all biblical theology. The context, in which God presides in the courtroom, argues for clearing away any pretences: if the sins are as scarlet, shall they appear white as snow in court? If they are as crimson cloth, shall they appear as wool, in court? By no means. And so Culver makes a strong case that Isaiah 1:18 must be the denial that sins can ever be anything but sinful.23 This does not do away with repentance and forgiveness, but it removes those notions from this particular verse. I think I am persuaded. This leads to that politically charged notion of theological presuppositions. All interpreters, translators foremost among them, interpret the Hebrew text based on a bewildering array of cues. Context, that is, their understanding of context, is necessarily a significant factor. When that understanding is shared by all, no difficulty arises, as in Judges 11:23. But where understanding differs, as in Isaiah 1:18, there is scholarly debate. This is as it should be. We must use our full arsenal of skills and convictions in approaching the biblical text, but we must do so with integrity. Of necessity, we all have theological presuppositions, whether they involve sin and forgiveness or the possibility of miracles. It would impoverish our scholarship to lay these aside. Context, and all the ways in which we understand it, is a critical tool in interpreting the text as we have it. To avoid special pleading, we must not fool ourselves into thinking we can dispense with theological presuppositions. But we can replace naive notions of objectivity with a more mature notion of scholarly integrity. 6 Conclusion This paper has been exploratory, probing into modality in Biblical Hebrew that goes beyond our usual expectations of marking. And this, indeed, may be the most valuable contribution: Modality in Biblical Hebrew is demonstrably more than the verbal system. We do find deontic modality within the volitional paradigm, but we also find it outside the verbal system, indicated by phenomena such as word order and particles. We find epistemic modality within and without the verbal system, and the same with evidential modality. The verbal system is only one part of expressing modality in Biblical Hebrew. 22 23 Duhm, 1892. Culver, 1969. 95 But although modality is clearly found outside verbal inflection, this is not to claim there is a grammatical system, or even specific grammatical markings of modality. There may be notional modality present that requires no grammatical marking in Biblical Hebrew, but upon translation to English requires an explicit modal rendering. Another result is the recognition that interrogative modality is not a discrete category. There is an overlap between interrogative modality and other kinds of modality, which may even account for many of the ‘unmarked questions’. Requests may have both deontic and interrogative modality. Exclamations may have both deductive and interrogative modality. As in English, where we can interchange question marks and exclamation points, Biblical Hebrew can behave similarly. Of particular interest to grammarians will be the observations on word order. Until now, verb-first yiqtol, or imperfect, clauses have nearly universally been considered to mark deontic modality. ‘May the Lord judge between us’ is, without a second thought, a verb-first clause. This is certainly common for jussive modality, but now we have seen that verb-medial clauses can also be an indication of modality. In most of our examples, it would seem the interrogative modality functions as a challenge: ‘Do you dare now turn away?’, ‘Are you to take possession of the land?’, ‘Are you doing nothing?!’ It may be that when the challenge inherently contrasts with the assumptions it is challenging, there is a natural alignment with syntactic signals for contrast, namely, fronted constituents, or verb-medial clauses. If so, is there then a semantic restriction, along the lines of Mitchell’s incredulity, sarcasm and irony, to what marked word order will indicate? Further research is required to determine how direct or indirect the correlation is between word order and modality. This leads to the whole concept of unmarked questions. In speech, unmarked sarcasm, that is, non-sarcastic in form but sarcastic in function, is the hallmark of dry humour. One is easily misled in attempting to interpret unmarked sarcasm. When speech is reduced to writing, vocal cues are lost, whereas other cues are gained, for example, orthography and punctuation. There are sufficient examples of unmarked questions in BH to document the phenomenon, and without doubt there are many more, still unrecognised, in the more difficult passages of the prophets and Job. This still remains a challenging area. Rhetorical questions can be a quick solution to a thorny theological problem. How can sins change colour and become white? If Isaiah 1:18 is a rhetorical question, then the problem disappears: the very thrust of the verse is that sins cannot ever change colour, or be anything but sinful. Rhetorical questions have as their very purpose to use rhetoric to challenge people’s assumptions. We’ve seen this multiple times with interrogative modality, particularly as it overlaps with other forms of modality. 96 As we interpret and translate the text, we must be aware of cues, both inside and outside the verbal system, to possible modal meanings. Grammarians must not claim that yiqtol is compatible with all manner of modality, but then provide no guidance as to how to identify the various kinds. May this paper be a step in the direction of providing such guidance. References Brongers, H. A., 1981, “Some remarks on the Biblical particle hªlōʾ ”, Oudtestamentische Studiën 21, pp. 177–189. Brown, F., Driver, S. and Briggs, C., 1907, The Brown-Driver-Briggs Hebrew and English Lexicon, Oxford: Clarendon Press. Callaham, S. N., 2010, Modality and the Biblical Hebrew Infinitive Absolute, Wiesbaden: Harrasowitz. Culver, R. D., 1969, “Isaiah 1:18 – declaration, exclamation or interrogation?”, Journal of the Evangelical Theological Society 12, pp. 133–141. Duhm, B., 1892, Das Buch Jesaja, Göttingen: Vandenhoek & Ruprecht. Gesenius, W., 1910, Gesenius’ Hebrew Grammar, 2nd English edn, E. Kautzsche and A. E. Cowley (eds), Oxford: Clarendon Press. Hatav, G., 1997, The Semantics of Aspect and Modality: Evidence from English and Biblical Hebrew, Amsterdam: John Benjamins. Joüon, P. and Muraoka, T., 2006, A Grammar of Biblical Hebrew, Rome: Pontifical Biblical Institute Press. Koehler, L. and Baumgartner, W., 2000, The Hebrew and Aramaic Lexicon of the Old Testament, Leiden: Brill. van der Merwe, C. H. J., Naudé, J. A. and Kroeze, J. H., 1999, A Biblical Hebrew Reference Grammar, Sheffield: Sheffield Academic. Mitchell, H. G., 1907, “The omission of the interrogative particle”, in R. F. Harper, F. Brown and G. F. Moore (eds), Old Testament and Semitic Studies in Memory of W. R. Harper, Chicago: University of Chicago Press, pp. 115–129. Moshavi, A., 2010, “ ‘Is that your voice, my son David?’: conducive questions in Biblical Hebrew”, Journal of Northwest Semitic Languages 36, pp. 65–81. Palmer, F. R. 2001. Mood and Modality, Cambridge: CUP. Williamson, H., 2006, Isaiah 1–5, London: T&T Clark. 97 The Shewa in the First of Two Identical Letters and the Compound Babylonian Vocalisation SHAI HEIJMANS University of Cambridge 1 Introduction1 The pronunciation of the shewa in Tiberian Hebrew is one of the most complicated topics in Hebrew phonology.2 One of the traditional categories into which this topic is subdivided relates to the shewa in the first of two identical, consecutive, letters (e.g. ‫‘ יְהַ לְ לֶ ֶ֑ ָך‬he will praise you’, Isa. 38:18), and the circumstances under which this shewa is vocal or silent. The aim of this article is to bring additional data to the discussion from a manuscript that is vocalised with the so-called ‘Compound Babylonian’ vocalisation system. Despite its name, the Compound Babylonian system in this manuscript represents, with very few exceptions, the Tiberian pronunciation tradition.3 2 Pronunciation of the shewa in the first of two identical letters The most widely used rules for distinguishing a vocal from a silent shewa are those formulated by Eliyahu Bachur (1469–1549), which state, inter alia, that a shewa in the first of two identical letters is vocal.4 A similar rule is brought by Rabbi Shelomo Almoli (c1490–1542), in his book Halikhot Shewa: 1 I would like to thank the editors of this volume for their valuable comments on an earlier draft of this article. 2 For a updated survey on this topic with additional bibliography, see Khan, 2013a. 3 The manuscript in question is the Codex Babylonicus Petropolitanus; the manuscript and its vocalisation will be surveyed in section 4 of this article. 4 See a restatement of these rules in Joüon and Muraoka, 2006, pp. 48–49. 98 two identical letters in one word – if the first is with a shewa, a pataḥ is always attached to it, e.g. ‫סובֲבים‬, ‫סוכֲכים‬, ‫שוטֲטו‬, except in Aramaic, where we find only a shewa [i.e. silent shewa], e.g., ‫[ ורעיוני לבְ בך‬Daniel 20:1].5 Vocal shewa in these circumstances is also reflected in the pronunciation of most Jewish communities in the Modern Era, both among Ashkenazim and among Sefaradim.6 The Tiberian treatises, however, seem to reflect a somewhat different rule. According to Sefer Diqduqe Haṭṭeʿamim, the nature of the shewa depends on the existence of a gaʿya in the preceding syllable: if there is a gaʿya, the shewa is vocal, for example, ‫‘ ִ ָּֽׁילְ לַ ָ֣ת הָ ר ָ֔ ִעים‬the wail of the shepherds’ (Zech. 11:3); if there isn’t, then the shewa is silent, for example, ‫י־אוֶן‬ ֶ֑ ָ ֵ‫‘ חִ קְ ק‬iniquitous decrees’ (Isa. 10:1).7 Several forms are listed as exceptions to this rule (e.g., ‫ִ֭ ִיקְ ָראנְ נִי‬ ‘they will call upon me’, Prov. 1:28; ‫‘ ְ ָּֽׁי ַַ֫כבְ ָ ָ֥דנְ ִני‬he glorifies me’, Ps. 50:23). In those exceptional cases the shewa follows the stressed syllable, which is deemed to be long, but is nevertheless silent. It should be noted that the term gaʿya in this passage does not denote a particular graphic sign; rather, it has the meaning of a vowel which is lengthened, either because of secondary stress or because of primary stress.8 A similar rule, including the exceptional cases mentioned above, is found in the Anonymous Treatise about the Shewa which was published by Kurt Levy.9 Here too, the shewa is vocal if it follows a gaʿya, and silent if it does not. However, the relevant passage is followed by an important addition, stating that if the shewa is preceded by a ḥolem, it is always vocal, whether there is a gaʿya or not; and also that if the shewa is preceded by a ḥireq, there are two exceptional cases where it is silent despite the fact that the ḥireq comes with a gaʿya.10 These exceptions are: ‫‘ ִ ָּֽׁרבְ ֵ֖בות אַ לְ פֵ ָ֥י יִ ְש ָר ֵ ָּֽׁאל׃‬ten thousand thousands of Israel’ (Num. 10:36) and ‫י־שמֶ ן‬ ֶ֑ ָ ‫‘ בְ ִ ָּֽׁרבְ ֵ֖בות ַ ָּֽׁנ ֲח ֵל‬with ten thousands rivers of oil’ (Mic. 6:7). In these two cases, despite the existence of a preceding gaʿya, the shewa is silent. A third source for the Tiberian pronunciation of the shewa in the first of two identical letters is the treatise Hidāyat al-Qāri.11 According to its author, the shewa is vocal if it is preceded by taṯqīl ‘weighting, burdening’.12 The author further states that the taṯqīl is caused by gaʿya, by ta’am (i.e. primary 5 See Yalon, 1945, p. 45; see also Yalon, 1963, pp. 88–89. For Ashkenazi Jews, see Yalon, 1963, p. 89; for Moroccan Jews, see Maman, 1994, p. 97 and Akun, 2010, pp. 192–194; for Syrian Jews, see Katz, 1981, p. 62; for the Jews of Djerba, see Katz, 1977, pp. 105–106; for Italian Jews, see Artom, 1947, p. 59. 7 See Dotan, 1967, chapter 5, pp. 115–116 and his analysis on pp. 189–192. 8 See Doton, 1967, §‫ב‬, p. 191 and p. 353 n. 18‫א‬. 9 Levy, 1936, text on p. ‫( טז‬translation on pp. 15*–16*). 10 These are on top of the exceptional cases mentioned in Sefer Diqduke Haṭṭeʿmim. 11 The author of this treatise is unknown, as are the place where and date when it was written. Eldar argues for Palestine, mid-11th century; he also attributes the work to Abu l-Faraj Harūn. See Eldar, 1994, pp. 19–43. 12 Eldar, 1987, pp. 15–16, lines 165–169; see also his Hebrew translation on p. 33. 6 99 stress) or if the preceding vowel is being ‘produced fully by a king’. Yalon argued, quite convincingly, that the term ‘king’ employed here refers to the ḥolem,13 meaning that when a shewa follows a ḥolem, it is always vocal.14 The fourth and last source that will be reviewed here is Kitāb faṣīḥ luġat al-ʿIbrāniyyīna (‘Book of elegance of the Hebrew language’) by Saadia Gaon (died in 942).15 According to Saadia, the general rule is that the shewa in the first of two identical letters is vocal; as examples he gives ‫יסובְ בו‬ ָ֥ ִ‫‘ ו‬and they prowl’ (Ps. 59:7), ֶׁ֙‫‘ ַויְעלְ ָּ֙להו‬and he made a gleaning of him’ (Judg. 20:45), ‫‘ וַיִ ְתרצְ ִׂ֤צו‬and they struggled together’ (Gen. 25:22), and ‫‘ דָ לְ לָ֥ ו‬they will diminish’ (Isa. 19:6). Then he limits the general rule by stating that the shewa is vocal “when the vowel that precedes the two letters is long (mamdūdah) ... but if the vowel is short (maqṣūrah), the shewa does not need to be vocal”, and as an example he gives the imperative form ‫‘ וְ שָ ְד ֵ֖דו‬and destroy!’ (Jer. 49:28).16 The principle that emerges from all four sources is the same: apart from a few exceptions, the pronunciation of the shewa in the first of two identical letters is conditioned by the quantity of the vowel that precedes it – when the vowel is long, the shewa is vocal, and when the vowel is short, the shewa is silent. The question that then arises is how to infer the length of a specific vowel from the vocalisation signs; how do we know, for example, that the ḥireq in one word is long, while the ḥireq in another word is short? 3 The gaʿya and the indication of vowel length in the Tiberian vocalisation system The standard Tiberian vocalisation system was designed, initially, to indicate two phonemic vowel quantities: vowels of ‘full’ length (qameṣ, pataḥ, ṣere, segol, ḥireq, qibbuṣ/šureq) on one hand, and vowels of ‘furtive’ length (the shewa and the ḥatafim) on the other hand.17 At a later point in time, however, it was deemed necessary to mark length differences within the full vowels (which is usually non-phonemic). For this purpose the gaʿya sign was added.18 However, different Tiberian schools seem to have had different traditions regarding the exact pronunciation of non-phonemic vowel lengths, resulting in considerable differences as to the placement 13 Yalon, 1963, pp. 86–87. His argumentation is based on various quotations from the medieval literature, where ḥolem is the ‘king of all kings’. 14 Eldar seems to reach the same conclusion as Yalon; see his analysis of this passage in Eldar, 1994, pp. 151–152. 15 Published by Dotan, 1997. 16 Dotan, 1997, vol. II, p. 466. See also Dotan’s analysis in vol. I, pp. 124–125. 17 See, e.g., the description in Gesenius’ Hebrew Grammar (Kautzsch, 1910, §8a–c, §10a–f). 18 See Yeivin, 2011, p. 211, §367. It is most probable that by the time of the addition of the gaʿya, the furtive vowels did not exist any more as such, i.e., the shewa and the ḥaṭef-pataḥ were pronounced as a regular pataḥ, the ḥaṭef-segol as a segol, and the ḥaṭef-qameṣ as a qameṣ; see Khan, 2013b, pp. 98–107. 100 of the gaʿya – both in matters of principle and regarding specific words.19 It should be emphasised that these inconsistencies exist not only between manuscripts, but also within manuscripts.20 The lack of uniformity regarding the insertion of the gaʿya is especially problematic in the case of the shewa in the first of two identical letters, where the length of the preceding vowel determines the pronunciation of the shewa. For example, in the Aleppo Codex we find the following forms with ḥolem before two consecutive identical letters, all of them in Isaiah: (1) ‫‘ סֹורֲ ִ ַ֗רים‬rebels’ (Isa. 1:23) (ḥolem without gaʿya, followed by a ḥaṭef-pataḥ) (2) ֶׁ֙‫‘ עָּֽׁ ֹורֲ רו‬they have laid bare’ (Isa. 23:13) (ḥolem with gaʿya, followed by a ḥaṭef-pataḥ) (3) ‫‘ ש ְמ ָ֥מֹות‬devastations’ (Isa. 61:4) (ḥolem without gaʿya, followed by a simple shewa) (4) ‫ֹוממ ָ֔ ֶתָך‬ ְ ‫‘ מֵ ָ֣ר‬when you lift yourself up’ (Isa. 33:3) (ḥolem with secondary cantillation mark – i.e. equivalent to gaʿya – followed by a simple shewa) Despite the notational differences in these four cases, we must assume that for our purposes all four forms exhibit one and the same pronunciation, long vowel followed by a vocal shewa.21 The situation becomes even more complicated in cases of other vowels, as can be seen in (5), for example. (5) ‫‘ ויתהללו‬and they will glory’ (Isa. 45:25), which is vocalised to: ‫ וְ ִ ָּֽׁיתהַ לְ לֵ֖ ו‬in Codex Aleppo ‫ וְ יִ ְת ַ ָּֽׁהלְ לֵ֖ ו‬in Codex New York (JTS 232 / ENA 346, 10th century) ‫ וְ יִ ְתהַ ֲללֵ֖ ו‬in Codex Cairo to the Prophets (11th century)22 In these examples we must assume that there were differences across the manuscripts which reflect differences in the actual pronunciation of the shewa: in Codex Aleppo, a silent shewa, and in the two other codices, a vocal shewa. We shall see additional examples of these inner Tiberian differences below. The best source on those differences is Kitāb al-Ḵilaf by Mishael Ben Uziel (Lipschütz, 1965). For differences on positioning of the gaʿyot in later manuscripts, see Cohen, 1982. 20 See the extensive material collected by Yeivin, 1968, chapters 11–20. 21 Since the ḥolem should be regarded as a long vowel almost by definition; see Dotan, 1967, pp. 256–257. 22 The variae lectiones are taken from the fourth apparatus of the Hebrew University Bible (Goshen-Gottstein, 1993, p. 210). 19 101 4 The compound vocalisation system and Codex Babylonicus Petropolitanus The Fikovich collection in St Petersburg contains a codex of the Later Prophets (shelf mark Heb. B3), which is vocalised with the so-called ‘Compound Babylonian vocalisation system’ and dates to 916 CE.23 Despite its vocalic inventory – which does not distinguish between pataḥ and segol – and some other Babylonian features,24 the general pronunciation tradition that is reflected in this manuscript is Tiberian.25 In fact, most Biblical manuscripts with Compound Babylonian vocalisation reflect, to some degree, Tiberian pronunciation.26 This is the result of a long process that saw the Jewish Babylonian communities adopting the Tiberian pronunciation tradition of the Bible, which they came to perceive as the only ‘correct’ pronunciation tradition of the text.27 The main characteristic of the Compound Babylonian vocalisation system is its three different ‘sets’ of vocalisation signs – in total sixteen signs, seen in table 1. Table 1. The Compound Babylonian vocalisation signs Set 1: Signs in an open, or a closed and stressed, syllable Set 2: Signs in an unstressed syllable closed by shewa Set 3: Signs in an unstressed syllable closed by dagesh pataḥsegol qameṣ ṣere ḥireq qibbuṣ / shureq ḥolem ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ – ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ ‫ב‬ ֒ – In addition, there is a seventeenth sign which denotes the shewa, both vocal and silent: ‫ב‬. ֒ We also find the Tiberian signs for dagesh and rafe, and the diacritical dots on the left and right arms of the shin. There is, however, no sign for a gaʿya in the compound vocalisation. To demonstrate the system, here are a few examples: ‫מי֒ ֒ד ֒כם‬ ֒ ‘from your hand’ (Isa. 1:12, Tiberian: ‫)מיֶדְ כֶם‬, ִ 23 The codex has been reprinted in facsimile; see Strack, 1876. E.g. the lack of furtive pataḥ. The main Babylonian features of this manuscript are listed in Yeivin, 1985, vol. I, p. 185. 25 The manuscript is categorised as ‘type II’ by Yeivin, meaning that “in principle, the manuscript reflects the Tiberian pronunciation tradition, with few Babylonian grammatical features” (Yeivin, 1985, vol. I, p. 91). 26 Out of several dozens of manuscripts that are vocalised with the compound system, Yeivin lists only one manuscript of ‘type V’ (i.e. old Babylonian pronunciation) – namely MS Kc 53 (p. 193) – and two manuscripts of ‘type IV’ (i.e. old-middle Babylonian pronunciation) – MS Ka 6 (p. 174) and MS Ka 53 (p. 179). 27 See Eldar, 1985, especially at pp. 229–231. 24 102 ‫הגּ֒בו֒ר‬ ֒ ‘the mighty man’ (Jer. 9:22, Tiberian: ‫)הַ גִ בֹור‬, ‫ש ֒ר ֒ך‬ ֒ ‘your cord’ (Ezek. 16:5, Tiberian ‫)שָ ֵרְך‬, ‫ס ֒רם‬ ֒‫א‬ ֒ ‫ּב‬ ֒ ‘when they are bound up’ (Hos. 10:10, Tiberian ‫)בְ אָ ְס ָרם‬. The interpretation of the compound signs as denoting vowels in syllables closed by shewa and dagesh, as shown in table 1, was first put forward by Luzzatto in 1846, followed by Rödiger in 1848 and by Pinsker in 1863.28 However, approximately a century later, in 1958, Bendavid argued that the compound signs denoted different vowel length,29 and his opinion was adopted by Yeivin in 1982, who explicitly rejected the ‘old’ opinion In forms like ‫שמר‬ ֒ ‫ת‬ ֒ , ‫קריב‬ ֒ ‫ת‬ ֒ , ‫לחן‬ ֒ ‫ש‬ ֒ , the signs in the first syllables ... do not denote unstressed syllables closed by shewa (as was previously assumed), but short vowels, while the non-compound signs are being used to denote long vowels or short vowels when stressed.30 This opinion was also adopted by Khan.31 We now return to the main subject of this article – the shewa. Since the compound system reflects vowel length,32 and since it represents, on the whole, the Tiberian pronunciation tradition, it is clear that the Codex Babylonicus Petropolitanus represents an independent source for the pronunciation of the shewa in the first of two consecutive letters in Tiberian.33 Here follows the material that I collected from the first three books in Codex Babylonicus Petropolitanus (hereinafter ‘Petropolitanus’) – Isaiah, Jeremiah and Ezekiel – compared with the two main Tiberian codices: Aleppo (‘A’) and Leningrad (‘L’).34 All words which have pataḥ/segol or ḥireq before a shewa in the first of two identical letters were collected. As will be seen, in most cases the length of the vowel is the same in Petropolitanus and in the Tiberian codices. There are cases, however, where Petropolitanus differs from the Tiberian codices, or where the Tiberian codices do not agree among themselves. 28 See Samuel David Luzzatto’s letter as printed in Polak, 1846, pp. 25–31; Rödiger, 1848; Pinsker, 1863, Hebrew part, p. 12. See also Ewald, 1849, pp. 160–172, who has a similar, but slightly different, description for the signs. 29 Bendavid, 1958, p. 16. 30 Yeivin, 1982, p. 43. In his monumental book on Babylonian vocalisation that was published in 1985 (but written several years earlier), Yeivin is less unequivocal, often using terms like ‫‘ חירק סגור שווא‬ḥireq closed by shewa’ for the sign ‫ ֜ב‬alongside ‫‘ חירק קצר‬short ḥireq’. 31 See e.g. Khan, 2013c. 32 It seems to me there can be little doubt that the system, in its initial stages, served to denote syllable structure, as set out in table 1, and only later – due to the fact that syllable structure and vowel length were largely overlapping – started to denote vowel length. It is not impossible that some of the Genizah fragments which are vocalised with the compound system still denote syllable structure. 33 Khan has used this feature of Codex Babylonicus Petropolitanus to demonstrate the long vowels in certain forms of the verbs haya and ḥaya; see Khan, 1994. 34 The material for A was collected from Cohen, 1996; Cohen, 2012; and Cohen, 2013. The material for L was collected from Dotan, 2001. 103 4.1 Long pataḥ In the following cases we find long pataḥ both in Petropolitanus and in the Tiberian codices A and L (one or both; occasionally A and L do not agree). Note that except for the form ‫חללי‬, all forms with long pataḥ are ones in which a dagesh has fallen away in the first of the two letters; the long vowel preserves the vocal shewa after this letter.35 (6) ֒‫‘ יהללך‬it will praise you’ (Isa. 38:18), A ‫יְ הַ ֲללֶ ֶ֑ ָך‬, but L ‫יְ הַ לְ לֶ ֶ֑ ָך‬ (7) ‫פ ֒ללי֒ם‬ ֒‫ת‬ ֒‫מ‬ ֒ ֒‫‘ ו‬and they pray’ (Isa. 45:20), A ‫ומ ְת ַפ ֲללִָ֔ ים‬ ָּֽׁ ִ , but L ‫ומ ְתפַלְ לִָ֔ ים‬ ִ (8) ‫שֿה‬ ֒ ‫ש‬ ֒ ‫‘ נ֒ ֒ג‬we shall grope’ (Isa. 59:10), A ‫נְ גַשֲ ָ ִׂ֤שה‬, L ‫נ ַ ְָּֽׁג ְש ָ ִׂ֤שה‬ (9) ֒‫ח ֒ללי‬ ֒ ‘the slain by [the Lord]’ (Isa. 66:16), A ‫ ַ ָּֽׁח ֲללֵ ָ֥י‬, L ‫ַ ָּֽׁחלְ לֵ ָ֥י‬ (10) ֒‫ח ֒ללי‬ ֒ ‘the slain of [the daughter of my people]’ (Jer. 8:23), A ‫חלֲלֵ ָ֥י‬, ָּֽׁ ַ L ‫ַ ָּֽׁחלְ לֵ ָ֥י‬ (11) ‫ה ֒ללּו‬ ֒ ‘praise!’ (Jer. 20:13), A and L ‫ַ ָּֽׁהלְ לֵ֖ ו‬ (12) ֒‫ח ֒ללי‬ ֒ ‘the slain by [the Lord]’ (Jer. 25:33), A ‫חלֲלֵ ִׂ֤י‬, ָּֽׁ ַ L ‫ַ ָּֽׁחלְ לֵ ִׂ֤י‬ (13) ‫פ ֒ללּו‬ ֒‫ת‬ ֒‫ה‬ ֒ ֒‫‘ ו‬and pray!’ (Jer. 29:7), A and L ‫וְ הִ ְת ַ ָּֽׁפלְ לָ֥ ו‬ (14) ֒‫ה ֒ללי‬ ֒‫ת‬ ֒‫ת‬ ֒ ‘you will boast’ (Jer. 49:4), A ֶׁ֙‫ת ְת ַ ָּֽׁהלֲלִ י‬, ִ L ֶׁ֙‫ִת ְת ַ ָּֽׁהלְ לִ י‬ (15) ֒‫ח ֒ללי‬ ֒ ‘the slain of [Israel]’ (Jer. 51:49), A and L ‫ַ ָּֽׁחלְ לֵ ָ֣י‬ (16) ֒‫ח ֒ללי‬ ֒ ‘the slain of [all the earth]’ (Jer. 51:49), A ‫חלֲלֵ ָ֥י‬, ָּֽׁ ַ L ‫ַ ָּֽׁחלְ לֵ ָ֥י‬ In the following two cases, however, we find long pataḥ in Petropolitanus, but short pataḥ in the two main Tiberian codices: (17) ֒‫ח ֒ללי‬ ֒ ‘the slain of [the sword]’ (Isa. 22:2), A and L ‫חַ לְ לֵי־‬ (18) ‫ה ֒ללּו‬ ֒‫ת‬ ֒ ‫‘ ו֒֒י‬and they will glory’ (Isa. 45:25), A and L ‫וְ ִ ָּֽׁי ְתהַ לְ לֵ֖ ו‬ In example (17), Petropolitanus has an independent primary stress in the word ‫( חללי‬viz. ‫ ָ֣חללי ָ֔חרב‬, in the equivalent Tiberian signs), while in A and L the word 35 Note that in examples (6), (19)–(23) and (32), the Babylonian shewa sign is omitted. In all these cases the shewa comes after a lamed, and the scribe of the codex probably wanted to avoid writing the shewa across the neck of the lamed. Writing it to the left or right of the neck is equally problematic, as it would then have made it look like a rafe sign. 104 ‫ חללי‬is attached to the next by maqqef (‫חללי־חרב‬ ָ֔ ). This can explain the ‘shortened’ pronunciation in A and L. As for example (18), there seems to be some inconsistency within good Tiberian manuscripts – as already mentioned above in example (5). 4.2 Short pataḥ In all cases of the words ‫‘ מַ עַלְ לֵיכֶם‬your deeds’ and ‫‘ מַ ַעלְ לֵיהֶ ם‬their deeds’,36 we find short pataḥ in Petropolitanus as well as in the Tiberian codices A and L.37 Here are a few examples: (19) ‫הם‬ ֒ ֒‫עללי‬ ֒‫מ‬ ֒ ֒‫( ו‬Isa. 3:8), A and L ֶׁ֙‫ומעַלְ לֵיהֶ ם‬ ָּֽׁ ַ (20) ‫עללי֒כ֒ם‬ ֒‫מ‬ ֒ (Jer. 4:4), A and L ‫ֵיכם׃‬ ָּֽׁ ֶ ‫מַ עַלְ ל‬ (21) ‫עללי֒כ֒ם‬ ֒‫מ‬ ֒ , in margin: ‫( הם כת׳‬Jer. 21:12), A and L ‫ מעלליהם‬Ketib, ‫ֵיכם׃‬ ָּֽׁ ֶ ‫ מַ עַלְ ל‬Qere The same is true for ‫‘ חַ לְ לֵיכֶם‬your slain’ and ‫‘ חַ לְ לֵיהֶ ם‬their slain’, as seen, for example, in (22) and (23).38 (22) ‫הם‬ ֒ ֒‫חללי‬ ֒ ֒‫( ו‬Isa. 34:3), A and L ‫ֵיהָ֣ם‬ ֶ ‫וְ חַ לְ ל‬ (23) ‫חללי֒כ֒ם‬ ֒ (Ezek. 6:4), A and L ‫חַ לְ לֵי ֶָ֔כם‬ In all the cases (19)–(23) we can safely assume that in both traditions there was no vowel between the two identical letters. There are, however, four cases where we find short pataḥ in Petropolitanus, but long pataḥ in either A or L: (24) :֒‫ק ֒ל ֒לוני‬ ֒‫מ‬ ֒ ‘they curse me’ (Jer. 15:10), A ‫מ ַ ָּֽׁקלְ ַ ָּֽׁלוְ נִ י׃‬, ְ L ‫ְמקַ לְ ַ ָּֽׁלוְ נִ י׃‬ (25) :֒‫ח ֒ל ֒ליך‬ ֒‫מ‬ ֒ ‘those who slay you’ (Ezek. 28:9), A ‫מחַ ל ֶ ֲָּֽׁליָך׃‬, ְ L ‫ְמחַ לְ ֶ ָּֽׁליָך׃‬ (26) ֒‫ח ֒ללי‬ ֒ ‘the slain of’ (Ezek. 21:34), A and L ‫ַ ָּֽׁחלְ לֵ ָ֣י‬ (27) ‫מ ֒לאי֒ם‬ ֒‫מ‬ ֒‫ה‬ ֒ ֒‫‘ ו‬and those who fill’ (Isa. 65:11), A ‫וְ ַ ָּֽׁה ֲממַ לְ ִ ֵ֖אים‬, L ‫וְ ַ ָּֽׁה ְממַ לְ ִ ֵ֖אים‬ In cases (24) and (25) the Babylonian tradition, unlike the Tiberian, has a dagesh in the lamed (as expected in a piel form), which in turn necessitates a In the books Isaiah, Jeremiah and Ezekiel we find thirteen occurrences of ‫ מעלליכם‬and four occurrences of ‫מעלליהם‬. 37 At least, there is no indication for a long pataḥ. 38 There are three occurrences of ‫ חלליכם‬in Ezekiel, and two occurrences of ‫ חלליהם‬in Isaiah and Ezekiel, one in each book. 36 105 short pataḥ before it. The Tiberian form without dagesh has a long pataḥ, which may indicate a kind of compensation for the lack of this expected dagesh. Case (26) appears in Petropolitanus as correction from ‫חללים‬, with the final mem crossed out. It is not impossible that the short pataḥ in Babylonian represents the Tiberian ḥaṭef-pataḥ under the ḥet in ‫ ֲחלָלִ ים‬, as is usual in Late Babylonian, and that this was not corrected when the consonantal form was corrected. 4.3 Long ḥireq I found one case of long ḥireq in Petropolitanus; A and L also have a long ḥireq in this case: (28) ‫ה ֒ללּו‬ ֒ ֒‫‘ ו‬and they will praise’ (Isa. 62:9), A ‫וְ ִ ָּֽׁה ֲללֵ֖ ו‬, L ‫וְ ִ ָּֽׁהלְ לֵ֖ ו‬ 4.4 Short ḥireq We find short ḥireq both in Petropolitanus and in the Tiberian codices A and L in the following forms: (29) ֒‫הנ֒ני‬ ֒ ‘here I am’ (Isa. 6:8), A and L ‫( הִ נְ ִנָ֥י‬frequent in all three books) (30) ֒‫קקי‬ ֒‫ח‬ ֒ ‘decrees of’ (Isa. 10:1), A and L ‫חִ קְ קֵ י‬ (31) ‫‘ ֒ילל֒תּ֒ה‬her wailing’ (Isa. 15:8), A and L ‫יִ לְ ל ָ֔ ָָתּה‬ (32) ֒‫‘ ֒צללי‬shadows of’ (Jer. 6:4), A and L ‫צִ לְ לֵי־‬ (33) ‫ממו֒ת‬ ֒‫ש‬ ֒ ‘waste of’ (Jer. 51:26), A and L ‫ִש ְמ ָ֥מֹות‬ There are, however, six cases where we find short ḥireq in Petropolitanus, but long ḥireq in either A or L (usually both): (34) ֒‫ה ֒ללּוך‬ ֒ ‘they praised you’ (Isa. 64:10), A ֶׁ֙‫ ִ ָּֽׁה ֲל ָּ֙לוָך‬, L ֶׁ֙‫ִ ָּֽׁהלְ ָּ֙לוָך‬ (35) ‫ממו֒ת‬ ֒‫ש‬ ֒ ‫‘ ֒ל‬to waste of’ (Jer. 25:12), A and L ‫לְ ִ ָּֽׁש ְמ ָ֥מֹות‬ (36) ‫‘ ו֒֒י ֒לל֒ת‬and the wail of’ (Jer. 25:36), A ‫וילֲלַ ֵ֖ת‬,ָּֽׁ ִ L ‫ִ ָּֽׁוילְ לַ ֵ֖ת‬ (37) ‫קנ֒נּו‬ ֒ ‘they made their nests’ (Ezek. 31:6), A ֶׁ֙‫קנֲנו‬,ָּֽׁ ִ L ֶׁ֙‫קִ נְנו‬ (38) ‫מֿה‬ ֒‫מ‬ ֒‫ש‬ ֒ ‫‘ ֒ל‬to waste’ (Ezek. 35:7), A ‫לְ ִ ָּֽׁשמ ָ ֲֵ֖מה‬, L ‫לְ ִ ָּֽׁש ְמ ָמֵ֖ה‬ (39) ‫ממו֒ת‬ ֒‫ש‬ ֒ ‘waste of’ (Ezek. 35:9), A ‫שמ ֲִׂ֤מֹות‬, ָּֽׁ ִ L ‫ִ ָּֽׁש ְמ ִׂ֤מֹות‬ 106 Example (34) has short ḥireq in Babylonian and long ḥireq in Tiberian because of the lack of dagesh in the latter; compare the note to examples (24) and (25) above. The short ḥireq in example (36) can be explained by the fact that the Tiberian rule wǝyi- > wī- did not operate in the Babylonian form. 4.5 Note on the form ‫ ִרבְ בֹות‬in Michah 6:7 An interesting case, although outside the three books which were checked for this study, is the form ‫‘ ִרבְ בֹות‬ten thousands of’ in Michah 6:7. This form is mentioned in the Anonymous Treatise about the Shewa as being exceptional, in that it has a long ḥireq followed by a silent shewa.39 It is interesting that in the Aleppo Codex the vocalisation explicitly contradicts this statement, as we find there the form vocalised as ‫( בְ ִ ָּֽׁרב ֲֵ֖בות‬in L we find the ambiguous form ‫)בְ ִ ָּֽׁרבְ ֵ֖בות‬. However, in Petropolitanus we find ‫ּב ֒רבֿבו֒ת‬ ֒ , with a long ḥireq but without the expected shewa sign above the first of the two consecutive bet letters. This exceptional vocalisation doesn’t seem to be coincidental, as the vocaliser of Petropolitanus invariably inserts the shewa signs above bet letters. The lack of shewa on this specific bet seems to express the pronunciation of the shewa as silent, in accordance with the statement of the Treatise about the Shewa. 4.6 Segol Finally, two forms should be mentioned which have a segol in Tiberian before two identical letters (this segol is represented in Babylonian, as expected, by pataḥ): (40) ֒‫ּב ֒ג ֒ללי‬ ֒ ‘in the dung of’ (Ezek. 4:12), A ֶׁ֙‫בְ ֶ ָּֽׁג ֲללֵי‬, L ֶׁ֙‫בְ ֶגָּֽׁלְ לֵי‬ (41) ֒‫‘ ֒ג ֒ללי‬the dung of’ (Ezek. 4:15), A ‫ ֶגלֲלֵ ָ֣י‬, L ‫גֶלְ לֵ ָ֣י‬ In the first case, we see a long vowel in both Petropolitanus and the Tiberian codices. In the second case, the vocalisation of the Aleppo Codex seems to reflect a long segol while the vocalisation of the Leningrad Codex seems to reflect a short segol. This inconsistency is reflected also in other excellent Tiberian manuscripts.40 Petropolitanus, like A (and unlike L), has a long pataḥ. It is not impossible, therefore, that this is an example of different schools inside the Tiberian pronunciation tradition. 39 See n. 9 above. The fourth apparatus of the Hebrew University Bible records twelve witnesses for ‫ גֶלְ לי‬and five for ‫( ֶגָּֽׁלְ לי‬Goshen-Gottstein and Talmon, 2004, p.‫)טו‬. 40 107 5 Summary This article has sought to demonstrate how Codex Babylonicus Petropolitanus can be used as an early, additional, source for our knowledge about the Tiberian pronunciation tradition, especially with regard to the pronunciation of the shewa under the first of two identical letters. Interpreting the Tiberian pronunciation tradition relies on reliable witnesses; the more reliable codices we have, the better our understanding of the system. Despite its Babylonian vocalisation, the pronunciation tradition that is reflected in Codex Babylonicus Petropolitanus is essentially Tiberian, and its Compound Babylonian vocalisation system has the advantage over the Tiberian system in having different signs for long and short variants of, among others, pataḥ and ḥireq. These, in turn, determine the nature of the shewa (vocal or silent) in cases of two consecutive identical letters. As a result, the nature of the shewa is much clearer in Petropolitanus than it is in the Tiberian codices. The analysis has shown that in the majority of cases there is a uniform tradition common to Petropolitanus and the best Tiberian codices regarding the length of pataḥ and ḥireq. However, there are cases where the Tiberian codices don’t agree among themselves, and Petropolitanus serves as an important witness for one group of manuscripts against the other. There are also cases where Petropolitanus differs from the Tiberian codices for other reasons, such as a slightly different pronunciation of other segments in the word; these have also been discussed. References Akun, N., 2010, The Hebrew Language of the Jews of Morocco: Phonetics (Communities of Tafilalet, Marrakesh and Maknès) (Hebrew, unpubl. diss., Hebrew University of Jerusalem). Artom, E. S., 1947, “Pronunciation of Hebrew among Italian Jews” (Hebrew), Leshonenu 15, pp. 52–61. Bendavid, A., 1958, “Minayin ha-ḥaluqa li-tenuot gedolot u-qeṭanot?”, Leshonenu 22, pp. 7–35, 110–136. Cohen, M., 1982, “Systems of light gaʿyot in medieval Biblical manuscripts and their importance for the history of the Tiberian systems of notation”, Textus 10, pp. 44–83. ——— (ed.), 1996, Mikraʾot Gedolot ‘Haketer’: Isaiah, Ramat Gan: Bar-Ilan University Press. ——— (ed.), 2012, Mikraʾot Gedolot ‘Haketer’: Jeremiah, Ramat Gan: Bar-Ilan University Press. ——— (ed.), 2013, Mikraʾot Gedolot ‘Haketer’: Ezekiel, Ramat Gan: Bar-Ilan University Press. Dotan, A., 1967, The Diqduqé Haṭṭĕʿamim of Ahăron Ben Moše Ben Ašér: With a Critical Edition of the Original Text from New Manuscripts (Hebrew), Jerusalem: The Academy of the Hebrew Language. 108 ——— (ed.), 1997, The Dawn of Hebrew Linguistics: The Book of Elegance of the Language of the Hebrews by Saadia Gaon (2 vols), Jerusalem: World Union of Jewish Studies. ———, 2001, Biblia Hebraica Leningradensia, Leiden: Brill. Eldar, I., 1985, “Torat ha-qeri’a ba-mikra”, Tarbiẓ 54, pp. 225–243. ———, 1987, “Mukhtaṣar Hidāyat al-Qāri”, Leshonenu 51, pp. 3–41. ———, 1994, The Art of Correct Reading of the Bible (Hebrew), Jerusalem: The Academy of the Hebrew Language. Ewald, H., 1849, “Die assyrisch-hebräische Punctation”, Jahrbücher der biblischen Wissenschaft 1, pp. 160–172. Goshen-Gottstein, M. H. (ed.), 1993, The Hebrew University Bible: The Book of Isaiah, Jersualem: Magnes Press. ——— and S. Talmon (eds), 2004, The Hebrew University Bible: The Book of Ezekiel, Jersualem: Magnes Press. Joüon, P. and Muraoka, T., 2006, A Grammar of Biblical Hebrew, Rome: Editrice Ponteficio Istituto Biblico. Katz, K., 1977, The Hebrew Language Tradition of the Community of Djerba (Tunisia): The Phonology and the Morphology of the Verb (Hebrew), Jerusalem: The Language Tradition Project. ———, 1981, The Hebrew Language Tradition of the Aleppo Community: The Phonology (Hebrew), Jerusalem: The Language Tradition Project. Kautzsch, E. (ed.), 1910, Gesenius’ Hebrew Grammar, 2nd English edn, A. E. Cowley (trans.), Oxford: Clarendon Press. Khan, G., 1994, “The pronunciation of the verbs haya and ḥaya in the Tiberian tradition of Biblical Hebrew”, in G. Goldenberg and S. Raz (eds), Semitic and Cushitic Studies, Wiesbaden: Harrassowitz, pp. 133–144. ———, 2013a, “Shewa: pre-modern Hebrew”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. III, pp. 543–554. ———, 2013b, A Short Introduction to the Tiberian Masoretic Bible and its Reading Tradition, 2nd edn, Piscataway: Georgias Press. ———, 2013c, “Vocalization, Babylonian”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. III, pp 953–963. Levy, K., 1936, Zur masoretischen Grammatik: Texte und Untersuchungen, Stuttgart: Kohlhammer. Lipschütz, L. (ed.), 1965, Kitāb al-Khilaf, Jerusalem: Magnes Press. Maman, A., 1994, “The reading tradition of the Jews of Tetouan: phonology of Biblical and Mishnaic Hebrew” (Hebrew), Massorot 1, pp. 51–120. Pinsker, S., 1863, Einleitung in das Babylonisch-Hebräische Punktationssystem, Vienna: Philipp Bendiner. Polak, G., 1846, Oostersche Wandelingen, Amsterdam: Proops. Rödiger, E., 1848, “Hebräische Handschriften mit eigenthümlicher Vocalbezeichnungen”, Allgemeine Literatur-Zeitung 169, cols 194–199. Strack, H. (ed.), 1876, Prophetarum Posteriorum Codex Babylonicus Petropolitanus, St Petersburg: Ricker. Yalon, H. (ed.), 1945, Rabbi Shelomo Almoli: Halikhot Shewa, Jerusalem: Mosad Harav Kuk. ———, 1963, “Gilgulei kelalei ha-shewa she-ba-otiyot ha-domot”, Leshonenu 27– 28, pp. 84–95. Yeivin, I., 1968, The Aleppo Codex of the Bible: A Study of its Vocalisation and Accentuation (Hebrew), Jerusalem: Magness Press. ———, 1982, “Ha-gaʿyot ve-tafkidan”, Leshonenu 46, pp. 39–56. 109 ———, 1985, The Hebrew Language Tradition as Reflected in the Babylonian Vocalization, Jerusalem: The Academy of the Hebrew Language. ———, 2011, The Biblical Masorah (Hebrew), Jerusalem: The Academy of the Hebrew Language. 110 ‫הֶ חָ כָם‬, but ‫הַ חָ כְ מָ ה‬: Some Notes on the Vocalisation of the Definite Article in Tiberian Hebrew DANIEL BIRNSTIEL Goethe Universität Frankfurt am Main 1 Introduction In the Tiberian reading tradition of Biblical Hebrew, the definite article consists of a prefixed - ַ‫ ה‬ha- followed by the gemination of the initial consonant, e.g. ‫ הַ מֶ לֶך‬/hammɛ́lɛk/ [ham.ˈmɛː.lɛχ] ‘the king’.1 There are two exceptions to this rule: firstly, in certain words with a word-initial consonant cluster marked by a shewa under the first consonant, the initial consonant is not geminated; and secondly, neither the guttural consonants nor /r/ can be geminated,2 and this then influences the vocalisation of the article. Although the sound laws involved are well known, to my knowledge no detailed description of the different historical stages has so far been given.3 This article offers a comprehensive account of the development of the vocalisation of the definite article before gutturals4 in Biblical Hebrew from the end of the Biblical period to the era of the Masoretes. 2 Prehistory of the definite article in Biblical Hebrew In Semitic languages, no definite article can be reconstructed for the protolanguage. Where they are attested, definite articles are independent innovations of the respective languages. They occur in modern Ethio-Semitic languages, especially Southern Neo-Ethio-Semitic, Modern South Arabian and several daughter languages of Common Semitic, namely North-West Semitic (only Aramaic and Canaanite), Ancient South Arabian and Arabic. In addition, several varieties of Neo-Aramaic have developed new articles after the 1 Unless otherwise indicated, all examples are based on the Westminster Leningrad Codex, available online at https://tanach.us/Tanach.xml. 2 There are some exceptional cases where geminated /r/ is attested in the Masoretic text; cf. Khan, 2013f, p. 386. 3 Blau, 1980 leaves a number of important cases unmentioned. 4 Unless stated otherwise, the term gutturals will be used to refer to the laryngeal and pharyngeal consonants and also /r/. 111 definite (emphatic) form of nouns (and adjectives) lost its determinative function, at least in Eastern Aramaic. In Common Semitic, definite articles become attested after the beginning of the first millennium BCE. Neither Amarna Canaanite (14th century BCE) nor Ugaritic (14th–12th century BCE) possess a definite article, although Tropper points to several examples of the Ugaritic presentative particle hn being used in an almost article like function.5 In Canaanite, a prefixed article h- is attested from the early first millennium BCE in Phoenician, and thereafter also in the other Canaanite dialects (Hebrew, Moabite,6 Ammonite and Edomite).7 In Aramaic, a suffixed definite article -ʾ is attested after the middle of the 9th century BCE, occurring in the bilingual inscription from Tell Fekheriye. Neither Sam’alian (9th–8th century BCE) nor the dialect of Deir ʿAllā (c800 BCE) possess a definite article.8 In Ancient South Arabian, suffixed definite articles (-n with singular, -hn with dual and plural nouns) appear from the beginning of the epigraphic record, around 1000 BCE.9 The picture presented by Arabic, where a prefixed definite article (ʾ)al- is attested in Classical and modern varieties, is more complicated; this is due to the previously unclear relationship between Arabic and the Ancient North Arabian languages. Recently, two of these dialects, Safaitic and Hismaic, have been identified as Arabic.10 Of these, Hismaic possesses no article, while Safaitic attests a plethora of forms (ʾl-, ʾ-, h-, hn-, Ø-). This means that varieties of Arabic also developed a definite article towards end of the first millennium BCE at the latest. Unlike the definite articles in Indo-European languages, which usually derive from demonstrative pronouns and generally form paradigms inflecting for gender, number and even case, the Common Semitic articles developed from presentative particles.11 Recent research has shown that the presentative particles from which the articles developed originally attached to attributive or nominalised adjectives, participles and demonstratives; only later did agreement rules cause it to be placed on both the head noun and its satellite in attributive constructions.12 The similarity in function and syntactic development make it likely that the parallel innovation of the definite article is an areal feature.13 5 Tropper, 2000, pp. 24–25. It already occurs in the Mesha Stele, dated to approximately 840 BCE; see Gzella, 2006, p. 15. 7 For examples from the epigraphic records, see Garr, 2004, pp. 87–89. 8 Tropper, 2000, pp. 21–22. 9 Stein, 2011, p. 1024. 10 Al-Jallad, forthcoming. 11 See Tropper, 2000, pp. 26–27 and Pat-El, 2009, pp. 40–42. Rubin, 2005, pp. 72–76 argues for a demonstrative origin. However, as shown by Pat-El, 2009, pp. 37–38, the elements in question do not show any similarity with demonstrative pronouns, lacking for example gender, number and case inflection. 12 Pat-El, 2009, pp. 42–46. 13 Huehnergard, 2005, p. 186. 6 112 3 The form of the definite article in Biblical Hebrew For Canaanite, the definite article may be reconstructed as *han- with subsequent assimilation of the n to the following consonant.14 As noted in section 1, in the Tiberian vocalisation tradition of Biblical Hebrew, the definite article usually consists of a prefixed - ַ‫ ה‬ha- followed by the gemination of the initial consonant. At some point, guttural consonants (including /r/) ceased to be geminated, and this degemination of the initial guttural resulted partially in quantitative changes in the vowel of the article, partially in qualitative changes in the vowel, and partially in no change at all. The exact distribution of the resulting variants of the article depends on several factors: the guttural in question, the quality of the following vowel and also the presence or absence of stress. In addition, in words which begin with an initial consonant cluster, where the first consonant is marked by a shewa in the Tiberian vocalisation, the initial consonant often does not undergo gemination – e.g. ‫ הַ ְמ ַרגלִ ים‬/hamragglīm/ [ham.ʀag.gaˈliːim] ‘the spies’, ‫ הַ ְילָדִ ים‬/haylādīm/ [hajlɔːˈðiːim]. Moreover, the initial glottal fricative of the article (h-) is syncopated when the article follows one of the three proclitic prepositions ‫ ב‬b- ‘in’, ‫ ל‬l- ‘to’ and ‫ כ‬k- ‘like’. This has no influence on its vocalisation. Regarding the gutturals, the following distribution rules were formulated by Gesenius. Before /ʾ/ and /r/, the pataḥ of the article is lengthened to a qameṣ. Before /ʿ/, the vowel is lengthened when the /ʿ/ is followed by a stressed qameṣ, but changes to a seghol when the /ʿ/ is followed by an unstressed qameṣ; in all other cases, including where the /ʿ/ is followed by a ḥatef qameṣ, the vowel is lengthened. Before /h/, the vowel is also lengthened when the /h/ is followed by a stressed qameṣ, but changes to a seghol when the /h/ is followed by an unstressed qameṣ; in all other cases, the article remains unchanged. Before /ḥ/, the vowel is changed to a seghol whenever /ḥ/ itself is followed by a qameṣ, whether accented or not, or by a ḥatef qameṣ, but remains unchanged in all other cases.15 Table 1 presents an overview of the distribution. A number of sporadically attested exceptions to these rules are mentioned by Gesenius as well as by Joüon and Muraoka.16 The lengthening of the vowel has been regarded as a case of compensatory lengthening. The cases where pataḥ remains unchanged are referred to by Gesenius as virtual strengthening;17 this phenomenon has also been called virtual gemination.18 There is, however, some difference regarding the application of these terms: while Gesenius distinguishes between virtual 14 Outside of Hebrew, evidence for gemination exists in Late Punic, where the geminated consonant may be written twice, e.g. ‫[ עממקם‬ammaqōm] ‘the place’; see Garr, 2004, p. 87. 15 Gesenius, 1910, pp. 110–112. 16 See Gesenius, 1910, p. 111; Joüon and Muraoka, 1991, p. 114. 17 Gesenius, 1910, pp. 76–77, 111. 18 Joüon and Muraoka, 1991, pp. 87–88, 113–114. 113 gemination following an unchanged vowel on the one hand and a modification of the vowel by lengthening or a change in quality on the other, others apparently regard a change in vowel quality as a case of virtual gemination.19 Table 1. Distribution of article variants before gutturals according to Gesenius Vowel following the guttural Guttural Stressed qameṣ Unstressed qameṣ ‫א‬ ‫ר‬ ‫ע‬ ‫ה‬ ‫ח‬ 4 Ḥaṭef qameṣ Other ָ‫ה‬ ?20 ָ‫ה‬ ַ‫ה‬ ָ‫ה‬ ָ‫ה‬ ָ‫ה‬ ֶ‫ה‬ ֶ‫ה‬ ֶ‫ה‬ ַ‫ה‬ The Tiberian vowel system The Tiberian vowel system21 distinguishes between vowels that are inherently long (/ɔ̄/ qameṣ, /ō/ ḥolem, /ē/ ṣere, /ū/ long shureq, /ī/ long ḥireq) and vowels of unspecified length (/a/ pataḥ, /ɛ/ seghol, /i/ ḥireq, /u/ qibbuṣ/shureq, /o/, /e/). Since all vowels in stressed or open syllables are long, the former can only occur in these two environments, whereas the latter are realised as long vowels in stressed or open syllables, but realised as short vowels in unstressed closed syllables. Most of these phonemes are represented by a single vowel sign. This means that means most vowels of unspecified length when occurring in a stressed or open syllable are phonetically indistinguishable from the corresponding long vowel phonemes, yet are represented by different vowel signs. However, two vowels of unspecified length, /e/ and /o/, are represented by different signs, depending on whether they occur in syllables bearing the main stress or not. In the first case, they are realised as [eː] and [oː] and are not only phonetically indistinguishable from the phonemes /ē/ and /ō/, but are also written with the same signs as these, ṣere and ḥolem. In the latter case, they are realised as [ɛ] and [ɔ], overlapping in quality with the phonemes /ɛ/ and /ɔ̄/ 19 See e.g. Joüon and Muraoka, 1991, pp. 77, 113–114; Huehnergard, 2015, p. 52. There are no examples of the definite article before /h/ followed by ḥatef qameṣ. This is not surprising, given the relative small number of lexemes derived from roots with initial /h/. In the infinitives of the hofʿal conjugation, ḥatef qameṣ is not usually to be expected. However, the only attested nominal form with an initial /h/ vocalised with ḥaṭef qameṣ is apparently such an infinitive, occurring in Lev. 26:34: ‫ ֵּ֚כל יְ ֵ ָ֣מי הֳשַ ָ֔ ָמה‬/kol ymē hošammɔ̄/ [ˈkol jiˈmeː hoːʃamˈmɔ:] ‘all the days of (its) being desolate’. The phrase reoccurs in the immediately following verse as ‫ כָל־יְ ֵ ָ֥מי הָ שַ ָ ֵ֖מה‬/kol-ymē hoššammɔ̄/ [kol-jiˈmeː hoʃʃamˈmɔː]. In the first attestation, the use of the ḥaṭef vowel serves likely to prevent the misinterpretation of qameṣ as a reflection of the long vowel /ɔ̄/. 21 This section summarises the descriptions by Khan, 2013a, pp. 85–107; Khan, 2013g; Khan, 2013h; Khan, 2013i; Khan, 2013j. 20 114 and written with the same signs as these, seghol and qameṣ. Moreover, due to a secondary stress, /o/ in this environment may be lengthened to [ɔː], becoming phonetically indistinguishable from the reflex of /ɔ̄/. In other words, the vocalic signs qameṣ, ḥolem, ṣere and seghol each represent reflexes of two different vowel phonemes. In addition to these vowels, there are also the ḥaṭef vowels, which are realised as short vowels with a quality identical to the vowel they are combined with; phonologically, they are usually realisations of shewa – that is, allophones of /Ø/ – which also has other realisations depending on the environment. Apart from this realisation as epenthetic vowels, ḥaṭef qameṣ and ḥaṭef seghol, when placed under non-guttural consonants, occasionally represent phonologically short vowels representing a syllable nucleus. In this context, they are often also written with the corresponding simple vowel signs. Both epenthetic as well as short ḥaṭef vowels can receive secondary stress and thus be lengthened. Table 2 lists the various vowel phonemes of the phonological system underlying the Tiberian vocalisation, together with their realisation in different phonetic environments and the respective vocalic signs. To this must be added the various realisation of /Ø/ shewa ([Ø], [a], [ɛ], [ɔ], [e], [o], [i], [u]). Table 2. The Tiberian vowel system and its orthographic representation Phoneme /a/ /ɛ/ /ɔ̄/ /e/ /ē/ /o/ /ō/ /i/ /ī/ /u/ /ū/ Conditioned allophones Stressed [aː] ( ַ◌) [ɛː] ( ֶ◌) [ɔː] ( ָ◌) [eː] ( ֵ◌) [eː] (‫ ◌ֵ י‬, ֵ◌) [oː] (◌), [ɔː]22 ( ֳ◌ [oː] (‫ ◌ו‬,◌) [iː] (◌ִ ) [iː] (‫) ִ◌י‬ [uː] (‫ ◌ו‬,◌) [uː] (‫)ו‬ Unstressed open [aː] ( ַ◌) [ɛː] ( ֶ◌) [ɔː] ( ָ◌) [ɛ] ( ֱ◌ , ֶ◌) [eː] (‫ ◌ֵ י‬, ֵ◌) , ָ◌) [ɔ] ( ֳ◌ , ָ◌) [oː] (‫ ◌ו‬,◌) [iː] (◌ִ ) [iː] (‫) ִ◌י‬ [uː] (‫ ◌ו‬,◌) [uː] (‫)ו‬ Unstressed closed [a] ( ַ◌) [ɛ] ( ֶ◌) – [ɛ] ( ֶ◌) – [ɔ] ( ָ◌) – [iː] (◌ִ ) – [u] (‫ ◌ו‬,◌) – The system of Tiberian vocalisation signs eventually replaced the Palestinian and Babylonian systems, however the pronunciation reflected by it fell into disuse; the surviving Ashkenazi and Sephardi pronunciation traditions go back to the Palestinian reading tradition, whereas the Yemenite pronunciation tradition is ultimately based on the Babylonian reading tradition.23 22 23 With secondary stress. Khan, 2013b, pp. 345–347. 115 The Western academic tradition is derived from the Sephardi pronunciation tradition24 and is based on a system of five vowels (a, e, i, o, u) that may be either short or long. It may be presented in a simplified form as in table 3.25 Table 3. The vocalisation system of Western academia Quality Quantity Long /ā/ ( ָ◌) /ē/ ( ֵ◌, ‫)◌ֵ י‬ /ī/ (‫ ִ◌י‬, ◌ִ ) /ō/ (◌, ◌‫)ו‬ /ū/ (‫)◌ו‬ a e i o u Short /a/ ( ַ◌) /e/ ( ֶ◌) /i/ (◌ִ ) /o/ ( ָ◌) /u/ (◌) Reduced /ă/ ( ֲ◌), /ə/ (◌ְ ) /ĕ/ ( ֱ◌), /ə/ (◌ְ ) /ŏ/ ( ֳ◌) Here, qameṣ is likewise used to represent two different phonemes, long /ā/ and short /o/. However, they are usually realised quite distinctively. They are also differentiated in terminology, the former being called simply qameṣ or qameṣ gadol, the latter qameṣ qaṭan or qameṣ ḥaṭuf. The difference between the Tiberian vowel system and the Western academic tradition regarding the definite article becomes quite evident from the examples in table 4. Table 4. Comparison between the Tiberian vowel system and the Western academic pronunciation Attestation Tiberian vocalisation Academic tradition26 Source ‫ֶ ָּֽׁהחָ כָ ִׂ֤ם‬ /hɛḥɔ̄kɔ̄m/ [ˌhɛː.ħɔː.ˈχɔːɔm] /hɛḥdɔ̄ʃīm/ [hɛː.ħɔðɔː.ˈʃiːim] /haḥokmɔ̄/ [ˌhaː.ħɔχ.ˈmɔː] /kol-haḥkɔ̄mīm/ [kɔl-ˌhaː.ħaχɔː.ˈmiːim] /heḥākām/ [ˌhe.xaː.ˈxaːm] /heḥŏdāʃīm/ [he.xŏ.daː.ˈʃiːm] /haḥokmā/ [ˌha.xox.ˈmaː] /kol-haḥăkāmīm/ [kol-ˌha.xă.xa:.ˈmiːm] Eccles. 2:19 ‘the wise man’ ‫הֶ חֳדָ ֜ ִשים‬ ‘the months’ ‫ַ ָּֽׁהחָ כְ ָ ָ֥מה‬ ‘the wisdom’ ‫ָל־ה ֲחכ ִָָ֔מים‬ ָ֣ ַ ‫כ‬ ‘all the wise men’ Neh. 10:34 2 Chr. 1:12 Ex. 36:4 In the Tiberian phonetic realisation, the vowels following /ḥ/ in the first three examples overlap in quality, being distinguished only partially, by length. In the academic pronunciation, the vowels in the first and last examples overlap in quality, as do the vowels in the two other examples. Also, the difference in the number of syllables is striking: in the academic tradition, the realisations of ‘the months’ and ‘all the wise men’ possess an additional syllable when compared with the Tiberian vocalisation. 24 Blau, 2010, p. 109. See e.g. Lambin, 1971, pp. 17–27. 26 This reflects the pronunciation of Biblical Hebrew in many institutions of higher learning, making the relevance of the following remark even more salient: “Much of our current knowledge of this reading tradition ... has not been incorporated so far into the standard textbooks of Biblical Hebrew used by students” (Khan, 2013a, v). 25 116 5 The point of departure for the development of the vocalisation of the definite article The degemination of gutturals and /r/ is a late feature in Biblical Hebrew. It must have occurred after the end of the Late Biblical period. It presupposes certain phonological developments and other phenomena which occurred in the spoken Hebrew of the Second Temple period. 5.1 The general weakening of the gutturals Originally, Hebrew possessed velar (or uvular) fricatives (/ḫ/, /ġ/) as well as pharyngeal fricatives (/ḥ/, /ʿ/). These were not distinguished in writing, since Hebrew used the Phoenician alphabet, where velar and pharyngeal fricatives had merged. The Greek transcriptions in the Septuagint show that these consonants were at least partially still distinguished in the 3rd century BCE, since they use χ and γ to represent the velar fricative /ḫ/ and /ġ/, although not consistently.27 Examples where the Septuagint has χ and γ representing velar sounds that later become pharyngeal are γαζα ‘Gaza’ (Josh. 15:47) for /ġazzā/ > ‫ ַעזָה‬/ʿazzɔ̄/; αχαζ ‘Ahaz’ (2 Kings 15:38) for /ʾāḫāz/ > ‫ אׇ חׇ ז‬/ʾɔ̄ḥɔ̄z/; γομορρα ‘Gomorra’ (Gen. 13:10) for /ġomorrā/ > ‫ עֲמ ָרה‬/ʿmōrɔ̄/; and χαρραν ‘Harran’ (Gen. 11:31) for /ḫarrān/ > ‫ חָ ָרן‬/ḥɔ̄rɔ̄n/. However, the latter also often appears as αρραν (e.g. Gen. 11:26), showing the instability of /ḥ/.28 The almost contemporary Dead Sea Scrolls and also inscriptions from Judaea similarly attest a weakening of the gutturals by either omitting or interchanging them.29 The Samaritan tradition, which has its roots in the Late Hebrew era, shows an extreme case of the weakening of the gutturals, resulting in their (almost) complete loss.30 Weakening of the gutturals is also attested in the Hebrew epigraphic record of certain vicinities in the Galilee.31 On the other hand, the Bar Kochba documents (2nd century CE) show almost no cases of guttural weakening. In the Greek transcriptions of Hebrew by Origen (3rd century CE), the existence and distinction of several gutturals may be inferred from the transcription of the vowels.32 Jerome (late 4th century CE) also clearly attests the preservation of different gutturals.33 This points not only to diachronic changes, but also to the synchronic coexistence of highly diverse spoken dialects and to differences between urban 27 Khan, 2013b, p. 342; Rendsburg, 2013, p. 102. This form occurs only referring to the proper name Ḥɔ̄rɔ̄n, not the geographic name. It may indicate that the distinction between velar and pharyngeal fricatives was less pronounced than often assumed and mainly preserved at least in writing regarding the better-known names, both proper names and geographical names. Nevertheless, the phonemes must have been distinguished at least until very close to the making of the Septuagint. 29 Khan, 2013b, p. 342. 30 Macuch, 1969, pp. 132–136; Mor, 2013, pp. 162–163. 31 Mor, 2013, p. 163. 32 Mor, 2013, pp. 164–165. 33 Brønno, 1970. 28 117 elites who are influenced by Greek and more rural speakers of Hebrew in the relevant era from around 300 BCE to around 300 CE.34 Not all attestations for preserved gutturals need to be interpreted as reflecting the living dialects, as they may reflect a ritual performance realisation during formal readings.35 This may be compared to Arabic, where, when reciting the Qurʾān, not only professional reciters but also many laypersons are able to realise the interdental fricatives /ṯ/ and /ḏ/, which in many dialects have either merged with the dental stops /t/ and /d/ or with the sibilants /s/ and /z/, and are also able to distinguish the phonemes /ḏ/̣ and /ḍ/, which have both become /ḏ/̣ in Bedouin or /ḍ/ in urban dialects. 5.2 The realisation of /r/ In the Greek of the Septuagint, geminated /r/ is still attested – e.g. τον αμορραιον ‘the Amorite (accusative sg)’ (Gen. 10:16); γομορρα ‘Gomorrah’ (Gen. 13:10); αρραν ‘Harran’ (Gen. 11:26); χαρραν ‘Harran’ (Gen. 11:31); τους χορραιους ‘the Horites (accusative pl)’ (Gen. 14:6); σαρρα ‘Sarah’ (Gen. 17:17). There are even examples in relatively late books; for example, αμορραιων ‘of the Amorites (genitive pl)’ (Neh. 9:8). In the Greek transcriptions of Hebrew in Origen’s Hexapla (3rd century CE) and also in the Latin transcriptions of Jerome (late 4th century CE), geminated /r/ is no longer attested in forms where it was found historically. For example, Origen has ουβαρεχ ‘and bless (msg)’ (Ps. 28:9), where the /r/ must originally have been geminated, since gemination of the middle consonant is the characteristic marker of the piʿel stem.36 The fact that /r/ shares the loss of gemination with the pharyngeal and laryngeal consonants makes it likely that it was realised as a uvular.37 Indeed, in Tiberian Hebrew, /r/ has a uvular realisation with an emphatic apico-alveolar roll [ṛ] as a conditioned allophone.38 In isolated cases, where the Tiberian tradition has a geminated /r/, this was realised as uvular trill [ʀ].39 Furthermore, the frequent loss of /r/ in the environment of gutturals40 should be taken as an additional indication for a uvular realisation. In this context, it is important to note that in the Samaritan tradition, the gemination of /r/ has been retained and /r/ is realised as an apical consonant.41 34 Khan, 2013b, p. 342; Mor, 2013, pp. 162–164. Mor, 2013, pp. 162–164. 36 Khan, 2013f, p. 388. 37 This was already being suggested in 1953 by Gumpertz; see Mor, 2013, p. 161. Although a guttural-like realisation was dismissed by others (e.g. Joüon and Muraoka, 1991, pp. 29–30), its primarily uvular realisation is clearly described in medieval Masoretic treatises; see Khan, 2013f, p. 384. 38 Khan, 2013i, p. 773. 39 Khan, 2013f, p. 386. 40 Khan, 2013f, p. 388. 41 Macuch, 1969, pp. 115–118; Stadel, 2017. 35 118 Since the laryngeals in Samaritan Hebrew have become considerably weakened and shifted in most cases to /ʾ/, the realisation of /r/ is probably not a secondary development under the influence of Arabic, but rather represents a more original realisation. 5.3 Consonant changes prior to degemination From what has been said above, the degemination of both gutturals and /r/ strongly suggests there was a uvular realisation of /r/. It also seems likely that the degemination follows the merger of the velar (or uvular) fricatives with the pharyngeal fricatives. If the degemination of the gutturals had preceded the merger of velar and pharyngeal fricatives, an even more variegated picture of conditioned allomorphs of the definite article might have been expected than was seen in table 1. The following phases may therefore be postulated: 1. /ḫ/ and /ḥ/ > /ḥ/; /ġ/ and /ʿ/ > /ʿ/ 2. /r/ [r] > [ʁ] To allow sufficient time for the merger of the velar and pharyngeal fricatives and the subsequent uvular realisation of /r/, the process of degemination is likely to have started no earlier than the end of the 3rd or the beginning of the 2nd century BCE. 5.4 Lengthening and reduction of vowels 5.4.1 The Canaanite shift In the development of Biblical Hebrew, several phases of vowel lengthening must be distinguished. The first of these concerns the so-called Canaanite sound shift during which original /ā́/ and /ā́/ < /áʾ/ (with loss of the laryngeal stop in syllable final position) shifted to /ṓ/. Examples are *lašā́nu > *lašṓnu (> ‫ לְ שון‬/lšōn/ [laˈʃoːon]) ‘tongue’; *ráʾsu > rā́šu > rṓšu (> ‫ ראש‬/rōš/ [ˈʁoːoʃ]) ‘head’. This change is very early; although it must be assumed to have occurred before the beginning of guttural degemination, it is relevant in so far as some monosyllabic nouns may have restored the original long /ā/ due to the levelling with the unaccented construct form, e.g. in ‫ עָב‬ʿāb > /ʿɔ̄b/ [ʕɔːɔv] ‘cloud’.42 5.4.2 Tonic lengthening The Canaanite shift was followed by the loss of final vowels successively on nouns in the construct state, finite verbs and nominal forms in the absolute 42 Blau, 1993, p. 35. 119 state. The dropping of these vowels resulted in the lengthening of short vowels that had originally stood in a stressed open penultimate syllable – e.g. *dayyánu > *dayyán > dayyā́n (> ‫ דַ יָן‬/dayyɔ̄n/ [dajˈjɔːɔn]) ‘judge’. This phenomenon is theoretically relevant given the abstract possibility of biradical nouns with original vowel /a/ having become /ā/ and thus prone to cause a change in the vowel of the definite article in certain cases. In this context also, the lengthening of stressed open /a/ in the pause forms of segholate nouns of the type qaṭl should be mentioned, e.g. málk ~ málek > mā́lɛk (> ‫ ָּֽׁ ָמלְֶך‬/mɔ̄lɛk/ [ˈmɔːlɛχ]) ‘king’. The long vowel here conditions the occurrence of the respective variant of the definite article. Pausal lengthening thus predates the degemination of gutturals. 5.4.3 Pretonic lengthening Later, short vowels in open syllables immediately preceding the accent were lengthened. This so-called pretonic lengthening did not influence all vowels, but rather depended on sonority and to a certain extent on the prosodic weight of the noun in question.43 It occurs, for example, in (*ḥakámu > *ḥakám >) *ḥakā́m > *ḥākā́m (> ‫ חָ כָם‬/ḥɔ̄kɔ̄m/ [ħɔːˈχɔːɔm]) ‘wise man’. This change has been regarded by many as due to Aramaic influence on the Hebrew spoken during the Second Temple period.44 Both the transcription of lengthened /i/ by eta in the Septuagint and the representation of lengthened /a/ in pretonic position by ʾalef in the Dead Sea Scrolls indicate that this sound shift indeed took place during the Second Temple period when Hebrew was still a spoken language. The dating of pretonic lengthening is important, since most cases of unstressed ָ◌ influencing the vocalisation of the article before /ʿ/, /h/ and /ḥ/ originated from pretonic lengthening of /a/. This means that pretonic lengthening must have occurred before the degemination of gutturals, or at least of /ʿ/, /h/ and /ḥ/. While other dates have been suggested,45 the evidence from the Septuagint as well as the Dead Sea Scrolls gives an important terminus ad quem for establishing the date of this lengthening. 5.4.4 (Pro)pretonic reductions In syllables further removed from the accent, /a/ in open syllables was reduced. Together with pretonic lengthening, this change accounts for the vocalic variations in the inflectional paradigms of nominals – e.g. /ḥɔ̄kɔ̄m/ [ħɔː.ˈχɔːɔm] ‘a wise man’ (msg) versus /hkɔ̄mīm/ [ħaχɔː.ˈmiːim] ‘wise men’ (mpl) – and thus for the distribution of the article variants across inflection patterns – /hɛḥɔ̄kɔ̄m/ [hɛː.ħɔː.ˈχɔːɔm] ‘the wise man’ (msg) versus /haḥkɔ̄mīm/ [haː.ħaχɔː.ˈmiːim] ‘the wise men’ (mpl). 43 Blau, 1993, p. 31; Blau, 2010, pp. 123–132; Khan, 2013e, p. 225. Blau, 1980, p. 33; Blau, 2010, pp. 128–129; Khan, 2013e, pp. 224–225. 45 Blau, 2010, pp. 126–129. 44 120 In addition, in most cases where short stressed vowels were occurring in open penultimate syllables, the stress moved to the ultimate syllable, usually resulting in a reduction of the short, now unstressed, vowel in the penultimate syllable. 5.4.5 Compensatory lengthening Compensatory lengthening occurs when a vowel is spread to occupy the slot vacated by the loss of a consonantal segment and presents a means to keep the previous prosodic structure intact.46 According to Blau, due to the occurrence of compensatory lengthening, a relative chronology can be established for the successive degemination of the gutturals:47 1. Degemination of /r/, /ʾ/ followed by compensatory lengthening of a preceding vowel: /VGG/ > /VːG/. 2. Degemination of /ʿ/, /h/ followed by compensatory lengthening of a preceding vowel in some cases: /VGG/ > /VːG/ or /VG/. 3. Degemination of /ḥ/, without compensatory lengthening of a preceding vowel: /VGG/ > /VG/. The attested exceptions from the established rules mentioned by Gesenius as well as by Joüon and Muraoka mainly involve nouns with initial /ʿ/.48 Blau emphasises that this lengthening is of a purely quantitative nature and the realisation of vowels in open syllables in the Tiberian vocalisation systems is a purely phonetic realisation.49 Khan, however, makes the important observation that compensatory lengthening was a recurring feature in Biblical Hebrew, and suggests that in the later pre-Masoretic stages a distinction must be made between an earlier phase where compensatory lengthening was paralleled or associated with a qualitative shift (/aGG/ > (/aːG/ >) /ɔ̄G/) and a later phase where no concomitant change in vowel quality was operative. He suggests that the slot of the lost consonantal segment may have been left empty for some time.50 This observation also indicates that the raising of /ā/ to /ɔ̄/ is a late feature, postdating at least the early stages of guttural degemination. The assumption that compensatory lengthening in the early phase was accompanied or followed by qualitative modification as well does not change the established relative chronology for the loss of gemination of the individual gutturals. 46 Khan, 2013c, p. 500. Blau, 2010, pp. 82–83. However, Blau, 1980, p. 35 posits two waves of degemination, differentiated by whether syllable structure did not yet allow short vowels in open syllables or already allowed them. 48 Gesenius, 1910, p. 111; Joüon and Muraoka, 1991, p. 114. 49 Blau, 2010, p. 82. However, Blau, 1980, pp. 37–39 argues very strongly for a quantitative and qualitative change. 50 Khan, 2013c, pp. 501–503. 47 121 5.5 Other phonological changes Neither the Law of Attenuation (an unaccented /a/ in a closed syllable before the accent shifts to an /i/) nor Philippi’s Law (an accented /i/ in a closed syllable, especially when penultimate, shifts to an /a/) are relevant for the development of the definite article before degeminated gutturals. Consequently, the dating of these does not matter here. In addition, the spirantisation of bgdkpt and the question of the anaptyctic vowels in segholates are irrelevant.51 It should also be noted that nouns derived from geminate roots in the derivational patterns qall, qill and qull were influenced by neither tonic lengthening nor compensatory lengthening;52 the loss of gemination of their final consonants did not lead to a perceived change in the prosodic structure of the syllable. 5.6 The vowel system in the Late Biblical or early Second Temple era Table 5 shows the vowel system which may be posited as existing at the time when the gutturals started to degeminate. Table 5. Vowel system when gutturals started to degeminate Quality a e o i u 6 Quantity53 Long ā<a ē < i, ay ō < u, aw ī<ī ū<ū Short a<a e<i o<u i<i u<u The degemination of the individual gutturals and subsequent vowel changes The preceding discussion shows that loss of gemination was preceded by various modifications of the following vowel, such as pretonic lengthening and propretonic reduction, and that this had already begun when Hebrew was still a spoken language. The occurrence of exceptions to the rules, mainly with nouns derived from roots with initial /ʿ/, together with the general observation 51 In the examples adduced, the anaptyctic vowel in segolates is assumed, but the spirantisation of bgdkpt is not. 52 On the lack of tonic lengthening, see Blau, 1993, p. 31; for the lack of compensatory lengthening, see Khan, 2013c, p. 503. 53 It is not certain that /e/ and /o/ should already be posited as independent phonemes by this point, rather than as allophones of /i/ (in the case of /e/) and /u/ (in the case of /o/); however, in this account they will be treated as phonemes. 122 that compensatory lengthening is not entirely regular, is indicative of spoken language. This is also true for the sporadically attested gemination of /r/. More precisely, these phenomena are indicative of a language undergoing change. Given that the Tiberian reading tradition is remarkably homogenous – even though it differs from the dialect (or dialects) underlying the consonant text and apparently makes recourse to different earlier dialect forms – the occurrence of these exceptions shows that the orally transmitted text became increasingly frozen as the language ceased to be spoken. The various stages of the loss of gemination accompanied by modification of the preceding vowel, or at least the initial stages of this process, may be some of the last changes connected to the use of Hebrew as a spoken language in antiquity. 6.1 Degemination /r/ and /ʾ/ When gemination of /r/ and /ʾ/ is lost and the preceding vowel is lengthened, this results in the definite article having the vowel /ā/. Since there are no conditioned variants of the definite article when it precedes /ʾ/ and /r/, it is not strictly necessary that pretonic lengthening was occurring or had already been completed. However, the transcription of lengthened /i/ by eta and the occurrence of geminate /r/ in the Septuagint make clear that pretonic lengthening had most likely already taken place, although this cannot be determined for all spoken Hebrew dialects. From the evidence of these two gutturals alone, it is impossible to determine if the loss of gemination was immediately followed by compensatory lengthening or whether the slot vacated by the lost consonantal segment remained empty for some time before the vowel was lengthened; likewise, it is impossible to determine if this lengthening was accompanied by a qualitive change immediately or when this took place.54 Note the following examples: /haʾ.ʾōr/ > /ha.ʾōr/ > /haa.ʾōr/ > /hā.ʾōr/ > ‫ הָ ֵ֖אֹור‬/hɔ̄ʾōr/ [hɔːˈʔoːoʁ] ‘the light’ (Gen. 1:4); /haʾ.ʾādām/ > /ha.ʾādām/ > /haa.ʾādām/ > /hā.ʾādām/ > ‫ ָּֽׁ ָהאָ ָ ֛דם‬/hɔ̄ʾɔ̄dɔ̄m/ [ˌhɔːʔɔːˈðɔːɔm] ‘the human being’ (Gen. 2:19); /haʾ.ʾiššā/ > /ha.ʾiššā/ > /haa.ʾiššā/ > /hā.ʾiššā/ > ‫ ָ ָּֽׁה ִא ָ ֵ֖שה‬/hɔ̄ʾiššɔ̄/ [ˌhɔːʔiʃˈʃɔː] ‘the woman’ (Gen. 3:2); /ʾet-har.rāʿā/ > /ʾet-ha.rāʿā/ > /ʾethaa.rāʿā/ > /ʾet-hā.rāʿā/ > ‫ אֶ ת־הָ ָרעָ ָ֥ה‬/ʾɛt-hɔ̄rɔ̄ʿɔ̄/ [ʔɛθ-hɔːʁɔːˈʕɔː] ‘the evil thing (fsg, direct object)’ (1 Sam. 6:9); /har.rōš/ > /ha.rōš/ > /haa.rōš/ > /hā.rōš/ > ‫ הָ ָּ֙ראש‬/hɔ̄rōš/ [hɔːˈʁoːoʃ] ‘the head’ (1 Sam. 13:17); /har.rīšōn/ > /ha.rīšōn/ > /haa.rīšōn/ > /hā.rīšōn/ > ‫אשֹון‬ ָ֔ ‫ הָ ִר‬/hɔ̄rīšōn/ [hɔːʁiːˈʃoːon] ‘the first’ (Ezra 7:9). Although several sporadic examples of geminated /r/ are attested, these do not include any cases with the definite article. There are also a few instances of initial /ʾ/ having been lost with the article showing compensatory lengthening.55 In these cases, ʾalef is occasionally preserved in writing, while at other 54 55 However, see the following section. Gesenius, 1910, pp. 110–111. 123 times it is lost: ֮‫ ִ ָּֽׁמן־הָ אזִקִ ים‬/min-hɔ̄zikkīm/ [ˌmin-hɔːziqˈqiːim] ‘from the chains’ (Jer. 40:4) < /min-haʾʾăzikkīm/; ‫סורים‬ ֵ֖ ִ ָ‫ ה‬/hɔ̄sūrīm/ [hɔːsuːˈʁiːim] ‘of the prisoners’ (Eccles. 4:14) < /haʾʾăsūrīm/, compare ֶׁ֙‫ֲסורים‬ ִ ‫ ַ ָּֽׁלא‬/laʾăsūrīm/ [ˌlaʔasuːʁˈiːim] ‘to prisoners’ (Isa. 49:9); ‫ ָ ָּֽׁה ַר ִ ֵ֖מים‬/hɔ̄rammīm/ [ˌhɔːʁamˈmiːim] ‘the Arameans’ (2 Chr. 22:5) < /haʾʾărammīm/, compare ‫ א ֲַר ִָ֔מים‬/ʾărammīm/ [ʔaʁamˈmiːim] ‘Arameans’ (2 Kings 9:15); and ‫ הָ שֲ פָּֽׁ ֹות‬/hɔ̄špōt/ [hɔː.ˈʃafoːoθ] ‘(gate) of the dung’ (Neh. 3:13) < /haʾʾašpōt/, compare ‫ הָ אַ ְש ַ֗פֹות‬/hāʾašpōt/ [hɔːʔaʃˈpoːoθ] ‘(gate) of the dung’ (Neh. 3:14). These examples include both cases where the vowel following the laryngeal stop was originally a short /a/ in an open unstressed syllable that underwent propretonic reduction (e.g. /haʾʾărammīm/) and one case of a short /a/ in a closed unstressed syllable where it should have been retained (/haʾʾašpōt/). In the former cases, it is easy to imagine that propretonic reduction leading to vowel elision and thus ultimately a syllable initial consonant cluster with /ʾ/ as the first consonant might result in the loss of the laryngeal; however, this does not occur usually, as can be seen from ‫ ָ ָּֽׁה ֲאג ִָָ֔גי‬/hɔ̄ʾgɔ̄gī/ [ˌhɔː.ʔaɣɔːˈɣiː] ‘the Agatite’ (Est. 8:3). Moreover, propretonic reduction did not lead to the sudden loss of a vowel, but is rather a slow process – *haʾʾagāgī > *haʾʾăgāgī > *haʾăgāgī > *haːʾăgāgī > haːʾØgāgī – with the insertion of an epenthetic vowel to break up the syllable initial cluster /ʾg/ in the Tiberian reading tradition: [hɔː.ʔaɣɔːˈɣiː]. These cases where an initial /ʾ/ is lost are not restricted to books written in Late Biblical Hebrew. However, the phenomenon of dropping the ʾalef also in writing seems to be; the additional cases of preserved ʾalef mentioned by Gesenius belong to Classical Biblical Hebrew.56 This seems to suggest two things. Firstly, the weakening of /ʾ/ in the spoken language may occasionally have gone further than mere degemination, leading instead to a complete loss of the glottal stop after compensatory lengthening took place. Secondly, degemination and loss of the glottal stop were an ongoing process during the canonisation of some of the later books, resulting in the phenomenon showing up even in the consonant text. This is especially easy to imagine with common ethnonyms (‘the Arameans’) or well-known landmarks such as ‘the dung gate’. 6.2 Degemination of /ʿ/ Degemination of /ʿ/ has two outcomes: where /ʿ/ is followed by an unstressed /ā/, the vowel of the definite article changes its quality, /a/ > /ɛ/; elsewhere, the vowel of the article undergoes compensatory lengthening, /a/ > /ā/. Examples of the first of these processes can be seen in /ʾel-haʿ.ʿāpā́ r/ > /ʾel-ha.ʿāpā́ r/ > /ʾel-hɛ.ʿāpā́ r/ > ‫ אֶ ל־הֶ ע ָ ָָּֽׁפר‬/ʾel-hɛʿɔ̄pɔ̄r/ [ʔɛl-hɛːʕɔːˈfɔːɔʁ] (Eccles. 3:20) ‘to the dust’; /haʿ.ʿānā́ n/ > /ha.ʿānā́ n/ > /hɛ.ʿānā́ n/ > ‫ ִ֠הֶ ָענָן‬/hɛʿɔ̄nɔ̄n/ [hɛːʕɔːˈnɔːɔn] ‘the 56 Gesenius, 1910, pp. 110–111. 124 cloud’ (Neh. 9:19). Examples of compensatory lengthening include /ʿadhaʿ.ʿɛ́rɛb/ > /ʿad-ha.ʿɛ́rɛb/ > /ʿad-haa.ʿɛ́rɛb/ > /ʿad-hā.ʿɛ́rɛb/ > ‫ עַד־הָ עֶ ֵֶ֖רב‬/ʿadhɔ̄ʿɛrɛb/ [ʕað-hɔːˈʕɛːʁɛv] ‘until the evening’ (Lev. 17:15); /ʿad-haʿ.ʿā́ rɛb/ > /ʿad-ha.ʿā́ rɛb/ > /ʿad-haa.ʿā́ rɛb/ > /ʿad-hā.ʿā́ rɛb/ > ‫ עַד־הָ עָ ֶֶ֑רב‬/ʿad-hɔ̄ʿɔ̄rɛb/ [ʕaðhɔːˈʕɔːʁɛv] ‘until the evening (pausa)’ (Lev. 15:16); /haʿ.ʿŭmārī́m/ > /ha.ʿŭmārī́m/ > /haa.ʿŭmārī́m/ > /hā.ʿŭmārī́m/ > ‫ ָ ָּֽׁהעֳמָ ִ ֛רים‬/hɔ̄ʿmɔ̄rīm/ [ˌhɔː.ʕɔmɔː.ˈʁiːim] ‘the sheaves’ (Ruth 2:15). This process may be explained as follows. In the first case, where there is a change in quality, the vowel following the guttural is /ā/, resulting mainly from pretonic lengthening of /a/; occasionally it is due to tonic or pausal lengthening or else continues an original /ā/. Following degemination, the slot of the lost consonantal segment remains unfilled for some time. Given the increased weakening of /ʿ/, the vowel of the definite article undergoes dissimilation: /aʿ.ʿā/ > /aØ.ʿā/ > /ɛ.ʿā/. This serves to prevent further weakening of the guttural and eventually the merger of the two original syllables; that is, it avoids a scenario such as /aʿ.ʿā/ > /aØ.ʿā/ > /a.ā/ > /aā/ > /ā/. In other environments, where degemination leads to compensatory lengthening, the slot of the lost consonantal segment remains likewise unfilled for some time before the vowel of the article spreads to the empty slot. The occurrence of a long /ā/ before the guttural likewise helps to prevent its loss between non-homorganic vowels; the weakening of /ʿ/ has not yet progressed sufficiently to facilitate its loss in front of a stressed homorganic vowel. The dissimilation57 from /a/ to /ɛ/ to prevent the loss of a laryngeal or pharyngeal between homorganic /a/ and /ā/ occurs not only with /ʿ/ and the remaining gutturals in this environment, but elsewhere as well, such as in the nominal derivation pattern qaṭṭāl – e.g. *lahhabat > *lɛhābā > ‫ לֶהָ בָ ה‬/lɛhɔ̄bɔ̄/ [lɛːhɔːˈvɔː] ‘flame’ and *naḥḥamat > *nɛḥāmā > ‫ נֶחָ מָ ה‬/nɛḥɔ̄mɔ̄/ [nɛːħɔːˈmɔː] ‘comfort’ – and possibly also in *ʾaḥḥāw > ‫ אֶ חָ יו‬/ʾɛḥɔ̄w/ [ʔɛːˈħɔːɔv] ‘his brothers’ and ʾaḥḥād > ‫ אֶ חָ ד‬/ʾɛḥɔ̄d/ [ʔɛːˈħɔːɔð] ‘one’.58 The fact that this dissimilation does not take place with the article before degeminated /ʾ/ shows that it is a feature connected to the ongoing, progressive weakening of the gutturals that dates to a later point than the loss of gemination of /ʾ/ following the definite article.59 57 According to Blau, 2010, p. 181 this should be regarded as a case of assimilation. However, this is only the case of assimilation of vowel quality in Tiberian Hebrew, and it is doubtful that both vowels would have been sufficiently raised to an equal height by this point in the historical development. Blau, 1980, pp. 35–36 also argues very strongly for assimilation. 58 Blau, 1980, p. 33–34; Blau, 1993, p. 38; Huehnergard, 2015, p. 52. 59 Blau, 1980, p. 36 argues that the change /a/ > /ɛ/ cannot be a case of dissimilation, because it did not occur following the degemination of /ʾ/. However, he overlooks the fact that in one phase, degemination may have caused compensatory lengthening in order to keep the prosodic structure intact, while in a subsequent phase, further weakening may have caused dissimilation in the environment /aGā/ > /ɛGā/, irrespective of whether the environment was the result of degemination of additional gutturals or original. 125 Dissimilation of /a/ > /ɛ/ also takes place in the interrogative particle h, which is generally vocalised with ḥaṭef pataḥ (hă-); before gutturals it is vocalised with pataḥ (ha-), unless the guttural is vocalised with qameṣ, in which case the vowel dissimilates to seghol (hɛ-). Since these gutturals were never geminated, it confirms that this is a phenomenon connected to the progressive weakening of all gutturals, which occurred only after the degemination of /ʾ/ following the definite article. It also suggests that ‫ אֶ חָ יו‬/ʾɛḥɔ̄w/ [ʔɛːˈħɔːɔv] ‘his brothers’ < *ʾaḥḥāw and ‫ אֶ חָ ד‬/ʾɛḥɔ̄d/ [ʔɛːˈħɔːɔð] ‘one’ < ʾaḥḥād may have only ever been virtually geminated; that is, for some reason they did not undergo pretonic lengthening.60 In ‫ ָ ָּֽׁהעֳמָ ִ ֛רים‬/hɔ̄ʿmɔ̄rīm/ [hɔː.ʕɔmɔː.ˈʁiːim] ‘the sheaves’, the /ʿ/ vocalised with ḥaṭef qameṣ was possibly still occupied by the reduced reflex of */u/ which was sufficiently distinct from the realisation of /ā/ to preclude the possible intervocalic loss of /ʿ/ in this environment. The exceptions from the rule attested for this consonant all have pataḥ as the vowel preceding the degeminated guttural – e.g. ֶׁ֙‫ ֶ ָּֽׁאת־הַ עֲבֹוט‬/ʾet-haʿbōṭ/ [ˌʔɛːθ-haː.ˈʕavoːoṭ] ‘the pledge (direct object)’ (Deut. 24:13); ‫הַ עִ וְ ִ ִׂ֤רים‬ /haʿiwrīm/ [haːʕivˈʁiːim] ‘the blind’ (2 Sam. 5:6); ‫ הִַ֭ עזֶבֶ ת‬/haʿōzɛbɛt/ [haːʕoːˈzɛːvɛθ] ‘the one who is leaving (fsg)’ (Prov. 2:17).61 This is an additional indication that compensatory lengthening did not occur immediately following degemination, but only a certain time later, suggesting that perhaps the maintenance of the original prosodic structure was regarded as a sufficient means to preserve the weakening guttural. The second qameṣ in ‫ הָ עָם‬/hɔ̄ʿɔ̄m/ [hɔːˈʕɔːɔm] ‘the people’ is not the expected outcome of the loss of final gemination (*ʿamm > *ʿam) but rather due to pause lengthening. A few nouns have turned their pausal forms into the form following the definite article, e.g. ‫ הָ ָּֽׁ ָא ֶרץ‬/hɔ̄ʾɔ̄rɛṣ/ [hɔːˈʔɔːʁɛṣ]. As it happens, ‫ הָ עָם‬is frequently found with disjunctive accents (more than 66 per cent of all attestations). 6.3 Degemination of /h/ Degemination of /h/ results initially in an empty slot following the loss of the consonant component. The dissimilation rule /aGā/ > /ɛGā/, operative since the previous stage, now affects also the environment /ahā/ < /ahhā/, causing the vowel of the definite article before /h/ followed by an unstressed /ā/ to dissimilate to /ɛ/. Examples are /ʾel-hah.hārīm/ > /ʾel-ha.hārīm/ > /ʾelhɛ.hārīm/ > ‫ל־הֶ הָ ִרים‬ ִ֠ ֶ‫ א‬/ʾel-hɛhɔ̄rīm/ [ʔɛl-hɛːhɔːˈʁiːim] ‘to the mountains’ (Ezek. 31:12); /hah.hāmōn/ > /ha.hāmōn/ > /hɛ.hāmōn/ > ‫ הֶ הָ ֶ֑מֹון‬/hɛhɔ̄mōn/ [hɛːhɔːˈmɔːɔn] ‘the multitude’ (2 Kings 25:11). 60 The form ‫ נֶאָ צות‬/nɛʾɔṣɔt/ [nɛːʔɔːˈṣɔːɔθ] ‘contempt (fpl)’ (Neh. 9:18, 26) should perhaps be ̄ ̄ regarded as a newly derived form or as a by-form of ‫ נְ אָ צָ ה‬/nʾɔ̄ṣɔ̄/ [nɔʔɔːˈṣɔː]; cf. Huehnergard, 2015, p. 52. 61 Gesenius, 1910, p. 111. 126 In other environments, the possibility of undergoing compensatory lengthening is becoming increasingly restricted. With degeminated /ʿ/, lengthening was possibly before any vowel following the guttural, but with /h/, this has become restricted to long, accented low and mid-high vowels; that is, it only occurs before stressed /ā/ and /ē/. After the context of unstressed /ā/ (where dissimilation happens), these two environments would seem to be the most likely to result in a further weakening of /h/, which would potentially lead to a complete loss of /h/ and then a possible syllable merger; that is, the following might hypothetically occur: /ahhā́ C/, /ahhḗC/ > /ahā́C/, /ahḗC/ > /aā́ C/, /aḗC/. To avoid this, the first /a/ is lengthened: /ahhā́ C/, /ahhḗC/ > /aːhā́C/, /aːhḗC/ > /āhā́ C/, /āhḗC/. Examples of this are ‫ הָ ָהֶ֑ר‬/hɔ̄hɔ̄r/ [hɔːˈhɔːɔʁ] ‘the mountain’ (Ex. 19:20); ‫ הָ ָה ֶָ֑רה‬/hɔ̄hɔ̄rɔ̄/ [hɔːˈhɔːʁɔː] ‘to the mountain (terminative)’ (Deut. 10:1); ‫ הָ הֵ ֒ם‬/hɔ̄hēm/ [hɔːˈheːem] ‘those (mpl)’ (Gen. 6:4); and ‫ הָ ֵ ָּֽׁהמָ ה‬/hɔ̄hēmmɔ̄/ [hɔːˈheːemmɔː] ‘those (mpl)’ (Jer. 14:15).62 Traditionally, compensatory lengthening has been regarded as only occurring regularly before stressed /ā/, with the other cases being regarded as exceptions.63 However, a closer look at the different stages of compensatory lengthening following the degemination of gutturals after the definite article reveals that what is happening is an ongoing restriction of the environments where lengthening (followed by qualitative change) can occur. Where the vowel following a degeminated /h/ is anything else – that is, where it is neither an unstressed /ā/ nor a stressed /ā/ or /ē/ – whether in a stressed open syllable or a closed syllable, the vowel of the definite article remains unchanged and the empty slot resulting from the loss of gemination remains unoccupied. For example, /hah.hēkāl/ > /ha.hēkāl/ > ‫ הַ הֵ יכָ ָ֥ל‬/hahēkɔ̄l/ [haːheːˈχɔːɔl] ‘the palace’ (1 Kings 6:17); /hah.hī́n/ > /ha.hī́n/ > ‫ הַ הִ י ֩ן‬/hahīn/ [haːˈhiːin] ‘the hin’ (Num. 28:14); /hah.hū́ / > /ha.hū́ / > ‫ הַ הוא‬/hahū/ [haːˈhuː] ‘that one (msg)’ (Isa. 2:11); /hah.hī́/ > /ha.hī́/ > ‫ הַ הִ י ֩א‬/hahī/ [haːˈhiː] ‘that one (fsg)’ (Dan. 12:1); /hah.hōlkī́m/ > /ha.hōlkī́m/ > ‫ הַ הלְ ִ ָּֽׁכים‬/hahōlkīm/ [haːhoːolˈχiːim] ‘those who walk (mpl)’ (Ex. 10:8). Historically, the third masculine plural personal pronoun hēm/hēmmɔ̄ developed from *hum; the vowel /u/ changed to the vowel /i/ by analogy with the third feminine plural form *hinnā, with a subsequent development of /i/ > /ē/. It would seems at first sight that this should be the phoneme /e/ of unspecific length, with the allophones [eː] in stressed and [ɛ] in unstressed environments. However, apart from a single attestation in Eccles. 3:18, a form [hɛm] does not occur. On the other hand, the third masculine plural possessive suffix is always /-hɛm/, despite having the same origin and likewise being always stressed, [hɛːm]. The development of the allophones of /e/ occurs late, and it seems that the independent and the suffix pronoun have ended up with different phonemes. However, if a thorough investigation were to show that the vowel in the independent personal pronoun is to be analysed as the conditioned allophone, the above environment can be equally restated as stressed [ā] and [ē]. 63 Blau, 1980, p. 37 discusses only /hɔhɔr/ [hɔːˈhɔːɔʁ], which he regards as a case of complete ̄ ̄ assimilation caused by the stress (/hɔ̄́r/). In his view, the form should actually have /a/ in the article. 62 127 6.4 Degemination of /ḥ/ The last guttural to lose gemination was /ḥ/. When occurring before an /ḥ/ followed by stressed or unstressed /ā/, the vowel of the definite article dissimilates. For example, this occurs in /wkol-haḥ.ḥārā́ š/ > /wkol-ha.ḥārā́ š/ > /wkolhɛ.ḥārā́ š/ > ‫ וְ כָל־הֶ חָ ָ ֵ֖רש‬/wkol-hɛḥɔ̄rɔ̄š/ [vaχɔl-hɛːħɔːˈʁɔːɔʃ] ‘and every craftsman’ (2 Kings 24:14); /haḥ.ḥāṣḗr/ > /ha.ḥāṣḗr/ > /he.ḥāṣḗr/ > ‫ הֶ חָ ֵצֵ֖ר‬/hɛḥɔ̄ṣēr/ [hɛːħɔːˈṣeːeʁ] ‘the court’ (Ex. 38:31); /haḥ.ḥārɛb/ > /ha.ḥārɛb/ > /hɛ.ḥārɛb/ > ‫ הֶ ָח ֶֶ֑רב‬/hɛḥɔ̄rɛb/ [hɛːˈħɔːʁɛv] ‘the sword’ (2 Sam. 11:25). Where /ḥ/ is vocalised with ḥaṭef qameṣ in Tiberian Hebrew, it may be assumed that the original vowel /u/ has been completely reduced to Ø by this point in time. In other words, it is the vowel on the following consonant which can trigger dissimilation of the vowel of the definite article. In most of the attested cases, the vowel on the second stem consonant is indeed /ā/ – e.g. /haḥ.ḥrābṓt/ > /ha.ḥrābṓt/ > /hɛ.ḥrābṓt/ > ‫ הֶ ח ֳָר ָּ֙בֹות‬/hɛḥrɔ̄bōt/ [hɛː.ħɔʁɔːvɔːɔθ] ‘the ruins’ (Ezek. 33:24); /haḥ.ḥdāšī́m/ > /ha.ḥdāšī́m/ > /hɛ.ḥdāšī́m/ > ‫הֶ חֳדָ ֜ ִשים‬ /hɛḥdɔ̄šīm/ [hɛː.ħɔðɔːˈʃiːim] ‘the months’ (Neh. 10:34). The only exception seems to be /haḥ.ḥrēbṓt/ > /ha.ḥrēbṓt/ > /hɛ.ḥrēbṓt/ > ‫ הֶ ח ֳֵר ֛בֹות‬/hɛḥrēbōt/ [hɛː.ħɔʁeːvɔːɔθ] ‘which are ruined (fpl)’ (Ezek. 36:35).64 This may possibly be an indication that following the loss of the ability to undergo compensatory lengthening, the environment conditioning the dissimilation has begun to be enlarged to include mid-high long vowels. Given that possibly around this time Hebrew ceased to be a spoken language, this state became frozen and the development discontinued. In all other environments, the vowel of the definite article remains unchanged following degemination. This is attested in /haḥ.ḥṓl/ > /ha.ḥṓl/ > ‫הַ ֶ֑חֹול‬ /haḥōl/ [haːˈħoːol] ‘the sand’ (Prov. 27:3); /haḥ.ḥóšɛk/ > /ha.ḥóšɛk/ > ‫הַ ָ֔חשֶ ְך‬ /haḥošɛk/ [haːˈħoːʃɛχ] ‘the darkness’ (Eccles. 11:8); /haḥ.ḥokmā/ > /ha.ḥokmā/ > ‫ ַ ָּֽׁהחָ כְ ָ ָ֥מה‬/haḥokmɔ̄/ [haːħoχˈmɔː] ‘the wisdom’ (1 Chr. 1:12); /ʾethaḥ.ḥélɛb/ > /ʾet-ha.ḥélɛb/ > > ‫ אֶ ת־הַ ֵ ֵ֖חלֶב‬/ʾet-ha.ḥelɛb/ [ʔɛθ-haːˈħeːlɛv] ‘the fat (direct object)’ (Lev. 7:31).65 Following this stage, /ā/ was raised further to /ɔ̄/, while /e/ and /o/ developed their respective allophones; this resulted in the qualitative overlap of /ɔ̄/ [ɔː] and /o/ [ɔ]/[ɔː]. Following the lengthening of all vowels in open syllables, including the virtually geminated /a/ in the definite article, the state prevailing in Tiberian Hebrew has been reached. It is not clear why the first consonant should be vocalized with ḥatef qameṣ. On the basis of the attested fsg /ḥrēbɔ̄/ [ħaʁeːˈvɔː] (e.g. in Lev. 7:10) one would have expected /hɛḥrēbōt/ [hɛː.ħaʁeːvɔːɔθ]. 65 There are a small number of attestations of what looks like compensatory lengthening involving /ḥ/, namely ‫כָל־הָ חַ י‬ ִ֠ ‫ומ‬ ִ /umikkol-hɔ̄ḥay/ [umikkɔl-ˈhɔːħaj] ‘and of every living thing’ (Gen. 6:19); ‫יטים‬ ָּֽׁ ִ ‫ וְ הָ ח ֲִר‬/whɔ̄ḥrīṭīm/ [vɔhɔː.ħaʁiːˈṭiːim] ‘and the girdles’ (Isa. 3:22); and ‫וְ הָ חַ מָ ִ ָּֽׁנים‬ /whɔ̄ḥammɔ̄nīm/ [vɔhɔː.ħam.mɔːˈniːim] ‘and the sun idols’ (Isa. 17:8); see Gesenius, 1910, p. 111. 64 128 Since all virtually geminated /a/ vowels have now undergone lengthening, lack of gemination of the gutturals can be said to result in compensatory lengthening in Tiberian Hebrew from a synchronic point of view, albeit accompanied by additional qualitative changes in certain cases. 7 Summary The different stages of the degemination of the gutturals and the development of variants of the definite article for different vocalic environments can be summarised as follows: 1. Degemination of /r/, /ʿ/ followed by compensatory lengthening /a/ > /ā/. 2. /aGā/ is no longer a stable environment, whether derived from degemination or some other source, resulting in dissimilation: /aGā/ > /ɛGā/. 3. /ʿ/ degeminates, with subsequent dissimilation of the vowel of the article (/a/ > /ɛ/) before a following unstressed /ā/; in all other cases, the vowel of the article is lengthened, /a/ > /ā/. 4. The environment for compensatory lengthening is becoming restricted to long stressed low and mid-high vowels (/ā́ /, /ḗ/). 5. /h/ degeminates, with subsequent dissimilation of the vowel of the article (/a/ > /ɛ/) before a following unstressed /ā/; before /ā́ / and /ḗ/, the vowel undergoes compensatory lengthening (/a/ > /ā/); in all other environments, the vowel is virtually geminated and undergoes no changes following degemination. 6. The environment permitting the dissimilation of /a/ > /ɛ/ is broadening beyond unstressed /ā/ to include stressed /ā́ / as well as long (unstressed) mid-high vowels (/ē/). 7. /ḥ/ degeminates, with subsequent dissimilation of the vowel of the article (/a/ > /ɛ/) before /ā/, /ā́ / and /ē/; in all other environments, the vowel is virtually geminated and undergoes no changes following degemination. 8. Virtually geminated /a/ is lengthened, as are all unstressed vowels in open syllables; certain vowels phonemes are raised and/or develop conditioned allophones, resulting eventually in the vowel system of Tiberian Hebrew. It seems likely that these changes, excluding the last one, took place at least partially while Hebrew was still a spoken language. The occurrence of various exceptions to the rules, as mentioned in the previous sections, are a hallmark of language change becoming frozen. 129 Table 6 presents a synopsis of the changes which the vowel of the article undergoes depending on the environment. It also highlights the gradual restriction of compensatory lengthening, with the development and spread of new environments conditioning different changes. Table 6. The changes of the vowel of the definite article Guttural r, ʾ ʿ h ḥ Environment Unstressed /ā/ a>ā a>ɛ a>ɛ a>ɛ Stressed /ā/ a>ā a>ā a>ā a>ɛ Unstressed /ē/ a>ā a>ā ? a>ɛ Stressed /ē/ a>ā a>ā a>ā a>a Other vowels a>ā a>ā a>a a>a The description of the synchronic distribution of the variants of the definite article will always be somewhat cumbersome. However, I believe that the description given here of the different changes in the vocalisation of the definite article as ways in which the spoken language dealt with the ongoing weakening of the gutturals and prevented their merger or loss, as happened in the Samaritan reading tradition, presents a convincing account of the diachronic processes involved. Further research is needed to verify the suggestions made in this paper. It is especially important to see whether similar or identical explanations can be applied to other environments affected by the loss of gemination where compensatory lengthening did or did not occur. This account has the distinct advantage of describing the dynamics involved in the development of the different conditioning environments. It also presents, I believe, a more accurate description of these environments by showing that vowels other than stressed and unstressed /ā/ may have been involved. References Al-Jallad, A., forthcoming, “What is Ancient North Arabian?”, in D. Birnstiel and N. Pat-El (eds), Re-Engaging Comparative Semitic and Arabic Studies. Wiesbaden: Harrassowitz. Blau, J., 1980, “Stages in the weakening of laryngeals/pharyngeals in Biblical Hebrew” (Hebrew), Lĕšonénu 41/1, pp. 32–39. ———, 1993, A Grammar of Biblical Hebrew, 2nd edn, Wiesbaden: Harrassowitz. ———, 2010, Biblical Hebrew Phonology and Morphology: An Introduction, Winona Lake: Eisenbrauns. Brønno, E., 1970, Die Aussprache der hebräischen Laryngale nach Zeugnissen des Hieronymus, Aarhus: Universitetsforlaget. Garr, W. R., 2004, Dialect Geography of Syria-Palestine, 1000-586 B.C.E., Winona Lake: Eisenbrauns. Gesenius, F. W., 1910, Hebrew Grammar, 2nd edn, Oxford: Clarendon Press. Gzella, H., 2006, “Die Entstehung des Artikels im Semitischen: eine ‘phönizische’ Perspektive”, JSS 51/1, pp. 1–18. 130 Huehnergard, J., 2005. “Features of Central Semitic”, in A. M. W. L. Gianto (ed.), Biblical and Oriental Essays in Memory of William L. Moran, Roma: Pontificio Istituto Biblico, pp. 155–203. ———, 2015, “Biblical Hebrew nominal patterns”, in J. M. Hutton and A. D. Rubin (eds), Epigraphy, Philology, and the Hebrew Bible: Methodological Perspectives on Philological and Comparative Study of the Hebrew Bible in Honor of Jo Ann Hackett, Atlanta: SBL Press, pp. 25–64. Joüon, P. and Muraoka, T., 1991, A Grammar of Biblical Hebrew, Rome: Biblical Institute Press. Khan, G., 2013a, A Short Introduction to the Tiberian Masoretic Bible and its Reading Tradition, 2nd edn, Piscataway: Georgias Press. ———, 2013b, “Biblical Hebrew: pronunciation traditions”, in G. Khan (ed.), 2013d, vol. I, pp. 341–352. ———, 2013c, “Compensatory lengthening”, in G. Khan (ed.), 2013d, vol. I, pp. 500–504. ——— (ed.), 2013d, Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill. ———, 2013e, “Pretonic lengthening: Biblical Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 224–229. ———, 2013f, “Resh: pre-Modern Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 384–389. ———, 2013g, “Shewa: pre-Modern Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 543–554. ———, 2013h, “Syllable structure: Biblical Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 666–676. ———, 2013i, “Tiberian reading tradition”, in G. Khan (ed.), 2013d, vol. III, pp. 769– 778. ———, 2013j, “Vowel length: Biblical Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 981–985. Lambdin, T. O., 1971, Introduction to Biblical Hebrew, Saddle River, N.J.: PrenticeHall. Macuch, R., 1969, Grammatik des samaritanischen Hebräisch, Berlin: de Gruyter. Mor, U., 2013, “Guttural consonants: pre-Masoretic”, in G. Khan (ed.), 2013d, vol. II, pp. 161–165. Pat-El, N., 2009, “The development of the Semitic definite article: a syntactic approach”, JSS 54/1, pp. 19–50. Rendsburg, G. A., 2013, “Phonology: Biblical Hebrew”, in G. Khan (ed.), 2013d, vol. III, pp. 100–109. Rubin, A. D., 2005, Studies in Semitic Grammaticalization, Winona Lake: Eisenbrauns. Stadel, C., 2017, “Gemination of /r/ in Samaritan Hebrew: a note on phonological diversity in Second Temple Period Hebrew”, Hebrew Studies 58/1, pp. 221–235. Stein, P., 2011, “Ancient South Arabian”, in S. Weninger (ed.), Semitic Languages: An International Handbook, Berlin: Mouton, pp. 1042–1073. Tropper, J., 2000, “Die Herausbildung des bestimmten Artikels im Semitischen”, JSS 46/1, pp. 1–31. 131 The Use of Dageš in the Non-Standard Tiberian Manuscripts of the Hebrew Bible from the Cairo Genizah SAMUEL BLAPP University of Zurich 1 Introduction1 The Tiberian Masoretes developed the Tiberian vowel and accentuation signs in order to represent their own inherited pronunciation and vocalisation tradition of biblical Hebrew (i.e. Standard Tiberian, abbreviated as ST in what follows). Besides the system of the Tiberian Masoretes, there were three other vocalisation systems (Babylonian, Palestinian and Samaritan), each of which had their own set of vowel and accentuation signs to represent their respective pronunciation traditions. While it became the preferred vocalisation system, after the death of the last Tiberian Masorete (Aharon ben Asher)2 the Tiberian vocalisation system and its pronunciation were separated, so that the Tiberian vowel signs were now used to represent pronunciation traditions other than the Tiberian.3 Biblical manuscripts where the Tiberian vowel signs are used to represent other pronunciation traditions will be referred to here as NonStandard Tiberian (NST) manuscripts.4 1 I am delighted to be able to congratulate Professor Geoffrey Khan on the occasion of his 60th birthday with this paper. I am indebted to him for his constant support as supervisor and his inspiring scholarship. This paper was written during my post-doctoral appointment at the University of Zurich funded by the Forschungskredit of the University of Zurich, grant. no. FK-16004. 2 See Ben-Hayyim, 2007, pp. 319–321. 3 Note that the Tiberian system was also used to vocalise the Mishnah and the Talmud, and is today the official vocalisation system used in Israel for Modern Hebrew. Its use, although defined by the Academy of the Hebrew Language, is no longer observed by the majority of Hebrew speakers. This is most likely due to colloquial speech (cf. Neudecker, 2017). 4 For the initial classification of manuscripts as NST, I used the Davis catalogue, which introduced this classification and also the term NST (Davis, 1978). Previous scholars have used the terms ‘Palestino-Tiberian’, ‘fuller Palestinian’, ‘extended Tiberian’, ‘non-traditional’, ‘nonconventional’ and ‘textus non-receptus’. Other terms used previously which connect to the Tiberian Masorete Ben-Naftali, however, are no longer justifiable, as the differences between him and the other Tiberian Masoretes were on minor issues and not at all as extensive as the differences in the NST manuscripts (Blapp, 2017, pp. 26–32). 132 The value of the NST manuscripts lies in the fact that they are primary witnesses for the pronunciation of the Hebrew language in the Middle Ages. As such they can help us to understand the development of the pronunciation of biblical Hebrew, and especially the developments which led to the loss of the ST pronunciation tradition. Additionally, they provide first-hand information about the use of the vocalisation system which can help us to explain and relate non-standard features in ST manuscripts, including the Leningrad Codex (abbreviated here as L),5 the Aleppo Codex (A) and manuscript BL Or 4445 (B).6 In this article I look at the use of the Tiberian dageš sign in NST manuscripts from the Cairo Genizah. The data presented here comes from my PhD thesis,7 and is based on six Non-Standard Tiberian Bible manuscripts from the Taylor-Schechter Genizah collection at Cambridge University Library. The manuscripts analysed here are T-S A11.1, T-S A12.1, T-S A12.10a, T-S A13.18, T-S A13.20 and T-S A13.35, all written on vellum between the 11th and the 13th centuries CE. I have used manuscripts from the Cairo Genizah here, since they are among the earliest witnesses for the use of the Tiberian system in a non-standard way. These manuscripts are most likely written in various different types of Oriental hand, and hence allow a glimpse into the Eastern manuscript tradition, as opposed to previous research on NST manuscripts which has mainly focused on medieval European manuscripts.8 Dageš, a single diacritical dot, has two main functions in ST: it indicates the doubling of a consonant, and it indicates the plosive pronunciation of the bgdkft letters (i.e. as /b/, /g/, /d/, /k/, /p/, /t/).9 The first function is often known as dageš forte, while the second function is dageš lene. These are Latin translations of the Hebrew terms dageš ḥazaq and dageš qal, which were first used by the medieval European Hebrew scholar David Qimḥi.10 There is a further function, dageš conjunctivum, which is not as well defined as the previous two and occurs much less frequently, though like dageš forte it can appear in all letters apart from the gutturals. Research on dageš conjunctivum is still very much undecided as to what it really indicates. Since it occurs in the first letter of a word, it is often argued that it has a conjunctive function (i.e. gemination due to prosody) or alternatively a disjunctive function (i.e. syllable division). Apart from a few minor differences, ST manuscripts use all types of dageš in the same way.11 5 These manuscripts are more formally: (L) National Library of Russia Evr. I B19a, see https://archive.org/details/Leningrad_Codex; (A) The Ben-Zvi Institute Jerusalem Israel Ms. 1, see http://www.aleppocodex.org/; (B) British Library Oriental Manuscript no. 4445, see http://www.bl.uk/manuscripts/FullDisplay.aspx?ref=Or_4445. 6 Blapp, forthcoming. 7 Blapp, 2017. 8 Blapp, 2017, pp. 26–32. 9 Golinets, 2017. 10 Golinets, 2017. 11 Golinets, 2017. 133 NST dageš12 is defined here as the use of the dageš sign in a way in which it is not used in ST manuscripts, in particular, in L.13 The most recent research into the history of the pronunciation of dageš shows that by the 11th century at the latest, dageš forte and dageš lene were both realised with gemination (i.e. as dageš forte), and thus their distinctive pronunciations had been lost.14 Since the NST manuscripts start to emerge around this time at the earliest, it is very likely that NST dageš was also realised with gemination. Such secondary gemination already occurs in isolated cases in ST manuscripts such as L,15 A or B. Some previous publications,16 however, did not differentiate between Tiberian vocalisation traditions, and thus missed the different developments of the usage of the Tiberian vowel signs and the implications of this.17 Additionally, it has to be noted that even the latest research on the use of dageš in Standard Tiberian is only conducted on the basis of a single manuscript, in most cases L. It has been shown, however, that L sometimes deviates from other ST manuscripts regarding the use of dageš.18 Thus, it will be necessary to compare a representative number of the available manuscripts, as well as medieval grammatical treatises, in order to gain a thorough understanding of dageš in the ST tradition in general, and to identify the tradition of marking dageš of each manuscript. Medieval European scholars such as David Qimḥi (11th–12th century CE) and Yequtiʾel ha-Naqdan (13th century CE) already mention the use of dageš in non-bgdkft letters in contexts which are not related to morphology; that is, 12 I will not deal with mappiq in this paper. Nor will I deal with NST dageš in bgdkft letters (i.e. plosive for spirant pronunciation), since its interpretation is not directly related to the scope of this paper (cf. Blapp, 2017, pp. 91–93, 115, 168, 194), and the evidence for it is very limited and thus its interpretation remains uncertain. NST dageš is marked in all bgdkft letters, and thus cannot be considered unilaterally to be influence from the Arabic speaking environment, since beth (bā in Arabic) did not have a corresponding spirant in the Arabic alphabet. 13 Since my data have been compared to L, for this paper (and my PhD) NST dageš is a dageš which specifically does not occur in L. 14 Professor G. Khan will soon publish a paper on the history of the pronunciation of dageš which he made available to me in draft form (Khan, forthcoming). In that paper he gives translations of the relevant passages from the medieval grammatical treatises on this issue. 15 Golinets, 2017. 16 Blake, 1943 and Prätorius, 1883. 17 These scholars did not distinguish between the ST and the NST manuscripts. They accepted the latter as of the same value as the ST manuscripts. As my research shows, the NST manuscripts no longer reflect the ST tradition, although they are vocalised with the Tiberian vowel signs. We can possibly use the NST manuscripts to explain developments in the ST tradition, but we cannot treat them equally when it comes to reliability, since the NST manuscripts reflect the individual pronunciation of each scribe, whereas the ST tradition reflects an inherited authoritative pronunciation tradition. Additionally, as I show in a forthcoming publication (Blapp, forthcoming), the use of the Tiberian vowel signs even in the ST manuscripts was not entirely consistent. Some of these cases reflect ST variants, whereas other cases reflect features which do not reflect the ST grammatical tradition. Thus, I believe that every manuscript has to be examined in terms of its purity in reflecting the ST tradition. The results of such an examination will show that every ST manuscript is different, and that developments within late ST manuscripts already show some non-standard features. 18 Golinets, 2017. 134 they mention NST dageš. Qimḥi says that all letters besides the bgdkft letters are pronounced either soft or hard according to whether or not they are preceded by a šəwa.19 Yequtiʾel ha-Naqdan explains that the pronunciation of dageš lene (‫)דגש קל‬,20 as he calls NST dageš, in the non-bgdkft letters was no longer heard, since people did not know how to pronounce it.21 Since dageš forte has a morphological or etymological implication,22 it is possible that Yequtiʾel chose the term dageš lene to refer to NST dageš because it has neither morphological nor etymological implications, and thus it is entirely secondary. This classification of Yequtiʾel, however, does not define the nature of the pronunciation of NS dageš (see below), since he and his contemporaries, as mentioned, no longer know how to pronounce any of the dageš at all. 2 The use of dageš in the NST manuscripts Examination of the data has shown that there are at least two different groups of manuscript in terms of the use of NST dageš. Based on an evaluation of their pronunciation features, the manuscripts in the corpus are divided into what I call the Tiberian-Palestinian23 manuscripts and a Tiberian-Babylonian manuscript, since the former show many Palestinian pronunciation features, whereas the latter shows many Babylonian pronunciation features.24 Based on 19 Chomsky, 2001, 18, §5b. Note that Qimḥi also refers in other places to the euphonic dageš, which supposedly occurs in pause or in letter with šəwa; see Chomsky, 2001, 18, §6b. 20 Like Qimḥi, Yequtiʾel ha-Naqdan uses the terms dageš ḥazaq (dageš forte) and dageš qal (dageš lene). While he does not mention Qimḥi, it is likely that Yequtiʾel borrowed these terms from him, since they were introduced in Qimḥi’s Sefer Mikhlol (cf. Golinets, 2017). 21 ‫ ורוב אנשי ארצנו לא ידעו להשמיע את הדגש‬... ‫אבל וַ֗ זַ֗ ַ֗ל ַ֗ט ַ֗מנַ֗ ַ֗ס ַ֗ק ַ֗ש הדגש הקל לא נשמע בהן ברוב מקומות כגון‬ ‫הקל הבא באותיות האלה‬, ‘but [in the letters] ‫ וַ֗ זַ֗ ַ֗ל ַ֗ט ַ֗מנַ֗ ַ֗ס ַ֗ק ַ֗ש‬dageš lene is not heard in them in many places, for example ‫( ְש ָּ֙ ָמר־לְ ָָ֔ך‬Ex. 34:11), ‫( וְ יִ ְש ְמ ָ֣עו‬Josh. 7:9), ֶׁ֙‫( נִ ְבקְ עו‬Gen. 7:11), ‫( וְ אַ ְח ָ ָּֽׁלמָ ה׃‬Ex. 28:19), ‫( ַרחְ ָ ֶ֑מּה‬Gen. 29:31) and ‫( בַ עְ ָ ָ֣לּה‬Deut. 24:4) [these examples can be found on p. 165 with the addition ‫‘ ודומיהם רבים‬and many similar cases’], and many people from our land do not know [how] to realise dageš lene which occurs in these letters’ (Yarqoni, 1985, pp. ‫ קה‬and 165, my translation). Note that Yarqoni does not mark the dageš about which Yequtiʾel is talking, and so it is not entirely clear to which dageš Yequtiʾel is referring. It is, however, most likely that in his manuscript dageš was marked, at least in the examples two to five, after the silent šəwa in one of the letters he mentioned earlier in the same sentence (i.e. ‫)וַ֗ זַ֗ ַ֗ל ַ֗ט ַ֗מנַ֗ ַ֗ס ַ֗ק ַ֗ש‬. In the first example it was most likely marked after the maqqef in lamed. 22 Golinets, 2017. 23 Note that the term ‘Palestino-Tiberian’ (e.g. Heijmans, 2017b) is the most commonly used term to refer to all these manuscripts in recent research. As I have shown in my PhD, however, this does not take into account that there are manuscripts with a majority of Babylonian features, and that the number of ST pronunciation features in any manuscript is significant, so that we cannot simply say that these manuscripts reflect the Palestinian pronunciation tradition with Tiberian vowel signs. The data is much more diverse (cf. e.g. Blapp, 2017, pp. 202–207). 24 Further research of the Palestinian tradition especially is needed in order to understand and situate it in the history of the development of the vocalisation systems of Hebrew. The Palestinian vocalisation tradition exhibits a significant number of Tiberian as well as Babylonian pronunciation features, which calls into question its status as an independent tradition (Blapp, 135 the evaluation of the vocalisation of these manuscripts, it is also possible to state that every manuscript reflects the pronunciation of its scribe, since none of these manuscripts exhibit the exact same vowel interchanges. While each manuscript is very different, the manuscripts of the Palestinian group share features such as frequent interchanges of qamaṣ with pataḥ and seghol with ṣere.25 The single manuscript from the Babylonian group mainly shows interchanges of pataḥ with seghol/ṣere.26 Despite the fact that we are able to identify these two groups, it is important to note that there are also a limited number of Palestinian features found in the Tiberian-Babylonian manuscript, as well as Babylonian features in the Tiberian-Palestinian manuscripts. In what follows the data are arranged according to the features of the preceding syllable (closed or open, in the preceding word or the same word) and the accent (after a disjunctive or conjunctive accent, when dageš occurs after an open syllable in word-initial position). Note that according to the ST tradition, dageš lene was only marked in bgdkft letters after closed syllables and after open syllables with a disjunctive accent.27 At a secondary level the data here are categorised according to the manuscript and the phoneme in which NST dageš occurs. I will only give a single example for each category because of the limitations of space.28 However it is important not only to see the distribution of NST dageš in the different letters, but also the respective frequency of occurrence, and therefore data is given in the appendices to provide an overview of the distribution and frequency of the features related to dageš in each manuscript. 2.1 The Tiberian-Palestinian manuscripts The frequent replacement of the vowel qamaṣ gadol with pataḥ and vice versa in these manuscripts, and similarly the frequent replacement of seghol with ṣere and vice versa, suggest a Palestinian background for the scribes, since the interchange of these sounds is an important characteristic of the Palestinian 2017, pp. 224–225). Note that I chose to place ‘Tiberian’ in front of ‘Palestinian’ and ‘Babylonian’ in naming these categories. I did this for two reasons: first, these manuscripts are vocalised with Tiberian vowel signs, and second, many features in both manuscript groups still reflect the ST pronunciation, and thus they seem to be only influenced by the ‘Palestinian’ and ‘Babylonian’ pronunciation traditions. 25 The reason for these interchanges is that qamaṣ gadol was pronounced like pataḥ and vice versa, and seghol was pronounced like ṣere and vice versa (Heijmans, 2017a). Note that in ST every vowel was pronounced differently (Khan, 2017a). 26 The reason for this interchange is that pataḥ and seghol, and sometimes ṣere, were pronounced alike (Khan, 2017b). 27 Thus Yequtiʾel ha-Naqdan’s use of the term dageš lene (see note 21) no longer reflects the ST tradition, and shows that he himself most likely no longer knew how to pronounce it properly. 28 See my PhD for the full list of examples (Blapp, 2017). 136 pronunciation tradition.29 Additionally, the data provided by Fassberg30 on the use of the Palestinian sign for dageš in the Palestinian reading tradition is similar to the use of NST dageš in the NST manuscripts. The Palestinian sign for dageš occurs where ST uses a dot to indicate dageš forte and sometimes dageš lene, dageš conjunctivum, and also mappiq, or to denote the diacritical dot above ST sin. It also occurs in letters other than the bgdkft letters when these are the first letters of a syllable (the syllable onset), both word-initially and word-internally, and this occurrence cannot be identified as any kind of ST dageš, mappiq or other diacritical dot.31 Most of the occurrences are either in the first letter (syllable onset) of a word-initial syllable, or else word-internal in the first letter of a syllable (syllable onset) which follows a closed syllable.32 Similarly, in the NST manuscripts analysed here, NST dageš is most frequently marked in the first letter of a syllable in either word-initial or wordinternal position (see the appendices for frequencies). The data33 below have been selected in order to give an overview of the use of NST dageš in the manuscripts in the corpus. In this paper, I will show that: • • • • In the Tiberian-Palestinian manuscripts, NST dageš is marked in the first letter following a syllable within the same word which is closed (e.g. ‫ נָחְ רֹו‬from T-S A11.1); see section 2.1.2. NST dageš does, however, also occur in the first letter of a word following a word with an open or closed final syllable (e.g. ‫ יִ ָגָּֽׁש‬from T-S A13.18); see section 2.1.1. Unlike in the Tiberian-Babylonian manuscript, there are only a few occurrences of NST dageš in the first letter of a syllable following an open syllable in the same word (e.g. ֶׁ֙‫ בַ שָ מַ י ִם‬from T-S A12.1); see section 2.1.3. There are also examples of NST dageš in the last letter of a word-final closed syllable which follows a vowel (e.g. ‫ בְ ָ֥יֹום‬from T-S A12.1); see section 2.1.4. 2.1.1 NST dageš in the initial letter of a word The following data have been selected out of a larger set of data, in order to outline the use of NST dageš in the first letter of a word. NST dageš in the first letter of a word-initial syllable can occur in all letters apart from the gutturals in the Tiberian-Palestinian manuscripts.34 ST dageš in an initial bgdkft 29 See e.g. Heijmans, 2017a. Fassberg, 1987, pp. 75–103. 31 Fassberg, 1987, pp. 78–89. 32 Fassberg, 1987, pp. 80, 85–86 (3a–c), 88 (2–3). 33 See section 1 for details of the corpus. 34 Note that in the Tiberian-Babylonian manuscript, NST dageš occurs sometimes in this position in the gutturals he and ḥeth (see section 2.2). NST mappiq, on the other hand, is marked in consonantal ʼalef regardless of the syllabic environment in both the Tiberian-Babylonian and the Tiberian-Palestinian manuscripts (Blapp, 2017, pp. 207–209). 30 137 letter (i.e. not a dǝḥiq)35 occurs only after a closed syllable or a final open syllable with a disjunctive accent. On the other hand, a dǝḥiq in ST manuscripts occurs only after an open syllable with conjunctive accent.36 The use of NST dageš in the corpus analysed here exceeds both of these limitations, since it occurs in initial positions after all kinds of syllables and accents (see sections 2.1.1.1–2.1.1.3 below). There is no evidence that stress influenced the placing of NST dageš. It rather marks syllable boundaries to guide the correct understanding and reading of the text. This marking was most likely accompanied by enforcing the pronunciation (i.e. gemination) of the letter with NST dageš. Based on the fact that either rafe or NST dageš is used on all letters which are the first letter of a word-initial syllable, it is likely that there was a phonetic distinction between NST dageš and rafe, and thus NST dageš is more than just a graphical syllable boundary marker.37 2.1.1.1 NST dageš after a closed syllable38 In the ST tradition, the bgdkft phonemes would receive dageš lene in these cases.39 T-S A11.1 ‫ וְ לא‬40 ‫לִ וי ָָ֣תן‬ ‫ֵץ‬ ֛ ‫י ְָאבֵ ר נ‬ (L ‫ ִ֭נָהָ ר ָ֣לא‬Job 40:23 ‘not’) (L ‫ ִת ְמ ָ֣ש ְך לִ וְ י ָ ָָ֣תן‬Job 40:25 ‘Leviathan’) (L ‫ ַ ָּֽׁיאֲבֶ ר־נֵ ֶ֑ץ‬Job 39:26 ‘the hawk will fly’) T-S A12.1 ‫זְרע ֶ ָּֽׁתיהָ ׃‬ ‫לְ נַפְ ֶ ָּֽׁשךָ׃‬ ‫מֵ בִ יָ֣ ש‬ ‫נִ פְ גָ ֶ֑שו‬ ‫סָ דִ יָ֣ ן‬ ‫יקים‬ ִ ‫צִַ֭ ִד‬ ‫ָ ָ֣קמו‬ ְ‫ָ֥של‬ (L ‫עֹותיהָ ׃‬ ָּֽׁ ֶ ‫אַמץ זְר‬ ֵ ַ֗ ‫ ֝ו ְת‬Prov. 31:17 ‘her arms’) (L ‫ מַ עֲדַ נִ ָ֣ים לְ נַפְ ֶ ָּֽׁשךָ׃‬Prov. 29:17 ‘to your (msg) soul’) (L ‫ ְ֝מ ֻׁ֯ש ַָ֗לח מֵ ִ ָ֥ביש‬Prov. 29:15 ‘he, who will bring shame to’) (L ‫ ְתכ ִ ָָ֣כים נִ פְ גָ ֶ֑שו‬Prov. 29:13 ‘they met together’) (L ‫ עִ ם־זִקְ נֵי־אָ ֶָּֽׁרץ׃ סָ ִ ָ֣דין‬Prov. 31:24 ‘linen garment’) (L ‫ תֹוע ֲַבָ֣ת צִַ֭ דִ יקִ ים‬Prov. 29:27 ‘righteous’) (L ‫ ת ֻׁ֯א ֵ ָּֽׁכל׃ ָ ָ֣קמו‬Prov. 31:28 ‘they rose up’) (L ‫ וְ גַ ֛ם של־‬Ruth 2:16 ‘to pull out’) For dǝḥiq see section 2.1.1.3 below. Ofer, 2017. 37 Khan, forthcoming. 38 I shall not distinguish whether a closed final syllable was stressed with a conjunctive or a disjunctive accent, since in either case the following bgdkft phoneme would already have received dageš in the ST tradition. 39 Golinets, 2017. 40 Note the extra conjunctive waw, which is not attested in the ST tradition. This is one of the rare cases where the NST consonantal text deviates from ST consonantal text, apart from defective and plene spellings. 35 36 138 T-S A12.10a41 ‫בֹורָך‬ ֶ ‫* ִמ‬ ‫*נְ כרי׃‬ []‫*צ‬ (L ‫בֹור ָך‬ ֶ֑ ֻׁ֯ ֶ ‫ֿה־מיִ ם ִמ‬ ָ֥ ַ ֵ‫ ְשת‬Prov. 5:15 ‘your cistern’) (L ‫ בְ ֵבָ֣ית נָכְ ָּֽׁ ִרי׃‬Prov. 5:10 ‘foreign’) (L ‫ כְ מַ ֵהָ֣ר צִ ָ֣פֹור‬Prov. 7:23 ‘bird’) T-S A13.18 ‫ְרֹועי‬ ָ֥ ִ ‫־ז‬ ‫טֹוב‬ ‫*לַיהוָה‬ ָּֽׁ ‫ִ ָּֽׁמי־‬ ‫ַ ָּֽׁנע ָ ֲָ֥רץ‬ ‫ֶ ָּֽׁסלָֿה׃‬ ‫ָ ָּֽׁצ ָ֣פֹון‬ ‫קְ ד ִ ָּֽׁשים׃‬ ‫יתי‬ ִ ‫ִש ִ ָ֣ו‬ (L ‫ְרֹועי‬ ָ֥ ִ ‫ אַף־ז‬Ps. 89:22 ‘my arm’) (L ‫ הַ שַ ָ ָּֽׁבת׃ ַ֗טֹוב‬Ps. 92:2 ‘good’) (L ‫ ַיע ֲָ֣ר ְך לַיהוָ ֶ֑ה‬Ps. 89:7 ‘to the Lord’) (L ‫ צְ בָ ַ֗אֹות ִ ָּֽׁמי־‬Ps. 89:9 ‘who’) (L ‫ ֵ ָ֣אל ִַ֭נע ֲָר ֻׁ֯ץ‬Ps. 89:8 ‘he, who is feared’) (L ‫ ֶנ ֱא ָ ָ֥מן ֶ ָּֽׁסלָה׃‬Ps. 89:38 ‘Selah’) (L ‫ ְיסַ ְד ָ ָּֽׁתם׃ צָ ָ֣פֹון‬Ps. 89:13 ‘north’) (L ‫ בִ קְ ַהָ֥ל קְ ד ִ ָּֽׁשים׃‬Ps. 89:6 ‘the saints’) (L ‫יתי‬ ִ ‫ ַו ַ֗תאמֶ ר ִש ִ ָ֣ו‬Ps. 89:20 ‘I have levelled’) T-S A13.35 ‫זֶ֑ את‬ ‫*מ ְש ַָ֗פט‬ ִ֝ ‫נִ ְש ְמ ָ֣דו‬ ‫ָצֶ֑ר‬ ‫קַ ְרנֵ ָ֥י‬ (L ‫ ל ַ ָָ֣דעַת ֶ֑זאת‬Ps. 73:16 ‘this (fsg)’) (L ‫ א ֵהִׂ֤ב ִמ ְש ַָ֗פט‬Ps. 37:28 ‘judgement’) (L ‫נִש ְמ ָ֣דו‬ ְ ‫ ָּֽׁ ִ֭ופ ְשעִ ים‬Ps. 37:38 ‘they (mpl) were destroyed’) (L ‫ יְ ָח ֶָ֣רף ָצֶ֑ר‬Ps. 74:10 ‘adversary’) (L ‫ וְ כָל־קַ ְרנֵ ָ֣י‬Ps. 75:11 ‘my horn’) 2.1.1.2 NST dageš after an open syllable with disjunctive accent42 In the ST tradition the bgdkft phonemes would receive dageš lene in these cases.43 T-S A12.1 ‫ז ֶֿ֛ה‬ ‫טִָ֭ עְ ָמֿה‬ ‫לָעַ ָ֥ד‬ ‫מֵ ִ ֵ֖איר‬ ‫נַפְ ֶ֑שֹו‬ (L ‫ וְ עַד־ ַָ֔עתָ ה זֶ ֛ה‬Ruth 2:7 ‘this (m)’) (L ‫עֹותיהָ ׃ טִָ֭ ע ֲָמֿה‬ ָּֽׁ ֶ ‫ זְר‬Prov. 31:18 ‘she perceived’) (L ‫ ֝ ִכ ְס ַ֗אֹו לָעַ ָ֥ד‬Prov. 29:14 ‘forever’) (L ‫ נִ פְ גָ ֶ֑שו ֵ ִׂ֤מ ִאיר־‬Prov. 29:13 ‘he, who gives light’) (L ‫ שֹונֵ ָ֣א נַפְ ֶ֑שֹו‬Prov. 29:24 ‘his soul’) T-S A12.10a ‫*מֹור‬ ַ‫*נִ לְ כ ַֻׁ֯דת‬ ‫ֶנגַע‬ (L ‫ ִמ ְש ָכ ִ ֶ֑בי ָ֥מר‬Prov. 7:17 ‘myrrh’) (L ָ‫י־פיָך ֝ ִנלְ ַַ֗כדְ ת‬ ֶ֑ ִ ‫ בְ ִא ְמ ֵר‬Prov. 6:2 ‘you are caught’) (L ‫ ַי ֲע ֶ ָּֽׁש ָנה׃ ֶ ָּֽׁנגַע־‬Prov. 6:33 ‘wound’) 41 The asterisk on examples below indicates that only the consonants are legible, while accentuation and vocalisation are not legible. 42 Yeivin lists similar cases as ‘conjunctive’ dageš (Yeivin, 1980, p. 292, §407). 43 Golinets, 2017. 139 T-S A13.18 ‫ז ְָ֣כר‬ ‫ָ֣לא‬ ‫*מ ַ֗אד‬ ְ֝ ‫ֶנ ֱא ֶ ָ֥מנֶת‬ ‫*סביבותיך׃‬ ְ ‫צִ נָ ֵֿ֖ה‬ ‫ְשאֵ ִ ָ֥תי‬ ‫שַ ֛סוהו‬ (L ‫ בֶ אֱמונ ֶ ָָּֽׁתךָ׃ ז ְָ֣כר‬Ps. 89:51 ‘remember! (msg)’) (L ‫ ובֶ ן־ ֝ ַעוְ ַָ֗לה ָ֣לא‬Ps. 89:23 ‘not’) (L ‫ יְ הוָ ֶ֑ה ְ֝מ ַ֗אד‬Ps. 92:6 ‘very’) ִַ֗ ‫ ֝ובְ ִר‬Ps. 89:29 ‘it (fsg), which is made firm’) (L ‫יתי ֶנ ֱא ֶ ָ֥מנֶת‬ (L ‫יבֹותיךָ׃‬ ָּֽׁ ֶ ִ‫ ֝ו ֱא ָּֽׁמונ ְָת ַָ֗ך ְסב‬Ps. 89:9 ‘round about you (msg)’); (L ‫ תֶ חְ ֶ ֶ֑סה צִ נָ ֵ֖ה‬Ps. 91:4 ‘shield’) (L ‫יך ְש ֻׁ֯ ֵא ִ ָ֥תי‬ ָ ‫ עֲבָ ֶ ֶ֑ד‬Ps. 89:51 ‘my bearing’) (L ‫ ְמחִ תָ ֿה׃ שִַ֭ סהו‬Ps. 89:42 ‘they plundered him’) 2.1.1.3 NST dageš after an open syllable with conjunctive accent (NST dǝḥiq?) In general, dǝḥiq is a case of dageš after a word with a final open syllable which is stressed on the penultima with a conjunctive accent.44 In ST it usually occurs in the initial stressed syllable of a new word following an unstressed seghol or qamaṣ.45 There may be further cases, but this is still uncertain – since scholarship on biblical Hebrew has not taken into account that Tiberian manuscripts have to be classified according to their stage of development of the Tiberian system,46 they have not yet been able to clearly define dǝḥiq. The use of dageš in the examples below thus may reflect an extension of the dǝḥiq rules, and we could define NST dǝḥiq as occurring after a conjunctive accent following all kind of vowels (the NST extension), while NST dageš occurs after closed syllables and open syllables with disjunctive accents. T-S A12.1 ‫ָ֣טֹוב‬ ‫־ל ַָמ ֵ֖ד ְִתי‬ ‫ָ֥י־נָּֽׁשֶ ר׃‬ ָ ֵ‫בְ נ‬ (L ‫י־טֹוב‬ ָ֣ ִ‫ כ‬Prov. 31:18 ‘good’) (L ‫ ָּֽׁלא־ל ַ ָָ֥מ ְד ִתי‬Prov. 30:3 ‘I learned’) (L ‫ֵי־נָּֽׁשֶ ר׃‬ ָ ‫ בְ נ‬Prov. 30:17 ‘the children of the eagle’) T-S A12.10a ‫* ַלגַנב‬ ‫נִ ְרוֶ ָ֣ה‬ (L ‫ לא־י ָָ֣בוזו ִַ֭ל ַגנָב‬Prov. 6:30 ‘(preposition +) the thief’) (L ‫ לְ כָ ִׂ֤ה ְנִרוֶ ָ֣ה‬Prov. 7:18 ‘let us take our fill’) T-S A13.18 ‫יִ ָגָּֽׁש‬ (L ‫ ָ֣לא י ִָגָּֽׁש׃‬Ps. 91:7 ‘it (msg) will approach’) 44 Ofer, 2017; Yeivin, 1980, pp. 289–290, §403. Ofer, 2017. 46 Blapp, forthcoming. Note that Yeivin gives further examples (Yeivin, 1980, pp. 290ff.), where it occurs after vowels other than seghol and qamaṣ, but it is not clear from which manuscripts he has taken them. This is a very important question, since he sometimes gives examples from late ST manuscripts such as the Cairo Codex of the Prophets, where the use of dageš is similar to the NST manuscripts. The Hebrew Bible manuscripts with Babylonian vocalisation (Yeivin, 1985) and those with Palestinian vocalisation (Revell, 1970) have already been categorised according to their stage of development. 45 140 T-S A13.35 ‫*לא‬ ‫לְ פִ י‬ ‫נִ ְקלֶ ֶֿ֑ה‬ (L ‫ בְ לִ ֶ֑בֹו ֵ֖לא‬Ps. 37:31 ‘not’) (L ‫ אֶ ְש ְמ ָ ָ֥רה לְ ִ ָ֥פי‬Ps. 39:2 ‘my mouth’) (L ‫ מָ לְ ָ֣או ִנקְ לֶ ֶ֑ה‬Ps. 38:8 ‘he, who burns’) NST dageš in the first letter of a syllable after a closed syllable word-internally In ST manuscripts, the bgdkft letters after a closed syllable word-internally would have receive dageš lene in many cases.47 2.1.2 T-S A11.1 ‫*יִ ְת ָלכָדו‬ (L ‫יִתל ָ ַָּֽׁכדו‬ ְ Job 38:30 ‘they are frozen’) ‫( *ולהצְ מיח‬L ַ‫ ו֝ לְ הַ צְ ִַ֗מיח‬Job 38:27 ‘to cause to grow’) ‫*נָחְ רֹו‬ (L ‫ נַחְ ָ֣רֹו‬Job 39:20 ‘his snorting’) T-S A12.1 ‫בְ בַ ְרזֶ ָ֣ל‬ ‫ִיבְ ָ ָּֽׁטח׃‬ ‫֝ ִי ְש ַ֗ ַמע‬ ‫־ת ְמנַ ֵ֖ע‬ ִ ‫מַ חְ ֶ֑סֹור‬ ‫וַיְ צַ ֩ו‬ ֶׁ֙‫חֶ ְלקַ ת‬ ‫לְ נַפְ ֶ ָּֽׁשךָ׃‬ (L ‫ בְ בַ ְרזֶ ָ֣ל‬Prov. 27:17 ‘with iron’) (L ‫ יִבְ ָ ָּֽׁטח׃‬Prov. 28:1 ‘he will trust’) (L ‫ ֝ ִי ְש ַ֗ ַמע‬Prov. 29:24 ‘he will hear’) (L ‫ל־ת ְמנַ ָ֥ע‬ ִ ַ‫ א‬Prov. 30:7 ‘do not deny’) (L ‫ מַ חְ ֶ֑סֹור‬Prov. 28:27 ‘lack’) (L ‫ וַיְ צַ ֩ו‬Ruth 2:15 ‘and he commanded’) (L ֶׁ֙‫ חֶ לְ קַ ת‬Ruth 3:18 ‘a portion of land’) (L ‫ לְ נַפְ ֶ ָּֽׁשָך׃‬Prov. 29:17 ‘to your (msg) soul’) T-S A12.10a ‫יִ ְמצַ א‬ (L ‫ ִי ְמ ָצֶ֑א‬Prov. 6:33 ‘he shall get’) T-S A13.18 ‫אֶ בְ טַ חְ ־‬ ‫ְ֜ונִ ְש ְמ ַ֗ ָחֿה‬ ‫תֶ חְ ֶ ֶ֑סֿה‬ ָ֥‫וָּֽׁ בְ ִצדְ ָּ֙ ָק ְתָך‬ ‫מַ חְ ְשבו ֶ ָּֽׁתיךָ׃‬ (L ‫ אֶ בְ טַ ח־‬Ps. 91:2 ‘I will trust’) (L ‫נִש ְמ ַ֗ ָחֿה‬ ְ ‫ ְ֝ו‬Ps. 90:14 ‘and let us rejoice’) (L ‫ תֶ חְ ֶ ֶ֑סה‬Ps. 91:4 ‘you (msg) will take refuge’) (L ָ֥‫ ובְ צִ ְדקָ ְת ָך‬Ps. 89:17 ‘and in your (msg) righteousness’) (L ‫ מַ חְ ְשב ֶֻׁ֯ ָּֽׁתיךָ׃‬Ps. 92:6 ‘your (msg) thoughts’) T-S A13.35 ‫*אבְ ָ ֶ֑טח‬ ‫נִ ְקלֶ ֶֿ֑ה‬ ‫מ ְזמֹור‬ ‫ומ ְתנַקֵ ם‬ ִ ‫נַפְ ִשי‬ (L ‫ אֶ בְ ָ ֶ֑טח‬Ps. 44:7 ‘I will trust’) (L ‫ נִ קְ לֶ ֶ֑ה‬Ps. 38:8 ‘he, who burns’) (L ‫ ִמז ְֵ֖מֹור‬Ps. 38:1 ‘Psalm’) (L ‫ומ ְתנ ֵ ַָּֽׁקם׃‬ ִ Ps. 44:17 ‘he, who avenges’) (L ‫ ַ֝נפְ ַ֗ ִשי‬Ps. 41:5 ‘my soul’) 47 Exceptions are the cases where the bgdkft letter is preceded by the so-called šəwa medium (Blau, 2010, 114–115, §3.5.6.3.6) which is, however, not a phonetic reality but rather a term used to describe this phenomenon (Blau, 2010, 115, §3.5.6.3.6n). 141 Exceptions: NST dageš in the first letter of a syllable after an open syllable word-internally There are only a few examples of this phenomenon in the Tiberian-Palestinian manuscripts where NST dageš occurs after open syllables within a word. There is no consistent theory of how this phenomenon could be explained. Two of these cases in T-S A12.1 ( ִ‫ בְ ַעלָ ֶ֑ה‬and ‫)ומעַלִ יָ֣ ם‬ ְ could be explained as the result of a change in the syllable structure, since ST silent šəwa could have been confused with the actual preceding vowel, thus resulting in a new syllable structure. A further example in T-S A12.1 occurs in a morpheme whose morphological form has been changed (‫)ו ֵַ֖ת ֵרד‬. All of the other examples are unique, and so it is also possible that they are mistakes or that dageš in these cases is just a speck on the vellum.48 The addition of rafe on some examples could support the theory that they were mistakenly marked with NST dageš.49 The statistics suggest that the marking of NST dageš in this context relates to sonority, since the majority of cases are attested in the sonorant letters lamed and mem. 2.1.3 T-S A11.1 ‫*ויחלֵק‬ (L ‫ י ָ ֵָ֣חלֶק‬Job 38:24 ‘it is divided’) T-S A12.1 ִ‫בְ ַעלָ ֶ֑ה‬ ‫ומעַלִ יָ֣ ם‬ ְ ֶׁ֙‫בַ שָ מַ י ִם‬ ‫ו ֵַ֖ת ֵרד‬ (L ‫ בַ עְ לָ ֶּ֑ה‬Prov. 31:11 ‘her husband’) (L ‫ ומַ עְ ִ ָ֥לים‬Prov. 28:27 ‘he, who hides’) (L ֮‫ בַ שָ מַ יִ ם‬Prov. 30:19 ‘in the skies’) (L ‫ ו ֵ ֵַ֖ת ֶרד‬Ruth 3:6 ‘and she went down’) T-S A12.10a ‫שוטר‬ ָ֥ ֵ (L ‫ ש ֵ ָ֥טר‬Prov. 6:7 ‘officer’) T-S A13.18 ‫ִי ְש ָּֽׁמרו‬ (L ‫יִש ָּֽׁמרו׃‬ ְ Ps. 89:32 ‘they will keep’) 2.1.4 NST dageš in the last letter of a word-final syllable NST dageš is marked only rarely in word-final syllables. The evidence shows that it mainly occurs in the sonorant letters lamed and mem. T-S A12.1 ‫בְ ָ֥יֹום‬ (L ‫ בְ יֹום־‬Ruth 4:5 ‘at a day’) T-S A13.35 ‫*כַל‬ ‫ְתבָ הְ ֵ ָּֽׁלם׃‬ (L ‫ כָל־הַ֝ יַ֗ ֹום‬Ps. 38:7 ‘all of’) (L ‫ ְתבַ ה ֵ ֲָּֽׁלם׃‬Ps. 83:16 ‘you (msg) will terrify them (mpl)’) 48 Such misleading specks have led to misinterpretations, for example in the editing of the Biblia Hebraica Stuttgartensia (Golinets, 2013, p. 248). 49 Morag, 1959, p. 225, §2.351. 142 2.2 The Tiberian-Babylonian manuscript T-S A13.20 As noted above, manuscript T-S A13.20 has been classified as Tiberian-Babylonian on the basis of its major vowel interchanges (i.e. replacement of pataḥ with seghol and vice versa).50 It is the only manuscript in the corpus examined here which shows a distinctive number of Babylonian features. Like in the Tiberian-Palestinian manuscripts, in the Tiberian-Babylonian manuscript the use of NST dageš in word-initial position after a closed syllable is more frequently attested than the use of NST dageš after a closed syllable word-internally, and indeed this latter is attested only sporadically in the Tiberian-Babylonian manuscript. A far greater number of examples of NST dageš occur after open syllables within a word. (As before, only a single example per letter and context is given here in the text; for frequency data, see appendix B.) 2.2.1 NST dageš in the first letter of a word-initial syllable Like in the Tiberian-Palestinian manuscripts, the marking of NST dageš in the first letter of a word-initial syllable is also attested in the Tiberian-Babylonian manuscript across a wide variety of letters. The number of cases, however, is lower than that of examples of NST dageš after an open syllable morphemeinternally (see appendix B). 2.2.1.1 NST dageš after a closed syllable T-S A13.20 ‫ו ָ ֶָ֑א ֶרץ‬ (L ‫ שָ ַ ָ֣מיִ ם ָואֶָ֑ ֶרץ‬Ps. 69:35 ‘and earth’) ‫ַחיִ יֶ֑ ם‬ (L ‫ ִמ ֵ ָ֣ספֶר חַ יִ ֶ֑ים‬Ps. 69:29 ‘of the living’) ‫*לָך‬ (L ִׂ֤‫ ־יְרושָ לָ ֶ֑ ִם לְ ָך‬Ps. 68:30 ‘for you (msg)’) ‫( שֵ עָ ֶ֑ר ִמתהַ ל ְֵך‬L ‫ שֵ עָ ֶ֑ר ִ֝מ ְתהַ ֵַ֗ל ְך‬Ps. 68:22 ‘he, who walks about’) ‫צִ וַה‬ (L ‫ נַפְ תָ ִ ָּֽׁלי׃ צִ וָ ָֻׁ֥֯ה‬Ps. 68:29 ‘he commanded’) ‫קֳ דְ ָ֣ק ְד‬ (L ‫יְבָ֥יו קָ ְד ָ֥קד‬ ָ ‫ ַ֫א‬Ps. 68:22 ‘scalp of’) ‫ָ ָּֽׁשי׃‬ (L ‫ ְמל ִ ָָ֣כים ָ ָּֽׁשי׃‬Ps. 68:30 ‘gift’) 2.2.1.2 NST dageš after an open syllable with disjunctive accent T-S A13.20 []‫ָּֽׁ ַּה‬ ‫ָו ֲאקֲ וֶ ָֿ֥ה‬ ‫יִ ָ֣סגו‬ ‫*לַמָ וָ֥ ת‬ ‫יבים‬ ָ֥ ִ ‫ֵ ָּֽׁמא‬ (L ‫ וָּֽׁ ְת ַפלְ ֵ ֶ֑טנִ י הַ ֵ ָּֽׁטה־‬Ps. 71:2 ‘stretch out! (msg)’) (L ‫ ָ ָּֽׁואַָ֫ ָֻׁ֥֯נושָ ֿה ָוא ֻׁ֯ ֲַקוֶ ָֿ֣ה‬Ps. 69:21 ‘and I hoped’) (L ‫ ַ֫ ַנפְ ִ ָ֥שי ִי ָ֣סגו‬Ps. 70:3 ‘they (mpl) will be turned’) (L ‫ ֲאדנָ ֶ֑י ֝ ַל ַ֗ ָמוֶת‬Ps. 68:21 ‘from death’) (L ‫יְבים‬ ָ֥ ֻׁ֯ ִ ‫ֶ֑יך מֵ א‬ ָ ‫ כְ ָֻׁ֯ל ֶב‬Ps. 68:24 ‘from the enemies’) 50 Blapp, 2017, pp. 149–151. For more details on the Babylonian vocalisation features, see Khan, 2017b. 143 2.2.1.3 NST dageš after an open syllable with conjunctive accent T-S A13.20 ‫מַ לְ ִ ָ֥כי‬ []‫כת‬ ִ ‫נִ סְ ַ ָ֣מ‬ (L ‫ אֵ ִ ֵ֖לי מַ לְ ִ ָ֣כי‬Ps. 68:25 ‘my king’) (L ‫ עָלֶ ִׂ֤יָך׀ נִ סְ ַ֬ ַמכְ ִתי‬Ps. 71:6 ‘I braced’) NST dageš in the first letter of a syllable after a closed syllable word-internally This feature is attested on a larger number of letters in the Tiberian-Palestinian manuscripts (see section 2.1.2), but it is not frequently attested in the TiberianBabylonian manuscript. 2.2.2 T-S A13.20 ‫ִתקְ ו ִ ֶָ֑תי‬ []‫*וכְ לִ ָ ֶ֑מ‬ ‫יש ָ ֶ֑מחו‬ ְ ִ‫ו‬ ‫נ ְָ֔ג ִנֶ֑ים‬ (L ‫ ִתקְ ו ִ ֶָ֑תי‬Ps. 71:5 ‘my hope’) (L ‫ וכְ לִ מָ ִ ֶ֑תי‬Ps. 69:20 ‘my disgrace’) (L ‫יִש ָ ֶ֑מחו‬ ְ Ps. 69:33 ‘and they will be glad’) (L ‫ נגְ ִ ֻׁ֯ ֶ֑נים‬Ps. 68:26 ‘those, who play music’) NST dageš in the first letter of a syllable after an open syllable word-internally This feature occurs only sporadically in the Tiberian-Palestinian manuscripts (see section 2.1.3), but it is very frequently attested in the Tiberian-Babylonian manuscript. 2.2.3 T-S A13.20 ‫יכֹותיךָ׃‬ ָ֣ ֶ ִ‫הַ ל‬ ‫ותפַלְ ֵ ֛ט ִני‬ ְ ‫יִ ְמחַ ץ‬ ‫ע ִָנִׂ֤י‬ ‫ְמבַ קְ שֶ יֶ֑ ָך‬ ‫אָ ִ ֶ֑שיב‬ (L ‫יך‬ ָ ‫ הֲלִ יכֹו ֶ ָֻׁ֣֯ת‬Ps. 68:25 ‘your ways’) (L ‫ וָּֽׁ ְתפַלְ ֵ ֶ֑טנִ י‬Ps. 71:2 ‘and you (msg) shall rescue me’) (L ֮‫ ִי ְמחַ ץ‬Ps. 68:22 ‘he will shatter’) (L ‫ עָנִ ָ֣י‬Ps. 70:6 ‘poor (msg)’) (L ‫יך‬ ָ ‫ל־מ ַ֫ ַבקְ ֶ ָ֥ש‬ ְ ‫ ָ ָּֽׁכ‬Ps. 70:5 ‘those, who seek you (msg)’) (L ‫ ִמבָ ָ ָ֣שן אָ ִ ֶ֑שיב‬Ps. 68:23 ‘I will bring’) 2.2.4 NST dageš in the last letter of a word-final syllable This feature occurs in the Tiberian-Babylonian manuscript only rarely, and then almost exclusively in the sonorants lamed and mem; this is also the case with the Tiberian-Palestinian manuscripts (see section 2.1.4). T-S A13.20 ְ‫יִ גדַ ל‬ ‫*כל‬ ָּֽׁ ַ ‫בְ ָ ָ֥דם‬ 144 (L ‫ֱֹלהים‬ ֶ֑ ִ ‫ יִגְ ַ ָ֣דל א‬Ps. 70:5 ‘he shall be great’) (L ‫ל־צֹור ָ ָּֽׁרי׃‬ ְ ‫ ָכ‬Ps. 69:20 ‘all of’) (L ‫ ְַ֫ב ָ ָ֥דם לְ ָ֥שֹון‬Ps. 68:24 ‘in blood’) 3 Conclusions The major difference in the use of NST dageš in the Tiberian-Palestinian and Tiberian-Babylonian manuscripts is within a word, since the latter marks it more frequently after open syllables within a word and only a few times after closed syllables within a word, whereas the Tiberian-Palestinian manuscripts attest NST dageš within a word mainly after closed syllables. The use of NST dageš in initial position is roughly the same in both groups, apart from the fact that the Tiberian-Babylonian manuscript attests it once in a guttural letter. The use of NST dageš in word-final letters is similar in both groups, attested only rarely and then in the letters lamed and mem. The major problem with the interpretation of NST dageš is that it is difficult to explain simply, since it occurs in many different contexts. While it is indeed difficult to interpret the findings above, they could all be interpreted as extensions of ST features. Thus, the marking of dageš after a closed syllable as well as after an open syllable with disjunctive accent in a non-bgdkft letter could indicate an extension of the use of so-called ST dageš lene. The use of NST dageš after an open syllable with non-final conjunctive accent could reflect an extension of dǝḥiq. Based on recent scholarship, however, I believe that it is very likely that NST reflects a phonetic dageš, which can only be identified as dageš forte.51 There are three points which support this thesis. First, there is no evidence that the use of NST dageš suggests that a letter could be realised in two different ways (i.e. with and without dageš) as was the case for the bgdkft letters in the ST tradition.52 Second, the marking of NST dageš indicates a phonetic realisation, as all other pronunciation features of these manuscripts do.53 And third, as Khan will show in a forthcoming publication, the dageš dot was always realised from at least the 11th century CE onwards with a doubling of the consonant, whether it was dageš forte or dageš lene.54 51 Cf. Fassberg, 1987, p. 79 n. 14 on the scholarly discussion of whether dageš euphonicum reflects gemination or not. Note that the data supplied by Bergsträsser on this issue in his grammar (Bergsträsser, 1918, pp. 64–69 (§10o–y)) are not exclusively from L or A, meaning that some of his examples do not reflect the ST tradition, e.g. Jer. 22:22 ַ‫ֶה־רוח‬ ָ֔ ‫( ִת ְרע‬Bergsträsser, 1918, p. 66, §10s), as opposed to L and A ַ‫ֶה־רוח‬ ָ֔ ‫ ; ִת ְרע‬and also Ps. 99:4 ‫( כֹונַ ָ֣נְ תָ מֵ ישָ ִ ֶ֑רים‬Bergsträsser, 1918, p. 66, §10r), in contrast to L and A ‫כֹונַ ָ֣נְ תָ מֵ ישָ ִ ֶ֑רים‬. It is, however, crucial to distinguish between the different stages of the development of manuscripts with Tiberian vocalisation, if the scope of our research concerns the ST tradition (Blapp, 2017, pp. 9–19; see also Blapp, forthcoming). 52 For instance, the letter beth was realised according to ST as /b/ when marked with dageš and as /v/ when no dageš was marked (in many cases, rafe was also added to indicate its fricative pronunciation as /v/). 53 See Blapp, 2017, pp. 206–207. Note that the use of rafe also supports this theory. It is used in the same context as NST dageš, and thus they could indicate the different realisations of the phonemes (i.e. geminated and ungeminated). This reflects a purely phonetic transcription of the reading tradition of each scribe, and is supported by the inconsistent use of the vocalisation signs. None of the use of the diacritical signs is entirely in accordance with the ST tradition, though the phonetic transcription sometimes bears vestiges of this (Blapp, 2017, 206–207). 54 Khan, forthcoming. 145 Due to the fact that NST dageš was always realised as dageš forte and also because the Tiberian Masoretic knowledge had been neglected or even lost after the death of the last Masorete in the late 10th century, there was no longer theoretical knowledge available of the different dageš, so that their distinctive pronunciations became detached from their original functions.55 Consequently NST dageš could be used where gemination for the sake of emphasis of the syllable boundary was needed. The main reason for using NST dageš was thus to ensure the precise reading of the text according to each scribe’s own pronunciation.56 This thesis is supported by the evidence presented above, since most of the examples of NST dageš occur in the word-initial letter or after closed syllables word-internally. Further evidence for the individuality of each manuscript can be found in other features of the NST manuscripts, such as the use of the vowel signs.57 Why the Tiberian-Babylonian manuscript in particular marks NST dageš after open syllables word-internally remains a subject for further investigation. On a different but no less important issue, I believe that we can no longer simply take the evidence from any manuscript with Tiberian vocalisation in order to do research on Tiberian biblical Hebrew, since the evidence might not represent the original standardised vocalisation system.58 I believe, rather, that we must examine and distinguish between different manuscripts and the traditions they reflect in order to gain a more precise picture of the development of the Tiberian vocalisation tradition.59 An important observation is that NST dageš throughout both classes and all the manuscripts is most frequently attested in the continuant sonorants lamed, mem and nun. It will be important to gather more data from NST as well as ST manuscripts in order to evaluate this observation in a more detailed way. 55 The NST manuscripts are one of the main witnesses for the loss of the knowledge of the ST reading tradition. Additionally, we have increasing activities to preserve the ST reading tradition in the late 10th century and early 11th century CE, since the last Tiberian Masorete, Aharon ben-Asher, died in the late 10th century (see Ben-Hayyim, 2007, pp. 319–321). Late ST manuscripts like L already exhibit a number of non-standard features, which could reflect the fact that its scribe no longer had primary knowledge of the ST tradition nor access to entirely reliable sources, so that he had to consult Masoretic treatises and his own pronunciation (see Blapp, forthcoming). 56 Khan, forthcoming. 57 Blapp, 2017, pp. 206–207. 58 Blapp, forthcoming. Neither Blake (Blake, 1943) nor Prätorius (Prätorius, 1883) distinguish between the different manuscripts. They often do not even indicate which manuscripts they are using. Israel Yeivin has remarked on the fact that there are different kinds of manuscript. However he still sometimes uses evidence from manuscripts with non-standard features in comparison to ST. Thus, for instance, he uses data from the Cairo Codex of the Prophets (e.g. Yeivin, 1980, pp. 286–287, §398) despite his conclusion that “In most of these ‘developed’ features, C shows relationship to the MSS of the ‘expanded’ Tiberian tradition, but in most features it resembles A” (Yeivin, 1980, p.20, §32). I believe that we have to be more cautious, and only use data from such manuscripts if we are comparing ST and NST data. We should not use this data as evidence for the ST tradition, since it reflects rather an intermediate stage between the ST and NST manuscripts. 59 Blapp, forthcoming. 146 It is possible that the high frequency of NST dageš on these three letters relates purely to their general high frequency, as they might appear more frequently than other letters in an appropriate context. A further possibility is that NST dageš in these letters indicates that their pronunciation had to be reinforced in order to preserve their proper pronunciation. Note that in the Tiberian-Babylonian tradition, NST dageš in medial position occurs almost exclusively after these three letters. All manuscripts of the Tiberian-Palestinian class use dageš similarly to the manuscripts with Palestinian vocalisation, apart from the use of dageš on guttural letters.60 References Ben-Hayyim, Z., 2007, “Ben-Asher, Aharon ben Moses”, in M. Berenbaum and F. Skolnik (eds), Encyclopaedia Judaica, Detroit: Macmillan Reference USA, vol. III, pp. 319–321. Bergsträsser, G., 1918, Hebräische Grammatik, Teil 1, Leipzig: F.C.W. Vogel. Blake, F. R., 1943, “The origin and development of the Hebrew dagesh”, JBL 62, pp. 89-107. Blapp, S., 2017, The Non-Standard Tiberian Hebrew Language Tradition according to Bible Manuscripts from the Cairo Genizah (unpubl. diss., University of Cambridge). ———, forthcoming, “The importance of a classification of the Standard Tiberian manuscripts”, in E. Attia-Kay, A. Perrot and S. Blapp (eds), Research Approaches in Hebrew Bible Manuscript Studies, Proceedings of the EAJS Laboratory 2016. Blau, J., 2010, Phonology and Morphology of Biblical Hebrew, An Introduction, Winona Lake: Eisenbrauns. Chomsky, W., 2001, David Kimḥi’s Hebrew Grammar (Mikhlol), Systematically Presented and Critically Annotated, New York City: Bloch Publishing Company. Davis, M., 1978, Hebrew Bible Manuscripts in the Cambridge Genizah Collections, vol. I: Taylor-Schechter Old Series and other Genizah Collections in Cambridge University Library, Cambridge: CUP. Fassberg, S., 1987, “Supralinear < and ^ in Palestinian pointed manuscripts of Hebrew and Aramaic from the Cairo Geniza”, in D. M. Golomb (ed.), ‘Working With No Data’: Semitic and Egyptian Studies Presented to Thomas O. Lambdin, Winona Lake: Eisenbrauns, pp. 75–103. Golinets, V., 2013, “Dageš, Mappiq, Specks on Vellum, and Editing of the Codex Leningradensis”, in R. G. Lehmann and A. E. Zernecke (eds), Kleine Untersuchungen zur Sprache des Alten Testaments und seiner Umwelt 15, Kamen: Hartmut Spenner, pp. 233–263. ———, 2017, “Dageš ”, in Khan et al. (eds), 2017, accessed 20 June 2017. Heijmans, Sh., 2017a, “Vocalization, Palestinian”, in Khan et al. (eds), 2017, accessed 21 June 2017. ———, 2017b, “Vocalization, Palestino-Tiberian”, in Khan et al. (eds), 2017, accessed 18 August 2017. Khan, G., 2017a, “Tiberian Reading Tradition”, in Khan et al. (eds), 2017, accessed 19 August 2017. 60 Cf. Fassberg, 1987, p. 84. 147 ———, 2017b, “Vocalization, Babylonian”, in Khan et al. (eds), 2017, accessed 19 August 2017. ———, forthcoming, “Remarks on the pronunciation of dageš in the Tiberian reading tradition of Biblical Hebrew”. ——— et al. (eds), 2017, Encylopedia of Hebrew Language and Linguistics, Leiden: Brill, http://referenceworks.brillonline.com/browse/encyclopedia-of-hebrew-language-and-linguistics. Morag, Sh., 1959, “The vocalization of Codex Reuchlinianus: is the ‘Pre-Masoretic’ Bible pre-Masoretic?”, JSS 4, pp. 216–237. Neudecker, H., 2017, “Vocalization of Modern Hebrew and colloquial pronunciation”, in Khan et al. (eds), 2017, accessed 15 August 2017. Ofer, Y., 2017, “Dǝḥiq”, in Khan et al. (eds), 2017, accessed 21 June 2017. Prätorius, F., 1883, “Ueber den Ursprung des Dagesch forte conjunctivum”, Zeitschrift für die alttestamentliche Wissenschaft 3, pp. 17–31. Revell, E. J., 1970, Hebrew Texts with Palestinian Vocalization, Toronto: Toronto University Press. Yarqoni, R., 1985, ʿEn ha-Qore by Yequtiʾel ha-Cohen (Hebrew) (unpubl. diss., Tel Aviv University). Yeivin, I., 1980, Introduction to the Tiberian Masorah, E. J. Revell (trans., ed.), Missoula, Mo.: Scholars Press. ———, 1985, The Hebrew Language Tradition as Reflected in the Babylonian Vocalization (‫( )מסורת הלשון העברית המשתקפת בניקוד הבבלי‬2 vols, Hebrew), Jerusalem: The Academy of the Hebrew Language. 148 Appendices A The use of NST in the Tiberian-Palestinian manuscripts61 T-S A11.1 NST dageš in initial ‫ז‬ NST dageš in medial ‫ז‬ NST dageš in initial ‫ט‬ NST dageš in medial ‫ט‬ NST dageš in initial ‫י‬ NST dageš in initial ‫ל‬ NST dageš in medial ‫ל‬ NST dageš in final ‫ל‬ NST dageš initial ‫מ‬ NST dageš in medial ‫מ‬ NST dageš in final ‫מ‬ NST dageš in initial ‫נ‬ NST dageš in medial ‫נ‬ NST dageš in initial ‫ס‬ NST dageš in medial ‫ס‬ NST dageš in initial ‫צ‬ NST dageš in medial ‫צ‬ NST dageš in initial ‫ק‬ NST dageš in medial ‫ק‬ NST dageš in medial ‫ר‬ NST dageš in initial ‫ש‬ NST dageš in initial ‫ש‬ NST dageš in medial ‫ש‬ NST dageš for rafe in the bgdkft letters NST dageš after originally closed syllable NST dageš after secondarily closed syllable NST dageš after two šəwas that merged NST dageš after original vocalic šəwa NST dageš after open syllable T-S 12.1 4 1 7 2 1 2 3 2 1 T-S T-S A12.10a A13.18 3 1 1 1 1 17 25 2 3 34 7 1 7 2 2 2 5 4 2 5 1 5 10 1 5 6 1 1 1 1 3 1 9 4 18 1 1 1 2 1 6 1 1 T-S A13.35 Total 1 8 1 8 5 1 48 6 1 51 19 3 24 4 4 3 13 5 4 6 2 1 10 4 9 1 2 2 1 2 8 2 4 2 1 1 1 7 2 1 5 12 40 2 4 2 1 1 2 2 61 Note that the manuscripts are of different sizes, and thus the amount of evidence (i.e. the number of possible deviating forms) increases with the size of the manuscript. For this paper, this difference is not the main issue, and thus I have not indicated the sizes of the manuscripts. For details, see Blapp, 2017, pp. 21–23. 149 B The use of NST in the Tiberian-Babylonian manuscript T-S A13.20 NST dageš in initial NST dageš in initial NST dageš in medial position after closed syllable NST dageš in medial position after open syllable NST dageš in initial NST dageš in initial NST dageš in medial position after closed syllable NST dageš in medial position after open syllable NST dageš in final NST dageš in initial NST dageš in medial position after closed syllable NST dageš in medial position after open syllable NST dageš in final NST dageš in initial NST dageš in medial position after closed syllable NST dageš in medial position after open syllable NST dageš in medial position of the 1sg suffix of verbal forms NST dageš in initial NST dageš in initial NST dageš in medial position after open syllable NST dageš in initial NST dageš in medial NST dageš after originally closed syllable NST dageš after secondary closed syllable NST dageš after two šəwas that merged Medial NST dageš after medial open syllable 150 ‫ח‬ ‫ט‬ ‫ט‬ ‫ט‬ ‫י‬ ‫ל‬ ‫ל‬ ‫ל‬ ‫ל‬ ‫מ‬ ‫מ‬ ‫מ‬ ‫מ‬ ‫נ‬ ‫נ‬ ‫נ‬ ‫נ‬ ‫צ‬ ‫ק‬ ‫ק‬ ‫ש‬ ‫ש‬ 1 1 1 1 2 12 1 7 7 18 2 12 2 2 2 2 3 2 3 1 1 3 6 1 1 16+ The Ashkenazic Hebrew of Nathan Nata Hannover’s Yeven Meṣula (1653) LILY KAHN University College London 1 Introduction1 This study will investigate the main grammatical features of Yeven Meṣula ‘Miry Depths’ or ‘Abyss of Despair’,2 a 17th century Hebrew historical work describing the events of the Chmielnicki Uprising that swept the Polish-Lithuanian Commonwealth in 1648–1649. Yeven Meṣula was written by the prominent Ashkenazic preacher and kabbalist Nathan Nata Hannover. Hannover was born and raised in Volhynia, a region in Eastern Europe corresponding to parts of present-day Poland, Ukraine and Belarus, but was forced to flee his homeland during the Chmielnicki Uprising and spent the next few years as an itinerant preacher in Poland, Germany and Holland. He wrote his account of the Chmielnicki pogroms during this period, and published it upon arriving in Venice in 1653. He subsequently travelled to Prague, and then settled in Jassy (present-day Iași in eastern Romania), where he became the head of the yeshiva and president of the rabbinical court. He remained in Jassy for approximately ten years, before relocating to Ungarisch Brod in Romania (presentday Uherský Brod in the Czech Republic), where he was killed by raiding Turkish soldiers in 1689.3 During his lifetime Hannover published three other works in addition to Yeven Meṣula: a homiletic sermon about the festival of Sukkot called Taʿame Sukka (Amsterdam, 1652), a Hebrew-German-Latin-Italian phrasebook called Safa Berura (Prague, 1660) and a collection of prayers according to the Lurianic kabbalistic rite called Shaʿare Ṣiyyon (Prague, 1662). He also wrote a collection of homiletical sermons on the Pentateuch which were never published. Hannover’s published writings had a long-lasting impact on Ashkenazic Jewry: his prayer collection Shaʿare Ṣiyyon enjoyed widespread popularity in Italy, Holland and Eastern Europe, and was reprinted in more than fifty 1 I am very grateful to Nadia Vidro and Esther-Miriam Wagner for their numerous insightful comments on a draft of this article. 2 A citation of Ps. 69:3. 3 See Halpern, 2007 for further details of Hannover’s life. 151 editions over the course of the 18th and 19th centuries. Likewise, Safa Berura was used among Jews for foreign language instruction until the 19th century. Hannover’s Yeven Meṣula is a relatively short work of 20 pages that provides an account of the 1648–1649 mass uprising of Ukrainian and Cossack peasants under the leadership of the Ukrainian Bogdan Chmielnicki against Polish rule in Ukraine. The uprising resulted in the destruction of many Ukrainian and Polish Jewish communities and the deaths of at least an estimated 18,000–20,000 Jews.4 Hannover’s work includes chronicles of the massacres that took place against the Jews in various places over the course of the two-year period between 1648 and 1649 in various locations throughout present-day Poland, Ukraine and Lithuania, such as Tulczyn, Zamość and Lwów/Lviv, as well as an account of the life of the Jews of the Kingdom of Poland. The work contains little information about Hannover’s personal experiences during the pogroms, although he did witness some of the events, but rather is based on eyewitness accounts and information gathered from others, both orally and from printed sources.5 Yeven Meṣula is a unique and ground-breaking piece of early modern Jewish historical writing,6 and has played a hugely influential role in Ashkenazic society and culture since its publication. The traumatic events of the Chmielnicki Uprising came to assume a central position in the Ashkenazic historical consciousness,7 and Hannover’s work dominated this consciousness well into the 20th century.8 It was reissued in its Hebrew and Yiddish versions in nearly every generation,9 and was also translated into a number of other languages, including French, German, Russian, Polish and English.10 The fact that Yeven Meṣula was the only source of information on the events of 1648– 1949 told from a Jewish perspective and accessible to readers without knowledge of Hebrew contributed to its authoritative status.11 Hannover’s text was also accepted as a reliable account of the pogroms by pioneering modern Jewish historians such as Heinrich Graetz and Simon Dubnow,12 and it remains an important historical source today, though it is no longer treated uncritically. Despite the prominent position which Yeven Meṣula has occupied in Central and Eastern European Jewish society and the importance which historians have accorded it as a key witness to the Chmielnicki Uprising, it has never been the subject of linguistic analysis. Given its status as a unique and influ4 Stampfer, 2003, p. 221. Halpern, 2007, p. 327. 6 Bartal, 2005, p. 7. 7 Stampfer, 2003; Ettinger, 2007; Stampfer, 2010. 8 Bacon, 2003, pp. 182–186. 9 Halpern, 2007, p. 327. 10 This study is based on the first edition of Yeven Meṣula (Hannover, 1653). 11 Bacon, 2003, p. 184. 12 Bacon, 2003, p. 183. 5 152 ential piece of early modern Ashkenazic Hebrew historical writing, examination of the grammatical composition of this text can shed valuable light on the 17th century Eastern European narrative and discursive use of the language. From a diachronic perspective, it can be instructive to analyse the influences of earlier strata of Hebrew on Hannover’s narrative and ascertain the extent to which it resembles the biblical, rabbinic and medieval forms of the language. It is also important to establish the relationship between Hannover’s 17th century historical writing and other forms of Central and Eastern European Hebrew which have been analysed, namely 19th century Maskilic Hebrew, Hasidic Hebrew and the language of the Kiṣur Shulḥan ʿAruḵ,13 as well as early modern and modern responsa literature.14 Comparison of Hannover’s writing with these other Central and Eastern European types of Hebrew is particularly important as it can help to ascertain the extent to which all of these authors were drawing on a shared Ashkenazic linguistic heritage which has not been adequately mapped. Moreover, in certain cases parallels can be observed between Yeven Meṣula and more distant Diaspora Jewish linguistic varieties such as medieval Ashkenazic writings, the Hebrew of Judaeo-Spanish speakers and Judaeo-Arabic, which can tentatively point towards possible broader trends. The present study thus seeks to provide an analysis of the characteristic orthographic, morphosyntactic and syntactic features of Hannover’s seminal narrative work and to place it within its diachronic context. Due to space constraints this study cannot provide an exhaustive survey of the linguistic features of Yeven Meṣula, but will give an overview of a number of representative features.15 It is hoped that this analysis will lead to a clearer understanding of the composition and chronological spread of Ashkenazic Hebrew and its relationship to other Diaspora forms of the language. 2 Orthography The orthography in the first edition of Yeven Meṣula is largely consistent with that of canonical forms of Hebrew, with a tendency to employ plene spelling in accordance with the post-biblical standard. The main area in which the spelling in Yeven Meṣula differs from that of earlier convention is in the widespread tendency to employ yod following ṣere in singular nouns with a 1cpl or 3msg possessive suffix, as illustrated in (1)–(3). This orthographic practice is likely rooted in the fact that in Ashkenazic Hebrew pronunciation, the vowel ṣere and the combination ṣere plus yod in stressed open syllables are both pronounced identically (generally as the diphthong [ej] or [aj]). The use of yod 13 Kahn, 2009; Kahn, 2012b; Kahn, 2015; Kahn, in press. Betzer, 2001. 15 Comparison of Hannover’s narrative work with his non-narrative writings is likewise beyond the scope of the present examination. 14 153 in these contexts suggests that the author’s own pronunciation had more impact on his orthography than the canonical written texts. The same phenomenon is widely attested in 19th century Hasidic Hebrew narrative, for the same reasons.16 The practice in both Eastern European forms of Hebrew may have been reinforced by the fact that some individual forms with non-standard yod are occasionally attested in medieval literature (for example, the form ‫עמינו‬ ʿamenu17 ‘our people’ shown in (1) below appears several times in the writing of the prominent 15th century biblical commentator Isaac Abarbanel). (1) ‫עמינו‬ ʿamenu ‘our people’18 (2) ‫מחניהו‬ maḥanehu ‘his camp’19 (3) ‫אדונינו המלך‬ ʾadonenu ham-meleḵ ‘our lord the king’20 3 3.1 Nominal morphosyntax Definite article with inseparable prepositions A common feature of Yeven Meṣula is the retention of the definite article following the inseparable preposition -‫ ל‬lǝ- ‘to, for’, as shown in (4)–(7). This type of construction contrasts with the standard in Biblical and Rabbinic Hebrew, where elision of the definite article following a prefixed preposition is the norm; cf. Biblical Hebrew ‫ הָ ֜ ִעיר‬haʿir ‘the town’21 vs ‫ ל ָ֔ ִָעיר‬laʿir ‘to the town’,22 and Mishnaic Hebrew ‫ הבית‬hab-bayit ‘the house’23 vs ‫ לבית‬lab-bayit ‘to the house’.24 In Biblical Hebrew there are only rare exceptions to this rule,25 and the same is true of Rabbinic Hebrew.26 However the phenomenon 16 See Kahn, 2015, pp. 20–22. The transcription system used in this study follows the Encyclopedia of Hebrew Language and Linguistics standard for post-biblical Hebrew; see Khan et al., 2013. 18 Hannover, 1653, p. 8. 19 Hannover, 1653, p. 15. 20 Hannover, 1653, p. 14. 21 Gen. 19:4. 22 1 Sam. 9:12. 23 Mishnah Ohalot 3:2. 24 Mishnah Negaʿim 13:3. 25 Joüon and Muraoka, 2009, p. 104. 26 Betzer, 2001, p. 86. 17 154 is a characteristic feature of prominent varieties of 19th century Eastern European Hebrew texts composed by Hasidic and Maskilic authors as well as Shlomo Ganzfried’s popular work of practical halachah (Jewish law), the Kiṣur Shulḥan ʿAruḵ,27 and is also attested in early modern and modern Ashkenazic and Sephardic responsa literature.28 The fact that the same phenomenon is commonly attested both in Hannover’s work and in these other varieties suggests that all of these Eastern European authors may have been drawing on a common Ashkenazic Hebrew legacy, which may in turn have had links to other forms of Diaspora Hebrew. This point will be discussed further throughout this study. (4) ‫להכומרי׳‬ lǝ-hak-komǝrim ‘to the priests’29 (5) ‫להדוכסים‬ lǝ-had-dukkasim ‘the dukes’30 (6) ‫להשר‬ lǝ-haś-śar ‘to the lord’31 (7) ‫להיונים‬ lǝ-hay-yǝwanim ‘the Ukrainians’32, 33 3.2 Indefinite article While Hebrew lacks a true indefinite article, Hannover regularly employs the numeral ‫ אחד‬ʾeḥad ‘one’ in this sense, with the meaning of ‘a’ or ‘a certain’, as in (8)–(11). While this use of the numeral has occasional precedent in Biblical Hebrew and other historical varieties of the language,34 these writings are unlikely to be the sole or chief source for Hannover as he utilises it much more systematically. Rather, any influence from earlier Hebrew texts is likely to 27 Kahn, in press. Betzer, 2001, p. 86. 29 Hannover, 1653, p. 4. 30 Hannover, 1653, p. 1. 31 Hannover, 1653, p. 3. 32 Hannover, 1653, p. 8. 33 The Hebrew word ‫ יונים‬yǝwanim literally means ‘Greeks’, but Hannover uses it as a label for ‘Ukrainians’. This is a metonym based on the Ukrainians’ Greek Orthodox faith; see Plokhy, 2015, p. 99. 34 Rubin, 2013b. 28 155 have received synchronic reinforcement from Hannover’s native Yiddish, which has a true indefinite article.35 As in the case of the definite article with prefixed prepositions discussed in section 3.1, the use of ‫ אחד‬ʾeḥad ‘one’ as an indefinite article is also a prominent feature of 19th century Eastern European varieties of Hebrew.36 Moreover, the existence of a similar use of the numeral ‘one’ is attested in medieval and later Judaeo-Arabic,37 which may suggest that there is a more widespread trend towards such a development in Semitic languages generally regardless of influence from a spoken substratum. (8) ‫ושם היה מושל ופקיד על העיר הנ״ל יהודי אחד ושמו זכרי״ה‬ wǝ-šam haya mošel u-p̄ aqid ʿal ha-ʿir hana‫״‬l yǝhudi ʾeḥad u-šmo zǝḵarya ‘and there was a governor and officer over the above-mentioned city, a certain Jew named Zechariah’38 (9) ‫ חזן א׳ ושמו ר׳ הירש‬39‫והיה בניה׳‬ wǝ-haya benehem ḥazzan ʾeḥad u-šmo reb hirš ‘and among them there was a certain cantor whose name was Reb Hirsh’40 (10) ‫ כמה ימי׳‬41‫והתארח אצל בעל הבית אחד‬ wǝ-hitʾareaḥ ʾeṣel baʿal hab-bayit ʾeḥad kama yamim ‘and he stayed with a certain home owner for a number of days’42 (11) ‫לקחו עשיר אחד לביתו‬ laqḥu ʿašir ʾeḥad lǝ-ḇeto ‘they took a rich man to his house’43 3.3 Definiteness discord in noun-adjective phrases Hannover’s writing typically exhibits definiteness concord between a noun and its associated adjective. However, in a significant minority of cases the noun takes the definite article but the associated attributive adjective does not, 35 Jacobs, 2005, p. 174. Kahn, in press. 37 Blau, 1980, p. 165; Wagner, 2010, p. 191. 38 Hannover, 1653, p. 2. 39 Sic; = ‫ביניה׳‬. 40 Hannover, 1653, p. 4. 41 Note the use of a definite construct chain as an indefinite noun. This is attributable to the fact that the phrase ‫ בעל הבית‬baʿal hab-bayit exists in Yiddish as an indefinite noun. The same phenomenon is widely attested in 19th century Hasidic Hebrew; see Kahn, 2015, pp. 62–63 for details. Similar constructions are also found in medieval Judaeo-Arabic; see Blau, 1980, p. 156. 42 Hannover, 1653, p. 20. 43 Hannover, 1653, p. 20. 36 156 as in (12)–(16). This phenomenon has occasional precedent in Biblical Hebrew,44 and appears more frequently in rabbinic literature.45 It is also a widespread feature of responsa literature46 and of 19th century Hasidic Hebrew.47 Hannover seems to have tended to employ it when the noun and adjective comprise a common collocation, as in (13) and (14), and therefore may have subconsciously regarded the phrase as a single unit. (12) ‫האשה חדשה אשר לקח‬ ha-ʾiša ḥadaša ʾašer laqaḥ ‘the new wife whom he had taken’48 (13) ‫הגזרה רעה‬ hag-gǝzera raʿa ‘the evil decree’49 (14) ‫הבשורה רעה‬ hab-bǝśora raʿa ‘the evil tidings’50 (15) ‫והנשים יפות לקחו לשפחו׳‬ wǝ-han-našim yap̄ ot laqḥu li-šp̄ aḥot ‘and they took the beautiful women as servant girls’51 (16) ‫החיל גדול של קאזקין‬ haḥayil gadol šel qozaqin ‘the great army of Cossacks’52 3.4 Non-standard definiteness of construct chains Hannover frequently forms definite construct chains by prefixing the definite article to the construct noun, as in (17)–(19). This differs from the biblical standard, in which the definite article in construct chains is prefixed to the absolute noun;53 this same convention has remained the norm in Mishnaic and 44 Waltke and O’Connor, 1990, p. 260; Williams, 2007, p. 31. Sarfatti, 1989, pp. 161–165; Pérez Fernández, 1999, pp. 26–27; Pat-El, 2009, pp. 35–36; Rubin, 2013a. 46 Betzer, 2001, p. 90. 47 Kahn, 2015, pp. 87–88. 48 Hannover, 1653, p. 2. 49 Hannover, 1653, p. 2. 50 Hannover, 1653, p. 4. 51 Hannover, 1653, p. 4. 52 Hannover, 1653, p. 8. 53 Williams, 2007, p. 8. 45 157 later varieties of Hebrew. However, Hannover’s usage has precedent in medieval and early modern responsa literature.54 Moreover, as in many of the other phenomena discussed in this study, it has a parallel in 19th century Eastern European forms of Hebrew.55 It is likely that the non-standard construction in all of these forms of Hebrew is attributable to influence from Yiddish, in which many of the construct chains in question exist independently as set phrases and which are made definite by placing the definite article at the beginning of the phrase.56 The same type of construction is also attested in the Hebrew writing of Judaeo-Spanish speakers.57 Because Judaeo-Spanish makes noun phrases definite by placing a definite article at the beginning of the phrase, as in Yiddish, the similarity between Hannover’s writing and that of the Judaeo-Spanish speakers suggests that in both cases the syntactic structures of the authors’ vernaculars had an influential role in the development of their Hebrew.58 (17) ‫האנשי מקומות‬ ha-ʾanše mǝqomot ‘the people of the places’59 (cf. standard Hebrew ‫ אנשי המקומות‬ʾanše ham-mǝqomot) (18) ‫הגבורי חיל‬ hag-gibbore ḥayil ‘the warriors’60 (cf. standard Hebrew ‫ גבורי החיל‬gibbore ha-ḥayil) (19) ‫הראש ישיבה‬ ha-roš yǝšiḇa ‘the head of the yeshiva’61 (cf. standard Hebrew ‫ ראש הישיבה‬roš hay-yǝšiḇa) This phenomenon extends to definite construct chains with a numeral: according to the standard Hebrew convention, the definite article in such constructions is prefixed to the absolute noun, but Hannover often prefixes it to the numeral, as in (20). This type of construction is also attested in medieval and 54 Betzer, 2001, p. 91. Kahn, in press. 56 See Kahn, 2015, pp. 60–61 and Kahn, in press for further details. 57 Bunis, 2013, pp. 50*–51*. 58 Note that a similar phenomenon is occasionally attested in medieval Judaeo-Arabic (see Blau, 1980, p. 157) but this seems to be much more restricted than that found in the Hebrew of Yiddish and Judaeo-Spanish speakers. 59 Hannover, 1653, p. 8. 60 Hannover, 1653, p. 8. 61 Hannover, 1653, p. 18. 55 158 later Judaeo-Arabic,62 which hints at the possibility of a more widespread internal Semitic developmental pattern requiring further investigation. (20) ‫השני שרי צבא‬ hašǝne śare ṣaḇa ‘the two army commanders’63 (cf. standard Hebrew ‫ שני שרי הצבא‬šǝne śare haṣ-ṣaḇa) It also extends to construct chains whose second member is a proper noun that would not be expected to take the definite article in any type of Hebrew. This particular usage, which is shown in (21)–(23), does not seem to have a clearly documented precedent in earlier or later forms of the language. Further research is required in order to ascertain whether it is attested in other varieties of Ashkenazic Hebrew. (21) ‫בכל המקומו׳ רוסי״א‬ bǝ-ḵol ham-mǝqomot rusya ‘in all the places of Russia’64 (cf. standard Hebrew ‫ בכל מקומות רוסיה‬bǝ-ḵol mǝqomot rusya) (22) ‫המלך פולי״ן‬ ham-meleḵ polin ‘the king of Poland’65 (cf. standard Hebrew ‫ מלך פולין‬meleḵ polin) (23) ‫בת המלך צרפת‬ bat ham-meleḵ ṣarp̄ at ‘the daughter of the king of France’66 (cf. standard Hebrew ‫ בת מלך צרפת‬bat meleḵ ṣarp̄ at) In addition, Hannover sometimes makes construct chains definite by prefixing the definite article to both the absolute noun and the construct noun, as in (24)– (27). This convention lacks precedent in the canonical forms of Hebrew, but is attested in the writing of the prominent 11th century commentator Rashi67 as well as in responsa literature.68 It is also a common feature of 19th century 62 Blau, 1980, p. 167; Wagner, 2010, pp. 206–210. Hannover, 1653, p. 16. 64 Hannover, 1653, p. 2. 65 Hannover, 1653, p. 17. 66 Hannover, 1653, p. 1. 67 Betzer, 2001, p. 108. 68 Betzer, 2001, p. 91–92. 63 159 Eastern European Hebrew,69 and in the writing of Ashkenazic Jerusalem community leader Joseph Rivlin.70 Taken together with the phenomena discussed previously in this study, this similarity may suggest that all of these Ashkenazic Hebrew authors were drawing on a shared linguistic heritage. (24) ‫השר הצבא‬ haś-śar haṣ-ṣaḇa ‘the army commander’71 (cf. standard Hebrew ‫ שר הצבא‬śar haṣ-ṣaḇa) (25) ‫על המפתן הבית‬ ʿal ham-mip̄ tan hab-bayit ‘on the threshold of the house’72 (cf. standard Hebrew ‫ על מפתן הבית‬ʿal mip̄ tan hab-bayit) (26) ‫השר העיר‬ haś-śar ha-ʿir ‘the city commander’73 (cf. standard Hebrew ‫ שר העיר‬śar ha-ʿir) (27) ‫הבעל הבית‬ hab-baʿal hab-bayit ‘the house owner’74 (cf. standard Hebrew ‫ בעל הבית‬baʿal hab-bayit) 3.5 Use of masculine plural ending in nun Hannover typically follows the biblical standard by employing the masculine plural noun ending ‫ים‬- -im on nouns and qoṭel forms. However, he sometimes opts for the variant ‫ין‬- -in, which is typical of Rabbinic Hebrew. The rabbinic variant is particularly common with qoṭel forms. This is illustrated in (28)– (31). Like many other aspects of Hannover’s writing, his fluctuation between the mem and nun endings has a direct parallel in 19th century Maskilic and Hasidic Hebrew.75 As in the other cases discussed in this study, this close resemblance between these various forms of Eastern European Hebrew points to the existence of a shared underlying variety of the language spanning several centuries. 69 Kahn, 2015, pp. 62–65; Kahn, in press. Wertheimer, 1975, pp. 159–160. 71 Hannover, 1653, p. 15. 72 Hannover, 1653, p. 7. 73 Hannover, 1653, p. 13. 74 Hannover, 1653, p. 18. 75 Kahn, 2012b, p. 185. 70 160 (28) ‫צדדין‬ ṣǝdadin ‘sides’76 (29) ‫שולחין‬ šolǝḥin ‘they send’77 (30) ‫הולכין‬ holǝḥin ‘they go’78 (31) ‫שותין‬ šotin ‘they drink’79 The use of the nun ending instead of the more frequently attested mem variant is not systematic. In some cases Hannover employs both endings on the same form within close proximity to each other, as in (32) and (33), which contain a nun and a mem respectively and are only five lines apart from each other in the text. This type of fluctuation between the mem and nun endings is also attested in medieval Ashkenazic copies of Hebrew manuscripts (e.g. the 14th century halachic code Arba’ah Turim),80 which suggests a much earlier origin for the phenomenon. (32) ‫והם היו פטורין מן מס המלך‬ wǝ-hem hayu pǝṭurin min mas ham-meleḵ ‘and they were exempt from the king’s tax’81 (33) ‫ולכן היו פטורים מן המס‬ wǝ-laḵen hayu pǝṭurim min ham-mas ‘and therefore they were exempt from the tax’82 The nun variant is particularly commonly attested on periphrastic verbs (see section 4.4), possibly because such verbs are a typical feature of Rabbinic Hebrew, and commonly appear with a nun ending in that form of the language. This is illustrated in (34) and (35): 76 Hannover, 1653, p. 3. Hannover, 1653, p. 3. 78 Hannover, 1653, p. 9. 79 Hannover, 1653, p. 12. 80 N. Vidro, personal communication. 81 Hannover, 1653, p. 1. 82 Hannover, 1653, p. 1. 77 161 (34) ‫ואם היו רוצים לילך לדרכם היו נותנין להם צדה לדרך‬ wǝ-ʾim hayu roṣim leleḵ lǝ-darkam hayu notǝnin lahem ṣeda lad-dereḵ ‘and if they wanted to go on their way, they would give them provisions for the road’83 (35) ‫והמחנה עם פולי״ן לא היו יודעין מה השמחה הזאת‬ wǝ-ham-maḥane ʿam polin lo hayu yodʿin ma haś-śimḥa haz-zot ‘and the Polish camp did not know what this rejoicing was for’84 However this is likewise inconsistent, so that periphrastic verbs are not uncommonly attested with the mem ending, as in (36) and (37): (36) ‫בכל המקומו׳ אשר היו מגיעים שמה‬ bǝ-ḵol ham-mǝqomot ʾašer hayu maggiʿim šamma ‘in all the places that they reached’85 (37) ‫והם היו יושבים בטח‬ wǝ-hem hayu yošǝḇim beṭaḥ ‘and they dwelled in safety’86 3.6 Long form numerals with feminine nouns Hannover’s writing exhibits a blurring of the gender distinction between long and short form numerals, whereby he frequently employs long form numerals in conjunction with feminine nouns; see examples (38)–(42). This differs from the standard convention in the canonical forms of Hebrew, which exhibit gender polarity with numerals (with the long forms employed in conjunction with masculine nouns, and the short forms employed in conjunction with feminine nouns). Like many of the other phenomena discussed in this study, this has a parallel in later Eastern European Hebrew writing.87 It may be ascribable to influence from the Yiddish vernacular, which has only one set of numerals that is used with nouns of all genders.88 As in several other cases discussed in this study, the same usage is also found further afield in Judaeo-Arabic,89 perhaps suggesting a more widespread tendency to shift away from gender polarity in Semitic languages. 83 Hannover, 1653, p. 20. Hannover, 1653, p. 11. 85 Hannover, 1653, p. 4. 86 Hannover, 1653, p. 3. 87 Wertheimer, 1975, p. 157; Kahn, 2015, pp. 137–139; Kahn, in press. 88 Katz, 1987, pp. 201–203. 89 Wagner, 2010, pp. 191–206. 84 162 (38) ‫ששה מאו׳ גבורי חיל‬ šišša meʾot gibbore ḥayil ‘six hundred warriors’90 (cf. standard Hebrew ‫ שש מאות גבורי חיל‬šeš meʾot gibbore ḥayil) (39) ‫ושני בנותיו‬ u-šne bǝnotaw ‘and his two daughters’91 (cf. standard Hebrew ‫ ושתי בנותיו‬u-šte bǝnotaw) (40) ‫שלשה שורות סוסים‬ šǝloša šurot susim ‘three rows of horses’92 (cf. standard Hebrew ‫ שלש שורות סוסים‬šaloš šurot susim) (41) ‫חמשה מאות אלף איש‬ ḥamišša meʿot ʾelep̄ ʾiš ‘five hundred thousand men’93 (cf. standard Hebrew ‫ חמש מאות אלף איש‬ḥameš meʿot ʾelep̄ ʾiš) (42) ‫יותר משבעה מאות קהילות‬ yoter miš-šiḇʿa meʾot qǝhillot ‘more than seven hundred communities’94 (cf. standard Hebrew ‫ יותר משבע מאות קהילות‬yoter miš-šeḇaʿ/šǝḇaʿ meʾot qǝhillot) 3.7 Avoidance of the dual The canonical varieties of Hebrew have a dual form of nouns used with paired body parts, time words and numerals; for example, ‫ יומי(י)ם‬yomayim ‘two days’, ‫ ח(ו)דשי(י)ם‬ḥodšayim ‘two months’ and ‫ אלפי(י)ם‬ʾalpayim ‘two thousand’. Hannover typically avoids the dual with reference to time words and numerals, instead using the numeral ‫ שני\שתי‬šǝne/šte ‘two’ in conjunction with a plural noun, as in (43)–(46). This practice can likewise be seen in 19th century Eastern European Hebrew.95 As in the case of the later writings, it is likely that Hannover’s avoidance of the dual is attributable to the fact that his Yiddish vernacular lacked such a form, instead using the plural in conjunction with the numeral ‘two’. Moreover, as in several instances discussed above, the 90 Hannover, 1653, p. 6. Hannover, 1653, p. 7. 92 Hannover, 1653, p. 9. 93 Hannover, 1653, p. 10. 94 Hannover, 1653, p. 14. 95 Kahn, 2015, pp. 51–53; Kahn, in press. 91 163 same phenomenon is also attested in Judaeo-Arabic96, where there is no clear influence from a substratum lacking the construction; this may hint at a more widespread developmental pattern common to certain Semitic languages. (43) ‫כשני אלפי׳ יהודי׳‬ ki-šne ʾalap̄ im yǝhudim ‘about two thousand Jews’97 (cf. standard Hebrew ‫ כאלפי(י)ם יהודים‬kǝ-ʾalpayim yǝhudim) (44) ‫ובשתי שעות ביום‬ u-bi-šte šaʿot bay-yom ‘and for two hours a day’98 (cf. standard Hebrew ‫ ובשעתי(י)ם ביום‬u-ḇi-šʿatayim bay-yom) (45) ‫שני ימים‬ šǝne yamim ‘two days’99 (cf. standard Hebrew ‫ יומי(י)ם‬yomayim) (46) ‫שני חדשים‬ šǝne ḥodašim ‘two months’100 (cf. standard Hebrew ‫ חדשי(י)ם‬ḥodšayim) There is only one example of a dual numeral in Yeven Meṣula, shown in (47). Note that this same phrase appears a few pages later in the more common plural construction, as shown in (48). (47) ‫מאתים אלף זהובים‬ matayim ʾelep̄ zǝhuḇim ‘two hundred thousand gold pieces’101 (48) ‫שני מאות אלף זהובים‬ šǝne me’ot ʾelep̄ zǝhuḇim ‘two hundred thousand gold pieces’102 96 Blau, 1980, p. 99. Hannover, 1653, p. 6. 98 Hannover, 1653, p. 9. 99 Hannover, 1653, p. 16. 100 Hannover, 1653, p. 10. 101 Hannover, 1653, p. 12. 102 Hannover, 1653, p. 16. 97 164 4 4.1 Verbal morphosyntax Use of wayyiqṭol Hannover very commonly constructs past narrative sequences by means of the quintessentially biblical wayyiqṭol form, as in (49)–(51). In this respect his writing resembles that of later Maskilic and Hasidic narrative literature, which likewise is replete with wayyiqṭol forms.103 Hannover’s use of this form, like that of the later Hasidic and Maskilic writers, is likely rooted in a desire to evoke in his readers echoes of the venerable biblical narrative tradition, thereby lending his writing an air of authority and significance.104 However Hannover employs the wayyiqṭol more systematically than his 19th century counterparts: while the Maskilic and Hasidic authors often round off a sequence of qaṭal forms with a single wayyiqṭol, which serves almost as a decorative flourish rather than an essential element of the verbal system, Hannover tends to employ it much more regularly. This suggests that he may have been more at ease with the function of the wayyiqṭol than the later authors were. Further research is required in order to ascertain whether other 17th century authors share this comparative familiarity with the biblical narrative preterite form. (Note, however, that Hannover does not employ the wayyiqṭol exclusively in his presentation of past narrative, but rather alternates it with the qaṭal; this will be discussed in section 4.2.) (49) ‫וישיבו לו יהיה כדבריך וילך חמיל י״מש עם כל חילו אל מלך הקדרי׳‬ way-yašiḇu lo yihye ki-dḇareḵa way-yeleḵ ḥmil yimaḥ šǝmo ʿim kol ḥelo ʾel meleḵ haq-qǝdarim ‘and they answered him, “may it be as you say”, and Chmielnicki – may his name be blotted out – went with his whole army to the king of the Tatars’105 (50) ‫ויכתירו לפאולוק שם ויעשו לו כסא של ברזל והושיבו עליו ויעש הסרדיוט כתר של‬ ‫ברזל בראשו‬ way-yaḵtiru lǝ-pawluq šam way-yaʿaśu lo kisse šel barzel wǝ-hošiḇu ʿalaw way-yaʿaś has-sardioṭ keter šel barzel bǝ-rošo ‘and they crowned Pawliuk king there and made an iron throne for him and set him upon it and the army officer put an iron crown on his head’106 103 Kahn, 2009, pp. 241–243; Kahn, 2012b, pp. 181–183; Kahn, 2015, pp. 172–174. See Kahn, 2012a for further discussion of this suggestion. 105 Hannover, 1653, p. 3. 106 Hannover, 1653, p. 2. 104 165 (51) ‫ויקומו וינוסו כולם ויעזבו את אהליהם את סוסיהם ואת חמוריהם וישליכו על הדרך‬ ‫כסף וזהב‬ way-yaqumu way-yanusu kulam way-yaʿazḇu ʾet ʾohalehem ʾet susehem wǝ-ʾet ḥamorehem way-yašliḵu ʿal had-dereḵ kesep̄ wǝ-zahaḇ ‘and they all arose and fled, and they abandoned their tents, their horses, and their donkeys, and they threw silver and gold on the road’107 In some cases, Hannover’s wayyiqṭol sequences may be introduced by the characteristically biblical construction wayehi, as in (52), which begins with wayehi and contains a sequence of another two wayyiqṭols. (52) ‫ויהי כשמוע הדוכסי׳ והשרים ויטב בעיניהם הדבר וימליכו עליהם למלך את קאז״ימר‬ ‫ירה בן שני של המלך ז״יגמונד‬ wa-yhi kišmoaʿ had-dukkasim wǝ-haś-śarim way-yiṭaḇ bǝ-ʿenehem had-daḇar way-yamliḵu ʿalehem lǝ-meleḵ ʾet qazimer yarum hodo ben šeni šel ham-meleḵ zigmund ‘and when the dukes and the lords heard, the matter was good in their eyes, and they made His Majesty Casimir the second son of King Sigmund, king over them’108 4.2 Use of qaṭal in narrative sequences While Hannover typically employs the wayyiqṭol in past narrative sequences, he occasionally employs qaṭal forms in such cases, as in (53)–(55). This type of sequence is ultimately traceable to Rabbinic Hebrew.109 Like many other features of Hannover’s writing, this fusion of biblical and rabbinic past narrative verbal structures is also a standard feature of 19th century Maskilic and Hasidic Hebrew.110 This practice of drawing on both the biblical and rabbinic methods of conveying past narrative in the same text may be a function of the author’s desire to adhere to the biblical historical narrative convention while simultaneously harbouring an intimate knowledge of the rabbinic model as well; this is likely to have been compounded by the fact that Hannover’s native Yiddish lacks a construction like the wayyiqṭol, rendering the rabbinic use of the qaṭal in past narrative intuitively more familiar. 107 Hannover, 1653, p. 11. Hannover, 1653, p. 14. 109 Pérez Fernández, 1999, pp. 115–116. 110 Kahn, 2009, pp. 87–89 and Kahn, 2015, p. 146 respectively. 108 166 (53) ‫וכן עשה אסף כל חילו רכבו ופרשיו והלך עם אשתו אל מקומות שיש לו אחורי הנהר‬ ‫ניפ״ור‬ wǝ-ḵen ʿaśa ʾasap̄ kol ḥelo riḵbo u-p̄ arašaw wǝ-halaḵ ʿim ʾišto ʾel mǝqomot šey-yeš lo ʾaḥore han-nahar niper ‘and thus he did; he gathered all his forces, his chariots and his horsemen, and he went with his wife to the places that he had behind the river Dnieper’111 (54) ‫ומשם נסעו לק״ק סטארי״דוב והרגו ביהודי׳ הרג רב‬ u-miš-šam nasʿu lǝ-qǝhilla qǝdoša staridub wǝ-hargu bay-yǝhudim hereg raḇ ‘and from there they travelled to the holy community of Starodub and killed many Jews’112 (55) ‫חתרו חתירה תחת העיר והכניסו הפוחזים בעיר בלילה והתחילו להרוג בעם‬ ḥatru ḥatira taḥat ha-ʿir wǝ-hiḵnisu hap-poḥazim ba-ʿir bal-layla wǝ-hitḥilu la-harog ba-ʿam ‘they tunnelled under the city and let the scoundrels into the city at night, and they started to kill the people’113 In many cases, Hannover’s qaṭal forms are preceded or followed by a wayyiqṭol, as in (56) and (57) respectively. (56) ‫ויערוך המלך מערכה גדולה ותקע אהלו בבית הכומרים‬ way-yaʿaroḵ ham-meleḵ maʿaraḵa gǝdola wǝ-taqaʿ ʾohalo bǝ-ḇet hak-komarim ‘and the king waged a large battle, and pitched his tent in the priest’s house’114 (57) ‫והיהודי הנ״ל ישב בשלחן אחר וחשב חשבנותיו ושמע הדבר וגילה הדבר להשר וישם‬ ‫השר לחמיל י״מש בבית האסורים‬ wǝ-ha-yǝhudi hana‫״‬l yašaḇ bǝ-šulḥan ʾaḥer wǝ-ḥašaḇ ḥešbonotaw wǝ-šamaʿ had-daḇar wǝ-gila had-daḇar lǝ-haś-śar way-yaśem haś-śar lǝ-ḥmil yimaḥ šǝmo bǝ-ḇet ha-ʾasurim ‘and the above-mentioned Jew sat at another table and made his calculations, and heard the matter and revealed the matter to the minister, and the minister put Chmielnicki – may his name be blotted out – in prison’115 Hannover also frequently initiates past narrative sequences with the typically biblical temporal construction wayehi plus a prefixed infinitive construct, and 111 Hannover, 1653, p. 2. Hannover, 1653, p. 11. 113 Hannover, 1653, p. 12. 114 Hannover, 1653, p. 17. 115 Hannover, 1653, p. 3. 112 167 then continues them with qaṭal forms, as in (58)–(60). This contrasts with Biblical Hebrew, in which wayehi is followed by wayyiqṭols.116 As in many other cases discussed in this study, this fusion of biblical and rabbinic usages has a direct parallel in 19th century Hasidic Hebrew.117 (58) ‫ויהי כשמוע הצורר חמי״ל י״מש עשה תחבולה ושלח ספרים אל השר הצבא‬ wa-yhi ki-šmoaʿ haṣ-ṣorer ḥmil yimaḥ šǝmo ʿaśa taḥbula wǝ-šalaḥ sǝp̄ arim ʾel haś-śar haṣ-ṣaḇa ‘and when the enemy Chmielnicki – may his name be blotted out – heard, he concocted a plot, and sent letters to the army commander’118 (59) ‫ויהי כשמוע אנשי העיר הדבר הזה הקדימו נעשה לנשמע‬ wa-yhi ki-šmoaʿ ʾanše ha-ʿir had-daḇar haz-ze hiqdimu naʿaśe lan-nišmaʿ ‘and when they heard this matter, they acted quickly’119 (60) ‫ויהי אחר הדברים האלה חזרו הקדרים והיונים לביתם‬ wa-yhi ʾaḥar had-dǝḇarim ha-’elle ḥazru haq-qǝdarim wǝ-hay-yǝwanim lǝ-ḇetam ‘and after these things, the Tatars and Ukrainians went home’120 Only rarely is a new narrative sequence introduced by a qaṭal of the root .‫ה‬.‫י‬.‫ה‬ h.y.h. instead of wayehi: (61) ‫והיה בתוכם איש אחד חכם ונבון‬ wǝ-haya bǝ-toḵam ʾiš ʾeḥad ḥaḵam wǝ-naḇon ‘and there was a clever and wise man among them’121 Often Hannover alternates between the wayyiqṭol and the qaṭal seemingly interchangeably in the same sequence, as illustrated in the following example: (62) ‫ונסעו משם ויצורו על ק״ק זאלקווי״א ובקשו לגשת אל החומה להעמיד סולמות‬ ‫וישפכו עליהם מים רותחין מן החומה וינוסו הפוחזים מפניהם‬ wǝ-nasʿu miš-šam way-yaṣuru ʿal qǝhilla qǝdoša zolqiewa u-ḇiqqǝšu lag-gešet ʾel ha-ḥoma lǝ-haʿamid sulamot way-yišpǝḵu ʿalehem mayim rotḥin min ha-ḥoma way-yanusu hap-poḥazim mip-pǝnehem ‘and they travelled from there and besieged the city of Żółkiew, and they tried to approach the wall in order to put up ladders, and they poured boiling water on them from the walls and the scoundrels fled from them’122 116 van der Merwe, Naudé and Kroeze, 1999, pp. 166–167. Kahn, 2015, pp. 176–177. 118 Hannover, 1653, p. 11. 119 Hannover, 1653, p. 14. 120 Hannover, 1653, p. 16. 121 Hannover, 1653, p. 9. 122 Hannover, 1653, p. 13. 117 168 4.3 Use of qaṭal with present reference Hannover occasionally employs the qaṭal form of stative qal root .‫ע‬.‫ד‬.‫ י‬y.d.ʿ. with present reference, as in (63) and (64). This is a characteristic feature of Biblical Hebrew,123 in contrast to Rabbinic Hebrew, which uses the qoṭel in such cases.124 Hannover’s usage has an exact parallel in later Maskilic Hebrew,125 as well as in Hasidic Hebrew, in which its use is likewise restricted to the root .‫ע‬.‫ד‬.‫ י‬y.d.ʿ.126 Further research is required in order to ascertain whether other 17th century Eastern European writers of Hebrew narrative employed this type of construction with a wider variety of roots, and that the lack of examples in Hannover’s text is due to its restricted size. (63) ‫אתה ידעת את האיש חמי״ל י״מש ואת מעשהו‬ ʾatta yadaʿta ʾet ha-ʾiš ḥmil yimaḥ šǝmo wǝ-et maʿaśehu ‘you know the man Chmielnicki – may his name be blotted out – and his deed’127 (64) ‫אתם ידעתם שעם פולין הם חזקים יותר ממנו‬ ʾattem yǝdaʿtem še-ʿam polin hem ḥazaqim yoter mimmennu ‘you know that the Polish people are stronger than us’128 4.4 Periphrastic verbal constructions for past progressive and habitual Hannover frequently employs a periphrastic verbal construction consisting of a qaṭal of the root .‫ה‬.‫י‬.‫ ה‬h.y.h. followed by a qoṭel to convey past progressive actions, as in (65)–(68). In some cases, the construction is used with stative verbs whose progressive sense is not evident in the English translation, as in (67) and (68). This type of construction is a characteristically post-biblical phenomenon; it appears frequently in Mishnaic Hebrew129 and in various types of medieval Hebrew texts.130 Hannover’s use of this construction can be contrasted with his use of the typically biblical wayyiqṭol discussed above. Like other elements of the verbal system in Yeven Meṣula, the use of the periphrastic construction has a direct parallel in 19th century Maskilic and Hasidic Hebrew.131 123 Waltke and O’Connor, 1990, pp. 364–373. Pérez Fernández, 1999, p. 133. 125 Kahn, 2009, pp. 90–91. 126 Kahn, 2015, pp. 151–152. 127 Hannover, 1653, p. 2. 128 Hannover, 1653, p. 3. 129 Pérez Fernández, 1999, pp. 108–109; Sharvit, 2004, p. 50; Mishor, 2013. 130 Rabin, 1968, p. 115; Sarfatti, 2003, p. 87; Rand, 2006, pp. 341–342. 131 Kahn, 2009, pp. 178–181; Kahn, 2015, p. 190. 124 169 (65) ‫] וסיפר חמי״ל‬...[ ‫ויהי היום היו יושבי׳ קאזקין חמי״ל י״מש ואוהביו במשתה היין‬ ‫י״מש לפני אוהביו‬ wa-yhi hay-yom hayu yošǝḇim qozaqin ḥmil yimaḥ šǝmo wǝ-ʾohaḇaw bǝ-mište hay-yayin […] wǝ-sipper ḥmil yimaḥ šǝmo li-p̄ ne ʾohaḇaw ‘and one day the Cossacks were sitting, Chmielnicki – may his name be blotted out – and his friends, at the wine banquet […] Chmielnicki – may his name be blotted out – said to his friends’132 (66) ‫בכל מקומו׳ שהיו יהודי׳ דרים שם ובמקומות הקאזקין שלא היו יהודים דרים שם‬ bǝ-ḵol mǝqomot še-hayu yǝhudim darim šam u-ḇi-mqomot haq-qozakin šel-lo hayu yǝhudim darim šam ‘in all the places where Jews were living, and in the places of the Cossacks, where Jews were not living’133 (67) ‫ויהי כשמוע הצורר חמי״ל י״מש את הדב׳ היה מתירא לנפשו‬ wa-yhi ki-šmoaʿ haṣ-ṣorer ḥmil yimaḥ šǝmo ʾet had-daḇar haya mityare lǝ-nap̄ šo ‘and when the enemy Chmielnicki heard the matter, he feared for his life’134 (68) ‫והשר ההו׳ היה מכיר את האיש‬ wǝ-haś-śar ha-hu haya makkir ʾet ha-ʾiš ‘and that lord knew the man’135 The construction can also be used to convey a habitual sense, as in (69) and (70). This is likewise a feature of Rabbinic Hebrew136 in addition to medieval forms of the language such as the piyyuṭim.137 Again, this is also a feature of 19th century Maskilic and Hasidic Hebrew.138 (69) ‫והיו נותנים לנערים אכיל׳ מקופה של צדקה‬ wǝ-hayu notǝnim lan-nǝʿarim ʾaḵila miq-quppa šel ṣǝdaqa ‘and they would give the boys food from the charity fund’139 132 Hannover, 1653, p. 3. Hannover, 1653, p. 16. 134 Hannover, 1653, p. 17. 135 Hannover, 1653, p. 2. 136 Pérez Fernández, 1999, pp. 108–109; Mishor, 2013. 137 Sáenz-Badillos, 1993, p. 210. 138 Kahn, 2009, pp. 181–182; Kahn, 2015, p. 189. 139 Hannover, 1653, p. 18. 133 170 (70) ‫ופרנסים דארבע הארצות היו בוררין להם דיינים‬ u-p̄ arnesim dǝ-ʾarbaʿ ha-ʾaraṣot hayu borǝrin lahem dayyanim ‘and community leaders of the Four Lands would choose judges for themselves’140 4.5 Verb-subject gender discord Hannover has a strong tendency to use the 3msg form of a qaṭal verb in conjunction with a feminine noun if the verb precedes the noun, as in (71)–(74). This has direct precedent in the Hebrew Bible.141 However, the fact that there are numerous instances of this phenomenon in the relatively short text of Yeven Meṣula suggests that, though the phenomenon is ultimately traceable to the Hebrew Bible, Hannover was not inspired solely by its occasional attestation there. This usage is not exhibited to the same extent in later Eastern European Hebrew writing, though it is sometimes found in Hasidic narrative literature.142 Further research on other types of early modern Eastern European Hebrew is needed in order to ascertain whether it was part of a more widespread tradition. (71) ‫והיה דירתו בעיר טשהארי״ן‬ wǝ-haya dirato bǝ-ʿir tšehirin ‘and he lived in the town of Czehryń ‘143 (72) ‫ומעולם היה שנאה גדולה בין הקדרים והיונים‬ u-me-ʿolam haya śinʾa gǝdola ben haq-qǝdarim wǝ-hay-yǝwanim ‘and there had always been a great hatred between the Tatars and the Ukrainians’144 (73) ‫ואם היה קהילה של חמישי׳ בעלי בתים היו מחזיקין לא פחות משלשים בחורים ונערים‬ wǝ-ʾim haya qǝhilla šel ḥamišša baʿale battim hayu maḥaziqin lo paḥot miš-šǝloša baḥurim u-nǝʿarim ‘and if there was a community of fifty house owners, they would maintain no less than thirty young men and boys’145 (74) ‫בא לפעמים עשיר אחד שהיה לו בת קטנה‬ ba li-p̄ ʿamim ʿašir ʾeḥad še-haya lo bat qǝtanna ‘there came sometimes a rich man who had a small daughter’146 140 Hannover, 1653, p. 20. Waltke and O’Connor, 1990, p. 109; Williams, 2007, p. 92. 142 Kahn, 2015, pp. 254–255. 143 Hannover, 1653, p. 2. 144 Hannover, 1653, p. 1. 145 Hannover, 1653, p. 18. 146 Hannover, 1653, p. 20. 141 171 5 5.1 Syntax Temporal constructions Hannover employs two different methods of forming temporal constructions. In some cases he uses the temporal conjunction ‫ כאשר‬kaʾašer ‘when’ or its prefixed variant -‫ כש‬kǝ-še- followed by a finite verb. The temporal construction may be introduced by wayehi. The following examples illustrate this. (75) ‫ויהי כאשר נסע הצורר חמי״ל י״מש עם מחניהו לכבוש ק״ק לובלי״ן הבירה ולא היה‬ ‫רק ארבע פרסאות מק״ק לובלי״ן בא אליו כתב המלך‬ wa-yhi kaʾašer nasaʿ haṣ-ṣorer ḥmil yimaḥ šǝmo ʿim maḥanehu li-ḵboš qǝhilla qǝdoša lublin hab-bira wǝ-lo haya raq 4 parsaʾot miq-qǝhilla qǝdoša lublin ba ʾelaw kǝtaḇ ham-meleḵ ‘and when the enemy Chmielnicki – may his name be blotted out – travelled with his camp to conquer the holy city of Lublin, the capital, and he was no more than four parsas from the holy city of Lublin, the king’s edict reached him’147 (76) ‫אבל הם לא חמלו עליהם כשנפלו עם פולין בידם‬ ʾaḇal hem lo ḥamlu ʿalehem kǝ-šen-nap̄ lu ʿam polin bǝ-yadam ‘but they did not have pity on them when the Poles fell into their hands’148 (77) ‫ויהי כאשר בא השר הנ״ל עם אשתו לעיר טש״הרין קבלו אותו אנשי המקום בשמחה‬ ‫גדולה‬ wa-yhi kaʾašer ba haś-śar hana‫״‬l ʿim ʾišto lǝ-ʾir tšehirin qibbǝlu ʾoto ʾanše ham-maqom bǝ-śimḥa gǝdola ‘and when the above-mentioned lord came with his wife to the town of Czehryń, the local people received him with great joy’149 However he also forms temporal constructions by means of an inseparable preposition prefixed to an infinitive construct, as in Biblical Hebrew, as in (78)–(80). Such temporal constructions are typically preceded by wayehi. The inseparable preposition -‫ כ‬kǝ- is used to denote the sense of ‘just after’, as in Biblical Hebrew. This type of construction is quite common, but is most frequently attested with the root .‫ע‬.‫מ‬.‫ ש‬š.m.ʿ. ‘hear’, as in the first two examples. This may suggest that the construction was not extremely productive for Hannover but rather that this particular collocation was an almost fossilised expression with which he was particularly familiar. Alternatively, it may simply 147 Hannover, 1653, p. 14. Hannover, 1653, p. 11. 149 Hannover, 1653, p. 2. 148 172 indicate that the expression ‘and when X heard’ is a high-frequency expression for a historical narrative such as Yeven Meṣula. (78) ‫ הנ״ל הדב׳ הזה ויחרד‬150‫ויהי כשמוע הדוכוס‬ wa-yhi ki-šmoaʿ had-dukkas hana‫״‬l had-daḇar haz-ze way-yeḥerad ‘and when the aforementioned duke heard this matter, he was afraid’151 ‫ויהי כשמוע חמיל י״מש שהדוכוס‬ ‫לנגדו שר הצבא שלו‬ wa-yhi ki-šmoaʿ ḥmil yimaḥ šǝmo še-had-dukkas wišniyeṣqi holeḵ wǝ-qareḇ ʾel maḥanehu […] šalaḥ lǝ-negdo śar haṣ-ṣaḇa šello ‘and when Chmielnicki – may his name be blotted out – heard that Duke Wiśniowiecki was approaching his camp […] he sent out his general’153 (79) ‫] שלח‬...[ ‫ווישני״עצקי הולך וקרב אל מחניהו‬ 152 (80) ‫ויהי כשבת המלך על כסא מלכותו כתב מיד ספרים אל הצורר חמיל י״מש שילך וישוב‬ ‫לביתו‬ wa-yhi kǝ-šeḇet ham-meleḵ ʿal kisse malḵuto kataḇ miy-yad sǝp̄ arim ʾel haṣ-ṣorer ḥmil yimaḥ šǝmo šey-yeleḵ wǝ-yašuḇ lǝ-ḇeto ‘and as soon as the king was sitting on his royal throne, he immediately wrote letters to the enemy Chmielnicki – may his name be blotted out – telling him to go home’154 5.2 Conditional clauses There are several real conditional clauses attested in Yeven Meṣula. Some have a future sense, as shown in (81) and (82), and the others have a past habitual sense, shown in (83) and (84). All protases are introduced by the subordinator ‫ אם‬ʾim ‘if’. The future conditionals have yiqṭol verbs in both the protasis and apodosis. Of the past habitual conditionals, the first is comprised of a periphrastic construction in both the protasis and apodosis, while the other has a qaṭal in the protasis and a periphrastic construction in the apodosis. All of these constructions are traceable to Mishnaic Hebrew.155 (81) ‫אם אנו נמתין עד שיבואו היונים לעיר יעשו בנו כלה ונחרצה‬ ʾim ʾanu namtin ʿad šey-yaḇoʾu hay-yǝwanim la-ʿir yaʿaśu banu kalla wǝ-neḥraṣa ‘if we wait until the Ukrainians arrive in the city, they will destroy us completely’156 Sic; = ‫דוכס‬. Hannover, 1653, p. 8. 152 Sic; = ‫דוכס‬. 153 Hannover, 1653, p. 8. 154 Hannover, 1653, p. 14. 155 Pérez Fernández, 1999, pp. 213–216. 156 Hannover, 1653, p. 4. 150 151 173 (82) ‫אם תשלחו יד בשרים ושמעו כל מלכי אדום וינקמו נקמתם מכל אחינו שבגולה‬ ʾim tišlǝḥu yad baś-śarim wǝ-šamʿu kol malḵe ʾedom wǝ-yinqǝmu niqmatam mik-kol ʾaḥenu šeb-bag-gola ‘if you lay a hand on the lords and all the Catholic kings hear of it, they will take revenge on all our brethren in exile’157 (83) ‫ואם היו רוצים לילך לדרכם היו נותנין להם צדה לדרך‬ wǝ-ʾim hayu roṣim leleḵ lǝ-darkam hayu notǝnin lahem ṣeda lad-dereḵ ‘and if they wanted to go on their way, they would give them provisions for the road’158 (84) ‫] היו מלבישים אותם‬...[ ‫ואם באו מארץ מרחקי׳ או ממקומו׳ אחרי׳‬ wǝ-ʾim baʾu meʾereṣ merḥaqim ʾo mim-mǝqomot ʾaḥerim […] hayu malbišim ʾotam ‘and if they came from a faraway land or from other places […] they would provide them with clothes’159 In one case, Hannover employs a fusion of biblical and post-biblical constructions in his real conditional: the apodosis is introduced by a yiqṭol, but this is prefixed by the waw-conjunctive, which echoes the biblical use of the wawconsecutive in real conditional apodoses.160 As discussed elsewhere in this study, this mix of biblical and rabbinic elements is a common feature of Hannover’s writing, and is also a common feature of 19th century Eastern European Hebrew, though this precise feature is not attested in Maskilic or Hasidic narrative literature. Further research is needed in order to ascertain whether it is an element of other types of Ashkenazic Hebrew. (85) ‫אם אנו הורגים לכולם ויחמלו עם פולי״ן על היונים‬ ʾim ʾanu horǝgim lǝ-ḵulam wǝ-yaḥmǝlu ʿam polin ʿal hay-yǝwanim ‘if we kill them all, the people of Poland will have pity for the Ukrainians’161 There is also an irreal conditional, with a verbless protasis introduced by ‫לולא‬ lule ‘if not’ and an apodosis with a qaṭal of the root .‫ה‬.‫י‬.‫ ה‬h.y.h., shown in (86). Interestingly, in contrast to the real past habitual conditionals shown above, this construction most closely resembles biblical irreal conditionals, which are likewise introduced by ‫ לולא‬lule ‘if not’.162 This is further evidence of the fusion of biblical and post-biblical elements present throughout Hannover’s text. 157 Hannover, 1653, p. 7. Hannover, 1653, p. 20. 159 Hannover, 1653, p. 20. 160 See Waltke and O’Connor, 1990, pp. 526–527. 161 Hannover, 1653, p. 11. 162 Waltke and O’Connor, 1990, 637–638. 158 174 (86) ‫כי לולא זאת לא היה תקומה חלילה לשארית ישר׳‬ ki lule zot lo haya tǝquma ḥalila li-šeʾerit yiśraʾel ‘and were it not for that, there would, God forbid, have been no survival for the remnant of Israel’163 5.3 Inconsistent use of the accusative marker ‫ את‬ʾet A characteristic feature of Hannover’s syntax is the inconsistent use of the accusative marker ‫ את‬ʾet. This marker is a standard feature of the biblical and rabbinic strata of Hebrew.164 However, it is commonly omitted in a variety of medieval Hebrew texts, including Rashi’s commentaries, the Sefer Ḥasidim, Spanish-Provençal Hebrew prose165 and Arabic translations.166 Rabin167 suggests that the medieval tendency to omit the particle is rooted in Paytanic Hebrew,168 and that this is itself based on Biblical Hebrew poetry, in which ‫את‬ ʾet is much less common than in biblical prose. Any such tendencies are likely to have been compounded by the fact that the medieval authors, like Hannover, spoke vernaculars lacking such a particle. As in the case of most other features discussed in this study, 19th century Hasidic Hebrew authors also frequently omit the particle.169 There are no clear patterns governing Hannover’s employment of the marker. It is likely that, as in the case of other varieties such as Hasidic Hebrew, which make use of the marker in a similarly inconsistent manner, Hannover consciously recognised it as an intrinsic element of the Hebrew prose style, but often unintentionally omitted it because such a form was not a feature of his Yiddish vernacular and therefore did not come naturally to him. Examples (87)–(89) illustrate cases where Hannover did employ the marker: (87) ‫ויהי כשמוע המלך והשרים את הדבר הזה היה כמצחק בעיניהם‬ wa-yhi ki-šmoaʿ ham-meleḵ wǝ-haś-śarim ʾet had-daḇar haz-ze haya kǝ-miṣḥaq bǝ-ʿenehem ‘and when the king and the minister heard this matter, it was like a joke to them’170 (88) ‫ולאבד את כל היהודי׳ ואת כל חיל עם פולין‬ u-lǝ-ʾabbed ʾet kol ha-yhudim wǝ-ʾet kol ḥel ʿam polin ‘and to destroy all the Jews and all the might of the people of Poland’171 163 Hannover, 1653, p. 5. Rabin, 2000, p. 117. 165 Rosén, 1995, pp. 64–66; Rabin, 2000, p. 117. 166 Goshen-Gottstein, 2006, p. 111. 167 Rabin, 2000, p. 117. 168 See Rand, 2006, pp. 258–259. 169 Kahn, 2015, pp. 280–282. 170 Hannover, 1653, p. 3. 171 Hannover, 1653, p. 5. 164 175 (89) ‫ובין כך ובין כך שלח המלך את השר אוסלי״נסקי משנה שלו אל המלך הקדרים‬ u-ḇen kaḵ u-ḇen kaḵ šalaḥ ham-meleḵ ʾet haś-śar oslinsqi mišne šello ʾel ham-meleḵ haq-qǝdarim ‘and meanwhile, the king sent his aide, the Lord Ossoliński, to the Tatar king’172 By contrast, (90)–(92) exemplify cases where he neglected to include it: (90) ‫ הדב׳ הזה ויחרד‬173‫ויהי כשמוע הדוכוס‬ wa-yhi ki-šmoaʿ had-duḵkas had-daḇar haz-ze way-yeḥerad ‘and when the duke heard this matter, he was afraid’174 (91) ‫לשמור העיר מן השונא‬ li-šmor ha-ʿir min haś-śone ‘to guard the town from the enemy’175 (92) ‫עד שלכדו המבצר ויהרגו כל היהודים‬ ʿad šel-laḵdu ham-miḇṣar way-yahargu kol ha-yhudim ‘until they captured the fortress and killed all the Jews’176 5.4 Use of -‫ ל‬lǝ- as accusative marker A striking and very common feature of Hannover’s writing is the use of the inseparable preposition -‫ ל‬lǝ- ‘to, for’ as a direct object marker. The preposition is attested with this function in conjunction with a variety of verbs and seems to be relatively productive, though its use is not uniform. A noteworthy aspect of this construction is that it seems to be used only with reference to animate objects and cities (which can be regarded as a sort of collective concentration of animate objects). Examples (93)–(96) illustrate this noteworthy construction. The phenomenon extends to the employment of -‫ ל‬lǝ- in conjunction with a pronominal suffix, as in (97). This feature has some precedent in Late Biblical Hebrew and Rabbinic Hebrew;177 in both cases it is thought to be ascribable to influence from Aramaic, in which -‫ ל‬lǝ- is a standard accusative marker.178 However, it does not appear to be a feature of medieval forms of Hebrew, which use the accusative marker ‫ את‬ʾet or leave direct objects unmarked.179 Notably, it also appears to be absent 172 Hannover, 1653, p. 16. Sic; = ‫דוכס‬. 174 Hannover, 1653, p. 8. 175 Hannover, 1653, p. 8. 176 Hannover, 1653, p. 12. 177 Gesenius, 2006, p. 366; Segal, 1927, p. 168. 178 Rabin, 2000, p. 117–118; see also Nicolae and Tropper, 2010, pp. 30–31 and Bar-Asher Siegal, 2013, pp. 201–202 for details of the particle in Aramaic. 179 Rabin, 2000, pp. 117–118. 173 176 from 19th century Eastern European forms of Hebrew, in contrast to many of the other constructions discussed in this study. The fact that Yeven Meṣula does not exhibit any direct grammatical influence from Aramaic 180 suggests that the historical basis for Hannover’s use of this construction is its appearance in Biblical and Rabbinic Hebrew. However, the fact that the use of -‫ ל‬lǝas an accusative marker is not a prominent feature of either of these strata of Hebrew, combined with the fact that Hannover’s restriction of the construction to animate objects lacks clear precedent in biblical or rabbinic literature, raise the possibility that the canonical strata are not the sole source of the phenomenon in Yeven Meṣula. Perhaps unexpectedly, the most direct parallel for Hannover’s usage can be found in the pre-modern Hebrew writing of JudaeoSpanish speakers from the Ottoman Empire, which exhibits precisely the same phenomenon, including the restriction to animate objects.181 This intriguingly specific apparent link between Hannover’s text and that of Ottoman JudaeoSpanish-speaking writers requires further investigation in order to ascertain the extent of the similarities between these two forms of Diaspora Hebrew. Likewise, further research needs to be done into the language of other early modern and modern Ashkenazic Hebrew textual sources in order to determine whether this phenomenon was rooted in a more widespread usage in Eastern Europe as well. (93) ‫והיה משפיל להדוכסים והשרים שהיו מדת היונים‬ wǝ-haya mašpil lǝ-had-dukkasim wǝ-haś-śarim še-hayu mid-dat hay-yǝwanim ‘and he would bring down the dukes who were of the Greek Orthodox religion’182 (94) ‫ויכתירו לפאולוק שם‬ way-yaḵtiru lǝ-pawluq šam ‘and they crowned Pawliuk king there’183 (95) ‫וחרב בתי תפלותם והרג לכומרי׳ שבהם‬ wǝ-ḥaraḇ batte tǝp̄ illotam wǝ-harag lak-komǝrim/lǝ-ḵomǝrim šeb-bahem ‘and he destroyed their churches and killed (the) priests that were in them’184 180 Aramaic features in Yeven Meṣula are limited to a number of set phrases such as ‫נטורי קרתא‬ naṭore qarta ‘guardians of the city’ (Hannover, 1653, p. 8) and the use of the possessive particle -‫ ד‬dǝ- ‘of’ on one occasion, ‫ חילו רכב ופרשים דעם פולי״ן‬ḥelo reḵeḇ u-p̄ arašim dǝ-ʿam polin ‘his Polish army, chariots and horsemen’ (Hannover, 1653, p. 10). 181 Bunis, 2013, p. 60*. 182 Hannover, 1653, p. 1. 183 Hannover, 1653, p. 2. 184 Hannover, 1653, p. 5. 177 (96) ‫ונתן לו העצה שאוהביו יוציאו לחמיל י״מש מבי׳ האסורי׳‬ wǝ-natan lo ha-ʿeṣa še-ʾohaḇaw yoṣiʾu li-ḥmil yimaḥ šǝmo mib-bet ha-ʾasurim ‘and gave him the advice that his friends should take Chmielnicki – may his name be blotted out – out of prison’185 (97) ‫ומהר אנו מביאים לכם אל אחיכם שבקוסטנטי״נא‬ u-maher ʾanu mǝḇiʾim laḵem ʾel ʾaḥeḵem šeb-bǝ-qostantina ‘and we shall quickly take you to your brothers who are in Constantinople’186 6 Conclusion The Hebrew of Yeven Meṣula exhibits a fusion of characteristically biblical features (the wayyiqṭol, stative qaṭals with present reference and temporal constructions composed of a prefixed infinitive construct) and typically rabbinic elements (the masculine plural in nun, the qaṭal in past narrative sequences and periphrastic verbal constructions), in many cases employing the biblical and rabbinic features alongside each other. It also contains a number of features without clear precedent in Biblical or Rabbinic Hebrew (the retention of the definite article with inseparable prepositions, the indefinite article, definiteness of construct nouns and doubly definite construct chains, the avoidance of the dual, and erratic use of the definite direct object marker), but which are attested in other Eastern European forms of the language, specifically the writings of 19th century Maskilic and Hasidic authors as well as the Kiṣur Shulḥan ʿAruḵ and rabbinic responsa literature. Moreover, at least one of these features (fluctuation between the nun and mem plural endings) is found in medieval Ashkenazic Hebrew. Some of them also have parallels in the Hebrew composed by Judaeo-Spanish speakers and, more distantly, in Judaeo-Arabic. Finally, Yeven Meṣula exhibits a single feature (the use of the prefixed preposition -‫ ל‬lǝ- ‘to, for’ as a definite direct object marker in addition to the standard ‫ את‬ʾet) whose closest parallel seems to be in the Hebrew of Ottoman Judaeo-Spanish speakers. The overall similarity between Yeven Meṣula and other Eastern European forms of Hebrew, particularly those composed by 19th century adherents of the Maskilic and Hasidic movements, suggests that all of these authors may have been heirs to a shared Ashkenazic variety of Hebrew whose roots stretch back to at least the 17th century and possibly much earlier. Further research is needed to determine the geographical and chronological boundaries of this form of Hebrew and establish its links with other types of early modern and modern Diaspora Hebrew, as well as more broadly with Judaeo-Arabic and other Semitic languages. 185 186 Hannover, 1653, p. 3. Hannover, 1653, p. 5. 178 References Bacon, G., 2003, “ ‘The House of Hannover’: Gezeirot Tah in modern Jewish historical writing”, Jewish History 17/2, pp. 179–206. Bar-Asher Siegal, E., 2013, Introduction to the Grammar of Jewish-Babylonian Aramaic, Münster: Ugarit-Verlag. Bartal, I., 2005, The Jews of Eastern Europe, 1772–1881, C. Naor (trans.), Philadelphia: University of Pennsylvania Press. Betzer, T., 2001, History of the Hebrew Language: The Medieval Division, Unit 7: Rabbinic Hebrew (Hebrew), Tel Aviv: Open University. Blau, J., 1980, A Grammar of Medieval Judaeo-Arabic (Hebrew), 2nd edn, Jerusalem: Magnes Press. Bunis, D. M., 2013, “ ‘Whole Hebrew’: a revised definition”, in I. Bartal (ed.), A Touch of Grace: Studies in Ashkenazic Culture, Women’s History, and the Languages of the Jews Presented to Chava Turniansky, Jerusalem: Zalman Shazar Center for Jewish History, pp. 37*–68*. Ettinger, S., 2007, “Chmielnicki, Bogdan”, in M. Berenbaum and F. Skolnik (eds), Encyclopedia Judaica, 2nd edn, Detroit: Macmillan Reference USA, vol. IV, pp. 653–656. Gesenius, W., 2006, Gesenius’ Hebrew Grammar, E. Kautzsch (ed.), A. E. Cowley (trans.), Mineola, NY: Dover Publications. Goshen-Gottstein, M., 2006, Syntax and Vocabulary of Mediaeval Hebrew as Influenced by Arabic (Hebrew), S. Assif and U. Melammed (revs), Jerusalem: BenZvi Institute. Halpern, I., 2007, “Hannover, Nathan Nata”, in M. Berenbaum and F. Skolnik (eds), Encyclopedia Judaica, 2nd edn, Detroit: Macmillan Reference USA, vol. VIII, pp. 326–327. Hannover, N. N., 1652, Taʿame Sukka (Hebrew), Amsterdam. ———, 1653, Yeven Meṣula (Hebrew), Venice. ———, 1660, Safa Berura (Hebrew), Prague. ———, 1662, Shaʿare Ṣiyyon (Hebrew), Prague. Jacobs, N. G., 2005, Yiddish: A Linguistic Introduction, Cambridge: CUP. Joüon, P. and Muraoka, T., 2009, A Grammar of Biblical Hebrew, 2nd edn, reprinted with corrections, Rome: Gregorian Biblical Press. Kahn, L., 2009, The Verbal System in Late Enlightenment Hebrew, Leiden: Brill. ———, 2012a, “Biblical grammatical elements in the nineteenth-century Hasidic Hebrew tale”, Jewish Studies, an Internet Journal 11, pp. 323–344. ———, 2012b, “Grammatical similarities between Hasidic and Maskilic Hebrew narratives”, Hebrew Studies 53, pp. 261–283. ———, 2015, A Grammar of the Eastern European Hasidic Hebrew Tale, Leiden: Brill. ———, in press, “The Kitsur Shulḥan ‘Arukh, Hasidic tale, and Maskilic literature as exemplars of Ashkenazic Hebrew”, JQR. Katz, D., 1987, Grammar of the Yiddish Language, London: Duckworth. ———, 1993, “The Phonology of Ashkenazic”, in L. Glinert (ed.), Hebrew in Ashkenaz: A Language in Exile, New York: OUP, pp. 46–87. Khan, G. et al. (eds), 2013, Encyclopedia of Hebrew Language and Linguistics (4 vols), Leiden: Brill. van der Merwe, C. H. J., Naudé, J. A. and Kroeze, J. H., 1999, A Biblical Hebrew Reference Grammar, Sheffield: Sheffield Academic Press. Mishor, M., 2013, “Mood and modality: Rabbinic Hebrew”, in Khan et al., 2013, vol. II, pp. 690–693. 179 Nicolae, D. and Tropper, J., 2010, Biblisch-Aramäisch Kompakt, Kamen: Hartmut Spenner. Pat-El, N., 2009, “The development of the Semitic definite article: a syntactic approach”, Journal of Semitic Studies 54/1, pp. 19–50. Pérez Fernández, M., 1999, An Introductory Grammar of Rabbinic Hebrew, J. Elwolde (trans.), Leiden: Brill. Plokhy, S., 2015, The Gates of Europe: A History of Ukraine, London: Penguin. Rabin, C., 1968, “The tense and mood system of the Hebrew of Sepher Ḥasidim”, in Fourth World Congress of Jewish Studies, Jerusalem: World Union of Jewish Studies, vol. II, pp. 113–116. ———, 2000, The Development of the Syntax of Post-Biblical Hebrew, Leiden: Brill. Rand, M., 2006, Introduction to the Grammar of Hebrew Poetry in Byzantine Palestine, Piscataway, NJ: Gorgias Press. Rosén, H. B., 1995, Hebrew at the Crossroads of Cultures: From Outgoing Antiquity to the Middle Ages, Leuven: Peeters. Rubin, A. D., 2013a, “Definite article: pre-modern Hebrew”, in Khan et al., 2013, vol. I, pp. 678–682. ———, 2013b, “Indefinite article”, in Khan et al., 2013, vol. II, p. 265. Sáenz-Badillos, A., 1993, A History of the Hebrew Language, J. Elwolde (trans.), Cambridge: CUP. Sarfatti, G. B., 1989, “Definiteness in noun-adjective phrases in Rabbinic Hebrew” (Hebrew), in M. Z. Kaddari and S. Sharvit (eds), Studies in the Hebrew Language and the Talmudic Literature Dedicated to the Memory of Dr. Menaḥem Moreshet (Hebrew), Ramat Gan: Bar-Ilan University Press, pp. 153–167. ———, 2003, History of the Hebrew Language: The Medieval Division, Unit 5: The Language of the Translators from Arabic (Hebrew), Tel Aviv: Open University. Segal, M. H., 1927, A Grammar of Mishnaic Hebrew, Oxford: Clarendon Press. Sharvit, S., 2004, History of the Hebrew Language: The Classical Division, Unit 3: Talmudic Hebrew (Hebrew), Tel Aviv: Open University. Stampfer, S., 2003, “What actually happened to the Jews of Ukraine in 1648?”, Jewish History 17/2, pp. 207–227. ———, 2010, “Gzeyres Takh Vetat”, in J. P. Edelstein et al. (eds), YIVO Encyclopedia of Jews in Eastern Europe, http://www.yivoencyclopedia.org/article.aspx/ Gzeyres_Takh_Vetat, accessed 17 May 2017. Wagner, E.-M., 2010, Linguistic Variety of Judaeo-Arabic in Letters from the Cairo Genizah, Leiden: Brill. Waltke, B. K. and O’Connor, M. P., 1990, An Introduction to Biblical Hebrew Syntax, Winona Lake, IN: Eisenbrauns. Wertheimer, Y., 1975, “On the study of 19th century Hebrew: based on an analysis of the language of Yosef Rivlin and M. L. Lilienblum” (Hebrew), in H. Z. Hirschberg (ed.), Vatiqin: Studies on the History of the Yishuv (Hebrew), Ramat Gan: Rivlin Institute, pp. 149–161. Williams, R. J., 2007, Williams’ Hebrew Syntax, 3rd edn, J. C. Beckman (rev.), Toronto: University of Toronto Press. 180 Medieval Jewish Exegetical Insights into the Use of Infinitive Absolute as the Equivalent of a Preceding Finite Form FIONA BLUMFIELD Hebrew University Various aspects of the origin, use and semantics of the infinitive absolute in Biblical Hebrew have puzzled modern grammatical researchers. Contemporary scholarship has offered a variety of solutions to the problem, but none seem entirely satisfactory. However the interpretation of the infinitive absolute was also considered by medieval exegetes including Rashi, Abraham Ibn Ezra and David Qimḥi, and while their interpretations are unfortunately often disregarded by modern scholars, they can be productively used for our own understanding of the infinitive absolute. Modern grammarians discuss verbal uses of the infinitive absolute. For example, Joüon and Muraoka comment on “the infinitive absolute as the equivalent of a preceding form” as follows: The infinitive absolute quite often (especially in the later books) continues a preceding form. The Waw, which usually precedes the infinitive absolute sometimes has the value of a simple ‘and’, and sometimes that of an ‘and (then)’ of succession. The reasons which have motivated the choice of the infinitive absolute are not clearly understood: sometimes there is probably a desire for variety or a stylistic affectation; sometimes the author wished to use a form with a vague subject like ‘one’ or ‘they’. The infinitive absolute virtually has the same temporal or modal value as the preceding verb. 1 Joüon and Muraoka bring examples of infinitive absolute after qaṭal, weqaṭal, yiqṭol, wayyiqṭol, jussive, imperative, participle and infinitive construct. We should take note of the parenthetical comment of Joüon and Muraoka, “especially in the later books”, when discussing this use of infinitive absolute as the equivalent of a preceding form. This diachronic aspect is especially emphasised by Ohad Cohen where he explains that the most frequent use (73.5%) of the infinitive absolute in the period of the First Temple was as the ‘tautological infinitive’ which emphasised the idea expressed in the verb. However, 1 Joüon and Muraoka, 1996, §123x. 181 in the Hebrew of the Second Temple period there are only nine occurrences of the tautological infinitive, whereas in 80% of the occurrences of this form in the Second Temple period it is used as a replacement for a conjugated verb; in most cases it is used as a sequential form and may come in sequence after various different verbal forms, so that it may be described as an “unmarked sequential form that receives its chronological and modal meaning from the context”.2 Steven Fassberg also states that the data show clearly that the phenomenon is less frequent in books from the First Temple Period than in compositions from the Second Temple Period and from the period of transition between the First and Second Temple Periods (Jeremiah).3 Given that this phenomenon is far more prevalent in the later books of the Old Testament, it is of great interest to note that scholars trace the origin of this usage to ancient Semitic languages. Steven Fassberg discusses this use of “the infinitive absolute with the conjunctive waw functioning as a finite verb, a use that is also known in El-Amarna, Ugaritic, and Phoenician”, and he asks the question why did the number of occurrences of the infinitive absolute as a finite verb increase in the late books of the Old Testament at a time when other uses of the infinitive absolute were disappearing? 4 Fassberg summarises the views of modern scholars to date, among whom we may note C. H. Gordon who connected the early northern Canaanite evidence with the data from the postexilic period and attributed the phenomenon in part to the ‘reunion of far-northern Jews with their Judean coreligionists during the Restoration’.5 Fassberg also cites Hammershaimb, who points out how rare this usage is in the Hebrew Bible (between 45 and 58 instances, depending on whether textual errors are taken into consideration or not) and states that “according to the old principle of always preferring the more complicated reading, I consider the infinitive to be the original form”.6 Fassberg is not convinced by the idea that the increased use of the infinitive absolute as finite verb reflects the spoken Hebrew of the period, because the phenomenon is “absent from Ben Sira, the Bar Kokhba letters and Tannaitic Hebrew, and it is infrequent in the Dead Sea Scrolls”.7 Fassberg concludes 2 Cohen, 2010, unit 22. Fassberg, 2008, p. 50. 4 Fassberg, 2008, pp. 49–50. 5 Fassberg, 2008, p. 54 and n. 37 there; see also Joüon and Muraoka, 1996, § 123x n. 1. 6 Hammershaimb, 1963, p. 91 n. 1. 7 Fassberg, 2008, p. 57. 3 182 that “the feature is a classicism”, that is, a borrowing from the First Temple Period, and that “some scribes chose to imitate this classicism. It, like other waw-consecutive syntagms, must have been viewed as elegant and elevated style”.8 Fassberg finds support for his theory in the high number of examples of the infinitive absolute as finite verb in Jeremiah, where the language reflects the beginning of a transitional stage between the Hebrew of the preexilic and the post-exilic periods, and as such, is full of classical as well as some post-classical features.9 Fassberg also cites the book of Esther (where there are comparatively many examples of waw plus infinitive absolute sequential to a finite verb) since “it has been demonstrated that its author deliberately imitated the language and literary motifs of the Joseph cycle, which was written in pure classical Hebrew”.10 The only example in Genesis of infinitive absolute as finite verb is in the Joseph cycle, Genesis 41:43, which we discuss in the context of medieval exegesis below. A noteworthy contribution to this discussion of infinitive absolute was made by A. Rubinstein,11 who is quoted by Waltke and O’Connor in their discussion of this phenomenon: The purpose served by this construction can just as well be served by other constructions; its use has been explained as an expression of a desire for stylistic variation, but this explanation does not account for its predominance in late Biblical Hebrew. Rubinstein suggests that it is the result of the disappearance of waw-consecutive forms in Late Hebrew: ‘it is at least significant that in the preponderant majority of our instances the [infinitive absolute] occurs precisely at the point where one would expect a transition to the appropriate consecutive form of the verb. His further thesis that the substitution belongs not to the original text but to the work of scribes and copyists lacks convincing evidence.’12 Overall we see that although modern scholars have demonstrated that the infinitive absolute form in place of a finite verb represents an ancient Semitic usage, opinions differ on the reasons for the increased use of this phenomenon in Late Biblical Hebrew. Although Fassberg’s argument that this is a classicism is convincing, we are still left with the ultimate question of why this construction is used in specific contexts. Rubinstein refers to this construction (finite verb continued by infinitive absolute) as VIA and argues that it occurs in forty-five passages.13 He states 8 Fassberg, 2008, p. 58. Fassberg, 2008, p. 58. 10 Fassberg, 2008, p. 58. 11 Rubinstein, 1952, pp. 362–367. 12 Waltke and O’Connor, 1990, pp. 595–596, §35.5.2, para b. 13 Rubinstein, 1952, p. 363. 9 183 that “there can be little doubt that the purpose which the VIA construction serves, is normally effected in Biblical Hebrew by other syntactical means” and he chooses six biblical verses which in his view best illustrate this idea.14 The present article seeks to examine the verses discussed by Rubinstein, as well as some other verses quoted in Joüon and Muraoka,15 and to examine comments of the medieval Jewish exegetes on these verses where they occur, since, in our view, the answers provided by contemporary philology are not totally satisfactory and the insights from medieval exegetes are useful despite the fact that they are not based on the critical methodology of today’s research. It could then be possible to use the linguistic insights of the medieval exegetes to shed light on the use of infinitive absolute sequential to a finite verb even in verses where the medieval comments are not available to us. We will see that these medieval comments shed light specifically on the semantic value of the use of infinitive absolute in context, regardless of the origin of the construction. The semantic value in such a context is often one of continuity and/or repetition. The context here may refer to the action expressed by the infinitive absolute itself, or to the permanence of the state brought about by the action expressed in the infinitive absolute, or even for a different action than that expressed by the infinitive absolute, but nevertheless for an action that is evident within the narrative context. We begin our study with a discussion of Isaiah 6:10: ‫יִשמָ ע ולְ בָ בֹו יָבִ ין‬ ְ ‫יִ ְראֶ ה בְ עֵינָיו ובְ אָ ְזנָיו‬-‫הָ ָעם הַ זֶה וְ אָ ְזנָיו הַ כְ בד וְ ֵעינָיו הָ ׁשַ ע פֶן‬-‫הַ ְׁשמן לֵב‬ :‫וָשָ ב וְ ָרפָא לֹו‬ There is an exegetical problem in understanding the three hifil imperatives in this verse – ‫הַ ְשמֵ ן‬, ‫ הַ כְ בֵ ד‬and ‫ – הָ שַ ע‬since one wonders why the prophet would be asked to ‘fatten’ the hearts of the people, ‘harden’ their ears and ‘smear’ their eyes. Why should the prophet be asked to increase their obstinacy? Rashi comments as follows: ‫ לִ בָ ם הֹולְֵך וְ הַ ְשמֵ ן ואזניו‬.‫ לשון הָ לֹוְך לשון פָעֹול‬.‫לִ בֹו‬-‫ וְ הַ כְ בֵ ד אֶ ת‬:)‫ (שמות ח יא‬:‫כמו‬ ‫הולכים הָ לֹוְך וְ הַ כְ בֵ ד ִמ ְשמֹו ַע‬ ‘Like Exodus 8:11: “and (he) hardened his heart”, an expression of continuity. Their heart continually becomes fatter and his ears are becoming heavier and heavier so that they do not hear.’16 Rashi explains that these forms are not imperatives but are hifil infinitive absolute forms; he labels them ‫לשון הָ לֹוְך לשון פָעֹול‬. They could have been taken 14 Rubinstein, 1952, p. 363. Joüon and Muraoka, 1996, §123x. 16 Translations of biblical verses and medieval Hebrew commentaries are partly based on those of the Jewish Study Bible (Berlin and Brettler, 2004) and Silberman’s Pentateuch with Rashi’s Commentary, but are also partly my own translations. 15 184 as hifil imperative forms (in the masculine singular) since these have the same morphological form as hifil infinitive absolute forms, but in the light of the exegetical problem mentioned above, Rashi prefers to avoid the exegetical problem and to take these forms as infinitive absolutes which express continuous action.17 The infinitive absolute forms in Isaiah 6:10 are not in the category of continuing a preceding form18 but rather in the category of being equivalent to a finite form at the beginning of a sentence,19 but the important point here (with regard to ensuing discussion) is the resort to infinitive absolute to express ‘continuity’. It is worth noting however that Ibn Ezra takes a different view from Rashi here and prefers to understand the three forms as imperatives because infinitives could not be the verbal form of a main clause followed by a subordinate clause introduced by ‫‘ פֶן‬lest’. Ibn Ezra implies that the people are being deprived of the ability to repent as punishment for their sins. David Qimḥi brings both views, that of Rashi as well as that of Ibn Ezra. In bringing the view that is similar to that of Rashi, Qimḥi relates to the imperative forms in Isaiah 6:9 as well as in Isaiah 6:10. Isaiah 6:9 reads as follows: :‫תֵ דָ עו‬-‫תָ בִ ינו ְּוראּו ָראֹו וְ אַ ל‬-‫וַיאמֶ ר לְֵך וְ אָ מַ ְרתָ ָלעָם הַ זֶה ִּׁש ְמעּו שָ מֹו ַע וְ אַ ל‬ ‘And He said, “Go, say to this people: Hear, indeed, but do not understand and see, indeed, but do not know.” ’ Qimḥi’s comment relates to verses 9 and 10 as follows: ‫ אתם‬:‫ אם תשמעו ותראו; כלומר‬:‫ כלומר‬,‫ויש לפרש ִש ְמעו ְוראו – ציווי במקום אית”ן‬ ‫ ואין אתם מבינים; ותראו נפלאות‬,‫שומעים באזניכם דברי הנביאים המוכיחים אתכם‬ ‫ ואין אתם משגיחים אלי ואתם‬,‫ לא תשימו לב‬:‫ ואין אתם יודעים; כלומר‬,‫הבורא‬ ‫ שלא תשמעו ולא תראו‬,‫משמינים בכוונה לבבכם ומכבידים אזניכם ומשיעים עיניכם‬ ‫ כי אינכם חפצים בתשובה‬,‫ולא תבינו ולא תדעו‬ ‘Some interpret “hear” and “see” as imperative forms functioning as yiqṭols, that is: if you hear and see; that is, you listen with your ears to the words of the prophets who rebuke you, but you do not understand, and you see the wonders of the Creator, but you do not know; that is, you do not lay it to heart and you do not pay attention to Me, but you fatten your hearts on purpose and make heavy your ears and smear your eyes, that you will not hear and will not see and will not understand and will not know, for you do not wish for repentance.’ 17 See Englander, 1939, pp. 391–392, 408–409. Englander explains the terminology and the grammatical elements in Rashi’s commentaries. He states that Rashi uses no special term for infinitive absolute but frequently identifies infinitive absolute by the term ‫ פָעֹול‬which “sometimes has the force of a customary or frequentative act”. 18 Joüon and Muraoka, 1996, §123x. 19 Joüon and Muraoka, 1996, §123u–v. 185 We see that Qimḥi understands the imperatives in verse 9 – ‫ ִש ְמעו‬and ‫ – ְוראו‬as well as the imperatives in verse 10 – ‫הַ ְשמֵ ן‬, ‫ הַ כְ בֵ ד‬and ‫ – הָ שַ ע‬as yiqṭol forms, and here we can see a small difference between Rashi and Qimḥi. Rashi understood ‘the heart of this people’ as the subject of the verb ‘becomes fat’, that is, ‘the heart of this people becomes fatter and fatter’, and similarly ‘his ears become heavier’ where ‘his ears’ are the subject of ‘become heavier’, and also ‘his eyes become smeared’.20 However, from the wording of Qimḥi’s comments, it seems that he understood ‘heart’ as the object of the verb ‘fatten’ and he took ‘this people’ as the subject of the verb ‘fatten’; he also took ‘your ears’ and ‘your eyes’ as the objects of the verbs ‘make heavy’ and ‘smear’. According to Qimḥi’s interpretation, the hifils would be causative hifils. Rashi opened his comment on Isaiah 6:10 by quoting Exodus 8:11: ‫לִ בו‬-‫ַוי ְַרא פ ְַרעה כִ י הָ יְ תָ ה הָ ְרוָחָ ה וְ הַ כְ בד אֶ ת‬ ‫וְ לא שָ מַ ע ֲאלֵהֶ ם‬ ‘And Pharaoh saw that there was relief and he hardened his heart and did not listen to them.’ In this verse, the infinitive absolute form ‫ וְ הַ כְ בֵ ד‬indeed fits the category discussed by Joüon and Muraoka of “the infinitive absolute as the equivalent of a preceding form”21 since it is sequential to the wayyiqṭol form ‫‘ ַוי ְַרא‬and he (Pharaoh) saw’. So let us study the comments of Rashi on this verse: :‫ כמו‬,‫לְ שֹון פָעֹול הוא‬ :‫ וַיִ סַ ע אַ בְ ָרם הָ לֹוְך וְ נָסֹו ַע הַ נֶגְ בָ ה‬:]‫[בראשית יב ט‬ :‫מֹואָ ב‬-‫בָ ּה וְ הַ כֹות אֶ ת‬-]‫ [ ַויַכו‬:]‫וכן [מלכים ב ג כד‬ ‫יִ שָ י בְ ִת ְתָך לֹו לֶחֶ ם וְ חֶ ֶרב וְ שָ אֹול לֹו‬-‫ לָמָ ה קְ שַ ְרתֶ ם ָעלַי אַ תָ ה ובֶ ן‬:]‫[שמואל א כב יג‬ :‫בֵ אלהים‬ :‫ ַו ַיכֵהו הָ ִאיש הַ כֵה ופָצ ַע‬:]‫[מלכים א כ לז‬ Rashi opens his comment here with the expression ‫לְ שֹון פָעֹול‬, which reflects his comment on Isaiah 6:10 – ‫ – לשון הָ לֹוְך לשון פָעֹול‬and may be translated as ‘an expression of continuity’. The full translation of Rashi’s comments on ‫ וְ הַ כְ בֵ ד‬here are as follows: ‘This is an expression of “continuity” like: Genesis 12:9: and Abraham journeyed continually towards the South; II Kings 3:24: and they smote Moab with great force; I Samuel 22:13: why have you conspired against me, you and the son of Jesse, by your giving him bread and a sword and you consulted G-d for him; I Kings 20:37: and the man smote him, smiting and wounding.’ 20 21 See Gesenius, 1910, §53d for this ‘inwardly transitive’ use of hifil. Joüon and Muraoka, 1996, §123x. 186 We see that Rashi brings examples from four different biblical verses. It is noteworthy that three of Rashi’s examples do not illustrate the use of infinitive absolute instead of an inflected form, but rather have infinitive absolute used after the finite verb for emphasis; the use of infinitive absolute after the finite verb may emphasise the nuance of continuity, and it seems to be this aspect of the function of infinitive absolute that Rashi is keen to highlight here: ‘Pharaoh continued hardening his heart’. Only in one of Rashi’s examples (I Samuel 22:13) is it possible that the infinitive absolute is used in place of the finite verb. We can therefore conclude at this stage that the comments of Rashi and Qimḥi on Isaiah 6:9–10 and of Rashi on Exodus 8:11 indicate clearly that these scholars discerned the desire to emphasise a nuance of continuity and also repetition in the context. However, this insight is not mentioned at all in the analysis presented by Rubinstein,22 where he cites Exodus 8:11 as his very first example to illustrate the “redundancy of function” of the VIA construction. He points out that in the Hebrew Pentateuch of the Samaritans we find the wayyiqṭol form ‫ ַויַכְ בֵ ד‬in Exodus 8:11 instead of the infinitive absolute form ‫וְ הַ כְ בֵ ד‬. Rubinstein brings a second example of VIA from Isaiah 37:18–19, where the infinitive absolute form ‫ וְ נָתן‬in verse 19 follows the qaṭal inflected form ‫ הֶ ח ֱִריבו‬in verse 18: ‫אלהיהם בָ אֵ ש‬-‫ וְ נָתֹׁ ן אֶ ת‬... ‫הָ א ֲָרצֹות‬-‫כָל‬-‫הח ֱִּריבּו מַ לְ כֵי אַ שור אֶ ת‬ ‘The kings of Assyria destroyed all the lands … and put their gods to fire.’ Rubinstein points out that the parallel text to the Isaiah in II Kings 19:17–18 has the finite form with waw and qaṭal ‫ וְ נ ְָתנו‬and the Dead Sea Isaiah scroll (DSI) has the wayyiqṭol ‫ויתנו‬. Rubinstein ascribes this use of infinitive absolute “to scribes or copyists, who resorted to it when they could not be certain of the form of a finite verb”.23 Such uncertainty might arise from a manuscript being indistinct in places, or if a copyist had before him divergent readings, or if he was writing from memory. Thus the reading with infinitive absolute in Isaiah 37:19 “may well have been a scribe’s way of resolving the difficulty created by alternative readings like those of DSI and II Kings”.24 Just as Rashi’s insights on Exodus 8:11 were not included in Rubinstein’s analysis, similarly Rashi’s comments on Isaiah 37:19 are not included. Here are Rashi’s comments on Isaiah 37:19: ‫אֶ ֶרץ מצרים (בראשית מא) לשון פָעֹול אָ מֹור זָכֹור‬-‫כָל‬-‫וְ נָתן כמו וְ נָתֹון אתֹו ַעל‬ 22 See discussion above; Rubinstein, 1952, pp. 362–367. Rubinstein, 1952, p. 365. 24 Rubinstein, 1952, p. 365. 23 187 Rashi has found another example of the use of the infinitive absolute form ‫ וְ ָנתֹון‬in Genesis 41:43, where the infinitive absolute form is used in place of the finite form following two wayyiqṭol forms as follows: ‫לֹו‬-‫ֲשר‬ ֶ ‫ַוי ְַרכב אתֹו בְ ִמ ְרכֶבֶ ת הַ ִמ ְשנֶה א‬ ‫וַיִּ קְ ְראּו לְ ָפנָיו אַ בְ ֵרְך‬ :‫אֶ ֶרץ ִמצְ ָריִ ם‬-‫וְ נָתֹון אתֹו עַל כָל‬ ‘And he made him to ride in the second chariot which he had and they called out before him Ἀvreḵ and they placed him over the whole land of Egypt.’ In his comment on Isaiah 37:19 where he quotes Genesis 41:43, Rashi uses his customary terminology ‫לשון פָעֹול אָ מֹור זָכֹור‬, which indicates that the use of infinitive absolute here concerns the nuance of continuity. This nuance of continuity certainly fits the context in Genesis 41:43, since the intention behind the appointment of Joseph was surely that he would be in charge of the whole land of Egypt on a continual basis; here, the idea of ‘continuity’ seems to apply more to the result of the action expressed by the infinitive absolute than to the actual process of appointment, although one could suggest that Joseph’s position would also depend on continued vigilance on the part of those who could influence events in Egypt at the time. Here again Rubinstein does not mention this insight in his comments: Similarly, if a copyist had before him the reading ‫ וַיִ קְ ְראו‬in Gen 41:43, he might be tempted to use the [infinitive absolute] ‫ וְ נָתֹון‬rather than commit himself to a finite tense, even if such a tense were present in the document before him; this would leave it to be inferred whether the subject of the verb-substitute is again Pharaoh as in the first verb of the series or is to be understood in an impersonal sense as in the second verb.25 Rubinstein picks up on the fact that the first verb – ‫ – ַוי ְַרכֵב‬is third person masculine singular ‘and he made to ride’ but the second verb is third person masculine plural – ‫‘ – ַו ִיקְ ְראו‬and they called out’, and the third verb in the sequence – ‫ – וְ נָתֹון‬is infinitive absolute. We saw above that Fassberg sees significance in the fact that the only example of infinitive absolute as finite verb in Genesis occurs in the Joseph cycle, since the author of Esther deliberately imitated the language of the Joseph cycle and we find the same infinitive absolute verb form ‫ וְ נָתֹון‬in Esther 6:8–9. Rashi’s use of the term ‫ זָכֹור‬is also significant because the learned reader will immediately recall Exodus 20:8: ‘Remember the Sabbath day to keep it holy.’ 25 Rubinstein, 1952, pp. 365–366. 188 ‫יֹום הַ שַ בָ ת לְ קַ ְדשֹו‬-‫זָכֹור אֶ ת‬ In this verse we note the use of the infinitive absolute form ‫ זָכֹור‬and not the imperative form ‫זְכר‬, and Rashi’s comment on Exodus 20:8 again highlights the continuous nature of remembering the Sabbath day: ‫ ַו ֵילְֶך ִאתָ ּה‬:‫ שמואל ב ג טז‬.‫ אָ כֹול וְ שָ תֹו כִ י מָ חָ ר נָמות‬:‫ ישעיה כב יג‬:‫ כמו‬,‫לשון פָעֹול הוא‬ ‫ ְתנו לֵב לִ זְכֹור תָ ִמיד את יום הַ שַ בָ ת שֶ ִאם נִ זְדַ מֵ ן לְ ָך חֵ פֶץ‬,‫ וכן פתרונו‬.‫ִאישָ ּה הָ לֹוְך ובָ כה‬ :‫ ְתהֵ א מַ ז ְִמינֹו לְ שַ בָ ת‬,‫ָיפֶה‬ ‘‫ זָכֹור‬is an expression of continuity, like Isaiah 22:13: “let us eat and drink, for tomorrow we shall die”; II Samuel 3:16: “and her husband went with her, weeping as he went”. This is the explanation; pay attention to remember continually the Sabbath day, that if something nice comes your way, set it aside for the Sabbath.’ We see here that Rashi compares ‫ זָכֹור‬with two examples. In Rashi’s first example (Isaiah 22:13), the two infinitive absolute forms ‫ אָ כֹול‬and ‫ וְ שָ תֹו‬are used in place of finite verbs, ‘let us eat and drink’. In Rashi’s second example (II Samuel 3:16), the infinitive absolute forms ‫ הָ לֹוְך ובָ כה‬coming after the finite verb ‫ ַו ֵילְֶך‬emphasize the nuance of continuity, ‘her husband went with her weeping as he went’ (Phalti went with Michal when David insisted on her return to him). It is noteworthy that the function of ‫ זָכֹור‬in Exodus 20:8 is not the same as the function of the infinitive absolute in either of the examples brought by Rashi, since ‫ זָכֹור‬in Exodus 20:8 comes in place of imperative, but it is the nuance of continuity and repetition that Rashi wishes to illustrate. Rubinstein’s third example to illustrate redundancy of function of the VIA construction is Jeremiah 19:13,26 but in the absence of any relevant medieval exegesis on this verse we will postpone discussion of it until later. Rubinstein’s fourth example27 cites Ecclesiastes 9:11, which employs the VIA construction, whereas similar passages in Ecclesiastes 4:1 and 4:7 do not. This lack of consistency is insufficient to discount the validity of the insights of the medieval commentators, and indeed we do find a comment by Rashi on Ecclesiastes 9:11 which accords with his comments cited earlier. In Ecclesiastes 9:11 we find the infinitive absolute form with waw ‫ וְ ָראה‬following the qaṭal ‫ שַ בְ ִתי‬as follows: ‫הַ שֶ מֶ ש כִ י לא לַקַ לִ ים הַ מֵ רֹוץ‬-‫ׁשַ ְב ִּתי וְ ָראֹׁ ה תַ חַ ת‬ ‘I have further observed under the sun that the race is not won by the swift.’ Rashi comments on ‫ שַ בְ ִתי וְ ָראה‬with the words ‫כמו זָכֹור‬. We have already seen the significance of the comment ‫זָכֹור‬, and accordingly Rashi here discerns that the Preacher’s wisdom was surely based on continuous and prolonged observation and contemplation. 26 27 Rubinstein, 1952, p. 364. Rubinstein, 1952, p. 364. 189 There is also a relevant comment by Rashi on Ecclesiastes 8:9, where the infinitive absolute ‫ וְ נָתֹון‬follows the qaṭal ‫יתי‬ ִ ‫ר ִא‬:ָ ‫מַ ע ֲֶשה אֲשֶ ר ַנע ֲָשה תַ חַ ת הַ שָ מֶ ש‬-‫לִ בִ י לְ כָל‬-‫יתי וְ נָתֹון אֶ ת‬ ִּ ‫זֶה ָר ִּא‬-‫כָל‬-‫אֶ ת‬ ‘I saw all this and I put my mind to all that is done under the sun.’ Rashi comments here that history repeats itself: “whenever a man overpowers another, it turns to his own harm; so it was with ʿAmaleq, with Pharoah and with Nevuḵadneʾṣṣar”. Prolonged contemplation by the Preacher leads him to the conclusion that his observations will never change. Rubinstein’s fifth example is from I Chronicles 5:20, and his sixth example – the final example of the six he chooses whereby “this redundancy of function can best be illustrated”28 – is Ezekiel 23:47. Once again, in the absence of relevant medieval comments we will postpone discussion of these verses until later. As stated above, Rubinstein29 argues that the VIA construction (waw plus infinitive absolute as equivalent of a preceding form) occurs in forty-five passages, and so far we have discussed three of his examples where we have also found medieval exegesis. Joüon and Muraoka30 also present a list of examples, and from examination of these passages it emerges that one may uphold the idea that the nuance of continuity and repetition may be discerned in the great majority of the examples, although not in all of them. I would like to discuss a selection of these examples, beginning with examples where we find medieval exegesis. In Exodus chapter 18 we read of Jethro, the father-in-law of Moses, who gives advice regarding the adjudication of disputes. Verse 22 reads as follows: ‫עֵת‬-‫הָ ָעם בְ כָל‬-‫וְ ׁשָ פְ טּו אֶ ת‬ ‫הַ דָ בָ ר הַ גָדל י ִָּביאּו אֵ לֶיָך‬-‫כָל‬ ‫הַ דָ בָ ר הַ קָ טן יִּ ְׁשפְ טּו הֵ ם‬-‫וְ ָכל‬ :‫וְ הָ קל מֵ ָעלֶיָך וְ נָשְֹׁ אּו ִאתָ ְך‬ ‘They shall judge the people at all times, and they shall bring every major matter to you, and every minor matter they shall judge, and it will ease from upon you, and they shall bear with you.’ Here we find the waw plus hifil infinitive absolute ‫ וְ הָ קֵ ל‬sequential to the yiqṭol ‫ יִ ְשפְ טו‬which is sequential to the yiqṭol ‫ יָבִ יאו‬which is sequential to the weqaṭal ‫וְ שָ פְ טו‬, so the five verbs in the verse present a sequence of weqaṭal, yiqṭol, yiqṭol, waw plus infinitive absolute and weqaṭal. Rashi comments on the waw plus infinitive absolute as follows: 28 Rubinstein, 1952, p. 363. Rubinstein, 1952, p. 363. 30 Joüon and Muraoka, 1996, §123x. 29 190 .‫ כד) לשון הווה‬:‫מֹואָ ב (מלכים ב ג‬-‫ יא) וְ הַ כֹות אֶ ת‬:‫לִ בֹו (לעיל ח‬-‫ כמו וְ הַ כְ בֵ ד אֶ ת‬,‫והקל‬ ‘‫ והקל‬denotes ongoing action,31 like the word ‫ וְ הַ כְ בֵ ד‬in the phrase “and he kept hardening his heart” (Exodus 8:11) and like the word ‫ וְ הַ כֹות‬in the phrase “and he kept smiting Midian” (II Kings 3:24).’ We may also note Rashi’s consistency here, since he refers back to our starting point of Exodus 8:11, ‘and Pharaoh continued hardening his heart’, where he also quoted II Kings 3:24. In the context of Exodus 18:22 it is absolutely clear that Jethro’s advice is intended for the long term! In Leviticus chapter 25 we read about the Sabbatical Years and the Year of the Jubilee. In verse 14 we read: :‫אָ חִ יו‬-‫תֹונו ִאיש אֶ ת‬-‫ ִּת ְמכְ רּו ִמ ְמכָר ַלע ֲִמיתֶ ָך אֹו קָ נֹׁה ִמיַד ע ֲִמיתֶ ָך אַ ל‬-‫וְ כִ י‬ ‘When you make a sale to your fellow or when you buy land from the hand of your fellow, do not victimise one another.’ In the Hebrew, the verb ‘when you buy’ is ‫קָ נה‬, which is qal infinitive absolute, apparently coming in place of yiqṭol since it is sequential to a yiqṭol ‫ ִת ְמכְ רו‬and not preceded by waw, since it follows ‫‘ אֹו‬or’. Here, the commentary by Ibn Ezra presents a different approach from that of Rashi. Ibn Ezra comments as follows on the infinitive absolute form ‫קָ נה‬: ‫ ורבים כמוהם‬.‫ וכן זכור את יום השבת‬.‫ ויחסר מקור (מקום) קניתם קנה‬.‫שם הפועל‬ Note the disputed reading32 – )‫מקור (מקום‬. We may translate as follows:33 ‘This word is a verbal noun. Scripture here employs brevity, instead of supplementing the infinitive with a verb in the indicative mood. The same construction appears in “Remember the Sabbath day” (Exodus 20:8) as well as in many other places.’ In other words, Ibn Ezra takes the view that we have here an elliptical expression, where we would have expected the infinitive absolute to be followed by the yiqṭol, ‫קָ נה ִתקְ נֶה‬. This use of infinitive absolute before or after a finite verb usually marks some kind of emphasis34 and not necessarily just a nuance of 31 See Englander, 1939, 408–409. Englander is aware of only one instance when the term ‫הווה‬ is used by Rashi for a participle form (Genesis 15:17) and he explains that “R very frequently applies the term ‫ הווה‬to various verbal forms when he believes that such forms have the force of customary action” (p. 392). Englander further explains that the term ‫“ הווה‬is applied very often to a form in the perfect and imperfect tense, to an infinitive absolute, or to a participle when R deems such forms to have the force of a frequentative or customary action, or to a verb that R deems to have the force of long continuance”. Furthermore, surprisingly, Rashi seldom refers to the participle as having the ‫ הווה‬force, but in his comments Rashi sometimes equates the force of an infinitive absolute (or of an imperfect or perfect) with the force of a participle. 32 Benyowitz, 2006, pp. 222–223 and n. 37 there. 33 Shachter, 1986, p. 138. 34 Joüon and Muraoka, 1996, §123. 191 continuity and repetition; note here the comment of Joüon and Muraoka: “it is only from the context that the nuance added by the infinitive can be deduced in each case”.35 However Ibn Ezra’s reference here to Exodus 20:8 ‘Remember the Sabbath day’ (see discussion above) may indicate that in this verse (Leviticus 25:14) he does indeed discern some nuance of continuity or repetition in the command not to oppress one’s brother in the actions of buying and selling. Ibn Ezra is consistent in this approach to the infinitive absolute because he comments also on Daniel 9:5, where the waw plus infinitive absolute ‫ וְ סֹור‬follows a series of qaṭals: ‫ומ ִמ ְשפָטֶ יָך‬ ִ ‫חָ טָ אנּו וְ עָ וִּ ינּו וְ ִּה ְרׁשָ עְ נּו ּומָ ָר ְדנּו וְ סֹור ִמ ִמצְ ֹותֶ יָך‬ ‘We have sinned and we have gone astray and we have acted wickedly and we have been rebellious and have turned aside from Your commandments and Your judgments.’ Here, Ibn Ezra comments as follows: ‫וסור שם הפועל כמו זכור את יום השבת והטעם סור סרנו‬ Ibn Ezra explains that the word ‫ וְ סֹור‬is the infinitive absolute form and has the same force as ‫ זָכֹור‬in Exodus 20:8 ‘Remember the Sabbath day’; as we saw above, Rashi explained that the use of infinitive absolute here implies the continuous nature of this command. Ibn Ezra also suggests that here too the infinitive absolute is elliptical and we should understand it as if it were followed by the qaṭal ‫‘ סֹור סַ ְרנו‬we have indeed (continually and repeatedly) turned aside’. Ibn Ezra’s approach is evident also in his comment on Zechariah 7:5, which is part of an oracle by Zechariah on the lack of value in fasting in the absence of social virtues. ‫ישי ובַ ְשבִ יעִ י וְ זֶה ִשבְ עִ ים‬ ִ ‫צַ ְמתם וְ סָ פֹוד בַ ח ֲִמ‬-‫הַ כהֲנִ ים לֵאמר כִ י‬-‫ ַעם הָ אָ ֶרץ וְ אֶ ל‬-‫ ָכל‬-‫ֱאמר אֶ ל‬ :‫שָ נָה הֲצֹום צַ ְמתנִי אָ נִ י‬ ‘Say to all the people of the land and to the priests, saying: when you fasted and mourned in the fifth (month) and in the seventh (month) even these seventy years, did you fast for Me, even for Me?’ The Hebrew has the qal qaṭal form ‫‘ צַ ְמתֶ ם‬you fasted’ followed by waw plus infinitive absolute ‫וְ סָ פֹוד‬, and again Ibn Ezra comments that ‫ וְ סָ פֹוד‬is an abbreviated expression and should be understood as ‫וספ ְַדתֶ ם סָ פֹוד‬ ְ qaṭal followed by infinitive absolute: ‫שם הפועל וככה הוא וספדתם ספוד והנה אחז דרך קצרה‬ 35 Joüon and Muraoka, 1996, §123d. 192 Ibn Ezra does not add a specific comment hinting at a nuance of continuity in the context here, but we do see the consistency of his approach to the occurrence of infinitive absolute forms without finite forms, and the context itself clearly implies that the people were accustomed to fasting on a regular basis, every fifth and seventh month. Jeremiah 14:1–10 describes a prayer to G-d to end the prolonged drought, and verse 5 describes the effect of this drought on the animal kingdom. :‫הָ יָה דֶ שֶ א‬-‫אַ ֶילֶת בַ ָשדֶ ה יָלְ דָ ה וְ עָ זֹוב כִ י לא‬-‫כִ י גַם‬ ‘Even the hind abandons her young in the open at birth, for there is not a blade of grass anywhere’.36 We see here waw plus infinitive absolute ‫ וְ עָזֹוב‬sequential to the qal qaṭal ‫יָלְ דָ ה‬ ‘she gives birth’, that is, ‘she gives birth and abandons’. David Qimḥi comments on ‫ וְ עָזֹוב‬as follows: ‫מקור במקום עבר כאילו אמר וְ ָעזְבָ ה וכן (יחזקאל א יד) וְ הַ חַ יֹות ָרצֹוא וָשֹוב כאילו אמר‬ :‫ָרצו וְ שָ בו‬ Qimḥi here explains that the infinitive absolute ‫ וְ עָזֹוב‬is used in place of qaṭal ‫ וְ ָעזְבָ ה‬and he compares this to Ezekiel 1:14, which describes the Living Beings (Ḥayyot) in Ezekiel’s vision as ‘running and returning like the appearance of a lightning flash’ and where there are two infinitive absolute forms ‫ ָרצֹוא‬and ‫ וָשֹוב‬instead of qaṭal forms ‫ ָרצו‬and ‫וְ שָ בו‬. In both contexts (Jeremiah and Ezekiel) we can discern the nuance of continuity and repetition. Jeremiah describes a prolonged and agonising period of drought, while Ezekiel describes the repeated movements of the Ḥayyot. Qimḥi himself does not say anything specific here about the nuance of continuity, but he does refer the reader to Ezekiel 1:14, for which Rashi describes the continuous and repeated movements of the Ḥayyot as follows: ‫וממַ הֲרֹות ַלחֲזֹור‬ ְ ‫פירשו רבותינו ָרצֹוא וָשֹוב כְ ַלהֲבַ ת הַ כִ בְ שָ ן שֶ יֹוצְ אָ ה תָ ִמיד ִמפִ י הַ כִ בְ שָ ן‬ ‫וְ לִ י ָכנֵס‬ ‘Our Sages explained (BT Ḥagigah 13b): “they would run and return like the flame of a furnace, which constantly shoots out of the mouth of the furnace and hastens back to enter”.’ It is now time to return to the examples of infinitive absolute as the equivalent of a preceding form brought by Rubinstein and Joüon and Muraoka where we did not find any medieval exegesis. We argue that it is possible to apply the insights of the medieval exegetes as described above to other examples in the biblical narrative. The medieval exegetes demonstrated a marked tendency to 36 McKane, 1986, p. 315. 193 discern a nuance of continuity and/or repetition in contexts where the infinitive absolute occurs. As the medieval comments are not based on the critical methodology of today’s research, it is possible to suggest that this nuance of continuity or repetition could apply to the action expressed by the infinitive absolute itself, but it could also apply to other actions or situations in the immediate context, including the situation which results from the action expressed by the infinitive absolute form. Rubinstein’s third example to illustrate redundancy of function of the VIA construction is Jeremiah 19:13 where he points out that Jeremiah 32:29 “uses finite verbs only to express a similar idea”.37 In the absence of any relevant medieval exegesis on Jeremiah 19:13, it is nevertheless feasible to apply the linguistic insight of continuity and repetition of action to the use of the hifil infinitive absolute form with waw ‫‘ וְ הַ סֵ ְך‬poured libations’ which follows the piel qaṭal ‫‘ קִ ְטרו‬made offerings’: ‫גַגתֵ יהֶ ם לְ כל צְ בָ א הַ שָ מַ יִ ם וְ הַ סְך ְנסָ כִ ים לאלהים אחרים‬-‫לְ כל הַ בָ ִתים אֲשֶ ר קִּ ְטרּו עַל‬ ‘all the houses where they made offerings on their roofs to the whole host of Heaven and poured libations to other gods’ The infinitive absolute form following the qaṭal here clearly refers to a prolonged and continuous period of repeated idolatrous behaviour by the Judeans for which Jeremiah rebukes the people, threatening to make these very houses (where idolatry had been practiced) into ‘Tophet’.38 Rubinstein sees significance in the fact that in Jeremiah 32:29, where this information is repeated, the infinitive absolute ‫ וְ הַ סֵ ְך‬is not used again, but the hifil qaṭal ‫וְ הִ ִסכו‬: ‫גַגֹותֵ יהֶ ם לַבַ ַעל וְ ִּה ִּסכּו נְ סָ כִ ים‬-‫אֲשֶ ר קִ ְטרו ַעל‬ However, here it could be argued that this is a repetition of information already given, a situation where qaṭal is customarily used.39 In any event, lack of complete consistency in use of infinitive absolute is not a convincing argument to disregard the medieval insights here. Rubinstein’s fifth example is from I Chronicles 5:20: ‫בָ ְטחו בֹו‬-‫ָזעֲקּו בַ ִמלְ חָ מָ ה וְ נַעְ תֹור לָהֶ ם כִ י‬ ‘for they cried to G-d in the battle and He was entreated by them, because they put their trust in Him.’ 37 Rubinstein, 1952, p. 364. See II Kings 23:10 when Josiah defiled Tophet, which was a large place of fire in the Valley of Ben-Hinnom where children were burned in the Moleḵ rite. 39 Cohen, 2009, unit 26. 38 194 In this case, the infinitive absolute nifal ‫ וְ נַעְ תֹור‬follows the qal qaṭal ‫ ָזעֲקו‬, and in the absence of relevant medieval comments we can apply the linguistic insight of continuity and repetition and suggest that after repeated calls for help by the East Jordanian tribes and their prolonged trust in Him, G-d allowed Himself to be entreated; this was a continuous and prolonged process rather than a quick one-time event. However in II Chronicles 33:13 we find similar phraseology in the account of the repentance of the wicked king Manasseh, and here we find wayyiqṭol rather than the VIA construction40 ‫וַיִ ְת ַפלֵל אֵ לָיו ַו ֵיעָתֶ ר‬ ‫(‘ לֹו‬Manasseh) prayed to Him and He allowed Himself to be entreated’. Here it is perhaps difficult to suggest that the linguistic insight of continuity and repetition could be applied to the former example of the prayer of the people but not to the latter example of the prayer of the wicked Manasseh, although just maybe there was a reluctance to credit Manasseh with as much sincerity as the East Jordanian tribes; in the immediately preceding verses (11–13) we learnt that Manasseh entreated the Lord only when he was in great distress, suffering imprisonment and great humiliation. Commentators have noted that this passage about Manasseh’s captivity and repentance in II Chronicles has no parallel in the book of Kings and may even be “pure invention”.41 Also, Peake comments that “the unqualified condemnation of the apostate Manasseh in 2Kg 21:10ff and by his contemporary Jeremiah (15:4) raises doubts about the historical probability of his repentance and restoration to divine favour”.42 Rubinstein’s sixth example (the final of the six he chooses whereby “this redundancy of function can best be illustrated”43) is Ezekiel 23:47, where the VIA ‫ ובָ ֵרא‬follows the qal weqaṭal ‫וְ ָרגְ מו‬. Rubinstein compares this to Joshua 17:15, where the same root letters occur in piel weqaṭal ָ‫ ובֵ ֵראת‬following the qal imperative ‫ ֲעלֵה‬. There is no relevant medieval exegesis here, but the very different contexts in Ezekiel and Joshua are relevant. Joshua chapter 17 relates the complaint of the Josephites to Joshua that their allotment was not sufficient for their numbers, and in Joshua 17:15 comes Joshua’s suggestion to the Josephites, ‘if you are a numerous people, go up to the forest country and clear an area for yourselves there’, where the Hebrew has qal imperative followed by weqaṭal ָ‫ ֲעלֵה ובֵ ֵראת‬to express a specific onetime immediate command. Ezekiel chapter 23 presents the harlotrous sisters Oholah and Oholibah as a metaphor for the idolatrous population of Samaria and Jerusalem who merit destruction. In verse 46, the VIA ‫ וְ נָתן‬follows the hifil imperative ‫ ;הַ ֲעלֵה‬then in verse 47, which is Rubinstein’s sixth example, the VIA ‫ ובָ ֵרא‬follows the weqaṭal ‫וְ ָרגְ מו‬: 40 Rubinstein, 1952, p. 364. Curtis and Madsen, 1910, pp. 497–498. 42 Peake, 1962, p. 368 n. 43 Rubinstein, 1952, p. 363. 41 195 ‫אֹותהֶ ן בְ חַ ְרבֹותָ ם‬ ְ ‫וְ ָרגְ מּו ֲעלֵיהֶ ן אֶ בֶ ן קָ הָ ל ּובָ רא‬ ‘Let the assembly pelt them with stones and cut them down with their swords.’ Here, in contrast to the passage in Joshua (where the weqaṭal ָ‫ ובֵ ֵראת‬expressed the one-time command to clear a forest), the VIA construction ‫ ובָ ֵרא‬suggests the permanent nature of the annihilation of the adulterous/idolatrous sisters. Of course, it is true that the forest would also be permanently cleared, but one might suggest that the medieval exegetes were sensitive to the implied constant vigilance required to keep Israel far from idolatry. Again, Rubinstein’s discussion takes no account of the linguistic insight suggested by the medieval exegetes to account for the use of infinitive absolute in place of a finite verb. Another example of the infinitive absolute occurs in Jeremiah chapter 37, where we learn of the imprisonment of the prophet Jeremiah during the reign of King Zedekiah. In 37:21 we learn that Jeremiah is placed in the court of the guard and receives a daily allowance of bread; in this verse the waw plus infinitive absolute ‫ וְ נָתן‬is sequential to the wayyiqṭol ‫ ַויַפְ קִ דו‬as follows: ‫לֶחֶ ם לַיֹום‬-‫יִ ְר ִמיָהו בַ חֲצַ ר הַ מַ טָ ָרה וְ נָתֹׁ ן לֹו כִ כַר‬-‫ַויַפְ קִּ דּו אֶ ת‬ ‘They placed Jeremiah in the court of the guard and gave him a daily loaf of bread.’ There is no relevant medieval exegesis on this verse, but the action of giving a daily allowance of bread was clearly a repeated and continuous action, although we can also observe that the use of infinitive absolute here does leave open the question of who exactly delivered the bread. The prophet Ḥaggai (chapter 1) describes desolate conditions after the Return from Exile, which prevented the people from embarking on the rebuilding of the Temple. Ḥaggai clearly describes a prolonged period of desolation, urging the people to embark on the rebuilding because then their condition would improve. In Ḥaggai 1:6 we find the hifil infinitive absolute and waw ‫ וְ הָ בֵ א‬sequential to the qaṭal ‫ז ְַרעְ תֶ ם‬, and then we find three more infinitive absolute forms in succession in place of the qaṭals ‫אָ כֹול‬, ‫ שָ תֹו‬and ‫לָבֹוש‬: :‫לְ חם לֹו‬-‫לְ שָ כְ ָרה לָבֹוׁש וְ אֵ ין‬-‫לְ ָשבְ עָה ׁשָ תֹו וְ אֵ ין‬-‫ז ְַרעְ תם הַ ְרבֵ ה וְ הָ בא ְמעָט אָ כֹול וְ אֵ ין‬ ‘You have sown much and bring in little. You eat without being satiated. You drink without getting your fill. You dress, and it has no warmth.’ There is no relevant medieval exegesis on the use of these infinitive absolute forms in Ḥaggai 1:6, but they can be understood as expressing continuity by analogy with the above-quoted medieval interpretations and given that the context was clearly one of a prolonged period of desolation. 196 So far this article has presented the insights of medieval Jewish exegetes on the use of infinitive absolute as the equivalent of a preceding finite form. We also observed that even in the absence of medieval comments, this insight is often relevant. However we cannot assert that this linguistic insight is in all cases applicable and we may illustrate this point with a couple of examples as follows. The cycle of Gideon stories (Judges chapters 6–8) includes Gideon’s battle against the Midianite camp when ‘they sounded the horns and smashed the jars’ (7:19) in an effort to frighten the enemy. In this verse, the waw plus infinitive absolute ‫‘ וְ נָפֹוץ‬and they smashed’ is sequential to the wayyiqṭol ‫וַיִ ְתקְ עו‬ ‘and they blew’. ‫וַיִּ ְתקְ עּו בַ שֹופָרֹות וְ נָפֹוץ הַ כַדִ ים‬ ‘They blew the trumpets and smashed the jugs.’ It is true that blowing the trumpets and smashing the jugs perhaps involved a series of repeated short actions, but overall this was clearly a one-time battle, so the reason for the use of the infinitive absolute here in the book of Judges is debatable. We may note here that Fassberg cites Judges 7:19 to illustrate Tur-Sinai’s theory that the original wording would have included two infinitive absolute forms functioning adverbially – ‫ַיִתקְ עו בַ שֹופָרֹות תָ קֹו ַע וְ נָפֹוץ‬ ְ ‫ – ו‬but the second infinitive absolute ‫ תָ קֹו ַע‬was subsequently deleted.44 Our second example is in the book of the prophet Zechariah. Zechariah 3:1–10 contains the fourth vision of Zechariah, which describes the purification of the high priest Joshua. In 3:4 we find the waw plus infinitive absolute ‫ וְ הַ לְ בֵ ש‬sequential to the qaṭal ‫הֶ עֱבַ ְר ִתי‬: ‫וַיאמֶ ר אֵ לָיו ְראֵ ה העֱבַ ְר ִּתי מֵ ָעלֶיָך עֲֹונֶָך וְ הַ לְ בׁש א ְתָך מַ ֲחלָצֹות‬ ‘[the angel] answered and said to those standing before him, “remove the filthy clothes from upon him” and he said to him: “see, I have removed your guilt from you, and I will clothe you with robes”.’ The use of the waw plus infinitive absolute ‫ וְ הַ לְ בֵ ש‬actually avoids the question of whether the angel or those standing before him clothe Joshua. “Indeed, so neutral a form is the infinitive absolute that in the case of Zechariah 3:4 it may even have the force of a passive, i.e. ‘and you shall be clothed’.”45 Here we could argue that the clothing of Joshua was a one-time action and that the nuance of continuity is absent. On the other hand, we could argue that Joshua would now be in a permanent state of purity. These latter examples (Judges 7:19 and Zechariah 3:4) illustrate that the medieval insights contribute greatly to our discussions and understanding of 44 45 Fassberg, 2008, p. 55 and n. 42 there; Tur-Sinai, 1954, p. 323. Rubinstein, 1952, p. 366. 197 the use of infinitive absolute in place of a finite verb, but their insights cannot be taken to answer all of our questions. We have seen that in many examples the nuance of continuity and/or repetition can certainly be discerned in a context where infinitive absolute is used in place of a finite verb and that it was the medieval Jewish exegetes who were sensitive to this feature. However, this insight cannot be definitively applied in every case. In conclusion it is our hope that the voices of the medieval exegetes of the biblical narrative will not be forgotten despite the fact that they lacked the critical methodologies of modern scholars. Modern scholars have indeed successfully traced the ancient origin of the use of infinitive absolute as the equivalent of a preceding form, but they have agreed neither on the reasons for its relatively frequent appearance in the biblical narratives of the Second Temple Period and for its less frequent appearance in the narratives of the First Temple Period, nor on the construction’s semantics. These modern scholars have, however, ignored the sensitivity and insights of the medieval exegetes who discerned the addition of a nuance of continuity and/or repetition in the context of the narratives concerned, if not in every case, nevertheless in a notable number of cases. References Benyowitz, A. R., 2006, Translation of Ibn Ezra’s Commentary on the Pentateuch, Jerusalem: n.p. Berlin, A. and Brettler, M. Z. (eds), 2004, The Jewish Study Bible, Oxford: OUP. Cohen, O., 2009, eTeacher Biblical Hebrew Notes for Teachers, Level C (unpubl. notes, Israel Institute for Biblical Studies). ———, 2010, eTeacher Biblical Hebrew Notes for Teachers, Level E (unpubl. notes, Israel Institute for Biblical Studies). Curtis, E. L. and Madsen A. A., 1910, A Critical and Exegetical Commentary on the Book of Chronicles, Edinburgh: T & T Clark. Englander, H., 1939, “Grammatical elements and terminology in Rashi”, HUCA 14, pp. 387–429. Fassberg, S. E., 2008, “The Infinitive Absolute as finite verb and Standard Literary Hebrew of the Second Temple Period”, in J. Joosten and J. S. Rey (eds), Conservatism and Innovation in the Hebrew Language of the Hellenistic Period. Proceedings of a Fourth International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, Leiden: Brill, pp. 47–60. Gesenius, W., 1910, Hebrew Grammar. As Edited and Enlarged by the Late E. Kautzsch, 2nd English edn, A. E. Cowley (rev.), Oxford: Clarendon Press. Hammershaimb, E., 1963, “On the so-called Infinitivus Absolutus in Hebrew”, in D. W. Thomas and W. D. McHardy (eds), Hebrew and Semitic Studies. Presented to Godfrey Rolles Driver in Celebration of his Seventieth Birthday 20 August 1962, Oxford: OUP, pp. 85–94. Joüon, P. S. and Muraoka, T., 1996, A Grammar of Biblical Hebrew (2 vols), Rome: Editrice Pontificio Istituto Biblico. McKane, W., 1986, A Critical and Exegetical Commentary on Jeremiah, vol. I, Edinburgh: T & T Clark. 198 Peake, A. S., 1962, Commentary on the Bible, London: Thomas Nelson & Sons. Rubinstein, A., 1952, “A finite verb continued by an infinitive absolute in biblical Hebrew”, Vetus Testamentum 2, pp. 362–367. Shachter, J. F. (trans.), 1986, The Commentary of Abraham ibn Ezra on the Pentateuch, vol 3: Leviticus. New Jersey: Ktav Publishing House. Silbermann, A. M., 1929, Pentateuch with Targum Onkelos, Haphtaroth and Rashi’s Commentary (5 vols), M. Rosenbaum (trans., ed.), Jerusalem: Routledge & Kegan Paul. Tur-Sinai, N. H., 1954, The Language and the Book (Hebrew), vol. I, Jerusalem: Bialik Institute. Waltke, B. K. and O’Connor, M., 1990, An Introduction to Biblical Hebrew Syntax, Winona Lake: Eisenbrauns. 199 Implementation as Innovation: The Arabic Terms Qiṣṣa and Ḵabar in Medieval Karaite Interpretation of Biblical Narrative and its Redaction History MEIRA POLLIACK Tel Aviv University 1 Preface1 The notion that the Hebrew Bible was formed, in its entirety, as the result of a lengthy and complex process of literary composition and redaction is typical of the era that ushered in modernity and its critical revision of religious dogmas, beginning around the middle of the 18th century. This era led the way to our current understanding that the biblical text embodies a collection of varied, contradictory and inter-polemical strands and sources, which reflect different currents of thought as well as circles and periods of creativity in the history of ancient Israel, put together by generations of redactors, until they reached their final canonised form. Modern criticism, especially from the late 19th century, introduced a new focus on biblical genres (and subgenres), and their connection to social functions (Sitz im Leben), such as the relationship between biblical law, the priesthood and Temple; biblical historiography, the royal court and its scribes; epic and popular story-tellers; prophetic circles and prophetic texts. Novel tools were developed for analysing these genres, especially for comparing them with those of other Semitic languages and Ancient Near Eastern texts which were discovered during the same period, and for tracing the influence of these on the biblical corpus. The history of modern biblical scholarship is generally well known and has been charted out in many surveys, guides and monographs.2 I mention it by 1 To Geoffrey, dear and admired scholar and teacher, may you prosper and double your wisdom ‫עד מאה ועשרים‬, leading us, your students, in its path. For you are to us, as always, the ideal embodiment of “the teacher who [if he] is indeed wise does not bid you to enter the house of his wisdom but rather leads you to the threshold of your mind” (Gibran Khalil Gibran, The Prophet). 2 For recent works relating to biblical genre and social functions, and a re-evaluation of Herman Gunkel’s pioneering notions see Carr, 2005; Kawashima, 2004; Polak, 1999; Polak, 2003; Polak, 2011; Schniedewind, 2004. 200 way of introduction, in order to stress that medieval Jewish exegetes of the Islamic world provide an interesting precedent in paving a unique path of understanding the Hebrew Bible as a distinct literary product. They too seem to have realised that its text was the outcome of a long and complex literary process, even though their comparative tools were far more limited, and confined to the Arabic and Aramaic languages and literatures alone. The medieval pairing of Bible and literature occurred due to the rapid change experienced by these Jews, especially between the 9th and the 12th centuries, through which they experienced a transition from oral modes of Jewish learning and study to written modes. This resulted from their encounter with a highly literate medieval Arabic culture at large, and their growing use of Arabic and Judaeo-Arabic (Arabic written in Hebrew letters) compositions, in day-to-day discourse and intellectual exchange. In various discussions, I have tried to elucidate the wider theory through which the Judaeo-Arabic exegetes conceived of the various components incorporated into the composition of the biblical text and its editing process.3 This article focuses on the use of the Arabic term qiṣṣa as a technical term for biblical ‘story/narrative’ in Judaeo-Arabic exegesis, and so contributes to a more detailed understanding of the medieval exegetes’ conceptualisation of the biblical story as a structuralthematic unit that forms part of the final form of the biblical text. As is well known, terms such as ‘literature’, ‘narrative’, ‘prose’, ‘text’, ‘literary text’ and ‘discourse’ hold different meanings in different periods, cultures and languages, but they also share a common denominator. In classical Arabic literature a ‘narrative’ is usually named by terms such as ḵabar, ḥikāya and qiṣṣa (pl. qiṣaṣ) and some biblical characters mentioned in the Qurʾān are already associated, generically, with particular ‘stories’. The plural qiṣaṣ designates narratives in the Qurʾān. It appears in Sura 3 verse 62, in reference to the stories of Jesus; in Sura 7 verse 176, in reference to the stories of the Prophet Muhammad, and most notably in Sura 12, the ‘Joseph Sura’, which actually begins with the self-referential observation “naquṣṣu ʿalayka aḥsan al-qiṣaṣ” (‘let us tell you the most beautiful of stories’). The narrative on the birth of Moses, reflected in Sura 28, gives that sura its name, Surat al-Qiṣaṣ. Verbal forms of the same root appear in other places. The term ḵabar, on the other hand, appears five times in the Qurʾān, in the broader sense of ‘news’, ‘information’, ‘rumours’ or ‘traditions’. In traditional Islamic Hadith literature, the term qiṣaṣ al-anbiyāʾ designates a subgenre of stories connected with Jewish characters, also known from the Bible or the Midrash.4 In the medieval period, the genre of story or novelette (short novel) gained a place of honour in Arabic and Judaeo-Arabic culture (as can be seen, for example, in texts such as One Thousand and One Nights). Jews for whom 3 See Polliack 2005; Polliack, 2008; Polliack, 2012. Examples of discussions of these common stories and elements are, for instance, Firestone, 1989; Firestone, 1990; Neuwirth, 2006. 4 201 Arabic gradually became the main language of speech and writing are known to have read genuine Arabic prose from an early period, as reflected by Judaeo-Arabic and Arabic script copies of parts of One Thousand and One Nights, Sirat Antar and Kalila wa-Dimna. It appears that Jews began to compose stories and novellas in Hebrew from this epoch due to the influence of Arabic and Judaeo-Arabic forms, and so innovated a genre that had hardly any precedent in ancient rabbinic literature.5 In their Judaeo-Arabic compositions, Jews used the term qiṣṣa as a generic title for literary works relating to biblical characters, as in ‫( קצת אברהם‬the story of Abraham) and ‫( קצת יוסף‬the story of Joseph). They also applied it to other figures from Jewish history, for example ‫( קצת חנה‬the story of Hanna and her seven sons). Parts of these stories have been preserved in fragments from the Cairo Genizah, including the opening term ‘… ‫’קצת‬. Such stories were also incorporated in various medieval anthologies, re-workings or adaptations, wider collections of stories, the most famous of which is Kitāb al-faraj baʿd al-shidda known in its medieval Hebrew translation as ‫( חיבור יפה מן הישועה‬The Book of Comfort) by Rabbi Nissim Gaon, or to give him his Arabic name, Ibn Shahīn. This composition, widely attested in the Genizah, incorporates narrative re-workings of Talmudic and Muslim sources, and its structure is reliant upon the established Arabic genre faraj baʿd al-shidda (‘success after hardship’).6 Such works show a fine level of combination between elements taken from aggadic literature, especially of the late midrash, often widened into full narrative plot structures, and elements from the Arabic story or novelette. They also demonstrate a new openness and receptiveness on the part of Jewish audiences to the independent story genre. This change occurred following the Jews’ initial adoption of Arabic literary models in Judaeo-Arabic works, and hence it permeated into their Hebrew compositions. 5 With regard to the way in which medieval stories, disconnected from Bible exegesis or halachic discussion, are differentiated as an independent literary genre from the genre of rabbinic Aggadah, which served as an illustration to exegesis or halachic discussion, consider the observations of Dan, 1974, pp. 15–16 (my emphasis): ‫דפי התלמוד והמדרשים מלאים וגדושים סיפורים … ואולם כל השפע הזה אינו אלא בן־לווייה‬ ‫ הסיפור הוא רק‬.‫לתכלית העיקרית של התלמוד והמדרש – פסיקת הלכה ומדרש הכתובים‬ ‫ עניינם של החכמים בסיפור לא היה‬,‫ ואינו תכלית לעצמה‬,‫ עיטור וסטייה מן העניין‬,‫ דוגמה‬,‫אילוסטראציה‬ ‫ מבחינה ספרותית התהליך העיקרי שהתחולל‬.‫רב … המצב השתנה שינוי גמור בראשית ימי הביניים‬ ‫ הוא התגבשותן‬,‫ בתקופת הגאונים ובראשית התרבות היהודית באירופה‬,‫במאות הראשונות של ימי הביניים‬ ‫ מדעי‬,‫ שירה‬,‫ פרשנות‬,‫ לשון‬,‫ מוסר‬,‫ עיון‬,‫של מסגרות ספרותיות נפרדות לחיבורים הדנים בענייני הלכה‬ ‫ כך נשברה האחדות החיצונית של הספרות התלמודית־מדרשית והחלו להתפתח תחומי יצירה‬.‫הטבע וכדומה‬ ‫ התפתחותו של הסיפור וגיבושו כדי סוג ספרותי עצמאי הוא אפוא חלק מתהליך‬.‫שונים כגופים עצמאיים‬ .‫ ובוודאי היתה השפעה הדדית בין תחומי הספרות השונים בהתפתחות זו‬,‫כולל‬ See also further discussions on this issue by Yasif, 1994, pp. 271–310; Hasson Kenat, 2012; Lavee, 2010; Lebedev, 1993; Tobi, 2010. 6 See Nissim, 1970 and cf. Rotman, 2010. 202 This general background makes it clearer why in the systemic Bible exegesis, which flourished as a distinct field of writing and expertise in medieval Judaeo-Arabic literature, commentators frequently use the term qiṣṣa in order to define wide discourse units in the biblical text (which nowadays we would term sagas, cycles, narratives, stories or plots). Often, these units have a connection to the life and actions of a particular biblical character (as they would in qiṣaṣ al-anbiyāʾ). Sometimes these exegetes also employ the term qiṣṣa in relation to smaller thematic units or narrative excursuses, which we would identify with chapters and subchapters. In the 10th–11th century commentaries of Saadiah Gaon and Shemuel ben Hofni Gaon and later Rabbanite exegetes, as well as in those by Salmon ben Yerhuam, Yefet ben ʿEli and later Karaite exegetes, we find many occurrences of the term qiṣṣa. It often serves as a proper title or as a heading of narratives about central biblical heroes – such as ‫קצת משה‬, ‫קצת מרים‬, ‫קצת יוסף מע אכותה‬,7 ‫קצת אבשלום‬, ‫ קצת יואב צרויה‬and ‫ – קצת שאול‬or narratives on more secondary characters – such as ‫קצת בני שאול‬ and ‫קצת יואש‬. In general, it seems that the Jewish exegetes of the Islamic milieu applied to the Hebrew Bible a wide literary category with which they were familiar both from general Arabic literature as well as from the Qurʾān and its exegesis, and even from Judaeo-Arabic literary, non-biblical compositions. Yet, this is not only a question of terminology transference. These exegetes also appear to have cast the biblical sources into a narrative mould and perceived it in terms of narrative and storytelling, which were highly developed in their Arabic milieu. Thus, qiṣṣa came to designate, in their working definition, a genre relaying credible historical or historiographical information, on the one hand, while still being fashioned by artistic design and ornament, on the other hand, in the ways of fiction. In this context, we should also go back to discuss the additional Arabic term aḵbār (the plural of ḵabar), which often appears in conjunction with qiṣaṣ in Islamic sources, designating ‘narrative’ in a wide sense. However there are times when in Islamic sources the terms remain distinct: qiṣṣa more often designates historical narrative materials, while ḵabar denotes general information related in tradition or oral tales. The Judaeo-Arabic sources appear to reflect some awareness of this distinction. They employ the term qiṣṣa in describing a thematic unit within the story, or the entire narrative unit and its boundaries, in connection with a specific biblical character (as explained above). The term ḵabar, however, may indeed describe more ephemeral traditions, probably oral, which were received by the biblical authors and transcribed or transformed into written form. 7 Note that in the Qurʾān (Sura 12:7) the reference is to ‫( ڍوسف وٳخوته‬Joseph ‘and his brothers’ as opposed to ‘with his brothers’). For manuscript sources of these and further examples, see Polliack, 2014. 203 2 Discussion of selected examples In the following, I offer a short discussion of a few examples of the use of qiṣṣa culled from various biblical commentaries by the great Karaite commentator, Yefet ben ʿEli, who was active in Jerusalem in the second part of the 10th century. Since Yefet’s work tends to anthologise the opinions of earlier and contemporary Karaite exegetes, as well as deliver his own analysis, it serves as an indication of the wider usage of the qiṣṣa terminology in this period. Nevertheless, an additional study into the surviving works of his contemporaries, which lies outside the limits of this article, is necessary in order to establish the wider frequency of these concepts as elaborated in Yefet’s work. 2.1 Moses narrative8 In the introduction to his commentary on Exodus, Yefet explains that the whole of Exodus includes the continued presentation of the history of the people of Israel, which began in the days of Noah and the patriarchs. Exodus continues a historic chain of events that also explains the dependence between fulfilling the covenant and promise made to Abraham (Genesis 15) and the test encompassing four hundred years of exile and enslavement. At a later stage, Yefet elaborates five major themes of the book of Exodus, which represent major narrative focuses, in his view, including: a. b. c. d. e. the background of the enslavement and its reasons; the details of the journeys of Israelites upon leaving Egypt; the revelation at Sinai and the plans for the tabernacle; the golden calf incident and the covenant with Israel; the building of the tabernacle and the Shekinah entering it. Yefet goes on to identify small narrative units that feed the wider narrative focuses, like circles within circles. Hence, he explains that Moses’ naming of his son Gershom (Exodus 2:22), due to his grief over losing connection with his native country and people, represents a wider narrative feature or theme, as follows: ‫אשתק לה אסם מן גנס קצתה ווגמה עלי מפארקה מולדה ועשירה ליערף אנהו כאן‬ ‫מגמום עלי מפארקתה‬ ‘The name (Gershom) is derived from the nature/type of his narrative (qiṣṣa, i.e. Moses’ personal story) and his sorrow over being severed from his clan and family, in order to indicate that he (Moses) was woeful over this separation.’ 8 For the Arabic original of the examples in this section, see MS RNL Yevr.-Arab I 0054; further discussion is in Polliack, 2014. 204 Beyond this insight into the ‘typical’ storyline of the Moses narrative, namely, one involving communal as well as personal exile, this comment also shows Yefet’s understanding of the methods of character portrayal in biblical narrative, which include the naming of offspring as an indirect expression of (and way of moulding) the character’s psychological situation, inner life and aspirations. At a later point in the commentary, Yefet refers to the whole narrative of the life of Moses with the expression ‫ קצת משה‬qiṣṣat moše. The personal story of Moses, and particularly the story of his birth, he explains, is a foreshadowing of the collective story of the Israelites. The narrator/editor (termed ‫מדַ וִ ן‬ mudawwin here and in many other works by Yefet) created a linking between the narrative of the life of Moses and the narrative of the salvation of Israel from Egypt.9 According to Yefet, the intricate interlacing of the two qiṣaṣ was intended to show that the birth of a saviour (Moses) will make good the promise made to Abraham in Genesis 15 (wherein it is revealed to the Patriarch that Israel will be redeemed from Egypt after four hundred years of slavery), as is partly explained in the following: ‫כאן אלמדוון אדכל פי וסט קצה ישראל קצה משה ליורי אן גזרת כל הבן הילוד זאלת‬ ‫בולאדה משה תם תמם קצה משה מקדמה למעאני יחתאג אליהא פימא בעד תם רגע‬ .‫אלמדוון אלי קצה ישראל ליורי כיף אסתחקו אלגאלה מן מצרים‬ ‘The narrator/editor inserted the story (qiṣṣa) of Moses in the middle of the story (qiṣṣa) of Israel to show how the edict concerning the (killing of the) firstborn male came to an end with the birth of Moses. Then he completed the story (qiṣṣa) of Moses as an introduction to issues he needed to relate later on. Then the narrator/editor went back to the story (qiṣṣa) of Israel to show how they were worthy of being redeemed from Egypt.’ 2.2 Jethro narrative10 Yefet explains, in a similar way, the incorporation of the story of Jethro (‘ ‫קצת‬ ‫ )’יתרו‬within the Moses/Exodus sequence, in his comment on Exodus 18:1: .‫ואכתלף אלעלמא פי תדוין הדה אלקצה פי הדא אלמוצע‬ ‘(Know that this story was written/included (dawana) in the Torah in order to relate some important ideas …) there is a dispute between scholars about the inclusion/writing (tadwīn) of this story (haḏihi alqiṣṣa) in this place.’ 9 On the narrator/editor’s use of truncating devices, which enable detailed and deliberate forms of elision (iḵtiṣar), see Polliack, 2012. 10 For the Arabic original of the examples in this section, see MS RNL Yevr.-Arab I 0054; further discussion is in Polliack, 2014. 205 Yefet then continues, in his lengthy commentary on the story of Jethro, to explain it as a separate unit, which originally existed on its own or as part of a separate narrative cycle, but that the narrator/editor (designated here again by the term mudawwin) incorporated the Jethro narrative in its current place within the Moses/Exodus sequence “for a specific reason”. Yefet partly elaborates on this reason by describing the disagreement among exegetes about the considerations that led the mudawwin to edit the story of Jethro into the main narrative strand (Israel’s salvation from Egypt) at this particular point, thus truncating it and inserting an episode that juts out because of its separate theme and hero. Commenting on Exodus 18:27 (“Then Moses let his father-in-law depart, and he went off to his own country”), Yefet emphasises that the words “and he went off to his own country” make reference to a matter which “the Torah (text) ‘skipped’ recounting at this place, and which was written/redacted elsewhere” (‫)מעני אכתצרתה אלתוראה האהנא ודונה פי מוצע אכר‬. The elided matter Yefet wishes to refer to seems to be the information recounted in Numbers 10:29– 33, for Yefet goes on, “and that is what he [Moses] says at the time of their setting out [to the Promised Land]” (‫“ – )והו קולה פי וקת רחילהם‬Moses said to Hobab son of Reuel the Midianite, Moses’ father-in-law, ‘We are setting out for the place which God said I will give to you. Come with us … and he said, no, I will go back to my own country and clan.’ ” So, it seems that the material which the ancient narrator/editor(s) had about the character of Jethro challenged them as to how it would best be disseminated in the Pentateuch. Their solution was to incorporate it as a subtheme of the story of the salvation from Egypt, in a deliberately truncated manner, and refrain from fully developing it as a separate continuous strand, perhaps in an additional composition or book, as they might have done in other circumstances. The “specific reason” for this seems to have been the paramount importance of the salvation theme in their editorial considerations. Hence, Jethro’s interactions with Moses as part of this theme are centred upon when necessary, while other details and stories about Jethro are omitted. Who is the redactor responsible for the distribution of the narrative material about the character Jethro between the Books of Exodus and Numbers? From Yefet’s discussion, it seems unlikely he thought Moses was describing himself in the verse above in the third person, especially since he attributes the elision to “the Torah (text)”. The ambiguity about the identity of the Jethro narrative’s narrator/editor(s) becomes even more intriguing in Yefet’s continued pursuit of the editing process behind the Exodus narrative materials. In this wider context, he explains that the biblical histories or traditions – that is, aḵbār such as those attested in the Books of Judges, Chronicles, Jeremiah and Samuel – were also subject to elision. In Samuel, in particular, Yefet points to a phrase which he says suggests, according to some scholars (he does not specify whom), the existence of a detailed tradition about Jethro’s acts of kindness 206 towards the Israelites, which was not included in the final form of the biblical text:11 ‫ולם תדכר אלתורה רגועה בל וגדנא פי אלאכבאר יקול ובְ נֵי קֵ ינִ י חתֵ ן משֶ ה עָלו מֵ עִ יר‬ ‫הַ ְתמָ ִרים [אֶ ת־בְ נֵי יְ הודָ ה ִמ ְדבַ ר יְהודָ ה אֲשֶ ר בְ ֶנגֶב ע ֲָרד ַו ֵילְֶך ַויֵשֶ ב אֶ ת־הָ עָם] (שופטים‬ ‫טז) פערף אנה רגע בעיאלה וסכן פי יריחו פבעד אנצראף יתרו צעדו אלי מדבר יתרוה‬,‫א‬ ‫ומ ְשפְ חֹות ספְ ִרים ישבו (י ְשבֵ י) יַעְ בֵ ץ ִת ְרע ִָתים ִש ְמע ִָתים שוכ ִָתים‬ ִ ‫ואקאמו ת ם וענהם קאל‬ ‫ נה) פקאל אלעלמא אן‬,‫ ב‬,‫ית־רכָב (דברי הימים א‬ ֵ ֵ‫הֵ מָ ה הַ קִ ינִ ים הַ בָ ִאים מֵ חַ מַ ת אֲבִ י ב‬ ‫ן־רכָב עמֵ ד לְ ָפנַי כָל־‬ ֵ ֶ‫קולה הֵ מָ ה הַ קִ י ִנים ישיר בה אלי בית הרכבים אלדין הם בני יֹונָדָ ב ב‬ ‫יט) ופי גמלה הדה אלאכבאר טהרת פצאיל יתרו ואולאדה אעני חבר‬,‫הַ י ִָמים (ירמיהו לה‬ ‫הקיני יונדב בן רכב ואעלם אן בעץ אלעלמא קאל אן יתרו פעל מע ישראל חסד גדול‬ ‫לם תדפעה אלינא אלאכבאר בשרח ונעלם דלך מן קול שאול וְ אַ תָ ה ע ִָשיתָ ה חֶ סֶ ד עִ ם־כָל־‬ .)‫ו‬,‫בְ נֵי יִ ְש ָראֵ ל בַ עֲלֹותָ ם ִמ ִמצְ ָריִ ם (שמואל א טו‬ ‘The Torah/Pentateuch does not mention/describe where he (Jethro) returned to (after accompanying Moses), but we find in the traditions/histories (al-aḵbār) that it is said, “the descendants of the Kenite (i.e. Jethro), Moses’ father-in-law, went up with the people of Judah from the city of Palms” (Judges 1:16). It thus lets it be known that he returned with his descendants and lived in Jericho. After Jethro passed away, they came to the desert, where they wandered and lived there, and about them it is said, “The families of the scribes that lived at Jabez, these are the Kenites.” (1 Chronicles 2:55). The scholars say that his saying “these are the Kenites” points to the house of the Rechabites, who are the sons of Jonadab son of Rechab (of whom it is said), “Jonadab son of Rechab shall not lack a descendant to stand before me for all time” (Jeremiah 35:19). All of these traditions/histories (al-aḵbār) reveal the noble qualities of Jethro and his children, that is, Heber the Kenite and Jonadab son of Rechab. You (the reader) should know that one of the scholars says that Jethro performed a great act of kindness, which the traditions/histories (alaḵbār) do not report to us in detail/explicitly, but of which we know from Saul saying, “For you showed kindness to all the people of Israel when they came up out of Egypt” (1 Samuel 15:6).’ Finally, in this context, we find Yefet’s insightful remark: ‫וקצה עמלק ועמידת הר סיני וסאיר כלמא תוסט דלך ללהדה אכבאר אתצלת בה‬ ‘The story (qiṣṣa) of Amalek and the standing at (Mt) Sinai (stories that come after the Jethro narrative in Exodus) and all that intercedes/comes between, are all related to these traditions/histories (aḵbār).’ In these examples concerning the Jethro narrative, when Yefet mentions aḵbār, he refers to traditions/histories preserved in the biblical historiographic 11 See MS RNL Yevr.-Arab I 0054. 207 literature, as distinct from stories/narrative units, for which he prefers the term qiṣṣa. These aḵbār, which were probably preserved and transmitted orally, were used by the mudawwin in the Pentateuch books of Exodus and Numbers as well as in other biblical books (Judges, Samuel, Jeremiah and Chronicles), which preserve data about the descendants of Jethro the Kenite, Moses’ fatherin-law. One source (Samuel) even suggests that the mudawwin excluded existing information related to a specific act of kindness that the Kenites showed the people of Israel when they came up out of Egypt, a matter which Yefet learns from Saul’s reported speech. These examples show that the differentiation between these terms, qiṣaṣ and aḵbār, as reflected in Yefet’s work, is close to the distinction in Arabic Muslim sources (see section 1). They also show the flexibility of the concept of the mudawwin that Yefet and his peers implement in their discussions about the crystallisation and redaction of the Bible. As far as they are concerned, these narrators/editors had diverse historical source materials at their disposal, which they recorded, copied or redacted in different contexts, and they did not necessarily concentrate them in one narrative sequence. The Arabic term aḵbār was suited, therefore, to their conceptualisation of the mainly oral ‘tales/traditions’, that is, the unhindered and yet unshaped narrative materials of ancient Israel, which were transmitted by generations of informants and storytellers, and which the biblical mudawwins committed to writing, fashioned or edited in different ways. These materials took on the shape of distinct stories (qiṣaṣ), as we know them from the biblical text, through the work of these narrators/editors. This may be the reason that Yefet sees no point in identifying the historical figure behind the mudawwin. If the raw materials, the aḵbār, about Jethro (and other biblical figures), were adapted into stories and segments of information, qiṣaṣ, truncated and spread out from the Pentateuch to Chronicles, it is clear why the mudawwin of the Jethro episodes cannot be a known biblical figure, such as Moses or Samuel. If Yefet were to pinpoint such a mudawwin at any given point he would undoubtedly be asked: How did he live for so long? So he preferred to leave his/their historical-biographical identity unknown and collective. These were the writers and redactors of the Bible text who worked behind its scenes, through history, fashioning traditions into stories and so creating the intricate (and sometimes strained at the seams) tapestry, which eventually became the canonised twenty-four books. The examples above and similar ones, relating to other biblical compositions, supply sufficient evidence for the way in which Yefet understood the aḵbār as a wide category, perhaps the widest, of raw ‘sources’ and not necessarily written ones, such as were common at the time of the biblical narrators/editors, who were active throughout biblical times. The preservation, transcription, documentation, filtering, organisation and shaping of the aḵbār as written stories – qiṣaṣ – were, in fact, the major tasks of the biblical mudawwins. 208 2.3 Abraham and Joseph narratives12 In the final three examples which conclude this article, this time from Yefet’s commentary on Genesis, we find the same distinction between the terms qiṣaṣ and aḵbār applied to the figures of Abraham and Joseph. Just as in the case of Jethro, where Yefet conceived of fluid traditions which became fashioned and fixed into story units and saw evidence for this in the wider biblical array regarding this character, so too in the case of these patriarchal characters. There are contexts in which qiṣṣa and ḵabar (the singular forms) are used as synonyms for one narrative unit. However, a close examination shows a different semantic hue. Thus, while explaining Genesis 20:1, Yefet ben ʿEli uses ḵabar alongside qiṣṣa when referencing the status conferred by the mudawwin to different narrative materials. .‫בעד פרג מן כבר סדום וכבר לוט רגע אלי קצה אברהם אללדי הו אלגרץ‬ ‘After completing the Sodom tradition/account (ḵabar) and the Lot tradition/account (ḵabar), he returns to the story (qiṣṣa) of Abraham, for he is the goal (of the narration).’ According to Yefet, this introductory verse shows that the story of Abraham is actually the guiding principle story, while the record of the events that happened to Lot in Sodom is an account/tradition rather than a full-fledged narrative. Thematically and narratively, its status is secondary to that of the guiding narrative, and it is shaped differently, and so remains less smooth and stylised in the biblical text. The same distinction is found in Yefet ben ʿEli comment on Genesis 25:19 (“These are the descendants of Isaac, Abraham’s son”). Yefet describes different stages in the work of the mudawwin, whose main purpose is to sift, organise and connect (yansuqu) the general traditions (aḵbār) into full-fledged stories (qiṣaṣ) about the patriarchs. In doing so, the mudawwin uses his anthological judgment to determine which stories to include and which to omit (iḵtaṣara) from the writing process of the living traditions (tadwīn al-aḵbār): ‘After he (the mudawwin) completed the stories of Abraham, he connected to them the stories of Isaac, and for this reason he narrated/edited them (dawwanahum) by (use of) the connective waw (“and these are the stories of Isaac son of Abraham”). And he omitted (iḵtaṣara) recording the (full) traditions/accounts/histories (aḵbār) of the sons of Qeturah and Ishmael since the purpose of the mudawwin was to connect the traditions of our (i.e., the Jews’) forefathers (wa-kāna ḡard almudawwin yansuqu aḵbār abāhatina). Furthermore, he did not wish to preoccupy us with the (fully-fashioned) stories (qiṣaṣ) of those (other 12 For the Arabic source of these examples, see MS RNL Yevr.-Arab B221 and B217. For further discussion of these examples, see Polliack, 2012 and Polliack, 2014. 209 descendants), who are like the stories of the rest of the world. Rather, he mentioned for us the stories of the forefathers which are of benefit to us (as Jews), and for this reason he elided (iḵtaṣara) mentioning (the stories of) those (others) and he mentioned (only) the stories of Isaac.’ In his comment on Genesis 47:13, Yefet analyses the structure of the wider Joseph narrative. According to his analysis, this narrative contains several stories narrated/edited by the Genesis mudawwin. One such story (qiṣṣa) in the wider narrative is that of the seven year famine that afflicted the Egyptians, and the long-term preparations for it (roughly, Gen. 41 and 47:14–26). Another story, which is inserted into this famine-story background, is that of the famine’s effect on Jacob’s sons in Canaan and their journey down to Egypt in search of food, consequently meeting their brother Joseph there (Gen. 42:1– 47:13 and 47:27–50:26). This is the foreground story. Yefet draws attention to a continuation technique (modern biblical study would define it as ‘resumption repetition’) used by the mudawwin in controlling the flow of information and connecting the two narratives. After the mudawwin finishes the account of Joseph’s brothers going down to Egypt, he ‘returns to conclude’ (‫)רגע יתמם‬ the background narrative on the happenings in Egypt itself during the seven years of famine (Gen. 47:14–26). Yefet considers this wider famine narrative as a background story that the mudawwin used and truncated in order to highlight the more important story materials about Joseph and his brothers – only after these materials are given their due place, is the wider background narrative duly resumed and briefly related. ‘He (the mudawwin) was preoccupied with the story of Joseph and his Brothers )‫ (כאן אשתגל פי קצّה יוסף מע אכותה‬from the words ‫וירא יעקב כי יש‬ ‫( שבר במצרים‬Gen. 42:1) up to here (Gen. 47:13, “‫)”ולחם אין בכל הארץ‬. Only after he (the mudawwin) ends their story (qiṣṣa), does he return to conclude the story (qiṣṣa) of Egypt during the years of famine (47:14–26). This is in order to demonstrate the difference between Jacob and his sons and Egypt and the rest, in that Jacob’s relative prosperity was retained (by God, despite the worldwide famine).’ Later on, in his comment on Genesis 47:27 (“‫)”וישב ישראל בארץ מצרים‬, Yefet continues to explain his line of thought: ‫בעד אן תמם קצה אלסבע סנין אללדי אערף כיף גרי מצרים פיהא רגע אלי גרצה‬ … ‫יעריפנא אכבאר אלאבא לינסקהא שיא בעד שיא‬ ‘After he (the mudawwin) ended the story (qiṣṣa) of the seven years of famine, in which he informed us of how the Egyptians managed, he returned to his goal, that is to let us know about the traditions/accounts (aḵbār) of the Patriarchs, so that they would be connected to each other (in sequence), one after the other.’ 210 Here too the emphasis is, as in Genesis 25:19, on the main goal guiding the narrator/editor in constructing narrative materials and sifting them out from more general traditions/accounts in order to inform his audience and properly arrange the tales/histories (aḵbār) of the patriarchs into individual stories (qiṣaṣ). 3 Conclusion As pointed out in section 1, it may well be that the Karaite distinction between qiṣṣa/qiṣaṣ and ḵabar/aḵbār was appropriated from Islamic Hadith literature or the genre of historiographic writing (taʾrīḵ), wherein aḵbār is used to describe traditions in the sense of ‘oral tales’. It is also possible that the notion of narrative materials going through a process of transformation into a written medium, thereby becoming structured stories (qiṣaṣ), was in vogue during the 9th–11th centuries. After all, this was the time of the editing and consolidation of the major Hadith collections.13 Nevertheless, there is no evidence in Islamic sources of the application of these terms to the question of the formation of the Qurʾān, not to its general study or exegesis. Were the Karaites aware of these trends in Arabic literary productivity, and did they try to apply them in their independent understanding to the formation of the Hebrew Bible? More research has to take place before we can give a definitive answer. It is certainly possible at this stage to suggest that this was not an accidental occurrence in the multicultural intellectual history of the medieval Islamic period. Jewish, Christian and Muslim notions of scripture and its formation interacted in various conscious and unconscious ways. Since we have no knowledge of the use of the terms aḵbār and qiṣaṣ in Islamic sources in reference to the Qurʾān’s formation as a text, this is yet another manifestation, in my view, of the Karaite movement’s immense innovativeness and radical thinking on the nature of scripture. I consider this aspect an independent development, therefore, in Karaite thought. Islamic terms in general, and qiṣaṣ and aḵbār in particular, are appropriated and remodelled at the same time, by the Karaites, and are used in an inventive way, as demonstrated above, in order to illuminate what for the Karaites became the only revealed source of Jewish tradition: the Hebrew Bible. Breaking away from rabbinic tradition enabled them to see its formation in historical-literary terms which were unwitnessed in Jewish tradition. Was the Karaites’ scientific and critical distancing from the biblical text the result of their philological training and tendencies? This is what Geoffrey Khan seemed to suggest to me on a memorable walk from his office at Sidg- 13 See further discussion of these aspects in Polliack, 2015 and Polliack, 2016. 211 wick Site to St John’s College in August 2013: “At heart they were true philologists; they were driven by their interest in how language works”, I recall as the general gist of his words.14 In that case, it was their linguistic training and commitment which channelled their vision into the fabric, the actual linguistic make-up, of the biblical text. Or was it their relative immersion in the Arabic intellectual milieu of their time, and their confidence within it, that freed them, as break-away Jews, to wander in new directions, a matter which I tend to see as no less instrumental in this development?15 However complex the answer, the works the Karaites left us are a clear testimony of the high level of interaction and interconnectedness between Arabic/Islamic and Hebrew-Aramaic/Jewish cultures, and the fertile notions and intellectual breakthroughs achieved due to this interaction and despite the tensions it undoubtedly aroused. In this article, I examined a mere segment of this complex interaction, which receives expression in the innovative Karaite application – or, as I would prefer to say, implementation – of two important Arabic terms, qiṣṣa/qiṣaṣ and ḵabar/aḵbār, in biblical exegesis. By ‘implementation’, I mean to emphasise an aspect of the transition, namely, that these terms are not just borrowed and applied simplistically (as calques), but are refined and adapted to the specific world-view and literary and intellectual needs of the Karaite Jews. This, in my view, is generally typical of the transference process of terminology and other lexical aspects from Arabic to Judaeo-Arabic, wherein the intercultural adaptation becomes effectively a tool for innovation. In this case, the novelty lies in the ability to recognise layers and strands of redaction in the biblical corpus, due to their perception through the lenses, as it were, of Arabic literature, an aspect that is not bound to these concepts in their immediate and original Arabic setting.16 It seems natural for us, in post-modern times, to use the term ‘story’ to describe biblical narratives connected with the characters of Jacob, Moses, Balaam, David or Ruth. The strong emphasis of biblical scholarship, especially since the second half of the 20th century, on the ‘art of biblical narrative’ and character portrayal has certainly contributed to this tendency. Nevertheless, the medieval period witnessed a development that should not be underestimated in the evolution of the Bible’s reception history and reception exegesis: for it was then that biblical texts began to be widely perceived, in Jewish circles (both Karaite and Rabbanite), as intentionally ‘fashioned’ stories or narrative units. This insight did not undermine, at first, the sacred or religious 14 Karaite innovativeness in structural thinking on the Bible’s linguistic and literary forms, and their proximity in certain respects to modern formalist and structural methods, has been stressed by Khan, 2000, pp. 128–133 and by Goldstein, 2010, pp. 466–469. 15 See Polliack, 2015; Polliack, 2016. 16 This issue has been partly addressed in Drory’s seminal 1988 Hebrew work on the early contacts between Arabic and Jewish literature. Nevertheless, more work is needed in refining her model in respect of specific genres, and in addressing the detailed mechanisms of innovation that operate in the intercultural zone, as I have tried to indicate here. See also Polliack, 1998. 212 value of the biblical texts, but rather re-positioned them in the literary consciousness of the time, which was adept in the treasures of prose, in an acceptable rhetorical and aesthetic framework. It was the Karaites, in particular, who developed a detailed theory of the biblical stories as the outcome of a selection process undertaken by narrators/editors, who sifted through many oral traditions, tales, histories and written records in order to create the Bible, similar to a literary collection of an anthological nature (diwān(. In their frequent usage of the Arabic term qiṣṣa, the Karaite exegetes wished to focus on the distinct narrative element and on its inner structural and textual meaning within the biblical sequence. They therefore tended to interpret a story from beginning to end and especially in relation to its plot. Another aspect of our discussion on the conceptualisation of biblical narrative is that it relates not only to the medieval recognition of the genre of prose narrative, as part of the textual biblical array, but also to furthering its prestige. For the new qiṣṣa terminology flourished in the context of a literate culture that increasingly valued, read and wrote stories. As is well known, unlike the Qurʾān, the Bible is not self-referential in commenting on its genres and literary composition. The literary consciousness that obviously existed during biblical times, without which such complex compositions and literary techniques would not have been possible, is not apparent in biblical sayings through which the Bible’s authors and editors reflect openly upon such issues. Moreover, the Bible does not have a distinct Hebrew word for ‘story’ as opposed to ‘song’, ‘law’ or ‘prophecy’, all of which have distinct Hebrew terms, designating literary types in the Bible itself. The book of Genesis utilises the title ‫ אלה תולדות‬for narrative materials which include genealogical lists, yet in most cases narrative is not defined in any way but flows naturally, as a routine account, using narrative formulae such as ‫ ויהי בימי הדברים האלה‬or ‫ויהי אחר‬. It is likely the Karaites sought to strengthen the narrative element within the Bible, as original and authentic to it, by introducing terms and notions which bolstered the usage of narrative, its role and its purpose as an integral part of the Hebrew Bible, and strengthened the place of this element within Jewish thought on the Bible. In Rabbinic Hebrew, we find terms such as maʿaseh and aggadah/haggadah designating non-legal issues or tales (either as found in the Bible or as composed by the sages themselves). However the noun familiar in modern Hebrew for ‘story’, sippur (and similarly ‘composition’, ḥibbur), is mostly documented in medieval Hebrew sources, possibly in imitation of Arabic or Judaeo-Arabic terms, precisely against the background and needs typical of the medieval era, as charted above.17 The new medieval hermeneutic also explains the relative ease with which the Judaeo-Arabic exegetes, most notably among them the Karaites, adopted the Arabic terms for story (qiṣṣa) and account/tradition/history (ḵabar) and applied them with no hesitation to Jewish scripture, as neutral terms. It was 17 For further discussion and references see Polliack, 2014, pp. 117–118 n. 24–26. 213 clear to them that their readers would find no fault in the matter, not only because the Bible and ancient rabbinic exegesis are bereft of a distinct generic term for ‘biblical narrative’, but also because of the centrality of this concept in their host culture and Judaeo-Arabic and Arabic literature in general. The Arabic generic terms for ‘story/narrative’ are used in Judaeo-Arabic Bible exegesis of the age, most extensively by Karaite commentators such as Qirqisani, Salmon, Yefet ben ʿEli, Yeshuah ben Yehudah and ʿAli ben Sulaiman, from the end of the 9th century until the late 13th century. The are also, though less intensively, found in the works of Rabbanite commentators, such as Saadiah Gaon, Shemuel ben Hofni Gaon, Shemuel al-Kinzi and Tanhum ben Joseph Ha-Yerushlami. All of these writers attest to the growing interest among the medieval Jewish literati of the Islamic milieu in the biblical ‘story’ as a distinct narrative genre, likened to the short story, romance, novelette or maqāma that they recognised in Arabic literature, and distinct from ancient midrash aggada. In parallel to the development of narratives in Judaeo-Arabic and Hebrew medieval literature, the medieval Bible exegetes erected a bridge to the biblical model of the story, by treating biblical narrative, first and foremost, as a fashioned literary unit, with an integral plot and major and minor characters, imbued with inner logic, artistic devices and structure. These same medieval exegetes also succeeded in conceptualising and categorising a historical-literary biblical genre, which is not defined as a distinctive genre in the Bible itself or in ancient rabbinic sources (excluding Hellenistic sources with which the medieval authors were generally unfamiliar). This was an era in which new narrative developments in Jewish literature occurred in conjunction with or as a continuation of breakthroughs in the understanding of biblical narrative among biblical exegetes. This understanding may have germinated in Jewish literature before the rise of Islam. Nevertheless, it was the encounter with Arabic literature and its strong inclination to stories and tales (folk and religious tales, as in the Hadith, or story anthologies and early novels such as A Thousand and One Nights), that heightened the narrative acumen and prestige of the story in Jewish Bible exegesis, as well as in Judaeo-Arabic and Hebrew works. These two trends (namely, exegesis of biblical stories as stories and story writing) contributed conjointly to the recognition of the art of biblical narrative in medieval times. They became fused into an understanding unique to medieval Judaeo-Arabic culture, that one of the most important achievements of the biblical endeavour, if not the most important one, is actually narrative (no less than law).18 It is not surprising, therefore, that the possible oral roots of biblical narrative were also beginning to become clearer in the same Ju- 18 This phenomenon will later also become a recognised feature of Jewish Bible exegesis in Spain and Provence, in the work of commentators such as David Kimḥi and Nahmanides. 214 daeo-Arabic milieu. No medieval treatise has yet been found to define the theories through which the Karaite exegetes tackled the question of authorship of the traditional materials or the raw oral tales )ḵabar/aḵbār( or the way these were transcribed in various biblical stories (qiṣṣa/qiṣaṣ) and strung together into books. In this sense, the Karaites were primarily philologists, as Geoffrey Khan so astutely pointed out to me in the above-mentioned conversation, and as he has demonstrated in his ground-breaking studies of their grammatical works. By ‘primarily philologists’, I mean that their discoveries on the Hebrew Bible were communicated mainly through a straightforward and sincere pursuit of the study of its language and text, in works devoted to the Bible per se. The Karaite exegetes referred very frequently to the mudawwin, also discussed above, the anonymous biblical authors/editors operating behind various biblical books and different parts therein. It is evident that there exists a connection, still to be further explored, between the functions of the mudawwin and the different ways in which these exegetes understood the crystallisation process of biblical narrative and its composite elements, as part of the Hebrew Bible’s redaction history. References Carr, D. M., 2005, Writing on the Tablet of the Heart: Origins of Scripture and Literature, Oxford: Oxford University Press. Dan, J., 1974, The Hebrew Story in the Middle Ages (Hebrew), Jerusalem: Keter. Drory, R., 1988, The Emergence of Jewish-Arabic Literary Contacts at the Beginning of the Tenth Century (Hebrew), Tel Aviv: ha-Kibuts ha-Meʼuhad. Firestone, R., 1989, “Abraham’s son as the intended sacrifice (al-dhabīḥ, Qurʾān 37: 99–113): issues in Qurʾānic exegesis”, JSS 34, pp. 95–131. ———, 1990, Journeys in Holy Lands: The Abraham-Ishmael Stories in Islamic Exegesis, Albany: State University of New York Press. Hasson Kenat, R., 2012, “The story of Zayd and Kaḥlāʾ – a folk story in a JudaeoArabic manuscript”, in L. Zack and A. Schippers (eds), Middle Arabic and Mixed Arabic, Leiden – Boston: Brill, pp. 145–156. Kawashima, R. S., 2004, Biblical Narrative and the Death of the Rhapsode, Bloomington: Indiana University Press. Khan, G., 2000, The Early Karaite Tradition of Hebrew Grammatical Thought, Leiden: Brill. Goldstein, M., 2010, “ ‘Arabic composition 101’ and the early development of JudaeoArabic Bible exegesis”, JSS 55/2, 451–478. Lavee, M., 2010, “Midrash and Aggada – activity and reception in the Islamic era”, in N. A. Stillman (ed.), Encyclopedia of Jews in the Islamic World (EJIW), Leiden: Brill, vol. III, pp. 413–416. Lebedev, V., 1993, “New material on the history of Jewish-Arabic literary relations in the Middle Ages”, World Congress of Jewish Studies 11, division D, vol. I, pp. 33–39. Neuwirth, N. 2006, “ ‘Oral scriptures’ in contact: the Qurʾanic story of the golden calf and its Biblical subtext between narrative, cult, and inter-communal debate” in S. Wild (ed.), Self-Referentiality in the Qurʾān, Wiesbaden: Harrasowitz, pp. 71–91. 215 Nissim ben Jacob ben Nissim ibn Shahin, 1970, Ḥibur Yafeh Meha-Yeshuʿah, H. Z. Hirschberg (transl.), Jerusalem: ha-Rav Kook. Polak, F. H., 1999, “The oral and the written: syntax, stylistics and the development of Biblical prose narrative”, Journal of the Ancient Near Eastern Society 26, pp. 59–105. ———, 2003, “Style is more than the person: sociolinguistics, literary culture and the distinction between written and oral narrative”, in I. Young (ed.), Biblical Hebrew: Studies in Chronology and Typology, London: T&T Clark, pp. 38–103. ———, 2011, “Book, scribe, and bard: oral discourse and written text in recent Biblical scholarship”, Prooftexts 31/1–2, pp. 118–140. Polliack, M., 1998, Genres in Judaeo-Arabic Literature, Tel Aviv: Tel Aviv University Press. ———, 2005, “Karaite conception of the Biblical narrator”, in J. Neusner and A. J. Avery-Peck (eds), Encyclopedia of Midrash: Biblical Interpretation in Formative Judaism, Leiden: Brill, vol. I, pp. 350–373. ———, 2008, “The ‘voice’ of the narrator and the ‘voice’ of the characters in the Bible commentaries of Yefet ben ʿEli”, in C. Cohen, V. Hurowitz, A. Hurvitz, Y. Muffs, B. Schwartz and J. Tigay (eds), Birkat Shalom – Studies in the Bible, Ancient Near Eastern Literature and Post-Biblical Judaism, Presented to Shalom M. Paul on the Occasion of his Seventieth Birthday, Winona Lake: Eisenbrauns, vol. II, pp. 891–915. ———, 2012, “ ‘The unseen joints of the text’: on the medieval Judaeo-Arabic concept of elision (iḫtiṣār) and its gap-filling functions in Biblical interpretation”, in A. Brenner and F. H. Polak (eds), Words, Ideas, Worlds: Biblical Essays in Honour of Yairah Amit, Sheffield: Sheffield Phoenix Press, pp. 179–205. ———, 2014, “Biblical narrative and the textualization of oral tradition: innovations in medieval Judaeo-Arabic Biblical exegesis” (Hebrew), in A. A. Hussein and A. Oettinger (eds), Ben ʿEver la-ʿArav, Contacts between Arabic Literature and Jewish Literature in the Middle Ages and Modern Times, vol. VI: A Collection of Studies dedicated to Prof. Yosi Tobi on the Occasion of His Retirement, Haifa: University of Haifa, pp. 109–152. ———, 2015, “The Karaite inversion of ‘written’ and ‘oral’ Torah in relation to the Islamic arch-models of Qurʾan and Hadith”, JSQ 22/3, pp. 243–302. ———, 2016, “Deconstructing the Dual Torah: a Jewish response to the Muslim model of scripture”, in M. Z. Cohen and A. Berlin (eds), Interpreting Scriptures in Judaism, Christianity and Islam: Overlapping Inquiries, Cambridge: CUP, pp.113–129. Rotman, D., 2010, “Folktales/folk literature”, in N. A. Stillman (ed.), Encyclopedia of Jews in the Islamic World (EJIW), Leiden: Brill, vol. II, pp. 215–218. Schniedewind, W. M., 2004, How the Bible Became a Book: The Textualization of Ancient Israel, Cambridge: CUP. Tobi, Y., 2010, “Literature, Hebrew prose (medieval)”, in N. A. Stillman (ed.), Encyclopedia of Jews in the Islamic World, (EJIW), Leiden: Brill, vol. III, pp. 266– 278. Yasif, E., 1994, The Hebrew Folk-Tale (Hebrew), Jerusalem: Byalik. 216 Patterns of Diffusion of Phonological Change in the North-Eastern Neo-Aramaic Dialect of Azran LIDIA NAPIORKOWSKA University of Warsaw 1 Introduction Phonological changes often appear as far from regular on the synchronic level since we can only observe one of the stages through which they gradually progress. The lack of regularity is especially perceptible in the field of language documentation, where newly recorded varieties may display a higher or lesser degree of diversity. Dealing with internal variation and ironing out the inconsistencies in order to arrive at a systematic grammatical description is certainly a task faced while documenting the dialects of Neo-Aramaic. Geoffrey Khan has been active in the field of Neo-Aramaic dialectology, especially of the North-Eastern subgroup (NENA), for at least twenty years, with his first NENA grammar of the Jewish dialect of Arbel appearing in 1999.1 It can safely be said that thanks to his publication of a series of articles and full-blown grammars, all based on extensive and meticulous fieldwork, as well as his management of the NENA Database Project, Geoffrey Khan has saved many of the dialects from falling into eternal oblivion. Moreover, as a result of his contribution to the field of Neo-Aramaic, combining language documentation with theoretical linguistics, we now have a fine-grained map of the features of the NENA subgroup contextualised within the linguistic diversity of the area. Geoffrey Khan’s work is, and with all certainty will remain, an exemplar of academic performance; with this humble inquiry into NENA phonology I wish to express my gratitude for all that I have had the privilege to learn from him during my time in Cambridge and beyond. In this paper, I will address the patterns of diffusion of phonological changes in the NENA dialect of Azran. Needless to say, any shortcomings are entirely my own. 1 Khan, 1999. 217 2 Community, variation and change The speakers of the NENA dialects, being a minority in the Middle East, have experienced displacement and extensive migration, and they have also been exposed to heavy contact with other linguistic varieties. These include other NENA dialects, on the one hand, and languages from entirely different families, such as Iranian, European and Turkic, on the other. No wonder that within any particular NENA community one may find a greater or lesser amount of internal variation. Nevertheless, this does not necessarily mean that a given variety should be considered as a mixture rather than as a genuine dialect; more likely, such variation is indicative of changes in progress, changes that the dialect is undergoing as a living, spoken variety. These changes are induced by different, sometimes competing, factors, and they result from both internal and external conditioning. The overall dialect profile may appear as irregular and the changes as unpredictable; however, by identifying particular factors and specifying the role played by them we can bring clarity to this seemingly irregular profile. Here I will make such an attempt, depicting the pattern of diffusion of some phonological changes in the NENA dialect of Azran which have been brought about through different mechanisms. 3 Phonological challenges in NENA Within the domain of phonology, two phenomena have proved to be especially challenging in NENA, namely phonological emphasis and palatalisation. Different typological outlines of emphasis for particular NENA dialects have been proposed, such as a largely segmental analysis for Alqosh or Jewish Betanure, and a suprasegmental analysis for Jewish and Christian Urmi.2 These different types of emphasis can be considered in the light of their geographical distribution, whereupon the influence of neighbouring languages can be seen. Thus the segmental type is found among NENA dialects from Iraq, which have been impacted by Arabic, whereas the suprasegmental type is characteristic of varieties from Iran, where the influence of Iranian and Turkic language varieties can be postulated.3 Similarly, there are several different outcomes from the process of palatalisation as it has applied in NENA varieties to the historical velar stops *k and *g. These include: 2 For Alqosh, see Coghill, 2003; for Jewish Betanure, see Mutzafi, 2008; for Jewish and Christian Urmi, see Khan, 2008a and Khan, 2016 respectively. 3 See Napiorkowska, 2015b. 218 • • • lack of palatalisation of k and g (e.g. Alqosh);4 slight palatalisation of the stops, resulting in k [kʲ] and g [gʲ] (e.g. Maha Khtaya d-Baz);5 heavy palatalisation, resulting in affrication of the former stops k and g, followed by the advancement of the place of articulation of the former affricates č and j, i.e. this includes a push-chain shift: k > č [t͡ʃʰ], g > j [d͡ʒ]; and consequently original č > c [t͡sʰ], j > ȷ [d͡z] (e.g. Christian Urmi, Sardārid).6 In the same vein that we may observe geographical clustering of dialects with a similar emphasis profile, so too palatalisation processes can be correlated with a dialect’s location. Palatalisation clearly gains intensity while moving from west (Alqosh, Betanure) to east (Urmi). An interesting case in this context is the Christian NENA dialect of Azran, now situated in north-eastern Iraq. Azran displays palatalisation of the historical velar stops; however, the process is nowhere as regular as in Christian Urmi. In Azran, rather than inducing a chain shift as described above, palatalisation results in mergers or near-mergers of phonemes. Moreover, the dialect also has an innovative high front rounded vowel /ü/, which is largely a reflex of the former back vowel *u, but the correspondence is, once more, far from consistent. The two developments may be included under one common term of fronting; however, the two display different degrees of variation. In what follows I will discuss the factors which may have led to the irregular distribution of fronting. The role that language contact might have played in shaping these features will also be considered. 4 4.1 Consonant fronting in Azran Palatalisation and affrication As noted in section 3, it is possible to observe how the changes in the pronunciation of the former velar stops progress across the NENA area. First, the velar stops are slightly palatalised, that is, a secondary articulation approaching the vowel /i/ or the glide /y/ is imposed on the consonant, resulting in k [kʲ] and g [gʲ]. This phenomenon is naturally triggered in environments where the segment following the velar stop is a palatal glide or a high front vowel /i/ or /e/, as can be seen in the Christian Barwar word kepa ‘stone’ [ˈkʲe:pʰæ ~ 4 See Coghill, 2003. The sound /k/ in Alqosh is in fact more fronted than the IPA [k] (Coghill, 2003, p. 31), but it is nevertheless not palatalised. I am grateful to Eleanor Coghill for pointing this out, as well as for her help with an earlier draft of this paper. 5 See Mutzafi, 2000. 6 For Christian Urmi, see Khan, 2016; for Sardārid, see Younansardaroud, 2001. 219 ˈcʲe:pʰæ].7 This patterning of segments has been analysed within the framework of Articulatory Phonology as a source of the development of the palatalisation of velar stops in Azran.8 The retiming of the articulatory gesture for the vowel /i/ influences the setting of the articulators for /k/, resulting in postalveolar /č/ and then further in alveolar /c/, so kipa [ˈkʲi:pʰa] > čipa [ˈt͡ʃʰi:pʰa] > cipa [ˈt͡sʰi:pʰa] ‘stone’. The same holds for the voiced stop /g/, developing in Azran into /j/ and /ȷ/, as in giba [ˈgʲi:ba] > jiba [ˈd͡ʒi:ba] > ȷiba [ˈd͡zi:ba] ‘side, flank’. 4.2 Variation and creation of near-mergers The palatalisation in Azran is found beyond the most widespread contexts mentioned above; that is, it is not restricted to cases where k or g precedes a high front vowel or a palatal glide. As has been mentioned, it is also not a regular process. In certain words, only the post-alveolar variant has been attested, e.g. *gnāhā > jnaha ‘fault, mistake’, not *ȷnaha; in a yet smaller number of words, exclusively the velar stops occur, e.g. guma ‘barn’ (from Kurdish), not *juma or *ȷuma. Interestingly, since the development of the alveolar segments in Azran does not induce a chain shift similar to one in Christian Urmi described in section 3, native Aramaic words feature the same segments as borrowings from Kurdish and Arabic, and the difference between them is neutralised in Azran, as can be seen in table 1. Table 1. Neutralisation of velar and post-alveolar segments in Azran9 Original word Azran Christian Urmi10 Aram. *kēp̄ ā ‘stone’ Kurd. čeltik ‘stalk of a crop’ Aram. *garmā ‘bone’ Kurd. ciwan ‘beautiful’ cipa cəlduč ȷarma ȷwanqa cipa čaltuc ȷarma jvank̭a Although no systematic phonological or morphological pattern can be observed which conditions the fronting in Azran, I would like to argue that the phenomenon is by no means random. 7 See Khan, 2008b, p. 30. This palatalisation phenomenon is widespread across languages. Compare, for instance, the alternation of the realisation of the English letter c, where the consonant is normally a velar stop before back vowels and an alveolar fricative before front ones, e.g. car [ˈkʰɑ:] vs ceiling [ˈsi:lɪŋ]. A similar morphophonemic correlation obtains in Polish, where [k] alternates with [t͡ʃ] in the corresponding environments, e.g. mak [ˈmak] ‘poppy seed’ vs maczek [ˈma.t͡ʃɛk] ‘poppy seed (diminutive)’ vs makowy [ma.ˈko.vɯ] ‘of poppy seed’. 8 See Napiorkowska, forthcoming. 9 The following abbreviations are used for language names in examples: Ar(abic); Aram(aic); Azer(i Turkish); CBarwar, Christian Barwar; CUrmi, Christian Urmi; Engl(ish); Pers(ian); Rus(sian). 10 A similar distribution of /č/ and /c/ and /j/ and /ȷ/ is also found in the dialects of Tazakand (Coghill, 2009) and Sardārid (Younansardaroud, 2001). 220 4.3 Distribution and frequency 4.3.1 Gangs As suggested above, the fronting of the former velar stops may have occurred in Azran first in the environment of high front vowels. This explanation addresses the phonetic nature of how the palatalisation would develop; however, the frequency of occurrence of the sounds /c/ and /ȷ/ is currently not significantly higher before high front vowels than before other segments. We find these sounds in words such as caḷu ‘bride’ < *kālō, cu-jyaje ‘every time’ < *kud-gaye (< Kurd. gaw), ȷanta ‘bag’ < Kurd. čente and ȷumile ‘he gathered (it)’ < *jumile (j-m-Ø < Ar. j-m-ʿ), all of which have back vowels following the relevant segment. The distribution of the fronted consonants /c/ and /ȷ/, rather than being conditioned solely by the nature of the following vowel, appears to be correlated with frequency of use. Joan Bybee has argued that certain high-frequency words with a particular phonological makeup form ‘gangs’ with other words which may not be entirely similar in structure, but which share certain phonological or morphological properties. Thus, for instance, the dialectal past tense of the English word sneak is snuck, by analogy with other verbs with an initial /s/ and a nasal, such as swim.11 Here, swim is as a high-frequency verb with original ablaut in the past tense, and it attracts new lexical members and forms a gang of items of similar phonological and morphological behaviour. I would like to suggest that a similar phenomenon of high-frequency words forming gangs may be responsible for the diffusion of the fronted realisation of the alveopalatal series in Azran. A possible scenario for the distribution of the fronted segments would entail, as the first step, the cross-linguistically and cross-NENA attested shift of a velar stop or post-alveolar affricate into an alveolar affricate before a high front vowel, i.e. the retiming of gestures (see above). The gangs would thus include: cipa ‘stone’ < *kēˉpā, ciri ‘autumn month’ < *tšīrī < *tišrī, cista ‘pocket’ < *kisṯā, citwa ‘thorny plant’ < *kiṯwā, etc. ȷiba ‘side, flank’ < *gībā, ȷizdan ‘wallet’ < Pers. *juzdan, ȷičče ‘he walked’ from ȷ-y-k (cf. j-w-j in CBarwar), etc. These gangs could be defined as short, mostly disyllabic, words with a front vowel following a former velar stop or a post-alveolar affricate, which has shifted to an alveolar affricate. At the second stage of gang formation, the tendency to render short words containing a (former) velar stop with an alveolar affricate would spread to other high-frequency words. These, in turn, 11 Bybee, 2001, p. 127. 221 would form gangs of their own, such as those below, containing a back vowel and a sonorant: calba ‘dog’ < *kalbā, caḷu ‘bride’ < *kālō, canna ‘cheek’ < Azer. čənə, etc. ȷanta ‘bag’ < Kurd. čente, ȷanȷar ‘threshing sleigh’, etc. The whole process is illustrated through the steps in table 2. Table 2. Process of the spreading of alveolar affricates in Azran Primary form Process Outcome kipa čipa retiming of the high front vowel retiming of the high front vowel diffusion of /c/ in gangs diffusion of /c/ by analogy diffusion of /c/ in a new gang čipa cipa ciri, cista, citwa, etc. calba caḷu, canna, etc. According to Bybee, high-frequency words are more prone to undergo progressive changes thanks to their occurrence rate, which through repetition allows them to be deeply entrenched into the speakers’ lexicon and cognitive system.12 The high occurrence rate and the process of gang formation would be responsible for the advanced fronting in this group of Azran words. 4.3.2 High-frequency and grammaticalised words In another group of high-frequency words in Azran, palatalisation is rather irregular, and no morphophonological similarity can be postulated as a motivation for gang formation. In many cases, these are grammatical or grammaticalised words, such as prepositions or particles, which often escape the usual rules of a language.13 Others are among the frequent adjectives and nouns. Among these words, some undergo fronting, but only to the alveolar region, while some others retain a velar consonant. In general, no variation of articulation within these words has been documented. Consider, for example: čəs- ‘at, chez’ (inflected) < *kes but ks- ‘at, chez’ (invariable) never cəsnever čs-, cs- čud- ‘every’ < *kulnever cudbut cu- ‘every’ in compounds, cf. cu-jyaye ‘every time’ above14 12 Bybee illustrates this point with the deletion of the final stops in common English words, such as and or just. See Bybee, 2001, pp. 10–12. 13 Such as rules about the prosodic minimum, as discussed in Hoberman, 2007. 14 Another factor at play in this particular word could be dissimilation, although it is also not a regular process. 222 düčta ‘place’ < *dukṯā never dücta guṛa ‘big’ < *gōrā never juṛa, ȷuṛa jărəj ‘have, must, modal particle’ < Azer. gərək never ȷărəȷ ju- ‘in, at’ < *gaw- never ȷu- kuma ‘black’ < *kōmā never čuma, cuma These items could be regarded as a special class of words whose high rate of occurrence, and often their grammatical function, grants them a special status in the lexicon. They would, thus, constitute a separate class within the speakers’ cognitive system, rendering them opaque to the sweeping phonological change. A situation where high-frequency words are an exception to regular phonological change is not uncommon in other languages. As an example from Neo-Aramaic, the case of Jewish Urmi may be cited, where common words such as +hudaa ‘Jew’ < *hūḏāyā or ida ‘hand’ < *ʾīḏā resisted the otherwise regular shift of *ṯ and *ḏ to /l/.15 Equally, one can consider the modern English word one which has a unique phonological shape and underwent a different shift from other words with the same historical vowel, such as stone.16 4.3.3 Low-frequency words One may connect the frequency rate of some of the Azran lexical items with the sociolinguistic realities of life. Former generations of speakers were mainly villagers, whose occupation was farming and husbandry. Over time, they were forced to give up their land and move from villages to towns in search of a safer life. As a result, nowadays the knowledge of many agricultural activities and processes survives only with the older members of the community. These former activities, as well as artefacts associated with them, are not very frequently referred to in current times, as they are less relevant to the present lifestyle. This is not to say that the words referring to these activities have disappeared from community discourse altogether, but rather they are characterised by a low frequency of use. We could consider that such lexical items have not been reached by the phonological shift of fronting: the rarer the occurrence of a lexeme, the rarer the opportunity to implement the phonological shift. Consider, for example, the following words, which have not been attested with any type of fronted 15 16 Khan, 2008a. Hopper and Traugott, 2003, p. 119. 223 articulation, or have only been marginally observed as having undergone the initial step of palatalisation: *garōpā > garüpa ‘oven shovel’ (marginally jarüpa, never ȷarüpa) *gədyā > gidya ‘young male goat’ (marginally jidya, never ȷədya) There are also similar words that are borrowings, although the lack of expected shift in them could alternatively be explained by their loan status (see section 4.3.4 below): guma ‘barn’ < Kurd. gom (never juma, ȷuma) koplina ‘wooden collar that forms part of a yoke’ < ? Kurd. qapal ‘stick’ (never čoplina, coplina) These words have been gathered from speakers who otherwise consistently render the former velar stops with /c/ and /ȷ/. It seems hardly reasonable to consider the relationship between the low occurrence rate of these words, their low relevance to modern life and their velar articulation as accidental. An apparent internal paradox may seem to arise from the analysis of the two groups of words above. On the one hand, we have very high-frequency words which resist phonological change; on the other, it is low-frequency words that prove immune to the same process. This seeming contradiction is solved by the fact that the two groups of words have a special position in speakers’ cognitive storage systems. A very high rate of occurrence allows words to maintain their present shape, since the constant input and use of these forms solidifies their current phonological form. A very low rate, in turn, forces speakers to put effort into memorising such words separately, along with all their peculiarities such as a conservative pronunciation. It is the special position which these words have that places them on a separate tier where fronting fails to operate. 4.3.4 Loan-words The last group of words with irregular renditions of the palatalisation of velars are loan-words. These undergo adaptation to the native phonology and morphology to varying degrees, thus palatalisation also displays irregular patterns. Loans from Kurdish in most cases undergo palatalisation (e.g. Kurd. kade > čade ‘type of pastry’, Kurd. ˉreng > ranja ‘colour’). In loans from Arabic, on the other hand, there is a tendency to retain the voiceless velar stop, whereas the voiced post-alveolar affricate is frequently rendered as /ȷ/ – for example, ʾakkid ‘sure, certain’ vs ȷensiya ‘citizenship’. This tendency includes words which are so well integrated with the language that they form verbs according to the native morphological patterns. 224 ʾakkid ‘sure, certain’ < Ar. (also as a verb ʾakkude ‘to make sure’) blok ‘block; a pile of banknotes’ < Engl. glas ~ gəlas ‘glass, cup’ < Engl. kábina ‘car, wagon’ < Engl. šárika ‘company, firm’ < Ar. (also as a verb šaruke ‘to team up’) markaz ‘city centre’ < Ar. šapka ‘hat’ < Rus. One may suggest that such high-frequency loans achieved a similar status to some of the native high-frequency words, that is, they became lexicalised with their phonological form and are stored in the part of the lexicon where the palatalisation process is not active. 4.4 The role of language contact It has been noted that palatalisation is a natural phonetic propensity of velar stops before high front vowels. However, if this were the sole and sufficient motivation for palatalisation, the shift of /k/ > /č ~ c/ and /g/ > /j ~ ȷ/ should surely be more widespread among the NENA dialects. However palatalisation appears to be limited to the north-eastern pocket of Iraq and to the Urmi region. This suggests that another factor is responsible for the diffusion of an otherwise natural phonetic tendency. This factor is most likely the input from the language varieties spoken in the area, such as Kurdish and Azeri Turkish, where palatalisation of velar stops and the presence of alveolar fricatives is widely attested,17 with this spreading as far as the Caucasus.18 The palatalisation process appears to be radiating from this region towards the west, affecting different varieties of NENA and gradually fading out. Which factors should then be regarded as the main sources of the diffusion of fronting in Azran? Some contact linguists assume an extreme stance in evaluating the role that language contact may play in inducing changes by stating that language features can be regarded as contact-induced only when it can be demonstrated that they would not have occurred without external influence. This extreme stance is not adopted here in evaluating the role of language contact in NENA; however, it does signal some considerations to be borne in mind. In particular, 17 18 Khan, 2016, p. 38; Kapeliuk, 2011, p. 739. Stilo, 1994; Chirikba, 2006; Johanson, 2006. 225 it places great emphasis on how much inherent potential for a given change already exists in a language. In this way, internal motivation is recognised along with external. Geoffrey Khan has often stressed that in NENA dialects, many changes are an effect of converging factors.19 The situation with the (post-)alveolar consonants in Azran presents itself as another example of inherent potential and external input combining to produce a phonetic change. It should be stressed that palatalisation in Azran is a change in progress, affecting different groups of words to varying degrees. It impacts very highfrequency and grammatical words to a limited degree, and has no effect on very low-frequency words. It mostly targets words with high- and middlerange rates of occurrence, including also many adapted loans, especially from Kurdish. These loans were probably borrowed into Azran together with the pronunciation of the donor language, strengthening the fronting process. On the other hand, Arabic, which does not undergo palatalisation or fronting of velars, provides a constant input of velar stops and affricates, with words containing the phonemes /k/ and /j/. To conclude this discussion of the fronting of velars, then, through ascribing particular roles to phonetics, contact and also frequency of use we have obtained a much clearer account of the seemingly random distribution of the velar, post-alveolar and alveolar consonants in Azran. 5 Vowel fronting As indicated in section 3, the fronting in Azran is not limited to consonants. Vowels have also undergone a development towards an articulation in the front of the mouth, resulting in the emergence of the high front rounded vowel /ü/ [ʉ~ʏ], undocumented in previous stages of Aramaic. The phoneme /ü/ generally corresponds to the *u of earlier Aramaic, both short and long,20 which is normally reflected as the high back rounded vowel /u/ in other dialects; for example, *būmā ‘owl’ > Christian Barwar buma, Azran büma. Nonetheless, not all instances of /ü/ are traceable to *u, and neither is every *u rendered in Azran as /ü/. In other words, the phonological shift is once again irregular. In the following, a proposal is outlined for the way in which this shift unfolded. 5.1 The shift *u > /ü/ The shift *u> /ü/ is a part of the larger push-chain shift in the vowel system of Azran, where the former vowel *o changed to /u/. Thus: 19 See, for example, Khan, 2008b, p. 18. This is different from the diphthong /uy/ of Christian Urmi, which developed from long vowels only. The short vowel remained as /u/ (Khan, 2016, p. 190); compare *ʾurḥā > Christian Urmi ʾurxa vs Azran ʾürxa ‘way, road’. 20 226 *o > /u/, e.g. *mōḥā ‘brain’ > Azran muxa (cf. CBarwar moxa) *u> /ü/, e.g. *nūnā ‘fish’ > Azran nüna (cf. CBarwar nune (pl)) The raising and fronting of the back vowels in Azran is most likely connected to the phenomenon of phonological emphasis. In emphatic contexts, most vowels tend to be pronounced with a constriction of the pharynx21 and a retraction of the tongue root. As a result, the back vowels /o/, /u/ and /a/ in Azran developed ‘back’ allophones for emphatic contexts, next to the existing ‘plain’ qualities. The appearance of the new vowel qualities resulted in quite a concentration of allophones in the back region of the mouth; as a remedy for the overload of back vowels, the chain shift of raising and fronting the basic, that is, non-emphatic vowel qualities took place in Azran.22 5.2 Other sources of /ü/ The high front rounded vowel is, however, found in other contexts, of which most can be traced to the sequence *iw in a closed syllable, or to historical *o. In the former case, the shift was regular, for example: k-t-w ‘to write’: CBarwar kθiwle vs Azran čtüle ‘he wrote’ Azran qariwa ‘best man; godfather’: CBarwar qariwta vs Azran qarüta ‘maid of honour; godmother’ In the case of /ü/ from *o, the shift was restricted. It appears in a handful of words, all of which share a high frequency of occurrence. It thus seems that, similar to the situation with consonant fronting, the vowel shift also correlates with frequency of word usage. Examples where *o became /ü/ in Azran include: glüla ‘round’ (cf. CBarwar galola) küpa ‘low’ (cf. CBarwar kopa) šxünta ‘hot’ (cf. CBarwar šaxina) There is one final source of the vowel /ü/, which is a high front vowel immediately next to a bilabial. This development may be explained as another case of assimilation, with retiming of the lip rounding to overlap with the pronunciation of the neighbouring vowel. It took place in some basic adjectives, but is by no means a productive process. Examples include the Azran words: 21 22 Khan, 2016, p. 50. Cf. Napiorkowska, 2015a, p. 38. 227 müḷana ‘green’ (cf. CBarwar milana ‘blue’) xamüma ‘hot, warm’ (cf. CBarwar xamima) xü̆wa ‘snake’ (cf. CBarwar xuwwe) But note that there are also many examples such as: bibiya ‘pupil’ (not bübiya) The development of /ü/ occurred also in the word for ‘threshing floor’, which would suggest that the shift was operative some time ago, when words such as this still occurred relatively frequently: büdra ‘threshing floor’ (cf. CBarwar bədra) 5.3 Lack of expected /ü/ As mentioned at the beginning of section 5, not all instances of *u became /ü/ in Azran. The words in which the shift failed to occur can be subdivided into three types. In the first type, the factor which blocks the sound change seems to be historical emphasis, where the pronunciation would be retracted towards the back of the mouth. The back articulation of a consonant or an entire word prevented the fronting of the vowel /u/, which in such contexts would appear with the back allophone. Examples of this type include: čawətra ‘lunch, midday meal’ (cf. CUrmi +cavutra) duxrana ‘feast of the saint’ (cf. CUrmi +duxrana) parušta ‘small flat stone’ (cf. CUrmi +parušta) ṱuṛa ‘mountain’ (cf. CUrmi +ṱuyra) qunya ‘water well’ qurba ‘close, near’ qurdaya ‘Kurd’ quya ‘strong, powerful’ rumxa ‘spear’ (cf. CUrmi +rumxa) 228 In modern Azran, none of these words contains emphasis, but traces of the former emphasis can be seen in the unaspirated articulation of the stop /ṱ/ – for example in ṱuṛa ‘mountain’ – and in the back articulation of the vowel /u/, especially between /q/ and /r/ – for example in qurba [ˈqʊɾ.ba ~ qʊɾ.ba] ‘close, ̱ near’. Other words in this group are plain, that is, non-emphatic, in both Christian Barwar and in Azran, but many appear as emphatic in Christian Urmi. The next group of words in which the shift failed to happen includes loanwords. Among loans in Azran one finds invariable adjectives, but also fully integrated items which serve as a base for further derivation. Examples include: kubba (pl. kubbe) ‘meatball’ < Ar. mašhūr (invariable) ‘famous’ < Ar. mufrad (invariable) ‘singular’ < Ar. muxtar (pl muxtare) ‘village chief’ < Kurd. sixur (pl sixure) ‘porcupine’ < Kurd. The lack of /ü/ here could be accounted for by the fact that, as mentioned earlier, loans have a special status in the lexicon and often do not conform to the rules of native phonology and morphology. The last group of words in which the vowel /u/ is kept against the general tendency of shifting to /ü/ encompasses, once again, high-frequency words, such as the most common nouns, prepositions, adverbs and discourse particles. For example: ʾurza ‘man’ (also the CBarwar form) dus, marginally düs ‘right, correct’ < Kurd. (cf. CBarwar dūz)23 ju- ‘in, at’ (cf. CBarwar gu-) čud- ‘every’ (cf. CBarwar kul-) xu- a discourse particle (cf. CBarwar xu- ~ xo-) We can, therefore, observe once more how the frequent occurrence of certain words renders them immune to a common sound shift and allows them to keep their more original phonological shape. 23 This could alternatively be treated as a recent loan, although it appears to be well-integrated with the language and takes part in derivational processes, e.g. duzzüta ‘truth’. 229 Among the high-frequency words without the expected /ü/, one can also discern a grammatical category, the past base and the resultative participle of II and III stems, which in Azran have /u/ as the thematic vowel: š-d-r ‘to send’ (II), past base šudər-, resultative participle šudra (cf. CBarwar, past base mšodər- ~ mšudər-, participle mšudra) g-d-l ‘to freeze’ (III), past base mujdəl-, resultative participle mujdəla (cf. CBarwar, past base mugðəl-, participle mugðəla) The vowel /u/ in the Azran forms is probably retained to create a uniform paradigm of inflection with a single thematic vowel across the stems. This is even more evident when we compare the forms of the infinitive of II and III stems, which also have the vowel /u/, but in this case as a result of the shift from an earlier *o (see section 5.1). Thus: š-d-r ‘to send’ (II), infinitive šadure (cf. CBarwar mšadore) g-d-l ‘to freeze’ (III), infinitive majdule (cf. CBarwar magðole) The Azran forms of the past base, resultative participle and infinitive are, accordingly, šudərre, šudra, šadure in stem II and mujdəlle, mujdəla, majdule in stem III. The lack of the shift *u > /ü/ in the verbal paradigm, although causing the two otherwise distinct historical vowels *o and *u to collapse here, nevertheless results in a clear inflectional paradigm with a single thematic vowel. The absence of the expected shift is in this case of a morphological, rather than lexical, nature. This demonstrates that phonological shift is by no means blind, but distinguishes between lexical and morphological categories. It also suggests that paradigm pressure may have the upper hand in allowing or disallowing changes over phonological processes. 5.4 Language contact As far as the motivation for the shift of *u to /ü/ is concerned, the internal factors mentioned above seem sufficient: the process of a chain shift, triggered by the accumulation of back quality allophones, pushed the original vowels higher and to the front of the mouth. Other sources of /ü/ may be explained as the process of retiming of gestures involved in articulation; for example, assimilation to an adjacent bilabial, or to a following glide in the case of *iw > /ü/. Similarly, Khan attributes the development /uy/ from a long *u in Christian Urmi to palatalisation, classifying it thus as an internal phonetic process.24 24 Khan, 2016, p. 191. 230 Cross-linguistically, the independent emergence of vowel fronting is attested in other languages. For example, in some modern accents of English, such as in the south-east of England, long /u/ has shifted to the front, and as a consequence, the words looking and licking are pronounced with almost the same vowel quality.25 Another example of the internal development of rounded nonback vowels comes from Modern Mandaic, another Neo-Aramaic language. Here, the entire vowel system underwent a process of raising in comparison to Classical Mandaic. As part of this, the system underwent a shift in the pronunciation of /u/ from [u] to a more fronted [ʉ].26 For these languages, an internal explanation of vowel fronting is generally accepted; that is, no language contact is postulated as a source of the emergence of the front rounded vowel. In this situation one should ask whether language contact is of any relevance to the development of /ü/ in Azran. The same high front rounded vowel is indeed attested in Azeri Turkish,27 as well as in certain dialects of Kurdish, such as Amedia, Zakho and Akre.28 However, it is difficult to maintain that any of these varieties induced the emergence of /ü/ in Azran, as none of them has been in steady and direct contact with the dialect. For the Azeri Turkish and Kurdish ü-quality to have an effect on Azran, we would have to postulate an areal feature of vowel fronting, similar to the situation with the areal palatalisation of velar consonants. There are, however, insufficient grounds for this claim. For many other NENA and Kurdish dialects around Azran, no üquality has been reported, and so it does not seem to have travelled by continual spreading across language varieties. This is very different from the diffusion of consonant fronting described in section 4 above, which clearly proceeds from the east to the west, gradually losing its influence over particular NENA dialects. Considering, therefore, the phonological factors, the crosslinguistic attestations and the geographical distribution of this feature, it can be concluded that so far there is no evidence for any external input in the development of the vowel /ü/ in Azran. Rather, the emergence of the vowel is motivated internally. 6 Conclusions In considering consonant and vowel fronting in Azran here, a number of classes of words where processes of palatalisation or vowel-fronting did or did not occur were established – high-frequency words, low-frequency words, loanwords, words containing emphasis, and so on. The classification of Azran words and the associated processes are by no means exhaustive, nor are they 25 Uffmann, 2013. Häberl, forthcoming. 27 Householder and Lofti, 1965. 28 MacKenzie, 1961. 26 231 deterministic. There are, of course, words which do not lend themselves to the above grouping – for example, purya ‘light’, where no vowel fronting took place despite favourable conditions, since it is a native word with no emphasis; or ȷanȷar ‘threshing sleigh’, where palatalisation to an alveolar affricate occurred even though the word designates an artefact no longer used or seen in the community. The classes of Azran words suggested above are rather intended to illustrate the different processes that have been involved in shaping the phonological profile of the dialect. These processes encompass phonetics (gesture retiming, push-chain shift and backing of segments), morphology (paradigm pressure) and language use (frequency and language contact), all of which contribute to the apparently irregular diffusion of the innovative features. However looking behind this apparent irregularity, we have seen the way in which each different process maps separately onto the lexicon. In addition, it has been argued that internal processes should be recognised as largely responsible for many of the observed shifts; the role of external language influence as a factor in changes in progress in Azran needs to be carefully evaluated. References Bybee, J., 2001, Phonology and Language Use, Cambridge: CUP. Chirikba, V. A., 2006, “The problem of the Caucasian Sprachbund”, in Y. Matras, A. McMahon and N. Vincent (eds), Linguistic Areas: Convergence in Historical and Typological Perspective, Houndmills: Palgrave, pp. 25–93. Coghill, E., 2003, The Neo-Aramaic Dialect of Alqosh (unpubl. diss., University of Cambridge). ———, 2009, “Four versions of a Neo-Aramaic children’s story”, ARAM 21, pp. 251– 280. Häberl, C., forthcoming, “Neo-Mandaic: the view from Iraq”. Hoberman, R. D., 2007, “Semitic triradicality or prosodic minimality? Evidence from sound change”, in C. L. Miller (ed.), Studies in Semitic and Afroasiatic Linguistics Presented to Gene B. Gragg, Chicago: The Oriental Institute of the University of Chicago, pp. 139–154. Hopper, P. and Traugott, E. C., 2003, Grammaticalization, 2nd edn, Cambridge: CUP. Householder, F. W., Jr. and Lofti, M., 1965, Basic Course in Azerbaijani, Bloomington: Indiana University. Johanson, L., 2006, “On the roles of Turkic in the Caucasus Area”, in Y. Matras, A. McMahon and N. Vincent (eds), Linguistic Areas: Convergence in Historical and Typological Perspective, Houndmills: Palgrave, pp. 160–181. Kapeliuk, O., 2011, “Language contact between Aramaic dialects and Iranian”, in S. Weninger et al. (eds), The Semitic Languages: An International Handbook, Berlin: Mouton, pp. 738–747. Khan, G., 1999, A Grammar of Neo-Aramaic: The Dialect of the Jews of Arbel, Leiden: Brill. ———, 2008a, The Jewish Neo-Aramaic Dialect of Urmi, Piscataway, NJ: Gorgias Press. ———, 2008b, The Neo-Aramaic Dialect of Barwar, Leiden: Brill. 232 ———, 2016, The Neo-Aramaic Dialect of the Assyrian Christians of Urmi, Leiden: Brill. MacKenzie, D. N., 1961, Kurdish Dialect Studies, vol. I, London: OUP. Mutzafi, H., 2000, “The Neo-Aramaic dialect of Maha Khtaya d-Baz: phonology, morphology and texts”, JSS 45/2, pp. 293–322. ———, 2008, The Jewish Neo-Aramaic Dialect of Betanure (Province of Dihok), Wiesbaden: Harrassowitz. Napiorkowska, L., 2015a, A Grammar of the Christian Neo-Aramaic Dialect of Diyana-Zariwaw, Leiden: Brill. ———, 2015b, “Diachronic perspective on the status of emphasis in Diyana-Zariwaw”, in G. Khan and L. Napiorkowska (eds), Neo-Aramaic in its Linguistic Setting, Piscataway, NJ: Gorgias Press, pp. 130–144. ———, forthcoming, “Handling variation in Azran Neo-Aramaic with Articulatory Phonology”. Payne Smith, J., 1999, A Compendious Syriac Dictionary: Founded upon the Thesaurus Syriacus of R. Payne Smith, Eugene: Wipf and Stock. Stilo, D. L., 1994, “Phonological systems in contact in Iranian and Transcaucasia”, in M. Marashi (ed.), Persian Studies in Northern America: Studies in Honor of Mohammad Ali Jazyrey, Bethesda-Maryland: Iranbooks, pp. 75–94. Uffmann, C., 2013, Move in the Right Direction: Variation, Change, and Phonological Representations (seminar, Queens’ College, University of Cambridge, 31 January 2013). Younansardaroud, H., 2001, Der neuostaramäische Dialekt von Särdä:rïd, Wiesbaden: Harrassowitz. 233 The Neo-Aramaic Dialect of Telkepe ELEANOR COGHILL Uppsala University 1 Introduction1 The dialect described here is a dialect of North-Eastern Neo-Aramaic (NENA) spoken by the Chaldean Catholic Christians of the town of Telkepe. It, and other Christian dialects, are known as sūraθ to their speakers. The Telkepe dialect is similar to the dialects of the surrounding Chaldean villages but distinct enough to require a separate description. It is generally well understood by other Iraqi Chaldeans, because the təlkəpnāyə (natives of Telkepe) have formed a large part of Chaldean communities in the diaspora, in Baghdad and Detroit especially. Telkepe [təlkepə] (Arabic Tall Kayf) is a small town situated at the southern end of the Mosul Plain, about fifteen kilometres north of the city of Mosul. Historically Christian, it gained a sizable Muslim population as well. In 2014, with the surge of Islamic State in Iraq, Telkepe was captured and almost all its Christian inhabitants were forced to flee. Telkepe has since been recaptured, but it remains to be seen how many will return. Telkepe is at the southern tip of a string of Neo-Aramaic–speaking villages leading north from Mosul: Telkepe, Baṭnāya, Baqopa, Tisqopa and Alqosh. To the south-east of Mosul there are three other Neo-Aramaic–speaking villages: Karimlesh, Qaraqosh/Baghdede and Bariṭle/Barṭille. Most of the inhabitants of these Neo-Aramaic–speaking villages belong to the Chaldean Catholic Church, but the inhabitants of Qaraqosh and Bariṭle adhere mainly to the Syriac Catholic Church and the Syriac Orthodox Church respectively. There are also Arabic and Kurdish speakers of various ethno-religious backgrounds living in the local area (especially Christians, Yezidis and Shabaks). 1 I would like to acknowledge with gratitude the speakers of the Telkepe dialect who have assisted me in my fieldwork, especially Amera Mattia-Marouf , Shawqi Talia, Mahir Awrahem, Haniya, Rania, Francis and Khalid. I would also like to thank Bishop Emanuel Shaleta, who helped me so much during my trips to Detroit. I also extend my thanks to the editors of this volume for their helpful suggestions. My deep gratitude goes especially to Geoffrey Khan, who introduced me to this wonderful language with its endless riches and who taught me to be a scholar. 234 The etymology of the name Telkepe is apparently ‘the mound of stones’ (Arab. tall ‘mound’, Aramaic kepə ‘stones’). This refers to the large archaeological tell at the edge of the village. It has not been excavated due to the village cemetery situated on it. According to Wilmshurst, the earliest mention of Telkepe is in an inscription commemorating the restoration of a nearby monastery in 1403 “by the residents of Telkepe”, and he suggests that Telkepe “may well have been founded as late as the fourteenth century”.2 Of course, the tell points to an ancient habitation on the site; it is not known what the name was of the Assyrian settlement now hidden under the tell. Formerly adhering to the Church of the East, Telkepe was one of the first villages to unite with the Catholic Church.3 According to Wilmshurst, there were Catholic missionaries in Telkepe in the 17th century and there were a significant number of converts by the end of the century.4 By the beginning of the 19th century, those in union with Rome were in the majority. Already in the 19th century Telkepe was the largest Christian village in the plain of Mosul and many of the clergy of the Chaldean Church were its sons. Its prominence in the Chaldean Church continues to this day. In the late 19th century, it had two churches, the churches of Saint Cyriacus and of the Virgin Mary;5 within a few decades the number grew to six. There are also several shrines.6 Telkepe is notable for its history of emigration, and communities of təlkəpnāyə are now found in all the major cities of Iraq, as well as abroad, especially in Detroit, Michigan. In Iraq the təlkəpnāyə are prominent in the management of hotels, while in Detroit they have predominantly worked in the grocery business. Emigration to Detroit began in the early 20th century, and the təlkəpnāyə are the largest group in the huge Chaldean community there.7 Until recently there was little published specifically on the dialect of Telkepe, although there were two articles by Sabar with texts and grammatical notes.8 More generally on the dialects of the area of the Mosul Plain, there are several early works providing information.9 Unfortunately these do not distinguish between the dialects of the area, which, though highly mutually intelligible, nevertheless are also clearly distinct in phonology, morphology, syntax and lexicon. 2 Wilmshurst, 2000, p. 223. The inscription was noted by Sachau, 1883, p. 361. Fiey, 1965, p. 360. 4 Wilmshurst, 2000, p. 224–226. 5 Sachau, 1883, p. 367. 6 Fiey, 1965, p. 369. 7 Sengstock, 2005. 8 Sabar, 1978 and Sabar, 1993. 9 Socin, 1882; Guidi, 1883; Sachau, 1895; Rhétoré, 1912; Maclean, 1895; Maclean, 1901. 3 235 More recently, studies have been published on individual dialects of this area, such as the varieties spoken in Tisqopa, Qaraqosh, Alqosh, Karimlesh and Bariṭle.10 In recent years I have also published a number of papers covering individual aspects of the dialect of Telkepe.11 We are fortunate in having a number of manuscripts of religious poetry composed in the dialects of the Mosul Plain,12 with the earliest dating to the 16th and 17th centuries. These early texts clearly show dialectal features of this region, while also exhibiting archaic features now lost, as well as lacking certain analytic verbal constructions which presumably developed later. They are therefore a priceless source for the historical development of the NENA dialects of this region.13 This study of the dialect of Telkepe was carried out as part of the NorthEastern Neo-Aramaic Project at Cambridge University, funded by the Arts and Humanities Research Board. Most of the fieldwork on which it is based was carried out during two fieldwork trips to Detroit in 2004 and 2007. Some other interviews were conducted in London and Chicago in 2006, while further interviews were also carried out by telephone. This paper will focus on the basic phonology, morphology and lexicon of the dialect, rather than the syntax, on which I have published elsewhere and which will also be treated in a separate monograph.14 I have tried here to keep to the same structure as in my other paper-length dialect descriptions, for maximum comparability.15 2 2.1 Phonology Phonemic inventories 2.1.1 Consonants The inventory of consonant phonemes in the dialect of Telkepe is given in table 1. Note the IPA values for the following symbols: č [ʧ], j [ʤ], ž [ʒ] (as an allophone of š), y [j], ġ [ɣ], ḥ [ħ], ʿ [ʕ], ʾ [ʔ]. Other symbols have their IPA values. Apart from ḥ, consonants with a dot under are the emphatic (velarised/ 10 See Rubba, 1993a and Rubba, 1993b for Tisqopa; Khan, 2002 for Qaraqosh; Coghill, 2004, Coghill, 2005 and Coghill, forthcoming-b for Alqosh; Borghero, 2008 for Karimlesh; and Mole, 2015 for Bariṭle. 11 See Coghill, 2008; Coghill, 2009; Coghill, 2010a; Coghill, 2010b; Coghill, 2014; Coghill, 2015. 12 See e.g. Pennacchietti, 1990; Poizat, 1990; Poizat, 1993; Mengozzi, 2002a; Mengozzi, 2002b; Mengozzi, 2011. 13 For diachronic studies using these texts as sources, see Mengozzi, 2012; Coghill, 2010b, pp. 377–379; Coghill, 2016, especially pp. 234–239, 268–282. 14 See Coghill, 2010a; Coghill, 2010b; Coghill, 2014. 15 See Coghill, 2013 on Peshabur; Coghill, forthcoming-b on Alqosh. 236 Stops/affricates plain voiceless voiced voiceless emphatic voiced Fricatives plain voiceless voiced voiceless emphatic voiced Nasals Lateral approximant Tap/trill plain emphatic Approximants p b t d ṭ f (v) θ ð s z ṣ č j c̣ k g š x ġ Laryngeal Pharyngeal Uvular Velar Palatal Post-alveolar Alveolar Dental Labio-dental Bilabial Table 1. Consonant inventory ʾ q ḥ h ð̣ m n l r ṛ w y ʿ pharyngealised) versions of the undotted consonant; for instance, the symbol ð̣ represents [ðˁ]. Unemphatic voiceless plosives are lightly aspirated, while emphatic or voiced stops are unaspirated: talθa ṭūṛå dəx [tʰɛlθæ] ‘the year before last’ [tˤuːrˤɒ] ‘mountain’ [dɘx] ‘how?’ Some phonemes are only found in loan-words, but are nevertheless common; for example /ð ̣/ occurs in words from Arabic. On the other hand, /v/ is only attested in the Kurdish loan-word šivānå ‘shepherd’. Voiced plosives and fricatives are devoiced in word-final position: mez [meːs] ‘table’ (K. mêz), primuz [priːmus] ‘primus stove’. This devoicing also occurs in Alqosh, and is an areal feature also found in the Qəltu-Arabic dialects of Mosul and Anatolia, as well as Kurdish dialects.16 The voicing is preserved when the word is followed by a suffix: mezā́t [meˑzæːt] ‘tables’, primuzā́t [priˑmuzæːt] ‘primus stoves’. 16 For the dialects of Mosul, see Jastrow, 1979, p. 41; for those of Anatolia, see Jastrow, 1978, p. 98; for Kurdish dialects, see Mackenzie, 1961, pp. 48–49. 237 2.1.2 Vowels There are nine vowel phonemes, five of them long and four short. The distinction between long and short is not phonemic in all environments. The phonemes /o/, /e/, and /i/ are usually realised as long, but are not marked as such in order to minimise the number of diacritics. The vowel phonemes are: Long vowels: /i/ /e/ /ā/ /o/ /ū/ Short vowels: /ə/ /a/ /å/ /u/ The most common realisations of these vowels (in the environment of nonemphatic consonants) are shown below. In an emphatic environment, they may be backed and lowered, at least in the onset. Long vowels may be realised as mid-long, or even short, in an unstressed syllable. /i/ /e/ /ā/ /o/ /ū/ /ə/ /a/ /u/ /å/ = = = = = = = = = [iː] [eː]; it is often diphthongised, with a lowering of the tongue, [e̞e̝] [æː] [oː] [uː]; before /y/ it may be realised as [yː]: rūyå [ˈryːjɒ] ‘grown up’ [ɪ] ~ [e̠] or a close-mid central vowel, [ɘ] [æ] or centralised to [ɜ]; in final position sometimes long, [æː] [u] or a more lax [ʊ] open back to mid central, slightly rounded, [ɒ], [ɐ] or [əʊ]: hallå [ˈhællɒ] ‘give her’, ʾiðå [ˈʔi:ðəʊ] ‘hand’ In an unstressed final open syllable, the length distinction of /a/–/ā/ and /u/– /ū/ is neutralised, and so only the following vowels occur: /i/, /e/, /o/, /ə/, /a/, /u/ and /å/, and the diphthong /ay/. In fact, /å/ only occurs in this position. What is unusual among NENA dialects is the presence of two distinct ‘a’ phonemes in final position, /a/ and /å/, where other NENA dialects have one: skinå qṭəllå ʾānå ‘knife’ ‘she killed’ ‘I’ skina qṭəlla ʾāna ‘her knife’ ‘they killed’ ‘those’ The realisation of these two vowels is quite distinct: final /a/ is a front vowel [æ], not normally centralised (unlike non-final /a/), with more tendency to be pronounced long as [æː]; /å/ is a back-central vowel, usually slightly rounded. Given the phonetic similarity of the former to the non-final /a/ phoneme, I have chosen to write them the same. Arguably, however, one could alternatively view å [ɒ] as an allophone of non-final /a/, given that when word stress is shifted on to it, it changes to /a/: 238 ʾarbå šātå ʾəkmå xənnå ‘four (m)’ ‘year’ ‘how many?’ ‘other’ ʾarbá꞊gūrə ‘four men’ šātá꞊xurtå ‘next year’ ʾəkmá꞊ʾarmonə ‘how many pomegranates?’ xənná꞊ʾaxonå ‘another brother’ There is one diphthong in Telkepe, normally only found in final open syllables (stressed or unstressed), usually a third person plural morpheme. It may also be found in certain Classical Syriac loans. /ay/ = [ɛy]; e.g. beθáy ‘their house’, kullay ‘all of them’, bassay! ‘enough for them’, wāway ‘they were’, way! (similar to German doch!), haymānūθå ‘faith’, suraytūθå ‘Christianity’ (< surāyå ‘Christian’) 2.2 Word stress Word stress is mostly penultimate, as is generally the case in Christian dialects of Iraq; e.g. mašəlxā́na ‘robber’, kəmšāqə́llə ‘he took it’. Non-penultimate stress can be found in specific verbal forms, e.g. mášəlxu ‘rob! (pl)’, kpāθə́xwālə ‘he used to open it’. As a result of this, stress is marginally phonemic: mbā́šəllə ‘cook it!’ mbāšə́llə ‘(that) he may cook it’ In this paper, word stress will only be marked where it is not penultimate. 2.3 Synchronic sound rules17 2.3.1 Assimilation Assimilation of consonants to each other is very common in Telkepe, as in other NENA dialects. It involves voicing, nasality, place of articulation and 17 Some forms and phrases in this paper are glossed, with a full list of abbreviations given in the Appendix; the Leipzig Glossing Rules are used where possible. Note, however, that, for economy, the NENA Present Base forms are not explicitly glossed as Present Base; all other verb forms are glossed with their category name. Thus the Present Base form k-šāqəl ‘he takes’ is glossed as [IND-go.3MS], while the Past Base form šqəl-lə ‘he took’ is glossed as [take.PASTL.3MS]. Words and morphemes are often combined in phrases containing a single stress: one element may be a clitic, but this is not necessarily the case. The long equals sign ‘=’ is used where the stress is on the second component, e.g. xā=xənnå ‘each other’. The short equals sign ‘꞊’ is used where the stress is on the first component, e.g. xoš꞊ʾixālå ‘good food’ and gāre꞊lə ‘it is a roof’. For affixes a simple hyphen is used. Note, however, that the distinction between affixes and clitics is somewhat blurred. For instance, the monoconsonantal prepositions (b-, land m-), as well as the genitive marker d-, are somewhere between. 239 emphatic spread. Usually a consonant assimilates to the following one (regressive/anticipatory assimilation), but in emphatic spread consonants before and after may be affected. Assimilation is very common with grammatical prefixes, and in these cases it will be indicated in the transcription; for example p-siyārå ‘in the car’ is underlyingly b- + siyārå. When it affects part of the root, on the other hand, assimilation will not be indicated in the orthography; e.g. xzelə ‘he saw’ is produced with a voiced initial consonant, as [ɣzeːlɘ] (compare kxāzə [kxaːzɘ] ‘he sees’). Assimilation, especially voicing assimilation, also commonly occurs over the word boundary, but such sandhi will not be indicated in the transcription. Assimilations which are shown in the examples in this section but which are normally ignored in my transcription will be put here in square brackets. Most consonants regularly assimilate to a following consonant in voicing: Underlying form b- + šāqəl [FUT- + take.S.3MS] k- + zad-ux [IND- + fear-S.1PL] kaləbθå bas dahå Assimilation p-šāqəl g-zadux kalə[p]θå ba[z] dahå ‘he will take’ ‘we fear’ ‘bitch’ ‘but now’ There are certain consonants that neither cause nor undergo voicing assimilation: the laryngeals /ʾ/ and /h/, the pharyngeal approximant /ʿ/ and the ‘sonorants’, that is, the nasals /m/ and /n/, the liquids /l/ and /r/ and the semivowels /y/ and /w/, as well as any emphatic counterparts of these. An emphatic consonant will normally make a neighbouring consonant emphatic also. Emphatic spread may also affect consonants not immediately adjacent: Underlying form qiṣ- + -tå [cut.RES.PTCP- + -FS] ṭūrå [mountain] ltexəd + ṭūrå [down + mountain] Assimilation qəṣṭå ‘cut’ ṭūṛå ‘mountain’ ltexəṭ=ṭūṛå ‘down the mountain’ The consonants /d/ and /b/ may, before a nasal, themselves become the equivalent nasal consonant, /n/ and /m/ respectively. This is obligatory with /b/ before /m/ and very common (though not obligatory) with the other combinations: Underlying form Assimilation b- + mašloxə [in- + rob.INF] m-mašloxə ‘robbing’ b- + nāpəl [FUT- + fall.S.3MS] m-nāpəl ‘he will fall’ b- + mez [in/on- + table] m-mez ‘on to the table’ ltexəd + mez [underneath + table] ltexən=mez ‘underneath the table’ 240 There are two cases where a consonant consistently assimilates to the following one in terms of its place of articulation: /n/ becomes an [m] before the bilabial /p/ (the sequence /nb/ is not attested) and /k/ is backed to [q] before uvular /q/: Underlying form npālå [fall.INF] k- + qem-ən [IND- + get_up-S.1MS] Assimilation [m]pālå ‘to fall’ qqemən ‘I (m) get up’ Sometimes a plosive assimilates to a following fricative, although this is not obligatory. In the following cases this results in total assimilation: Underlying form yom-əd + šabθå [day-CST + week/Saturday] k- + xašw-an [IND- + think-S.1FS] kud + θe-li [when + come.PAST-L.1SG] Assimilation yoməš=šabθå ‘Saturday’ x-xašwan ‘I (f) think’ ku[θ] θeli ‘when I came’ 2.3.2 Secondary gemination There is a tendency (but not a rule), where a short vowel is in an open syllable, for the syllable to be closed by means of the gemination of the following consonant. The main cases of this are presented below: *la꞊ + piš(ən) lappəš [NEG꞊ + there_is_left] la꞊ + zilə la꞊zzilə [NEG꞊ + PRSP.3MS] k- + zālə kəzzālə [IND- + go.3MS] kəm- + (ʾ)āxəl-lə kəmmāxəllə [PST_PFV- + eat.S.3MS-L.3MS] ‘there is/are no … left’ ‘he is not going to’ ‘he goes’ ‘he ate it’ 2.3.3 Plosivisation of interdentals adjacent to /l/ As is common across NENA, there is a tendency for an interdental fricative adjacent to an /l/ to become a stop: Underlying form yalðå (ylð I) məθ- + -lə Output yaldå mətlə ‘she may give birth’ ‘he died’ 241 2.3.4 Vowel length alternations A selection of the synchronic vowel alternations in this dialect are presented here. Syllable closure, through the addition of a suffix, usually results in the shortening of a vowel: /ā/ to /a/, /i/ to /ə/, /ū/ to /u/ and /o/ to /o/ ~ /u/ ~ /a/: Open syllable Closed syllable ʾazālå ‘going’ ʾazaltå ‘going’ (msg active participle) (fsg active participle) pθixå ‘open (msg)’ pθəx-tå ‘open (fsg)’ yarūqå ‘green (msg)’ yaruq-tå ‘green (fsg)’ qṭol ‘kill!’ qṭol-li ~ qṭal-li ‘kill me!’ komå ‘black (msg)’ kum-tə ‘black (fsg)’ šaxlopə ‘to change’ šaxlap-tå ‘changing’ (infinitive) (fsg infinitive) Vowel lengthening also takes place, in a similar way as in Alqosh, either through the opening of a syllable or when a suffix is added that places the vowel in a non-final open syllable: čangal k-xāzə θelə p-kāθu ‘fork’ ‘he sees’ ‘he came’ ‘he will write’ čangāli k-xāzela θə́lelan p-kāθūlə ‘my fork’ ‘he sees her’ ‘there came to us’ ‘he will write it’ Both short final ‘a’ vowels shift to /ā/ under the latter condition: k-xāza k-šaqlå ‘they see’ ‘she takes’ k-xāzālə ‘they see him’ k-šaqlālə ‘she takes him’ Vowel shortening often takes place when the stress is shifted from an open syllable making it pretonic: ṭāle gūdå ‘to him’ ‘wall’ ṭaláy gudānə ‘to them’ ‘walls’ Vowels are also shortened when moved to a stressed position before two or more syllables: b-zālux ‘you (msg) will go’ k-šaqlūtu ‘you (pl) take’ θelə ‘he came’ 242 b-záloxu ‘you (pl) will go’ k-šaqlə́tūlə ‘you (pl) take him’ θə́lelan ‘there came to us’ The shift /e/ > /ə/, seen in the previous example, is morphologically conditioned; it does not occur before the anterior suffix: θeli 2.4 ‘I came’ θéwāli ‘I came (remote past)’ Historical developments 2.4.1 Beḡaḏkep̄ aṯ and other consonant changes As in NENA generally, the plosive and fricative allophones of Late Aramaic *b, *g, *d, *k and *t have for the most part become separate phonemes.18 As usual, *p is the exception to this: its allophones have merged as plosive /p/: Stop *b → b *g → g *d → d *k → k *p → p *t → t Fricative *ḇ → w *ḡ → ʾ *ḏ → ð *ḵ → x *p̄ → p *ṯ → θ Examples of fricative reflexes šwāwå ‘neighbour’ (Syr. šəḇāḇā) raʾolå ‘valley’ (Syr. rāḡōlā) ʾiðå ‘hand’ (Syr. ʾīḏā) rakixå ‘soft’ (Syr. rakkīḵā) ʾuprå ‘soil’ (Syr. ʿap̄ rā) māθå ‘village’ (Syr. māṯā) As indicated above, Telkepe, like other dialects of the Mosul Plain, is among those dialects which have preserved *ṯ and *ḏ as interdentals (/θ/ and /ð/), rather than merging them with the dental stops. Original *ḥ has merged with *ḵ as /x/, e.g. xəṭṭə ‘wheat’ (< ḥeṭṭē), as in most but not all NENA dialects.19 Original *ʿ and *ḡ have generally merged, both shifting to /ʾ/, e.g. ʾamṛå ‘wool’ (Syr. ʿamrā) and šʾārå ‘fuel, kindling’ (Syr. šəḡārā ‘kindling’). Immediately before or after a consonant, the resultant /ʾ/ may have been elided, e.g. ṣubetå ‘finger’ (< *ṣubəʾta < *ṣubeʿta, Syr. ṣeḇʿəṯā), ṭəmå ‘taste’ (< *ṭəmʾa < *ṭemʿā, Syr. ṭaʿmā), xāta ‘thorn-bush’ (< *xaʾta < *xaḡta, Syr. ḥāḡtā). Two cases where it was not elided are paʾlå ‘labourer’ (Syr. pāʿlā) and pəʾlə ‘radishes’ (< *peḡlē, Syr. puḡlē ~ paḡlē). Apparent exceptions to these sound shifts, where the original sounds are preserved (as ḥ, ʿ and ġ), are usually borrowings from Classical Syriac (see section 2.4.3). Historical gemination of consonants has mostly been lost where it followed /a/, e.g. yāmå ‘sea’ (Syr. yammā), rābå ‘big’ (Syr. rabbā), rakixå ‘soft’ (Syr. rakkiḵā), mzābən ‘he may sell’ (Syr. məzabbēn).20 Gemination loss and the resultant presence of single post-vocalic plosives is one of the reasons for the phonemicisation of the plosive-fricative distinction in NENA. 18 In Syriac, these consonants were realised as fricatives when they occurred after a vowel, unless they were geminated (when they were realised as a plosive). In all other positions they were realised as plosives. 19 The two are merged as /ḥ/ in the dialects of Hertevin (Jastrow, 1988, p. 6), Umra (Hobrack, 2000, p. 22–24) and Derabün (my own fieldwork data). 20 In stressed syllables the vowel was lengthened in compensation for the loss of gemination. 243 2.4.2 Vowel changes The vowel phonology of Telkepe is relatively conservative within NENA, except that the old diphthongs have been monophthongised. The reconstructed proto-forms in what follows are based on Syriac forms, as well as those of other NENA dialects. Original *ō (as in the eastern pronunciation of Classical Syriac) is preserved as /o/, e.g. *rāḡōlā > raʾolå ‘valley’ and *bərōnā > brona ‘boy, son’. Original *ē is preserved as /e/ [eː] in non-final position and as /ə/ in final position, e.g. *rēšā > rešå ‘head’ and *ḥāzē > xāzə ‘he may see’. In its preservation of *ō and *ē, Telkepe resembles most other dialects native to northern Iraq and much of the Hakkari province in Turkey.21 The old diphthong *aw (< *aw, *aḇ and *ap̄ , where *a in some cases < *ā) is also realised as /o/: *gawzā > gozå *zaḇnā > zonå *ṭlāp̄ ḥē > ṭloxə ‘walnut’ ‘time’ ‘lentils’ This matches what is found for other documented Mosul Plain dialects: Alqosh, Tisqopa (e.g. zon- < *zawn- Present Base ‘sell’), Bariṭle (e.g. goṛa < *gawra < *gaḇrā ‘man’), and Qaraqosh.22 The old diphthong *ay (and *āy) has been monophthongised in Telkepe and merged with *ē in most positions (non-final and stressed final), unlike in Alqosh, where it is monophthongised but kept distinct as /ɛ/,23 or other dialects such as Peshabur, where it is preserved as a diphthong /ay/ [ɛi]:24 *bayṯā > TK beθå Alq. bɛθa *payšā > TK pešå Alq. pɛša *xzay > TK xze Alq. xzɛ Pesh. bayθa ‘house’ Pesh. payša ‘she may become’ Pesh. xzay ‘see (pl)!’ In a final, unstressed, open syllable, *ay is also monophthongised, but as /a/: *xāzay > TK xāza *šqəl-lay > TK šqəlla *ʾannay > TK ʾāna 21 Alq. xāzɛ Pesh. xāzay ‘they may see’ Alq. šqəllɛ Pesh. šqəllay ‘they took’ Alq. ʾānɛ ‘those’ This contrasts with some dialects of eastern Hakkari, such as Jilu, and the Christian dialects of Urmi, in which *ē has in many cases shifted to /i/ and *ō to /u/ (i.e. *rēšā > riša ‘head’ and *bərōnā > bruna ‘boy, son’). For Jilu, see Fox, 1997, pp. 17–18, 127; for Urmi, see Khan, 2016, pp. 186–87, 190–91. 22 For Alqosh, see Coghill, 2004, p. 78; for Tisqopa, see Rubba, 1993a, p. 175; for Bariṭle, see Mole, 2015, p. 112; and for Qaraqosh, see Khan, 2002, p. 54. 23 Coghill, 2004, p. 78. 24 Coghill, 2013, p. 39. 244 There are a few cases where *ay in a final unstressed syllable is realised as /e/. These are the feminine imperatives of verba tertiae /y/ in derivations II, III and Q, which usually end in /e/, even though this goes back to unstressed *ay, which should be realised as /a/. This exception results presumably from analogy with the forms in derivation I, which end in /e/ (e.g. xzé ‘see (f)!’). *mšā́ṛay (šṛy II) > TK mšā́ṛe ‘begin (f)!’ *máḥkay (ḥky III) > TK máḥke ‘speak (f)!’ An exception to the exception is meθa ‘bring (fsg)!’ (< *mayθay, ʾθy III), suggesting that the analogy is not made consistently. The historical 3pl pronominal suffix *-ayhən-25 has become a diphthong -ay, e.g. beθáy ‘their house’ (compare Alq. bɛθɛ́y). In some forms the suffix does not take the stress, but the diphthong remains: e.g. kúllay ‘all of them’, mə́nnay ‘from them’ and ʾarbáθnay ‘four of them’. Telkepe may be contrasted with another dialect of the northern Mosul Plain, namely Tisqopa. In this dialect *ay has also generally merged with *ē to /e/, e.g. *mayθa > meθa ‘she may die’.26 On the other hand, final unstressed *ay is preserved as a diphthong: kxāzey ‘they see’, 3pl L-suffix -ley.27 The existence of two ‘a’ vowel qualities in this dialect has already been mentioned. The back /å/ vowel, found only in unstressed final open syllables, is usually a reflex of original *a < *ā28 in final position, as found in nominal and adjectival inflection and some pronouns, e.g. nāšå ‘person’, māθå ‘village’, skināθå ‘knives’, rābå ‘big’ and ʾāwå ‘that (m)’, as well as in the anterior suffix -wå (< *-(h)wā), when word-final. The front /a/ vowel in unstressed final position is usually a reflex of original *ay, as mentioned above. Both ‘a’ vowels, however, also go back to original *-ah < *-āh, but in different morphological contexts. The 3fsg possessive suffix on nouns and prepositions, *-ah < *-āh, is realised as -a, e.g. barāna ‘her ram’. The 3fsg L-suffix, *-l-ah < *-l-āh, on the other hand, is realised as -lå. I have elsewhere suggested that already in early NENA the /h/ was lost in the L-suffix, but retained in the possessive suffix in order to disambiguate it from the nominal inflection *-a.29 Various dialects preserved this distinction in different ways: some by preserving the /h/ or by reinforcing it as a pharyngeal /ḥ/. In Telkepe, the /h/ was lost, but the vowel quality distinguished the possessive suffix -a from the nominal inflection -å, which now had a back vowel. 25 See Hoberman, 1988, p. 565 for this reconstruction. Rubba, 1993a, p. 176. 27 Rubba, 1993a, pp. 71–72. 28 The original Aramaic ending was -ā, but across NENA it is normally a short -a. A 12th century source for early NENA also suggests a short vowel; see Khan, 2008b, p. 97. 29 See Coghill, 2008, pp. 91–97. 26 245 2.4.3 Borrowed phonemes The following consonants are introduced into Telkepe Neo-Aramaic primarily through loan-words from neighbouring languages, mainly Kurdish and Arabic: ð̣ č f j (< Arab.) manð̣ofə ‘to clean’ (< K. and Iraqi Arab.) čāyi ‘tea’, čangal ‘fork’, ču꞊ ‘no’ (< Arab.) flān- ‘such and such’, fyāṛå ‘to fly’ (< Arab. and K.) jullə ‘clothes’, mjawobə ‘to answer’ The following consonant is found only marginally: v (< K.) šivānå ‘shepherd’ (the native synonym maṛəʾyānå is also used) The sounds /ʿ/, /ḥ/ and /ġ/ (i.e. ḡ), which mostly underwent sound changes in the native lexicon, have been reintroduced into the language through loanwords from Arabic and Classical Syriac; e.g. ʿāṣərtå ‘evening’ (< Arab. ʿaṣr), yaʿqu ‘Jacob’, ḥaqquθå ‘truth’ (< Arab. ḥaqq), ḥaššå ‘suffering, Passion’ (< Syr. ḥaššā), ġliṭå ‘wrong’ (< Arab. ġlṭ i ‘to err’), paġrå ‘body (of Christ)’ (< Syr. paḡrā). 3 3.1 Morphology Pronouns In table 2 are the independent personal pronouns as well as the pronominal suffixes which can be affixed to nouns (with possessive function) and to prepositions. Table 2. Personal pronouns Independent pronouns 3 msg fsg pl 2 msg fsg pl 1 sg pl ʾāwu ʾāyi ʾani ~ ʾāni ʾāyət ʾāyat ʾaxtu ʾānå ʾaxni ‘he’ ‘she’ ‘they’ ‘you (msg)’ ‘you (fsg)’ ‘you (pl)’ ‘I’ ‘we’ Pronominal suffixes -e -a -áy -ux -ax -óxu -i -an beθe beθa beθáy beθux beθax beθóxu beθi beθan ‘his house’ ‘her house’ ‘their house’ ‘your (fsg) house’ ‘your (msg) house’ ‘your (pl) house’ ‘my house’ ‘our house’ In the third person singular possessive suffixes, Telkepe contrasts with some other dialects of the Mosul Plain (Tisqopa, Alqosh, Karimlesh and Qaraqosh), 246 which, instead of losing the *h of 3msg *-eh and 3fsg *-ah, have strengthened it to a pharyngeal, -əḥ and -aḥ.30 The independent possessive pronouns are formed on the stem diy-, e.g. diyi ‘mine’. These are typically, though not only, used predicatively, e.g. lelə diyoxu [NEG.COP:3MS POSS:2PL] ‘It is not yours (pl)’. Table 3 gives the demonstrative pronouns in both their independent and attributive forms. Table 3. Demonstrative pronouns Near deixis Far/absent deixis Independent sg ʾāyi ~ ʾāði Attributive ʾāyi ~ ʾaθ pl ʾāni ʾan ~ ʾāni Independent msg ʾāwå fsg ʾāyå pl ʾāna Attributive ʾāwå ~ ʾo ʾāyå ~ ʾe ʾāna The attributive forms usually form a stress phrase with the following noun. The stress may fall either on the demonstrative or the noun. As shown in the table, Telkepe, similarly to Alqosh and Qaraqosh,31 has only two distinctions in deixis: e.g. ʾan꞊nāšə ‘these people’ vs ʾāna꞊nāšə ‘those people’. Contrast this with dialects further north, such as Peshabur, which distinguish between ‘near’ and ‘far’ (both of which can be pointed towards) and ‘absent’ deixis (where the direction is unknown or irrelevant); the masculine singular forms in Peshabur are ʾawwa ‘this (here)’, ʾawāḥa ‘that (there)’ and ʾāwa ‘that (absent/past time)’.32 The reflexive pronoun is formed from gyānå ‘soul, self’ with possessive suffixes, e.g. la꞊maʿiq-at gyān-ax! [not꞊bother-S.2FS self-2FS] ‘Don’t bother yourself (f)’. Reciprocity can be expressed with ʾə́ġðāðə or xā=xənna (with feminine form ġðā=xurta/xərta) [one=other] ‘each other’, e.g. ʿənna xā=xənnå [help.PAST:3PL one.M=other.M] ‘They (m or mixed) helped each other’, ʿənna ġðā=xərtå ‘They (f) helped each other’. 3.2 Nouns Masculine nouns usually end in -å, e.g. gorå ‘man’, kalbå ‘dog’ and kθāwå ‘book’. Feminine nouns usually end in -Tå, that is, either -tå (< *-ta) or -θå (< *-ṯa), e.g. sustå ‘mare’, šabθå ‘week’ and betå ‘egg’. There are also some unmarked feminine nouns, which end in -å: yəmmå ‘mother’, dūkå ‘place’, ʾaqərwå ‘scorpion’, ʾanānå ‘cloud’, ʾarå ‘earth’, ʾəzzå ‘goat’, ʿalmå ‘world’, ʿəddānå ‘time’ and berå ‘well’. 30 See Coghill, 2008, pp. 96–97 for an explanation for the various developments these suffixes have undergone. 31 Deixis in Alqosh is described in Coghill, 2004, pp. 112–113; and the system of Qaraqosh in Khan, 2002, pp. 81–82. 32 In Peshabur, furthermore, greater distance in far deixis can be indicated by lengthening the stressed syllable: ʾawāāḥa or ʾawaʾḥa ‘that one, way over there’. For a description of deixis in Peshabur, see Coghill, 2013, pp. 97–100. 247 Nouns with other endings may be masculine or feminine: e.g. gārə (m) ‘roof’, lelə (f) ‘night’, xūwə (f) ‘snake’, məndi (m) ‘thing’, kālu (f) ‘bride’. Female beings (animals or humans) are always feminine, e.g. yəmmå ‘mother’, as are most place names, e.g. baġdad ‘Baghdad’, təlkepə ‘Telkepe’. The feminine endings -Tå and -iθå are often used for derivations that, in relation to the source noun, are female, singulative or diminutive, e.g. qāṭu (m) ‘tomcat’, qaṭuθå (f) ‘female cat’; məzzə (pl) ‘hairs’, məzzetå (f) ‘(single) hair’; quprānå (m) ‘shelter, booth’, qupraniθå (f) ‘small shelter, booth’. As in Alqosh, -u also occurs as a diminutive suffix, especially in hypocoristic names, e.g. sotu ‘little old lady’ (< sotå ‘old woman, granny’), maxxu ‘Mike’ (< mixāʾíl ‘Michael’), and šammu ‘Sam’ (< šmuʾél ‘Samuel’). Another diminutive suffix is -onå; the examples elicited with this suffix were of animals and people with disabilities, e.g. kalbonå ‘little dog’, səmyonå ‘blind man’. There are eight plural suffixes, whose distribution is lexically defined. The plural -ā́t is borrowed: it derives from Arabic -āt but occurs not only with Arabic loans but also with European loans (perhaps via Arabic). The suffixes, along with examples, are: -ə -ānə -āθå -awāθå -wāθå -yāθå -āCe -ā́t torå (m) ‘bull’, pl torə; ʾabəštå (f) ‘raisin’, pl ʾabišə gūdå (m) ‘wall’, pl gudānə; dūkå (f) ‘place’, pl dukānə ʾaqlå (f) ‘leg’, pl ʾaqlāθå; šišəltå (f) ‘chain’, pl šəšlāθå deṛå (m) ‘monastery’, pl deṛawāθå; ʿammå (m) ‘paternal uncle’, pl ʿammawāθå nāšå (m) ‘person’, pl našwāθå; səpθå (f) ‘lip’, pl səpwāθå ʾitotå (f) ‘party’, pl ʾitoyāθå; xawərθå (f) ‘(female) friend’, pl xawəryāθå təllå (m) ‘hill’, pl təllālə; səkθå (f) ‘ploughshare’, pl səkkākə mez (m) ‘table’, pl mezā́t (also mezə); primuz (m) ‘primus stove’, pl primuzā́t Alqosh and Qaraqosh have also borrowed the Arabic plural -ā́t, but in those dialects it has lost the stress, in line with the native penultimate stress, e.g. Alq. maḥállə ‘town quarter’, pl maḥállat.33 There exist irregularities in the plurals of some common words. Some are the same as in Alqosh, e.g. gorå ‘man’, pl gūrə (Alq. gūrə), while others are different, e.g. ʾaxonå ‘brother’, pl ʾaxawāθå (Alq. ʾaxunwāθa). Both bronå ‘boy, son’ and brātå ‘girl, daughter’ show nasal assimilation in their plurals, mnonə (Alq. bnonə) and mnāθå (Alq. bnāθa) respectively. The following are entirely irregular: gā ‘time’, pl gāyi; and ʿaji ‘child’ (< Mosul Arab.), pl ʿajāyå. 33 Coghill, 2005; Coghill, forthcoming-c. 248 3.3 Adjectives As generally in NENA, adjectives show at most a three-way distinction in gender and number: masculine singular, feminine singular and common plural. There are four patterns of inflection in Telkepe adjectives, shown in table 4. The first three patterns vary only in the feminine inflection; the last class is uninflected (invariable). Table 4. Adjective inflections Pattern Masculine Feminine Plural 1 2 3 4 -å -å -å -Ø -Tå -ə -Tə -Ø -ə -ə -ə -Ø Inflectional patterns 2 and 3 are only used for certain very restricted sets of adjectives. Pattern 4 is used with certain Arabic loan adjectives. All other adjectives take inflectional pattern 1, which is the original Aramaic inflection. Adjectives which take a particular inflectional pattern tend to follow certain morphological and derivational patterns, which will also be discussed here. Adjectives taking inflectional pattern 1 include derivations ending in -ānå or -(n)āyå, as well as the following common adjectival patterns: CCiCå, CaCiCå, CaCCiCå, CaCūCå and CaCāCå. Adjectives taking inflectional pattern 2 include certain loan-words, of both Arabic and Kurdish origin. The feminine inflection -ə is borrowed from vernacular Arabic, and is identical to the (native) plural inflection. As in other dialects, such as Alqosh,34 some adjectives which take this pattern belong to the lexical field of disabilities; e.g. ṭaršå (m), ṭaršə (f), ṭaršə (pl) ‘deaf’ (< Arab.). They also all have a stem of the form CaCC-. Other attested adjectives taking this inflection are: randå ‘fine’ (< K.), xarså ‘dumb’ (< Arab.), baṛšå ‘albino’, ʿarjå ‘lame’ (< Arab.), zarqå ‘blue’ (< Arab.), sahlå ‘easy’ (< Arab.) and ṣaʿbå ‘difficult’ (< Arab.). Inflectional pattern 3 is a mixed inflection, where the feminine is doubly marked in a combination of -Tå (the native inflection of pattern 1) and -ə (the borrowed Arabic form of pattern 2), resulting in -Tə. Adjectives taking this inflection are all of Aramaic origin. This inflection is, to the author’s knowledge, not yet attested in other dialects; in Alqosh, for instance, the same words take inflectional pattern 2. What these adjectives have in common is unusual or unique consonant-vowel patterns: none of the common adjectival patterns occur in this group.35 The attested members of this group, next to a representation of their consonant-vowel patterns, are: 34 Coghill, 2004, pp. 282–283. Note that it is the patterns that are unusual, in that few adjectives appear in them. The adjectives themselves are common. 35 249 rābå (m), rabθə (f), rābə (pl) ‘big’ ṭāwå (m), ṭotə (f), ṭāwə (pl) ‘good’ xāθå (m), xaθtə (f), xāθə (pl) ‘new’ CoCå zorå (m), zurtə (f), zorə (pl) ‘small’ komå (m), kumtə (f), komə (pl) ‘black’ CCoCå smoqå (m), smuqtə (f), smoqə (pl) ‘red’ CCāCå xwārå (m), xwartə (f), xwārə (pl) ‘white’ CāCå Inflectional pattern 4 consists of no inflection at all. Adjectives following this pattern are probably recent borrowings from Arabic, which have not been adapted to Aramaic morphology or phonology, e.g. ð̣aʿíf ‘weak, thin’ (< Arab.), ð̣aʿíf (f), ð̣aʿíf (pl). Other examples of unadapted uninflected adjectives are: lā́-ṣaḥ ‘ill’ (< Arab.), ʾarzan ‘cheap’ (< K.), ʾagran ‘expensive’ (< K.), rəṣāṣi ‘grey’ (< Arab.), qahwāyi ~ qahwāʾi ‘brown’ (< Arab.), qə́rməzi ‘purple’ (< Arab.), ʾaṣlaʿ ‘bald’ (< Arab.). The lack of agreement is illustrated by the following examples: šuqtå qə́rməzi ‘a purple shirt (f)’, ʾāni ð̣aʿíf ‘the weak ones’. The loan-word xoš ‘good’ is also invariable, but is different to the other words here in that it precedes the noun: xoš꞊ʾixālå ‘good food’. 3.4 Annexation constructions A genitive relationship between two (or more) nouns is usually expressed by means of the head-marking (construct) suffix -əd, e.g. yoməd=daʿwå ‘the day of the wedding’ (cf. yomå ‘day’). Two irregular forms are bərt ‘son of’ (cf. bronå ‘son’) and bərtəd ‘daughter of’ (cf. brātå ‘daughter’). The older dependent (genitive) marker d- is also found, especially when the possessor is predicated, e.g. wāwå d-gūrə [PST.COP.3PL GEN-men] ‘they were the men’s’. The /d/ consonant of both morphemes undergoes anticipatory assimilation (see section 2.3.1), e.g. nāšəz=zāxu ‘the people of Zakho’ (cf. nāšə ‘people’) and ʾāyi betå, k-oyå t-kepå [this egg IND-be.3FS GEN-stone] ‘this egg, it is of stone’. The old Aramaic apocopate construct is preserved in the following productive prefixes: bi- ‘house of’, mar- ‘owner of’, e.g. bi-kālu ‘the family of the bride’ and mar-beθå ‘house-owner’. Measurements of quantity are usually simply placed in juxtaposition with the noun, e.g. ġða꞊maṭamiθå məšxå ‘a spoonful of oil’, tətté꞊tanayāθå sūraθ ‘two words of Surath’. 3.5 Numerals and the indefinite article The numerals 1–10 are given in table 5. These numerals, and only these numerals, inflect for gender to agree with the noun modified. Before a noun the 250 stress is usually shifted onto the final syllable and any /å/ replaced by /a/, e.g. ʾarbé꞊ʾənšə ‘four women’ and ʾarbá꞊gūrə ‘four men’ (see section 2.1.2). Sometimes the stressed vowel is lengthened, e.g. ʾəšwā́꞊ʾənšə ‘seven women’. The forms for ‘one’ undergo shortening when used attributively: xa꞊ (m) and ġða꞊ (f). Table 5. Independent numerals (1–10) m f one two three four five xāʾ ġðāʾ treʾ tətte ṭlāθå ṭəllaθ ʾarbå ʾarbe xamšå ʾəštå xamməš ʾəššət six seven eight šoʾå ʾəšwå tmanyå təšʾå tmāne təššå nine ten ʾəsṛå ʾəssar The indefinite specific article (expressing ‘a certain’) is identical to the attributive numeral ‘one’ and thus also inflects for gender: xa꞊ (m) and ġða꞊ (f). The numerals 11–19 are: xadesar, tresar, təltāsar, ʾarbāsar, xamšāsar, ʾəštāsar, šoʾāsar, tmanyāsar, čāsar. The multiples of ten are: ʾəsri ‘twenty’, ṭlāθi, ʾarbi, xamši, ʾəšti, šoʾi ~ šuʾi, tmāni, təšʾi. ‘Hundred’ is ʾuṃṃå and ‘thousand’ is ʾalpå. Combinations of tens and units are ordered with the unit first; note that this order varies across NENA dialects. Stress is placed on the final syllable of the unit: ʾarbá-w꞊əsri ‘twenty-four’. Cardinal numerals, as in other NENA dialects, are expressed by annexation constructions (see section 3.4), but also with gender agreement, e.g. gorå dətreʾ ‘second man’, baxtå t-tətte ‘second woman’. 3.6 Verbs 3.6.1 Derivational patterns and verbal bases There are five main verb derivation patterns (binyānīm): four triradical and one quadriradical. Derivations I, II and III are derived from earlier Aramaic pəʿal, paʿʿel and aphʿel derivations respectively. Derivation II2 is a variant of II found with roots where the last two radicals are the same (e.g. √xll): in this derivation the original gemination of paʿʿel is preserved. The bases used in the verbal system are formed according to the derivation (see table 6). They are: the Present Base, Past Base, Imperative, Infinitive, Resultative Participle, and Active Participle. Like some other NENA dialects (including Alqosh and Qaraqosh), Telkepe has acquired new derivations, borrowed from Arabic, in particular the Ct- derivation (with infixed -t- after the first radical), borrowed from the Arabic eighth derivation, and the St- derivation (with prefixed st-), borrowed from the Arabic tenth derivation.36 Their existence as independent derivations is 36 In personal correspondence, David Enochs reports of a further borrowed derivation used by Telkepe speakers living in America, namely the Arabic fifth derivation, loaned along with the Arabic verb mny v ‘to wish’, where the t- prefix is also transferred into the Telkepe forms. The precise paradigm still needs to be confirmed, however, so it has not been listed here. 251 Table 6. Verbal bases Present Base Present Base 3msg Past Base Imperative Infinitive Res. Ptcp. m Res. Ptcp. f Act. Ptcp. I II II2 III Q qṭl ‘to kill’ qaṭlqāṭəl qṭəlqṭol qṭālå qṭilå qṭəltå qaṭālå bšl ‘to cook’ mbašlmbāšəl mbušəl(m)bāšəl (m)bašolə mbušlå mbušaltå mbašlānå xll ‘to wash’ mxallmxalləl mxulləl(m)xalləl (m)xallolə mxullå mxullaltå ? šlx ‘to rob’ mašəlxmašləx mušləxmašləx mašloxə mušəlxå mušlaxtå mašəlxānå šxlp ‘to change’ mšaxəlpmšaxləp mšuxləp(m)šaxləp (m)šaxlopə mšuxəlpå mšuxlaptå mšaxəlpānå undermined somewhat by the fact that they are only found with borrowed Arabic verbs. Nevertheless, like the other derivations they have their own paradigms, even if these show some variation, as shown in table 7.37 Table 7. Arabic verbal bases Present Base Present Base 3msg Past Base Imperative Infinitive Res. Ptcp. m Res. Ptcp. f Act. Ptcp. Ct- St- ḥrm Ct- ‘to respect’, ḥfl Ct- ‘to celebrate’, xlf Ct- ‘to differ’ maḥtarmməḥtaflmaxtəlfmaḥtarəm muḥtərəmməḥtəfəl? maḥtaromə muḥtərmå muḥtaramtå ? ʿml St- ‘to use’ məstaʿaml- ~ məstaʿməl- ~ məstaʿəml? mustəʿməl? məstaʿmolə mustəʿəmlå mustaʿmaltå ? The Present, Past and Imperative bases are inflected for person and used as verb forms themselves. The main person indexes are the S- and L-suffixes (see table 8). The Infinitive and the Resultative and Active Participles, as nominal/adjectival forms, require auxiliary verbs such as the copula to lend them verbal force. 37 More detail on Arabic loan derivations in Telkepe and other dialects can be found in Coghill, 2015. 252 Table 8. Verb inflection paradigms 3 msg fsg pl 2 msg fsg pl 1 msg fsg pl S-suffixes Present Base with S-suffixes L-suffixes Past Base with Past Base with L-suffixes S-suffixes — -å -i -ət -at -ū́ tu -ən -an -ux šāqəl šaqlå šaqli šaqlət šaqlat šaqlūtu šaqlən šaqlan šaqlux -lə -lå -la -lux -lax -loxu -li -li -lan šqəllə šqəllå šqəlla šqəllux šqəllax šqə́lloxu šqəlli šqəlli šqəllan šqil šqilå šqili ? ? ? šqilən šqilan šqilux The Present Base takes S-suffixes to index the subject and may take L-suffixes to index an object: k-šaql-ux IND-take-S.1PL ‘we take’ k-šaql-ux-la IND-take-S.1PL-L.3PL ‘we take them’ The Past Base takes L-suffixes to index the subject and may take S-suffixes to index a feminine or plural third person pronominal object: šqəl-lan take.PAST-L.1PL ‘we took’ šqil-i-lan take.PAST-S.3PL-L.1PL ‘we took them’ The Past Base is also used in a passive construction, where it takes S-suffixes to index the subject (and no L-suffixes). This can be elicited from certain older speakers, but has not been documented in spontaneous speech.38 It expresses a passive: the examples offered by speakers all have present perfect aspect, but it is not known whether it is restricted to this function: (1) sayārətt-i mzubn-å car(f)-1SG sell.PAST-S.3FS ‘My car has been sold.’ As in Alqosh, L-suffixes undergo regressive assimilation to a previous consonant: to the final /n/ of a root and to a final rhotic (/r/ or /ṛ/): *zwən + li > zwənni *gwər + lə > *gwərrə > gwerə ‘I bought’ ‘he married’ 38 The Past Base with S-suffixes is used to express a passive in some other NENA dialects; see Coghill, 2016, pp. 268–269. It seems, however, to be undergoing a general decline in favour of analytical passive constructions: in the closely related dialect of Alqosh it is not productive but survives only in fixed idioms and proverbs; see Coghill 2004, pp. 191–192. 253 They also assimilate to the final consonants of S-suffixes, i.e. /n/ and /t/: *k-šaqlən + lux > k-šaqlənnux ‘I (m) take you (msg)’ *k-šaqlat + li > k-šaqlatti ‘you (fsg) take me’ Note that /t/ as part of a root does not trigger assimilation: fətlə ‘it passed’ (fyt I). There is another set of suffixes, B-suffixes, which are found predominantly attached to the existential particle ʾiθ in a form which expresses ‘to be able’ (see section 3.6.5). These have the same form as L-suffixes, except with the /l/ replaced by /b/, e.g. -bə (3msg) and -bå (3fsg). The main verb forms of NENA, Telkepe included, originate in Late Aramaic participles. The Present Base derives from the Late Eastern Aramaic active participle and the Past Base from the passive participle. The S- and Lsuffixes have quite different historical origins. The S-suffixes originate in gender and number inflection of the participles which merged with enclitic first and second pronouns. The L-suffixes originate in the Late Aramaic dative preposition l- with pronominal suffixes attached. This preposition flagged direct as well as indirect objects of the active participle construction. With the passive participle it flagged firstly experiencers (with verbs such as ‘to hear’), then was extended to all agents.39 The B-suffixes have a similar origin, except that they were formed on the locative/instrumental preposition b-. 3.6.2 Tense-aspect-mood categories and verbal modifiers The Past Base inflected with L-suffixes expresses the past perfective: this includes present perfect aspect, e.g. šqəl-li ‘I took’, ‘I have taken’. The inflected Present Base may occur without a prefix as the present subjunctive, in which case it expresses deontic modality, or forms part of a verbal complement. Other tense-aspect-mood (TAM) values are expressed by means of prefixes on the Present Base or an auxiliary (pseudo-)verb with or without the complementiser d=, as shown in table 9. As in other NENA dialects, the past perfective prefix kəm- always co-occurs with object suffixes: kəm-Present Base normally serves in place of Past Base forms, when an object needs to be indexed, as only 3fsg and 3pl objects may be indexed on the Past Base. The prefixes k-, b- and šud- follow the normal rules or tendencies of assimilation (see section 2.3.1), as in the following examples: k- + bāxə > gbāxə ‘he weeps’ b- + pāyəš > ppāyəš ‘he will be’ b- + maθyāli > mmaθyāli ‘she will bring to me’ 39 See Coghill, 2016 for a description of the development of the NENA verbal system and accompanying alignment change in the language. 254 Table 9. TAM modifiers of Present Base forms Modifier Main function In combination Translation ØØkb- (~ bəd-) šud= kəmzišwoq/šoq d= xoš d= lāzəm/garag zil-S/zi-L jussive complement indicative future jussive past perfective prospective jussive cohortative necessitive prospective yalpå kəbå d=yalpå k-yalpå b-yalpå ~ bəd-yalpå šud=yalpå kəm-yalpā-lə zi-yalpå šoq d=yalpå xoš d=yalpux lāzəm yalpå zilå yalpå ‘let her learn’ ‘she wants to learn’ ‘she learns’ ‘she will learn’ ‘let her learn’ ‘she learned it’ ‘she is going to learn’ ‘let her learn’ ‘let us learn’ ‘she must learn’ ‘she is going to learn’ Prefixes also follow the rules of syllable structure, disallowing CCC, so that when the addition of an affix causes a consonant cluster, an epenthetic vowel, ə, is usually inserted to break it up: k- + mbāšəl > kəmbāšəl ‘he cooks’ When kəm- or b- (> m-) is prefixed to a stem beginning with mC, one /m/ is elided. This can cause ambiguity: kəm- + mbāšəllå kə- + mbāšəllå b- + mbāšəl Ø- + mbāšəl > > > > kəmbāšəllå kəmbāšəllå mbāšəl mbāšəl ‘he cooked it (f)’ ‘he cooks it (f)’ ‘he will cook’ ‘he may cook’ Another common feature is the loss of /ʾ/ after a prefix ending in a consonant, e.g. bd-āwəð ‘he will make’ (< *bəd-ʾāwəð). It is not always consistent, e.g. kəm-amrannax ~ kəm-ʾamrannax ‘I (f) said to you (f)’. Verbs formed on the Present and Past Bases may take an affix -wå (-wā-) directly after the base, or after the S-suffix, if there is one, but before any L-suffix. This shifts the time reference (further) into the past: present subjunctive darsən ‘I (m) may study’, past subjunctive darsənwå ‘I (m) might study’; present indicative k-āθa ‘they come’, past habitual k-āθāwå ‘they used to come’; past perfective məθlə ‘he died, he has died’, remote past perfective mə́θwālə ‘he had died’. In Telkepe the past habitual usually takes the indicative prefix, unlike in Alqosh,40 but it sometimes occurs without, in which case it is indistinguishable from the past subjunctive, e.g. nablíwāla [nabl-i-wā-la] l-ḥarub [take-S.3PL-ANT-L.3PL to-war] ‘they used to take them to war’. 40 Coghill, 2004, p. 139. 255 The Imperative is inflected for singular (-Ø) and plural (-u), e.g. pθox ‘open (sg)!’, pθūx-u ‘open (pl)!’ (cf. Alq. pθox, pəθx-u). Verba tertiae /y/ in all derivations also distinguish between masculine and feminine singular, as does the irregular verb ʾzl I ‘to go’. The Imperative takes initial stress, as in many other NENA dialects, e.g. mášəlx-u ‘rob (pl)!’. As in Alqosh, the Imperative is sometimes combined with a particle di- ~ də-, adding some kind of emphasis, e.g. di-pθox šubbak! ‘Come on, open a window!’. A similar particle (in form and function) is found in Kurmanji and Qəltu Arabic.41 Verbs are negated by the preposed negator particle la꞊, which takes stress. For negated imperatives there is a suppletive construction, namely the inflected Present Base with no further prefixes, e.g. la꞊dārət [not꞊put:S.2MS] qeså b-nuqbəd dəbborə ‘Don’t put a stick in a hornet’s nest!’. The auxiliary verb pyš I can be used in various tenses, aspects and moods (more in its sense ‘to become’ than ‘to be’) to express a dynamic passive, e.g. malkå lāzəm pāyəš qṭilå [king necessary become.S.3MS kill.RES.PTCP.MS] ‘The king must be killed’. 3.6.3 Weak verbs The following are some of the less predictable weak classes of verbs. Verba primae /ʾ/ fall into two groups. In type 1, the /ʾ/ is not necessarily elided and the verbs conjugate as strong verbs, e.g. k-ʾārəq ~ k-ārəq ‘he runs (ʾrq I ‘to run’). Type 2, which is weak, includes ʾxl I ‘to eat’, ʾmr I ‘to say’, ʾsq I ‘to climb’, ʾṣṛ I ‘to tie’, ʾθy I ‘to come’, ʾwð I ‘to make, do’, ʾwr I ‘to enter’ and ʾtw I ‘to sit’, as well as the irregular verb ʾzl I ‘to go’.42 When these verbs are used with the indicative prefix k-, the /ʾ/ is always elided. There is, however, no change of vowel: kāxəl ‘he eats’ (cf. Alq. kixəl). In Past Base forms and the Resultative Participle, the first radical is elided: xəl-li ‘I ate’, xilå ‘eaten (m)’. For the Imperative we find ʾixul (sg), ʾəxlu (pl) ‘eat!’, ʾimor (sg), ʾəmru (pl) ‘say!’ (cf. Alq. mor, muru). Infinitives begin with (ʾ)i: ʾixālå ‘to eat’, ʾimārå ‘to say’.43 Verba tertiae /y/ behave much like in other NENA dialects, for instance with a msg/fsg distinction in the Imperative: k-xāzə ‘he sees’, k-xazyå ‘she sees’, k-xāzotu ‘you (pl) see’, xzelə ‘he saw’, Resultative Participle xəzyå (msg), xziθå (fsg), xəzyə (pl) ‘seen’, Imperative xzi (msg), xze (fsg), xzo (pl) ‘see!’, Infinitive xzāyå ‘to see’. 41 Jastrow, 1978, pp. 310–311. Membership of this class varies somewhat from that of Alqosh, where ʾwr I is type 1, and some other verbs that are type 2 in Telkepe are primae /y/ in Alqosh. See Coghill, 2004, pp. 143, 146. 43 The initial glottal stop is elided after the preposition b-, as in the progressive construction, e.g. ʾilə b-ixālå ‘he is eating’. 42 256 3.6.4 Irregular verbs The irregular verb ʾzl I ‘to go’ has a suppletive Present Base stem zā́- inflected with L-suffixes, e.g. zālə ‘he may go’, zā́loxu ‘you (pl) may go’. This is used with all Present Base TAM modifiers (unlike in Alqosh where the indicative has a different stem), e.g. b-zālə ‘he will go’ and šud=zālə ‘let him go’. It also takes the anterior suffix, e.g. zá-wā-li ‘I used to go’. After indicative k- a shwa is inserted, often followed by gemination: kə-zālə ~ kə-zzālə ‘he goes’. There is a three-way distinction in the Imperative: si (msg), se (fsg) and so (pl) ‘go!’. This verb also has a special form based on the Past Base (zil-/zi-) inflected with a mixture of S- and L-suffixes. It may be used as an independent verb with immediate future reference, e.g. zilə l-šūqå ‘He’s about to go to the shops’, or as an auxiliary marking prospective aspect, e.g. zilə zālə šl-šūqå [PRSP:3MS go:L.3MS to-market] ‘He’s going to go to the shops’. In the latter sense it may also occur as a particle, eroded to zi- ~ si-, e.g. zi-zālə l-šūqå [PRSP-go:L.3MS to-market] ‘He’s going to go to the shops’.44 Other irregular verbs, with some examples, are the following: ʾθy I ‘to come’ has Present Base ʾāθə ‘he may come’, ʾaθyå ‘she may come’, k-āθə ‘he comes’, k-aθyå ‘she comes’, bd-āθə ‘he will come’, št-aθyå ‘let her come’. The Past Base is θe-, e.g. θeli ‘I came’. There is a suppletive Imperative hayyu ~ hay (sg), hayyo (pl) ‘come!’, and the Infinitive is ʾiθāyå ‘to come’. bʾy I ‘to want’ behaves as a regular tertiae /y/ verb, with /ʾ/ unelided, except for the Present Base with k-, which has the irregular stem kəb-; contrast baʾyå ‘she may want’ with kəbå ‘she wants’. hwy I ‘to be’ is a regular verb of the verba tertiae /y/, apart from the lack of a Past Base form (except in the meaning of ‘to be born’) and the changes that prefixes make to the Present Base forms: hāwə ‘he may be’, k-āwə ‘he is (generally)’, pt-āwə ‘he will be’, t-āwə ‘that he may be’. yðʾ I ‘to know’ has an irregular Present Base stem with k-, namely kəð- ~ keð-, e.g. yaðux ‘we may know’, kəðux ~ keðux ‘we know’. The final radical /ʾ/ is elided, or in some cases treated like /y/: yað-i ~ yað-a ‘they may know’. ywl I ‘to give’ has an irregular Present Base stem: yāwəl ‘he may give’, yāw-i ‘they may give’. After the kəm- prefix, this is sometimes altered to -ewəl-/-ew-, e.g. kəmm-ewəl-lå ‘he gave to her’. The 44 See Coghill, 2010b and Coghill, 2012 for the forms, functions and development of this form in the Mosul Plain dialects. 257 /y/ is elided in Past Base forms and the Resultative Participle: wəlli ‘I gave’, wilå ‘given (m)’. The Imperative is irregular: hal (sg), hallu (pl) ‘give!’. 3.6.5 Copulas and other pseudo-verbs Telkepe has a Present Copula and a Past Copula, both available in independent form (occurring before the predicate) and enclitic form. Both may also be negated, in which case the copula stands before the predicate: ʾilå ʾāxå, ʾāxa꞊lå wāwå ʾāxå, ʾāxå꞊wāwå lelå ʾāxå la꞊wāwå ʾāxå ‘she is here’ ‘she was here’ ‘she is not here’ ‘she was not here’ These copulas are ‘pseudo-verbs’, that is, they take special inflection unlike normal verbs. Other TAM values are expressed with hwy I ‘to be’ or pyš I ‘to become, be’, e.g. purṭenå, k-āwə smoqå [flea(m) INF-be.3MS red.MS] ‘The flea, it’s (generally) red’, hāwotun brixə [be:2PL blessed.PL] ‘May you (pl) be blessed’, hwi/poš ṭāwå [be.IMP.MS/be.IMP.SG good.MS] ‘Be (msg/sg) good!’. The copula paradigms are presented in table 10. Table 10. Copulas Present Present independent enclitic 3 msg ʾilə fsg pl 2 msg fsg pl ʾilå ʾila ʾiwət ~ ʾit ʾiwat ~ ʾit ʾiwotu ~ ʾitu 1 msg ʾiwən ~ ʾin fsg ʾiwan ~ ʾin pl ʾiwux ~ ʾix Negative Present Past Past independent enclitic Negative Past ꞊ilə lelə lelå lela lewət ~ let lewat ~ let léwotu ~ letu lewən ~ len (ʾi)wənwå lewan ~ len (ʾi)wanwå lewux ~ lex (ʾi)wuxwå ꞊wewå ~ ꞊wāwə ꞊wāwå ꞊wāwå ꞊wətwå ꞊watwå ꞊wútuwå la꞊wewå ꞊ilå ꞊ila ꞊iwət ~ ꞊it ꞊iwat ~ ꞊it ꞊iwotu ~ ꞊itu ꞊iwən ~ ꞊in ꞊iwan ~ ꞊in ꞊iwux ~ ꞊ix (ʾi)wewå ~ (ʾi)wāwə (ʾi)wāwå (ʾi)wāwå (ʾi)wətwå (ʾi)watwå (ʾi)wútuwå ꞊wənwå ꞊wanwå ꞊wuxwå la꞊wənwå la꞊wanwå la꞊wuxwå la꞊wāwå la꞊wāwå la꞊wətwå la꞊watwå la꞊wotuwå The /i/ of the Present Copula merges with a final vowel of the predicate: dəx꞊ilə ‘how is he?’, ʾāxå + ꞊ilə > ʾāxa꞊lə ‘he is here’, gārə + ꞊ilə > gāre꞊lə ‘it is a roof’. Telkepe is relatively unusual among NENA dialects in using ʾilə as an unbound copula preceding the predicate as well as in clitic form.45 In many other dialects it only occurs as an enclitic, and there is a separate deictic copula The ʾilə copula may still occur in unbound form, taking its own stress, in the Christian dialect of Barwar, typically between the subject and predicate; see Khan, 2008a, pp. 181, 622, 625– 628. Deictic functions are, however, expressed by the deictic copula hole. 45 258 which covers some of the functions of Telkepe ʾilə, for instance expressing the present progressive in combination with the infinitive. Further north this is usually holə or a variant thereof (ʾolə in Tisqopa, wolə in Alqosh), while in the eastern Mosul Plain one finds kilə.46 Compare the Telkepe present progressive expression ʾiwan bə-syāqå [PRS.COP.1FS in-drive.INF] ‘I am driving’ with Alqosh wo-la kās-i bə-mrāʾa [DEIC.COP-3FS stomach-1SG in-hurt.INF] ‘My stomach is hurting.’ Presumably unbound ʾilə existed in the common ancestors of the dialects, but a cliticised form arose and the unbound variant eventually disappeared in most. The distinct deictic copulas, holə and kilə, would then be innovative forms that were never adopted in Telkepe. The first probably derives from a deictic element plus -ilə; the second from the indicative present prefix k- plus -ilə. The purely deictic functions of these copulas may be expressed in Telkepe by combinations of the demonstratives ʾāyi ‘this’ and ʾāwå ‘that (msg)’ with the enclitic copula, e.g. ʾāyi꞊wan ‘Here I (f) am!’ and ʾāwa꞊lə ‘There he is!’. The copulas and verbs ‘to be’ (hwy I, pyš I) are used in a variety of analytic verb forms. For example, they may be combined with the Resultative Participle to express perfect or stative aspect: (2) ʾilə ʾəθyå ta maxrowə. PRS.COP.3MS come.RES.PTCP.MS for destroy.INF ‘He has come to destroy.’ (3) wewå dmixå. PST.COP.3MS sleep.RES.PTCP.MS ‘He was asleep.’ (4) baġdad lewan xziθå. Baghdad NEG.PRS.COP.1FS see.RES.PTCP.FS ‘Baghdad, I haven’t seen.’ Such constructions may also express passive voice, in which case the preposition l- ‘to’ may mark the agent: (5) ʾilə xilå. eat.RES.PTCP.MS ‘It has been eaten.’ or ‘He has eaten.’ PRS.COP.3MS (6) ʾilə mulpå l-polus. teach.RES.PTCP.MS to-Paul ‘He has been taught by Paul.’ PRS.COP.3MS 46 This Qaraqosh form is from Khan, 2002, p. 128; the same form is also found in Karimlesh (Roberta Borghero, personal communication) and Bariṭle (Kristine Mole, personal communication). 259 The copulas or verbs ‘to be’ may also be combined with the Active Participle, in which case they express a kind of scheduled future: (7) bd-aθy-at ṣaprå? – laʾ, ʾiwan palaṭṭå. FUT-come-2FS tomorrow no, PRS.COP.1FS go_out.ACT.PTCP.FS ‘Will you come tomorrow? – No, I’m going out.’ With the Infinitive prefixed by b- ‘in’, they express a present progressive: (8) ʾiwan b-ixālå PRS.COP.1FS in-eat.INF ‘I’m eating.’ The deictic copulas may be combined with the inflected Past Base to emphasize the here-and-now: (9) ʾāyi꞊wat mṭe-lax! this꞊PRS.COP.2FS arrive.PAST-L.2FS ‘Here you are, arrived!’, i.e. ‘You’re already here!’ (10) ʾā́wa꞊lə θe-lə! that꞊PRS.COP.3MS come.PAST-L.3MS ‘There he is, just come!’ Other pseudo-verbs are formed from the existential particle ʾiθ ~ ʾiθən ‘there is/are’ and its negated equivalent leθ ~ leθən ‘there is/are not’. The corresponding past forms are ʾəθwå ‘there was/were’ and laθwå ‘there was/were not’. With L-suffixes, these express possessive predication, that is, ‘to have’. As in Alqosh, the sequence *tl is realised as /tt/. Some examples are: ʾəttə [EXIST:L.3MS] ‘he has’, lattux [NEG.EXIST:L.2MS] ‘you (m) don’t have’, ʾə́θ-wā-lan [EXIST-ANT-L.1PL] ‘we had’. With B-suffixes (see section 3.6.1), the existential particle expresses ability or location. In this form the /θ/ is elided before the /b/. Some examples are: ʾibə [EXIST:B.3MS] ‘he can’, ʾə́θ-wā-bə [EXIST-ANT-B.3MS] ‘he couldn’t’, le-ba t=palṭ-i [NEG.EXIST-B.3PL COMP=get_out-S.3PL] ‘they can’t get out’, le-bə taṭawwur [NEG.EXIST-B.3MS development] ‘there’s been no development in it’. Both L- and B-suffixes can also be combined with the 3msg Present Base form of hwy I ‘to be’ to express other TAM values, e.g. d=la꞊hāwe-bə də=mḥārək [COMP=not-be.S.3MS-B.3MS COMP=move.S.3MS] ‘so that he would not be able to move’. 260 L-suffixes are also combined with various 3msg Past Base verbs, expressing (dis-)possession/affectedness, e.g. θə́le-lan nāšə [come.PAST:L.3MS-L.1PL people] ‘people have come to us’ (i.e. ‘we have guests’).47 Some other pseudo-verbs are the following: bass- ‘it’s enough for’ is inflected with the possessive suffixes, e.g. bassa! ‘It’s enough for her!’. baʿd- ‘to be still X’ is inflected with the possessive suffixes, e.g. baʿde tāmå ‘He is still there’. xəšt- ‘to resemble, to be like’ takes the same person inflection as the Land B-suffixes, e.g. xəšt-a ʾənglezāyə ‘They resemble English people’ and xəšt-å qaqwānå ‘She is like a partridge’ (i.e. she is beautiful). 4 Lexicon Presented in this section are the main members of some restricted lexical sets, as well as common words which are known to vary between dialects of NENA. 4.1 Prepositions Prepositions, as the name suggests, always precede the noun or noun phrase. They are formed in various ways, with some meanings being represented by two or more forms (e.g. l-, rešəd, rəš ‘on’ and m-, mən ‘from’). When they govern personal pronouns, prepositions take the possessive suffixes (see table 2), and they have special stems for this. In the lists below, the attachable stem is given, attached to the 3msg suffix, e.g. mənn-e ‘from him’. Some prepositions consist of only a single consonant in their basic form, and this must be attached to another word. This often assimilates to a following consonant, or takes an epenthetic vowel before a consonant cluster. lbm- ʾəll-e bgāw-e mənn-e ‘to’, ‘on’, ‘about’, agent of passive ‘in’, ‘into’, ‘at’, ‘on’, ‘with’ (instrumental) ‘from’ Other prepositions are independent words, though often unstressed. 47 See Coghill, 2016, pp. 210–211 and Coghill, forthcoming-a for more examples and discussion. 261 baθər ~ baθər mben ~ benaθ ta mən barqul ~ darqul wəl baθṛ-e benāθ-e ṭāl-e mənn-e b/darqul diy-e – ‘after’, ‘behind’ ‘between’ ‘to, for’ ‘from’ ‘opposite, against’ ‘until’ Other prepositions end in the construct suffix, -əd. The /d/ of the suffix usually assimilates to a following consonant, as described in section 2.3.1, e.g. bgāwəṣ=ṣomå ‘in Lent’. Some of these prepositions are derived from nouns, e.g. p-palgəd ‘in the middle of’ < b- ‘in’ + palgå ‘half’ + -əd. Others, such as ʾəmm-əd ‘with’ (originally ʾəm- ‘with’), have presumably acquired the ending -əd by analogy. xwāθəd ʾəmməd (l-)xoθəd dormadāṛəd xawəð̣rānəd p-palgəd ltexəd ~ ltex m- xwāθ-e ʾəmm-e (l-)xoθ-e dormadāṛ-e xawəð̣rān-e p-palg-e ltex mənn-e ‘like’ ‘with’ ‘under’ ‘around’ ‘around’ ‘in the middle of’ ‘below’ Some prepositions have forms both with and without the -əd suffix. go ~ (b-)gāwəd geb ~ gebəd l-rəš ~ l-rešəd qam ~ qāməd (b)gāw-e geb-e reš-e qām-e ‘in’ ‘beside’, ‘at the house of’ (French chez) ‘on’ ‘before’ A different type of preposition is the particle dla꞊ ‘without’, formed from the genitive marker d- and the negator la꞊. When a demonstrative pronoun or deictic adverb (e.g. ʾāxå ‘here’) beginning in /ʾ/ follows a preposition, the genitive marker d- is sometimes inserted between the two and the /ʾ/ is elided; e.g. l-d-aθ꞊beθå [to-GEN-this.SG꞊house] ‘to this house’, mən d-o꞊gūda xənnå [from GEN-that.MS꞊wall(m) other.MS] ‘from the other wall’, mən d-āni [from GEN-these] ‘from these’, wəl d-āyå ʾetå [up_to GEN-that.FS church(f)] ‘up to that church’, mən d-āxå [from GEN-here] ‘from here’.48 48 See Gutman, 2016, pp. 282–289 for an analysis of genitive d- and its development in NENA. 262 4.2 Interrogatives man ~ māni ~ mani ‘who?’, mahå ~ mā ‘what?’, ʾemå ‘which’, ʾekå ‘where, whither?’, ʾiman ‘when?’, dəx ‘how?’, ʾukmå ~ ʾəkmå ‘how many?’, māqå ‘how much?’, qāyi ~ qay ‘why?’, ta-mahå ‘how come?, why?’ 4.3 Conjunctions u ‘and’, lo ‘or’, ʾaw ‘or’, yā ‘or’, fa ‘so, for, you see’, bas ‘but’, lākən ‘but’, d= ‘which, who; that’ (relativiser, complementiser), kud ‘when’, ʾən ‘if’, wəl ‘until’, tad= (<ta-d= ‘for’ + complementiser) ‘so that’, ‘in order to’. 4.4 Miscellaneous hādax ‘thus’, ham ‘also’, har ‘just, exactly; always, constantly’, bas ‘only, just’, baʿad ‘still’, lappəš ‘no longer’, ʾégahå ~ ʾega ‘then, at that time’, ʾāxå ‘here’, tāmå ‘there’, lʾel ‘above’, ltex ‘below’, təmmal ‘yesterday’, ṣapra ‘tomorrow’, ʾomå xənnå ‘the day before yesterday’, mxuškå ‘in the morning’, kabirå ‘much, a lot, very’, kabirə ‘many’, xaṣṣå ~ xa꞊qəṣṣå ‘a little’, qəṣṣa ‘little, few, not often’, tərwaθ- ‘the two of, both of’, nxθ I ‘to go down’, ʾsq I ‘to go up’, pyš I ‘become’. 5 Syntax Syntax will be covered in a monograph to be published on the Telkepe dialect, but some syntactic features have already been discussed in various papers, in particular ditransitive constructions, differential object marking and grammatical relations.49 6 ‘Weddings’ (glossed text) The following text was recorded by the author in Detroit in 2004 with an elderly lady who grew up in Telkepe. Note that SMALL CAPS indicates the nuclear stress in the intonation phrase, while | marks the intonation phrase boundary. 1. kud GGORIWÅ,| nāšə P-QAMEΘÅ,| kud k-gor-i-wå nāš-ə b-qameθå when IND-marry-S.3PL-ANT person-PL in-before ‘When they used to marry, people, formerly,’ 49 See Coghill, 2010a for a presentation of ditransitive constructions; Coghill, 2014 for differential object marking; and Coghill, forthcoming-a on grammatical relations. Coghill, 2016, pp. 12–13, 145–146, 210–211, 226, 236, 270, 285 also deals with some aspects of syntax in Telkepe. 263 2. kud꞊ġðā L-BEΘA kāθāwå kəmbarxiwālå.| kud꞊ġðā l-beθ-a k-āθā-wå kə-mbarx-i-wā-lå each꞊one.F to-house-3FS IND-come.S.3PL-ANT IND-bless-S.3PL-ANT-L.3FS ‘each (bride), they would come to her house and bless her.’ 3. ʾānå pəšli šātå ṬLƏBTÅ,| ʾānå pəš-li šātå I remain.PAST-L.1SG year ‘I remained engaged for a year,’ 4. u ʾiman d=ʾāθewå GEBAN,| u ʾiman d=ʾāθe-wå geb-an and when REL-come.S.3MS-ANT chez-1PL ‘and whenever he came around to ours,’ 5. la꞊maḥəkyanwå ʾəmme u ʾARQANWÅ.| la꞊maḥəky-an-wå ʾəmm-e u ʾarq-an-wå not꞊speak-S.1FS-ANT with-3MS and run-S.1FS-ANT ‘I didn’t speak with him, but I would run away.’ 6. záwāli GEBÁY.| la-ʾatwanwå MAḤƏKYAN꞊ƏM(ME).| zá-wā-li geb-áy la-ʾatw-an-wå maḥəky-an꞊əm(m-e)50 go-ANT-L.1SG chez-3PL not-sit-S.1FS-ANT speak-S.1FS꞊with-3MS ‘I used to go to them. I didn’t sit and talk with him.’ 7. dahå <?> ʾiwotu bəxzāyå mā꞊ʾiθ BƏBRĀYÅ.| dahå <?> ʾiwotu bə-xzāyå mā꞊ʾiθ now <?> PRS.COP.2PL in-see.INF what꞊EXIST ‘Now <?> you see what is happening.’ 8. yā ʾĀLAHA꞊lloxu.| kfahmūtu m꞊in BIMĀRÅ?| YAʿNI.| yā ʾālaha꞊ll-oxu k-fahm-ūtu m꞊in b-imārå? yaʿni O God꞊on-2PL IND-understand-S.2PL what꞊PRS.COP.1SG in-say.INF it.means ‘O God be upon you. You understand what I’m saying? So-so.’ ṭləb-tå| engage.RES.PTCP-FS bə-brāyå in-happen.INF [Interviewer: ‘How old were you when you got married?’] 9. ʿumri wewå … tmanesar ŠƏNNƏ.| ʿumr-i wewå tmanesar ŠƏNNƏ age-1SG PST.COP.3MS eighteen years ‘I was … eighteen years old.’ 10. liʾan bābi MƏ́ΘWĀLƏ,| liʾan bāb-i mə́θ-wā-lə because father-1SG die.PAST-ANT-L.3MS ‘Because, my father had died.’ 50 The speaker stops before finishing the word: (me) is a reconstruction of the end of the word. 264 11. wanwå ṭləbtå꞊w bābi MƏ́ΘWĀLƏ,| wanwå ṭləb-tå꞊w bāb-i mə́θ-wā-lə PST.COP.1FS engage.RES.PTCP-FS꞊and father-1SG die.PAST-ANT-L.3MS ‘When I was engaged, my father had already died.’ 12. pəšlan ʾARBÉ꞊šənnə.| pəš-lan ʾarbé꞊šənnə remain.PAST-L.1PL four.F꞊years(f) ‘We remained four years (thus)(?)’ 13. p-qameθå la꞊mbarxíwå ʾƏLLÅ …| qameθå꞊wāwå lə-TRESAR꞊šənnə,| b-qameθå la꞊mbarx-í-wå ʾəllå …| qameθå꞊wāwå lə-tresar꞊šənnə. in-before not꞊bless-S.3PL-ANT except before꞊PST.COP.3FS to-twelve꞊years ‘Before, they didn’t bless/marry you except … Formerly, it was at twelve years,’ 14. baθər mə-TRESAR꞊šənnə,| w-EngeitherEng ʾARBĀSAR.| baθər mə-tresar꞊šənnə w-either ʾarbāsar after from-twelve꞊years and-… fourteen ‘after twelve years or fourteen.’ 15. d-arbāsar. ʾānå ʿumri ʾarbāsar mətlə BĀBI.| d-arbāsar ʾānå ʿumr-i ʾarbāsar mət-lə bāb-i GEN-fourteen I age-1SG fourteen die.PAST-L.3MS father-1SG ‘The fourteenth (year). Myself, my age was fourteen when my father died.’ 16. pəšli ṬLƏBTÅ,| pəš-li become.PAST-L.1SG ‘I got engaged.’ 17. yaʿnə wanwå … xwāθəd=ʾARBE꞊šənnə,| yaʿnə wanwå xwāθəd=ʾarbe꞊šənnə it.means PST.COP.1FS like=four꞊years ‘I mean, I was … around four [sic] years’ 18. yaʿnə KƏBÉWĀLI.| yaʿnə k-əbé-wā-li it.means IND-want.S.3MS-ANT-L.1SG ‘I mean, he was in love with me.’ 19. la꞊muḥkeli ʾƏMME!| u LA꞊MUḤKELƏ ʾəmmi!| ʾE.| la꞊muḥke-li ʾəmm-e u la꞊muḥke-lə ʾəmm-i ʾe not꞊speak.PAST-L.3MS with-3MS and not꞊speak.PAST-L.3MS with-1SG yes ‘I didn’t speak with him! And he didn’t speak with me. Yes.’ ṭləb-tå engage.RES.PTCP-FS 265 20. HĀDAX꞊wuxwå 21. bə-ḥtišām꞊u laθwå ʾəθwå yaʿnə … bə-ḥtišām꞊u laθ-wå ʾəθ-wå yaʿnə in-decency꞊and NEG.EXIST-ANT EXIST-ANT it.means ‘With decency and there wasn’t – there was, I mean …’ 22. ʾəθwå ʾadab KABIRÅ geban.| ʾəθ-wå ʾadab kabirå geb-an EXIST-ANT manners much chez-1PL ‘There were good (lit. a lot of) manners among us.’ 23. kud θela kəmbarxilan bgāwəd=BEΘÅ.| kud θe-la kə-mbarx-i-lan bgāwəd=beθå when come.PAST-L.3PL IND-bless-S.3PL-L.1PL in=house ‘When they came, they blessed us in the home.’ 24. u qameθå ʾiman kālu D=GORĀWÅ,| u qameθå ʾiman kālu d=gor-ā-wå and formerly when bride COMP=marry-S.3FS-ANT ‘And formerly, whenever a bride got married,’ 25. kmarəkwíwālå L-SUSTÅ.| k-marəkw-í-wā-lå l-sustå IND-make_ride-S.3PL-ANT-L.3FS on-mare ‘they had her ride on a mare.’ 26. kmarəkwíwālå l-sustå꞊w k-marəkw-i-wā-lå l-sustå꞊w IND-make_ride-S.3PL-ANT-L.3FS on-mare꞊and ‘They made her ride on a mare and’ 27. gdārāwå xa꞊ʿaji ZORÅ qāma.| k-dārā-wå xa꞊ʿaji zorå IND-put.S.3PL-ANT a.M-child small.M ‘put an infant in front of her.’ 28. yaʿnə,| ṣaprå mmaθyå YĀLƏ.| ʾE.| nišan yaʿnə ṣaprå b-maθy-å yālə ʾe. sign it.means tomorrow FUT-bear-S.3FS children yes ‘A sign, you see. Tomorrow she will bear children. Yes.’ 266 yaʿnə.| hādax꞊wuxwå yaʿnə thus꞊PST.COP.1PL it.means ‘That’s what we were like, you see.’ NIŠAN qām-a before-3FS 29. u katwiwå bgāwəd=BƏGNŪNƏ,| b-āyå QURNIΘÅ,| u k-atw-i-wå bgāwəd=bəgnūnə b-āyå qurniθå and IND-sit-S.3PL-ANT in=bridal_chamber in-that.F corner ‘And they sat in a bridal chamber, in that corner,’ 30. koðíwālå xa꞊məndi <?> xwāθəd=BƏGNŪNƏ.| k-oð-í-wā-lå xa꞊məndi <?> xwāθəd=bəgnūnə IND-make-S.3PL-ANT-L.3FS a-thing <?> like=bridal_chamber ‘they made it/for her something <?> like a bridal chamber.’ 31. ʾað꞊bəgnūnə katwāwå šabθå kullå, kālu BGĀWA.| ʾað꞊bəgnūnə k-atw-ā-wå šabθå kull-å kālu bgāw-a this꞊bridal_chamber IND-sit-S.3FS-ANT week all-3FS bride in-3FS ‘This bridal chamber, she sat a whole week in it, the bride.’ 32. leθ MAḤKOYƏ,| u knaxpāwå d=AXLĀWÅ.| leθ maḥkoyə u k-naxp-ā-wå d=axl-ā-wå NEG.EXIST speak.INF and IND-be_shy-S.3FS-ANT COMP=eat-S.3FS-ANT ‘There was no speaking. And she was too shy to eat.’ 33. knaxpāwå ta-d=AXLĀWÅ.| k-naxp-ā-wå ta-d=axl-ā-wå IND-be_shy-S.3FS-ANT for-COMP=eat-S.3FS-ANT ‘She was too shy to eat.’ 34. ʾe … ʾiwewå yaʿnə zonānət=QAMEΘÅ,| ʾe … ʾiwewå yaʿnə zonān-ət=QAMEΘÅ yes PST.COP.3PL it.means times-CST=before ‘Yes … they were the old times.’ 35. baʿdén … duni KƏMBADLÅ,| baʿdén duni kə-mbadl-å later world IND-change-S.3FS ‘Later, the world changes.’ 36. w-ilå kamri MṬUWERÅ yaʿnə,| DAHÅ,| w-ilå k-amr-i mṭuwər-lå yaʿnə dahå and-PRS.COP.3FS IND-say-S.3PL develop.PAST-L.3FS it.means now ‘And they say it’s progressed, you see, now.’ 37. u ʾiwotu bəxzāyå mā=ʾIΘƏN.| ʾĀYI꞊LÅ.| u ʾiwotu bə-xzāyå mā=ʾIΘƏN ʾāyi꞊lå and PRS.COP.2PL in-see.INF what=EXIST this꞊PRS.COP.3FS ‘And you see what there is. That’s it.’ 267 References Borghero, R., 2008, “The verbal system of the Neo-Aramaic dialect of Karimlesh”, in G. Khan (ed.), Neo-Aramaic Dialect Studies, Piscataway, NJ: Gorgias, pp. 81– 89. Coghill, E., 2004, The Neo-Aramaic Dialect of Alqosh (unpubl. diss., University of Cambridge). ———, 2005, “The morphology and distribution of noun plurals in the Neo-Aramaic dialect of Alqosh”, in A. Mengozzi (ed.), Studi Afroasiatici. XI Incontro Italiano di Linguistica Camitosemitica, Milan: FrancoAngeli, pp. 337–348. ———, 2008, “Some notable features in North-Eastern Neo-Aramaic dialects of Iraq”, in G. Khan (ed.), Neo-Aramaic Dialect Studies, Piscataway, NJ: Gorgias, pp. 91–104. ———, 2009, “Four versions of a Neo-Aramaic children’s story”, ARAM Periodical 21, pp. 251–280. ———, 2010a, “Ditransitive constructions in the Neo-Aramaic dialect of Telkepe”, in A. Malchukov, M. Haspelmath and B. Comrie (eds), Studies in Ditransitive Constructions: A Comparative Handbook, Berlin: De Gruyter Mouton, pp. 221– 242. ———, 2010b, “The grammaticalization of prospective aspect in a group of NeoAramaic dialects”, Diachronica 27/3, pp. 359–410. ———, 2012, “Parallels in the grammaticalisation of Neo-Aramaic zil- and Arabic raḥ- and a possible contact scenario”, in D. Eades (ed.), Grammaticalization in Semitic, Oxford: OUP, pp. 127–144. ———, 2013, “The Neo-Aramaic dialect of Peshabur”, in R. Kuty, U. Seeger and S. Talay (eds), Nicht nur mit Engelszungen: Beiträge zur semitischen Dialektologie. Festschrift für Werner Arnold zum 60. Geburtstag, Wiesbaden: Harrassowitz, pp. 37–48. ———, 2014, “Differential object marking in Neo-Aramaic”, Linguistics 52/2, pp. 335–364. ———, 2015, “Borrowing of verbal derivational morphology between Semitic languages: the case of Arabic verb derivations in Neo-Aramaic”, in F. Gardani, P. Arkadiev and N. Amiridze (eds), Borrowed Morphology, Berlin: De Gruyter Mouton, pp. 83–107. ———, 2016, The Rise and Fall of Ergativity in Aramaic: Cycles of Alignment Change, Oxford: OUP. ———, forthcoming-a, “Grammatical relations in Neo-Aramaic”, in A. WitzlackMakarevich and B. Bickel (eds), Argument Selectors: New Perspective on Grammatical Relations, Amsterdam: John Benjamins. ———, forthcoming-b, “The Neo-Aramaic dialect of Alqosh”, in S. Fassberg and S. Hopkins (eds), Handbook of Neo-Aramaic. ———, forthcoming-c, “North-Eastern Neo-Aramaic and language contact”, in A. P. Grant (ed.), The Oxford Handbook of Language Contact, New York: OUP. Fiey, J.-M., 1965, Assyrie chrétienne, contribution á l’étude de l’histoire et de la géographie ecclésiastiques et monastiques du nord de l’Iraq, vol. I, Beirut: Impr. catholique. Fox, S. E., 1997, The Neo-Aramaic Dialect of Jilu, Wiesbaden: Harrassowitz. Guidi, I., 1883, “Beiträge zur Kenntniss des neu-aramäischen Fellîḥî-Dialektes”, Zeitschrift der Deutschen Morgenländischen Gesellschaft 37, pp. 293–318. Gutman, A., 2016, Attributive Constructions in North-Eastern Neo-Aramaic: Areal, Typological and Historical Perspectives (unpubl. diss., University of Konstanz). 268 Hoberman, R. D., 1988, “The history of the modern Aramaic pronouns and pronominal suffixes”, Journal of the American Oriental Society 108/4, pp. 557–575. Hobrack, S., 2000, Der neuaramäische Dialekt von Umra (Dereköyü): Laut- und Formenlehre–Texte–Glossar (unpubl. diss., Friedrich-Alexander-Universität Erlangen-Nürnberg). Jastrow, O., 1978. Die mesopotamisch-arabischen qǝltu-Dialekte, vol. I: Phonologie und Morphologie, Franz Steiner: Wiesbaden. ———, 1979, “Zur arabischen Mundart von Mossul”, Zeitschrift für arabische Linguistik 2, pp. 36–75. ———, 1988, Der neuaramäische Dialekt von Hertevin (Provinz Siirt), Wiesbaden: Harrassowitz. Khan, G., 2002, The Neo-Aramaic Dialect of Qaraqosh, Leiden: Brill. ———, 2008a, The Neo-Aramaic Dialect of Barwar, vol. I: Grammar, Leiden: Brill. ———, 2008b, “The Syriac words in the Kitāb al-Mustaʿīnī in the Arcadian Library”, in C. Burnett (ed.), Ibn Baklarish’s Book of Simples, Oxford: OUP, pp. 95–104. ———, 2016, The Neo-Aramaic Dialect of the Assyrian Christians of Urmi, Leiden: Brill. Mackenzie, D. N., 1961, Kurdish Dialect Studies I, London: OUP. Maclean, A. J., 1895, Grammar of the Dialects of Vernacular Syriac, Cambridge: CUP. ———, 1901, A Dictionary of the Dialects of Vernacular Syriac as Spoken by the Eastern Syrians of Kurdistan, North-West Persia, and the Plain of Mosul, Oxford: Clarendon Press. Mengozzi, A., 2002a, Israel of Alqosh and Joseph of Telkepe, A Story in a Truthful Language: Religious Poems in Vernacular Syriac (North Iraq, 17th Century), vol. I: Text and Glossary, Leuven: Peeters. ———, 2002b, Israel of Alqosh and Joseph of Telkepe, A Story in a Truthful Language: Religious Poems in Vernacular Syriac (North Iraq, 17th Century), vol. II: Introduction and Translation, Leuven: Peeters. ———, 2011, Religious Poetry in Vernacular Syriac from Northern Iraq (17th–20th Centuries): An Anthology, E. Braida, S. Destefanis and S. Talia (contrib.), Leuven: Peeters. ———, 2012, “The contribution of early Christian vernacular poetry from northern Iraq to Neo-Aramaic dialectology: preliminary remarks on the verbal system”, ARAM Periodical 24, pp. 25–40. Mole, K., 2015, “The three /r/s of Baṛtəḷḷa”, in G. Khan and L. Napiorkowska (eds), Neo-Aramaic and its Linguistic Context, Piscataway, NJ: Gorgias, pp. 110–129. Pennacchietti, F. A., 1990, “Due pagine da un manoscritto inedito di una poesia religiosa neoaramaica di Yawsip Ğemandi (XVII sec.)”, in A. Vivian (ed.), Biblische und judaistische Studien. Festschrift für Paolo Sacchi, Frankfurt am Main: Peter Lang, pp. 691–709. Poizat, B., 1990, “La complainte sur la peste de Pioz”, in W. Heinrichs (ed.), Studies in Neo-Aramaic, Atlanta: Scholars Press, pp. 161–179, 203–207. ———, 1993, “La peste de Pioz. Suite et fin”, in R. Contini, F.A. Pennacchietti and M. Tosco (eds), Semitica: Serta philologica Constantino Tsereteli dicata, Turin: Zamorani, pp. 227–272. Rhétoré, J., 1912, Grammaire de la langue Soureth ou Chaldéen vulgaire, selon le dialecte de la plaine de Mossoul et des pays adjacents, Mossoul: Imprimerie des Pères Dominicains. Rubba, J., 1993a, Discontinuous Morphology in Modern Aramaic (unpubl. diss., University of California, San Diego). 269 ———, 1993b, “Forms derived from verbal roots in Tisqoopa Modern Aramaic”, in R. Contini, F. Pennacchietti and M. Tosco (eds), Semitica: Serta philologica Constantino Tsereteli dicata, Turin: Zamorani, pp. 115–126. Sabar, Y., 1978, “From Tel-Kēpe (‘a pile of stones’) in Iraqi Kurdistan to Providence, Rhode Island: the story of a Chaldean immigrant to the United States of America in 1927”, Journal of the American Oriental Society 98/4, pp. 410–415. ———, 1993, “A folktale and folk songs in the Christian Neo-Aramaic dialect of TelKēpe (Northern Iraq)”, in R. Contini, F. Pennacchietti and M. Tosco (eds), Semitica: Serta philologica Constantino Tsereteli dicata, Turin: Zamorani, pp. 289– 297. Sachau, C. E., 1883, Reise in Syrien und Mesopotamien, Leipzig: Brockhaus. ———, 1895, Skizze des Fellichi-Dialekts von Mosul, Berlin: Königlichen Akademie der Wissenschaften. Sengstock, M. C., 2005, Chaldeans in Michigan, East Lansing: Michigan State University Press. Socin, A., 1882, Die neuaramäischen Dialekte von Urmia bis Mosul. Tübingen. Wilmshurst, D., 2000, The Ecclesiastical Organisation of the Church of the East, 1318–1913, Leuven: Peeters. Appendix: Abbreviations and glosses I, II, II2, III, Q Ct-, St-, Ti, v, viii, x = ꞊ | <?> SMALL CAPS ACT.PTCP Alq. ANT Arab. B COMP COP CST EXIST F FS FUT GEN IND INF K. L M 270 NENA verbal derivation patterns NENA verbal derivation patterns borrowed from Arabic Arabic verbal derivation patterns links two words or morphemes in a phrase with a single stress on the second component (including but not limited to proclitics) links two words or morphemes in a phrase with a single stress on the first component (including but not limited to enclitics) intonation phrase boundary inaudible speech nuclear stress in intonation phrase active participle Alqosh dialect anterior (shifting the time reference back, glossing -wå~-wā) Arabic B-suffix complementiser copula construct state suffix -əd existential (particle) feminine feminine singular future (tense) genitive marker dindicative infinitive Kurdish L-suffix masculine MS NEG PAST Pesh. PL PRS PRSP PST PST_PFV REL RES.PTCP S SG Syr. TK masculine singular negator/negated Past Base Peshabur dialect plural present (tense) prospective (aspect) past (tense) past perfective (glossing kəm-) relativiser resultative participle S-suffix singular Classical Syriac Telkepe dialect 271 ‘The King and the Wazir’: A Folk-Tale in the Jewish North-Eastern Neo-Aramaic Dialect of Zakho OZ ALONI Middlebury College At the centre of this article is the transcription and translation of a folk-tale told in the Jewish North-Eastern Neo-Aramaic (NENA) dialect of Zakho (member of the lišána dèni group of dialects). This is a rather unusual folktale, since it is built around a relatively uncommon motif in folk-literature, the motif of gender transformation. The folk-tale, told by Ḥabuba Messusani, was recorded as part of a Jewish Zakho NENA audio database project, which now comprises approximately 150 hours of audio recordings of native speakers of that dialect, in various spoken genres. It was Professor Geoffrey Khan, who first encouraged me to start this project in 2010, stressing the importance of the documentation and study of the NENA dialects.1 I wish to express my gratitude to him for that.2 1 The folk-tales of the Jews of Zakho The NENA-speaking Jewish community of Zakho (Iraqi Kurdistan) migrated collectively to Israel in 1951, together with the other Jews of Iraqi Kurdistan, carrying with it, so to speak, its unique language, culture, customs and exceptionally rich oral heritage.3 An essential part of that oral heritage is the large and complex corpus of folk-tales. This draws on both Jewish and Kurdish folklore: many of its tales bear distinctive Jewish characteristics, while others belong in the general regional repertoire. Telling folk-tales, and listening to 1 See Khan, 2007, p. 1: “The description of these dialects is of immense importance for Semitic philology. The dialects exhibit linguistic developments that are not only interesting in their own right but also present illuminating parallels to developments in earlier Semitic.” 2 I also wish to thank Batia Aloni, Yoel Perez and Zadok Alon for their comments. Special thanks are extended to my mother Batia Aloni for the help in proofreading the Neo-Aramaic transcription. I thank Ḥabuba Messusani for the many hours of recording sessions, of which this folk-tale is a fraction. 3 For the history and culture of the Jews of Zakho and Kurdistan, see Gavish, 2010; Ben-Yaacob, 1981; Brauer, 1993; Zaken, 2007; Aloni, 2014. 272 them, was a very common and popular shared pastime of the communities of Kurdistan. The very same folk-tales, in different versions, with additions, omissions or creative embellishments – all depending on the taste (and talent) of the tellers and their audience – could be told throughout Kurdistan, and in all of its different languages and dialects. The practice of storytelling continued in the Jewish-Kurdish communities in Israel: the senior members of the Zakho community in Jerusalem tell of the regular gatherings in a diwàn, a drawing room of a home of one of the elders of the community, for the purpose of listening and telling stories. Zakho folk-tales vary in length from relatively short ones, like the one presented here, to very long ones capable of filling several long consecutive winter evenings – oral novels, one may call them. Folk-tales are a social institution that plays a role in the forming and maintaining of the Zakho communal identity. They also take part in intergenerational communication: in a society that experienced a deep intergenerational gap brought about by the sharp transition to modern Israel,4 folk-tales (and other oral genres) are a mode of contact between the generation of the grandparents and their grandchildren.5 2 ‘The King and the Wazir’: Synopsis A king and his wazir go out to explore their town, wearing ordinary clothes. After crossing a bridge, the wazir’s horse breaks into gallop, leaving the king alone. The king arrives at a river, and he sits down in order to eat and rest. He plays with his ring, and it falls into the water. The king dives into the water in order to recover his ring, and when he gets out, yímmed ṃáya ‘the mother of the water’ (a water spirit) hits him on the head, and he is transformed into a woman. As he sees his reflection in the water, he realises that he is now a very beautiful woman. Some fishermen who are passing by take the beautiful woman, with the intention of marrying her to the son of their own king. The king and queen are astounded by the woman’s beauty, and their son the prince 4 See Sabar, 1975. About the social changes within the community caused by the migration, see Gavish, 2010, pp. 316–336. 5 Published Jewish Zakho folk-tales are: Socin, 1882, pp. 159–168, pp. 219–223; Polotsky, 1967, two episodes from a ‘novel’; Alon and Meehan, 1979; Avinery, 1978; Avinery, 1988, pp. 48–65; Zaken, 1997; Shilo, 2014, a collection of 14 folk-tales written originally in NENA (not transcribed from a recording), which I edited; Aloni, 2014, pp. 65–79. An important collection of oral literature of the Jews of Kurdistan, though only in English, is Sabar, 1982. The most important collection of folk-tales in the Jewish NENA dialect of Zakho is yet unpublished. It is a corpus of 33 stories recorded from Mamo (‘uncle’) Yona Gabbay Zaqen, father of the teller of our present folk-tale, Ḥabuba Messusani. Mamo Yona (Zakho 1867–Jerusalem 1970), an exceptional bearer and performer of the rich tradition of the Jews of Kurdistan and a wellknown storyteller throughout Iraqi Kurdistan, was recorded during 1964 by Professor Yona Sabar for the Hebrew University’s Jewish Language Traditions Project (Mifʿal Masorot HaLlašon, see Fellman, 1978). Only a small portion of this material has been published, in Sabar, 2005: Mamo Yona’s own life story, narrated by him. 273 falls in love with her. The woman and the prince get married and have three children. To celebrate the third birth, the king throws a seheràne (an outdoor celebration) for all his people. The woman goes to the riverside in order to look again for her lost ring (the king’s ring). She sees the ring in the water, and gets into the river to take it. The mother of the water comes again, hits her on the head, and the woman becomes a man once more, the king. He does not know what to do next. In the meantime, the wazir, who had fallen from his horse, is found by some hunters who realise that he is an important man, seeing his beautiful clothes and horse. He does not remember who he is, as he has lost his memory. The hunters take him to a hospital, where he is taken care of for one year. A professor takes him home to be his servant, and eventually the wazir becomes like a son to him. One day the wazir is riding his horse, the horse again gallops, and the wazir falls from his horse at the same place where he had fallen before. He regains his memory. The wazir and his adoptive father go to the wazir’s home, but his wife does not recognise him. She suggests that they should go to the imam, and he will decide whether the wazir is her husband or not. The king also comes back to his home. His wife does not believe that he is her husband, so he also waits for the imam to come on Friday. The imam, who turns out to be Bahlul, the king’s brother, decrees that the king is the king and that the wazir is the wazir, and he sends them back to their homes. The prince, who had been married to the woman who the king became, searches for his wife everywhere. Eventually he arrives in the town of the king and the wazir. He goes to the imam and tells him about his lost wife. The imam tells the prince that his wife is not lost, she is a king. The king demands that the prince give him the children that he bore as a woman, and tells the whole story of his transformation. The imam decrees that the prince will keep those children, since the king has other children who he had earlier fathered as a man. The king and the prince both return to their homes. 3 The motif of gender transformation Many of the motifs6 that appear in our story are known from other literary and folk traditions. To list but a few: the king and his wazir go out wearing ordinary clothes (motif K1812.17 ‘king in disguise to spy out his kingdom’); the king dropped his ring in water and then recovered it (K1812.17 ‘Solomon’s power to hold kingdom dependent on ring; drops it in water’); yímmed ṃáya 6 As classified by Thompson, 1955–1958. Motifs numbers and titles discussed here are taken from Thompson’s classification. For the concept of motif in folklore, and critiques thereof, see Dundes, 1962; Ben-Amos, 1980; Ben-Amos, 1995. 274 ‘the mother of the water’ (motif F420 ‘water spirits’);7 the king looks at his reflection in the water after having been transformed and sees an extraordinarily beautiful woman (motif T11.5.1 ‘falling in love with one’s own reflection in water. (Narcissus.)’).8 But the most surprising motif in our folk-tale, and one which plays a fundamental role in its structure, is certainly motif D10 ‘transformation to person of different sex’.9 Motif D10 is relatively uncommon in literary and folk traditions cross-culturally. In both written and oral literature, it is predominantly found in narratives from the Indian cultural space,10 though it is not restricted to it. Many of its other occurrences in oral folk-literature come from the Middle-East – Egypt,11 Turkey,12 the Jews of Iraqi Kurdistan13 and the Jews of Yemen14 – although it appears in non–Middle Eastern traditions as well.15 Only one occurrence of motif D10 is to be found in classical Jewish literature. That is in a story of a poor widower whose wife left him a nursing baby. The widower could not afford a wet nurse, and by way of miracle he gained breasts and fed his son himself.16 Perhaps the most well-known occurrence of D10 in Western culture is the Greek myth of Tiresias, the blind prophet who, as a punishment from Hera for hurting a pair of copulating snakes, spends seven years as a woman and gives birth to children. After encountering another pair of copulating snakes and spearing them, he is released from his punishment. Having the experience of being both a man and a woman, Tiresias is asked to judge in an argument between Zeus and his wife Hera: who has more pleasure in sexual relations, men or women? Tiresias agrees with Zeus’ opinion, and says that women’s enjoyment is ten times greater. An Indian story from the Mahabharata, the story of King Bhangaswana,17 shares many plot elements with our folk-tale. King Bhangaswana is punished by Indra for not including him in a sacrificial ceremony. He is transformed into a woman while bathing in a lake. Bhangaswana had one hundred sons as 7 In his index, Noy (Neuman, 1954, p. 395) refers to Ginzberg, 1925, pp. 87, 204, who lists several occurrences of water spirits in Jewish literature. Ginzberg mentions the belief, also found in Greek literature, that “water is the adobe of demons”. 8 See also motif J1791.6.1. 9 Similar relevant motifs are: D10.2 ‘change of sex after crossing water’; D12 ‘transformation: man to woman’; D695 ‘man transformed to woman has children’; T578 ‘pregnant man’. 10 For a thorough overview of the sources, see Brown, 1927; Penzer, 1927. 11 El-Shamy, 1980, pp. 33–38. 12 Walker and Uysal, 1992, pp. 241–243. 13 In addition to our folk-tale, tales number 3932, 13471 and 16376 at the Israel Folktale Archives Named in Honor of Dov Noy (IFA), University of Haifa. 14 Tale number 1235 at IFA. 15 For instance it is found in Benin, China, the French-speaking region of Canadia, India, Inuit regions and Ireland. See Thompson, 1955–1958, vol. II, pp. 8–9; Thompson and Balys, 1958, p. 97. 16 Babylonian Talmud, Shabbat 53b. Noy (Neuman, 1954, p. 281) gives several cases of male embryo transformed into female in the womb. 17 Ganguli, c1900, book 13, §12, pp. 35–38. 275 a man and one hundred sons as a woman. They all slew one another in a battle incited by Indra. When Indra pardons Bhangaswana, now living as an ascetic woman, he asks which of the children should be resurrected. Bhangaswana replies that those he had as a woman should be resurrected, since the affection of a woman to her children is greater than that of a man. Highly pleased by the woman’s truthfulness, Indra resurrects all two hundred children. He then gives Bhangaswana the choice of being a man or a woman, but Bhangaswana chooses to remain a woman, since the pleasure a woman finds in sexual relations is greater than that of a man. The many print and manuscript versions of the Arabian Nights include four stories which containing the motif of a change of gender: ‘The Enchanted Spring’, ‘Hasan the King of Egypt’, ‘Warlock and the Young Cook of Baghdad’, and ‘Shahab al-Din’.18 The latter two correspond to international taletype ATU 681 ‘relativity of time’19 (previously known as tale-type AT 681 ‘king in a bath; years of experience in a moment’20). ‘Hasan the King of Egypt’ is reminiscent of an Egyptian oral tale.21 In ‘Warlock and the Young Cook of Baghdad’ a transformed vizier gets married and gives birth to seven children; the transformed vizier of ‘Hasan the King of Egypt’ gives birth to only a single child. In all four stories the change of sex is by means of dipping in water. The oldest of the Middle-Eastern manifestation of the motif is the one of the tale of Khurafa (Ḥadith Khurafa).22 In its most elaborate version, in the book Al-Fākhir by 9th century writer Al-Mufaḍḍal ibn Salama, Khurafa, taken prisoner by three jinns, hears the following story told by a man: the man was transformed into a woman after being trapped in a particular well; he then got married and gave birth to two children; after some time he went back to the same well, was transformed back into a man, got married again and had two more children.23 The final story that will be mentioned here is possibly the earliest recorded folk-tale of the Jews of Zakho. It also includes the transformation of men into women in proximity to water – in this case, the transformation of two men. This is a Jewish Zakho NENA text recorded by Socin as early as 1870 from Pineḥas of Zakho,24 which recounts the story of the two brothers ʿAli and 18 Stories number 191, 545, 412 and 435 in Marzolph, Leeuwen and Wassouf, 2004. Uther, 2004, vol. I, p. 373; see also Marzolph, Leeuwen and Wassouf, 2004, p. 797. 20 Aarne and Thompson, 1961, p. 238. 21 El-Shamy, 1980, p. 33–38, mentioned above in n. 11. 22 See Drory, 1994, where she claims that Ḥadith Khurafa was one of the earliest “attempts to legitimize fiction in classical Arabic literature”. See also Marzolph, Leeuwen and Wassouf, 2004, p. 616. 23 This story is classified by El-Shamy, 2004, p. 378, as tale-type 705B ‘ “I have begotten children from my loins, and from my womb!”: Khurâfah’s experience’, where he lists more of its occurrences. 24 Sabar, 2002b, p. 613, suggests that this is Pineḥas Čilmèro. 19 276 ʿAmar.25 Sabar has published an updated version of this story, written in language as if it were told in the 1950s, together with a commentary on the linguistic differences between the two versions.26 In this story, the son of ʿAmar and his friend go hunting. They chase after a gazelle for three days, and on the third day they reach a river. The gazelle leaps over it and says to them, “Stop following me. God will, if you are men, you will become women; if you are women, you will become men!”.27 They marry men and live as women for seven years. One of them gives birth to a triplet of boys, and the other to a triplet of girls. One day they dress as men, take their horses, and ride to find the gazelle. Again they chase after her for three days, and then reach a river. The Gazelle leaps again and says the same words, and the two are transformed back into men and return to their homes. Almost all of the stories mentioned here present a curious coupling: the proximity of motif D10 to water. Indeed, in his article about the motif in Indian literature, Brown lists “bathing in an enchanted pool or stream”28 as the first of five means by which a change of sex is effected,29 and Penzer, after providing an overview of cases of sex transformation “by a magic pill, seal or plant, or merely by mutual agreement with a superhuman being”,30 writes that “as the motif travelled westward it seems that water became the more usual medium”.31 One more element of our story should be commented on: the name of the imam, Bahlul. The character of Bahlul, or Behlül Dane – the clever brother, or son, of caliph Harun Al-Rashid – is well known from many folk-tales, especially those originating in eastern Turkey.32 A whole sub-genre of folk-tales features him. In all of them he seems at first like a simpleton, or pretends to be one, but eventually he proves his mental and moral superiority over everyone, including the caliph. One of the many Behlül Dane stories is particularly relevant to our folk-tale. In the story ‘Behlül Dane Teaches God’s Time versus Human Time’,33 the caliph Harun Reşit is sceptical when he hears Behlül Dane saying, ‘I have a God whose one hour is equivalent to a thousand of our hours’. When entering the toilet with a kettle of water Harun Reşit has a vision in which he lives as a woman for years, gets married and has children. He then wakes up to discover himself still in his toilet. 25 Socin, 1882. Sabar, 2002b. 27 Sabar, 2002b, p. 625. 28 Brown, 1927, p. 4. 29 The other four are: curse or blessing of a deity; exchanging sex with a Yakṣa, “a creature that is unique in possessing the power to make this remarkable exchange”; by magic; by the power of righteousness or in consequence of wickedness. See Brown, 1927, pp. 4–5. 30 Penzer, 1927, p. 224. 31 Penzer, 1927, p. 224. 32 See Walker and Uysal, 1966, p. 296. 33 Told by Hacı Mehmet Sivri in 1974; see Walker and Uysal, 1992, pp. 241–243. 26 277 4 Báxtox ḥakòma-la ‘your wife is a king’: Gender boundaries and perplexity Many scholars have commented on the cultural and social unrest and anxiety that undermining gender boundaries may create.34 In our folk-tale, confusion generated by the focal point of motif D10 – the notion that breaking gender boundaries is possible, even by magic – permeates through many of the narrative elements. A latent sense of confusion is everywhere: in the plot and the reasoning of its events, in the words and the actions of the characters, in the narration, even in the language of the folk-tale. From the very first event in the storyline, obscurity is present. The wazir’s horse breaks into a gallop for no apparent reason. He then falls from it, loses his memory, and spends several years under another identity. The king is transformed into a woman by a water spirit, gets married and has children. He has not done anything to enrage the water spirit which could have caused this unwelcome transformation.35 What is the reason for or purpose of these ordeals? Do they come as a punishment, or in order to teach some lesson? In many of the other stories built around these motifs, some rationale for the tormenting adventures undergone by the characters is given: they are either punished by enraged gods or spirits, or taught a lesson after showing disbelief. Not in our folk-tale. The king and the wazir’s long and harsh ordeals come and then go away with no apparent motive nor benefit of a lesson learned. Even when their period of transformation is done and they regain their original identity, there are hardships involved – the disbelief of the wives, the king torn away from the children he gave birth to as a woman, the prince losing his beloved wife – and no greater power, position, wealth or wisdom – no compensation – is gained. This is a Kafkaesque folk-tale, almost as Kafkaesque as Kafka’s own Metamorphosis, where the suffering of the protagonists is left unexplained and unresolved. The words of the king after being transformed back into a man in his second encounter with the mother of the water, where we would expect him to rejoice at having recovered his identity, are (45) wi-má-b-ozə́n ʾə-nàqla?| … lá-k-iʾən ma34 For example, “Cross-dressing is about gender confusion.” About this sentence, taken from Marjorie Garber’s book Vested Interests: Cross-Dressing and Cultural Anxiety (1992, p. 390), Tova Rosen, 2003, pp. 149–150, writes: “If clothing is a language, then cross-dressing poses a gender riddle. Clothes are intended both to cover and to reveal; they hide the body’s sexual signs and, at the same time, signify the binarism of the sexes. The concealed anatomical differences are replaced by a culturally determined gendered symbolism of clothing. Thus, in texts, as well as in life, clothing functions as a code for sexual (and other) differences. Moreover, the language of clothing does not only encode ‘masculinity’ or ‘femininity’, but rather points to the very constructedness of gender categories. Cross-dressing, on the other hand, manifests the discontinuity between the sexual body and the cultural gender and, thus, offers a challenge to easy notions of binarism.” Also, Meiri, 2011, pp. 164–165: “Transsexuality evokes categorical and epistemic crises more than any other form of crossing of gender.… transsexuality, in its visibility, holds in itself the various anxieties evoked by different forms of crossing of gender” (my translation). 35 About gender transformation as unexpected and unwelcome, see Brown, 1927, pp. 6–9. 278 ʾòzən.| ‘Oh, what shall I do now? … I do not know what to do.’ His confusion is evident, and is growing: (46) la-k-íʾa ma-ʾòza,| ta-máni ʾáza ʾámra ʾána ḥakòma-wán.| ta-máni ʾámra ʾána bax-ḥakòma-wán.| ‘She does not know what to do, to whom would she go [and] say “I am the king”? To whom would she say “I am the wife of the king”?’ This reaction of the king, his manhood restored, seems even more helpless than his reaction to his first transformation, where he simply wore his old man’s clothes and was taken away by the fishermen. The peak of confusion and loss of identity in the story is found in the secondary character, the wazir. When he is found by the hunters after he has fallen from his horse, the following short dialogue takes place: (51) là-g-maḥké,| lahè la-lá,| g-əmríle màni-wət?| g-émer là-k-iʾen, wéle pṣìʿa.| m-èka wét? g-émer là-k-iʾen.| ‘He does not speak, not “yes” [and] not “no”, they say to him “who are you?” He says, “I don’t know”, he is wounded. “Where are you from?” He says, “I don’t know”.’ The wazir’s words are at variance with his appearance, a tension between his external identity markers and his own lack of identity: he is recognised by the hunters as being an important person by his clothing and horse, but the external aspects of his identity do not help him when he loses his sense of self. The atmosphere of confusion is not created by the events of the storyline alone; stylistic features of the narrative contribute to it as well. For instance, the characters are namelessness. Only one character, who appears towards the end of the story, has a name: the imam Bahlul.36 The lack of given names, which is a well-known characteristic of fairy-tales in itself, contributes to the confusion of the listener due to the identity transformations in our folk-tale. Furthermore, the confusion is aggravated. Our folk-tale contains three kings (the main character; the father of the prince; and the prince, who is also referred to as king), three queens (the wife of the main character; the mother of the prince; and the woman who used to be king, who is referred to as queen after marrying the prince), and three women (the main character; the wazir’s wife; the main character’s wife). These sets of characters are referred to as ‘the king’, ‘the queen’ and ‘the woman’ respectively, without specification. It seems that even the teller of the story herself is partaking in the general bafflement. The following episode occurs just before the wazir goes out for the ride which will bring about the regaining of his memory: (55) ʾáwa| qə́mle xà-yoma,| g-ə́mri wéle ḥakòmda,| ʾə́tle ṭèra.| ḥakóma dóhun mə̀tle.| ʾə́tle ṭéra g-mandèle.| ‘He rose one day, they say there’s a king, which has a bird. Their king died. He has a bird which they throw.’ This episode, which seems incoherent and has no clear ties to preceding or subsequent events, is located at a 36 It is interesting to note that the imam plays a role of clarifying and restoring order. The children of the wazir are also given names, Mirza-Maḥamad Aḥmad and Fatma, but these characters play no role in the story; the knowledge of their names is used as proof of identity. That is, once again, names have a role in restoring order. 279 crucial point of the storyline, just before all the entanglements of the story begin to be resolved. Gender transformation spreads confusion and chaos even in the grammatical structure of the language of the folk-tale: at the points of transformation, as well as when the king later recounts his experiences, the use of referential elements with specified gender – pronouns and conjugations – becomes unclear. Grammatical elements of the ‘wrong’ gender are used both before and after a transformation takes place. For example, in (44)–(46): pə́šla gòra.| qə́mla lwišíla júlle dìda| mxéla l-ʾúrxa ‘She became a man … She rose [and] wore her clothes and started walking.’ And also: (79) báxtox ḥakòma-la.| ‘Your wife is a king’; (80) k-xáze gòra híle,| ‘He [=the king] sees it is her [feminine, =the king’s] husband’; (81) g-émer yalúnkəd mà?| ʾa[he]t-gòra wə́t!| ṃàṭo| yalúnke mesə́nnu-làx?| ‘He [=the husband] says [to the king]: “Children of what? You are a man! How will I bring you [feminine] the children?” ’. The same grammatical confusion occurs in other places in our folktale as well.37 5 ‘The King and the Wazir’: The text This folk-tale,38 ‘The King and the Wazir’, told by Ḥabuba Messusoni, was recorded on 7 January 2013 at Ḥabuba’s home in Jerusalem’s Katamonim neighbourhood, where many of the Jewish immigrants from Kurdistan settled when arriving in 1951. Ḥabuba was born in Zakho in 1936 and came to Jerusalem in 1951. As mentioned, she is the daughter of the famous storyteller Mamo Yona Gabbay.39 Present in the recording session were Ḥabuba Messusani (HM), Batia Aloni (BA), Professor Geoffrey Khan (GK), and myself. The transcription system used here is the one used by Professor Khan in his NENA grammars. In addition to the standard Semitic consonant and vowel signs, intonation signs are employed: a superscript vertical line (a|) indicates an intonation unit boundary; a grave accent (à) indicates the main stress in an intonation unit; acute accents (á) indicate secondary stresses in an intonation unit. Words or phrases in Modern Hebrew are written between superscript capital H letters (H…H). The English translation is as literal as possible; tenses are kept as in the NENA text, at the expense of standard English style.40 The recording ID is HM130107T4 00:04-12:16. This linguistic abnormality appears also in the story of the brothers ʿAli and ʿAmar; see Socin, 1882, p. 164, line 6; Sabar, 2002b, p. 621, no. 51. 38 This folk-tale clearly belongs to the genre of fairy-tale (Märchen). It presents the genre’s distinctive characteristics: unknown time and place of happening, nameless protagonists, archetypical characters, miraculous incidents and supernatural beings. That being said, keep in mind Dundes’ assertion (1964, p. 252): “… thus far in the illustrious history of the discipline [=folkloristics], not so much as one genre has been completely defined.” 39 See n. 5 above. 40 For a study of Jewish Zakho NENA narrative syntax, see Cohen, 2012, pp. 237–357. 37 280 (1) HM: Hhayá mélexH xá ḥakòma| u- HM: There was a king, a king, and a wazir. wazìra.| (2) ḥakóma g-émer ta-wazíra dìde,| 41 | |H H| d -áx xàzax má hìle maṣàv bážer dèni.| (3) b-lóšax júlle də́d Hragìl,H| hàdxa,| júlləd dàrwiše,| b-áx zàvrax.| The king says to his wazir, ‘Let us see what is the situation of our town. We shall wear these ordinary clothes [lit. clothes of regular], like that, beggars’ clothes, we shall go [and] wander around.’ (4) g-émer[r]e-go-ʾèni.| He says to him, ‘upon my eyes’.42 (5) g-émer náblax xa-ġolàma He says, ‘Shall we take a servant with us?’, he says, ‘No’. ʾə́mman,| g-émer là.| (6) ṭʾón xápča ʾawàye,| ʾixàla,| u-drí go-kə́sta dìdox, | Carry some things, food, and put [them] in your bag, (7) ʾá[hə]t go-mahíne dídox, ʾàna go- you on [lit. in] your horse, I on [lit. in] my horse. Both of us will mahíne dídi| kútran b-áx. go. (8) [m]pə́qlu básər gə̀šra,| They went out, [and right] after the bridge, (9) mahíne dəd wàzir| dhə̀rra.|43 ʾí u- the wazir’s horse broke into gallop. I and di44 she ran and ran and ran and ran and took him [=the wazir] with her, until a distance [lit. way] of some five kilometres [where] she dropped him. dì u-ʾrə́qla u-ʾrə́qla u-ʾrə́qla uʾrə́qla u-qam-nablále ʾèmma,| hìl| ʾúrxət-HʾezeH xamšá HkelométerH qam-mamp[ə]làle.| (10) pə́šle ḥákoma Hlevàd,H| lá-k-iʾe ʾéka ʾàl,| ʾéka lá ʾàzəl.|45 The king was left [lit. became] alone, he does not know where he should go, where he should not go.46 41 Contraction of the interjection de. Idiomatic expression meaning ‘I will fulfill your request’. 43 The Modern Hebrew root dhr is used here with NENA morphology. 44 Sabar, 2002a, p. 141: “day-day-day: sounds describing speed of racing animals”. 45 Note the use of two allomorphic forms of the same verb within one sentence: ʾàl, ʾàzəl. 46 Idiomatic expression meaning ‘he did not know where to go’, ‘he was utterly perplexed’. 42 281 (11) zə̀lle.| He started walking [lit. he went]. (12) zə́lle47 xzéle xá,| xawòra.| xawóra k-íʾət mà-yle?| He went47 [and] saw a river. Do you know what is xawóra? (13) GK: … he… GK: … Yes … (14) HM: xawòra,| Hnàhar.H| | (15) xzéle-xa xawòra, rùwwa. HM: xawóra, a river. | (16) qə́mle túle ž -dáw… tàma. 48 He saw a river, [a] big [one]. | (17) šlə́xle ḥášak dídox49 Hnaʿalà…H| qundáre dìde,| dréle ʾáqle goṃàya,| mopə́qle xápča ʾixála xə̀lle,| mopə́qle józi díde ʾúzlele xa-qàhwa,| mtoʿə́lle bə́d| ʾasə́qsa dìde hàdxa.| ʾasə́qsa díde mpélla He rose [and] sat down upon that … there. He took off, excuse my language,49 his shoes, [and] put his feet in the water. He took out some food [and] ate, took out his coffee kettle [and] made himself a coffee, he played with his ring, like that. His ring fell into the water. go-ṃàya.| (18) wày g-émer| mpə̀lla| ʾátta lá-k-iʾən ‘Oh!’ he says, ‘It fell, now I do not know where, what I shall do, ʿéka má b-òzen,| d-lá ʾasə̀qsa.| without a ring.’ He rose, took off qə̀mle,| šlə́xle júlle dìde| u-g-émer his clothes, and he says, ‘I shall b-àn,| kóšən go-ṃàya,| zéʾli ʾéka go, go down into the water, [since] I know where it had fallen. mpə̀lla.| mapqə̀nna.| I shall bring it out.’ (19) mpə́qle, yímmed ṃáya50 sèla.| mxéla-[ʾəl]le xá… hə̀nna|51 47 [When] he went out [of the water], the Mother of the Water came. She struck him with one … this,51 rašòma52 upon his head. She turned him into such a girl, This repetition of a word or phrase is a typical stylistic feature of Jewish Zakho NENA narration. It usually appears at the beginning of an episode in the narrative. 48 Contraction of rəš-. 49 Sabar, 2002a, p. 169, on ḥàšak dōxun: “All present/of you excluded (said after saying a dirty word)”. 50 Sabar, 2002a, p. 177: “a female ghost that dwells in the river”. 51 The literal meaning of hənna is ‘this’ or ‘this thing’. Pragmatically it is used for several ̀ functions: a substitute for a word that the speaker is unable to remember (sometimes the speaker will add the forgotten word immediately thereafter); an anaphoric pronoun referring back to an object or a concept mentioned earlier; an abbreviation replacing an idea that all participants know it refers to; and as a euphemistic substitute for words that the speaker wishes to avoid saying. hə̀nna is translated as italicised ‘this’ throughout the English translation. 282 rašóma52 go-rèše,| qam-ʾozále xà H baḥorà,H| lá-g-hanélox ʾə̀bba men[xət].| ḥakòma pə́šle H you could not stare enough at [lit. you would not enjoy (i.e. be satisfied) to stare at her]. The king became a young woman. baḥurà.H| (20) k-xáze gyàne,| bràta-le!| xà sqélta! He sees himself [=his reflection in the river], he is a woman! So lá-g-hanèlox ʾə̀bba.| beautiful! You could not enjoy [staring enough] at. (21) [m]pə́qle l-wàrya,| júllet gùre-lu táma. lúšle júlle dìde| túle ltàma.|53 (22) sèlu,| ʾánya| də́d g-dóqi hə̀nna| šabakvàne g-ə́be dóqi g-doqí nunyàsa.| k-xáze ʾé HbaḥuráH | hádxa sqə̀lta,| g-ə́mri wáḷḷa bə́r ḥakóma dèni,| hay-ṭḷá[ha] šə́nne wélu bə-zvára xa-HbaḥuráH ṭàḷe,| xa-sqə̀lta,| xa-bràta u-là| g-ṛáẓe bəd-čù-xa.| He went out [of the water], men’s clothes were there. He wore his clothes. He sat there. Came, these, who catch this, fishermen, they want to catch, they catch fish. They see this so beautiful girl, they say, ‘Indeed the son of our king, for three years they have been searching [lit. turning around] for a girl for him, a beautiful [girl] [or: a beauty], a girl, and he is not satisfied with anyone.’ (23) BA: ʿaqə́le la-qṭéʾle ʾəl-čù-xa.| BA: His mind was not cut on anyone [=He was not satisfied with anyone]. (24) HM: ʾéha b-nabláxla HʾulàyH HM: ‘This one [=the girl], we shall take her [to him], perhaps he would be satisfied with her.’ ṛaẓe-ʾə́bba.| (25) qə́mlu sèlu,|54 sèlu,| qam-nablíla qămáye kəz-ḥakòma, yímme ubàbe,| qam-… g-ə̀mri,| ʾéha ġe[r]… ʾé ġèr-məndi-la| go-Hkól They rose [and] came, they came,54 they took her first to the king, his mother and father, they say, ‘That [girl] is something different, in the entire world there is not [a girl] like her, she is even 52 Sabar, 2002a, p. 292: “vertical hand used as cursing sign; a blow with open hand on top of the head (to indicate disdain, disapproval …)”. Also appears in Rivlin, 1959, pp. 226, 240. 53 Verbal forms and pronouns in this sentence are masculine. The woman is still referred to as a man here. 54 See n. 47 above. 283 ha-ʿolámH lez-moxwà[sa]| bəs55sqə́lta-la mə́n ráḥel ʾəmmènu ʾafə́llu.| (26) Htòv.H| məsélu Hyèled,H ʾéne…| qam-xazèla, ʿšə́qle ʾə̀lla, qamʾebèla.| | | more beautiful than Rachel our Mother’.56 Good. They brought the child [=the prince]. His eyes … he saw her, he fell in love with her, he loved [or: wanted] her. (27) zə́llu məsélu qám|-barxíla ʾə̀lle,| u- They went [and] brought [and] married them [lit. they blessed her ʾáy šàta,| smə̀xla.| [h]wélela xato him], and in that year she bròna.| šátəd…| pə̀šla,| báser tré became pregnant. She gave birth šə̀nne,| smə̀xla, hwélela xa-bróna to a son [lit. a son was born to her]. A year … she stayed [=she xə̀t.| báser tré tḷá[ha] šə̀nne| did not become pregnant for one smə́xla hwélela xa-bróna xə̀t hay- year, and then] After two years she became pregnant [again] and tḷàha.| gave birth to another son. After two [or] three years she became pregnant [again and] gave birth to another son, that’s three. (28) qə́mlu HʾanšeyH-bàžer,| ʾo ḥakóma They rose, the people of the city, the king said, he says, ‘I shall do a mə̀rre,| g-émer b-ózen| seheràne.’57 Do you know what is seheràne.|57 k-íʾət má-yla a seheràne? | seheràne? (29) GK: mm (30) HM: mà-yla? GK: Mm. | HM: What is it? (31) GK: Hmesibà.H| GK: A party. (32) BA: Hnaxon.H (33) HM: seheráne nápqax ʾə́l-e… BA: Right. | HM: Seheràne, we go out to the … (34) BA: Hmesibà.H| BA: A party. (35) GK: Hpìknik.H| GK: A picnic. š > s due to the following consonant. Rachel the Matriarch. 57 Sabar, 2002a, p. 237: “communal procession and picnic in the country side (during Passover or Succoth Holidays)”. 55 56 284 HM: … picnic. (36) HM: … Hpə̀knək.H| (37) [m]pə́qlu b-seheràne,| u-bnablə́nna báxti u-yalúnke dìdi, They went out for the seheràne, ‘and I shall take my wife and my children, I will give all of the food kúlle ʾixàla| ʾána b-yáwən ta-náš to the people of the city, for free. They should come at my expense, bàžer,| bàlaš.| ʾáse ʾəl-xəšbòni,| because my daughter-in-law gave čukun-kálsi [h]wélela hay-tḷà[ha] birth to three boys.’ bnóne.| (38) [m]pə̀qlu.| They went out. (39) kàlse-ši,| HmalkàH-la,| …wéle H H kéter b-rèša. His daughter-in-law, she is also a queen, [she has] a crown on her head. | (40) zə̀llu,| wélu, ʾaw-yòma| xə̀llu,| štèlu,| kùllu| welu bə-rqàza| uḍòla| u-zə̀rne u| u-mád| g-ə́be| bʾ[w]ázat| faràḥe.| They went, they were, on that day they ate, they drank, everyone were dancing, and ḍoḷa and zurne,58 and whatever is necessary for a celebrations [lit. whatever is needed in making celebrations]. (41) ʾéha séla xa-hə́nna b-rèša,| g- That one [=the woman], some this came into her head, she says [to ə́mra wàḷḷa b-azána kəz-gəván herself], ‘Indeed, I shall go to the H nàhar.H ʾasə́qsa dídi mpə́lwala riverside. My ring had fallen tàma.| u-ʾasə́qsa lá xəzyàli.| qam- there. And I did not find [lit. see] the ring. That Mother of the Water ʾozáli ʾe-yímmed ṃáya HbaḥuràH|. made [=turned] me into a girl.’ | (42) zə́lla l-tàma,| zə́lla l-táma59 ʾèna,| báz monə́xla bəd-ṃàya| ʾéna nẓə́rra bə[d]-ʾasə̀qsa.| qamxazyàla.| (43) wáy! g-ə̀mra| wáḷḷa wéla ʾasə́qsa ʾasə́qsət ḥakòme-la. p-košàna. | | She went there, she went there,59 her eye, she only looked at the water, her eye caught a glance of her ring. She saw it. Oh! She says, ‘Indeed here is the ring!’ It is the ring of the king. ‘I shall go down [there]’. 58 The zurne, a conical wind instrument with a double reed, similarly to the western oboe, is played together with a large double-headed bass drum, the ḍoḷa, during weddings and other happy occasions. 59 See n. 47 above. 285 (44) šlixíla júlle dìda, šlixíla júlle g-ə́ba šáqla tabàʿat, séla yímmed ṃàya,| | | dìda, kùšla. kùšla, H |60 H| mxéla-la xá| rašòma,| pə́šla ḥakòma.| pə́šla gòra.| (45) wi-má-b-ozə́n ʾə-nàqla?| júlləd baxtàsa ʾísən!| lá-k-iʾən maʾòzən.|62 She took off her clothes, she took off her clothes, she went down [into the water]. She went down [into the water],60 she wants to take the ring, the Mother of the Water came, she hit her with a rašòma,61 she became the king. She became a man. ‘Oh what shall I do now [lit. this time]? There are women’s clothes! I do not know what to do.’62 (46) qə́mla lwišíla júlle dìda| mxéla l- She rose [and] wore her clothes ʾúrxa b-[ʾ]àqle u-dí u-dí u-dí u-dí and started walking [lit. hit the road by legs] and onwards she u-sèla.| la-k-íʾa ma-ʾòza,| ta-máni came. She does not know what to ʾáza ʾámra ʾána ḥakòma-wán.| ta- do, to whom would she go [and] say ‘I am the king’? To whom máni ʾámra ʾána bax-ḥakòmawould she say ‘I am the wife of wán.| the king’? (47) lá-k-iʾa mà-[ʾ]oza,| ʾə́tla tḷá[ha] bnóne mə̀nne. H |63 H H tóv mṭèla, | ʿaxšávH ʾáya b-šoqànna,| sélan kəz-wàzir.| (48) wázir sèlu, ʾànya| də̀d| g-èzi,| gdóqi hə̀nna ṭère. nəšàre. | | | | (49) BA: nəč ̣àre.| | mux-ḥakòma-le wázir, xá-kma júlle sqìle-ʾəlle,| ʾe mahíne, wele- 60 The wazir, they came, those [people] that go [and] catch this, birds. Hunters. BA: Hunters. (50) HM: g-él g-mènxi,| ʾò| xá nàša,| mpíla l-tàm.| She does not know what to do. She has three sons from him.63 Good, she arrived, now we shall leave her, we come [lit. came] to the wazir. HM: He walks, they look. [They see] this, one man, he is like [=he looks like] a king, the wazir, some beautiful clothes he has, and a horse [lit. that horse], he [the wazir] had fallen there [lit. he is fallen there]. See n. 47 above. See n. 52 above. 62 The verbal forms in (45) with which the king refers to himself are masculine. 63 Unlike in (45), where the king is referred to using masculine forms, in (46) and (47) he is referred to using feminine forms. 61 286 (51) là-g-maḥké,| la-hè la-lá,| g-əmríle He does not speak, not ‘yes’ [and] màni-wət?| g-émer là-k-iʾen, wéle not ‘no’, they say to him, ‘who are you?’ He says, ‘I don’t know’, he pṣìʿa.|64 m-èka wét? g-émer là-k- is wounded. ‘Where are you iʾen.| HzikarónH díde zə̀lla.|65 la-k- from?’ He says, ‘I don’t know’. His memory was gone [lit. went]. táxer čù-məndi̇ . He does not remember anything. (52) qə́mlu qam-nablìle,| qam-daréle gó,| ʾe hə̀nna,| gó xastaxàna,| mə́rru ta-dáw…| e dóktor g-émer ʾŏ ̀ h! ʾó xà náša rúwwa-le,| qamxazáxle wele-mpíla mən-mahìne,| msàdərre,| mtàpəl66 ʾə́bbe.| They rose and took him, they put him in a, this, in a hospital, they said to that … eh doctor, he [=one of the hunters] says, ‘Oh! This is a great [=important] man, we saw him [he had] fallen down from a horse, fix him, treat him. (53) mtopə̀lle66 pə́šle gó...| xastaxàna| He treated him, he stayed in the H ʾézeH xá, xá šàta.| g-mbaqríle m- hospital for about one year. They ask him ‘where are you from?’ He èka wét,| g-émer là-k-iʾən,| ʾéka b- says ‘I don’t know’. ‘Where will you go?’ ‘I don’t know’. He àt?| là-k-iʾən,| pə́šle l-tàma.| stayed there. (54) xà,| muxwàsox| profèsor|67 g-émer One, like yourself, a professor,67 says, ‘Come stay with me, I will ysálox68 kə̀sli| b-yà[wə]nnox| give you food [and] drink, I have ʾixàla| štàya,| ʾə́tli šùla,| ʾúzli work [for you], do some work for | | me, do whatever you like.’ He xápča šùla, mád g-ə́bət ʾòz. gsays, ‘all right’. He does not know émer hàwwa.| lá-k-iʾe čù-məndi.| anything. He rose one day, they say there’s wéle ḥakòma, ʾə́tle ṭèra. ḥakóma a king, which has a bird. Their king died. He has a bird which dóhun mə̀tle.| ʾə́tle ṭéra gthey throw. | mandèle. (55) ʾáwa| qə́mle xà-yoma,| g-ə́mri | | 64 The Modern Hebrew root pṣʿ is used here with NENA morphology. Verb in the feminine form, although HzikarónH is masculine. 66 The Modern Hebrew root ṭpl is used here with NENA morphology. Since the historical emphasis of the consonant ṭ is not retained in Modern Hebrew, it is pronounced as t by Ḥabuba. 67 Directed to Professor Khan. 68 Dativus ethicus. 65 287 (56) ʾóha rkúle mahíne dìde,| mahíne He [the wazir] rode his horse, his horse galloped, galloped, díde dhə̀rra, dhə̀rra, dhə̀rra,|69 galloped. Where he had fallen, he ʾə́ka mpə̀lle| mpə́lle xa-gar-xét ʾəl- fell there again. But [when] he fell, nothing happened to him, he tàm.| HʾavalH-mpə̀lle,| la-brélele remembered. čù-məndi,| txə̀rre.| | | (57) wáy!-g-èmer| ʾána wàzir wéli| ké- ‘Wow!’ he says, ‘I was a wazir! le ḥakòma? ʾéka zə̀lle? ʾána pə́šli Where is the king? Where has he gone? I became already old, what H kvàr| mevugàr,| zakèn,H| mà-bwill I say? Would [lit. where amrən?| ʾéka p-šaqláli bàxti? la- would] my wife take me [back]? She wouldn’t take me [back], she k-šaqlàli,| HkvárH la-g-bàli!| ʾána doesn’t love [or: want] me wə́l pə̀šli…| la-g-mhéməna ʾə̀bbi| anymore. Indeed I became … She díwən [=dəd hiwən] ʾána wàzir!| won’t believe me that I am the wazir!’ (58) séle ʾəl-bèsa,| kəz-bàbe,| kəz-daw- He came home, to his father, to his that did such and bábe d-qam-hənnə̀lle,|70 g-emə̀rre,| that father of 70 such for him, he says to him, mà qə́ṣṭa?| g-émer ḥàl| u-qə́ṣṭa ‘What is the story?’, he says, ‘My story [lit. situation and story] is dídi hàdxa wèla.| dídi u-dədthus. Of mine and of the king. The | | ḥakòma. ḥakóma zə́lle b-xá ʾàl, king went to one side, I do not lá-k-iʾen ʾéka zə̀lle,| u-ʾána zə́lli b- know where he went, and I went to another [lit. one] side [=we xà-ʾal.| separated].’ (59) g-émer de-qú sà bròni,| k-taxréten He [the father] says, ‘So go ahead [lit. rise come] my son, do you ʾèka-wət,| go-d-éma bàžer?| gremember where you were?’ He émer hè.| k-taxrə́tte šə́mmed says, ‘Yes.’ ‘Do you remember | | | bèsox, k-iʾə̀tte? g-émer hè. qu-d- the name of your home, do you know it?’ He says ‘Yes.’ ‘So let’s àx| b-ásən ʾə̀mmox.| go [lit. rise that we shall go], I’ll come with you.’ (60) šqə́lle ʾáwa u-báxte, làtle yalúnke,| ʾó pə́šle mux-bròne.| se- 69 He took his wife [lit. he took himself and his wife], he doesn’t have children, he [the wazir] was [lit. became] like a son to him [lit. The Modern Hebrew root dhr is used here with NENA morphology. The irregular root h-nn-l in derived from hə̀nna; see n. 51 above. Sabar, 2002a, p. 151: “to say this and that; to do this and that, have intercourse …”. 70 288 d-áx b-ásən ʾèmmox,| zə́lle ʾə̀mme.| his son]. ‘Let’s go [lit. go that we shall go], I’ll come with you.’ He went with him. He went with him,71 they knocked (61) zə́lle ʾə̀mme,|71 mtoqtə́qlu [b]dàrga,| [m]pə́qla xa-xəddàmta,| – on the door, a maid opened – he has money, he is a wazir, he ʾə́tle pàre,| wázir hìle,| k-šáqəl receives [lit. take] a salary, his màʿaš,| báxte k-šáqla màʿaš,| – g- wife receives [lit. take] the [=his] salary – she [=the maid] tells him əmrále màni-wət ʾàhət?| g-émer ‘Who are you?’, he says, ‘I am the ʾána wàzir wə́n,| ʾó bésa dìdi-le.| wazir, this house is mine.’ (62) g-ə́mra wày!| zə́lla məŕra ta-báxte She says, ‘Huh?!’ She went [and] said to his wife, she says, ‘One g-əmra-xa-šəzàna wəl-sèle, gmadman indeed came, he is émer| ʾána wàzir wə́n,| ʾó bésa saying “I am the wazir, this house | is mine.” ’ dìdi-le. (63) g-ə́mra màʾurre,| máʾurre xázyan She [the wife] says, ‘Show him in, show him in [and] I’ll see what ʾèma šəzàna.| k-xazyá-le la-gmadman [this is].’ She sees him yaʾàle.| [and] she doesn’t know [=recognise] him. (64) g-emə́rra ʾáhat bàxti wát,| šə́mmed bróni, mirza-maḥàmadíle,| šə́mmed bróni xèt,| ʾàḥmadíle,| šə́mmed bràti| fàṭma-le.| ʾàna| ḥàl| u-qə́sta dìdi hádxa-la.| He tells her, ‘You are my wife, the name of my son is MirzaMaḥamad, the name of my other son is Aḥmad, the name of my daughter is Fatma. I, this is my story [lit. my situation and story is thus].’ (65) g-ə́mrale ḥmòl,| tú tamà,| xà ʾála.| She tells him, ‘Wait, sit over there, aside. I’ll take you to the nablánnox kəz-ʾìmam.| hăkanimam. If the imam says that you ʾìmam mə̀rre də[d]| HbeʾemétH are my husband, [you are my] husband, [if] not, [then] not, you ʾá[hə]t gòri wét,| góri, láʾ làʾ| are not my husband.’ lèwət góri.| (66) g-emə́rra Hbəssèder.H| He tells her, ‘OK.’ (67) ḥákoma šíne ṭréle ṭréle ʾáw The king also, he rode and rode that king. He also came. He came, he came72 he arrived home. He ḥakòma,| séle ʾàp-awa.| séle, 71 See n. 47 above. 289 séle72 mṭéle ʾəl bèsa.| séle g-pásxa came, the maid opened the door, he says, ‘I am the king, I … that is dárga xəddàmta,| g-émer ʾána my wife.’ | ḥakóma wə̀n. ʾána… ʾáya bàxtila.| (68) ʾə́lla g-əmrá, lèwan ʾána báxtox,| Indeed she replies, ‘I am not your ʾáhət wət-píša ġèr šəkə́l,| lá-welox wife, you changed [lit. you became a different shape], you hàdxa!| ʾátta-wal pə́šlox ġèr were not like that! Now you | indeed became [of] different this! hə̀nna! ʾána là-gə-mhémenan I do not believe you.’ He tells her, ʾə́bbox.| g-émerra Htòv.H| ‘OK.’ (69) ʾáp-awa zə́lle qam-matùle, ʾéka wàzir,| qam-matwíle xàzre.| He also went, [someone] sat him down where the wazir [was], they sat him down next to him. (70) yóm ʾəròta,| yóm ʾəròta-g-əmri b- ‘Friday, [on] Friday our imam will áse ʾímam dèni.| ímam déni ʾáwa come. Our imam he will decree. He knows. He has prophecy. He b-qàṭeʾ.| k-ìʾe.| ʾə́tle Hnevuʾà.H k- sees whether he is really the king.’ xáza ʾákan d-íle Hbe-ʾemètH ḥakóma.| (71) wáḷḷa k-èse,| ʾímam dóhun yóm ʾərròta, k-xáze bàhlul-íle, ʾaxón | | ḥakóm,| k-xàze ʾàwa-le.| Indeed, their imam comes [on] Friday, he [=the king] sees it is Bahlul, the king’s brother. He [=the king] sees it is him. They tell him, ‘Indeed, you know, our this … our case [lit. trial] we dèni,… mišpát déni qambrought to you. Because we are mesáxla73 kə̀slox.| HkíH là-mṣax.| not able [to decide whether] that | | | [man] is the king [and] this [is ʾòha, ḥakòma-le, ʾó wàzir-ile. the] wazir. You, say [=tell us the ʾàhət| màr,| psóx jəzúka74 b-qúrʿan answer], open a booklet74 in your | | dìdox kan-díle wàzir kan-díle Quran, whether he is the wazir | [and] whether he is the king.’ ḥakòma. (72) g-əmríle wáḷḷa k-iʾèt,| ʾé hə̀nna| |H 72 H See n. 47 above. Verb in the feminine form, although HmišpátH is masculine. This may be because NENA šarìʾəta/šərʿəta ‘trial, judgment’ is feminine. 74 Sabar, 2002a, p. 127: “booklet (of religious or magic nature)”. 73 290 (73) g-émer ʾó wázir-ile u-ʾó ḥakòma- He says, ‘That is the wazir and that is the king, go back to your le,| dʾórun l-bés gyanòxun.| homes.’ (74) qam-nabə́lle ʾáwa l-bèse| u-ʾáwa l-bèse.| He led them, him to his home and him to his home [=he led each one of them to his home]. (75) ʾó bə́r ḥakòma,| də́d wéla bàxte,| That son of the king, that she75 was his wife, that entire kúlla ʾáy seheràne| pə́šla ʿázaya seheràne76 turned into mourning ʾèlle.| g-ṭáʾe báxte zə̀lla,| u-zàʿla| upon him. He is looking for his u-zàʿla| u-,| la šúqle xá dùksa,| híl wife [but] she is gone, and she has disappeared and disappeared ʾamèrika zə́lle!| and … He did not leave [out even] one place, he went all the way to America! (76) čú dúkka lá šúqle híle b-ṭáʾya ʾə̀lla. čú-xxa lá k-ìʾe lé xə́zya bàxta.| | | He did not leave [out even] one place, he is searching for her. No one knows, [no one] had seen a woman. (77) xzélu xá góra ḥakòma| zə̀lle.| They had seen one man, a king. He [already] went [away]. He mṭèle l-d-áy bážer. mṭéle l-d-áy [=the husband] arrived in that city, bàžer,| ʾéka b-àl?| zə́lle ʾə́l hə̀nna,| where should he go? He went to this, to the imam, to the mosque. kəz-ʾìmam,| kəz-jèmaʿ.| | He tells him, ‘My son, what is your request?’ He says, ‘This is my story [lit. my situation and dìdi| hàdxa wéla.| qam-xazéla xá story was thus]. They [=the H fishermen] saw one girl on the baḥuráH rə́š,| bastád Hnàhar,H| river bank, they brought her to qam-meséla ṭàli| u-qam-gorə̀nna| me, and I married her, and I have u-[ʾə]tlí ṭlá[ha] bnóne mə̀nna,| u- three sons from her, and my wife has disappeared!’ zàʿla báxti!| (78) g-emə́rre bròni| má HbakašáH dídox hìla?| g-émer ḥàl| u-qə̀sta| (79) g-émer là záʿla báxtox,| báxtox ḥàl u-qə́sta hàdxa-la, báxtox ḥakòma-la.| ʾátta mnablə́nnox | 75 76 | He says, ‘Your wife has not disappeared, your wife that is her story [lit. the situation and story is thus], your wife is a king, now I Meaning, the king who turned into a woman. See n. 57 above. 291 kə̀sle,| u-, ʾàwa| b-qaṭéʾla šərʿə́ta dìdox.| shall take you to him, and, he will decree [lit. cut] your judgement.’ (80) g-émərre d-àx.| zə́lle qam-nabə́lle.| He tells him, ‘Let’s go.’ He went and led him. He [=the king] sees it k-xáze gòra77 híle,| ʾáwa k-íʾe, is her [=the king’s] husband. He wéle báxta gòra77 híle.| g-əmrále78 [=the king] knows, he was a woman, this is [=was] her kèlu yalúnke dídi?| g-əbànnu!|79 husband. She [=the king] tells him, ‘Where are my children? I want79 them!’ (81) g-émer yalúnkəd mà?| ʾa[he]tgòra wə́t!| ṃàṭo| yalúnke mesə́nnu-làx?|80 He [=the husband] says, ‘Children of what? You are a man! How will I bring you80 the children?’ (82) g-ə́mra ḥàl| u-qə́sta dídi hàdxa-la.| She [=the king] says, ‘This is my ʾána| mpə̀lla| ʾasə̀qsa| dìdi,| hádxa story [lit. my situation and story is thus]. I, my ring fell, I twisted it qam-mazvərànna| mpə́la go[around my finger] like that, it fell | | ṃàya, séla yímmed ṃàya mxélali into the water, the Mother of the Water came, struck me with a xá rašòma| qam-ʾózali HbaḥùraH.| rašòma82 [and] turned [lit. made] | | qam-gorànnox, ʾiláha wə́lleli me into a girl. I married you, God |81 gave me three sons from you.’ ṭlà[ha] bnóne mə́nnox. (83) ʾúzlox seheràne,| sèli,| ʾéni nẓə́rra- You made a seheràne,83 I came, my eye caught a glace of my ring, [ʾe]l ʾasə́qsa dìdi,| ʾasə́qsa dədit is a ring of diamond, of, H yahalòmH híla,| də́d,| jawàhar.| diamond. I bent down in order [lit. I want] (84) kə́pli g-ə́ban šaqlànna,| séla ʾày | | to take it, that Mother of the yímmed ṃáya mxélali xá rašòma Water came, struck me with a qam-ʾozali xá-gar xə́t gòra.|84 rašòma [and] turned [lit. made] me again into a man. 77 The feminine possessive pronoun -a refers to the king. Feminine verbal form. 79 This verb, uttered by the king, is in the feminine form. 80 Feminine pronoun. 81 All forms in (82) referring to the king are feminine. 82 See n. 52 above. 83 See n. 57 above. 84 All forms in (84) referring to the king are feminine. 78 292 (85) ʾána ḥakòma-wən,| k-xázət ʾàxxa.| I am a king, you see here. Now, I want85 my children, whatever the ʾe-náqla g-éban85 yalúnke dìdi,| imam says [lit. said]. He says [lit. mád mə́rre ʾìman,| mə́rre ṭáli-ilu,| said] they are for me or they are H H for you [=he will decree either]. ʾo -ṭàlox hílu.| (86) g-emə́rra HgamH-ʾà[h]at zə́llax| H gamH-yalùnke yawə̀nnu-lax?| ʾilà[ha]-la qabə́lla mə́nnax.| (87) sèle-kəz ʾímam| ʾímam g-èmer,| ʾá[h]at ʾə́tlax yalùnke,| ʾàwa| – yalúnke dìde hílu.| ʾàni| yálunke díde ṭàle,| yalúnke dídax ṭàlax,| sí bròni,| ʾílaha-ha[w]e ʾə̀mmox,| sí gór xa-xèta.|86 (88) há ʾèha wéla,| ʾáwa zə́lle l-bèse,| ʾó séle l-bèse. zéhu g-ə́bet xaxèt?| |H H (89) BA: kúd šmiʾále xà[y]e…| He tell her [=the king], ‘First [lit. also] you went away, and [now you want that] I will give you the children as well?! God will not permit this! [lit. God will not accept it from you; =this is a violation of the divine justice].’ He came to the imam, the imam says, ‘You [=the king] [already] have children, he [=the prince] – those are his children. They, his children are for him [=should stay with him], your children are for you. Go my son, may God be with you, go and marry another.’86 Here, this is it, he went to his home, [and the other] one went to his home. That’s it, would you like another one [=story]? BA: [May] whoever has heard it live … (90) HM: …xà[y]e,| kud-là šmiʾále…|87 HM: … live, whoever has not heard it … [also live].87 Would g-ə́bet xa-xèt?| you like another one? References Aarne, A. A. and Thompson, S., 1961, The Types of the Folktale: A Classification and Bibliography, 2nd edn, Helsinki: Suomalainen Tiedeakatemia. Alon, J. and Meehan, C., 1979, “The boy whose tunic stuck to him: a folktale in the Jewish Neo-Aramaic dialect of Zakho (Iraqi Kurdistan)”, Israel Oriental Studies 9, pp. 174–203. 85 Feminine verbal form. All forms in (86) and (87) referring to the king are feminine. 87 A common ending formula in NENA folk-tales. 86 293 Aloni, O., 2014, The Neo-Aramaic Speaking Jewish Community of Zakho – A Survey of the Oral Culture, Saarbrücken: Lambert. Avinery, I., 1978, “Folktale in the Neo-Aramaic dialect of the Jews of Zakho”, Journal of the American Oriental Society 98, pp. 92–96. ———, 1988, Ha-Niv Ha-ʿArami Šel Yehude Zaxo: Ṭeqstim Be-Tseruf Targum ʿIvri, Mavo U-Millon, Jerusalem: The Israel Academy of Sciences and Humanities. Ben-Amos, D., 1980, “The concept of motif in folklore”, in V. J. Newall (ed.), Folklore Studies in the Twentieth Century: Proceedings of the Centenary Conference of the Folklore Society, Woodbridge: Brewer and Totowa: Rowman and Littlefield, pp. 17–36. ———, 1995, “Are there any motifs in folklore?”, in F. Trommler (ed.), Thematics Reconsidered: Essays in Honor of Horst S. Daemmrich, Amsterdam: Rodopi, pp. 71–85. Ben-Yaacob, A., 1981, Qhilot Yehudey Kurdistan (Hebrew), 2nd edn, Jerusalem: Qiryat Sefer. Brauer, E., 1993, The Jews of Kurdistan, R. Patai (ed., comp.), Detroit: Wayne State University Press. Brown, W. N., 1927, “Change of sex as a Hindu story motif”, Journal of the American Oriental Society 47, pp. 3–24. Cohen, E., 2012, The Syntax of Neo-Aramaic: The Jewish Dialect of Zakho, Piscataway: Gorgias Press. Drory, R., 1994, “Three attempts to legitimize fiction in classical Arabic literature”, Jerusalem Studies in Arabic and Islam 18, pp. 146–164. Dundes, A., 1962, “From etic to emic units in the structural study of folktales”, The Journal of American Folklore 75, pp. 95–105. ———, 1964, “Texture, text, and context”, Southern Folklore Quarterly 28, pp. 251– 265. El-Shamy, H. M., 1980, Folktales of Egypt, Chicago: The University of Chicago Press. ———, 2004, Types of the Folktale in the Arab World: A Demographically Oriented Tale-Type Index, Bloomington: Indiana University Press. Fellman, K., 1978, Qaṭalog Ha-Ttiʿud Ha-Mmuqlat (Hebrew), Jerusalem: Hebrew University of Jerusalem. Ganguli, K. M. (trans.), c1900, The Mahabharata of Krishna-Dwaipayana Vyasa Translated into English Prose from the Original Sanskrit Text, Calcutta: Oriental Publishing Co. Garber, M., 1992, Vested Interests: Cross-Dressing and Cultural Anxiety, New York: Routledge. Gavish, H., 2010, Unwitting Zionists: The Jewish Community of Zakho in Iraqi Kurdistan, Detroit: Wayne State University Press. Ginzberg, L., 1925, The Legends of the Jews, vol. V, Philadelphia: Jewish Publication Society of America. Khan, G., 2007, “The North-Eastern Neo-Aramaic dialects”, Journal of Semitic Studies 52/1, pp. 1–20. Marzolph, U., Leeuwen, R. and Wassouf, H., 2004, The Arabian Nights Encyclopedia, Santa Barbara: ABC-CLIO. Meiri, S., 2011, Any Sex You Can Do, I Can Do Better: Cross-Gender in Narrative Cinema, Tel Aviv: Hakibbutz Hameuchad. Neuman (Noy), D., 1954, Motif-Index of Talmudic-Midrashic Literature (unpubl. diss., Indiana University). Penzer, N. M., 1927, The Ocean of Story: Being C. H. Tawney’s Translation of Somadeva’s Kathā Sarit Sāgara, vol. VII, London: Chas. J. Sawyer. 294 Polotsky, H. J., 1967, “Eastern Neo-Aramaic: Urmia and Zakho”, in F. Rosenthal (ed.), An Aramaic Handbook, Wiesbaden: Harrassowitz, part II/1, pp. 69–77, part II/2, pp. 97–111. Rivlin, J. J., 1959, Širat Yehude Ha-Ttargum: Pirkqe ʿAlila U-Gevura Be-Fi Yehude Kurdistan, Jerusalem: Mosad Byaliḳ. Rosen, T., 2003, Unveiling Eve: Reading Gender in Medieval Hebrew Literature, Philadelphia: University of Pennsylvania Press. Sabar, Y., 1975, “The impact of Israeli Hebrew on the Neo-Aramaic dialect of the Kurdish Jews of Zakho: a case of language shift”, Hebrew Union College Annual 46, pp. 489–508. ———, 1982, The Folk Literature of the Kurdistani Jews: An Anthology, New Haven: Yale University Press. ———, 2002a, A Jewish Neo-Aramaic Dictionary, Dialects of Amidya, Dihok, Nerwa and Zakho, Northeastern Iraq, Wiesbaden: Harrassowitz. ———, 2002b, “The story of the brothers Ali and Amar in the Jewish dialect of Zakho, based on Albert Socin’s text from 1870, transcribed anew as if it were told ca. 1950, and emphasizing the linguistic changes that occurred in the dialect since Socin’s time”, in W. Arnold and H. Bobzin (eds), “Sprich doch mit deinen Knechten aramäisch, wir verstehen es!”: 60 Beiträge zur Semitistik; Festschrift für Otto Jastrow zum 60. Geburtstag, Wiesbaden: Harrassowitz, pp. 613–627. ———, 2005, “Yona Gabbay, a Jewish peddler’s life story from Iraqi Kurdistan: as narrated by him in his Jewish Neo-Aramaic dialect of Zakho (four episodes)”, Mediterranean Language Review 16, pp. 167–220. Shilo, V., 2014, Lishana Deni – A Bilingual Anthology of Folktales of the Jewish Community of Zakho (Hebrew), O. Aloni (ed.), Jerusalem: Minerva Press. Socin, A., 1882, Die neuaramäischen Dialekte von Urmia bis Mosul: Texte und Übersetzung, Tübingen: H. Laupp. Thompson, S., 1955–1958, Motif-Index of Folk-Literature: A Classification of Narrative Elements in Folktales, Ballads, Myths, Fables, Mediaeval Romances, Exempla, Fabliaux, Jest-Books, and Local Legends, Bloomington: Indiana University Press. ——— and Balys, J., 1958, The Oral Tales of India, Bloomington: Indiana University Press. Uther, H.-J., 2004, The Types of International Folktales: A Classification and Bibliography, Helsinki: Suomalainen Tiedeakatemia. Walker, W. S. and Uysal, A. E., 1966, Tales Alive in Turkey, Cambridge, MA: Harvard University Press. ———, 1992, More Tales Alive in Turkey, Lubbock: Texas Tech University Press. Zaken, M., 1997, “The fate of inventors – a folktale in the Neo-Aramaic of the Jews of Zakho” (Hebrew), in M. Bar-Asher (ed.), Massorot 9–11: Studies in Language Traditions and Jewish Languages, Volume in Honor of Gideon Goldenberg, Jerusalem: Department of Hebrew Language, Hebrew University of Jerusalem, pp. 383–395. ———, 2007, Jewish Subjects and Their Tribal Chieftains, Leiden: Brill. 295 PART 2 Texts, Scribes and the Making of Books and Documents Crossing Palaeographical Borders: Bi-Alphabetical Scribes and the Development of Hebrew Script – The Case of the Maghrebi Cursive JUDITH OLSZOWY-SCHLANGER École Pratique des Hautes Études-PSL and Institut de Recherche et d’Histoire des Textes-CNRS In the middle of the 11th century in Jerusalem, Abū al-Faraj Furqān ibn Assad or, to use his Hebrew name, Joshua ben Judah, a scholar of fame and the leader of the Karaite house of study, faced a dilemma.1 A wealthy patron, most probably Daʾūd ibn ʿImrān, an influential merchant from Fustat, ordered from Joshua ben Judah a commentary on the Pentateuch, but omitted to specify whether he wished his book, composed in Arabic, to be written in Hebrew or in Arabic characters. Joshua ben Judah mentioned this problem to an anonymous middleman who was by all evidence in charge of providing the Egyptian patron and keen bibliophile with books from Jerusalem. The bookish go-between wrote to the patron explaining that the question of the choice of script was hindering Joshua ben Judah from sending off the finished work. His letter, sent from Jerusalem and written in the Arabic language using the Arabic alphabet, must have reached its destination. It was recycled in the 12th century, this time by a Rabbanite Jew who wrote a Hebrew-Arabic dictionary of the Mishnah in Hebrew characters on the blank verso and between the lines of the original Arabic letter. After a long circulation, the letter was deposited in the Genizah, to be finally discovered, studied and published by Professor Geoffrey Khan,2 to whom this essay is dedicated with gratitude and affection. What seems to be just an anecdote in fact has serious bearings on the question of script used by Jewish scribes in the medieval East. The discovery and publication of the letter by Geoffrey Khan in 1993 has provided new elements for our understanding of the possible motivations behind the choice of Hebrew 1 Some ideas of this paper were included into the John Coffin Memorial Lecture presented at King’s College, University of London, on 27 May 2017. I thank Miriam Wagner and Ronny Vollandt for their comments and suggestions. All images reproduced by kind permission of the Syndics of Cambridge University Library. 2 T-S K 25.230, in Khan, 1993a. 299 or Arabic script to copy Jewish books. More importantly, this letter shows that Jewish scribes possessed the skills to write with equal ease in Hebrew and in Arabic. The purpose of this present essay is to reflect on the possibility that training in and simultaneous practice of two writing systems had implications on the development of Hebrew script itself, and should be considered as a factor in palaeographical research. The impact of bi-alphabetism on the creation and development of specific subtypes or styles of Hebrew script will be illustrated through the example of the specific script style used from the beginning of the 11th century by the professional trading network of Maghrebi merchants. 1 Jewish scribes writing in Arabic script It is well known that medieval Jews spoke the vernacular languages of their non-Jewish environment as their mother tongue, reserving Hebrew for education, prayer and reading. In contrast to their total acculturation to the spoken linguistic environment, the written production of the Jews was usually conveyed in Hebrew script. However, the study of the evidence provided by the Cairo Genizah suggests that the scribal reality in Eastern communities was much more complex than is suggested by this clear-cut oral/written distinction. The letter concerning Joshua ben Judah’s commentary reveals an indiscriminate use of Hebrew and Arabic scripts by the same scribe. The choice does not depend on the scribe’s skill or the lack of it; it is left to the preference of the reader, in this case a wealthy patron. Geoffrey Khan explains that this indiscriminate use of Hebrew and Arabic scripts was particularly frequent among the Karaites, who even transliterated the Hebrew Bible into Arabic characters.3 More open to the literary influences of the Arab majority culture4 and adepts of the Aristotelian view of script as an arbitrary – and thus interchangeable – reflection of the linguistic reality, the Karaites were indeed not averse to writing in Arabic script.5 Although Rabbanite Jews displayed less readiness to use Arabic script, knowledge of the Arabic script was nonetheless also common among mainstream Jews. Genizah documents indeed show that medieval Jewish scribes, Karaites and Rabbanites alike, functioned not only in a situation of linguistic diglossia but also within a bi-alphabetical graphic culture. They learned Arabic script alongside Hebrew, and used it for reading Arabic books as well as for writing literary and pragmatic texts. This rubbing shoulders, or rather calami, with non-Jewish writings, scribes and scripts had a lasting impact on stylistic aspects of the Hebrew script itself. 3 Khan, 1993a, pp. 138–140. For the Karaite Bibles in Arabic script, see especially Hoerning, 1889; Khan, 1990. 4 Drory, 2000a, pp. 135–136; Drory, 2000b, pp. 107–108. 5 Ben Shammai, 1982; Khan, 1997; Olszowy-Schlanger, 1997. 300 The first issue to consider is the capacity and willingness of Jewish scribes to employ non-Jewish scripts, and the extent of their use. Medieval Jewish communities, or at least their male component, were literate.6 Urbanised, book-centred in their religious rituals, involved in bureaucratically managed economic and legal activities, for medieval Jews reading and writing skills were simply vital. They inherited from antiquity the tradition of schooling for young boys which made basic literacy, be it scientific, religious or pragmatic, quite widespread across the population. Mentions of schools and teachers – either privately hired or employed by the community as a whole – abound in medieval sources. The teaching methods can be reconstructed from Rabbinic texts as well as from the hundreds of extant children’s primers and writing exercises found in the Cairo Genizah. The pedagogical methods inherited from classical antiquity insisted on the recognition of the forms of letters. This was achieved by learning their shape and ductus (the number, direction and mutual relationship of the pen movements involved in producing a letter).7 It is true that the main purpose of the Jewish elementary writing education was to acquire skills in reading and writing Hebrew, and more specifically in reading and writing the square, calligraphic register of the script, the one found in Bible scrolls and display codices. However, there is ample evidence that Jewish individuals also learned to read and write in Arabic script. At the turn of the 10th and 11th centuries, Hai ben Sherira Gaon is explicit about allowing the teaching of the Arabic alphabet (in addition, of course, to the study of the Torah) in Jewish primary schools.8 And indeed, the Cairo Genizah shows that at least some Jewish children learned Arabic writing in parallel with attending a Hebrew school. In a letter studied by Shlomo Dov Goitein, a father begs his son’s Jewish teacher’s indulgence for delays caused by overlapping Arabic lessons,9 and a large number of writing exercises in Arabic script, including learning the alphabet, are preserved in the Genizah, sometimes with Hebrew equivalents (see figure 1). Members of the Jewish elite were aware that Arabic literacy skills were essential for climbing social ladders. On the Western side of the Mediterranean, Judah ibn Tibbon, a Granada-born physician and translator of Jewish philosophical works in 12th century Lunel, admonishes his son Samuel in his unflattering ethical will: 6 For a general overview, see Reif, 1990. There are different scholarly opinions as to the true level of Jewish literacy in the Middle Ages. The actual percentage of literate Jews in medieval communities is impossible to establish, as is, incidentally, the precise number of Jews in general. In my understanding, a literate society is not necessarily a society in which everyone or nearly everyone has access to and masters literacy and numeracy skills, but rather a society in which such skills constitute a necessary foundation for economic and religious activities. 7 Goitein, 1962, especially pp. 35–41; Goitein, 1967–1988, particularly vol. II, pp. 74–76; Narkiss, 1972; Olszowy-Schlanger, 2003. 8 Quoted by Assaf, 1931, pp. 4–5. 9 Goitein, 1962, p. 35. 301 Figure 1. T-S K 5.31 r. and v.: Children’s exercises in Hebrew and Arabic script; the Arabic writing seems more fluent than the Hebrew Also in the matter of Arabic writing that you started to learn seven years ago, I had to encourage you constantly, but you have never listened to my advice. But you know well that the greatest among our people have not achieved their greatness and elevated status but thanks to the Arabic writing. 10 This doting father insisted on his son learning Arabic calligraphy, in addition to his lessons in Hebrew calligraphy for which the father paid the considerable sum of 30 gold dinars a year. It seems that, unlike Hebrew whose study required a teacher, for learning Arabic calligraphy Samuel was supposed to train by copying Arabic books. Judah also encouraged Samuel to read Arabic books, both translation of the Bible and scientific works that were accessible in Judah’s private library. Indeed, it was not uncommon for Jewish bibliophiles to own and study books in Arabic. Fragments of works such as Kalila wa-Dimna (see T-S Ar. 51.60), mentions of copies of the Arabian Nights11 and fragments of Arabic scientific works were found in the Cairo Genizah.12 Even in Christian Spain, Arabic books were commissioned or even copied by Jewish individuals.13 Arabic literacy of the Jews in Dār al-Islām went much further than a mere acquaintance with the Arabic script and reading works of Arabic literature in their original graphic form: Jewish scribes copied manuscripts in both Hebrew 10 Abrahams, 1976, p. 84. Goitein, 1959. 12 Khan, 1986. 13 See van Koningsveld, 1992, especially pp. 89–93. 11 302 and Arabic scripts. Evidently, there must have been a demand for Jewish books in Arabic script, and therefore we can conclude that there was a community of competent readers, some of whom preferred to read the Arabic rather than the Hebrew graphic forms. As we saw, the choice of Hebrew or Arabic script was indifferent on the production end, and depended only on the personal choice of the book’s prime reader – the ordering and paying patron. This arbitrary use of Hebrew or Arabic script as an individual’s choice is also attested in documentary Genizah writings.14 As was mentioned, Karaite scribes produced manuscripts of the Hebrew Bible transliterated into Arabic script (but retaining their Tiberian vowels) to facilitate its study, and a number of Jewish books written in the Arabic language have been preserved in manuscripts in either Hebrew or Arabic script. As shown by Geoffrey Khan, the Karaites used the two scripts indiscriminately between the 11th and 13th centuries (with a preference for Hebrew script from the 14th century onwards). For example, the aforementioned commentary on the Pentateuch of Joshua ben Judah has been preserved in at least five manuscripts in Arabic script and seven in Hebrew script.15 The use of Arabic script was, however, by no means restricted to Karaite book production. Several books written in the Arabic language by Rabbanite authors, in particular scientific and medical compositions, were rendered in both Hebrew and Arabic scripts. Although some specifically Jewish texts, such as early Bible translations and glossaries, were transmitted in Arabic script in addition to Hebrew,16 the former was primarily used in writings addressed to a broader public that included non-Jewish readers, such as medical and philosophical texts. For example, Masarjawayh (c700) wrote his medical tractates in Arabic letters.17 The works of another pioneer of Judaeo-Arabic literature, David al-Muqammaṣ, active in the 9th century, also exhibit clear indications of being originally composed in Arabic script.18 Similarly, Isaac Israeli, who was active around the turn of the 10th century, wrote his philosophical and medical compositions in Arabic characters.19 His pupil Dunash b. Tamim (first half of the 10th century) continued this custom in a treatise on the armillary sphere.20 A particularly interesting case is the manuscript transmission of the medical and philosophical works of Moses Maimonides. There is no question that Maimonides, with his official functions as the court physician to Saladin’s In a private letter, a certain Mūsā b. Jacob instructs the addressee to reply in Hebrew script because he prefers it (wa-yakūn al-khaṭṭ ʿibrānī); see CUL Or. 1080 J 42, in Gil, 1997, vol. III, p. 304. I thank Ronny Vollandt for this reference. 15 For this, and for the use of Hebrew and Arabic scripts for copies of the works of other Karaite authors, see Khan, 1993a, p. 137. 16 See Zucker, 1959, p. 50; Hopkins, 2002, pp. 369–374. 17 Gil, 2004, pp. 297–298. 18 See Stroumsa 1989, p. 36. 19 See Altmann and Stern, 1958, pp. 3–4. 20 Stern 1954–1956, pp. 373–382. 14 303 family, was perfectly acquainted with Arabic and could use it in writing. His medical works in particular were of universal interest, and they are found in manuscripts in both Hebrew and Arabic scripts. Maimonides’s Tractate on Poisons and Their Antidotes is preserved in at least six manuscripts in Arabic script and four in Hebrew script.21 His multilingual herbalist encyclopaedia, Book of the Names of Drugs, is preserved only in Arabic script, in a manuscript copied by a famous 13th century Muslim botanist, Ibn al-Baytar.22 Moreover, the entries of this encyclopaedia are organised according to the order of the Arabic alphabet. It is possible that Maimonides preferred to write in Hebrew script; the autograph drafts of his medical works found in the Genizah are all in Hebrew script,23 and this includes the drafts of works written for his Muslim patrons. For example, T-S Ar. 44.79 and T-S Misc. 34.24 are fragments of Maimonides’s draft of Fī l-Jimāʿ, a tractate on sexual intercourse, aphrodisiacs and healthy diet written for a nephew of Saladin, the woman-loving AlMuẓaffar ʿUmar ibn Nūr ad-Din, who asked his trusted physician to provide him with a guide to help increase his sexual potency.24 None of the autographs is in Arabic script, although it has been suggested that Maimonides sometimes used Arabic script for corrections and marginal glosses – in a Genizah fragment of the Epitomes of Galen (T-S Ar. 21.112 and Gaster 1019) studied by Simon Hopkins, marginal additions which seem to be by Maimonides’s hand are in Arabic script.25 In any case, while Maimonides may have preferred to draft his works in Hebrew script, at least those destined for Muslim patrons or the general Jewish and non-Jewish public must have been immediately copied in Arabic script. Even when such fair copies were destined for non-Jews, those who prepared the initial master copy must have been Jewish scribes well acquainted with Hebrew script, and able to read the spidery and ligatured informal handwriting of the master. A manuscript tradition in both the Hebrew and the Arabic alphabets is also attested for Maimonides’s works written explicitly for a Jewish audience. The famous Guide for the Perplexed, the author’s draft of which, in Hebrew script, has been found among Genizah fragments,26 was rapidly circulating in both scripts. This is particularly important because, according to the Iraqi philosopher and physician ʿAbd al-Laṭīf al-Baghdādī, who visited Maimonides in Egypt in 1191, the author forbade and cursed in anticipation those who would be tempted to transliterate his book into Arabic script.27 However, the prohi21 Rosner, 1998, p. 32. Meyerhof, 1940, pp. lvii–lxi; Ferre, 2009, p. 20. 23 Hopkins, 2005, p. 91. 24 Stern, 1956, vol. III, pp. 17–21; Backer and Polliack, 2001, p. 416; Hopkins, 2005, p. 93. 25 Hopkins, 1994. 26 The fragments from Maimonides’s draft of the Guide have been published and studied by several scholars; see the list of fragments and their publication in Sirat and Di Donato, 2011, pp. 70–71. See also Sirat, 2014. 27 Munk, 1842, p. 27 n. 1; Hopkins, 2005, p. 91. 22 304 bition was short lived (or, less likely, al-Baghdādī learned the Hebrew alphabet28), since the Iraqi scholar was able to read the book soon after it was finished. It seems that, whether Maimonides accepted it or not, copies in Arabic characters circulated freely and some are still extant.29 When Samuel ibn Tibbon undertook the translation of the Guide into Hebrew in Provence, he wrote to Maimonides in Egypt complaining that the model that reached him and served for the first translation differed from the original author’s version. He claimed that his model was a retro-transliteration back into Hebrew characters from a copy in Arabic script. This script manipulation resulted in many errors and misinterpretations.30 Maimonides’s preference in writing his drafts may serve as an example that Jewish scribes on the whole preferred Hebrew script. From the letter of Samuel ibn Tibbon we gather that it was the Arabic script used for a copy of the Guide that was to blame for the text’s corruption and mistakes. A similar criticism of the Arabic script as responsible for errors in the transmission of ancient knowledge – this time of the astrological compendium of Ptolemy, the Tetrabiblos – appears in the 13th century encyclopaedia, the Midrash haHokhmah (‘the lesson of wisdom’) by Judah ben Solomon ha-Cohen, a Spanish Jew who probably worked in Sicily at the time of Frederick II. Judah ben Solomon stressed the superiority of the Hebrew letters because he believed them to reflect the movement of the planets and to convey esoteric and scientific truth. But it was the graphic nature of the Arabic script, and in particular the absence in the manuscripts of the diacritical points to distinguish such letters as bāʾ, yāʾ, nūn, tāʾ and ṯāʾ, that he found responsible for the errors in text transmission: For the Arabic script leads to falsities. It has many letters whose form is identical, and which are distinguished only by dots. Scribes often err in placing the dots and this leads to misunderstandings.31 Despite this evident esteem for Hebrew, scribes were often so familiar with both alphabets that Hebrew and Arabic can be found side by side in the same manuscript, on the same page and line. Scribes writing in Arabic script could easily switch to Hebrew script, for example to write a biblical quotation within the Arabic body of the text; for instance, several manuscripts of the Guide in Arabic characters quote the Bible in Hebrew characters. Although some scholars argue that these bilingual and bi-alphabetical manuscripts were destined for study sessions between Jewish and non-Jewish intellectuals, it seems more 28 As suggested by Hopkins, 2005, p. 91. Vajda, 1960. 30 Lichtenberg, 1859, part II, p. 27a, col. 2: ‫כי הראשונה כאשר הודעתי אל כבוד אדוננו נראה ממנה שנכצבה‬ ‫מספר כתוב ערבי או מספר שנכתב הוא מספר כתוב ערבי ולזה רבו טעיותיו כאשר יראה בה אדוננו‬. See Vajda, 1960; Hopkins, 2005, p. 91; Sirat and Di Donato, 2011, p. 60. 31 Oxford, Bodleian Library, MS Mich. 551, ff. 178 v.–179 r., quoted in Langerman, 2000, p. 380. 29 305 likely that they were destined for Jewish readers who were used to reading in Arabic script, but preferred to keep the Biblical text in its own characters. In any case, if we take as an example T-S Ar. 18(1).141,32 a fragment of the Guide, the Arabic and Hebrew parts were copied by the same scribe, with the same ink, and by all evidence at the same time and place. There is no question that the scribe writing in Arabic left spaces to be filled in later with Hebrew script. The writing flows easily and the passage from one script to another is so fluent and natural that, at first glance, it is difficult to distinguish the quotations in Hebrew script (see figure 2). Similar instances of a natural flow of writing in Hebrew and Arabic can be found in many legal documents and letters.33 Figure 2. T-S Ar. 18(1).141 (13th century, Egypt): a page from a manuscript of the Guide of the Perplexed, in Arabic script with quotations in Hebrew script 32 Several leaves from the same manuscript of the Guide in Arabic characters with quotations in Hebrew have been found in the Cairo Genizah: T-S NS 306.252 and T-S Ar. 18(1).141. See Hopkins, 1985, p. 713; Hopkins, 2005, p. 92; Khan, 1990, p. 2; Sirat and Di Donato, 2011, p. 61. 33 For example, the 12th century booksellers’ records in BL Or. 10656.5 (see Allony, 2006, no. 78). 306 2 The impact of bi-alphabetism on Hebrew script: A cognitive approach The fluency of use of Arabic and Hebrew scripts in the same text indicates that its scribes mastered both writing systems, and probably did so at an early stage of their writing education. Just like people who learn to speak a language early or late in life can be often recognised by their accent, early writing education has an impact on the way scribes trace the graphic forms. There are native and non-native writers just as there are native and non-native speakers. The fluent use of both Hebrew and Arabic scripts indicates that Jewish scribal culture was not segregated from that of their neighbours. It has been argued that such a shared scribal culture had a lasting impact on the Hebrew script, on the formation of various local script-types and styles and on the visual aspects of Jewish books and documents in general. However, the precise ways and mechanisms through which non-Jewish scripts impacted on the development of Hebrew script have not yet been sufficiently studied by Hebrew palaeographers. It was Colette Sirat who first observed and studied the visual impact of non-Hebrew scripts on Jewish books and scribes. In her book Écriture et civilisation, published in 1976, she noted that the Hebrew alphabet, just like other material characteristics of Jewish books, evolved in parallel with the contemporary scripts of the non-Jewish environment, so that the style of non-Jewish script is reflected in Hebrew script.34 This impact of non-Jewish scribal cultures on Hebrew books and script was subsequently studied by Malachi Beit-Arié in his Panizzi Lectures series in London, in 1993. He pointed out that Hebrew manuscripts produced in various places of the Muslim or Latin world differ in their graphic and technological features, and that they are “moulded by the different places where they were made”.35 The sharing of techniques of parchment making, ink, composition of quires, patterns of layout, and so on made these manuscripts into real “cross cultural agents”. Beit-Arié also discussed the influence of non-Jewish scripts on the appearance of Hebrew script, but warned that the possibility that contemporarily shared or similar writing styles and techniques of production in different cultures of the same area do not necessarily mean intellectual scribal borrowings, but might have been independent outcomes of common aesthetic and technical impulses of the Zeitgeist. 36 The influence of the writing implement – calamus for Hebrew writings in the Orient, Spain and parts of Italy, and quill in Northern Europe – can also be seen as responsible for some specific features, where similarities with nonJewish scripts result from being executed with the same instrument. 34 Sirat, 1976, p. 4; my translation. Beit-Arié, 1993, p. 7. 36 Beit-Arié, 1993, p. 13. 35 307 However it would seem that the impact on Hebrew script not only results from watching and imitating the aesthetics of the contemporary visual culture, nor of the external aspects of the page or the outer shapes of letters. As stated above, Jewish scribes were often trained in early life to be able to write in Arabic script. It seems to me that it is the memory of the gestures necessary to trace a letter as well as the economy and ergonomics of these movements which contribute to the fact that Hebrew script can be so evidently similar to the non-Hebrew one. It is this intimate relationship – defined for Latin by Jean Mallon as being between the hand and the eye in the production of the graphic act (“le concours de la main et de l’oeil dans l’acte graphique”37) – that accounts for the production of forms of Hebrew script which echo the features of the non-Hebrew script of the majority culture. Observed empirically by palaeographers, the importance of body memory or sensorimotor memory has been studied by anthropologists and cognitive psychologists who include writing among those human activities which are executed with a high degree of automatism. In recent years, the cerebral function related to the sensorimotor memory and the graphic act have received renewed attention from cognitive psychologists concerned in particular about the negative impact of the computer keyboard on the acquisition of literacy skills among young children. Scholars such as Marieke Longcamp and JeanLuc Velay, working on the impact of graphic motricity on the process of reading, have observed that during the act of writing and the act of recognising letters (reading), the neuronal information which determines the order of writing strokes composing the characters is encoded – ‘memorised’ – in a specific zone in the brain (the premotor cortex of the left hemisphere for right-handed people, and the right hemisphere for left-handed people). Letters are represented in the brain not only through vision but also through the gesture or the mental, subconscious simulation of the movements that one makes when writing. This sensorimotor memory is activated in the process of writing, just as it is when someone forgets a phone number or the spelling of a word and retrieves it by mechanically dialling or writing, trusting that the hand ‘remembers’ it better than the conscious efforts to recall. The extensive neural network responsible for the complex sensorimotor memory involved in writing and reading is usually established in early education.38 The gestures used in writing, the ductus of the letters and their groups are acquired through practice and repetition. When we take, for example, T-S 12.710 (see figure 3), we realise that the person who reused this 10th century letter as scrap paper for training in the Arabic script repeated over and over again the same conjunction of letters, the same gestures and their sequence, so that the chain of movements would become fluent and automatic. 37 38 Mallon, 1952, p. 33. Longcamp, 2003; Velay and Longcamp, 2005. 308 Figure 3. T-S 12.710: Arabic writing trials on a reused Hebrew document; the word ‫ ذلك‬is repeated nine times Automatisation and reliance on sensorimotor memory is stronger when the writer is at ease with the action of writing. A proficient writer concentrates on the meaning of the message that he or she wants to convey, rather than on the graphic act itself. It is as if the hand is guided by an ‘automatic pilot’ while the conscious brain activity formulates the message to be conveyed in the writing. The best calligraphers, those able to reproduce perfectly their models, are those who are able to focus on the very act of writing, on shaping the letters, instead of analysing in real time the text’s message. On the opposite end of the scale, the less strictly controlled automatic writing allows the scribe’s mind to dwell on the meaning and gives his or her hand the freedom to trace the movements it ‘prefers’.39 Cognitive scientists agree that this subconscious 39 In palaeographical terms, calligraphically executed, controlled writings are closer to their model whereas informal writings stray the most from the models and contain the imprint of the scribe’s idiosyncratic features. The discrepancy between the stereotype models and a scribe’s individuality has been analysed by palaeographers (for Hebrew script in particular, by Malachi Beit-Arié, 1990) who distinguish accordingly between the typological entity of the ‘script’, shared by a group of scribes trained to follow the same models in the same chronological, geographical and cultural context, and the script’s idiosyncratic realisation by individual scribes, referred to as ‘hand’ or ‘handwriting’. The difference between the ‘ideal’ model and its actual realisation was defined by Geoffrey Khan (1993b), in terms inspired by Chomskian linguistics, as ‘competence’ versus ‘performance’. The actual ‘performance’ depends on many factors: on the individual’s perception of the model (visual perception and comprehension of the image by the central nervous system) and his/her capacity to reproduce what he/she perceives (motor 309 state based on memorised gestures lasts for a few seconds, with consciousness retaking control intermittently. What happens, then, when scribes used to writing in Arabic write in Hebrew script? It is likely that the scribes’ sensorimotor memories lead them to reproduce subconsciously the sequence of movements that they have learned and employ with frequency when they write in Arabic script. Indeed, bi-alphabetical Jewish scribes may prefer in some circumstances the fluent and ligatured movements of Arabic script because they are used to tracing such forms, having acquired them through long practice and repetition. The recourse to the gestures needed to trace Arabic letters and words may be even more natural when they allow for greater comfort and speed of writing, important when the intended text is a personal letter or private notes, which do not involve a controlled calligraphy. Indeed, it is in ordinary, personal writings when individual scribes seek gestural commodity and ease and also speed that one finds a close similarity of form between Hebrew and Arabic scripts. The forms thus created by an individual but influential scribe can become a stylistic model, followed and reproduced by colleagues, disciples, family and the broader professional and intellectual milieu. With time, individual features can become characteristic of a group of scribes and of a distinctive style of script. Thus the capacity of scribes to function in two different working systems may have played a role in developments and changes of the Hebrew script through time, especially in the development of less formal, personal or documentary script registers. 3 A case study: The Maghrebi merchants’ script The documentary register of Hebrew script used in the Near East and North Africa and up to the Iberian Peninsula was based on the script variety developed in the chancelleries of the Talmudic academies in Iraq. By the late 10th century, this script had spread to the diaspora communities over which the Iraqi Geonim extended their control and with which they maintained active epistolary exchange. The chancellery script of the Babylonian Geonim was devised for letters which were sent as far as Egypt, Tunisia, Spain and even northern France. This script had a double function: it was economical but also clear, elegant and distinctive. The economy of space and time in writing was achieved by the reduction of the horizontal strokes and the resulting predominance of vertical movements, the reduction of the number of strokes, the reduction of the size of letters, the rounded movements and finally a few ligatures. The salient morphological feature of this documentary script is the skills and external conditions: body position, writing material and implements). But a particularly important factor is the writing behaviour that the scribe has learned in the past and applies in a quasi-automatic way. 310 kappa shape of the aleph, which goes on to appear in almost all documentary and cursive registers of Hebrew script. The speed of writing, however, did not prevent the letters from displaying a decorum worthy of the Rabbinic scholarly leaders.40 Functional, but also a mark of prestige, the chancellery script was imitated and used in the diaspora communities for their legal documents and other pragmatic writings. The Iraqi chancellery script was the model for the script of the Maghrebi merchants, but this script also possesses novel cursive features, which, as we shall see, reflect the impact of the ductus and form of contemporary Arabic characters. The earliest evidence of the Maghrebi script stems from Tunisia in the early 11th century. It is constituted by a corpus of letters preserved in the Cairo Genizah, written by merchants who were involved in international trade between Spain, North Africa, Sicily and Egypt, and from there further east to Abbasid Iraq and Iran on the silk road and, in the 12th century, southwards to Yemen and India. The influence of the script of the Babylonian Geonim on the Maghrebi script reflects the close and well-attested links between Maghrebi merchants and the Rabbinic yeshivot in Iraq. Derived from the Babylonian chancellery style, the script of the merchants’ letters nonetheless served a different function. Epistolary exchange was the backbone of international trade. Letters were not only essential for conveying vital information, but played a cohesive role by maintaining close ties between the members of the trading network.41 The letters they exchanged in great quantities needed to be written fast.42 The Genizah letter T-S 8J28.12, written in the early 11th century by Moshe ben Shmuel ibn Jāmiʿ in Gabes in Tunisia to the merchant potentate and scholar Joseph ibn ʿAwkal in Fustat repeats three times that the letter was written in a hurry.43 The writer explains that his previous communication, accompanied by sums of money for the Babylonian Gaon Hai son of Sherira, to be transmitted through the intermediary of Joseph ibn ʿAwkal of Fustat, was sent in an equally hurriedly manner, because the sender had to take advantage of a merchant caravan passing through Gabes on its way to Fustat. The Maghrebi merchant travelling with the caravan to whom the money was entrusted, Shmuel ben Abraham al-Tāhirtī, could not stop for longer in Gabes, because he wanted to reach the next stop before the Shabbat. Thus, the sending of letters depended on the vicissitudes of passing caravans. No wonder that the merchants needed a script register which allowed for speedy writing. At the same time as being able to be written quickly, the script the merchants used was legible and immediately recognisable, if not to all potential Jewish readers, at least to the members of their exclusive merchants’ network. 40 On the script of the Babylonian chancelleries, see Olszowy-Schlanger, 2014. Goldberg, 2012a; Goldberg, 2012b. 42 Wagner, 2017. 43 See Mann, 1931, vol. I, pp. 140–141; Gil, 1997, vol. II, no. 142. 41 311 Like the Italian mercantesca script of medieval trading republics, the script of the Jewish Maghrebi merchants was a highly cursive and distinctive register with its own identifiable ductus and style. From North Africa, this script spread to Muslim Spain, where it developed into a fully fledged ‘Sephardi’ cursive which in turn was transferred and further modified in the Christian north of the Iberian Peninsula, and, in the 12th century, spread to southern France and Egypt. A palaeographical analysis of the 11th century Maghrebi mercantile script (and its Iberian avatars) reveals that some letter forms are reminiscent of the ductus and shapes of the documentary Arabic script that the Jewish merchants certainly knew and practiced. In this respect, three graphic features are particularly relevant: ligatures between letters and between letter components; unprecedented numbers of allographs; and ‘nesting’. The presence of these features is facilitated by the use of a pointed calamus which allows writers to easily trace soft, rounded and looped lines. To illustrate these features, we take as an example manuscript T-S 10J9.26.44 This is a Judaeo-Arabic letter of the 1040s, written to Joseph ben Jacob ibn Awkal in Fustat. The letter was written in the chancellery of Abraham ibn ʿAṭāʾ, the Nagid (leader) of the Jewish community in Qayrawan, probably by Joseph ben Berakhyah, who was a member of a prominent Qayrawan family and acted as the Nagid’s helper.45 3.1 Ligatures Traditionally, in Hebrew script the letters are written well separated, one from another. In the calligraphic square scripts used in Torah scrolls for liturgical public reading, the distance between the letters in a word should be of a hair’sbreadth (BT Menachot 29b). Such prescriptions do not apply to pragmatic writings, but in most documentary registers, letters in a word are written separately. However, the need to save time and space in some script styles leads to writing letters close together and, when possible, in a single movement, without taking the implement off the support. Some basic ligatures are attested in the chancellery script of the Geonim, such as the nexus of aleph-lamed (attested also in the square calligraphic scripts) and nun followed by vav at the end of words.46 It is, however, only in the Maghrebi merchants’ script that ligatures affect a large number of letters and can modify their shapes drastically. The choice of options for the ligatures is graphically and culturally determined. Interestingly, some of the Hebrew ligatures imitate the ductus proper to Arabic script. 44 See Gil, 1997, vol. II, no. 163. See Gil, 1997, vol. II, no. 146. 46 Olszowy-Schlanger, 2014, p. 191. 45 312 Figure 4. ‫אלבאקי‬: aleph-lamed written as a nexus To begin with, the Maghrebi script contains the aleph-lamed nexus found in other types and registers of medieval Hebrew script, as shown in figure 4. However, the Maghrebi script contains ligatures which are attested for the first time. The most prominent concern allographs of the letters final he, final kaph, final pe, and qoph following a letter whose last stroke ends on the base-line (see figures 5a–d respectively). In these ligatures, these letters are written with one stroke of the calamus, an extension of the last stroke of the previous letter. In order to trace these letters with one movement, their morphology is modified: it is loop-shaped. The loop-shaped allographs often appear alongside the more standard forms of the same letters in the same document, as can be seen in figure 6. a. b. c. d. Figure 5. a. ‫ציאנתה וסעאדתה וכלאיתה‬: the ligature with final he; b. ‫לך‬: the ligature with final kaph; c. ‫באלכלף‬: the ligature with final pe; d. ‫אלבאקי‬: the ligature with qoph after a letter that ends on the base-line a. b. Figure 6. a. ‫להא‬: non-ligatured he; b. ‫עאנקתה‬: non-ligatured qoph Another frequent ligature involves the letter yod. In addition to being written as a small vertical stroke placed at the head-line, yod is sometimes written as an extension of the previous stroke, without detaching the calamus from the writing surface, and is thus linked with the preceding letter by the so-called ‘invisible line’ (see figure 7). 313 a. b. Figure 7. a. Ṣade followed by yod; b. Ṣade-yod ligature 3.2 Allographs The reduction of ‘unproductive’ strokes and movements of the calamus when tracing letters, as seen in the ligatures, is also applied within a letter, where components may be linked in a different place from that required by the usual ductus. The letter qoph, for example, is often composed of a lobe to which the descender is attached at the level of the base-line, as in figure 8; whereas in the more usual ductus, also attested in the same document, the descender is attached to the top of the lobe and descends from the head-line, as seen in figure 6b. The usual ductus requires lifting the calamus and moving the hand from the base-line back to the top of the letter and then down again, while the alternative ductus dispenses with this ‘unproductive’ movement. It has already been seen in figure 5d that when ligatured to the previous letter, the qoph takes the form of a loop. Thus, in the same document, one finds three different shapes – three allographs – of the letter qoph. Figure 8. ‫קאל‬: a cursive ductus of the qoph with its descender attached to the rounded lobe on the level of the base-line The letter aleph has two allographs. One is kappa-shaped, similar to the eastern documentary script inherited from the Iraqi chancellery tradition (see figure 6 for examples). The other, written in a single movement, has a rounded downstroke on the right side of the letter and a base almost parallel to the baseline (see figure 8). The presence of the base allows for this allograph to be attached to the following letter. Allographs of aleph and qoph similar to those found in this Maghrebi script are subsequently attested in Sephardi script. 3.3 ‘Nesting’ and writing at different levels of the line One of the characteristic features of the Maghrebi and also Sephardi Hebrew scripts is a lack of a clearly defined head-line and base-line. Neighbouring 314 letters are placed at different levels in the line of writing, often one underlying another. This feature is particularly developed in cursive scripts from the Iberian Peninsula from the 13th century onwards, but a modest form of ‘nesting’ is present in the Maghrebi script, as can be seen in figure 9. In that example, the rounded bases of tav, ṣade and nun project to under the following letter, and in the sequences ṣade-mem and nun-tav-he, the long bases cause the shortening of the right-hand downstrokes. Figure 9. ‫תצ׳מנתה‬: projecting bases of tav, ṣade and nun, and shortened downstrokes in the sequences ṣade-mem and nun-tav-he 3.4 The Maghrebi merchants’ script and Arabic script Together with the soft and rounded lines used in writing the Maghrebi script, ligatures, allographs and nesting all seem to have penetrated Hebrew script from contemporary Arabic graphic culture. Linking the letters in a word and writing them at different levels of the line of writing are obvious features of contemporary Arabic script in both calligraphic and documentary variants. Allographs are the prominent orthographic feature of Arabic, with different forms being written in different, standard places: the beginning, the middle and the end of a word. Whereas in the Hebrew alphabet only five letters have an alternative form when written at the end of the word (kaph, mem, nun, pe and ṣade), in the Maghrebi documentary script, like in Arabic script, other letters can be involved in this process. The use of a particular form in a specific place in a word (beginning, middle or end) is not consistent, but the very idea of such a wide scope of allographs parallels Arabic orthography. Moreover, the ductus of several characters of the Maghrebi script resembles the ductus of Arabic letters. Thus, the way of tracing the ligatured Maghrebi allograph of the aleph is identical to the way of tracing Arabic ʿayin (see figure 10); the ductus of the qoph is that of Arabic ṣād (see figure 11); and the ligatured, looped he is traced in the same way as Arabic waw (see figure 12). vs Figure 10. Maghrebi aleph versus Arabic ʿayin (T-S Ar. 5.1, of 1031) 315 vs Figure 11. Two forms of Maghrebi qoph versus Arabic ṣād (T-S Ar. 5.1) vs Figure 12. Maghrebi he versus Arabic waw (T-S Ar. 5.1) This graphic similarity between the ductus and shape of the Hebrew and Arabic letters is purely external. There is no phonetical or grammatical correspondence between the Hebrew and Arabic letters with confusingly identical ductus. It is not the underlying linguistic or phonetic closeness but rather the spatial relationship between the neighbouring strokes and their frequency in certain graphic configurations that create this identical graphic effect. 4 Conclusion Just like other members of the Jewish literate elites, including scribes of books copied in both Hebrew and Arabic, the Maghrebi merchants were highly proficient readers and writers of texts in Arabic script. It is this intimate familiarity with the Arabic writing system and the memory and execution of the partly subconscious and automatised writing gestures – together with the script’s rapidity and the lack of conscious calligraphic effort on the part of the scribes – which account for the application of the movements of Arabic script when writing in Hebrew. This cognitive process works at the level of an individual writer and would seem to be at the origin of a stylistic development of the Hebrew script, a specific merchants’ cursive of the Muslim West. References Abrahams, I., 1976, Hebrew Ethical Wills, Philadelphia: Jewish Publication Society of America. Allony, N., 2006, The Jewish Library in the Middle Ages: Book Lists from the Cairo Genizah (Hebrew), M. Frenkel, H. Ben-Shammai, and M. Sokolow (eds), Jerusalem: Ben-Zvi Institute. Altmann, A. and Stern, S. M., 1958, Isaac Israeli: A Neoplatonic Philosopher of the Early Tenth Century, Oxford: OUP. 316 Assaf, S., 1931, Source-Book for the History of Jewish Education (Hebrew), vol. II, Tel Aviv: Devir. Baker, C. F. and Polliack, M. R., 2001, Arabic and Judaeo-Arabic Manuscripts in the Cambridge Genizah Collections, Cambridge: CUP. Beit-Arié, M., 1990, “Stéréotypes et individualité dans les écritures des copistes hébraïques du Moyen Âge”, in C. Sirat, J. Irigoin and É. Poulle (eds), L’Écriture: le cerveau, l’œil et la main, Turnhout: Brepols, pp. 201–219. ———, 1993, Hebrew Manuscripts of East and West: Towards a Comparative Codicology (Panizzi Lectures), London: British Library. Ben-Shammai, H., 1982, “Hebrew in Arabic script – Qirqisānī’s view”, in S. R. Brunswick (ed.), Studies in Judaica, Karaitica and Islamica: Presented to L. Nemoy on his Eightieth Birthday, Ramat-Gan: Bar-Ilan University, pp. 116–126. Drory, R., 2000a, Models and Contacts: Arabic Literature and its Impact on Medieval Jewish Culture, Leiden: Brill. ———, 2000b, “Le rôle de la littérature karaïte dans l’histoire de la littérature juive au Xe siècle”, Revues des études juives 159, pp. 99–111. Ferre, L., 2009, “Dissemination of Maimonides’s medical writings in the Middle Ages”, in C. Fraenkel (ed.), Traditions of Maimonideanism, Leiden: Brill, pp. 17– 31. Gil, M., 1997, In the Kingdom of Ishmael (Hebrew, 4 vols), Tel Aviv: Tel Aviv University. ———, 2004, Jews in Islamic Countries in the Middle Ages, Leiden: Brill. Goitein, S. D., 1959, “The oldest documentary evidence for the title of Alf Laila waLaila”, JAOS 78, pp. 301–302. ———, 1962, Jewish Education in Muslim Countries (Hebrew), Jerusalem: Ben-Zvi Institute. ———, 1967–1988, A Mediterranean Society: The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Geniza (5 vols), Berkeley: University of California. Goldberg, J. L., 2012a, Trade and Institutions in the Medieval Mediterranean: The Geniza Merchants and their Business World, Cambridge: CUP. ———, 2012b, “The use and abuse of the commercial letters from the Cairo Geniza”, Journal of Medieval History 38, pp. 127–154. Hoerning, R., 1889, Six Karaite Manuscripts of Portions of the Hebrew Bible in Arabic Characters, London: Williams and Norgate. Hopkins, S., 1985, “Two new Maimonidean autographs in the John Rylands University Library”, Bulletin of the John Rylands Library 67, pp. 710–735. ———, 1994, “A new autograph fragment of Maimonides’s Epitomes of Galen (De locis affectis)”, Bulletin of the School of Oriental and African Studies 57/1, pp. 126–132. ———, 2002, “On the Vorlage of an early Judaeo-Arabic translation of Proverbs”, Jerusalem Studies of Arabic and Islam 27, pp. 369–374. ———, 2005, “The languages of Maimonides”, in G. Tamer (ed.), The Trias of Maimonides: Jewish, Arabic and Ancient Culture of Knowledge / Die Trias des Maimonides: Jüdische, arabische und antike Wissenskultur, Berlin: De Gruyter, pp. 85–106. Khan, G., 1986, “The Arabic fragments in the Cambridge Genizah Collections”, Manuscripts of the Middle East 1, pp. 54–61. ———, 1990, Karaite Bible Manuscripts from the Cairo Genizah, Cambridge: CUP. ———, 1993a, “On the question of script in Karaite manuscripts: new evidence from the Genizah”, Bulletin of John Rylands Library 75/3, pp. 133–142. 317 ———,1993b, Bills, Letters and Deeds: Arabic Papyri of the 7th to the 11th Centuries AD, London: Nour Foundation and OUP. ———, 1997, “Abū’l Faraj Hārūn and the early Karaite grammatical tradition”, JJS 43, pp. 314–334. van Koningsveld, P. S., 1992, “Andalusian-Arabic manuscripts from Christian Spain: a comparative intercultural approach”, Israel Oriental Studies 12, pp. 75–110. Langerman, Y. Tz., 2000, “Some remarks on Judah ben Solomon ha-Cohen and his encyclopedia, Midrash ha-ḥokhmah”, in S. Harvey (ed.), The Medieval Hebrew Encyclopedias of Science and Philosophy: Proceedings of the Bar-Ilan University Conference, Dordrecht: Kluver, pp. 371–389. Lichtenberg, A., 1969, Qovetz teshuvot ha-rambam ve-teshuvotav, Farnborough: Gregg. Longcamp, M., 2003, Étude comportementale et neurofonctionnelle des interactions perceptivo-motrices dans la perception visuelle de lettres. Notre manière d’écrire influence-t-elle notre manière de lire? (unpubl. diss., Université de la Méditerranée–Aix-Marseille II). Mallon, J., 1952, Paléographie romaine, Madrid: Instituto Antonio de Nebrija de Filología. Mann, J., 1931, Texts and Studies in Jewish History and Literature, vol. I, Cincinnati: Hebrew Union College Press. Meyerhof, M. (ed., transl.), 1940, Un glossaire de matière médicale de Maïmonide, Cairo: Institut d’Egypte. Munk, S., 1842, “Notice sur Joseph b. Iehouda ou Aboul’Hadjâdj Yousouf ben Yaʾhya al-Sabti al-Maghrebi, disciple de Maimonide”, Journal asiatique 14 (juillet), pp. 4–70. Narkiss, B., 1972, “Illuminated Hebrew children’s books from medieval Egypt”, Scripta Hierosolymitana 24, pp. 58–71. Olszowy-Schlanger, J., 1997, “Karaite linguistics: the ‘renaissance’ of the Hebrew language among early Karaite Jews, and contemporary linguistic theories”, Beiträge zur Geschichte des Sprachwissenschaft 7, pp. 81–100. ———, 2003, “Learning to read and write in medieval Egypt: children’s exercise books from the Cairo Geniza”, Journal of Semitic Studies 48, pp. 47–69. ———, 2014, “Early Babylonian ‘documentary’ script: diplomatic and palaeographical study of two Geonic letters from the British Library Cairo Genizah Collection”, in N. de Lange and J. Olszowy-Schlanger (eds), Manuscrits hébreux et arabes. Mélanges en l’honneur de Colette Sirat, Turnhout: Brepols, pp. 177–195. Reif, S. C., 1990, “Aspects of mediaeval Jewish literacy”, in R. McKitterick (ed.), The Uses of Literacy in Early Mediaeval Europe, Cambridge: CUP, pp. 134–155. Rosner, F., 1998, The Medical Legacy of Moses Maimonides, New York: Ktav. Sirat, C., 1976, Écriture et Civilisations, M. Dukan (collab.), Paris: Centre National de la Recherche Scientifique. ———, 2014, “Maimonides by his own hand”, Archives d’histoire doctrinale et litteraire du moyen âge 81, pp. 7–38. ——— and Di Donato, S., 2011, Maïmonide et les brouillons autographes du Dalâlat al-ḥâʾirîn (Guide des égarés), Paris: Libraire Philosophique J. Vrin. Stern, S. M., 1954–1956, “A treatise on the armillary sphere by Dunas ibn Tamim”, in Homenaje a Millás-Vallicrosa (2 vols), Barcelona: Consejo Superior de Investigaciones Científicas, vol. II, pp. 373–382. ———, 1956, Maimonidis Commentarius im Mischnam e Codicibus Hunt. 117 et Pococke 295 in Bibliotheca Bodleiana Oxoniensi Servatis et 72–73 Bibliothecae Sassooniensis Letchworth, Copenhagen: E. Munksgaard. 318 Stroumsa, S., 1989, Dāwūd ibn Marwān al-Muqammiṣ’s Twenty Chapters: ʿIshrūn maqāla, Leiden: Brill. Vajda, G., 1960, “Un abrégé chrétien du ‘Guide des égarés’ de Moïse Maïmonide”, Journal asiatique 1960, pp. 115–136. Velay, J.-L. and Longcamp, M., 2005, “Clavier ou stylo: comment apprendre à écrire?”, Cerveau & psycho 11, pp. 30–35. Wagner, E.-M., 2017, “The socio-linguistics of Judaeo-Arabic mercantile writing”, in E.-M. Wagner, B. Beinhoff and B. Outhwaite (eds), Merchants of Innovation. The Languages of Traders, Berlin: de Gruyter-Mouton, pp. 68–86. Zucker, M., 1959, Rav Saadya Gaon’s Translation of the Torah; Exegesis, Halakha and Polemics in R. Saadya’s Translation of the Torah (Hebrew), New York: Feldheim. 319 Beyond the Leningrad Codex: Samuel b. Jacob in the Cairo Genizah BENJAMIN M. OUTHWAITE University of Cambridge Samuel b. Jacob is a scribe of the early 11th century whose name is not as familiar to most Bible scholars as that of his greatest creation, the earliest complete copy of the Hebrew Bible, Russian National Library (RNL) Evr. I B19a, the famous Codex Leningrad of the Firkovich Collection.1 Samuel b. Jacob’s name is etched, emblazoned even, throughout the illuminated carpet pages of the volume, but these conspicuous examples of the medieval artisan’s trade are reproduced far less often than the main work itself – the biblical text. The consonants, vowels and accents of RNL Evr. I B19a form the base text for the most widely used of scholarly editions – Kittel’s Biblia Hebraica (BHK, in its third edition), the German Bible Society’s Biblia Hebraica Stuttgartensia (BHS), Dotan’s Biblia Hebraica Leningradensia (BHL), the new Biblia Hebraica Quinta (BHQ) and increasingly vital online tools such as the Westminster Leningrad Codex (WLC) – but the name of its scribe is probably unknown to most who use these editions.2 From the pristine pages of BHS, BHK, BHL, BHQ or the XML of the WLC it is impossible to get the flavour of the sheer high quality of the original manuscript; it is only when you look at the careful layout of the parchment leaves or the embellishment of the micrography on the carpet pages that you can really appreciate the standard of workmanship of the scribe Samuel b. Jacob. It is fitting, therefore, that such a sumptuous volume should be from a scholarly perspective also the most important codex of the Hebrew Bible that we possess. It is a masterpiece of the medieval Masoretic art. 1 I am grateful to my colleagues in the Genizah Research Unit, particularly Dr Kim Phillips and Dr Samuel Blapp, for stimulating conversations around the conclusions reached in this paper, as well as for the feedback I received from a number of seminar and conference participants at home and abroad following my presentations on aspects of the texts featured here. The Leningrad Codex itself can now be viewed easily online on the National Library of Israel’s Ktiv digital library, in stately black and white digitised microfilm form, with restrictive conditions: http://web.nli.org.il/sites/NLIS/en/ManuScript/, item 151623, accessed 13 December 2017. 2 One of many tools now to be found on the internet for biblical scholars, the Westminster Leningrad Codex reproduces the bare text, vowels and accents of B19a in useful XML. It is found at https://tanach.us/Tanach.xml; accessed 12 December 2017. 320 Comparatively little is known about this valuable codex’s production however. The name of its scribe is rarely quoted alongside the text he produced, its exact date is uncertain, and even its place of composition often misreported. Little thought has generally been given to who produced such a masterpiece, for whom and why. This is a gap that can be filled, relatively easily, by reading the colophons of the book. Yet, they are rarely reproduced alongside the text, or, when they are published, interest tends to focus on the gold-illuminated carpet page colophons, rather than the highly informative and lengthy plain colophon of the manuscript.3 As I shall show in this paper, other documentary sources can shine a light on the context of the book’s production and the background of its talented scribe, but the first page to turn to for anyone who has an interest in the book, its production and its ownership is folio 1 r. of RNL Evr. I B19a, the plain ownership colophon, composed and written by Samuel b. Jacob. The colophon on f. 1 r. is lengthy, taking up most of the page; it is longer, for instance, than the Moses b. Asher colophon in the Cairo Codex of the Prophets.4 Its purpose is principally to indicate the ownership of the volume, as well as to record who copied it, where and when. It also serves as a long encomium on the owner and an opportunity for the scribe to show off his mastery of poetic Hebrew prose. It is worth reading in full not just for the details it preserves, but also for the quality of Samuel’s original writing; he was a skilled and versatile scribe.5 The plain colophon RNL Evr. I B19a, Codex Leningrad, folio 1 r. ‫זה המחזור מקרא שלם נכתב ונגמר בנקודות ובמוסרות ומוגה יפה במדינת‬ ::: ‫מצרים‬ ‫ונשלם בחדש סיון שלשנת ארבעת אלפים ושבע מאות ושבעים שנה לבריאת עולם‬ ‫והיא שנת אלף וארבע מאות וארבעים וארבעה לגלות המלך יהויכין והיא שנת‬ ‫אלף‬ ‫ושלוש מאות ותשע עשרה שנה למלכות יונים שהיא למנין {שטרות} ולפסיקת‬ ‫הנבואה‬ ‫ והיא שנת שלוש מאות ותשעים‬.‫והיא שנת תשע מאות וארבעים לחרבן בית שני‬ ‫מה שזכה מבורך בן יוסף בן נתנאל‬ :‫ותשע למלכות קרן זעירה‬ ‫הידוע בן יזדאד הכהן ועשה אתו לעצמו להגות בו מעמלו ומיגיע כפיו ומיזיעת אפו‬ .1 .2 .3 .4 .5 .6 .7 3 An exception is Freedman et al., 1998, which reproduces every page of the codex in glorious detail. If you don’t already own a copy, however, then you need to be rich to acquire one. 4 Though this includes several lines added to the original ownership colophon by a subsequent owner of RNL Evr. I B19a, as seen below. 5 A complete edition of the colophon, transcribed and translated into French, can be found in Beit-Arié, Sirat and Glatzer, 1997, pp. 115–116. For the edition given here, I relied on digital images and the facsimile volume, and my readings differ occasionally from those of Beit-Arié, Sirat and Glatzer. 321 ‫‪.8‬‬ ‫‪.9‬‬ ‫‪.10‬‬ ‫‪.11‬‬ ‫‪.12‬‬ ‫‪.13‬‬ ‫‪.14‬‬ ‫‪.15‬‬ ‫‪.16‬‬ ‫‪.17‬‬ ‫‪.18‬‬ ‫‪.19‬‬ ‫‪.20‬‬ ‫‪.21‬‬ ‫‪.22‬‬ ‫‪.23‬‬ ‫‪.24‬‬ ‫יהי רצון מלפני יהוה שיחזיק אותו בתורתו ויאמץ אותו במצותיו ויחכמו בדקדוק‬ ‫דתו ויאיר עיניו בתעודתו ויזכה לבנין ביתו וינחילו יוצרו שני חיי עולמים חיי‬ ‫העולם‬ ‫הזה וחיי העולם הבא וְ ָיגֵן ברחמיו הרבים עליו ויפרש עליו ועל זרעו סכת שלום‪:‬‬ ‫ויזכה אתו להגות בתורתו תמיד ויזכה לקים כל התורה מקרא ודקדוקי מקרא‪ :‬ויקים‬ ‫לא ימושו מפיך ומפי זרעך ומפי זרע זרעך אמר יהוה מעתה ועד עולם‪ :‬ויזכה‬ ‫לזרע חי וקים ונכון בתורה ובמצות ובמעשים טובים ולחיים ארכים שאין בהם‬ ‫חטא ועון‬ ‫וינחילו יוצרו חן וחסד והון ועשר וכבוד ועטרת תורה ותפארת תושיה ויזכה‬ ‫להוד […]‬ ‫ולהדרה ולעזה ולעזוזה וינצל מכל עברה וצרה וזעם וחרון אף ומכל נזק ים ויבשה‬ ‫ויראה זרע‬ ‫ויאריך ימים כאב המון הזקן אשר בא בימים ויצליח מאד בכל מעשים מכשרים‬ ‫כיחיד אשר‬ ‫מצא מאה שערים ופדות והצלח כְ פָץ איה השה בניאומים ועשר וכבוד כחו[ז]ה‬ ‫סלם מצב‬ ‫במורא ממקומים וחן וחסד והוד כמבית האסורים יצא למשול בעמים ונסי נסים‬ ‫כשנעשו‬ ‫למוציאים בן בכור מבין ענמים‪ :‬וברית עולם כנחשבה לו צדקה לדורות עולם‬ ‫וגודל שמע‬ ‫כשעמדו לו שמש וירח ביום מימים‪ :‬וטוב עם יהוה ועם אנשים כנענה בהקריבו‬ ‫טלה תמים‪:‬‬ ‫ובינת דבר כמנגן בשירות בנעימים‪ :‬וחכמה ושכל כנחכם מכל חכמים‪ :‬וישועה‬ ‫גדולה‬ ‫כעל פי דברו זרמו גשמים‪ :‬וינצל מכל צרה וצוקה כָאמר לא זה הדרך לגדודי‬ ‫ארמים‪:‬‬ ‫אמן כן יהי רצון מצור עולמים ושלום רב מבורך זה יועצם ויותמם ויורם ויוחטם‬ ‫ויונעם‬ ‫לעולם אמן ואמן באלהי אמן ומלך עולם ‪:‬ס‪:‬‬ ‫‪Addition to the colophon in a later hand6‬‬ ‫‪ .25‬קנה זה המצחף כ׳ג׳ק׳מ׳ור׳ אדוננו מצליח הכהן ראש ישיבת גאון יעקב יברכהו‬ ‫אלהינו וישמרהו‬ ‫‪ .26‬בר כ׳ג׳ק׳מ׳ור׳ אדוננו שלמה הכהן ראש ישיבת גאון יעקב זלחה״ה נין הגאונים‬ ‫בממונו לנפשו ומאת‬ ‫‪ .27‬כ׳ג׳ק׳ המשכיל החכם והנבון יוסף הנודע בן כוגך בחדש תמוז שנת אתמ״ו לשטרות‬ ‫האלהים יזכהו להגות‬ ‫‪6‬‬ ‫‪This purchase and ownership note was added to the colophon in the same style as the original‬‬ ‫‪when Maṣliaḥ Gaʾon, head of the Palestinian Academy in the 12th century, purchased the co‬‬‫‪dex. Subsequent owners of codices often write their own colophons, but Maṣliaḥ added himself‬‬ ‫‪to the chain of ownership in the original colophon, probably because of the significance and‬‬ ‫‪obvious value of the book. For more on this purchase, and the previous ownership by Ibn Ku‬‬‫‪chek, see Outhwaite, 2017.‬‬ ‫‪322‬‬ ‫חצר דלך לוי‬ ‫ בו הוא ובניו וזרעם כל ימיהם לקיים לא ימושו מפיך וג׳‬.28 ‫ חצר מנשה הכהן ביר׳ יעקב נ״ע חלפון הלוי ביר׳ מנשה נ״ע‬.29 ‫הלוי בר יפת הלוי נ״ע‬ 1. This codex7 of the complete Bible was written, furnished with vocalisation and masora,8 and carefully checked9 in Fusṭāṭ.10 2. And it was completed in the month of Sivan of the year four thousand and seven hundred and seventy of Creation, 3. and which is the year one thousand and four hundred and forty-four of the Exile of King Jehoiachin, and which is the year one thousand 4. and three hundred and nineteen of the Kingdom of the Greeks, which is the reckoning of {documents}11 and of the Ceasing of Prophecy, 5. and which is the year nine hundred and forty of the Destruction of the Second Temple, and which is the year three hundred and ninety6. nine of the Kingdom of the Little Horn.12 This has been rightfully acquired13 by Mevoraḵ ben Joseph ben Netanʾel 7 The word translated as ‘codex’ is Hebrew ‫מחזור‬, which has come to mean a prayer-book for the festivals in Modern Hebrew. In its earliest usage, however, it appears to have the sense of ‘book, codex’ in general, as opposed to sefer, which is ‘scroll’. On the earliest uses of ‫ מחזור‬in this sense, see Glatzer, 1989, pp. 260–263. It can be found in a number of biblical colophons, including the Moses b. Asher colophon in the Cairo Codex of the Prophets: ‫אני משה בן אשר‬ ‫‘ כתבתי זה המחזור שלמקרא‬I, Moses b. Asher, have written this codex of scripture’. Others prefer the loan-word miṣḥaf (‫מצחף‬, Arabic ‫مصحف‬, maṣḥaf or muṣḥaf) for the same meaning. Indeed, Samuel himself uses ‫ מצחף‬in the carpet page on f. 474 r. of Leningrad, ‫כתבתי ונקדתי ומס׳ זה המצחף‬ ‘I have written, vocalised and provided the masora of this codex’ (abbreviating ‫ ומסרתי‬to keep a straight left-hand margin). 8 Samuel uses the plural, which he spells with a ‫ ו‬after the initial ‫מ‬, perhaps reflecting the back pronunciation of the qameṣ. It is an unusual spelling that I cannot find replicated elsewhere. I hesitate to suggest it is an error this early in the colophon. 9 This is the hufʿal participle of ‫נגּה‬, ‘to check, revise’, and refers to authoritative Bibles, those copied or corrected against a certain textual tradition. Yeivin, 1980, p. 138 describes ‫ מוגה‬as “any sort of carefully corrected text”. The idea of ‘checked’ implying ‘checked and corrected’ texts is present already in the Talmud, e.g. BT Ketubbot 19b ‫ספר שאינו מוגה‬, ‘a Torah scroll that has not been checked’, which may only be kept for thirty days without having any corrections made. The most accurate translation would perhaps be ‘copy-edited’ or ‘proof-read’, but it sounds too anachronistic in the context of 11th century Fusṭāṭ. 10 Some other translations of the opening of the colophon read ‘Cairo’ – e.g. Lebedev in Freedman et al., 1998, p. xxii; Würthwein, 2014, p. 254 – but Samuel is definitely referring to Fusṭāṭ. The phrase occurs, for instance, in a divorce deed, CUL T-S 8.154, in which a man from Alexandria (No Amon) divorces his wife from Fusṭāṭ (Medinat Miṣrayim) in Taṭay in 1052. Stern, 2017, p. 199 suggests that RNL Evr. I B19a was written “probably in the land of Israel”, but that is at odds with Samuel’s statement in the colophon. 11 The word ‫‘ שטרות‬documents’ has been erased in the text, perhaps deliberately. 12 This is the Hijra, the Islamic era, and the term ‘little horn’ is a belittling epithet for Islam derived from Daniel 7:8, “I considered the horns, and, behold, there came up among them another little horn”. 13 The verb ‫ זכה‬has a range of meanings, and is usually used in the sense of ‘to be worthy of’ in blessings and other pieces of laudatory prose (as often, for instance, in this colophon). But it may also have the technical sense of rightful ownership or possession – e.g. Mišna Bava Meṣiʿa 323 7. who is known as Ibn Yazdād ha-Kohen, and he had it made for himself,14 to study it, out of [the proceeds of] his own labour, the toil of his hands and the sweat of his brow. 8. May it be the will of the LORD15 to encourage him through His Torah, to strengthen him through His commandments and to make him wise through the fine points of 9. His law. And may He give light to his eyes through His testimony. And may he award him with the [re]building of His house. And may his Creator bestow upon him the life of two worlds: the life of this world, 10. and the life of the world to come. May He protect16 him with His abundant mercy and may He spread over him and over his offspring a Booth of Peace. 11. And may He grant him to always study His Torah, and may He grant the fulfilling of the Torah, scripture and fine points of scripture.17 May He establish 12. “they shall not depart from your mouth, or from the mouth of your children, or from your children’s children, said the LORD, from now and forever” [Isaiah 59:21]. And may he enjoy 13. offspring living, enduring and secure in Torah and in the commandments and in good deeds, and long18 life without sin or transgression. 1:3, ‫‘ ואמר אני זכיתי בה זכה בה‬He said “I have acquired it”, then he has acquired it’ – which is probably also the case here in this ownership colophon. Beit-Arié, Sirat and Glatzer, 1997, p. 115 translate it as ‘a eu le mérite de faire’, taking ‫( ועשה‬line 7) as subordinate to ‫זכה‬. The same coordinate phrase occurs in the ownership colophon of the Cairo Codex of the Prophets, ‫‘ מה שזכה יעבץ בן שלמה נח נפש ועשה אותו‬which is owned by Yaʿbeṣ b. Solomon – his rest be easy – and he made it …’. 14 Literally, of course, ‘he made it for himself’, but the sense is that he commissioned it, rather than bought or acquired it otherwise. Exactly the same phrase occurs in the Yaʿbeṣ b. Solomon colophon of the Cairo Codex of the Prophets, lines 2–4, ‫ועשה אותו לעצמו להגות בו מעמלו ומיגיע‬ ‫כפיו ומזיעת אפו‬. Note that the Cairo colophon uses the expected ‫ז(י)עה‬, ‘sweat’, appropriated from Genesis 3:19, whereas Samuel’s ‫ מיזיעת‬is, at best, a hybrid form from ‫יזע‬, Ezekiel 44:18, and ‫ ֵזעָה‬. 15 Samuel uses the full form of the divine name, which is unusual for a piece of non-biblical prose. The Cairo Codex of the Prophets ownership colophon uses an abbreviated form in the phrase ‫‘ לחזות בנועם ייי‬to behold the beauty of the L[ord]’ (Psalms 27:4), as does the colophon of RNL Evr. II B17, ‫‘ אל י[ח]שב לי יוי עון‬let the L[ord] not impute me blame’ (a reworking of Psalms 32:2), which is dated 929 CE. 16 The vocalisation on ‫ ויגן‬appears to be original, added by Samuel when he wrote it and highlighting a difficult form that occurs only three times in the Bible (in Isaiah 31:5 and twice in Zechariah). Sporadic vocalisation in otherwise unvocalised Medieval Hebrew texts often occurs on obscure biblical forms or ambiguous words; see Outhwaite, 2000, pp. 6–7. 17 The term ‫ מקרא‬often implies the “study of scripture” and the phrase ‫ דקדוקי המקרא‬has the sense of “the fine points of scripture as established by detailed investigation” in early grammatical or Masoretic works; see Khan, 2000, pp. 14–15, 17–19. 18 Samuel employs a number of defective forms, such as ‫( אר(ו)כים‬line 13), ‫( א(ו)תו‬lines 7, 11) and ‫( קי(י)ם‬line 13), suggestive of a deliberately biblicising style to his prose here. 324 14. And may his Creator award him grace, kindness, wealth, riches, honour, and a crown of Torah and a diadem of success. And may he be granted the glory of the […],19 15. its majesty, its strength and its might. May he be saved from all fury, hostility, indignation and wrath, and from all harm by sea and dry land. May he see offspring 16. and may He lengthen his days like the “Father of Many”,20 who was old and stricken in age,21 and grant him great success in all proper deeds, like the only one22 who 17. found “one hundred measures”,23 and redemption and success, like the one who spoke24 the words “where is the sheep?”,25 and wealth and honour like the one who had a vision of a ladder set up26 18. in the most dreadful27 of places, and grace and kindness and splendour like the one who came out of the prison to rule over people,28 and miraculous miracles like those that were performed 19. for those who brought out the firstborn son from among the Anamites.29 And an everlasting covenant like the one who was reckoned for righteousness for eternal generations,30 and greatness of reputation 20. like the one31 for whom the sun and moon stood still on one particular day. And eminence with the LORD and with men, like the one32 who was answered when he sacrificed a whole lamb. 19 Beit-Arié, Sirat and Glatzer, 1997, p. 115 read ‫ ויזכה להודה‬and translate it as ‘qu’il ait le mérite de sa grâce’, but unless it is a highly elongated ‫ ה‬at the end of the line, another word should follow ‫להוד‬. 20 The patriarch Abraham, from Genesis 17:5. 21 Genesis 24:1. 22 Isaac, from Genesis 22:2. 23 Isaac, from Genesis 26:12. 24 Samuel b. Jacob vocalises ‫ כְ פָץ‬because of the unusual form and syntax. It is commonplace in classical Hebrew piyyuṭ, however, reflecting a payṭan-like use of the preposition -‫ כ‬with a finite verb, together with the apocopation of a final-he verb (‫ פוץ > פצה‬in fact) in the 3rd person masculine singular suffix conjugation. This particular verb, ‫פץ‬, is very common in Hebrew poetry of the Byzantine period; see Rand, 2006, pp. 136–138, 422–427. Another example of payṭanlike language is ‫ניאומים‬, a plural of ‫נאום‬, where the yod perhaps reflects a pronunciation niyūmīm, with a glide replacing the glottal stop. The extensive use of biblical allusion in this colophon is also typical of poetic language. It suggests that Samuel was probably quite familiar with piyyuṭ and, as a professional scribe, he may well have copied poetry or liturgy as well as Bible codices in order to make ends meet. 25 Isaac, from Genesis 22:7. 26 Jacob, from Genesis 28:12. 27 Beit-Arié, Sirat and Glatzer, 1997, p. 115 reconstruct ‫‘ [במור]א ממקומים‬dans le plus redoubtable des lieux’. The allusion is to Genesis 28:17, ‫‘ מה נורא המקום הזה‬how dreadful is this place?’. I think the reading ‫ במורא‬is visible, but it could conceivably be ‫בנורא‬. 28 Joseph, alluding to Ecclesiastes 4:14. 29 Through synecdoche, the Egyptians; Genesis 10:13 and 1 Chronicles 1:11. 30 Phineas, from Psalms 106:31, though Samuel has ‫ צדקה‬for the Psalmist’s ‫לצדקה‬. 31 Joshua, from Joshua 10:12–13. 32 Samuel, from 1 Samuel 7:9. 325 21. And the understanding of matters like the player33 of songs in delight. And wisdom and understanding like the one who was wiser than all wise men.34 And a great salvation 22. like the one on whose word the rains poured down.35 And may he be saved from all distress and anguish, like the one36 who said “This is not the way” to the bands of Aramaeans.37 23. Amen. So may it be the will of the Rock of Ages. Great peace. This Mevoraḵ – may he be made strong and upright, be raised up, calm and pleasant. 24. For ever, amen and amen, by the God of Truth and the Eternal King :o:38 Lines 1 to 24 are the original ownership colophon of the codex. A further note of ownership was added when the book changed hands in the 12th century. 25. The honourable, great, holy, our master and our teacher our lord Maṣliaḥ ha-Kohen Head of the Academy of the Pride of Jacob purchased this book – may our God bless him and protect him – 26. son of the honourable, great, holy, our master and our teacher our lord Solomon ha-Kohen Head of the Academy of the Pride of Jacob – may his memory be for a blessing for the life of the next world39 – descendant of geʾonim, with his own money for himself from 27. the honourable, great, holy, the wise and distinguished teacher Joseph who is known as Ibn Kuchek, in the month of Tammuz, the year 1446 of Documents. May God allow him to study 28. it, he, his children and their offspring, all their days to fulfil [the commandment] “they shall not depart from your mouth, etc.”.40 29. Attending, Manasseh ha-Kohen son of the scholar Jacob – his rest be in Eden. Ḥalfon ha-Levi son of the scholar Manasseh – his rest be in Eden. Attending it, Levi ha-Levi son of Japheth ha-Levi – his rest be in Eden.41 The original colophon is in the hand of Samuel b. Jacob himself. This is clear from a comparison of the handwriting with the rest of the manuscript, and can also be seen from the use at the end of the colophon of his distinctive textual marker :o: (see figure 1), which he uses to delimit Masoretic sections 33 David, from 1 Samuel 18:10 and elsewhere. Solomon, alluding to 1 Kings 3:12. 35 Elijah, from 1 Kings 18:41–45. 36 Elisha, from 2 Kings 6:19. 37 2 Kings 6:23. 38 Samuel’s siglum :o: marks the end of the original colophon. 39 An abbreviation of ‫זכרונו לברכה לחיי העולם הבא‬. 40 Isaiah 59:21. 41 These three are witnesses to the sale, and all had links to Maṣliaḥ Gaʾon and the Palestinian Yešiva. 34 326 throughout the masora magna of the book (and which also occurs in the illuminated carpet pages). Different copyists of the medieval Masoretic Bible use different sigla, such as simple circles or more elaborate combinations of circle and lines, but Samuel’s is rare enough – though not unique to him – that we can with confidence take this as his own work here too. Indeed, there is no reason to doubt it. Figure 1. Samuel b. Jacob’s distinctive siglum is used to divide Masoretic notes in another codex of the Torah that he copied, which now exists as scattered leaves in the Cairo Genizah (detail of CUL T-S A3.35 verso)42 From the colophon, we can see that Samuel was working in Fusṭāṭ at this stage of his life. In using ‫ מדינת מצרים‬medinat miṣrayim instead of ‘Fusṭāṭ’ – ‫פסטאט‬ or ‫( מצר‬Arabic, miṣr) – Samuel is endeavouring to use only Hebrew vocabulary in his colophon. This is a feature typical of Hebrew high prose style of this era, which tends to avoid using foreign terms or toponyms if vocabulary from the Bible can be used instead.43 Fusṭāṭ was, since the Fāṭimid founding of Cairo, no longer the capital of Islamic Egypt. It remained, however, the administrative centre for decades after Cairo’s emergence. For the Jews, it was the social, economic and cultural centre of Egypt.44 Consequently, a scribe could expect to find commissions for scribal work in Fusṭāṭ from among the elite of the city, its community leaders and merchant princes. Samuel gives the month of completion of his work as the Jewish month of Sivan (May–June), and then – to lend it more weight and significance – the year according to five different systems of reckoning, not all Jewish. They are: 4770 of the Creation (‫)לבריאת עולם‬, which equates to 1010 CE; 1444 of the Exile of King Jehoiachin (‫)לגלות המלך יהויכין‬, which is in the range 1006–1014 CE;45 1319 of the Seleucid Era (‫)למלכות יונים שהיא למנין שטרות‬, which equates to 1008 CE; 940 of the Destruction of the Second Temple (‫)לחרבן בית שני‬, 42 Image reproduced by kind permission of the Syndics of Cambridge University Library. As Golb, 1965, pp. 269–270 describes it, “The Jews probably began using Biblical Hebrew names for Egyptian cities during this period of nationalistic reorientation [i.e. after the revolt of the 2nd century CE] … This practice of giving Hebrew names to Egyptian cities was evidently kept up after the arrival of Islam and through the beginning of the Fatimid period.” On the avoidance of loan-words in Hebrew of this period, see Outhwaite, 2000, pp. 98–99. 44 Bareket, 2017, pp. 88–89. 45 The exile took place in 597 BCE; see Freedman, 1992, s.v. Jehoiachin. Rabbinic sources date it to 430 BCE, but it can be dated further back in 437/8 BCE in medieval documents (OlszowySchlanger, 1998, pp. 162–163). It’s an important event in Jewish chronology, because it helps to set the date for the destruction of the First Temple eleven years later. 43 327 which is in the range 1008–1010 CE;46 399 of the Hijra (‫)קרן זעירה‬, which is 1009 CE. Which to follow? The Jews in Egypt were more accustomed to using the Seleucid or Creation dates than others.47 The Seleucid is probably reliable, as it is so commonly used. A date of 1008–1009 CE seems reasonable, as it comes closest to the Seleucid, Destruction and Hijra date ranges. We should not expect absolute calendrical exactitude in a pre-modern text, particularly when the writer was probably not a specialist in the discipline. Egregious errors of dating can be easily found in Genizah legal documents, such as a court record from Fusṭāṭ where the writer has recorded the date as 4994 of Creation (‫)ד׳ אלפים ותשע מאות ותשעים וארבע‬, when he probably meant to write the year 4794 (‫)ושבע מאות‬, that is, 1034 CE – an error of 200 years!48 Most interesting is the fact that the colophon uses all five systems of dating, including the Exile of King Jehoiachin, which is a system used mainly in Karaite legal documents.49 Samuel means to be comprehensive in his colophon, but in adding the reckoning according to the Jehoiachin exile, he is probably conforming to the affiliation of the owner of the Bible. Mevoraḵ b. Joseph b. Netanʾel known as Ibn Yazdād ha-Kohen was the commissioning owner of RNL Evr. I B19a, as not only the plain colophon (f. 1 r.) but also a number of the carpet pages make richly clear (e.g., ff. 474 r., 475 v. and 478 v., among others).50 The family was originally from Persia: the patronymic Ibn Yazdād is of Persian origin, meaning ‘God has given’, that is, ‘God’s gift’. An Ibn Yazdād, perhaps Mevoraḵ’s father, appears in commercial correspondence from the Genizah early in the 11th century, whence it seems he is based in Egypt and plays a role in Mediterranean trade, associating with the great merchant Joseph b. Jacob Ibn ʿAwkal.51 46 Although the destruction of the Second Temple took place in 70 CE, the calculations of the elapsed time differ in Jewish sources, with the result that the reckoning from the Destruction effectively covers a range of several years, 68–70 CE; see Friedman, 1980, vol. I, pp. 104–106. 47 Goitein, 1967–1993, vol. I, p. 355. The Seleucid system of dating, ‘the Era of Documents’ or ‘of the Greeks’, was the preferred system of the Babylonian Jews and, given their cultural dominance, it became common in documents from the Genizah too; see Friedman, 1980, p. 106. The suggestion in Würthwein, 2014, p. 254 that the Hijra date “is probably the most reliable because the writer lived in an Islamic country” goes too far. Hijra dating, though found in Genizah texts (and, indeed, see the document discussed below), is used far less than the Jewish systems of Creation and the Seleucid Era. 48 CUL T-S 8J6.8. The handwriting and language are clearly of the 11th century. The error was probably not strictly mathematical or calendrical, but scribal: anticipating the ‫ תשעים‬a few words later, he wrote ‫ תשע‬instead of ‫שבע‬. 49 And mainly, it seems, in Karaite deeds from Ramla; Olszowy-Schlanger, 1998, pp. 162–163. 50 Although Samuel spells Ibn Yazdād’s name ‫ יזדאד‬on most occasions, he does write it once ‫אזדאד‬, in the central text on the carpet page f. 474 r. Medieval Jewish writers like variety, and will often switch between the Hebrew and Arabic versions of people’s names in the same document, but given that Samuel spells it ‫ יזדאד‬everywhere else, it is probably an unfortunate error by the scribe, which is explainable perhaps by the unfamiliarity of the name. 51 It could alternatively be our Mevoraḵ himself. He is mentioned in a letter sent from Tunisia to Ibn ʿAwkal in Fusṭāṭ around 1000 CE, “What disturbed me most was your failure to pay to Ibn Yazdād and Salāma, the son-in-law of Furayj, the sum that I asked you to pay them” (translation from Goitein, 1973, p. 31). Ibn ʿAwkal’s roots also lay in Persia. 328 The owner Mevoraḵ was therefore himself probably a major merchant, and had accumulated both personal wealth and social status. Evidence of this, beyond his commissioning of an expensive codex, is that he was appointed around 1019 CE to oversee the two supervisors of an inheritance, ensuring the safeguarding of a substantial sum in trade goods for the minor son of a Jewish merchant.52 The Genizah has preserved a letter that Mevoraḵ wrote in his role as overseer to one of the deceased merchant’s business partners.53 Further correspondence, written by the supervisors themselves, reveals Mevoraḵ’s name in Arabic to be Abū l-Ḥusayn al-Mubārak ibn Yūsuf ibn Yazdād.54 Goitein believes that Mevoraḵ was a Karaite; Bareket, who also wrote about the case, suggests that there is not enough evidence to be sure.55 A Karaite legal deed from 1004 CE, however, shows that Joseph Ibn Yazdād, Mevoraḵ’s father, also held a position of trust in Karaite society, as there he is named as one of three supervisors of a Persian Karaite merchant’s accumulated possessions.56 This strongly suggests that the Ibn Yazdād family were themselves Persian Karaites, and it explains therefore Samuel b. Jacob’s use of the very rare reckoning from the Exile of King Jehoiachin in the ownership colophon. Samuel’s expertise as a scribe of the Bible is evident on every page of RNL Evr. I B19a, from the careful layout to the micrographic conceits of the carpet pages.57 We know from the several colophons that he copied the text and also added the vocalisation and the masora, performing alone a task often carried out by a scribe (sofer) and a punctuator (naqdan) working together.58 Moreover, he appears to have produced, at the very least, the micrography on the carpet pages, and may well have added the gold illumination to them too.59 It seems that this was a remarkable piece of solo entrepreneurship, producing single-handedly a luxury edition of the Hebrew Bible. Samuel’s client, Mevoraḵ, would have wanted his personal wealth to be reflected in the quality of 52 A number of letters and legal documents from this case have been recovered in the Genizah. It eventually reached litigation in the Muslim and Jewish courts when the orphan came to his majority and purposefully sought to recover his father’s assets. See Bareket, 1998, pp. 124– 136; Goitein, 1967–1993, vol. III, pp. 293–295. 53 CUL T-S 10J30.7. 54 CUL T-S 16.27. 55 Bareket, 1998, p. 125 n. 8. 56 CUL T-S 16.171. 57 Clearly inspired by Islamic book decoration, Samuel’s carpet pages are a remarkable artistic achievement. Stern, 2017, p. 84 has described them as “among the most complex in the history of the Jewish book”. 58 He states this explicitly in the plain colophon, and the colophons on ff. 474 r. and 479 r. It is not unusual for a single scribe to produce the different text layers of a Masoretic Bible, e.g. RNL Evr. II B39, which was written, vocalised and provided with masora by Joseph b. Jacob the Maghribi, or RNL Evr. II B115, which was written and vocalised by one Moses b. Hillel. 59 The micrography is in a hand very similar to Samuel’s and the sections, biblical verses and so on are often delimited by his :o: siglum, or various elaborations of it. Furthermore, one can take ‫ חקקתי‬in the micrographic colophon on f. 477 r. as referring to the creation of the micrographic embellishments themselves: ‫‘ אני שמואל חקקתי למבורך יחיה‬I Samuel have inscribed this for Mevoraḵ – long may he live’. 329 the volume: the care taken in its copying, its size, its embellishment. There is no doubt it would have been a costly book to produce – in raw materials alone it would have taken at least 120 sheep for the 491 leaves.60 There is no record in the colophons of how much the volume cost, but we are fortunate to have a record of Samuel’s later scribal activity that shows how costly it might have been. CUL T-S 10J5.15 is a legal document in Judaeo-Arabic from the Cairo Genizah on the terms of payment for copying books of the Hebrew Bible. It was edited in full by Elinoar Bareket in her 1995 book Jewish Leadership in Fustat, where she notes that it was an agreement with a scribe by the name of Samuel b. Jacob.61 The document is dated 1021 CE, and Samuel was to be paid twenty-five dinars by Salāma ibn Saʿīd ibn Ṣaḡīr for copying ‫אלתמניה‬ ‫‘ אספאר אלנבי ואלכתיב‬eight books of the Prophets and the Writings’, including all the vocalisation and masora (‫)ונקטהא ומאסרתהא‬. Not only was he to do all this on his own, but, showing the same entrepreneurship that produced Codex Leningrad, he would also bind and cover the book. It was to be copied in the same style as a Torah in the possession of Abū Isḥāq Ibrāhīm ibn Ḥujayj, which Samuel had previously produced, and Ibrāhīm would receive two dinars from this agreement. Agreement between Samuel b. Jacob and Salāma ibn Saʿīd ibn Ṣaḡīr CUL T-S 10J5.15 ‫ בשם יי׳‬.1 ‫ יקול שמואל בן יעקוב אני ואקפת סלאמה בן סעיד בן צגיר עלי‬.2 ‫ נסך אלתמניה אספאר אלנבי ואלכתיב בכטהא ונקטהא ומאסרתהא‬.3 ‫ וכתב שרוט פי אול כל גזו ופי אכרה ותגלידהא ועמל זנפילגה‬.4 ]‫ להא ויכון דלך עלי אלנמודג אלדי ענד אבו אסחק אברהים [בן‬.5 ‫ חגיג והי אלתורה אלדי נסכתהא לה לא זאיד ענהא ולא נאקץ‬.6 ]‫ [כ]מא תקדם בה אלשרט ואלאגרה פי דלך עלי גמיע מא תקד[ם‬.7 ‫ בה אלוצף כ׳ה׳ דינארא גיאדא ואזנה‬.8 ‫ קבצ׳ת מן דלך דינארין ודלך עלי <יד> סיידי אבו נצר סלאמה‬.9 ‫ בן סעיד בן צגיר איידה אללה והו אלמתולי לנסך הדה אלאגרה‬.10 ‫ לאבי אסחק אברהים בן חגיג איידה אללה‬.11 ‫ ודלך ללנצף מן שעבאן סנה אתני עשר וארבע מאיה‬.12 60 This is assuming that one complete sheepskin gives two bifolia of the quarto size used in RNL Evr. I B19a. There are 491 leaves in the book, arranged in quinions (quires of five bifolia, as is standard in eastern Jewish manuscripts), though the volume has been subject to later repair that has disordered some of the last leaves; see Beit-Arié, Sirat and Glatzer, 1997, p. 114. 61 Bareket, 1995, pp. 204–205, with a translation into Hebrew. She did not make the connection with the scribe of RNL Evr. I B19a, which is understandable since she is an historian, not a biblical scholar. It was Y. Ofer who first made the link with the scribe of the Leningrad Codex in his study of Samuel b. Jacob’s copy of Saʿadya’s Tafsīr, noting the fact in a footnote; see Ofer, 1999, p. 197 n. 23. I first discussed this document and its wider significance for the background of Codex Leningrad in Outhwaite, 2016. 330 1. In the name of the LORD 2. Samuel b. Jacob says: “Behold, I hereby make an agreement with Salāma ibn Saʿīd ibn Ṣaḡīr concerning 3. a copy of the eight books of the Prophets and the Writings,62 with its script, its vowel points and its masora,63 4. and the signs at the beginning and end of each section,64 and its binding and the manufacture of the case65 5. for it. And it should be according to the exemplar66 that is in Abū Isḥāq Ibrāhīm ibn 6. Ḥujayj’s possession – that is the Torah that I copied for him – nothing more and nothing less, 7. according to what is stipulated in this agreement. And the fee for this for all that is stipulated 8. in this specification is 25 dinars of full weight. 9. I have received from this two dinars, which came from the hand of my lord Abū Naṣr Salāma 10. ibn Saʿīd ibn Ṣaḡīr – God give him strength – who is directing the copying. This fee is 11. for Isḥāq Ibrāhīm ibn Ḥujayj – God give him strength. 12. And this is Mid-Šaʿbān, the year 412.67 62 The Hebrew Bible is traditionally divided into the Torah (Genesis–Deuteronomy), the Prophets and the Writings (the Hagiographa). The Prophets can be divided into the four books of the Former Prophets (Joshua–Kings) and the four books of the Latter Prophets (Isaiah, Jeremiah, Ezekiel and the Twelve). 63 This division of the three main tasks of the biblical copyist – the script, vowels and masora – is reflected in many of Samuel’s colophons, e.g. RNL Evr. I B19a f. 479 r., ‫שמואל בן יעקב כתב‬ ‫‘ ונקד ומסר‬Samuel b. Jacob wrote, vocalised and provided the masora’. 64 Bareket, 1995, p. 204 understands these signs as denoting the open and closed sections of the text, the parašiyyot petuḥot and setumot: ‫נראה שהכוונה לסימון הפרשיות הסגורות והפתוחות‬. The Arabic word is ‫‘ جزء‬section, part’. This could refer to the open and closed paragraphs (parašiyyot or pisqot), but perhaps refers to the individual biblical books; i.e. the note of the total number of verses found after each book in Codex Leningrad. Or maybe it refers to the physical parts of the volume, the quires (though in Arabic this is usually karārīs). Quires have signs at the beginning and end of each part, to ensure the volume is bound in the correct order. In Codex Leningrad, the quires are marked with a catchword at the end of each and a quire number at the beginning: e.g. f. 10 v., the last page of the first quire, has the catchword ‫מת‬, anticipating the first word on f. 11 r., the first page of the second quire, numbered ‫ ב‬in the top margin. All the quires in the book are so marked. Given that the instructions in the agreement have already referred to the masora, which probably include the parašiyyot and other paratextual features, it seems logical that this separate instruction may denote something more to do with the manufacture, binding and completion of the volume, i.e. the correct ordering of the quires. 65 Bareket, 1995, p. 204 cannot read the last word clearly and guesses ‫ארפילוה‬, but it is to be read ‫זנפילגה‬, Arabic zanfalīja (though here either with metathesis of the vowels, or yod denoting ʿimāla), meaning a rigid case for a book, a slipcase; see Goitein, 1967–1993, vol. IV, pp. 387 n. 191, 463 n. 229. 66 ‘Exemplar, pattern, model’, ‫)نموذج( נמודג‬, is a Persian loan-word into Arabic, perhaps reflecting the eastern origins of this type of scribal practice, copying from model codices. 67 The night of 14/15 Šaʿbān, which equals November 1021 CE. As this is purely a financial agreement and in Judaeo-Arabic, the parties use the Hijra reckoning. 331 Like Mevoraḵ b. Joseph a decade or more before, Abū Naṣr Salāma ibn Saʿīd ibn Ṣaḡīr commissioned a copy of the Masoretic Bible, presumably for his own use. While we do not today possess a Bible with an ownership colophon for Salāma, Samuel copied at least one other partial copy of the Bible like this. The manuscript described by Richard Gottheil as number 27 in his famous 1905 article “Some Hebrew manuscripts in Cairo” is a copy of the Former Prophets with a colophon by Samuel b. Jacob. Gottheil 27 seems to have been written for a Yaḥyā b. Jacob.68 Salāma ibn Saʿīd ibn Ṣaḡīr (Arabic, ‘little’), or Solomon b. Saʿadya in the Hebrew version of his name, was a leading financier and philanthropist in Fusṭāṭ in the first quarter of the 11th century.69 He was a respected figure, serving as a trustee of orphans and as a fundraiser for the Jerusalem Yešiva.70 Unlike Mevoraḵ, he was almost certainly a Rabbanite and not a Karaite, as his connections to the Jerusalem Academy suggest. Nevertheless, as a man of substance – in social, political and economic terms – he had a desire to possess a fine copy of the Bible, produced by one of Fusṭāṭ’s leading scribes. The sum of money to be paid for producing this copy of the Prophets and Writings is considerable, 25 dinars. Goitein states that the average price of a complete codex of the Bible in the Classical Genizah Period was 20 dinars.71 He further estimates that an average monthly income might have been 2 dinars, and a yearly rent on a middle-class home, 5–6 dinars.72 Thus 25 dinars for Samuel’s partial copy of the Bible seems like a suitably high price for what presumably would have been a luxurious volume, like Leningrad.73 Salāma ibn Saʿīd probably kept the Bible for himself, rather than dedicating it to a synagogue, as often occurred, since around 1110 CE we find, in a letter by Nathan ha-Kohen b. Mevoraḵ from Ashqelon, that a widow, known as ‘the daughter of Ibn Ṣaḡīr’ (‫)בנת בן צגיר‬, had a number of biblical codices in her 68 For the colophon of Gottheil 27, see Gottheil, 1905, pp. 636–637. A further description of the manuscript is in Beit-Arié, Sirat and Glatzer, 1997, p. 118. This manuscript is now in private hands. 69 CUL T-S 18J2.16 is a deed written in 1026 by Abraham the son of the Palestinian Gaʾon Solomon b. Judah appointing a guardian for an orphan. Salāma ibn Ṣaḡīr is named as overseer of the guardianship. Since the deed is in Hebrew, Salāma is introduced with the Hebrew version of his name: ‫‘ שלמה בר סעדיה הידוע בן צגיר‬Solomon b. Saʿadya who is known as Ibn Ṣaḡīr’. 70 Gil, 1992, pp. 254, 428, 602, 609. The Genizah has preserved documents referring to him from the period 1021–1026 CE, though he was possibly still active in the 1060s; see CUL T-S 10J20.13, edited in Gil, 1997, vol. IV, pp. 673–675. 71 Goitein, 1967–1993, vol. I, p. 259. This is perhaps at the more luxury end of the market; Gil, 1992, p. 234 suggests 12 to 13 dinars. 72 Goitein, 1967–1993, vol. I, p. 358, vol. IV, pp. 94–95. 73 We know how magnificent RNL Evr. I B19a looks, but we can also see that Samuel put a similar effort into his separate books of the prophets. Gottheil 27 is described as “magnificently written in beautiful characters, three columns to the page, plentiful Masora” and it is embellished with “Bible verses in letters of heroic size and in golden rims” and “gold borders”; see Gottheil, 1905, pp. 636–637. 332 possession, including a volume containing the Former Prophets and a volume containing the Latter Prophets and Writings.74 In the agreement, Samuel testifies that he produced a Torah for Abū Isḥaq Ibrahīm ibn Ḥujayj, and that this will be the model for the Prophets and Writings that he will produce for Salāma ibn Saʿīd. Perhaps Salāma had seen the copy produced for Ibrahīm and decided that he wanted one as magnificent for himself. Certainly Salāma and Ibrahīm knew each other, as both were major figures in Fusṭāṭ: Salāma was a money changer with links to the Palestinian Academy in Jerusalem and Abū Isḥaq Ibrahīm ibn Ḥujayj, known usually by his Hebrew name Abraham ha-Kohen b. Haggai, was a parnas – an administrator of the public charity – for the Palestinian leadership in the town.75 In this role he was known for helping newcomers, visitors and scholars, a fact that is celebrated in public acclamation of his name.76 The agreement delivers two dinars of Samuel’s fee into the hands of Abraham. Perhaps this was a fee for the use of the book, or maybe it was money that Samuel owed to Abraham and was thereby paying off. It is quite possible that Samuel owed a debt to Abraham ha-Kohen b. Haggai, though maybe more moral than financial, since we find him seeking his help in a letter that was probably sent some years before. CUL T-S 10J10.4 is a letter from Samuel b. Jacob to Abraham b. Haggai seeking his charitable support. This document was edited in full by the historian Mark Cohen in 2005, who stated that Samuel was writing to Abraham “supplicating his help and appealing to his reputation as a generous benefactor”.77 He inferred that the writer was a “recent arrival” and that his use of Hebrew suggested “that he hailed from a European country”. Like Bareket, Cohen did not recognise the name of Samuel b. Jacob as that of the scribe of Codex Leningrad. But the combination of ‘Samuel b. Jacob’ (in itself, not a common combination of names in the Genizah world) and ‘Abraham b. Haggai’, as found in the legal agreement, ensures that we are dealing with that very man here. Letter from Samuel b. Jacob to Abraham b. Haggai CUL T-S 10J10.4 ‫בשם רחום‬ ‫ויאחז צדיק דרכו וטהר יד׳ וג׳‬ .1 .2 74 Now damaged, they are sold for 12 dinars. Gil points to the possible connection between her ownership of the volumes and the deed T-S 10J5.15; see Gil, 1992, p. 234–235 and n. 8. The letter is CUL T-S 10J5.21, edited in Gil, 1983, vol. III, pp. 484–486. 75 Gil, 1992, p. 613. A letter from the Palestinian Gaʾon Solomon b. Judah is addressed to him in Hebrew and Arabic script, giving both the Hebrew and Arabic versions of his name side by side; Bodleian Libraries MS Heb. C.28.44, edited in Gil, 1983, vol. II, pp. 280–283. 76 For instance, the letter edited by Mann, which tells Abraham that the congregation in the synagogue at Damascus blessed him over the Torah scrolls for his kindness to a traveller from Byzantium; see Mann, 1920–1922, vol. I, p. 104, vol. II, pp. 113–115. 77 Cohen, 2005b, pp. 66–67. 333 ‫‪.3‬‬ ‫‪.4‬‬ ‫‪.5‬‬ ‫‪.6‬‬ ‫‪.7‬‬ ‫‪.8‬‬ ‫‪.9‬‬ ‫‪.10‬‬ ‫‪.11‬‬ ‫‪.12‬‬ ‫‪.13‬‬ ‫‪.14‬‬ ‫‪.15‬‬ ‫‪.16‬‬ ‫‪.17‬‬ ‫‪.18‬‬ ‫‪Margin‬‬ ‫‪.1‬‬ ‫‪.2‬‬ ‫‪.3‬‬ ‫‪.4‬‬ ‫‪.5‬‬ ‫‪.6‬‬ ‫‪.7‬‬ ‫‪.8‬‬ ‫שלום שלום לאיש השלום מארץ השלום שלום שלום‬ ‫[כ]לט פניו באדרתו ויגלום והנה חלמתי חלום הוא כב׳‬ ‫גד׳ קד׳ מר׳ ור׳ אברהם הזקן היקר החכם והנבון […]‬ ‫ירא שמים האוהב תורה ובעליה ומתנתו רחבה‬ ‫בסתר [וב]גלוי לכל יברכו בכל מכל וכל בן כב׳ גד׳‬ ‫קד׳ מר׳ ור׳ חגי נ׳׳בג אודיע לאדוני הזקן היקר כי אני‬ ‫מן הזוכרים מעשיך ומתגעגיע בטובותיך ומעת‬ ‫שהגעתי אל הנה מתאוה לראות את אדוני הזקן ובכל <עת>‬ ‫מתפלל אני עליך ולא באתי אל הנה אלא מתוך […]‬ ‫גדול וטובות אדוני הזקן מ[…]‬ ‫אליו כווסתו הנאה ומעת בא[…]‬ ‫מכלום אדם ואם יעש[…]‬ ‫בדבר שאתפרנס […]‬ ‫אפילו פרוטה […]‬ ‫מהק׳ ב׳ ה׳ […]‬ ‫לו אח[…]‬ ‫ויבשריהו במחילה וסליחה‬ ‫[…]ה כל מה שהוא עושה עמי‬ ‫כקרבן כליל ע[ל] גבי המזבח ויזכה לחזות בנועם‬ ‫ייי ולבקר בהיכלו ואילולי שאני יודיע ווסתו וחסדו וטובו‬ ‫לא הטרחתי עליו ומנוחת אב[‪ ..‬ו]אין בידי דבר ואין לאל ידי‬ ‫ימציאך הקב׳׳ה חן וחסד לפניו ולפני כל רואיך עקב שלום‬ ‫תלמידו שמואל בריבי יעקב הרב‬ ‫בריבי שמואל הרב נ׳׳בג‬ ‫‪1. In the merciful Name.‬‬ ‫‪2. “The righteous will hold fast to their way, and the clean of hands”,‬‬ ‫‪etc.78‬‬ ‫‪3. Many greetings to the man of peace from the land of peace, many‬‬ ‫‪greetings.‬‬ ‫‪4. [When he] wrapped his face in his cloak and rolled it up.79 Now, I‬‬ ‫‪have had a dream:80 it is the honourable,‬‬ ‫‪5. great, holy, our master and teacher Abraham the precious elder, the‬‬ ‫]…[ ‪wise and understanding‬‬ ‫‪, which makes it difficult to be‬יד ‪A quotation from Job 17:9. There is a hole in the paper after‬‬ ‫‪ rather‬יד׳ וג׳ ‪certain of the reading, but it looks like Samuel abbreviated the end of the verse to‬‬ ‫‪.‬ידים ‪than writing‬‬ ‫‪ ‘he wrapped his face in his mantle’, and‬וילט פניו באדרתו ‪79 This line merges 1 Kings 19:13,‬‬ ‫‪ ‘and Elijah took his mantle and wrapped it together’, piv‬יקח אליהו את־אדרתו ויגלם ‪2 Kings 2:8,‬‬‫‪ ‘his mantle’, which occurs in both. Cohen, 2005b, p. 66 instead‬אדרתו ‪oting around the form‬‬ ‫‪, ‘[…] his face and reveal them’.‬גלה ‪ as the verb‬ויגלום ‪takes‬‬ ‫‪80 Genesis 37:9.‬‬ ‫‪78‬‬ ‫‪334‬‬ 6. fears Heaven, who loves Torah and its masters, and whose offering is generous 7. (whether) in secret (or) revealed to everyone. May He bless him in every way, son of the honourable, great, 8. holy, our master and teacher Haggai – his rest be in Eden. I inform my lord the precious elder that I 9. am one of those who recall your [good] deeds and I long for your welfare, and from the moment 10. that I arrived here I have desired to see my lord the elder and at every moment 11. I pray for you. And I only came here out of great [need?], 12. and the good [deeds] of my lord the elder […] 13. to him according to his worthy custom. And from the time of [my arrival …] 14. from any man. And if he should do […] 15. with something with which I might sustain myself […] 16. even a coin […] 17. from the Holy One Blessed Be He […] 18. to him […] Margin 1. and may He gladden him with a pardon and forgiveness […] 2. […] whatever he does with me 3. like a complete sacrifice upon the altar, and may he merit seeing the delightfulness 4. of the LORD and to visit his temple. And were it not that I know his customary behaviour, his kindness and his generosity 5. then I would not have troubled him and the rest of [… And] I have nothing and I am helpless. 6. May the Holy One Blessed Be He let you find grace and kindness before Him and before all who see you. A reward of peace.81 7. His scholar82 Samuel son of the scholar Jacob ha-Rav 8. son of the scholar Samuel ha-Rav, his rest be in the garden of Eden.83 The phrase ʿeqev šalom is a ʿalāma, a motto attached to correspondence, adopted from the Islamic practice. High-ranking individuals might have their own, or correspondents would use those of the leader of the time, or the recipient of the letter, to show loyalty. 82 The ‘his’ is a gesture of humility, rather than marking a previous relationship between the two. Talmid ‘disciple, scholar’ seems to change meaning over the Geonic period. Originally referring to as-yet-unordained scholars of the Academy (Mann, 1920–1922, vol. I, p. 54 n. 2), it comes more to denote a scholar who has acquired his learning outside of the academies, in the midrašim of eminent scholars; see Goitein, 1967–1993, vol. V, p. 266, “select few who attained a high degree of scholarship while studying with renowned masters far away from the seats of yeshivas”. We might call them ‘independent scholars’. 83 Samuel’s father, Jacob, is still alive, but his grandfather, Samuel, is dead. Samuel was named after his paternal grandfather, as was the usual custom among the Jews of the Islamic world (Goitein, 1967–1993, vol. III, pp. 6–7). This is evidence that he was probably not from Europe. 81 335 This letter is asking for favour, charity or financial support. It is a common genre in the Genizah, where travellers, immigrants and the indigent were often forced to turn to local sources of funding. Samuel informs Abraham that he is a new arrival (in Fusṭāṭ, where Abraham was based), having arrived here not by choice but probably out of ‘great [need]’ (line 11), and he now needs further assistance. It evidently dates from the days before Samuel became a scribe capable of charging twenty-five dinars to produce a glorious copy of the Bible. Abraham b. Haggai was a natural figure to turn to.84 He had a reputation for helping scholars of the Torah (i.e. pious men, line 6), perhaps in a role as a patron of the arts, and as noted above he was a parnas, one of several at the time, who supervised charitable collections and looked after the community’s charitable foundations.85 In the previous document, the agreement, it is clear that Abraham was known to Samuel – or rather the other way around, as their differing social status demanded. This letter, where Samuel appears to know Abraham only by reputation, must predate the agreement of 1021 by many years, since in the meantime Samuel became acquainted with Abraham and copied a Pentateuch for him. By 1021, the relationship appears to be that of patron and client, and this letter is probably evidence of the first steps that Samuel took in securing the patronage of the wealthy Abraham.86 Though Samuel is in need at the time of writing this letter, because he has left his previous place of residence, he is not without a potential source of livelihood, as his appellation talmid ‘scholar’ suggests that he had enjoyed the benefit of an education at the feet of a sage or in a house of study. Certainly his later accomplishments strongly suggest a solid education in the Bible and masora, as well as a thorough knowledge of contemporary literary Hebrew. He is educated and only requires assistance to get on his feet in the new town, through the patronage of a wealthy and connected figure. Where might Samuel have come from? Cohen believes him to be European, since he wrote his letter to Abraham in Hebrew. But this proves nothing of his origins, only of his education. Indeed, the letter is well constructed and mixes biblical phraseology in a playful manner alongside the characteristic contemporary language of letters. Letters seeking favour from social superiors are often written in Hebrew; it reflected favourably on the education of the writer and the respect he held for the addressee. In the 10th and 11th centuries Hebrew was particularly in vogue as an idiom of communication, and many letters between officials, scholars and other people of status in the community are written in the Holy Language.87 84 CUL T-S 13J23.9 is another request for help sent to Abraham, from a cantor called Yaʿīš b. Sahl al-Nahrawānī, a Persian. 85 Abraham’s brother Isaac, a community leader, served in a similar role. On the parnasim in general, see Gil, 1976, pp. 47–53; Cohen, 2005a, pp. 211–216. 86 On reading ‘letters of appeal’ as requests for patronage, see Cohen, 2005a, pp. 174–188. 87 On the use of Hebrew, as opposed to Judaeo-Arabic, for communicative purposes in the classical Genizah period, see Outhwaite, 2013. 336 Other details of the letter suggest that Samuel was not from a European land: the use of the title talmid and the use of an ʿalāma and an opening invocation; the fact he was named after his grandfather, which as mentioned earlier was common in the Islamic lands of the Genizah; and the layout of his letter, leaving a clear margin at the top and on the right-hand side, again very common in letters from the Genizah world.88 The importance of this letter is not just that it shows that Samuel b. Jacob was probably not a native of Fusṭāṭ, but that it also throws light on his ancestry. We could certainly have guessed that his grandfather’s name was Samuel, but this confirms it. More importantly, however, it shows that both his father and grandfather had a title after their name, ha-Rav. The use of rav before people’s names is just a common courtesy, and it usually occurs in the form of the phrase ‫ מרנא ורבנא‬or ‫מורנו ורבנו‬, or just abbreviated to ‫ מ׳ ור׳‬or similar, and is a polite appellation used for most anyone, ‘master and teacher’. When the noun rav follows the name, however, it appears to have been a specific title, bestowed originally by the Babylonian Yešivot on those who could function as a jurisconsult in the Maghrebi Jewish communities.89 Šemarya b. Elḥanan, head of the Jews in Egypt in the late 10th century, and the leading member of the Babylonian party in Fusṭāṭ, had the title ‫‘ הרב הראש‬the Great Rav’ after his name. His origins are obscure, though he probably spent time in Babylon and North Africa before settling in Fusṭāṭ. Following him, a number of leading Maghrebi figures in Egypt have the appellation ha-Rav, including Nahray b. Nissim from Qayrawān, the merchant-scholar who led the Babylonians, and the great scholar Judah ha-Kohen b. Joseph, who was often simply known as Ha-Rav.90 Given the connection of the title ha-Rav with the Maghreb and the Babylonian congregation, Samuel b. Jacob’s roots probably therefore lay in North Africa, in the Babylonian sphere of influence.91 With both his father and grandfather styled ha-Rav, he came from a line of scholars who had also achieved prominence in their community. This would account for the high standard of educational attainment that Samuel had clearly acquired. His own appellation of talmid implies advanced study at a college or with a sage, as 88 Compare Byzantine Hebrew letters, for instance, which write across the full width of the page and do not imitate the basmala in their opening; see Outhwaite, 2009, p. 198. 89 Bareket, 2017, p. 190, says: “According to Goitein, this title was only popular amongst the Maghrebi community and was not acceptable in Egypt. Hence, only figures from Maghrebi origin, such as Nehorai [=Nahray] ben Nissim, carried this title.” Cohen, 1980, p. 103 agrees: “Nahray functioned as professional jurisconsult to the Jewish community of Old Cairo [=Fusṭāṭ]. That is to say, like the Muslim mufti, he issued legal opinions, although not as a judge. Such authorities were styled ‘the rav’ (ha-rav), ‘the master’, a title peculiar to Jewish scholars from the Muslim west.” 90 Cohen, 1980, pp. 102–105. 91 Samuel’s familiarity with the Babylonian masora is well attested in other copies of the Bible he produced such as Lm (Gottheil 14, now locked away in private hands) and the fragments that have been recently discovered in the Cambridge Genizah Collection; see Phillips, 2016, pp. 289–291. 337 befits the child of such a line, but his lack of the title ha-Rav suggests that he was not so successful politically. Indeed, we should not expect him to be writing Bibles for a living had he been more successful in community affairs.92 On top of giving us the single most important complete codex of the medieval Hebrew Bible, Samuel b. Jacob has left behind a considerable legacy of professional work and a growing number of documentary sources.93 As a result, he is arguably the most important medieval Jewish scribe that we know of, though, until recently, we knew very little about the man himself or how he worked. From a close reading of the plain colophon in the Leningrad Codex and of the documentary sources in the Cairo Genizah presented above, we can now fill out some of the context in which he produced his work and begin to appreciate what sort of role he played. The documents suggest he came from a family affiliated with the Babylonian Academies, and that his immediate ancestors had accrued some prominence in communities of Maghrebi Jews (the title ha-Rav). He himself had a good education, which continued into adulthood with an intellectual apprenticeship of some kind, probably in North Africa, though not necessarily so (the title talmid). His Medieval Hebrew idiom is accomplished and imaginative. If he was not a writer of piyyuṭ himself, then he was probably very familiar with the genre. He sought, and obtained, the patronage of a leading member of the Rabbanite Palestinian community in Egypt, but he accepted commissions from members of the Maghrebi (specifically, Persian-Maghrebi) Karaite community. Clearly, his services were sought after from across the religio-political spectrum in Egypt – Babylonian and Palestinian, Karaite and Rabbanite. And he achieved enough reputation in his field that he could charge a considerable sum for the production of a beautiful model codex of the Bible, of the kind that would have been an ostentatious, tangible sign of its owner’s piety, good taste and wealth. And this is an important point to make. The Fusṭāṭ community, and other communities like it, produced and cultivated professional scribes of the calibre of Samuel b. 92 Further evidence of his family’s prominence can be found in another letter, CUL T-S 13J15.13, which is thanking a leading Karaite courtier, Abū Saʿd b. Sahl al-Tustarī. It is written and signed by one ‫יעקב הרב בר׳ שמואל הרב בר׳ אברהם הרב‬, Jacob ha-Rav son of the scholar Samuel ha-Rav son of the scholar Abraham ha-Rav (who also uses the same ʿalāma as Samuel in his letter, ‫)עקב שלום‬. This is probably Samuel b. Jacob’s father, and it supplies the name of his greatgrandfather, Abraham, who also held the title Rav. The letter, which was perhaps never sent (as it leaves a blank space to fill in the recipient’s Hebrew name), was originally edited by Scheiber, though he made no connection with Samuel b. Jacob the scribe; see Scheiber, 1969, pp. 215– 218. There are other traces in the documentary record of Samuel b. Jacob and of his family, but they do not add much more to the story for now. For instance, his name appears as the scribe of a divorce settlement (CUL T-S 10J27.12) in 1009 CE, and the same Samuel b. Jacob witnessed a deed of indemnity (CUL T-S 24.11) in 1002 CE. 93 His works of biblical copying include RNL Evr. I B19a, Gottheil 14 (Lm), Gottheil 27, a copy of the Bible with Saʿadya’s Tafsīr and an innovative masora (mostly in the Russian National Library as RNL Evr. II C1, but with leaves in the Genizah as well), further leaves of Bibles in the Taylor-Schechter Collection of the Cairo Genizah, as well as further items in the Russian National Library. 338 Jacob, but they would not have existed without a class of wealthy, pious and scholarly individuals, who could extend patronage to them and commission the works of art that they laboured over. References Bareket, E., 1995, Jewish Leadership in Fustat (Hebrew), Tel Aviv: Diaspora Research Institute, Tel Aviv University. ———, 1998, “Parašat Avraham ben Šemuʾel ha-Sefaradi ʿal šemona mismaḵim min ha-Geniza” (Hebrew), in E. Fleischer, M. A. Friedman and J. A. Kraemer (eds), Mas’at Moshe: Studies in Jewish and Islamic Culture Presented to Moshe Gil, Jerusalem: Mosad Bialik, pp. 124–136. ———, 2017, Eli ben Amram and his Companions: Jewish Leadership in the Eleventh-Century Mediterranean Basin, Eastbourne: Sussex Academic Press. Beit-Arié, M., Sirat, C. and Glatzer, M., 1997, Codices Hebraicis litteris exarati quo tempore scripti fuerint exhibentes, vol. I: Jusqu’à 1020, Turnhout: Brepols. Cohen, M. R., 1980, Jewish Self-Government in Medieval Egypt: The Origins of the Office of Head of the Jews, ca. 1065–1126, Princeton: Princeton University Press. ———, 2005a, Poverty and Charity in the Jewish Community of Medieval Egypt, Princeton: Princeton University Press. ———, 2005b, The Voice of the Poor in the Middle Ages: An Anthology of Documents from the Cairo Geniza, Princeton: Princeton University Press. Freedman, D. N. (ed.), 1992, The Anchor Bible Dictionary (6 vols), New York: Doubleday. ——— et al. (eds), 1998, The Leningrad Codex: A Facsimile Edition, Grand Rapids: Wm. B. Eerdmans and Leiden: Brill. Friedman, M., 1980, Jewish Marriage in Palestine: A Cairo Geniza Study (2 vols), Tel Aviv: Tel Aviv University and New York: Jewish Theological Seminary of America. Gil, M., 1976, Documents of the Jewish Pious Foundations from the Cairo Geniza (Hebrew), Leiden: Brill. ———, 1983, Palestine During the First Muslim Period (634–1099) (Hebrew; 3 vols), Tel Aviv: University of Tel Aviv and the Ministry of Defence. ———, 1992, A History of Palestine, 634–1099, Cambridge: CUP. ———, 1997, In the kingdom of Ishmael (Hebrew; 4 vols), Jerusalem: Mosad Bialik. Glatzer, M., 1989, “The Aleppo Codex: codicological and palaeographical aspects” (Hebrew), Sefunot 19/4, pp. 167–276. Goitein, S. D., 1967–1993, A Mediterranean Society: The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Geniza (5 vols with an index volume by P. Sanders), Berkeley: University of California Press. ———, 1973, Letters of Medieval Jewish Traders, Princeton: Princeton University Press. Golb, N., 1965, “The topography of the Jews of medieval Egypt”, Journal of Near Eastern Studies 24/3, pp. 251–270. Gottheil, R., 1905, “Some Hebrew manuscripts in Cairo”, JQR 17/4, pp. 609–655. Khan, G., 2000, The Early Karaite Tradition of Hebrew Grammatical Thought, Leiden: Brill. 339 Mann, J., 1920–1922, The Jews in Egypt and in Palestine under the Fāṭimid Caliphs: A Contribution to their Political and Communal History Based Chiefly on Genizah Material Hitherto Unpublished (2 vols), London: OUP. Ofer, Y., 1999, “A Masoretic reworking of ‘Maḥberet Menaḥem’ ” (Hebrew), Leshonenu 62/4, pp. 189–255. Olszowy-Schlanger, J., 1998, Karaite Marriage Documents from the Cairo Geniza: Legal Tradition and Community Life in Mediaeval Egypt and Palestine, Leiden: Brill. Outhwaite, B. M., 2000, A Descriptive Grammar of the Medieval Hebrew of the Cairo Geniza Letters (unpubl. diss., University of Cambridge). ———, 2009, “Byzantium and Byzantines in the Cairo Genizah: new and old sources”, in N. De Lange, J. Krivoruchko and C. Boyd-Taylor (eds), Jewish Reception of Greek Bible Versions, Tübingen: Mohr Siebeck, pp. 182–220. ———, 2013, “Lines of communication: medieval Hebrew letters of the eleventh century”, in E.-M. Wagner, B. Outhwaite and B. Beinhoff (eds), Scribes as Agents of Language Change, Berlin: De Gruyter, pp. 183–198. ———, 2016, “Samuel ben Jacob: the Leningrad Codex B19a and T-S 10J5.15”, Fragment of the Month January 2016, http://www.lib.cam.ac.uk/collections/ departments/taylor-schechter-genizah-research-unit/fragment-month/fragmentmonth-5. ———, 2017, “The first owners of the Leningrad Codex: T-S 10J30.7”, Fragment of the Month November 2017, http://www.lib.cam.ac.uk/collections/departments/ taylor-schechter-genizah-research-unit/fragment-month/fragment-month-17. Phillips, K., 2016, “The Masora Magna of two biblical fragments from the Cairo Genizah, and the unusual practice of the scribe behind the Leningrad Codex”, Tyndale Bulletin 67/2, pp. 287–307. Rand, M., 2006, Introduction to the Grammar of Hebrew Poetry in Byzantine Palestine, Piscataway, NJ: Gorgias Press. Scheiber, A., 1969, “Further letters addressed to Abraham Ben Yashar”, JQR 59, pp. 215–221. Stern, D., 2017, The Jewish Bible: A Material History, Seattle: University of Washington Press. Würthwein, E., 2014, The Text of the Old Testament: An Introduction to the Biblia Hebraica, 3rd edn, A. A. Fischer (rev.), E. F. Rhodes (trans.), Grand Rapids, MI: Wm. B. Eerdmans. Yeivin, I., 1980, Introduction to the Tiberian Masorah, Missoula, MT: Scholars Press. 340 Arabic Vocalisation in Judaeo-Arabic Grammars of Classical Arabic NADIA VIDRO University College London Many Judaeo-Arabic texts use Arabic vocalisation signs. In the vast majority of such texts, vocalisation is sporadic, and rarely includes case endings, tanwīn or other elements typical of fully vocalised classical Arabic texts. A much smaller group of Judaeo-Arabic texts – most if not all of which were originally composed in Arabic script and later transcribed into Hebrew characters – are consistently vocalised with Arabic signs. Examples include Judaeo-Arabic fragments of the Qurʾān (Halle DMG Arab 5),1 of al-Ḥarīrī’s Maqāmāt (L-G Ar. 2.73), and of medical (Mosseri I.126.2, IX.124, X.30.1) and grammatical works (T-S NS 301.25). In addition, there are fully vocalised manuscripts that use a combination of Tiberian vowels with Arabic signs such as waṣla and tanwīn, which are not found in the Tiberian system; for example, a complete copy of the Qurʾān in Judaeo-Arabic (Ox. Bodl. Hunt. 529). Few studies of Arabic vocalisation in medieval Judaeo-Arabic texts exist. E. Rödiger included a relatively detailed analysis of the Arabic vocalisation in his description of the Judaeo-Arabic Qurʾān fragment Halle DMG Arab 5, highlighting a number of instances of non-standard vocalisation.2 Recently, E.-M. Wagner has studied Arabic vocalisation marks in Judaeo-Arabic letters and legal documents written by Ḥalfon b. Manasse, an early 12th century Jewish court scribe. Wagner suggests that this scribe may have become familiar with Arabic vocalisation practices through copying Arabic books into Hebrew characters, subsequently pioneering the use of Arabic signs in Jewish documentary texts.3 A study of Arabic vocalisation in Judaeo-Arabic texts is important for a number of reasons. Firstly, by identifying instances of non-classical vocalism such study can contribute to our knowledge of the phonology of medieval Arabic, in both its Jewish and its Muslim varieties, given that vocalisation marks 1 For a transcription (without vowels) and a facsimile of this manuscript, see Paudice, 2009, pp. 230–239, 252–257. For a study of the manuscript, see Rödiger, 1860. 2 Rödiger, 1860, pp. 485–489. 3 Wagner, forthcoming. See Ox. Bodl. Heb. e.74.1–6, a Muslim letter formulary, transliterated into Hebrew by Ḥalfon b. Manasse. 341 in texts transcribed into Hebrew characters could have been copied from Arabic Vorlagen.4 Secondly, it can inform our ideas on medieval Jewish education in Classical Arabic and its scribal conventions, shedding light on the level of Jews’ knowledge of Arabic vocalisation rules, the kinds of people who might have had this knowledge, the periods when Arabic vocalisation marks were used by Jewish scribes in texts of different types, and the role Judaeo-Arabic texts consistently vocalised with Arabic vocalisation signs might have had as teaching materials for learning Classical Arabic pronunciation and vocalisation rules. To answer the latter set of questions, a systematic study of JudaeoArabic manuscripts with Arabic vocalisation signs is required, based on a corpus of sources that includes texts that were transcribed from Arabic as well as those that were originally written in Judaeo-Arabic. This article makes a small contribution to the programme of research outlined above by analysing the Arabic vocalisation in Judaeo-Arabic manuscripts transcribed from Arabic Vorlagen, based on my work on a corpus of Classical Arabic grammars copied in Hebrew characters and preserved in the Cairo Genizah and in the Firkovich Collections in the National Library of Russia.5 The article consists of an edition of a grammatical fragment vocalised with Arabic signs, accompanied by a study of its spelling and vocalisation in the context of linguistic features reflected in other Judaeo-Arabic grammars of Classical Arabic and vocalised Judaeo-Arabic texts. 1 T-S NS 301.25 T-S NS 301.25 is a well-preserved one-folio fragment measuring 20.5cm x 12.5cm. The folio carries two unrelated texts: on recto, a grammar of Classical Arabic is copied in Judaeo-Arabic, in a 12th–13th century Egyptian handwriting;6 on verso, in a different Egyptian 12th–13th century hand, there is a dirge for a communal official who bore the title Nagid.7 The grammar on T-S NS 301.25 recto has been identified by Dr Almog Kasher from Bar-Ilan University as a passage from Kitāb al-Jumal fī al-Naḥw by Abū al-Qāsim ʿAbd al-Raḥmān b. Isḥāq al-Zajjājī, a 10th century Arab grammarian.8 Kitāb al-Jumal fī al-Naḥw is an introduction to Classical Arabic 4 For studies of Judaeo-Arabic Genizah fragments vocalised with Tiberian vocalisation signs see Blau and Hopkins, 1998, pp. 195–254; Khan, 1992, pp. 105–111; Khan, 2010, pp. 201–218. 5 I thank Dr José Martínez Delgado (University of Granada) for drawing my attention to the sources in the Firkovich Collections. I thank Dr Almog Kasher (Bar-Ilan University) for his comments on this article, as well as his cooperation and expert advice on the Arabic grammatical tradition. For studies of Judaeo-Arabic grammars of Classical Arabic, see Basal, 2010 and Vidro and Kasher, 2014. 6 I thank Dr Amir Ashur of Tel Aviv University for assessing the manuscript’s handwriting. 7 Published in Allony, 1991, pp. 460–461. I thank Dr Michael Rand for his help with the poem. 8 Kitāb al-Jumal fī al-Naḥw is edited in Cheneb, 1927 and Al-Ḥamad, 1996. See also Sezgin, 1984, pp. 88–94; Zabara, 2005; Binaghi, 2015. 342 grammar written for beginners, in which Al-Zajjājī presents the rules of grammar accompanied by multiple examples and explains grammatical terminology. Numerous Arabic script copies of and commentaries on Kitāb al-Jumal exist, testifying to its popularity in the Muslim world, especially in al-Andalus.9 Kitāb al-Jumal was well known to Andalusian Jewish grammarians, as is shown by quotations from it identified in Jonah ibn Janāḥ’s Kitāb al-Lumaʿ and in Isaac ibn Barūn’s Kitāb al-Muwāzana bayn al-Lugha al-ʿIbrāniyya wal-ʿArabiyya.10 Copied in the 12th–13th century, T-S NS 301.25 is one of the earliest surviving manuscripts of Kitāb al-Jumal.11 The preserved text belongs to The Chapter on Knowing the Markers of Inflection (bāb maʿrifat ʿalāmāt al-iʿrāb) and forms the closing section of the chapter.12 Below this, The Chapter On Verbs is announced but is not copied, leaving a large empty space at the bottom of the page. The text is consistently vocalised with Arabic signs. 2 Edition13 ْ‫ ומَא‬14‫ם יَפْעَלَא ולَםْ יَפْעَלُו‬ ْ َ‫ וגَמْעَהَא נَחוَ קَולِךَ ל‬.1 ‫َאב‬ ُ ‫ אשْבَהَ ד َِל َך פَג ِْמ‬.2 ِ ‫ْמאת אלאٔעْר‬ ِ ‫יע עَלא‬ ‫ אَרْבَעَ עَשרהَ עَלאْמהً אَרْבَעُ ללרَפْעُ וَכَמْסﹲ‬.3 16 ‫ ואתْנתَאןِ ללגَזْם‬15‫ ללנَצْבِ וَת َْלתَ ללכَפ ِْצי‬.4 َ‫ וגَמِיעُ מَאْ יُערَבُ בِ ِה אٔלْכَלאْםُ דסْעَהُ אَשיא‬.5 َ ‫ חَרכَאתٍ וَהי א‬18ُ‫ ת ََלת‬.6 ‫ٔלצמَהُ ואٔלפَתْחَהُ ואَלכَסר ُה‬ 17 ‫ וَאליٓאُ ואٔלףُ וَאלנُון‬21‫ אלוَאו‬20َ‫ אחרُףٍ וَהי‬19‫ וَאَרْבَ ُע‬.7 9 Binaghi, 2015, pp. 339–348. See Becker, 1998, pp. 44–46, 57 and Becker, 2005, pp. 66–67. 11 The earliest identified copy in the Arabic script is dated 1207 CE (Binaghi, 2015, p. 173). 12 Cheneb, 1927, pp. 18–21, esp. p. 21; Al-Ḥamad, 1996, p. 3–6, esp. p. 6. 13 A transcription of this fragment, without vocalisation signs and identification, can be found on the Friedberg Genizah Project (FGP) website, https://fjms.genizah.org/. 14 In Cheneb, 1927, p. 21 and Al-Ḥamad, 1996, p. 6 the elision of the final nun in 2fsg verbs is also mentioned, exemplified by ‫لم تفعلي‬. This passage is worded and placed slightly differently in the editions. The omission of this passage in the early Judaeo-Arabic copy, together with its instability in the editions, suggests that it is a gloss which made its way into the main body of the text. 15 ‫ ללכَפْצי‬and ‫ללגזמי‬, corrected to ‫( ללגَזْם‬see n. 16), with the plene spelling of the short /i/ of the ِ genitive, may have originated in the process of dictation or of ‘inner dictation’ when copying from a model. Alternatively, the spelling ‫ ללכפצי‬could be explained by the graphic similarity between the Arabic ‫ ض‬and ‫ضى‬. 16 Originally ‫ללגזמי‬, corrected to ‫ללגזם‬. 17 This vowel sign is barely legible and uncertain. 18 Originally ‫תלאת‬, corrected to ‫תלת‬. 19 The expected form is ‫ארבעה‬. 20 This vowel sign is barely legible and uncertain. 21 Ink traces are preserved above the final waw, but the vowel is uncertain. 10 343 24 23 22 ٌ ]‫ [ו‬.8 ِ‫חדף וסכُון َלאْ יכُוןُ מُעْרِבَ פِי שיא ِמן אלכَלאْם‬ 25 ‫ ِא ّלא בَאח َِד הَדה אל ٔאשْיَא‬.9 ‫ באב אלאפעאל‬.10 3 3.1 Analysis Spelling and vocalisation reflecting non-standard pronunciation Although T-S NS 301.25 is a copy of a grammar of Classical Arabic, and its spelling and vocalisation were undoubtedly intended to represent Classical Arabic, some of its readings indicate non-standard pronunciations. These include: a. ‫( פَגْ ؚמי ُע‬l. 2), with a sukūn instead of the expected fatḥa for the Classical Arabic fa-jamīʿu, probably reflects a sandhi-type elision of the short /a/.26 b. The numeral three is vocalised َ‫( וَת ֻׁَ֯לת‬l. 4) in place of the Classical Arabic wa-ṯalāṯun. In the second occurrence of the same numeral in line 6, the initially written ‫ תלאת‬is corrected by overwriting to ُ‫תَלَת‬. c. The spelling ُ‫( דסْעَה‬l. 5) instead of tisʿatu reflects a voiced or an unaspirated pronunciation of /t/. The same pronunciation is attested in medieval Judaeo-Arabic letters from the Maghreb in the spelling of the name Tustarī as ‫דסתרי‬.27 d. In ‫( בَאחَדؚ‬l. 9) the preposition ‫ ב‬is vocalised ba- instead of bi-. Similar vocalisation can be found in Halle DMG Arab 5, a Judaeo-Arabic Qurʾān fragment vocalised with Arabic vowel signs, in which fatḥa is occasionally marked where kasra is expected in Classical Arabic, especially but not exclusively on the prepositions ‫ ב‬and ‫ל‬: ‫בَמَא‬, ‫בَעבאדה‬, ْ‫לَמَן‬, ّ‫לَכל‬.28 In Judaeo-Arabic texts vocalised with Tiberian signs, the preposition ‫ ב‬is occasionally vocalised with a shewa: ‫( בְ דַ אך‬T-S Ar. 53.12 r.), ‫בְ מַ א‬, ‫( בְ כֵאר‬T-S Ar. 53.12 v.). Inasmuch as the main sound value of shewa in the Tiberian reading tradition is a short /a/, and since the phonetic conditions in the above given examples are not conducive to realising the shewa as short /i/, it has been assumed that the vocalisation of the Judaeo-Arabic preposition ‫ ב‬with a shewa either 22 Ink traces are preserved above the final nun, but the vowel is uncertain. The sukūn is partially rubbed and is uncertain. 24 The final aleph may have been crossed out. 25 Two dots are visible above the aleph and the ḥet. 26 Cf. Woidich, 1991, pp. 1632–1633. 27 See Wagner, 2010, p. 35 and n. 23 there. 28 Rödiger, 1860, pp. 487–488. 23 344 reflects the Palestinian substrate pronunciation, in which the shewa stands for a short /e/, or is a Hebraism.29 The vocalisation of this preposition with a fatḥa found in manuscripts with Arabic vowel signs may hint that the intended value of the shewa here is, in fact, a short /a/ reflecting the reading ba-, possibly by hypercorrection.30 3.2 Inflectional vowels The majority of case endings in T-S NS 301.25 are correct. Exceptions are: a. ‫וג ْמעَהَא‬ َ (l. 1) should probably have the genitive case marker /i/ and not the accusative /a/. Although the preceding text is missing, the phrase according to the editions is31 ‫وحذف النون ايضا عالمة للجزم في تثنية االفعال وجمعها‬ ‘The elision of the nun is also a marker of jazm, in the dual and plural verb forms.’ It is likely that the reading in our fragment was the same, as is supported by the preserved examples ‫( לَםْ יَפْעَלَא ולَםْ יَפْעَלُו‬l. 1). If so, the genitive case ending is expected after the preposition ‫פי‬. b. ُ‫( ללרَפْע‬l. 3), where the genitive rather than the nominative ending is expected after the preposition. c. َ‫( לَאْ יכُוןُ מُע ِْרב‬l. 8), where a fatḥa on the second radical, and a nunated nominative ending -un are expected: muʿrabun.32 The active participle form muʿrib is highly unlikely in this context and appears to be a mistake. The accusative ending may be due to an erroneous parsing of ‫ יכון‬as ‘incomplete’ kāna and of ‫ מערב‬as its object. Confusion in the marking of case endings is also attested in a Kufan grammatical primer preserved in T-S Ar. 31.254, T-S 24.31 and T-S AS 155.132,33 where the name ʿAbd Allāh after a preposition is occasionally vocalised with a fatḥa – for example, ‫( עלי עבדَ אללה‬T-S 24.31 r.) – as well as in the Qurʾān fragment Halle DMG Arab 5 and in Judaeo-Arabic texts with Tiberian vocalisation.34 29 Khan, 2010, p. 209; Khan, 1992, pp. 110–111. For examples of substituting /a/ for the Classical Arabic /i/ by hypercorrection, see Khan, 2010, p. 206. 31 Cheneb, 1927, p. 21; Al-Ḥamad, 1996, p. 6. 32 Cf. Cheneb, 1927, p. 21; Al-Ḥamad, 1996, p. 6. 33 Edited and analysed in Vidro and Kasher, 2014. 34 Rödiger, 1860, p. 487; Khan, 2010, p. 205; Blau and Hopkins, 1988, p. 469, §26. 30 345 3.3 The marking of long vowels In a number of cases in T-S NS 301.25, long /ā/ is represented by an aleph vocalised with a sukūn; for example, ْ‫( מَא‬l. 1, l. 5), ِ‫( עَלאْמאת‬l. 2), ً‫( עَלאْמה‬l. 3), ُ‫( אٔלכَלאْם‬l. 5), ‫ْם‬ ِ ‫( אלכَלא‬l. 8) and ْ‫( לَא‬l. 8, example uncertain). This spelling is found in about half of the cases of long /ā/ in the fragment; in the rest of the cases the aleph is unvocalised, and other long vowels are never marked with a sukūn on the respective matres lectionis. The marking of all three matres lectionis with a sukūn is attested in Islamic manuscripts,35 and was known to Jewish scribes. It is used in Judaeo-Arabic fragments L-G Ar. 2.3, 2.4, 2.10, 2.142 of Kitāb al-Afʿāl Ḏawāt Ḥurūf al-Līn by Judah Hayyūj – for example, ٌ‫( בَאْב‬L-G Ar. 2.3 v.) and ُ‫( פאקُוْל‬L-G Ar. 2.4 v.) – as well as in an Arabic script Pentateuch commentary by Abū al-Faraj Furqān in BL Or. 2545, where a sukūn can be found on matres lectionis both in the original Arabic words (mainly for the long /ī/ and /ū/) – for example, ‫( َرسُ ْو ًال‬BL Or. 2545, f. 8 v.) – and in transliterations of Hebrew words – for example, ‫ مغ ّمزيْن‬for ‫מגמזין‬ (BL Or. 2545, f. 87 r.).36 This function of the sukūn was carried over to the Tiberian shewa in some Judaeo-Arabic texts, such as a copy of the Qurʾān in Hebrew characters in Ox. Bodl. Hunt. 529, where most long /ī/ and /ū/ vowels are represented by yod or waw with a shewa while the aleph of the long /ā/ is left unvocalised – for example, ַ ‫רבّ ִ ؐאל ָעאל ִַמיְ ן‬,ַ ‫ ו ִַאיָאָך נ ְַסתַ עִ יְ ן‬and ִ‫ַ׳יְר אؐלמַ גְ ׳צוְ ב‬ ִ ‫ג‬ (f. 1 v.) – and a liturgical fragment T-S Ar. 8.3, where the aleph of the long /ā/ is the only mater lectionis vocalised with the shewa – for example, ‫דַ ְאר‬ (f. 13 r.).37 3.4 The marking of the initial hamzat al-qaṭʿ and hamzat alwaṣl Only hamzat al-qaṭʿ is found in T-S NS 301.25, written on top of the aleph ‫ٔא‬ irrespective of its vowel, as seen in ‫( אלאٔשْיَא‬al-ʾašyāʾ, l. 9) vs. ِ‫( אלאٔעْרَאב‬alʾiʿrābi, l. 2). The hamza is marked inconsistently and is missing in such forms as ُ‫( אَרْבَע‬l. 3), َ‫( אَשיא‬l. 5), ٍ‫( אחרُף‬l. 7) and َ‫אשבَה‬ ْ (l. 2). On the other hand, it is used a number of times on the aleph of the definite article after a word ending in a vowel, where it is not pronounced according to the rules of Classical Arَ ‫( וَהי א‬l. 6). A parallel pheabic: ُ‫( וג َِמיעُ מَאْ יُערَבُ ِב ِה אٔלْכَלאْם‬l. 5) and ُ‫ٔלצמَהُ ואٔלפَתْחَה‬ nomenon, understood in secondary literature as pseudo-Classical or morphophonemic spelling, is attested in Judaeo-Arabic texts with Tiberian vocalisation signs, where alif al-waṣla after a vowel is often vocalised as if it were pronounced as a glottal stop: ‫( פִ י אַ לחִ כְ ִמה‬T-S Ar. 53.12 v.).38 35 Cf. Wright, 1996, vol. I, p. 13, §10 rem. See Tirosh-Becker, 1998, pp. 383, 386. 37 See Khan, 1992, pp. 108–109 and n. 20 there. 38 See Khan, 2010, p. 205. See also Blau and Hopkins, 1988, p. 239, §14.2. 36 346 3.5 Nunation39 The marking of nunated vowels in T-S NS 301.25 is largely in accord with Classical Arabic norms, with the exception of some cases where non-nunated vowels are found instead, for example: (l. 3) ُ‫אَרْבَעَ עَשרהَ עَלאْמהً אَרْבَעُ ללרَפْע‬ (l. 8) ‫לَאْ יכُוןُ מُעْרِבَ פِי שיא‬ In other grammars, too, tanwīn is occasionally unmarked where it is clearly intended. Thus, in T-S Ar. 5.45 the forms ُ‫זיד‬, ُ‫ עמר‬and ُ‫ בכר‬stand for Zaydun, ʿAmrun and Bakrun: ُ‫ודכלה אלתנוין כקולך זידُ ועמר‬ ‘It has the tanwīn, e.g. Zaydun (ُ‫ )זיד‬and ʿAmrun (ُ‫)עמר‬.’ (T-S Ar. 5.45, P1 r.) ُ‫אעלם אן אלרפע פי אלאסם אלואחד יכון בשיין באלצמה ואלואו פאלצמה זידُ ועמר‬ ‫ובכרُ ומא אשבה דלך עלאמה אלרפע פי הדה אל אסמא צמה אכרהא ובעד אלצמה תנוין‬ ‘Take note that the nominative case of single nouns is (expressed) by two things: the ḍamma and the waw. The examples of ḍamma are Zaydun (ُ‫ )זיד‬and ʿAmrun (ُ‫ )עמר‬and Bakrun (ُ‫)בכר‬, etc. The marker of the nominative in these nouns is the ḍamma at the end and the tanwīn after the ḍamma.’ (T-S Ar. 5.45, P1 v.) In the Kufan grammatical primer tanwīn ḍamma is never marked:40 for example, ُ‫( קאם זיד‬T-S Ar. 31.254 r.) and ُ‫( עלי אכיך תובُ גדיד‬T-S 24.31 r.). Both tanwīn fatḥa and tanwīn kasra are found in the fragments alongside their non-nunated counterparts, but the signs are used indiscriminately: ‫( רפעת עבדً אללה‬T-S Ar. 31.254 r.) vs. ‫( רפעת עבדَ אללה‬T-S 24.31 r.); ‫אללה‬ ً‫( לקית עבד‬T-S Ar. 31.254 r.) ِ vs. ‫אללה‬ ‫ד‬ ُ ‫עב‬ ‫לקיני‬ (T-S Ar. 31.254 r.); ‫אלפא‬ ‫בכסר‬ (T-S 24.31 v.). At the top of ٍ ٍ T-S Ar. 31.254 short discontinuous passages of Arabic grammar are copied in Arabic script.41 In these passages a similar confusion between nunated and non-nunated vowels can be detected: tanwīn ḍamma is not used, whereas tanwīn fatḥa and tanwīn kasra are invariably used at the end of words irrespective of their syntactic position, as well as for final non-inflectional vowels: ُ ‫عمر; بين‬ ‫سا ُم الكالم‬ ً ُ‫س وغال ُم; ضربً زيد‬ ُ ‫وفر‬ َ ‫اق‬ َ ‫ثالث اِس ُم وفع ُل وحرف; فاالس ُم قولكً رج ُل‬ ‫الفاع ٍل‬ On tanwīn in Judaeo-Arabic texts see Baneth, 1945–1946; Blau, 1980, 153–154; Blau, 1955; Wagner, 2010, pp. 175–188. 40 Vidro and Kasher, 2014, p. 206. 41 See Vidro and Kasher, 2014, pp. 176–177. 39 347 The lack of tanwīn ḍamma in these grammatical fragments resembles the vocalisation of the Qurʾān fragment Halle DMG Arab 5, in which tanwīn fatḥa and tanwīn kasra are marked as expected, whereas tanwīn ḍamma is not attested and the simple ḍamma is used instead.42 When tanwīn is marked, its graphic representation varies somewhat among different Judaeo-Arabic grammars of Classical Arabic. In the section on orthographic signs in T-S Ar. 5.45, P1 r., tanwīn is recorded as two oblique strokes, as in figure 1. Figure 1. Tanwīn as two oblique strokes (T-S Ar. 5.45, P1 r.)43 Unsurprisingly, this sign placed above or below the final consonant is used in the corpus for tanwīn fatḥa and tanwīn kasra respectively. For tanwīn ḍamma more variants are attested. The most common one is a ḍamma with an oblique stroke to the left, as in figure 2; in more cursive notation, the stroke connects to the ḍamma’s tail (see, e.g., SPB RNL Evr Arab II 185, f. 4 r.).44 Figure 2. Tanwīn ḍamma as ḍamma with oblique stroke (T-S NS 301.25)45 Tanwīn ḍamma can also be written with a double ḍamma, occasionally accompanied by the Hebrew qubbuṣ, as in figure 3. Figure 3. Tanwīn ḍamma as double ḍamma with Hebrew qubbuṣ (T-S Ar. 31.30 v.)46 42 Rödiger, 1860, p. 486. Image courtesy of the Syndics of Cambridge University Library. 44 Image available on Ktiv, the International Collection of Digitized Hebrew Manuscripts, http://web.nli.org.il/sites/NLIS/en/ManuScript/, item 159468, accessed 6 July 2017. 45 Image courtesy of the Syndics of Cambridge University Library. 46 Image courtesy of the Syndics of Cambridge University Library. 43 348 The writing of tanwīn ḍamma with two ḍammas one on top of the other, mentioned in Muslim treatises on Arabic orthography,47 has not been found in Judaeo-Arabic grammars but can be seen in a Judaeo-Arabic copy of the Qurʾān copied in Iraq or Iran in 1575–1625 (see, for example, Ox. Bodl. Hunt. 529, f. 2 v.).48 In addition to the tanwīn sign, nun or aleph in combination with simple vowels can be used to indicate tanwīn in all three cases. Examples of nun are: ‘Abū Zaydin’ (T-S Ar. 5.45, P1 v.) ‫אבו זידِן‬ ‫ אלכפץ יכון מע אל תנוין ואלכסר בלא תנוין מתל זידן בתנוין‬... ‘… the genitive case is with tanwīn and kasra is without tanwīn, for example, Zaydin (‫ )זידן‬has tanwīn.’ (T-S Ar. 5.45, P1 r.) Examples of aleph are: ‫אלרפע פי קולך זידא ואלנצב פי קולך זידא ואלכסר והו אל גר פי קולך זידِא‬ ‘An example of the nominative is Zaydun (‫)זידא‬, an example of the accusative is Zaydan (‫)זידא‬, and the example of /i/, which is the genitive, is Zaydin (‫)זידِא‬.’ (T-S Ar. 5.45, P1 v.) ‫אלחרכאת פתח והו אלנצב כקולך זידאَ או רפע והו אלצם כקולך זידُא או כפץ והו‬ ‫זידא‬ ِ ‫אלכסר כקולך‬ ‘The vowels are: /a/ which is the accusative, e.g. Zaydan (َ‫)זידא‬, or the nominative, which is /u/, e.g. Zaydun (‫)זידُא‬, or genitive, which is /i/, e.g. Zaydin (‫)זידא‬.’ (T-S Ar. 5.45, P1 r.) ِ The writing of the tanwīn with an aleph can also be found in the example ‫קאם‬ ‫( זידא‬T-S Ar. 5.45, P1 r., for the Classical Arabic qāma Zaydun), where aleph should probably be interpreted not as a hypercorrection but as a marker of the tanwīn but not of the case ending.49 3.6 Function of the text T-S NS 301.25 is unique in the corpus of Classical Arabic grammars in Judaeo-Arabic in that it is consistently vocalised with Arabic vocalisation signs. In all other grammars, Arabic vocalisation is used but is sporadic. This may hint at the fragment’s function. Al-Zajjājī’s Kitāb al-Jumal was composed in See Abū al-ʿAbbās Aḥmad Al-Qalqašandī (Egypt, 1355–1418), Kitāb Ṣubḥ al-Aʿšā (Shams al-Dīn, 1987, p. 161). 48 Image available at Digital Bodleian, https://digital.bodleian.ox.ac.uk/inquire/p/0673a6098fa3-40f2-b372-23099ab76822, accessed on 22 June 2016. 49 See also Blau, 1955 on the use of aleph to indicate nunation but not case in certain types of nominal sentences. 47 349 order to provide learners with basic knowledge of the Classical Arabic language and grammar,50 and was traditionally used in the classroom for beginning students.51 It is clearly with the same purpose – that of learning the basics of Classical Arabic and its grammar – that this fragment was transcoded into Hebrew characters. That the single currently identified part of this grammar in Hebrew characters is the chapter on inflection, and the following chapter on verbs was not copied even though enough space remained on the page to do so, may indicate that only a portion of this book was transcribed and vocalised, possibly as a vocalisation exercise. Indeed, it seems fitting to use a basic text on grammatical cases, which mainly deals with vowels and ends with a summary of all case markers, as teaching material on the topic of Arabic vocalisation and as a sample text to practice one’s vocalising skills. The imperfect vocalisation of the fragment may indicate that this is not an expert’s work to be copied by future students, but the product of a learner who has not yet attained full mastery of this subject. 4 Conclusions In this article I have edited and analysed a Judaeo-Arabic fragment of Abū alQāsim ʿAbd al-Raḥmān b. Isḥāq al-Zajjājī’s basic grammar of Classical Arabic, Kitāb al-Jumal fī al-Naḥw, preserved in T-S NS 301.25 and consistently vocalised with Arabic vowel signs. T-S NS 301.25 was undoubtedly intended to represent Classical Arabic, but nonetheless its spelling and vocalisation hint at the scribe’s substrate pronunciation and imperfect knowledge of the Arabic case system. The present analysis complements earlier studies of Judaeo-Arabic fragments vocalised with Tiberian vowel signs and describes different ways of indicating vowel length and nunation, which are not regularly marked in manuscripts with Tiberian vocalisation or in those sporadically vocalised with Arabic signs. It is suggested that the fragment is a vocalisation exercise performed by a learner of Classical Arabic and its grammar. References Al-Ḥamad, ʿA. T. (ed.), 1996, Kitāb al-Jumal fī al-Naḥw, 5th edn, Beirut: Al-Risāla. Allony, N., 1991, “Ten poems from the Cairo Genizah”, in N. Allony, Studies in Medieval Philology and Literature (Hebrew), J. Tobi, R. Attal and Sh. Morag (eds), Jerusalem: Ben Zvi Institute, vol. IV, pp. 449–477. Baneth, D. H., 1945–1946, “The tanwīn and its development into a separate word in Judaeo-Arabic texts” (Hebrew), Bulletin of the Jewish Palestine Exploration Society 12, pp. 141–153. 50 51 Binaghi, 2015, pp. 158–159. Carter, 2017; Binaghi, 2015, pp. 155–156, 158–159. 350 Basal, N., 2010, “Mediaeval Jewish and Muslim cultures: an anonymous Judaeo-Arabic adaptation of Ibn Jinnī’s al-Lumaʿ ”, Jerusalem Studies in Arabic and Islam 37, pp. 223–263. Becker, D., 1998, Arabic Sources of R. Jonah ibn Janāḥ’s Grammar, Tel Aviv: Tel Aviv University. Becker, D., 2005, Arabic Sources of Isaac ben Barūn’s Book of Comparison between the Hebrew and the Arabic Languages, Tel Aviv: Tel Aviv University. Binaghi, F., 2015, La postérité andalouse du Ğumal d’al-Zağğāğī (unpubl. diss., Université d’Aix-Marseille). Blau, J., 1955, “The accusative ʾalif as an indication of subject and predicate in JudeoArabic texts” (Hebrew), Tarbiz 25, pp. 27–35. ———, 1980, A Grammar of Medieval Judaeo-Arabic, 2nd edn, Jerusalem: Magnes Press. ——— and Hopkins, S., 1988, “A vocalized Judaeo-Arabic letter”, in J. Blau, Studies in Middle Arabic and its Judaeo-Arabic Variety, Jerusalem: Magnes Press, pp. 195–254. Carter, M. G., 2017, “Grammatical tradition: history”, in L. Edzard and R. de Jong (managing eds), Encyclopedia of Arabic Language and Linguistics, Leiden: Brill, http://dx.doi.org/10.1163/1570-6699_eall_EALL_COM_vol2_0028, first published online 2011, accessed 23 July 2017. Cheneb, M. (ed.), 1927, Az-Zağğaği, Al-Ğomal. Accompagné du Commentaire des vers-témoins, Algiers: J. Carbonel and Paris: E. Champion. Khan, G., 1992, “The function of the shewa sign in vocalized Judaeo-Arabic texts from the Genizah”, in J. Blau and S. C Reif (eds), Genizah Research after Ninety Years. The Case of Judaeo-Arabic, Cambridge: CUP, pp. 105–111. ———, 2010, “Vocalised Judaeo-Arabic manuscripts in the Cairo Genizah”, in B. Outhwaite and S. Bhayro (eds), ‘From a Sacred Source’: Genizah Studies in Honour of Professor Stefan C. Reif, Leiden: Brill, pp. 201–218. Paudice, A., 2009, “On three extant sources of the Qurʾan transcribed in Hebrew”, EJJS 2/2, pp. 213–257. Rödiger, E., 1860, “Mittheilungen zur Handschriftenkunde. Über ein Koran-Fragment in hebräischer Schrift”, ZDMG 14, pp. 485–501. Sezgin, F., 1984, Geschichte des Arabischen Schrifttums, vol. IX, Leiden: Brill. Shams al-Dīn, M. Ḥ. (ed.), 1987, Ṣubḥ al-Aʿšā fī Ṣināʿat al-Inšā, Taʾlīf Aḥmad ibn ʿĀlī Al-Qalqašandī, vol. III, Beirut: Dār al-Kutub al-ʿIlmiyya. Tirosh-Becker, O., 1998, “Linguistic study of a Rabbinic quotation embedded in a Karaite Commentary on Exodus”, Scripta Hierosolymitana 37, pp. 380–407. Vidro, N. and Kasher, A., 2014, “How medieval Jews studied Classical Arabic grammar: a Kūfan primer from the Cairo Genizah”, JSAI 41, pp. 173–244. Wagner, E.-M., 2010, Linguistic Variety of Judaeo-Arabic in Letters from the Cairo Genizah, Leiden: Brill. ———, forthcoming, “Script-switching between Hebrew and Arabic scripts in documents from the Cairo Genizah”, Intellectual History of the Islamicate World, special issue on allographic traditions, G. Kiraz & S. Schmittke (eds). Woidich, M., 1991, “Short /a/ in Cairo Arabic morphology”, in A. S. Kaye (ed.), Semitic Studies in Honor of Wolf Leslau on the Occasion of his Eighty-Fifth Birthday, Harrassowitz: Wiesbaden, vol. II, pp. 1632–1651. Wright, W., 1996, A Grammar of the Arabic Language (2 vols), 3rd edn, Beirut: Librarie du Liban. Zabara, H., 2005, Perspectives on Tenth Century Arabic Grammar Al-Zajjaji’s ‘Jumal’ (unpubl. diss., Georgetown University). 351 The Structural and Linguistic Features of Three Hebrew Begging Letters from the Cairo Genizah ESTARA J ARRANT University of Cambridge 1 Introduction1 Genizah begging letters are informal petitions in Judaeo-Arabic or Hebrew which were composed by ‘foreigners’ or by individuals who had fallen upon hardship, and which request practical assistance.2 The majority of research on Genizah petitionary correspondence focuses on formal petitions, and the begging letters remain an under-studied category of epistolary literature.3 The extant research on begging letters has focused on the Judaeo-Arabic specimens, and has not explored the form and features of their Hebrew counterparts in equal depth.4 Furthermore, the formulaic structure and linguistic features of the Hebrew begging letters are particularly under-studied.5 Hebrew begging 1 I am deeply honoured to contribute to this volume in honour of my dear supervisor and mentor Professor Geoffrey Khan. I also wish to express my gratitude to Dr Ben Outhwaite, who guided me in selecting this set of letters, patiently read multiple drafts of this article and provided many helpful comments and corrections. 2 We may indeed consider begging letters to be an informal variation on the petitionary genre, since their purpose is to make a request, and since they tend to contain the basic formulaic elements found in more complex petitions. Mark Cohen has established that many of these letters were written by foreigners in Egypt and Palestine; see Cohen, 2005b and Cohen, 2006. Cohen, 2005b, pp. 174–188 also discusses in more depth why begging letters should be considered part of the petitionary genre. 3 Two relevant articles about begging letters are Cohen, 2000 and Cohen, 2005a. Concerning the broader category of petitionary documents in the Genizah, relevant articles include Goitein, 1954; Stern, 1962; Richards, 1992; Rustow, 2010. Geoffrey Khan has also published on the epistolary form of Arabic petitionary texts; see Khan, 1990a and Khan, 1990b. 4 Cohen’s two volumes contribute a detailed discussion of the role of begging letters in shaping our understanding of the history of the poor and foreigners in the early medieval Middle East; see Cohen, 2005b and Cohen, 2006. Some of the documents he mentions in those volumes are in Hebrew. 5 The main articles which examine Hebrew begging letters and certain aspects of their language are Scheiber, 1981 and Bareket, 2002. Outhwaite, 2009 researches in depth the linguistic features of Hebrew Genizah correspondence by Byzantine authors, but does not focus on begging letters. 352 letters form an intrinsic part of Genizah correspondence, and thus they should be more thoroughly researched in order to clarify how they are situated amongst their Judaeo-Arabic counterparts. Even more importantly, these Hebrew documents provide a window into the average layperson’s use of Hebrew as a language for correspondence, and are thus a valuable source of linguistic data. Here I seek to make a contribution towards this end. I wish to complement previous scholarship by focusing on the linguistic features and formulaic structure of some additional Hebrew begging letters, contextualising them with those which have already been published by Bareket and Scheiber.6 I first present an edition of three unedited Hebrew begging letters, and then I explore their internal structure, their linguistic register and their authors’ skill with Hebrew. I note which specific features are found in wider, more formal Genizah correspondence, as well as which features appear to be unique to these particular three letters, especially with regard to formulaic structure. The reader should be aware that while my analysis and definitive comments concern only the specific documents presented in this study, I would argue that it is reasonable to extrapolate the general trends of the features of these letters onto the genre of Hebrew begging letters as a whole. My analysis found that for these three documents (and, by tentative extension, most Hebrew begging letters), the language is in general the same as in the more formal medieval Hebrew correspondence of the Geonim, only simpler and less skilful. Likewise, while their formulaic structure is similar to that of more formal petitions, it is simpler in nature, and adheres to a specific order of elements which are specially tailored for begging.7 2 2.1 Text editions Mosseri II.98.1 Material: paper (folded many times). Recto: 19 lines. Verso: 4 lines. 2.1.1 Transcription8 Recto …[‫ ש שלום רב לאוהבי תורתיך ו]אק‬.1 ‫ שלומות רבות וישועות קרובות‬.2 6 Bareket, 2001; Scheiber, 1981. Such assertions will be tentative, however, as a full study of the Hebrew begging letter corpus has not been conducted. 8 Text in brackets has been reconstructed; I only reconstruct the text where there is sufficient contextual and/or physical evidence for a reconstruction. Ellipses indicate a lacuna or a missing section of text. It is difficult to tell where each line truly ends, and whether it has been cut off prematurely. The reader should note that lines without ellipses still may be incomplete. 7 353 …‫ וחנינה ויד ושם ומזל גבוה וקי‬.3 ]‫ כבוד גדו קדו מר ורבי אור עינינ[ו‬.4 …‫ ר ישועה החכם והנבון הירא א‬.5 …]‫ אלהינו יברכהו וישמרהו ויעודדהו [ו‬.6 … ‫ לשם ותהילה ויזכהו לראות שמחת‬.7 ‫ אמן ויזכהו לראות ביאת גואל ידוע‬.8 ‫ לך מרי ר ישועה לא כתבתי זו הכתב‬.9 ]‫ ב בושת פנים כי האל יודע אלו היה לי [כסף‬.10 ]‫ לא כתבתי אצלך ואני לא אחפוץ מ[כבודך אלא‬.11 …]‫ ב ריטלין לחם כי יש לי מן השבת [ה‬.12 [‫ בלא לחם ואם אשב יום אח[ר] ולא או]כל‬.13 … ‫ אפחד שלא אפול בחולי ובאמו]נ[ה על‬.14 …[‫ של[א] בטובתי כתבתי אצל ]כ[בו]ד[ך ו]א‬.15 ‫א]מ[… מן הרעב לא אכתו[ב] על אדם‬.16 … ‫ ובאמונה אל תאשימני כ[י] האל יודע‬.17 … ‫ כתבתי אצל כבודך זו ה[אג[רת אלא‬.18 …‫ איתי עצמי ב[ר]עב כ‬.19 Verso ‫ … אפילו פרוטה ואני אשב ב ימים‬.1 ‫ … ולא אכלתי לחם ולא אבקש מכבודך‬.2 ‫ [אלא ב רי]טלין וֿהקבֿה ישלם שכרך בעולם‬.3 ‫ [הזה וב]עולם הבא ושלומך ישגא ויפרה עד‬.4 Translation9 2.1.2 Recto Great peace to the lovers of your Torah, and … Much peace and imminent salvation, and mercy and remembrance and great fortune … Honourable, great and revered teacher and my master, light of [our] eyes 5. Rabbi Yeshuaʿ, the wise and the clever, fearer of … 6. our God. May he be blessed, guarded, encouraged [and] … 7. for remembrance and praise. And may he be granted the joy of … 8. Amen. And may he be granted to see the coming of the redeemer. Let it be known 9. to you, my teacher Rabbi Yeshuaʿ, I did not write this letter 10. out of shame, because God knows if I had [money] 11. I would not write to you and I seek nothing from [your honour except] 1. 2. 3. 4. 9 Text which has been reconstructed in the transcription appears in the translation inside square brackets. Text in parentheses in the translation indicates my clarification. 354 12. two raṭls of bread, which I have [not had] since Shabbat … 13. without bread, and if I go for [another] day and do not [eat] 14. I fear that I may fall ill. And truly … 15. That it is [not] of my goodness that I wrote to your honour … 16. … from hunger. And I will not [write] to man10 17. and truly, do not blame me, [because] God knows … 18. I did not write this [letter] to your honour but … 19. to myself in hunger …11 Verso 2.1.3 1. … even a cent and I will dwell for two days 2. … and I did not eat bread and I will not ask from your honour 3. [(anything) except two raṭls]. May the Holy One, blessed be He, complete your gain in [this] world 4. [and in] the world to come. And may your peace grow and prosper more. Textual notes Recto 1. The first shin in the line is probably a mistake; the author apparently wished to indent the first line of the letter. The line itself quotes Psalms 119:165, and the rafes over the bet and the first tav of ‫ לאוהבי תורתיך‬mark it as a quotation. Such marks are generally ornamental,12 and perhaps they also function to indicate a biblical quote without using an actual introduction beforehand (in contrast, the last letter of this study, T-S Misc. 28.18, uses the Judaeo-Arabic ‫ כמא קאל‬to indicate that a biblical quotation is going to follow). These rafes may be a secondary addition, as the ink is lighter than the surrounding text; however, close inspection reveals that it is still the same colour as faded portions of letters throughout the document. The word ‫ אוהבי‬is spelled plene, contrary to its defective spelling in Codex Leningradensis; this is not unexpected. 2.–3. The irregular use of ‫ות‬- in ‫ שלומות‬is the preferred spelling in Genizah letters, as opposed to the more biblical ‫שלומים‬.13 4. The honorifics appearing on this line are nearly universally used in Genizah letters from this period. As such, they appear here in abbreviated form, with dashes that appear similar to the rafe sign. The first set of 10 Lines 16–19 are damaged, so only a tentative translation can be made here. There are more letters beyond the end of what appears in the transcription, but they are for all purposes indecipherable. 12 Outhwaite, 2000, p. 15. 13 Outhwaite, 2001, p. 216. 11 355 diacritics indicates the full phrase ‫כבוד גדולת וקדושת‬. Such honorifics and their abbreviations are usual features of Genizah correspondence,14 although the forms of this abbreviation can vary.15 5. The line probably ends in some form of ‫אדון‬, perhaps ‫אדונינו‬, but it was not reconstructed in order to avoid prescriptiveness. 6. The aleph and the lamed in ‫ אלהינו‬have been joined together in a ligature – a common orthographic feature in both Hebrew and Judaeo-Arabic Genizah correspondence. This ligature appears in all three of our letters here. ‫הו‬- is placed at the end of each of the verbs to indicate that Rabbi Yeshuaʿ is their direct object. Its appearance here in the florid opening of the letter is appropriate.16 7. ‫ לשם ותהילה‬is difficult to translate because the previous line is damaged. I have chosen to respect the previous line’s context as much as possible, and to translate this phrase as the end of the sentence, with the lamed appearing as a preposition to the noun ‫שם‬. This noun also appears in line 3, where I translate it as ‘remembrance’, and it is likely that it has the same meaning here. 8. The writer’s use of ‫ לראות ביאת גואל‬is noteworthy because he uses the gerund form of ‫ בוא‬but drops the definite article for the rest of the construct phrase. In comparison, T-S 12.258 (line 5 of the marginal sidewriting) has the phrase ‫וישום ביאת הגואל בימיו‬, which contains the definite article.17 9. The author’s ordering of the demonstrative pronoun ‫ זו‬here – ‫– זו הכתב‬ is noteworthy.18 In this instance, the gender of the demonstrative pronoun does not match the noun which it governs. This construction is not anomalous in Genizah letters. Such gender confusion is a common feature in medieval Hebrew, and in some instances this construction can occur without the definite article.19 Importantly, scholarship has accepted that this particular construction is an Arabicism, corresponding to the construction -‫ ;هذا ال‬and as Outhwaite says, “it is certainly very pervasive, since in Genizah correspondence it can be found in letters by those who show no other Arabic influence in their language”.20 Finally, 14 Outhwaite, 2009, p. 186. Outhwaite, 2013, p. 3. 16 Outhwaite, 2013, p. 4. 17 This manuscript was published by Bareket, 2001, pp. 381–383. 18 We see the same construction again on line 18 (recto) – -‫ – זו ה‬but in that case the attached noun has been damaged; however the gender matches, as ‫( אגרת‬reconstructed) is feminine. 19 Outhwaite, 2001, p. 201. 20 Outhwaite, 2001, p. 201. 15 356 note the use in this line of ‫ כתב‬as opposed to ‫מכתב‬. Both words are common in post-biblical Hebrew, but ‫ מכתב‬appears more often than ‫כתב‬.21 This letter tends to utilise predominately post-biblical nouns, so the use of ‫ כתב‬here is a slight deviation from this trend. In letters by Jews whose vernacular is Arabic, this word, which is closer to ‫كتاب‬, is preferred over the Hebrew ‫מכתב‬.22 10. The preposition ‫ ב‬of ‫ בבושת פנים‬is separated from ‫ בושת‬by a visible space on the manuscript. Note that for ‫ אלו‬the writer gives the only vowel sign in the entire manuscript. Outhwaite notes that ‫ אלו‬is less common in Genizah epistolary Hebrew than ‫אלה‬, and that a marked minority of writers prefer ‫אלו‬.23 11. The use of the prefix conjugation for ‫ חפץ‬does not necessarily indicate the future tense; it is common for this conjugation to “express a wide range of modal nuances”.24 I attempt to show such nuance in my translation above. Also note my reconstruction at the end of the line: ‫מכבודך‬ ‫אלא‬, which is the likely ending to the line had the document not been torn, although ‫ מכבודך כי אם‬is also a possibility.25 The preposition ‫אצלך‬ is rather awkward: one would expect the writer to use ‫ אליך‬or ‫ לך‬instead. This may be a way of using distancing language (for more detail on ‘distancing language’, see my comment on line 15 below). Note that on line 18 the author uses a similar construction (‫)כתבתי אצל כבודך‬. Perhaps ‫( אצל‬either with the suffix ‫ך‬- or else preceding ‫ )כבודך‬serves as a means of avoiding direct address26 and of formalising the tone of the letter. Regardless, it is a testament to the author’s awkward grasp of Hebrew. 12. ‫ ריטלין‬lacks the dual form even though it is paired with the numeral ‫ב‬. The use of the dual, however, is rare in this period, and “we often find that the dual is not used where it could be expected in BH or RH, being replaced by the number ‘two’ and the plural noun”.27 The ‫ין‬- suffix for the plural is typical for words which do not have a Hebrew origin or which were introduced to Hebrew after the biblical period (this particular root comes from the Arabic ‫)رطل‬.28 An important point which clarifies the meaning of the entire document is that the phrase ‫ יש לי‬here 21 Outhwaite, 2009, p. 195. Outhwaite, 2009, p. 195. 23 Outhwaite, 2001, p. 199–200. 24 Outhwaite, 2001, p. 225. 25 Thanks to Dr Ben Outhwaite for suggesting this possibility. 26 Although on line 9 he does write … ‫לך … לא כתבתי‬. 27 Outhwaite, 2000, p. 73. 28 Outhwaite, 2013, p. 4 notes that the ‫ין‬- suffix is a common feature in Gaonic Hebrew. The historical context behind mentions of bread in many begging letters, and this specific measurement, is to be found in detail in Cohen, 2005a, pp. 408–414. See Lane, 1984, vol. I, pp. 1101– 1102 for the usage of the Arabic root. 22 357 refers not to bread, but to the amount of time the beggar has been without bread (i.e. ‘I have gone [x] number of days since Shabbat without bread’). We can assume he has gone two days without the required food based on line 1 of the verso which could be understood to mean ‘I will go two days without even a cent’s worth of bread’. 14. With regard to ‫‘ אפחד שלא אפול בחולי‬I fear that I may fall ill’, in Arabic it is possible to express in a subordinate clause a desire that something not happen by inserting ‫ ال‬between the two verbs, and it is to be noted that this ‫ ال‬does not negate the verb but instead serves to express the negative desire of the sentence.29 Here we have this construction calqued in Hebrew, thus the translation ‘I fear that I may (i.e. lest) I fall ill.’ 15. A comment on ‫‘ כתבתי אצל כבודך‬I wrote to your honour’. By distancing himself from his addressee through using an honorific such as ‘your honour’, the author is writing in a higher and more formal register.30 The phrase ‫ שלא בטובתי‬is reminiscent of ‫ בעל כרח‬/ ‫על כרח‬, a phrase found in Rabbinic Hebrew: ‘it is against my will’.31 The phrase here likely conveys a similar meaning: the author feels forced to write this begging letter. This seems to connect, thematically, to the idea that the author is not writing ‘out of shame’. The theme of ‘shame’ is pervasive in Genizah begging letters, and by emphasising that he would rather not have written, the author in effect reduces the ‘shame’ involved in writing a begging letter.32 Verso 4. 29 I have reconstructed the phrase ‘in this world and’ because it typically precedes ‘in the world to come’ in constructions such as this. Note that the letter does not contain a signature at the end. Normally one would conclude that a letter without a signature is a draft, but in the case of begging letters this conclusion is questionable. Being informal correspondence, it likely needed no real signature; signatures with names are infrequent or rare for our published corpus of Hebrew begging letters. Harrell and Brunot, 2004, p. 155 describe this construction fully, though one should note that this construction is not restricted to Moroccan Arabic. Thanks to Dr Ben Outhwaite for this suggestion. 30 For an additional example, see T-S 8J13.5, line 3, in which distancing language is also employed in the blessing section of the letter. 31 Cf. Jastrow, 1903, p. 666. Thanks to Dr Ben Outhwaite for this suggestion. 32 Bareket, 2001, p. 363 mentions the pervasive desire to avoid shame in the begging letters, and there is an extensive discussion of this matter in Cohen, 2005a, especially p. 185. 358 2.2 T-S 8J13.533 Material: paper. Recto: 18 lines. Verso: blank. Transcription34 2.2.1 … ‫בשם אל רחום וחנון‬ .1 …‫ [אל] אדוני הזקן המכובד והמ‬.2 … ]‫ ויאריך ימיו בטוב ושנותי[ו‬.3 … ]‫ לטובה ויברך את כל מעש[יו‬.4 35 … ]‫ מפעליו ו[יז]כה ויראה בנ[ים‬.5 … ‫ וגם יחיה וישמר וינצור ויפרה‬.6 … ‫ יאריך ימיו ירבה שנותיו יום‬.7 …‫אודיע לאדוני הזקן ה‬ .8 36 … ]‫ כי באתי ממקום רחוק עד ה[נה‬.9 37 … ]‫ והייתי מן הנותנים ועושי ח[סד‬.10 … ]‫ כל עשרי וכל ממוני ונשאר]תי‬.11 …‫ אדוני הזקן וטוב כי ש‬.12 … ]‫ עתה רחם עלי כמנהגך ה[טוב‬.13 … [‫ ייי ועליך שמתי בטחו]ני‬.14 …‫ ואני [נ]כרי ולכן שא‬.15 … [‫ מאשר נתן לך ייי ]אלהי‬.16 … ]‫ ויכפיל ממ[ו]נך וית[ן‬.17 … ]‫ייי אלהי אבו[תינו‬ .18 2.1.2 Translation In the name of God, the Merciful and Compassionate … [To] my lord the honourable elder and … May his days be lengthened with goodness and his years … with pleasantness. And may he (God) bless all his (the addressee’s) deeds … 5. and his actions, and may he merit to see (have) posterity … 6. And may he be guarded and kept and multiplied … 7. and may his days be long, also may he be given (long)38 life … 1. 2. 3. 4. 33 Cohen, 2006, p. 51 has published a translation of this letter. This article presents a closer, line-by-line edition of the text. 34 This document has spacing, and that spacing has been retained in the transcription. Ellipses indicate a lacuna and letters in square brackets are reconstructed. In the translation, curved brackets indicate my clarification. Note that in the manuscript, the Tetragrammaton is abbreviated with three yods in a triangle. I have represented this here as ‫ייי‬. 35 I agree with Cohen’s reconstruction of the last word of this line as ‫בנים‬. 36 Cohen has provided plausible reconstructions of the end of this line as either ‫ הנה‬or ‫עד המקום‬ ‫הזה‬. While I have been more reserved in reconstructing the end of the line, Cohen’s second suggestion is most likely. 37 The end of the line is probably ‫ חסד בממונם ואבד‬but in order to avoid prescriptiveness I have not reconstructed it. 38 ‘Long’ is not in the letter, but it is implied. 359 8. To inform my lord the elder … 9. Because I came from a faraway place to [here] … 10. and I was one of the givers, and of those who practise [compassion] … 11. all of my wealth and all of my money and I was left … 12. My lord the wise and good, for … 13. now have compassion upon me according to your custom … 14. in God and you I put [my] trust … 15. and I am a foreigner, and therefore … 16. From what God has given you … 17. And may He increase your wealth and may He give … 18. Lord, God of [our fathers] … 2.1.3 Textual notes 2. There are at least two adjectives written here to describe the elder, but only one, ‫מכובד‬, is fully preserved enough to include in the translation. Later, on line 12, ‫ הזקן‬functions as an adjective for ‘wise’.39 Note that ‫ זקן‬in this letter is a title, designating an ‘elder’; it does not function as an adjective describing the addressee as ‘old’.40 3. Cohen translates this line as ‘may He lengthen your days with goodness and your years with pleasantness’41 but I wish to draw attention to the fact that the blessing, while still directed towards the addressee, is written in third person, that is, ‘may He lengthen his (the addressee’s) days …’. Distancing language such as the use of the third person in the opening blessings of letters, alongside the presence of the imperfect with jussive force, adds to the formality of the text. This is also seen in the letter discussed in section 2.1 above, where the beggar – speaking of the addressee – writes ‘May he be blessed, guarded, encouraged’ (line 6). Writing in the third person in this manner is a common feature in epistolary Hebrew.42 I transcribe the last letter of the line as a waw in order to adhere to the third person pattern which continues to this point. 4. I have reconstructed the last word of this line as ‫ מעשיו‬to remain consistent with the rest of the text (it is symmetrical with ‫ מפעליו‬in line 5). However, an alternative would be to reconstruct the end of the line as ‫מעשה יבין‬.43 The use of the direct object marker ‫ את‬is not an unusual 39 Cohen translates ‫ הזקן‬as a noun, thus ‘the good elder’, but it is possible to read ‫ הזקן‬as an adjective, as I have done in my translation. 40 Bareket, 1999, pp. 41–43 describes the use of this title in the Genizah. 41 Cohen, 2006, p. 51. 42 Asher and Outhwaite, 2014, p. 208. 43 Thanks to Dr Ben Outhwaite for this particular suggestion. 360 feature in Genizah documents,44 but it appears to be an unusual feature in Hebrew begging letters. This line, ]‫לטובה ויברך את כל מעש[יו‬, contains the only occurrence of ‫ את‬in our three letters, and it occurs very infrequently in those begging letters which have already been published, specifically in T-S 12.354 and T-S 8.24. In most instances it is included when it occurs within the context of a biblical quotation or allusion. That applies here in this letter as well, because even though the blessing itself is not a biblical quote, it does allude to the phrase ‫יברך את כל מעשי‬ ‫ ידיך‬in Deuteronomy 15:6, which we will see as an actual quote in line 1 of the final letter of this study, T-S Misc. 28.18, transcribed in section 2.3.1. 8. The spacing here very clearly indicates the transition from the opening blessing section of the letter to the addressing and petitioning sections. This may be considered evidence that the authors adhered to an order of specific sections, especially as the letter discussed in section 2.3 below also contains similar formatting. 13. Note that the author avoids using the possessive ‫ של‬here (as in ‫)שלך‬, instead opting for an attached pronoun without ‫ של‬to indicate possession: ‫כמנהגך‬, instead of ‫כמנהג שלך‬. 14. Due to the lacuna we cannot know for certain whether the Tetragrammaton here belongs to a previous clause or sentence, but I have translated it as belonging to the clause of this line. Note the fronted focus of ‫ עליך‬in the clause (and ‫ ייי‬if it belongs with this clause). This places the emphasis on the addressee himself. It could have been written as ‫שמתי‬ ‫ בטחוני בייי ועליך‬and although the meaning would be the same, the discourse structure would have altered the emphasis and placed it upon the author – in other words, it would read as ‘I place my trust in you and God’, as opposed to ‘in God and you I place my trust’. The use of this construction in this letter serves the function of placing pressure on the addressee to acquiesce to the request. 15. The word ‫ נכרי‬is what is visible on the manuscript, but the first letter is slightly damaged. Within the context, the translation ‘I am a foreigner’ is sensible. Cohen notes in his edition that the author of this letter is a foreigner, possibly European, basing this on the phrase ‫כי באתי ממקום‬ ‫‘ רחוק‬I came from a faraway place’ in line 9. Cohen leaves the phrase ‫‘ אני נכרי‬I am a foreigner’ out of his translation, even though it is apparent in the text of the letter. Thus the letter itself, in two places, explicitly confirms Cohen’s argument that begging letters can be from literal ‘foreigners’. 44 For an overview of the use of ‫ את‬in Genizah letters, see Outhwaite, 2001, pp. 213–214. 361 2.3 T-S Misc. 28.1845 Material: paper. Recto: 12 lines. Verso: blank. 2.3.1 Transcription ‫ לתת מט[ר אר[צך [בעתו ולברך] את כל מעש]ה[ ידיך [ו]הלוית ג[וים ר[בים ואתה לא‬.1 ‫ויתקיים עליכם כל הברכות ואימ[ו]רות יש בארבע‬ ‫ תלוה לחיים ואמרו אמן‬.2 ‫דעו אדונינו‬ ‫ פינות העולם כולם יבואו ויוחלו עליכם ועל בניכם ועל בתיכם‬.3 ]‫ ואחינו ישראל כי אני איש מן ארץ רחוקה עני ודל ואחפוץ אליך לא יש לי ש[כר‬.4 … ‫ הצפינה ולא יש לי צידה לדרך ואני בטח בייי ובישראל אם ייי יתין בלובכים‬.5 ]‫ ורחמים בעיניכם ויעזריני בזה דרך רחוקה כי אני אליך בארצ]י[ ובמולדתי ודעו א[חינו‬.6 ‫ ישראל כי טוב שם משמן טוב וזה לא יכשל בישראל כי היום א[נ]י הנה שבעים יום לא‬.7 ‫ פרוטה אחד ולא פת לחם ב… ישראל והינה כניסה וקהל וחזן ופרנס לא יכשל א[ל] ישראל‬.8 ‫ כי קאל … ב ה… וג׳ וכ]מ[א קאל ]ו[אהבתה לרע]ך[ וג׳ וכמא קאל‬.9 ‫ וצד]קה[ תצ]יל[ ממ]ות[ וכמ[א ק]אל וחי אחיך עמ]ך[ וכמ]א[ ]ק[אל פ]ת[ח ]תפת[ח את ידך‬.10 ‫ ל]א[חיך וכמה זה אתם בששון ושמ]חה[ מייי והוא שובע ולא יש עליכם גלות‬.11 ‫ כמו פי ארץ פרץ לא יש זה טוב ולא יכשל ושלום על כל יש[ראל] ששששששש‬.12 2.3.2 Translation 1. … give [rain to your land in its season, and bless] all the deeds of your hands, [and] you will lend to [many nations] and you will not 2. borrow forever, amen. And may all the blessings come upon you. And may, from the four 3. corners of the earth, all of them, (the blessings) come and be applied to you and to your sons and daughters. Know, our masters 4. And our brothers Israel, that I am a man from a faraway place, poor and destitute, and I want to go, (but) I do not have the [fare] 5. (for the) boat. And I do not have provisions for the road, and I trust in God and in Israel. If God will put in your hearts … 6. and mercy in your eyes and help me on this long journey. Because I want to go to [my] land, and the land of my birth. Know [our brothers] 7. Israel, that a (good) name is better than good oil. And this will not fail in Israel, for today [I] am here seventy days. I do not have 8. a cent and not a piece of bread … Israel. And here is a synagogue and a community and a ḥazzan and a parnas. And the God of Israel will not falter 9. As it is quoted … as it is said, ‘Love your neighbour’, and as it is quoted, 45 This document makes use of spacing to delineate formulaic sections, and that spacing has been retained in the transcription. Parentheses are used for clarification in the translation. 362 10. ‘and righteousness delivers from death’, and as it is quoted, ‘and your brother may live with you’, and as it is quoted, ‘open your hand46 11. to your brother’. And what joy and happiness may you have from God, and He satisfies. And may none of you be exiled 12. as in the land of disaster (where) there is no goodness. And Israel will not falter, and peace (be) upon [Israel]. 2.3.3 Textual notes 1. The phrase ‘you will lend to many nations and not borrow’ is quoting Deuteronomy 15:6, but it is especially noteworthy that the author uses the verb from the root ‫ לוה‬for ‘lend’, rather than ‫עבט‬, which is what is used in the actual verse. The root used in the letter is common in Rabbinic Hebrew, and is synonymous with ‫עבט‬, so its use does not appear to be controversial or interpretive.47 It is likely that this author was much more familiar with that root and so used it here, perhaps indicating that he was more concerned with the meaning that the verse conveyed than with the correct writing of the biblical text. 2. The use of ‫ יש‬as opposed to ‫ אשר‬or -‫ ש‬here is awkward, and may indicate an unfamiliarity with the more usual syntax of epistolary Hebrew. ‫ש‬/‫ אשר‬is not an uncommon relativiser in Genizah epistolary Hebrew, and in this letter it appears that constructions using ‫ יש‬are influenced by a foreign idiom.48 This is an indicator that this writer was perhaps not accustomed to writing in Hebrew. 3. Note the plural for ‘your daughters’, which instead of the expected ‫ בנותיכם‬is ‫בתיכם‬. I would not translate it as ‘houses’; it seems to be parallel with ‘sons’, and therefore ‘daughters’ is a more sensible translation. 3.–4. This contains the only example in our three letters of ‫ כי‬as a complementiser (‫‘ דעו אדונינו ואחינו ישראל כי אני איש מן ארץ רחוקה‬know our lords and our brothers Israel that I am a man from a faraway land’). Of the complementisers, both ‫ כי‬and ‫ ש‬are used in broad Genizah correspondence as a general characteristic, so it is noteworthy that this instance of ‫ כי‬is the only occurrence of a complementiser in the three letters which 46 Much of this line has been reconstructed, but to ease reading I have not put reconstruction brackets in the English translation. 47 See Jastrow, 1903, p. 697 (entry under ‫לוי‬, ‫ )לוה‬for the extensive rabbinic usage of this root. 48 See the note for line 4. Also, see Outhwaite, 2000, p. 42 for a note on the usage of ‫ אשר‬in the Genizah epistolary corpus. He describes the use of this particle as ‘interchangeable’ with its shortened form. 363 are discussed here (it is repeated on lines 6–7).49 The phrase used here, ‫ כי‬+ ‫‘ ידע‬know that’, is a preferred construction in Genizah correspondence.50 4. The phrase ‫ לא יש לי‬is awkward: we would expect ‫אין לי‬, yet ‫ לא יש לי‬is a favoured phrase of this author. This construction may be due to influence from Spanish: the same phrase occurs in T-S 16.100 (line 20), a letter which originated in Spain. Such a construction could be considered as similar to the Spanish no tengo + object (‘I do not have’ + object).51 This makes sense of the pattern in general terms: the negation comes first, and it precedes the particle ‫יש‬, here a replacement for what would be the first person present tense verb in Spanish.52 While it is entirely possible that this is the case, it also appears that the author is not entirely certain how to use the particle ‫יש‬, as he uses the construction -‫ יש ב‬on line 2 and ‫ יש זה‬on line 12; these are patterns for which I do not have a plausible explanation. Another possibility is that the phrase ‫ לא יש‬comes from the Arabic ‫ليس‬, which is semantically identical. Given the evidence in other letters for this phrase resulting from Spanish influence, as well as the amount of Arabic influence shown in this particular document, both possibilities seem equally plausible.53 5. In this line, we find ‫ הצפינה‬for an expected ‫ – הספינה‬that is, with ‫ צ‬for an expected ‫ – ס‬suggesting the influence of the author’s vernacular on his pronunciation of the Hebrew sibilant. There are also two issues with regard to the verbs on this line. First is ‫אני בטח‬, which is noteworthy because ‫ בטח‬is not conjugated (or if conjugated, is written defectively, which in this instance is unusual). Next, he spells the verb ‫ נתן‬in plene form (‫)יתין‬, and this plene spelling of the ṣere is a “pronounced characteristic” of Genizah letters.54 When it comes to the word ‘your hearts’, what is written in the document appears to involve the exchange of waw for bet (i.e. ‫ לובכים‬instead of ‫ ;)לבבכם‬the two would have a similar sound in this instance, where the author is using a ‫ ו‬in place of the fricative ‫ב‬. 49 Outhwaite, 2001, pp. 205–206. Outhwaite, 2001, p. 206, referring to Outhwaite, 2000, p. 45, notes: “In Genizah letters generally, it can be seen that verbs denoting speaking, knowing and understanding … prefer the complementizer ‫ כי‬whereas other verbs, those in particular with a volitive or directive aspect … tend to take the complementizer ‫ש‬.” 51 Yahalom, 1999. Thanks to Dr Ben Outhwaite for bringing this to my attention. 52 One does not need to use an explicit pronoun yo ‘I’ here; the verb is conjugated as first person singular. 53 Outhwaite explores the possibility of this being a Greek construction, as the same construction also occurs in a Byzantine scribe’s documents; see Outhwaite, 2009, p. 198, especially n. 55. I would argue that, as it shows up in the writing of speakers of both Spanish and Greek backgrounds, it is most likely an indicator that the author’s native tongue is, or is influenced by, a non-Semitic language. 54 Outhwaite, 2001, p. 217. 50 364 Alternatively, the ‫ ו‬may reflect a vowel which has been partially assimilated to the following consonant ‫ב‬. More specifically, an i vowel here preceding the labial ‫ ב‬would be labialised under assimilation to the consonant, thereby shifting to u.55 7. The author begins this line with a quote from Ecclesiastes 7:1, noting that ‘a good name is better than fine oil’. I interpret the next clause to mean that Israel’s greatness will not fail the writer’s expectations with regard to his request. His use of the word ‫ הנה‬here as opposed to ‫הינה‬ (which occurs on the next line) is probably unremarkable, but this particular spelling has been linked to the Arabic word ‫هنا‬.56 8. Juxtaposing his request for assistance with his estimation of the Jewish community makes his request seem even more reasonable; that is, ‘surely Israel will not falter in supporting me’. Note the author’s use of the Arabicism ‫כניסה‬, which is a Hebraicised derivation from ‫ كنيسة‬and is frequent in Genizah correspondence.57 9. This line is a quotation, the only legible part of which is a quote from Leviticus 19:18. 10. This line is also a mixture of quotations from the Tanakh, separated by the quotation indicator. The first quote comes from Proverbs 10:2 or 11:4 (the phrase occurs in both places), the second from Leviticus 25:36 and the third (which spills over onto line 11) from Deuteronomy 15:11. It does not appear that these quotes are marked by any diacritics, but they would not need to be so marked since they are introduced in the text as quotations.58 12. The use of ‫ כמו‬here is noteworthy. It is not a common conjunction, especially without the prefix -‫ש‬, and is rare in other forms of Hebrew.59 Equally, its meaning here in the sentence is unclear. The phrase ‫ארץ פרץ‬ is rhymed prose, with ‫ פרץ‬denoting disaster or trouble.60 The translation is difficult, but the general meaning is a blessing: a hope that the addressee(s) will not suffer exile in a place where tribulation occurs. 55 Waw/bet interchange is also attested in the Dead Sea Scrolls, for the same reasons we see it here. Reymond, 2014 notes this phenomenon (see p. 70 for a specific example). He also points to instances where “a following bilabial (/b/, /m/, /p/) or resh causes the shift from an /i/, /e/, or /a/ vowel to an /o/ or /u/ vowel” (p. 174), which may be what is occurring here. Qimron, 1986 also describes the assimilation of bilabials in the Dead Sea Scrolls (pp. 39–40). 56 Outhwaite, 2001, p. 213. 57 Outhwaite, 2009, p. 188. 58 Thanks to David Sklare for help in identifying the biblical passages on this very damaged line. 59 Outhwaite, 2000, p. 48. 60 For a discussion of the root ‫ פרץ‬and its connotations, see Jastrow, 1903, pp. 1237–1238. 365 3 Analysis and further commentary I am primarily concerned with three aspects of these documents’ internal features: a. their macro-structure: their formulaic features and the order of their internal elements; b. their linguistic register: the location and extent of ‘biblicising’ or ‘rabbinicising’ features in the language, whether this appears in the form of quotations or in a mixture of syntactic and lexical elements; c. their Hebrew: the skill demonstrated in the use of the language and the presence of external linguistic influences upon it. In this section I show that the letters are comprised of a definite formulaic structure which is simple but consistent and well-suited to the purposes of begging. They also display a register of medieval Hebrew which contains similar features found in Gaonic correspondence, yet is less sophisticated and less complex. In particular, my linguistic findings complement Bareket’s conceptual understanding of the formulaic features in the begging letters.61 With regard to structure, the documents themselves can be split into a sixpart formulaic structure: 1. 2. 3. 4. 5. 6. opening biblical quotation or bismillah blessings and honorifics direct addressing of the recipient justification for writing and defence of motives request closing blessings and signature This six-part structure can be grouped into three obvious subsections: the opening, which consists of parts 1–3 (though parts 1–3 can vary in their order somewhat); the body, which consists of parts 4 and 5; and the closing, part 6. All three of the letters discussed here follow this six-part structure, and it tentatively appears that the documents published by Bareket and Scheiber also adhere to this ‘template’ (if one dares use so strong a word to describe informal letters such as these). In our letters, the six-part structure appears as follows: Opening: • Mosseri II.98.1: recto lines 1–8 • T-S 8J13.5: lines 1–7 • T-S Misc. 28.18: lines 1–3 61 Bareket, 2001, especially pp. 362–363. 366 Body: • Mosseri II.98.1: recto line 9–beginning of verso line 3 • T-S 8J13.5: lines 8–15 • T-S Misc. 28.18: end of line 3–line 10 Closing: • Mosseri II.98.1: verso lines 3–4 • T-S 8J13.5: 16–18 • T-S Misc. 28.18: 11–12 The discussion in sections 3.1 and 3.2 focuses on the formulaic structure and linguistic register of the opening and the body of the letters. The closings of these texts are very simple in their structure and warrant a separate discussion in section 3.3. 3.1 Formulaic structure These medieval Hebrew begging letters belong to the array of petitionary literature of the 11th century Middle East, which consists mainly of petitions in Arabic and Judaeo-Arabic. Thus we can understand their features more clearly by contextualising them within this petitioning culture. Arabic petitions have a definite formulaic structure, the specific elements of which shifted and developed over the centuries and, especially in the Fatimid period, this structure consisted in general of specific formulaic blessings surrounding an inner core of “exposition, request, and motivation”.62 Our documents, too, follow a specific formulaic order of elements. However, it would be a mistake to assume that these documents were directly inspired by, or in direct imitation of, the structure of Islamic/Judaeo-Arabic petitions.63 It seems much more likely that the structure of Hebrew begging letters was loosely influenced by their more formal Judaeo-Arabic counterparts, and more strongly determined by individual need. Furthermore, a poor petitioner would not necessarily have had easy access to the rich petitionary models exhibited by the Islamic/Judaeo-Arabic texts.64 One may argue that some (or indeed many) begging letters were written by scribes on behalf of the beggar, and thus would allow the beggar to 62 Khan, 1990b, p. 8. Cohen himself notes that “the Geniza [begging] letters were supplications, in the style of Muslim-Arabic petitions to rulers or other dignitaries” (Cohen, 2005a, p. 418). But he also asserts that they “conform even less consistently to the eight-part structure of the Islamic petition” which was elucidated by Khan (Cohen, 2000, p. 448). While they belong to the same genre as Islamic petitions, Jewish, and especially Hebrew, begging letters must be considered on their own merits. 64 T-S 8J13.5 is unusual: the author must have been an educated beggar. It was written in a practised hand and contains elegant vocabulary and syntax, both of which indicate this man was a scribe and/or educated. Thus we cannot paint the situation with a broad brush. Educated people also fell on hard times and needed to request assistance. 63 367 express himself in a higher register. This is possible, but in the case of the three documents here, it seems unlikely. Given the uneven spread of ability with Hebrew, the frequent impingement of non-Hebrew vernacular on the Hebrew text, the highly personal and intimate nature of the letters, and (with the exception of T-S 8J13.5) informal handwriting, it appears far more likely that these documents were penned by the beggars themselves. Thus I see begging letters as only indirectly influenced by the Islamic petitionary style and only loosely connected to the strictures of a formal petition. Ultimately they are tethered by the general practice of a petitioning culture, but remain freer in terms of formulaic structure, eloquence and content. 3.1.1 Opening Each of the documents begins either with a biblical quotation (Mosseri II.98.1 and T-S Misc. 28.18) or a version of the bismillah (T-S 8J13.5).65 A couple of the already published manuscripts also begin with a shortened form of the bismillah (T-S 18J4.4 and BL Or. 4856.1 have ‫ בשמך רחמנא‬at their beginnings). If biblical quotations are included in this section, they are usually, but not always, marked with diacritics or formatted differently from the rest of the text. Mosseri II.98.1, as noted above, places rafes over the relevant bgdkft consonants in the quotation from Psalms.66 The presence of such graphical indicators for an opening passage serves to set this section of the opening apart in an almost formulaic manner, which is strong evidence of adherence to a particular structure. Furthermore, the nearly universal tendency in these letters to use abbreviations for the honorifics indicates a formulaic structure simply by the presence of the easily recognised shorthand. In the letters published by Bareket and Scheiber, abbreviated honorifics occur in BL Or. 4856.1, T-S 18J4.4, L-G Misc. 39,67 T-S 13J17.9 and T-S 12.354. The direct addressing of the letter’s recipient tends to mark the transition point between the opening blessings and the body. This has an effect on the format of the letter, as the opening blessings are usually separated from the address by a blank space,68 although sometimes the letters do not adhere to a 65 The top of T-S Misc. 28.18 is cut off, so it is possible that this document began with the bismillah or further biblical quotations. The important point is that it is normal for biblical quotations and/or the bismillah to appear close to the beginning of these letters. Bareket, 2001, p. 362 notes that blessings and praises typically accompany the quotation or the bismillah. 66 Other examples from the published texts: BL Or. 4856.1 marks the scriptures at the beginning of the letter by adhering to the formatting of the quoted scripture, as well as using occasional diacritics to ensure the reading is correct; and T-S 8J16.29 has four lines of marked scripture beginning the letter, separated from the rest of the letter by obvious spacing. 67 Previously Westminster Misc. 39. The Westminster Collection is now the Lewis-Gibson Genizah Collection, and is shared by the Cambridge University Library and the Bodleian Libraries, University of Oxford. 68 In the already published begging letters, obvious spacing appears to occur in BL Or. 4856.1, T-S 18J4.4, ENA 2808.31, T-S 13J7.9 and T-S 12.24. In T-S NS 325.184 the addressee appears in the opening, and though it is not reflected in Scheiber’s edition, the image shows spacing in line 4 after ‫אמן‬. The body of the letter then follows. 368 strict formula of spacing (this inconsistency is unsurprising for informal begging letters).69 In the letters described here, this spacing is very evident in TS 8J13.5 and T-S Misc. 28.18. Variation is possible, as sometimes the addressee can be mentioned twice in the opening section: first with the blessings, and then in the part where he is addressed directly. This occurs in two of the letters here, Mosseri II.98.1 and T-S 8J13.5. In both instances, it is clear that the first mention of the addressee does not serve as the direct address because third person reference is used at this point; the following mention is the direct address, where second person reference is used. For example in Mosseri II.98.1, the opening refers to the addressee in the third person – ‫‘ יברכהו וישמרהו‬may he be blessed and guarded’ – and the same occurs in T-S 8J13.5 – ]‫‘ ויברך את כל מעש[יו‬may you bless all of his deeds’; these mentions of the addressee thus do not serve as direct addresses. The major indentation in the line of the text typically occurs before the second, direct mention of the addressee. This direct address both closes the opening and flows seamlessly into the body of the text. 3.1.2 Body While the opening is the lengthiest section, the body of the document is succinct. Here the register, syntax and vocabulary shift dramatically from florid, allusion-based language to a more direct and original register, though it is often interspersed with reinforcing quotations and allusions where necessary. Because this section contains many individual details, it is the most variable and the least formulaic of all the sections. The body is comprised of two parts: a justification or statement in defence of writing, and the actual request itself. In the justification the beggar defends his intentions and describes why he is writing, often emphasising his hesitancy to write a letter. This primes him to make his request elegantly. All three of the letters discussed here show this definite pattern. The justification section appears first in the formulaic order. It tends to not only justify the author’s reason for writing, but also serves to admonish the addressee to respond favourably to the request. Bareket has also noted this admonishing trend in her documents, though she does not say that the trend belongs to a justification section, instead indicating that the admonishing tone reminds the addressee of the commandment of charity.70 For the three letters here, however, it seems that this admonishment comprises an actual formulaic feature in their structure. In Mosseri II.98.1 the author spends quite a few lines justifying his reasons for writing and declaring his proper intentions. This is evident in these phrases: 69 For example, this spacing is absent from Mosseri II.98.1 discussed here, and from T-S 12.354, T-S 8J16.29 and L-G Misc. 39. 70 Bareket, 2001, p. 363. 369 ‘I did not write this letter to you out of shame, because God knows if I had [money] I would not write to you’ (recto, lines 9–12) ‘and truly, do not blame me, [because] God knows …’ (recto, line 17) T-S 8J13.5 contains two justification clauses: ‘because I came from a faraway place’ and ‘because I was one of the givers, and of those who practise [compassion]’ (lines 9 and 10). In other words, while the author is asking for support now, he is making it clear that he used to be someone who gave support to others (which justifies his current request). T-S Misc. 28.18 also contains a lengthy justification section, mainly in lines 4–7 (with the theme revisited in line 10). First, the author describes himself as a foreigner, ‘poor and destitute’, but he also couches the justification in the phrases of ‘I trust in God and in Israel’, ‘this will not fail in Israel’ and ‘the God of Israel will not falter’, thus reminding the community of its duty to help those like him. The climax of his justification lies in the scriptural quotations in line 10: after all, what is more admonishing than to quote a scripture which says, ‘open your hand to your brother’? The justification section of the letters is typically followed by the request – T-S Misc. 28.18 is unique in this regard, with the request mixed in with the justification. However, the requests (what we can see of them) are not made in an overt fashion. Their tone is indirect, which is sensible, as the general tone of begging letters is to follow a sense of decorum and avoid overt begging.71 In Mosseri II.98.1 the actual request is located on the verso, where he writes ‘even a cent[’s worth of bread] and I will dwell for two days’. In T-S 8J13.5 the request itself has been destroyed, but we can see the beginning in line 13, ‘now have compassion upon me according to your custom’, with the following request having been lost. T-S Misc. 28.18 is quite damaged, but we do have part of the request, which is made indirectly: ‘I do not have a cent and not a piece of bread … And here is a synagogue and a community and a ḥazzan and a parnas. And the God of Israel will not falter.’ Importantly, in no legible place do any of these three authors include a statement with a verb which would make a direct request. 3.2 Linguistic register These begging letters are written in medieval Hebrew, but within that language the register sometimes shifts to favour more biblicising or rabbinicising features. The most obvious changes in tone occur in the openings and closings of each document. The innermost parts of the letters are the most original and the least dependent upon external textual syntax, allowing us to glimpse the Exemplary of this desire to avoid begging is Mosseri II.98.1, where it states, ‘I did not write to you out of shame’; Cohen, 2005a, pp. 174–188 extensively discusses the desire to avoid shame in asking for assistance. 71 370 foreign influence of the particular author’s vernacular, whether it be Arabic or another language.72 As a general comment, it appears that the syntax and the vocabulary of these documents, when not influenced by the author’s mother tongue, correspond to a simplified form of what might be called Gaonic Hebrew.73 We have a mixture of Biblical Hebrew syntax and of plene and defectiva spellings, post-biblical terminology, and the use of particular archaisms, and all of these features are also found in Gaonic correspondence. However, the begging letters employ these elements in a less complex and masterful fashion, often make ‘errors’ owing to an influence of the author’s mother tongue (which is not Hebrew) upon the written Hebrew, and adhere to a formulaic, but less formal, format. 3.2.1 Opening The opening blessings tend to contain a biblical quotation, but even if they do not, they are usually written in a register which imitates biblical syntax.74 An obvious indicator of this syntax is the employment of imperfect with jussive force.75 Mosseri II.98.1 uses this technique quite heavily, and although its particular use of the jussive is not especially biblical, it is still formalised writing. Lines 6–8 are the most exemplary: …]‫ אלהינו יברכהו וישמרהו ויעודדהו [ו‬.6 … ‫ לשם ותהילה ויזכהו לראות שמחת‬.7 ‫ אמן ויזכהו לראות ביאת גואל ידוע‬.8 T-S 8J13.5 combines the imperfect with jussive force quite frequently in the opening lines: … ]‫ ויאריך ימיו בטוב ושנותי[ו‬.3 … ]‫ לטובה ויברך את כל מעש[יו‬.4 … ]‫ מפעליו ו[יז]כה ויראה בנ[ים‬.5 … ‫ וגם יחיה וישמר וינצור ויפרה‬.6 … ‫ יאריך ימיו ירבה שנותיו יום‬.7 72 However, the reader should note that I have been very careful to not speculate about the supposed vernaculars of the authors of these begging letters in this article. One can present the evidence, but one must be careful not to assert what cannot be known. What we do know is that certain external linguistic features are apparent in these letters: we cannot ultimately say for certain what language(s) the authors spoke. 73 For an overview of this kind of Hebrew, see Outhwaite, 2013. His PhD dissertation (2000) is a comprehensive study of its linguistic features. 74 This feature is found in other, more formal correspondence; for an example of a formal document written in Hebrew in this period that has an opening which uses biblical syntax and allusions, see T-S 20.173 in Ashur and Outhwaite, 2014. 75 For a more comprehensive discussion of the use of the jussive, the imperfect with jussive force and the waw-consecutive, see Niccacci, 2013, a general survey on waw-consecutive, and the references therein; also Outhwaite, 2013, pp. 4–5, which shows that Hebrew letters in this period tend to use the waw-consecutive and the jussive in imitation of older forms of Hebrew. 371 Although the syntax of T-S Misc. 28.18 is markedly different and relatively influenced by Arabic, the author still employs these features in the blessing, beginning after the space on line 2; ‫ יבואו ויוחלו‬is the most overt instance: ‫ויתקיים עליכם כל הברכות ואימ[ו]רות יש בארבע‬ ‫ תלוה לחיים ואמרו אמן‬.2 ‫דעו אדונינו‬ ‫ פינות העולם כולם יבואו ויוחלו עליכם ועל בניכם ועל בתיכם‬.3 Post-biblical language is common throughout the letters, but it is most apparent in the blessings and the honorifics surrounding the addressee (most telling are the titles ‫אדוני‬, ‫זקן‬, ‫ נבון‬and ‫)חכם‬. Specifically rabbinic vocabulary, however, is apparent in T-S Misc. 28.18, where the author switches from the more common biblical verb for ‘lending’, ‫עבט‬, and instead uses ‫לוה‬, a verb that is extremely common in rabbinic literature. As noted in section 2.3.3, the meanings of the verbs are synonymous, and so it appears that the author was using the vocabulary which was more familiar to him: he knew the concept of the quotation, but utilised the particular verb because of its familiarity. It is noteworthy that despite the use of rabbinic vocabulary, T-S Misc. 28.18 does not contain formal, rabbinicised titles in the opening address. This may be because the document was written to a community, but even when he describes the community in line 8 (‫)והינה כניסה וקהל וחזן ופר[נ]ס‬, he does not include any descriptors which would be overtly rabbinicised in nature. 3.2.2 Body Because the content of the body section is highly individualised, the writers are less dependent upon formal constructions except when they wish to use an allusion to justify their reasons for writing. The lack of dependence upon formulaic syntactic and lexical structures makes the language of the writers less standardised. In our letters, we see a few errors in the language where it is sometimes possible to see the author’s mother tongue shining through the syntax of the poorly constructed Hebrew. The most evident example is T-S Misc. 28.18, which contains a few examples of written code-switching into Arabic. While the letters themselves contain a conglomeration of biblical and postbiblical vocabulary, certain authors tend to use more of one type than another. Mosseri II.98.1, for example, uses mostly post-biblical Hebrew nouns and other vocabulary (examples include ‫חנינה‬, ‫יד ושם‬, ‫מזל‬, ‫גבוה‬, ‫ אפילו‬and ‫)גואל‬.76 The use of ‫ בלא‬as opposed to ‫ בלי‬in this letter does not seem to carry any significance when it comes to word selection or register. While T-S 8J13.5 also contains mainly post-biblical vocabulary, the author chooses a very elegant word, ‫אודיע‬, when he addresses the recipient for the first time (line 8). This archaism significantly adds to the formality of the tone of the letter, which is, 76 In determining whether a particular word or root was ‘rabbinic’, ‘biblical’ or ‘post-biblical’, I relied on three sources: the Bar Ilan Responsa database; Jastrow, 1903; and Outhwaite, 2000. 372 on the whole, the most skilled of the three. T-S Misc. 28.18 has the most diverse vocabulary of the three letters, with multiple biblical quotations, and it also includes Arabic phrases and constructions which I have tentatively traced back to the influence of a non-Semitic language. Overall, it appears that these letters are interspersed with biblical and postbiblical elements. Biblical Hebrew is frequently used rhetorically, especially in quotations and in the opening syntax of the letters, whereas post-biblical Hebrew in an informal register seems to be the baseline and the structure to which the authors orient in their general language. This indicates that the authors are by and large familiar with Hebrew as a literary language, mostly through an oral medium, but they are not highly educated and so do not have a complex mastery of the higher, more formal or more poetic registers of Hebrew that are seen in letters written by the Geonim and paytanim. 3.3 Brief notes on the closing section The brevity and lack of grandiose formulaic features are the most notable aspects of the closing section of these letters. Rarely more than a couple of lines, it consists of a transitionary blessing flowing from the body of the text. The addressee may or may not be mentioned again, but further blessings are conferred on him (again using third person reference). Rarely are the endings of the letters preserved enough to show the full signature.77 The syntax is much like that of the opening section of the letters, as it relies on biblical and rabbinic allusions to confer the final blessings upon the addressee. 4 Conclusion In this brief study I have presented an edition of three Hebrew begging letters and have thoroughly commented upon their internal structure, their linguistic register and the authors’ skills in Hebrew. I compared my observations with the features of those Hebrew begging letters which have already been published, and noted areas where the underlying native tongue of the writer shows through the literary Hebrew. Though three specific letters have been the focus here, it seems possible that the findings of my analysis may apply to other contemporary Hebrew begging letters. I wish to emphasise that more research should be done in order to confirm, on a wider and more certain scale, the conclusions which I have laid forth here concerning the structure and language of Hebrew begging letters. Through this analysis it seems clear that the language and structure of these letters are related to formal correspondence in Gaonic Hebrew and Judaeo77 The repeated shins at the end of T-S Misc. 28.18 are graphic line-fillers, and are not used as a signature. 373 Arabic, but are only loosely influenced by the Islamic petitionary model. These Hebrew begging letters consistently follow a six-part structure, the most important sections being the justification and the request. The justification typically precedes and seems to be necessary for the request. The request is couched in indirect language and is made in a florid manner, which indicates an influence from more formal petitions. The profile of their features thus makes it apparent that Hebrew begging letters are a medium which is influenced by the higher levels of Gaonic correspondence. Moreover, they show that even the average layperson was able to utilise this style of language in a formulaic and persuasive manner. Their structural features, combined with the unique and sometimes varied levels of register, not only define such letters as a distinct type of Genizah correspondence, but also situate them solidly within the rich continuum of petitionary practices in the early medieval Middle East. References Ashur, A. and Outhwaite, B., 2014, “Between Egypt and Yemen in the Cairo Genizah”, Journal of Islamic Manuscripts 5/2–3, pp. 198–219. Bareket, E., 1999, Fustat on the Nile: The Jewish Elite in Medieval Egypt, Leiden: Brill. ———, 2001, “ ‘Thou shalt surely open thy hand unto thy poor and needy brother’: letters requesting financial aid from the Genizah” (Hebrew), in Y. Hoffman (ed.), Studies in Judaica, Tel Aviv: Tel Aviv University, pp. 359–389. Cohen, M., 2000, “Four Judaeo-Arabic petitions of the poor from the Cairo Geniza”, Jerusalem Studies in Arabic and Islam 24, pp. 446–471. ———, 2005a, “Feeding the poor and clothing the naked: the Cairo Geniza”, Journal of Interdisciplinary History 35/3, pp. 407–421. ———, 2005b, Poverty and Charity in the Jewish Community of Medieval Egypt, Princeton, NJ: Princeton University Press. ———, 2006, The Voice of the Poor in the Middle Ages: An Anthology of Documents from the Cairo Geniza, Princeton, NJ: Princeton University Press. Goitein, S. D., 1954, “Petitions to Fatimid caliphs in the Cairo Geniza”, JQR 45/1, pp. 30–38. Harrell, R. S. and Brunot, L., 2004, A Short Reference Grammar of Moroccan Arabic (With Audio CD) with an Appendix of Texts in Urban Moroccan Arabic by Louis Brunot, Washington, DC: Georgetown University Press. Jastrow, M., 1903, Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature, London: Luzac & Co. Khan, G., 1990a, “A petition to the Fāṭimid caliph al-ʿĀmir”, Journal of the Royal Asiatic Society 122, pp. 44–54. ———, 1990b, “The historical development of medieval Arabic petitions”, Bulletin of the School of Oriental and African Studies 53, pp. 8–30. Lane, E. W, 1984, Arabic-English Lexicon, Cambridge, UK: The Islamic Texts Society. Niccacci, A., 2013, “Consecutive waw”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. I, pp. 569-572. 374 Outhwaite, B., 2000, A Descriptive Grammar of the Medieval Hebrew of the Cairo Genizah Letters (unpubl. diss., University of Cambridge). ———, 2001, “Karaite epistolary Hebrew: the letters of Ṭoviyyah ben Moshe”, in G. Khan (ed.), Exegesis and Grammar in Medieval Karaite Texts, Oxford: OUP, pp. 195–234. ———, 2009, “Byzantium and Byzantines in the Cairo Genizah: new and old sources”, in N. de Lange, J. Krivoruchko and C. Boyd-Taylor (eds), Jewish Reception of Greek Bible Versions: Studies in their Use in Late Antiquity and the Middle Ages, Tübingen: Mohr Siebeck pp. 182–220. ———, 2013, “Gaonic correspondence”, in G. Khan (ed.), Encyclopedia of Hebrew Language and Linguistics, Leiden: Brill, vol. II, pp. 2–6. Qimron, E., 1986, The Hebrew of the Dead Sea Scrolls, Georgia: Society of Biblical Literature. Reymond, E. D., 2014, Qumran Hebrew: An Overview of Orthography, Phonology and Morphology, Georgia: Society of Biblical Literature. Richards, D. S., 1992, “A petition for an iqtaʿ addressed to Saladin or al-ʿĀdil”, Bulletin of the School of Oriental and African Studies 55, pp. 100–105. Rustow, M., 2010, “A petition to a woman at the Fatimid court (413–414 A.H./1022– 23 C.E.)”, Bulletin of the School of Oriental and African Studies 73/1, pp. 1–27. Scheiber, A., 1981, “Begging letters from the Genizah” (Hebrew), in A. Scheiber, Geniza Studies, Hildesheim: Georg Olms, pp. 75–84 (Hebrew section). Stern, S. M., 1962, “Three petitions of the Fatimid period”, Oriens 15, pp. 172–209. Yahalom, J., 1999, “The Muño letters: the work of a village scribe from northern Spain”, Sefunot: Studies and Sources on the History of the Jewish Communities in the East 1999, pp. 23–40. 375 Birds of a Feather? Arabic Scribal Conventions in Christian and Jewish Arabic ESTHER-MIRIAM WAGNER Woolf Institute and University of Cambridge 1 Confessional varieties: Judaeo-Arabic, Christian Arabic and Muslim Arabic The founding father of the field of Middle Arabic, Joshua Blau, wrote the first proper grammar of ‘Christian Arabic’, and also coined the term ‘Judaeo-Arabic’ in his numerous works on Jewish varieties of Arabic.1 Much of his work propagated the separation between confessional dialects, rendered in statements such as: “the writers of Judaeo-Arabic, themselves, in fact, had the feeling they were writing in a separate language”.2 Blau’s works were pioneering and inspired many scholars to engage with Middle Arabic sources, but his general ideas about the nature of Middle Arabic and his ideas of segregation between the different confessional varieties of Arabic have been increasingly viewed with caution. In recent years, various scholars have argued quite strongly against a separation of Arabic dialects according to confessional lines. Johannes Den Heijer, for example, who works mainly on Christian Arabic texts, suggests that religious sociolects should essentially be viewed as registers, comparable to those used in medical and philosophical texts.3 Similar reservations have been expressed by Holes, who finds little foundation for convincing confessional difference in modern spoken dialects.4 Shohat has disputed the concept of Judaeo-Arabic itself in very strong terms and called it a nationalist projection onto the Arabic of Jews, which mirrors “the persistence of the ‘Arab versus Jew’ dichotomy”5 and “reflects an undergirding investment in dislocating Arab-Jews from their Arab past”.6 1 See Blau, 1966–1967; Blau, 1980; Blau, 1981; Blau, 1988. This paper follows Geoffrey Khan’s recommendation to use the term Judaeo-Arabic only to designate Arabic written in Hebrew script; see Khan, 2007, p. 526. 2 Blau, 1988, p. 102. 3 Den Heijer, 2012. 4 Holes, in press. 5 Shohat, 2015, p. 14. 6 Shohat, 2015, p. 64. 376 Ideology appears to play an important role in the way confessional varieties of Arabic are being described.7 In a general article on how ideologies create perceptions of linguistic difference, Gal and Irvine have contended that “speakers and observers notice, justify and rationalize linguistic differences, placing them within larger ideological frames … sometimes exaggerating or even creating linguistic differentiation”.8 The distinctness of Judaeo-Arabic in particular as opposed to other confessional varieties of Middle Arabic often seems overly emphasised. Features that are part of a greater substandard continuum are sometimes described as typical Jewish phenomena, when they can also be found in other confessional varieties. Having academically grown up with Blau’s work, many of his ideas guided my early research on Judaeo-Arabic, and over years of studying Middle Arabic documents I have nurtured the impression that Judaeo-Arabic is more progressive than Christian Arabic in the way it takes up colloquial forms, and more inventive in creating its own register, and that it is more removed from contemporary Muslim norms than Christian Arabic. The question that arises for me now, after critically engaging with the concept of confessional varieties, is whether this impression can actually be verified through linguistic analyses, and how much of this difference and the specific inventory of JudaeoArabic is potentially due to writing in Hebrew script. To approach this, I will compare two sets of texts from Jewish and Christian authors in the following for their similarities and differences. I have chosen to concentrate on mercantile correspondence, specifically letters from the Ottoman period. The corpora I have chosen are suitable for three different reasons. Firstly, as documentary sources they can be reliably dated and have not been subject to copying and editing processes. Secondly, mercantile correspondence is often linguistically more progressive than other sorts of texts, as I will describe below, and therefore linguistic change can be monitored more easily in business writing. Thirdly, work on this time period could potentially open up further avenues of comparative research, as there are literary texts available written in Garshuni and in Judaeo-Arabic from the same time period; this will enable further study of the differences between Jewish and Christian Arabic, focusing on whether the use of non-Arabic scripts by writers of the two confessional varieties yield different results from this study, in which the Christian writers use Arabic script, while the Jewish writers employ Hebrew script. 2 Features in 18th and 19th century Christian Arabic and Judaeo-Arabic mercantile letters In many text genres, the Arabic used in medieval Christian and Jewish sources is very close to what would be considered the standard Muslim Arabic of the 7 8 Wagner, in press. Gal and Irvine, 1995, pp. 992–993. 377 time. To detect difference in medieval material, one would ideally rely on corpus linguistics carried out on an extensive compilation of sources, as statistical analyses would enable us to detect variations that might go unnoticed otherwise. In contrast, an investigation of Ottoman material is more fruitful even in smaller samples of text, as we see much more variation in this late material as opposed to the earlier sources. This article thus focuses on the investigation of variation between the Arabic used in different confessional groups in the Ottoman period, concentrating on mercantile correspondence. These business letters are very suitable for purposes of comparison for various reasons. First, operating in a linguistic mercantile continuum comprising members of all three Abrahamic religions, traders by and large appear to avoid non-Arabic forms to a greater extent than, for example, dignitaries, which allows us to concentrate on the Arabic by itself, without being distracted by Hebrew which is commonly used in Jewish Arabic literature.9 Secondly, the language employed by traders in many countries displays particular features that cannot be found in other text genres. Linguistic forms in mercantile letters are often closer to the spoken language, and specific words and phrases cross the threshold of codification – that is, they are written down for the first time – in a mercantile context. The literacy of traders has been coined ‘pragmatic literacy’ by Parkes,10 as they write in particular circumstances – they typically compose their correspondence efficiently, in great quantity and fast. The socially open linguistic networks and their geographic mobility add to this particular kind of literacy induced by certain frameworks of education and writing purpose, and facilitate the introduction of progressive language forms and linguistic levelling. 11 All these aspects make traders’ writings very worthy subjects of linguistic study.12 An additional advantage in focusing on mercantile writings is the availability of suitable materials. Extant corpora of both Christian and Jewish traders’ letters from roughly the same period of time have been preserved, and with an increasing focus on Ottoman Arabic documentary materials in current Arabist research, we should soon be able to compare Christian and Jewish materials with contemporary Muslim sources of the same text genre.13 More 9 Wagner and Connolly, 2018 have shown that mercantile correspondence contains only a few Hebrew words and phrases as opposed to contemporary communal correspondence. They have also demonstrated that individuals would vary the Hebrew content of their letters according to the genre: the same writers might use extensive Hebrew phrases in their correspondence for communal, political or religious purposes, but avoid Hebrew in their mercantile letters. 10 Parkes, 1973. 11 For a study of closed and open network linguistic analyses, see Alcolado Carnicero, 2017. 12 The role of the middle class, and in particular traders, in the course of emergence of vernacular languages has been explored in a book edited by Wagner, Beinhoff and Outhwaite, 2017; see also Wagner, 2013 and Wagner, 2017. 13 In addition to the work on the Prize Papers and the Genizah collections, discussed below, other documentary corpora being worked on include diaries and letters from the Gotha Research Library, which are currently being studied and prepared for publication by Boris Liebrenz and Kristina Richardson, and Garshuni letters edited by George Kiraz. Samples of these and many 378 than a hundred Jewish letters from the late 18th and early 19th centuries have been preserved in the Cairo Genizah, scattered in the different Genizah collections, with over-proportionally large batches of correspondence preserved in the collections of the Alliance Israélite Universelle (AIU), the Rylands Genizah Collection and the Mosseri Collection. The letters come from a Jewish trade network operating in Egypt.14 From a slightly earlier time, several dozen Christian traders’ letters have been preserved in the Prize Papers collection in The National Archives in Kew, London.15 These letters were written in the years 1758 and 1759 by Egyptian Christian traders to their business partners in Egypt. The compared corpora are thus only separated by half a century. Apart from the similar geographical and chronological contexts, there is an even tighter link between the two different corpora: in the Prize Papers, the Egyptian-based company Francis and Sons is mentioned, and this is also the name of a firm receiving and sending large numbers of letters in the Genizah collections. The basis for an examination of linguistic features in Jewish documentary writing is thus this corpus of Judaeo-Arabic letters from the Cairo Genizah, comprising the ten letters listed in table 1.16 Table 1. Judaeo-Arabic letters in the corpus (RGC = Rylands Genizah Collection) Classmark Sender Addressee L-G Misc. 24 T-S 10J16.35 T-S 10J19.22 T-S 10J19.24 AIU VIIE 132 T-S NS 99.23 RGC A 803 RGC L 205 RCG A 701 T-S NS 99.38 Pinto Vasuarez Abraham Hamān & Gabriel Ḥefez Nissim Mašiš Unknown Unknown Solomon Ḥayyim & Abraham Jizana Nissim Sabbāḥ David b. Na’īm Raḥamīm Abzardil Abraham Hamān and Gabriel Ḥefez Elijah Saʿd and Jacob Šalom Mercado Karo & Simeon Fransis Mercado Karo & Simeon Fransis Mercado Karo & Simeon Fransis Mercado Karo & Simeon Fransis Mercado Ḥayyim Abraham ha-Levi Jacob Yabets Moses b. Na’īm Moses b. Na’īm Mercado Karo and Simeon Fransis These letters exhibit quite a bit of linguistic variation, with certain pieces containing many more colloquial and Middle Arabic features than others. For other Ottoman Arabic documents will be included in a forthcoming handbook and reader of Ottoman Arabic, which is being prepared by a large number of scholars from the fields of Arabic, Turkish and Ottoman Studies, to be edited by Esther-Miriam Wagner. 14 The letters are part of a project started by Geoffrey Khan, which Esther-Miriam Wagner and Mohamed Ahmed have now joined, as a result of which editions of about fifty letters will be published in the near future. Some of these Late Judaeo-Arabic letters have already been edited, and appear in: Khan, 1991; Khan, 1992; Khan, 2006; Khan, 2014; Wagner and Ahmed, forthcoming. 15 This correspondence is currently studied by Esther-Miriam Wagner and Mohamed Ahmed, to be published within the next three years. 16 The transcriptions of T-S and AIU letters in the table were originally compiled by Geoffrey Khan. I transcribed the L-G letter. With Mohamed Ahmed, I also transcribed the Rylands letters, for an edition of the three; see Wagner and Ahmed, forthcoming. 379 example, some letters spell final long /ā/ with ‫א‬, as in Standard Arabic norms, but others use ‫ה‬, reflecting the shortened pronunciation of final long vowels. Or, T-S 10J19.24 shows some rather unusual features which set it apart from most of the other correspondence. It spells some forms going back to Classical Arabic ‫ ذ‬with ‫ז‬, as in ‫ אלזי‬T-S 10J19.24/25 ‘which’ or ‫ זליך‬T-S 10J19.24/8 ‘this’,17 while other examples going equally back to Classical Arabic ‫ ذ‬have ‫ד‬, as in ‫ דלווקת‬T-S 10J19.24/10 ‘now’. In this last case, obviously the colloquial form is written as it is pronounced, with [d], while the forms which are not a regular part of the colloquial repertoire, ‫ אלזי‬and ‫זליך‬, are spelled somewhat reflecting their Classical Arabic pronunciation. This rich Jewish material is compared with a corpus of ten Christian business letters from the Prize Papers collection. These letters are all kept under the classmark HCA 32/212 in The National Archives. Some of them have individual classmarks on the sheets of paper, but because many of the letters were removed from bigger envelopes and opened for the very first time for me when I initially visited The National Archives and viewed the Arabic materials, the majority have not received further classmarks within the generic classmark for the box in which they were kept. Mohamed Ahmed and I, while working on a volume to publish the letters, have given provisional classmarks to the letters. The ten letters on which the comparison here is based are listed in table 2 – note that the last two have individual classmarks. Table 2. Christian Arabic letters in the corpus No. Sender Addressee 1 3 4 11 17 20 23 24 27 (NAL HCA 32/212 E25) 29 (NAL HCA 32/212 E23) Gerges Faranjī Ni‘mat Allāh Da’ūd ‘Īsā Zal’ūm Anton Ḵayr Yūsuf Baktī Anton Ḵayr Yūsuf Baktī Yūsuf Baktī Anton Ḵayr Anton Ḵayr Yūsuf Ni‘mat Allāh al-Šāmī Ni‘mat Allāh al-Šāmī Yūsuf and Faḍl Allāh Elias Mesk Yūsuf Yūsuf Buṭrus Demetri Ḵayr Nicola For the purposes of the comparison between Jewish and Christian letters, I concentrate here on a limited number of common features only, specifically on those which can be explored in the limited space available here, although I understand that the topic would merit a much larger and more comprehensive 17 In medieval Judaeo-Arabic epistolary writing, the Classical Arabic system of near and far deixis has been abandoned and there is a clear differentiation between pronominal demonstratives and attributive demonstratives as well as a possible distinction between anaphoric and cataphoric demonstratives. The demonstrative hāḏā is mainly used as an attributival demonstrative, while ḏālika serves almost exclusively pronominally, in this specific example with the meaning ‘this’ rather than Classical Arabic ‘that’. 380 study. The linguistic phenomena gathered here, however, already give a good impression of the difference between Jewish and Christian mercantile writing, and help us to understand the commonalities and differences encountered in Arabic materials written by members of different confessions. The point of reference will be an artificial, presumed Standard Arabic of the Ottoman period, which is close to Classical Arabic and Modern Standard Arabic. This approach is admittedly flawed and will inevitably draw criticism, as the time period of the materials precedes the nahḍa, during which ideas of normative grammar informed by Classical Arabic were again superimposed on Arabic. Yet there is no real alternative, as for now Classical Arabic or Modern Standard Arabic are the only varieties with a prescriptive, fixed set of rules, described in grammar books, against which any other variety can be measured and compared. 2.1 Otiose ʾalif Otiose ʾalif occurs very commonly in the Christian letters, but not once in the Judaeo-Arabic corpus – for example, ‫ ارسلوا‬NAL HCA 32/212.23/11 ‘you should send!’ vs ‫ תרסלו‬T-S 10J16.35/10 ‘you should send’. Clearly, Arabic script norms here hold a stronger sway over the Christian material than over the Hebrew script Judaeo-Arabic. 2.2 Vocalism In the Judaeo-Arabic, colloquial vocalisations, and also vocalisations that may not occur in colloquial Christian or Muslim Egyptian Arabic but only in Jewish Egyptian Arabic, are frequently spelled: ‫ כתיבנא‬L-G Misc. 24/3 ‘we wrote’; ‫ צועוב‬Ryland Genizah Collection A 803/5 ‘difficult to bear’; ‫חוצור‬ ַ֗ T-S NS 99.23/15 ‘it was present’; ‫ יחובו‬T-S NS 99.23/16 ‘they would like’; ‫תכון וצליתך‬ T-S 10J19.24/28 ‘it will have arrived to you’. Such colloquial forms are not found in the Christian texts, with a few exceptions that may also be due to a shift of verbal forms, such as ‫ يصالكم‬NAL HCA 32/212.17/page1verso9 ‘it will arrive to you’. The question which then arises is: Did spoken Jewish Egyptian Arabic, or the reading tradition of written substandard Jewish Egyptian Arabic, have more idiosyncratic vocalisation patterns than contemporary Christian Egyptian Arabic varieties, or is it simply not expressed in script in the latter? If we turn to the modern context, Holes has stated that there is no evidence of observable differences in speech between modern Coptic Christians and Egyptian Muslims.18 Rosenbaum, however, describes a preference for /u/ over Standard dialect /i/ in Modern Jewish Egyptian Arabic, such as in fuʿul for 18 Holes, in press. 381 fiʿil.19 However, is this distinctly differing behaviour in Jewish and Christian speech a consequence of post-nationalist linguist behaviour, or can we extrapolate from the modern situation to say that pre-nationalism Egyptian Christian and Jewish speech were different? Hary offers an explanation supporting the former thesis. He postulates that Modern Jewish fuʿul preserves an older urban Cairene, which has been lost in the other, non-Jewish dialects.20 Purely descriptively, the written Jewish Arabic of the investigated time period renders such deviating vocalisms whereas the Christian Arabic does not. This may be due to Hebrew script, or perhaps to the fact that Jewish Egyptian at the time already followed more obviously differing vocalisation – that is, older, inherited patterns – than contemporary Christian and Muslim Arabic. 2.3 Plene spelling of short vowels Plene spelling of short vowels is a very regular feature of the investigated Judaeo-Arabic letters: ‫ באין‬Ryland Genizah Collection A 803/4 ‘that’; ‫עילמיכום‬ Ryland Genizah Collection A 803/10 ‘your knowledge’; ‫ אלמוחבין‬Ryland Genizah Collection L 205/6 ‘the beloved’; ‫ מעא‬T-S 10J19.24/4 ‘with’; ‫אחנה כונה‬ T-S NS 99.23/8 ‘we were’; ‫ כאם יום‬L-G Misc. 24/21 ‘some days’; ‫ נאכוד‬T-S 10J16.35/25 ‘we will take’; ‫ ארסילנהא‬T-S 10J19.24/12 ‘we sent it’. In the Christian letters, we also occasionally find short vowels in plene spelling, although much less commonly than in the Judaeo-Arabic letters: ‫معاه‬ NAL HCA 32/212.29/recto8 ‘with it’; ‫ بكام يوم‬NAL HCA 32/212.20/recto4 ‘in a few days’.21 The much more frequent plene spelling of short vowels in Judaeo-Arabic probably has its roots in the Hebrew alphabet, perhaps also in the transferral of orthographic conventions from Hebrew, which commonly spells certain short vowels.22 2.4 Spelling of Classical Arabic long vowels Long vowels are sometimes spelled defectively in the Jewish corpus and only very occasionally in the Christian letters, so the phenomenon appears to be somewhat more frequent in the Judaeo-Arabic letters: ‫ مقيض‬NAL HCA 32/212.3/10 ‘exchanger’, Standard Arabic ‫ ערפנכום ;مقايض‬T-S NS 99.23 ‘we informed you’, Standard Arabic ‫ חיסבכום ;عرفناكم‬T-S NS 99.23/36 ‘your account’, Standard Arabic ‫حسابكم‬. Many of these examples probably reflect shortened pronunciation in the vernacular. In some of the Judaeo-Arabic letters, the vast majority of final Classical Arabic long /ā/ are spelled with ‫ה‬, for example in Rylands Genizah Collection 19 Rosenbaum, 2002, p. 37. Hary, 2009, p. 23. 21 This may be an attempt to differentiate the kam graphically from the second person plural suffix. 22 Hary, 1996, p. 732 has called this “hebraized orthography”. 20 382 A 803 and T-S NS 99.23. Exceptions are usually forms of the third person singular feminine suffix, which in order to avoid being spelled with two consecutive ‫ ה‬retain the Classical Judaeo-Arabic ‫א‬. The majority of Judaeo-Arabic letters in this corpus, however, show a preference for ‫א‬, mirroring Classical Judaeo-Arabic, for example T-S 10J19.22 and Rylands Genizah Collection L 205. In the Christian Arabic corpus, hāʾ for Classical Arabic long /ā/ occurs only very occasionally and seems restricted to particular words, for example ‫ في هذه الوقت‬NAL HCA 32/212.11/25 ‘in this time’. This particular feature of Judaeo-Arabic appears to be related to the use of Hebrew script, and potentially due to the influence of Hebrew orthography. 2.5 Spelling of hamza We might expect spelling of hamza to be more frequent in the Arabic script material than in the Hebrew script material, but in fact there is no hamza by itself in either of the corpora, and it is usually dropped or replaced by yāʾ and wāw when it would be sitting on them in Classical Arabic, which corresponds to its colloquial pronunciation. 2.6 Interdental fricatives The marking of the Standard Arabic interdental fricatives /ḏ/ and /ṯ/ varies in both corpora. In the Arabic script corpus, they are often spelled as dāl and tāʾ, as in ‫ دكرتوه‬NAL HCA 32/212.17/1verso3 ‘you mentioned it’, ‫ تم‬NAL HCA 32/212.1/2 ‘then’ or ‫ كترة‬NAL HCA 32/212.29/2 ‘magnitude’, but we also find Classical Arabic spelling indicated: ‫ هذا الذي‬NAL HCA 32/212.1/9 ‘that which’ and ‫ ثم‬NAL HCA 32/212.17/1verso3 ‘then’. In the Judaeo-Arabic letters, there is also marking of the interdental fricatives, indicated by a stroke or dot above the letter in particular words: ‫ אלדַ֗ י‬T-S NS 99.23/6 and T-S NS 99.38/5 ‘which’ and ‫ דַ֗ אלך‬T-S 10J16.35/margin3 and T-S NS 99.38/8 ‘that’. In other cases, there seems to be no marking: ‫ תלת‬T-S NS 99.23/11 ‘a third’ and ‫ תלתה‬Rylands Genizah Collection A 701/9 ‘three’. Those words which receive marking – that is, the demonstrative and relative pronouns above – seem to be part of a lexicon that is distinctly marked as ‘high standard’, as these forms do not occur in the vernacular. Both the Jewish and the Christian letters thus seem partially to be influenced by Standard Arabic norms, but also by the colloquial pronunciation. 2.7 Tāʾ marbūṭa In both corpora, tāʾ marbūṭa is not always indicated by dots: ‫ בפצה‬T-S NS 99.38/20 ‘for faḍḍa’ and ‫ بالسالمه‬NAL HCA 32/212.27/6 ‘safely’. Similarly, we often find tāʾ maftūḥa for tāʾ marbūṭa in both corpora, such as in ‫מאע סלמת‬ ‫ תעאלא‬T-S NS 99.38/15 ‘with God’s (lit. the exalted’s) protection’, ‫בקלת‬ 383 ‫ גוואבכום‬T-S NS 99.23/18 ‘scarcity of your reply’ and ‫ وعينت حب الرمان‬NAL HCA 32/212.17/1verso3 ‘a sample of pomegranate kernels’. This occurs also outside of traditional ʾiḍāfa constructions: ‫אל כַ֗ בר אלדַ֗ י טלע בקולת אן וקע צולח‬ ‫ שאפי מן אסטמבול‬T-S NS 99.38/23–24 ‘the news that came up saying the peace is close from Istanbul’, ‫ חוואלת תאנייא‬L-G Misc. 24/12–13 ‘a second money order’, and ‫ الخمست المدكورة‬NAL HCA 32/212.4/71 ‘the aforementioned five’, although in all these examples there is a tight syntactic connection between the word ending in tāʾ maftūḥa/tāʾ marbūṭa and the following word. A phenomenon only found in the Christian corpus is tāʾ marbūṭa for tāʾ, such as in ‫ اندفعة‬NAL HCA 32/212.17/1verso20 ‘they were paid’. This feature appears to be a specific phenomenon restricted to Arabic script. 2.8 Tafḵīm and tarqīq Both corpora show examples of tafḵīm and tarqīq, however the phenomena occur more commonly in the Judaeo-Arabic letters: ‫ بخصر‬NAL HCA 32/212.1/18 ‘I make a loss’ and ‫ خصرو‬NAL HCA 32/212.4/48 ‘they made a loss’, from the root ḵ-s-r ‘to make a loss’; ‫ נכלסו‬Rylands Genizah Collection A 701/5, from the root ḵ-l-ṣ ‘to release, buy up, settle a bill’; ‫ ותרוצו‬Rylands Genizah Collection L 205/5 ‘you should return’, from the root r-d-d ‘return’; ‫ צורעה‬T-S NS 99.23/43 ‘quickly’, from the root s-r-ʿ ‘quick’; ‫ טערפו‬T-S NS 99.23/9 ‘you should know’;‫ איש רסלטו‬T-S 10J19.22/14 ‘anything you sent’. Tafḵīm and tarqīq therefore seem to be more closely associated with the use of Hebrew script and the accompanying greater removal from Arabic script norms. 2.9 Nunation A feature found in both corpora is the spelling of Arabic accusative nunation -an with final nūn: ‫ حقن‬NAL HCA 32/212.4/20 ‘indeed’; ‫ חאלן‬T-S NS 99.38/11 ‘now’; ‫ דאימן‬AIU VIIE 132/33 ‘always’. 2.10 Colloquial spelling of pronouns and pronominal suffixes Pronouns and pronominal suffixes in colloquial spelling occur almost exclusively in the Judaeo-Arabic corpus. Most common are colloquial or hybrid spellings of the first person plural – such as ‫ אחנה‬T-S NS 99.23/8 and 10 ‘we’, ‫ וחנה‬T-S NS 99.23/13 ‘and we’ and ‫ נחנא‬L-G Misc. 24/5 ‘we’ – and of the second person plural – such as ‫ אנתו‬T-S NS 99.23/9 ‘you’. The suffix of the third person singular masculine is often spelled as ‫ו‬-, e.g. ‫ מעו‬T-S 10J19.22/10 ‘with him’; ‫ אוגרתו‬Rylands Genizah Collection A 701 ‘his fee’; ‫ תאריכו‬T-S NS 99.38/3 and 5 ‘today (literally: its history)’. Beside these and examples of standard orthography, we also find hybrid spellings such as ‫ אנהו‬AIU VIIE 132/margin7 ‘that he’. The inclusion of these vernacular pronouns is probably 384 encouraged by Hebrew spelling, as in the case of the third person pronominal suffix, or generally caused by the use of Hebrew script. In the Christian Arabic corpus such forms are much rarer, but do occur, as in ‫ صحبتو‬NAL HCA 32/212.23/verso8 ‘with it’. In some of the examples, identification of the forms is difficult because both hāʾ and wāw may be represented with a fairly similar squiggle of the pen. 2.11 Verb As in almost all Middle Arabic texts, the letters display shortening of the third person plural verbal endings -ūna to -ū, for example ‫ يطرحوا‬NAL HCA 32/212.27/top margin1 ‘they are throwing them away’. Yet, in contrast to the Jewish material, the Christian letters also occasionally feature examples of the long form, such as ‫ يسلمون‬NAL HCA 32/212.24/page4line6 ‘they send’. The second person plural ending is changed in analogy with the third person plural from -tum to -tū, such as in ‫ שרחתו‬T-S 10J16.35/5 ‘you explained’; ‫ اتسلمتوا‬NAL HCA 32/212.1/6 ‘you have received’. As is visible in the latter example, both Christian and Jewish Arabic texts also commonly show the colloquial stem itfaʿʿala. In the corpora investigated here, the dichotomy nifʿil – nifʿilu appears only in the Jewish letters, for example ‫ לם ענדי מה נטול‬Rylands Genizah Collection A 803/11 and L-G Misc. 24/25 ‘I have nothing [to report] to prolong [this letter]’; ‫ נרסלו‬L-G Misc. 24/12 ‘we will send’; ‫ ובנסתארגו‬L-G Misc. 24/5 ‘we are looking forward’. This may be a heritage of Maghrebian influence on spoken Jewish Egyptian dialects in the Middle Ages, as suggested by Blau, or reflect the retention of older original Egyptian forms in the Jewish dialects, as put forward by Blanc and Hary.23 Examples of other non-standard verbal forms, pertaining to vocalisation patterns and assimilation, occur only in the Judaeo-Arabic letters and appear to be related to the use of Hebrew script: ‫ וכַ֗ ברית‬T-S 10J19.24/28 ‘they (the ships) brought news’; ‫ קבטושי‬T-S NS 99.23/66 ‘you have received anything’. 2.12 Bi-imperfect Bi-imperfect forms occur very commonly in both corpora: ‫ ما بيعرف‬NAL HCA 32/212.1/31 ‘he does not know’; ‫ الن كده عمرى ما بعرف ان كان بخصر واال بكسب‬NAL HCA 32/212.1/16–17 ‘because otherwise I will never know in my lifetime whether I am making a loss or a profit’; ‫ بيتوجه‬NAL HCA 32/212.20/recto5 ‘it will be leaving’; ‫ לאן בנסמע סמעאת‬Rylands Genizah Collection L 205/10 ‘because we keep on hearing rumours’; ‫ בינבאע‬T-S 10J16.35/margin5 ‘we are buying’; ‫ ובנשופוה‬T-S NS 99.23/62 ‘we are watching’. 23 See Blau, 1981, pp. 56–64; Blanc, 1974; Hary, 1992, p. 278. 385 2.13 Auxiliary verbs We find similar use of auxiliary verbs in both corpora, again demonstrating that there are connections between the registers used in Jewish and Christian letters. Derivatives of the verb baqā are employed, like in colloquial Arabic, and can for example be found in: ‫ نبقا نوجه لكم‬NAL HCA 32/212.17/1verso15 ‘it will arrive to you’; ‫ وبقى معلومي‬NAL HCA 32/212.1/5 ‘it became known to me’; ‫ نبقا نعرفكم‬NAL HCA 32/212.29/13 ‘we will inform you’; ‫יבקא אל עריל‬ ‫ יכסיב‬T-S 10J19.24/7 ‘the Christian will continue making profit’; ‫פי בקא עמלנא‬ T-S 10J19.24/12 ‘we continued making’; ‫ תבקו תלקו באלכם‬T-S 10J16.35/14 ‘keep on directing your attention’; ‫ בקה וצלכם‬AIU VIIE 132/7 ‘it continued arriving to you’. In a very similar use, we find derivatives of the verb ṣāra: for example, ‫ צאר מעלומנא‬T-S 10J16.35/5 and AIU VIIE 132/17 ‘it became known to us’; ‫ صار معلومي‬NAL HCA 32/212.17/1verso18 ‘it became known to me’. 2.14 Negation The use of negative markers varies greatly in different registers of Arabic, and much can be inferred from the occurrence and distribution of certain particles in particular constructions. The negation particle lam, for example, appears as a particular substandard register marker in various Middle Arabic varieties.24 According to prescriptive Classical Arabic grammar, lam must be combined with the apocopate to negate the past only. In Ottoman Arabic literary and documentary texts written by members of all confessions, lam appears to have enlarged its functions considerably. As I have shown elsewhere, it becomes the main and almost exclusive negation particle in 18th–19th century epistolary Judaeo-Arabic, where it is used as both a verbal and a nominal negation particle and takes over the functions of most other negation particles.25 It also appears in Muslim sources,26 and in the early 20th century travelogue written by al-Jarādī.27 In the Judaeo-Arabic corpus investigated here, lam is used abundantly in all letters, for both verbal negation of all tenses and for nominal negation: ‫לם‬ ‫ הייא‬L-G Misc. 24/18 ‘it is not’; ‫ לם תרסיל‬T-S 10J19.24/28 ‘do not send’; ‫לם‬ ‫ פכרתו פינה‬T-S NS 99.23/38 ‘you did not think of us’. The negation particle mā, however, only occurs once in a negation phrase in the Judaeo-Arabic corpus, with a verb which also shows the colloquial negation marker -š: ‫ מה תגיבוש‬Rylands Genizah Collection A 803/9 ‘do not blame it on us’. In the Christian material, lam does occur in many of the letters in all sorts of negation phrases, just as in the Jewish sources. It is found, however, much 24 For lam used in mercantile Judaeo-Arabic, see Wagner, 2013. For a more general discussion of lam in spoken and written varieties of Arabic, see Wagner, 2010, pp. 141–150. 25 Wagner, 2017. 26 See Lentin, 2008. 27 Mittwoch, 1926. I am indebted to G. Rex Smith for introducing me to this text. 386 less frequently than in the Jewish letters: ‫ ولم معنا زمان‬NAL HCA 32/212.3/13 ‘we do not have time’; ‫ لم ناخد في بالنا‬NAL HCA 32/212.23/19 ‘we do not care’; ‫ لم وجدنا شي‬NAL HCA 32/212.24/page2verso3 ‘we did not find anything’. The negation particle mā, however, which is very rare in the Jewish letters, occurs more frequently: ‫ ما معي خبر‬NAL HCA 32/212.1/9 ‘I had no news’; ‫ما بعرف‬ NAL HCA 32/212.1/16 ‘I do not know’; ‫ ما يصالكم‬NAL HCA 32/212.23/1verso16 ‘it will not arrive’. To summarise, in the Christian letters, lam is used to negate past and present, and, just as in Late Judaeo-Arabic and other forms of Ottoman Arabic, it can also be found as a nominal negation particle. In the investigated corpora, however, it appears to be much more common in the Judaeo-Arabic than in the Christian sources. The Christian traders use the negation particle mā more frequently; it is only found once in the Judaeo-Arabic corpus. This analysis becomes even more interesting when we look at letters written by Christian church dignitaries, which are also preserved in the Prize Papers collection. There many more cases of lam occur; see, for example, the inlay in NAL HCA 32/212.46.1, which shows a large number of phrases with lam:28 ‫ لم استعنيتم‬NAL HCA 32/212.46.1 inlay/recto 7 ‘you did not care’; ‫ولم اتآ‬ ‫ في بالكم فكرا ً قط‬NAL HCA 32/212.46.1 inlay/recto 9 ‘it did not occur to you at all’; ‫ ولم اريد‬NAL HCA 32/212.46.1 inlay/recto 7 ‘I do not want’; ‫وال علم عنه ولم‬ ‫ من اين هي هذه الخيبه‬NAL HCA 32/212.46.1 inlay/recto right margin 19 ‘no knowledge of it, and no [knowledge] where this failure [had come] from’. It thus appears that with regard to this feature, the mercantile Judaeo-Arabic register is closer to the register used by church dignitaries than to the contemporary Christian mercantile register. 2.15 Colloquial words and phrases Both corpora display a large number of words and phrases which are part of the colloquial lexicon, and there is no discernibly difference between the Jewish and Christian letters in their use of the vernacular: ‫ منشان‬NAL HCA 32/212.29/recto8 ‘concerning’; ‫ حكم‬NAL HCA 32/212.29/recto9 ‘concerning’; ‫ يكون خاطركم مطمن‬NAL HCA 32/212.11/52 ‘you may calm yourself’; ‫מן‬ ‫ שאן‬AIU VIIE 132/17 ‘concerning’; ‫ חוכם‬T-S NS 99.38/28 ‘concerning’; ‫ דלווקת‬T-S 10J19.24/7 and 10 ‘now’; ‫ ומנערפוש‬L-G Misc. 24/15 ‘we do not know’; ‫ והלבת‬T-S 10J19.22/21 ‘probably’; ‫ فعلي خيرا‬NAL HCA 32/212.17/page1verso25 ‘hopefully’; ‫ معرفيني‬NAL HCA 32/212.1/28 ‘you tell me’ (which appears to reflect the colloquial Egyptian form maʿraf-ni and maʿrif-ni ‘you tell me’). More complex colloquial phrases often appear in emotional statements, for example expressing anger, where the register change creates a dramatic effect: ‫ ما بيعرف فين عقله‬NAL HCA 32/212.1v/3–4 ‘he does not know where his mind 28 These and more examples can be found in Wagner and Ahmed, 2017, pp. 396–397. 387 is’; ‫ الن كده عمرى ما بعرف ان كان بخصر واال بكسب‬NAL HCA 32/212.1/16–17 ‘because otherwise I will never know in my lifetime whether I am making a loss or a profit’; ‫ לם אחנה כותע וואלה אחנה מבצרווייה‬T-S NS 99.23/23 ‘we are neither one-armed nor blind’; ‫ מה תגיבוש‬Rylands Genizah Collection A 803/9 ‘do not blame it on us’. Emotions such as anger are a common socio-linguistic trigger and facilitate the introduction of vernacular expressions. 3 Conclusions The analyses done for this contribution have revealed a fairly complex picture. On the one hand, orthographically Judaeo-Arabic appears to be more removed from Standard Arabic than the Christian letters. Some of these linguistic differences between the Christian and Jewish letters can be traced back to the use of script. The frequently found otiose ʾalif in Christian letters is obviously a norm inherited from Arabic scribal traditions, which has been lost in documentary Judaeo-Arabic. Plene spelled short vowels, defective spelling of Classical Arabic long vowels, ‫ ה‬for Classical Arabic long /ā/ and tafḵīm and tarqīq in the Judaeo-Arabic correspondence appear to be aided by Hebrew script and Hebrew orthographical conventions. Tāʾ marbūṭa for tāʾ, however, only occurs in Christian letters and appears to be associated with Arabic script. With other phenomena it is less clear how much they are influenced by the script or by the underlying dialect, such as the vocalism patterns found in the Judaeo-Arabic corpus. Potentially we may infer from the examples found in the texts that the vocalism found in Early Modern Jewish speech was different from that used in Christian and Muslim speech, following older inherited Cairene speech patterns abandoned by the other communities, but this needs to be investigated further. Some phenomena are shared, such as spelling of interdental fricatives or of the nunation -an with final nūn. Features restricted to the Judaeo-Arabic letters include the colloquial spellings of pronouns and pronominal suffixes, perhaps influenced by Hebrew orthography, and the dichotomy nifʿil – nifʿilu, which may indicate a morphological difference between the forms used in spoken Jewish and Christian Egyptian Arabic. Only Christian letters show the long form of the third person plural verbal endings. While the Judaeo-Arabic letters may be further removed from Classical Arabic on an orthographical level, and related to that also with regard to some morphological features, it is difficult to maintain the same for syntactical phenomena, although we may see a difference in frequency. The bi-imperfect is a frequent occurrence in both corpora. The negation particle lam in non-Classical uses, including that of nominal negation, occurs in both Christian and Jewish letters, although looking just at mercantile correspondence, it is much 388 more common in the Jewish letters. However although this uneven distribution may be seen as an example of a Judaeo-Arabic more removed from the standard than its Christian counterpart, further study of letters written between Christian dignitaries reveals that lam is an extremely common part of the register used in that correspondence, too. As for lexical peculiarities, both corpora display a rich share of colloquial phenomena, and no discernible difference in the inclusion of vernacular forms can be found. Returning to the initial question of whether Judaeo-Arabic is linguistically more progressive than Christian Arabic in its inclusion of colloquial forms, and more inventive in the creation of its own register, while more removed from contemporary prescriptive Arabic than Christian Arabic, this study reveals a heterogeneous picture. The Judaeo-Arabic letters indeed display greater variation with regard to orthographical and some morphological features, which may create the impression of greater removal from Standard Arabic standards, but the overall picture is much less pronounced when we focus on a syntactical and lexical level. All of the features mentioned thus need to be investigated in greater depth, ideally by means of corpus linguistics. Further investigations should also focus on literary sources, which even in the Ottoman period seems to adhere more closely to the prescriptive norms of Arabic than the documentary sources. Optimal material for comparison with the above material would be contemporary Judaeo-Arabic texts, such as the Qissat Hannah (T-S Ar. 54.63)29 or the Townsman and Fellah (preserved in a post-medieval copy in AIU VII.C.16),30 which both come from the 18th–19th century. Interesting also would be an in-depth study on Christian texts written in Garshuni, to see whether the use of alphabets other than Arabic produces similar effects in Jewish and in Christian texts. Appropriate sources for comparative purposes would be contemporary to the above-mentioned texts, such as The Story of Zayana the King’s Daughter who Converted 1,130,000 Souls, which has been dated to the 18th century (preserved under CUL Or. 1125).31 Potentially, although outside of the mercantile arena, another excellent corpus for comparison would be the 19th century Patriarch letters written in Garshuni, which are currently being examined by George Kiraz. A comparison of all these materials will allow us comprehensive insights into the commonalities and differences of the historical varieties of Arabic as used by members of the three Abrahamic religions. 29 This text was edited and described linguistically in depth in Schorreel, 2011. This text was edited in Goitein, 1972. It is dated there to the 17th century, but from the script it is clear that it must be a later text. 31 I thank Dr Chip Coakley for acquainting me with the relevant Garshuni materials. Chip is currently compiling a catalogue of Syriac and Garshuni manuscripts in Cambridge University Library, where the above-mentioned source is described. 30 389 References Alcolado Carnicero, J. M., 2017, “Bridges of innovation and change: the English language around the networks of the Mercery of London”, in E.-M. Wagner, B. Beinhoff and B. Outhwaite (eds), Merchants of Innovation. The Languages of Traders, Berlin: de Gruyter-Mouton, pp. 40–67. Blanc, H., 1974, “The nekteb – nektebu imperfect in a variety of Cairene Arabic”, Israel Oriental Studies 4, pp. 206–226. Blau, J., 1966–1967, A Grammar of Christian Arabic Based Mainly on South-Palestinian Texts from the First Millennium, Louvain: Secrétariat du Corpus SCO. ———, 1980, Diqduq ha-ʿAravit ha-Yehudit Šel Yeme ha-Benayim, Jerusalem: Magnes Press. ———, 1981, The Emergence and Linguistic Background of Judaeo-Arabic, 2nd edn, Oxford: OUP. ———, 1988, Studies in Middle Arabic and its Judaeo-Arabic Variety, Jerusalem: Magnes Press. Den Heijer, J., 2012, “Remarques sur la langue de quelques textes copto-arabes médiévaux”, in J. Lentin and J. Grand’Henry (eds), Moyen arabe et variétés mixtes de l’arabe à travers l’histoire, Louvain-la-Neuve: Peeters Press, pp. 113–139. Gal, S. and Irvine, J. T., 1995, “The boundaries of languages and disciplines: how ideologies construct difference”, Social Research 62/4, pp. 967–1001. Goitein, S. D., 1972, “Townsman and fellah: a Geniza text from the 17th century”, Asian and African Studies 8, pp. 257–261. Hary, B., 1992, Multiglossia in Judaeo-Arabic, Leiden: Brill. ———, 1996, “The adaptations of Hebrew script”, in P. T. Daniels and W. Bright (eds), The World’s Writing Systems, New York: OUP, pp. 727–734. ———, 2009, Translating Religion: Linguistic Analysis of Judaeo-Arabic Sacred Texts from Egypt, Leiden: Brill. Holes, C., in press, “Confessional varieties”, in E. Al-Wer and U. Horesh (eds), Handbook of Arabic Sociolinguistics, London: Routledge. Khan, G., 1991, “A study of the Judaeo-Arabic of late Genizah documents and its comparison with Classical Judaeo-Arabic”, Sefunot 20, pp. 223–234. ———, 1992, “Notes on the grammar of a late Egyptian Judaeo-Arabic text”, Jerusalem Studies in Arabic and Islam 15, pp. 220–239. ———, 2006, “A Judaeo-Arabic commercial letter from early nineteenth century Egypt”, Ginzei Qedem 2, pp. 37–59. ———, 2007, “Judaeo-Arabic”, in K. Versteegh (ed.), Encyclopedia of Arabic Language and Linguistics, Leiden: Brill, vol. II, pp. 526–536. ———, 2014, “A Judaeo-Arabic document from Ottoman Egypt in the Rylands Genizah Collection”, in R. Smithuis and P. S. Alexander (eds), From Cairo to Manchester: Studies in the Rylands Genizah Fragments, Oxford: OUP, pp. 233–248. Lentin, J., 2008, “Unité et diversité du moyen arabe au Machreq et au Maghreb: quelques données d’après des textes d’époche tardive (16ème–19ème siècles)”, in J. Lentin and J. Grand’Henry (eds), Moyen arabe et variétés mixtes de l’arabe à travers l’histoire, Louvain-la-Neuve: Peeters Press, pp. 305–319. Mittwoch, E., c1926, Aus dem Jemen: Hermann Burchardts letzte Reise durch Südarabien. Leipzig: Deutsche Morgenländische Gesellschaft. Parkes, M., 1973, “The literacy of the laity”, in D. Daiches and A. Thorlby (eds), The Mediaeval World, London: Aldus Books, pp. 555–577. Rosenbaum, G., 2002, “The Arabic dialect of Jews in modern Egypt”, Bulletin of the Israeli Academic Center in Cairo 25, pp. 35–46. 390 Schorreel, T., 2011, An Edition and Translation of T-S Ar. 54.63 with a Grammatical Analysis of the Text (unpubl. diss., University of Gent). Shohat, E., 2015, “The question of Judaeo-Arabic”, Arab Studies Journal 23/1, pp. 14–76. Wagner, E.-M., 2010, Linguistic Variety of Judaeo-Arabic in Letters from the Cairo Genizah, Leiden: Brill. ———, 2013, “Challenges of multiglossia: the emergence of substandard Judaeo-Arabic registers”, in E.-M. Wagner, B. Outhwaite and B. Beinhoff (eds), Scribes as Agents of Language Change, Berlin: de Gruyter-Mouton, pp. 259–273. ———, 2017, “The socio-linguistics of Judaeo-Arabic mercantile writing”, in E.-M. Wagner, B. Beinhoff and B. Outhwaite (eds), Merchants of Innovation. The Languages of Traders, Berlin: de Gruyter-Mouton, pp. 68–86. ———, in press, “Judaeo-Arabic language or Jewish Arabic sociolect? Linguistic terminology between linguistics and ideology”, in L. Kahn and M. Geller (eds), Jewish Languages, Leiden: Brill. ——— and Ahmed, M., 2017, “From Tuscany to Egypt: eighteenth century Arabic letters in the Prize Paper Collections”, Journal of Semitic Studies 62/2, pp. 389– 412. ——— and Ahmed, M., forthcoming, “Nineteenth century mercantile correspondence in the Rylands Genizah Collections”. ———, Beinhoff, B. and Outhwaite, B. (eds), 2017, Merchants of Innovation. The Languages of Traders, Berlin: de Gruyter-Mouton. ——— and Connolly, M., 2018, “Code-switching in Judaeo-Arabic documents from the Cairo Geniza”, Multilingua 37/1, pp. 1–23. 391 A 19th Century CE Egyptian Judaeo-Arabic Folk Narrative: Text, Translation and Grammatical Notes MAGDALEN M. CONNOLLY University of Cambridge 1 Introduction The manuscript BnF Hébreu 5831 (dated to 1839) contains, amongst other material, three Egyptian Judaeo-Arabic (henceforth JA) tales,2 depicting fictional events in the life of Abraham ibn ʿEzra (c1089–1167), the renowned Jewish Biblical scholar and polymath. This edition focuses on the third of these tales in which Abraham ibn ʿEzra, brought from Cairo by two students at the urgent behest of a Rabbi, saves the life of the Rabbi’s son and secures the freedom of a Jewish community. While the literary contents of this tale are doubtless worthy of exploration,3 the purpose of this paper is to provide a new transcription of the original text, with simultaneous transcoding into Arabic script,4 an English translation, and a commentary on the text’s notable linguistic features. 1 This manuscript was kindly made available to me by the Département de la reproduction at the Bibliothèque nationale de France, Paris. As of 2016, the manuscript is available to view online at http://archivesetmanuscrits.bnf.fr/ by searching for Hébreu 583. 2 These three tales may be found in BnF Hébreu 583, ff. 135 r.–141 r. The first tale begins on ff. 135 r.–137 v.; ff. 137 v.–139 v. comprise the second tale; and ff. 139 v. 19–141 r. 20 the third tale, on which this paper is centred. A Hebrew edition of the three tales from the manuscript BnF Hébreu 583 was first produced by Yitshak Avishur (1992). Avishur’s interest in the folk narrative appears to have been predominantly literary and historical, and consequently his transcription does not reflect the true state of the text’s orthographic features. A new transcription and translation is required for the manuscript to be of use to broader audiences – linguistic, literary and historical. This paper, therefore, acts as a supplement to the existing edition by Avishur. 3 Of particular interest in this folk-tale is the accusation of blood libel – directed throughout the Middle Ages at European Jewish communities by Christians – here levelled at a Christian community. It was not uncommon in Europe during the medieval period for Jews to be accused of kidnapping and murdering Christian children for ritual sacrifice. This particular folk-tale bears a striking resemblance to the tale of William of Norwich, perpetuated by Thomas of Monmouth, in which it was suggested that each year, Jews would decide upon a country from which to take a Christian child for human sacrifice. 4 In transcoding the text into Arabic script, I hope to make this Judaeo-Arabic text with its many noteworthy linguistic features available to a wider audience interested in varieties of Middle 392 There are significantly fewer extant texts, across a narrower range of genres, which are composed in JA and datable to the Ottoman period, compared with the preceding periods. Letters and folk narratives, however, were among those genres which continued to be written in JA. Of the JA letters5 and folk narratives that survive from this period, most are dated to the mid-to-late 18th and early 19th centuries, and are written in ‘late JA’.6 Prior to the 15th century, JA was primarily based upon the imitation and emulation of Classical Arabic (henceforth CA) orthographic, morphological and syntactic principles. From approximately the 15th century onwards, this general adherence to the strictures of CA appears to have been gradually permeated with colloquial phonetic features, an increase in Hebrew and Aramaic influences and the standardisation of pre-existing pseudo-literary features. The varied linguistic elements of the folk narratives found in BnF Hébreu 583 conform to the general principles – in so much as these have been established7 – of late JA, while provoking questions regarding the assumed origins and influences of some of these features. 2 2.1 Text, transcoding and translation Notes on the critical edition The text has been as faithfully rendered as possible, including all diacritics and orthographic idiosyncrasies found in the original manuscript. The JA text has then been transcoded into Arabic letters, grapheme-for-grapheme, except where the text uses Hebrew words. No adjustments or amendments have been made to the text in its transcoded form. As for the English translation, any additions that have been made to aid comprehension and readability are enclosed in parentheses. Arabic. In so doing, I follow the practice pioneered by Diem, 2014, kindly suggested to me by Dr Esther-Miriam Wagner. 5 Prof. Geoffrey Khan has produced a number of critical editions of JA letters from the Ottoman period; see Khan, 1991; Khan, 1992; Khan, 2006; Khan, 2013. These not only comprise transcriptions and translations of late JA letters, but also include detailed analyses of each text’s linguistic features. These editions are invaluable to furthering the study of JA, and late JA in particular, and provide the template on which this paper is based. 6 Late JA is generally dated to the 15th–19th centuries; see Khan, 2007, pp. 526–529 and Hary, 1997, pp. 199–203 for discussions regarding the chronology of JA. 7 Studies of late JA texts have tended to focus on letters (see Khan, 1991; Khan, 1992; Khan, 2006; Khan, 2013; Wagner, 2010), šarḥ (see Hary, 1992; Hary, 2009) and folk-tales (HassonKenat, 2016; Ørum, 2017). Very limited linguistic analysis has been conducted in relation to other genres of late JA literary and documentary texts. Caution is, therefore, advisable when speaking of the ‘principles’ or ‘standards’ of late JA. 393 2.2 The critical edition Translation Transcoding Text 139 v. 139 v. 19 They also recounted that in one of the towns of the uncircumcised, every ‫ ايضه اخبرو ان كان في بلد‬19 ‫من بالد ال ערלים كانو كل‬ 20 year during their festival, they would take one of the Jews, making him a sacrifice for the(ir) idols. ‫ سنه في عيدهوم ياخدو واحد‬20 ‫ياودي يعملوه קורבן لل ע״ז‬ ‫איצה אכַ֗ ברו אן כאן ַ֗פי‬ ַ֗ 19 ‫בלד מן בלאד אל ערלים‬ ‫כאנו כל‬ ‫ סנה פַ֗ י עידהום יאכַ֗ דו‬20 ‫ואחד יאודי יעמלוה קורבן‬ 9 ‫לל ע״ז‬ 139 v. 8 21 (Every year), the Jews would cast lots for the children of the Jews in order that ‫ وكانو ال ياود يعملو גורל‬21 ‫ וכאנו אל יאוד יעמלו גורל‬21 ‫عال اووالد ال ياود الجل ما‬ ‫עלא אוולאד אל יאוד לאגַ֗ ל‬ ‫מא‬ 22 they might know who would be made a sacrifice the following year, so that he might take ‫ يعرفو مين الذي ينعمل‬22 ‫קורבן ال سنه ال اتييه الجل‬ ‫ان يطلع‬ 23 his expenses from the uncircumcised during the year. When the appointed time came, 140 r. 1 they would take him in a great procession and all that he asked for would be granted. One year, 2 they cast the lot (and) it fell on the son of the Rabbi of that town. He was 3 twenty years old and the Rabbi had no other (children) besides him. This boy had ‫יערפו מין אלדַ֗ י ינעמל‬ ַ֗ 22 ‫קורבן אל סנה אל אתייה‬ ‫לאגַ֗ ל אן יטלע‬ ‫ مصروفو من عند ال ערלים‬23 ‫מצרופו מן ענד אל ערלים‬ ַ֗ 23 ‫طول ال سنه ⸱ وحين يجي‬ ‫טול אל סנה ⸱ וחין יגַ֗ י אל‬ ‫ال ميعاد‬ ‫מיעאד‬ 140 r. 140 r. ‫ ياخدوه بموكب عظيم وكل‬1 ‫الذي يطلوب ينول ⸱ في سنه‬ ‫من ال سنين‬ ‫עטים וכל‬ ַ֗ ‫ יאכַ֗ דוה במוכב‬1 ‫אלדַ֗ י יטלוב ינול ⸱ פַ֗ י סנה‬ ‫מן אל סנין‬ ‫ عملو ال גַ֗ ורל طلع عال ابن‬2 ‫ עמלו אל גַ֗ ורל טלע עלא‬2 ‫אבן אל רב בתאע דַ֗ אלך אל ال רב بتاع ذالك ال بلد وكان‬ ‫בלד וכאן‬ ‫ عومرو عشرين سنه ولم كان‬3 ‫ עומרו עשרין סנה ולם כאן‬3 ‫אל ראב ענדו כַ֗ לאפַ֗ ו ⸱ וכאן ال ראב عندو خالفو ⸱ وكان‬ ‫ذالك ال ولد‬ ‫דַ֗ אלך אל ולד‬ 8 The term ‫‘ ערלים‬uncircumcised’ (sg ‫ )ע ֵָרל‬is a Jewish term used exclusively to refer to nonJews, generally Christians and Muslims; see Jastrow, 2005, p. 1119. Here, the term is used to refer to Christians. 9 ‫ע״ז‬: ‘idolatry’ (‫)עבודה זרה‬. 394 ‫נטיר ַ֗פאל דונייא ‪ 4‬ليس لهو نظير فال دونييا من‬ ‫‪ 4‬ליס להו ַ֗‬ ‫ال فصاحا وال قراييه‬ ‫מן אל ַ֗פצאחא ואל קראייה‬ ‫وخالفو في حين الذي‬ ‫לאפו ַ֗פי חין אלדַ֗ י‬ ‫וכַ֗ ַ֗‬ ‫‪4 no equal in the world in‬‬ ‫‪terms of eloquence and the‬‬ ‫‪recitation (of the Torah),‬‬ ‫‪and so on, so when‬‬ ‫‪ 5‬טלע אל גַ֗ ורל עלא אל ולד ‪ 5‬طلع ال גַ֗ ורל عال ال ولد قامو‬ ‫קאמו באל בוכא ואל צייאח بال بوكا وال صيياح وال‬ ‫نوواح وال حوزن‬ ‫ואל נוואח ואל חוזן‬ ‫‪5 the lot fell on the boy, they‬‬ ‫‪(all) began weeping and‬‬ ‫‪wailing and mourning and‬‬ ‫‪grieving.‬‬ ‫‪ 6‬וכאן מיעאד אלדַ֗ י יאכַ֗ דו אל ‪ 6‬وكان ميعاد الذي ياخدو ال‬ ‫ولد يقربوه يحكوم اوول‬ ‫ולד יקרבוה יחכום אוול‬ ‫לילת פסח في ~‬ ‫לילת פסח ַ֗פי ~‬ ‫‪6 (The) date on which they‬‬ ‫‪would take the boy and‬‬ ‫‪present him as a sacrifice‬‬ ‫‪was decided as the first‬‬ ‫‪night of Passover.‬‬ ‫אלתפת אל ראב לאתנין מן ‪ 7‬التفت ال ראב التنين من ال‬ ‫ַ֗‬ ‫‪7‬‬ ‫תלמידים وقال لهوم تعرفو‬ ‫אל תלמידים וקאל להום‬ ‫تروحو مصر‬ ‫תערפו תרוחו מצר‬ ‫‪7 The Rabbi turned to two of‬‬ ‫‪the students and said to‬‬ ‫‪them ‘You know you will‬‬ ‫‪go to Cairo‬‬ ‫‪ 8‬לענד אל רב אברהם ןַ֗ עזרא ‪ 8‬لعند ال רב אברהם ןַ֗ עזרא‬ ‫עא״ס תסלמו להו האדַ֗ א אל עא״ס تسلمو لهو هاذا ال‬ ‫جوواب وتعرفو‬ ‫גַ֗ וואב ותערפו‬ ‫‪8 to the place of Rabbi Abra‬‬‫‪ham ibn ʿEzra, peace be‬‬ ‫‪upon him! You will deliver‬‬ ‫‪this letter to him and you‬‬ ‫)‪will inform (him‬‬ ‫‪9 of this appeal. As for the‬‬ ‫‪people of your home(s),‬‬ ‫‪we will support them‬‬ ‫‪ 11‬בלד למצר תלת אושהור ‪ 11‬بلد لمصر تلت اوشهور‬ ‫רוואח ותלת אושהור מגַ֗ יי ⸱ روواح وتلت اوشهور مجيي‬ ‫⸱ وهلبت‬ ‫והלבת‬ ‫ַ֗‬ ‫‪11 town to Cairo it is three‬‬ ‫‪months going and three‬‬ ‫‪months coming back. No‬‬ ‫‪doubt‬‬ ‫‪12 you will stay in Cairo for‬‬ ‫‪a month. It will be seven‬‬ ‫’‪months until you return.‬‬ ‫‪ 9‬פי האדַ֗ י אל דעווה ומן גיהת ‪ 9‬في هاذى ال دعووه ومن‬ ‫جيهت اهل منزلكوم نحن‬ ‫נצרופ‬ ‫ַ֗‬ ‫אהל מנזלכום נחן‬ ‫نصروف عليهوم‬ ‫עליהום‬ ‫‪10‬‬ ‫‪10 until you return, on the‬‬ ‫‪ 10‬لحين ما تحضرو وبشرط لم‬ ‫תחצרו ובשרט‬ ‫‪ 10‬לחין מא ַ֗‬ ‫‪condition that you do not‬‬ ‫تتعووقو فال طريق ⸱ الن من‬ ‫לם תתעווקו ַ֗פאל טריק ⸱‬ ‫‪tarry on the road. For from‬‬ ‫هاذى ال‬ ‫לאן מן האדַ֗ י אל‬ ‫‪this‬‬ ‫‪ 12‬תקעודו ַ֗פי מצר שהר יציר ‪ 12‬تقعودو في مصر شهر‬ ‫يصير سبعت اوشهور لحين‬ ‫סבעת אושהור לחין מא‬ ‫ما تحضرو في‬ ‫תחצרו פַ֗ י‬ ‫ַ֗‬ ‫‪13 They replied, ‘We hear‬‬ ‫‪ 13‬קאלו להו סמיע מוטיע יא ‪ 13‬قالو لهو سميع موطيع يا‬ ‫‪you and are obedient, O‬‬ ‫סיידנא ַ֗פי כתב להום גַ֗ וואב سييدنا في كتب لهوم جوواب‬ ‫‪our master.’ So, he wrote a‬‬ ‫واتووجهو‬ ‫ואתווגַ֗ הו‬ ‫‪letter for them and they set‬‬ ‫‪off‬‬ ‫‪I am grateful to Dr Nadia Vidro for her suggested translation of this sentence.‬‬ ‫‪395‬‬ ‫‪10‬‬ 14 for Cairo. After three months, they arrived in Cairo and they happened upon a poor man11 15 walking in the road. They asked him, ‘Where is the house of Abraham ibn ʿEzra?’ He replied, ‫ لمصر من بعد تلت‬14 ⸱ ‫اوشهورحوضرو في مصر‬ ‫وجدو واحد עני‬ ‫ למצר מן בעד תלת‬14 ⸱ ‫חוצרו פַ֗ י מצר‬ ַ֗ ‫אושהור‬ ‫וגַ֗ דו ואחד עני‬ ‫ ماشي فال سكه سالوه فين‬15 ‫ מאשי ַ֗פאל סכה סאלוה ַ֗פין‬15 ‫بيت ال רב אברהם ןַ֗ עזרא‬ ‫בית אל רב אברהם ןַ֗ עזרא‬ ‫في قال لهوم‬ ‫ַ֗פי קאל להום‬ 16 ‘I am he!’ They gave him the letter. Then he said to them, ‘There’s nothing for it! ‫ هوا انا ⸱ اعطو لهو ال‬16 ‫جوواب ⸱ في قال لهوم لم فيه‬ ‫باس‬ ‫ הוא אנא ⸱ אעטו להו אל‬16 ‫גַ֗ וואב ⸱ ַ֗פי קאל להום לם‬ ‫ַ֗פיה באס‬ 17 With God’s help, I will come with you and I will answer this appeal and all will be well.’ He then ‫ בע״ה نتووجو معاكوم‬17 ‫نقضي هاذى ال دعووه ولم‬ ‫يكون اال خير ⸱ في‬ ‫ נתווגַ֗ ה מעאכום‬12‫ בע״ה‬17 ‫נקצי האדַ֗ י אל דעווה ולם‬ ַ֗ ‫יכון אלא כַ֗ יר ⸱ ַ֗פי‬ 18 took them to his home (where) they stayed with him. After a month, they said to him, ‘O our master, ‫ اخدهوم اال منزلو وقعدو‬18 ‫ אכַ֗ דהום אלא מנזלו וקעדו‬18 ‫عندو ⸱ وبعد شهر قالو لهو يا‬ ‫ענדו ⸱ ובעד שהר קאלו‬ ‫سييدنا‬ ‫להו יא סיידנא‬ ‫ نريدو نتووجهو الجل ال‬19 ‫دعووه تتمها في قال لهوم ال‬ ‫ראב لم‬ 20 no longer address me with ‫ عودتو تخاطبوني في شان‬20 regard to this matter. I will ‫ذالك ⸱ انه وقت ما نريد‬ go when I see fit.’ So, they ‫نتووجه فقعدو‬ stayed 19 we wish to go so that you can see to the appeal.’ The Rabbi replied, ‘You should ‫ נרידו נתווגַ֗ הו לאגַ֗ ל אל‬19 ‫דעווה תתמהא פי קאל‬ ‫להום אל ראב לם‬ ‫ עודתו תכַ֗ אטבוני פַ֗ י שאן‬20 ‫דַ֗ אלך ⸱ אנה וקת מא נריד‬ ‫נתווגַ֗ ה ַ֗פקעדו‬ 21 until the night of the eve of Passover. After the Rabbi did the Chametz check,13 they went ‫ لלילת ערב פסח من بعد ما‬21 ‫ ללילת ערב פסח מן בעד‬21 ‫מא עמל אל רב בדיקת חמץ عمل ال רב בדיקת חמץ‬ ‫واراحو‬ ‫ואראחו‬ 22 for recitation. They sat for four hours during the night until the recitation was finished. ‫ في قراييه قعدو الربع سعات‬22 ‫من ال ليل لحين ما فرغيت‬ ‫ال قراييه‬ ‫ ַ֗פי קראייה קעדו לארבע‬22 ‫סעאת מן אל ליל לחין מא‬ ‫ַ֗פרגִ ית אל קראייה‬ ‫ואחד‬: ‘one’ translated here as ‘a man’. ‫בע״ה‬: ‘with the help of God’ (‫)בעזרת השם‬. 13 Chametz refers to food that contains grains that have been mixed with water and left to rise. In Judaism, it is forbidden to eat any products containing chametz from the day before Passover until the end of Passover. During this period, only matzōt – unleavened bread – is consumed. On the day before Passover, all chametz food must be removed from the house; thus, the ‘chametz check’. 11 12 396 23 Then, they distributed the kaʿk14 with sesame oil. They gave two kaʿk to the Rabbi and two kaʿk to each of the ‫ في فرقو كعك بسيرج في‬23 ‫اعطو لل ראב كعكتين وال‬ ‫תלמידים كل‬ 24 students. Then, they set out to (tend to) their business. The Rabbi Abraham ibn ʿEzra went on his way, and the 25 students ‫ واحد كعكتين واتووجهو‬24 ‫ ואחד כעכתין ואתווגַ֗ הו‬24 ‫לחאלהום פַ֗ י אכַ֗ ד טריקו אל لحالكوم في اخد طريقو ال‬ ‫אב ןַ֗ עזרא وال‬ ַ֗ ‫רב‬ ‫אב ןַ֗ עזרא ואל‬ ַ֗ ‫רב‬ 140 v. 1 students accompanied him and they set off (together). Instead of going to the house (from which) the current 2 decree (came), they went off in the direction of the wilderness. The students exclaimed, ‘O our master, where 3 are we going? This (seems to us like) we are still in the wilderness!’ The Rabbi replied, ‘Stay 4 close to me.’15 So they stayed where they were. He called out a Name (but) not till dawn broke over them (did they realise that) 5 they were in the town from which the appeal came.16 The students saw the town ‫ ַ֗פי פרקו כעך בסירג ַ֗פי‬23 ‫אעטו לל ראב כעכתין ואל‬ ‫תלמידים כל‬ ‫ תלמידים‬25 ‫ תלמידים‬25 140 v. 140 v. ‫ תלמידים معو واتووجهو في‬1 ‫عوواض ما يتووجهو لل بيت‬ ‫حوكم ال‬ ‫ תלמידים מעו ואתווגַ֗ הו פַ֗ י‬1 ‫עוואץ מא יתווגַ֗ הו לל בית‬ ַ֗ ‫חוכם אל‬ ‫ جاري طلعو ناحييت ال خليه‬2 ‫ גַ֗ ארי טלעו נאחיית אל כַ֗ ליה‬2 ‫في قالو ال תלמידים يا‬ ‫ַ֗פי קאלו אל תלמידים יא‬ ‫سييدنا الين‬ ‫סיידנא לאין‬ ‫ متووجهين نحن دا احنا بقينا‬3 ‫فال خليه في قال لهوم ال‬ ‫ראב امسكو‬ ‫ מתווגַ֗ הין נחן דא אחנא‬3 ‫בקינא ַ֗פאל כַ֗ לֵיה פַ֗ י קאל‬ ‫להום אל ראב אמסכו‬ ‫ في طرفي في مسكو طرفو‬4 ‫טרפו וקרי‬ ַ֗ ‫ פַ֗ י טרפַ֗ י פַ֗ י מסכו‬4 ‫وقرى שם لم طلع عليهوم ال‬ ‫שם לם טלע עליהום אל‬ ‫فجر اال‬ ‫ַ֗פגַ֗ ר אלא‬ ‫ وهوم فال بلد بتاع ال دعووه‬5 ‫⸱ في نظرو ال תלמידים ال‬ ‫بلد‬ ‫ והום ַ֗פאל בלד בתאע אל‬5 ‫נטרו אל‬ ַ֗ ‫דעווה ⸱ ַ֗פי‬ ‫תלמידים אל בלד‬ ‫כעך‬: In Arabic, ‫ كعك \ كحك‬kaʿk/kaḥk ‘cookies of flour, butter, and sometimes a sweet filling or a dusting of sugar, baked for special occasions’; Hinds and Badawi, 1986, p. 737. In light of the context in which these ‘cookies’ are consumed in this tale, however, it is possible that kaʿk here refers not to celebratory cookies but to matzōt, the unleavened bread consumed during Passover (see n. 13). For this reason, I have merely transcribed the JA noun, rather than translating it into English. 15 ‫טרפי‬ ַ֗ ‫אמסכו ַ֗פי‬: literally reads ‘keep my place!’ 16 ‫פאל בלד בתאע אל דעווה‬: ַ֗ literally reads ‘in the town of the appeal’. 14 397 ‫‪ 6‬ואתעגַ֗ בו ⸱ ואחד יקול האדַ֗ י ‪ 6‬واتعجبو ⸱ واحد يقول هاذا ال‬ ‫بلدنا وواحد يقول لم هيا بلدنا‬ ‫אל בלדנא וואחד יקול לם‬ ‫היא בלדנא‬ ‫!‪6 and they were astonished‬‬ ‫‪One (of them) said, ‘This‬‬ ‫‪is our town!’ But the other‬‬ ‫‪exclaimed, ‘this is not our‬‬ ‫‪town,‬‬ ‫‪ 7‬נחן ַ֗פי מצר ואל רב אברהם ‪ 7‬نحن في مصر وال רב‬ ‫אברהם ماشي قودامهوم‬ ‫מאשי קודאמהום לחין מא‬ ‫لحين ما وصلو لبيت‬ ‫וצלו לבית‬ ‫’!‪7 we’re (still) in Cairo‬‬ ‫‪Rabbi Abraham walked in‬‬ ‫‪front of them until they ar‬‬‫‪rived at the house of the‬‬ ‫‪ 8‬אל חכַ֗ ם פַ֗ י טרקו עלא אל‬ ‫באב ⸱ ַ֗פי טלע אל חכַ֗ ם‬ ‫ַ֗פתח אל באב‬ ‫ואקפ‬ ‫ַ֗‬ ‫‪ 9‬וגַ֗ ד אל רב אברהם‬ ‫עלא אל באב פַ֗ י כַ֗ מנו אנו‬ ‫סַ אֵ ל ⸱ פַ֗ י קאל‬ ‫‪ 8‬ال חכַ֗ ם في طرقو عال ال‬ ‫باب ⸱ في طلع ال חכַ֗ ם فتح‬ ‫ال باب‬ ‫‪ 9‬وجد ال רב אברהם واقف‬ ‫عال ال باب في خمنو انو‬ ‫سال ⸱ في قال‬ ‫‪8 sage. They knocked on the‬‬ ‫‪door. The sage came down‬‬ ‫‪(and) opened the door‬‬ ‫‪9 and found Rabbi Abraham‬‬ ‫‪standing on his doorstep‬‬ ‫‪(and) he assumed that he‬‬ ‫‪was a beggar. So, the sage‬‬ ‫‪ 10‬להו אל חכַ֗ ם מא תריד לם ‪ 10‬لهو ال חכַ֗ ם ما تريد لم تعلم‬ ‫תעלם בחאלנא פַ֗ י קאל להו بحالنا في قال لهو ال רב‬ ‫אברהם‬ ‫אל רב אברהם‬ ‫‪10 said to him, ‘What do you‬‬ ‫‪want? Do you not know of‬‬ ‫‪our situation?’ Rabbi‬‬ ‫‪Abraham replied,‬‬ ‫‪ 11‬נעם עלם בדעוותך ולאכן ‪ 11‬نعم علم بدعووتك والكن‬ ‫اتكل عال هللا ס״ו وهوا يعمل‬ ‫אתכל עלא אללה ס״ו והוא‬ ‫لك נסים‬ ‫יעמל לך נסים‬ ‫‪11 ‘Yes, (I) know about your‬‬ ‫‪appeal but you should trust‬‬ ‫‪in God, He is exalted! He‬‬ ‫‪will make miracles for‬‬ ‫’!‪you‬‬ ‫‪12 The sage turned and found‬‬ ‫‪the two students whom he‬‬ ‫‪had sent to Cairo‬‬ ‫‪ 15‬قدم عليه واخدو ودخل بو ال‬ ‫بيت ⸱ في ما بعد ما صلو فال‬ ‫كنيس‬ ‫‪15 he introduced himself and‬‬ ‫‪they took (him) and he en‬‬‫‪tered the house with him.‬‬ ‫‪After they had prayed in‬‬ ‫‪the synagogue,‬‬ ‫‪ 12‬التفت ال חכַ֗ ם وجد ال انتين‬ ‫אלתפת אל חכַ֗ ם וגַ֗ ד אל‬ ‫ַ֗‬ ‫‪12‬‬ ‫ال תלמידים الذي كان‬ ‫אתנין אל תלמידים אלדַ֗ י‬ ‫ارسلهوم لمصر‬ ‫כאן ארסלהום למצר‬ ‫‪ 13‬לל רב אברהם ןַ֗ עזרא ⸱ פַ֗ י ‪ 13‬لل רב אברהם ןַ֗ עזרא ⸱ في ‪13 (in search of) Rabbi Abra-‬‬ ‫‪ham ibn ʿEzra. He greeted‬‬ ‫سلم عليهوم وقال لهوم احكو‬ ‫סלם עליהום וקאל להום‬ ‫‪them and said to them,‬‬ ‫لي ما جرا‬ ‫אחכו לי מא גַ֗ רא‬ ‫’!‪‘Tell me what happened‬‬ ‫‪ַ֗ 14‬פי קאלו להו יא סיידנא אל ‪ 14‬في قالو لهو يا سييدنا ال רב ‪14 They replied, ‘O our mas-‬‬ ‫‪ter, it is Rabbi Abraham‬‬ ‫אב’ ןַ֗ עזרא هوا الذي واقف‬ ‫רב אב’ ןַ֗ עזרא הוא אלדַ֗ י‬ ‫‪ibn ʿEzra who stands in‬‬ ‫قوصادك في‬ ‫ואקפ קוצאדך פַ֗ י‬ ‫ַ֗‬ ‫‪front of you!’ So,‬‬ ‫‪ 15‬קדם עליה ואכַ֗ דו ודכַ֗ ל בו‬ ‫אל בית ⸱ ַ֗פי מא בעד מא‬ ‫צלו ַ֗פאל כניס‬ ‫‪398‬‬ ‫‪ 16‬חכו אל תלמידים לל סי’ ‪ 16‬حكو ال תלמידים لل سي’‬ ‫חכַ֗ ם بالذي جرا وفرجو لهو‬ ‫ופרגו להו‬ ‫חכַ֗ ם באלדַ֗ י גַ֗ רא ַ֗‬ ‫ال كعك بسيرج‬ ‫אל כעך בסירג‬ ‫‪16 the students told the mas‬‬‫‪ter, (the) rabbi about what‬‬ ‫‪had happened and they‬‬ ‫‪showed him the kaʿk with‬‬ ‫‪sesame oil‬‬ ‫‪ 17‬אלדַ֗ י מעאהום ⸱ ואתעגַ֗ ב ‪ 17‬الذي معاهوم ⸱ واتعجب ال‬ ‫רב אבַ֗ ןַ֗‬ ‫ַ֗ח عال ذالك ⸱ وال ַ֗‬ ‫רב‬ ‫אל ח‪ ′‬עלא דַ֗ אלך ⸱ ואל ַ֗‬ ‫עזרא ע״ה‬ ‫אב ןַ֗ עזרא ע״ה‬ ‫ַ֗‬ ‫‪17 which (they had) with‬‬ ‫‪them. The rabbi was‬‬ ‫‪amazed at this. (Mean‬‬‫‪while,) Rabbi Abraham ibn‬‬ ‫‪ʿEzra, may God help him,‬‬ ‫‪18 spoke to the boy who was‬‬ ‫‪intended to be taken and‬‬ ‫‪made an offering to the‬‬ ‫‪idols, ‘When they come‬‬ ‫‪19 in the procession to take‬‬ ‫‪you, tell them that I will be‬‬ ‫‪with you. And when they‬‬ ‫‪say‬‬ ‫‪20 to you, “(Whatever) you‬‬ ‫‪wish for, you will be‬‬ ‫‪granted”, say to them,‬‬ ‫‪“what(ever) my companion‬‬ ‫‪desires.” ’ The boy replied,‬‬ ‫‪21 ‘I hear (and) am obedi‬‬‫‪ent.’ After only two hours,‬‬ ‫‪the uncircumcised came in‬‬ ‫‪a great procession‬‬ ‫‪22 in order to seize the boy‬‬ ‫‪(and) parade him through‬‬ ‫‪the town. The boy said to‬‬ ‫‪them,‬‬ ‫‪23 ‘Take me and my com‬‬‫‪panion with me and‬‬ ‫‪what(ever) is done to me,‬‬ ‫‪shall be done to‬‬ ‫‪24 my companion.’ The un‬‬‫‪circumcised replied, ‘We‬‬ ‫‪have one (already) but if‬‬ ‫‪you have given us two,‬‬ ‫‪ 18‬קאל לל ולד אלדַ֗ י נאוויין‬ ‫יאכַ֗ דוה יעמלוה קורבן לל‬ ‫ע״ז ⸱ חין מא יגַ֗ ו‬ ‫‪ 18‬قال لل ولد الذي ناوويين‬ ‫ياخدوه يعملوه קורבן لل ע״ז‬ ‫⸱ حين ما يجو‬ ‫‪ 19‬באל מוכב יאכַ֗ דוך קול‬ ‫להום אן נכון אנא מעאך ⸱‬ ‫וחין מא יקולו‬ ‫‪ 19‬بال موكب ياخدوك قول‬ ‫لهوم ان نكون انا معاك ⸱‬ ‫وحين ما يقولو‬ ‫‪ 20‬لك اتمنا توعطا قول لهوم‬ ‫‪ 20‬לך אתמנא תועטא קול‬ ‫الذي يتمنا رفيقي ⸱ في قال‬ ‫רפיקי ⸱‬ ‫להום אלדַ֗ י יתמנא ַ֗‬ ‫لهو ال ولد‬ ‫ַ֗פי קאל להו אל ולד‬ ‫‪ 21‬سميع موطيع ⸱ في بعد‬ ‫‪ 21‬סמיע מוטיע ⸱ ַ֗פי בעד‬ ‫ساعتين اال وجو ال ערלים‬ ‫סאעתין אלא וגַ֗ ו אל ערלים‬ ‫بموكب عظيم‬ ‫עטים‬ ‫במוכב ַ֗‬ ‫‪ 22‬לאגַ֗ ל אנהום יאכַ֗ דו אל ולד ‪ 22‬الجل انهوم ياخدو ال ولد‬ ‫يووكبو بو فال بلد ⸱ في قال‬ ‫יווכבו בו ַ֗פאל בלד ⸱ פַ֗ י‬ ‫لهوم ال‬ ‫קאל להום אל‬ ‫ורפיקי מעי ‪ 23‬ولد خدوني انا ورفيقي معي‬ ‫‪ 23‬ולד כַ֗ דוני אנא ַ֗‬ ‫ואלדַ֗ י יתאתא עלייה יתאתא والذي يتاتا علييه يتاتا عال‬ ‫עלא‬ ‫רפיקי ⸱ ַ֗פי קאלו אל‬ ‫‪ַ֗ 24‬‬ ‫ערלים אחנא לנא ואחד‬ ‫ואדַ֗ א כאן תעטונא אתנין‬ ‫‪ 24‬رفيقي ⸱ في قالو ال ערלים‬ ‫احنا لنا واخد واذا كان‬ ‫تعطونا اتنين‬ ‫‪141 r.‬‬ ‫‪141 r.‬‬ ‫‪ 1‬אחסן ואחסן ⸱ פַ֗ י רכב אל‬ ‫רב אב‪ ′‬ןַ֗ עזרא עא״ס הוא‬ ‫ואל ולד פַ֗ י קלב‬ ‫‪ 1‬احسن واحسن ⸱ في ركب ال‬ ‫אב ןַ֗ עזרא עא״ס هوا‬ ‫רב ַ֗‬ ‫وال ولد في قلب‬ ‫‪399‬‬ ‫‪141 r.‬‬ ‫‪1 so much the better!’ So,‬‬ ‫‪Rabbi Abraham ibn ʿEzra‬‬ ‫‪rode, him and the boy, in‬‬ ‫‪the centre of‬‬ ‫‪ 2‬אל תכַ֗ תרוואן וקאלו להום‬ ‫אל ערלים אתמנו תעטו ⸱‬ ‫ַ֗פי קאל להום אל רב‬ ‫‪ 2‬ال تخترووان وقالو لهوم ال‬ ‫ערלים اتمنو تعطو ⸱ في قال‬ ‫لهوم ال רב‬ ‫‪2 the sedan chair. The uncir‬‬‫‪cumcised said to them,‬‬ ‫‪‘(Whatever) you wish for,‬‬ ‫‪you will be granted.’ To‬‬ ‫‪which Rabbi‬‬ ‫‪ 3‬עב’ ןַ֗ עזרא אתמנא אן‬ ‫תגַ֗ עלו אל קסיס אל כביר‬ ‫ַ֗פי זכיבה ותרבטו‬ ‫‪ 3‬עב’ ןַ֗ עזרא اتمنا ان تجعلو‬ ‫ال قسيس ال كبير في زكيبه‬ ‫وتربطو‬ ‫‪3 Abraham ibn ʿEzra replied,‬‬ ‫‪‘I wish you to put the high‬‬ ‫‪priest into a large gunny‬‬ ‫‪sack and bind‬‬ ‫‪ 4‬فوم ال زكيبه فال تخترووان‬ ‫‪ 4‬פַ֗ ום אל זכיבה ַ֗פאל‬ ‫لحين ينتم ال موكب ⸱ في‬ ‫תכַ֗ תרוואן לחין ינתם אל‬ ‫قالو ال ערלים‬ ‫מוכב ⸱ ַ֗פי קאלו אל ערלים‬ ‫‪4 the opening of the sack to‬‬ ‫‪the sedan chair until the‬‬ ‫‪procession is over.’ The‬‬ ‫‪uncircumcised said,‬‬ ‫‪ 5‬סמיע מוטיע ַ֗פי אכַ֗ דו אל‬ ‫קסיס אל כביר וחטוה ַ֗פי‬ ‫זכיבה ורבטו‬ ‫‪ 5‬سميع وموطيع في اخدو ال‬ ‫قسيس ال كبير وحطوه في‬ ‫زكيبه وربطو‬ ‫’‪5 ‘We hear and are obedient.‬‬ ‫‪So, they took the high‬‬ ‫‪priest and lowered him into‬‬ ‫‪a sack and they tied‬‬ ‫‪ 6‬פַ֗ ום אל זכיבה ַ֗פאל תכַ֗ ת‬ ‫רוואן‪ 17‬ואנגַ֗ ר אל מוכב‬ ‫ולפַ֗ ו אל בלד ואראחו‬ ‫‪ 6‬فوم ال زكيبه فال تخترووان‬ ‫وانجر ال موكب ولفو ال بلد‬ ‫واراحو‬ ‫‪ 7‬אל כניסה ⸱ ַ֗פי סאלוהום‬ ‫איש תרידו תאכלו פַ֗ י טלב‬ ‫מנהום אל ראב‬ ‫‪ 7‬ال كنيسه ⸱ في سالوهوم ايش‬ ‫تريدو تاكلو في طلب منهوم‬ ‫ال ראב‬ ‫‪6 the opening of the sack to‬‬ ‫‪the sedan chair. The pro‬‬‫‪cession was swept along as‬‬ ‫‪they went around the town.‬‬ ‫‪Then, they went‬‬ ‫‪7 (to) the church. They asked‬‬ ‫‪them, ‘What do you want‬‬ ‫‪to eat?’ Rabbi Abraham‬‬ ‫‪ibn ʿEzra asked them‬‬ ‫‪8 for two chickens. So, they‬‬ ‫‪brought two chickens for‬‬ ‫‪them. Then Rabbi Abra‬‬‫‪ham went into the‬‬ ‫‪9 church and he summoned‬‬ ‫‪(the) large idol(s), which‬‬ ‫)‪they make the sacrifice(s‬‬ ‫‪to. He said, ‘O,‬‬ ‫‪10 bastard, get down from‬‬ ‫‪your place and sharpen‬‬ ‫‪th(is) knife!’ The idol got‬‬ ‫‪down and sat, sharpening‬‬ ‫‪11 the knife. All of the uncir‬‬‫‪cumcised were speechless‬‬ ‫‪and great fear descended‬‬ ‫‪upon them.‬‬ ‫‪ 8‬אברהם فرختين في جابو‬ ‫‪ 8‬אברהם ַ֗פרכַ֗ תין ַ֗פי גאבו‬ ‫لهوم فرختين ⸱ في طلع ال‬ ‫להום ַ֗פרכַ֗ תין ⸱ פַ֗ י טלע אל‬ ‫ראב אברהם لل‬ ‫ראב אברהם לל‬ ‫‪ 9‬כניסה ונדה על ע״ז אל‬ ‫כבירה אלדַ֗ י ביעמלו להא‬ ‫אל קורבאן וקאל יא‬ ‫‪ 9‬كنيسه ونده על ע״ז ال كبيره‬ ‫الذي بيعملو لها ال קורבאן‬ ‫وقال يا‬ ‫‪ 10‬ממזר אנזל מן מכאנך וסן‬ ‫אל סכינה ⸱ ַ֗פי נזל אל ע״ז‬ ‫וקעד יסן‬ ‫‪ 10‬ממזר انزل من مكانك وسن‬ ‫ال سكينه ⸱ في نزل ال ע״ז‬ ‫وقعد يسن‬ ‫‪ 11‬אל סכינה ַ֗פי אנבהתו גַ֗ מיע ‪ 11‬ال سكينه في انبهتو جميع ال‬ ‫ערלים ونزل عليهوم ال فزع‬ ‫אל ערלים ונזל עליהום אל‬ ‫ال‬ ‫ַ֗פזע אל‬ ‫‪: ‘sedan chair’ is written here in two parts, whereas elsewhere it reads as a single‬תכַ֗ ת רוואן‬ ‫‪).‬תכַ֗ תרוואן( ‪word‬‬ ‫‪17‬‬ ‫‪400‬‬ ‫עטים ⸱ פַ֗ י מא בעד נדה אל ‪ 12‬عظيم ⸱ في ما بعد نده ال רב‬ ‫‪ַ֗ 12‬‬ ‫אב’ על ממזרתه وقال لها‬ ‫רב אב’ על ממזרתה וקאל‬ ‫انزلي‬ ‫להא אנזלי‬ ‫‪ 13‬מן מכאניך ולעי אל נאר‬ ‫לאגַ֗ ל מא נטבוך אל ַ֗פראך‬ ‫ַ֗פי נזלית אל‬ ‫‪ 13‬من مكانيك ولعي ال نار‬ ‫الجل ما نطبوخ ال فراخ في‬ ‫نزليت ال‬ ‫‪12 Afterwards, Rabbi Abra‬‬‫‪ham ibn ʿEzra summoned‬‬ ‫‪his bitch, saying to her,‬‬ ‫‪‘Get down‬‬ ‫‪13 from your place and kin‬‬‫‪dle the fire so that we may‬‬ ‫‪cook the chickens!’ The‬‬ ‫‪bitch‬‬ ‫תנפוך אל ‪ 14‬ממזרתه وقعديت تنفوخ ال‬ ‫‪ 14‬ממזרתה וקעדית ַ֗‬ ‫نار ⸱ في حين ما شافو ال‬ ‫שאפו אל‬ ‫ַ֗‬ ‫נאר ⸱ פַ֗ י חין מא‬ ‫ערלים‬ ‫ערלים‬ ‫‪ 15‬كذالك وقعيت قلوبهوم‬ ‫‪ 15‬כדַ֗ אלך וקעית קלובהום‬ ‫وغشييت עיניهوم وقالو يا‬ ‫וגִ שיית עיניהום וקאלו יא‬ ‫سييد ال ياود ارفع‬ ‫ארפע‬ ‫סייד אל יאוד ַ֗‬ ‫‪15 this, she stilled their hearts‬‬ ‫‪and darkened their eyes.‬‬ ‫‪They exclaimed, ‘O, lord‬‬ ‫‪of the Jews, dispel‬‬ ‫‪ 16‬גִ ַ֗צבך עננא ורגע לנא אל ‪ 16‬غضبك عننا ورجع لنا ال‬ ‫اصنام اال مكانهوم وخود ال‬ ‫אצנאם אלא מכאנהום וכַ֗ וד‬ ‫ياودي معك‬ ‫אל יאודי מעך‬ ‫‪16 your anger towards us and‬‬ ‫‪return the idols to their‬‬ ‫‪places for us, and take the‬‬ ‫‪Jew with you‬‬ ‫‪ 17‬ואתווגַ֗ ה ונחן נכתב לכום‬ ‫ַ֗פרמאן אן מן אל יום לם‬ ‫עודנא נטלבו‬ ‫‪ 17‬واتووجه ونحن نكتب لكوم‬ ‫فرمان ان من ال يوم لم‬ ‫عودنا نطلبو‬ ‫‪17 and go! We will write an‬‬ ‫‪edict for you that from to‬‬‫‪day we will never again‬‬ ‫‪claim‬‬ ‫‪ 18‬ואחד כל סנה אבדן פַ֗ י אל‬ ‫חין כתבו להום ַ֗פרמאן‬ ‫וכַ֗ תמו ואעטו‬ ‫‪ 18‬واحد كل سنه ابدن في ال‬ ‫حين كتبو لهوم فرمان وختمو‬ ‫واعطو‬ ‫‪14 descended and sat, blow‬‬‫‪ing (on) the fire. When the‬‬ ‫‪uncircumcised saw‬‬ ‫‪18 one (of your people), each‬‬ ‫‪year.’ Then, they wrote an‬‬ ‫)‪edict for them, signed (it‬‬ ‫)‪and gave (it‬‬ ‫‪ 19‬لهوم واخدو طريقهوم‬ ‫‪ 19‬להום ואכַ֗ דו טריקהום‬ ‫واتووجهو لمنزلهوم وعملو‬ ‫ואתווגהו למנזלהום ועמלו‬ ‫عيد פסח لم لهو‬ ‫עיד פסח לם להו‬ ‫‪19 to them. Then they went‬‬ ‫‪on their way, setting off for‬‬ ‫‪their home (where) they‬‬ ‫‪made a Passover festival,‬‬ ‫‪the like of which‬‬ ‫‪ 20‬نظير ابدن واتووجه ال ראב‬ ‫נטיר אבדן ואתווגה אל‬ ‫‪ַ֗ 20‬‬ ‫ראב לבלדו זכַ֗ ותו יגִ ן עלינו لبلدو זכַ֗ ותו יגִ ן עלינו אמן‬ ‫כי״ר ∵‬ ‫אמן כי״ר‪∵ 18‬‬ ‫‪20 has never been seen.‬‬ ‫‪Then, Rabbi Abraham ibn‬‬ ‫‪ʿEzra returned to his town.‬‬ ‫!‪May His virtue protect us‬‬ ‫!‪Amen. His will be done‬‬ ‫‪).‬כן יהי רצון( ’!‪: ‘His will be done‬כי״ר‬ ‫‪401‬‬ ‫‪18‬‬ 3 Grammatical notes All examples from the text referred to in this section are examined in relation to their Classical Arabic (CA) and, where applicable, their Modern Cairene Arabic (MCA) forms. 3.1 3.1.1 Orthography and phonology The diacritics 3.1.1.1 Peh for fāʾ In BnF Hébreu 583, ninety-five per cent of occurrences of peh denoting fāʾ are marked with a supra-linear dot. Unlike its Hebrew counterpart,19 the Arabic grapheme fāʾ has only one reflex: a voiceless labio-dental fricative /f/. Moreover, peh is only employed to denote fāʾ and no other Arabic graphemes, in all periods of JA writing. Therefore, the phonetic value so often ascribed to the diacritic – either to differentiate between allophones of a phoneme, as in the case of gimel for ğīm, or to distinguish between independent graphemes, as in the case of dalet for dāl and ḏāl – is void in this context. The function of the diacritic above peh appears to be stylistic rather than phonetic:20 it may result from the imitation of the Arabic grapheme fāʾ (‫)ف‬. The emulation of Arabic fāʾ is also evident in the writing of peh in word-final position, where it is found in initial/medial form (‫)פ‬, ַ֗ which closely resembles the final form of its Arabic cognate (‫ ;)ف‬for example, ‫ואקפ‬ ַ֗ ‘standing’ (140 v. 9, 14; CA: ‫واقف‬ wāqifun); ‫נצרופ‬ ַ֗ ‘we will pay’ (140 r. 9; CA: ‫ نصرف‬naṣrifu). 3.1.1.2 Gimel for ğīm and ġayn The marking of gimel for ğīm with a sub- or supra-linear dot emerged as a common orthographic practice during the period of classical JA (c10th–15th centuries). The use of the diacritic to denote ğīm has been interpreted by scholars of JA as indicative of the fronted pronunciation of ğīm. The perceived phonetic value of the diacritic was compounded by the observation that this orthographic practice prevailed between the 12th and 17th centuries,21 after 19 In Hebrew, peh has a plosive and a fricative allophone. The latter occurs when preceded by a vowel. 20 It has been suggested that the use of the diacritic above peh for fāʾ may be attributed to the continuation of the raphe, a supra-linear dash introduced in Hebrew by the Tiberian Masoretes to differentiate fricative (with raphe) from plosive (with dageš lene) allophones of the six bgdkft letters. However, the inclusion of the diacritic above peh for fāʾ is not evident in JA manuscripts until the 18th century. Why, when the script used in JA was the Hebrew script, would the raphe have not been included above peh for fāʾ in JA texts from the 10th century onwards? The form of the diacritic is also worth considering in this context; in JA texts, the diacritic found above peh for fāʾ is most commonly a dot (in keeping with the Arabic practice), rather than a dash (as is the case with the raphe). 21 Blanc, 1981, p. 191. 402 which gimel for ğīm was increasingly left unadorned.22 By the early 19th century, the diacritic is said to be omitted entirely from gimel denoting ğīm, a phenomenon that is generally understood to reflect the velar plosive pronunciation characteristic of MCA.23 The use of the diacritic in this 19th century folk narrative challenges this assumption: a supra-linear dot appears above gimel for ğīm in 80.5 per cent of occurrences of the grapheme, e.g. ‫‘ לאגַ֗ ל‬in order to, so that’ (139 v. 21, 22; 140 r. 19; 140 v. 22; 141 r. 13; CA: li-ʾağli); ‫‘ גַ֗ וואב‬a letter’ (140 r. 8, 13, 16; CA: ğawābun); ‫‘ ואתעגַ֗ בו‬and they were astonished!’ (140 v. 6; CA: taʿağğaba). Much like the graphical imitation of peh, however, the supra-linear diacritic appears predominantly to be a stylistic device, used in imitation of the graphical form of its Arabic cognate (‫)ج‬, rather than a marker of fronted pronunciation.24 A further corroboration of this interpretation is found in the application of the diacritic to gimel in the Hebrew noun ‫‘ אל גַ֗ ורל‬the lot’ (140 r. 2, 5), where it is preceded by the JA definite article. Gimel is one of six Hebrew plosives (the others being bet ‫ב‬, dalet ‫ד‬, kaf ‫כ‬, peh ‫ פ‬and tav ‫ )ת‬which are fricativised post-vocalically. In this context, however, gimel follows the consonant lamed /l/, and consequently the expected realisation of gimel would be plosive [g]. The Arabic grapheme ġayn, also denoted with the Hebrew grapheme gimel, is consistently marked with a sub-linear diacritic in this folk-tale, e.g. ‫‘ ַ֗פרגִ ית‬it was finished’ (140 r. 22); ‫‘ וגִ שיית‬and she covered’ (141 r. 15). 3.1.1.3 Dalet for ḏāl The Hebrew grapheme dalet is used to denote both the Arabic graphemes dāl and ḏāl. The latter is sometimes differentiated from the former by the use of a supra-linear diacritic in late JA literary and documentary texts. In the narrative BnF Hébreu 583, the diacritic only appears above dalet for ḏāl in CA forms,25 e.g. ‫‘ האדַ֗ א‬this’ (140 r. 8; CA: hāḏā); ‫‘ אלדַ֗ י‬that, which’ (139 v. 22; 140 r. 1, 4, 6; 140 v. 12, 14, 16, 17, 18, 20, 23; 141 r. 9; CA: ʾallaḏī; MCA: ʾillī), and is omitted in more quotidian words, e.g. ‫‘ אכַ֗ דהום‬he took them’ (140 r. 18; CA: A diachronic analysis of the use of the diacritic in relation to gimel for ğīm in letters and folk narratives spanning from the 9th to the 19th centuries suggests that this practice was less uniform than has been suggested in previous scholarship. The diacritic is found to be an unreliable source from which to reconstruct the historical development of the phonetic realisations of ğīm. See Connolly, in press. 23 Blanc, 1981, pp. 189, 191; Hary, 1996, p. 154. 24 This is not to rule out the possibility of the diacritic’s phonetic value, but to emphasise its stylistic function, which has heretofore been overlooked; see Connolly, in press for a more detailed discussion of the use of the diacritic in relation to gimel for ğīm. 25 The practice of writing a diacritic above dalet for ḏāl in words of CA origin is noted in Khan, 1992, p. 238. In my analysis of letters and folk narratives from the 15th–19th centuries, it has become apparent that the inclusion of a diacritic above dalet denoting ḏāl, as with the writing of peh for fāʾ, is absent in 15th–16th century material and emerges only in the 17th(?)–18th century letters and folk-tales. Furthermore, the custom observed by Khan and evident in this 19th century folk-tale of including the diacritic above dalet for ḏāl only in words of CA origin is also consistently found in contemporaneous letters and folk-tales. 22 403 ʾaḫaḏa). This appears to reflect the phonetic shift of ḏāl in colloquial forms of Arabic from a voiced dental fricative /ð/ to a voiced alveolar plosive /d/. However, it must not be presumed that the inclusion of the diacritic above dalet for ḏāl in CA forms is indicative of the voiced dental fricative /ð/ reflex. In the late JA manuscript CUL T-S Ar. 54.63, a fully vocalised folk narrative entitled Qiṣṣat Ḥanna, the JA relative pronoun ‫ אלדי‬ʾldy ‘which’ is vocalised as ‫ אֵ לַדִ י‬ʾeladdī (CUL T-S Ar. 54.63, 1v.; CA: ʾallaḏī; MCA: ʾillī).26 The Hebrew pointing of the JA relative pronoun suggests that the fricative realisation of CA ḏāl was not necessarily retained in the reading tradition, and that the diacritic may indeed be a stylistic device, employed to evoke the graphical form of its Arabic counterpart ḏāl (‫)ذ‬. 3.1.1.4 Kaf for ḫāʾ and kāf A diacritic appears above kaf for ḫāʾ in initial/medial form throughout the text, e.g. ‫לאפו‬ ַ֗ ַ֗‫‘ כ‬besides him’ (140 r. 3; CA: ḫilāfu-hu); ‫‘ ַ֗פרכַ֗ תין‬two chickens’ (141 r. 8; CA: farḫatayn). In final form the diacritic is omitted, e.g. ‫‘ נטבוך אל ַ֗פראך‬we may cook the chickens’ (141 r. 13; CA: naṭbuḫu al-firāḫa); ‫תנפוך‬ ַ֗ ‘blowing’ (141 r. 14; CA: tanaffuḫun).27 The Arabic grapheme kāf, also denoted with the Hebrew grapheme kaf, is unmarked in its initial/medial form – e.g. ‫‘ מוכב‬a procession’ (140 r. 1; 140 v. 19, 21; 141 r. 4; CA: maukibun); ‫‘ כניסה‬a church’ (141 r. 7, 9; CA: kanīsatun) – but contains a central dot in word-final position, e.g. ‫‘ דַ֗ אלך‬that’ (140 r. 2, 3, 20; 140 v. 17; CA: ḏālika); ‫‘ כעך‬cookie’ (140 r. 23; 140 v. 16; MCA: kaʿk). The distinctions drawn between these two graphemes through the different uses of the diacritic mirror the graphical forms of their Arabic counterparts. Kaf for ḫāʾ ( ַ֗‫ )כ‬with its supra-linear dot imitates the graphical form of ḫāʾ (‫)خ‬, whereas kaf for kāf (‫כ‬/‫)ך‬, which is unmarked in initial/medial form but takes a central marker in word-final position, evokes the graphical forms of Arabic kāf (‫كـ‬/‫)ك‬. 3.1.1.5 Ṣadeh for ḍād The Arabic grapheme ḍād is consistently represented with a ṣadeh and supralinear dot, in keeping with the orthographic practice established during the classical JA period, e.g. ‫‘ גִ ַ֗צבך‬your anger’ (141 r. 16; CA: ġaḍab). 26 This example comes from Khan, 2010, p. 213. As regards the use of the diacritic above the kaf in the commonly occurring noun ‫חכם‬ ַ֗ ‫‘ אל‬the sage; rabbi’, two interpretations present themselves: (i) this may be a Hebrew loan word which has an established use in colloquial Arabic (‫ حاخام‬ḥāḫām ‘rabbi’) and is thus considered to be a JA word, in which kaf denotes ḫāʾ; or (ii) the JA graphical imitation of the Arabic ḫāʾ (i.e. ַ֗‫כ‬ for ‫ )خ‬is used to denote the Hebrew kaf, which is fricativised here due to its post-vocalic position. The latter interpretation seems more probable in light of the re-occurrence of kaf with a diacritic in a Hebrew word, ‫זכותו‬ ַ֗ ‘his good deed’ (141 r. 20) in which kaf is also fricativised. 27 404 3.1.1.6 Tet for ẓāʾ As with ṣadeh for ḍād, the orthographic representation of ẓāʾ with tet and a supra-linear diacritic (‫)ט‬, ַ֗ initiated during the classical JA period, is continued without exception in this late JA text, e.g. ‫עטים‬ ַ֗ ‘great’ (140 r. 1; 140 v. 21; ַ֗ ‘equal’ (140 r. 4; 141 r. 20; CA: naẓīrun). The 141 r. 12; CA: ʿaẓīmun); ‫נטיר‬ representation of both ḍād and ẓāʾ in classical JA and this late JA text is driven by the imitation of the graphical forms of their Arabic orthographic cognates (‫ ض‬and ‫ ظ‬respectively). The determining factor in the use of gimel for ġayn, dalet for ḏāl and kaf for ḫāʾ with a diacritic is commonly attributed to each of these grapheme’s phonetic congruence with the fricative Hebrew allophones of gimel [ɣ], dalet [ð] and kaf [χ], respectively.28 While the present study does not refute the fundamental essence of this interpretation, it argues that the inclusion of the diacritic was influenced not by the Hebrew phonemes but – as with ṣadeh for ḍād and tet for ẓāʾ – by the graphical form of their respective Arabic cognates. The re-evaluation of the diacritic’s functions as primarily practical and stylistic rather than phonetic is founded on the inclusion of a supra-linear dot above peh for fāʾ in this and other late JA texts in which the diacritic serves no perceptible phonetic function, on the erratic use of the diacritic in relation to gimel for ğīm29 and on the relatively recent consistent inclusion of the diacritic above peh for fāʾ, dalet for ḏāl and kaf for ḫāʾ. While the continuation of classical JA spelling practices in late JA texts is generally acknowledged, it has often been suggested that Arabic itself has little or no influence on late JA.30 Here we find evidence to suggest that written Arabic practices played a more significant role in the JA orthographic innovations of the Ottoman period than has previously been suggested. 3.1.2 Doubling of vav and yod The graphemes vav and yod are often written twice to indicate the Arabic consonants waw and yāʾ, differentiating them from the long vowels /ū/ and /ī/ and scriptio plena of short vowels /u/ and /i/. This practice, common in late JA texts, was influenced by Rabbinic Hebrew orthographic practices. In initial position, consonantal waw is always represented with a single vav in this text, e.g. ‫‘ ואחד‬one’ (139 v. 20; 140 r. 14, 24; 140 v. 6, 6, 24; 141 r. 18; CA: wāḥid); ‫‘ וגַ֗ ד‬he found’ (140 v. 9, 12; CA: wağada). However, in medial form consonantal waw is routinely denoted with double vav, e.g. ‫‘ אוולאד‬children’ (139 v. 21; CA: ʾawlādun); ‫עוואץ מא‬ ַ֗ ‘instead’ (140 v. 1; CA: ʿiwāḍa; MCA: ʿuwāḍ mā). To a lesser extent, medial consonantal yāʾ is also denoted with a double yod, e.g. ‫‘ דונייא‬world’ (140 r. 4; CA: dunyā). 28 Blau, 1981, pp. 34–35; Khan, 2016, pp. 24–25. See Connolly, in press. 30 Khan, 2007, pp. 527–528; Wagner, 2010, pp. 233–234. 29 405 3.1.3 Scriptio plena The Hebrew graphemes ʾalef, yod and vav are used with varying degrees of frequency in the text to express the Arabic short vowels fatḥa /a/, kasra /i/ and ḍamma /u/, respectively. 3.1.3.1 Scriptio plena of CA short /a/ The denoting of the short vowel /a/ with ʾalef is relatively rare in this text. It occurs with the independent preposition ‫‘ מע‬with’ (CA: maʿa) three times when a suffix pronoun is attached to it, e.g. ‫‘ מעאהום‬with them’ (140 v. 17; CA: maʿa-hum); ‫‘ מעאך‬with you (sg)’ (140 v. 19; CA: maʿa-ka).31 This reflects the MCA practice in which the second short vowel /a/ of the preposition is lengthened to /ā/ with the addition of a pronoun suffix (MCA: maʿā).32 3.1.3.2 Scriptio plena of short /i/ The use of yod to indicate the Arabic sub-linear vowel kasra occurs frequently in the 3fsg suffix conjugation, representing the vowel shift /a/ > /e/ in the 3fsg verbal suffix conjugation which is characteristic of MCA, e.g. ‫‘ ַ֗פרגִ ית‬it was empty’ (140 r. 22; CA: faraġat); ‫‘ וקעדית‬and she sat’ (141 r. 14; CA: qaʿadat). It is also present before the 2fsg suffix pronoun, again indicating MCA pronunciation (see section 3.2.6.2.1), e.g. ‫‘ מכאניך‬your (f.) place’ (141 r. 13; CA: makānu-ki). Plene yod also denotes the short vowel kasra in the noun of ‫ומן‬ ‫‘ גיהת‬on behalf of’ (140 r. 9; CA: ğiha). The CA relative pronoun man ‘the one who’ is written with yod in this JA text, e.g. ‫‘ מין‬the one who’ (139 v. 22), indicating the MCA pronunciation.33 3.1.3.3 Scriptio plena of short /u/ The most common form of matres lectionis is the use of vav to represent the short vowel ḍamma,34 e.g. ‫‘ ואל חוזן‬and grieving’ (140 r. 5; CA: al-ḥuzn); ‫עומרו‬ ‘his age’ (140 r. 3; CA: ʿumru-hu); ‫‘ קודאמהום‬in front of them’ (140 v. 7; CA: quddāma); ‫‘ ַ֗פום‬mouth’ (141 r. 4, 6; CA: famun/fum, construct state; MCA: fumm). In some instances, the plene spelling of the short vowel reveals a shift in pronunciation. In the following example, the vowel shift /a/ > /u/ reflects a change from CA to MCA pronunciation: ‫‘ אושהור‬months’ (140 r. 11, 11, 12, 14; CA: ʾašhur; MCA: ʾušhur). Scriptio plena vav is employed in the 3mpl independent (CA: hum) and 3mpl and 2mpl suffix pronouns (CA: -hum/-him, -kum) consistently throughout the text, irrespective of the preceding vowels, e.g. ‫‘ עליהום‬on them’ (140 r. 31 The independent preposition ‫‘ מע‬with’ (CA: maʿa) occurs as frequently without the MCAinfluenced ʾalef as with it, in this text. 32 Hinds and Badawi, 1986, p. 828. 33 Hinds and Badawi, 1986, p. 842. 34 Plene spelling of the short /u/ vowel is also very common in JA letters of all periods; see Wagner, 2010, p. 53. 406 9; 140 v. 4, 13; 141 r. 11; CA: ʿalay-him); ‫‘ מנזלכום‬your home’ (140 r. 9; CA: manzilu-kum). The 3msg pronoun suffix is also written with a vav representing the short vowel ḍamma when suffixed to the preposition l- (CA: li-), e.g. ‫‘ להו‬to him’ (140 r. 4, 8, 13, 16, 18; CA: la-hu). This may reflect the pronunciation practice (which is not represented in CA orthography) in which the 3msg pronoun suffix is long (/-hū/, /-hī/) after an open syllable.35 3.1.4 Fy for faThe CA bound particle fa- functions as a connective particle between two clauses, denoting either a temporal or consequential connection.36 The CA bound particle is written in this JA text as an independent entity,37 constituted of peh and yod, e.g. ‫(‘ ַ֗פי כתב‬then) he wrote’ (140 r. 13; CA: fa-kataba); ‫ַ֗פי סלם‬ ‫(‘ עליהום‬then) he greeted them’ (140 v. 13; CA: fa-salima).38 The classical JA representative functions of yod – in which yod denotes its orthographic cognates yāʾ /ī/ and /y/ and also ʾalif maqṣūra /ā/ – diminish in late JA compositions to denote only yāʾ /ī/and /y/, leaving ʾalif maqṣūra to be represented with ʾalef or hey, e.g. ‫עלה‬/‫‘ עלא‬on, upon’ (classical JA: ‫ ;עלי‬CA: ʿalā). However, in the writing of the conjunction fa-, the short vowel /a/ is denoted in plene with yod. This phenomenon is indicative of ʾimāla – the raising of the long or short vowel /a/ to /e/ or /ɛ/ – in which yod, representing the plene spelling of the short /e/ vowel, denotes a shift in pronunciation.39 This interpretation is corroborated by the Hebrew vocalisation in the late JA folk narrative CUL T-S Ar. 54.63, in which the bound particle fa- is vocalised with the Hebrew vowel ḫireq /i/: ‫‘ פִ צַ אר‬and he came’ fi-ṣār (CUL T-S Ar. 54.63, 4r.; CA: fa-ṣāra).40 3.1.5 ʾAlif maqṣūra and ʾalif mamdūda ʾAlef is routinely deployed to denote ʾalif maqṣūra bi-ṣūrat al-yāʾ in this JA text, e.g. ‫‘ עלא‬on, about, concerning’ (139 v. 21; 140 r. 2, 5; 140 v. 8, 9, 11, 17, 23; CA:ʿalā); ‫‘ אלא‬to, towards’ (140 r. 18; 141 r. 16; CA: ʾilā).41 In contemporaneous letters, it is common for final ʾalif, ʾalif maqṣūra and ʾalif mamdūda to be denoted with either ʾalef or hey in a phenomenon attributed to Hebrew and Aramaic influence. 35 Fischer, 2002, p. 142, §268. Wright, 1967, vol. 1, pp. 290–291, §366. 37 Thus, CA fa- ‘so, thus’ becomes indistinguishable from the independent preposition fī ‘in, during’ (JA: ‫ ;פי‬CA: fī). Differentiation between the two entities depends, therefore, on the surrounding context. 38 There is only one exception to this phenomenon, where the particle fa- appears in accordance with CA convention, that is, without the yod and conjoined to the following verb: ‫‘ ַ֗פקעדו‬So, they sat …’ (140 r. 20; CA: fa-qaʿada). 39 See Wagner, 2010, p. 63, §4.5.6 for examples of this representation of fa- in late JA letters; see Blau, 1981, p. 74 for yod and ʾalef denoting yāʾ. 40 Khan, 2010, p. 213. 41 When the 1csg pronoun suffix is attached to the preposition ʿalā, it is denoted with a yod and hey, e.g. ‫‘ עלייה‬on me’ (140 v. 23; CA: ʿalayya). 36 407 The ʾalif mamdūda ending is also represented with ʾalef in this text, e.g. ‫‘ קאמו באל בוכא‬they started weeping’ (140 r. 5; CA: al-bukāʾu). As such, ʾalif mamdūda becomes indistinguishable from ʾalif maqṣūra bi-ṣūrat al-yāʾ.42 While there are discernible elements of the external influences of Hebrew and Aramaic, it is possible that the main motivation for the use of ʾalef for ʾalif maqṣūra bi-ṣūrat al-yāʾ is phonetic. Unlike many contemporaneous texts, hey does not represent ʾalif in this folk narrative (except in the case of tanwīn ʾalif, see section 3.1.7). 3.1.6 Dagger ʾalif Occurrences of CA dagger ʾalif – a supra-linear ‘dagger’ ( ٰ ) placed above a consonantal grapheme, taking the place of the plene ʾalif 43 – in this JA folk narrative are written in plene withʾalef,44 e.g. ‫‘ דַ֗ אלך‬that’ (140 r. 2, 3, 20; 140 v. 17; CA: ḏālika); ‫‘ האדַ֗ א‬this’ (140 r. 8; CA: hāḏā); ‫‘ ולאכן‬but’ (140 v. 11; CA: wa-lākin). 3.1.7 Tanwīn The CA tanwīn ʾalif indefinite adverbial marker is expressed with both the Hebrew graphemes hey and nun in this late JA text, e.g. ‫איצה‬ ַ֗ ‘also’ (139 v. 19; CA: ʾayḍan); ‫‘ אבדן‬never, ever’ (141 r. 18, 20; CA: ʾabadan). 3.1.8 Separation of units 3.1.8.1 The definite article The JA definite article is written consistently throughout the text separately from the noun it modifies, e.g. ‫‘ אל קסיס‬the priest’ (141 r. 3; CA: al-qissīsu); ‫‘ אל סכינה‬the knife’ (141 r. 10, 11; CA: al-sikkīnatu); ‫‘ אל נאר‬the fire’ (141 r. 13; CA: al-nāru). This is a well-established phenomenon in both literary and documentary late JA texts.45 3.1.8.2 Fy + definite article When preceding the definite article, the independent preposition ‫ ַ֗פי‬fy ‘in, during’ (CA: fī) loses the yod and becomes prefixed to the ʾalef-lamed ligature, e.g. ‫‘ פאל בלד‬in the town’ (140 v. 5, 22; CA: fī al-baladi); ‫‘ פאל תכַ֗ תרוואן‬in the sedan chair’ (141 r. 4, 6; CA: fī al-taḫtarawāni).46 42 Blau, 1981, p. 74. Wright, 1967, vol. 1, pp. 9–10, §6a. 44 The numerical form ṯalāṯ ‘three’ appears three times in this JA folk-tale written as ‫‘ תלת‬three’ (140 r. 11, 11, 14). In CA, ‘three’ may be written either with an ʾalif or dagger ʾalif (CA: ṯalāṯ). Its form here may either be regarded as a defective spelling in which the ʾalif has been omitted, or as an imitation of the CA form in which the dagger ʾalif would have been included. 45 Khan, 1992, p. 231; Khan, 2006, p. 51; Khan, 2013, p. 243. 46 The assimilation of the independent pronoun fy and the definite article is recorded as occurring infrequently in some 11th century Maġribi and 15th–16th century letters; see Wagner, 2010, p. 66. The appearance of this phenomenon in BnF Hébreu 583 is very consistent. 43 408 3.2 3.2.1 Morphology Suffix conjugation 3.2.1.1 Otiose ʾalif In CA, the 3mpl suffix conjugation, the 3mpl and 2mpl subjunctive and jussive forms of the prefix conjugation, and the mpl imperative form are written with an otiose (also referred to as ‘silent’) ʾalif, e.g. ‫ َكتَبُوا‬katabū ‘they wrote’; ‫ يَ ْكتُبُوا‬yaktubū ‘they (should) write’.47 In a phenomenon common to classical JA documentary48 and late JA texts alike, the otiose ʾalif is omitted in all 3mpl suffix conjugations, e.g. ‫‘ אכַ֗ ברו‬they related’ (139 v. 19; CA: ʾaḫbarū); ‫עמלו‬ ‘they made’ (140 r. 2; 141 r. 19; CA: ʿamilū). 3.2.1.2 Prosthetic ʾalif The fifth form of the suffix conjugation is prefixed with an additional ʾalif as is customary in MCA, e.g. ‫‘ ואתווגַ֗ הו‬and they set off’ (140 r. 13, 24; 140 v. 1; CA: tawağğaha; MCA: ʾitwaggah); ‫‘ ואתעגַ֗ בו‬and they were astonished!’ (140 v. 6; CA: taʿağğaba; MCA: ʾitʿaggib).49 3.2.2 Prefix conjugation 3.2.2.1 Niktib–niktibū This literary text contains instances of both n-type and a-type verbal inflection. Occurrences of 1csg n-type verbal inflection may be seen in examples (1) and (2), while 1cpl n-type verbal forms are seen in examples (3) and (4): (1) ‫נכון אנא מעאך‬ ‘I will be with you’ (140 v. 19) (2) ‫אנה וקת מא נריד נתווגָ֗ ה‬ ‘When I want to, I will go’ (140 r. 20) (3) ‫ובעד שהר קאלו להו יא סיידנא נרידו נתווגָ֗ הו‬ ‘After a month, they said to him, “O our master, we wish to go …” ’ (140 r. 19) (4) ‫לם עודנא נטלבו ואחד כל סנה אבדן‬ ‘We will never again claim one (of you) each year’ (141 r. 17) 47 Examples from Fischer, 2002, p. 7, §7.2. Wagner, 2010, p. 47. 49 Hinds and Badawi, 1986, pp. 563, 925. 48 409 Examples (5)–(7) constitute instances of a-type verbal inflection as they occur in the text. In examples (5) and (6), 1cpl form appears without the plural suffix vav in adherence with CA convention. In example (7) we find the sole instance of 1csg a-type verbal inflection: (5) ‫נצרופ‬ ָ֗ ‫נחן‬ ‘We will support …’ (140 r. 9) (6) ‫ונחן נכתב לכום ַ֗פרמאן‬ ‘And we will write a decree for you’ (141 r. 17) (7) ‫אתמנא אן תגַ֗ עלו אל קסיס אל כביר פַ֗ י זכיבה‬ ‘I wish you to put the high priest into a gunny sack …’ (141 r. 3) As is evident from examples (1), (5) and (6), 1csg n-type and 1cpl a-type verbs are often accompanied by an independent pronoun.50 The inclusion of the independent pronoun – either post-positionally, as in example (1), or pre-positionally, as in examples (5) and (6) – is borne of a desire to avoid the semantic confusion which may arise from using 1csg n-type and 1cpl a-type verbal forms (which are identical in form) alongside one another in the same text. As Khan remarks, the co-existence of n-type and a-type verbal inflection in JA written texts, as seen here, also reflects the concurrent uses of these verbal forms in Jewish spoken varieties of Arabic.51 3.2.2.2 Bi-imperfect There is only one instance of the bi-imperfect in this late JA folk-tale. It is used to express habitual action: (8) ‫על ע״ז אל כבירה אלדַ֗ י ביעמלו להא אל קורבאן‬ ‘And he summoned (the) large idol which they make the offering(s) to …’ (141 r. 9) 3.2.3 Ğāʾa + biThe MCA form gāb ‘to bring’ occurs in the text, once. This phonogenetic verbal form, comprising the CA verb ğāʾa ‘to come’ + the bound preposition bi- ‘with, in’, is common in JA documentary texts of all periods:52 (9) 50 ‫ַ֗פי גאבו להום ַ֗פרכַ֗ תין‬ ‘So, they brought them two chickens’ (141 r. 8) I am indebted to Dr Nadia Vidro for this observation. Khan, 2006, p. 58; cf. Blanc, 1974, pp. 209–212; Blau, 1981, pp. 58–60. 52 Wagner, 2010, p. 108; Esseesy, 2010, p. 225. 51 410 3.2.4 Vowel shifts This text reveals two explicit deviations from CA and MCA verbal vocalisation, made evident by the plene spelling of one or both short vowels: ‫נצרופ‬ ַ֗ ‫נחן‬ ‫‘ עליהום‬we will pay them’ (140 r. 9; CA: naṣrifu; MCA: ṣaraf, yiṣrif); and ‫חוצרו‬ ַ֗ ‘they arrived’ (literally, ‘they were present’; 140 r. 14; CA: ḥaḍara; MCA: ḥaḍar, yaḥḍar). The plene spelling of short vowels is prevalent in late JA literary and documentary texts (see section 3.1.3).53 Yet the vowel changes they indicate – such as /a/ or /i/ to /u/ found here – do not always reflect the situation in MCA. Rather, they may be regarded as features unique to Egyptian Arabicspeaking Jews.54 3.2.5 Negation 3.2.5.1 Verbal negation The CA particle lam is used for all forms of verbal negation, irrespective of tense and form. In the following examples, for instance, the particle negates a suffix conjugation form, in (10), and non-jussive forms of the prefix conjugation, in (11) and (12), all of which connote the future. (10) ‫ַ֗פי קאל להום אל ראב לם עודתו תכַ֗ אטבוני ַ֗פי שאן דַ֗ אלך‬ ‘Then the Rabbi said to them, “You will no longer address me about this matter!” ’ (140 r. 19-20) (11) ‫ובשרט לם תתעווקו ַ֗פאל טריק‬ ‘… and on the condition that you do not tarry on the way’ (140 r. 10) (12) ‫ולם יכון אלא כיר‬ ‘… and all will be well’ (140 r. 17; CA: lam yakun) 3.2.5.2 Nominal negation The use of the particle lam as a verbal negative marker of the jussive to indicate the past tense extends well beyond its limited capacity in CA in this late JA text. As well as being used with all verbal forms (see section 3.2.5.1), it serves as the negative marker of nominal sentences: (13) ‫לם היא בלדנא‬ ‘It is not our town!’ (140 v. 6) (14) ‫נטיר אבדן‬ ַ֗ ‫ועמלו עיד פסח לם להו‬ ‘They made a Passover festival, the like of which has never been seen’ (141 r. 19-20) 53 54 See also Hary, 1997, pp. 212–216. Hary, 2009, p. 101. 411 The varied functions and frequency of the negative particle lam evident in this text are found in JA and Middle Arabic texts alike.55 Lam appears to have been viewed as a literary feature and, as such, was employed to raise the register of the narrative in both documentary56 and, as is evident here, literary texts. 3.2.6 Pronominals 3.2.6.1 Independent personal pronouns 3.2.6.1.1 First person common plural pronoun The MCA (ʾiḥnā) and CA forms (naḥnu) of the 1cpl independent pronoun occur interchangeably throughout the text, e.g. ‫‘ אחנא‬we’ (140 v. 3, 24); ‫נחן‬ ‘we’ (140 r. 9; 140 v. 3, 7; 141 r. 17). 3.2.6.1.2 Third person masculine and feminine pronouns The spelling of the 3msg and 3fsg independent pronouns in this text obscures their morphological form. Both independent pronouns are written with wordfinal ʾalef – ‫‘ הוא‬he’ and ‫‘ היא‬she’ – which may be interpreted as either a phonetic plene spelling of the short vowel /a/, characteristic of the CA independent pronouns huwwa ‘he’ and hiyya ‘she’; or else as the Hebrew independent pronouns ‫‘ הוא‬he’ (hūʾ) and ‫‘ היא‬she’ (hīʾ). 3.2.6.2 Suffix pronouns 3.2.6.2.1 Second person feminine singular suffix The use of mater lectionis yod in conjunction with the 2fsg suffix pronoun ‫יכ‬is indicative of the colloquial pronunciation of this suffix pronoun. Whereas in CA the 2fsg pronoun is pronounced with a final short vowel /i/, -ki ‘your’, in MCA /i/ or /ī/ precedes the kāf: -ik/-īk ‘your’, e.g. ‫‘ מכאניך‬your (f.) place’ (141 r. 13; CA: makānu-ki; MCA: makān-ik). 3.2.6.2.2 Third person masculine singular suffix after a consonant When written after a consonant, the 3msg pronoun suffix is denoted phonetically – according to MCA pronunciation – with Hebrew vav in this late JA literary text, e.g. ‫‘ ענדו‬with him’ (140 r. 3, 18; CA: ʿinda-hu; MCA:ʿandu); ‫‘ עומרו‬his age’ (140 r. 3; CA: ʿumru-hu; MCA: ʿumr-u). 3.2.6.2.3 Third person masculine singular suffix after a (long) vowel The Hebrew grapheme hey supplants vav as the representative of the 3msg direct object suffix pronoun when written after a (long) vowel, e.g. ‫יקרבוה‬ 55 A Muslim Middle Arabic example of the frequent and varied use of lam as a nominal and verbal negative particle is found in the 19th century text by al-Ğarādī in Mittwoch, 1926. 56 Wagner, 2010, pp. 135, 144. 412 ‘they sacrifice him’ (140 r. 6; CA: yuqarribūna-hu); ‫‘ סאלוה‬they asked him’ (140 r. 15; CA: sāʾalū-hu). The use of hey – more commonly associated with classical JA orthographic practices and directly influenced by CA orthography – avoids any confusion that may arise with the use of an additional vav after the 3mpl suffix -ū. 3.2.6.3 Interrogative pronouns The MCA interrogative pronoun fēn ‘where?’ (CA: ʾayna) appears once in the text in direct speech: (15) ‫ָ֗פין בית אל רב אברהם ןַ֗ עזרא‬ ‘Where is the house of R. Abraham ibn ʿEzra?’ (140 r. 15) Also in direct speech, the colloquial Arabic interrogative pronoun ʾēš ‘what’,57 which corresponds to MCA ʾēh, occurs: (16) ‫ַ֗פי סאלוהום איש תרידו תאכלו‬ ‘They asked them, “What would you like to eat?” ’ (141 r. 7) 3.2.6.4 Demonstrative pronouns Direct deixis is expressed in this JA folk narrative with the demonstrative pronouns ‫ האדַ֗ א‬hʾḏʾ and ‫ האדַ֗ י‬hʾḏy. The msg demonstrative pronoun ‫ האדַ֗ א‬hʾḏʾ emulates its CA counterpart hāḏā, while the fsg form ‫ האדַ֗ י‬hʾḏy may either be a reference to the archaic demonstrative pronoun hāḏī,58 or – more probably – a pseudo-literary feature, resulting from the amalgamation of the CA fsg form hāḏihi and the fsg MCA form dī. In contemporaneous JA letters, the CA-derived demonstrative pronouns have been almost completely supplanted by the colloquial demonstrative pronoun ‫ די‬dy ‘this’ (MCA: dī), confirming the use of these forms in this folk narrative as a literary or archaising device intended to elevate the register of the narrative. The MCA msg demonstrative pronoun ‫‘ דא‬this’ (140 v. 3) appears once in this text, used in a presentative manner. The appearance of dʾ is conspicuous. It has often been noted that in 17th century Middle Arabic literary and early 18th century JA documentary texts, the form ‫ די‬dy occurs to the exclusion of the contemporary msg form. As such, it has been argued that dy is an earlier, gender-indifferent version of the contemporary form.59 3.2.6.5 JA relative pronoun Unlike its CA counterpart (ʾallaḏī) but akin to its MCA equivalent (ʾillī), the JA relative pronoun ‫ אלדַ֗ י‬ʾldy/ʾlḏy, frequent in both classical JA documentary 57 The colloquial interrogative pronoun ʾēš ‘what’ is no longer used in daily speech in MCA. It has been superseded by the contracted form ʾēh ‘what’, while ʾēš remains the preserve of “proverbs and set phrases” according to Hinds and Badawi, 1986, p. 46. 58 See Fischer, 2002, p. 145, §274. 59 See Wagner, 2010, pp. 125–126. 413 texts and late JA texts alike, does not inflect for gender or number in this folk narrative: (17) ‫אלדי כאן ארסלהום למצר‬ ָ֗ ‫וגַ֗ ד אל אתנין אל תלמידים‬ ‘… he found the two students whom he had sent to Cairo’ (140 v. 12; CA: ʾallaḏīna) (18) ‫אלדי ביעמלו להא אל קוראבן‬ ָ֗ ‫ונדה על ע״ז אל כבירה‬ ‘And he summoned the large idol which they make the offering(s) to …’ (141 r. 9; CA: ʾallatī) 3.2.7 Construct state The MCA particle bitāʿ is used interchangeably with the construct state to denote possession: (19) ‫ַ֗פי סנה מן אל סנין עמלו אל גַ֗ ורל טלע עלא אבן אל רב בתאע דַ֗ אלך אל בלד‬ ‘One year, they cast the lot (and) it fell on the son of the Rabbi of that town’ (140 r. 2) (20) 3.3 3.3.1 ‫והום ַ֗פאל בלד בתאע אל דעווה‬ ‘And they were in the town from which the appeal came’ (140 v. 5) 60 Syntax Subordination 3.3.1.1 Adverbial clauses and complex subordinators Complex subordinators, composed of prepositions or conjunctions which take a complementiser to introduce an adverbial clause, are more prevalent than their ‘simple’ counterparts in this literary text. The following examples show complex subordinators acting as the head of a temporal adverbial clause, in (21), and a purposive adverbial clause, in (22). (21) ‫קעדו לארבע סעאת מן אל ליל לחין מא ַ֗פרגִ ית אל קראייה ַ֗פי פרקו כעך בסירג‬ ‘They sat for four hours during the night until the recitation was finished, then they distributed the cookies with sesame oil …’(140 r. 22–23) In MCA, the particle bitāʿ inflects for gender and number (in construct: bitāʿ (m), bitāʿit (f) and bitūʿ (pl)). In other instances in the text, the noun ‫‘ בלד‬town’, which may be either feminine or masculine in CA is referred to as feminine (as would be expected in MCA), e.g. ‫האדי אל בלד‬ ַ֗ ‘this town’ (140 r. 10–11); ‫‘ לם היא בלדנא‬it is not our town’ (140 v. 6). It is unclear as to whether ‫‘ בלד‬town’ was considered masculine or feminine in this context and, therefore, whether or not the construct particle bitāʿ is invariable in this text. 60 414 (22) ‫עטים לאגָ֗ ל אנהום יאכַ֗ דו אל ולד יווכבו בו פַ֗ אל בלד‬ ַ֗ ‫וגַ֗ ו אל ערלים במוכב‬ ‘The uncircumcised came in a great procession in order to take the boy (to) parade him through the town’ (140 v. 21–22) 3.3.1.2 Opacity As is reflected in examples (21) and (22) above, ‫( אן‬ʾn/ʾnn) – the JA equivalent of the CA complementisers ʾan ‘to’ and ʾanna ‘that’ – is used interchangeably with the free relative particle ‫( מא‬mʾ) in complex constructions introducing adverbial clauses. The syntactic functions of CA mā include that of free relative particle and complementiser, the identification of which is context-dependent: (23) baʿda māʾ rāʾa (a) ‘After what he saw’ (free relative particle) (b) ‘After he saw …’ (complementiser) In BnF Hébreu 583, however, the opacity regarding these two functions of māʾ appears to affect the use of the invariable JA relative pronoun ʾldy/ʾlḏy. In (24), the subordinator ḥyn occurs with the complementiser mʾ, twice. In the following example, (25), however, mʾ is replaced by the relative pronoun ʾldy/ʾlḏy, functioning as a complementiser:61 (24) ‫חין מא יגַ֗ ו באל מוכב יאכַ֗ דוך קול להום אן נכון אנא מעאך ⸱ וחין מא יקולו לך אתמנא‬ ‫רפיקי‬ ַ֗ ‫תועטא קול להום אלדַ֗ י יתמנא‬ ‘When they come in the procession to take you, tell them that I will be accompanying you. And when they say to you, “What(ever) you wish will be granted”, say to them, “Whatever my companion wishes!” ’ (140 v. 18–20) (25) ‫ָ֗פי חין אל ָ֗די טלע אל גַ֗ ורל עלא אל ולד קאמו באל בוכא ואל צייאח ואל נוואח ואל חוזן‬ ‘When the lot fell on the boy, they began weeping and wailing and mourning and grieving’ (140 r. 4) 3.3.1.3 Analogical extension In all forms of Arabic, the preposition baʿda may function as the head of a prepositional phrase with nominal dependents, as in (26), or as the head of an adverbial subordinate clause, as in (27). As has already been noted, adverbial subordinators in this text are often written in constructions which take a complementiser, which marks the shift of the construction from head of a prepositional phrase to head of an adverbial clause. 61 Further evidence of the effects of the opacity regarding the function of mʾ on the relative pronoun is found in the following example: ‫באלדי גַ֗ רא‬ ַ֗ ‫ …‘ חכו אל תלמידים לל סי’ חכַ֗ ם‬the students told Mr. Rabbi about what had happened to them’ (140 v. 16; CA: bi-mā ğarā). 415 (26) ‫חוצרו פי מצר‬ ַ֗ ‫מן בעד תלת אושהור‬ ‘After three months, they arrived in Cairo’ (140 r. 13) (27) ‫ַ֗פקעדו ללילת ערב פסח מן בעד מא עמל אל רב בדיקת חמץ ואראחו פַ֗ י קראייה קעדו‬ ‫לארבע סעאת מן אל ליל‬ ‘So, they stayed for the night of the eve of Passover. After the Rabbi did the chametz check, they went for recitation (and) they stayed (there) for four hours during the night.’ (140 r. 20–22) In a phenomenon analogous to that described above, the adverbial phrase fy mʾ bʿd (CA: fīmā baʿdu) ‘afterwards’, used in example (28) in accordance with its CA syntactic function, also appears in construction with the complementiser mʾ as head of a temporal adverbial clause in (29). (28) ‫ָ֗פי מא בעד נדה אל רב אב’ על ממזרתה וקאל להא אנזלי מן מכאניך ולעי אל נאר לאגַ֗ ל‬ ‫מא נטבוך אל ַ֗פראך‬ ‘Afterwards, Rabbi Abraham summoned his female bastard, saying to her, “Get down from your place and light the fire in order that we may cook the chicken.” ’ (141 r. 12–13) (29) ‫ָ֗פי מא בעד מא צלו ַ֗פאל כניס חכו אל תלמידים לל סי’ חכַ֗ ם באלדַ֗ י גַ֗ רא‬ ‘After they had prayed in the synagogue, the students told Mr. Rabbi about what had happened’ (140 v. 15–17) Constructions taking a complementiser to introduce adverbial clauses, which occur frequently in this and other 19th century folk narratives, are relatively rare in contemporaneous JA letters. They should be regarded as literary devices, added to enrich the ‘literary flavour’ of the narrative. 3.3.1.4 Relative clauses 3.3.1.4.1 Double use of relative pronouns In a phenomenon that only occurs once in the text, the MCA relative pronoun mīn ‘the one who’ is immediately followed by the JA relative pronoun ʾldy/ʾlḏy ‘that, which’. The JA relative pronoun was perhaps added to raise the register of the text after the use of a colloquial form. (30) ‫אלדי ינעמל קורבן אל סנה אל אתייה‬ ָ֗ ‫לאגַ֗ ל מא יערפַ֗ ו מין‬ ‘… in order that they might know the one who was to be sacrificed the following year’ (139 v. 22) 3.3.1.4.2 Indefinite noun and relative pronoun In the following example, the JA relative pronoun ‫ אלדַ֗ י‬ʾldy/ʾlḏy is preceded by an indefinite noun, which appears to be semantically definite. The omission 416 of the definite article may have simply been an error on the part of the author or scribe. (31) ‫אלדי יאכַ֗ דו אל ולד יקרבוה יחכום אוול לילת פסח‬ ָ֗ ‫וכאן מיעאד‬ ‘(The) date on which they would take the boy (and) present him as a sacrifice was decided as the first night of Passover’ (140 r. 6) 3.3.2 Definite article + noun + pronoun suffix In two instances, the definite article and a pronoun suffix are both employed to qualify the same noun: (32) ‫ַ֗פי נזלית אל ממזרתה‬ ‘His female bastard descended’ (141 r. 13-14) (33) ‫ואחד יקול האדַ֗ י אל בלדנא‬ ‘One (of them) said, “This is our town!” ’ (140 v. 6) 3.3.3 Prepositions 3.3.3.1 Directional lThe bound preposition li- has superseded the independent preposition ʾilā as the particle used to express spatial location in this text, e.g. ‫‘ ואתווגַ֗ הו למצר‬and they set off towards Cairo’ (140 r. 13–14); ‫‘ וצלו לבית אל חכַ֗ ם‬they arrived at the house of the rabbi’ (140 r. 7); ‫‘ מן האדַ֗ י אל בלד למצר‬from this town to Cairo’ (140 r. 10–11). 3.4 Hebrew content 3.4.1 Hebrew vocalisation There are two separate instances in which a JA noun is vocalised with Hebrew vowels: ‫‘ ַ֗פאל כַ֗ לֵיה‬in the wilderness’ (140 v. 3; CA: ḫalāʾ); and ‫‘ סַ אֵ ל‬a beggar’ (140 v. 9; CA: sāʾil). It is probable that in the latter case, vocalisation is included to avoid the misunderstanding that may arise with the otherwise homographic form ‫‘ סאל‬to ask’ (140 r. 15; 141 r. 7; CA: saʾala), or to indicate the omission of the hamza and presence of an /i/ vowel. 3.4.2 Hebrew lexicon The Hebrew content of this JA folk narrative accounts for 8.8 per cent of the total text, the most frequent manifestation of which is in nominal form. The Hebrew nouns often refer to Jewish religious events or concepts, e.g. ‫פסח‬ ‘Passover’ (140 r. 6, 21; 141 r. 19); ‫(‘ בדיקת חמץ‬the) chametz check’ (140 r. 21); ‫‘ ערלים‬uncircumcised’ (139 v. 19, 23; 140 v. 21, 24; 141 r. 2, 4, 11, 14); 417 and ‫רב‬/‫‘ ראב‬Rabbi’ (140 r. 2, 7, 8, 15, 19, 23; 140 v. 3). However more quotidian words are also used, spanning from the prosaic – e.g. ‫‘ ערב‬eve’ (140 r. 21) – to the profane – e.g. ‫‘ ממזר‬bastard’ (141 r. 10). Hebrew nouns are consistently qualified with the JA definite article ‫ אל‬rather than the Hebrew equivalent, e.g. ‫‘ אל גַ֗ ורל‬the lot’ (140 r. 2, 5); ‫אל תלמידים‬ ‘the students’ (140 r. 7, 23; 140 v. 2, 5, 12, 16); ‫‘ אל ערלים‬the uncircumcised’ (139 v. 19, 23; 140 v. 21, 24; 141 r. 2, 4, 11, 14). In one case, the Arabic 3msg suffix pronoun, denoted with a hey, is used to modify the Hebrew noun: ‫ממזרתה‬ ‘his female bastard’ (141 r. 12, 13–14). A small number of nominal forms in the folk narrative may be interpreted as either Arabic or Hebrew, e.g. ‫‘ לילת‬night of’ (140 r. 6, 21); ‫‘ חכַ֗ ם‬rabbi, sage’ (140 v. 8, 10, 12, 16; see also n. 27); ‫קורבאן‬/‫‘ קורבן‬a sacrifice, offering’ (139 v. 20, 22; 140 v. 18; 141 r. 9). These visual diamorphs have been included in the Hebrew word count. 3.4.3 Hebrew preposition The Hebrew preposition ‫( על‬ʿal) ‘on, by’ is used twice in this folk narrative after the verb ‫‘ נדה‬to summon’ (MCA: nadah) in place of the Arabic preposition ʿalā/li- ‘on, upon’/’to, for’: (34) ‫ונדה על ע״ז אל כבירה‬ ‘And he summoned the large idol’ (141 r. 9) (35) ‫ַ֗פי מא בעד נדה אל רב אב’ על ממזרתה‬ ‘Afterwards, the Rabbi Abraham summoned his female bastard …’ (141 r. 12) This switch into Hebrew following an MCA verb may be influenced by the fact that the object of the verb, immediately following the Hebrew preposition, is in both cases a Hebrew noun. 4 Summary As is evident in the spelling of consonantal waw and yā with double vav and yod, the not insubstantial Hebrew code-switching and the occurrence of niktib–niktibū forms, this folk narrative contains a number of features often referred to as characteristic of late JA. However, the quantity of these elements is overshadowed by the continuation of classical JA features, of limited CA influence, Middle Arabic practices and contemporaneous Arabic dialectal interference. From the plene spelling of short vowels and the denoting of the 3msg pronoun suffix with vav to the presence of the colloquial verb gāb ‘to bring’ and the fifth form’s prosthetic ʾalef, this text reveals numerous colloquial features 418 that are characteristic of MCA. Yet, in the presence of the JA relative pronoun, CA-influenced demonstrative pronouns and complex adverbial subordinators, the text also displays a preoccupation with raising the register above the quotidian, an aspiration which is achieved through the use of CA-influenced pseudo-literary features. Furthermore, the use of the diacritics and the consonantal representations indicate both a continuation of classical JA spelling practices – for example, in the representation of ḍād with ṣadeh and a supra-linear diacritic – and the enduring, albeit limited, influence of contemporaneous Arabic orthographic practices evident in the innovative application of the diacritic to graphemes such as peh for fāʾ, dalet for ḏāl and kaf for ḫāʾ, in imitation of the physical form of their Arabic equivalents. References Avishur, Y., 1992, “New folk-tales about Abraham b. ʿEzra (and his sons) from Egypt and Iraq” (Hebrew), in I. Levin and M. Itzhaki (eds), Studies in the Works of Abraham b. ʿEzra (Hebrew), Tel Aviv: Tel Aviv University Press, pp. 163–192. Blanc, H., 1974, “The nekteb – nektebu imperfect in a variety of Cairene Arabic”, Israel Oriental Studies 4, pp. 206–226. ———, 1981, “Egyptian Arabic in the seventeenth century: notes on the Judeo-Arabic passages of Darxe no’am (Venice, 1697)”, in Sh. Morag, I. Ben-Ami and N. A. Stillman (eds), Studies in Judaism and Islam: Presented to Shelomo Dev Goitein on the Occasion of his Eightieth Birthday by his Students, Colleagues and Friends, Jerusalem: Magnes Press, pp. 185–202. Blau, J. 1981. The Emergence and Linguistic Background of Judaeo-Arabic: A Study of the Origins of Middle Arabic, Jerusalem: Ben-Zvi Institute. Connolly, M. M., in press, “Revisiting the question of ğīm from the perspective of Judaeo-Arabic”, JSS. Diem, W., 2014, “Ägyptisch-Arabisch im 17. Jahrhundert. Die arabischen Zeugenaussagen in Mordechai ha-Levis Sefer Darḵe noʿam (Venedig 1697)”, Mediterranean Language Review 21, pp. 1–89. Esseesy, M., 2010, Grammaticalization of Arabic Prepositions and Subordinators: A Corpus-Based Study, Leiden: Brill. Fischer, W., 2002, A Grammar of Classical Arabic, New Haven: Yale University Press. Hary, B., 1992, Multiglossia in Judaeo-Arabic, Leiden: Brill. ———, 1996, “The ǧīm-gīm in colloquial urban Egyptian Arabic”, Israel Oriental Studies 16, pp. 153–168. ———, 1997, “On later and modern Egyptian Judaeo-Arabic”, in A. Afsaruddin and A. H. M. Zahniser (eds), Humanism, Culture, and Language in the Near East: Studies in Honor of Georg Krotkoff, Winona Lake, IN: Eisenbrauns, pp. 199–224. ———, 2009, Translating Religion: Linguistic Analysis of Judaeo-Arabic Sacred Texts from Egypt, Leiden: Brill. Hasson-Kenat, R., 2016, New Texts Written in Late Judaeo-Arabic from the Firkovitch Collection: Classification, Description and Sample Texts (Hebrew, unpubl. diss., Hebrew University of Jerusalem). 419 Hinds, M. and Badawi, E.-S., 1986, A Dictionary of Egyptian Arabic: Arabic-English, Beirut: Librairie du Liban. Jastrow, M., 2005, A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature, Peabody, MA: Hendrickson Publishers. Khan, G., 1991, “A linguistic analysis of the Judaeo-Arabic of late Genizah documents and its comparison with Classical Judaeo-Arabic” (Hebrew), Sefunot 20, pp. 223–234. ———, 1992, “Notes on the grammar of a late Judaeo-Arabic text”, Jerusalem Studies in Arabic and Islam 15, pp. 220–239. ———, 2006, “A Judaeo-Arabic commercial letter from early nineteenth century Egypt”, Ginzei Qedem 2, pp. 37–58. ———, 2007, “Judaeo-Arabic”, in K. Versteegh et al. (eds), Encyclopedia of Arabic Language and Linguistics, Leiden: Brill, vol. II, pp. 526–536. ———, 2010, “Vocalised Judaeo-Arabic manuscripts in the Cairo Genizah”, in S. Bhayro and B. Outhwaite (eds), From a Sacred Source: Genizah Studies in Honour of Professor Stefan C. Reif, Leiden: Brill, pp. 201–217. ———, 2013, “A Judaeo-Arabic document from Ottoman Egypt in the Rylands Genizah Collection”, in R. Smithuis and P. S. Alexander (eds), From Cairo to Manchester: Studies in the Rylands Genizah Fragments, Oxford: OUP, pp. 233–248. ———, 2016, “Judeo-Arabic”, in L. Kahn and A. D. Rubin (eds), Handbook of Jewish Languages, Leiden: Brill, pp. 22–63. Mittwoch, E., 1926, Aus dem Jemen. Hermann Burchardts Letze Reise durch Südarabien. Leipzig: Deutsche Morgenländische Gesellschaft. Ørum, O. G., 2017, ʾUṣṣit il-Gumguma or “The Story of the Skull”: With Parallel Versions, Translation and Linguistic Analysis of Three 19th-Century Judaeo-Arabic Manuscripts from Egypt, Leiden: Brill. Wagner, E.-M., 2010, Linguistic Variety of Judaeo-Arabic in Letters from the Cairo Genizah. Leiden: Brill. Wright, W., 1967, A Grammar of the Arabic Language: Translated from the German of Caspari and Edited with Numerous Additions and Corrections by W. Wright (2 vols). London: CUP. 420 Popular Renditions of Hebrew Hymns in 19th Century Yemen: How a Crudely Formed, Vocalised Manuscript Codex Can Provide Insights into the Local Pronunciation and Practice of Prayer REBECCA J. W. JEFFERSON University of Florida 1 Introduction The Isser and Rae Price Library of Judaica recently acquired a Yemenite manuscript codex, which was purchased online from Israel Mizrahi, the owner of the Mizrahi Book Store in New York.1 Mizrahi obtained the codex through an estate sale upon the death of a private collector from Morocco whose library had included many Yemenite books. Beyond these particulars, however, no other provenance information is available for this fascinating object. The bookseller’s inventory simply described the item as a “c1830 Yemen Manuscript of Kinot & Burial Procedures”. The Price Library of Judaica, with over 110,000 circulating items, is the leading Jewish studies research collection in the south-eastern United States. In addition, it has a growing special collections department boasting thousands of rare, early-modern printed books, periodicals, pamphlets and ephemera; unique archival collections from Florida, Latin America and the Caribbean, and a small collection of 19th century manuscripts from India, North Africa and most recently the Yemen.2 The Yemenite manuscript discussed here was purchased initially as a gift to the Library for use in classes related to the history of Jewish books and libraries, and to demonstrate to students the differences and similarities between early and late Hebrew manuscript traditions. The c1830 Yemenite manuscript codex (see figure 1) includes traditional lamentations (qinot) and penitential hymns (seliḥot) for the Ninth of Av and Israel Mizrahi’s Sephardi bookstore, the Mizrahi Book Store, is based in New York and has an online presence at http://stores.ebay.com/Jewish-Bookshop. For more information about Mizrahi, see http://forward.com/culture/books/307727/the-sephardic-bibliophile-of-brooklyn/. 2 For more information about the Library, see http://cms.uflib.ufl.edu/judaica/. 1 421 Figure 1. The Yemenite codex for personal mourning, as well as texts regarding funerary practices and burial procedures. The texts are vocalised throughout with Babylonian vowel signs. The codex itself is a curious and crudely constructed object: the first few pages of its text are repeated again on folios 2–3; additional pages on different paper and in different scripts have been added in between the main quires; and pages Figure 2. Parchment leaf used to fortify the binding 422 exhibiting pen trials and a get (divorce writ) formulary appear towards the back of the book. Hiding in between one of the folios and bound into the leaves is a colourful piece of patterned red cloth; some loose green threads are found in the binding next to the Kaddish prayer, and some plant leaves – probably sorghum – are wedged between two pages, and there is also what appears to be traces of flour in between the folds of several folios. Inside a cracked leather and loose sack cloth binding we see older, crumbled manuscript leaves being used to fortify the binding. One of these is a parchment leaf containing text from the Mishnah (see figures 2 and 3).3 This article, in addition to providing an analysis of the object itself and a general overview of the manuscript’s content and vocalisation, will examine a sample text from the codex, the classical Hebrew hymn ‫ שני חיי‬by Solomon Ibn Gabirol. This text will be analysed for its layout, poetic notation, orthography, vocalisation, morphology and syntax to compare the differences between this Yemenite edition and the standard, published version. In addition, the Yemenite edition of ‫ שני חיי‬will be compared to earlier vocalised manuscript versions of the poem from the Cairo Genizah,4 revealing similar practices in the everyday recitation of Hebrew hymns among the Jewish communities of Western Asia. Figure 3. Hidden leaves in the binding 3 The codex can be viewed online at http://ufdc.ufl.edu/AA00058758/00001, and the finding aid for the physical item is at http://www.library.ufl.edu/spec/manuscript/guides/yemenite.htm. 4 These versions are taken from Jefferson, 2004, pp. 14–49. 423 Figure 4. The tre lune watermark 2 Codicology: Materials Most of the leaves in this codex are produced from good quality late 18th– early 19th century Italian paper from the Galvani mills in Venice. The Galvani paper is identified by its tell-tale tre lune (three crescent) watermark and VG initial counterfoil (for Valentino Galvani, the company’s founder), many of which are found in the gutter margins of this codex (see figures 4–6). The crescent symbol is identifiable in paper produced by Genoese mills as early as 1520, but it was developed largely with export to the Ottoman Empire in mind, and it is found in Islamic documents and manuscripts occurring in endless variations. Tre lune watermarks found together with the counterfoil initials of VG or AG signify that the paper was produced in the Galvani paper mills in the province of Friuli between Venice and Trieste by either Valentino Galvani (the founder, who died in 1810) or his son Antonio or grandson Andrea, who continued the family business until the mid-19th century.5 The majority of the leaves in this codex are cut to a similar size and shape; the paper quality is thick and fibrous, with seven chain lines running horizontally at one inch apart. However, quire 7 is distinct from most of the rest of the 5 Walz, 2011, p. 88. 424 Figure 5. Later curly moon version of the tre lune watermark codex: its fibres are much smoother, the page size is smaller and the watermarks show much more elaborate, curlier moon faces. This type of tre lune Figure 6. VG initial counterfoil 425 watermark is also found in the Galvani papers but of a slightly later date.6 The paper bifolium at the end of the book, while on thicker, less smooth paper, also has the curlier-shaped crescent watermark. The pages in the first quire and the end two quires are in poorer physical condition than those in the middle quires, which are still very well preserved. The paper in the middle quires is of a lighter colour, whereas the paper in the outer quires is darker, and at the back of the codex the paper exhibits signs of water damage. 3 Codicology: Binding and quires The entire manuscript codex is contained within one modern cardboard binding, which has modern paper end leaves, and evidence of a former sack cloth and leather binding now only exists at the back of the codex. The manuscript itself comprises 111 manuscript leaves (or folios), of which 110 are handmade paper – 104 leaves are laid papers with horizontal chain lines, 6 leaves (at the front and back of the codex) have no visible lines and appear to be made from woven paper – and one leaf (an insert) is made from modern, commercially produced lined paper. The leather and sack cloth binding at the back contains approximately six leaves of paper (which are severely damaged and crumble to the touch) and one leaf of parchment. Figure 7. Inserted pages 6 Walz, 2011, p. 87. The paper itself began to be described by this prevalent watermark tre lune, and by the 1820s it had developed from a simple crescent shape to a moon face. 426 Two outer leaves at the beginning of the manuscript codex are conjoined to two leaves at the end, and these keep the entire manuscript together. Inside these outer leaves, the rest of the codex is constructed in a haphazard manner with eight distinctly bound sections plus three additional leaves. The first quire is comprised of the following sections (which are all sewn together with a thread going down the spine): two bifolia (4 folios) which surround the other three inner sections; one senion (6 bifolia or 12 folios) sewn through the gutter of the middle two pages; and a bifolium (2 folios) and an octonion (8 bifolia or 16 folios) sewn together through the middle pages. The second quire simply consists of one octonion sewn through the middle pages. Quire 3 has one ternion (3 bifolia or 6 folios) sewn through the middle. The fourth quire has a senion (6 bifolia or 12 folios) together with 2 singletons (f. 7 and f. 9) tipped in. The fifth quire is formed by one bifolium which comprises the outer leaves, inside of which are two binions (8 folios) attached to another bifolium. The sixth quire has a bifolium joined to a binion (which is sewn down the middle) attached to another bifolium (8 folios). Quire 7 was originally a ternion, although the end leaf is missing and is replaced by a modern, lined paper (see figure 7). The last quire is a quarternion (4 bifolia or 8 folios) which is sewn through the middle pages. There are three additional leaves after the last quire: two form a bifolium, and one originally joined to an outer leaf at the beginning of the book which is now missing. Variations in the size of the quires is something that is also found in the Cairo Genizah poetry manuscripts.7 However, as no systematic and comprehensive codicological study of Cairo Genizah manuscripts has been undertaken yet, it is difficult to say how widespread such variability is within those manuscripts. The findings and statistics of the major Hebrew manuscript codicology project, SfarData, are based on complete or near-complete dated Hebrew manuscript codices, of which there are only a small number in the Genizah. The SfarData project shows that most Hebrew manuscript codices have some variability in the size of the quires. Such variability was due to the exigencies of ensuring that the content fit the pages or due to there being multiple working stages in the process of the codex construction.8 Nevertheless, every codex examined had a clear dominant structure; for example, the dominant structure found in Oriental Hebrew manuscripts is the quinion.9 The lack of an overall dominant structure in this Yemenite codex further supports the idea that it was an informal user-produced manuscript. 7 See the reconstruction of MS JTS ENA 641 in Jefferson, 2004, p. 28, where the extant leaves enable us to deduce that part of the original medieval codex quires consisted of a ternion, a quinion and a quaternion. 8 Beit-Arié, 2012, p. 28. SfarData, The Codicological Data-Base of the Hebrew Palaeography Project at The Israel Academy of Sciences and Humanities, is online at http://sfardata.nli.org.il/ sfardatanew/home.aspx. 9 Beit-Arié, 2012, p. 30. 427 Figure 8. Binding structure The quires are bound together using three main cords, as well as several long strings threaded through the folds on the spine, which are then tied to one central knot (see figure 8). The three main cords are threaded through the sackcloth binding and glued to the outside where they would have been covered by the leather binding (see figure 9). 4 Contents of the codex The first section of the first quire contains a selection of qinot (lamentations) for the Ninth of Av, mostly from the Yemenite and Sephardi rites. In the case of the first qinah ‫אש תוקד בקרבי‬, this is a hymn known to all the major liturgical rites. The scribe, however, erroneously ascribes the poem to Judah Ha-Levi Figure 9. Cords used for binding the codex 428 when it was written by Abraham Ibn Ezra. A few hymns are found only in the Yemenite rite, such as ‫ אז בבית שבונו‬and ‫על נהרות בבל‬.10 Some of the hymns do not seem to be part of the traditional Ninth of Av liturgy, but are known from the diwanim (major poetic collections) of Judah Ha-Levi and Abraham Ibn Ezra; these include, for example, the hymns ‫ יונה מה לך‬and ‫אמרה ציון איך‬. Following that section, a bifolium of two conjoined pages has been inserted into the main quire. Its contents include selections of biblical text from the Book of Prophets entitled ‫‘( נחמה‬comfort’), followed by a pizmon by Judah Ha-Levi ‫יעלו לאלף ולרבבה כבני ציון‬, a list of poetry titles, a Judaeo-Arabic astronomical text concerning the lunar mansions, prognostications by quivering, another poem and finally more lists of poems.11 In the second section of the first quire, we find a selection of penitentiary hymns (seliḥot) all beginning with the title ‫אל מלך‬. These hymns continue through to quire 5, where we find some crudely written text, which could be writing practice, entitled ‫תמאם בן אדם לעול‬. More penitentiary hymns follow in the next section, but these are all in Aramaic and titled ‫ מרנות‬or ‫מרן‬. The Aramaic hymns continue through quires 6 and 7, and the final three pages of quire 7 are written in Judaeo-Arabic and deal with the lunar mansions, with the Hebrew title ‫ובה מולד השנה‬. Quire 8 begins with the title ‫‘( קינות שאומרים על המתים‬lamentations to recite over the dead’), and it also contains the Yemenite Baladi version of the Kaddish prayers,12 followed by more lamentations, each with the heading ‫מספד‬. Folio 7 verso of quire 8 contains a get (divorce law) formulary based on the text in Mishneh Torah, Sefer Nashim, Halakhot Girushin 4:12. The date ‫שנת‬ ‫‘ אלפא ותשע מאה וכך שנין לשטרות‬one thousand nine hundred and so-and-so years according to the Seleucid Era’ is also supplied. This date stands for the entire period 1900–1999 SE (1588–1687 CE), enabling the precise date to be inserted as necessary. By comparison, a divorce formulary in a 16th century Yemenite prayer book (JTS MS ENA 2249) does not supply a date range but rather keeps the statement open as follows: ‫בשנא דהוא כך וכך לירח שלשנת כך וכך‬. Evidence for an older date being used in a more modern prayer book is found in the Karaite tradition, and it suggests that the scribe was emphasising the importance of accurately transmitting long-established legal traditions.13 On the same page at the bottom, we find the words ‫‘( נסיון הדיו וקלמוס‬pen and ink trial’) several times. Subsequent pages continue to present piyyutim and prayers, ‫יעלו אלאלף‬ ‫ ולרבבה‬by Judah Ha-Levi, for example, which is also marked with the word ‫‘( נסיון‬trial’) at the head of the page. Other pages of writing include Mishnaic 10 Liturgical rites and hymns are listed, and sometimes transcribed and sung, on the National Library of Israel website and database ‫( אתר הפיוט והתפילה‬http://web.nli.org.il/sites/nlis/he/song). 11 These texts were identified by Nadia Vidro at UCL. A similar astronomical text concerning the lunar mansion is found in Kunitzsch and Langermann, 2003, p. 166. 12 A comparable version is available at https://kadishyatom.net/2009/‫בלדי‬-‫תימן‬-‫נוסח‬-‫יתום‬-‫קדיש‬/. 13 See Olszowy-Schlanger, 1998, p. 117. 429 texts (Tractate Avot) and some Talmudic passages, followed by unidentified text in Judaeo-Arabic (in a lighter ink and smaller script). The final page, which wraps around the entire codex and joins to the first pages of the codex, contains more ‫ אל מלך‬hymns in the same ink and script as the beginning of the book. The binding, as mentioned previously, contains older manuscripts used to fortify the structure. This was a known practice of Yemenite bookbinders who regarded the binding as a type of genizah.14 One piece is quite legible: a leaf of parchment containing text from Pirkei Avot, Mishnah 1:4–8, with Babylonian vocalisation. The other six leaves are unfortunately irretrievable: they comprise very crumbled papers with severely damaged and largely illegible text, although some words in Aramaic and some biblical verses in Hebrew are apparent. Overall, the contents of the book suggest that the scribe was copying his text directly from the old Yemenite Baladi (local) prayer book, the Sefer Tiklāl.15 In addition to prayers and liturgical poems, the Sefer Tiklāl also included elements from Saadia Gaon’s Siddur, the Haggadah, Megillat Antiochus, Tractate Avoth, Halakhic compendia and calendric tables, all of which point to the idea that these manuscript prayer books additionally served as a sort of religious almanac or guidebook on religious conduct.16 Thus, our Yemenite codex may have once formed part of a larger codex, or it may have been copied from an older text for practice. At the time it was written in around 1830, most Yemenite synagogues had adopted the Sefardi rite and only a handful clung to the older rite.17 5 Handwriting A number of scribes have worked on the pages in this book. The scribe who copied the qinah poems in the first section of the codex has a different handwriting to the scribe who copied the seliḥot. The differences are to be found in the size of the letters and the number of words placed on a line. The scribe who wrote the non-liturgical materials (lunar mansions and prognostications) had smaller handwriting than the copyist of the poems, and he uses a distinctive ‘alef-lamed ligature. Towards the end of the codex, where there is also non-liturgical material, the script again becomes dense, with more words fitting on a line. 14 Krupp, 2014, p. 287. Krupp discusses the wealth of old European manuscript fragments discovered in the bindings of a Yemenite manuscript collection. 15 Hubarah, 1964, pp. 221b, 222a. 16 Kohler, 1897, pp. 234–235. Kohler describes the Yemenite Siddur found in MS Gaster Codex 4 with its many additional materials as a type of “religious almanac”. 17 Klorman, 2014, p. 22. 430 Figure 10. Pen and ink trials notation Here and there throughout the codex, we find a scribe testing his pen. At the end of folio 70, the words confirming the efficacy of a pen and ink trial ‫ נסיון דדיו והקולמוס סימן טוב‬appear at the bottom of the page written in smaller characters in ink (see figure 10). The writer of the selection of biblical passages in the first quire has a hand that is childlike; the copyist of the Aramaic poems in quires 5–7 has handwriting similar to the writer of quire 1. The additional hymns, prayers and Judaeo-Arabic texts at the back of the codex are also written in two or three separate hands. All the scribes throughout the codex added catchwords to the bottom of the pages to assist with the binding. In the first section of the first quire, a catchword appears on every page (which isn’t necessary for ordering the pages); subsequent sections, however, only have the catchword at the bottom of the verso. Some of the scribes embellish the titles ‫ קינות‬or ‫ סליחה‬or ‫ אל מלך‬with decorative patterns. Later hands have added small notes in pencil, and on the verso of folio 1 a comment has been written in blue ink. 6 Insertions The existence of multiple readers/owners is attested by the addition of explanatory notes, such as ‫ אני יהודה הלוי ברבי שמואל‬added to the beginning of the piyyut ‫אבאר קצת פלאי אלהים‬, which inform the reader about the poem’s author and acrostic, and which are added in a different ink or smaller script. Other 431 Figure 11. Plant leaves found in the codex smaller notes on the content have been added to the margins here and there throughout the codex. In between folios 38 and 39 of the codex we find a small bunch of dried leaves, which look like sorghum (see figure 11). This plant is native to the Yemen and was used for baking breads like lahoh. The presence of these leaves together with a white substance (flour?) in the gutters of folios 47–48 (all part of the same quire) suggest that one of its owners may have been involved in the production or sale of wheat or bread. Yemenite breads feature regularly in Jacob Sapir’s account of his visit to Yemen in the 1850s.18 Another ‘foreign’ insert is found between folios 77 and 78, where we find a small torn page from a printed text of zemirot (this may have belonged to the later Moroccan owner) which perhaps served as a bookmark. A page of text on modern commercial paper is sewn into the codex at the end of the seventh quire. The text appears to be a continuation of some calendrical text in Judaeo-Arabic, and it may have been added in later to replace the missing folio from the back of the quire; however it has been placed in the wrong way around. In quire 8, in between folios 3 and 4, next to the Kaddish prayers, are some green and red threads. Towards the back of quire 8, between folios 7 and 8, where the text has a practice get formulary, several pen trials and the continuation of the poem from the previous folio, a piece of cloth is bound into the quire. It is mostly red with repeated thin green stripes and geometrical patterns 18 See e.g. the description in Lavon, 1997, p. 70. 432 Figure 12. Inserted strip of cloth (crosses and key shapes) in a beige colour (see figure 12). In colour and design it could be derived from a Yemenite or Moroccan textile. 7 The penitentiary hymn ‫שני חיי‬ In order to look more closely at the ways in which the hymns in this Yemenite codex were popularly recalled and read, we will examine a sample text, the penitentiary hymn (seliḥah) entitled ‫‘( שני חיי‬These are the Days of My Life’), by the medieval Spanish poet Solomon Ibn Gabirol. The penitentiary hymn was introduced in the pre-classical period of piyyut (around 5th–6th century CE) to accompany the fixed penitential prayers. During the Andalusian period of piyyut (10th–12th century CE), the emotional charge of the High Holidays led to an increased desire for audience participation to which the poets all responded by composing seliḥot with the addition of popular choruses or refrains (pizmonim). Solomon Ibn Gabirol, for example, composed at least ten of these hymns, according to Jarden’s edition of his liturgical poems. The old traditional Yemenite prayer books according to the Baladi (local) rite were replete with seliḥot. The poem ‫ שני חיי‬has lines of verse divided into three metrical units: two units have three long syllables and rhyme with each other; the third has six long syllables and carries the rhyme of that strophe; short vowels like vocal shewa and the ḥaṭefim are disregarded in the metre. The internal rhymes and rhythms in this poem create a memorable and compelling sound. A refrain (pizmon) based on the biblical verse ‫‘( כי גר אנכי עמך‬for I am a stranger with 433 thee …’, Ps. 39:12) is regularly repeated throughout the hymn. The initial letters in the first three lines of verse spell out the identity of the poet, “Solomon”. 7.1 Manuscript and printed editions of ‫שני חיי‬ Ibn Gabirol’s ‫ שני חיי‬was included in the Tlemcen (North African) mahzor (festival prayer-book) for the Days of Awe, in the Romanian and Lithuanian liturgical rites, and it is found in numerous anthologies of seliḥot.19 The Israeli scholar Dov Jarden produced a modern critical edition in his two-volume edition of Solomon Ibn Gabirol’s liturgical poetry.20 The critical edition does not follow any one manuscript version exclusively but rather presents an idealised version of each text based on a thorough analysis and comparison of all the known manuscript editions (although greater weight is given where a substantial number of manuscripts are in agreement). Jarden was able to trace 49 manuscript versions of ‫שני חיי‬.21 Another previously unknown medieval copy was discovered by Jefferson in CUL MS T-S AS 116.338; this version, however, is badly damaged and only sporadically vocalised.22 Jefferson has also identified another copy of the poem in a Yemenite manuscript from 1472 CE (JTS MS ENA 2250).23 Of the manuscripts on Jarden’s list, twelve are from the Cairo Genizah collections, and four of these are fully vocalised: CUL MS T-S NS 299.185, an informal copy on re-used paper; JTS MS ENA 641, ff. 7–8, from a collection of seliḥot originally bound into multiple quires of varying lengths; CUL MS T-S NS 299.98, a version from a collection of seliḥot, with more attention paid to presentation; and JTS MS ENA 3239, f. 26, a neat copy derived, perhaps, from the same line of transmission as T-S NS 299.98. All four vocalised Genizah versions appear to stem from closely related lines of transmission. This is evident in shared textual variations, vocalisation traits and variations of syntax. Similarities can also be found between the 19th century Yemenite manuscript discussed here and JTS MS ENA 2249 (the first part of the above-mentioned JTS MS ENA 2250) from the 15th century, which may prove a long and extensive line of transmission for what appear to be more commonly used and copied versions of the poem. 7.2 Aids to reading the hymn ‫שני חיי‬ Like all the other seliḥot in this Yemenite codex, the hymn ‫ שני חיי‬is supplied with the heading ‫“ אל מלך‬God, the King”, a shortened form of the phrase ‫אל‬ 19 Davidson, 1961, vol. III, no. 1961. Jarden, 1971–1972, vol. II, no. 220. 21 Jarden, 1971–1972, vol. II, p. 696. 22 Jefferson, 2004, p. 16. 23 JTS MS ENA Collection, 2250, folio 249. This version is also vocalised with supra-linear Babylonian vowel signs. 20 434 Figure 13. The text of ‫שני חיי‬ :‫“ מלך יושב‬God, the King Who sits” from the introduction to the ‘Thirteen Attributes’, which form the core of all the seliḥah prayers.24 In this codex, Ibn Gabirol’s seliḥah appears in quire 4, ff. 4–5 or ff. 62–63 of the whole manuscript (see figure 13). It follows another seliḥah, ‫שנותיו ספו בדלות וקלות‬, also by Solomon Ibn Gabirol. In addition to the ‫ אל מלך‬title, which indicates the start of a new hymn, the scribe helps the reader know how to accentuate the pauses in the poetic text by placing a dot and a space after each line of verse, and a colon at the end of each stanza together with the abbreviation ‫ פז‬pizmon (chorus/refrain). For aesthetic purposes, he employs either three dots, a line or an expanded final letter as a space filler to keep the one-inch margins around the text even, as no ruling or guiding lines have been drawn. The scribe’s methods of textual layout are similar to those used in the medieval manuscripts of the Cairo Genizah.25 7.3 Transcription of ‫שני חיי‬ The supra-linear signs used to represent the Babylonian vowel signs in this transcription are as follows: ‫( א‬ḥireq), ‫( א‬ṣere), ‫( א‬shewa or ḥaṭefim), ‫( ָ֔א‬ḥolem), ‫( א‬pataḥ or seghol), ‫( ַ֬א‬qameṣ), ‫( ֜א‬šureq). 24 25 Elbogen, 1993, p. 178. See e.g. ‘Aids to reading’ in Jefferson, 2004, p. 23. 435 ‫‪Folio 62, recto26‬‬ ‫‪ .4‬אל מלך‬ ‫‪ .5‬שני חיי ו֝ מאויי לריק לריק ספו֝ בחוָ֔ בוָ֔ תי‬ ‫‪ .6‬למי אפנה ו֜ מי אענה וחטאתי סביבוָ֔ תי ַ֬האל‬ ‫‪ .7‬חמוָ֔ ל וחסד גמוָ֔ ל ואם ַ֬רבו֜ משו֜ בוָ֔ תי כי גר‬ ‫‪ַ֬ .8‬אנכי ע ַ֬מך תוָ֔ ַ֬שב ככַ֬ ל אבוָ֔ ַ֬תי‪ :‬זַ֬ ד‬ ‫לא נכבַ֬ ש והוַ֬ תוָ֔‬ ‫לא נכנע ו ָ֔‬ ‫‪ .9‬יצרי על יוָ֔ צרי ו ָ֔‬ ‫כראש פתן בקרב דבש ו֜ מה בצעו֜‬ ‫ָ֔‬ ‫‪ .10‬בתאוַ֬ תוָ֔‬ ‫לא‬ ‫לא חו֜ ַ֬בש ו ָ֔‬ ‫לא זוָ֔ ר ו ָ֔‬ ‫‪ .11‬עלי פצעו֜ אשר ָ֔‬ ‫יח ַ֬בש‬ ‫שלך כבוָ֔ ר ַ֬שאוָ֔ ן ו ַ֬שם ַ֬‬ ‫‪ .12‬נמ ַ֬לך ביוָ֔ ם יו֜ ַ֬‬ ‫לא זַ֬ כר כבן נכַ֬ ר ביוָ֔ ם ַ֬ע ַ֬פר וגו֜ ש יל ַ֬בש‬ ‫‪ .13‬ו ָ֔‬ ‫נטו֜ י ואני כַ֬ עשב אי ַ֬בש בקו֜ ם‬ ‫‪ .14‬יַ֬ מי כצל ַ֬‬ ‫‪ .15‬שדי ל ַ֬בנוָ֔ תי מה יהיו֜ תשו֜ בוָ֔ תי פז‬ ‫‪ .16‬עזוב‬ ‫‪Folio 62, verso27‬‬ ‫יבב נתיב שוָ֔ בַ֬ ב ורו֜ ץ כצבי והנַ֬ צל‬ ‫‪ .1‬עזוָ֔ ב ל ַ֬‬ ‫‪ .2‬פשוָ֔ ט הבל לבו֜ ש אבל וה ַ֬אזוָ֔ ר וה ַ֬אצל וקו֜ ם‬ ‫‪ .3‬ליל כבן חיל ואל תישן שנת ַ֬עצל ולין‬ ‫‪ .4‬בוָ֔ דד ומתנוָ֔ דד ו֜ בבכי ַ֬אזנ ַ֬ך צלצל והתנחם‬ ‫‪ .5‬ו ַ֬אז תרו֜ ַ֬חם ואו֜ לי נפשך תצל ו֜ מה יצדק‬ ‫‪ַ֬ .6‬אבק דק ואיך יו֜ כל להנַ֬ צל כציץ יַ֬ ַ֬צא‬ ‫‪ .7‬וי ַ֬מל ויברח כצל ו֜ מה יתרוָ֔ ן לחרבוָ֔ תי‬ ‫‪ .8‬לא ַ֬לפי וריבבוָ֔ ַ֬תי פז גדוָ֔ ל‬ ‫קמתי אדוָ֔ ן‬ ‫‪ .9‬חסד אשר יַ֬ סד גַ֬ למי ור ַ֬‬ ‫עלם והו֜ א נ ַ֬צב לעו֜ ַ֬מתי‬ ‫‪ .10‬עוָ֔ ַ֬לם אשר נ ַ֬‬ ‫‪ .11‬ארנן ואתחנן בכַ֬ ל שבתי וקי ַ֬מתי‬ ‫‪ .12‬פדו֜ ת מהר ולב טהר והעבר את כל ַ֬מתי‬ ‫‪ .13‬עין פקח ושועי קח ואל תפקוָ֔ ד עלו֜ ַ֬מתי‬ ‫דמתי‬ ‫‪ .14‬רעה אוָ֔ תי עדי מוָ֔ תי ליוָ֔ ם שו֜ בי לא ַ֬‬ ‫‪ .15‬ותיטיב סוָ֔ ף ביוָ֔ ם תאסוָ֔ ף ל ַ֬ך רו֜ חי‬ ‫‪ .16‬ונשמתי‬ ‫‪Folio 63, recto‬‬ ‫‪ .1‬ונש ַ֬מתי מבין נתיבוָ֔ תי ויוָ֔ דע ַ֬מחשבוָ֔ תי‬ ‫‪.‬בקרב ‪26 Line 7: The vocaliser seems to have added two vowel signs above the letter bet in‬‬ ‫‪,‬יח ַ֬בש ‪Line 9: The vocalisation isn’t clear here – there are two dots above the yod in‬‬ ‫‪ַ֬ but these‬‬ ‫‪are probably marked in error.‬‬ ‫‪, perhaps indicating that it has‬ל ַ֬בנוָ֔ תי ‪Line 12: The scribe has placed a small letter over the bet in‬‬ ‫‪been copied in error.‬‬ ‫‪ is not clear as the sign is blurred from other marks on the page.‬והנַ֬ צל ‪27 Line 1: The ṣere in‬‬ ‫‪436‬‬ 7.4 Textual variants between the Yemenite edition and other editions The first variant in this Yemenite edition of ‫ שני חיי‬is the erroneous copying of the words ‫‘( לריק‬for nothing’) twice in the first line of verse. A second variant is the use of the phrase ‫‘( ו֜ מי אענה‬and to whom shall I reply’) on f. 62 r. 3 where all the other MS versions examined here, including the model printed text, have the phrase ‫‘( ומָ ה אֶ ֱע ֶנה‬and how shall I reply’). On f. 62 r. 6, the scribe has written ‫‘( על יוָ֔ צרי‬on my creation’) where the model version has the poetic form ‫‘( ֲעלֵי‬upon’). The same formulation is found in JTS MS ENA 2250, f. 249. One case where our Yemenite manuscript and the other manuscript versions are all in agreement is the use of the words ‫יח ַ֬בש‬ ַ֬ ‫‘( ו ַ֬שם‬and there imprisoned’) in the verse of f. 62 r. 9, in contrast to the published ‘model’ text which has ‫‘( וְ גַם יָחְ בָ ש‬and also imprisoned’). On f. 62 r. 12, the scribe has used the phrase ‫‘( ל ַ֬בנוָ֔ תי‬for my daughters’) where the model text has ‫‘( לְ ִריבֹותָ י‬for my quarrels’), but a letter has been placed underneath the bet perhaps signifying that this is an error. In addition, it appears as though he has vocalised the suffixes of these words with a ḥireq rather than with the qameṣ needed to signify the plural: ‫‘( בקו֜ ם שדי לבַ֬ נוָ֔ תי מה יהיו֜ תשו֜ בוָ֔ תי‬when the Lord rises for my daughters what will be my replies’). It is possible that the curved strokes of the qameṣ were made so quickly they appear like dots. On f. 62 v. 2, the scribe has used the words ‫‘( וה ַ֬אזוָ֔ ר וה ַ֬אצל‬and an area/band and a throne’) where the model text has ‫‘( וָּֽׁ הִ ָנזֵר וְ הֵ אָ צֵ ל‬and a crown and a throne’). This variant, although spelt as ‫ והיאזור‬with an additional yod, also occurs in CUL MS T-S NS 299.185. On f. 62 v. 4, the scribe provides the present tense form ‫‘( ומתנוָ֔ דד‬and staggering’) where the published edition has ‫‘( וְ הִ תנֹודֵ ד‬and staggered’). However, the present tense form does occur in all the other medieval Cairo Genizah manuscript versions of this text, and also in the Yemenite Codex JTS MS ENA 2250, f. 249. On three occasions, words that appear in the model text are missing from the Yemenite manuscript. In two of these cases, the missing words are necessary to the sense of the verse, and in all three cases the additional syllables are needed for maintaining the prosody. The first of these missing words is ‫וגם‬ (‘and also’), where the model has ‫‘( וגם רקמתי‬and also upon my flesh’); ‫ וגם‬is likewise missing from JTS MS ENA 641, ff. 7–8. The word ‫‘( לך‬go’) is absent from the text at the beginning of f. 62 v. 11; ‫‘( לך ארנן‬go to Arnon’) is attested in both the JTS medieval version and in the model text. The third missing word is the imperative form ‫‘( צור‬create’), absent in f. 63 r. 1 but found at the beginning of the line of verse ‫‘( צור מבין נתיבותי‬create from my paths’) in both the JTS manuscript and the published version. Finally, two verses appearing on f. 62 v. 12–13 are placed in reverse order to the way they appear in both the JTS medieval manuscript and the printed version, where what corresponds to line 13 here comes before line 12. 437 7.5 Orthography The scribe consistently uses waw as a vowel letter where the model version prefers the defectiva spellings without waw. The use of plene spellings is standard in Rabbinic Hebrew (RH) orthography, and it occurs regularly in the medieval manuscript versions of this hymn from the Cairo Genizah,28 and in the 15th century Yemenite manuscript JTS MS ENA 2550, f. 249. Thus, for example, all the manuscript versions have the imperatives ‫‘( חמוָ֔ ל‬take pity’), ‫‘( גמוָ֔ ל‬bestow’) and ‫‘( עזוָ֔ ב‬leave’) spelt with a waw, and also the pu’al forms ‫‘( תרו֜ ַ֬חם‬you will find compassion’), ‫‘( חו֜ ַ֬בש‬covered’) and ‫שלך‬ ַ֬ ֜‫‘( יו‬be discarded’) have waw for the ‘u’ vowel, where the model text has the defective spellings with qibbuṣ. In addition, the Yemenite scribe spells ‫‘( לבב‬heart’) with a yod, ‫( ליבַ֬ ב‬f. 62 v. 1). 7.6 Vocalisation The vocalisation signs appear to be in the same ink as the consonantal text and all words are vocalised in this poem, except for one on f. 62 r. 3: ‫‘( וחטאתי‬my sins’). There are no diacritics, and the scribe/vocaliser consistently uses the shewa sign for the ḥaṭefim, which is a graphical convention found in medieval Hebrew manuscripts.29 Vocal shewa is marked regularly, except for once in ‫‘( ו֜ בבכי‬and the tears of’, f. 62 v. 4) and again in ‫עלם‬ ַ֬ ‫‘( נ‬disappeared’, f. 62 v. 10). In a similar way, silent shewa is mostly left unmarked, except in one phrase, ‫‘( יוָ֔ צרי‬my creation’, f. 62 r. 6). Consistent with the modern Yemenite reading tradition, the scribe/vocaliser uses a single vowel sign for pataḥ and seghol (marked in this transcription with a symbol resembling the small ‘ayin shape of the Babylonian equivalent of pataḥ, the miֿptaḥ pumma sign).30 Lastly, on a number of occasions, the scribe/vocaliser places the ḥireq sign above the following letter, usually (but not always) where yod is used as a vowel letter – for example, in ‫‘( לריק‬for nothing’, f. 62 r. 2) – or following consonantal yod, as in ‫‘( יתרוָ֔ ן‬advantage’ or ‘benefit’, f. 62 v. 7). In a few places, we find that the scribe/vocaliser has used the pataḥ vowel sign where qameṣ is expected. Most of these examples occur where the model text produces a pausal form (see section 7.7 below). However, in three cases, the unexpected use of pataḥ cannot be explained by the scribe/vocaliser’s inconsistency in marking the pause. These examples are: in the last syllable of ‫לא נכנע‬ ָ֔ ‫‘( ו‬and did not surrender’, f. 62 r. 6), in the last syllable of ‫‘( ַ֬אבק‬dust’, f. 62 v. 6), and above ḥet in ‫‘( לחרבוָ֔ תי‬to my swords’, f. 62 v. 7). Similarly, we find two instances in this text where the scribe/vocaliser has used the qameṣ sign and not the expected pataḥ: first, on f. 62 r. 9 in the final syllables of the 28 See e.g. Jefferson, 2004, p. 20. See Jefferson, 2004, p. 266. 30 Khan, 2013, p. 955. 29 438 first and third words, ‫שלך‬ ַ֬ ֜‫‘( נמ ַ֬לך ביוָ֔ ם יו‬succeed on his being overthrown’); and then again on the first syllable of the final word in this text, ‫‘( ַ֬מחשבוָ֔ תי‬my thoughts’), but here the vowel sign is faint. The scribe/vocaliser has employed the ṣere sign in an unusual way twice in the lines of verse between f. 62 v. 3–4: firstly, in the phrase ‫‘( ולין‬and sleep’) where one expects ḥireq, and secondly in the final syllable of ‫‘( והתנחם‬and consoled himself’), instead of qameṣ. The first word appears in a section of the poem in which most of the words have ṣere in the final syllable; the second word appears at the start of the next line of verse, and so perhaps the scribe/vocaliser ‘heard’ the ‘e’ vowel dominating throughout these particular words. An unexpected use of ṣere also appears on the next line (f. 62 v. 5) in ‫ואו֜ לי‬ (‘and perhaps’) where pataḥ is expected. The text also reveals one example where ṣere has been used for ḥolem, in the word ‫‘( ַ֬אנכי‬I’, f. 62 r. 5). This may reflect the closeness of the ṣere and ḥolem in the Yemenite pronunciation tradition.31 In the last line of verse (f. 63 r. 1), it appears that the vocaliser may have simply placed his ṣere and ḥireq vowels in the wrong order: ‫‘( מבין‬understands’). In another singular case within this text, we find ḥolem used in place of qameṣ in ‫ גדוָ֔ ל‬on f. 62 v. 8 where the model text has the imperative ‫‘( גָּֽׁ דָ ל‬grow/increase’). Again, this may reflect the close quality of the vowels in the Yemenite pronunciation of ḥolem and qameṣ.32 7.7 Pausal forms The vocaliser vocalises pausal forms inconsistently. Thus, for example, he omits to use them at the end of the first three verses of the first stanza (f. 62 r. 1–5) – ‫‘( בחוָ֔ בוָ֔ תי‬with my iniquities’), ‫‘( סביבוָ֔ תי‬all about me’) and ‫‘( משו֜ בוָ֔ תי‬my thoughts’) – but he does use it for the final verse of that stanza – ‫‘( אבוָ֔ ַ֬תי‬my fathers’). This same pattern of vocalising is also prevalent amongst the Cairo Genizah manuscript versions of this hymn. 7.8 Morphology This sample text from the Yemenite codex provides one example where the scribe/vocaliser has vocalised the form of the triliteral root in such a way that it presents a different form of the verb than the one appearing in the model text. Thus, ‫ יסד‬has been vocalised on f. 62 v. 9 as though it were the pa’al form, ‫‘( יַ֬ סד‬he founded’), rather than the pi’el form ‫‘( יִ סֵ ד‬it was founded’) present in the model text. 31 32 Khan, 2013, p. 956. Morag, 1963, p. 92. 439 7.9 Syntax The Yemenite scribe copied the phrase ‫‘( ורו֜ ץ כצבי והנַ֬ צל‬and run like a hart and save yourself’) on f. 62 v. 1 where the model text has ‫‘( וְ רוץ כַצְ בִ י לְ הִ נָצֵ ל‬and run like a hart to save yourself’). The version in the Yemenite codex is also found in all the Genizah manuscripts consulted here. In another example where the Yemenite scribe produces variant syntax, we find on f. 62 v. 8 the phrase ‫‘( לא ַ֬לפי וריבבוָ֔ ַ֬תי‬to my thousands and my tens of thousands’). The published, standard edition of the poem, however, has ‫‘( וְ ַל ֲא ָלפַי ולְ ִרבבֹותָ י‬and to my thousands and to my tens of thousands’). The version in the codex is also found in JTS MS ENA 3239, while a different version, ‫ולא ַ֬לפי וריבבוָ֔ ַ֬תי‬, occurs in the Yemenite JTS MS ENA 2550, f. 249. These manuscript versions ignore the prosodic rules which require the additional syllables provided by the addition of waw conjunctive and by the addition of the preposition ‫‘ ל‬for’. 8 Conclusion: The writers, readers and owners Our Yemenite codex was a much-used book with multiple scribes, readers and owners. The different scribes/vocalisers, with at least five of them, can be detected in the variable handwriting found across the quires. Later readers of these texts are apparent in the notes they added in the margins, which are also rendered in at least three separate hands. And at least one of the owners has left physical traces within the codex: an additional page of modern printed paper and the non-textual physical items tucked in between the pages, as well as the later addition of a cardboard outer binding (which may have been added in Morocco or the United States). Thus, the codex is something of a conundrum: a medieval-style manuscript on 19th century Venetian paper surrounded by modern cardboard. The sewing of the quires and spine reveals a mixture of informal and formal techniques, including stab and slip stitches; the manuscripts used to fortify the binding reflect the old Yemenite practice of bookbinding, and the irregular sizes of the quires recall common codicological practices hinted at by the leftover fragments of the Cairo Genizah. The text itself has older and newer elements: notes in the margins and above poems are suggestive of a later reader/interpolator; whereas the sections of the Aramaic and Judaeo-Arabic text recall the liturgical practices of the traditional Yemenite Baladi rite, and parts of the codex (particularly the sample poetic text examined here) echo earlier nonstandard readings found in medieval Cairo Genizah manuscripts, as well as some non-standard readings in an earlier Yemenite codex from the 15th century. Indeed, the non-standard vocalisation traits described above confirm that a high degree of orality surrounded the reproduction of Hebrew hymns throughout the ages and across Jewish cultures, and that greater latitude was continually exercised in the rendering of poetic texts as opposed to biblical or formal religious texts. 440 References Beit-Arié, M., 2012, Hebrew Codicology: Historical and Comparative Typology of Hebrew Medieval Codices based on the Documentation of the Extant Dated Manuscripts from a Quantitative Approach (pre-publication version 0.1, http://web.nli.org.il/sites/NLI/English/collections/manuscripts/hebrewcodicology/ Pages/default.aspx). Davidson, I. (ed.), 1961, Thesaurus of Medieval Hebrew Poetry (4 vols, Hebrew), New York: Ktav. Elbogen, I., 1993, Jewish Liturgy: A Comprehensive History, R. P. Scheindlin (trans.), Philadelphia: Jewish Publication Society. Hubarah, Y. (ed.), 1964, Ha-Tiklal: Sidur Tefilah Kadmon Ke-minhag Yehude Teman ʻal pi ketav yad Tiklal Yaḥya Elbasiri... (Hebrew), Jerusalem: Y. Hubarah. Jarden, D. (ed.), 1971–1972, The Liturgical Poetry of Rabbi Solomon Ibn Gabirol (2 vols, Hebrew), Jerusalem: American Academy for Jewish Research. Jefferson, R., 2004, Popular Renditions of Hebrew Hymns in the Middle Ages: Based on a Selection of Vocalized Liturgical Poems by Solomon Ibn Gabirol from the Cairo Genizah (unpubl. diss., University of Cambridge). Khan, G., 2013, “Vocalization, Babylonian”, in G. Khan (ed.), The Encyclopedia of Hebrew Language and Linguistics, Boston: Brill, vol. III, pp. 953–963. Klorman, B. Z. E., 2014, Traditional Society in Transition: The Yemeni Jewish Experience, Leiden: Brill. Kohler, K., 1897, “Review of ‘The Haggadah according to the Rite of Yemen’ by William H. Greenburg”, American Journal of Semitic Languages and Literatures 13/3, pp. 234–239. Krupp, M., 2014, “European fragments in the spines of the book collection of a Yemenite community”, in A. Lehnardt and J. Olszowy-Schlanger (eds), Books Within Books: New Discoveries in Old Book Bindings, Leiden: Brill, pp. 287– 298. Kunitzsch, P. and Langermann, T., 2003, “A star table from medieval Yemen”, Centaurus 45, pp. 159–174. Lavon, Y. (ed. and trans.), 1997, My Foosteps Echo: The Yemen Journal of Rabbi Yaakov Sapir, Israel: Targum Press. Morag, S., 1963, The Hebrew Language Tradition of the Yemenite Jews (Hebrew), Jerusalem: The Academy of the Hebrew Language. Olszowy-Schlanger, J., 1998, Karaite Marriage Documents from the Cairo Geniza: Legal Tradition and Community Life in Mediaeval Egypt and Palestine, Leiden: Brill. Walz, T., 2011, “The paper trade of Egypt and the Sudan in the eighteenth and nineteenth centuries and its re-export to the Bilad as-Sudan”, in G. Kratli and G. Lydon (eds), The Trans-Saharan Book Trade: Manuscript Culture, Arabic Literacy and Intellectual History in Muslim Africa, Leiden: Brill, pp. 73–108. 441 The Status Quaestionis of Research on the Arabic Bible RONNY VOLLANDT Ludwig-Maximilians-Universität, Munich In what follows, I seek to offer a status quaestionis of research on the Arabic Bible. As a newly emerging field of academic research, it has a need to clearly define itself and to develop methodological standards. This is necessary not least to close scholarly lacunae and produce new, seminal perspectives on the field. Many questions as to the origins of biblical versions in Arabic, their various text types, their Vorlagen and translation strategies, their geographical, chronological and denominational distribution, as well as to the ways they were produced, disseminated and consumed can, for the time being, only be answered tentatively. This contribution thus attempts to bring together different strands of a dynamic field, which has received considerable momentum since the turn of the new millennium. It lies in the nature of posing a status quaestionis to be descriptive and programmatic at the same time. I have relegated tangential discussions, as interesting they may to be, to the footnotes, which are at times quite lengthy. Much of the recent scholarship draws from understudied primary sources. All of these sources share a common denominator: they bear witness to the attempts of various communities to realign the biblical text with a new era in a time of profound political, social and cultural change, on the one hand, and to the need for comprehensible versions in Arabic, the new vernacular, on the other. Further, they attest to a great variety of textual traditions and a mobility, partly intercommunal, of these traditions. The total number of manuscripts containing Arabic versions of the Bible, which can only be an estimate as there is no comprehensive union catalogue, amounts to about ten thousand items.1 While only a very small, almost minute, portion of this corpus has been duly identified or published in critical or semi1 In contrast to the situation with Arabic Bible manuscripts, inventories exist for other corpora. For the Septuagint, for example, we have Rahlfs, 1914; Aland, 1975; Rahlfs and Fraenkel, 2003; and Septuaginta-Unternehmen der Akademie der Wissenschaften zu Göttingen, 2012 (https://rep.adw-goe.de/handle/11858/00-001S-0000-0022-A30C-8, accessed January 2018). For the Greek New Testament, see Gregory, 1908; Aland, Welte, Köster and Junack, 1994; and also the New Testament Virtual Manuscript Room of the Institut für Neutestamentliche Textforschung (http://ntvmr.uni-muenster.de/liste/, accessed January 2018). For the Bible in Syriac, see Peshitta Institute, 1961. 442 critical editions, the corpus is now increasingly available online as digital images. This corpus, if we may call it such, is quite diverse. It encompasses books of the Hebrew Bible or Old Testament, and also of the New Testament. There are also deutero-canonical books. Of the manuscripts, some resisted the corruption of time as intact codices, while a not insignificant number only survive in a fragmentary state or as objects of reuse, today kept in public or ecclesiastic collections all over the world. Their time of production ranges from the 9th to the 20th centuries. Furthermore, some Arabic versions are of Jewish provenance, others of Christian or Samaritan origin. Each group created and maintained a corpus of biblical translations into Arabic, based on the various source texts (the Masoretic Text and the Greek New Testament, as well as Greek, Syriac, Coptic and Latin versions). In the course of history, these communities demonstrate a great variety in their translations and in a certain way they cultivated this. While previous research has tended to obscure the horizontal ties of translations, there is no doubt that specific translations were used by and transmitted among two or all three denominations at the same time.2 Not only did Jews, Christians and Samaritans share parts of their scriptures, they also often used the same translations, and they examined each other’s translations with curiosity and attention to detail. This nexus can be seen in multidimensional personal relationships, scholarly modes of interaction and circulations of biblical texts. In the course of diffusion from one contemporaneous cultural context to another, translations were often significantly transformed and adapted to the setting of the receptor community. Texts are passed on to and take root in contexts different from those in which they emerged, and thereby assume new meaning without being completely cut off from their original context. Changes also occurred in diachronic transmission over time. Older registers of Arabic or particular translation techniques may not have been understood by later copyists and readers and this therefore necessitated textual changes. As a result, there is an astonishing plurality of biblical versions in Arabic; the number of versions far exceeds all other translation traditions. The appearances of these manuscripts 2 Scholarly interaction between Jews and Christians in the realm of biblical translation long preceded the Islamicate world, and is known to have existed since Antiquity. These early interactions – which can be seen, for example, in the work of Origen – are important, since they underlie later discourses and would seem to provide a way of looking at the later interconnected scriptural history, after the Islamic expansion inaugurated the shift of Late Antiquity into the Umayyad and Abbasid periods; see for example Vollandt, 2016a. Further examples of scriptural entanglement can also be adduced from the Old Testament Peshitta found in the Syriac Christian tradition. There are many cases in which this Syriac text provides a translation in harmony with the usual Jewish interpretation of early Late Antiquity. The source of many, if not most, of these interpretations can be found in the targumim, which themselves represent Jewish versions of the Hebrew Bible translated into Aramaic dialects; see Brock, 1979b; Brock, 1982; and Brock, 1995. In the book of Proverbs, there is even a case where the Syriac text coincides almost verbatim with the Jewish Aramaic targum. It has been suggested that the Peshitta text of this book reflects a Jewish translation from northern Babylonia, which also provided the basis for the surviving Jewish targumim of Proverbs; see Weitzman, 1999. 443 are manifold, and to this one can add further distinctions as to the script used (Arabic, Hebrew, Syriac and rarely Greek), the writing materials, layout, the translation’s intended use, and so on. These Arabic versions of the Bible were shunned by the dominant strain of biblical scholars in the late 19th century and for much of the 20th century, as they were perceived to be of no value for textual criticism. In the view of these scholars, not only did these versions lack the primacy of age compared to earlier versions in Greek or Syriac, most versions were in fact of a tertiary rank, translated translations as it were.3 What is more, Arabic ceased to play a significant role in biblical scholarship after the discovery of ancient East Semitic and North-West Semitic languages, such as Akkadian and Ugaritic, in the course of the 19th and 20th centuries. Research on this topic likewise remained rare in Judaic Studies until the 1980s, although the scholars of the Wissenschaft des Judentums in 19th century Europe had dedicated a notable part of their research to the Arabic literature of the Jews in the medieval Islamicate world and their biblical versions in Arabic.4 Furthermore, the scholarly study of Arabic philology in Europe and elsewhere, in particular in the 20th century, gradually converged with the study of Islam. 5 Once these became almost synonymous, anything Jewish or Christian in that language was sidelined in this disciplinary setting as marginal.6 An additional reason for the marginalisation of Arabic versions of the Bible was an almost arbitrary historic preconception. Unlike in the fields of philosophy or the natural sciences, scholars of scripture held a belief that the Islamicate world of Late Antiquity led to, at best, an intellectual stagnation or, at worst, a complete state of tabula rasa among Jewish, Christian or Samaritan communities now under new rulers. The idea that Late Antique scriptural heritage could be seen as flourishing or even achieving an unprecedented moment of originality under Islamic hegemony and in Arabic, the literary koine, would have disrupted this preconception.7 3 Earlier scholars were less dismissive, as will be seen below. See Cohen, 1994, pp. 3–14; Polliack, 2006; Stillman, 2010. 5 Arabic studies, as well as Oriental studies generally, emerged as an ancillary to Biblical studies; see discussion in Bobzin, 1998. Arabic scholarship before the 20th century is far from centered predominantly on Islamic texts, as is demonstrated by the many examples in Toomer, 1996, Jones, 1988 and Fück, 1955. As a further illustration, Polaschegg, 2004, p. 83 points out that Paulus, 1790–1791 covers a variety of topics, from Muslim pilgrim reports and Hebrew ostraca to Syriac and Samaritan chronicles. The same holds true for Eichhorn, 1777–1786. In her view, “the attested, both phobic and affirmative, obsession with Islam” in Oriental studies is a phenomenon of the 20th century (p. 97). 6 Christian Oriental philologies, which could occupy themselves with Syriac or Christian-Arabic literature, have become a rare luxury at European universities. 7 Recent studies, such as those of Peter Brown and his students, have stressed the continuity between the Late Antique and Early Islamic periods; see also Fowden, 2014. Their perspective can be extended to exegetical practices and literary production among the communities now under Muslim rule. For a Jewish context, this can be illustrated by Geoffrey Khan’s studies on the Masorah in Khan, 2013. For a reassessment of the literary production in Greek, see Mavroudi, 2015 and Mango, 1999. 4 444 The Arabic Bible had thus become orphaned within scholarship, and remained so for most of last century. 1 Contested origins We cannot rule out the possibility that Jewish and Christian communities in the Arabian Peninsula orally translated parts of the Bible into Arabic in liturgical settings; indeed, this a likely scenario,8 and in fact the Qurʾān and early traditionalist literature offer us the strongest evidence for this.9 Biblical narratives often resound in this corpus as a kind of subtext, in a manner that presupposes a great familiarity with them on the part of the new community of believers that was coalescing in the formative years of Islam. They serve as a point of reference to corroborate Qurʾānic prophetology. Direct quotations, however, are strikingly absent. If we leave aside this intricate interplay with the text of the Qurʾān, the Bible makes its first appearance in Arabic in the writings of Christian apologists in the 8th century CE, although manuscripts of complete translations into Arabic are only found from the 9th century. It appears that social changes in the status of non-Muslims, which resulted from the ʿAbbāsid revolution (750 CE), encouraged the composition and textualisation of Arabic versions of the Bible.10 In contrast to the Umayyad period, by this time more extensive conversion to the new faith, Islam, appears to have taken place.11 Non-Muslims were becoming a well-defined legal category, whose rights and obligations were regulated by dhimma, a state of protection.12 These prescriptions regarding non-Muslim subjects in the first ʿAbbāsid century had obvious implications for the Jewish, Christian and Samaritan communities and their literature. The increased number of apologetic tracts, on the one hand, and the strenuous effort to translate biblical books, on the other, reflect these communities’ collective endeavours to respond to this new setting. In addition, committing Arabic biblical translations to writing was surely fostered by the general tendency, from the rise of the ʿAbbāsid caliphate onwards, to write down oral traditions. 8 See Vollandt, 2015, pp. 40–51; Griffith, 2013, pp. 7–53. Cf. Griffith, 2013. Apart from the Qurʾān itself and post-Qurʾānic literature, evidence of biblical knowledge among Arabs before the arrival of Islam can also be found in early Arabic poetry, for example that of Umayya ibn Abī al-Ṣalt, a contemporary of the Prophet; see Seidensticker, 2011 and Sinai, 2011. His poems draw heavily on episodes from biblical history and exemplify how biblical traditions were recounted in Arabic in the immediate milieu of the Qurʾān. Compare also the Christian poet ʿAdī ibn Zayd al-ʿIbādī (al-Ḥīra, died c600); see Dmitriev, 2009 and Toral-Niehoff, 2008. 10 As first pointed out by Griffith, 1985. 11 Bulliet, 1979. 12 See Levy-Rubin, 2011, which has now replaced older studies on this subject: Tritton, 1970; Fattal, 1958; and Noth, 1987. 9 445 While there is no corroboration of the existence of written translations before the rise of Islam or during its formative centuries, the view that such written translations must have existed had and continues to have its advocates.13 The premise that the prevalence of biblical motifs among early Muslims, alluded to above, must be linked to written translations that circulated is flawed.14 It presumes that the Qurʾānic revelation unfolded in a total rupture from the Late Antique scriptural heritage, creating intentionally or unintentionally an artificial amnesia of previous exegetical and translational practices. The seeming silent vacuum that this view postulates had to be bridged by assuming pre-existing translations into Arabic. What is more, such an assertion expresses interest in staking possessory claims. It carries inherently the charge of epigonality. Early Muslim literature, first and foremost the Qurʾān, becomes the mere product of Christian or Jewish influence.15 Turning now to the history of early modern and modern research on the Arabic Bible is too long to survey here in detail, so I will give only the main characteristics. It was in the form of early printings that European scholars first became aware of Arabic versions of the Bible on a large scale. Specimens of this achievement include the 1516 polyglot Psalter of Agostino Giustiniani, as well as the later Paris Polyglot and the London Polyglot.16 With minor exceptions, print editions remained the main focus of attention for scholars interested in Arabic biblical translations for much of the next two centuries.17 13 E.g. Baumstark, 1931; Baumstark, 1934. See also Rhode, 1921, p. 14; Algermissen, 1933, pp. 10–13; al-Maqdisī, 1933; Peters, 1936; Peters, 1940; Peters, 1942; Khoury, 1972, p. 258; Khoury, 1989; Shahid, 1984, p. 440; and Newby, 1988, p. 67, who even speaks of a “flood of translations”. All of these authors, often following Baumstark’s line of argumentation or referring to him directly, have no doubt that pre-Islamic translations existed. M. C. A. Macdonald and Corriente independently dated the Violet fragment (on which see below) to the pre-Islamic era – see Macdonald, 2009, vol. I, pp. 100–102, vol. III, pp. 50, 68 n62; and Corriente, 2007. In a paper written later, Macdonald reconsidered his earlier dating and corrected it to the Islamic period; see Macdonald, 2008. Kashouh, 2011, pp. 162–165 argues for a pre-Islamic dating of one of the oldest surviving manuscripts containing the Gospels and the Pauline Epistles (Vatican, BAV, MS Ar. 13) and suggested Najrān as its possible place of emergence. This scenario, however, was convincingly rejected by Griffith, 2013, pp. 114–118. 14 This premise is maintained to varying degrees by by Abbott, 1967, vol. II, p. 257; Cheikho, 1912–1923, vol. I, p. 254; Newby, 1988, p. 67; and Sprenger, 1869, p. 132. Possibly it is also implied in Samir, 2008, p. 159. 15 On the claimed epigonality of the Qurʾān, see for example Griffith, 2013, pp. 7–8, where he quotes Massignon’s famous dictum that the Qurʾān is nothing but “une edition arabe tronquée de la Bible”. Cf. Neuwirth, 2010, pp. 42–44. 16 Giustiniani, 1516; Lejay, 1628–1645; Walton, 1653–1657. 17 For example, the studies of Saadiah’s Tafsīr by Tychsen, 1782 and Schwarzstein, 1886 are based entirely on the Polyglots. The copy that Schwarzstein used had the shelf mark Df. 118 at the Badische Landesbibliothek Karlsruhe (http://ipac.blb-karlsruhe.de/index.php?img_id= 283346;nav_id=283351;cat_id=1;scroll=0) before it perished in the turmoil of war after 1942. Similarly limited to the texts in the Polyglots were studies of al-ʿAlam’s translation of the Prophets, such as Gesenius, 1820–1821 for Isaiah; Cornill, 1886 for Ezekiel; Wald, 1784, Gehman, 1925 and Löfgren, 1936 for Daniel; S. M. Reynolds, 1943 for Zechariah; Ryssel, 1885 for Micah; Reinke, 1867 for Nahum; and Reinke, 1868 for Haggai. ʿAbdallāh ibn al-Faḍl alAnṭāki’s version of the Psalms, found in the Polyglots, was studied by Döderlein, 1778–1779. 446 A new approach was inaugurated by Guidi’s 1888 study of the Arabic and Ethiopian versions of the Gospels and Vaccari’s studies in the early 1920s on the Arabic versions of the Prophets.18 They were the first to produce a comprehensive inventory of the available manuscripts.19 Both also introduced a comparative method for examining textual evidence. After an initial classification according to their Vorlagen, the versions were grouped in subcategories based on different branches of transmission. In particular, Guidi drew attention to the fact that textual changes frequently occur within a given version over the course of time, whether as the product of secondary revisions, adaptations or linguistic development. As a consequence, he stressed the necessity of introducing an additional distinction according to text types. As obvious as it may seem now, this approach was unprecedented and had never been applied to biblical translations into Arabic. Subsequently, Graf followed this method in his epochal Geschichte der christlichen arabischen Literatur.20 This work, while it is in need of additions and corrections in the light of recent research, remains the major reference for Christian Arabic versions of the Bible.21 Then in the 1960s, microfilms of the Arabic manuscripts of St. Catherine’s Monastery became widely available, after Kenneth W. Clark led an expedition to the Middle East under the auspices of the Library of Congress and its partners in 1949 to microfilm old manuscripts in various libraries of the Middle East, the largest and most pivotal of which was that at St. Catherine’s. Among these manuscripts were some early biblical codices that were unknown to Graf, and this prompted renewed scholarship in the study of the Arabic Bible of Christian provenance.22 In a similar manner, the major factor that has impacted recent scholarship on Jewish versions has been the hitherto unprecedented availability of new manuscript collections, such as the Cairo Genizah (from the 1980s onwards) and the Firkovich collections (after the fall of the Iron Curtain). Vaccari’s, Guidi’s and Graf’s comparative method has also been employed in a number of specialised inquiries focused on particular books.23 After an 18 Guidi, 1888; Vaccari, 1920; Vaccari, 1921; Vaccari, 1922. A number of authors before them described and analysed selected manuscripts: Adler, 1783– 1784; Paulus, 1789; Gildemeister, 1865. However, their work was rather dependent on the sources at their immediate disposal. 20 Graf, 1944, vol. I, pp. 85–195. 21 Samir, 1982 offered constructive criticism and suggestions for improving Graf’s work. 22 For example, see the four volumes of Staal, 1984; van Koningsveld, 1975; Leemhuis, 1984; Leemhuis, 1989; Leemhuis, Klijn and van Gelder, 1986; Drint, 1997; Drint, 1999. Some photographs of manuscripts in Graf’s Nachlass, kept at the Bayerische Staatsbibliothek in Munich, indicate that he only had very limited access to manuscript collections in the Sinai, and then only at a late stage in his life. Graf seems to have derived his information on Sinaitic manuscripts mostly from the Studia sinaitica series. Apart from that, Graf was acquainted with a number of membra disjecta from Sinai that circulated in Europe, and especially in Germany, as for instance the Grote collection, on which see Tarras, forthcoming. 23 For example, Rhode, 1921 examined the Pentateuch translations employed in the Coptic Church; Polliack, 1997 has looked at Qaraite translations of the Pentateuch; Vollandt, 2015 19 447 initial classification according to their Vorlagen, all of these studies group each version into subcategories. However research on the diverse versions which have been found has taken different directions, which will be briefly surveyed in the following sections. 2 Textual criticism The value of Arabic versions of the Bible for textual criticism was generally accepted by the Republic of Letters and is stressed, for example, in the 1524 publication Oratio de laudibus & utilitate trium linguarum by Robert Wakefield, the founding father of Hebrew studies in Renaissance England.24 Franciscus Junius the Elder, too, relied on Arabic versions for his New Testament textual criticism. In 1578 he issued a Latin translation of an Arabic translation of the Acts of the Apostles and the Epistles to the Corinthians, to which he appended a discussion of the value of the Arabic variants for the received Greek text.25 His translation of and comments on 1 John, Galatians and Hebrews remained in manuscript. A similar approach was pursued by Petrus Kirstenius in his 1608 book, Vitae quatuor Evangelistarum, ex antiquissimo codice Arabico Caesario, and also in his later Notae in Evangelium S. Matthaei, ex collatione textuum Arabicorum Aegyptiacorum, Hebraeorum, Syriacorum, Graecorum, Latinorum.26 Similar statements of the importance of Arabic versions are found in Gabriel Sionita and Joannes Hesronita’s Grammatica Arabica Maronitarum and in the preface to Thomas Erpenius’ Pentateuchus Mosis Arabicè.27 Arabic versions were also occasionally used in critical editions of the Hebrew and Greek Scriptures in the 19th century. Holmes’ 1822 work, Vetus Testamentum ex versione Septuaginta interpretum acknowledges its use of four unspecified Arabic versions from the Bodleian Library. Von Tischendorf employed Arabic versions of the Gospels to improve his 1849 edition of the Novum Testamentum Graece.28 The manuscripts he used include Vatican, BAV, MS Ar. 13 (referred to with the siglum ArVat), which is an early Arabic version of the Gospels, and St. Petersburg, Institute of Oriental Manuscripts, analyses the Pentateuch generally; Löfgren, 1936 and Hjälm, 2016 focus on Arabic versions of the book of Daniel. Other examples include Knutsson, 1974 for Judges; Madros, 1984 for Psalms; Samaan, 1994 for Ecclesiasticus; Bengtsson, 1995 and Bengtsson, 2003 for Ruth; Blackburn, 1999 for Job; and Kashouh, 2011 for the Gospels. 24 Wakefield, 1989. 25 Junius, 1578. For details, see Hamilton, 1985, p. 81; Smitskamp, 1976, pp. 119–120; de Nave, 1986, pp. 100–101. 26 Kirstenius, 1608; Kirstenius, 1611. Petrus Kirstenius based his investigations on Vienna, ÖNB, MS N.F. 97, which contains an Arabic translation of the Gospels with many text-critical marginalia. His personal copy of the preface and epilogue of this manuscript are now Hamburg, Stadtbibliothek, MS Or. 27. 27 Sionita and Hesronita, 1616, sig. aii r–aiiv; Erpenius, 1622. 28 Von Tischendorf, 1849. 448 MS D 226 (with siglum ArPet). A full list of Arabic manuscripts used is found in the third volume of the Novum Testamentum Graece published by Gregory in 1894.29 The first four editions of Alford’s New Testament also presented variants taken from Arabic translations.30 However, following criticism that they were “of very little use in the present stage of critical investigation of the text”, later editions dispensed with them.31 The apparatus of Kittel’s Biblia Hebraica also contains references to the Arabic parts of the Paris Polyglot.32 However, 19th and 20th century biblical scholarship was generally unfavourably disposed to the study of biblical versions in Arabic. De Lagarde wrote that “there are more Arabic translations of the Gospels than theologians pressed with more urgent priorities can care about; their value is comparatively limited”.33 In the same vein, Nestle, in the Realencyklopädie für protestantische Theologie und Kirche, conjectured that Arabic versions “are not worth much for biblical criticism and exegesis, because, with only minor exceptions, they are secondary translations”.34 Margoliouth called them “of the slightest possible importance”.35 Tregelles added: “The Arabic versions existing in MS. exhibit very various forms: it appears as if alterations had been made in the different countries in which they had been used; hence it appears an endless task to discriminate amongst them precisely.”36 Writing in 1957, Roberts had no doubt that biblical translations in Arabic are “at most of secondary value for the study of the biblical text”.37 On the other hand, also in the 20th century, Levin allows that Arabic translations have a “textual significance”.38 Peters stresses the text-critical importance of some translations, such as that by al-Ḥārith b. Sinān for the Hexapla, Origen’s opus magnum of the Greek Old Testament, in which he laid out six parallel columns with different texts across each or that by Isaac Velasquez for the Latin Gospels.39 The value of al-ʿAlam’s translation for the study of the Alexandrian text type has been pointed out frequently.40 29 Gregory, 1894. Alford, 1859–1870. 31 Empson, 1851, p. 29; see Davidson, 1852. 32 Kittel, 1905–1906. 33 De Lagard, 1864, p. 1: “Arabische übersetzungen der evangelien giebt es mehr, als der mit drängenden arbeiten überhäuften theologie lieb sein kann, ihr werth is verhältnissmässig gering [sic]”. 34 Nestle, 1897, p. 91: “Für die Biblische Kritik und Exegese haben sie nur wenig Wert, da sie mit wenigen Ausnahmen Tochterübersetzungen sind”. 35 As quoted in Jellicoe, 1968, p. 267. 36 Tregelles, 1893, p. 1615. 37 Roberts, 1957, p. 1, cols 1200–1201: “bestenfalls nur von zweitrangigem Wert für die Untersuchung des Bibeltextes”. 38 Levin, 1938, p. 1: “textgeschichtliche Bedeutung”. 39 Peters, 1942. 40 See e.g. Graf, 1944, vol. I, pp. 132–133; Vaccari, 1921; Gesenius, 1820–1821, vol. I, pp. 98– 106; Cornill, 1886, pp. 49–56; Wald, 1784; Gehman, 1925; Löfgren, 1936; S. M. Reynolds, 1943; Ryssel, 1885; Reinke, 1867, pp. 65–70; Reinke, 1868, pp. 34–37. See also the discussion in Wevers, 1970, pp. 8–11. 30 449 3 Printed editions A number of printed editions exist, some of them from the early modern period. As mentioned in section 1, the orientalist Agostino Giustiniani (1470– 1536), who was bishop of Nebbio, issued a polyglot Psalter in 1516.41 Its eight columns (whence its designation Psalterium Octaplum) contain the Masoretic Text, a Latin translation of that text, the Vulgate, the Septuagint, an Arabic version, the Targum Onkelos with a Latin paraphrase and running scholia. Eliezer Soncino published a polyglot Pentateuch in 1546 that became known as the Constantinople Polyglot, which contained the Hebrew text accompanied by Targum Onkelos, Saadiah’s Tafsīr, a Judaeo-Persian translation by Jacob ben Josef Ṭāwūs and Rashi’s commentary.42 Eliezer’s edition of Saadiah’s Tafsīr holds the distinction of being the earliest printed Judaeo-Arabic text as well as the first printed Arabic Pentateuch. However despite the importance of Eliezer’s undertaking, few details are known concerning its production, the size of the print run or its distribution.43 A psalter in Karshūnī and Syriac was published in 1610 at the Monastery of Mar Antonius, Quzḥayya, supervised and sponsored by Sergius al-Rizzī, alumnus of the Maronite College in Rome.44 Another important work is the Pentateuchus Mosis Arabicè, printed in 1622; it is of rather modest size in comparison to the Constantinople and Paris Polyglots, which it falls between chronologically. However Erpenius, its initiator, was one of the foremost Arabists of his time.45 His edition was based on a manuscript that had been owned by Scaliger, today kept under the shelf-mark Leiden, University Library, MS Or. 236, which represents later North African traditions of Judaeo-Arabic biblical translation, generally grouped under the term shurūḥ (sing. sharḥ). In 1632 the sixth volume of the Paris Polyglot appeared, with the Peshiṭta and a translation on the left-hand pages and the Arabic text with Latin on the right-hand pages.46 The Arabic version is based on MS Paris, Ar. 1.47 The London Polyglot, dedicated to Oliver Cromwell, was the last and greatest of the Polyglots, being completed in 1657.48 The Arabic portions, however, simply reproduced the text of the Paris Polyglot, with minor changes. In 1867, Paul de Lagarde edited Leiden, University Library, MS Or. 377, a manuscript that contains Saadiah’s translation of the Pentateuch in a Christian See Vercellin, 2001, pp. 70–75. On Giustiniani’s biography, see Bobzin, 1990. Soncino, 1546. 43 Prestigious parchment copies are attested; see Freimann, 1901, p. 56. The edition was popular in Egypt, as illustrated by the numerous fragments in the Genizah, e.g. CUL MSS T-S NS 214.74, 266.61, 267.5, 267.210, 269.32, etc. 44 Al-Rizzī, 1610. 45 Erpenius, 1622. A full-scale biography of Erpenius is not available. The fullest account is found in Juynboll, 1931, pp. 59–118, but this is in need of additions and corrections in the light of recent research. 46 Lejay, 1628–1645. Cf. Vollandt, 2012. 47 On this manuscript see Vollandt, 2016b. 48 Walton, 1653–1657. On the London Polyglot, see Toomer, 1996, pp. 202–210; Miller, 2001. 41 42 450 recension.49 Gibson and Stenij provided the first editions of manuscripts that originated from St. Catherine’s Monastery at Sinai, both exhibiting parts of the Pauline Epistles.50 With the microfilms having become available in the 1960s, a number of editions based on this collection ensued, with editions of: MS Sinai, Ar. 151, containing the Acts, Pauline and Catholic Epistles; the book of Ben Sira in MS Sinai, Ar. 155; and MS Sinai, Ar. 589 with the Apocalypse of Baruch and IV Ezra.51 More recently, further editions from Eastern manuscripts have included, for instance, Arbache’s edition of Luke from MS Sinai, Ar. 72 and Monferrer-Sala’s edition of Philemon from MS Vatican, BAV, Ar. 13.52 Further, Monferrer-Sala has edited Revelation and the three Johannine Epistles of MS Madrid, El Escorial, Ar. 1625; Bonhome Pulido has edited Galatians from the same manuscript; and Potthast edited Romans from MS Madrid, BNE, Or. 497153 – all of these are Andalusian versions. The editors have usually appended their editions with studies of the Vorlagen as well as the linguistic features of the texts. However, all of them share one characteristic: they are based on a single manuscript and flatten complex histories of transmission, often manifested in quite a substantial number of copies, into something rather one-dimensional. Another approach has been proposed by Knutson in his edition of three Arabic versions of the Book of Judges in 1976.54 He added a critical apparatus that includes readings from all other known manuscripts. The same is the case for Bengtsson’s edition of the Arabic versions of Ruth.55 Both Samaritan-Arabic versions of the Pentateuch were published by Shehadeh in 1989 and 2002 – the older, which he calls the Old Arabic Translation of the Samaritan Pentateuch (and which dates to before the second half of the 13th century) and the later, revised version of Abū Saʿīd (13th century Egypt).56 Importatly, the edition once again takes into account all known copies of the texts. Among recent editions, equally comprehensive in their use of sources, we find Yefet ben ʿElī’s commentary and translation on the books of Ruth, Jonah, Hosea and Obadiah, Esther, Jeremiah, and Proverbs, and also passages from the book of Genesis.57 The critical edition of Hibat Allāh ibn al-’Assāl’s translation of the Gospels was published by Samuel Moawad in 2014, including the text-critical 49 De Lagarde, 1867, vol. I. See also Hughes, 1914 and Vollandt, 2015, p. 225. Gibson, 1894; Stenij, 1901. 51 For example, see the four volumes of Staal, 1984; Frank, 1974; Leemhuis, Klijn and van Gelder, 1986; Drint, 1997. 52 Arbache, 2012; Monferrer-Sala, 2015. 53 Monferrer-Sala, 2000; Monferrer-Sala, 2002; Bonhome Pulido, 2013; Potthast, 2011. 54 Knutsson, 1974. 55 Bengtsson, 1995. 56 Shehadeh, 1989; Shehadeh, 2002. 57 See Butbul, 2003 for the book of Ruth; Andruss, 2007 for Jonah; Polliack and Schlossberg, 2001 for Obadiah; Polliack and Schlossberg, 2009 for Hosea; Wechsler, 2008 for Esther; Sabih, 2009 for Jeremiah; Sasson, 2016 for Proverbs; Ben-Shammai, Sklare, Batat, Butbul and Strousma, 2000, for Genesis; and Zawanowska, 2012 also for Genesis. The second volume of Sasson’s work is to appear, as is a work by Sadan on Yefet’s book of Job. 50 451 marginalia and based on the oldest eight manuscripts.58 A team, consisting of C. Adang, Miriam L. Hjälm, J. P. Monferrer-Sala, S. Schmidtke and myself is presently working on a critical edition of an early Melkite version of the Pentateuch in Arabic that was produced by al-Ḥārith b. Sinān b. Sunbaṭ alḤarrānī. He used as the basis of his translation the fifth column of the Hexapla, which contained Origen’s revision of the text of the Septuagint, and endowed his translation with a set of introductory chapters and a hexaplaric apparatus. Sara Schulthess has suggested in recent years new ways of editing in a digital age.59 In two sample projects, the texts of 1 Corinthians, as found in MS Vatican, Ar. 13 (http://tarsian.vital-it.ch/) and the Acts, the Catholics letters and the Pauline letters in Greek, Latin and Arabic, as found in MS Venice, Marciana, Gr. Z. 11 (http://humarec-viewer.vital-it.ch/), have been made available to the scholarly public. Despite this quite active development in editing over the past few decades, it needs to be stressed that there is still much to do. For example, there are no critical editions of some the most central and influential translations, such as Saadiah’s Tafsīr on the Pentateuch.60 A complication emerges for translations with no attributions. While some translations have a clear authorial voice – such as in the translations of Saadiah (882–942), al-Harith ibn Sinan (first half of the 10th century) or Bishr ibn alSirri (9th century), to name just a few – for most translations there was probably never a single translator. A large majority of translations are anonymous and of a provenance that only further research will perhaps discern. In addition, their textuality is fluid due to a combination of authorial anonymity and a high degree of variation. Sometimes differences between the extant manuscripts of a translation are so great that we are obliged to view them as representing separate versions or redactions. Occasionally these versions or recensions are so different that, even while showing clear textual affinities, it is impossible to imagine how they could go back to a single original. We have to see them as representing separate manifestations of an underlying oral transmission. A comparison of surviving manuscripts in order to identify and eliminate those features of their texts which pertain to scribal interference rather than to an authorial figure, traditionally the task of an editor, is virtually impossible. Trying to reverse “the process of transmission and restore the words of the ancients as closely as possible to their original form” seems a futile task.61 The term mouvance has been used to describe this textual mobility, and it calls for a particular editorial practice, carried out, for example, in a 58 Most of these Gospels represent Family (La) in Kashouh, 2011, pp. 258–274; see also Moawad, 2014, p. xxxix. On the marginalia, see Vollandt, 2015. 59 Schulthess, 2012; Schulthess, 2013; Clivaz, Schulthess and Sankar, 2017. 60 A critical edition has been announced by Eliezer Schlossberg (Bar-Ilan University). New sources towards this new edition have been presented by Avishur, 1992 and Blau, 1998. 61 L. D. Reynolds and Wilson, 1974, p. 212. 452 non-stemmatic, synoptic way.62 It remains to be seen whether digital editions can also prove to be the right tool for these traditions. 4 Language It is clear that there is no single direction of research in modern scholarship on the Arabic Bible. Biblical scholarship, as has been shown above, completely rejected the usefulness of studying Arabic translations for gaining insights into the text(s) of the Bible itself. By contrast, a number of scholars, including Levin, Knutsson, Bengtsson and, most recently, Dikken, have proposed concentrating on the Middle Arabic features in these texts.63 This approach is best captured in Knutsson’s statement that the text-critical aspect of the Arabic versions of the Pentateuch is only one amongst many. … The Arabic versions deserve to be examined from several other viewpoints, of which the purely linguistic one is the most important. 64 The Violet fragment, a bilingual Greek-Arabic fragment of Psalm 78:20–31, 51–61 (LXX 77) which was found at the end of the 19th century in the Qubbat al-Khazna (the Treasure Dome) in the compound of the Umayyad mosque in Damascus and was first published by Violet in 1901, instigated not a little scholarly attention.65 Yet of greatest interest was the fact that the Arabic column was written in Greek letters, and it was therefore of linguistic importance.66 The same fragment in Arabic script would have hardly caused comparable consideration. A particular strand of studies on Saadiah’s Tafsīr is similarly primarily concerned with the linguistic aspects of this work.67 Early ‘non-Saadianic’ translations, and also Jewish translations that emerged from the 14th century onwards (the so-called shurūḥ), have increasingly attracted the attention of scholars during recent decades, and the same holds true for them.68 62 On mouvance see Zumthor, 1972. This idea can be compared to other textual traditions with shared characteristics and their synoptic editions, e.g. Schäfer and Becker, 1995; Schäfer, Schlüter and von Mutius, 1981; or Bumke, 1996. 63 Levin, 1938; Knutsson, 1974; Bengtsson, 1995; Dikken, 2012. 64 Knutsson, 1974, p. 4. 65 It is known as the Violet Fragment in honour of its discoverer. See Violet, 1901. 66 Mavroudi, 2008; Hopkins, 1984, pp. 1–2; Blau, 2002, p. 68; Blau, 1967, vol I, p. 31 (his grammar usually excludes Bible translations); Haddad, 1992; M. C. A. Macdonald, 2009, vol. I, pp. 100–102, vol. III, pp. 50, 68 n 62; M. C. A. Macdonald, 2008; Corriente, 2007. See also alJallad, forthcoming. 67 E.g. Blau, 1998; Zewi, 1997a; Zewi, 1997b; Zewi, 2000; Zewi, 2001; Zewi, 2002; Zewi, 2014; Zewi, 2016. 68 The first fragment, CUL T-S Ar. 53.8, an early fragment exhibiting a translation of the book of Proverbs, was published by Blau, 1992. The discovery of additional fragments has been announced; a study of their possible Vorlage was furnished by Hopkins, 2002. Extant translations from this period cover most of the Torah and, to a lesser degree, portions of the Prophets and 453 The linguistic aspect is certainly important, and should not be neglected. However, biblical translations often follow a grammar of their own, which is governed by a wish to imitate the exalted source text and maintain a high degree of literalism in the translation. Furthermore, this literalism played a functional role in instruction and emerged directly out of the didactic need to mirror the source text as closely as possible, as is the case for biblical translations in other languages as well. The language used could often be described as a professional translation language, in which frequently the choices made by translators are overly literal, imitative of the source text and expressed in a grammatically perplexing Arabic style that was not employed in any other literary genre. This means that these biblical translations reflect only a rather specific register of Middle Arabic. Thus the concentration on the linguistic aspects alone limits and undermines the historical significance of these translations. 5 Translation techniques It has been emphasised in the past that a focus by scholars on linguistic features or on textual criticism often comes at the expense of analysing translation techniques and studying how a particular version is embedded in the broader context of the related theological, exegetical and grammatical traditions of which biblical versions have always been an inextricable part.69 Equally neglected in research has been their embeddedness in liturgy, education or apologetics.70 These contexts must have strongly conditioned the strategies which translators used to transfer particular structures, proper names or concepts from the source language into the target language. Polliack’s 1997 book on Qaraite translations not only offers a clear methodology for describing translation techniques, but also situates Arabic Bible translations in a larger exegetical context, in which they, together with running commentaries and linguistic thought (grammars and dictionaries), form what can be called an exegetical triangle.71 While earlier research addresses the underlying principles of translating only sparsely and selectively, Polliack starts from the relationship between translation and Vorlage, and describes in a systematic way the strategy employed by the translator to transfer particular structures, concepts or ideas from the source language into the target language. Writings; see Blau and Hopkins, 2017. As an illustration of the linguistic approach with regard to the shurūḥ, see Hary, 2009. 69 There is a great deal of literature, too abundant to cover here, on the translation techniques of other biblical versions. As examples, see Barr, 1979 for general discussion; Aejmelaeus, 1982 and Tov, 1979 for the Greek Bible; Brock, 1979a, Brock, 1983 and Szpek, 1992 for the Syriac Bible. 70 Griffith, 2013 is a recent and rare exception. 71 Polliack, 1997, pp. 3–22. 454 Polliack has served as a valuable model for my own study on the Arabic versions of the Pentateuch, and also for Hjälm’s on the book of Daniel and Bengtsson’s on the book of Ruth.72 The translation techniques examined in the above-mentioned studies are organised under the headings of syntax, vocabulary, particles and morphemes, and style. The first three of these address decreasing textual units of the source text: from single verses or a cluster of verses, to single lexemes in a verse and smaller grammatical units. Style concentrates on stylistic modifications and paraphrases, such as additions, omissions and substitutions. 6 Paratextual approach Some studies go beyond a mere critical presentation of the text of a given translation, its language or translation technique.73 The introductions of translators, if they are available, serve as an important source for their intention and strategies. Saadiah’s introductions, for example, have received much attention.74 Al-Ḥārith b. Sinān added an introductory tractate to his translation, referred to as a risāla ‘epistle’, in which he goes into great detail about the earlier Jewish biblical translations.75 After retelling the narrative from the Letter of Aristeas, al-Ḥārith goes on to describe Origen’s Hexapla, including its arrangement in columns, the content of each column and the text-critical apparatus used to indicate variants between the Septuagint and the translations of Aquila, Symmachus and Theodotion. In this work there are also introductions to each book of the Pentateuch, which, following the custom of Syro-hexaplaric manuscripts, provide a short summary of the contents. Some manuscripts contain the original hexaplaric readings, including the so-called Aristarchian symbols.76 Important information about the use of manuscripts can also be contained in non-authorial paratexts, such as prefaces by readers,77 edificatory proems,78 72 Vollandt, 2015; Hjälm, 2016; Bengtsson, 2003. For a similar approach to the Greek Bible, see Wallraff and Andrist, 2015. 74 E.g. Stroumsa, 2007; Ben-Shammai, 2000; and Polliack, 1997, pp. 77–90. 75 The introduction has been partly edited and translated into Latin. See Aldrich, 1692; Hody, 1705, pp. 622–625; White, 1779, pp. 8–29. 76 On the marginalia, see Monferrer-Sala, 2017. 77 See Vollandt, 2015, pp. 9–11; Vollandt, 2016a; and Vollandt, forthcoming. 78 Some Christian manuscripts of Saadiah’s translation are preceded by an edificatory proem which elaborates on abrogation of Mosaic Law (Arab. al-sharīʿa al-musawiyya) – that is, the abrogation of the Torah – by the New Testament; it also contains a short summary of the contents, referred to as the ‘study guide’ (Arab. dallāl). The manuscripts close with an account, called an ‘epilogue’ (Arab. al-khātima) on how the Hebrew scriptures were handed down in an authoritative, unbroken line of transmitters, until they were eventually translated into a variety of languages and thus became corrupted. On this, see Vollandt, 2016c. 73 455 liturgical notes79 or text-critical marginalia.80 Glosses and notes can often illuminate the use of the manuscript; for example, liturgical marks might demonstrate their use in liturgy, or annotations by readers might be suggestive of a private study Bible. As manuscripts are physical objects that continue to exist through time, they are disseminated and consumed in ways which are also socially, economically and intellectually determined. This leaves traces on them, so that one might say that scripture grows along with its readers and through interaction with the reading community. For this reason, paratexts must be an import direction in future research. Paratextual elements provide contextualising evidence that is largely absent from the texts themselves. 7 Outlook One of the major impediments which currently holds back research on the Arabic Bible – whether that research focuses on textual criticism, critical editions, the linguistic features of the text, translation techniques or paratextual features – is the difficulty scholars have in becoming aware of and accessing the relevant manuscripts. To study any particular biblical book, every scholar needs to begin with the cumbersome and time-consuming task of sifting through the manuscript material, which demands a fair amount of detective work and archival skill. This process would be assisted by an online union catalogue of Arabic Bible manuscripts, which organises the translations into a clavis based on the biblical books they cover. To illustrate what this would involve, consider MS Paris, BnF, Ar. 1. The manuscript contains an almost complete Old Testament. A union catalogue would need to capture and describe in detail all textual as well as codicological units. On the one hand, the data on codicological features would include the multiple colophons from Ramaḍān 992 AH to Muḥarram 993 AH (which correspond to the period between September 1584 and January 1585 CE), and would describe the four different scribes who can be distinguished. On the other hand, the data on textual units would include a list of all books contained in the copy, including additional information on paratexts (for example, the introductions to the Pentateuch on ff. 1 v.–3 r. and the book of Daniel on ff. 346 v.–347 v.) and the extensive glosses. The catalogue would provide, into the clavis, all the basic information on the particular translations 79 See Baumstark, 1929–1930, and more recently, Zaki, 2017. In 1252, al-Asʿad Abū al-Faraj Hibatallāh ibn al-ʿAssāl produced a critical revision of the Arabic Gospels that were in use among the Copts. Al-Asʿad noted the variants among the different manuscripts, and also gave text-critical notes, in the margins of his text with a set of marks, called ʿalāmāt ‘sigla’ in Arabic; cf. D. B. Macdonald, 1904. The text-critical notes were included in Moawad, 2014. See Vollandt, 2016a and Vollandt, forthcoming for further manuscripts containing such an apparatus. 80 456 exhibited in the manuscript; for example, that the manuscript contains a version of Saadiah Gaon’s translation of the Pentateuch (ff. 3 v.–83 v.) and an anonymous translation of 2 Maccabees (ff. 429 v.–439 r.). A comprehensive searchable online database of all known manuscripts is currently being developed by the Biblia Arabica team in Munich.81 This highly accessible database will enable users to identify, locate and cite any translation (through the clavis) or one of its manuscript embodiments (through the union catalogue), with full documentation and without demanding expert knowledge to sift through existing catalogues and inventories. In addition to information on the text of the translation, the union catalogue will collect – as described above – codicological and paleographical data, and also describe paratextual elements (translators’ introductions, prefaces by later readers or copyists, interpretative or text-critical glosses, liturgical marks, ownership) in great detail, enabling scholars to browse through the catalogue to answer various research questions. This data will establish a firm (eventually definitive) corpus, which will enable quantitative research. The database will allow navigation diachronically (e.g. search for all identified manuscripts of a particular translation, from the earliest to the latest copy) and synchronically (e.g. search for all translations of a particular biblical book, be it of Jewish, Christian or Samaritan provenance). A digital catalogue and clavis such as these for Arabic Bible manuscripts will provide a single starting point for manuscript research and direct users to each of the repositories offering manuscript images, cataloguing information and related bibliography.82 The quantitative data gathered in the union catalogue and clavis will, furthermore, invite the investigation of chronologies and canonisation processes, thus revisiting what has been observed on a much smaller and less systematic scale in the study of particular groups of manuscripts. It will make it possible to determine when a particular translation emerged (earliest dated copy) and until when remained in use (latest dated copy). Scholars of the Arabic Bible believe that communities employed multiple and complementary versions side by side, with no rivalry among them. There does not seem to have ever been a binding canon of translations. This assumption will now be able to be 81 The URL will be www.biblia-arabica.com/clavis. Biblia Arabica – The Bible in Arabic, is a DFG-DIP funded project, co-directed by Camilla Adang (Tel Aviv University), Meira Polliack (Tel Aviv University), Andreas Kaplony (LMU, Munich) and myself (LMU, Munich), as well as formerly Sabine Schmidtke (now IAS, Princeton). Current members of the team in Munich, involved in the development of the database, are Nathan Gibson, Peter Tarras and Vevian Zaki, and previously also Miriam L. Hjälm. 82 The Biblia Arabica team in Munich, together with partners in Tel Aviv and independent collaborators, has been preparing an online Bibliography of the Arabic Bible: A Classified and Annotated History of Scholarship (http://biblia-arabica.com/bibl). This not only updates the items cited by Graf relating to Christian Arabic translations, but also includes Jewish and Samaritan translations and the post-Qurʾānic Muslim reception of the Bible. Each bibliographic item has an entry displaying a full reference, a summary of the content, the manuscripts mentioned, a digital identifier (Uniform Resource Identifier or URI) and links to open-access online versions of the item where available. 457 tested in a quantitative manner, since it seems that some versions have survived in many more exemplars than others, suggesting perhaps that each community had its preferred and quasi-canonical version. The union catalogue and associated clavis are important for another reason. They will future-proof and make durable collective efforts in research. By making these resources online and open-access, they also become available to scholars in related fields, such as Biblical Studies, Judaic Studies, Islamic and Arabic Studies and the study of Eastern Christianity. This will integrate individual findings into a much larger scholarly context, and thus create new synergies for future research. Finally, the union catalogue and clavis will encourage diverse heritage communities beyond the religious and linguistic divides to access the cultural archive of biblical translations into Arabic. References Abbott, N., 1967, Studies in Arabic Literary Papyri, Chicago: University of Chicago Press. Adler, J. G. C., 1783–1784, Kurze Übersicht seiner biblisch-kritischen Reise nach Rom, Altona: J. D. A. Eckhardt. Aejmelaeus, A., 1982, Parataxis in the Septuagint: A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch, Helsinki: Suomalainen Tiedeakatemia. Aland, K., 1975, Repertorium der griechischen christlichen Papyri I. Biblische Papyri: Altes Testament, Neues Testament, Varia, Apokryphen, Berlin: Walter de Gruyter. ———, Welte, M., Köster, B. and Junack, K., 1994, Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments, Berlin: De Gruyter. Aldrich, H., 1692, “Testimonia aliquot Scriptorum Orientalium de LXXII interpretibus eorumque Versione,” in H. Aldrich, Aristeae Historia LXXII interpretum accessere veterum testimonia de eorum versione, Oxford: e Theatro Sheldoniano, pp. 131–144. Alford, H. (ed.), 1859–1870, The Greek Testament, London: Rivingtons. Algermissen, E., 1933, Die Pentateuchzitate Ibn Ḥazms: Ein Beitrag zur Geschichte der arabischen Bibelübersetzungen (unpubl. diss., Münster). al-Jallad, A., forthcoming, The Damascus Psalm Fragment: Middle Arabic and the Legacy of Old Ḥigāzī. al-Maqdisī, A.,1933, “Naql al-Kutub al-Muqaddasa ilā al-ʿArabiyya qabla al-Islām”, al-Mashriq 31, pp. 1–12. Andruss, J. H., 2007, The Judaeo-Arabic Commentary on Jonah by the Karaite Japheth Ben Eli: Introduction and Translation (unpubl. diss, The Ohio State University), http://rave.ohiolink.edu/etdc/view?acc_num=osu1232800114. Arbache, S., 2012, L’Évangile arabe selon Luc, Bruxelles: Éditions Safran, http://hdl.handle.net/2078.1/156889. Avishur, Y., 1992, “Some new sources for the study of the text and language of Saadya’s translation of the Pentateuch into Judaeo-Arabic”, in J. Blau and S. Reif (eds), Genizah Research after Ninety Years, the Case of Judaeo-Arabic: Papers Read at the Third Congress of the Society for Judaeo-Arabic Studies, Cambridge: CUP, pp. 5–13. 458 Barr, J., 1979, The Typology of Literalism in Ancient Biblical Translations, Göttingen: Vandenhoek & Ruprecht. Baumstark, A., 1929–1930, “Die sonntägliche Evangelienlesung im vorbyzantinischen Jerusalem”, Byzantinische Zeitschrift 30, pp. 350–359. ———, 1931, “Das Problem eines vor-islamischen christlich-kirchlichen Schrifttums in arabischer Sprache”, Islamica 4, pp. 565–566. ———, 1934, “Arabische Übersetzung eines altsyrischen Evangelientextes und die in Sure 21:105 zitierte Psalmenübersetzung”, Oriens Christianus 9/3, pp. 165– 188. Bengtsson, P. Å., 1995, Two Arabic Versions of the Book of Ruth: Text Edition and Language Studies, Lund: Lund University Press. ———, 2003, Translation Techniques in Two Syro-Arabic Versions of Ruth, Stockholm: Almqvist & Wiksell. Ben-Shammai, H., 2000, ‫ “ההקדמה הגדולה” ו”ההקדמה הקטנה” לתרגום‬:‫חדשים גם ישנים‬ ‫רס”ג לתורה‬, Tarbiz 69, pp. 199–210. ———, Sklare, D. E., Batat, E., Butbul, S. and Strousma, S. (eds), 2000, “Yefet ben ʿEli, commentary on Genesis: a sample critical edition” (Hebrew), in Judaeo-Arabic Manuscripts in the Firkovitch Collections: Yefet ben ʿEli al-Basri, Commentary on Genesis: A Sample Catalogue; Texts and Studies, pp. 79–179, Jerusalem: Ben-Zvi Institute. Blackburn, S., 1999, The Early Arabic Versions of Job (First Millennium C.E.) (unpubl. diss., University of St Andrews). Blau, J., 1967, A Grammar of Christian Arabic, Louvain: Secrétariat du Corpus SCO. ———, 1992, “On a fragment of the oldest Judaeo-Arabic Bible translation extant”, in J. Blau and S. Reif (eds), Genizah Research after Ninety Years: The Case of Judaeo-Arabic, Cambridge: CUP, pp. 31–39. ———, 1998, “Saadya Gaon’s Pentateuch translation in light of an early-eleventhcentury Egyptian manuscript” (Hebrew), Leshonenu 11, p. 111. ———, 2002, A Handbook of Early Middle Arabic, Jerusalem: Max Schloessinger Memorial Foundation and the Hebrew University of Jerusalem. ——— and Hopkins, S., 2017, Early Judaeo-Arabic in Phonetic Spelling: Texts from the First Millennium (2 vols), Jerusalem: Ben Zvi Institute. Bobzin, H., 1990, “Agostino Giustiniani (1470–1536) und seine Bedeutung für die Geschichte der Arabistik”, in W. Diem and A. Falaturi (eds), XXIV. Deutscher Orientalistentag vom 26. bis 30. September 1988 in Köln. Ausgewählte Vorträge, Stuttgart: F. Steiner Verlag Wiesbaden, pp. 131–139. ———, 1998, “Vom Sinn des Arabischstudiums im Sprachkanon der Philologia Sacra”, Hallesche Beiträge zur Orientwissenschaft 24, pp. 21–32. Bonhome Pulido, L., 2013, “A fragmentary Arabic version of the Epistle of Galatians kept in the royal library of the monastery of the Escorial”, in S. K. Samir and J. P. Monferrer-Sala (eds), Graeco-Latina et Orientalia: Studia in honorem Angeli Urbani heptagenarii, Cordoba: CNERU, vol. II, 35–47. Brock, S., 1979a, “Aspects of translation technique in antiquity”, Greek-Roman and Byzantine Studies 20, pp. 69–87. ———, 1979b, “Jewish tradition in Syriac sources,” JJS 30, pp. 212–232. ———, 1982, “Some Syriac legends concerning Moses”, JJS 33, pp. 237–255. ———, 1983, “Towards a history of Syriac translation technique”, in R. Levenant (ed.), III Symposium Syriacum 1980. Les contacts du monde syriaque avec les autres cultures, Rome: Pont. Institutum Studiorum Orientalium, pp. 1–14. ———, 1995, “A Palestinian Targum feature in Syriac”, JJS 46, pp. 271–282. Bulliet, R. W., 1979, Conversion to Islam in the Medieval Period: An Essay in Quantitative History, Cambridge, MA: Harvard University Press. 459 Bumke, J., 1996, Die vier Fassungen der “Nibelungenklage”: Untersuchungen zur Überlieferungsgeschichte und Textkritik der höfischen Epik im 13. Jahrhundert, Berlin: De Gruyter. Butbul, S., 2003, “The commentary of Yefet ben ʿEli the Karaite on the book of Ruth” (Hebrew), Sefunot: Studies and Sources on the History of the Jewish Communities in the East N.S. 8/23, pp. 459–571. Cheikho, L., 1912–1923, Al-Naṣrāniyya wa-ādābuhā bayna ʿArab al-Jāhiliyya, Beirut: Imprimerie catholique. Clivaz, C., Schulthess, S. and Sankar, M., 2017, “Editing New Testament Arabic manuscripts in a TEI-base: fostering close reading in Digital Humanities”, Journal of Data Mining and Digital Humanities, https://hal.archives-ouvertes.fr/hal01280627. Cohen, M., 1994, Under Crescent and Cross: The Jews in the Middle Ages, Princeton: Princeton University Press. Cornill, C. H., 1886, Das Buch des Propheten Ezechiel, Leipzig: J. C. Hinrichs. Corriente, F., 2007, “The psalter fragment from the Umayyad Mosque of Damascus. a birth certificate of Nabaṭī-Arabic”, in J. P. Monferrer-Sala (ed.), Eastern Crossroads: Essays on Medieval Christian Legacy, Piscataway, NJ: Gorgias Press, pp. 303–321. Davidson, S., 1852, A Treatise on Biblical Criticism: Exhibiting a Systematic View of that Science, Edinburgh: A. and C. Black. Dikken, B. J., 2012, “Some remarks about Middle Arabic and Saʿadya Gaon’s Arabic translation of the Pentateuch in manuscripts of Jewish, Samaritan, Coptic Christian, and Muslim provenance”, in L. Zack and A. Schippers (eds), Middle Arabic and Mixed Arabic: Diachrony and Synchrony, Leiden: Brill, pp. 51–81. Dmitriev, K., 2009, “An early Christian Arabic account of the creation of the world”, in A. Neuwirth, N. Sinai and M. Marx (eds), The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu, Leiden: Brill, pp. 349–387. Döderlein, J. C., 1778–1779, “Von arabischen Psaltern”, Repertorium für Biblische und Morgenländische Litteratur 2, pp. 151–179 and 4, pp. 57–96. Drint, A., 1997, The Mount Sinai Arabic Version of IV Ezra, Louvain: Secrétariat du Corpus SCO. ———, 1999, “Some notes on the Arabic versions of IV Ezra and the Apocalypse of Baruch in MS Mount Sinai Arabic 589”, Parole de l’Orient 24, pp. 165–178. Eichhorn, J. G., 1777–1786, Repertorium für Biblische und Morgenländische Litteratur, Leipzig: Weidmann. Empson, W. (ed.), 1851, “The Greek Text of the New Testament”, Edinburgh Review 94/191, pp. 1–46. Erpenius, T., 1622, Pentateuchus Mosis Arabicè, Leiden: Erpeniana. Fattal, A., 1958, Le statut légal des non-Musulmans en pays d’islam, Beirut: Imprimerie catholique. Fowden, G., 2014, Before and After Muḥammad: The First Millennium Refocused, Princeton: Princeton University Press. Frank, R. M., 1974, The Wisdom of Jesus Ben Sirach (Sinai ar. 155 IX–X cent.), Louvain: Secrétariat du Corpus SCO. Freimann, A., 1901, “Die hebräischen Pergamentdrucke”, Zeitschrift für Hebräische Bibliographie 15, pp. 46–57. Fück, J., 1955, Die arabischen Studien in Europa bis in den Anfang des 20. Jahrhunderts, Leipzig: Harrassowitz. Gehman, H. S., 1925, “The ‘polyglot’ Arabic text of Daniel and its affinities”, Journal of Biblical Literature 44, pp. 327–352. 460 Gesenius, W., 1820–1821, Der Prophet Jesaia: uebersetzt und mit einem vollständigen philologisch-kritischen und historischen Commentar begleitet, Leipzig: Friedr. Christ. Wilh. Vogel. Gibson, M. D. (ed.), 1894, An Arabic Version of the Epistles of St Paul to the Romans, Corinthians, Galatians, with Part of the Epistle to the Ephesians, from a Ninth Century MS. in the Convent of St Catharine on Mount Sinai, Cambridge: C. J. Clay and Sons. Gildemeister, J., 1865, De Evangeliis in arabicum e simplici Syriaca translates, Bonn: Adolphum Marcum Giustiniani, A., 1516, Psalterium hebraeum, graecum, arabicum et chaldaeum, cum tribus latinis interpretationibus et glossis, Genoa: Petrus Paulus Porrus. Graf, G., 1944, Geschichte der christlichen arabischen Literatur (5 vols), Città del Vaticano: Biblioteca apostolica vaticana. Gregory, C. R., 1894, Novum Testamentum Graece, Leipzig: Giesecke & Devrient. ———, 1908, Die griechischen Handschriften des Neuen Testaments, Leipzig: J. C. Hinrichs. Griffith, S. H., 1985, “The Gospel in Arabic: an inquiry into its appearance in the first Abbasid century”, Oriens Christianus 69, pp. 126–167. ———, 2013, The Bible in Arabic: The Scriptures of the “People of the Book” in the Language of Islam, Princeton: Princeton University Press. Guidi, I., 1888, Le traduzioni degli Evangeli in arabo e in etiopico, Rome: Accademia dei Lincei. Haddad, R., 1992, “La phonétique de l’arabe chrétien vers 700”, in P. Canivet and J.-P. Rey-Coquais, La Syrie, de Byzance à l’Islam: VIIe–VIIIe siècles, Damascus: Institut français de Damas, pp. 159–164. Hamilton, A., 1985, William Bedwell: The Arabist 1563–1632, Leiden: Brill. Hary, B. H., 2009, Translating Religion: Linguistic Analysis of Judeo-Arabic Sacred Texts from Egypt, Leiden: Brill. Hjälm, M. L., 2016, Christian Arabic Versions of Daniel: A Comparative Study of Early MSS and Translation Techniques in MSS Sinai Ar. 1 and 2, Leiden: Brill. Hody, H., 1705, De Bibliorum textibus originalibus, versionibus graecis & latina vulgata, Oxford: e Theatro Sheldoniano. Holmes, R., 1822, Vetus Testamentum ex versione Septuaginta interpretum, Glasgow: Academy Press. Hopkins, S., 1984, Studies in the Grammar of Early Arabic: Based upon Papyri Datable to Before 300 A.H. / 912 A.D., Oxford: OUP. ———, 2002, “On the Vorlage of an early Judaeo-Arabic translation of Proverbs”, Jerusalem Studies of Arabic and Islam 27, pp. 369–374. Hughes, J. C., 1914, De Lagardes Ausgabe der arabischen Übersetzung des Pentateuchs (Cod. Leiden arab. 377), Leipzig: J. C. Hinrichs. Jellicoe, S., 1968, The Septuagint and Modern Study, Oxford: Clarendon Press. Jones, J. R., 1988, Learning Arabic in Renaissance Europe (1505–1624) (unpubl. diss, University of London). Junius, F., 1578, Sanctorum Apostolorum Acta, ex Arabica translatione Latine reddita: Addita obscurorum aliquot difficiliumque locorum, Heidelberg. Juynboll, W. M. C., 1931, Zeventiende-eeuwsche beoefenaars van het Arabisch in Nederland, Utrecht: Kemink en Zoon. Kashouh, H., 2011, The Arabic Versions of the Gospels: The Manuscripts and their families, Berlin: De Gruyter. Khan, G., 2013, A Short Introduction to the Tiberian Masoretic Bible and its Reading Tradition, Piscataway, NJ: Gorgias Press. 461 Khoury, R. G., 1972, Wahb b. Munabbih, vol. I: Der Heidelberger Papyrus PSR Heid Arab 23. Leben und Werk des Dichters, Wiesbaden: Harrassowitz. ———, 1989, “Quelques réflexions sur la première ou les premières Bibles arabes”, in T. Fahd (ed.), L’Arabie préislamique et son environnement historique et culturel, Leiden: Brill, pp. 549–561. Kirstenius, P., 1608, Vitae quatuor Evangelistarum, ex antiquissimo codice Arabico Caesario, Breslau: Published by the author. ———, 1611, Notae in Evangelium S. Matthaei, ex collatione textuum Arabicorum Aegyptiacorum, Hebraeorum, Syriacorum, Graecorum, Latinorum, Breslau: Published by the author. Kittel, R., 1905–1906, Biblia Hebraica, Leipzig: Hinrichs. Knutsson, B., 1974, Studies in the Text and Language of Three Syriac-Arabic Versions of the Book of Judicum with Special Reference to the Middle Arabic Elements, Leiden: Brill. van Koningsveld, P. S., 1975, “An Arabic manuscript of the Apocalypse of Baruch”, Journal for the Study of Judaism 6, pp. 205–207. de Lagarde, P., 1864, Die vier Evangelien Arabisch, Leipzig: F. A. Brockhaus. ———, 1867, Materialien zur Kritik und Geschichte des Pentateuchs, Leipzig: B. G. Teubner. Leemhuis, F., 1984, “The Mount Sinai Arabic version of the Apocalypse of Baruch”, in Actes du deuxième congrès international d’études arabes chrétiennes, Rome: Pontificium Institutum Studiorum Orientalium, pp. 73–79. ———, 1989, “The Arabic version of the Apocalypse of Baruch: a Christian text?”, Journal for the Study of Pseudoepigrapha 2/4, pp. 19–26. ———, Klijn, A. F. J. and van Gelder, G. J. H. (eds), 1986, The Arabic Text of the Apocalypse of Baruch: Edited and Translated with a Parallel Translation of the Syriac Text, Leiden: Brill. Levin, B., 1938, Die griechisch-arabische Evangelien-Übersetzung. Vat. Borg. ar. 95 und Ber. orient. oct. 1108, Uppsala: Almquist & Wiksells Boktrycheri. Levy-Rubin, M., 2011, Non-Muslims in the Early Islamic Empire: From Surrender to Coexistence, Cambridge: CUP. Löfgren, O., 1936, Studien zu den arabischen Danielübersetzungen, mit besonderer Berücksichtigung der christlichen Texte. Nebst einem Beitrag zur Kritik des Peschittatextes, Uppsala: A.B. Lundequistska Bokhandeln. Macdonald, D. B., 1904, “Ibn al-ʿAssāl’s Arabic version of the Gospels”, in E. Saavedra (ed.), Homenaje á D. Francisco Codera en su Jubilación del Profesorado Zaragoza: M. Escar, pp. 375–392. Macdonald, M. C. A., 2008, “Old Arabic (Epigraphic)”, in K. Versteegh (ed.), Encyclopedia of Arabic Language and Linguistics, Leiden: Brill, vol. III, pp. 464–477 Macdonald, M. C. A., 2008, “Old Arabic (Epigraphic)”, in K. Versteegh (ed.), Encyclopedia of Arabic Language and Linguistics, Leiden: Brill, vol. III, pp. 464–477. ———, 2009, Literacy and Identity in pre-Islamic Arabia, Aldershot: Ashgate. Madros, P., 1984, Six Arabic Translations of the Psalms: Problems of Exegesis and Philology, Rome: Pontificia Commissio Biblica. Mango, C., 1999, “Greek culture in Palestine after the Arab conquests”, in G. Cavallo, G. de Gregorio and M. Maniaci (eds), Scritture, libri e testi nelle aree privinciali di Bisanzio: Atti del seminario di Erice (18-25 settembre 1988), Spoleto: Centro Italiano di Studi sull’Alto Medioevo, pp. 149–160. Mavroudi, M., 2008, “Arabic words in Greek letters: the Violet Fragment and more”, in J. Lentin and J. Grand’Henry (eds), Moyen arabe et variétés mixtes de l’arabe à travers l’histoire, Louvain: Peeters, pp. 321–354. 462 ———, 2015, “Greek language and education under early Islam”, in B. Sadeghi, A. Q. Ahmad, A. Silverstein and R. Hoyland (eds), Islamic Cultures, Islamic Contexts. Essays in Honor of Professor Patricia Crone, Leiden: Brill, pp. 295–342. Miller, P. N., 2001, “The ‘antiquarianization’ of biblical scholarship and the London Polyglot Bible”, Journal of the History of Ideas 62, pp. 463–482. Moawad, S., 2014, Al-Asʿad Abū Al-Faraǧ Hibat Allāh Ibn Al-ʿAssāl: Die Arabische Übersetzung der vier Evangelien. / ‫ الترجمة العربية‬:‫األسعد أبو الفرج هبة هللا بن العسال‬ ٢٠١٤ ،‫ مدرسة اإلسكندرية‬:‫ القاهرة‬.‫ تحقيق ومقدمة‬،‫لألناجيل األربعة‬, Cairo: Alexandria School. Monferrer-Sala, J. P., 2000, “Una versión árabe del ‘Apocalipsis’ contenida en el Cod. Ár. 1625 de la Biblioteca de El Escorial: estudio y edición”, Qurṭuba 5, pp. 147– 180. ———, 2002, “An eastern Arabic version of the three Epistles of Saint John, Codex Ar. 1625, kept in the monastery of El Escorial, Madrid”, Parole de l’Orient 27, pp. 27–49. ———, 2015, “The Pauline Epistle to Philemon from Codex Vatican Arabic 13 (ninth century CE): transcription and study”, JSS 2, pp. 341–371. ———, 2017, Lecturas hexaplares y posthexaplares en la versión árabe del libro de Números de al-Ḥāriṯ b. Sinān b. Sunbāṭ (c. s. X AD), Madrid: Editorial Sindéresis. de Nave, F. (ed.), 1986, Philologia Arabica. Arabische Studiën en Drukken in de Nederlanden in de 16de en 17de Eeuw, Antwerp: Museum Plantin-Moretus. Nestle, E., 1897, “Bibelübersetzungen, arabische”, in Realencyklopädie für Protestantische Theologie und Kirche, 3rd edn, Leipzig: J. C. Hinrichs, vol. III, pp. 90–95. Neuwirth, A., 2010, Der Koran als Text der Spätantike: Ein europäischer Zugang, Berlin: Islam Verlag. Newby, G. D., 1988, A History of the Jews of Arabia: From Ancient Times to their Eclipse under Islam, Columbia: University of South Carolina Press. Noth, A., 1987, “Die ‘Bedingungen ʿUmars (aš-Šurūṭ al-’umariyya)’ unter einem anderen Aspekt gelesen”, Jerusalem Studies in Arabic and Islam 9, pp. 290–315. Paulus, H. E. G., 1789, Commentatio critica, exhibens e Bibliotheca Oxoniensi Bodleiana Specimina versionum Pentateuchi septem Arabicarum, nondum editarum, cum observationibus, Jena: apud Bibliop. Academicum. ———, 1790–1791, Neues Repertorium für Biblische und Morgenländische Litteratur, Jena: Cuno. Peshitta Institute, 1961, List of Old Testament Peshitta Manuscripts, Leiden: E. J. Brill. Peters, C., 1936, “Proben eines bedeutsamen arabischen Evangelien-Textes”, Oriens Christianus 11/3, pp. 188–211. ———, 1940, “Psalm 149 in Zitaten islamischer Autoren,” Biblica 21, pp. 138–151. ———, 1942, “Grundsätzliche Bemerkungen zur Frage der arabischen Bibeltexte”, Rivista degli Studi Orientali 20, pp. 129–143. Polaschegg, A., 2004, Der andere Orientalismus. Regeln deutsch-morgenländischer Imagination im 19. Jahrhundert, Berlin: De Gruyter. Polliack, M., 1997, The Karaite Tradition of Arabic Bible Translation: A Linguistic and Exegetical Study of Karaite Translations of the Pentateuch from the Tenth and Eleventh Centuries C.E., Leiden: Brill. ———, 2006, “Medieval Judaeo-Arabic literature”, in J. Neusner, A. J. Avery-Peck and W. S. Green (eds), Encyclopaedia of Judaism, Leiden: Brill, http://referenceworks.brillonline.com/browse/encyclopaedia-of-judaism. ——— and Schlossberg, E. (eds), 2001, “Yefet ben Eli’s translation of the book of Obadiah” (Hebrew), Peʿamim: Studies in Oriental Jewry 89, pp. 61-82. 463 ——— and Schlossberg, E. (eds), 2009, Yefet ben ʻEli’s Commentary on Hosea (Hebrew), Ramat Gan: Bar-Ilan University Press. Potthast, D., 2011, “Die andalusische Übersetzung des Römerbriefs”, Collectanea Christiana Orientalia 8, pp. 65–108. Rahlfs, A., 1914, Verzeichnis der griechischen Handschriften des Alten Testaments, Berlin: Weidmann. ——— and Fraenkel, D., 2003, Verzeichnis der griechischen Handschriften des Alten Testaments. Band I,1: Die Überlieferung bis zum VIII. Jahrhundert, Göttingen: Vandenhoeck & Ruprecht. Reinke, L., 1867, Zur Kritik der älteren Versionen des Propheten Nahum, Münster: Niemann. ———, 1868, Der Prophet Haggai: Einleitung, Grundtext und Uebersetzung, nebst einem vollständigen philologisch-kritischen und historischen Commentar, Münster: Niemann. Reynolds, L. D. and Wilson, N. G., 1974, Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature, 2nd edn, Oxford: Clarendon Press. Reynolds, S. M., 1943, “Al-Alam’s version of Zechariah”, Muslim World 33, pp. 273– 275. Rhode, J. F., 1921, The Arabic Versions of the Pentateuch in the Church of Egypt: A Study from Manuscript Sources (IX–XVII Century), St. Louis: B. Herder. al-Rizzī, S., 1610, Psalmi Davidis cum canticis veteris et novi Testamenti, Syriacè et Arabicè, sed charactere Syriaco minori, nam alter majusculus est : Praemittitur epistola Sergii Risii, Maronicae archiepiscopi Damasceni, Monte Libano. Roberts, B. J., 1957, Die Religion in Geschichte und Gegenwart: Handwörterbuch für Theologie und Religionswissenschaft, 3rd edn, Tübingen: Mohr Siebeck. Ryssel, V., 1885, “Die arabische Übersetzung des Micha in der Pariser und Londoner Polyglotte”, ZAW 5, pp. 102–138. Sabih, J. A. (ed.), 2009, Japheth Ben Ali’s Book of Jeremiah: A Critical Edition and Linguistic Analysis of the Judaeo-Arabic Translation, London: Equinox. Samaan, K. W., 1994, Sept traductions Arabes de Ben Sira, Frankfurt: Peter Lang. Samir, S. K., 1982, “Pour une nouvelle histoire de la littérature arabe des Chrétiens”, in S. K. Samir (ed.), Actes du premier congrès international d’études arabes chrétiennes (Goslar, septembre 1980), Rome: PIO, pp. 259–286. ———, 2008, “The theological Christian influence on the Qurʾān”, in G. S. Reynolds (ed.), The Qurʾān in its Historical Context, New York: Routledge, pp. 141–162. Sasson, I., 2016, The Arabic Translation and Commentary of Yefet Ben ʻEli on the Book of Proverbs, vol. I: Introduction and Edition, Leiden: Brill, http://dx.doi.org/10.1163/9789004317499. Schäfer, P. and Becker, H.-J. (eds), 1995, Synopse of Talmud Yerushalmi, Tübingen: Mohr. Schäfer, P., Schlüter, M. and von Mutius, H. G. (eds), 1981, Synopse zur HekhalotLiteratur, Tübingen: Mohr. Schulthess, S., 2012, “Les manuscrits du Nouveau Testament, le monde arabe et le digital: l’èmergence d’un discours hybride”, in C. Clivaz, J. Meizoz, F. Vallotton and J. Verheyden (eds), Reading Tomorrow. From Ancient Manuscripts to the Digital Era / Lire demain. Des manuscrits antiques à l’ère digitale, Lausanne: Presses polytechniques et universitaires romandes, pp. 333–344. ———, 2013, “The role of the internet in New Testament textual criticism: the example of the Arabic manuscripts of the New Testament”, in C. Clivaz, A. Gregory and D. Hamidovic (eds), Digital Humanities in Biblical, Early Jewish and Early Christian Studies, Leiden: Brill, pp. 71–82 464 Schwarzstein, J., 1886, Targum Arvi: Die arabische Interpretation des Pentateuchs von Rabbi Saadiah hagaon, Karlsruhe: A. J. Hofmann. Seidensticker, T., 2011, “Die Authentizität der Umaiya Ibn Abî s-Salt zugeschriebenen Gedichte II”, ZDMG 161, pp. 39–68. Septuaginta-Unternehmen der Akademie der Wissenschaften zu Göttingen (eds), 2012, Offizielles Verzeichnis der Rahlfs-Sigeln, https://rep.adwgoe.de/handle/11858/00-001S-0000-0022-A30C-8. Shahid, I., 1984, Byzantium and the Arabs in the Fifth Century, Washington, DC: Dumbarton Oaks Research Library and Collection. Shehadeh, H. (ed.), 1989, The Arabic Translation of the Samaritan Pentateuch, vol. I: Genesis–Exodus, Jerusalem: The Israel Academy of Sciences and Humanities. ——— (ed.), 2002, The Arabic Translation of the Samaritan Pentateuch, vol. II: Leviticus, Numbers, Deuteronomy, Jerusalem: The Israel Academy of Sciences and Humanities. Sionita, G. and Hesronita, J., 1616, Grammatica Arabica Maronitarum, Paris: Savariana. Sinai, N., 2011, “Religious poetry from the Qurʾanic milieu: Umayya b. Abī l-Ṣalt on the fate of the Thamūd,” Bulletin of SOAS 74, pp. 397–416. Smitskamp, R., 1976, Philologia Orientalis: A Description of Books Illustrating the Study and Printing of Oriental Languages in Europe, Leiden: Brill. Soncino, E. b. G., 1546, Pentateuchus hebraeus cum versione persica, Targum et commentariis R. Salomonis, paraphrasi arabica R. Saadi Gaonis, Constantinople: Soncino. Sprenger, A., 1869, Das Leben und die Lehre des Mohammed, Berlin: Nicolaische Verlagsbuchhandlung. Staal, H., 1984, Mt. Sinai Arabic Codex 151, Leuven: Peeters. Stenij, E., 1901, Die altarabische Übersetzung der Briefe an die Hebräer, an die Römer und an die Corinther. Aus einem in St. Petersburg befindlichen Codex Tischendorfs vom Jahre 892 n. Chr., Helsingfors: Frenckellska TryckeriAktiebolaget. Stillman, N. A., 2010, “Academic study of Islamicate Jewry”, in N. A. Stillman (ed.), Encyclopedia of Jews in the Islamic World, Leiden: Brill, http://referenceworks.brillonline.com/browse/encyclopedia-of-jews-in-the-islamic-world/. Stroumsa, S., 2007, “A literary genre as an historical document: on Saadia’s introductions to his Bible commentaries”, in M. M. Bar-Asher, S. Hopkins, S. Stroumsa and B. Chiesa (eds), A Word Fitly Spoken: Studies in Mediaeval Exegesis of the Hebrew Bible and the Qur’ān presented to Haggai Ben-Shammai, Jerusalem: Ben Zvi Institute, pp. 193–204. Szpek, H. M., 1992, Translation Technique of the Peshitta to Job: A Model for Evaluating a Text with Documentation from the Peshitta to Job, Atlanta: Scholars Press. Tarras, P., forthcoming, “Friedrich Grote (1861-1921): a collector of Oriental manuscripts”. von Tischendorf, L. F. C., 1849, Novum Testamentum Graece, Leipzig: Adolph Winter. Toomer, G. J., 1996, Eastern Wisedome and Learning: The Study of Arabic in Seventeenth-Century England, Oxford: Clarendon Press. Toral-Niehoff, I., 2008, “Eine arabische poetische Gestaltung des Sündenfalls: das vorislamische Schöpfungsgedicht von ʿAdī b. Zayd”, in D. Hartwig, W. Homolka, M. J. Marx and A. Neuwirth (eds), “Im vollen Licht der Geschichte”: Die Wissenschaft des Judentums und die Anfänge der kritischen Koranforschung, Würzburg: Ergon, pp. 235–256. 465 Tov, E., 1979, “Loan-words, homophony and transliterations in the Septuagint”, Biblica 60, pp. 216–236. Tregelles, S. P., 1893, “Arabic Versions”, in A Dictionary of the Bible Comprising its Antiquities, Biography, Geography, and Natural History, 2nd edn, London: John Murray, vol. III, pp. 1614–1616. Tritton, A. S., 1970, The Caliphs and their Non-Muslim Subjects: A Critical Study of the Covenant of ʿUmar, London: Frank Cass. Tychsen, O. G., 1782, “Ueber die Quelle aus welcher die Handschrift der Arabischen Version in den Polyglotten geflossen ist”, Repertorium für Biblische und Morgenländische Litteratur 11, pp. 95–110. Vaccari, A., 1920, “Le versioni arabe dei Profeti”, Biblica 1, pp. 266–268. ———, 1921, “Le versioni arabe dei Profeti”, Biblica 2, pp. 401–424. ———, 1922, “Le versioni arabe dei Profeti”, Biblica 3, pp. 401–423. Vercellin, G, 2001, Venezia e l’origine della stampa in caratteri arabi, Padova: Il Poligrafo. Violet, B., 1901, “Ein zweisprachiges Psalmfragment aus Damascus”, Orientalische Literaturzeitung 4, pp. 384–403, 425–441, 475–488. Vollandt, R., 2012, “The Arabic Pentateuch of the Paris Polyglot: Saadiah Gaon’s advent to the republic of letters”, in S. Binay and S. Leder (eds), Translating the Bible into Arabic: Historical, Text-Critical, and Literary Aspects, Beirut: OrientInstitut Beirut, pp. 19–35. ———, 2015, Arabic Versions of the Pentateuch: A Comparative Study of Jewish, Christian, and Muslim Sources, Leiden: Brill. ———, 2016a, “The conundrum of scriptural plurality: the Arabic Bible, polyglots, and medieval predecessors of biblical criticism”, in A. Piquer Otero and P. A. Torijano Morales (eds), The Text of the Hebrew Bible and its Editions: Studies in Celebration of the Fifth Centennial of the Complutensian Polyglot, Leiden: Brill, pp. 56–85. ———, 2016b, “Making quires speak. An analysis of Arabic multi-block Bibles and the Creation of a Canon”, Intellectual History of the Islamicate World 4, pp. 173– 209. ———, 2016c, “An unknown medieval Coptic Hebraism? On a momentous junction of Jewish and Coptic biblical studies”, in A. Grafton and G. Most (eds), Canonical Texts and Scholarly Practices: A Global Comparative Approach, Cambridge: CUP, pp. 183–199. ———, forthcoming, “Flawed Biblical translations into Arabic and how to correct them: a Copt and a Jew study Saadiah’s Tafsīr”. Wakefield, R., 1989, On the Three Languages (1524), G. Lloyd Jones (ed. and transl.), Binghamton, NY: Medieval & Renaissance Texts & Studies. Wald, S. G., 1784, “Über die arabische Übersetzung des Daniel in den Polyglotten”, Repertorium für biblische und morgenländische Litteratur 14, pp. 204–210. Wallraff, M. and Andrist, P., 2015, “Paratexts of the Bible: a new research project on Greek textual transmission”, Early Christianity 2, pp. 237–243. Walton, B. (ed.), 1653–1657, Biblia sacra polyglotta, London: Thomas Roycroft. Wechsler, M. G. (ed.), 2008, The Arabic Translation and Commentary of Yefet Ben ʻEli the Karaite on the Book of Esther: Edition, Translation, and Introduction, Leiden: Brill. Weitzman, M., 1999, The Syriac Version of the Old Testament: An Introduction, Cambridge: CUP. Wevers, J. W., 1970, “The Arabic versions of Genesis and the Septuagint”, Bulletin of the International Organization for Septuagint and Cognate Studies 3, pp. 8– 11. 466 White, J., 1779, A Letter to the Right Reverend the Lord Bishop of London, Suggesting a Plan for a New Edition of the LXX. To which are added Specimens of some inedited Eastern Versions made from the Greek, and a Sketch of a Chart of Greek Mss. By the Rev. Joseph White, B.D. Fellow of Wadham College, Laudian Professor of Arabic, one of His Majesty’s Preachers at Whitehall; and Editor and Translator of the Syriac Philoxenian Version of the Gospels, Oxford: Clarendon Press. Zaki, V., 2017, “The Textual History of the Arabic Pauline Epistles: One Version, Three Recensions, Six Manuscripts,” in Hjälm, Miriam L. (ed.), Senses of Scripture, Treasures of Tradition: The Bible in Arabic among Jews, Christians and Muslims (Biblia arabica, 5), 392–424, Leiden: Brill. Zawanowska, M. (ed.), 2012, The Arabic Translation and Commentary of Yefet Ben ʻEli the Karaite on the Abraham Narratives (Genesis 11:10-25:18), vol. I: Introduction and Text Edition, Leiden: Brill. Zewi, T., 1997a, “Relative clauses in Saadya Gaon’s translation of the Pentateuch”, Annali (Istituto Universitario Orientale) 57, pp. 426–449. ———, 1997b, “Saadia Gaon’s translations to nominal sentences in the Pentateuch” (Hebrew), Massorot 9–11, pp. 511–26. ———, 2000, “Relative clauses in Saadya Gaon’s translation of the Pentateuch according to MS. St. Petersburg”, in M. Mifsud (ed.), Proceedings of the Third International Conference of AIDA Held in Malta on 29 March - 2 April, 1998, Paris: Association Internationale de Dialectologie Arabe, pp. 179–183. ———, 2001, “Energicus in Saadya Gaon’s Translation of the Pentateuch”, in A. Zaborski (ed.), New Data and New Methods in Afroasiatic Linguistics - Robert Hetzron in Memoriam, Wiesbaden: Harrassowitz, pp. 223–230. ———, 2002, “Syntactic questions in Saadya Gaon’s translation of the Pentateuch” (Hebrew), Massorot 12, pp. 103–115. ———, 2014, “Proper names in the Samaritan version of Saadya Gaon’s translation of the Pentateuch” (Hebrew), Ben ʿEver La-ʿArav 6, pp. 97–108. ———, 2016, “Proper nouns in the Samaritan version of Saadya Gaon’s translation of the Pentateuch”, JSS 61, pp. 139–155. Zumthor, P., 1972, Essai de poétique médiévale, Paris: Seuil. 467 ACTA UNIVERSITATIS UPSALIENSIS Studia Semitica Upsaliensia 30 Editor: Eleanor Coghill 1. Al Hamdânî, Kitâb al-Ǧauharatain al-‘atîqatain herausgegeben und übersetzt von Christopher Toll. 1968. 2. The Aramaic Portions of the Pesiqta de Rab Kahana according to MS Marshall Or. 24, the oldest known manuscript of the Pesiqta de Rab Kahana, with English Translation, commentary and Introduction by Gerhard Svedlund. 1974. 3. Kitāb al-Luma‘ fi-n-Naḥw d’Ibn Ǧinnī édité par Hadi M. Kechrida. 1976. 4. Ibn Ḥazm al-Andalusī, Kitāb al-’axlāq wa-s-siyar ou Risāla fī mudāwāt an-nufūs wa-tahḏīb al-’axlāq wa-z-zuhd fī raḏā’il. Introduction, édition critique, remarques par Eva Riad. 1980. 5. Joseph ha-Kohen, Sefer ‘Emeq ha-Bakha (The Vale of Tears), with the chronicle of the anonymous Corrector. Introduction, critical edition, comments by Karin Almbladh. 1981. 6. Olof Pedersén, Archives and Libraries in the City of Assur. A Survey of the Material from the German Excavations. Part I. 1985. 7. Karin Almbladh, Studies in the Book of Jonah. 1986. 8. Olof Pedersén, Archives and Libraries in the City of Assur. A Survey of the Material from the German Excavations. Part II. 1986. 9. Witold Witakowski, The Syriac Chronicle of Pseudo-Dionysius of Tel-Maḥrē. A Study in the History of Historiography. 1987. 10. Bo Isaksson, Studies in the Language of Qoheleth. With Special Emphasis on the Verbal System. 1987. 11. Eva Riad, Studies in the Syriac Preface. 1988. 12. Mats Eskhult, Studies in Verbal Aspect and Narrative Technique in Biblical Hebrew Prose. 1990. 13. Mikael Persenius, The Manuscripts of Parts 1 and 2 of Shams al-‘ulūm by Nashwān al-Ḥimyarī. A Study of Their Relationship. 1997. 14. Björn Frennesson, “In a Common Rejoicing”. Liturgical Communion with Angels in Qumran. 1999. 15. Gidon Avraham, A Bridge of Words. A Term List Based on the Study and Classification of Compounding Operations in Avot Yeshurun’s Later Poetry (1974–1992) Concerning the Notion of (Home/House). 1999. 16. Johann E. Erbes, The Peshitta and the Versions. A Study of the Peshitta Variants in Joshua 1–5 in Relation to Their Equivalents in the Ancient Versions. 1999. 17. Anette Månsson, Passage to a new wor(l)d. Exile and restoration in Mahmoud Darwish’s writings 1960–1995. 2003. 18. Aziz Tezel, Comparative Etymological Studies in the Western Neo-Syriac (Ṭūrōyo). With Special Reference to Homonyms, Related Words and Borrowings with Cultural Signification. 2003. ACTA UNIVERSITATIS UPSALIENSIS Studia Semitica Upsaliensia 30 Editor: Eleanor Coghill 19. Torkel Lindquist, A War of Words: From Lod to Twin Towers. Defining Terrorism in Arab and Israeli Newspapers 1972–1996 (2001) – A Study in Propaganda, Semantics and Pragmatics. 2003. 20. Assad Sauma, Gregory Bar-Hebraeus’s Commentary on the Book of Kings from His Storehouse of Mysteries. A Critical Edition with an English Translation, Introduction and Notes. 2003. 21. Astrid Ottosson Bitar, “I Can Do Nothing against the Wish of the Pen”. Studies in the Short Stories of Widād Sakākīnī. 2005. 22 Hallvard Hagelia, The Tel Dan inscription. A critical investigation of recent research on its palaeography and philology. 2006. 23 Tal Davidovich, The Mystery of the House of Royal Women. Royal Pīlagšīm as Secondary Wives in the Old Testament. 2007. 24 The Professorship of Semitic Languages at Uppsala University 400 years. Jubilee Volume from a Symposium held at the University Hall, 21–23 September 2005. Edited by Bo Isaksson, Mats Eskhult & Gail Ramsay. 2007. 25. Søren Holst, Verbs and War Scroll. Studies in the Hebrew Verbal System and the Qumran War Scroll. 2008. 26. Ablahad Lahdo, The Arabic Dialect of Tillo in the Region of Siirt (South-eastern Turkey). 2009. 27. Sina Tezel, Arabic Borrowings in Ṣūrayt/Ṭūrōyo within the Framework of Phonological Correspondences. In Comparison with other Semitic Languages. 2011. 28. Jakob Andersson, Kingship in the Early Mesopotamian Onomasticon 2800–2200 BCE. 2012. 29. Ablahad Lahdo, The Arabic Dialect of Tillo in the Region of Siirt (South-eastern Turkey). Textbook. 2016. 30. Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan. Edited by Nadia Vidro, Ronny Vollandt, Esther-Miriam Wagner and Judith Olszowy-Schlanger. 2018. ISSN 0585-5535 ISBN 978-91-513-0290-4 Distributor: Uppsala University Library, Box 510, SE-751 20 Uppsala, Sweden www.uu.se, acta@ub.uu.se