Scribal Habits in Near Eastern Manuscript Traditions 9781463241964

Citation preview

Scribal Habits in Near Eastern Manuscript Traditions

Texts and Studies (Third Series)

23 Series Editor H. A. G. Houghton

Editorial Board Jeff W. Childers Alba Fedeli Viktor Golinets Christina M. Kreinecker Gregory S. Paulson Peter J. Williams

Texts      and Studies is a series of monographs devoted to the study of Biblical and Patristic texts. Maintaining the highest scholarly standards, the series includes critical editions, studies of primary sources, and analyses of textual traditions.

Scribal Habits in Near Eastern Manuscript Traditions

Edited by

George Anton Kiraz Sabine Schmidtke

gp 2020

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com 2020 Copyright © by Gorgias Press LLC All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC.

‫ܝܐ‬

1

2020

ISBN 978-1-4632-4195-7

ISSN 1935-6927

Library of Congress Cataloging-in-Publication Data Names: Kiraz, George Anton, editor. | Schmidtke, Sabine, editor. Title: Scribal habits in Near Eastern manuscript traditions / edited by George Anton Kiraz, Sabine Schmidtke. Description: Piscataway, NJ : Gorgias Press, 2020. | Series: Texts and studies, 1935-6927 ; Third series 23 | Includes bibliographical references. | Summary: "This volume brings together contributions by scholars focussing on peritextual elements as found in Middle Eastern manuscripts: dots and various other symbols that mark vowels, intonation, readings aids, and other textual markers; marginal notes and sigla that provide additional explanatory content akin to but substantially different from our modern notes and endnotes; images and illustrations that present additional material not found in the main text. These elements and additional layers to the main body of the text are crucial for our understanding of the text's transmission history as well as scribal habits"-- Provided by publisher. Identifiers: LCCN 2020051104 (print) | LCCN 2020051105 (ebook) | ISBN 9781463241957 (hardback) | ISBN 9781463241964 (ebook) Subjects: LCSH: Paratext--Middle East--History. | Marginalia--Middle East--History. | Middle Eastern literature--Criticism, Textual. | Manuscripts--Middle East--Editing. | Scribes--Middle East--History. | Transmission of texts--Middle East--History. Classification: LCC Z242.P37 S37 2020 (print) | LCC Z242.P37 (ebook) | DDC 002.0956--dc23 LC record available at https://lccn.loc.gov/2020051104 LC ebook record available at https://lccn.loc.gov/2020051105

Printed in the United States of America

TABLE OF CONTENTS Contributors ............................................................................................ vii Preface ....................................................................................................... xi Connecting the Dots: Using Diaeresis as a Source of Information about Scribal Practices in Byzantine Egypt ................... 1 Elizabeth Buchanan Marginalia as Traces of Changing Knowledge Culture: The Circulation of Taqwīm Texts in the Late Mamluk Sultanate ... 33 Fien De Block The Manuscripts of Arabic Popular Siyar and Sīrat Sayf ibn Dhī Yazan ............................................................................................... 47 Zuzana Gažákovà A Portable Majlis: On Publishing Reliable Editions in Ottoman Manuscript Culture........................................................................ 69 Aslihan Gürbüzel Chapter Divisions and the Interpretation and Transmission of the Tosefta ........................................................................................ 83 Binyamin Katzoff The Second-Hand Scribe: The Intellectual Environment of the Production of a Unique Tosefta Fragment from the Levant ............................................................................................107 Binyamin Katzoff Peritextual Encoding for the Metatron / Yahoel Theme in the Kabbalistic Sefer Ha-Ot, or “Book of the Sign,” by R. Abraham Abulafia (1240–1292) ................................................125 Aryeh M. Krawczyk Reading and Remembering in the Medieval Near East: The Syriac Shemohē Book (aka. the Syriac “Masorah”)...................141 Jonathan Loopstra v

vi

SCRIBAL HABITS

Annotations in the Earliest Medieval Hebrew Bible Manuscripts ....................................................................................... 167 Elvira Martín-Contreras An Illuminating Scribe: The ʿArza-dasht of Jaʿfar Bāysunghurī and Its Wealth of Information ..................................................189 Shiva Mihan Annotation Practices in a Syriac Exegetical Collection (MS Vat. Syr. 103) ........................................................................................225 Marion Pragt Scribes and the Book of Revelation in Eastern New Testaments ........................................................................................ 247 T. C. Schmidt On the Sumerian Glossographic Tradition ......................................277 Szilvia Sövegjártó Can Manuscript Headings Prove that there were Arabic Gospels before the Qurʾān? ......................................................289 Robert Turnbull

CONTRIBUTORS Elizabeth F. Buchanan received her D.Phil. from the University of Oxford, United Kingdom, in 2015, with a thesis entitled Debt in Late Antique Egypt, 400-700 CE – Approaches to a Time in Transition, which she is currently preparing for publication. She has been employed as an Assistant Professor at the University of Findlay since the fall of 2016, and teaches introduction to history, the global history series, and the history of the late antique and medieval eastern Mediterranean as well as Latin and Greek. Commencing in 2018, Dr. Buchanan created a new Museum Studies minor for undergraduates. She has spoken on and published several articles on debt and rural community relations in late antique Egypt. Dr. Buchanan’s research interests focus on credit and the late antique economy, legal and religious rhetoric in the ancient world, and the transmission of legal and commercial information through scribal practices.

Fien De Block is a PhD researcher of Arabic and Middle Eastern studies at the University of Ghent and a fellow of the Flemish Research fund. His research focusses on the shifting boundaries within the science of the stars in the late fifteenth century sultanate of Cairo, also known as the Mamluk Sultanate and is supervised by prof. dr. Jo Van Steenbergen and prof. dr. Steven Vanden Broecke. Before this he graduated from Ghent university as a MA in Philosophy (2013, MA thesis on the epistemology of the ikhwān al-ṣafā’) and as an MA in Arabic and Islamic Studies (2015, MA thesis on Ibn al-Shāṭir’s work as an Islamic timekeeper in 14th century Syria).

Zuzana Gažáková (1977) obtained her Ph.D. from the Faculty of Arts at Comenius University in Bratislava in 2006, where she works as an Assistant Professor. Her dissertation thesis examined Sīrat Sayf ibn Dhī Yazan, which she studied from manuscripts. In her research, she focuses on Arab popular and modern literature and vii

viii

SCRIBAL HABITS

translation. She was a Research Fellow at the University of Heidelberg, Department of Semitic Studies, in 2007, and at the University of Vienna, Department of Oriental Studies, in 2014.

Aslıhan Gürbüzel is Assistant Professor of Ottoman Studies at the Institute of Islamic Studies at McGill University. Her research focuses on early modern Ottoman Sufism with an emphasis on the role of Sufi orders in intellectual and cultural production. Her forthcoming book, based on her dissertation, explores notions of changing notions of political participation in early modern Sufi literature. She has also written about the role of Sufi networks, particularly the Mawlawī order, in providing non-madrasa education to a broad sector of urban populations, and in producing and circulating essential literature in subjects varying from rhetoric to medicine. Aryeh Krawczyk has an MA in the Study of Religions at the Jagiellonian University of Cracow, Poland, and a Ph.D. in Hebrew Manuscripts at Adam Mickiewicz University of Poznan, Poland. In 2018 he published a book based on his dissertation: “Sefer ha-Ot, Book of the Sign – Endophasia, the Inner Speech and Heathoscopy, the Seeing of Oneself in the work of r. Abraham Abulafia (1240-1292)”. He is an associate scholar at the Jewish Historical Institute of Warsaw, Poland. As of 2019 he studies at the Or Jonatan Rabbinical Kollel in Hamburg, Germany and collaborates with the Maimonidean Centre of Advanced Studies at Hamburg University.

Elvira Martín-Contreras is Tenured Research Fellow at the Institute of Languages and Cultures of the Mediterranean and the Near East (ILC-CSIC). She is a specialist in Masora – i. e. the marginal annotations placed in the medieval Hebrew Biblical manuscripts, and classic rabbinic exegesis. Her research is focused on the textual transmission and the reception of the Hebrew Bible text attested in the rabbinic literature and the marginal annotations. She is author of several monographs and many articles, including Masora. La transmisión de la Tradición de la Biblia Hebrea (Navarra 2010). She is coeditor of The Text of the Hebrew Bible. From the Rabbis to the Masoretes (Vandenhoeck & Ruprecht 2014).

Jonathan Loopstra is Professor and Chair of the History Department at the University of Northwestern-St. Paul. Dr. Loopstra has received a Ph.D from The Catholic University of America in Wash-

CONTRIBUTORS

ix

ington DC, a Masters of Studies (Mst) from Oxford University, and a MA from Trinity Divinity School. His publications include Job. According to the Peshitta Version with English Translation (Gorgias Press, 2017), An East Syrian Manuscript of the Syriac ‘Masora’ Dated to 899 CE (Gorgias Press, 2014-15), and The Patristic ‘Masora.’ A Study of Patristic Collections in Syriac Handbooks from the Near East, CSCO 685, syr. 265 (Peeters, 2020).

Marion Pragt is a PhD candidate of the FWO Research Foundation Flanders at the Louvain Centre for Eastern and Oriental Christianity, KU Leuven, Belgium. Her dissertation examines how the Song of Songs was interpreted in Syriac commentaries and florilegia and in what ways compilers transmitted and adapted the interpretations of their predecessors. Her research interests include the history of biblical interpretation, Christianity in the late antique and medieval Mediterranean world and the translation and textual transmission of Greek and Syriac literature. She has published on Syriac biblical interpretation and the reception of Gregory of Nyssa’s Homilies on the Song of Songs. T.C. Schmidt received his PhD from Yale University in 2020 and is currently Visiting Assisting Professor of New Testament and Early Christianity at Fairfield University. His research interests include manuscript studies, eschatological thought, the formation of the New Testament and other issues relevant to the early church and later eastern Christian traditions. Schmidt has published the following volumes: Cassiodorus, Gregory the Great, and Others: Writings on the Apocalypse. Fathers of the Church series. Washington DC: CUA Press, 2020/21 (with Francis X. Gumerlock and Mark DelCogliano); Revelation 1–3 in Christian Arabic Commentary: John’s First Vision and the Letters to the Seven Churches. New York, NY: Fordham University Press, 2019 (with Stephen Davis and Shawqi Talia); Hippolytus of Rome: Commentary on Daniel and ‘Chronicon’. Piscataway, NJ: Gorgias Press, 2017.

Szilvia Sövegjártó (PhD 2017 Jena) is postdoctoral researcher of Assyriology at the Center for the Study of Manuscript Cultures, University of Hamburg specialized in the Sumerian and Akkadian language and literature of the 2nd millennium BCE. Her dissertation project at the Friedrich Schiller University, Jena treated the glosses of Sumerian literary manuscripts of the Old Babylonian period. She

x

SCRIBAL HABITS

worked as a postdoctoral researcher in the project “Forms of presence and tradition of Sumerian royal hymns” lead by Prof. Dr. Kai lämmerhirt at the University of Heidelberg. Her current project focuses on the colophons in literary, lexical and scientific manuscripts of the 3rd and 2nd millennium BCE.

Robert Turnbull studied Astrophysics and Mathematics at Monash University before working in Geodyamics in the Monash Cluster Computing research group. Robert then studied the Master of Divinity at Ridley College in Melbourne where he received the award for the top graduating student of his year. He is completing his PhD studies looking into a family of Arabic manuscripts of the Gospels under the supervision of Alexander Treiger (Dalhousie University) and Michael Bird (Ridley College). He currently works as a Research Data Specialist at Melbourne University.

PREFACE Up to the middle of the previous century, philology—in its broadest sense—focused on the literary productions of the ancient world.1 Scholars painstakingly produced critical editions comparing dozens of manuscripts. They analyzed manuscript genealogies and produced theories of transmission history. Most importantly, they concentrated on determining the original text with little attention to the actors who actually transmitted these texts in the forms we have received them. Other paratextual elements such as margin glosses, commentaries, or comments resulting from collations with different copies were largely ignored. Scribes were relegated to a secondary position. Users, readers, and owners held a tertiary position. Their only hope of appearing in our stories was to produce an interesting factoid in a colophon or to scribble a date or a name that is of interest. As such, we only heard part of the story. This is not to say that eighteenth and nineteenth century philology did not recognize the role of the scribe. Johann Jakob Griesbach (1745–1812) is a case in point. His first canon in resolving the synoptic problem of the Gospels defined the maxim brevior lectio, Brevior lectio, nisi testium vetustorum et gravium auctoritate penitus destituatur, praeferenda est verbosiori. Librarii enim multo proniores ad addendum fuerunt, quam ad omittendum. A shorter reading, unless it stands completely without the support of ancient and important witnesses, is to be preferred to a

James Turner, Philology, The Forgotten Origins of the Modern Humanities (Princeton, 2014). 1

xi

xii

SCRIBAL HABITS more verbose one. Copyists [italics ours] were much more inclined to add than to omit.2

Griesbach was not the first to recognize scribes as literary actors. During the seventh century, Jacob of Edessa (640–708) was quite strict with his own scribes, I prohibit all those who copy the books which I have translated or composed from changing, in their own will, anything, either in the writing or in the dotting.3

But it is not only until the 1960s when we begin to hear of scribal habits giving the scribes the spotlight that they deserve. This was advocated first by the New Testament textual critic E. C. Colwell and then followed by J. R. Royse.4 We began to hear of “scribal behavior” and how that affected text transmission. Is the scribe disciplined or sloppy? Is the scribe concerned with transmitting the meaning of the text or a literal, formal representation of the exemplar at hand? Colwell and Royse laid the foundations of our modern understanding of scribal habits. Yet, they too focused solely on the primary literary element found in the manuscripts, the literary text in question. During the past few decades, there has been an increased interest in scribal habits. Firstly, the subject is no longer exclusively connected to New Testament textual criticism. But more importantly, scribal habits scholarship began to consider paratextuality. The manuscript—as material culture—and its actors (plural) now stand in the spotlight. The original author of the literary work being transmitted, when we speak of scribal habits, is now relegatTranslation from P. J. Williams, Brevior lectio - history of the concept, post (Nov. 12, 2005) at http://evangelicaltextualcriticism.blogspot.com/ 2005/11/brevior-lectio-history-of-concept.html. 3 Jacob of Edessa, Letter on Orthography ‫ܗ‬. 4 E. C. Clowell, “Scribal Habits in Early Papyri: A Study in the Corruption of the Text” in J. P. Hyatt (ed.), The Bible in Modern Scholarship (Nashville, 1965; republished in 1969) 370–389; J. R. Royse, “Scribal Habits in the Transmission of New Testament Texts” in W. D. O’Flahherty (ed.), The Critical Study of Sacred Texts (Berkeley, 1979) 139–161. 2

PREFACE

xiii

ed to a tertiary position. This gave rise to New Philology whereby the entirety of the manuscript content formed the scholar’s opinion about any text.5 The manuscript, as an object, becomes the stage. Its nonliterary elements—medium (tablet, scroll, papyrus, codex) and constitutive facets (sheets, quires, columns, ink color)—become the set. But how about the actors? The original scribe, the readers who left annotations, the commentators who provided marginal notes and sometimes longer commentaries, and the dotters who disambiguated readings. Then there are the illuminators, binders, and a chain of users, all of whom left traces in colophons and added other paratextual elements. Obviously, the original author wrote the play but it is these actors who performed it. This is how the manuscript became a living document. It is the play that we aim to bring to the spotlight in this volume. The papers presented here are the result of two workshops held at the Institute of Advanced Study, Princeton, in 2018 and 2019. The first focused on paratextual elements, in other words, everything but the main texts; the second focused on the main texts but from the point of view of our actors, the scribes and the paratext contributors, not the original author. Our aim in both workshops was, in the words of J. R. Royse, to “virtually look over the scribe’s shoulder” in order to bring out the full story of the manuscript. The volume consists of fourteen papers. They span various cultures of the Middle East and demonstrate a cross cultural collective of scribal habits. Linguistically, the current authors cover Arabic, Hebrew, Persian, Syriac, and Turkish—one paper even Sumerian. With regards to the paratextual objects under discussion, they range from macro-elements (chapter headings and chapter divisions) to micro-elements (dotting and diacritical marks) and anything in between (marginalia and supralinear annotations). For New Philology, see, e.g., Snapshots of Evolving Traditions: Jewish and Christian Manuscript Culture,Textual Fluidity, and New Philology. Edited by Liv Ingeborg Lied and Hugo Lundhaug (Texte und Untersuchungen zur Geschichte der altchristlichen Literatur 175. Berlin: De Gryuter, 2017). 5

xiv

SCRIBAL HABITS

The long discussions during our workshops demonstrated that scribal habits transcend language and faith. The main literary authors are distinctly Jewish, Christian, or Muslim. But the acts of our actors—the scribes, readers, owners, and users—transcend faith and linguistic traditions. Our actors performed. Their craft allowed an interdisciplinary band of scholars—the paratext critics—to produce this volume.

George A. Kiraz and Sabine Schmidtke

CONNECTING THE DOTS: USING DIAERESIS AS A SOURCE OF INFORMATION ABOUT SCRIBAL PRACTICES IN BYZANTINE EGYPT ELIZABETH BUCHANAN This article makes the argument that although diaeresis, a set of two dots placed above certain vowels in Greek writing, has not been well-studied in recent years because of a perception that its use is idiomatic and not useful, a close study of this practice in a contained set of documents, specifically, the documents in the archive of Dioscorus of Aphrodito (c. 520585 CE) prepared by notaries or contract writers, can be useful in providing dating criteria, information about scribal training and networks, and the distribution of legal clauses in sixth-century CE Aphrodito, Egypt. Different notaries and contract writers used diaeresis differently, however, some consistent patterns can be seen. For example, all the notaries and contract writers whose names should take diaeresis used it for their own names, if not necessarily those of other parties to the document. Amanuenses also used it for their names. Witnesses used it inconsistently, even for their own names when it would have been appropriate. In addition, the use of diaeresis increased between the 520s and the 580s, and the number of words that routinely took it also expanded. This may be related to an expanded use of diaeresis by imperial authorities when they distributed clauses, such as the dating clause, to the notaries. Although most notaries used diaeresis for appropriate names of imperial officials in dating clauses, some did not use it in clauses such as the oath clause for the names of the same imperial officials. This suggests that the 1

2

ELIZABETH BUCHANAN notaries were slower to update standard clauses even when using diaeresis in the dating clause. Diaeresis may thus help us better understand both scribal educational practices and the mechanisms for distributing contract clauses.

A diaeresis is a set of two dots placed above certain vowels. Scholars of classical Greek, including Eduard Schwyzer, Leslie Threatte, E.G. Turner and Francis Gignac, after analysing classical and late antique documents, identified three reasons for using the diaeresis, including: (a) to mark when either an iota or upsilon stand at the beginning of a word, (b) to mark two consecutive vowels not forming a diphthong, and (c) in foreign names.1 Although occasionally attested in the early Roman period, it became more common after the second century CE.

An example of a diaresis in Victor’s name–from P. Flor. III 280.292 For discussion on the diaeresis in ancient or late antique Greek, see Ruth Barbour, Greek Literary Hands, A.D. 400–1600 (Oxford: Clarendon Press, 1981), xxix; Francis Thomas Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, Vol. 1, Phonology (Milan: Istituto Editoriale Cisalpino, 1976), 205–207; Eduard Schwyzer, Griechische Grammatik, Auf der Grundlage von Karl Brugmanns Griechischer Grammatik (Munich: C.H. Beck’sche Verlagsbuchhandlung, 1939), 199–200; Leslie Threatte, The Grammar of Attic Inscriptions, Vol. 1 Phonology (Berlin: Walter de Gruyter, 1980), 94–97; E.G. Turner, Greek Manuscripts of the Ancient World, 2nd ed., ed. P.J. Parsons (London: Bull. Inst. Class. Stud. London 46, 1987), 12– 13; Raffaella Cribiore, Writing, Teachers, and Students in Graeco-Roman Egypt (Atlanta: Scholars Press, 1996), 83–84. 2 http://www.misha.fr/papyrus_bipab/pages_html/P_Flor_III_280 .html. Firenze, Biblioteca Medicea Laurenziana, P. Flor. III 280. Su con1

CONNECTING THE DOTS

3

Very little work, however, has been done in recent years on the use of the diaeresis in the largest extant collection of late antique documents: the papyri of Egypt. There are at least two reasons for this. First, and perhaps most importantly, the use of the diaeresis in Byzantine Egypt papyrological texts appears idiomatic, used on some words within a document and not on others. For example, the contract writer, Abraamios, in the text of P. Cair. Masp. III 67296 (535),3 used a diaeresis on “Flavius” in the dating clause (67296.1) and for a senior official (67296.3), on υἱός (67296.7,16,18 — the latter two were written by the amanuensis of the subscribing parties), and on ὑμεῖς in one location (67296.14) but not in another location (67296.12), nor on ὑπὲρ (67296.15, 19 — the last by an amanuensis), ὑποκειμένων (67296.14), or “Flavius Justinian” when used in the oath clause (67296.6). The nomikos Pilatus, in the text of P. Cair. Masp. I 67095 (548), used a diaeresis on “Psais” (67095.6), ὑπὲρ (67095.10, 13, 18), and ὑπογράφοντας (67095.18), but not on the indiction (67095.3,16 — both abbreviated), ὑδραγωγοῦ (67095.13), or ὑπογρ̣αφῆς (67095.18).4 The amanuensis in Pilatus’s document above used a diaeresis on “Psais” (67095.19) and εἰδότος (67095.21 — misspelling it as ἰδότης), but not on ὑπὲρ (67095.21). The second reason, which is perhaps related to the first reason, is that editors are not always accurate in reporting diaereses. For example, the editor for P. Cair. Masp. I 67100.27 does not show a diaeresis on Victor’s name, but a photograph of the papyri clearly shows it. The editor of P. Cair. Masp. II 67128.23 shows the word χρυσοχοικῷ̣ with a diaeresis over the omicron iota, but the photograph does not show one. In all fairness, diaeresis marks are not easy to decipher. Because of the difficulties of deciphering the existence of the dots and the relative lack of information on ancient usages, there may be some reliance on contemporaneous usage to determine where the dots exist. The editor of P. Cair. Masp. I cessione del Ministero per i Beni e le Attività Culturali. E’vietata ogni ulteriore riproduzione con qualsiasi mezzo. 3 See Giovanni Ruffini, Life in an Egyptian Village in Late Antiquity (Cambridge: Cambridge University Press, 2018), 99, for a discussion of the document. 4 Ibid., 91, 109.

4

ELIZABETH BUCHANAN

67094, for example, shows it on words that are no longer extant.5 It may be reasonable to assume that words frequently taking a diaeresis had it, but this practice complicates any analysis of the reasons and implications of the diaresis. The diaeresis is not only reflected in late antique documents, but also in late antique education exercises. Raffaella Cribiore, in her book Writing, Teachers, and Students in Graeco-Roman Egypt, notes that the diaeresis is frequently indicated in school exercises, starting in around the second century CE.6 Citing Turner, she says that it can be placed over a vowel to indicate that that vowel is pronounced as a separate syllable; it can also appear over an initial or final vowel, most often on iota or upsilon. She further states that the first use, over a vowel to indicate that it is pronounced as a separate syllable, is rarely used, even by teachers. The second use, over the initial or final vowel, is more common, although frequently misused or used inconsistently.7 Finally, she states that there is evidence in Byzantine Egypt that the letters iota and upsilon may have been taught with a diaeresis, although this does not explain the inconsistent use of the diaeresis on these letters. Why should we care about the use of the diaeresis? I am proposing in this paper that a careful examination of the use of the diaeresis in a controlled set of documents—specifically, the documents in the archive of Dioscorus of Aphrodito, prepared by notaries or contract writers—can be useful in providing dating criteria, information about scribal training and scribal networks, and the distribution of legal clauses in sixth-century8 Aphrodito, Egypt. Dioscorus of Aphrodito was a legally trained landowner and village official who lived in Aphrodito, Egypt, from about 520 to perhaps 585, with a short period of residency in Antinoopolis from 565 P. Cair. Masp. I 67094 (552/3).1 (Flavius) and 11 (Ioannos). Cribiore, Writing, Teachers, and Students, 83–84. For a general discussion of education in the Thebaid, see Laura Cavero, Poems in Context: Greek Poetry in the Egyptian Thebaid 200–600 AD (Berlin: De Gruyter, 2008) and R.A. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity (Berkeley: University of California Press, 1988). 7 Cribiore, Writing, Teachers, and Students, 84. 8 All dates in this paper are CE, unless otherwise noted. 5 6

CONNECTING THE DOTS

5

until about 573.9 His status after 573 is unclear. Jean-Luc Fournet has proposed, based on the remaining documents in the archive, that Dioscorus may have died or retired to a monastery around 573, leaving his wife, Sophia, in charge of their estates.10 Dioscorus and his family (including perhaps his spouse) left an archive that is comprised of approximately 680 edited documents, the majority of which are legal documents.11 The legal documents from 520 until 565 are largely comprised of personal legal transactions involving Dioscorus’s family and documents, such as petitions, tax receipts and surety agreements, arising from Dioscorus’s and his father’s responsibilities as village officials. The documents dating from the period in which he lived in Antinoopolis, a larger city nearby, are comprised of legal documents, some written in his own hand, plus poems that he wrote. These documents appear to be samples of legal formats because they do not involve Dioscorus or known members of his family, and they generally lack a notarial clause. There are a few documents post-573, including a divorce contract and a few leases and tax receipts. Based on a poem written by Dioscorus to Victor the Prefect in 566/7 asking for an appointment as a notary,12 and on a slightly later poem written by him to John the notary (nomikos) thanking him for bringFor information about Dioscorus of Aphrodito, see Jean-Luc Fournet and Caroline Magdelaine, eds., Les archives de Dioscore d’Aphrodité cent ans après leur découverte: histoire et culture dans l’Egypte byzantine. Actes du Colloque de Strasbourg, 8–10 décembre 2005 (Paris: De Boccard, 2008); James Keenan, “The Aphrodite Papyri and Village Life in Byzantine Egypt,” Bulletin de la Société d’Archéologie Copte 26 (1984): 51–63; Leslie S.B. MacCoull, Dioscorus of Aphrodito: His Work and His World (Berkeley: University of California Press, 1988); Constantin Zuckerman, Du village à l’empire: autour du régistre fiscal d’Aphroditô (Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2004). 10 Jean-Luc Fournet, “Archive ou archives de Dioscore? Les dernières années des ‘archives de Dioscore,’” in Les archives de Dioscore d’Aphrodité cent ans après leur découverte, actes du Colloque de Strasbourg, 8–10 décembre 2005, ed. Jean-Luc Fournet (Paris: De Boccard, 2008), 17–30. 11 More data on this archive is available at “Dioskoros,” Trismegistos, accessed April 29, 2018, www.trismegistos.org/archive/72. 12 P. Cair. Masp. II 67131 (566/7). 9

6

ELIZABETH BUCHANAN

ing ‘the inexperienced’ Dioscorus in,13 it is widely believed that he worked on John’s staff.14 Although he is known to have written fiftysix legal documents in his own hand,15 I am unaware of any legal document in which Dioscorus identified himself as a notary or contract writer. In this paper, I will use the term ‘notary’ to refer to what in Greek is a “nomikos” and in Latin is a “tabellion.”16 These are officials who were legally required to verify the intentions of their clients who came for the preparation of legal documents; their documents were then given deference when challenged.17 There was a distinction made in Byzantine Egypt between “nomikoi” and “notarioi.” This latter group are attested in papyri dated to the fourth through eighth centuries, most frequently in the sixth and seventh centuries. They appear in both Greek and Coptic documents. A notarios appears to be an official assigned to a city, estate or senior official such as a pagarch who was responsible for conveying or executing the instructions of their employer18 and for participating in the collection of taxes.19 These notarioi rarely prepared private legal documents, although they may have acted as witnesses.20 The term “notaries” in this paper does not apply to officials with the title of notarioi. P. Lit. Lond. 100F (ca. 567), translated by MacCoull, 85. See, for example, MacCoull, 85. 15 Lucio Del Corso, “La Scritture di Dioscoro,” in Les archives de Dioscore d’Aphrodité cent ans après leur découverte: histoire et culture dans l’Egypte byzantine. Actes du Colloque de Strasbourg, 8–10 décembre 2005, ed. Jean-Luc Fournet (Paris: De Boccard, 2008), 89–115, 91–94. 16 These terms appear to be used simultaneously for the same official. See P. Cair. Masp. III 67283, in which Pilatus is described both as a tabellion (page 2, line 15) and as a nomikos (page 2, line 19). 17 N. Jus. 73 (538). 18 See, for example, P. Apoll. 20 (VII), P. Apoll. 26 (VII), P. Oxy. 61 4131 (600). 19 For example, see P. Lond. V 1746 (VII-VIII), SB VI 9144 (post 641), SB XVI 13018 (714), SB XXII 15716 (mid-VII). 20 For examples of notaries as witnesses, see P. Coll. Youtie II 92 (569), P. Gen. IV 181 (early VII), P. Koeln. III 156 (late VI) — which has two notaries as witnesses but is written by another individual. 13 14

CONNECTING THE DOTS

7

In addition to notaries, there were also contract writers (“sumbolaiographoi”), but it is unclear what the difference was in the sixth century between notaries (nomikoi/tabellions) and contract writers (sumbolaiographoi). They appear to have prepared similar types of documents. For example, Isakios in 514 wrote a contract for shepherd and guard services, identifying himself as a notary.21 In that same year, he wrote a land rental agreement, also identifying himself as a notary.22 Abraamios wrote similar documents in roughly the same period. He frequently did not provide a title for himself;23 but when he did, it was always as a contract writer.24 The fact that individuals generally identified themselves as either notaries or contract writers but not as both implies that there was a difference between the positions.25 Unfortunately, I am aware of no evidence concerning the nature of the difference. Many of the documents in Dioscorus’s archive, however, identify a notary or contract writer. There are fifteen identified notaries or contract writers who executed legal documents in Dioscorus’ archive for Aphrodito, plus one unidentified notary, for a total of eighty-three documents.26 With the exception of Abraamios, son of Apollo, all from Aphrodito identified themselves as notaries, either as nomikoi or tabellions. Abraamios identified himself consistently as a contract writer and never as a notary. There are seven notaries or contract writers identified in Dioscorus’s archive for Antinoopolis, on a total of nine documents.27 All but one of the document preparers for the Antinoopolis documents identified themselves as contract writers rather than notaries, if they provided P. Cair. Masp. I 67001 (514). P. Flor. III 279 (514). 23 See, for example, SB XXVI 16529 (526), P. Michael 43 (526), and P. Mich. XIII 670 (527). 24 See, for example, P. Cair. Masp. III 67296 (535), P. Cair. Masp. III 67327 (539), and P. Cair. Masp. I 67112 (544/545). 25 An exception to the general rule applies to Pilatus, who generally referred to himself as a notary, with one exception found in P. Michael 42A (566). 26 See Appendix, attached. 27 Ibid. 21 22

8

ELIZABETH BUCHANAN

a title at all. The only notary for Antinoopolis for whom we have an extant document from Dioscorus’s archive prepared a settlement document dated to 527–547, during which Apollo, father of Dioscorus, was present in Antinoopolis and signed as a witness.28 There are many more legal documents in Dioscorus’s archive beyond the documents described above; but for the purposes of this research, I have only looked at the documents with extant notarial clauses that identified the notary or contract writer. All of the notaries and contract writers whose documents are found in Dioscorus’s archive used the diaeresis, to a greater or lesser extent. Fortunately, there are several for whom we have more than one document, so their use of the diaeresis can be studied in more than one document. This is useful because the specific words in which the diaeresis often appears do not occur in all documents, and many documents are so fragmentary that any diaeresis may have been lost. There are also changes that occurred over time. Based on extant published documents, the following document preparers executed more than one document: Victor, son of Apollo (eight documents); Isakios (eight documents); Abraamios, son of Apollo29 (twenty-one documents); Kuros 130 (eight documents); Pilatus, son of Apollo (thirteen documents); Hermauos (five documents); Kuros 2 (four documents); and Constantinus (three documents). Before moving to the discussion of the ways in which the document preparers used the diaeresis, I found it helpful to divide each document into three sections: the dating clause, the main body of the document, and the execution and witness clauses. This is useful beP. Mich. XIII 659 (527–547). See the discussion in Ruffini, Life in an Egyptian Village, 54–56, in which he proposes that Apollo was representing his niece and her husband, parties in the settlement discussion. 29 The editor to P. Mich. XIII 670 stated that the notary in that document is the same as in P. Michael 43 (25), 45 (76), P. Cair. Masp. 296 (21), 307 (17), 308 (17); he believes that Abraamios, son of Apollo, is the same as the notary that the editor of P. Michael believed was Abraam, son of Apygchios. 30 See the discussion in Giovanni Ruffini, A Prosopography of Byzantine Aphrodito (Ann Arbor: University of Michigan Press, 2011), 336–337 on Kuros 8, and 340 on Kuros 25 (whom I have called Kuros 1 and 2). 28

CONNECTING THE DOTS

9

cause it is not unusual for the same words to be written with a diaeresis in the dating clause but not in the main body of the contract, and for the same names to be written with a diaeresis in the execution and witness clauses but not in the main body of the contract, or vice versa. P. Cair. Masp. III 67296, which I discussed earlier, illustrates this in that Abraamios used a diaeresis on the “Flavius” in the dating clause but not in the imperial oath. As will be discussed towards the end of this paper, these differences are important to our understanding of how the documents were created. To frame this discussion, I have divided it into discussions of specific writing practices, starting with the most common uses of the diaeresis and moving into less common uses.

NAMES

The most common use of the diaeresis in the archive is for (a) names or patronyms beginning in iota (such as Ioannos or Isakios), (b) names or patronyms beginning with omicron-upsilon and iota (such as Victor: Oυϊκτορος), or (c) names or patronyms with a series of internal vowels that are adjacent but not diphthongs, such as Psais or Tasaitos. According to the classical description of the diaeresis, the first use would fall into the category of words beginning with iota or upsilon, and the second and third to the category of adjacent vowels that do not constitute a diphthong. The most important factor, however, appears to be that it is a person’s name. When a notary’s name began with either iota or the omicron-upsilon and iota combination, he always used a diaeresis on his own name. This is true even if he did not use a diaeresis on other names. For example, the notary Isakios, who used a diaeresis for his own name, did not use it for Isak, the name of a party in the body of one of his texts.31 See the table below for notaries with names taking diareses.

(P. Flor. III 279 - Ἰσὰκ (279.3, 20, 24 — the last two are party subscriptions). 31

10

ELIZABETH BUCHANAN Table 1: Notaries and contract writers with names normally bearing diaereses and their incidence

Name of the notary or contract writer Victor, son of Apollo

He spelled his name ‘Oυϊκτορος’ – those who spelled their name ‘Biktoros’ or “Biktor’ did not use diaeresis.

Isakios

Jeremias, son of Victor Amais, son of Abraam

Documents they prepared with diaeresis on their name

Documents they prepared without diaeresis on their name None

Documents in which diaeresis on their name is unclear None

P. Cair. Masp. I 67001.49, P. Flor. III 279.25, P. Cair. Masp. III 67328.pg3.35, P. Cair. Masp. I 67114.frB.7, P. Lond. V 1844.22, P. Cair. Masp. II 67264.frB P. Cair. Masp. III 67283.pg2.15,19 P. Cair. Masp. II 67127.30

None

P. Cair. Masp. II 67262; the first part of the signature is missing

None

P. Cair. Masp. II 67133.632 None

P. Cair. Masp. I 67100.26, P. Flor. III 280.29, P. Cair. Masp. III 67306.21, P. Flor. III 281.22, P. Cair. Masp. III 67301.38, P. Michael 51.30 (also P. Thomas 28.30)

None

Most witnesses and amanuenses also used a diaeresis for their names or patronyms. The use appears to increase both over the course of the sixth century and with the education level of the wit-

The editor sees a diaeresis on this name. I have gone back and forth on whether I see anything. 32

CONNECTING THE DOTS

11

ness or amanuensis. I have created separate tables for the use of the diaeresis by amanuenses and witnesses because many witnesses did not ascribe to the document in their own handwriting—it was done by an amanuensis. All the amanuenses, by definition, wrote in their own handwriting. The table for witnesses only contains information on witnesses that signed for themselves. I did not include all witnesses and amanuenses because there are many of them, often with names that would not take diaeresis; but a sample taken over the period of the archives helps to illustrate the use of the diaeresis by witnesses and amanuenses. Table 2: Sample of the use of the diaeresis by amanuenses

Name of the amanuensis

Date

Docu-

Ioannes, son of Beskouis

514

Abraam, son of Biktor

514

P. Cair. Masp. I 67001.40, 44

Theoteknos, son of Psais33

First half VIc

P. Michael 51.25, 26 (also P. Thomas 28.25, 26)

ment

P. Cair. Masp. I 67001.36, 37, 39

Use of diaeresis

Status Indicators

Used on his first name and on the patronyms of two of his clients, Iosephios and Psais; did not use on ὑπὲρ Used on ὑπὲρ but not on the patronym of a client, Psais, although the last letter is missing Did not use on Psais but abbreviated the name with a line over the last letter; he did use it on ὑπέρ

Deacon(?)

Aurelius

Flavius, reader

See also P. Ross. Georg III 36.24 (537) and P. Cair. Masp. II 67127.23 (544). He is an amanuensis and writes his patronym in the same way as above. He did clearly use diaeresis in his patronym in P. Flor. III 281.20 (517). 33

12

ELIZABETH BUCHANAN

Ioannos, son of Biktor

549

P. Cair. Masp. II 67129.31

Psellos, son of Isakios

553

Dioscorus, son of Iosephius

564/5 or 549/50

P. Cair. Masp. III 67303.22, 23 P. Hamb. I 68.46

David, son of Andreas

584/5, 599/600

Dioscorus, son of Iosephius

566

P. Michael 42A.35, 3934

P. Mich. XIII 664.40, 42

Used on this name, but it is unclear whether he used it on his client’s name, Psais; the document is abraded Used on his patronym and on his client’s name, Psais Used on his patronym and his client’s name, Iacob Used it for his own patronym but not for his client’s name, Iakob Used it on his client’s name, Ioudith, daughter of Ioannos; his name would not have taken it

Aurelius

Aurelius Priest Priest

Priest

Table 3: Sample of the use of the diaeresis by witnesses who signed for themselves Name of the witness Isakios, son of Iosephius Iakob, son of Biktor

Date 515 515

Document P. Cair. Masp. III 67306.18 P. Cair. Masp. III 67306.19

Use of diaeresis Did not use on his name

Used on his first name

Status Indicators Aurelius, no profession given Deacon

There is no photograph available for this document, so I did not personally review it. I am relying on the editor. 34

CONNECTING THE DOTS Enoch, son of Ieremias

530

Ioannes, son First of Abraamios half VIc Constantine, 526– son of Ioannos 548 Psais, son of 526– 548 Apollo Mathias, son 544 of Iosephius Ioannes, son of Biktor Menas, son of Iosephius

544 549

P. Cair. Masp. III 67301.36,37 P. Michael 51.26

P. Bingen 130.29 P. Bingen 130.31 P. Cair. Masp. II 67127.25 P. Cair. Masp. II 67127.27 P. Cair. Masp. II 67129.32, 33 P. Hamb. I 68.47

Psais, son of 564/5 or Apollo 549/50 566 P. Michael Paulos, son 42A.44, 45 of Ioannes Isakios, son of Ioannes – declarant

573

P. Cair. Masp. I 67121

Did not use either on his patronym or on ὑπὲρ Used on his first name

13 Aurelius, no profession given Aurelius

Used on his patronym Used on his first name Did not use on his patronym

Aurelius

Used on his patronym

Flavius

Priest

Aurelius

Did not use on his name

Aurelius

Used on his name

Priest

Used diaeresis for the word ϋποθηκης but not for his patronym Used for name and patronym

Aurelius Aurelius, doctor

You can see from these tables that all the listed amanuenses, except for possibly Theoteknos, son of Psais, used a diaeresis on their own names, and most used it for their clients as well. It is very difficult to determine whether Theoteknos, son of Psais used a diaseresis on Psais because he abbreviated his patronym and used a line over the final omicron. This line could, and often does, represent an abbreviation, but in this case, it could also substitute as a diaeresis. Witnesses were more inconsistent in their use of the diaresis. About half used it for their names, but there were exceptions. The witness Palous, son

14

ELIZABETH BUCHANAN

of Ioannos in P. Michael 42A used a diaeresis for the word ϋποθηκης in his witness clause (P. Michael 42A, 44), but not for his own patronym Ioannos (42A.45).35 A witness in P. Vat. Aphrod. 4, Makarios, son of Kollouthos, did not use it on his name, which would not ordinarily take it, but did use it for his profession (ἰατρός, doctor) and for the word “witness” (μαρτυρῶ).36 The latter word should not have taken diaeresis because it is not a word beginning in iota or upsilon or a word with adjacent vowels not pronounced as a diphthong. This may indicate that the witness or amanuensis saw a diaeresis used on a word in the document and copied that use without understanding the common reasons for the use.

WORDS OTHER THAN NAMES

The second most common use of the diaeresis is for certain words (other than names) beginning with iota or upsilon or with adjacent vowels that are not a diphthong. Some words were frequently written with a diaeresis. Notaries and contract writers used a diaeresis above the upsilon-iota for the word υἱός (son) in almost all cases.37 It should be noted that the word υἱός is rarely used. Most patronyms are reflected in the genitive, but sometimes, perhaps to stress the relationship, υἱός will be used. Another word that is frequently shown with diaeresis is ἴδιος. Because of the frequency with which these words appear with a diaeresis, it is likely that they were taught with a diaeresis. Other words vary considerably as to whether a diaeresis is used. “Indiction” is an example of a word in which the usage varied significantly. By way of background, “indiction” was often written in an abbreviated form (“ind”) with a stylized ‘X’ after it. In the early years of the sixth century, the abbreviated form rarely took a No photo is available for this document, so I could not tell if someone else wrote the word ϋποθηκης; but the edition does not indicate a change of hand mid-clause. 36 P. Vat. Aphrod. 4 (late VI). 4 frC.22. 37 An exception is P. Cair. Masp. III 67325 IVr.27 (585), but the word υἱός in that line is added above another line and is crammed into a small space. The subsequent line, which also uses υἱός but is not crammed into a small space, used a diaeresis in the word. 35

CONNECTING THE DOTS

15

diaeresis. Over time, it became more common to see the entire word written out, and the presence of a diaeresis became more common for both the whole word and its abbreviated form. A related issue involves the word “Flavius.” “Flavius” was often abbreviated in texts to “Fl.” With this abbreviation, there are no vowel chains for which to use a diaeresis. Later, the entire word started to appear in dating clauses and sometimes for senior officials. The diaeresis appeared more frequently in these cases. Other words, like ὑπάρχοντα, ὑπατεία, ὑπέρ, and ὑπογράφης show inconsistent use of the diaeresis in the early sixth century. The word ὑπέρ in particular is difficult to assess because it is sometimes written by the notary but is more often written by a witness or amanuensis. In the latter situation, it is less likely to have a diaeresis. The usage appears to increase by the mid-sixth century, and it also appears to have been more common in Antinoopolis than Aphrodito, although this could also be a factor of the time period; most of the Antinoopolis documents are later in the century. See table for usage of the diaeresis on certain words. Table 4: Usage of the diaeresis on certain common words Word

υἱός - son

Found with Diaeresis

P. Flor. III 280.3,7 (514), P. Cair. Masp. I 67001.6,7 (514), P. Flor. III 281.5 (517), P. Cair. Masp. III 67328, pg 3.28 (521), P. Cair. Masp. III 67301.6,7 (530), P. Cair. Masp. III 67296.7,16,18 (535), P. Cair. Masp. II 67128.10 (547), P. Cair. Masp. III 67303.9 (553), P. Lond. V 1692a.5 (555), P. Mich. XIII 668.1 (555/540), P. Michael 40.30, 61 (559/545), P. Michael 46.5 (559), P. Cair. Masp. I 67109.8 (565), P. Michael 42A.2,12 (566), P. Cair. Masp. I 67121.4

Found without Diaeresis P. Cair. Masp. III 67325 IVr.27 (585 – this line is crammed up above another line; the following line uses it for the same word)

16

ὑπέρ – on behalf of

ἴδιος - same

ELIZABETH BUCHANAN (573), P. Mich. XIII 664.8,11 (599/600), P. Cair. Masp. III 67325 IV.7,42,45 (585), P. Cair. Masp. I 67115.13, 14 (midVI), P. Cair. Masp. II 67154r.4 (midVI) P. Flor. III 280.10,11 (514), P. Flor. III 279.18 (514), P. Cair. Masp. I 67001.34,44 (514), P. Cair. Masp. III 67300.2, 11 (527), P. Cair. Masp. III 67301.24, 29 (530), P. Ross. Georg. III 36.17,24 (537), P. Cair. Masp. III 67327.43 (539), P. Cair. Masp. II 67127.24 (544), P. Ross. Georg. III 37.10,18 (545), P. Cair. Masp. II 67128.13, 29 (547), P. Cair. Masp. I 67095.10,13,18 (548), P. Cair. Masp. I 67116.3,8 (548), P. Lond. V 1661.11,12,19 (553), P. Lond. V 1692a.19 (555), P. Michael 46.16, 25 (559), P. Cair. Masp. I 67109.43,47 (565), P. Cair. Masp. I 67110.40,47 (565), P. Michael 42A.4 (566), P. Michael 59.10 (VI), P. Mich. XIII 664.36 (599/600), P. Hamb. I 68.31, 32, 38(564/5, 549/550) P. Cair. Masp. I 67001.32 (514), P. Michael 43.7 (526), P. Cair. Masp. III 67300.10 (527), P. Cair. Masp. III 67301.22 (530), P. Heid. V 351.8 (534/5),

P. Cair. Masp. I 67100.17 (506), P. Lond. V 1701.7, 12 (510, 525, 540), P. Cair. Masp. I 67001.29 (514), P. Cair. Masp. III 67328.20 (521), SB XXVI 16529.27 (526), P. Michael 43.15, 18 (526), P. Cair. Masp. III 67300.20? (527), P. Cair. Masp III 67301.37 (530 – witness clause). P. Michael 51.22 (1st half VI), P. Koeln. X 421.34 (524–545), P. Cair. Masp. III 67308.5 (526–541), P. Bingen 130.27 (526–548), P. Cair. Masp. III 67296.15, 19 (535), P. Cair. Masp. III 67327.3,9,17,24 (539), P. Cair. Masp. I 67112.26 (544/5 – amanuensis), P. Lond. V 1661.25 (553 – amanuensis), P. Lond. V1692a.23 (555 – amanuensis), P. Hamb. I 68.47 (564/5, 549/50 – amanuensis), PSI VIII 936.2,3 (1st half VI), P. Cair. Masp. II 67251.2 (549), P. Michael 46.15 (559), P. Straus. I 46.27 (566) P. Michael 45.4 (540)

CONNECTING THE DOTS

Indiction – the fifteen-year tax cycle; often used to date documents

P. Ross. Georg. III 36.16 (537), P. Lond. V 1696.12 (1st half VI), P. Cair. Masp. III 67303.19 (553), P. Lond. V 1661.21 (553), P. Cair. Masp. II 67161.4 (566), P. Michael 40.36 (559/545), P. Michael 46.12 (559), P. Hamb. I 68.12 (564/5), P. Cair. Masp. I 67109.34 (565), P. Mich. XIII 664.25, 30 (584/5, 599/600), P. Vat. Aphrod. 4.frC.12 (lateVI), P. Vat Aphrod 5.frC.3 (lateVI), P. Mich. XIII 663.19 (VI), P. Mich. XIII 671.11 (midVI) P. Flor. III 280.2, 16 (514), P. Flor. III 281.3,10 (517), SB V 8029.1 (537), P. Cair. Masp. III 67327.3 (539), P. Cair. Masp. III 67303.3 (553), P. Lond. V 1661.14 (553), P. Lond. V 1692a.10 (555), P. Michael 46.8 (559), P. Michael 40.28 (559/545), P. Flor. III 286.A.4 (552), P. Michael 46.8 (559), P. Hamb. I 68.14 (564/5, 549/50), P. Cair. Masp. I 67109.18 (565), P. Cair. Masp. I 67110.18 (565), P. Michael 43.5,15 (572), P. Mich. XIII 664.27 (584/5, 599/600), P. Cair. Masp. III 67325 IVr.40 (585) – earlier entry very abraded

17

P. Flor. III 279.2,8,17 (514), P. Cair. Masp. I 67001.22 (514), P. Cair. Masp. III 67306.1 (515), P. Cair. Masp. III 67328.p3.12 (521), P. Cair. Masp. III 67307.6, 12 (524/539), SB XXVI 16529.2 (526), P. Cair. Masp. III 67300.1,7 (527), P. Mich, XIII 670.11 (527), P. Cair. Masp. III 67301.4,26 (530), P. Heid. V 351.4 (534/5), P. Cair. Masp. III 67296.1 (535), P. Ross. Georg. III 36.2 (537), SB V 8029.2 (537), P. Cair. Masp. III 67327.9,11,42 (539) – very abraded, P. Cair. Masp. III 67308.3,4 (1st half VI), PSI VIII 937.3,6 (1st half VI), P. Cair. Masp. II 67133.4,5 (542), P. Ross. Georg. III 37.2 (545), P. Cair. Masp. II 67128.22 (547), P. Cair.

18

ELIZABETH BUCHANAN

P. Cair. Masp. III 67328.1 (521), P. Cair. Masp. III 67301.1 (530), P. Ross. Georg. III 36.1 (537), SB V 8029.1 (537), P. Ross. Georg. III 37.1 (545), P. Cair. Masp. II 67128.4 (547), P. Cair. Masp. II 67129.1 (549), P. Cair. Masp. III 67303.2 (553), P. Lond. V 1661.3 (553), P. Cair. Masp. I 67110.1 (565) P. Flor. III 281.2 (517), P. φλαυἱός – a Cair. Masp. III 67301.1 title of respect; when written in (530), P. Cair. Masp. III 67296.1,3 (535), SB V full 8029.25 (537), P. Michael 45.1 (540), P. Cair. Masp. II 67127.6 (544), P. Cair. Masp. I 67094.4 (552/3), P. Lond. V 1692a.5 (555), P. Cair. Masp. I 67109.1 (565), P. Cair. Masp. III 67325.2 (585) P. Cair. Masp. I 67094.7 ὑπογράφης (552/3), P. Michael 40.20 subscription (559/545), P. Cair. Masp. I 67109.43,44 (565), P. Mich. XIII 664.19, 36 (599/600) ὑπατεία - consulship

ὑπάρχοντα property

P. Cair. Masp. I 67100.12 (506), P. Flor. III 279.8 (514), P. Cair. Masp. III

Masp. I 67116.3,6 (548), P. Cair. Masp. II 67129.2,17,29 (549), P. Michael 46.2 (559) P. Flor. III 280.1 (514), P. Cair. Masp. III 67306.1 (515), P. Cair. Masp. III 67300.1 – very abraded (527), P. Cair. Masp. II 67252.1 (538), P. Lond. V 1692a.2 – very abraded (555), P. Michael 46.2 (559), P. Cair. Masp. I 67121.1 (573), P. Michael 48.1 (572) P. Cair. Masp. III 67296.6 (535) – although it has a line over the omicron, which may indicate an abbreviation or perhaps diaeresis

P. Cair. Masp. I 67095.3 (518), P. Cair. Masp. III 67307.13 (524/539), SB XX 15202.18 (1st half VI), PSI VIII 936.5 (1st half VI), PSI VIII 937.6 (1st half VI), P.Lond. V 1701.7,8 (1st half VI), P. Cair. Masp. II 67128.30 (547), P. Michael 40.52 (559/545), P. Ross. Georg. V 37.12 (midVI), P. Mich. XIII 663.29 (VI) SB XXVI 16529.10 (526), P. Michael 46.8 (559), P. Michael 40.55,56

CONNECTING THE DOTS

χρυσοχοικός – goldsmith; diaeresis found over the omicron-iota

67301.14 (530), P. Cair. Masp. II 67127.17 (544), P. Cair. Masp. III 67303.13 (553), P. Lond. V 1661.21 (553), P. Lond. V 1692a.19 (555), P. Mich. XIII 672.7 (557?), P. Ross. Georg. III 36.12 (537), P. Michael 48.16 (572), P. Mich. XIII 664.10, 38 (599/600), P. Vat. Aphrod. 4.fr.c10,11 (later VI), P. Vat. Aphrod. 5.fr.c12 (later VI) P. Lond. V 1701.4 (510, 525, 540), P. Mich. XIII 664.16 (599/600)

19

(559/545), SB XIV 11855.13 (547?), P. Cair. Masp. III 67325.18 (585), P. Mich. III 663.32 (VI)

P. Cair. Masp II 67128.23 (547), P. Cair. Masp. I 67116.5 (548), P. Lond. V 1692a.18 (555)

There appear to have been at least four reasons for the differences, perhaps interconnected. First, certain words were generally abbreviated. Examples include “indiction” and “Flavius.” When abbreviated, the word “indiction” was often written with a symbol behind it. When these words were abbreviated, they were rarely used with diaeresis. Exceptions became more common in the later sixth century, as did use of the entire word. It would appear, however, that the abbreviated form of “indiction” was learned as a block and was slower to change. The word “Flavius” was frequently abbreviated to “Fl,” and when abbreviated would not take a diaeresis. As can be seen from the chart, when it was written in full, it generally took a diaeresis. Second, there is a wide difference in how notaries and contract writers, even in the same location and period, used the diaeresis. Abraamos, son of Apollo, contract-writer, rarely used it, except in a handful of words. He has twenty-one extant documents, and those with secure dates span the period from 524 through 544/5. He used a diaeresis on the word ἴδιος with some regularity.38 He 38

P. Michael 43.7, P. Heid. V 351.8, P. Lond. V 1696.12.

20

ELIZABETH BUCHANAN

occasionally used it for other words. For example, he used a diaeresis on “Flavius” in the dating clause, and also when applied to a party.39 As discussed above, this document also contains an oath to the Emperor Flavius Justinian, whose name and title were not written with a diaeresis, even though it does appear earlier in that same document in the dating clause.40 Abraamios can be contrasted with Victor, son of Apollo. Victor’s extant documents with secure dates range from 506 to 530, overlapping with those of Abraamios. Victor regularly used a diaeresis for his own name, and for the names of the parties when they should use it.41 In the body of his documents, Victor used the diaeresis for words such as ὑπάρχοντα,42 υἱός,43 and ὑπὲρ.44 With one exception (P. Cair. Masp. III 67301.1, in a dating clause), he did not use a diaeresis on “Flavius” or on most other words beginning with iota or upsilon. He did, however, usually abbreviate “Flavius,” which would explain the lack of a diaeresis. Thus, while Victor was a contemporary of Abraamios, he used the diaeresis more regularly, although not as much as later writers. Pilatus’s use of the diaeresis between 547 and 552 is consistent with that of his colleagues. He used it on customary words for the time, such as υἱός45 and ὑπὲρ.46 He also used it on Coptic names to indicate two consecutive vowels not forming a diphthong, as in Tασαϊτος47 and Σαϊαϊτε,48 as well as the more common Psais.49 In a document dated 553, he used the diaeresis on even more words.50 P. Cair. Masp. III 67296.1,3 (535). P. Cair. Masp. III 67296.6. 41 P. Flor. III 281 (517) – Psais (281.6); and Ioannos (281.5,13). 42 P. Cair. Masp. III 67301.14 (530). 43 P. Flor. III 280.3 (514), P. Flor. III 281.5 (517), P. Cair. Masp. III 67301.6 (530). 44 P. Flor. III 280.10, P. Cair. Masp. III 67301.24. 45 P. Cair. Masp. IΙ 67128.10 (547). 46 P. Cair. Masp. IΙ 67128.13, 30, 34; P. Cair. Masp. I 67095.10 (548). 47 P. Cair. Masp. IΙ 67128.7 (547); P. Cair. Masp. ΙI 67129.6 (549). 48 P. Cair. Masp. II 67144.3 (VI). 49 P. Cair. Masp. I 67095.19 (548); P. Cair. Masp. III 67303.4, 22 (553). 50 See P. Cair. Masp. III 67303 (553). 39 40

CONNECTING THE DOTS

21

In that document, Pilatus used the diaeresis on the Greek words for “consulship” (ὑπατεία), “Justinian” and “indiction.” In no extant document did Pilatus use a diaeresis on the word “Flavius,” partly because he generally abbreviated the word; but even when writing it out, he did not use a diaeresis on it. After 553, he used more diaereses if the document was complex or involved more senior people, and fewer if the document was simple. By way of example, in a receipt for a rent payment to a landlord, he only used a diaraesis for “Ioannos,” the patronymic of the landlord, and for υἱός, but not for “indiction.”51 However, in a document for the sale of land in 559/545, Pilatus used the diaeresis for the names of the parties beginning in iota, and the words “indiction,” ἴδιος, and ϊσχυρός.52 He did not, however, use it on ὑπογραφῆς, ὑποκειμένων, ὑπαρχόντων, or ἰδικῶς.53 So while Pilatus increased the number of words on which he used the diaeresis over time, he did not use it on all words beginning with iota or upsilon. Third, there are differences over time. The previous section discussed the use of the diaeresis by notaries in the early to midsixth century. The usage of the diaeresis increased over the span of the sixth century. An example of a later notary is Constantinus, who regularly described himself as a tabellion.54 There are three extant documents from Constantinus dated to the later sixth century. In these, he made extensive use of the diaeresis, putting it on most words beginning in iota or upsilon, including names. Examples from P. Mich. XIII 664 include: ὕπαρχον (664.10, 38), χρυσοχοικος (664.16), ἴδιος (664.25, 30), indiction (664.27), ὑπογράφης (664.19, 36), ὑπὲρ (664.36), ὑπέρκειμαι (664.37), ἰδικῶς (664.38), and ὑποθήκη (664.39). Fourth, there appears to be a difference between the use of the diaeresis in dating clauses and in the body of the text. Not all P. Mich. XIII 668 (555/540).1. P. Michael 40.4, 18, 19, 20, 25, 28, 36, 40, 48, 61 (559/545) (no photograph available for this document). 53 P. Michael 40.52, 55, 55, 57 (559/545) (no photograph available for this document). 54 P. Hamb. IV 265.13 (late VI); P. Mich. XIII 664.50 (584/5, 599/600); P. Vat. Aphrod. 4 frC, 33 (late VI). 51 52

22

ELIZABETH BUCHANAN

the documents in this archive with known notaries or contract writers still have dating clauses extant. Of the documents with dating clauses, some are so abraded that it is difficult to tell whether the words which should take a diaeresis had one. Of the documents with extant dating clauses in relatively good shape, almost all use a diaeresis on at least one appropriate word in the clause. The words on which the diaeresis should have been used changed slightly over the sixth century as the principal dating clause changed from one referring to consulships to one referring to the reign of the governing emperor. In the earlier clauses, we would expect to see a diaeresis on the word for consulship (ὑπατεία) and perhaps on the names of any consuls beginning with iota or omicron, and on “indiction” if not abbreviated. In the later clauses, we would expect to see a diaeresis on “Flavius” (which was generally written out in full) as well as the names of any emperors beginning with iota or omicron, such as Justin (Ioustin) or Justinian (Ioustinian), and on “indiction.” But even when the dating clause used a diaeresis, other parts of the document may not have used it on a similar word. For example, as noted earlier, Abraamos, son of Apollo, used a diaeresis on “Flavius” in the dating clause, and also when applied to a party.55 This document also contains an oath to the Emperor Flavius Justinian, whose name and title were not written with a diaeresis, even though the word “Flavius” appeared earlier in that same document in the dating clause.56 Similarly, Abraamios used a diaeresis on the “Flavius” in the dating clause but not on Ἰουστίνου̣ (the emperor Justin).57 He then put an oath clause in the document in which he did not use a diaeresis for “Flavius Justinian.”58 The notary Amais used a diaeresis on “indiction” in the dating clause but not on the same word later in the body of the document.59 This shows that the dating clause and other clauses, such as the oath and indiction clauses, were treated differently. The dating clause was P. Cair. Masp. III 67296.1, 3 (535). P. Cair. Masp. III 67296.6. 57 P. Michael 45.1 (540). 58 P. Michael 45.17 (540). 59 Compare lines 5 and 16 in P. Cair. Masp. II 67127 (544). 55 56

CONNECTING THE DOTS

23

delivered annually by the imperial government. Other clauses were saved and probably re-used by notaries and contract writers.

CONCLUSION

In conclusion, although the use of the diaeresis is inconsistent, there are patterns that can be detected in sets of documents, such as those in the archive of Dioscorus. These patterns can be used to help identify document preparers and narrow the period in which a document was written. The patterns may also prove helpful regarding our understanding of the education of writers. Some people who learned to write either weren’t trained in the diaeresis or did not feel comfortable using it. Others were more comfortable with it. The latter group often, but not always, included the more accomplished writers: the notaries, amanuenses, priests, deacons, and senior officials. This may mean that the diaeresis was taught later in the writing education or that there were separate schools, perhaps depending upon social position. Finally, the use of the diaeresis may help to provide markers for the development and dissemination of clauses, such as the dating clause and imperial oath clauses. The notaries and contract writers examined in this paper used the diaeresis in most of the dating clauses. Many, especially in the early part of the sixth century, did not use the diaeresis on words used later in the body of the text, even if they may have used the diaeresis on those same words in the dating clause. This seems to indicate that imperial authorities, perhaps on a provincial basis, used the diaeresis in specific clauses that were distributed to the notaries, and the notaries and contract writers used the clause with the diaeresis. The non-use of diaeresis in other clauses, such as oathclauses, suggests that these were standard clauses in the possession of the preparer, which were not changed even with use of the diaresis in the dating clause. The diaeresis may help us understand these mechanisms better. The purpose of this paper has been to show that there are patterns in the use of the diaeresis and that these patterns can be analysed. There are many questions left unanswered, but I hope that this paper has been successful in raising the possibilities that an analysis of diaereses in an archive may offer.

24

ELIZABETH BUCHANAN

BIBLIOGRAPHY

Ast, Rodney. “Writing and the City in Later Roman Egypt. Towards a Social History of the Ancient ‘Scribe.’” CHS Research Bulletin 4, no. 1 (2015). https://nrs.harvard.edu/urn-e:hlnc.essay:AstR.Writing_in_the_City_in_Later_Roman_Egypt.2016 [accessed 19 July 2020].

Barbour, Ruth. Greek Literary Hands, A.D. 400–1600. Oxford: Clarendon Press, 1981.

Bucking, Scott. “On the Training of Documentary Scribes in Roman, Byzantine, and Early Islamic Egypt: A Contextualized Assessment of the Greek Evidence.” Zeitschrift für Papyrologie und Epigraphik 159 (2007): 229–247. Cavero, Laura. Poems in Context: Greek Poetry in the Egyptian Thebaid 200–600 AD. Berlin: De Gruyter, 2008.

Cribiore, Raffaella. Writing, Teachers, and Students in Graeco-Roman Egypt. Atlanta: Scholars Press, 1996. Diethart, J.M. and K.A. Worp, ed. Notarsunterschriften im Byzantinischen Agypten. Vienna: Hollinek, 1986.

Fournet, Jean-Luc, and Caroline Magdelaine, ed. Les archives de Dioscore d’Aphrodité cent ans après leur découverte: histoire et culture dans l’Egypte byzantine. Actes du Colloque de Strasbourg, 8–10 décembre 2005. Paris: De Boccard, 2008. ———. “Le système des intermédiaires dans les reçus fiscaux byzantins et ses implications chronologiques sur le dossier de Dioscore d’Aphrodité.” Archiv für Papyrusforschung und verwandte Gebiete - Beihefte 46.2 (2000): 233–247.

Gignac, Francis Thomas. A Grammar of the Greek Papyri of the Roman and Byzantine Periods. Vol. 1: Phonology. Milan: Istituto Editoriale Cisalpino, 1976.

Kaster, R.A. Guardians of Language: The Grammarians and Society in Late Antiquity. Berkeley: University of California Press, 1988.

Keenan, James G. “The Aphrodite Papyri and Village Life in Byzantine Egypt.” Bulletin de la Société d’Archéologie Copte 26 (1984): 51–63.

CONNECTING THE DOTS

25

MacCoull, Leslie S. B. Dioscorus of Aphrodito: His Work and His World. Berkeley: University of California Press, 1988.

Maspero, Jean, ed. Papyrus grecs d'époque byzantine, Catalogue général des antiquités égyptiennes du Musée du Caire. Vols. 1–3. Cairo: Institut Français d'Archéologie Orientale, 1911–1916.

Ruffini, Giovanni. A Prosopography of Byzantine Aphrodito. Ann Arbor: University of Michigan Press, 2011.

———. Life in an Egyptian Village in Late Antiquity. Cambridge: Cambridge University Press, 2018.

Schwyzer, Eduard. Griechische Grammatik, Auf der Grundlage von Karl Brugmanns Griechischer Grammatik. Munich: C.H. Beck’sche Verlagsbuchhandlung, 1939.

Threatte, Leslie. The Grammar of Attic Inscriptions. Vol. 1: Phonology. Berlin: De Gruyter, 1980.

Turner, E.G. Greek Manuscripts of the Ancient World, 2nd ed. Edited by P.J. Parsons. London: Bull. Inst. Class. Stud. London 46, 1987. Zuckerman, Constantin. Du village à l’Empire: Autour du régistre fiscal d’Aphroditô. Paris: Association des amis du Centre d’histoire et civilisation de Byzance, 2004.

26

ELIZABETH BUCHANAN

City Aphrodito

APPENDIX: LIST OF DOCUMENT PREPARERS Name of Preparer Victor, son of Apollo

Documents with dates P. Cair. Masp. I 67100 (506) – rental of a palm grove P. Flor. III 280 (514) – debt acknowledgement

P. Cair. Masp. III 67306 (515) – receipt for a debt

P. Flor. III 281 (517) – land rental

P. Cair. Masp. III 67301 (530) – land rental

P. Michael 51 (first half VI) – exchange of land

P. Cair. Masp. II 67288 (VI) – list of taxpayers (unknown author) Isakios

P. Cair. Masp. II 67150 (VI) – list of landowners (unknown author) P. Cair. Masp. I 67001 (514) – contract for shepherd and field guards P. Flor. III 279 (514) – land rental

P. Cair. Masp. III 67328 (521) – register of surety agreements – at least one written by Isakios SB XX 14669 (524)60 – Isakios is mentioned once providing a payment for a monastery

P. Flor. III 297 (525)61 – tax receipts

60

Zuckerman, Village a l’Empire, 33–34.

CONNECTING THE DOTS

27

– Isakios is mentioned twice as providing payments on behalf of people

Pap. Congr. XXV (Ann Arbor 2007) p. 399–400 (514–527) – fragment of a land rental contract P. Cair. Masp. I 6711462 (526/6, 511/2) – land rental

P. Lond. V 1844 (early VI) – debt acknowledgement – probably short form

P. Cair. Masp. II 67264 (c. 539/540) – partial tax receipt

Abraamios, son of Apollo

P. Cair. Masp. II 67262 (VI) – only the signature block remains P. Cair. Masp. III 67307 (524/539) – receipt for dues of money and in kind SB XXVI 16529 (526) – lease of land, payment in grain

P. Michael 43 (526)63 – lease of farm – no picture available P. Mich. XIII 670 (527) – loan of grain

P. Heid. V 351 (534/5) – rental of land Ibid., 247–271. Ibid., 29. 63 The editor to P. Mich. XIII 670 noted that the notary of that document was the same as the notary in P. Michael 43, P. Michael 45, P. Cair. Masp. III 67296, P. Cair. Masp. III 67307, P. Cair. Masp. 67308. 61 62

28

ELIZABETH BUCHANAN P. Cair. Masp. III 67296 (535) – surety

P. Cair. Masp. III 67335 (c. 537) – tax receipt

P. Cair. Masp. II 67252 (538) – surety P. Cair. Masp. III 67327 (539) – tax receipts P. Michael 45 (540) – sale of pasture

P. Cair. Masp. I 67112 (544/545) – rental of pasture SB XX 15202 (1st half VI) – advance sale of wine

P. Cair. Masp. III 67308 (1st half VI) – ack. of debt for grain – short form PSI VIII 934 (522–553) – land rental

PSI VIII 936 (1st half VI) – receipt for rent (Picture very poor) PSI VIII 937 (1s half VI) – receipt for rent

P. Bingen 130 (526–548) – division of property

P. Cair. Masp. II 67259 (1st half VI) – fragments of a contract P. Koeln. X 421 (524–545) – testament

P. Lond. V 1696 (1st half VI) – land lease P. Lond. V 1701 (510,525,540) – advance sale of wine

CONNECTING THE DOTS Jeremias, son of Victor (?)

29

P. Cair. Masp. II 67133 (542) – receipt for wheat

P. Cair. Masp. III 67283 (before 547) – report to Empress Theodora – Jeremias is referenced in this document Amais, son of P. Cair. Masp. II 67127 Abraamios (544) – acknowledgement of debt for 1/3 solidus provided for wool Pilatus, son of Apollo P. Cair. Masp. III 67283 (before 547) – report to Empress Theodora – Pilatus signed the document for himself and as an amanuensis P. Cair. Masp. II 67128 (547) – short form lease

P. Cair. Masp. I 67095 (548) – surety agreement

P. Cair. Masp. I 67116 (548) – rental of land for flax

P. Cair. Masp. IΙ 67129 (549) – short form lease P. Cair. Masp. II 67251 (549) – short form lease

P. Cair. Masp. I 67094 (552/3) – surety agreement P. Cair. Masp. III 67303 (553) – wagon rental

P. Lond. V 1661 (553) – contract for tax collection

P. Mich. XIII 668 (555/540) – rent receipt P. Flor. III 298 (557) – tax receipts

30

ELIZABETH BUCHANAN P. Cair. Masp. III 67325 R Fol. II (559/560) – tax accounts

P. Michael 4064 (559/545) – sale of land (photo not available)

P. Michael 42A (566) IOU and mortgage – signed by Pilates (photo not available) SB XX 15018 (VI) – receipt

Hermauos

P. Cair. Masp. II 67144 (VI) – partial account – Pilatus is acting as a receiver probably transmitting taxes P. Hamb. I 68 (564/5 or 549/50) – land rental P. Lond. V 1692a (555) – land lease P. Mich. XIII 672 (557?) – end of a contract with a mortgage

P. Michael 46 (559) – lease of share of farm Anouphis Kuros 1

P. Mich. XIII 671 (mid VI) – guarantee for repayment of deposit P. Cair. Masp. I 67121 (573) – divorce contract P. Cair. Masp. III 67300 (527) – land rental P. Ross. Georg. III 36 (537) – land lease P. Ross. Georg. III 37 (545) – debt for barley

Note to line 81: the di’emou signature is not the same handwriting as the person who wrote the text. He could have employed a clerk; see Bell’s note to P. Lond.1661, line 29. 64

CONNECTING THE DOTS

31

SB XIV 11855 (also P. Berl.Brash. 27)(547?) – land rental to Apollo, son of Dioscorus P. Ross. Geor. V 37 (midVI) – contract for sale or debt?

P. Flor. III 288 (VI) – surety agreement

P. Flor. III 286 (552) – land rental Kuros 2

P. Cair. Masp. II 67234 (VI) – fragment of a contract P. Cair. Masp. I 67109 (565) – rental of land of Apollo, son of Dioscorus’s heirs

P. Cair. Masp. I 67110 (565) – rental of a potter’s workshop P. Michael 48 (572)65 – lease of pasture Constantinus

David, son of Constantinus Mathaiu Theodosios 65

en 130.

P. Michael 59 (VI) – end of lease P. Hamb. IV 265 (late 500s) – rental P. Mich. XIII 66466 (584/5. 599/600) – sale of parts to house

P. Vat. Aphrod. 4 (later VI) – sale of house P. Cair. Masp. III 67325 (folio IV) (585) – land rental P. Vat. Aphrod. 5 (later VI) – house sale P. Mich. XIII 663 (VI) - sale of parts of house

Victor, son of Kollouthis, is a partner of Phoibammon in P. Bing-

This has many of the same witnesses as P. Mich. 667/669 and P. Lond V 1899. 66

32

Antinoopolis

ELIZABETH BUCHANAN Apollo, son of Pilatus No known documents; he is a witness with the title of public tabellion in P. Vat. Aphrod. 4 (4.frC.10.20) unknown P. Cair. Masp. I 67115 (527–553) – partial sale? of land Dios P. Mich. XIII 659 (527–547) settlement of litigation Kollouthos, son of SB V 8029 (537) - surety Phoibammon Elias 1 P. Cair. Masp. II 67154r (527–565) – divorce contract Elias 2 P. Koeln III 156 (582–602) – credit acknowledgement Victor (Photo not P. Koeln. XIV 589 (VI) – rental? available) Kosmos P. Straus. I 46 (566) – guarantee of meat delivery P. Straus. I 47 (566) – guarantees of meat

Pekusios

P. Straus. I 50 (566) – guarantees of meat P. Cair. Masp. II 67161 (566) – power of attorney

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE CULTURE: THE CIRCULATION OF TAQWĪM TEXTS IN THE LATE MAMLUK SULTANATE FIEN DE BLOCK This paper discusses the importance of peritexts in the understanding of the evolution of the science of the stars in the fifteenth century Mamluk Sultanate on the basis of the marginalia present in the collection of taqwīm-works from this period, in particular the taqwīm-works authored by the Islamic timekeeper Ibn al-Majdī1 that were written down in the late fifteenth century. Until now, these manuscripts have only been studied with regard to their primary texts, leaving aside the peritextual material. This unilateral focus was motivated by a preoccupation with the “content” of the text, which was considered clearly distinguishable from its material “context”. Especially for mathematical texts, the presumption that the content can be studied without taking into account the context is widespread, as mathematical problems are generally considered not to be historical artefacts or products of the historical context. This paper challenges this assumption. showing that taking into account the context via the marginalia of a text, provides us with a 1 There are several copies of these treatises compiled in the Ottoman Sultanate too, but I will not discuss these manuscripts within this paper, because I have focussed on the material circulation of these texts in the context of the Mamlūk Sultanate.

33

34

FIEN DE BLOCK much more rich and interesting understanding of texts and their shifting role in knowledge cultures.

In this paper I discuss the importance of peritexts in the understanding of a knowledge culture and its shifting disciplinary boundaries. I show how we can trace the evolution of the science of the stars in the fifteenth century Mamluk Sultanate on the basis of the marginalia present in the collection of taqwīm works from this period.2 More specifically, I will discuss the taqwīm works authored by the Islamic timekeeper Ibn al-Majdī3 that were written down in the late fifteenth century. While these manuscripts circulated widely in the later fifteenth century, they have up until now only been studied with regard to their primary texts, leaving aside the peritextual material. This unilateral focus was motivated by a preoccupation with the “content” of the text, which was considered clearly distinguishable from its material “context.” Especially for mathematical texts, the presumption that the content can be studied without taking into account the context is widespread, as mathematical problems are generally considered not to be historical artefacts or products of the historical context.4 Throughout this paper, I challenge this assumption, showing that taking the context into account via the marginalia of a text provides us with a much more rich and I use the term “science of the stars” because I consider it to be an instrumental term for investigating these practices. This term does not correspond to any contemporary disciplinary category, but it does correspond to the actor’s term ‘ilm al-nujūm, which was used in a rather general sense. I am borrowing this term from studies of Robert Westman, who encountered similar problems in his research concerning late medieval astronomy and astrology in Europe. Robert Westman, The Copernican Question (Berkeley: University of California Press, 2011). 3 There are several copies of these treatises compiled in the Ottoman Sultanate, too, but I will not discuss these manuscripts within this paper because I am focusing on the material circulation of these texts in the context of the Mamlūk Sultanate. 4 Karine Chemla, “On Mathematical Problems as Historically Determined Artifacts: Reflections Inspired by Sources from Ancient China,” Historia Mathematica, no. 36 (2009): 213–246. 2

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE

35

interesting understanding of texts and their shifting role in knowledge cultures.

IBN AL-MAJDĪ’S CORPUS OF TAQWĪM TEXTS AND ITS

Ibn al-Majdī was a late-fourteenth- and early-fifteenth-century scholar who specialised in the science of the stars (ʿilm al-nujūm) and who lived in Cairo, the capital of the Mamluk Sultanate. He worked at the al-Azhar mosque as a muwaqqit (Islamic timekeeper) and was head of the teachers at the Jānibakiyya madrasa.5 As a teacher-timekeeper, Ibn al-Majdī wrote several didactical treatises on timekeeping, of which a substantial part focused on the practice of “taqwīm.” Taqwīm is a term that was particularly used by Muslim astronomers for the practice of determining the real positions of the sun, the moon and the five planets6 in a specific year. In a zīj, or manual for timekeeping, the mean position of the planets was calculated; from this, the real position could be derived by means of an equation.7 This practice, as well as the tables resulting from it, were referred to as taqwīm. Apart from the didactical works on taqwīm in manuscripts TM 82, DM 405 and MM 44, a lot of actual taqwīm tables based on the methods discussed in these didactical works survive. The ones discussed here are manuscripts MM 25, MM 26 and MM 85, which are taqwīm tables for the sun and the moon.8 Previous studies of Ibn al-Majdī’s taqwīm texts share three big narratives about these sources. The first is the central role of APPROACH IN EARLIER RESEARCH

François Charette, “Ibn Al‐Majdī: Shihāb Al‐Dīn Abū Al‐ʿAbbās Aḥmad Ibn Rajab Ibn Ṭaybughā Al‐Majdī Al‐Shāfiʿī,” in The Biographical Encyclopedia of Astronomers (New York: Springer, 2007). 6 I.e. Saturn, Jupiter, Mars, Venus and Mercury. 7 D.M. Varisco, “Takwim,” in Encyclopaedia of Islam, 2nd ed., ed. P. Bearman et al. (Brill Online: Brill, 2012) http://referenceworks.brillonline.com/entries/encyclopaedia-of-islam2/takwim-COM_1158?s.num=0&s. f.s2_parent=s.f.book.encyclopaedia-ofislam-2&s.q=Takwim. 8 There are taqwīm tables for the other five planets, too, but the context of this paper does not allow us to discuss them here. 5

36

FIEN DE BLOCK

Ibn al-Majdī himself as the one and only author of these treatises. King and Kennedy ascribe all the sources discussed in this paper to the early-fifteenth-century author.9 A second narrative is the theoretical character of these texts. As the practice of compiling these tables required rather difficult mathematical calculations, Ibn al-Majdı̄ in his didactical works developed an alternative way to calculate the real position of the sun, the moon and the planets. That is, he developed auxiliary tables for compiling ephemerides (the Latin/English equivalent of taqwīm).10 These auxiliary tables consisted in two sections, as discussed by King and Kennedy: The first, the “extended” one (Ar. mabsuta), gives the number of days from the Hijra epoch to the initial day of each big period tabulated. Opposite each day entry are the corresponding noon positions of mean, anomaly, and center. The second, the “summed” one (majmuca) gives the number of days elapsed from the beginning of the big period to the first day of the successive anomalistic periods. The three entries opposite each day number show the changes in the mean, anomaly, and center during the particular number of anomalistic periods.11

On the basis of these tables, the real position of the planets could be easily calculated. As such, Ibn al-Majdī’s work provided an alternative method for the calculation of the true position of the seven planets. This method, however, never became more important or widespread than its Ptolemaic counterpart and was hence considered to have been a theoretical frivolity. The third narrative states that this was a method used for the same reasons as the Ptolemaic method. The mathematical problem, it is assumed, is not linked to the specific situation. It always stays the same and thus invariably gives rise to the same practices.12 David A. King and E.S. Kennedy, “Ibn Al‐Majdī’s Tables for Calculating Ephemerides,” Journal for the History of Arabic Science 4 (1980): 48–68. 10 Ibid. 11 Ibid., 53. 12 Ibid. 9

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE

37

These three narratives are related and all stem from the same view of scientific practices. In these studies, scientificmathematical practices are considered to be independent from the place and time in which they developed. Mathematical problems are presumed to be universal, immaterial and independent of their context; hence mathematical texts are considered to be merely particular, material instances of these universal ideas. This is reflected in the divide between the content and carrier of a text, where the content is universal and the material aspects are the particular, accidental local carriers of it. Although this approach has an intuitive appeal, this dichotomy brings up a lot of problems. These problems have led several voices in recent research to reject the dichotomy and to look at texts in a different, more material way. One of those voices is Karine Chemla’s. Chemla is a historian of science who states that the problems with a clear-cut distinction between the content and carrier of a text are manifold. First, a unique definition of the “content” of a text can never be given. Interpretation always depends on who is reading or writing the text and in which circumstances—information we get, among other sources, from the peritext. Furthermore, texts are always written in the process of carrying out intellectual activities. Chemla argues that texts are artifacts, elaborated in the course of the practice to which they belonged.13 Therefore, even for mathematical texts, peritextual data are not a mere appendix to the texts but an integral part of them. In their research on Ibn al-Majdī’s taqwīm treatises, David King and Edward Kennedy focus on his technique of calculating ephemerides/taqwīm through auxiliary tables. King and Kennedy compare this technique to the usual Ptolemaic way of calculating ephemerides by converting them both into present day mathematical notation. Concluding that both methods were more or less mathematically equivalent, they assume that their use was equivalent too.14 This mathematical notation is but one example of how this view Karine Chemla, ed., History of Science, History of Text, Boston Studies in the Philosophy of Science 238 (Dordrecht: Springer, 2004). 14 King and Kennedy, “Ibn Al‐Majdī’s Tables for Calculating Ephemerides,” 48. 13

38

FIEN DE BLOCK

on scientific practices is linked to the neglect of the peritextual features of the text. In this article, I discuss the collection of taqwīm treatises authored by Ibn al-Majdī and shed light on the other ways in which this view has influenced our understanding of the science of the stars in the late Mamluk Sultanate. More particulary, I will discuss a section of Ibn al-Majdī’s works, the section that is available at the national library of Egypt.15 I will limit myself to those manuscripts that were written in the fifteenth century itself and thus bear the peritextual marks of their circulation in that period.

ANALYSIS OF THE PERITEXTS OF THE TAQWĪM WORKS

In what follows, I will look at the peritexts found on Mss. Cairo (Dār al-Kutub) ṬM 82,1, DM 405, MM 44, 2, MM 85, 1, MM 25 and MM 26 and compare them to the three narratives mentioned above: the role of Ibn al-Majdī, the theoretical character of the works and the idea that a mathematical problem is invariably bound to the same practices. The role of Ibn al-Majdī If we take a look at the colophons of Ibn al-Majdī’s texts in light of his authorship, it is striking that most of the works are actually written down around or shortly after Ibn al-Majdī’s death in 1447/850. The dates found in the colophons and marginal notes of the texts are the following: Shelf mark TM 82 DM 405

MM 44, 2 (=fol 22r27v)

Title Ghunyat al-fahīm wa-l-ṭarīq ilā hall al-taqwīm Kitāb al-durr al-yatīm fī tashīl ṣanāʿāt al-taqwīm Kitāb al-durr al-yatīm

Date in colophon 870 AH 850 AH 850AH

The National Library of Egypt has the largest collection of astrological and astronomical manuscripts written in Arabic in the world. The library has more manuscripts ascribed to Ibn al-Majdī than any other institution. 15

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE MM 85, 1 (= fol.1r-2v) MM 25 MM 26

Solar tables from al-durr alyatīm Lunar Tables from al-durr alyatīm Lunar tables from al-durr alyatīm

39

850 AH 825 AH 850 AH

This means that only one of these texts was actually written during the life of Ibn al-Majdī himself. This is manuscript MM 25, which contains lunar tables based on the method discussed in the work “al-Durr al-Yatīm.” However, MM 25 contains a note which suggests that this manuscript, too, is not an autograph. It says: ‫هذه تعاديل القمر باصول �ن يو�� حَساب �ن ا���دي و�ي ��ط الع��مة عبد‬ .‫العز�� الوفاي تغمدهم ا�� ّٰ� ����تھ واسكنھم فسيخ جنته آم�ن‬

These are the equations for the moon based on the (parameters of) Ibn Yūnus for the calculation of Ibn al-Majdi. They were written down by ʿAbd al-Azīz al-Wafa’ī, may God protect them with His grace and accommodate them in his spacious paradise. Amen.16

This work appears to have been written down by al-Wafā’ī, a famous contemporary of Ibn al-Majdī who worked as a timekeeper and astronomer in the Mamluk Sultanate. Al-Wafā’ī died twenty years after Ibn al-Majdī.17 In the small colophon at the end of this text, we find a reference to the latter as al-muṣannaf, or “the compiler.” The colophon says: Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā Ibn al-Majdī, “Lunar Equation Tables Based on Al-Durr Al-Yatim” (AH 825), MM 25, Dar al-Kutub wa-al-watha’iq al-qawmiyya. 17Al-Wafā’ī was a fifteenth-century astronomer and a pupil of Ibn alMajdī. He worked in the Mu’ayyad mosque of Cairo as a muwaqqit, or timekeeper. Muḥammad ibn ʻAbd al-Raḥmān al-Sakhāwī, “539 - ‘Abd AlAzīz Bin Muḥammad Bin Muḥammad Al-Wafā’ī Al-‘Izz Abū Al-Faḍl Wa Abū Al-Fawā’Id,” in Al-Ḍaw’ Al-Lāmi‘ Li-Ahl Al-Qarn Al-Tāsi‘, vol. II (Beirut: Dar al-kutub al-’ilmiyya, 2004), 348. 16

40

FIEN DE BLOCK ‫��ت ا��رر �ي الشمس والقمر��ط ا��ستاذ الع��مة عبد العز�� الوفائى‬ ‫وهو ناقل من خط ا��صنف ����ه ا�� ّٰ� تعا�ى‬

‘al-Durar’ with regard to the sun and the moon ends here, from the writing of the compiler, in the hand of the very learned ‘Abd al-Aziz al-Wafa’ī – and this was transmitted from the manuscript of the muṣannaf, with the mercy of God, who is exalted.18

Just like this work, the peritexts of the other taqwīm tables do not suggest that they were copies from tables that were necessarily originally written by Ibn al-Majdī, but rather refer to him as the source of inspiration for the method used. MM 85, for example, only refers to al-Durr al-Yatīm indirectly with a similar title: “Kitāb al-durr al-nadhīm fī ḥall al-taqwīm.” Nevertheless, the work has been catalogued as a copy of ‘al-Durr al-Yatīm’ written by Ibn al-Majdī himself.19 However, this work seems to consist of a collection of tables compiled by another scholar who used Ibn al-Majdī’s technique for auxiliary tables. In addition to the reference to Ibn al-Majdī’s method, the title page of the text also refers to the zīj, or manual for timekeeping, that was written for Ulugh Beg and of which a recension was made for Cairo in this period by Ibn al-Majdī’s famous colleague Abd al-Fatḥ al-Ṣūfī. 20 No explicit reference to Ibn al-Majdī is made in the entire manuscript. The same is true for MM 26. This short manuscript does not have a title page and only contains a couple of tables, above which is noted that they are calculated on the basis of the method explained in al-Durr alYatīm. The didactical treatise Ghunyat al-fahīm (TM 82) does explicitly refer to Ibn al-Majdī as the author. As this was written down after his death, this must have been done by a copyist. However, the Ibn al-Majdī, MM25,fol 2V. David A. King, A Survey of the Scientific Manuscripts Available in the Egyptian National Library (Cairo: American Research Center in Egypt, 1986). 20 Ibn al-Majdī, MM25,fol 2V. 18 19

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE

41

name of this copyist is not mentioned. Whereas the name of a copyists is absent, this and the other didactical treatises contain a lot of possessor’s notes. In general, however, they concern scholars whom I could not find in any contemporary biographical dictionaries or chronicles. When we focus on the fifteenth-century notes in Ghunyat al-fahim, we find a waqf, or bequest note. This note is not dated, but the handwriting is the same as the handwriting of many of the notes in the manuscript in which the writer comments on calculations for the year 870 AH.21 ‫���د �ن ���د الصدق اللي�ى ع�� ا�� ّٰ� وقفا‬... �ّٰ �� ��‫من كتب العبد الفق‬ ‫�� ���يع ا��س���ن‬ ‫وو��ه ���ل ا���ن‬

From the books of ʿAbd al-Faqīr li-l-Lāh … Muḥammad bin Muḥammad al-Sadaq al-Laythī, may God forgive [him], as a waqf from him to all the Muslims and his son Kamāl al-Dīn.

In DM 405, one of the copies of the other didactical work Kitāb al-Durr al Yatīm, no explicit mention of Ibn al-Majdī is made; but the copyist, too, refers to the title of the work al-Durr alYatīm. The manuscript does not have a real title page or colophon, but it contains many notes around the tables.22 Here, too, an unknown possessor or transmitter of the manuscript is mentioned: .‫مل� هذا ال��اب الفق�� ا�ى ا�� ّٰ� تعا�ى ا��د العال ع�� عنه‬

This work was possession of al-faqīr ilā Allāh ta’ālā Aḥmad (al‘āl), may he be forgiven.23 Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā Ibn al-Majdī, “Ghunyat Al-Fahīm Wa-l-Ṭarīq Ilā Ḥall Al-Taqwīm” (AH 870), Dar al-Kutub wa-al-watha’iq al-qawmiyya. 22 Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā Ibn al-Majdī, ‘Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm’ (ca. AH 850), DM 405, Cairo Dar al-Kutub wa-al-watha’iq al-qawmiyya. 23 Ibid. 21

42

FIEN DE BLOCK

In MM 44, Ibn al-Majdī is mentioned by name, but not in the part of the manuscript which contains a fragment of al-Durr al Yatīm (ff. 22v-27v). The text is bound together with some tables for the planet Mars based on Ibn al-Majdī’s method. In this section, the name of Ibn al Majdī seems to have been added in a different hand later on.24 Does all this mean that these works are not Ibn al-Majdi’s? Not necessarily so. It does, however, indicate that this new method for the calculation of ephemerides was not only the theoretical idea of one astronomer but was embodied in a broader movement active from the later fifteenth century on. It is, of course, not the case that earlier researchers were not aware of this; but this was considered irrelevant to the mathematical content of the text. As King and Kennedy treated mathematical problems to be culturally invariant, the fact that many of these sources were copied or even created after the death of Ibn al-Majdī does not really make a difference if the mathematical content fits the theory. If however, like Karin Chemla, we consider these texts to be historical artifacts, the question arises as to whether the changing historical situation affected the copying of these texts in this period. In the next paragraph, I will take a look at the actual manuscripts in which these colophons were found.

THEORETICAL TREATISES? THE USE OF TAQWĪM TEXTS AS

These copies of didactical treatises and anonymously compiled tables referring to Ibn al-Majdī’s work all contain a lot of notes from users. These notes can be small indications that the calculations in the tables are correct or incorrect, saḥḥ or ghalaṭ, or references to corrections or ṣawāba. A few times, the reader mentions the opinion of his shaykh in the margins.25 At the end of manuscript TM 82, a very practical note is added, clarifying how you TOOLS

Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā Ibn al-Majdī, ‘Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm’ (AH 850), DM 44,2, Cairo Dar al-Kutub wa-al-watha’iq al-qawmiyya. 25 Ibid. 24

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE

43

should observe the astronomical phenomenon of a conjunction. These notes indicate that these texts were not merely calculations, but that the tables were actively used as tools.26 This is visible in a very direct way in MM 44, where fragments of al-Durr al-Yatīm are immediately followed by a paper version of a sine quadrant. The sine quadrant is an astronomical instrument with which one can measure angles and their sine and cosine.27 This paper instrument at the back of the manuscript shows how these texts were actively used and reused. Intended Use Considering both the absence of the names of the copyists and the possessors’ references to rather unknown scholars, these texts seem to have been used as utensils, rather than being mere mathematical curiosities. As taqwīm texts in the Mamluk Sultanate were often used together or were even part of a zīj, or manual for timekeeping, King and Kennedy seem to assume that this was also the case for Ibn al-Majdī’s taqwīm corpus.28 If we look at the notes in the manuscripts, however, as well as the contemporary texts with which these manuscripts are bound together, this does not appear to be the main tendency. For example, the users of TM 82 seemed to have been particularly interested in planetary conjunctions and aspects, which had no direct relation to timekeeping. In this same manuscript, the last page contains a page-long note on specifically this topic. This note might not be a mere note but, instead, a page of another treatise that was added to the manuscript. The handwriting does, however, match the handwriting of many of the notes in the previous pages of the manuscript.29 Ibn al-Majdī, “Ghunyat Al-Fahīm Wa-l-Ṭarīq Ilā Ḥall AlTaqwīm.” 27 Ibn al-Majdī, “Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt AlTaqwīm,” AH 850. 28 Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā Ibn al-Majdī, “al-Jāmiʿ al-Mufīd fī Bayān Uṣūl al-Taqwīm wa-l-Mawālīd” (Cairo, A.H 880), Academia 48, University Library Leiden, fol. 3a. 29 Ibn al-Majdī, “Ghunyat Al-Fahīm Wa-l-Ṭarīq Ilā Ḥall AlTaqwīm.” 26

44

FIEN DE BLOCK

Similar to the extra page added to TM 82, several of the other manuscripts are also bound together with other texts. The logic behind the compilation of majmūʿs in the Mamluk Sultanate has not been studied thoroughly, but we can at least be sure that texts that were bound together circulated together at a certain point. MM 85 is the most interesting source in this respect as it is bound together with a short medical-astrological treatise. Ibn al-Majdī’s tables are immediately followed by an astrological wheel in which every day of the lunar month is linked to specific parts of the body. The rest of the text is rather unclear, but it refers to the parameters of Ulugh Beg’s Sultani zīj, of which a recension was made for Cairo by Ibn al-Majdī’s colleague al-Sūfī. Bound together with a didactical work on the calculation of ephemerides, it is probable that the actual positions of the planets in the taqwīm tables were used in order to say something about the health of the people for which the wheel was used—for example, on the basis of their time of birth or death.

CONCLUSIONS

Whereas this collection of late-fifteenth-century taqwim texts has always been considered the material bearer of one scholar’s theoretical alternative, a first look at the peritext of this particular set of texts sheds light on its actual use. When we consider their peritexts, and thus the context of circulation of the manuscripts, these texts were far more than mathematical curiosities. If we return to the three narratives about Ibn al-Majdī’s taqwīm works in earlier research, we can state that a material approach makes us question the role of the author here. Ibn al-Majdī may have written important texts on the calculation of ephemerides, but the use and application of the texts goes beyond him and his scholarly network. Second, these texts were used as instruments: as astronomical instruments, but also as didactical instruments, which is demonstrated by the notes, and as actual astronomical paper instruments. Thirdly, these texts were likely not only used for timekeeping practices but also for astrological purposes. This, however, is a hypothesis for further research. Returning to Chemla’s view on the materiality of mathematical texts, we can state that a preliminary study of the peritexts in Ibn al-Majdī’s taqwīm corpus clearly shows how a mathematical problem like the calculation of the real positions of the sun, moon and seven

MARGINALIA AS TRACES OF CHANGING KNOWLEDGE

45

planets does not automatically result in the same practices. Whereas taqwīm was, in general, used for timekeeping, the alternative method for ephemerides developed by Ibn al-Majdī seems to have been picked up in the late fifteenth century by seemingly less well known scholars who engaged in the calculation of conjunctions and horoscopic information in the first place. This meaning and importance of Ibn al-Majdī’s texts and legacy can only be grasped through the study of peritextual material.

BIBLIOGRAPHY

Charette, François. “Ibn Al‐Majdī: Shihāb Al‐Dīn Abū Al‐ʿAbbās Aḥmad Ibn Rajab Ibn Ṭaybughā Al‐Majdī Al‐Shāfiʿī.” In The Biographical Encyclopedia of Astronomers, edited by Thomas Hockey, V. Trimble, Th.R. Williams, K. Bracher, R. Jarrell, J.D. Marche, F.J. Ragep. New York: Springer, 2007. Chemla, Karine, ed. History of Science, History of Text. Boston Studies in the Philosophy of Science 238. Dordrecht: Springer, 2004.

———. “On Mathematical Problems as Historically Determined Artifacts: Reflections Inspired by Sources from Ancient China.” Historia Mathematica, no. 36 (2009): 213–246. Ibn al-Majdī, Shihāb al‐Dīn Abū al‐ʿAbbās Aḥmad ibn Rajab ibn Ṭaybughā. “Ghunyat Al-Fahīm Wa-l-Ṭarīq Ilā Ḥall AlTaqwīm.” Cairo, AH 870. Dar al-Kutub wa-al-watha’iq alqawmiyya.

———. “Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm.” Cairo, ca. AH 850. MM 26,1. Cairo Dar al-Kutub wa-alwatha’iq al-qawmiyya.

———. “Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm.” Ca. AH 850. DM 405. Cairo Dar al-Kutub wa-al-watha’iq alqawmiyya.

———. “Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm.” Cairo, AH 850. DM 44,2. Cairo Dar al-Kutub wa-al-watha’iq al-qawmiyya.

46

FIEN DE BLOCK

———. “Kitāb Al-Durr Al-Yatīm Fī Tasḥīl Ṣanāʿāt Al-Taqwīm.” Cairo, ca. AH 850. MM 85,1. Cairo Dar al-Kutub wa-alwatha’iq al-qawmiyya.

———. “Lunar Equation Tables Based on Al-Durr Al-Yatim.” Cairo, AH 825. MM 25. Dar al-Kutub wa-al-watha’iq alqawmiyya.

King, David A. A Survey of the Scientific Manuscripts Available in the Egyptian National Library. Cairo: American Research Center in Egypt, 1986.

King, David A., and E.S. Kennedy. “Ibn Al‐Majdī’s Tables for Calculating Ephemerides.” Journal for the History of Arabic Science 4 (1980): 48–68.

Sakhāwī, Muḥammad ibn ʻAbd al-Raḥmān al-. “539 - ‘Abd Al-Azīz Bin Muḥammad Bin Muḥammad Al-Wafā’ī Al-‘Izz Abū AlFaḍl Wa Abū Al-Fawā’Id.” In Al-Ḍaw’ Al-Lāmi‘ Li-Ahl AlQarn Al-Tāsi‘, II:348. Beirut: Dar al-kutub al-’ilmiyya, 2004.

Varisco, D.M. “Takwim.” In Encyclopaedia of Islam, second edition, edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, and W.P. Heinrichs. Brill Online: Brill, 2012. http://referenceworks.brillonline.com/entries/encyclopaedia-ofislam2/takwimCOM_1158?s.num=0&s.f.s2_parent=s.f.book. encyclopaedia-of-islam-2&s.q=Takwim.

THE MANUSCRIPTS OF ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN1 ZUZANA GAŽÁKOVÀ Manuscripts are the most authentic sources of Arabic popular epics (sīra šaʿbīya, pl. siyar) that are accessible to us nowadays since the number of audio recordings is low, which bears witness to the diminished status of the oral narrative tradition. This contribution focuses on specific features of these voluminous handwritten documents, such as their general appearance, their title pages, their script, the opening of the sīra, the occasional vocalization, the subdivision of the text, the colophons, and the layout. In addition, other details behind the production of these texts – such as the considerable discrepancies in the sīra’s plot and the scribe’s occasional negligence – will be commented upon. It is also possible to distinguish between manuscripts which were used most probably as performance prompts (or as a certain aide-mémoire) and those probably compiled for private use. Here, the social context of this type of literature will be taken into account. The presented findings primarily stem from the scrutiny of manuscripts of Sīrat Sayf ibn Dhī Yazan. The manuscripts of other siyar (e.g., Sīrat Dhāt al-Himma, Sīrat al-Iskandar Dhū ʾl-Qarnayn, etc.) will also be discussed to demonstrate certain features.

Arabic popular epics (sīra šaʿbīya) can be regarded as a genre of Arabic literature which originated and developed within a flourishing 1

This article was written as a part of the VEGA 1/0541/16 grant project.

47

48

ZUZANA GAŽÁKOVÀ

tradition of oral storytelling. The most authentic sources of these heroic narratives are manuscripts that, today, represent the only “hard evidence” of this centuries-long tradition; the form of oralgenerated composition diminished significantly in the twentieth century, and the number of recordings is rather low. Nevertheless, hundreds of manuscript volumes of these pseudohistorical works of battle and romance have been preserved. They remain scattered over various collections, housed in public and university libraries, and are primarily located in Europe; the most notable of these are the manuscript collections of Staatsbibliothek zu Berlin (Preuβischer Kulturbesitz), which were described by W. Ahlwardt in his enormous catalog. Ahlwardt cataloged the Arabic manuscripts that existed at the time in the “Königliche Bibliothek zu Berlin,” and the list of popular siyar and qiṣaṣ fills almost the entire eighth volume.2 In addition, Berlin’s Staatsbibliothek contains other valuable manuscripts of popular stories. Noteworthy manuscripts of Arabic popular epics can also be found in the Universitäts- und Forschungsbibliothek in Gotha, the Universitätsbibiothek in Leipzig, the Bibliothèque Nazionale de France in Paris, the Egyptian National Library (Dār al-kutub) in Cairo and elsewhere around the world. It must be emphasized that this truly impressive amount of literary material has not yet been sufficiently covered in books on Arab literary history.3 This contribution is supposed to be a more corpora-based expansion of a previous article,4 and it will focus Wilhelm Ahlwardt, Verzeichniss der arabischen Handschriften der königlichen Bibliothek zu Berlin, [= Die Handschriften-Verzeichnisse der königlichen Bibliothek zu Berlin], Vol. 8 (Berlin: Asher, 1896), 69–461. 3 Cf. Jan Just Witkam, “Johann Gottfried Wetzstein’s Manuscripts Containing Arab Popular Stories,” in Manuscripts, Politics and Oriental Studies. Life and Collections of Johann Gottfried Wetzstein (1815–1905) in Context, ed. Boris Liebrenz and Christoph Rauch (Leiden, Boston: Brill, 2019), 121. 4 Zuzana Gažáková, “Arabic Epics.” In Comparative Oriental Manuscript Studies: An Introduction, ed. Alessandro Bausi (Hamburg: Tredition, 2015) 397–402. Highly inspiring research on the manuscripts of popular siyar can be seen in the work of Claudia Ott, Metamorphosen des Epos: Sīrat Muğāhidīn (Sīrat al-Amīra Dāt al-Himma) zwischen Mündlichkeit und Schriftlichkeit (Leiden: CNWS Publications, 2003); Thomas Herzog, Geschichte und 2

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 49 (albeit not comprehensively) on the scribal practices and processes of the manuscript copying of Arabic popular siyar. Owing to the fact that the present author has been studying various aspects of Sīrat Sayf ibn Dhī Yazan for more than ten years, this contribution will mainly focus on the manuscripts of this popular epic.

GENERAL CHARACTERISTICS

The written versions of popular siyar are extraordinary long. Some copies of Sīrat Dhāt al-Himma have almost 12,000 folios (Ms. Berlin 9149 We. 434–4835), although there are also medium-length and short versions as well as considerably abridged “concise versions” (e.g. Ms. Leipzig Vollers 630, dated to 1784, which covers the whole Sīrat Sayf ibn Dhī Yazan in just 193 folios). The differences and relationships between the very long and medium-length manuscripts are not clear, yet it seems obvious that the longest versions should be the most elaborated and contain various additional information. This is, however, not the rule; texts of medium length often offer interesting details which do not appear in the longer texts.6 The shortened, concise versions contain brief, and sometimes even extremely schematic, summaries of the content, which are only comprehensible (and even then with difficulty) to people familiar with the sīra in detail; these served as a sort of narrator’s prompt or aide-mémoire. Sometimes a scribe left some brief information about the character of the composition in the title or in the introduction to the narration, as is the case with the previously

Imaginaire: Entstehung, Űberlieferung und Bedeutung der Sirat Baibars in ihrem sozio-politischen Kontex. (Wiesbaden: Harrassovitz, 2007); and Jan Just Witkam “Johann Gottfried Wetzstein’s Manuscripts Containing Arab Popular Stories.” In Manuscripts, Politics and Oriental Studies: Life and Collections of Johann Gottfried Wetzstein (1815–1905) in Context, ed. Boris Liebrenz and Christoph Rauch (Leiden, Boston: Brill, 2019), 116–139. 5 Cf. Claudia Ott, Metamorphosen des Epos: Sīrat al-Muğāhidīn, 51. 6 Jérôme Lentin, “Variétés d´arabe dans des manuscrits syriens du Roman de Baybars et histoire du texte,” in Lectures du Roman de Baybars, ed. Jean-Claude Garcin (Marseille: Editions Parenthéses, 2003), 94.

50

ZUZANA GAŽÁKOVÀ

mentioned Ms. Sīrat Sayf Leipzig Vollers 630, 193ff.7 Here the information about the concise character of the composition appears right after the transitional formula al-baʿdīya (wa baʿd), fol. 2a: ���‫��� ا�� ّٰ� ا����ن ا��ح�� | وبه ��تع�ن وص�� ا�� ّٰ� ��� سيدنا ���د | و‬

‫ا�� وا��ابه ا��ع�ن وبعد فهذه | س��ت ا��ل� سيف ذال��ل م�ن وا�ح‬ ‫من ���| ��ح زا�� اتفق انه ك�ن �ي قد�� ا��مان‬

In the name of God, the most Gracious the most Merciful, whom we ask for help, blessings of God be upon our Master Muḥammad and upon his family and all his companions. Having said so, this is Sīrat al-Malik Sayf Dhu ’l-Yazal8 of clear text, without excessive explanation. It happened that there was in the distant past…

APPEARANCE

The general impression is that the majority of these texts are rather modest in appearance and have apparent traces of manipulation. They often demonstrate that the process of copying was imperfect and that they are less elaborate than other texts from the established literary and scholarly canon. In light of this situation, it seems that their creators enjoyed more freedom. However, despite their variability, they did share a large number of common features or processes. Boris Liebrenz has characterized these texts as being of simple (if not to say cheap) appearance; they were tattered and worn from constant reading. They usually appear in the form of booklets of a small size—fifteen to twenty centimeters long and ten

The manuscript is digitized and accessible online via: https://www.refaiya.uni-leipzig.de/receive/RefaiyaBook_islamhs_00002082. 8 There is a distortion of the hero’s name (Yazan>Yazal). The same phenomenon can be observed in other manuscripts of this sīra. In some cases, the name varies between al-Yazan and al-Yazal within one folio, e.g., Ms. Or. 4643 London, the British Library, Oriental and India Office Collections, f. 3a. 7

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 51 centimeters wide—that are simply bound.9 There is, for instance, a sixty-volume set of Sīrat ʿAntar (Ms. Berlin 9123 We. 901–960) and a twenty–six volume set of Sīrat Baybars (Ms. Berlin 9155 We. 561–58610). The present author has also come across an eighteenvolume set of Sīrat Sayf ibn Dhī Yazan (Ms. Cairo 13524, 89ff., 69ff., 78ff., 75ff., 72ff., 77ff., 86ff., 79ff., 88ff., 78.ff., 90ff., 98ff., 92ff., 80ff., 71ff., 80ff., 85ff., and 75ff. dated in 1187). Each volume of these multi-volume sets contains between forty and one hundred folios. The reasonable explanations behind this rather inflated number of volumes for a single text seem to be practical: a slim volume allowed for easier manipulation. It was also technically easier to produce a small volume than a larger one because many texts came about as a result of dictation. Economic reasons, however, should also not be underestimated. It was possible for a scribe to price each volume individually, and a book lender could lend out many parts of a text at the same time.11 The producer of these booklets followed a strategy by which the final portion of each manuscript would break off the story at a particularly thrilling moment so that the reader would be tempted to borrow the following volume. This can be seen in Ms. 3883 of Sīrat al-Muğāhidīn Paris, BnF. In the last folio of this volume (f. 79a), there is a long remark written by an annoyed reader dated to Šaʿbān, 4th (?), 1076 (approx. 1666). The reader comments on the division of the text into volumes, and his commentary is accompanied by many curses. It is evident that he was thoroughly bothered by the sudden breaking off of the story. He complains that the text was split up for the sake of money, Boris Liebrenz, “The Library of Aḥmad al-Rabbāṭ. Books and Their Audience in 12th to 13th/18th to 19th Century Syria,” in Marginal Perspectives on Early Modern Ottoman Culture: Missionaries, Travellers, Booksellers, ed. Ralf Elger and Ute Petruschka (Halle: Zentrum für Interdisziplinäre Regionalstudien, 2013) 17–59, 22. 10 For the physical description of these texts, cf.: Wilhelm Ahlwardt, Verzeichniss der arabischen Handschriften, Vol. 8, Grosse Romane, 80–84; 114–120. 11 Boris Liebrenz, “The Library of Aḥmad al-Rabbāṭ,” 26. 9

52

ZUZANA GAŽÁKOVÀ

and he exclusively blames the owner of the manuscript for this situation.12 Oh, possessor of this book, may God not have mercy upon you! What shame you are among all storytellers! You are more shameful than even ʿUqba and Šūmadris, because you end your book with the most unlikely and meanest passage. And this is only to trick your customers so that you can take as much (money) from them as you would like! Otherwise, you would not finish at this place.13 Didn’t you hear the saying of the Prophet, peace be upon him: He who cheats us is not from us? And this is the work of a cheater, and the reader of this book needs to take what belongs to him as all the people cannot read the story from the beginning till the end. Jesus Christ is angry with you alongside aš-Šaykh an-Nikkīḥ14 and the apostles and John the Baptist.

This was written by a poor and miserable sinner and duffer in front of the knowing and all-powerful God……. Ibn ʿAbdallāh on a Friday night, the 4th of the month Šaʿbān 1076.

Such a multi-volume set is not usually the product of a single process of copying; it is pieced together from fragments of varying size, age, and style by the owners of the manuscripts, who sometimes took great pains to collect the fragments, put them in the right order, and fill the gaps between them, sometimes even in their own handwriting.15 The period of constructing such manuscripts could cover several centuries (e.g., Ms. Sīrat al-Muğāhidīn, Paris, BnF, Arabe 3859–3892 contains a fragment from 1430, while its latest contribution is from 1808). Even individual volumes were often recreated or reconstructed from tattered fragments of older ones in order to create a new volume that could be put into circulaCf. Claudia Ott, “From the Coffeehouse into the Manuscript.” In Oriente Moderno 83, no.2 (2003), 443–451, 450–451. 13 Ibid., 451. 14 The epithet of ʿUqba, the hero of this sīra. Cf. Claudia Ott, Metamorphosen des Epos, 92. 15 Boris Liebrenz, “The Library of Aḥmad al-Rabbāṭ,” 27. 12

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 53 tion and thus make a profit (e.g., Ms. Sīrat Sayf Gotha orient. A 2411). The layout and handwriting of this volume change substantially several times (cf. f. 6a, f. 11a, f. 23a), and from f. 84b the main text is enclosed in a red rule-border. Even catchwords at the bottom of verso pages have often been adjusted to fit into the “new” structure, a new catchword being written next to the original (e.g., ff. 10b–11a) while the old one is simply crossed out. The missing text between such fragments is usually furnished in the margins (e.g., Ms. Sīrat Sayf Gotha orient. A 2410, f. 8b). In contrast to the manuscripts discussed so far, there are some examples of complete copies.16 These have a consistent and homogenous text that covers the hero’s life from birth to death in several written volumes, presumably, but not necessarily, written by a single hand, e.g., Ms. Sīrat Sayf London, the British Library, Or. 4643 (dated to 1736; complete text, 271ff.). The title of the manuscript even suggests the character of the text: Hādhā ’l-kitāb sīrat al-malik Sayf Dhū ’l-Yazan ʿalā ’t-tamām wa ’l-kamāl (This is a complete book of the sīra about King Sayf Dhū ’l-Yazan). Sometimes the text that is declared to be complete abruptly and for unknown reasons ends shortly before the conclusion with closing formulas and invocations to God that provide information about the very end of the narration: e.g., Ms. Sīrat Sayf, the National Egyptian Library, 4592, two volumes (315ff. and 312 ff.). The last folios of the text are extremely abbreviated and condensed, as if the scribe needed hastily to come to the end (2nd volume, ff. 311b –312a).

THE DATE OF ORIGIN

Unfortunately, most of the available manuscripts of popular siyar do not contain an explicit mention of when they were copied.

The relativity of the term “complete” in relation to Arabic popular epics can be demonstrated in the experience of Professor Stephan Procházka from the Oriental Institute in Vienna: when a popular storyteller in Urfa was recently asked for a performance of the complete Sīrat az-Zīr Sālim, he answered with a question about which “complete” version was desired—the two-hour one, the evening-long one, or the twoday one. 16

54

ZUZANA GAŽÁKOVÀ

Based on an evaluation of the content, paper, binding, and writing, their dating is a very difficult and sensitive affair. In general, they represent relatively late versions from approximately the seventeenth to eighteenth centuries, although their content is obviously older. References to the written works appear as early as in the twelfth century—for instance, in the polemic work Ifḥām al-Yahūd of the mathematician and physician as-Samawʾal al-Maghribī (1130–1180);17 meanwhile, the oldest dated manuscripts come from the fifteenth century.18 There is, however, a certain tendency to regard the manuscripts of Arabic popular literature as being older than they actually are: Ms. Tübingen MA VI 32 of Ḥikāyat ʿUmar an-Nuʿmān had been dated to the fifteenth century, yet the latest research has shown that it was composed later, probably around 1640. This might also be the case for the oldest fragment of Sīrat Sayf ibn Dhī Yazan, which is believed to be represented by the first volume of Ms. 148 (CXLVII, 79ff.) from the Ambrosiana Library in Milano, tentatively dated, according to O. Löfgren’s and R. Traini’s catalog, to the sixteenth century.19

LANGUAGE20

Manuscripts of popular siyar are categorized as Middle Arabic texts.21 Regarding the linguistic structure of these documents, a range of literary material was written in various Middle Arabic subvarieties (more or less classicizing or colloquializing language). This 17 As-Samawʾal al-Maghribī, Ifḥām al-Yahūd (Cairo: Dār al-hidāya, 1986), 51–52. 18 Peter Heath, “Sīra šaʿbīya,” in Encyclopaedia of Islam, CD-Rom edition (Leiden: Brill, 1999). 19 For a further physical description of the manuscript, see Oscar Löfgren and Renato Traini, Catalogue of the Arabic Manuscripts in the Bibliotheca Ambrosiana, vol. 1 (Vicenza: Neri Pozza, 1975), 83. 20 Throughout the paper, all manuscripts are followed as closely as possible in order to faithfully note all linguistic features. 21 It is not necessary to resume the discussion about what is meant by “Middle Arabic.” In recent years, there has been remarkable progress in this field, including the creation of the International Association for the Study of Middle and Mixed Arabic (AIMA) (2004).

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 55 material includes both passages written in formulaic rhymed prose (sağʿ) and dialogues, often recorded spontaneously and freely due to the fact that it was difficult to orthographically represent the vivid communication between the characters according to classical norms. In this respect, the manuscripts of popular siyar contain the features of Arabic dialects, forms that correspond to Classical Arabic, and “pseudo-corrections;” and, as such, they are valuable for determining when the influence of Middle Arabic texts became so strong that authors attempted to imitate them rather than use Classical Arabic. It goes without saying that the orthography of such texts is often open to different interpretations. One can, however, read a text written in a dialect correctly only if one already knows how that particular dialect is pronounced; this is a rarely acknowledged but serious problem in the case of older dialects.

SCRIPT

The overwhelming majority of the sīra’s texts are written in rather scrawled than carefully written naskh, in some cases with apparent traces of ruqʿa. This style of writing often demonstrates that the process of copying was imperfect and unprofessional. As a result, some texts are only legible with difficulty and with prior knowledge of the content. Only a few manuscripts of popular siyar written in maghribī have been examined by the author of this article. These include a 287ff. fragment entitled Qiṣṣat Dhū ’l-Qarnayn, dated to 1744, from the Mamma Haidara Commemorative Library of Timbuktu.22 The other two examples are the manuscripts of Sīrat Sayf ibn Dhī Yazan: Ms. 91 from the National Egyptian Library and Ms. Or. 1680a and 1680b from Leiden University Library.23 They both bear traces of vocalization, which is better executed in the latter case. A closer look at the The manuscript is digitized and accessible online: https://www.wdl.org/en/item/9674/view/1/1/ [10.9.2020]. 23 Both manuscripts (Ms. 91 from the National Egyptian Library and Ms. Leiden Or. 1680a) are briefly described in the present author’s forthcoming article: “The Episode of King Dhū Yazan and the Kaʿba According to Several Manuscripts and Early Arabic Books on History.” 22

56

ZUZANA GAŽÁKOVÀ

vowel system of this text reveals numerous anomalies (e.g., an omitted inflection of nouns by the attachment of a sukūn and case endings that are not the ones required by classical norms), and often the vocalizaton is redundant (e.g., َ‫) ﻛَﺎن‬. As Faustina Doufikar Aerts acknowledges in her study dedicated to the vocalized texts of Sīrat alIskandar, it is possible that the scribe simply intended to “vocambellish” the text.24 In addition, thanks to the almost full correspondance between the texts, a larger omission (saqṭa) was discovered in MS. Leiden Or. 1680a (fol.2b), which is a quite common feature in these voluminous materials. In Ms. Cairo 92, this concerns the text from fol. 2a/lines 15–28 to fol. 2b/line 14. This might prompt the speculation that the omission happened by accident during the process of hasty mechanical copying, when a scribe may have simply overlooked a folio without even noticing it. Despite the fact that it is clear that the two manuscripts are related to each other, it is very difficult to assess the relationship between these copies. It is nevertheless apparent that both originated from the same branch of the manuscript tradition of Sīrat Sayf ibn Dhī Yazan, which can be identified as North African due to the maghribī script.

THE TITLE PAGE

Traditionally, the title page of Arabic manuscripts contains the title of the work and possibly other information pertaining to its authorship and transmission. The majority of sīra title pages are rather simple, often written in negligent style and containing only the name of the narration in a somewhat concise form and the proper number of the volume. Graphically, this information usually appears written in black ink, sometimes combined with red in the form of an upward triangle; the script of the title does not usually differ from the script used for the main text. For more about the vocalized script in Arab popular epics, cf. Faustina Doufikar-Aerts, “Ġarāʾib or ʿAğāʾib, that’s the Question: Vocalised Script in Two Arabic Romances of Alexander.” In Moyen arabe et variétés mixtes de l’arabe à travers l’historie. Actes du Premier Colloque International (Louvainla-Neuve, 10–14 mai 2004), ed. Jérôme Lentin and Jacques Grand’Henry (Louvain: Université Catholique de Louvain, 2008), 165–179. 24

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 57 The story is considered to be a tale (qiṣṣa), biography (sīra), saga (malḥama), or even a tree or possibly a genealogy (šağara): e.g., Ms. Berlin 9163 We. 618 entitled ‫“ اﻷول ﻣﻦ اﻟﺸﺠﺮه اﻟﻈﺎھﺮﯾﮫ‬The first [volume] from aẓ-Ẓāhirīya tree,”25 followed logically by Ms. 9163 Berlin We. 619 entitled: ‫“ اﻟﺜﺎﻧﻲ ﻣﻦ اﻟﺸﺠﺮه اﻟﻈﺎھﺮﯾﮫ‬The second [volume] from aẓ-Ẓāhirīya tree.”26 Aẓ-Ẓāhirīya was a common designation for Sīrat aẓ-Ẓāhir Baybars. Sometimes a title-folio contains brief information about the possessor of the document, ex libris (min kutub), e.g., Ms. Berlin We. 643: ‫ا��ول من مل� سيف من ابتداء نوح �ليه الس��م من كتب شيخ‬ ‫عبد ا�� ّٰ� البغدادي‬

The first [volume] about King Sayf from the beginning of Nūḥ, peace be upon him, from the books of šaykh ʿAbdallāh al-Baghdādī

ʿAbdallāh al-Baghdādī was apparently an important figure in the market of popular epics in Damascus. Another example of this kind can be seen in Ms. Leipzig Vollers 630: -

‫هذه س��ة مل� سيف ا�ن ذل��ل من او��ا ا�ى ا��ها ابتدا وا���ا‬

‫و�ي من كتب ا��اج أ��د ا��باط ا��ل�ي‬

This is Sīrat malik Sayf ibn Dhu’l-Yazal from the beginning till the end, and it is from the books of al-Hāğğ Aḥmad ar-Rabāṭ al-Ḥalabī.

The manuscript is digitized and accessible online: https://digital.staatsbibliothek-berlin.de/werkansicht/?PPN=PPN663633 508&PHYSID=PHYS_0003 [10.9.2020]. 26 The manuscript is digitized and accessible online: https://digital.staatsbibliothekberlin.de/werkansicht?PPN=PPN672 771845&PHYSID=PHYS_0001&DMDID= [10.9.2020]. 25

58

ZUZANA GAŽÁKOVÀ

Here the title contains a reassurance that it is a complete volume “from the beginning till the end” (min awwalihā ilā ākhirihā, ibtidā wa ʾntihā). Moreover, the owner can be identified as a certain Aḥmad ar-Rabāṭ al-Ḥalabī ’š-Šaqīfātī ’n-Naqšbandī ’š-Šāfiʿī, known simply as ar-Rabāṭ, who was a well-known owner of a private library in Damascus in the eighteenth and nineteenth centuries.27 In the sphere of popular literature, Jan Just Witkam has rightfully pointed out, the owners or collectors of the manuscripts tended to acquire the position of authority in the transmission of the stories even more than the copyists.28 Another typical and frequent phrase that usually appears in the sīra’s titles is ʿalā ’t-tamām wa ’l-kamāl (“completely and entirely”) as the most valuable texts were naturally complete ones, e.g. Ms. Sīrat Sayf Gotha orient A 2401:29 ‫هذا هو ا��زء ا��ول من س��ة أبو ا��مصار وسائق النيل من أرض‬

‫ا��بشة إ�ى هذه ا��يار ��� ا���ام وال����ل‬

This is the first complete and entire part of the sīra of Abū ’lAmṣār and about the one who diverted the River Nile from Ethiopia into these regions.

However, some initial folios of popular siyar offer much more. Lengthy titles can often include a short summary of the text’s plot, accompanied by various typical formulas (ḥamdala, ṣalwala) as well as by the names of the ascribed author and the scribe, e.g. Ms. Sīrat al-malik at-Tubbaʿ, Paris BnF, Ar. No. 3821, f.1a. The initial as well as the final folios of manuscripts of popular epics frequently contain notes of readership and various remarks from readers. They are often written in the least legible script, and they usually have a formulaic character. The shorter and less formal Cf. Boris Liebrenz, “The Library of Aḥmad al-Rabbāṭ.” Jan Just Witkam, “Johann Gottfried Wetzstein’s Manuscripts Containing Arab Popular Stories,” 128. 29 The manuscript is digitized and accessible online: https://archive.thulb.uni-jena.de/ufb/receive/ufb_cbu_00006113 [10.9.2020]. 27 28

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 59 formula, ‫“ ﻧﻈﺮ ﻓﯿﮫ وﺗﺄﻣﻞ ﻣﻌﺎﻧﯿﮫ‬has looked in it (the particular volume) and considered its content,” is usually followed by a name. The socalled muṭālaʿa seems a bit more formal and longer, probably because ṭālaʿa denoted a single silent reading of the text. It may contain phrases in which God is asked for forgiveness, as well as the author’s name and the date, e.g. Ms. Sīrat al-malik at-Tubbaʿ Paris, BnF Ar. No. 3821. Here muṭālaʿa appears on the margins of the colophon and reads as follows: ��‫طالع هذه الس��ة من ابتدا إ�ى آ��ها را�ي عفو ربه ا��نان الفق‬

‫�ليل جور ��ي تفتخنان و�� ا��رحوم �ليل ع�بان غفر ا�� ّٰ� ذنو��ما‬

١١٦١ ‫ �ي ��ر ��اد آ�� سنة‬،‫ آم�ن‬،‫و��يع ا��س���ن‬

This sīra was read from the beginning till the end by the one who hopes for forgiveness from his generous Lord, poor Khalīl Ğūrbağī Taftakhnān, the son of the deceased Khalīl ʿAzbān. May God forgive their sins and the sins of all Muslims, amen, in the month of Ğumād ākhir in the year 1161.

In addition to formulaic readers’ notes, there are often various opinion from readers,30 e.g. Ms. Leipzig Vollers 627 of Sīrat alMalik al-Badrnār ibn an-Nahrawān, f.1a31: ‫هذه الس��ة ب�� فا��ة ك���ا بعض ليا�ي الشته طو ي�� سقعة بارده‬

This sīra is useless. It is like some winter nights—long, chilly, and cold.

THE PRESENTATION OF THE SĪRA

The overwhelming majority of Arabic manuscripts begin with the invocation of God, the formula known as the basmala. Accordingly, it is also at the beginning of all sīra manuscripts. It is usually folCf. the above quoted reader’s remark from Ms. 3883 of Sīrat alMuğāhidīn, f. 79a. 31 The manuscript is digitized and accessible online: https:// www.refaiya.uni-leipzig.de/receive/RefaiyaBook_islamhs_00002804 [10.9.2020]. 30

60

ZUZANA GAŽÁKOVÀ

lowed by a religious dictum that contains other typical formulas (ḥamdala, ḥawqala, taṣliya, etc.) as well as direct appeals to the audience, convincing them of the attractiveness and credibility of the depicted events. It was up to the scribe or the narrator to decide its length. In some cases, it is very brief, e.g. Ms. Berlin 9163 We. 618. Here the text of Sīrat Baybars starts in media res right after the basmala, which is followed by a short pious invocation (fol. 1b): ‫��� ا�� ّٰ� ر��ن رح�� به ��تع�ن | قال ا��اوي أن ا��ل� الصا�� أيوب‬ ...‫نام �ى بعض اليا�ي‬

In the name of God, the most Gracious the most Merciful, whom we ask for help, the narrator said that one night King aṣ-Ṣāliḥ Ayyūb slept...

In other siyar, it could take a considerably longer time till the narration actually started. The narrative may be preceded by the more elaborated religious dictum, which is then followed by the presentation of the work and its main characters. Sometimes there is a sort of a short introduction containing legends about ancestors, which may go back even to the creation of the Earth, e.g. Ms. Berlin 9121 We. 643 of Sīrat Sayf. Here, after the basmala, there is a well-developed and thematically ordered episode which contains the basic guidelines and fundamental ideas of Islamic theology. It starts with the creation of the Earth and the early history of mankind. This part apparently comes from another genre of Arabic popular literature: qiṣaṣ al-anbiyāʾ (“tales of the prophets”). On the basis of Ms. Berlin 9121 We. 643, it is possible to demonstrate how the Arabic popular sīra overlaps with popular religious narratives. The process is quite simple: a part of qiṣaṣ al-anbiyāʾ (cca. fol. 2a –7a) was inserted into the sīra, where it expressed its own thematical agenda. It was quite common for popular narrators to imitate in this way the model of learned historiography written in Classical Arabic. In addition, these initial stories slowly proceeded toward the description of the origins of the Dhū Yazan dynasty, depicting various tribal genealogies and the king himself as a direct descendant of the Prophet Nūḥ. The fact that the initial parts of the sīra come from qiṣaṣ al-anbiyāʾ is emphasized by the fact that the proficient transmitter Ibn ʿAbbās is indicated as the authority (qāla Ibn ʿAbbās, f. 2b, line 3) instead of Abū ’l-Maʿālī, to whom Sīrat Sayf

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 61 ibn Dhī Yazan is usually ascribed. The striking resemblance between these texts is evident when comparing the opening of Ms. Berlin 9121 We. 643 with the initial parts of Isaac Eisenberg’s edition of Kisāʾī’s Stories of the Prophets.32 The resemblance is so great that otherwise incomprehensible phrases in the manuscript can be understood with the help of Eisenberg’s edition.

PAGE LAYOUT AND CONTENT ARRANGEMENT

The body of the sīra’s text within a small volume (or a ğuzʾ) is not usually further divided into segments. The text is, however, visually structured with the use of emphasized words (formulas) written in larger letters, usually in red or bold black so that they immediately catch the readers’ attention. These were inserted into the text whenever suitable, such as when it was necessary to switch between the different narrative levels or to introduce a new storyline or a piece of poetry. They represent remnants of real performances and invite the reader to remember that the narrator is always present. The most frequent formulas include qāla ’r-rāwī (“the narrator said”), wa yarğiʿ al-faṣl ilā (“and the chapter returns to” with the following specification of the particular moment in the narration), and wa kāna ’s-sabab fī dhālika (“and the reason for that was”). These formulas were regularly inserted into the text after the execution of the whole document (especially in cases when a different color was used for formulas), and the scribe left blank spaces in the text for their sake. It is possible to find unfinished manuscripts with such blank spaces, e.g. Ms. Sīrat Sayf, Gotha orient A 2412.33 Here it is probably the formula qāla ’r-rāwī which is missing, and all blank spaces in the text are followed by another formula yā sādāt (“Oh, you gentlemen”).

32

Vita prophetarum, Muḥammed ben ʿAbdallāh al-Kisaʾi, ex codicibus, qui in Monaco Bonna, Ludg. Batav. Lipsia et Gothana asservantur, ed. Isaac Eisenberg, (Leiden: Brill, 1922), 5–6. 33 The manuscript is digitized and accessible online: https://archive.thulb.uni-jena.de/ufb/receive/ufb_cbu_00007494 [10.9.2020].

62

ZUZANA GAŽÁKOVÀ

The features of typical scriptural transmission and copying procedures, such as various scribal errors and corrections, occur frequently. Most corrections and emendations are done very simply, by crossing out the misplaced words and writing the correct phrases beside the main body of the text. Owing to their enormous length, these texts contain many contradictions and discrepancies in the details of the story and the arrangement of events. Minor inaccuracies commonly appear in the names and especially the lineages of the main protagonists, in numerals, and in place names. From among the more serious confusions, it is worth mentioning the treatment of the wicked viziers Saqardīs and Saqardiyūs, who are the principal enemies of Sayf ibn Dhī Yazan. In Ms. London, the British Library, Or 4643 (271ff.), f. 231b, they are both hanged, only to reappear in the narration two folios later as if nothing had happened. They eventually disappear from the narration apparently unharmed.34 These and many other similar discrepancies originated during the process of textualizing, which must have been quite complex. The scribe could write the text on the basis of several hearings or source materials. Due to the fact that the narrations were lengthy, lively, and rich in action, the storytellers’ audiences were not likely to worry much about logical coherence, as they simply did not have enough time for a detailed analysis of the events. They were most probably not bothered by such irregularities. On the other hand, such contradictions would surely disturb the readers of these handwritten volumes.

COLOPHONS

A substantial part of the manuscripts of popular siyar is incomplete: the text is abruptly interrupted and lost, and consequently colophons are missing. Nevertheless, the complete manuscripts of Arabic siyar do end with the copyist’s colophon, which usually has the shape of one or even two upturned triangles, e.g. Ms. Sīrat Sayf London, the British Library, Or. 4643, f. 271a. Mariana Klar, “A Study of MS BM 4274 Or. in Relation to the Printed Text of Sīrat Saif b. Dhī Yazan” (MPhil diss., Pembroke College, 1994), 34. 34

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 63 Because of the fact that popular siyar are usually undated and anonymous, the colophons of these documents do not contain the usual information that appears in the colophons of the texts from the literary canon. The information is often limited to some typical formulas indicating the end of the text. Note the case of Ms. Sīrat Sayf Cairo 4592 (two volumes, 2nd volume, f. 311b), which runs as follows: ‫وهذا ما ا���ى الينا | من الس��ة البديعة الفر��ة | وما �� ل��ل� سيف‬

‫وما سار | ��� صن ّف راوي هذه الس��ة وهو | ابو ا��عا�ي راوي س��ة‬

�ّٰ ��‫ا��مصار | وسايق الن ّيل من أرض ا��بشة إ�ي | هذه ا��يار ر��ة ا‬

| ‫تعا�ي | �ل��م ا��ع�ن و��� من م�� | من اموات ا��س���ن و�لينا اذا‬

���‫�دنا ال��م رب العا���ن وص�� | ا�� ّٰ� ��� سيدنا ���د الن�ي ا���يّ | و‬ .‫ا�� و��به وس� ّ� ام�ن | وا���د �� ّٰ� رب العا���ن‬

...and this is what came down to us from this wonderful and unique sīra and what happened to King Sayf according to the composition of the narrator of this sīra, Abū ’l-Maʿālī, the narrator of sīrat al-Amṣār about the one who diverted the Nile from the country of Ethiopia into these regions. May God have mercy with all of them, and with all dead Muslims, and with us when we return to them, the Lord of the Universe. Blessings of God be upon our Muḥammad, the illiterate Prophet, and upon his family and his companions, amen. Praise be to God, the Lord of the Universe.

Fortunately, there are also more informative colophons. In Ms. Sīrat Sayf, Leipzig Vollers 630, f. 193b, the colophon is not distinguished from the main text; and the scribe revealed his name, the date of the composition, and even the name of the patron. This colophon reveals that the text was written by a certain Muṣṭafā ibn al-Ḥāğğ Yūsuf, who was commissioned by as-Sayyid Muḥammad al-Ḥamawī, giving as the time of copying the fifteenth day of the month of Rabīʿ al-Anwar of the year 1198 AH (1748):

64

ZUZANA GAŽÁKOVÀ ��‫خط مصط‬ ّ ‫وهذا ما ا���ى من كتـ]بـه[ ���� السيد ���د ا���وي من‬

،۱۱٩٨ ‫سنة‬۱٥ ‫ ���ر ربيع ا��نور‬،‫عفا ا�� ّٰ� عنه‬،‫ا�ن ا��اج يوسف‬

‫��ت‬

More elaborated colophons give away considerably more information about manuscripts and the works they contain, e.g. Ms. Sīrat al-malik at-Tubbaʿ Paris, BnF Ar. No. 382. In this case, the colophon offers a sort of summary of the content with various textual formulas (ḥamdala, ṣalwala, šahāda, ḥawlaqa, etc.); the name of the scribe; the date of the composition; and the prayers in which God is asked to grant forgiveness to the scribe, the reader, the listener, the owner, and the author. An interestingly composite colophon appears in Ms. Sīrat Sayf Leiden Or. 1680a, which covers ff. 322a–322b. Apart from various textual formulas, there appears the name of the scribe (Aḥmad ibn Būqāsim al-Mīʿādī) as well as some additional information about him, including that he was of al-Mālikī madhhab and of al-Ašʿarī school—all this in an upturned triangle. Other information then follows, including the date of the manuscript’s composition— Thursday, the 25th day of the month of Šawwāl of the year 1259— as well as prayers in which God is asked to grant forgiveness to the scribe; to his parents; to all Muslims, male and female, living or dead; and to all readers and scribes. In the end, the scribe even adds some verses containing material frequently encountered in texts from the literary canon. The poetry reads as follows (the text is transcribed for practical reasons in naskh as the manuscript is written in maghribī): I wrote some poetry and realized that without any doubts

‫شعر كتبت وقد أيقنت �� شك ا��ي‬

My hand would pass away and the book would remain

‫ستفنا ��ي ويبقا ��ا��ا‬

No doubt that the Lord will question [my hand] tomorrow

‫و�� شك أن ا�� ّٰ� سايلها �دا‬

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 65 I wish I knew what the answer would be Either comfort and rest in paradise

‫فيا ليت شعري ما يكون جوا��ا‬

Or the unbearable torment of hell

‫فاما نع�� �ي ا��نان ورا�ة‬ ‫فاما جح�� �� يطاق �ذا��ا‬

CONCLUSION

This contribution has provided a brief outline of the characteristic features of the manuscripts of Arabic popular siyar. Special attention was dedicated to a description of their appearance: the titlefolios, layout, script, and colophons. Generally, the manuscripts of popular siyar share these common features. There are, however, some splendid exceptions of illuminated manuscripts of great beauty and intricacy showing evidence of considerable skill and expense in their production. An example of such a well-executed manuscript is Ms. or. fol. 2195 from Berlin Staatsbibliothek, entitled Sīrat al-Iskandar ibn Fīlibus al-Yūnānī,35 which does not belong to the Wetzstein’s collections and which the library acquired only in 1896. It contains a considerable quantity of vivid miniatures that represent the various scenes of the story. Despite the fact that illuminated manuscripts of popular siyar are a rare phenomenon, there is another exception: the famous Ms. Tübingen MA VI 32 of Ḥikāyat ʿUmar an-Nuʿmān (from the second volume of Sīrat Alf Layla wa Layla36), which straddles the divide between the realm of popular siyar and the Arabian Nights and which resembles the sīra genre not The manuscript is digitized and accessible online: https://digital.staatsbibliothekberlin.de/werkansicht?PPN=PPN632 210885&PHYSID=PHYS_0007&DMDID= [10.9.2020]. 36 The manuscript is digitized and accessible online: http://idb.ub.uni-tuebingen.de/opendigi/MaVI32 [10.9.2020]. 35

66

ZUZANA GAŽÁKOVÀ

only because of the term sīra in its title but also due to its resemblance to Sīrat Muğāhidīn in its content.37 Both these examples prove that there are still unanswered questions about the production of these voluminous texts. For now, this contribution can conclude with a narrator’s comment from Ms. Berlin Staatsbibliothek We. 643 of Sīrat Sayf (f. 2v): ‫قال ا��والف ر��ه ا�� ّٰ� تعا�ي �� أق�� الك��م ��ن ��حنا يطول و خفنا‬ ‫من ز يادة ا����ل لك� سامع‬

The author, may the Almighty God’s mercy be upon him, spoke, and then he shortened his speech, for our explanation is getting very long, and we are afraid that listeners will be bored.

BIBLIOGRAPHY

Ahlwardt, Wilhelm. Verzeichniss der arabischen Handschriften der königlichen Bibliothek zu Berlin, [= Die Handschriften-Verzeichnisse der königlichen Bibliothek zu Berlin]. Volume 8. Berlin: Asher, 1896.

Doufikar-Aerts, Faustina. “Ġarāʾib or ʿAğāʾib, that’s the Question: Vocalised Script in Two Arabic Romances of Alexander.” In Moyen arabe et variétés mixtes de l’arabe à travers l’historie. Actes du Premier Colloque International (Louvain-la-Neuve, 10–14 mai 2004), edited by Jérôme Lentin and Jacques Grand’Henry. Louvain: Université Catholique de Louvain, 2008. Gažáková, Zuzana. “Arabic Epics.” In Comparative Oriental Manuscript Studies: An Introduction, edited by Alessandro Bausi. Hamburg: Tredition, 2015.

Heath, Peter. “Sīra šaʿbīya.” In Encyclopaedia of Islam, CD-Rom edition. Leiden: Brill, 1999.

37

Cf. Claudia Ott, Metamorphosen des Epos, 113–121.

ARABIC POPULAR SIYAR AND SĪRAT SAYF IBN DHĪ YAZAN 67 Herzog, Thomas. Geschichte und Imaginaire: Entstehung, Űberlieferung und Bedeutung der Sirat Baibars in ihrem sozio-politischen Kontex. Wiesbaden: Harrassovitz, 2007.

Klar, Mariana. “A Study of MS BM 4274 Or. in Relation to the Printed Text of Sīrat Saif b. Dhī Yazan.” MPhil diss., Pembroke College, 1994.

Lentin, Jérôme. “Variétés d´arabe dans des manuscrits syriens du Roman de Baybars et histoire du texte.” In Lectures du Roman de Baybars, edited by Jean-Claude Garcin. Marseille: Editions Parenthéses, 2003: 91–111. Liebrenz, Boris. “The Library of Aḥmad al-Rabbāṭ. Books and Their Audience in 12th to 13th/18th to 19th Century Syria.” In Marginal Perspectives on Early Modern Ottoman Culture: Missionaries, Travellers, Booksellers, edited by Ralf Elger and Ute Petruschka. Halle: Zentrum für Interdisziplinäre Regionalstudien, 2013. Löfgren, Oscar and Traini, Renato. Catalogue of the Arabic Manuscripts in the Bibliotheca Ambrosiana. Vol. 1 Antico fondo and medio fondo, Vicenza: Neri Pozza, 1975.

Ott, Claudia. “From the Coffeehouse into the Manuscript.” In Oriente Moderno 83 no.2/2003, edited by Giovanni Canova (special issue): 443–451.

Ott, Claudia. Metamorphosen des Epos: Sīrat Muğāhidīn (Sīrat al-Amīra Dāt al-Himma) zwischen Mündlichkeit und Schriftlichkeit. Leiden: CNWS Publications, 2003. As-Samawʾal al-Maghribī. Ifḥām al-Yahūd. Cairo: Dār al-hidāya, 1986.

Vita prophetarum, Muḥammed ben ʿAbdallāh al-Kisaʾi, ex codicibus, qui in Monaco Bonna, Ludg. Batav. Lipsia et Gothana asservantur, Isaac Eisenberg (ed.), Lugduni Batavorum: E. J. Brill, 1922.

Witkam, Jan Just. “Johann Gottfried Wetzstein’s Manuscripts Containing Arab Popular Stories.” In Manuscripts, Politics and Oriental Studies. Life and Collections of Johann Gottfried Wetzstein (1815– 1905) in Context, edited by Boris Liebrenz and Christoph Rauch. Leiden, Boston: Brill, 2019: 116–139.

A PORTABLE MAJLIS: ON PUBLISHING RELIABLE EDITIONS IN OTTOMAN MANUSCRIPT CULTURE ASLIHAN GÜRBÜZEL This paper describes an Ottoman practice of textual criticism that I dub “the portable majlis.” Mimicking the social reading majlis, textual authorities used paratextual space to engage critically with given texts and situate them within the broader canon. The work of these textual authorities, who operated as middle authors between the author and the copyist of the text, was recognized and sought out by the reading public. Following the work of one such middle author, İsmāʿīl Ḥaḳḳı Bursevī (d. 1725), I show that his efforts as a reliable textual critic were broadly recognized by Ottoman readers and copyists who prized his edition of texts as distinctly reliable versions. The paper suggests that the practice of authoring portable majlises as a form of publication became widespread at the turn of the eighteenth century due to two larger shifts in Ottoman reading culture: the rise of deep reading and the proliferation of new reading spaces, such as the public library. Both of these developments presupposed individual readers. It was the task of the middle author to provide guidance for the individual reader by reproducing the social reading majlis on the page.

When we buy a print book today, or download an ebook, we often flip past the title page quickly, without paying much attention to the details about the publishing company or the place of publication. Yet we also instinctively know that these pages are essential for the life of our book: these pages establish that we read the 69

70

ASLIHAN GÜRBÜZEL

words in the exact form and sequence in which the author intended them to appear before us. Her words reach us as she wants us to see them: no corruptions, no variations. In other words, these pages represent a social and legal procedure that culminates in a contract between the publisher and the author that guarantees the authenticity and the fixity of the text.1 Fixity, publication, impersonal transmission, and even editorial work are practices we tend to ascribe exclusively to print cultures. By contrast, manuscript cultures are considered to lack accuracy, veracity, and universal reliability. However, this common misconception is based less on the actual study of manuscript culture than on a misinformed need to amplify its contrast with print culture. A classic example of this approach is Elisabeth Eisenstein’s The Printing Press as an Agent of Change, which epitomizes the discourse on print and manuscript that prevailed until recently. According to Eisenstein, the technology of printing alone paved the way for the Renaissance, Reformation, and Scientific Revolution, therefore accomplishing nothing less than the success of Western civilization. In addition to enabling widespread publication, print is credited with the creation of accurate, standardized texts that allowed for a shared and reliable experience among readers. Print was objective, manuscript subjective; print produced accuracy, manuscript mere approximation. Is it true that before the widespread adoption of printing technology, readers wandered in a haze of confusion? Were their intellectual activities doomed from the start because their world was littered with corrupt, incomplete texts, and even with outright forgeries? To answer these questions, we need to move away from technological determinism and focus on the social and cultural practices that surrounded the book, be it in print or manuscript The author wishes to thank participants of the workshop “Peritext, Dots, and Marginalia in Middle Eastern Manuscripts” for their feedback on an earlier version of this paper. 1 For the formation of the legal aspects of authorship and its implications for modern authorship, see Ronan Deazley, Martin Kretschmer, and Lionel Bently, Privilege and Property: Essays on the History of Copyright (Cambridge, UK: OpenBook, 2010).

A PORTABLE MAJLIS

71

form. Historians of print have recently documented that early printed books could hardly qualify as uniform and standard. There was a gradual, historical process by which authors, readers, and commercial printers negotiated the norms of press and took legal steps to give printing culture its present shape. In other words, it was not the technological tool but the social-cultural practices shaping its usage that determined the unity and standardization of texts. The lessons from the history of early print apply to the study of manuscripts. It was the social and intellectual practices that surrounded publishing and reading, rather than the technology alone, that determined the quality and reliability of books in a manuscript culture.2 Many important manuscript works, especially those of canonical status, circulated clad in editorial paratexts that attest to an authoritative assessment of the text—in other words, to editorship. In this paper, I introduce a form of textual editorship that became widespread in Ottoman manuscript culture as of the early eighteenth century. The editor used marginalia, and occasionally even the space in between lines, to create the textual equivalent of a reading majlis. These notes, intended for public circulation, served as guidance for the reader concerning the reliability, meaning, and cultural relevance of the text. The editor’s expertise and editorial contribution were recognized by copyists, who reproduced the text with the utmost attention to these editorial marginalia. In other words, the marginal notes that appear like personal reading notes at first sight highlight a world of publication in which texts circulated not as bare texts, but as distinct editions. Paratextual editing for publication purposes elucidates a key strategy of standardization and verification that came to prominence in Ottoman manuscript culture as of the early eighteenth century.

For a recent iteration of this debate, see Elizabeth Eisenstein, “An Unacknowledged Revolution Revisited” and Adrian Johns, “How to Acknowledge a Revolution.” The American Historical Review 107/1 (2002): 87–105; 106–125. 2

72

ASLIHAN GÜRBÜZEL

CHOOSING AN EDITION IN THE AGE OF MANUSCRIPT

Manuscript readers and copyists were perpetually aware of the existence of variations, major or minor, between different versions of a manuscript book. Commentaries and later paratextual notes often discussed these variations as part of the work of textual criticism. Some variations were scribal errors, some were trivial differences, yet others were significant enough to qualify a work as a distinct “edition.” Edition consciousness was an important part of textual criticism, where the “editor” clarified on which version of a text the reader should rely and the reasons behind that choice. A wellknown example of a multiple-edition manuscript is Al-Fiqh alAkbar, an authoritative creed attributed to the founder of the Hanafi school of law, Abu Hanifa (d. 767). The creed circulated in two different manuscript versions, each going back to a different student of Abu Hanifa. The differences were not minor and touched upon a major theological debate. One version of AlFiqh al-Akbar declared that the prophet’s parents died unbelievers, since they died when Muhammad was very young and had not yet received the revelation. This declaration was unacceptable to many early modern Ottomans, since they could not bear to imagine that the prophet of Islam was unable to save the souls of his own parents. Other versions of the text were less contentious, leaving this part out. The theological debate about whether the prophet’s parents were granted salvation, therefore, inevitably hinged on which manuscript version of Abu Hanifa’s creed was authentic. In the seventeenth century, when Ottoman preachers had a major disagreement about the fate of the prophet’s parents, one of the reasons that the debate was prolonged was that both positions could be supported by the same text, only different versions; each side continued to argue that theirs was the “authentic” version of the text. For early modern readers, the debate was never fully resolved.3 For a summary of the seventeenth century debates, see Abdülehad Nuri, Te'dibu’l-Mütemerridin, MS Süleymaniye, Fatih 5293, ff 278a-315a; Kâtib Çelebi, Keşfü’ẓ- Ẓunün Ed. Şerafeddin Yaltkaya et al, Beirut: Dār alKutub al-ʻIlmīyah, 1992., vol II, p 1287. For modern discussions of the 3

A PORTABLE MAJLIS

73

This example of Al-Fiqh al-Akbar’s versions illustrates that choosing a reliable version was a crucial task before beginning to read a text. The task was often tackled in a social reading circle, as were other challenges of reading. The “ideal” Ottoman reader prior to the eighteenth century was a social reader. His choice of a reliable edition was arbitrated by his reading community: fellow madrasa-goers, the reading circle at the mosque or at the Sufi lodge, and later on at the coffeehouse or at the weekly or biweekly majlises one attended at the dwellings of the local ‘ulama or poets. Social reading practices informed not only which version of a text one was to take as authoritative, but also the cultural and intellectual meaning of the chosen version. They answered many key questions about the text: what are the important points of which one should make a mental note to deploy in future conversations or writing? What are the fine points of philology that need to be taken away from the text? Which opinions, while seeming neutral, are in fact expressions of disagreement with other known authorities in the field?4 If the reader wanted to read both a textual discussion about the reliability of a manuscript and instruction on how to absorb it, he consulted lengthy commentaries, most of which carried few topical signposts to help the reader navigate the book. The ideal reader of the eighteenth century, in contrast, sought guidance that was short, accessible, and amenable to individual study. Two developments altered the culture of manuscript edition in this period. First, readers were encouraged to engage with texts individually. They were encouraged to adopt methods of deep reading (mütalaa) to parse given texts using philological and analytical text, see Eric Ormsby, “The Faith of Pharaoh. A Disputed Question in Islamic Theology,” SI 98, no.9 (2004): 5–28. 4 Walter Andrews and Mehmet Kalpaklı, The Age of Beloveds: Love and the Beloved in Early Modern Ottoman and European Culture and Society (Durham: Duke University Press, 2005); Helen Pfeifer, “Encounter After the Conquest: Scholarly Gatherings in 16th-Century Ottoman Damascus,” International Journal of Middle East Studies 47/2 (2015): 219–39. The social nature of reading is vividly captured in Tülün Değirmenci, “Bir Kitabı Kaç Kişi Okur?,” Tarih ve Toplum: Yeni Yaklaşımlar 13 (2011), pp. 7-43.

74

ASLIHAN GÜRBÜZEL

tools. Khaled el-Rouayheb documents the rise of deep reading in Ottoman scholarly circles as of the late seventeenth century, underlining the growing importance of state examinations for acquiring scholarly posts as the main motivation for developing a practice of individual reading.5 Second, the rise of free-standing libraries transformed the culture of manuscript edition and circulation. Unlike earlier libraries, these new institutions were not annexes to madrasas, mosques, or lodges, but intellectual and social spaces that were endowed for the benefit of the reading public alone. İstanbul’s literati flocked to the new free-standing public libraries to read, write, and socialize. The infatuation with the new libraries quickly spread to the provinces. It was now in these libraries that learned Ottoman men read, copied, or deposited their manuscripts as a way of acquiring and exhibiting literacy.6 The new ideal reader possessed knowledge of the philological and analytical tools to assess a text and its cultural relevance. Given the increasing availability of texts, he favored shorter yet penetrative annotations over lengthier commentaries. An important manifestation of this transformation was the increasing practice of textual criticism in a new format that I dub the portable majlis. Portable majlises made extensive yet economical use of paratextual space to engage in a critical dialogue with the author of the text. The editor used marginalia, and occasionally even the space in between lines, to create the textual equivalent of a reading majlis. This mode of textual criticism provided guidance to the reader and invested the manuscript with the credibility of having been inspected by an authoritative annotator. Unlike the lengthy commentarial For the analysis of deep reading practices, see El-Rouayheb, Islamic Intellectual History in the Seventeenth Century: Scholarly Currents in the Ottoman Empire and the Maghreb (New York, NY: Cambridge University Press, 2015), 97–128. 6 For Ottoman libraries in general, see İsmail Erünsal, Ottoman Libraries: A Survey of the History, Development and Organization of Ottoman Foundation Libraries (Cambridge, MA: Harvard University Press, 2008). For freestanding public libraries of the eighteenth century, see Yavuz Sezer, “The Architecture of Bibliophilia: Eighteenth Century Ottoman Libraries.” (PhD Diss., MIT, 2016). 5

A PORTABLE MAJLIS

75

dialogues and annotations, the editor of a portable majlis was selfconsciously brief and accessible. The new reader sought after these edited texts as manifested in the paratexts—some examples of which I discuss below—as well as in anecdotal evidence. For instance, eighteenth-century readers often recommended to each other the annotations of certain individuals, recognizing annotators as authors. Similarly, library donors annotated works in a way that was clearly intended for the public eye.7 In short, the new reader preferred manuscripts verified and streamlined by a middle author. Middle authors and the portable majlises that they prepared were sought out by Ottoman manuscript readers and copyists. In what follows, I turn to a brief description of one interesting example of this cultural practice of manuscript verification: a successful edition of an iconically eloquent text that was circulated with consistent attention to the interventions of the middle author. This particular middle author was İsmāʿīl Ḥaḳḳı Bursevī (d. 1725), one of the most prolific Sufi authors of the eighteenth-century Ottoman Empire. In addition to lengthy commentaries on the Quran and the Masnavi, for which he was known, Bursevi undertook systematic philological criticism of classical works of Ottoman Turkish.8 One of these classical texts was Pearls of the Crown by Veysi al-Üskübi Sezer, Architecture of Bibliophilia, 226. Sami Arslan refers to ‘public’ annotations by Carullah Efendi that were meant for readers who would visit his library. See Arslan, “Der-Kenar’ın Gölgesinde: Carullah Efendi’nin Tasavvuf Kitaplarına Düştüğü Notlar,” in Osmanlı Kitap Kültürü, ed. Berat Açıl (İstanbul: İlem Kitaplığı, 2005), 205–254. 8 Tobias Heinzelmann’s thorough study of the philological work of İsmail Bursevi is an exception. See his “Anfänge einer türkischen Philologie? Bursalı İsmāʿīl Ḥaḳḳı kopiert und kommentiert Yazıcıoġlı Meḥmeds Muḥammedīye,” in Buchkultur im Nahen Osten des 17. und 18. Jahrhunderts, ed. Tobias Heinzelmann and Henning Sievert (Bern: Peter Lang, 2010): 99–150. Heinzelmann thoroughly documents Bursevī’s efforts to establish an authoritative version of a highly popular religious work, the Muhammediye of Yazicizade Mehmed (d. 1451). On Bursevi’s life and ouvre, see Günay Alpay Kut, “İsmāʿīl Ḥaḳḳı,” in Encyclopaedia of Islam, 2nd ed., ed. P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs (Leiden: Brill, 2008). 7

76

ASLIHAN GÜRBÜZEL

(d.1628), a work on the life of the prophet that was read, above all, for its mastery of fine Ottoman prose in a heavily Persianized vein. Contemporaries and posterity considered Veysi to be a master of this prose technique, bestowing on him the epithet “the Firdawsi of our time.” To be likened to Firdawsi, the author of one of the greatest Persian epics, was a grand and enviable achievement. Ottoman men of letters aspired to imitate Veysi’s highly prized technique and eloquence. Since no standard handbooks or textbooks were available to produce this prose, readers studied fine Persianized prose by scrutinizing good examples of it, such as Pearls of the Crown. Pearls thus functioned not only as a biography of the prophet, but also—and arguably, even more—as a philological guide to Ottoman high prose.9 It was therefore a claim to outstanding philological knowledge to annotate this text for readers, a claim that Bursevī was not shy to make. To be sure, annotating a text with an eye to the possibility of others reading was not an entirely novel phenomenon; rather, it was the eighteenth-century editor’s claim to authorship and its recognition by readers and copyists that was novel. The editor marks his authorship throughout the paratextual annotations. İsmāʿīl Ḥaḳḳı Bursevī signs most of his brief marginal notes with “Ḥaḳḳı,” for example, and copyists reproduce these signatures. In addition to signing his notes, Bursevī marks the distinguishing and rewarding features of his editorial efforts with a note that he adduces right after the colophon. Before [I made my] amendments, this book was full of glaring mistakes and omissions […]. Some literate people ventured to amend it before me and commented upon some of its enig-

On the development of a stylized, Persianate prose style as the marker of distinction among Ottoman bureaucratic class, see Emine Ekin Tuşalp-Atiyas, “Political Literacy and the Politics of Eloquence: Ottoman Scribal Community in the Seventeenth Century” (Ph.D. Diss., Harvard University, 2013). On Veysī and his importance for Ottoman prose composition, see Th. Menzel and Edith Gülçin Ambros, “Weysī,” in Encyclopedia of Islam, 2nd ed., ed. P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs (Leiden: Brill, 2008). 9

A PORTABLE MAJLIS

77

mas. Hence yet another melody has been added to the discord, clouding the right melody [further], causing grief and ruining [the harmony] as acridity ruins even the sweetest [things]. I have laboured to amend it, hoping for [otherworldly] returns.10

This note leaves no doubt that Bursevī’s editorial notes were neither intended nor perceived as private note-taking; the edition was meant to be public and legible. Bursevī underlines the distinct nature of his edition as “harmonious,” purged not only of mistaken interpretations and reproductions but also of any note that, while true, was not useful to the reader. Did Bursevi’s cacophony analogy refer to lengthy commentaries on the text, to haphazard annotations made without sharing his concerns for pithiness and wide readership, or to both? Regardless of the reference he had in mind, as a middle author he advertised his own edition, emphasizing the reliability and quality of his annotations. In improving the reader’s experience, he used paratextual notes to enter a critical dialogue with the base text instead of overwriting it, even at instances when he claimed the base text had been corrupted by scribes. İsmail Hakkı Bursevī’s editions thereby rendered the reader’s experience of the text dialogical, much like in an actual majlis. In addition to enhancing the text with additional credibility and enveloping it in a dialogical, majlis-like reading experience, what was it in İsmāʿīl Ḥaḳḳı Bursevī’s editorial work that appealed to readers? Most of Bursevī’s interventions help the reader notice the linguistic mastery of the base text, particularly by helping to penetrate the logic behind Veysi’s meticulous word choices. By far the majority of the editorial marginalia are on lexicography. Veysi’s prose abounded with vivid and playful analogies and connotations, the key to which was the deployment of an extremely rich vocabuMS Hamburg Cod.Or. 322, f 153a. For reproductions of the original colophon and the notes that surround it, see Aslıhan Gürbüzel, “Ein Manuskript mit drei Kolophonen: Cod. orient. 322: A Triple-colophon Manuscript: Cod. orient. 322,” in “Special Issue: Catalogue for the Exhibition ‘Wonders of Creation – Ottoman Manuscripts in Hamburg Collections’ at the State and University Library Hamburg,” ed. Janina Karolewski and Yavuz Köse, Manuscript Studies 9 (2016), 123–128. 10

78

ASLIHAN GÜRBÜZEL

lary, including some words seldom used in simpler Ottoman prose. Many readers would read the text for the purpose of absorbing such linguistic mastery, and Bursevī’s goal as an editor was to help the reader notice the philological finery by way of signed marginal notes. In these notes, he singled out select words within the text and copied a select meaning from the relevant dictionary entry in Arabic or Persian. Allusions to terminology are explained with reference to the relevant literature that the reader can consult. For instance, one marginal note refers the reader to al-Farabi’s work on music, another to a book of marvels (ajaib), should the reader want to read more about a legendary emerald that repels flies.11 Much less common, yet still present, are notes on the content of the work. Most of these content notes are cross-references to other works on the biography of the prophet that are recommended to the reader desiring further study. The editor did not see himself merely as an aide to the author, but also as his corrector. Some corrections are of pietistic etiquette. For instance, the base text claims that when Muhammad moved to Medina from Mecca, his companion Ayyub al-Ansari hosted him in his house, allocating the first floor of his two-story home to the prophet. The author notes the living arrangement in passing, without much attention to the hierarchy implied therein. Bursevī, however, notes that it was the Prophet’s own request to live below his companion, “Otherwise, it would be inconceivable for a companion to be so disrespectful to live above the prophet.” Bursevī’s dialogical corrections served not only to mark the disputable sections of the text, but also to underline the editor’s authority to evaluate the text, particularly his philological credentials. Therefore, his signed marginalia adopt a more critical tone in correcting grammatical errors than in debating pious etiquette. In correcting the content or the propriety, Bursevī resorts to a wellknown strategy of decorum. After fixing the textual error, Bursevī comments that a scholar of Veysi’s calibre was clearly above making the mistake in question. It must have been the copyist who, in copying out the text, corrupted Veysi’s original wording. By invokMS İstanbul Research Institute ŞR 189, f 37b, same note on MS Istanbul University T 1521 39b. 11

A PORTABLE MAJLIS

79

ing the possibility of scribal errors, the annotations could disagree with the text at hand without disrespecting the author’s scholarly persona. In correcting grammar or other linguistic features, though, Bursevī’s editorial marginalia resort to no such polite formula.12 The marginalia that criticize the author’s use of a grammatical particles or form are direct and unencumbered by excuses, simply declaring the base text mistaken. These two types of marginalia provide an important insight into the authority claims of middle authors and how they conveyed their authority beyond their personal circles, within the larger reading community. The middle author used the paratextual space of annotations to exhibit his philological mastery, as well as his learning on the subject at hand. Yet it was his proficiency in the former, more so than the latter, that warranted the authority of his annotations. Did Bursevī’s hard work and self-promotion bear fruit? Manuscript evidence establishes that his editorial authorship was widely acknowledged and appreciated. I have been able to locate three reproductions of Bursevī’s edition of Pearls, mostly by chance, since catalogues do not typically pay systematic attention to middle authors.13 These three copies suggest the existence of further copies of the same edition. Two of the copies end with Bursevī’s colophon, which dates Bursevī’s rectification of the book (waqa’a ıslah haze’l-kitab) to 1707. Because of the ending and the signed marginalia, modern researchers have mistaken Ḥaḳḳı for a mere copyist of Veysi’s work, failing to notice the larger world of publishing and edition that surrounded the manuscript.14 Thanks to the third copy, MS İstanbul Research Institute ŞR 189, f 62a // MS Istanbul University T 1521, f 63b; ŞR 189, f 62a // Istanbul University T 1521, 63b. 13 In addition to the Hamburg copy, the following manuscripts copy Bursevī’s edition with striking consistency: MS Istanbul University, Turkish Manuscripts (T) 1521; MS İstanbul Research Institute, Şevket Rado Collection 189. I thank Nir Shafir for drawing my attention to the latter copy. 14 Nuran Öztürk presents the Istanbul University copy as Bursevī’s personal copy, whereas the catalog for the Pera Museum manuscript collection mentions Bursevī as the copyist. See Nuran Öztürk, “Siyer Türü ve 12

80

ASLIHAN GÜRBÜZEL

we are made aware of this larger world, as it copies not only the colophon, but Hakkı’s marginalia to the colophon cited above. This note is missing from the other two copies, which implies the existence of at least one more copy in circulation, perhaps more. Of the three manuscripts, two were copied in Bursa, where İsmail Hakkı was based as the preacher of the Grand Mosque (Ulu Camii) at the time he edited the text. The Pera Museum copy was endowed to a Sufi lodge, the Ashrafiyya, seven years after Bursevī completed the edition. The Hamburg copy was copied out by a madrasa professor named Hammamizade Sayyid Mustafa, son of alHajj Musa in October 1780. Whether the edition reached beyond Bursa is a question that awaits the discovery of further copies of Bursevī’s edition. Yet it is clear from these two dated copies that Bursevī’s edition reached beyond his personal circles, where it was preserved with utmost attention to his paratextual notes. Furthermore, the edition was recyled and copied for a long time, through nearly the entire span of the eighteenth century.

CONCLUSION

Early modern reading was a social affair. Rather than encountering texts in a vacuum, readers encountered texts through social majlises or lengthy commentaries that provided guidance on how to understand and how to situate the manuscript’s text and discussed the reliability of the text. An important shift took place in Ottoman reading culture in the eighteenth century; guidance come to be provided through paratextual editorial work, which continued to pay homage to the reading majlis in its insistence on the dialogical form. Despite mimicking the majlis, however, the new portable majlis was meant to meet the needs of an individual reader singlehandedly. The broadly recognized practice of embedding texts within a critical, comprehensive, yet accessible apparatus shows that manuscript cultures developed effective practices of textual standardization that served to create a shared experience among the wider Siyer-i Veysī” (Ph.D. Diss., Erciyes University, 1997), 49; Günay Kut and Commission, İstanbul Araştırmaları Enstitüsü Yazma Eserler Kataloğu, vol. I, 246.

A PORTABLE MAJLIS

81

reading community. Studying these cultural practices helps us move beyond seeing manuscript cultures as belonging to the dark and confused age that came before the technological light of print.

BIBLIOGRAPHY

Andrews, Walter and Mehmet Kalpaklı. The Age of Beloveds: Love and the Beloved in Early Modern Ottoman and European Culture and Society. Durham: Duke University Press, 2005.

Arslan, Sami. “Der-Kenar’ın Gölgesinde: Carullah Efendi’nin Tasavvuf Kitaplarına Düştüğü Notlar.” In Osmanlı Kitap Kültürü, edited by Berat Açıl. İstanbul: İlem Kitaplığı, 2005.

Deazley, Ronan, Martin Kretschmer, and Lionel Bently. Privilege and Property: Essays on the History of Copyright. Cambridge, UK: OpenBook, 2010. Değirmenci, Tülün. “Bir Kitabı Kaç Kişi Okur?,” Tarih ve Toplum: Yeni Yaklaşımlar 13 (2011), pp. 7–43.

Eisenstein, Elizabeth. “An Unacknowledged Revolution Revisited.” The American Historical Review 107/1 (2002): 87–105.

Erünsal, İsmail. Ottoman Libraries: A Survey of the History, Development and Organization of Ottoman Foundation Libraries. Cambridge, MA: Harvard University Press, 2008.

Gürbüzel, Aslıhan. “Ein Manuskript mit drei Kolophonen: Cod. orient. 322: A Triple-colophon Manuscript: Cod. orient. 322.” In “Special Issue: Catalogue for the Exhibition ‘Wonders of Creation – Ottoman Manuscripts in Hamburg Collections’ at the State and University Library Hamburg,” edited by Janina Karolewski and Yavuz Köse, Manuscript Studies 9 (2016). Heinzelmann, Tobias. “Anfänge einer türkischen Philologie? Bursalı İsmāʿīl Ḥaḳḳı kopiert und kommentiert Yazıcıoġlı Meḥmeds Muḥammedīye.” In Buchkultur im Nahen Osten des 17. und 18. Jahrhunderts, edited by Tobias Heinzelmann and Henning Sievert. Bern: Peter Lang, 2010.

Johns, Adrian. “How to Acknowledge a Revolution.” The American Historical Review 107/1 (2002): 106–125.

82

ASLIHAN GÜRBÜZEL

Kut, Günay Alpay. “Ismāʿīl Ḥaḳḳī.” In Encyclopaedia of Islam, second edition, edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs. Leiden: Brill, 2008.

Menzel, Th. and Edith Gülçin Ambros. “Weysī.” In Encyclopedia of Islam, second edition, edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs. Leiden: Brill, 2008.

Ormsby, Eric. “The Faith of Pharaoh. A Disputed Question in Islamic Theology.” SI 98, no.9 (2004): 5–28. Öztürk, Nuran. “Siyer Türü ve Siyer-i Veysī.” Ph.D. Diss., Erciyes University, 1997.

Pfeifer, Helen. “Encounter After the Conquest: Scholarly Gatherings in 16th-Century Ottoman Damascus.” International Journal of Middle East Studies 47/2 (2015): 219–39.

El-Rouayheb. Islamic Intellectual History in the Seventeenth Century: Scholarly Currents in the Ottoman Empire and the Maghreb. New York, NY: Cambridge University Press, 2015.

Sezer, Yavuz. “The Architecture of Bibliophilia: Eighteenth Century Ottoman Libraries.” PhD Diss., MIT, 2016. Tuşalp-Atiyas, Emine Ekin. “Political Literacy and the Politics of Eloquence: Ottoman Scribal Community in the Seventeenth Century.” Ph.D. Diss., Harvard University, 2013.

CHAPTER DIVISIONS AND THE INTERPRETATION AND TRANSMISSION OF THE T OSEFTA

BINYAMIN KATZOFF∗ The Tosefta is a compilation of Jewish law composed in Palestine in the third century. It is closely related to the Mishna, another compilation of Jewish law produced in the same place and time, which became the most important legal work of its period in rabbinic circles. The Tosefta contains a considerable amount of interchange between the textual witnesses in chapter divisions. In most cases, the Erfurt manuscript (hereafter E) has an additional chapter division that is not found in the other witnesses, and the additional chapter division in E matches, or closely approximates, the corresponding chapter division in the Mishna. This indicates that the E tradition was transmitted by a learned scribe, for which the Tosefta was not an independent work, but one that was to be studied together with the Mishna, hence, internal divisions of the Mishna were transferred to the Tosefta.

The Tosefta is a compilation of Jewish law that was composed in Palestine in the third century CE and is comprised of predominantly legal sayings of sages who lived from the second century BCE to This research was supported by the Israel Science Foundation (Grant number 1292/18), and by the Beit Shalom Fund, Kyoto, Japan. ∗

83

84

BINYAMIN KATZOFF

the mid-third century CE.1 The Tosefta is closely related to the Mishna, another compilation of Jewish law produced in the same place and time. The Mishna quickly became the most important legal work of its period in rabbinic circles, eventually becoming the basis for both the Babylonian Talmud and the Palestinian Talmud. While the Tosefta never achieved the canonical status of the Mishna, rabbis of late antiquity did use material from it in their discussions in both the Babylonian Talmud and the Palestinian Talmud, and leading scholars of the Babylonian academies of the early middle ages, the Geonim, clearly viewed it as an important and authoritative work.2 The earliest extant manuscripts of the Tosefta are those written in the Levant in the tenth century and discovered in fragmentary form in the Cairo genizah and in book bindings in early modern municipal archives across Europe.3 Almost all the extant copies of the Tosefta written before the twelfth century originated in the Levant. Only from the twelfth century on do we have a significant See Jacob N. Epstein, Mevo’ot le-sifrut ha-tanna’im (Jerusalem: Magness, 1957), 241–251; Chanoch Albeck, Meḥkarim ba-baraita ve-tosefta veyaḥasan la-talmud (Jerusalem: Mosad Harav Kuk, 1969), 137; Abraham Goldberg, “The Tosefta—Companion to the Mishna,” in The Literature of the Sages, ed. Shmuel Safrai (Leiden: Brill, 1987), 283–284; Paul Mandel, “The Tosefta,” in The Cambridge History of Judaism, ed. S. Katz (Cambridge: Cambridge University Press, 2006), 316–321. 2 See Saul Lieberman, Tosefeth Rishonim, vol. II (Jerusalem: Mosad Harav Kuk, 1939), 7–15; Binyamin Katzoff, “The Tannaitic Sources in Halachot Psukot,” Shenaton Ha-Mishpat Ha-Ivri 25 (2008): 199–216; Binyamin Katzoff, “Did Rav Natronai Gaon Use the Tosefta in His Responsa,” Alei Sefer 20 (2008): 17–27. 3 See, for example, MS Norcia Archivio Storico Comunale Fr. ebr 1, Perani, “Il più antico frammento,” 261–265; Guenter Stemberger, “I frammenti della Tosefta di Norcia, e il loro contributo alla studio della tradizione testuale,” in La Genizah italiana, ed. Mauro Perani (Bologna: Il Mulino, 1999), 267–273; Binyamin Katzoff, “Bookbindings of Tosefta from Norcia and their Place in the Textual Tradition,” Jewish Studies Internet Journal 15 (2019): 1–17. Some of the fragments may have been copied as early as the ninth century; see, for example, MS CUL T-S AS 78.205. 1

CHAPTER DIVISIONS AND THE TOSEFTA

85

number of manuscripts written in Europe.4 The manuscripts of the Tosefta are divided into two distinct textual traditions, extensive exemplars of which are the Erfurt manuscript, on the one hand, and the Vienna manuscript and the editio princeps (Venice, 1521) on the other. Both textual traditions can be detected in early eastern Mediterranean manuscripts.5 Similar to other rabbinic halakhic compositions, the Tosefta is also divided into smaller hierarchical units; these units are noted in a variety of ways within the existent manuscripts, though there are considerable discrepancies as to how each manuscript divides the text. Internal divisions and internal titles, like other elements of peritext, “control one’s whole reading of the text,” in the words of Genette.6 In this paper, I will argue that the differences in how each manuscript divides the text reflect the way the Tosefta was studied and, in the eyes of those who introduced these divisions, the way it should be understood and interpreted. I will begin with a description of the subdivisions found in the Tosefta and the different ways in which these divisions are noted in the manuscripts. Following this, I will analyze the discrepancies between the manuscripts concerning chapter divisions, how these divisions impact the interpretation of the text, and what we can learn from them concerning the milieu in which the Tosefta was studied and transmitted.7 These include the three broad scope manuscripts of the Tosefta: MS Berlin, Staatsbibliothek, Or. 20 1220 (formerly MS Erfurt, E); MS Wein, Osterreichische Nationalbibliothek, Heb. 20 (V); MS London, British Library, Add. 27,296 (L). 5 On the division into two textual traditions, see Robert Brody, Mishnah and Tosefta Studies (Jerusalem: Magness, 2014), 45; Robert Brody, Mishnah and Tosefta Ketubbot: Text, Exegesis, and Redaction (Jerusalem: Magness, 2015), 20. 6 Gerard Genette, Paratexts: Thresholds of Interpretation, trans. Jane E. Lewin (Cambridge: Cambridge University Press, 1997), 2. Genette’s work here is based on Philippe Lejeune, Le Pacte Autobiographique (Paris: Seuil, 1975), 45. 7 A detailed description of the textual findings concerning the discrepancies chapter divisions in extant manuscripts can be found in Bin4

86

BINYAMIN KATZOFF

MARKING THE INTERNAL DIVISION OF THE T OSEFTA

Similar to other elements of the peritext, the internal division of a composition is meant to aid those reading the text or referring to it in various ways.8 For example, the internal division can indicate a thematic division of the composition into various subjects. A different form of internal division might divide the composition into sub-units of fixed lengths in order to facilitate study and memory. Indeed, quite a few works in antiquity were originally transmitted without any internal division or with minimal divisions. Over the course of time, medieval tradents and copyists added increasing levels of sub-divisions.9 Rabbinic compositions also developed in a similar manner. We can trace increasing sub-division in rabbinic compositions transmitted in the medieval period, as well as the variety of means gradually developed by scribes to note the hierarchical levels of sub-divisions which internally divide the text. Before I discuss the internal division of the Tosefta and how this division is noted, I want to bring into the discussion a distincyamin Katzoff, “Chapters of the Tosefta: The Origin and Significance of the Differences in the Chapter Divisions,” Alei Sefer 28 (2018): 7–30. In this article, I focus on describing the various means through which chapter divisions are noted in the manuscripts, and I analyze the impact the different divisions have on the meaning of the text. 8 The use of internal division is also dependent on the nature of the composition. A text that is essentially one unit, such as some medieval epics, may contain no internal divisions and no subheadings. Conversely, in a book that is composed from a collection of ungrouped independent thoughts, every passage may be a chapter in itself with its own heading. Similarly, the function and the means for noting a subdivision in a composition transmitted orally may be different from a written work. See Genette, Paratexts, 294–296; Shlomo Naeh, “The Structure and the Division of Torat Kohanim (B): Parashot, Perakim, Halakhot,” Tarbiz 69 (2000), 64–65. 9 See Jocelyn Small, Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity, (London: Routledge, 1997), 12–15; Frederick G. Kenyon, Books and Readers in Ancient Greece and Rome (Oxford: Clarendon Press, 1951), 67–69; John Van Sickle, “The Book-Roll and Some Conventions of the Poetic Book,” Arethusa 13 (1980), 9–12.

CHAPTER DIVISIONS AND THE TOSEFTA

87

tion made by Genette concerning subordinate divisions and titles that correspond to a contrast in authorial positioning. Genette distinguishes between what he calls “thematic titling,” in which the title refers to something about the content of the textual unit and expresses a demonstrative stance on the part of the author toward his work, and what he calls “designative titling,” which expresses a more restrained stance of the author.10 While Genette’s distinction is clearly correct, in my opinion there is benefit in presenting the distinction in a slightly different, more complex manner. At one end of the spectrum of how expressive a title is, we find titles that describe the content of the textual unit, thus making it clear that whoever wrote the title was directing the reader to a particular understanding of the main content of the text. At the other end of the spectrum stand texts that contain no internal divisions or secondary titling, leaving the reader himself to analyze the text and to determine the relationship between the content of its various parts. Between these two ends stands a broad spectrum of types of divisions and titles through which the person dividing the text and inserting the title can transmit various levels of information to the reader. For example, a title that specifies not only the division itself but also the serial number of the textual unit can provide the reader with information about the hierarchical structure of the composition and the place of the current text within the larger unit. Similarly, headings placed in a larger font, those with a larger space between the separated units of text, or those accompanied by special decorations suggest to the reader that the heading represents a more significant break between the units.11 In other words, instead of the sharp distinction between “thematic titling” and all other forms of titling, I believe it more useful to describe the type of information and the level of emphasis expressed in the various types of headings. As we shall see below, the headings used in manuGenette, Paratexts, 315. See Malachi Beit Arie, Hebrew Codicology, Historical and Comparative Typology of Medieval Hebrew Codices Based on the Documentation of the Extant Dated Manuscripts until 1540 (Preprint internet version 0.9, 2018), 362; Naeh, “The Structure,” 78–83. 10 11

88

BINYAMIN KATZOFF

scripts of the Tosefta present a variety of ways of emphasizing the internal division. The Tosefta, like the Mishna, contains three main levels of division. Here it may be useful to compare the sub-division of the Tosefta with that of Justinian’s Digest, the most important work of Roman law, whose compilation was completed in 533 CE, about two and a half centuries after that of the Tosefta. The sub-division of the Digest is not the same as that of the Tosefta, but the comparison may help clarify the nature of that used in the Tosefta. The Digest is composed of fifty so-called “books” of roughly equal size. The division into books does not reflect a division based on content but rather a division enacted in order to establish books of relatively uniform length. The “books” were originally grouped into seven so-called “parts.” The vast majority of the “books” are divided internally into what are called “titles,” of greatly varying length, each devoted to a single topic and named accordingly. There are altogether four hundred and thirty-two titles.12 The Tosefta is also composed of books, each of which is called a masekhet, or tractate. There are about sixty tractates, grouped into six “orders.” However, unlike in the Digest, the first two levels of division of the Tosefta—the division into tractates and the grouping of the tractates together into “orders”—are based on the subject matter.13 Hence, the length of the tractates and orders varies signif-

Each “title” is composed of so-called ‘fragments,’ that is, quotations from works of jurists of the classical period of Roman law, the first century BCE to the third century CE. In the medieval period, the “fragments” were further divided into paragraphs. I will not address the last two divisions. 13 On the division of the Mishnah and Tosefta into orders/tractates/chapters, see Jacob N. Epstein, Introduction to the Mishnaic Text (Jerusalem: Magness, 1964), 980–1005; Guenter Stemberger, Introduction to the Talmud and Midrash, translated by M. Bockmuehl (Edinburgh: T&T Clark, 1996), 109–123; Ishay Rosen Zvi, “Introduction to the Mishna,” in The Classic Rabbinic Literature of Eretz Israel: Introduction and Studies, ed. M. Kahana (Jerusalem: Yad Ben Zvi, 2018), 4–8. 12

CHAPTER DIVISIONS AND THE TOSEFTA

89

icantly.14 At these two levels, there are no differences among the various manuscripts of the Tosefta; and, aside from some slight variations, the division of the Tosefta into tractates and orders is identical to that of the Mishna.15 The third level of division in the Tosefta, the division of the tractates into chapters, is quite different from the first two levels. The immediately obvious difference is that while some chapter divisions reflect differences in subject matter, other chapter divisions clearly do not. The latter type, those that do not reflect the subject matter, are termed “external” by some recent scholars and appear, at least in part, to derive from technical considerations. That is to say, in some places, subjects extend over several chapters, while in others, chapters include several only loosely related subjects. At times, it seems that the motivation to divide a chapter into two is simply to standardize the length of chapters in order to facilitate study and memorization.16 A second difference between the division into orders and tractates and that into chapters is that the chapter divisions of the Tosefta are independent of and do not correlate with those of the For example, Pesachim—the tractate dealing with the laws of Passover, which includes ten chapters—and Rosh Hashana—the tractate on the New Year rites, which includes four chapters (or fewer, depending on the various MS; see in the following)—are grouped together with other tractates into an “order” devoted to festivals. 15 All tractates found in the Mishna are also in the Tosefta, with the exception of Avot, Midot, Tamid and Kinnim. In the Tosefta, Kelim is sub-divided into three tractates, whereas in the Mishna it is not. See Stemberger, Introduction to the Talmud, 150; Mandel, “The Tosefta,” 317. 16 For example, the laws of reading the Shema are spread out in the first two chapters of Tosefta Berakhot, without any clear distinction as to how this split was made. See Saul Lieberman, Tosefta, vol. I (New York: JTSA Press, 1955), 1–11. In contrast, the last chapter of Berakhot contains collections of laws concerning various subjects not particularly connected one to the other, including blessings recited over seeing unusual phenomena, blessings over the performance of commandments, prayers in dangerous places and other topics. On this, see Lieberman, Tosefta, 32– 40. See also Epstein, Introduction to the Mishnaic Text, 994–995. 14

90

BINYAMIN KATZOFF

Mishna. The Tosefta is divided into about 420 chapters, whereas the Mishna contains about 100 more chapters. Thus, at this level, it is clear that the division into chapters of one composition was not based on the other.17 The third difference—and this will be my focus below—has to do with the consistency and inconsistency of chapter divisions across the various manuscripts of the Tosefta. Whereas the division into tractate and order is consistent between manuscripts, when it comes to the division into chapters, there is some variety among the manuscripts of the Tosefta. In nineteen places, a text presented as a single chapter in some manuscripts is divided into two or three chapters in other textual witnesses. Thus nineteen chapters in the one textual grouping comprise thirty-nine chapters in the other.18 My argument is that these additional chapter divisions were the result neither of inaccurate copying nor of arbitrary decisions but are rather the product of editorial activity, mainly in one of the textual traditions of the Tosefta; and with proper analysis, we can detect the motives for these additional divisions. Before I begin to analyze the variances in chapter divisions, I would like to point out that the manuscripts of the Tosefta contain different types of subheadings, which differ in terms of both the words used to signify the division and the manner in which these words are presented. These various subheadings emphasize different aspects of the internal division of the text and, accordingly, a different degree of involvement on the part of the person choosing to include these words and to present them in this manner when transmitting or copying the text.19 The various types of headings The Tosefta contains approximately 1.5 times as many words as the Mishna, and thus the average length of a chapter in the Tosefta is nearly twice that in the Mishna. 18 Epstein briefly mentioned the phenomenon without presenting his findings and without discussing their meaning; see Mevo’ot le-tanna’im, 262. For a full list of the differences in chapter division between Tosefta manuscripts, see Katzoff, “Chapters of the Tosefta,” 23–24. 19 Or, in a case in which the scribe was copying titles from an earlier manuscript, the involvement of the scribe who originally established the titles found in this manuscript. 17

CHAPTER DIVISIONS AND THE TOSEFTA

91

help the copyists express the hierarchy between the levels of division, which, as stated, reflect, in some cases, different reasons for the division. To illustrate, I will present here a few examples, and I will begin at the level of the third division, the division of the tractates into chapters. This division is less central than the other, more foundational divisions described above, both due to the fact that, from the perspective of the internal hierarchy of the text, it reflects a secondary, lower level of division, and also due to the fact that, in the cases in which one topic is divided into multiple chapters, such division is based on technical considerations unrelated to the content. Accordingly, in many cases, especially in relatively early manuscripts, this division is presented in a relatively minor fashion in two common ways. The first is through the words slik pirka, Aramaic for “end of chapter.”20 These words are found inserted into the sequence of the line of text with a small blank space before and after, in the same font type and size as the text itself; see, for example, illustration #1. The second method for dividing chapters is to place the chapter number in the same place and style; see, for example, the words “chapter two” in illustration #2. According to Genette, both of these are cases of “designative titling.” However, the second type of heading, that which cites the chapter number, presents more information to the reader than the heading, which merely notes the end of the previous chapter. This information is related to the hierarchical structure of the composition and reveals the place of the unit within the composition. While the heading does not directly relate to the content of the text, the information does aid the reader in understanding the relationship between the As is common in manuscripts of Palestinian origin. Only at the ends of some of the chapters in the editio princeps do we find a variation which originated in the Babylonian Talmud, hadran alakh, which can roughly be translated as “may we return to you.” See David Rosenthal, “On the Contribution of the Italian Genizah to the Text of the Mishnah, Babylonian Talmud and Palestinian Talmud,” in The Italian Genizah, ed. J. Tabory and A. David (Jerusalem: Orhot Press, 1988), 107; Yaakov Sussmann, “The Ashkenazi Yerushalmi MS — ‘Sefer Yerushalmi’,” Tarbiz 65 (1996), 48. 20

92

BINYAMIN KATZOFF

various parts of the text. Thus it represents a greater involvement of the copyist in interpreting the composition.21 MS CUL T-S E2.134: Tosefta Kelim, Bava Metsia, 8:3 – 9:422

Illustration #1: ‘‫ = ’סליק פרקה‬End of Chapter

Some manuscripts combine these two types of headings; for instance, “end of chapter four.” See, for example, MS CUL T-S NS 329.452 22 Reproduced with the generous permission of the Syndics of Cambridge University Library. According to the assessment of Dr. Edna Engel, the manuscript was copied in the eleventh century, either in Palestine, Egypt or Syria. I want to thank Dr. Engel for providing her assessments regarding the time and place in which the fragments mentioned in this article were copied. See Yaakov Sussmann, Thesaurus of Talmudic Manuscripts (Jerusalem: Yad Ben Zvi, 2012), I, 205. 21

CHAPTER DIVISIONS AND THE TOSEFTA

93

MS CUL Or. 1080.2.1: Tosefta Berachot, 1:6 – 2:1123

Illustration #2: ‘‫ = ’פרק ב‬Chapter 2

Other manuscripts use various means to place an even greater emphasis on the internal division. One way is to write the heading in the same font and size, but to place it in an independent line. In other words, an entire line is left empty save for the heading noting the internal division of the text. Such headings are found in some manuscripts even for the division into chapters. However, this technique is Reproduced with the generous permission of the Syndics of Cambridge University Library. According to the assessment of Dr. Edna Engel, the manuscript was copied in either the eleventh or twelfth century in Oriental script. See Sussmann, Thesaurus of Talmudic Manuscripts, I, 171. 23

94

BINYAMIN KATZOFF

more commonly used for headings of larger divisions, such as the division of orders into tractates, as is found in the following example: MS Ox. Heb. b. 13.4: Tosefta Tvul Yom – Oqtzim24

Illustration #3: ‘‫ = ’סליק פרקא ומסכתא‬End of chapter and tractate ‘‫ = ’מסכת עוקצין‬Tractate Oqtzim

In this case, the copyist placed two subheadings one after the other. The first subheading notes the completion of the last chapter of Tractate Tevul Yom, which means the completion of the tractate as well. To note this division, he writes only “end of chapter and tractate.” These he writes in the same font and size, and in the same line, as the last three words of the chapter; and he leaves the rest of the line, before and after the heading, empty. Here, the heading is more graphically emphasized than those we saw above that also note the end of a chapter, where only the space for an average word was left between the heading and the text. This heading does not note the number of the chapter that was just completed. Using Genette’s terminology, we would call this a “designative title” that in a minimal fashion graphically expresses a small amount of information. The second heading notes the beginning of the next Reproduced with the generous permission of the Bodleian Libraries, University of Oxford. According to the assessment of Dr. Edna Engel, the manuscript was copied in the eleventh century in Oriental script. See Sussmann, Thesaurus of Talmudic Manuscripts, I, 112. 24

CHAPTER DIVISIONS AND THE TOSEFTA

95

tractate and includes the words “Tractate Oqtzim,” which refers to a tractate dealing with the laws of purity and impurity of stems. These words are written on an independent line left empty save for the heading, which creates a slightly stronger emphasis than that placed on the previous heading. This heading provides a name to the text that follows and thus relates to the reader the central theme of the text.25 This is what Genette would call “thematic titling,” and the copyist uses graphic means to emphasize its greater importance. Another way to emphasize a heading is to use a font larger than the text itself and to place it in its own line. Such headings are sometimes found at the lower levels of division, between chapters and tractates. But they are more common in noting the highest level of division, that between orders, as is found in the following example: MS JTS ENA 340.1: Tosefta Chagiga 326

Illustration #4: ‘‫ = ’סליק פרקא‬End of chapter ‘‫ = ’וסדר מועד‬And Order of Festivals

The name of the tractate itself is much older and was created by an early oral transmitter of the text who wished to note the tractate’s central content. The copyist accepted this early name and chose the appropriate graphic means to include it as the tractate’s name. 26 Courtesy of The Library of The Jewish Theological Seminary, New York. According to the assessment of Dr. Edna Engel, the manuscript was copied in the fourteenth century, in Sephardi script. See Sussman, Thesaurus of Talmudic Manuscripts, II, 571. 25

96

BINYAMIN KATZOFF

Here too the copyist presents two subheadings noting internal division, one after the other. The first subheading notes the end of the last chapter of Tractate Chagiga, which is also the end of the entire tractate. To note this division, the copyist writes only slik pirka, Aramaic for “end of chapter.” Since Tractate Chagiga is the last tractate in Order Mo’ed (festivals), this is also the conclusion of the entire Order. The copyist notes this with a second subheading, “And (end of) Order of Mo’ed.” Both subheadings appear in an enlarged font, each in its own line with blank space to both the left and the right. In addition, the copyist left space both above and below. This subheading notes the highest level of internal division, the division between Order Mo’ed and Order Nashim (Women); thus from the perspective of the internal hierarchy of the text, this is the highest level of internal titling. The subheading notes the name of the concluding unit of text and transmits to the reader the main theme of the text just completed. This is a central “thematic title,” and the copyist expresses this in clear graphic form.

THE VARIATIONS IN CHAPTER DIVISIONS IN T OSEFTA

In contrast to the Mishna, in which the division into chapters is mentioned already in the Talmud and whose origin is thus quite early, there is no early evidence as to the division of the Tosefta into chapters. According to Epstein, the division of the Tosefta tractates into chapters is later than its editing.27 Nevertheless, there is no reason to date the division of the Tosefta into chapters to a late period of its transmission; in close to 95 percent of cases, all full and partial manuscripts (geniza and bookbinding fragments) agree on this division. This near-complete agreement indicates that the division into chapters precedes the writing of all extant copies of the Tosefta. The differences in chapter division between the manuscripts in all likelihood stems from late processes in the transmission of the Tosefta’s textual witnesses. This is the phenomenon I focus on below. MANUSCRIPTS AND THEIR SIGNIFICANCE

27

See Epstein, Mevo’ot le-tanna’im, 262.

CHAPTER DIVISIONS AND THE TOSEFTA

97

A late division of one chapter into two or merging of two into one could occur simply in error. However, late chapter divisions or mergings could also result from a conscious decision of the transmitter to deviate from the division he had before him, based on his conclusion that a division different from the one before him would serve the reader better. This sort of intervention could reveal something about how the transmitter understood the text. To determine whether such new divisions or mergings were deliberate, we need to search for a logical reason as to why a transmitter would change the received division. An examination of all chapter divisions in all the direct available textual witnesses of the Tosefta, including the major manuscripts and the fragments from the Cairo Geniza and those found in medieval bookbindings, reveals that in eighty percent of the places where there is chapter-division variance, the variance is between the E textual tradition and all the other textual witnesses. The variation in chapter divisions, then, is not the result of the random variation that always occurs in manuscript copying. Rather, it is related to the nature of the different traditions, and it is on this that I now wish to focus. Of the sixteen instances where chapter divisions vary between the E tradition and other manuscripts, fourteen of them—that is, over eighty-five percent—are cases in which E divides a unit into multiple chapters while the other manuscripts do not. In all but one of these, the additional chapter division matches, or closely approximates, the corresponding chapter division in the Mishna. Relatively simple examples are those in which both works, the Mishna and the Tosefta, treat a series of subjects in the same order; and while the Vienna tradition of the Tosefta includes the whole series in a single chapter, the Mishna and the E tradition of the Tosefta divide the material into several chapters and do so identically. Here is one example from Tosefta Megilla Chapter Two.28

In the following example, I will not go into the actual content; the table itself will illustrate the phenomenon. 28

98

BINYAMIN KATZOFF

Tosefta Megilla Ch. 2: V29 He who reads the scroll of Esther backwards has not fulfilled his obligation and so too for the Hallel and so too for the Prayer …

Mishna Megilla Ch. 2–3 [Ch. 2] He who reads the scroll of Esther backwards has not fulfilled his obligation.

If he read it by heart he has not fulfilled his obligation … If he read it in a foreign language, those who speak the language fulfill their obligation … They fulfill their obligation only if it is written in Assyrian script, in Hebrew, on a scroll, in ink. All are obligated for the reading of the scroll, Priests, Levites, and Israelites … Women, slaves, and minors are exempt, and they do not fulfill the obligation of the community ... The requirement is to read from the beginning to end, the words of R. Meir, R. Yehuda says from ‘There was a certain Jew …’

If he read it by heart or in a foreign language, he has not fulfilled his obligation … It may be read in a foreign language to those who speak the language … He has not fulfilled his obligation unless it is written in Assyrian script, on a scroll and in ink … From what point does a person read the scroll and fulfill his obligation R. Meir says the whole thing, R. Yehuda says from ‘there was a certain Jew …’ All are eligible to read the scroll except for a deaf person, an imbecile, and a minor …

The entire day of waving is prohibited … And the entire night is valid for harvesting the omer and for offering of the sacrificial parts ...

The entire day is valid for reading the scroll and for reciting the Hallel … The entire night is valid for harvesting the omer and for offering the sacrificial parts … This is the general rule: any act whose commandment applies by day is valid the whole day and any act whose commandment applies by night is valid the whole night …

This is the general rule: any act whose commandment applies by day is valid the whole day and any act whose commandment applies by night is valid the whole night … E only: End of chapter If they sold a synagogue they should not buy a street. R. Yehuda said: When is this so? Only when the charity collectors did not make a stipulation with them … Coverings for one scroll may be used for a different scroll but they may not be used for other purposes … R. Eleazar b. R. Zadok pur-

[Ch. 3] Townsfolk who sold a street may buy with its proceeds a synagogue; a synagogue, they may buy an ark … They may sell a synagogue only on the condition that if they want they may take it back …

The same is found in L, in MS Bologna, Archivio di Stato, Fr. Ebr. 8, and in the editio princeps. 29

CHAPTER DIVISIONS AND THE TOSEFTA chased the synagogue of the Alexandrians and did with it as he wanted … Synagogues, they should not behave frivolously inside them … They should not eat or drink in them … R. Yehuda said: When is this so? Only when they are in use but when they are destroyed they should be left alone and grass let to grow in them because of anguish.

99

R. Yehuda says: they may sell it for a courtyard and the buyer may do with it whatever he wants … A synagogue which was destroyed they do not … twist ropes in it and spread out nets in it …If grass grew in it, they do not cut it down because of anguish.

In the Vienna tradition, the first part of the chapter treats the recitation of the Scroll of Esther on the festival of Purim and corresponds to the second chapter of the Mishna. The second part of the same chapter in the Vienna tradition deals with objects consecrated for some sacred purpose and corresponds to the third chapter of the Mishna. In the E tradition of the Tosefta, the two subjects are divided into two chapters by use of the words salik pirka, Aramaic for “end of chapter,” exactly as in the Mishna, as we can see in illustrations 5–6. MS Berlin, Staatsbibliothek, Or. Fol. 20 1220: Tosefta Megila 2–330

MS Wien, Osterreichische Nationalbibliothek, Heb. 20: Tosefta Megila 231

Illustrations #5–6: Only in MS Berlin ‘‫ = ’סל פיר‬End of chapter

This illustration and illustration #7 are provided courtesy of the Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, Orientabteilung. See Sussmann, Thesaurus of Talmudic Manuscripts, II, 653. 31 This illustration and illustration #8 are provided courtesy of the Osterreichische Nationalbibliothek. See Sussmann, Thesaurus of Talmudic Manuscripts, I, 5. 30

100

BINYAMIN KATZOFF

One might object and suggest that this division does not indicate the dependence of the E tradition on the Mishna. The decision to divide the material in this manner could have been made independently by the author of the Mishna and by the transmitter of the E tradition. This is theoretically possible, but unlikely. Many chapters in the Tosefta include several only loosely related subjects and could have been divided thematically. Chapter divisions were added to these units almost exclusively in places where there is an identical division in the Mishna. It is unlikely that this is a coincidence. Rather, it points to the influence the Mishna had on the decision to divide these particular chapters but not others. The next example is more complicated but also more significant. The first chapter of Tosefta Rosh Hashanah and the first two chapters of Mishna Rosh Hashana contain a series of rules concerning the testimony of witnesses who saw the new moon.32 The sequence of these rules is interrupted in both compositions by a narrative explaining the background to a present rule: Tosefta Rosh Hashanah Ch. 1: V33 Mishna Rosh Hashanah Ch. 2 At first, testimony about the new moon was accepted from anyone. Once, the Boethusians hired two witnesses to come and mislead the sages … One of them came gave his testimony and departed. Then the second one came and said: I was coming up at Ma’aleh Adumim and I saw it crouching between two rocks … I heard that the Boethusians were planning to confuse the sages … E only: End of chapter

32According

If [the judges] do not know [the witness], others are sent with him to testify about him.

to rabbinic tradition, there was a period during which the calendar, a hybrid lunar-solar calendar, was determined month by month following the testimony of witnesses who asserted that they saw the new moon. 33 The same is found in L and in the editio princeps.

CHAPTER DIVISIONS AND THE TOSEFTA If [the judges] do not know [the witness], [character] witnesses about him are sent along with him, even on the Sabbath.

101

At first, testimony about the new moon was accepted from anyone. After the sectarians corrupted it was ordained that it would be accepted only from those who were known.

The Tosefta starts with the explanatory narrative: “At first, testimony about the new moon was accepted from anyone. Once, the Boethusians [whom the rabbis considered to be sectarian] hired two witnesses to come and mislead the sages….” The result was the new—and, for the Tosefta, the present—rule: “If [the judges] do not know [the witness], [character] witnesses about him are sent along with him.” First the explanatory narrative, then the accepted rule. In the Mishna, at the beginning of the second chapter, the same rule precedes the explanatory narrative, which is found in a very abbreviated form. Only in the E manuscript is the chapter divided, right between the narrative and the rule, again using the Aramaic phrase slik pirka, as we can see in illustrations 7–8. This division creates a Tosefta chapter which begins with the words “If [the judges] do not know [the witness],” just as the second chapter of the Mishna begins. However, the different order of rule and narrative in the Tosefta results in a harsh separation between the explanatory narrative in chapter one and the rule, which now appears in chapter two, even though they are clearly meant to be read together, as they indeed are in the Mishna and in the Vienna tradition of the Tosefta. Matching the Tosefta text to that of the Mishna was clearly more important to the scribe of the E tradition, than the internal coherence of the Tosefta text. MS Berlin, Staatsbibliothek, Or. Fol. 20 1220: Tosefta Rosh Hashanah 1–2

102

BINYAMIN KATZOFF MS Wien, Osterreichische Nationalbibliothek, Heb. 20: Tosefta Rosh Hashanah 1

Illustrations #7–8: Only in MS Berlin ‘‫ = ’סליק פירקא‬End of chapter

The tendency in the E tradition to add chapter divisions in the Tosefta so that the division of the text better corresponds with the chapter divisions in the Mishna tells us something about the nature of the transmission of the E tradition. In this textual tradition, the Tosefta was transmitted at some point by a learned scribe who did not merely copy mechanically from his source but paid attention to the content as he wrote—and, I would argue, often with an eye to the corresponding Mishna. For him, the Tosefta was not an independent work but one that was to be studied together with the Mishna, the more primary work; so he transferred the internal divisions of the Mishna into the Tosefta. The chapter divisions found in the E tradition, therefore, do not represent the original structure of the Tosefta. To return to Genette’s work, he argued that the most essential of the peritext’s properties is functionality: to ensure that the text is interpreted consistently with purpose of the author, or of whoever introduced the peritext. To place my findings into that context, I maintain that the addition of chapter divisions in the Tosefta that match those in the Mishna serves an interpretative agenda that has significant impact on understanding the contents of the Tosefta. Commentators as early as the Middle Ages—and, in their wake, critical scholars of our times—have deliberated about the relation of the Tosefta to the Mishna. Some asserted that the editor of the Tosefta knew the Mishna, accepted its rulings, and included in his work supplementary material expanding on the positions taken in the Mishna. According to this view, the Tosefta has to be read as relating specifically to rulings in the Mishna and as accepting its authority. Conversely, there have been scholars who claim that at least parts of the Tosefta are independent rulings that do not know

CHAPTER DIVISIONS AND THE TOSEFTA

103

the Mishna or do not address it and put forth halakhic positions contradictory to it.34 This methodological issue will obviously have significant interpretative implications, as I and other scholars have demonstrated concerning many rulings in the Tosefta.35 The additional chapter divisions in the E tradition of the Tosefta, I claim, reflect a reading of the Tosefta as closely bound to the Mishna. Whoever made these additions revealed the way he read the Tosefta and subtly directed the reader to follow his interpretative premises. I mentioned earlier the distinction between thematic titles, which say something about the content of the text and express a demonstrative stance on the part of the author, and designative titles, which express a more restrained stance of the author. In light of this definition, the words slik pirka, “end of chapter,” are certainly a designative title. They say nothing explicit about the content of the text. However, in a most interesting way, they do deliver a subtle message about how the text is to be understood. Whoever added these words did take an active stance towards the text. Here, the peritext influences the understanding of the text.36 The phenomenon I have described and its interpretative implications add insight, I believe, into the nature of the text traditions of the Tosefta, a fundamental issue among scholars of the Tosefta and perhaps of other rabbinic works as well. It contributes to our understanding of how these traditions were shaped and reshaped in various study environments at various times, in different places and contexts in the Middle Ages. And it especially emphaSee Stemberger, Introduction to the Talmud, 152–155; Shamma Friedman, Tosefta Atiqta Pesah Rishon (Ramat Gan: Bar Ilan University Press, 2002); Judith Hauptman, Rereading the Mishnah: A New Approach to Ancient Jewish Texts (Tübingen: Mohr Siebeck, 2005); Brody, Mishnah and Tosefta Studies, 111–154. 35 See, for example, Binyamin Katzoff, “A Story in Three Contexts: The Redaction of a Toseftan Pericope,” AJSR 38 (2014): 109–127. 36While this phenomenon appears in many copies of the Tosefta, it is particularly prevalent in the E tradition. While the Erfurt manuscript was indeed written in eleventh- or twelfth-century Germany, it represents a textual tradition documented from the eastern Mediterranean region of the tenth century. 34

104

BINYAMIN KATZOFF

sizes the importance of attention to the elements accompanying the text, the peritext, for understanding the text and the way the text was understood in the past.

BIBLIOGRAPHY

Albeck, Chanoch. Meḥkarim ba-baraita ve-tosefta ve-yaḥasan la-talmud. Jerusalem: Mosad Harav Kuk, 1969.

Beit Arie, Malachi. Hebrew Codicology, Historical and Comparative Typology of Medieval Hebrew Codices based on the Documentation of the Extant Dated Manuscripts until 1540. Preprint internet version 0.9, April 2018.

Brody, Robert. Mishnah and Tosefta Studies. Jerusalem: Magness, 2014.

———. Mishnah and Tosefta Ketubbot: Text, Exegesis, and Redaction. Jerusalem: Magness, 2015.

Epstein, Jacob N. Mevo’ot le-sifrut ha-tanna’im. Jerusalem: Magness, 1957. ———. Introduction to the Mishnaic Text. Jerusalem: Magness, 1964.

Friedman, Shamma. Tosefta Atiqta Pesah Rishon. Ramat Gan: Bar Ilan University Press, 2002.

Genette, Gerard. Paratexts: Thresholds of Interpretation. Translated by Jane E. Lewin. Cambridge: Cambridge University Press, 1997.

Goldberg, Abraham. “The Tosefta – Companion to the Mishna.” In The Literature of the Sages, edited by Shmual Safrai. Leiden: Brill, 1987. Hauptman, Judith. Rereading the Mishnah: A New Approach to Ancient Jewish Texts. Tübingen: Mohr Siebeck, 2005.

Katzoff, Binyamin. “Did Rav Natronai Gaon Use the Tosefta in His Responsa.” Alei Sefer 20 (2008): 17–27.

———. “The Tannaitic Sources in Halachot Psukot.” Shenaton HaMishpat Ha-Ivri 25 (2008): 199–216.

———. “A Story in Three Contexts: The Redaction of a Toseftan Pericope.” AJSR 38 (2014): 109–127.

CHAPTER DIVISIONS AND THE TOSEFTA

105

———. “Chapters of the Tosefta: The Origin and Significance of the Differences in the Chapter Divisions.” Alei Sefer 28 (2018): 7–30. ———. “Bookbindings of Tosefta from Norcia and their Place in the Textual Tradition.” Jewish Studies Internet Journal 15 (2019): 1–17.

Kenyon, Frederick G. Books and Readers in Ancient Greece and Rome. Oxford: Clarendon Press, 1951. Lejeune, Philippe. Le Pacte Autobiographique. Paris: Seuil, 1975.

Lieberman, Saul. Tosefeth Rishonim. Jerusalem: Mosad Harav Kuk, 1939. ———. Tosefta. New York: JTSA Press, 1955.

Mandel, Paul. “The Tosefta.” In The Cambridge History of Judaism, edited by S. Katz. Cambridge: Cambridge University Press, 2006.

Naeh, Shlomo. “The Structure and the Division of Torat Kohanin (B): Parashot, Perakim, Halakhot.” Tarbiz 69 (2000): 59–104.

Perani, Mauro. “Il più antico frammento della Genizah italiana: la Tosefta di Norcia (ca.1000 e.v.) – rilievi codicologici e paleografici.” In La Genizah italiana, edited by Mauro Perani. Bologna: Il Mulino, 1999.

Rosen Zvi, Ishay. “Introduction to the Mishna.” In The Classic Rabbinic Literature of Eretz Israel: Introduction and Studies, edited by M. Kahana. Jerusalem: Yad Ben Zvi, 2018. Rosenthal, David. “On the Contribution of the Italian Genizah to the Text of the Mishnah, Babylonian Talmud and Palestinian Talmud.” In The Italian Genizah, edited by J. Tabory and A. David. Jerusalem: Orhot Press, 1988. Small, Jocelyn. Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity. London: Routledge, 1997.

Stemberger, Guenter. Introduction to the Talmud and Midrash. Translated by M. Bockmuehl. Edinburgh: T&T Clark, 1996.

106

BINYAMIN KATZOFF

———. “I frammenti della Tosefta di Norcia, e il loro contributo alla studio della tradizione testuale.” In La Genizah italiana, edited by Mauro Perani. Bologna: Il Mulino, 1999. Sussmann, Yaakov. “The Ashkenazi Yerushalmi MS – ‘Sefer Yerushalmi’.” Tarbiz 65 (1996): 37–63.

———. Thesaurus of Talmudic Manuscripts, Introductions & Indices. Jerusalem: Yad Ben Zvi, 2012.

Van Sickle, John. “The Book-Roll and Some Conventions of the Poetic Book.” Arethusa 13 (1980): 5–42.

THE SECOND-HAND SCRIBE: THE INTELLECTUAL ENVIRONMENT OF THE PRODUCTION OF A UNIQUE TOSEFTA FRAGMENT FROM THE LEVANT BINYAMIN KATZOFF∗ A few leaves of the Tosefta copied in the eastern Mediterranean at the end of the eleventh century or in the first half of the twelfth century CE were intensively emended by a second-hand scribe who operated at the same time. This unusual phenomenon, not present in any of the Tosefta’s other MSS, is related to the modus operandi of the second-hand scribe and his intellectual environment, and can reveal his motivation for correcting the text, and his attitude to the composition he had before him and to his role as one of its producers. Thus, the activity of the second-hand scribe, or other scribes who operated in a similar way, has significantly influenced the Tosefta’s textual tradition, and affected our understanding of the text’s transmissional history at the beginning of the second millennium at the Levant.

What is the nature of the intensive emendations found on a few fragments of the Tosefta manuscript copied in the Levant? What do they teach us about the second scribe who so intensely emended I would like to thank Prof. Judith Olszowy-Schlanger for her comments on some of the issues discussed in the article. This research was supported by the Israel Science Foundation (Grant number 1292/18) and by the Beit Shalom Fund, Kyoto, Japan. ∗

107

108

BINYAMIN KATZOFF

these leaves and about the intellectual environment in which he operated? What was his attitude to the composition he had before him and to his role as one of its producers, and how did his activity affect our understanding of the text’s transmission history? These questions arise from a few leaves of the Tosefta, a compilation of Jewish law in Hebrew redacted in Palestine in the third century of this era.1 The leaves were discovered in the Cairo Geniza, a depository for discarded Hebrew manuscripts in a synagogue in old Cairo, and are now scattered throughout several libraries in Europe and North America. These fragments exhibit some unusual phenomena not present in any of the Tosefta’s full extant manuscripts or other partial fragments—phenomena related to the modus operandi of this second scribe and his intellectual environment. Understanding this scribe’s activity is critical in revealing how he, or other scribes operating in a similar way, may have shaped the subsequent transmission of the texts they copied.2 Following a short description of these fragments, I will discuss the nature of the emendations that appear in some of the leaves. Subsequently, I will look into what we can learn from these emendations concerning the identity of the second scribe and the intellectual environment in which he operated. Finally, I will focus on the emender’s attitude to his role and the influence he had on the Tosefta’s textual tradition. The fragments which I will discuss all come from one codex, probably originally containing at least parts of the Tosefta for tractates Demai, Ma’asrot and Yom Tov, and at least parts of the Palestinian Talmud for tractates Yom Tov and Rosh Hashana. Some of the leaves of Tosefta Yom Tov are covered with intense emendations, and I will

Some aspects of the Tosefta’s redaction and transmission are discussed in my article “Chapter Divisions and the Interpretation and Transmission of the Tosefta” in this volume. 2 See Malachi Beit Arie, Hebrew Codicology: Historical and Comparative Typology of Medieval Hebrew Codices based on the Documentation of the Extant Dated Manuscripts until 1540 (preprint internet version 0.11, May 2019), 401. 1

THE SECOND-HAND SCRIBE

109

Illustration #1: Emendations above lines and in the margins. MS CUL T-S F17.34 1v: Tosefta Yom Tov 4:9 – 4:113

focus only on these in the following discussion.4 Both the base text of this manuscript and the emendations by the second scribe were Reproduced with the generous permission of the Syndics of Cambridge University Library. 3

110

BINYAMIN KATZOFF

written in the eastern Mediterranean (Palestine, Egypt or Syria) at the end of the eleventh century or in the first half of the twelfth century CE.5

THE NATURE OF THE EMENDATIONS

Tosefta manuscripts generally do not contain many marginal notes; and in order to describe the nature of the emendations on these leaves of Tosefta Yom Tov, we should first differentiate between different types of marginal notes and their origins. Some notes, which can be called interpretive notes, aim to explain some aspect of the text’s content. For example, they might provide a translation of a word in a foreign language, explain a complex or rarely used legal term, or further discuss the text.6 We can assume that these interpretive notes were added in the margins by a learned reader writing comments for himself for future study, or perhaps for someone else who would be interested in studying the text in the future. Other notes, which can be called text emendations, aim to present a different, perhaps better text than that presented by the first-hand scribe. Different motivations may have led to the The leaves of Tosefta tractate Yom Tov are MS CUL T-S F 17.15 (1v left column ,1r right column), and MS CUL T-S F 17.34 (1r – 1v), and they contain text from Tosefta Yom Tov 3:16 – 4:11 (end of the tractate; Saul Lieberman, Tosefta, (New York: JTSA Press, 1955), II, 297, l. 57 – 304, l. 55); see Yaakov Sussman, Thesaurus of Talmudic Manuscripts, Introductions & Indices (Jerusalem: Yad Ben Zvi, 2012), I, 282–283. On the text of the Tosefta for the other tractates and on the origin and distribution of the different text traditions of the Tosefta in light of these fragments, see my forthcoming article “Confluence of Traditions in the Levant.” 5 I want to thank Dr. Edna Engel for providing her assessment regarding the time and place in which the fragments mentioned in this article were copied. 6 See, for example, MS Ox. Heb. C.18/39 1r (copied around the 10th century in Palestine, Egypt, Syria, or Iraq; Sussman, Thesaurus of Talmudic Manuscripts, I, 89) for Tosefta Bava Batra 11:7 (Lieberman, Tosefta, IV, 168, l. 33); in this passage, the text contains the relatively less common ‫'גזרי‬ '‫דינין‬, and a second-hand scribe added in the margin '‫'פסק דין‬, the more used term for verdict. 4

THE SECOND-HAND SCRIBE

111

emending scribe’s suggestions concerning how to change the base text.7 Some emendations would have aimed to correct what in the view of the emender was a mistaken transcription by the first-hand scribe, who did not copy the text he had before him with adequate precision. In these cases, the second-hand scribe, just like the firsthand scribe, is interested in presenting the text that was before the first-hand scribe, which in his opinion was not transmitted with precision. Such emendations could be performed by a second scribe checking the accuracy of the first scribe, by the first scribe himself reviewing his work, or even by someone studying the text as part of his own program of study.8 At other times, rather than correcting an error, the secondhand scribe is interested in presenting a version of the text that differs from the text on which the first-hand scribe based his transcription. There are different circumstances under which such emendations can occur. One such circumstance occurs when an experienced scholar studies the text and, after concluding that the current text suffers from internal problems, emends the text’s precise wording such that it accords with his understanding of the content and with his own thoughts concerning what the text should be. He may or may not have been concerned with the question of whether this alternative version is historically authentic. Alternatively, these emendations can occur when the second-hand scribe has before him another version of the text, one that differs from the text from which the first-hand scribe transcribed. Again, these emendations could have been executed by a second scribe See Beit Arie, Hebrew Codicology, 374–376. See, for example, MS British Library Add. 27.296 21v (copied in Spain at the end of the 14th century or at the beginning of the 15th century; Sussman, Thesaurus of Talmudic Manuscripts, I, 126) for Tosefta Shabbat 15:2 (Lieberman, Tosefta, II, 68, l. 5); in this passage, the text reads '‫'רשב”ג בן גמ‬. The letters '‫ 'רשב”ג‬are an acronym for “Rabbi Simon Son of Gamliel”; therefore, the words “‫”בן גמ‬, or “Son of Gamliel,” are a mistaken repetition. Rightly, then, a second scribe drew a line through the middle of the words “‫ ”בן גמ‬to indicate that they should be omitted from the text. In this case, there is no need to compare the text to a different version in order to decide that the base text is mistaken. 7 8

112

BINYAMIN KATZOFF

going over the first scribe’s work, or by the first scribe himself comparing his original transcription, copied from one copy of the composition, with the text found in another copy. In these cases, the emendation is not a result of the second-hand scribe’s own conclusion that there is a problem with the text presented by the first-hand scribe, and the new version is not a reflection of the second-hand’s scribe’s personal reshaping of the text. Rather, both the motivation to change the text and the new version are a result of an already existing variant of the text available to the second scribe. In light of these distinctions, I now turn to analyze the notes found on the leaves of the Tosefta Yom Tov. However, before proceeding we need to understand the characteristics of the first-hand text. Saul Lieberman, the scholar who produced the first and, to this day, the only critical edition of the Tosefta, has already noticed that the textual tradition presented in the work of the first scribe of the manuscript for these pages is very close to that found in the London manuscript of the Tosefta.9 This close resemblance is evident from many similar unique readings and, even more clearly, from a few shared secondary readings; there is no reason to elaborate on this phenomenon here. A careful comparison of the version preserved in the original text with that preserved in the emendations demonstrates that these emendations were not based on the second scribe’s opinion that the first scribe had erred in transcribing the text before him, nor on the second scribe’s own opinions concerning how he thought the text should read. Rather, these emendations are consistently based on readings found in the Vienna manuscript tradition.10 In other words, this scribe was using one textual tradition, that found in the Vienna manuscript, to emend a base version proMS British Library Add. 27.296. See Saul Lieberman, Tosefta KiFshutah (New York: JTSA Press, 1955), V, 986; Robert Brody, Mishnah and Tosefta Studies (Jerusalem: Magness, 2014), 75–76. The London manuscript was copied in Spain at the end of the 14th century or at the beginning of the 15th century. 10 MS Osterreichische Nationalbibliothek Heb. 20. The Vienna manuscript was copied in Spain at the end of the 13th century or at the beginning of the 14th century. 9

THE SECOND-HAND SCRIBE

113

duced by the first scribe, which belonged to the textual tradition preserved in the London manuscript. The following are two examples taken from the first two lines of the following illustration: Illustration #2: MS CUL T-S F17.34 1v: Tosefta Yom Tov 4:9

At the beginning of the first line, on the right side in the base text, the first scribe wrote “‫“( ”לא יוליכו לו‬they should not bring it to him”), in accordance with the London manuscript (illustration #3). The second scribe then drew a line in the middle of the word “‫ ”יוליכו‬and wrote above it “‫”יביאו‬, a word with a different root that essentially means the same thing. Clearly the purpose of such an emendation is to bring the text into accordance with the reading of the Vienna manuscript (illustration #4). Illustration #3: MS British Library, Add. 27.296 59v: Tosefta Yom Tov 4:911

Illustrations #4: MS Vienna, Austrian National Library, Cod. Hebr. 20 94v: Tosefta Yom Tov 4:912

This illustration and illustration #5 are provided courtesy of the British Library Board, London. 12 This illustration and illustration #6 are provided courtesy of the Austrian National Library, Vienna. 11

114

BINYAMIN KATZOFF

In the middle of the next line, in the base text, the first scribe wrote “‫בפני‬,” which means “in the presence of” and is in accordance with the London tradition (illustration #5). The second scribe again drew a line through the middle of the word “‫ ”בפני‬and wrote above it “‫”על פי‬, which means “on the instruction of” in accordance with the reading of the Vienna manuscript (illustration #6). Again, these essentially mean the same thing. Illustration #5: MS British Library, Add. 27.296 59v: Tosefta Yom Tov 4:9

Illustration #6: MS Vienna, Austrian National Library, Cod. Hebr. 20 94v: Tosefta Yom Tov 4:9

The phenomenon is clear. The version known to the first scribe is obviously that of the London manuscript, whereas that known to the second scribe is similar to that found in the Vienna manuscript. And here it is important to emphasize again that the second scribe emends the text not only in places where the legal content of his tradition differs from the legal content of the tradition copied by the first scribe. In some of these instances, like those presented above, the same content is simply expressed in two different ways. The second scribe’s motivation is not only to make sure that the user of this manuscript will understand the correct legal ruling; he is also motivated to preserve the linguistic style of the manuscript which is before him, the reading of which he prefers. I want now to examine the behaviour of this second-hand scribe in relation to the different approaches of medieval rabbinic sages towards textual variants in general and towards emending texts specifically. By the end of the first millennium in Babylonia, the main rabbinic center of learning, the Geonim—the leading rabbis who headed the rabbinic academies in Babylonia—paid considerable attention to the existence of different readings of talmudic passages and refrained from any attempt to systematically deter-

THE SECOND-HAND SCRIBE

115

mine which reading was correct.13 A similar situation prevailed in the eleventh and twelfth centuries in the centers of learning in North Africa and Spain: rabbinic scholars were well aware of the existence of variant textual readings but did not relate frequently to the issue of textual emendation.14 In contrast, in Franco-German lands there exists evidence of intense preoccupation with the issue of textual emendation. In particular, we hear rabbis denouncing those who do emend texts. Some scholars ascribe an enactment against emending texts to Rabbenu Gershom, Me’or Hagolah, a leading sage active in Germany at the end of the tenth century. R. Shlomo Yitzchaki, Rashi, the leading French commentator who operated during the second half of the eleventh century, certainly paid considerable attention to the issue of emending talmudic texts. One of the most famous opponents of emending texts was R. Jacob b. Meir, known as Rabbenu Tam, one of the most important of the Tosafists, those Franco-German commentators who followed Rashi. Rabbenu Tam distinguished between two types of emendations. In his opinion, it was acceptable to present an alternative version during a discussion of the text’s original version so that the reader could be aware of both versions. This is, in his opinion, how Rashi acted.15 Presenting See Robert Brody, The Geonim of Babylonia and the Shaping of Medieval Jewish Culture (New Haven: Yale University Press, 2013), 156–161; Robert Brody, “Sifrut Ha-Geonim ve-ha-Tekst ha-Talmudi,” in Meḥqerei Talmud: Qovetz Meḥqarim ba-Talmud u-ba-Teḥumim Govelim, ed. Yaakov Sussman and David Rosenthal (Jerusalem: Magness, 1990), 237–240; Yaakov Spiegel, Chapters in the History of the Jewish Book: Scholars and their Annotations (Ramat Gan: Bar Ilan University Press, 1996), 79–80; Uziel Fuchs, Geonic Talmud: The Attitude of Babylonian Geonim to the Text of the Babylonian Talmud (Jerusalem: Herzog College, 2017), 140–146. 14 See Spiegel, The Jewish Book, 81-95. On the sparse evidence concerning the emendations of already copied manuscripts, see Beit Arie, Hebrew Codicology, 493; Colette Sirat, Hebrew Manuscripts of the Middle Ages (Cambridge: Cambridge University Press, 2002), 217, 286–287. 15 See Israel Ta-Shma, Talmudic Commentary in Europe and North Africa (Jerusalem: Magness, 1999), 46–48; Avraham Grossman, The Early Sages of France: Their Lives, Leadership and Works (Jerusalem: Magness, 1996), 219– 13

116

BINYAMIN KATZOFF

an alternative version while erasing the original one was, according to the Rabbenu Tam, forbidden. The second-hand scribe that emended this fragment operated during this period in the East, in a cultural milieu different from that in which Rabbenu Tam and his disputants were active. There was, as I stated above, awareness in eastern lands as to the existence of talmudic variants. However, comments aimed directly at the phenomenon of textual emendations are rare. This secondhand scribe behaved in exactly the way Rabbenu Tam vehemently denounced. He drew a line through the original version and wrote above it the new reading, such that it would have been difficult for a subsequent reader to even read the original. And even if the reader could make out the original reading, this way of emending the text relates clearly to the reader that what he is seeing are not two legitimate readings that he must evaluate; rather, one version, that produced by the second-hand scribe, is the correct reading, whereas the version produced by the original scribe is in error and is not worthy of attention.

WHO WAS THE SECOND SCRIBE

Determining the identity of a second-hand scribe is frequently difficult or impossible. In only a few instances do we find that the second-hand scribe wrote a colophon describing his work and including details concerning his identity. Marginal emendations and notations were often written under conditions in which the secondhand scribe could not shape the text like he could have when writing in the center of the page, where there is proper space to write the letters and sequences of words. Furthermore, the second-hand scribe often did not write all that much material, leaving us insufficient data by which to describe the style of his writing or to garner any information concerning his identity. This is the case with regard to the second-hand scribe that emended these leaves of Tosefta 220; Spiegel, The Jewish Book, 101–142. Indeed, manuscripts often present an alternative version next to the original one using the words “an alternative version,” or “other manuscripts read,” as Rabbenu Tam demanded. See Beit Arie, Hebrew Codicology, 377.

THE SECOND-HAND SCRIBE

117

Yom Tov. We simply do not have any specific information with regard to his identity, beyond a general description that emerges from paleographic analysis of his writing; this indicates that, like the manuscript’s main scribe, he operated in the East towards the end of the eleventh century or the beginning of the twelfth. Nevertheless, despite the paucity of available biographical information, it is worth examining the nature of his activity in light of the dispute among scholars in recent years concerning the production of rabbinic manuscripts during the medieval period. Malachi Beit Arye, one of the most prominent scholars of the Hebrew book, claimed that unlike the institutional and centralized character of medieval Latin, Greek, and, to some extent, Arabic book production, the medieval Hebrew book was initiated and produced primarily by individuals for private use. In his view, not only did Jewish establishments such as centers of learning, synagogues, or community authorities not instigate or finance the production of Hebrew manuscripts, but there were also no significant collective or commercial multi-scribed ateliers which dealt with producing and marketing books. Rather, book production was performed by individuals, either professional scribes hired by individuals or learned individuals copying for their own use.16 In contrast, other scholars argued for a substantial amount of more systematic production of manuscripts. Judith OlszowySchlanger has shown that there were teams of copyists working together in an organized process of copying, probably with some intention towards commerce. Judah Galinsky has argued that there was a significant amount of cases in which a professional scribe copied works that had a broad readership without having an order in place, in anticipation of finding a buyer interested in purchasing it after the book had been copied.17 See Malachi Beit Arie, “The Individual Nature of Hebrew Book Production and Consumption,” in Manuscrits Hébreux et Arabes: Melanges en L’Honneur de Colette Sirat, ed. Nicholas de Lange and Judith OlszowySchlanger (Turnhout: Brepols, 2014); Beit Arie, Hebrew Codicology, 71–89. 17 See Judah Galinsky, “The Riddle of the Colophoned Manuscripts without the Name of the Patron Who Commissioned the Work,” forthcoming. 16

118

BINYAMIN KATZOFF

In light of this background, we can now ask which model best fits the nature of the work of the second-hand scribe. None of the many notes found between the lines or in the margins can be classified as “interpretive notes.” In other words, these notes were not produced by a learned individual intent on improving the manuscript for future study. All of these notes fit in the category I have called “text emendations,” emendations that do not reflect the personal considerations of the second-hand scribe but rather are attempts to bring the already copied text in line with another textual tradition available to him. These notes are the result of a comparison between two copies of the Tosefta, undertaken with the goal of presenting one of these texts as preferable, even in cases where the differences are only stylistic and have no impact on the text’s content. Such activity attests to the dedication of multiple resources; it provides evidence for the investment of considerable time and the possession of more than one copy of a text, even when the main motivation for such an effort is not related to the text’s content but to the precise documentation of its linguistic style. In principle, it is possible that an extremely dedicated individual learning the text would dedicate his time and energy to such an endeavor. However, I want to suggest that the possession of multiple copies of a composition and the dedication of time towards emending a text’s linguistic style would be more likely to exist in the context of systematic and commercial book production than in a private setting. In such a setting, there is a clear motivation to produce the best copy of the text possible; and there is also a financial justification for such an intense emendation process.18 An emended copy of a text, a copy which presents consistently reliable textual readings, could be worth far more to an advanced student The paleographic characteristics of the small amount of writing left by the second-hand scribe do not allow us to determine whether the emendations were made by the same scribe who wrote the base text or by another scribe who operated roughly in the same time and place. If they were made by a second scribe, then we could characterize the production of this manuscript as indicating a certain type of collaborative work; this would support the possibility that the manuscript was produced by some organized group of scribes and not just by an individual for private use. 18

THE SECOND-HAND SCRIBE

119

who would appreciate the great care taken to produce such a precise text. It is also worth considering the possibility that such emendations were made with the goal that the text would serve as the copy from which subsequent copies of the text were to be made. Fascinating evidence for such activity, albeit in a different time and place, can be found in the words of those who printed the first copy of the Talmud Yerushalmi in Venice in 1523 and who used as their base text MS Leiden Cod. Or. 4720, Scaliger 3. According to these printers, before printing, the manuscript was emended based on three other copies of the Yerushalmi available to them. Indeed, we can see that there are many emendations on this manuscript, likely made by more than one emender.19 It is thus possible that the emendation of this copy of Tosefta Yom Tov was intended to prepare it to be the source of future copies of the text made by this emender himself or by the person who initiated the emendations.

THE INTELLECTUAL ENVIRONMENT OF THE SECOND

These findings aid us in considering the intellectual environment in which the second scribe was active. The first issue we can address is the textual traditions present in the scholarly environment in which he operated. The second scribe, as stated, operated in the same time and place as the fragment’s first scribe. The fact that the base text and the text from which the emendations were taken were present in the same time and place is the earliest evidence of the co-existence of what were already, in the second scribe’s eyes, two distinct traditions of the Tosefta. Prior to the analysis of this fragment, we knew of the existence of multiple copies of the Tosefta in the same time and place from incidental discussions of various issues in the writings of a number of medieval sages both from the east and from Europe. However, the variants between these tradiSCRIBE

See Talmud Yerushalmi, Venice 1523, Nidah, end of chapter 4, 51:2; Yaakov Sussmann, “Introduction,” in Talmud Yerushalmi According to Ms. Or. 4720 (Scal. 3) of the Leiden University Library, with Restorations and Corrections (Jerusalem: Magness, 1999), 22–26. 19

120

BINYAMIN KATZOFF

tions are only sporadic and isolated, and they usually present minor stylistic variations.20 Additional evidence regarding multiple copies of Tosefta manuscripts is found in a letter in the Geniza that, in the estimation of Dr. Edna Engel, was written between the years 1250– 1350 in Egypt or Iraq: And when I came to Egypt, the sages who were in Egypt sought the Tosefta that I had and found it to be like the Toseftot that they had; and I was doubtful about this matter until I returned to Damascus. And I found there about six Toseftot that were brought from Iraq, and I examined all of them; and they are all copies of the same version, which is the same version as my copy.21

Here, too, it seems that the variants between the multiple copies were, at the most, pretty minor. This geniza fragment is the only solid evidence for the existence of two distinct traditions in the same time and place as early as the 11th or 12th century; it is evidence of the richness of the literary environment in which the second scribe operated.22 Another noteworthy aspect of the second scribe's work is the very fact that he attempted to systematically correct one textual tradition according to another. This phenomenon has been documented on a few occasions in other rabbinic compositions;23 but as See Binyamin Katzoff, “Was There a Separate Transmission of Tosefta Tohorot? Evidence from Rishonim and from Fragments of the Cairo Geniza and ‘European Geniza’,” Hebrew Union College Annual 88 (2017), 41–52. 21 MS New York, Jewish Theological Seminary, Schechter 27. 22 On the significance of this phenomenon for our understanding of the origin and distribution of the textual traditions known to us from the end of the medieval period in Europe, see my forthcoming article “Confluence of Traditions in the Levant.” 23 See Beit Arie, Hebrew Codicology, 508–511. On correcting manuscripts in light of other versions in some circles of Muslim copiers, see Johannes Pedersen, The Arabic Book, trans. Geoffrey French (Princeton: Princeton University Press, 1984), 46–47; Ella Landau-Tasseron, “Had20

THE SECOND-HAND SCRIBE

121

relates to the Tosefta, it is found only in this fragment. In other words, many copies of the Tosefta contain textual emendations here and there; and in some cases, these emendations were made in accordance with another version of the text. But a systematic and continuous project to correct a base text according to a different tradition is found in the Tosefta only in these leaves. It is possible that most Tosefta manuscripts were not intensely emended based on different textual traditions due to medieval rabbinic scholars’ belief that the precise text of many rabbinic compositions was not completely fixed down to the level of exact words. Rather, medieval sages accepted the fact that different manuscripts of a work transmitted identical content in nearly similar language. To use a phrase uttered by several medieval sages—for example, Rav Hayya Gaon, head of the Pumbeditha academy in Baghdad at the beginning of the 11th century—these were cases of “two different wordings with the same meaning.”24 This may be a result of their understanding that rabbinic literature was originally transmitted orally and that this means of transmission left the text somewhat susceptible to the language of those transmitting it. Furthermore, in certain learning environments, scholars and copyists would have believed that it was permissible, and perhaps even an obligation, to continue to improve the language of the composition during its transmission.25 The second scribe in this leaf from the Tosefta, who went to the trouble of emending the text even in places where the difference between the two traditions was only stylistic, seems not to have shared the opinion of many other copyists concerning the nature of the divergent textual traditions and their relationship one to the other. Rather, this scribe perceived the differences between the manuscript traditions to be meaningful and worthy of note. ith,” in Islam, History, Religion, Culture, ed. Meir M. Bar-Asher (Jerusalem: Magness, 2017), 199. 24 See Lewin, Otzar Ha-Geonim, I, part B (Interpretations), 53; Fuchs, Geonic Talmud, 126-132. 25 See Rosenthal, “The History of the Text;” Friedman, “The Origin of Textual Variants.”

122

BINYAMIN KATZOFF

The last issue related to these leaves of the Tosefta that I wish to discuss is the implications of the creation of the new textual readings by the second-hand scribe for our ability to use stemmatic analysis of the various manuscripts of the Tosefta to determine the Tosefta’s earliest or even original form. In this case, we can easily see that the second copyist corrected one tradition according to a different tradition. But such work cannot always be detected. If other copyists in this region emended one textual tradition based on another, then we should expect a considerable amount of contamination within other manuscripts as well. For example, if a copyist used this manuscript from the Geniza as the base text from which he was to copy another manuscript of the Tosefta—as I suggested earlier may have been the very goal of these emendations—and if he accepted the second-hand scribe's corrections, the result would be a new hybrid version, one that would pose an immense difficulty for modern scholars trying to trace the history of the text.26 I am not claiming that conventional stemmatic analysis of the Tosefta’s textual traditions is useless. It may provide weighty considerations in favor of certain readings. However, the phenomenon that I have noted here indicates the caution and moderation required when employing this type of analysis to the manuscripts of the Tosefta. In conclusion, even a small number of leaves from one manuscript of the Tosefta can teach us a great deal concerning the modus operandi of their second scribe. This scribe was motivated not only by a desire to make sure that the user of this manuscript would understand the correct legal ruling; he was also interested in preserving the linguistic style of the manuscript that was before him, whose reading he prefers. The way he marked the text makes it clear to the reader that the two readings are not equivalent in terms of their legitimacy. Rather, one reading is correct—that which he adds as an emendation—whereas the reading copied by the first scribe is errant and not worth reading at all. I have suggested that the possession of multiple copies of the Tosefta and the dedication to correct what are only stylistic differences is best understood in the context of a systematic creation of Concerning the possibility that the London manuscript was influenced by other textual traditions see Brody, Mishnah and Tosefta Studies, 70–73. 26

THE SECOND-HAND SCRIBE

123

manuscripts for commercial purposes. In such a setting, there is a clear motivation to create the best possible copy of the text; and there is also financial justification for the time needed to engage in such an intensive project of emendation. We should also consider the possibility that this manuscript was emended in order to prepare it to serve as a copy from which subsequent copies of the Tosefta could be made. This geniza fragment is also the only solid evidence as to the existence of two distinct traditions of the Tosefta in the same time and place as early as the 11th or 12th century; it is a testimony to the richness of the literary environment in which the second scribe operated. The attempt to emend one branch of the text according to another, uncommon in the copying of Tosefta in this period, demonstrates that the second scribe did not share the tolerance for different traditions with identical content that was common among some medieval sages. Instead, this scribe perceived even stylistic differences to be meaningful. This phenomenon also points to the possibility that there may be a considerable amount of contamination within manuscripts of the Tosefta, at least those copied in a similar time and place; and it raises questions as to the reliability of stemmatic analysis of the Tosefta.

BIBLIOGRAPHY

Beit Arie, Malachi. “The Individual Nature of Hebrew Book Production and Consumption.” In Manuscrits Hébreux et Arabes: Melanges en L’Honneur de Colette Sirat, edited by Nicholas de Lange et Judith Olszowy-Schlanger. Turnhout: Brepols, 2014.

Beit Arie, Malachi. Hebrew Codicology: Historical and Comparative Typology of Medieval Hebrew Codices based on the Documentation of the Extant Dated Manuscripts until 1540. Preprint internet version 0.11, May 2019.

Brody, Robert. “Sifrut Ha-Geonim ve-ha-Tekst ha-Talmudi.” In Meḥqerei Talmud: Qovetz Meḥqarim ba-Talmud u-ba-Teḥumim Govelim, edited by Yaakov Sussman and David Rosenthal. Jerusalem: Magness, 1990,

Brody, Robert. The Geonim of Babylonia and the Shaping of Medieval Jewish Culture. New Haven: Yale University Press, 2013. Brody, Robert. Mishnah and Tosefta Studies. Jerusalem: Magness, 2014.

124

BINYAMIN KATZOFF

Fuchs, Uziel. Geonic Talmud: The Attitude of Babylonian Geonim to the Text of the Babylonian Talmud. Jerusalem: Herzog College, 2017. Galinsky, Judah. “The Riddle of the Colophoned Manuscripts without the Name of the Patron Who Commissioned the Work.” Forthcoming.

Grossman, Avraham. The Early Sages of France: Their Lives, Leadership and Works. Jerusalem: Magness, 1996. Katzoff, Binyamin. “Was There a Separate Transmission of Tosefta Tohorot? Evidence from Rishonim and from Fragments of the Cairo Geniza and ‘European Geniza’.” Hebrew Union College Annual 88 (2017): 29–54. ———. “Confluence of Traditions in the Levant.” Forthcoming

Landau-Tasseron, Ella. “Hadith.” In Islam, History, Religion, Culture, edited by Meir M. Bar-Asher. Jerusalem: Magness, 2017.

Lewin, Binyamin, Otzar Ha-Geonim, Thesaurus of the Gaonic Responsa and Commentaries, Haifa: n/a, 1928 Lieberman, Saul. Tosefta. New York: JTSA Press, 1955.

———. Tosefta Ki-Fshutah. New York: JTSA Press, 1955.

Rosenthal, Eliezer Shimshon. “The History of the Text and Problems of Redaction in the Study of the Babylonian Talmud.” Tarbiz 57 (1988): 1–36.

Sirat, Colette. Hebrew Manuscripts of the Middle Ages. Cambridge: Cambridge University Press, 2002. Spiegel, Yaakov. Chapters in the History of the Jewish Book: Scholars and their Annotations. Ramat Gan: Bar Ilan University Press, 1996.

Sussmann, Yaakov. Thesaurus of Talmudic Manuscripts, Introductions & Indices. Jerusalem: Yad Ben Zvi, 2012.

———. “Introduction.” In Talmud Yerushalmi According to Ms. Or. 4720 (Scal. 3) of the Leiden University Library, With Restorations and Corrections. Jerusalem: The Academy of the Hebrew Language, 2001. Ta-Shma, Israel. Talmudic Commentary in Europe and North Africa. Jerusalem: Magness, 1999.

PERITEXTUAL ENCODING FOR THE M ETATRON / YAHOEL THEME IN THE KABBALISTIC S EFER H A-OT , OR “BOOK OF THE SIGN,” BY R. ABRAHAM ABULAFIA (1240–1292) ARYEH M. KRAWCZYK The purpose of this paper is to present the peritextual encoding applied by the thirteenth-century kabbalist R. Abraham Abulafia in his prophetic treatise entitled Sefer ha-Ot, or “Book of the Sign.” Each of the extant manuscripts of this work provide similar examples of such peritext values. Significantly, the majority of them were deliberately designed by the author to convey the esoteric notion of the angel Metatron, known in this Kabbalistic source by his alternative name, Yahoel. My aim is to present this sophisticated structure parallel to the plain narrative. Illustrative material to corroborate the analyses has been applied from one of the most representative manuscripts of Sefer haOt, Munich 409, courtesy of Bayern City Library.

Kabbalistic manuscripts are often filled with peritextual values, diagrams, charts and sketches. Among the most well-known Kabbalists, R. Abraham Abulafia (1240–1292) was a dominant figure within the school of prophetic-ecstatic Kabbalah of the thirteenth century who notoriously used peritextual encoding connected to

125

126

ARYEH M. KRAWCZYK

numerical meanings (gematrias) to convey mystical messages.1 One of his books, Sefer ha-Ot, or “Book of the Sign,” is special exactly because of the permanent usage of peritextual hints pointing at a parallel narrative. The purpose of this paper is to present some of these cases from “Book of the Sign,” especially those related to the semantic field connected with the notion of the angel Metatron, known also as Yahoel. This mystical figure plays a singular role in the general system of R. Abulafia’s thought and is mentioned explicitly or implicitly throughout the text of Sefer ha-Ot. The Kabbalistic work Sefer ha-Ot survives in eight manuscripts.2 It is not a typical work of R. Abulafia; most of the content is related to apocalyptic visions regarding the redemption of the Jewish people from the bondage of Edom, represented by the pope and Rome. The main figure is Zecharyahu, R. Abulafia’s alter ego who experiences several mystical encounters with the angel Yahoel, or Metatron. Apart from the plain narrative, this theme is conveyed in several concealed modes of coding. Crucial for our discussion here is the fact that the text is filled with paratextual/peritextual values designed to encode specific esoteric meanings—mostly numerical (gematrical). The author’s clear intention was that the majority of such hints should relate to the name YHVH (gematrical value 26) and its multiples: The biography of this Kabbalist has been presented in a plethora of publications. See, for example, Moshe Idel, The Mystical Experience in Abraham Abulafia (Albany: State University of New York Press, 1985). More recently, see Aryeh Krawczyk, Krew, orchidea, atrament : endofazja i heautoskopia w “Sefer ha-Ot”, “Księdze Znaku” R. Abrahama Abulafii (12401292) : edycja krytyczna, tłumaczenie tekstu (Warsaw: Żydowski Instytut Historyczny im. Emanuela Ringelbluma, 2018). 2 Presented illustrations are taken from Ms. Munich 409, Bayern City Library, provided here with permission of the owner. A detailed overview of the manuscripts and a summary of the book has been discussed in one of my previous publications: A. Krawczyk, “Sefer ha-Ot: Preliminary Insights on a Critical Edition,” Jewish History Quarterly 2 (2015) (254): 281–314. Some of the data from that paper have been incorporated here in a revised version. 1

PERITEXTUAL ENCODING IN THE SEFER HA-OT

127

1) 52 = 2 x 26: the value for Yahoel¸ an angel identified with Metatron 2) 78 = 3 x 26: value of Ana YHVH and/or Ani VHV – names read during Hallel and Hoshana Rabah 3) 91 = 3.5 x 26 = 7 x 13 = value for yichud of the names YHVH (26) and Adonai (65), which combined gives the name YAHDVNHY (91).

Paratextual values3 found in the manuscripts are designed to lead towards polysemantic gematrical patterns inherent in the text, where the key matrix is comprised of the numbers 26, 52, 78 and 91. We may propose, therefore, that what unifies Sefer ha-Ot throughout all of its complex strata consequently builds up to the coding chain for Yahoel and YAHDVNHY, a semantic field that is submerged in the literal text and plotted even deeper on a gematrical level.4 The peritextual values in Sefer ha-Ot are: - Dots indicating that the gematrical meaning is to precede the literal meaning of the verse - Dots above letters indicating a special meaning assigned to them - Notes about kri u-ktiv - Enlarged/diminished letters - Marginal notes In the following three sections of this paper, I would like to present the paratextual values related to the gematria of 26/52, 78 and 91, respectively. These are three representative, though not exhaustive, examples regarding the leading Metatron/Yahoel theme.

Followed by D. Abrams’s insights on textological methodology to be applied to Hebrew manuscript analysis. See Daniel Abrams, “Kabbalistic Paratext,” Kabbalah 26 (2012): pp. 7–24, and the monumental Daniel Abrams, Kabbalistic Manuscripts and Textual Theory: Methodologies of Textual Scholarship and Editorial Practice in the Study of Jewish Mysticism (Jerusalem: Magnes Press, 2010). 4 These paratextual values are best visible in the Hebrew edition, from which excerpts are provided here. 3

128

ARYEH M. KRAWCZYK

GEMATRIA OF 26 AND 52

Meaz or Mehaz ? How is Yahoel hidden in the first verse of Sefer ha-Ot?

PARATEXT CASE: a word annotated with a dot above a he that was written in place of an alef. This paratext conveys the following encoded messages: - Encoding of gematria 52 for Yahoel, Eliyahu, Ben, YHVH YHVH, etc. - Allusion to the biblical verse concerning Moshe’s staff, a famous kri and ktiv case in the Torah text. In the passage, m-h-z is to be read as m-z-h. As per the Masoretic note, this is to be pronounced as ma-ze (“what is it”): “YHVH spoke to him: what is it in your hand? He said: a staff.”5 Quote from Sefer ha-Ot “I have the speech of YHVH7 from then—from the day I got the knowledge of His Name until today. Still I will make myseparated6

Ex. 4:2. I follow the view of T. Sikora, who proposes translating kadosh and its derivatives not using theologically biased terms like “saint, sanctified” but using more precise meanings of “separation and discernment, uniqueness”; this is well supported by logic provided in Rashi’s commentary on Lev. 19:1–2: kedoshim tihyu, “you shall be separated.” Methodologically, such an application is also corroborated in Oskar Goldberg, Die Wirklichkeit der Hebräer (Berlin: David Verlag, 1925; Wiesbaden: Harrassowitz, 2005); see this source for the most well-known scholarly approach. 7 Et pi YHVH is a fixed Tanachic phrase. See: Num. 14:41; 20:24; 22:18; 22:28; 24:13, Deut. 1:26; 1:43; 9:23, 1 Sam. 12:14; 12:15: 15:24, I Krl. 1:13. Although it would literally mean “the mouth of YHVH”. we follow the Onkelos indication to render it like memra da-Hashem and gzerat memra da-Hashem. Thus, there are three options for translation: I have separated the speech of YHVH I have separated the mouth of YHVH With my lips I have separated YHVH I would like to thank Ewa Gordon for hinting at the aforementioned biblical data. 5 6

PERITEXTUAL ENCODING IN THE SEFER HA-OT

129

self unique in His Name and through His uniqueness I will revive in truth.”

f. 11a

The first stanza is immediately striking for its extraordinary spelling of the term mehaz, which is supposed to be read as meaz “from now.” H. Hames states: “Given the importance of letter notation for Abulafia, and taking into account the possibility of scribal error, the fact that this same spelling appears in the first two stanzas of the work, would seem to suggest that it is meaningful. Given Abulafia’s preponderance8 for playing with letters, it is possible that there is a double entendre here (or perhaps a qri’ and ktiv) – read ‘then’, but understand ‘from heh zayin’ referring to year 5007 (1247) or when permutated ‘from the year zeh’ in other words from the twelfth year (zayin = 7, heh = 5). This could possibly be a reference to his childhood implying that at the age of 12, in 1252, he began studying with his father, or perhaps refer to the visions in late 1270 when he was first awakened to the potential of the divine names, really closer to eleven rather than twelve years after his first awakening in 1260.” 9 I assume H. Hames meant here “propensity.” Harvey Hames, “Three in One or One That is Three: On the Dating of Abraham Abulafia’s Sefer ha-Ot,” Revue des études juives 165, no. 1–2 (2006), 182. 8 9

130

ARYEH M. KRAWCZYK

I agree with Hames that the case under discussion is an obvious instance of kri and ktiv; a claim of scribal error is unsubstantiated, as all manuscripts follow this intentional writing. However, we disagree on the proposed meanings, as such a spelling is most certainly placed here to hint at a gematria of 52 (=Yahoel, 2 x 26, Ben, Eliyahu, etc.) and, more importantly, to yet another classic kri and ktiv stemming from the Torah, namely m-h-z. Per the Masoretic note, this is to be read as ma-ze (“what is it”): “YHVH spoke to him: what is it in your hand? He said: a staff.”10 In light of this observation, the most plausible, appropriate strategies for decoding are through the juxtaposition of parallel Torah and Sefer ha-Ot meanings and through tracing the modelling of Zecharyahu on the figure of Moses. When these strategies are applied, decoding of the passage includes the following: - In the Torah, Moses speaks with the Name (26; anagrammatic relationship between Moshe and Hashem: ‫)משה השם‬. In Sefer ha-Ot, Zecharyahu speaks with Yahoel (52 = 2 x 26) in his prophetic vision. - Moses embarks on his mission to free the people of Israel from Egypt and the Pharaoh. In Sefer ha-Ot, Zecharyahu sets off on a similar enterprise to free the Jews from the bondage of Rome/Edom and the Pope. - The gematria for mateh (“staff”) is 54 (52+2); in line with R. Abulafia’s methods, one could perform experimental substitution of this word and instead insert be-Yahoel11 (“in/through/with the Ex. 4:2. Obviously, one could insert here Ben, Eliyahu, Libecha etc., but this is not the place to build up some elaborated exegesis. Rather, I am following a plausible method of revealing the hidden data through experiential understanding, as likely intended by the author of Sefer ha-Ot. However, other variants are likewise justified. Also, notions of Yahoel and “the staff” resembling the shape of the sign of his forehead are of importance. This interpretation is corroborated by other sources, too. For example, see chapter 5 in Agata Paluch, Megalleh ʿAmuqot: The EnochMeṭaṭron Tradition in the Kabbalah of Nathan Neṭa Shapira of Kraków (1585– 1633) (Los Angeles: Cherub Press, 2014), esp. pp. 179–188 and the literature cited therein (mainly M. Idel, as noted on p. 33 and on) in which the 10 11

PERITEXTUAL ENCODING IN THE SEFER HA-OT

131

assistance of Yahoel”) or be-YHVH YHVH (“in/through/with the assistance of YHVH YHVH”). Let us apply this substitution and translate Ex. 4:2 in a Kabbalistic vein, as follows: “YHVH (26) spoke to him: Yahoel (52) in your might.12 He said: through Yahoel (52)” or “YHVH (26) spoke to him: YHVH YHVH (52) in your might. He said: through YHVH YHVH (52).”13 - As Moses is given the Torah on Sinai, so, too, Zecharyahu receives the call to write the “Book of the Sign,” which he receives by inspiration in, among other places, Messina: mi-Sinai.14 - The notion of “staff/sceptre” is central in Sefer ha-Ot, as the sign itself is described as “resembling the staff.” However, the terms used throughout the book are makel and sharbit, as if to stress even more within the Rambam’s mode of “revelation through concealment”15 that mateh is hinted at by a peritextual and gematrical relationship of Metatron to Moses’s staff (or rod) is discussed (pp. 179– 182, n. 86). R. Natan Nata Spira (or Shapiro) utilizes notions stemming from the chasidei ashkenaz tradition that is the common source of R. Abulafia’s ideas. However, an explicit connection between the gematria of 54 (mateh) and 2+52 (be-Yahoel) within the context of Ex. 4:2 is not indicated in Paluch’s study. 12 A translation of be-yadeicha, lit. “in your hand,” as “in your might” is corroborated by the Rambam’s proposals, as presented in Moreh Nevuchim. 13 Let us keep in mind that in the preceding verse, Ex. 3:14, Moses encounters for the first time Ehyeh asher Ehyeh, the Name that is itself dual in its inherent dynamics. See also Paluch, Megalleh ʿAmuqot, 85–88, where the reference to the gematria of 52 in the context of the early chasidei ashkenaz tradition of Commentary on the 70 Names of Metatron is discussed. Again, as this source must have impacted R. Abulafia greatly, it further corroborates the present analysis pertaining specifically to Sefer ha-Ot. 14 Themes pointing directly to an analogy for R. Abulafia-Moses with reference to the very act of prophesying have been discussed in Aryeh Krawczyk, “Zdejmowanie zasłony języka. Tora prorocka Mojżesza i prorokowanie Torą Abulafii,” Hermaion 2 (2013): 28–38. 15 Adopted aptly by R. Abulafia in many places in his theoretical works: for example, Sitrei Torah. See Elliot Wolfson, Abraham Abulafia—

132

ARYEH M. KRAWCZYK

level of coding and points to an even more important figure: Metatron.16

GEMATRIA OF 78

Ani va-H u or Ani Vaho? The secret of H allel on Sukkot and

the unification of Name 72 with YHVH PARATEXT CASE: enlarged font in the word ko (=25), immediately followed by Ani VHV. Both expressions hint at Metatron/Yahoel/Naar. Quote from Sefer ha-Ot “Thus spoke about this17 YHVH, Elohei Israel: Do not fear the enemy, because ‘I, VHV’ we will fight him to save you from his hand.” Kabbalist and Prophet: Hermeneutics, Theosophy and Theurgy (Los Angeles: Cherub Press, 2000), 23 and 83–86. 16 See also Robert Sagerman, The Serpent Kills or the Serpent Gives Life: The Kabbalist Abraham Abulafia’s Response to Christianity (Leiden: Brill, 2011), 212–216 and especially p. 214, where exactly the same passage from Ex. 4:3 about “the staff” is discussed by R. Abulafia in his Sitrei Torah, p. 33: “And it was said (Ex. 4:3), ‘Put your hand out and seize it by the tail.’ And he put out his hand, and caught it, and it became a rod in his hand (‫)בכפו‬. [This should be read] ‘In twenty-six (‫)בכ”ף ו”ו‬,’ because with six in his hand (‫ )בו”ו בכפ”ו‬are, ‘In his hand are six (‫)בכ”ף ו”ו‬.’” Note that “six” is representative of Metatron in many Kabbalistic and midrashic traditions. 17 Note the enlarged font in the word ko (=25). Following the im hakollel rule, it has a gematria of 26. See also R. Abulafia’s Sefer ha-Yashar, where ko is also presented in a special manner, in a similar, revelatory context. Yet again, this is corroborated by the common strata of chasidei ashkenaz, as interpreted by R. Natan Spira: see Paluch, Megalleh ʿAmuqot, 182, in the same vein. From our perspective, the evident source in the Torah to connect all these ideas (ko, Metatron, dual-subjectivity, teleological aspect of YHVH, etc.) is Gen. 22:5 from Akedah, read during each daily shachrit: Vaani ve-hanaar nelcha ad ko (“I and the youth will go towards ko [there].” Yet another epithet for Metatron/Yahoel is Naar, “The Youth.”

PERITEXTUAL ENCODING IN THE SEFER HA-OT

133

f. 23a and 23b

The word ko, following the im ha-kollel rule, has a gematria of 26. Ani and Vaho are the 37th and 1st triplets of the Name 72,18 respectively, which together total in gematria to 61 + 17 = 78. According to Rashi, the gematrical equivalence of these triplets to the expression Ana YHVH (52 + 26) enabled the substitution of the latter in the ritual of Hoshanot during Hallel on Sukkot.19 Thus, instead of standard version of Hallel, where the version Ana YHVH hoshiya na is recited, one might say during the Hallel on Sukkot: Ani Vaho hoshiya na. “Me-VHV” or Ani Vaho clearly indicates the unification of the subject, chanting the Hoshanot together with the Name Name 72 is derived from three verses of Ex. 14:19–21, each of them consisting of 72 letters. Combining each of 216 (72 x 3) letters according to the method described by Rashi brings forth the mystical Name 72. 19 See: Machzor for Sukot, Artscroll, New York, 2013 pp. 118–119 and p. 364. See also: TB, Mishnah Sukkah 4:5. 18

134

ARYEH M. KRAWCZYK

72. Furthermore, this connection repeats itself on a higher level: the structure of gematria for Ana YHVH (52 + 26) shows that Ana is a substitute for Yahoel (=52) in his unification relationship with YHVH (=26). Also, the order of the triplet invoked is not deprived of significant meaning. Ani = 37, which in mispar katan gematrical method refers to chochmah (“wisdom”), the second sefirah; and Vaho = 1, which refers both to the mystic/Kabbalist that gains this particular wisdom related to prophecy and to the one that provides this wisdom, that is, Metatron/Yahoel. To recapitulate, this idea is best presented in the chart below:

GEMATRIA OF 91 Unification of Written (26) and Spoken (65) Name PARATEXT CASE: dots over letters forming the name Adonai, interwoven into the name YHVH. Quote from Sefer ha-Ot “For20 your salvation we hope, YAHDVNHY” Note that li-yeshuatchah is split: a double yud appears and a final he is added with a mark of nekudah meal ot. Such intentional script points also to the Name YaH. 20

PERITEXTUAL ENCODING IN THE SEFER HA-OT

135

f. 31b

This gematria of 91 is also hinted at in a paratextual, special division of the word va-tetze, where tze gives the value of 91; this is one among several other instances of hidden encoding of this value in Sefer ha-Ot.

Additionally, paratextual value is visible in how the word liyeshuatchah is split: a double yud appears and a final he is added with a mark of nekudah meal ot. Such intentional script points also to the Name YaH. YAHDVNHY is the unification of names Adonai and YHVH, the former being the kri of the latter’s ktiv. It is present nowadays in almost every Sephardic siddur based on kavvanot and yichudim.21 Its gematria, 91, is the ratio of 7 x 13 or 3.5 x 26 and is identical with SaEl (“El saves”), Amen (the conclusion for each of the brachot), and the rashei tevot of the key verse of Ashrei (Ps. 145:16), relating to the letters Peh (“mouth”), Alef (initial for Adonai), Yud (initial for YHVH): P oteach E t Yadeicha. In order to relate this data to the endophatic theme permeating Sefer ha-Ot, one should recall the opening words of the ouvre: et pi “the mouth.” Association between the gematria 91 for the Name YAHDVNHY and Peh (“mouth”) is corroborated by the significant fragment from the Zohar. Though we obviously cannot assume any direct and mutual influence of the 21

For example, Siddur Avodat ha-Shem, Yerushalaim 5758.

136

ARYEH M. KRAWCZYK

Zohar (or the traditions preceding it) on R. Abulafia’s work, nevertheless it is worth to quoting here at length an interesting elaboration on the meaning of this very name: “The sword of the Holy One Blessed Be He - the sword of judgment in the spiritual realms - is hinted at in the very name YHVH. The Yud represents the handle of the sword and also the sefirot of Keter and Malchut. The Vav is the body of the sword, the six directions of the world, Tiferet, combining judgment with mercy. The two letters He symbolize the two sharp cutting edges (pipiyot),22 of the sword, two mouths (peh and peh) – the higher mouth, Malchut, and the mouth of the judge in this world. It is written regarding these two mouths: ‘Pursue justice, true justice, in order that you may live, and inherit the land which the Lord your G-d gives you’23. The repetition of the word ‘justice’ in the text refers to the two decisions involved in justice: the decision of the court in the spiritual realms and the concurrent decision of the court in the physical realm. From this we see that there is judgment involved in everything, no matter how small the matter, as we have learnt ‘No-one bruises their finger in this world unless it was so decreed against him in heaven’.24 The sheath of the sword of justice is the name Adonai, representing Malchut, when combined with the merciful holy name YHVH. The unity of these is the meditative state of union with the infinite symbolized by the names when spelled together: YAHDVNHY.”25 Elliot Wolfson aptly highlights the relationship of the Name YAHDVNHY discussed in the Zohar with the contemplative teachings of R. Abulafia, where the emphasis on auditory and verbal experiences relates his actions directly to the realm of Names

See the analogical context of pipiyot in Ps. 149:6, which is read on daily and Shabbat shachrit in pesukei dezimra section. 23 Deut. 16:20. 24 TB Chulin 7b. 25Zohar, Shoftim 274b. Based on translation by R. Simcha Shmuel Treister, http://www.chabad.org/kabbalah/article_cdo/aid/380122/jewish/The-Sword-of-Judgment.htm [accessed January 13, 2019]. 22

PERITEXTUAL ENCODING IN THE SEFER HA-OT

137

and their recombinations.26 Referring to the Name, the text comments:27 “When the lower splendor, Adonai [i.e. the tenth gradation or Shechinah] joins with the supernal splendor, YHVH [i.e., the sixth gradation or Tiferet] the hidden name [i.e. YAHDVNHY, the combination of the two names]28 is produced, which the true prophets know and [by means of which] they [visually] contemplate the supernal splendor.” Additionally, linkage of YHVH to the sixth sefirah points to Metatron and the letter Vav that represents him,29 which is also initial to the Name 72.30 In order to summarize, the paratextual notions indicating the gematria of 26, 78 and 91 and the mystical

However, we also find in R. Abulafia’s output descriptions of teachings on visualization of the letters of the name YAHDVNHY and their anthropomorphic representations—for example, the vision of 22 thousand letters/warriors accompanying Yahoel from the final section of Sefer ha-Ot. 27 Zohar 1:110b. 28 See Mark Verman, “The Development of Yihudim in Spanish Kabbalah,” in “Proceedings of the 3rd International Conference on the History of Jewish Mysticism: The Age of the Zohar,” ed. J. Dan. Special Issue, Jerusalem Studies in Jewish Thought 8 (1989): 25–41, where one can find further analysis of yihud appearing in R. Yosef Gikatilla’s Ginnat Egoz and the corpus of Tikkunei Zohar. Further gematria 91 examples, like sukkah and malach, are presented there in connection to Metatron. 29 Note the description of the sign on the forehead of Yahoel, resembling “the staff” or “scepter” and “separating between blood and ink” that appears in the last section of Sefer ha-Ot. 30 More on the sources of these relations in Elliot Wolfson, Along the Path: Studies in Kabbalistic Myth, Symbolism, and Hermeneutics (New York: SUNY Press, 1995), 150. Regarding the linkage of the sixfold and dual-triangle symbol of Magen David and the Name 72 with the Metatron theme, see: Gershom Scholem, Kabbalah (Jerusalem: Keter Publishing House, 1974), 366. On the connection of Magen David and YHVH with the reference to the gematria for barach 222 (2 x 111), see Aryeh Krawczyk, “Uwagi o błogosławieństwie, czyli zasada 222,” Ex nihilo 2, no. 6 (2011): 47–56. 26

138

ARYEH M. KRAWCZYK

meanings conveyed in R. Abulafia’s opus are best exposed via the following scheme:

BIBLIOGRAPHY

Abrams, Daniel. Kabbalistic Manuscripts and Textual Theory: Methodologies of Textual Scholarship and Editorial Practice in the Study of Jewish Mysticism. Jerusalem: Magnes Press, 2010.

———. “Kabbalistic Paratext.” Kabbalah 26 (2012): 7–24.

Goldberg, Oskar. Die Wirklichkeit der Hebräer. Berlin: David Verlag, 1925; Wiesbaden: Harrassowitz, 2005.

Hames, Harvey. “Three in One or One That is Three: On the Dating of Abraham Abulafia’s Sefer ha-Ot.” Revue des études juives 165, no.1–2 (2006): 179–189. Idel, Moshe. The Mystical Experience in Abraham Abulafia. Albany: State University of New York Press, 1988. Krawczyk, Aryeh. Krew, orchidea, atrament : endofazja i heautoskopia w “Sefer ha-Ot”, “Księdze Znaku” R. Abrahama Abulafii (12401292) : edycja krytyczna, tłumaczenie tekstu. Warsaw: Żydowski Instytut Historyczny im. Emanuela Ringelbluma, 2018.

———. “Uwagi o błogosławieństwie, czyli zasada 222.” Ex nihilo 2, no. 6 (2011): 47–56.

PERITEXTUAL ENCODING IN THE SEFER HA-OT

139

———. “Sefer ha-Ot: Preliminary Insights on a Critical Edition.” Jewish History Quarterly 2 (2015): 281–314.

———. “Zdejmowanie zasłony języka. Tora prorocka Mojżesza i prorokowanie Torą Abulafii.” Hermaion 2 (2013): 28–38. Paluch, Agata. Megalleh ʿAmuqot: The Enoch-Meṭaṭron Tradition in the Kabbalah of Nathan Neṭa Shapira of Kraków (1585–1633). Los Angeles: Cherub Press, 2014.

Sagerman, Robert. The Serpent Kills or the Serpent Gives Life. Leiden: Brill, 2011.

Scholem, Gershom. Kabbalah. Jerusalem: Keter Publishing House, 1974.

Verman, Mark. “The Development of Yihudim in Spanish Kabbalah.” In “Proceedings of the Third International Conference on the History of Jewish Mysticism: The Age of the Zohar.” Edited by J. Dan. Special Issue, Jerusalem Studies in Jewish Thought 8 (1989): 25–41. Wolfson, Elliot. Along the Path: Studies in Kabbalistic Myth, Symbolism, and Hermeneutics. New York: SUNY Press, 1995. ———. Abraham Abulafia—Kabbalist and Prophet: Hermeneutics, Theosophy and Theurgy. Los Angeles: Cherub Press, 2000.

READING AND REMEMBERING IN THE MEDIEVAL NEAR EAST: THE SYRIAC S HEMOHĒ BOOK (AKA. THE SYRIAC “MASORAH”) JONATHAN LOOPSTRA Between the sixth to thirteenth centuries, Syriac-heritage Christian communities developed an innovative type of manuscript commonly known to scholars today as the Syriac “Masorah.” Syriac-heritage scribes, however, knew these manuscripts as booklets of shemohē, roughly translated as vocalized “words.” True to this title, manuscripts of this genre consist mostly of individual words or short phrases taken from the Bible or other works of Syriac literature. The thousands of excerpts, or “sample texts,” in these manuscripts occur without versification with the result that most have not been studied in great detail. Yet, these texts are some of our earliest sources for the vocalization and punctuation of many Syriac words, and, as such, are of great value. This paper will introduce the general reader to this curious genre. Given the paucity of readily accessible studies of these manuscripts and a number of misunderstandings in previous work on this material, this paper aims to provide a basic, yet accessible introduction to these shemohē books. Though these manuscripts are complex, they tell us a great deal about how the Syriac Bible and other works were read, recited aloud, and studied in the Medieval Near East. This genre represents a significant, yet under-explored development in the history of education. The author hopes that this paper will help the nonspecialist appreciate the variety of ways these Syriac manuscripts were used. 141

142

JONATHAN LOOPSTRA

The type of manuscript we are considering incorporates all three features named in the title of our 2018 Institute for Advanced Study colloquium: “Dots, Marginalia and Peritexts.” Though this Syriac genre has often been overlooked in contemporary scholarship, we have more manuscripts of this type than we have manuscripts of better-known works in Syriac, such as the Demonstrations of Aphrahat, the Book of Steps, or even certain homilies of famous writers such as Ephrem and Jacob of Serugh. To date, most introductions to this genre have been written by specialists for specialists.1 Yet, this genre represents a significant milestone in the historical development of the Syriac Bible and in how the Syriac language was taught and transmitted. This article is an attempt to provide what one hopes will be a more accessible, basic introduction to these fascinating texts from antiquity. Though scholars have labelled this genre the Syriac “Masora,” the native Syriac-Christian title, “collections of vocalized words (shemohē),” more accurately communicates the pedagogical purpose of these works. To be exact, manuscripts of this genre are composed of list after list of excerpts with added vowels, glosses, and other reading marks. The main purpose of these compilations was to help the reader pronounce and interpret select vocalized words, or shemohē, from across a spectrum of different sources: including works of patristics, grammar, exegesis, lexicography, and the Bible. Since Syriac was not originally written with vowels, these excerpts include some of the earliest examples of fully vocalized words in the Syriac language. These are also some of the first texts to include the fully-developed spectrum of West Syrian reading dots and accents used to mark interpretation and intonation. Furthermore, intermixed among these thousands of difficult words are definitions, interpretive notes, alternative vocalizations, and other peritextual These specialized introductions have also tended to focus only on certain parts of these compilations and not the whole. For example, there is an excellent survey of this material as it relates to the New Testament in Juckel, “The ‘Syriac Masora’ and the New Testament Peshitta,” 107–21. Such specific surveys are necessarily limited. Much of the material in this article is taken from the author’s new study. See, Loopstra, The Patristic “Masora”. 1

READING AND REMEMBERING

143

marks—all reminders of how the reader was expected to recite or explain the text. Though this genre of shemohē book has been largely overlooked in modern scholarship, this was not the case in the Medieval Near East. Manuscripts have been preserved from both EastSyrian (Church of the East) and West-Syrian (‘Jacobite’) communities. Most of these date to the short period between the late ninth to early twelfth centuries, making them roughly contemporary with the period in which Jewish Masoretes and Muslim scholars were helping their own communities to read and interpret their scriptures. Since we know so little about this Syriac material, possible parallels between these traditions have not been followed through. When most people open a book they expect to find ideas expressed in complete sentences, whether these sentences together constitute the text of the Bible, Chaucer’s Canterbury Tales, or the latest novel. But tucked away in many great libraries around the world, whether in London, Paris, Rome, or Damascus, are manuscripts in the Aramaic dialect of Syriac for which this is not true. Rather than including whole sentences, these shemohē books consist, line upon line, of individual words or phrases. And these lists of vocalized words often continue for hundreds of pages. We can see an example of one of these lists in the following image taken from a manuscript in London known as BL (British Library) Additional MS 7183, possibly dated to 1031-2 CE. Opening this manuscript to the biblical book of Acts, we find the following word list on the verso of folio 82.

144

JONATHAN LOOPSTRA

Fig. 1. BL Additional MS 7183, fol. 82v b 2–8 © The British Library Board

The Syriac words in this manuscript image are set down below with English translation.2 Note the added ellipses (…), indicating where the biblical text has been omitted before the next excerpt appears. Acts 1:1–6 line 2: ‫( ܟܬܒܐ ܦܕܡܝܐ ܟܬܒܬ ܐܘ ܬܐܘܦܝ�܆‬I wrote the first book, O Theophilus …) line 3: ‫( ܕܫܪܝ ܡܪܢ‬Our Lord began …) line 4: ‫( ܠܡܥܒܕ ܘܠܡ�ܦܘ܇‬to do and to teach …)

English translation is taken from Kitchen, trans., The Book of Acts According to the Syriac Peshiṭta Version with English Translation. 2

READING AND REMEMBERING

145

line 5: ‫( ܐܣܬܠܩ‬he ascended …) line 5: ‫( ܕܦܩܕ‬he had commanded …) ̈ line 5: .‫ܘܒܝܘܡܝ ܐ�ܒܥܝܢ‬ (during the forty days …) line 6: ‫( ܐܟܠ‬he ate …) line 6: ‫( ܦܩܕ‬he directed …) line 6: ‫( ܢܦܪܩܘܢ‬to depart …) line 7: ‫( ܐܥܡܕ‬baptized …) line 7: ‫( ܬܥܡܕܘܢ‬you shall be baptized …) line 8: ‫( ܙܒܢܐ ܡܦܢܐ‬the time you will be restoring [mapnē] …) ܿ marginal note: ‫ܡܦܢܐ ܐܢܬ‬ ܼ ‫( ܐܝܬ‬there is, he is bringing back [mapnē])

As we see, there are no complete sentences in the above list; it continues, in the same fashion, line after line, for over a hundred more pages in the shemohē book. Moreover, these words appear in this manuscript in the same sequence as they appear in the Bible, but often with significant gaps between these excerpts. But why would a scribe fill a manuscript such as BL Additional MS 7183 with thousands of words and phrases, all excerpted in the same order as they appear in the full work? By themselves, apart from the underlying narrative, these ‘disembodied’ excerpts make little sense. In this list, there is no story, no plot, no sentence, no connection between subject and object; there are not even numbered verses to help the reader place these excerpts in the appropriate context. So, why bother to create such a list in the first place?

THE PROBLEM

The major factor leading to the development of this genre was the ambiguity inherent in early written Syriac. As with so many other Semitic languages, Syriac began as a consonantary system, not an alphabetic one. Vowels and other pronunciation marks were added only centuries later. This means that a significant percentage of words in Syriac are homonyms or homographs.3 As any student of this language knows, without additional helps such as vowels or It has been estimated that homographs make up at least ten percent of the words in modern Syriac lexicons. Kiraz, The Syriac Dot, 31. 3

146

JONATHAN LOOPSTRA

dots, it is easy to confuse words or verb forms. Take, for instance, the following phrase written in English and without vowel marks: “FR GD S LVD”

Should we read the consonants GD as ‘good’, ‘gad,’ or ‘God’? Can we understand LVD as ‘lived,’ ‘livid,’ ‘laved,’4 or ‘loved’? Of course, if we already know that this unvocalized excerpt is taken from the Gospel of John, we might be able to identify the string of consonants as excerpted from John 3:16: “FOR GOD SO LOVED [THE WORLD].”

In a similar way, Syriac verbal forms can be especially tricky to disambiguate without extra help. Lacking added vowels or dots, one form of a verb can often look exactly like another. In fact, we have already seen this potential for ambiguity in the word list from BL Additional MS 7183 above (Fig. 1.). Notice that there appears to have been some confusion about how to vocalize the consonants in the word ‫ܡܦܢܐ‬, marked by a reddish-colored symbol in line 8 towards the bottom of the page. The verb ‫ ܡܦܢܐ‬derives from a form of the root ‫ܦܢܝ‬, ‘to turn’ or ‘return.’5 If you look carefully at Figure 1, you can see that a later user—not the original copyist—has added vowels to these consonants in a marginal note with a vowel above the pe

: “there is ܰ mpanē (‫)ܡܦ ܷܢܐ‬.” However, to the left of this note, in the list of words, the original copyist has preferred a slightly different vocaliܰ zation, indicated by a dot and also a triangular-like mark (◌) above the mem : mapnē (‫) ܰܡܦ ܷܢܐ‬. Though at first glance the difference between these two vocalizations seems small (mpanē versus mapnē), in practice this change results in a variance of interpretation. In this case, adding an ‘a-vowel’ to the pe

, results in mpanē, what ‘Laved’ is in fact a word, the past tense of ‘lave,’ to wash. This paper will avoid adding other phonological dots found in these manuscripts. Ignored here, therefore, are large dots known as rukkokho and quššoyo, used on certain Syriac letters, known as the ‘bgdkpt letters,’ to express unaspirated or aspirated sounds. I have used in this paper to express the consonantal form of the word, without added vowels. 4 5

READING AND REMEMBERING

147

is known grammatically as a Paʿel participle: ‘he is giving back,’ ‘repaying,’ or ‘granting.’6 Conversely, adding an ‘a-vowel’ to the first letter, the mem , produces mapnē, a causative Aphʿel participle: ‘he makes to return,’ or ‘he restores.’7 Either way, there was apparently enough of a difference to merit the addition of the marginal note to this shemohē book by a later user. But we can also see in this example a second problem that faced the reader. These supplementary vowels and dots were later additions; they were not originally part of many Syriac texts. In fact, it took centuries for Syriac vowels and punctuation marks to develop.8 Consequently, once added, these marks could be open to debate. In the centuries between when the consonants were composed and the vowels added, different scribes, schools, or regions had developed their own preferred vocalization. Some of the biggest differences were between the two great divisions of Syriac Christianity: the East and West Syrian religious communities. Over time, each of these traditions had gradually developed its own peculiar preferences in style of vocalization, punctuation, and even the spoken recitation of the Bible. Though originally consonantal, the Syriac Bible had gradually accumulated layers upon layers of these traditions of supplementary vowels, punctuation, and accent marks. And here lies the significance of the Syriac shemohē book: these manuscripts became platforms where certain vocalized words—including those that posed challenges or were possibly contentious—could be collected, assessed, and passed down.

THE SOLUTION: THE SYRIAC “S HEMOHĒ BOOK”

It is in this context that the genre of shemohē book developed, a creative answer both to the ambiguities discussed above and to other problems faced by the reader of Syriac. And it is significant that we find examples of this genre in both of the main divisions of Syriac Sokoloff, A Syriac Lexicon, 1206. Sokoloff, A Syriac Lexicon, 1206. 8 Kiraz, The Syriac Dot; Coakley, “The Five Greek Vowel Signs,” 307–325. 6 7

148

JONATHAN LOOPSTRA

Christianity; both communities, after all, faced similar linguistic issues. Though it appears that this genre developed first in EastSyrian Christianity,9 we know of only one manuscript, dated to 899 CE, which has been preserved in this Eastern tradition.10 On the other hand, the vast majority of these shemohē books are preserved in manuscripts of the West-Syrian tradition, and most of these were written between the tenth and thirteenth centuries, a period known as the Syriac Renaissance.11 It is perhaps no coincidence that a genre used primarily for teaching and learning would have had its heyday in a period renowned for a rebirth of learning and scholarship. As we saw in the example from the Book of Acts above, the vast majority of the word lists in these shemohē books include excerpts from the Syriac Bible known as the Peshiṭta version. But there are also other lists. Some are taken from a more literal Syriac version of the New Testament, the Ḥarqlean, and yet others from the writings of key fourth- and fifth-century thinkers, such as Gregory of Nazianzus, Basil of Caesarea, Severus of Antioch, and Pseudo Dionysius. In addition, most shemohē books from the period of the Syriac Renaissance include a type of appendix, consisting of a series of grammatical, philological, lexicographical, and exegetical tracts. Moreover, we now know that this genre was eventually applied to other literary works in Syriac. For example, one manuscript includes word lists taken from the works of Syriac-heritage poets such as Ephrem (4th c.), Jacob of Serugh (5th c.), and Isaac of Antioch (5th c.).12 Some lists are taken from the stories of famous monks, such as St. Antony and Abba Bishoi, yet others from narraIn his Book of Governors, the ninth-century bishop Thomas of Marga tells the memorable story of the seventh-century monk ʿEnanishoʿ, who wrote “[the correct pronunciation] of the words and readings…in the writings of the Fathers.” Budge, ed., The Book of Governors, 1:80. 10 This manuscript, BL Additional 12138, has now been published in Loopstra, An East-Syrian Manuscript of the Syriac ‘Masora’. 2 Vols. 11 Teule, et al, eds. The Syriac Renaissance. 12 See the addition to the Damascus manuscript, Dam. Syr. Orthodox Patriarchate 12/22, fols. 267r–305v. Dolabani, “Catalogue des manuscrits de la Bibliothèque du Patriarcat Syrien Orthodoxe à Ḥomṣ,” ‫ܝܗ–ܝܛ‬. 9

READING AND REMEMBERING

149

tives of saints and martyrs, such as St. Barbara and the Forty Martyrs of Sebaste. Other word lists would have helped the reader make sense of important theological works, such as Basil’s treatise on the seven days of Creation, known as the Hexaemeron. All said, though this Syriac genre appears to have first developed in the sixth and seventh centuries to help users read difficult or ambiguous words in the Bible, by the thirteenth and fourteenth centuries the shemohē book had become a highly integrated and complex tool, one that helped users read not only biblical works, but a wide spectrum of the Syriac literary canon. It is perhaps best to think of these shemohē books as memory aides, used alongside full manuscripts of literary works to remind the user of issues or questions regarding pronunciation, vocalization, and interpretation. This abbreviated composition makes these shemohē books quite challenging for modern scholars. As we have seen above (Fig. 1), these lists are extremely concise and implicit, having been designed for readers who already had some background in the material and had been taught how to use these lists. If it is true that a pedagogical component was integral to these shemohē books, then we are missing at least half the conversation— the unwritten, spoken part of the teaching process. These books might also indicate that these underlying literary works were studied in whole, from beginning to end, in depth, since it would be very difficult to start in the middle of the shemohē lists, given the absence of versification. Yet, because we have well over a dozen of these manuscripts, we can unravel at least part of the riddle of how these shemohē books were used. Doing so gives us a genuine appreciation and admiration for the high level of attention given by Syriac-heritage scribes to their Bible and other writings.

USING THE S HEMOHĒ BOOK

As short-hand guides, there are many possible ways these shemohē books would have been used. In the following pages, we will look briefly at five of the most significant reasons this genre would have benefitted a reader. 1) As Guides to Vowels and Pronunciation Perhaps the most common use for these shemohē books was as a guide to the vocalization of a word or short phrase. Such infor-

150

JONATHAN LOOPSTRA

mation would have been invaluable precisely because many early Syriac manuscripts lacked vowels. Of course, knowledge of how to vocalize or pronounce a passage would have been especially important for someone preparing to read the Scriptures aloud, such as a deacon or a priest.13 It would also have been necessary for the reader to pronounce the passage in the manner embraced by his own church, whether East or West Syrian. The shemohē book offered a solution to these issues. We have seen, in the passage from Acts 1:6 above (Fig. 1), a later user has added a note in the margin concerning another way to vocalize the consonants ‫ܡܦܢܐ‬: “there is ‘mpanē’”. In consequence, we find that BL Additional 7183 provides two possible vocalizations for this one consonantal form: 1) the preferred reading set down by the copyist in the word list, and 2) the marginal variant added by a later user. Interestingly, other shemohē books likewise include comments on this same word, and some give the readings preferred by certain renowned scribes. In Vatican siriaca 152 (hereafter Vat. sir. 152), dated to 980 CE, the copyist writes out the preferred vocalization of a famous scribe Ṭubono, abbreviated as ‫( ܛܘ‬Fig. 2).

Fig. 2. Vat. sir. MS 152, fol. 118v © 2019, Biblioteca Apostolica Vaticana

The compilers of these manuscripts frequently use the abbreviation “is found,” ‫)ܡܫܬܟܚ( ܡܫ‬, for variants that the compiler may have located in other shemohē books. In Figure 3 (beIt is perhaps for this reason that many of these manuscripts were owned by deacons and priests. 13

READING AND REMEMBERING

151

low), the compiler of a late manuscript from Jerusalem, St. Mark’s ܰ 42 (17th century), places the variant mpanē (‫ )ܡܦܢܶܐ‬immediately after ܿ ܵ his preferred reading mapnē (‫) ܼܡܦܢܐ‬, vocalized here with dots. Note, that both readings appear in the text of the manuscript and not in the margin. In this case, it is possible that this copyist is setting down a variant vocalization he has “found” in other shemohē books (though we cannot be completely certain of this). Regardless, it is clear that the compiler knew of both possible vocalizations of this one word.

Fig. 3. St. Marks’ MS 42, fol. 78r © 2019, Hill Monastic Manuscript Library

Moreover, based on available evidence in these shemohē books, it would appear that the two Syriac-heritage communities were divided on how to vocalize the word ‫ ܡܦܢܐ‬in Acts 1:6. West-Syrian shemohē books unanimously prefer aܰ causative sense, as Christ ‘restoring’/’making return’ (mapnē ‫ ) ̇ܡܦܢܶܐ‬the Kingdom to Israel. On the other hand, our one East-Syrian manuscript indicates ܿ a preference that Christ was going to ‘repay’/‘grant’ (mpanē ‫ܡܦ ܵܢܐ‬ ܼ ) the Kingdom (Fig. 4).14

Fig. 4. BL Additional MS 12138, fol. 266r

Remarkably, we find this same debate over the vocalization of this word echoed in Syriac biblical commentaries from the period when these shemohē books were still being written. One prolific thirSee possible options for English translations in Sokoloff, A Syriac Lexicon, 1206. Images from this manuscript are taken from the published facsimile in Loopstra, An East-Syrian Manuscript of the Syriac ‘Masora’. 14

152

JONATHAN LOOPSTRA

teenth-century West Syrian commentator, Bar ʿEbroyo, interprets this word as follows:15 ‘In this time you will be making return,’ [written] with a ptoḥo ܶ ܰ ̇ As in Ps. 126:1 [Syr.125] ‘When (‘a-vowel’) mīm [mapnē ‫]ܡܦ ̣ܢܐ‬. the Lord restores the fortunes of Zion.’ But the Nestorians, ܰ 16 As in Ps. 16:5 ‘As you [write] with a ptoḥo pe [mpanē ‫]ܡܦܢܶܐ‬. grant my inheritance.’

Bar ʿEbroyo, therefore, is clearly echoing the same variants in vocalization that we also find in these shemohē books. Moreover, by using the Psalter to illustrate his preferred West-Syrian vocalization ܰ of this word in the book of Acts (mapnē ‫) ̇ܡܦܢܶܐ‬, Bar ʿEbroyo shows an awareness of these seemingly-minute details of vocalization across the Bible. Though such a debate about whether to place a vowel over one consonant or another may seem trivial today, such concerns were not necessarily trifling to the compilers of these shemohē books. 2) As ‘All-in-One’ Handbooks to the Bible These shemohē books not only served as guides to pronunciation, but they could also function as something akin to an ‘all-in-one’ handbook to the Bible and the Syriac language. Most of these compilations include a type of appendix, which consists of tracts that covered diverse subjects ranging from grammar and lexicography to history and theology. The types and number of tracts often Klamroth, ed., Gregorii Abulfaragii Bar Ebhraya in actus Apostolorum et epistolas catholicas adnotationes, 3. .‫ ܒܦܬܚ ܡܝܡ‬:‫ ܐܢ ܒܗܢܐ ܙܒܢܐ ܡܦܢܐ ܐܢܬ‬.‫ܡܪܢ‬ 15

‫ ܕܐܢܬ ܗܘ‬.‫ ܐܝܟ ܗܝ‬.‫ ܒܦܬܚ ܦܐ‬.‫ ܢܣܛܘ�ܝܢܘ‬.‫ ܕܟܕ ̇ܡܦܢܐ ܡܪܝܐ ܫܒܝܬܐ ܕܨܗܝܘܢ‬.‫ܐܝܟ ܗܝ‬ .‫̇ܡܦܢܐ ܠܝ ܝܪܬܘܬܝ‬ Interestingly, though both of these passages from the Psalter translate with the Syriac verb , these are two entirely different verbal roots in the original Hebrew. Psalm 126:1 uses the verb ‫שׁוּב‬, ‘to return;’ Psalm 16:5 has the verb �‫תּ ַמ‬,ָ ‘to grasp’ or ‘support.’ In other words, Bar ʿEbroyo’s comparison between these passages can only be made in Syriac. He is also right to see a difference in semantics between these passages, though they are translated into Syriac with the same root verb. 16

READING AND REMEMBERING

153

vary from manuscript to manuscript. But in the hands of a skilled user, these tracts effectively turned a shemohē book into a shorthand reference work. Using this one manuscript, the reader could look up facts and figures related to the Syriac Bible and Syriac Christianity, while also perusing the vocalized word lists towards the front of the book. The result would have been a highly portable, compact companion-volume to Bible—a handbook. One of the most common of the tracts in the back of these handbooks is a grammatical work listing the various ways one could vocalize certain consonantal homonyms in Syriac. When we turn to this “Tract on Homonyms” in Vat. sir. 152, we find that the copyist has actually included the verb discussed above ‫ܡܦܢܐ‬. On folio 197v of this tract, he lists three possible vocalizations for this word, one after the other: ‫( ܶܡ ̣ܦ ܳܢܐ‬mēpno) [Peʿal infinitive, ‘to turn’], ܰ ‫( ̇ܡ ̣ܦܢܶܐ‬mapnē) [Aphʿel participle, ‘he made return’], ܰ̇ ܶ and ‫( ܡܦܢܐ‬mpanē) [Paʿel participle, ‘he grants’].

In other words, should our West-Syrian reader of Acts 1:6 have wanted to recall the various ways ‫ ܡܦܢܐ‬could be vocalized, all he had to do was flip to the back of his shemohē book. But there are many other helpful tracts in the back of these shemohē books as well. Perhaps our diligent reader was working through the Book of Acts and wanted to know the number of sentences in this book—there was a tract for that. According to this tract, Acts should have 4,149 versicles. Or, perhaps, our reader wanted to figure out how and where each of the prophets and apostles died—there was a tract for that as well. If he flipped to the back of his shemohē book, our diligent reader would discover that Peter was “the head of the Apostles; he was crucified upside down in Rome.” This information is concise and to the point, in keeping with the abbreviated style of these handbooks. But say he needed a little help remembering specific weights and measures such as the mina or the liter used in the Bible—yes, there was a tract for that. Again, if our reader wanted to dig a little deeper and explain to his friends or students the meaning of the original Greek or Hebrew words—there was a tract for that. Working through the lexicographical tract in the back of his shemohē book, our industrious reader could determine that the Greek word praxis (πρᾶξις), used

154

JONATHAN LOOPSTRA

as a title for Book of Acts in his Syriac Bible (‫)ܦܪܟܣܝܣ‬, means in Syriac ‘acts’ (‫)ܣܘܥ�ܢܐ‬. Moving on, he finds that euangelion (εὐαγγέλιον ‫ )ܐܘܢܓܠܝܘܢ‬can be translated into Syriac as ‘good news’ (‫)ܣܒܪܬܐ‬, katholikós (καθολικός ‫ )ܩܬܠܝܩܐ‬as ‘universal’ (‫)ܟܘܠܢܝܬܐ‬, or parakletos (παράκλητος ‫ )ܦܪܩܠܝܛܐ‬as ‘advocate’ (‫)ܒܥܝܐ‬, and so on. This fascinating pair of tracts titled “Interpretations of Hebrew and Greek Words” attest to the highly developed state of lexicography in Syriac-heritage communities by the tenth century (Fig. 5).

Fig. 5. St. Marks’ MS 42, fol. 140v © 2019, Hill Monastic Manuscript Library

Though there are many other tracts in the back of these shemohē books, one more tract bears mentioning: “On Heresies.” In his reading of Acts, our reader might want to know more about what Jews believe about God. Or, he might be curious to grasp how the theology of his East-Syrian neighbor differs from his own WestSyrian beliefs. Yes, there was a tract for that as well. Several shemohē books include a short work designed to help the reader recall very ܶ ‫) ܷܐ‬. Here, our diligent basic distinctions between ‘heresies’ (‫�ܣ ܺܝܣ‬ reader learns that Jews believe there is “One God in one person (prosopon πρόσωπον).”17 He also learns that his East-Syrian neighbor believes with “Diodore, ܳ Theodore, and Nestorius … in two ܳ Sons, ܳ ܽ ܰ one18 of nature (kyono ‫ )ܟܝܢܐ‬and another of grace (ṭaybutho ‫)ܛܝܒܘܬ ܐ‬.” He could read on, moreover, to learn about Manicheans, Arians, or other sects. 17 18

St. Marks’ MS 42, fol. 141r. St. Marks’ MS 42, fol. 141v.

READING AND REMEMBERING

155

Though we have examined only a few of the tracts in back of these handbooks, we can already see how useful this material could be for Syriac-heritage readers. Not only could the hundreds of pages of vocalized and dotted words in the first part of these books have helped the reader recall how to vocalize select excerpts, but various tracts in the appendix provided additional helpful information though in abbreviated form. Brought together in one large codex, this varied material allowed shemohē books to function as fairly comprehensive sources for knowledge of the Syriac Bible and Syriac language. 3) As Guides to Punctuation and Accent Marks These shemohē books were used as well to aid the reader when reciting the Scriptures aloud. Given our modern, relatively-literate age, we often forget that most Christians in the past approached the Scriptures through the spoken, not the written word.19 How the Scriptures were read aloud to the congregation would have been especially important. In a passage in his grammar, Bar ʿEbroyo (13th c.) discusses the need for these additional dots.20 In every language the hearer can distinguish by hearing (alone) the various meanings of one and the same phrase … but simply by changes of modulation (in the voice). Syrian writers, therefore, who laid the firm foundations of their speech, devised a system and arranged point-symbols for the accents, so that these various tones, each of which indicated a particular meaning, could be pronounced by the reader through sight just as they are recognized by the speaker through hearing.

Carol Harrison writes that most “early Christian texts which we now read were originally spoken, rather than written, and were intended for hearers, rather than readers. They sounded, resonated, and impressed themselves upon the mind and memory through the ear rather than the eye.” Harrison, The Art of Listening in the Early Church, 1. 20 This translation taken directly from Segal, The Diacritical Point and the Accents in Syriac, 61. 19

156

JONATHAN LOOPSTRA

In a similar way, the compilers of these shemohē books also appear to have understood the need to properly punctuate and intone the Scriptures aloud. For this reason, dots were added to the thousands of words in these shemohē books. These included medium-sized dots used for punctuation, such as stops and pauses, and larger dots used to inform the reader where to raise or lower one’s voice or where to stress a syllable. These punctuation and accent marks can consist either of single or multiple dots, and they can be placed above, on, or below the line. As one can imagine, the resulting combination of dots could be bewildering, even to the initiated; some Syriac scribes took to labelling this bewildering mixture a “net of dots.”21 Moreover, as was true for vowels, most of these larger punctuation and accent marks were developed only after the Bible had been translated into Syriac, and this meant that these dots could be open for debate as well. Because there were eventually well over twenty of these dotted-marks, it might have been relatively easy for a reader to forget or confuse these dots in his reading. For this reason, the copyists of these shemohē books sometimes felt the need to remind their users of the presence of a punctuation or accent mark. They generally do this by writing the abbreviated name of the mark either in the margin or in the word list, immediately following the actual dots of the mark. Putting this idea into an English-language perspective, imagine a scribe who adds the abbreviation “Excl.” after an exclamation mark (!) to remind his readers how to properly recite Matthew 12:34: “You brood of vipers! Excl.”. In a similar manner, the copyists of these shemohē books include abbreviations to ensure their readers note key features of Syriac punctuation. We can illustrate this phenomenon in Syriac with a look at two of these marks. One of the most common punctuation marks is the double-dotted taḥtoyo (‫) ܆‬, used to divide a sentence and also to mark a pause.22 In keeping with our study of the Book of Acts in Vat. sir. 152, we find that the abbreviation ‫( ܬܚܬ‬taḥtoyo) immediate21 22

Joseph bar Malkun (12th c.) wrote a treatise by this title. Kiraz, Syriac Orthography, 148.

READING AND REMEMBERING ly follows the name Aeneas in Acts 9:34: “Aeneas

‫)ܬܚܬ‬.” Placed in context, the entire verse would read, [Simeon said to him,] “Aeneas heals you.]23

157

‫ ܆‬tḥt ( ‫ܐܢܝܐ܆‬

‫ ܆‬taḥtoyo” [Jesus the Messiah

Apparently, the copyist was so eager for his reader to remember the taḥtoyo that he provided this abbreviation as an additional cue, even after he had already written the double dots signifying the pause. Interestingly, our East-Syrian shemohē book likewise includes only this one word, Aeneas, for Acts 9:34, though with a slightly different mark. Whereas West-Syrian scribes used the taḥtoyo as a way to note pauses, East-Syrian scribes had developed a different three-pointed mark, the taḥtoyo da-taloto ( _ ), to indicate pauses or to possibly raise intonation (Fig. 6).

Fig. 6. BL Additional MS 12138, fol. 269r

Once again, we find that these shemohē books serve to remind their readers of features distinct to the copyists’ own church tradition. Most West-Syrian books remind the reader to include after Aeneas the double-dotted taḥtoyo (‫) ܆‬, while our only copy of the EastSyrian shemohē book reminds the reader of the taḥtoyo da-taloto ( _). Another mark that is likewise reiterated in these shemohē books is the mshalono (‫)ܡܫ�ܢܐ‬, which serves as a type of question marker. Because Syriac lacks a basic particle used to mark questions, such as we have in Hebrew (-‫ ) ֲה‬and in Arabic (‫)ھل‬, many sentences can be taken either as a statement or rhetorical question.24 As one can imagine, this could be confusing. In Job 38:28, for instance, the Syriac can The marks [ ] here denote sections of the text not included in the word lists. Kitchen, trans., The Book of Acts According toܰ the Syriac Peshiṭta ܽ ‫ܐܡܪ ܶܠܗ ܶܫ‬ ܶ ‫ܡܥܘܢ܂ ܐ ܺ� ܰܝܐ܆ ܰܡ‬ ܰ ‫ܶܘ‬ Version with English Translation, 65. ‫ܐܣܐ ܳܠ ܼܟ �ܶ ܽܫܘܥ‬ ܺ . Note that the editors of the Antioch Bible here, have also placed ܳ‫ܡܫܝܚܐ‬ ܰ a taḥtoyo as punctuation after the word Aeneas (‫)ܐ ܺ� ܰܝܐ܆‬. 24 For more background, see Segal, The Diacritical Point and Accents in Syriac, 68. 23

‫܂‬

158

JONATHAN LOOPSTRA

be read either as a statement, “there is a father (‫ )ܐܝܬ ܐܒܐ‬for the rain,” or as a rhetorical question, “does the rain have a father?”25 If you think about it, there is a danger that the former reading—the statement—might take listeners in an unorthodox direction: Is the Bible really declaring that the rain has a father? It is perhaps for this reason that almost every West-Syrian shemohē book inserts for this passage in Job either the abbreviation ‫< ܡܫ‬msh>, or the full word ‫ܡܫ�ܢܐ‬, after the double dots of the mshalono (Fig. 7). Fig. 7. Dam. Patriarchate Syr. MS 7/16, fol. 38r © 2019, Syrian Orthodox Patriarchate Damascus

The goal was to remind the user that this verse should be read as a question (usually a rhetorical question).26 And as we saw with the taḥtoyo punctuation mark above, such reminders were added even though the dots of the mshalono were also written down in the same excerpt. The copyists of these shemohē books apparently wanted to ensure their users read Job 38:28 as a rhetorical question, not a statement. Moreover, such information is particularly valuable for us, because these mshalono question marks in shemohē books were sometimes placed on verses where our English Bibles do not read the same passage as a question. Syriac-heritage copyists may have interpreted these passages as questions solely based on their knowledge of the Syriac, without recourse to the original Hebrew or Greek. This means that these shemohē books sometimes directed their users to read passages in a sense that would be quite different from the way we read these same verses in English. Take Numbers 36:3, for instance. Most modern English Bibles translate this passage as a statement: “But if they become wives in any of the tribes of Israel … it shall be taken away from the lot of our inherLoopstra, trans., The Book of Job According to the Syriac Peshitta Version, with English Translation, 216-217. 26 Kiraz, Orthography, 136. 25

READING AND REMEMBERING

159

itance.”27 Yet, the copyists of these shemohē books generally remind the reader of the presence of a mshalono here, thus a rhetorical question: “… shall it be taken away from the lot of our inheritance?”28 4) As Repositories of Variant Readings and Comments by Later Readers While there can be, and often are, similarities between shemohē books, individual copyists appear to have had a significant degree of freedom to choose which words from within certain phrases they wanted to include in their own handbook. Between manuscripts, words can vary by spelling, vocalization, and other specifics. Curiously, as we have seen in Figure 3 above, copyists often make it clear that they are aware of alternative readings by recording these variants in their own word lists. In other words, these shemohē books appear to have functioned more as repositories for different readings than as highly regulated and standardized master copies. We see this most frequently when two shemohē manuscripts (e.g. ‘MS 1’ and ‘MS 2’) have different spellings or vocalization for the same word. Not infrequently, the copyist of MS 1 will list an alternative vocalization or spelling for the word (Z) in the margin. But what appears as the variant in MS 1 appears as the main reading in MS 2. Likewise, the copyist of MS 2 can list a variant (X) that appears only in the body, not in the margin, of MS 1. In other words, the spelling of a word in one manuscript is sometimes listed as the variant in another manuscript, and vice versa. MS 1

MS 2

X

Z

Z

X

This is the reading in the Antioch Bible, which lacks many of these dotted marks. Cook, The Syriac Peshitta Version, with English Translation. Numbers, 23. 28 See Vat. sir. 152, fol. 22v. 27

160

JONATHAN LOOPSTRA

Though somewhat confusing, this complexity shows us that copyists would have been aware of multiple possible options for how they could vocalize, punctuate, or spell these words. As we have already seen above (Fig. 3), these variants are often labelled by the copyists “is found” (‫ܡܫ‬, or ‫ ;)ܡܫܬܟܚ‬as it happens, this is the same two-letter abbreviation used for the mshalono. Variants labelled in one book as “is found” often appear as the preferred reading—not the variant—in other shemohē books. In other words, we might conclude that copyists seem to have compared multiple shemohē books and were aware of readings in other manuscripts, even as they were writing down their own word lists. As an example, we can take a passage from Acts 8:27: “Candace, the queen of Ethiopia.”29 Some shemohē books spell Candace ܰ ‫( ܰܩ‬Qandaq), as does the St. Mark’s manuscript below. as ‫ܢܕܩ‬ Fig. 8. St. Marks’ MS 42, fol. 79r © 2019, Hill Monastic Manuscript Library

ܰ ‫ܰܩ‬ Yet, other manuscripts, such as Vat. sir. 152, record ‫ܐܢܕܐܩܝܣ‬ (Qandaqīs), a spelling that is closer to the original Greek (Κανδάκης).30 Though the Vatican shemohē book has the longer ܺ ܰ ‫ ) ܰܩ‬as its preferred text, the same copyist reading Qandaqīs (‫ܐܢܕܐܩܝܣ‬ ܰ ‫ ) ܰܩ‬in the margin. It would lists the shorter alternative Qandaq (‫ܢܕܩ‬ seem that the compiler of Vat. sir. 152 was aware of multiple spellings for the name Candace, but chose to include the longer reading, Qandaqīs, in his list of vocalized words, leaving the alternative reading at the margin of the manuscript. These shemohē books, therefore, appear to have been less a means of providing an exactly uniform set of exemplars than a resource for collecting preferences and sifting through variants. Many of the readings in these word lists were not carried forward into later biblical manuscripts. In fact, many of these words, even those ܺ

Kitchen, trans., The Book of Acts According to the Syriac Peshiṭta Version with English Translation, 57. 30 Vat. sir. 152, fol. 120v. 29

READING AND REMEMBERING

161

found in all shemohē books, do not appear in printed Syriac Bibles today. For example, while most shemohē books read “Yes, Jesus (‫ ”)ܝܫܘܥ‬in Matthew 9:28, modern published Syriac Bibles read instead “Yes, Our Lord (‫)ܡܪܢ‬.”31 To illustrate, we have included below three other curious variants, among many others, that never made it into most modern printed Syriac Bibles. Mark 10:52 Mark 14:54 John 4:42

Shemohē Manuscripts “Go (‫”)ܙܠ‬ ̈ “guards (‫”)ܕܚܫܐ‬

“we have believed (‫”)ܗܝܡܢ‬

Printed Syriac Bible (Antioch Bible) “See (‫”)ܚܙܝ‬ ̈ “the servants (‫”)ܡܫܡܫܢܐ‬32 “we have heard (‫ ”)ܫܡܥܢ‬33

So, while there is much that is consistent across these shemohē books, compilers do not appear to have had as their goal a single, highly-standardized text. There is a fair amount of multiformity and diversity between manuscripts. Instead, these compilations were often used as a platform to record variants that may have been found in other shemohē books or in full biblical manuscripts. 5) As Repositories of Later Scribbles, Illustrations, and Doodles Given everything that has been said, it might not be surprising to find that later users often recorded their own notes and insights in these shemohē books. Users even record whether or not a reading in their shemohē book is ‘correct’ (ḥtith ‫ )ܚܬܝܬ‬or ‘not correct’ (lo ḥtith ‫)� ܚܬܝܬ‬. It is not unusual to find that users sometimes weigh in over something as ostensibly minute as whether a word should be ܳ ܰ vocalized with an ‘a’-vowel (◌) or an ‘o’-vowel (◌). Childers, trans., The Gospel of Matthew According to the Syriac Peshitta Version with English Translation, 48-49. 32 Childers, trans., The Gospel of Mark According to the Syriac Peshitta Version with English Translation, 82-83; 114-115. 33Childers, trans., The Gospel of John According to the Syriac Peshitta Version with English Translation, 24-25. For more on these alternative readings in these books, see Loopstra, “Le Nouveau Testament dans les manuscrits syriaques massorétiques. Ou en sommes-nous?” 181-201. 31

162

JONATHAN LOOPSTRA

We can find an example of this in the Christological passage Hebrews 2:9: “[Christ] is the one that was humbled (‫)ܕܡܟ‬.” Shemohē books vary on whether ‘humbled’ should be vocalized as ܳ ), with an ‘o’-vowel. ܰ ), with an ‘a’-vowel, or dmok (‫ܕܡܟ‬ dmak (‫ܕܡܟ‬ The compiler of the Vatican manuscript gives both possible readings, though a later reader has used a red pen to indicate that the reading listed in the margin of this manuscript, with the ‘a’-vowel ܳ dmok), is ܰ dmak), is ‘not correct’ and the other reading (‫ܕܡܟ‬ (‫ܕܡܟ‬ 34 ‘correct.’ Oddly enough, what appears to be our user’s preferred reading (the word not in the margin) is not the vocalization one finds in most printed Syriac Bibles today.35 Clearly, these shemohē books are also valuable for what they tell us about the later scribes who used these compilations, regarding their preferences and concerns. Finally, some of these shemohē books include incredibly beautiful images.36 One of the most striking pictures includes a miniature of Christ between Moses and Paul, remarkable for its imagery and color (in Barbarini MS 118). Likewise, a manuscript in Mosul includes images of Moses, Cyril, and even Ezekiel’s chariot (in Mosul MS 16). In fact, the copyists of these books frequently refer to their works as “treasures.”37 Though this language may be in part rhetorical, this term may very practically reflect the real expense of money and time that would have been involved in creating these massive volumes. As one can imagine, it would have been an extremely tedious and time consuming process to fill up five hundred or so pages with carefully vocalized and punctuated words. Hundreds of thousands of vowels and dots of various sizes would have had to have been placed in these shemohē books, all very carefully and by hand. Vat. sir. 152, fol. 137v. King, and Walters, trans., Hebrews & General Epistles According to the Syriac Peshitta Version with English Translation, 74-75. This edition vocalizes ܰ ). as dmak (‫ܕܡܟ‬ 36 The Barbarini 118 manuscript includes Christ between Moses and Paul on folio 98v. Mosul MS 16, includes an image of Moses and Cyril. You can see this page in color through the Brigham Young University hosted site: https://archive.org/stream/STC1641/Manuscript%2041#page/n3/mode/1up. 37 See colophon of Vat. Sir. 152, for an example. 34 35

READING AND REMEMBERING

163

This being the case, we also find amusing doodles in these shemohē books, left for posterity by various readers. A good example occurs in BL Additional MS 7183, the manuscript with which we began our exploratory journey into this genre. Throughout the first part of this manuscript, a creative doodler has sketched interlaced patterns and humorous bird scenes. We can find on folio 29 an artful scene of two birds pecking at either side of a disembodied head (see Fig. 10 below). It may be that this doodler was simply bored. Or perhaps, through a stretch of the imagination, we might suppose his doodles were prompted by the marginal note below his sketch: “there are 2,553 versicles in the Book of Job.”38 Our doodler would have had a long way to go in order to make it through Job, let alone the rest of the Syriac Bible. Moreover, another note, at the bottom of the page, records that Job 6:5 is “the first mshalono [rhetorical question] of Job.”39 With so much to learn, and so many possible mistakes to make, perhaps our doodler was a bit overwhelmed, deciding to take a break from these word lists to sketch. Whatever the case, these drawings remind us that although studying the Bible with these costly shemohē books would have been hard work, some readers still took the time to leave parts of their own personalities in what were otherwise fairly tedious lists of thousands of words, vowels, and dots.

We find this same number of versicles for Job mentioned in a tract towards the back of these manuscripts. 39 The marginal note here refers to the sentence “Does the wild ass bray over the grass, or the ox bellow over the fodder?” Loopstra, trans., The Book of Job According to the Syriac Peshitta Bible with English Translation, 33. Though this is not technically the ‘first question in Job,’ it is the first such mark to be listed in the collection of words from Job in this shemohē book. 38

164

JONATHAN LOOPSTRA

Fig. 10. BL Additional MS 7183, fol. 29r © The British Library Board

READING AND REMEMBERING

BIBLIOGRAPHY

165

Budge, E. W., ed. The Book of Governors: The Historia Monastica of Thomas Bishop of Marga. 2 vols. London, 1893. Childers, Jeff, and James Prather, trans. The Gospel of John According to the Syriac Peshitta Version with English Translation. Piscataway, NJ: Gorgias Press, 2014. Childers, Jeff, trans. The Gospel of Mark According to the Syriac Peshitta Version with English Translation. Piscataway, NJ: Gorgias Press, 2012. __________. The Gospel of Matthew According to the Syriac Peshitta Version with English Translation. Piscataway, NJ: Gorgias Press, 2012. Coakley, James. “When Where the Five Greek Vowel Signs Introduced into Writing?” JSS 56 (2011): 307–325. Dolabani, F. Y., René Lavenant, Sebastian Brock, and Samir Khalil Samir. “Catalogue des manuscrits de la Bibliothèque du Patriarcat Syrien Orthodoxe à Ḥomṣ (auj. à Damas).” ParOr 19 (1994): 555–661. Harrison, Carol. The Art of Listening in the Early Church. Oxford: Oxford University Press, 2013. Juckel, Andreas. “The ‘Syriac Masora’ and the New Testament Peshitta.” Pages 107–21 in The Peshitta: Its Use in Literature and Liturgy, ed. Bas ter Haar Romeny. Leiden, 2006. King, Daniel, and Edward Walters, trans., Hebrews & General Epistles According to the Syriac Peshitta Version with English Translation. Piscataway, NJ: Gorgias Press, 2016. Kiraz, George. Orthography. Vol. 1 of Turras Mamlla; A Grammar of the Syriac Language. Piscataway, NJ: Gorgias Press, 2012. __________.The Syriac Dot. Piscataway, NJ: Gorgias Press, 2015. Kitchen, Robert A, trans. The Book of Acts According to the Syriac Peshitta Version, with English Translation. Piscataway, NJ: Gorgias Press, 2014. Klamroth, Martin, ed. Gregorii Abulfaragii Bar Ebhraya in actus Apostolorum et epistolas catholicas adnotationes. Göttingen: Dieterichsche Universitätsbuchhandlung, 1878.

166

JONATHAN LOOPSTRA

Loopstra, Jonathan. The Book of Job According to the Syriac Peshitta Version, with English Translation. Piscataway, NJ: Gorgias Press, 2016. __________. An East-Syrian Manuscript of the Syriac ‘Masora’. 2 Vols. Piscataway, NJ: Gorgias Press, 2014–15. __________. “Le Nouveau Testament dans les manuscrits syriaques massorétiques. Ou en sommes-nous?” Pages 181-201 in Le Nouveau Testament en syriaque, ed. J.-C. Haelewyck. Études Syriaques 15. Paris, 2017. __________. “Reading the Bible with the Taḥtāyā da-Tlāta.” In From Ancient Manuscripts to Modern Dictionaries: select studies in Hebrew, Aramaic, and Greek, edited by Tarsee Li and Keith Dyer, 109-137. Piscataway, NJ: Gorgias, 2017. __________. The Patristic ‘Masora’. A Study of Patristic Reading Traditions in the Near East. Corpus scriptorium christianorum orientalium (CSCO) 698; Scriptorus Syri Tomus 265. Louvain, 2020. Martin, Jean Pierre Paulin. “Tradition karkaphienne, ou la massore chez les Syriens.” JA ser. 6 vol. 14 (1869): 245–379. Sokoloff, Michael. A Syriac Lexicon. Piscataway, NJ, 2009. Teule, Herman G. B. and Carmen Fotescu Tauwinkl, Robert Bas ter Haar Romeny, and Jan J. van Ginkel, eds. The Syriac Renaissance. Eastern Christian Studies 9. Leuven, 2010.

ANNOTATIONS IN THE EARLIEST MEDIEVAL HEBREW BIBLE MANUSCRIPTS ELVIRA MARTÍN-CONTRERAS The earliest extant medieval Hebrew Bible manuscripts (late ninth and mid-eleventh) feature several innovations with respect to the Hebrew biblical textual witnesses preserved from earlier periods (the biblical scrolls found in the Qumran caves, third century BCE - first century CE). One of those features is the annotations to the biblical text found in every folio of most of those manuscripts. This article provides a comprehensive overview of these annotations: their main characteristics (location, format, presentation, content), their function, their purpose, and their origin and authorship. To end, a brief state of the art and some ideas for future research are offered.

The earliest extant medieval Hebrew Bible manuscripts (dated between the late ninth and mid-eleventh centuries and produced in the Near East—Egypt, Palestine, Syria, Tunisia and Iraq)1 feature several innovations with respect to the Hebrew biblical textual witnesses preserved from earlier periods (the biblical scrolls found in the Qumran caves, dated between the third century BCE and the first century CE). The first of these is the change in the manuscript For a description of the dated codices, twenty-one in total, see Malachi Beit-Arie, Colette Sirat, and Mordechai Glatzer, Codices hebraicis litteris exarati quo tempore scripti fuerint exhibentes (Turnhout: Brepols, 1997–2002), vol. I–III. 1

167

168

ELVIRA MARTIN-CONTRERAS

form:2 the scroll was the format previously used to write the biblical text. Since halakhah (Jewish religious law) forbids a biblical scroll from containing any form of punctuation, the Torah scrolls used in public reading in the synagogue contained only the consonants of the Hebrew text.3 The earliest extant medieval Hebrew Bible manuscripts are in the form of a codex—that is, the form of a book. For the Jews, the transition from the scroll to the codex took place several centuries later than for most everyone else in the Mediterranean area. Without any textual witnesses from earlier centuries, it is only possible to speculate about the reasons for this transition. Possible explanations include the wide use of the codex throughout the area and the importance of the Islamic context in which these Jews lived.4 The adoption of the new codex form, which was not subject to religious laws, was conducive to other innovations: the page layout introduced new elements that could now be placed alongside the consonantal text. The pages of the codices have a format different from that of the Torah scrolls. The text, arranged in two or three columns and written in square letters, bears some resemblance to text found in scrolls, but the wide margins around the text and the new elements accompanying the consonantal text in the codex make it something completely different. These elements, which cannot be written on a Torah scroll according to halakhah, are vowel points, cantillation marks, and annotations to the text. This article is devoted to the third of these elements.5 David Stern, The Jewish Bible: A Material History (Seattle and London: University of Washington Press, 2017). 3 This is similar to the way in which the biblical text is still written on scrolls used in public reading in the synagogue. 4 Geoffrey Khan, A Short Introduction to the Tiberian Masoretic Bible and Its Reading Tradition (Piscataway, NJ: Gorgias Press, 2012), 6. 5 These annotations are found in manuscripts vocalized in the three systems of Hebrew vocalization, i.e. Palestinian, Babylonian and Tiberian. The Tiberian system and its annotations eventually prevailed over the others; as such, it has a larger number of textual testimonies and is there2

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

169

ANNOTATIONS TO THE BIBLICAL TEXT6 Location The annotations are mainly found in the intercolumnar, outer, top and bottom margins of each folio (hereafter “marginal annotations”), thus sharing the page with the text arranged in the columns (fig. 1),7 which is the actual biblical text. These annotations may also be found at the end of either the individual biblical books or the collection of biblical books.

fore the best known. Consequently, this article will concentrate on the annotations found in manuscripts using this system. 6 In the field of Jewish studies and biblical studies, one technical term refers to these annotations as a group: Masora (in its narrow sense). The denominations “Masora Parva” (for the annotations placed in the intercolumnar and outer margins) and “Masora Magna” (for the annotations placed in the top and bottom margins) are also used. However, I deliberately avoid the use of these terms, bearing in mind that: (a) the term Masora has two meanings: its broader meaning includes vowels and accents, which could create confusion; (b) the terms Masora Parva and Masora Magna merely express an external-technical division of the annotations; (c) this article discusses a new methodological proposal for the study of these annotations; and (d) readers of this article may come from other fields. 7 To illustrate this article, images were chosen from four manuscripts attributed or related to the Ben Asher family, considered responsible for the textual tradition that was ultimately accepted: Or 4445 (http://www.bl.uk/manuscripts/FullDisplay.aspx?ref=Or_4445), Cairo Codex of the Prophets (photographs of C held by the Masora team at the CSIC), Aleppo Codex (http://www.aleppocodex.org/newsite/index.html) and Leningrad Codex (https://archive.org/details/Leningrad_Codex/ page/n952/mode/2up).

170

ELVIRA MARTIN-CONTRERAS

Fig. 1. The London Codex BL MS Or 4445, f. 57r

Format The form in which these annotations appear varies according to their position on the page. As a general rule, the marginal annotations and those placed at the end of each biblical book are written in tiny script (fig. 2, margins and end of the center column), while those at the end of a collection of biblical books are written in a larger script, although not as large as that of the biblical text (fig. 2, left column).

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

Fig. 2. Leningrad codex, f. 463r, Neh 13:19b-13:31 and masoretic lists

171

The annotations are most often written in straight lines, though they are sometimes also written in figured patterns. The annotations in figured patterns can be found in any of the margins of the folio, but they are especially common in the upper margin. The forms are mainly simple geometric shapes, such as triangles, semicircles, zigzags, circles or a combination thereof (fig. 3). Vegetal

172

ELVIRA MARTIN-CONTRERAS

motifs and, exceptionally, other forms (e.g. a six-pointed star, a house shape, etc.) also appear.8

Fig. 3. Cairo Codex of the Prophets, p. 30, Josh 12:7–13:2 In later biblical manuscripts, these annotations have much more complex and sophisticated designs. See Elvira Martín-Contreras, “The Image at the Service of the Text: figured Masorah in the Biblical Hebrew Manuscript BH Mss1,” Sefarad 76 (2016): 55–74. 8

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

173

The annotations at the end of a collection of books can, as an exception, also be written in figured patterns, and they appear as part of the design of the so-called carpet pages (fig. 4).

Fig. 4. Leningrad Codex (Firkovich B19), fols. 489v-490r

Function Usually, each marginal annotation is linked to one or more words of the biblical text written on the same folio. A graphic symbol—a small circle that scholars today call the circellus—is often placed over one word or between two or more words of the biblical text (fig. 5). The circellus operates as a notice, informing the reader that extra information on the word can be found in a marginal annotation.

Fig. 5. Aleppo Codex, Isa 3:4–5:2. Courtesy of the Ben-Zvi Institute, Jerusalem. Photographer: Ardon Bar Hama.

A word in the biblical text may have a circellus but no corresponding annotation; on the contrary, there may be no circellus over a word that does have an annotation with related information. Annotations are connected to their lemmas through the following: a) their placement next to the line of the text—this is the case with the annotations placed in the intercolumnar margins (fig. 6, lateral margin); and b) the repetition of the lemma in the annota-

174

ELVIRA MARTIN-CONTRERAS

tion itself. The latter technique is typically used for those annotations written in the top and bottom margins as well as for annotations found at the end of a biblical book or collection of biblical books (fig. 6, upper margin).

Fig. 6. Aleppo Codex, Judg 6:25. Courtesy of the Ben-Zvi Institute, Jerusalem. Photographer: Ardon Bar Hama.

As noted above, the notes and lists are generally related to a word contained in the biblical text of the same folio. However, on some occasions, long lists in the upper and lower margins do not fit on the corresponding page and appear on the following page and, at times, at the end of biblical books. Presentation As general rule, the information contained in the annotations is presented in a short, concise way (generally expressed using abbreviations) and, on many occasions, elliptically—that is, part of the information remains implicit. However, the way the information is provided can vary according to the location of the annotations. Those placed in the intercolumnar and outer margins have the briefest form, with the words often represented only by their initial letters. The ultimate expression of this ellipsis would be annotations that only give a Hebrew letter used as a numeral (fig. 6). There is no standardized form for these abbreviations; that is, the same word may be represented in several different ways. Neither is there a single way of expressing similar information: meth-

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

175

ods can vary between manuscripts and, sometimes, even within the same manuscript. The annotations placed in the top and bottom margins provide information in a more detailed and expanded way while maintaining conciseness and the use of abbreviations. Usually, the lemma—the repetition of the word or words from the biblical text to which the annotation refers—comes first (as in fig. 7, ‫קנּ ֹא‬, ַ qanoʾ, “I’ve been jealous,” from 1 Kings 19:10); it is then followed by the heading, which usually contains a numerical reference and mention of a particular text feature (“[is written] four times [in the whole Bible], twice [with] defective spelling [‫]קנֹּא‬, ַ and twice [with] plene spelling [‫)”]קנּוֹא‬. ַ Finally, after these elements appear the references to the biblical verses by means of catchwords (one or more words from the verse being referred to: “and their references are: 1 Kgs 19:10, 19:14, Josh 24:19, and Nah 1:2”).9

Fig. 7. Cairo Codex, p. 224, 1 Kings 19:10

The information contained in the annotations placed at the end of a collection of biblical books is arranged in a list format (fig. 8). The list is composed of the heading (where the group is delimited and the phenomenon enunciated), lemma, and catchwords. For example, one of the lists placed at the end of the Leningrad Codex (fig. 8) starts with the following heading: “list of single words which is unique with [accent] milʿel [penultimate stress] and in the rest of the Bible has [accent] milraʿ [final stress]”; it is followed by the first lemma—the word we-hishquita ‘water it’—and its catchwords: the three first words from the verse Deut 11:10. Chapter and verse enumeration did not exist in Hebrew Bible manuscripts until the 14th century. 9

176

ELVIRA MARTIN-CONTRERAS

Fig. 8, Leningrad Codex, p. 951, masoretic lists

In addition to the use of letters according to their numerical value and of abbreviated forms, other devices are also used in the annotations. Some include mnemonic devices, such as acronyms, designed to assist the user in remembering the information—for example, the mnemonic signs attached to the verses listing the names of Zelophehad’s five daughters (Num 26:3; 27:1; 36:11; Josh 17:3). Each mnemonic sequence—made up of the initial letters of the words used—contains the correct order of the names in each of the texts. For example (fig. 9), the mnemonic sequence in the annotation to Num 26:3 is ‫וֹח ׄמוֹ‬ ׄ ‫ ׄמ‬, ‫“ מחלה( ׄמ‬Mahlah”), ‫“ ונעה( וֹ‬and Noah”), ‫ׄח‬ ( ‫“ חגלה‬Hoglah”), ‫“ מלכה( ׄמ‬Milcah”), ‫“ ותרצה( וֹ‬and Tirzah”).

Fig. 9. Leningrad Codex (Firkovich B19), f. 91v, Num 26:3

The annotations are characterized by the widespread use of technical terms. Although some of these terms are already used in rabbinic literature, most of them had not been attested previously. The language used to express this information is a mixture of rabbinic Hebrew and Aramaic. Some annotations contain references to the verses first in Hebrew and then in Aramaic; on some occasions, mnemotechnical phrases in Aramaic are used to record

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

177

the information. Each Aramaic word corresponds in some fashion (not necessarily literally) to a Hebrew word in the verses containing the term to which the note refers.10 For example, in an annotation to 1 Sam 2:26 in the Aleppo Codex, the following phrase appears in Aramaic: “and the chosen lad ran and found wisdom” (fig. 10). This catchy phrase would have reminded the reader of the five occurrences of the word wa-tob, “a good,” in which it is vocalized with a qames under the letter waw (‫)וָ טוֹב‬: “and the lad”= 1 Sam 2:26; “the chosen one”= 1 Sam 9:26; “ran” = Gen 18:7; “and he found” = 1 Chr 4:40; “wisdom” = 1 Kgs 10:7.

Fig. 10, Aleppo Codex, 1 Sam 2:26. Courtesy of the Ben-Zvi Institute, Jerusalem. Photographer: Ardon Bar Hama.

Finally, the information given in these annotations can be articulated around either a specific word or a groups of words that share a common feature in the text. In the case of a specific word, the person who annotated often chooses to comment upon some aspect of a particular word, either in relation to its specific occurrence in the annotated passage or in relation to its occurrence in other places in the biblical text. In the case of a group of words with a common feature, annotations are used to draw the reader’s attention to a shared or common phenomenon among a number of different biblical passages. David Marcus, Scribal Wit: Aramaic Mnemonics in the Leningrad Codex (Piscataway, NJ: Gorgias Press, 2013). 10

178

ELVIRA MARTIN-CONTRERAS

Content Scholars have generally assumed that the content of these notes is largely numerical. However, except in the case of the calculations included in the annotations at the end of each biblical book, the enumeration is only one part of the information provided in these annotations. Along with this enumeration, additional information is provided—whether implicit or explicit—about various aspects of the biblical text with which they appear. These aspects are discussed below, under letters (a) through (i). (a) Spelling. Many of the annotations refer to the way a word is written. The most common are those indicating whether a word is written plene or defective—in other words, with or without the matres lectionis11 yôd and waw. Or, to a lesser extent, the annotations may also indicate the absence or presence of other letters and words. They may also refer to other spelling particularities, such as deviations from the usual writing, different graphic possibilities for a single word, the reversal of letters, and orthographic irregularities (an inverted nûn, dotted words, large and small letters, suspended letters, special letters, etc.). (b) Differences between what is written in the consonantal text and what is read. The biblical text contains numerous words that show a discrepancy between how they are written and how they should be read. In many of these cases, annotations mention these discrepancies and, usually, present the way in which they should be read. The medieval Masoretic text also contains several words that are not written in the text but must be read, as well as the opposite—words that are written in the biblical text but should not be read. The annotations on both phenomena state what word must be read or should not be read. For example, the word banaw, “his sons,” in 2 Kings 19:37 is not written in the biblical text apart from its vowels and accent, but a marginal annotation informs the reader that the word must be read even though it is not written (fig. 11). In the spelling of Hebrew, matres lectionis refers to the use of certain consonants to indicate a vowel. 11

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

179

Fig. 11. Leningrad Codex (Firkovich B19), f. 216v, 2 Kings 19:37

(c) Grammar. There are also a large number of grammatical annotations. Some deal with the gender and number of the words. Others state whether a word is a common noun, a proper noun or a naming place. There are also annotations that refer to or list multiple function words (conjunctions, pronouns, grammatical particles and prepositions). Many other annotations focus on syntactical questions. These may include indications of the common preposition that governs a verb and its exceptions; annotations that draw attention to peculiar constructions, usually where a syntactic oddity occurs (repetition of one word, odd sequences of particles, lack of direct or indirect object, lack of syntactical agreement, etc.); or annotations on combinations of words. (d) Morphology. Annotations that gather all the morphological variants of a root (most words in Hebrew have a three-letter root). (e) Semantic. Most of the semantic annotations indicate the different meanings of words in the Hebrew Bible that have identical spellings. There are also annotations that point out an unusual meaning (for example, a marginal annotation to the word Kenaʿanim in Job 40:30 in the Leningrad Codex: “three [times] meaning ‘merchant’”12). Others state the language in which a word is written as a way to clarify its meaning (for example, a marginal annotation to the word safar, “he had numbered,” in 2 Sam 24:10 in the Aleppo The other two instances are Gen 38:2 and Zech 14:21. In the rest of the cases, this word means “Canaanite.” 12

180

ELVIRA MARTIN-CONTRERAS

Codex says that this is the only time that the word is not in Aramaic). (f) Vocalization. These annotations usually indicate when a single word appears with a different vocalization or provide vocalization rules for when that word is affected by accents or some consonants (for example, gutturals). (g) Accentuation/stress. Annotations can indicate the position of the accent or stressed syllable within a word; others point out a sequence of accents within a verse. They can also indicate where one specific accent follows another. (h) Variants or alternative readings. These annotations gather the opinions of various schools and the different traditions related to the biblical text, such as the so-called differences between Ben Asher and Ben Naphtali13 or the differences between Eastern (Babylonian) and Western (Palestinian) Jews. They also give alternative readings in accordance with other recognized manuscripts (such as the Mugah Codex and Hilleli Codex) or only point out that a variant, other reading, or different opinion exists without mentioning the source. (i) Calculations. The annotations that appear at the end of each biblical book always contain numerical information (fig. 12), such as the total number of verses or of words and letters in each of the books.

This refers to the discrepancies between two of the most important Tiberian Masoretes, who advocated different traditions of the biblical text. 13

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

Fig. 12. Cairo Codex, end of the book of Ezekiel

181

Lists with the number of verses and sections can also be found in the annotations at the end of a collection of biblical books. Moreover, more sporadically, annotations with purely numerical content are found in the margins, such as those that indicate the midpoint of the book in terms of verses (fig. 13), words, or letter count.

Fig. 13. Cairo Codex, p.220, Mic 3:12, midpoint of the book of Micah [in terms of verses]

182

ELVIRA MARTIN-CONTRERAS

Purpose The primary purpose of these annotations is the precise preservation of the holy text.14 Beyond this general function, however, the contents of the annotations show proved useful in copying manuscripts and reading and interpreting the biblical text. The annotations that contain calculations, for example, make it possible to confirm whether the copied text is full and complete. The numerous annotations with spelling content are precise instructions about the exact way in which a word should be written—despite its oddity, apparent error, or the fact that it’s written in another way in a different passage—so that it is not corrected or homogenized when it is copied. The semantic annotations are directly relevant for the interpretation of the text. They can help to confirm an unusual meaning of a form or to distinguish the meaning when the same word has several meanings. The annotations that contain information about vowels and accents also have an exegetic function—for instance, in disambiguating identical verbal forms. Other annotations have a clear function for reading, such as those that mark a section break for public reading. Finally, some annotations have multiple functions, such as those that include the differences between the written and read form, with a double function for reading and copying. Although they indicate how a word should be read, they also indicate the exact way in which it should be copied, without being changed. Some of the lists in these annotations have also been shown to function as biblical concordances. This function predates the first Latin concordances by several centuries.15 The annotations containing variants from other manuscripts and various sources appear to indicate an intention to provide a compilation, giving as much information as possible. Although this Aron Dotan, “Masorah,” in Encyclopaedia Judaica, 2nd ed., ed. Michael Berenbaum and Fred Skolnik (Detroit, MI: Macmillan Reference USA, 2007), 612. 15 Bella Hass Weinberg, “Index Structures in Early Hebrew Biblical Word Lists,” The Indexer 22:4 (2001): 178–186. 14

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

183

information differs from the text it is linked to, the latter is not corrected. Bearing in mind the characteristics and phenomena noted, it is possible to see a clear intention to preserve the minorities—the exceptions, peculiarities and irregularities—and to prevent them from being changed. Origin and authorship The earliest material evidence for this kind of annotation and textual information dates from the ninth and tenth centuries. However, most contemporary scholars assume that this practice developed in the centuries preceding the earliest extant codices, around the sixth to seventh centuries CE. It is also assumed that the Masoretes compiled the textual information and wrote it down in the annotations. But who were the Masoretes? Almost nothing is known about them. The annotations are often anonymous. However, the few Masoretes mentioned by name in some annotations—among others, “the teachers of Tiberias,” “the men of Tiberias,” “Pinḥas, the head of the Academy” and Ben Asher and Ben Naftali—also date to a later period, i.e. the ninth to tenth centuries CE.16 It is only possible to speculate about the prehistory of these annotations before the extant codices, but this is not really necessary. Some information does exist concerning who added the annotations to some manuscripts. According to the colophons in some of these manuscripts, the annotations may have been penned (a) by the person who copied the text in the writing area—as in the case of the Leningrad Codex, composed by a single individual, Samuel ben Yaakov; or (b) by others who took part in the production of the manuscript, such as the 929 CE Pentateuch in St Petersburg (Evr. II, B 17). This scribe was Salomon ha-Levi b. R. Buya, while Some names found in two grammatical treatises also indicate a later era. Only the names mentioned in an annotation at the end of the Sefer Oklah lists in the second part of the Halle manuscript could be dated to the 2nd century CE. Elvira Martín-Contreras, “Masora and Masoretic Interpretation,” in The Oxford Encyclopedia of Biblical Interpretation, 2 vols., ed. Steven L. McKenzie (Oxford: Oxford University Press, 2013), 542–550. 16

184

ELVIRA MARTIN-CONTRERAS

the person who added the vocalization, the accentuation marks and the annotations was Ephraim b. Rabbi Buyaʿa. Plurality The fixed location of the annotations in the manuscripts may suggest that they were part of a planned layout and, therefore, that the contents of the annotations are always the same in all the biblical manuscripts. However, a comparison of the annotations in the manuscripts published to date from Cairo (C), Aleppo (A), and Leningrad (L) reveals the numerous differences between the annotations in these codices. These differences are manifested in the number of annotations, the way the information is provided (more or less explicit, different terminology, different abbreviations, etc.) and the content of the annotations. Thus, C, A and L may include the same notation with some variation; some manuscripts may include a specific annotation while others do not. In some cases, while referring to the same word, each manuscript focuses on different aspects of the word; this leads to different annotations, most of which are not contradictory or wrong. These differences show that each codex has its own set of marginal annotations and that no two sets of annotations are the same. State of the art and future research In the past, these annotations have been studied as a secondary element of the biblical text, not on their own merits. In the absence of direct textual testimony to these textual annotations outside of the earliest Hebrew Bible manuscripts, the theories used to explain the transmission of the Hebrew biblical consonantal text have also been used to explain the origin, transmission and function of the marginal annotations. In other words, these annotations have been studied only for their content, while other aspects have been left aside.

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

185

The works published over the last few decades in the field of Masoretic studies17 have notably expanded our knowledge of these annotations; and, as a result, some ideas about them and their history have been called into question. Moreover, many and important differences have been identified among the manuscripts studied to date, not only in the way the annotations articulate information, but also within the content of the annotations. All this not only confirms the plural nature of the annotations, but it has also raised new questions about the origins and transmission of these notes, the relationship between the biblical text and the marginal annotation in each manuscript, and the function of the marginal annotations beyond their primary purpose. Given these new findings, it appears that the methodological approach used to date is not the most suitable method to handle the diversity and plurality of these annotations. It is necessary to explore other ways of approaching them. This will mean reexamining the role played by these annotations in each manuscript and identifying and evaluating the differences between these annotations in multiple biblical codices. My proposal18 is to study these annotations in the context of the manuscripts containing them. In this model, each manuscript is considered a single witness, and the external evidence of the manuscripts—formerly excluded—is incorporated into the analysis. Moreover, the annotations are approached in their medieval manuscript culture.19 In short, this proposed method of study will take into consideration that these annotations are part of a medieval text See Elvira Martín-Contreras, “The Current State of Masoretic Studies” Sefarad 73 (2013): 423–458. 18 See Elvira Martín-Contreras, “Rethinking the Annotations Placed in the Margins of the Medieval Hebrew Biblical Manuscripts,” (in press). 19 Manuscript culture is a way of representing the world in accord with contemporary—as opposed to historical—perception; see Stephen G. Nichols, “What is a Manuscript Culture?” in The Medieval Manuscript Book: Cultural Approaches, ed. Michael Johnston and Michael Van Dussen, Cambridge Studies in Medieval Literature 94 (Cambridge: Cambridge University Press, 2015). 17

186

ELVIRA MARTIN-CONTRERAS

transmitted in manuscript form—in the codex—and that they make up one of the texts of a multiple-text manuscript.20 In this model, the notion of paratextuality21 as it was applied in medieval manuscript culture22 will be taken into consideration. These marginal annotations fall under the category of biblical paratext; consequently, they may have affected the reading of the text. Hopefully, this approach can point to some new possibilities for our understanding of these intriguing marginal annotations.

BIBLIOGRAPHY

Beit-Arie, Malachi, Colette Sirat, and Mordechai Glatzer. Codices hebraicis litteris exarati quo tempore scripti fuerint exhibentes. Turnhout: Brepols, 1997–2002.

Cooper, Charlotte E. “What is Medieval Paratext?” Marginalia 19 (2014): 37–50.

Dotan, Aron. “Masorah.” In Encyclopaedia Judaica, second edition, edited by Michael Berenbaum and Fred Skolnik. Detroit, MI: Macmillan Reference USA, 2007.

Friedrich, Michael, and Cosima Shwarke (eds.). One-Volume Libraries: Composite and Multiple-Text Manuscripts. Studies in Manuscript Cultures 9. Berlin, Boston: De Gruyter, 2016.

Genette, Gerard. Paratexts: Thresholds of Interpretation. Translated by Jane E. Lewin. Cambridge: Cambridge University Press: 1997. “This term designates a codicological unit ‘worked’ in a single operation’ (Gumbert) with two or more texts or a ‘production unit’ resulting from one production process delimited in time and space”; see Michael Friedrich and Cosima Shwarke (eds.), One-Volume Libraries: Composite and Multiple-Text Manuscripts, Studies in Manuscript Cultures 9 (Berlin, Boston: De Gruyter, 2016), 15–16. 21 Gerard Genette, Paratexts: Thresholds of Interpretation, trans. Jane E. Lewin (Cambridge: Cambridge University Press: 1997). 22 See Charlotte E. Cooper, “What is Medieval Paratext?,” Marginalia 19 (2014): 37–50. 20

ANNOTATIONS IN MEDIEVAL HEBREW BIBLES

187

Khan, Geoffrey. A Short Introduction to the Tiberian Masoretic Bible and Its Reading Tradition. Piscataway, NJ: Gorgias Press, 2012.

Marcus, David. Scribal Wit: Aramaic Mnemonics in the Leningrad Codex. Piscataway, NJ: Gorgias Press, 2013. Martín-Contreras, Elvira. “The Current State of Masoretic Studies.” Sefarad 73 (2013): 423–458.

———. “Masora and Masoretic Interpretation.” In The Oxford Encyclopedia of Biblical Interpretation. 2 volumes. Edited by Steven L. McKenzie. Oxford: Oxford University Press, 2013. ———. “The Image at the Service of the Text: Figured Masorah in the Biblical Hebrew Manuscript BH Mss1.” Sefarad 76 (2016): 55–74. ———. “Rethinking the Annotations Placed in the Margins of the Medieval Hebrew Biblical Manuscripts.” In press.

Nichols, Stephen G. “What is a Manuscript Culture?” In The Medieval Manuscript Book: Cultural Approaches, edited by Michael Johnston and Michael Van Dussen. Cambridge Studies in Medieval Literature 94. Cambridge: Cambridge University Press, 2015. Stern, David. The Jewish Bible: A Material History. Seattle and London: University of Washington Press, 2017.

Weinberg, Bella Hass. “Index Structures in Early Hebrew Biblical Word Lists.” The Indexer 22:4 (2001) 178–186.

AN ILLUMINATING SCRIBE: THE ʿARZADASHT OF JAʿFAR BĀYSUNGHURĪ AND ITS WEALTH OF INFORMATION SHIVA MIHAN Our knowledge of royal libraries is even less than our knowledge of their artists and is limited to information found in a few art-historical sources and tazkiras. Exceptionally, an art-historical document, written in 830/1427 and entitled the ʿArża-dāsht, gives us unique insights into the production progress of Persian manuscripts at the atelier of the Timurid Prince Baysunghur (1399–1433) in Herat. This single folio provides rare evidence of the artistic activities and operation of a Timurid workshop and offers a wealth of tantalising information, including the speed and sequence of manuscript production.

I will open this essay with a review of the literature, after which I will discuss the meaning of the ʿArża-dāsht’s title and then briefly describe the document’s physical traits. Then I will turn to the order in which the paragraphs should be read, according to which I provide a complete translation with explanations for the meaning of some obscure terms. I next consider the date, mainly on the basis of the colophons in the manuscripts mentioned by Jaʿfar, the library’s head, and estimate the work rates of the scribes.1 This article builds on some aspects of the subject discussed in my PhD dissertation. For further information about topics not treated in detail here, see Shiva Mihan, “Timurid Manuscript Production: The Scholar1

189

190

SHIVA MIHAN

INTRODUCTION

The Timurid dynasty, founded by Timur (1336–1405), ruled in Iran from 1370 to 1507.2 Under his successor Shahrukh (r. 1405–1447), the capital moved from Samarqand to its centre in Herat, which became a cradle of cultural activity for the remainder of the century. Art and architecture flourished to the point that this period is generally regarded as the pinnacle of cultural production. Shahrukh’s sons received a highly advanced education, and among them were noted bibliophiles and great patrons of art, literature and science. Shahrukh’s third son and heir apparent, Prince Baysunghur (1397–1433), established his celebrated library and atelier in his residence at the Bāgh-i Safīd (White Garden) in Herat around 1420.3 Our limited knowledge about this multifaceted establishment derives mainly from a petition-report that bears no signature but, judging by its contents, is addressed to the prince. It was almost certainly written by his chief librarian and project overseer, Jaʿfar Tabrīzī (active 1413–1452),4 who is also widely known by the sobriquet al-Baysunghuri. This fragment is now preserved in the Topkapi Palace Library, Album H. 2153, f. 98r (fig. 1). Based on the information it provides, we know that it was written in the second half of Ramadan 830 (July 1427). From Jaʿfar’s petition-report, known as the ʿArża-dāsht, we understand that the Garden accommodated several other buildings, some of which were under construction at that time. Those included a ṣūrat-khāna (picture-house), a kutub-khāna (books-house/library) and a kulang-khāna (crane-house). We know the kutub-khāna was the atelier founded for copying and decorating manuscripts; according to Jaʿfar, it was founded as a work-place for painters and scribes who had recently ship and Aesthetics of Prince Bāysunghur’s Royal Atelier (1420–1435)” (Ph.D. diss., University of Cambridge, 2018a), ch. II. 2 For an overview of the dynasty, see H.R. Roemer, “The Successors of Timur,” in The Cambridge History of Iran, ed. Peter Jackson and Lawrence Lockhart (Cambridge: Cambrige University Press, 1986). 3 Khwāndamīr, Tārīkh-i Ḥabīb al-siyar, ed. J. Humāʾī (Tehran, 1380/2001), 3:623. 4 Mehdi Bayani, Ahvāl va āthār-i khushnivīsān (Tehran, 1363/1984), 1:118.

AN ILLUMINATING SCRIBE

191

moved in. We can deduce from some verses in the Jung-i Marāthī, a book of elegies composed on Baysunghur’s death, that the ṣūratkhāna was a picture gallery containing figurative wall paintings.5 From its name and the prince’s passion for falconry, the kulang-khāna must have been a place for breeding and raising cranes as reared game for hunting. The ʿArża-dāsht, as “the only known report of the ‘work in progress’ in a Timurid workshop,”6 provides a snapshot of the situation at a particular moment in time; without a precursor or a follow up, it raises as many questions as it answers. These issues include the sequence and logical connection between the paragraphs, its precise dating, the meaning of some ambiguous technical terms, and the correct identification of the manuscripts and artists mentioned. It also provides clues to questions of more fundamental importance, such as the speed and sequence of manuscript production.

HISTORIOGRAPHY

First introduced by Zeki Velidi Toğan in the fourth Congress of Turkish History held in Ankara in 1948, the document was not published until 1976 in Istanbul.7 A year after that, Parsay Quds published the document in the journal Hunar va Mardum as “a document related to artistic activities in the Timurid period in Baysunghur’s kitābkhāna in Herat”; in the journal, he discussed Baysunghur’s library and the historical sources that provide further information about it. He dated the document based on the completion dates of four manuscripts and discussed the data it provides regarding the names and number of artists and manuscripts, portable objects and

The only copy of this significant source is preserved in the Tabriz National Library, no. 2967. 6 Yves Porter, “Models, Sketches, and Pounced Drawings in the Diez Albums,” in The Diez Albums: Contexts and Contents, ed. Julia Gonnella et al. (Leiden, Boston: Brill, 2017), 373. 7 Mustafa Kemal Özergin, “Temürlü sanatına ait eski bir Belge: Tebrizli Ca’fer’in bir arzı,” Sanat Tarihi Yıllığı 6 (1976): 471–518. 5

192

SHIVA MIHAN

architecture.8 Peter Andrews, an expert in tents, translated the tent section and elucidated the technical terminology in 1978.9 The following year (1979), Aslanapa translated the ʿArża-dāsht’s paragraphs on bookbinding.10 Thomas Lentz studied the ʿArża-dāsht in his PhD dissertation (1985) as part of album H. 2153. His analysis covers the artists and ongoing projects, and he identifies four manuscripts (one of them incorrectly).11 He provides a terminus post quem of 830 for the ʿArża-dāsht on the basis of the completion dates for Dublin Gulistān and Vienna Humāy u Humāyūn.12 Bernard O’Kane translated the section on buildings (ʿimārāt) in his book on the architecture of Khurasan (1987) and discussed the data pertinent to architectural projects and collaboration between craftsmen. He did not specify the author of the “report” and said only that it was written by “the head of a Timurid library,” who referred to himself as banda-yi kamtar va ẕarra-yi aḥqar. O’Kane referred to “what must have been royal patronage” but stopped short of naming the patron. However, he stated that the extensive library staff “could only have been employed by a ruler or one of his sons”; and when discussing the demarcation of labour between

Ahad Pārsāy Quds, “Sanadī marbūṭ ba faʿāliyyat-hā-yi hunarī-i dura-yi Timūrī dar kitābkhāna-yi Bāysunghurī-i Harat,” Hunar va Mardum 15:175 (1356/1977): 42–50. 9 Peter A. Andrews, “The Tents of Timur: An Examination of Reports on the Qurıltay at Samarkand, 1404,” in Arts of the Eurasian Steppelands, ed. Ph. Denwood (London: SOAS, 1978), 143–87. 10 Oktay Aslanapa, “The Art of Bookbinding,” in The Arts of the Book in Central Asia, 14th–16th Centuries, ed. Basil Gray & O.F. Akimushkin (London: Shambhala Publications, 1979), 59–92, esp. 60. 11 The manuscripts Lentz identifies are: Shāhnāma (Tehran, Golestan Palace Library), Gulistān (Dublin, Chester Beatty Library), Tārīkh-i Ṭabarī (St. Petersburg, Public Library), Dīvān of Khwājū Kirmānī (Humāy-u Humāyūn, Vienna, National Library). Thomas W. Lentz, “Painting at Herat under Baysunghur ibn Shah Rukh” (Ph.D. diss., Harvard University, 1985), 149. I shall argue that the last one is a misidentification. 12 Ibid., 148–54. 8

AN ILLUMINATING SCRIBE

193

different teams of craftsmen, he provided evidence from Gauharshād’s mosques in Herat and Mashhad.13 In 1989, Wheeler Thackston published a complete translation of the entire document in his book A Century of Princes, together with several other petitions and documents. He identified most of the manuscripts mentioned in the document and provided useful explanations for some unclear items and terms in the notes.14 Later, several other scholars used and discussed fragments of the ʿArża-dāsht for different purposes, especially to clarify the structure of Baysunghur’s courtly foundation for artistic activities (mainly referring to it as the kitābkhāna) and, in a few cases, to date the document. Akimushkin’s article on the prince’s workshop usefully discusses the size and characteristics of such petitions in the Timurid era.15 In 2001, Thackston published a revised translation with a different order of reading and minor alterations to the wording.16 Sergei Tourkin provided a somewhat different translation of the first section while making a case for a revised order of reading the ʿArża-dāsht (2003).17

THE ʿARŻA-DĀSHT : REPORT OR PETITION

There have been some scholarly differences over how to translate the ʿArża-dāsht and over its purpose. Akimushkin believes that “it is a kind of official report (something like a modern formal acBernard O’Kane, Timurid Architecture in Khurasan (Costa Mesa, CA: Mazda Publishers, 1987), 40–42. 14 Wheeler M. Thackston, A Century of Princes: Sources on Timurid History and Art (Cambridge, MA: The Aga Khan Program for Islamic Architecture, 1989), 323–27. 15 Oleg F. Akimushkin, “The Library-Workshop (Kitābkhāna) of Bāysunghur-Mīrzā in Herat”, Manuscripta Orientalia, 3, no. 1 (1997): 14–24, esp. 22. It is also discussed in I. Afshar, “Arża-dāsht-i mulāzimān-i kitabḫana mansub ba Jaʿfar-i Tabrīzī,” in Barg-i bībargī (Tehran, 1378/1999), 277–85. See also Porter, “Models, Sketches, and Pounced Drawings,” 356–57. 16 Wheeler M. Thackston, Album Prefaces and Other Documents on the History of Calligraphers and Painters (Leiden, Boston, Cologne: Brill 2001), 43–46. 17 Sergey Tourkin, “Another Look on the Petition (‘Arḍadāšt) by Jaʿfar Bāysunġurī Addressed to His Patron Bāysunġur b. Šāhruḫ b. Tīmūr,” Manuscripta Orientalia, 9, no. 3 (2003): 34–38. 13

194

SHIVA MIHAN

count).”18 Tourkin discusses it at some length in his article: “In all studies and translations published to date, the document is called in English ‘petition’, while nothing is petitioned, or asked in it, except for the eternal power of the addressee which is nothing more but a formal prayer. In [the] Persian original, the document is entitled ʿArża-dāsht, and a more exact translation into English of its title should be ‘report’.”19 Thackston changed the title from “petition” (1989) to “report” (2001) in his revised translation.20 The verb ʿarż(a) dāshtan means to state and explain a current situation to someone of a higher status or to present a complaint or request to a higher authority. The term ʿarża-dāsht and its close equivalents, ʿarż-dāsht, ʿarż-i ḥāl, ʿarż al-ḥāl and ʿarīża, were used either for requesting financial aid or making a petition.21 Evidently Jaʿfar was the overseer of work on manuscripts and other portable objects, constructions, and tents; and in all likelihood, he was the controller of the purse strings. The ʿArża-dāsht is both a report of ongoing projects and an implicit petition, which makes the title he gave to his missive doubly meaningful. The closest Jaʿfar came to an overt request was for Baysunghur’s presence, which he expressed in elevated language as follows: “Petition from the most humble servants of the royal library whose eyes are as expectant of the dust from the hooves of the regal steed as the ears of those who fast are for the cry of Allāhu akbar, and whose joyful and gleeful shout of ‘praise be onto God, who hath taken away sorrow from us! Verily our Lord is ready to forgive and to reward’ [Qurʾan, 35:34] reaches the apex of the celestial sphere.”22 Akimushkin, “The Library-Workshop,” 15. Tourkin, “Another Look,” 34. 20 Tourkin presented a version of his paper in 2000, and it reached publication only in 2003. See n. 3. 21 Other examples of ʿarża-dāshts as petitions are found in album H. 2153. Some are explicitly requesting physical materials, as in ff. 98v and 119v, and some express non-materialistic requests, as in f. 137r. See Thackston, Album Prefaces, 47. Bihzād’s petition to Shāh Tahmāsp is found in Ebadollah Bahari, Bihzad, Master of Persian Painting (London: I.B. Tauris, 1996), Appendix 1. 22 Thackston, Album Prefaces, 43. 18 19

AN ILLUMINATING SCRIBE

195

We do well to regard the document as an artfully constructed piece by a man of refined culture—a calligrapher, an illuminator and a poet. The first section begins with a passage of elevated prose and ends with a poem of invocation by Jaʿfar. It does indeed contain, as has been noted by others, a progress report for the works in and around the atelier; however, there are a few further points to notice. Firstly, the section on the manuscripts is organised not around the works, but around individuals. Although there must have been a much larger number of people working at the atelier, the ʿArża-dāsht only reports on the main artists who were very well known to the patron.23 There seems to be a strong possibility of hierarchy, as we know that the most important artists are named at the top. Secondly, there is a sense that Jaʿfar is indicating where the atelier’s resources have been expended. This is endorsed by the fact that the three paragraphs on the kishtīs, i.e. the atelier’s gold consumption,24 were separated off from the other paragraphs dedicated to manuscripts. Could this not be a subtle request for gold supplies? We cannot agree with Tourkin that nothing is requested except in the opening of the “Report.” In the opening paragraph, a parallel is drawn between the hunger of those fasting during Ramadan and the hunger of the atelier staff for the prince’s presence. Metaphorically, the entire atelier might have been in some need of the absent prince—possibly to approve their activities, but also to refresh their supplies. Yves Porter discusses the absence of apprentices and junior artists in the ʿArża-dāsht: “The activity of the kitābkhāna also had to cope with the training of younger artists and to elaborate didactic materials for this purpose. This ‘pedagogic’ facet does not appear at all in the ʿArża-dāsht.” Porter, “Models, Sketches, and Pounced Drawings,” 357. It is obvious that there would have been other unmentioned staff taking care of less specialist jobs, such as lapis grinders. The fact that there were 75 painters dyeing a tent gives an indication of the scale of the atelier. 24 Kishtī was a ship-like container used by artists for measuring and storing gold. See Shiva Mihan, “On the Meaning of a Fifteenth-Century Technical Term in a Timurid Document Associated with Prince Baysonghor’s Library in Herat,” Iran 54, no. 2 (2016): 15–20. 23

196

SHIVA MIHAN

To sum up, the ʿArża-dāsht can be understood as a progress “report,” an itemized record of who has done what; but it is also an indirect record of payments to the artists and artisans and a subtle “petition” for precious supplies, framed by prayers for Baysunghur’s presence. One way or another, the atelier’s generous benefactor was sorely missed. After the prince’s death, years later, Jaʿfar would mourn poetically: “O, illuminator! No more will your problem be solved with gold.”25

PHYSICAL APPEARANCE

The ʿArża-dāsht is written on a light, creamy paper measuring 135 x 460 mm.26 Written in black nastaʿlīq, the document is mainly arranged in three columns and comprises an opening, 3 sections and 30 paragraphs. Following the introduction (4 lines), the first section comprises 23 paragraphs containing the names of 24 artists, 9 manuscripts and two objects; the second section contains 6 paragraphs27 on royal constructions, and the third section (9 lines) on a trellis tent is written in a single paragraph. Paragraphs in the first section are for the most part dedicated to named artists; and in the architecture section, they are dedicated to distinct constructions. The appearance of petitions varied in different periods, but the typical format remained almost unchanged: a rectangular scroll-like folio often long and rolled.28 One of the oldest petitions, dated circa Tabriz National Library, no. 2967, f. 3r. Jaʿfar beautifully deployed the word “solve” (hall) in a figure of speech, both for solving the illuminator’s problem and for making a gold solution to apply in manuscripts. ‫ای ﻣﺬھّﺐ ﻣﺸﮑﻠﺖ از زر ﻧﮕﺮدد ﺣﻞ دﮔﺮ‬ 26 The paper is very likely a product of Herat. See Özergin, “Temürlü sanatına ait eski bir Belge,” 487–88. 27 Previous scholars have tended to divide it into 7 paragraphs. See the “Order of Reading” below. 28 Decrees sometimes had a similar appearance. For example, a Mongol decree dated 630/1232–33 granting a village in Ardabil appears in the same long scroll format. See Gottfried Herrmann, “Ein früher persischer Erlass,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 144, no. 2 (1994): 284–300. Decrees from the Ilkhanid period are found in Paul Pelliot, “Les documents mongols du Musée de Teherān,” Athār-é Īrān 1 (1936): 37–44. 25

AN ILLUMINATING SCRIBE

197

806/1403 and written in the early nastaʿlīq script, is known as the ʿarża-dāsht of Shaykh Dursūn, now in the Topkapı Palace Library.29 Almost all such documents include the name of God at the top, even if shortened to hū. Then the decree or scroll begins, generally with the name of the intended recipient and a laudatory passage.30 In his article on Baysunghur’s Kitābkhāna (1997), Akimushkin highlighted the fact that the ʿArża-dāsht lacks bismillāh at the beginning and noted that the absence of the sender and receiver’s names in an official correspondence could indicate that the heading and footing were damaged in the course of time and therefore cut off. Alternatively, they may have been simply cut off as specimens of Jaʿfar’s handwriting (qiṭʿa) and perhaps set into an album for a connoisseur.31 This significant document must have been dated as well, and it is not certain if it was shortened for one of the reasons given by Akimushkin or if it was trimmed at the top and bottom to accommodate its placement in the Topkapı album. The latter possibility agrees with the cavalier manner in which other pieces were inserted into album H. 2153. Considering Akimushkin’s suggestion that the length of such documents was one ẕarʿ (cubit), which is 49.9 cm, this would imply that the ʿArża-dāsht has lost approximately four centimetres of its initial length, for whatever reason.32 Four centimeLajos Fekete and György Hazai, eds., Einführung in die persische Paläographie: 101 persische Dokumente = Madḫal-i pāliʼūgrāfī-i fārsī : ṣad u jak sanad-i fārsī (Budapest: Akadémiai Kiadó, 1977), Plate 2; his article also includes many other examples of decrees and petitions from different periods. In Plate 1, another decree is seen, from Mīrānshāh to Sulṭān Sanjar and the officials of Nakhjavān to inform them of a village bestowed upon Shaykh Dursūn as suyurghal, dated 27 Ramadan 798/4 July 1396. 30 The images of two long scrolls are reproduced in Norah M. Titley, Persian Miniature Painting and Its Influence in the Art of Turkey and India (London: British Library Publishing Division, 1983), 20–21. 31 Akimushkin, “The Library-Workshop,” 22. 32 Ibid. I have not come across examples of scrolls measuring one cubit in length. Herrmann’s book contains 28 examples of scrolls, but there is no obvious preferred length; indeed, one is 46 cm. long and another 45 cm. (numbered 14 and 10, respectively). Gottfried Herrmann, Persische Urkunden der Mongolenzeit: Text- und Bildteil (Wiesbaden: Harrassowitz, 2004). 29

198

SHIVA MIHAN

tres would have been sufficient space for both Prince Baysunghur’s name and titles and Jaʿfar’s signature and date. In any case, despite the absence of the author’s signature and the recipient’s name, the data given in the document reveal their identity beyond question, whereas the date remains obscure and requires some argumentation. I have discussed the order in which the document should be read elsewhere and suggested that the best order to read the ʿArżadāsht would not be totally row-wise or column-wise, but rather scroll-wise: a different reading sequence taking into account cohesive sections and the reading of the three columns from right to left, which logically follows from the scroll format of the document and the way it would be read while unrolling.33

DATING THE ʿARŻA-DĀSHT

Proposals for the date of the ʿArża-dāsht range between 830 and 833, but I argue that a more definite dating is possible through a reconsideration of the information provided in the document. Parsay Quds suggests Ramadan 830, based on the first paragraph of the ʿArża-dāsht, in which Jaʿfar addresses the prince in laudatory language: “Petition from the most humble servants of the royal library, whose eyes are as expectant of the dust from the hooves of the regal steed as the ears of those who fast are for the cry of Allahu akbar….”34 This specification of the month relies upon the assumption that it is a reference to the breaking of the fast of Ramadan, rather than merely a literary device expressing eagerness. Parsay Quds suggested the year should be 830 (1427) according to the completion date of Gulistān (830) and Rasāyil (830), as well as two Shāhnāmas in progress (831 and 833).35 At the other end of the range, in his article about Baysunghur’s library, David Roxburgh more or less accepts the date 833/1430, previousFor the full discussion of the order of reading as well as a revised translation and identification of the manuscripts, see Mihan, “Timurid Manuscript Production,” 22–26. 34 Thackston, Album Prefaces, 43. 35 The date for the Shāhnāma copied by Muḥammad Muṭahhar is 833, but Parsay Quds incorrectly states 831. 33

AN ILLUMINATING SCRIBE

199

ly suggested by Thackston,36 but modifies it slightly to propose a date either late in 832 or early in 833.37 In the first line after the opening paragraph, the author begins with an account of Amīr Khalīl’s work on the Gulistān of Saʿdī, which is now in the Chester Beatty Library (Per. 119), Dublin, dated 830/1427. “Amīr Khalīl has finished the waves in two sea scenes of the Gulistān and will begin to apply colour.”38 Elsewhere, he states that Khwāja ʿAṭā, who was the rulings maker, had finished his work on the Gulistān; and Khwāja Ghiyāth al-Dīn was involved in producing a scene in the same manuscript.39 The scribe, Jaʿfar Baysunghuri, does not mention the month in the colophon of the Chester Beatty Gulistān, but it is clear that he was done with the text by the time of the ʿArża-dāsht because he does not mention it under his own works in progress in the final paragraphs of the first section. The production of the rulings was also finished, and the illustrator and illuminator were completing the remaining parts of the manuscript’s decoration. Thomas Lentz seems to suggest as a “tentative date” 830 or 831. He certainly uses the completion date of the Gulistān to provide a terminus post quem for the document.40 He also mentions the comThackston, Album Prefaces, 44, n. 15. David J. Roxburgh, “Baysunghur’s Library: Questions Related to Its Chronology and Production,” Journal of Social Affairs 18, no. 72 (2001): 11–41, esp. 24–26. 38 Thackston, Album Prefaces, 44. 39 Making rulings was a workshop profession in its own right. See N. Māyil Haravī, Kitābʹārāyī dar tamaddun-i Islāmī: majmūʻa-yi rasāʼil dar zamīna-yi khūshnivīsī, murakkab-sāzī, kaghaz-garī, tazhīb va tajlīd: ba inẓimām-i farhang-i vāzhigān-i niẓām-i kitāb-ārā’ī (Mashhad, 1372/1993), 607. 40 The fact that Lentz uses the Gulistān’s date to provide a tentative date of 830 or 831 for the ʿArża-dāsht shows that, unlike Thackston, he does not consider that the account of that manuscript was referring to subsequent “repair” work; he states that Ghiyāth al-Dīn was “revising an architectural subject that was rejected from the Gulistān.” Lentz, “Painting at Herat,” 151. He sees no problem with the Golestan Palace Shāhnāma being completed three years after the putative date of the Arża-dāsht (830): “It is entirely possible that the deluxe surviving Shāhnāma took three years to complete,” p. 150. 36 37

200

SHIVA MIHAN

pletion date of the Humāy-u Humāyūn of Khwājū Kirmānī in Vienna (831/1427–28),41 but he does not seem to have been aware of the other manuscript of Khwājū Kirmānī, the Kulliyyāt copied by Muḥammad Muṭahhar in 829/1426 in Herat, now at the Malek National Library, no. 5963.42 In fact, the Dīvān of Khwājū in the ʿArżadāsht refers to the Malek manuscript, not the Vienna one. Humāy-u Humāyūn is one of Khwājū Kirmānī’s five mathnavīs and could not be referred to by Jaʿfar as the Dīvān. In addition, the accounts of Maulānā Shams and Maḥmūd working on the groundwork of one and 10 lauḥs of the Dīvān, respectively, confirm that the manuscript in hand was not the Humāy-u Humāyūn. The latter begins with a subsequently added sarlauḥ on the first folio (in the shikasta script) and does not contain any other sarlauḥ, lauḥ or illumination, except for 69 small inscription boxes appearing as headings to introduce the next passage of poetry. We understand from the ʿArża-dāsht that the copying of the Dīvān was finished and the rulings were made, but the decorations were in progress.43 By contrast, the Malek Library copy of the Dīvān contains 14 illuminated sarlauḥs, all extensively gilded. Although the number of illuminations (14) exceeds the amount referred to by Jaʿfar in relation to Maḥmūd (10), it is very likely that four further sarlauḥs were done by Maulānā Shams (one of them is mentioned in the account of a kishtī). It is noticeable that the last three headings are smaller along the top edge by about one third. We can now turn to other manuscripts that shed further light on the dating question. The scribe, Muḥammad Muṭahhar, copied three manuscripts for Baysunghur’s library: the Ẓafar-nāma, completed in Shaʿbān 828; the Dīvān of Khwājū Kirmānī, completed in Shaʿbān 829; and a dual-text volume containing the Shāhnāma and Humāy-u Humāyūn of Khwājū Kirmānī (Herat, 831/August 1427), Nationalbibliothek, Vienna, cod. NF. 382. See Lentz, “Painting at Herat,” 149. 42 Lentz states that Khwājū’s work mentioned by Jaʿfar is “possibly” the Humāy-u Humāyūn in Vienna. Lentz (1985); ibid., 149. 43 The ʿArża-dāsht describes this sarlauḥ at the stage of bi būm risānīdan. The term means “to complete the primary steps before applying colours,” including copying the design and inscription, outlining the patterns, in some cases applying gold and burnishing it, and preparing the work for the next phase of illumination. 41

AN ILLUMINATING SCRIBE

201

Khamsa. The latter was completed in 833/1430, but the preface of the Shāhnāma (dated 829) suggests that the scribe could have begun his work in late 829 or early 830.44 According to the ʿArża-dāsht, two Shāhnāma copies were in the process of production at that time: the Shāhnāma of the Golestan Palace Library, ms. no. 716 (hereafter, GB), and the dual-text manuscript of the Malek Library, ms. no. 6031 (hereafter, MB), containing the Shāhnāma and Niẓāmī’s Khamsa. Both manuscripts were completed in 833/1430 and were only in the early stages of preparation at the time of the ʿArża-dāsht.45 Jaʿfar states that he himself had transcribed three and a half quires of the Golestan Shāhnāma. With eight folios in each quire and pages set in thirty-one rows and six columns, this means he could have written no more than around 5,000 verses by then. On the other hand, according to Jaʿfar, the scribe of the Malek manuscript, Muḥammad Muṭahhar, had transcribed 25,000 verses of the Shāhnāma. If we reckon the whole poetry of the Baysunghuri edition to be about 60,000 verses, Muḥammad Muṭahhar was not yet half way through the Shāhnāma and still had to start the second part of the codex, the Khamsa of Niẓāmī (around 30,000 verses). Both GB and MB share the Baysunghuri Preface, dated 829/1426, which indicates the finishing date of the compilation of the Preface.46 Presumably, the transcription of the new edition began directly afterwards, with the unillustrated volume, MB. The starting date of copying MB, therefore, should have been around late 829 or, at the latest, early 830. Before that, as noted above, Muṭahhar was working on the Dīvān of Khwājū. Bearing in mind the number of verses copied by the time of the ʿArża-dāsht (25,000), its date can be estimated to have been 830/1427, especialFor further discussion, see Mihan, “Timurid Manuscript Production,” chapter V. 45 A comprehensive account of the Malek Library manuscript is found in Shiva Mihan, “The Baysunghuri Manuscript in the Malek Library,” in Shahnama Studies III: The Reception of the Shahnama, ed. Charles Melville and Gabrielle van den Berg (Leiden, Boston: Brill, 2018), 373–419. 46 I point out the discrepancies between the two copies in detail in Mihan, “Timurid Manuscript Production,” chapter V. 44

202

SHIVA MIHAN

ly considering the fact that Jaʿfar himself had finished with the Gulistān (830/1427) only recently (the rulings having been in progress), was 1/12th of the way through GB (to finish in 833), and had started copying the Nuzhat al-arvāḥ, as he testifies.47 The chart below displays the manuscript titles mentioned in the ʿArża-dāsht and the stage of progress. The completion dates in the last column help us to establish the date of the ʿArża-dāsht itself. Table.1 – Progress report and completion dates of the manuscripts mentioned in the ʿArża-dāsht.

Manuscript title Gulistān49

dāsht

Account(s) in the ʿArżaAmīr Khalīl has finished waves in two sea scenes of the Gulistān and will begin to apply colour.

Completion date of identified manuscript48 830/1426–27

Jaʿfar copied the Naṣāʾiḥ-i Iskandar (a selected text from the Sirr alAsrār in Arabic) in 829/1426 (Chester Beatty Library, Ar. 4183). On my research trip to Afghanistan in 2016, I was notified that a well-known antique dealer had sold three Baysunghuri manuscripts over the past decades: a copy of the Ẕakhīra-yi Khārazm-shāhī, dated c. 831–32; a Dīvān of Kamāl Khujandī, dated 830; and a Tārīkh including the shamsa of Baysunghur’s ex libris, which lacked the last folio(s). To what extent his intriguing information is accurate, I cannot be certain. However, I have discussed a copy of the Dīvān of Kamāl Khujandī, copied by Jaʿfar alBāysunghurī (British Library, Or. 15395) with no date or place provided in the colophon; see ibid., 176–77. One would wonder if the Tārīkh manuscript with an illuminated Shamsa, lacking the final few folios, is one and the same as the Tārīkh-i Vaṣṣāf, now at the Keir Collection in Dallas Museum of Art. See an account of that manuscript in Shiva Mihan, “Tārīkh-i Vaṣṣāf: A Treasure from Prince Bāysunghur’s Library,” in Nāma-yi Bāysunghur, ed. K. Afzali, no. 1 (Herat, 2019), 91–116. 48 Completion date is most often the date a scribe finished transcription. In general, manuscripts could well be under production long after the text was copied; in the case of Baysunghur’s atelier, though, as the ʿArża-dāsht attests, the calligraphy and transcription were usually done in parallel to ruling, illumination and illustrations. 49 Gulistān of Saʿdī, Per. 119, Chester Beatty Library. 47

AN ILLUMINATING SCRIBE

Dīvān of Khwājū50

Tārīkh (Zubdat al-tawārīkh)51

At present [Ghiyāth al-Dīn] is busy with [redoing] one architectural scene in the Gulistān that was rejected. Maulānā Shihāb (…) is busy with another architectural scene of the Gulistān. Khwāja ʿAtā has finished the [decorative] elements of the Gulistān. Maḥmūd has completed the groundwork for seven out of ten cartouches for the Dīvān of Khwājū and is busy with the rest. Khwāja ʿAtā, the ruling maker, has finished (…) the Dīvān of Khwājū (…). Maulānā Shams has finished one kishtī and has done the groundwork on one cartouche for the Dīvān of Khwājū. Khwāja ʿAtā, the ruling maker, has finished Maulānā Saʿd alDīn’s Tārīkh (…). Khwāja ʿAtā (…) has done the groundwork for two out of three lauḥs in the Tārīkh that Maulānā Saʿd al-Dīn has copied.

203

Shaʿbān 829/JuneJuly 1426

Undated (c. 829/1426)

Kulliyyāt of Khwājū Kirmānī, Malek Library, no. 5963. Zubdat al-tawārīkh, St Petersburg National Library of Russia, Dorn 268. Although it does not include the scribe’s signature, I have argued elsewhere that the scribe was very likely Saʿd al-Mashhadī and that it is the same Tārīkh mentioned in the ʿArża-dāsht. For my discussion and more about the enigmatic figure of Saʿd al-Mashhadī, see Shiva Mihan, “Hidden from Scholarly Eyes for a Century: An Unknown Bāysunghurī Manuscript Sheds New Light on His Court and Library,” Journal of the Royal Asiatic Society 29, no. 1 (2019): 51–73. 50 51

204

SHIVA MIHAN

Shahnāma [+ Khamsa] (MB)52 Shahnāma (GB)53

Maulānā Muḥammad Muṭahhar has finished writing 25,000 verses in the transcription of the Shahnāma. Maulānā ʿAlī (…) has begun to work on the design of the Shāhnāma Preface. Maulānā Shihāb has applied gold to the Preface,54 four cartouches, the finials and the Preface’s illustration and has outlined the eight lobes of the shamsa within the Preface. Maulānā Qavām al-Dīn has completed the arabesque (islīmī) margins of the front cover for the Shahnāma’s binding, and has taken up the chisel for the revelry scene on the doublure. The background is nearly one-third done, the back, flap and flapspine have been attached, and the groove has been drawn. Khwāja ʿAtā, the ruling maker, (…) is busy with the Shahnāma.55 [Jaʿfar] has finished writing three and a half quires of the Shahnāma (…).

Shaʿbān 833/April– May 1430 (end of the Shahnama) 5 Jumādā I, 833/30 January 1430

Shāhnāma and Khamsa, Malek Library, ms. no. 6031. Shāhnāma, Golestan Palace Library, ms. no. 716. 54 He is very likely referring to the entire Preface (dībācha) here and the gilding between the lines. 55 The rulings of the Shāhnāma could well refer to MB; but since the rulings of GB are more elaborate, I believe this is an account of work on GB. 52 53

AN ILLUMINATING SCRIBE Rasāyil56

Rasāyil-i khaṭṭ-i Khwāja

Tārīkh-i Ṭabarī57 Nuzhat alarwāḥ59

Khwāja Ghiyāth al-Dīn has progressed in two scenes of the Rasāyil to the point of [working on] the faces and another scene is nearly there. Maulānā Shams is one section short of completing the facsimile (naql) of the calligraphy treatises of the late Khwāja. Hājjī Maḥmūd has done the groundwork and outlining of the front doublure for the binding of the facsimile of the treatises (Rasāyil) and is busy drawing outlines. Khwāja Maḥmūd has finished the front and back [boards] of the binding for the Khwāja’s calligraphy treatises and is busy with the flap and the flap-spine. Maulānā Quṭb has copied ten quires of the Tārīkh-i Ṭabarī.58 [Jaʿfar] has begun the Nuzhat al-arwāḥ.

205 Ramadan and 1 Shawwāl 830/June and 26 July, 1427 Unknown

20 Jumādā II, 833/21 March 1430 Lost; formerly in the Yeni Cami Library

Known as the Baysunghur Anthology, Berenson Collection, Villa I Tatti, Florence. In my forthcoming article, I have argued this manuscript is not complete and have hypothesised that dispersed folios in three other collections were once parts of the Rasāyil. See Shiva Mihan, “The Rasāyil: Reconstruction of a Dismembered Masterpiece from Prince Bāysunghur’s Atelier.” Muqarnas, forthcoming. 57 Tarjuma-yi Tārīkh-i Ṭabarī by Bal‘amī, St Petersburg National Library of Russia, PNS 49. 58 Depending on the quire size, ten sections were somewhere between 80 to 100 folios. This means he was done with approximately 1/6th of the 500 folios in the manuscript. 59 I have not yet been able to locate Jaʿfar’s copy of the Nuzhat al-arvāḥ. It was in Yeni Cami Library up until around 1908. See Clément Huart, Les Calligraphes et les Miniaturistes de l’Orient Musulman (Paris: E. Leroux, 1908), 97. See Thackston, Album Prefaces, 44. See also Lentz, “Painting at Herat,” 150. 56

206

SHIVA MIHAN

Thackston suggests that everything datable in the “report” indicates a date of circa 833.60 The chart above, however, shows that the manuscripts dated c. 833–34 were in the early phases of production, either being copied or having their beginning folios illuminated. On the other hand, the manuscripts dated 829–30 were done with copying and in their final phases of preparation. Ergo, this document most likely dates back to 830/1427. The other fact that fortifies this conclusion is that there is no mention of subsequently completed manuscripts in the ʿArża-dāsht; and if it was written in late 832 or 833, not only is it unlikely that Jaʿfar would have mentioned the long-completed Gulistān and Dīvān of Khwājū, but it is unlikely he would have made no mention of the following manuscripts, each of which would have been in progress at that time. Table.2 – Manuscripts completed in and after 833/1430 but not mentioned in the ʿArża-dāsht.

Later mss.

Scribe

Kalīla-u

Shams

Completion date

Kunūz al-wadīʾa + Al-faraj baʿd al-shidda62

Saʿd al-Mashhadī (Saʿd al-Dīn)

28 Rabīʿ II, 833

Tārīkh-i Jahān-gushāy64

Jaʿfar

Saʿd al-Mashhadī

834

Jaʿfar

Shaʿbān 834

Dimna61

Kalīla-u Dimna63 Tārīkh-i

Iṣfahānī65

Muḥarram 833

late Rabīʿ I, 834

Thackston, Album Prefaces, 44, n. 15. Kalīla-u Dimna, Topkapi Palace Library, R. 1022. 62 Kunūz al-wadīʿa min rumūz al-ẕarīʿa ilā makārim al-sharīʿa by Ẓāfir b. Shams al-dīn Ḥasan, and (a translation of) Al-Faraj baʿd al-shidda wa al-żīqa by Ḥusayn b. Asʿad b. al-Ḥusayn al-Dihistānī, Yeni Cami Library (Suleymaniye), ms. no. 937. This manuscript carries the ex libris of Bāysunghur and is still bound in its original binding, which bears the prince’s name. I have discussed this neglected codex in Mihan, “Timurid Manuscript Production”. 63 Kalīla-u Dimna, Topkapi Palace Library, H. 362. 64 Tārīkh-i Jahān-gushāy, St Petersburg National Library of Russia, PNS. 233. Bāysunghur’s ex libris at the beginning of this manuscript is pasted over. 60 61

AN ILLUMINATING SCRIBE Kulliyyāt-i Khwāja ʿImād66 Laylī-u

Majnūn67

Chahār Maqāla68

Tārīkh-i Vaṣṣāf69

Aẓhar

26 Dhu’l-ḥijja 834

al- Sulṭānī

5 Rabīʿ I, 835

Jaʿfar

unknown

207

835 835

I have argued that the earlier date, 830, is preferable to the later estimates of 832 or 833 on the grounds of a close analysis of the identification of the manuscripts mentioned. Given Jaʿfar’s allusion to the Ramadan fast in the opening, it is possible to propose a more specific date for the document, namely Ramadan 830 (as Parsay Quds originally suggested). What is more, we know that Baysunghur left Herat on the first of Shaʿbān to accompany his father to Samarqand. When they reached Balkh, at Ulugh Beg’s request, Shahrukh dismissed Baysunghur, who returned to Herat, settling there on 21 Ramadan.70 The ʿArża-dāsht was very likely written in the middle of that month, not long before Baysunghur’s return to Herat on 21 Ramadan 830.

Tārīkh sinī mulūk al-ʿarż wa’l-anbiyāʾ by Hamza b. Ḥasan Iṣfahānī, known as Tārīkh-i Iṣfahānī, British Library, Or. 2773. 66 Kulliyyāt of ʿImād al-Dīn Faqīh Kirmānī, Oxford Bodleian Library, Elliott 210. 67 Laylī-u Majnūn by Niẓāmī, Metropolitan Museum of Art, 1994.232. 68 Majmaʿ al-navādir by Aḥmad b. ʿUmar b. ʿAlī Niẓāmī ʿArūżī Samarqandī, known as Chahār Maqāla, Turkish and Islamic Art Museum (TIEM), ms. no. 1954. 69 Tārīkh-i Vaṣṣāf by ʿAbd Allāh b. Fażl Allāh Shīrāzī, known as Vaṣṣāf al-Ḥażrat, The Keir Collection, Dallas Museum of Art (DMA), VII: 62. 70 ‘Abd al-Razzāq Samarqandī. Maṭlaʻ-i saʻdayn va majmaʻ-i baḥrayn, ed. ʻA. Navāʾī (Tehran, 1383/2004): 2/1:386. Khwāndamīr, Tārīkh-i Ḥabīb al-siyar, ed. J. Humāʾī (Tehran, 1380/2001), 3:617. 65

208

SHIVA MIHAN

AN ILLUMINATING SCRIBE

209

210

SHIVA MIHAN

SCRIBAL WORK RATES

Another way to estimate the date of the ʿArża-dāsht is to examine those manuscripts being copied at the time of the report for which Jaʿfar supplied detailed information on progress and for which we have specific completion dates in the manuscripts that have come down to us, allowing us to construct estimates for when copying might have begun. In such cases we can also derive the apparent rates at which the named scribes worked. I use months in the Gregorian calendar for the computations in what follows. 1– [Tarjuma-yi] Tārīkh-i Ṭabarī: assume that Quṭb al-Dīn began work on Tārīkh-i Ṭabarī in September 1426, soon after he completed Tāj al-maʿāthir,71 and that he worked at a constant rate until the colophon date of 21 March, 1430 (Norouz!). That would mean he took 43 months for the entire project. Since the manuscript contains 62.5 quires in total, Quṭb’s average work rate on the Tārīkh-i Ṭabarī would have been 1.4 quires per month (62.5 quires ÷ 43 months = 1.4 quires/month). According to Jaʿfar, Maulānā Quṭb had copied ten quires out of 62.5 quires of the Tārīkh-i Ṭabarī by the time of the report, which means he would have been 1/6th of the way through after 7 months. If he started copying in September 1426, the 1/6th mark would have fallen in April 1427/Jumādā II, 830 (3 months prior to Ramadan 830). Since we believe the ʿArża-dāsht was written in Ramadan 830, we suspect Quṭb al-Dīn began a few months later than we first suggested. 2– Malek Library codex containing the Shāhnāma and Khamsa: To calculate the date Muḥammad Muṭahhar could have started copying the Shāhnāma part, we must first compute the completion date for his previous works: the Dīvān of Khwājū Kirmānī (Malek Library, no. 5963) and a dual-text codex containing the Ẓafar-nāma of Shāmī and the Ẕayl-i Ẓafar-nāma of Ḥāfiẓ Abrū (Nuruosmaniye, Suleymaniye Library, ms. no. 3267). We suppose he began copying Tāj al-maʿāthir fī Tārīkh by Tāj al-Dīn Muḥammad b. Ḥasan Niẓāmī al-Nishābūrī, completed on Thursday, 25 Shawwal 829/31 August 1426, St. Petersburg State University, ms. no. 578. 71

AN ILLUMINATING SCRIBE

211

the Dīvān of Khwājū soon after he completed the Ẓafar-nāma on 9 Shaʿbān 828/26 June 1425. 3– The Kulliyyāt of Khwājū has lost a few folios from the end, but it must have contained around 1275 pages in total.72 The text consists of 14 treatises, each beginning with a sarlauḥ and ending with a colophon. Most colophons simply mark the end of the section, with two exceptions. The first (page 924) provides the date of copying, and the second (page 1070) provides the scribe’s identity, date and place of completion. The latter gives the date as 1 Shaʿbān 829/7 June 1426. The total number of verses up to this colophon is 34075 verses, which the scribe completed in about 11 months, beginning in late June 1425.73 This means his work rate was around 100 verses a day. Based on this rate, we can calculate the completion date of the entire manuscript. Pages 1071 to 1275 contain the equivalent of 5125 verses, which should have taken around another 50 days to complete.74 Let’s assume that directly after completing the Dīvān of Khwājū, Muḥammad Muṭahhar began copying MB in early Shaʿbān 829/mid-June 1426 and worked on MB at a constant rate until the colophon date, May 1430. That means he took 47 months for the entire project (Shāhnāma and Khamsa), 90,000 verses total. At the time of the ʿArża-dāsht, he had copied 25,000 out of 90,000 verses, which means he would have been working on the project for 13 months so far (i.e. 47 months × 25,000 ÷ 90,000). On that basis, we would estimate the date for the ʿArża-dāsht to be mid-July 1427/Ramadan 830. Given all the approximations involved, this is Since the manuscript is paginated, I refer to page numbers rather than folio numbers. 73 The page layout varies between two and four columns in the manuscript, with some pages in prose. Pages 1 to 773 contain 25 verses in two columns; pages 775 to 1070 contain 50 verses per page (25 rows in four columns); and pages 1071 to 1275 contain 25 verses per page (or the equivalent in prose). 74 I have argued elsewhere that he added the prose treatises to the end of the Dīvān some time after the ʿArża-dāsht. Mihan, “Timurid Manuscript Production,” 94–95. 72

212

SHIVA MIHAN

a remarkably accurate match for our hypothesis regarding the date when Jaʿfar wrote his petition-report. Muḥammad Muṭahhar’s average work rate on MB was 1.4 quires per month, which we compute as follows: in 47 months he completed 1017 pages with 16 pages per quire, that is, 63.5 quires (1017 ÷ 16). 63.5 quires ÷ 47 months = roughly 1.35 quires per month. We can convert that figure to verses per day as follows: there are 30 days in a month and an average of 90 verses per page and 16 pages per quire, so the average number of verses per day equals 1.35 × 90 × 16 ÷ 30 = 65 verses per day. This is a drop in absolute verse count relative to his work on the Dīvān (100 verses per day), which is explicable considering the increase in the quality of calligraphy and the reduction in pen size in MB. 4– GB: At the time of the ʿArża-dāsht, Jaʿfar records that he was 1/12th of the way through. The period from July 1427/Ramadan 830 (ʿArża-dāsht date) to GB’s completion date of 5 Jumādā I, 833/30 January 1430 is 33 months. If Jaʿfar had worked at a constant rate, he would have begun around 3 months prior to the ʿArża-dāsht, i.e. in April 1427/Jumādā II, 830. This does not contradict his having begun work after completing the Gulistān (for which we have only the year of completion, 830). Jaʿfar’s average work rate on GB was 1.2 quires per month, computed as follows: 690 pages ÷ 16 pages per quire = 43 quires in total, which he completed in 36 months, then 43 ÷ 36 = 1.2 quires per month. We can convert that figure to verses per day using the same computation as above: 1.2 × 90 × 16 ÷ 30 = 60 verses per day. This is marginally slower than the work rate of Muḥammad Muṭahhar, but we know that Jaʿfar was also working on other manuscripts in parallel, including Nuzhat al-arwāḥ; and he also had supervisory duties to fulfil.

AN ILLUMINATING SCRIBE

213

Table. 3 – Manuscripts enabling an estimation of scribal work rates. Work and scribe in the ʿArża-

dāsht

Shāhnāma copied by Muḥammad Muṭahhar (MB) Shāhnāma copied by Jaʿfar Tabrīzī (GB) [Tarjuma-yi] Tārīkh-i Ṭabarī copied by Quṭb al-dīn

Completion date (colophon)

Estimate for the time at which copying began

Scribal work rate 1.35 quires per month

5 Jumādā I, 833/30 Jan 1430.

Following completion of the Dīvān of Khwājū on 1 Shaʿbān, 829/7 June 1426 (i.e. early Shaʿbān 829/mid June 1426).75 Some time in 830 after completing the Gulistān.

Soon after 25 Shawwāl 829/31 Aug 1426, since that is the colophon date for his previous work, Tāj al-maʿāthir.76

1.4 quires per month

Shaʿbān 833/April– May 1430

20 Jumādā II 833/21 March 1430

1.2 quires per month

The estimates calculated above, ranging from 60 to 100 verses per day, show that the average copying rate of a royal scribe was around one page of a large folio manuscript each day. As mentioned above, this rate must be considered along with other factors, such as scribes copying more than one work simultaneously; scribes working as illuminators as well, or having other responsibilities at the workshop (supervision, in the case of Jaʿfar); the familiarity of scribes with the text in hand; the luxuriousness of the manuscript; and the quality of calligraphy required.77 This rate is supKulliyyāt Khwājū Kirmani, no. 5963, Tehran Malek Library. What initially was the Dīvān of Khwājū became a Kulliyyāt when the prose treatises were appended at a later date. See Mihan, “Timurid Manuscript Production,” chapter III, 88–95. There is a colophon at the end of the Dīvān (poetry section), dated 1 Shaʿbān 829. 76 Tāj al-maʿāthir, no. 578, State University, St. Petersburg. 77 No need to mention the obvious possibilities of the scribes having a few days off occasionally, or becoming unwell once in a while. 75

214

SHIVA MIHAN

ported by another petition found in the same album, H. 2153, f. 137v, which gives a similar average number of verses in a day: 80 verses for mathnavīs and 50 for ghazals. “Apprentices are as yet beginners: if they make haste in writing it impedes their progress. Now, since an order to be obeyed by all the world has been issued, of mathnavī’s eighty bayts, while of ghazals fifty bayts can be written well in a day”.78 The increase in experience of the professional scribes seems to have been counterbalanced by the demands of producing high quality, princely commissions. The figures estimated from the ʿArża-dāsht are only based on a few examples and cannot be taken to be uniformly applicable, but they can be set alongside other efforts to discuss scribal rates of work. One useful comparison can be found in Sheila Blair’s book, Islamic Calligraphy, where she discusses the average work of Shāh Maḥmūd Nishābūrī on a copy of Niẓāmī’s Khamsa for Shāh Tahmāsp. He copied 400 pages (formatted as 21 lines and four columns) in three and half years, which is the equivalent of 20 to 30 verses per day.79 According to the Bābur-nāma, his predecessor, Sulṭān ʿAlī Mashhadī, also copied at a similar daily rate: 30 lines for Sultan Ḥusayn and 20 lines for Amīr Alīshīr Navā’ī.80 A study of the scribal work rates on the Quran manuscripts by Francoise Déroche provides a time estimate for copying the Quran. Déroche states: “If we take what I called the standard Ottoman Qur’ān as the basis for an estimate, it means that the last mentioned calligrapher [Kayiszadeh Hafiz Osman Efendi] wrote daily from 3 to 4 pages of 15 lines,” which is 45 to 60 verses.81 A rate Thackston, Album Prefaces, 48. Sheila Blair, Islamic Calligraphy (Edinburgh: Edinburgh University Press, 2006), 434. For the details of the story, see ʿAbd al-Razzāq Samarqandī (1383/2004), 2/1:589–90. See also Khwāndamīr (1380/2001), 3:616–17. 80 Thackston, A Century of Princes, 267. 81 Francoise Déroche, “The Copyists’ Working Pace: Some Remarks towards a Reflexion on the Economy of the Book in the Islamic World,” in Theoretical Approaches to the Transmission and Edition of Oriental Manuscripts: Proceedings of a symposium held in Istanbul March 28–30, 2001, eds. Judith Pfeiffer and Manfred Kropp (Beirut: Orient-Institut, 2007), 209. 78 79

AN ILLUMINATING SCRIBE

215

closer to the copying speeds at Baysunghur’s workshop is found in the Javāhir al-Akhbār by Pīr Budāq Munshī Qazvīnī, which refers to Yāqūt Mu‘taṣimī (d. 1298) copying 100 Arabic verses a day for the caliph and 70 verses for the viziers.82 Akimushkin et al. state that in general, the normal daily production was between 160 and 210 verses; and for works in prose, it varied between 50 and 100 lines.83 Muhammad Isa Waley discusses copyists’ working pace in his 2018 article, in which he studies the colophons of several Mathnavī manuscripts from various libraries. He concludes that the average working pace in cases that were not unusual (i.e. not having been prepared in haste or as a working copy for the scribe’s own use) was between 130 to 250 verses a day.84 Jean Chardin, who travelled in Iran in the 17th century, reports that scribes could hardly earn a sufficiency and states that if they were real experts and worked with no interruption, they could not copy more than 500 to 600 verses (couplets) per day.85 However, scribes could copy at a much faster speed in unusual instances. Francis Richard mentions a beautiful manuscript transcribed in the hand of ‘Alī b. Muḥammad in 820/1417, commissioned by Sayyid Shams al-Din Bukhārī (BnF, Persan 266). The scribe copied 273 folios of 100 verses per folio to finish in only two weeks.86 Isa Waley provides another such example: Ḥāfiẓ Déroche, “The Copyists’ Working Pace,” 209, quoting Akimushkin et al. , 49. 83 Akimushkin et al., 49. 84 Muhammad Isa Waley, “How Long Did It Take to Copy an Islamic Manuscript? Some New Data from Multi-Colophon Manuscripts,” in Research Articles and Studies in Honour of Iraj Afshar, ed. Ibrahim Chabbouh and Francoise Déroche (London: Al-Furqan Islamic Heritage Foundation, 2018), 174–194. 85 John Chardin, Voyage du Chevalier Chardin en Perse et autres lieux de l’Orient (Paris: Le Normant, 1811), 281–82. Both Déroche and Isa Waley quote Chardin in their articles. Déroche, “The Copyists’ Working Pace,” 203, and Isa Waley, “How Long,” 190. 86 Francis Richard, Catalogue des Manuscrits Persans (Paris: Bibliothèque Nationale, 1989), 277. Déroche, “The Copyists’ Working Pace,” 210, states that the scribe copied 450 lines per day. Isa Waley, “How Long,” 82

216

SHIVA MIHAN

Muḥammad Muḥsin of Sind copied an average of 900 verses per day (BL, Or. 16170).87 One might wonder about these figures in relation to the speed of writing recorded for Maulānā Maʿrūf Baghdādī, who used to routinely copy 500 verses a day and in one instance is recorded as copying 1500 verses between dawn and dusk in a single day. First, Maʿrūf was the master of masters: he was the teacher of Shams alDīn Muḥammad Haravī, Bāysunghur’s calligraphy teacher. In Qāżī Aḥmad Qumī’s words, he “was the coryphaeus of the calligraphers of his time and a rarity of the ages.”88 Secondly, the fact that he could routinely copy 500 verses a day is recorded as indicative of his extraordinary skill. The copying of 1500 verses “executed in all elegance and perfection” in a single day was so unprecedented that a huge crowd gathered in the city square to witness such a marvel, and “Mīrzā Iskandar bestowed numerous gifts upon him.”89 The Gulistān-i Hunar emphasises that transcribing so many verses in a day was a unique skill and could take place only rarely. “It is reported that for a bet (ba-da‘vā), Maulānā Sīmī in one day composed and wrote two thousand verses, which is beyond the capacity of any poet or scribe. For the text on his signet ring he wrote the following verse and had the engraver cut it: One day, in praise of the shah of pure nature Sīmī recited and wrote two thousand verses…”90

It is clear from Qāżī Aḥmad’s account that performances of such superhuman productivity occurred exceptionally and, even for a 190, refers to the same manuscript but exaggerates a little to assume its scribe copied 2000 lines a day. 87 Isa Waley, “How Long,” 188–89. He also mentions an Ottoman scribe who could write 1000 lines a day. See his n. 24 for references. 88 Aḥmad Ibrāhīmī Ḥusaynī, Calligraphers and Painters, trans. V. Minorsky (Washington, D.C.: Smithsonian Institution, 1959), 64–65. Samarqandī, Maṭlaʻ-i saʻdayn va majmaʻ-i baḥrayn, ed. ʻA. Navā’ī (Tehran, 1383/2004), 289. 89 Ḥusaynī, Calligraphers and Painters, 65. 90 Qāżi Aḥmad Qumī, Gulistān-i Hunar, ed. A. Suhaylī Khwānsārī (Tehran, 1383/2005), 58–59. Translated by Minorsky in Ḥusaynī, Calligraphers and Painters, 125.

AN ILLUMINATING SCRIBE

217

master calligrapher like Sīmī, only when he sought to demonstrate his unique prowess. In short, the ʿArża-dāsht, long recognised as an extremely valuable source for better understanding the work carried out at the famous atelier of Prince Baysunghur, is still susceptible to new analysis and provides information that has not been fully exhausted in previous studies. In particular, more precise identification of the manuscripts to which it refers helps clarify the time at which it was written as a report on the progress of the leading artists in the atelier and a request for new resources. Furthermore, more precise dating allows us to formulate examples of the speed of the artists’ work and to reassess the scale of the projects being undertaken in the atelier at this specific moment in time.

BIBLIOGRAPHY

Afshār, Īraj. “Arża-dāsht-i mulāzimān-i kitābkhāna mansūb ba Ja‘far-i Tabrīzī.” In Barg-i bībargī. Tehran, 1378/1999. Akimushkin, Oleg F. “The Library-Workshop (Kitābkhāna) of Bāysunghur-Mīrzā in Herat.” Manuscripta Orientalia 3, no. 1 (1997): 14–24.

Akimushkin, Oleg F., A.B. Khalidov, E.A. Rezvan, Edward Tyomkin. De Bagdad à Ispahan. Paris: Electa, 1994.

Andrews, Peter A. “The Tents of Timur: An Examination of Reports on the Qurıltay at Samarkand, 1404.” In Arts of the Eurasian Steppelands, edited by Philip Denwood. London: SOAS, 1978.

Aslanapa, Oktay. “The Art of Bookbinding.” In The Arts of the Book in Central Asia, 14th–16th Centuries, edited by Basil Gray & O.F. Akimushkin. London: Shambhala Publications, 1979.

Bahari, Ebadollah. Bihzad, Master of Persian Painting. London: I.B. Tauris, 1996.

Bayānī, Mehdi. Ahvāl va āthār-i khushnivīsān. Tehran, 1363/1984.

Blair, Sheila. Islamic Calligraphy. Edinburgh: Edinburgh University Press, 2006.

Chardin, John. Voyage du Chevalier Chardin en Perse et autres lieux de l’Orient. Paris: Le Normant, 1811.

218

SHIVA MIHAN

Déroche, Francoise. “The Copyists’ Working Pace: Some Remarks towards a Reflexion on the Economy of the Book in the Islamic World.” In Theoretical Approaches to the Transmission and Edition of Oriental Manuscripts: Proceedings of a symposium held in Istanbul March 28–30, 2001, edited by Judith Pfeiffer and Manfred Kropp. Beirut: Orient-Institut, 2007. Fekete, Lajos and György Hazai, eds. Einführung in die persische Paläographie: 101 persische Dokumente = Madḫalii pālʼūgrāfī-ji fārsī : ṣad u jak sanad-i fārsī. Budapest: Akadémiai Kiadó, 1977.

Herrmann, Gottfried. “Ein früher persischer Erlass.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 144, no. 2 (1994): 284– 300.

———. Persische Urkunden der Mongolenzeit: Text- und Bildteil. Wiesbaden: Harrassowitz, 2004.

Huart, Clément. Les Calligraphes et les Miniaturistes de l’Orient Musulman. Paris: E. Leroux, 1908.

Ḥusaynī, Aḥmad Ibrāhīmī. Calligraphers and Painters. Translated by V. Minorsky. Washington, D.C.: Smithsonian Institution, 1959.

Isa Waley, Muhammad. “How Long Did It Take to Copy an Islamic Manuscript? Some New Data from Multi-Colophon Manuscripts.” In Research Articles and Studies in Honour of Iraj Afshar, edited by Ibrahim Chabbouh and Francoise Déroche. London: Al-Furqan Islamic Heritage Foundation, 2018.

Khwāndamīr. Ḥabīb al-sīyar. Edited by M. Dabīr Sīyāqī. Tehran, 1353/1974.

———. Tārīkh-i Ḥabīb al-siyar. edited by J. Humā’ī. Tehran, 1380/2001.

Lentz, Thomas W. “Painting at Herat under Baysunghur ibn Shah Rukh.” Ph.D. diss., Harvard University, 1985.

Māyil Haravī, Najīb. Kitābʹārāyī dar tamaddun-i Islāmī: majmūʻa-yi rasāʼil dar zamīna-yi khūshnivīsī, murakkab-sāzī, kaghaz-garī, tazhīb va tajlīd: ba inẓimām-i farhang-i vāzhigān-i niẓām-i kitāb-ārā’ī. Mashhad, 1372/1993.

AN ILLUMINATING SCRIBE

219

Mihan, Shiva. “On the Meaning of a Fifteenth-Century Technical Term in a Timurid Document Associated with Prince Baysonghor’s Library in Herat.” Iran 54, no. 2 (2016): 15–20.

———. “The Baysunghuri Manuscript in the Malek Library.” In Shahnama Studies III: The Reception of the Shahnama, edited by Charles P. Melville & Gabrielle van den Berg. Leiden, Boston: Brill, 2018: 373–419. ———. “Timurid Manuscript Production: The Scholarship and Aesthetics of Prince Bāysunghur’s Royal Atelier (1420–1435).” Ph.D. diss., University of Cambridge, 2018.

———. “Hidden from Scholarly Eyes for a Century: An Unknown Bāysunghurī Manuscript Sheds New Light on His Court and Library.” Journal of the Royal Asiatic Society 29, no. 1 (2019): 51–73.

———. “Tārīkh-i Vaṣṣāf: A Treasure from Prince Bāysunghur’s Library.” In Nāma-yi Bāysunghur, edited by K. Afzali. Herat, 2019: 91–116.

———. “The Rasāyil: Reconstruction of a Dismembered Masterpiece from Prince Bāysunghur’s Atelier.” Muqarnas, forthcoming.

O’Kane, Bernard. Timurid Architecture in Khurasan. Costa Mesa, CA: Mazda Publishers, 1987.

Özergin, Mustafa Kemal. “Temürlü sanatına ait eski bir Belge: Tebrizli Ca’fer’in bir arzı.” Sanat Tarihi Yıllığı 6 (1976): 471– 518.

Pārsāy Quds, Aḥad. “Sanadī marbūṭ ba faʿāliyyat-hā-yi hunarī-i dura-yi Timūrī dar kitābkhāna-yi Bāysunghurī-i Harat.” Hunar va Mardum 15:175 (1356/1977): 42–50.

Pelliot, Paul. “Les documents mongols du Musée de Teherān.” Athār-é Īrān 1 (1936): 37–44.

Porter, Yves. “Models, Sketches, and Pounced Drawings in the Diez Albums.” In The Diez Albums: Contexts and Contents, edited by Julia Gonnella, Friederike Weis, and Christoph Rauch. Leiden, Boston: Brill, 2017.

220

SHIVA MIHAN

Qāżi Aḥmad Qumī. Gulistān-i hunar. Edited by A. Suhaylī Khwānsārī. Tehran, 1383/2005.

Richard, Francis. Catalogue des Manuscrits Persans. Paris: Bibliothèque Nationale, 1989. Roemer, H.R. “The Successors of Timur.” In The Cambridge History of Iran, edited by Lawrence Lockhart and Peter Jackson. Cambridge: Cambridge University Press, 1986.

Roxburgh, David J. “Baysunghur’s Library: Questions Related to Its Chronology and Production.” Journal of Social Affairs 18, no. 72 (2001): 11–41. Samarqandī, ʿAbd al-Razzāq. Maṭlaʻ-i saʻdayn va majmaʻ-i baḥrayn. Edited by ʻA. Navā’ī. Tehran, 1383/2004.

Thackston, Wheeler M. A Century of Princes: Sources on Timurid History and Art. Cambridge, MA: The Aga Khan Program for Islamic Architecture, 1989.

———. Album Prefaces and Other Documents on the History of Calligraphers and Painters. Leiden, Boston, Cologne: Brill, 2001.

Titley, Norah M. Persian Miniature Painting and Its Influence in the Art of Turkey and India. London: British Library Publishing Division, 1983.

Tourkin, Sergey. “Another Look on the Petition (‘Arḍadāšt) by Ja‘far Bāysunġurī Addressed to His Patron Bāysunġur b. Šāhruḫ b. Tīmūr.” Manuscripta Orientalia 9, no. 3 (2003): 34–38.

ANNOTATION PRACTICES IN A SYRIAC EXEGETICAL COLLECTION (MS VAT. SYR. 103) MARION PRAGT This paper explores the annotation practices of the Collection of Simeon, a Syriac compilation of commentaries on scripture from the ninth century, based on its oldest manuscript, Vat. Syr. 103. The colophons and notes preserved in the manuscript are investigated for what they may reveal

about the history of the Collection’s composition and use. By way of case-study, the paper then focuses on three

marginal notes attributed to Hippolytus that have been added by the manuscript’s scribe and annotator Simeon. The paper examines the function of these extracts in the margin of the Collection’s commentary on the Song of Songs and the interests they may reflect. It is argued that the notes add to the Miaphysite character of the Collection and shed light on the reception of and continued interaction with Greek Christian exegesis in the Syriac tradition.

“I, Simeon, have added out of my diligence all explanations which are inscribed in the margin of the book.”1 This is how the copyist and annotator Simeon presented himself in the colophon of the socalled Collection of Simeon, a copiously annotated Syriac collection of commentaries on scripture. This paper focuses on the peritextual elements preserved in its oldest manuscript and Simeon’s auto1

Vat. Syr. 103 f. 371r.

221

222

MARION PRAGT

graph: Vat. Syr. 103.2 As such, the paper is inspired by the growing attention for the study of texts as material objects and “reception artefacts.”3 In this paper, I first investigate the colophons and notes preserved in the manuscript for what they may reveal about the aims with which the Collection was composed and how it was read and used. By way of case study, I then examine several of the marginal notes added to the commentary on the Song of Songs. For this commentary, the Collection of Simeon mainly depends on Gregory of Nyssa’s Homilies on the Song of Songs. Gregory’s work was translated into Syriac in the late fifth or sixth century and circulated in both full and abridged versions, which have not yet been edited or fully studied. This paper concentrates on three extracts from Hippolytus that Simeon added. These extracts were also transmitted in Syriac outside of the Collection of Simeon and have in that form been The Collection of Simeon has also been preserved in one eleventhcentury and, partly, in three modern manuscripts. Dirk Kruisheer, “Reconstructing Jacob of Edessa’s Scholia,” in The Book of Genesis in Jewish and Oriental Christian Interpretation. A Collection of Essays, ed. Judith Frishman and Lucas Van Rompay, TEG 5 (Leuven: Peeters, 1997), 188 n. 5; Bas ter Haar Romeny, “The Identity Formation of Syrian Orthodox Christians as Reflected in Two Exegetical Collections,” Parole de l’Orient 29 (2004), 108. 3 Book history, the transmission and use of manuscripts and their material aspects have been studied especially in the context of Western medieval studies. For example, for an overview of perspectives on medieval manuscript cultures, see Michael Johnston and Michael van Dussen (eds.), The Medieval Manuscript Book. Cultural Approaches (Cambridge: Cambridge University Press, 2015). For a recent study on Jewish and Christian manuscripts approached from the perspective of Material or New Philology, see Liv Lied Ingeborg and Hugo Lundhaug, “Studying Snapshots. On Manuscript Culture, Textual Fluidity, and New Philology”, in Snapshots of Evolving Traditions: Jewish and Christian Manuscript Culture, Textual Fluidity, and New Philology, ed. Liv Ingeborg Lied and Hugo Lundhaug (Berlin: De Gruyter, 2017), 1–19. Furthermore, with specific regard to Syriac, Heal provides an extensive bibliography on Deir al-Surian and the study of Syriac manuscripts in Kristian S. Heal, “Catalogues and the Poetics of Syriac Manuscript Cultures,” Hugoye: Journal of Syriac Studies 20.2 (2017), 375–417. 2

ANNOTATION PRACTICES

223

studied as early witnesses of the history of Christian doctrine. However, this paper examines their function in the margin of the Collection’s commentary on the Song and the interests they may reflect.

1 THE COLOPHON AND NOTES OF VAT. SYR. 103: TRACES OF THE MANUSCRIPT'S COMPOSITION AND USE

The Collection of Simeon contains a commentary on difficult words in the Old Testament, based mainly on the writings of Ephrem and Jacob of Edessa, and an interpretation of the New Testament largely relying on John Chrysostom.4 While the Collection may, in this respect, be regarded as a “multiple-text manuscript” designed to bring together material from diverse origins,5 at the same time it follows a Vat. Syr. 103 f. 1v. Dirk Kruisheer, “Ephrem, Jacob of Edessa, and the Monk Severus. An Analysis of Ms. Vat. Syr. 103, ff. 1–72,” in Symposium Syriacum VII, ed. René Lavenant (Rome: Pontificio Instituto Orientale, 1998), 600. Vat. Syr. 103 is available among the digitized manuscripts of the Vatican Library: https://digi.vatlib.it/view/MSS_Vat.Syr.103.pt.2. For an overview of the study of Syriac manuscripts, see Françoise BriquelChatonnet, Pier Giorgio Borbone, and Ewa Balicka-Witakowska, “Syriac Codicology,” in Comparative Oriental Manuscript Studies: An Introduction, ed. Alessandro Bausi et al. (Hamburg: Tredition, 2015), 252–266. 5 The term “multiple-text manuscript” was introduced by Michael Friedrich and Cosima Schwarke (eds.), One-Volume Libraries: Composite and Multiple-Text Manuscripts (Berlin: De Gruyter, 2016), 8–11, 15–16. See also Jan Peter Gumbert, Codicologische eenheden. Opzet voor een terminologie (Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen, 2004) and Erik Kwakkel, “Towards a Terminology for the Analysis of Composite Manuscripts”, Gazette du livre médiéval 41 (2002): 12–19. A “multipletext manuscript” is defined as a “production unit” planned by the same persons in a single process to create a new work out of two or more texts. Friedrich and Schwarke oppose this to composite manuscripts, where material that was created independently is bound together. Whereas the term “multiple-text manuscript” thus refers to the level of production, the term “composite manuscript” refers to the level of circulation or reception. Both categories are closely related, as Gumbert already noted, for a copy of a composite manuscript (“convoluut” in his terminology) be4

224

MARION PRAGT

clear organising principle. It consists of commentaries on selected books from the Old and New Testament, which are structured around biblical verses followed by brief interpretations.6 The main text of the Collection is attributed to the monk Severus, who completed his work in the Monastery of the Holy Barbara on the mountain of Edessa in 861 CE.7 Nonetheless, the precise role Severus played in the composition of the Collection is uncertain. He may have united the material based on the works of Ephrem and Jacob, as suggested by Anton Baumstark, who described Simeon as someone “der vielmehr seine gesamte anscheinende Gelehrsamkeit einheitlich aus dritter oder vierter Hand geschöpft haben wird.”8 Alternatively, the commentary attributed to Severus may have been an already existing compilation to which he only made further changes.9 On the next comes a multiple-text manuscript (“verzamelhandschrift”). Gumbert, Codicologische eenheden, 5–6. 6 For the Old Testament, the books commented upon are the Pentateuch, Job, Joshua, Judges, the books of Samuel and Kings, Hosea, Amos, Micah, Joel, Obadiah, Zachariah, Malachi, Ezekiel, Jeremiah, Lamentations, Song of Songs, Daniel, Isaiah, Proverbs, and Ecclesiastes. For the New Testament, the books commented upon are Romans, 1 and 2 Corinthians, Galatians, Ephesians, Philippians, Colossians, 1 and 2 Thessalonians, 1 and 2 Timothy, Titus, Philemon and Hebrews, as well as the gospels of Matthew and John. 7 See the extensive colophon in Vat. Syr. 103 f. 371r. Severus is prė ). sented as a lover of toil (‫ܪܚܡ ܥܡܐܠ‬ 8 Anton Baumstark, Geschichte der syrischen Literatur mit Ausschluß der christlich-palästinensischen Texte (Bonn: A. Marcus und E. Webers Verlag, 1922), 279. 9 The possibility of seeing Severus’s commentary as a work with multiple layers is explored by Bas ter Haar Romeny, “Ephrem and Jacob of Edessa in the Commentary of the Monk Severus,” in Malphono w-Rabo d-Malphone: Studies in Honor of Sebastian P. Brock, ed. George A. Kiraz (Piscataway, NJ: Gorgias Press, 2008), 535–557. See also Kruisheer, “Ephrem, Jacob of Edessa, and the Monk Severus,” 600; Bas ter Haar Romeny, “The Identity Formation of Syrian Orthodox Christians,” 107–108; Bas ter Haar Romeny, “Greek or Syriac? Chapters in the Establishment of a Syrian Orthodox Exegetical Tradition,” in Studia Patristica XLI: Papers Presented at the Fourteenth International Conference on Patristic Studies Held in

ANNOTATION PRACTICES

225

level, longer additions to the commentary of Severus were made in the running text. Dirk Kruisheer has shown that the Commentary on the Octateuch by Jacob of Edessa was inserted in between Severus’s compilation based on Ephrem and Jacob of Edessa.10 This means that Severus’s work was interspersed with the full commentary of a single author. Furthermore, Bas ter Haar Romeny has shown that after the commentary on Isaiah attributed to Ephrem, an abbreviated version of Cyril’s Commentary of Isaiah was inserted.11 Moreover, the main text of Vat. Syr. 103 is enriched with additional material appearing in the margins. As seen above, these marginal notes were added by Simeon, who copied and expanded the compilation of Severus: This is the book of Simeon, the monk of Ḥisn Manṣur from the monastery of the Seven Martyrs near the ancient city of Perrhe, which he wrote for his own benefit and that of all who come across it.12 I, Simeon, have added out of my diligence all explanations which are inscribed in the margin of the book. Let everyone who reads (this) also pray for me, that our Lord

Oxford 2003, ed. Frances Young, Mark Edwards and Paul Parvis (Leuven: Peeters, 2006), 89–96. 10 Kruisheer, “Ephrem, Jacob of Edessa, and the Monk Severus,” 600–603. This is made clear by Vat. Syr. 103 f. 66v, where it is noted that “the commentaries of Mar Jacob on the five books of Moses and on the book of Job and of Joshua bar Nun and of Judges are finished.” 11 Bas ter Haar Romeny, “The Greek vs. the Peshitta in a West Syrian Exegetical Collection,” in The Peshitta: Its Use in Literature and Liturgy: Papers Read at the Third Peshitta Symposium, ed. Bas ter Haar Romeny (Leiden: Brill, 2006), 305; Bas ter Haar Romeny, “The Peshitta of Isaiah. Evidence from the Syriac Fathers,” in Text, Translation and Tradition: Studies on the Peshitta and its Use in the Syriac Tradition Presented to Konrad D. Jenner on the Occasion of his Sixty-Fifth Birthday, ed. Wido Van Peursen and Bas ter Haar Romeny (Leiden: Brill, 2006), 155. 12 On the vocabulary of “encountering” (‫ )ܦܓܥ‬a codex, instead of merely reading one, see Heal, “Catalogues and the Poetics,” 390.

226

MARION PRAGT may have mercy on me and may correct my errors and may not find fault.13

Vat. Syr. 103 is not dated. However, it is possible to determine its terminus post quem as the colophon dates Severus’s completion of his work to the year 861. Moreover, a note in the manuscript shows that it was brought to the monastery of Mary the Mother of God, also known as Deir al-Surian, the Monastery of the Syrians in the Egyptian desert.14 The monastery's famous manuscript collection, chiefly associated with the activities of its abbott Moses of Nisibis, has been described by Sebastian Brock.15 Both before and after Moses, books were donated to the monastery; and it is of these smaller-scale donations that the Collection of Simeon appears to have been part. In fact, Vat. Syr. 103 contains a donation note which mentions a certain Simeon of Takrit: Simeon of Takrit, priest and monk, has bequeathed this book of commentaries to the monastery of Mary the Mother of God which is in the desert of Scetis, for the absolution of his debts and the forgiveness of his sins. I ask each reader and listener that they will pray for me, so that he will find mercy and be spared, like the robber on the cross and like the sinful woman of the house of Simeon the Pharisee, through the prayers of Mary the Mother of God and all the saints. Amen.16

Vat. Syr. 103 f. 371r. The preceding part of the colophon describes the work of Severus. 14 Vat Syr. 103 f. 371v. 15 Sebastian Brock, “Without Moses of Nisibis, Where Would We Be? Some Reflections on the Transmission of Syriac Literature,” Journal of Eastern Christian Studies 56 (2004), 15–24. See also Hugh G. Evelyn-White, The Monasteries of the Wadi ‘n Natrun: The History of the Monasteries of Nitria and of Scetis, Vol. 2 (New York: The Metropolitan Museum of Art Egyptian Expedition, 1932), 337–338. 16 This is the first part of the donation note included in Vat. Syr. 103 f. 371v. On the connection between the monastery and monks and merchants from Takrit, see Evelyn-White, The Monasteries of the Wadi ‘n Natrun, Vol. 2, 439–440; Sebastian Brock and Lucas Van Rompay, Catalogue of the 13

ANNOTATION PRACTICES

227

This note is undated, and the precise time of the manuscript’s arrival in Deir al-Surian is therefore unknown.17 Nonetheless, the names “Simeon” and “Simeon of Takrit” appear in the colophons and donation notes of several ninth-century manuscripts of Deir alSurian. For example, a manuscript containing the lives of martyrs and holy women and men—dated by Wright to the ninth century—belonged to a priest named Simeon of Takrit and was donated to the monastery library upon his death.18 Also, a Simeon of the monastery of the Holy Mar Solomon near Doluk is mentioned as the benefactor of a series of Funeral Services and Funeral Sermons written in 877 and donated to the library.19 Moreover, Simeon from the monastery of the Holy Mar Solomon near Doluk is also mentioned in a collection of saints’ lives dated to 875 as someone who possessed and handed down eighteen books, which were donated to Deir al-Surian.20 It is possible that some or all of these Simeons are the same person as the Simeon of Takrit mentioned in Vat. Syr. 103, which would then place the manuscript’s arrival at Deir alSurian in the late ninth century. This, in turn, would suggest that “our” Simeon, the scribe and scholiast, worked on Vat. Syr. 103 slightly before this period. The manuscript seems to have belonged to Deir al-Surian for several centuries. Interest in the Collection of Simeon continued as it was copied in 1081 in the Monastery of the Mother of God in

Syriac Manuscripts and Fragments in the Library of Deir al-Surian, Wadi alNatrun (Egypt) (Leuven: Peeters, 2014), xiv. 17 Brock and Van Rompay, Catalogue of the Syriac Manuscripts, 74. 18 BL Add. 14649. William Wright, Catalogue of Syriac Manuscripts in the British Museum Acquired Since the Year 1838 (London: British Museum, 1870–1872), Vol. 3, 1110. 19 BL Add. 17130. Wright, Catalogue, Vol. 1, 393. Evelyn-White describes him as hailing from Takrit. Evelyn-White, The Monasteries of the Wadi ‘n Natrun, 442. 20 BL Add. 14650. Wright, Catalogue, Vol. 3, 1107. Evelyn-White indeed identifies him with the Simeon of Takrit of Vat. Syr. 103. Evelyn White, The Monasteries of the Wadi ‘n Natrun, 442.

228

MARION PRAGT

Gazarta near Alexandria, most likely on the basis of Vat. Syr. 103.21 Indeed, part of the colophon of Vat. Syr. 103—which urges “everyone who takes it and profits from it or wishes to copy it” to take over that note and pray for its author—is included in the eleventhcentury copy.22 Furthermore, Vat. Syr. 103 contains traces of a now almost illegible reader’s note, of which the Assemani have made a partial transcription. On that basis, the following fragmentary translation may be given: This book, a commentary on the Old and the New (Testament), belongs to the holy Monastery of the Mother of God which is in the desert of Scetis. No one has the right to take it away (...) I beg you, my brothers and masters, pray for me, John the monk of the great and venerable Monastery of Mar Gabriel from the blessed village of Beth Sbirina.23

This note indicates that Vat. Syr. 103 attracted the attention of at least one of the visitors and pilgrims who came to Deir al-Surian and left traces of their presence in the library’s manuscripts.24 A note in a later hand indicates that Vat. Syr. 103 was brought to

BL Add. 12144. Wright, Catalogue, Vol. 2, 908–914. This manuscript represents a further stage in the transmission of the Collection of Simeon as many of Simeon’s marginal notes were taken over and included in its main text. 22 The scribe of BL Add. 12144, Samuel, interestingly compares his own three years of copying to the hard work of a painter who copied an ̈ image and “applied pigments in abundance” (‫ܕܣܡܐܢܐ‬ ‫)ܘܣܪܚ ܣܘܓܐ‬, f. 233v. Also see Wright, Catalogue, 913. 23 Vat. Syr. 103 f. 1r. J.S. Assemanus and S.E. Assemanus, Bibliothecae Apostolicae Vaticanae Codicum Manuscriptorum Catalogus (Rome, 1758–1759), Part I, Vol. III, 27. 24 Heal draws attention to three similar readers’ notes by visitors from Tur ʿAbdin; two fifteenth-century notes by visitors from Beth Sbirina and one eleventh-century note by a monk from Basbirina. Heal, “Catalogues and the Poetics,” 388. 21

ANNOTATION PRACTICES

229

Aleppo in 1646.25 Eventually, the manuscript arrived in the Vatican in 1716.26

2 MARGINAL ANNOTATION IN THE C OLLECTION OF

SIMEON

The marginal notes in Vat. Syr. 103 serve to comment on, add to and correct the main text. They also provide cross-references and point the reader to parallel texts27 and are occasionally used by Simeon to express his opinion on a passage.28 Furthermore, marginal annotation was used to correct mistakes in the running text—for instance, when a section was written in the wrong place.29 The notes stand out visually, appearing in red and sometimes green ornamented frames.30 This paper proposes that studying the marginalia of the Collection of Simeon is important for understanding its profile. In the main text of the Collection, Ephrem and Jacob of Edessa take pride of place.31 Yet the margins provide a broader and more varied outlook in which authors such as Severus of Antioch and Cyril of Alexandria are often cited.32 Vat. Syr. 103 f. 247r. Assemani, Bibliothecae Apostolicae Vaticanae Codicum Manuscriptorum Catalogus, Part I, Vol. III, 28. 27 For example, between Isaiah and 2 Kings. Vat. Syr. 103 f. 206v. 28 For example, the commentator signaled his approval of a passage by adding the brief comment “beautiful thought” in the margin. Vat. Syr. 103 f. 238r. 29 Vat. Syr. 103 f. 2r and 351r. 30 That these frames may have been drawn before the text was added is suggested by Vat. Syr. 103 f. 303r, where several “text boxes” appear in the upper, inner and outer margins, one of which is left empty. 31 Romeny, “The Identity Formation of Syrian Orthodox Christians,” 115–117, contrasts the focus on Syriac authors of the Collection of Simeon to the earlier “London Collection” (Ms. BL Add. 12168), which concentrates on Greek authors and in which biblical citations often follow the Syro-Hexapla instead of the Peshitta. 32 In what is now the second volume of the Collection of Simeon, fifty attributed marginal commentaries occur, twenty-one of which are ascribed to Syriac authors and twenty-nine to Greek authors. 25 26

230

MARION PRAGT

Moreover, besides its running commentaries and marginal notes, the Collection of Simeon also contains signs and symbols. Overall, no consistent system of sign use has yet been identified. The traditional, text-critical signs appear rarely; rather, the signs appear to have functioned chiefly as signes-de-renvoi, which indicate on what part of the main text a specific marginal note comments, a practice especially useful on heavily annotated pages. Some signs may have been of Simeon’s own invention, such as the several variations of crosses and rosettes he employs, while others also appear in Syriac philosophical works and the Syro-Hexapla.33 Thus, the signs used by Simeon show the Collection’s affiliation to late antique scholarly practices and the writing of commentaries on authoritative texts 2.1 Marginal notes in the commentary on the Song: Extracts from Hippolytus The section on the Song stands out among the commentaries in the Collection of Simeon because it is not attributed to a Greek or Syriac author. Instead, it is described as having been compiled from the interpretation of the “divine teachers.” Ceslas Van den Eynde established that it mainly depends on Gregory of Nyssa’s Homilies on the Song.34 The commentary cites most verses of the Song, folFor example, in Vat. Syr. 158, dated to ca. 800 CE and containing a translation of and commentary on Porphyry’s Isagoge as well as works by and commentaries on Aristotle. Similar signs are also used in the margins of the Syro-Hexapla (Codex Ambrosianus) to indicate what readings have variants among the minor versions. 34 Gregory’s lengthy Homilies were transmitted in Syriac as a commentary (‫)ܦܘܫܩܐ‬. The abbreviated commentary on the Song from the Collection of Simeon depends on the full Syriac translation of Gregory’s Homilies. Ceslas Van den Eynde, La version syriaque du Commentaire de Grégoire de Nysse sur le Cantique des Cantiques: ses origines, ses témoins, son influence (Louvain: Bureaux du Muséon, 1939), 50–56. In the Syriac manuscript tradition, Gregory’s Homilies on the Song are preserved together with the interpretations of a certain Symmachus, an otherwise unknown Greek author. This is also the case for the Collection of Simeon. As Gregory’s Homilies reach as far as Song 6:9, Symmachus’s interpretations were included to cover Song 6:10 to 8:14. For Symmachus’s commentary and its two abbreviated 33

ANNOTATION PRACTICES

231

lowed by brief interpretations taken from Gregory, which illustrate the spiritual significance of the Song’s words and phrases. Simeon has added thirty-six marginal notes to the commentary. Twentyfive of these deal with citations from scripture and offer variant readings—for example, from the Syro-Hexapla—or etymological explanations, displaying a philological interest. The other eleven notes add new material or further elucidate parts of the Song not fully explained in the main text. Three of these notes go back to Hippolytus, one is attributed to Daniel of Salah, and the remaining seven are of yet unknown origin. The next sections will consider the extracts from Hippolytus—two with attribution and one without—in their connection to the main text of the commentary, their wider Syriac dissemination, and the interests they may reflect. 2.2 The spices of the Song In Song 4:13–14, the bride is praised by being associated with all kinds of spices and perfumes—among them myrrh and aloe — which receive a brief interpretation in the Collection: Myrrh and aloe with the best of spices. For myrrh and aloe refer to participation in the burial, but by means of the smell of the perfumes (they refer) to purification in which no licentiousness is mixed.35

Simeon has marked the term “myrrh” with a sign in the shape of a cross, with small circles on both horizontal sides, perhaps similar to the ὡραῖον used to highlight noteworthy passages in late antique texts.36 He has added the following marginal commentary: versions (in “our” Vat. Syr. 103, as well as in BL Add. 12168), see Van den Eynde, La version syriaque, 77–95 (Syriac text) and 103–126 (Latin translation). 35 Vat. Syr. 103 fol. 179r. 36 On the ὡραῖον, see Kathleen McNamee, “Sigla in Late Greek Literary Papyri,” in Signes dans les textes, textes sur les signes: Érudition, lecture et écriture dans le monde gréco-romain, ed. Gabriel Nocchi Macedo and Maria Chiara Scappaticcio (Liège: Presses universitaires de Liège, 2017), 133, and Eva Steinová, Notam superponere studui. The Use of Annotation Symbols in the Early Middle Ages (Turnhout: Brepols, 2019), 218.

232

MARION PRAGT By Hippolytus from the commentary on the Song of Songs. When Joseph desired these spices, he became a counselor of God. When the virgin Mary was consecrated with these, she conceived the Word in her womb. O new mysteries and truth newly demonstrated! O new economy! O great and ineffable mysteries!37

As a fuller version of Hippolytus’ Commentary on the Song of Songs is extant in Georgian, it is possible to contextualise this extract.38 Hippolytus mentions Joseph and Mary as part of a long list of biblical figures whom he associates with spices, perfume and ointment.39 The marginal note attributes two characteristics to Joseph: he is said to have desired spices and is named a “counselor of God.” An explanation of this latter characteristic may be given on the basis of the Georgian version, where Joseph is presented as a Vat. Syr. 103 f. 179r. The Georgian text (based on an Armenian version, which in turn was translated from the Greek) is the most complete version in which Hippolytus’s work is extant. It was recently studied and translated into English by Yancy Smith, The Mystery of Anointing: Hippolytus' Commentary on the Song of Songs in Social and Critical Contexts. Texts, Translations, and Comprehensive Study (Piscataway: Gorgias Press, 2015). My references to the Georgian version follow his translation. Smith takes the view that Hippolytus’s work functioned in the setting of a Roman Easter liturgy of baptism and anointing, during which new members were initiated and received instruction. Smith, The Mystery of Anointing, 90-95, 174-179, 238-247, 249-300. Hippolytus’s Commentary on the Song was previously edited and translated into Latin in Gérard Garitte, Traités d’Hippolyte sur David et Goliath, sur le Cantique des cantiques et sur l’Antéchrist, CSCO 263-264 (Louvain: Peeters, 1965). 39 It could be argued that Hippolytus meant to refer to Joseph of Arimathea here, who is introduced as a counselor in Mark 15:42–43 and Luke 23:50–51 and is described as treating Jesus’ body with myrrh in John 19:39–40. However, Hippolytus generally lists his biblical figures in the order in which they appear in scripture. Joseph is the first out of four New Testament characters listed—followed by Mary, Martha and Judas— which makes it likely that he should be identified with Joseph of Nazareth. 37 38

ANNOTATION PRACTICES

233

“divine consultant.”40 The image of Joseph as consulting with God fits his portrayal in the gospel of Matthew. At three key instances during the narrative, Joseph receives and follows divine advice: to accept that Mary’s son was conceived by the holy spirit and to name the child Jesus, to flee to Egypt, and, eventually, to return.41 Joseph’s association with spices is more difficult to explain. For this, it is necessary to look into the wider context in which the list of biblical figures including Joseph occurs. Hippolytus is concerned with explaining Song 1:3, which reads: “Your name is aromatic oil poured out.”42 He makes the following parallel: in the same way that perfume kept in a container does not emit any fragrance, so too God’s Word was kept hidden before the incarnation. Yet once Christ was “poured out” from the Father, from heaven to earth, he became known to everyone, like perfume can only delight people with its fragrance once it is poured out. Thus, in Hippolytus’s view, aromatic oil refers to Christ; and longing for this perfume signifies a desire for Christ.43 Regarding Mary, a similar parallel is at work; the perfume and spices here refer to the incarnation.44 The appearance of spices in the marginal comment may have prompted Simeon to add it, by association, to the interpretation of the spices and perfumes of Song 4. Specifically, it may be that myrrh, which is used in the treatment of deceased bodies and is especially associated with Christ’s burial, reminded Simeon of this passage in which spices and Christ are related as well. 2.3 Seeking Christ In Song 5:2–6, the bride describes herself as being asleep while her heart was awake, unable to reach her beloved: “I sought him but I In Cant. 2.27. Smith, The Mystery of Anointing, 457 and Garitte, Traités d’Hippolyte, vol. 2, 29 l. 24–25 (consultator Dei). 41 Respectively, Matt. 1:19–25, Matt. 2:13–14 and Matt. 2:19–21. 42 In Cant. 2.5- 2.32. Smith, The Mystery of Anointing, 447-459. 43 “Now what was that anointing oil, except Christ himself?” Smith, The Mystery of Anointing, 459. 44 As Hippolytus noted in his interpretation of Song 1:3: “See the aroma of anointing oil spread out; it was taken into the womb and created a new[ly] begotten human.” Smith, The Mystery of Anointing, 450. 40

234

MARION PRAGT

did not find him.” Gregory of Nyssa interprets this seeking without finding as the soul’s searching for knowledge of and union with the divine. This longing is never quite fulfilled; or rather, the fulfillment consists in the continuous process of seeking and following.45 This idea is summarised in the main text of the Collection of Simeon. The marginal note then offers a different interpretation of the same passage: By Hippolytus. The women went by night seeking him in the grave. “I sought him,” she said, “but I did not find him.” “But why,” said the angel, “are you seeking the living among the dead?” We did not find him there, for that tomb was not his place, but heaven. “Why did you search him who sits above the Cherubim?”46

Hippolytus sees the Song’s bride as a type of the women who came to look for Christ’s crucified body before the break of day, whom he identifies with Martha and Mary.47 As the main text of the Collection does not associate Song 5 with the search for the risen and ascended Christ, Simeon may have regarded this extract as an attractive addition to the compilation. Hermannus Langerbeck, Gregorii Nysseni In Canticum Canticorum (Leiden: Brill, 1960), 353 l. 11–359 l. 4. Gregory reads the pericope with the account of Moses’s encounter with God as his intertext. Just as Moses did not see God’s face but only his back, according to Exodus 33, so also the bride could not fully grasp the bridegroom. The references to Moses are not taken up in the Collection of Simeon. 46 Vat. Syr. 103 f. 179v. As far as I am able to tell, Simeon used no sign here, perhaps because the marginal note does not comment upon a single word or phrase of the main text but offers an alternative explanation of the complete searching account of Song 5:2–6. 47 Smith, The Mystery of Anointing, 526-529. See also Luke 24:5. On the important role Hippolytus attributes to Martha and Mary as first witnesses of the resurrection and his interpretation of the text as the synagogue searching for and failing to find Christ, see Allie M. Ernst, Martha from the Margins: The Authority of Martha in Early Christian Tradition (Leiden: Brill, 2009), 98–102. 45

ANNOTATION PRACTICES

235

2.4 The leap and the divine economy In Song 2:8-9 the bride announces the arrival of her beloved, as he comes jumping over the mountains and leaping over the hills. The Collection of Simeon explains the verse with reference to the coming of Christ: The voice of my beloved. His voice and not the sight of his face: the message of the prophets on the coming of Christ, which was proclaimed before. See, this one comes. The divine economy of Christ, as it was revealed to us in the gospel. While he jumps over the mountains and leaps over the hills. Because he adapted himself to a gazelle as he sees everything, but to a fawn of the hind because he tramples and destroys the evil powers, which are allegorically called mountains and hills.48 For these mountains of Bethel indicate the exalted and heavenly way of life.49

Although the marginal comment Simeon added to this passage is unattributed, it is part of Hippolytus’s Commentary on the Song:50 What is the leap? The Word which jumped from heaven to the womb of the virgin, and jumped from the belly to the cross, and from the cross to the grave.51

The main text of the Collection of Simeon already interprets the bridegroom’s arrival as the coming of Christ in the flesh. This idea is continued in the marginal note, where the image of leaping is used to refer more specifically to Christ’s birth, crucifixion and burial. On Gregory’s association of Christ with a gazelle and a deer, see Richard A. Norris, Gregory of Nyssa: Homilies on the Song of Songs (Atlanta: Society of Biblical Literature, 2012), 153 n. 2 and 155 n. 5. 49 Vat. Syr. 103 f. 177v. 50 For the longer, Georgian text, see Smith, The Mystery of Anointing, 517-518. 51 Vat. Syr. 103 f. 177v fifth note. The note is marked with a sign in the shape of a horizontal line with two smaller, vertical lines at either end. The part of the main text on which the note comments—“while he jumps”—is marked with a similar sign, a horizontal line with two smaller, vertical lines at its right side. 48

236

MARION PRAGT OF THE C OLLECTION OF S IMEON

3 THE SYRIAC HIPPOLYTUS EXTRACTS IN AND OUTSIDE

The three Hippolytan extracts on the Song are not only part of the Collection of Simeon but appear in similar form in Ms. BL Add. 12156.52 This manuscript contains a collection of “many arguments from the holy fathers,” which has been dated to before 562.53 There, the Hippolytus extracts are part of a collection described by Ignaz Rucker as the Florilegium Edessenum.54 This collection consists of extracts from the works of Christian Greek authors to argue, for example, that Christ is truly God and that Mary gave birth to the Word of God. This shows that extracts from Hippolytus’s Commentary on the Song were known in Syriac from the sixth century onwards and that they may first have appeared in the context of doc-

For an overview of the extracts according to Vat. Syr. 103 and Add. 12156, see the Appendix. 53 Ms. BL Add. 12156 f. 70r. Wright, Catalogue, Vol. 2, 639–648. Also see Luise Abramowski, “Zur geplanten Ausgabe von Brit. Mus. add. 12156,” in Texte und Textkritik. Eine Aufsatzsammlung, ed. Jürgen Dummer (Berlin: Akademie-Verlag, 1987), 23–28. 54 A schematic overview of the contents of the Florilegium Edessenum (f. 69r–79v) and a brief interpretation of the doctrinal background of the statements is given by Ignaz Rucker, Florilegium Edessenum anonymum (München: Verlag der Bayerischen Akademie der Wissenschaften, 1933), x–xiii. The three Hippolytus extracts are listed as his fragments 11–13. Rucker, Florilegium Edessenum, 8–10. Previously, the Hippolytus extracts on the Song from BL Add. 12156 were also published in Paul de Lagarde, Analecta Syriaca (Leipzig: Teubner, 1858), 87, and J.B. Pitra, Analecta Sacra Spicilegio Solesmensi parata, (Paris: A. Jouby et Roger, Bibliopolis, 1883), 40–41 (Syriac text) and 310 (Latin translation). Apart from the three extracts of BL Add. 12156, both Pitra and Rucker also attribute a longer explanation of the Song to Hippolytus, which they take from BL Add. 12144, the eleventh-century copy of Vat. Syr. 103. However, the passage they have in mind actually belongs to the abbreviated version of Gregory’s Homilies on the Song, which appears in the Collection of Simeon. Thus, it goes back to Gregory, not to Hippolytus. 52

ANNOTATION PRACTICES

237

trinal debate against Dyophysites and Chalcedonians.55 Furthermore, the extracts do not seem to originate from a full Syriac version of Hippolytus’s Commentary on the Song but perhaps rather represent material which had already been translated from the Greek in an excerpted form at an earlier stage. Thus, in this instance, Simeon’s notes reproduce previously selected material but present it in a new, exegetical context. The relationship between the extracts as they appear in Simeon’s version in Vat. Syr. 103 and in BL Add. 12156 remains to be explained, as the attestations of the extracts differ with regard to their order, delineation and precise vocabulary. Comparison with the Georgian version of Hippolytus’s Commentary shows that the three extracts form part of his interpretations of Song 1:3, 2:8 and 3:1. In Add. 12156, they are listed in this order. The first extract, on the aromatic oil poured out, refers to the incarnation and Mary’s conceiving of the Word. Next, the leap signifies the journey of the heavenly Word from the incarnation to his crucifixion and burial. While in the second note the Word ended in the grave, the third extract takes the now empty tomb as its point of departure and describes Christ as enthroned above the Cherubim. However, in the Collection of Simeon, only the explanation of the leap is added to the same verse for which Hippolytus originally designed it. As such, it becomes the first extract from Hippolytus in the margin of the commentary on the Song. The extract on the spices is used in a different context altogether, as it is added to the catalogue of spices with which the Collection interprets Song Several other works and fragments attributed to Hippolytus are known in Syriac. Brock edited and translated into English several hitherto unknown fragments, commented on their rather literal style of translation, and assigned them to the sixth and seventh centuries. Sebastian Brock, “Some New Syriac Texts Attributed to Hippolytus,” Le Muséon 94, no. 1– 2 (1981), 177–200. A Syriac version of Hippolytus’s commentary on Daniel was edited and translated into French by André de Halleux, “Une version syriaque révisée du commentaire d’Hippolyte sur Suzanne,” Le Muséon 101, no. 3–4 (1988), 297–341. He studied the version in greater detail and tentatively dated it to the sixth century in Halleux, “Hippolyte en version syriaque,” Le Muséon 102, no. 1–2 (1989), 35, 41–42. 55

238

MARION PRAGT

4:13–14. Lastly, in both Song 3:1 and Song 5:6, the bride exclaims: “I sought him but I did not find him.” The extract on the women searching for Christ goes back to Hippolytus’s interpretation of Song 3. In the Collection of Simeon, however, the searching account of Song 3 is only partially included, apparently to avoid repetition; and the extract from Hippolytus is added to the similar passage of Song 5.56 In both Add. 12156 and Vat. Syr. 103, the extract on the spices contains the attribution “by Hippolytus from the Commentary on the Song of Songs.” Add. 12156 refers to the incarnation as “truth justly demonstrated” (‫)ܕܟܐܢܐܝܬ‬, whereas Vat. Syr. 103 reads “truth newly demonstrated,” perhaps parallel to the new mysteries and new economy it also mentions. The extract on the leap in Add. 12156 begins as follows: “And after other (words) from the same: O new economy! O great mysteries! See, my cousin comes leaping.”57 In Simeon’s version of the extract, this introduction is absent; there is no attribution, and the note contains no reference to the bridegroom’s arrival, perhaps because Song 2:8 is already cited in the main text. Also, the reference to the economy and mysteries appears instead at the end of the extract on the spices, perhaps because that note already contained similar exclamations.58 Moreover, the extract on the leap also contains different vocabulary. In contrast to 12156, the Collection uses not only ‫ ܡܪܒܥܐ‬but also ‫ ܟܪܣܐ‬to indicate Mary’s womb. Also, in Add. 12156 the Word is said to jump from the cross into the abyss (‫)ܬܗܘܡܐ‬, whereas the Collection of Simeon reads “grave” (‫)ܩܒܪܐ‬.59 The latter may perhaps have occurred by analogy with the extract in which the women search for Christ in his tomb. Cf. Vat. Syr. 103 f. 178r and f. 179v. “Cousin” (‫ )ܒܪ ܕܕܝ‬here seems to have rendered ἀδελφιδός, the term of endearment with which the Song’s woman addresses her beloved several times, according to OG Song. 58 Moreover, there the mysteries are not only said to be “great” but ̈ also “ineffable” (‫ܡܬܡܠܠܢܐ‬ ‫)ܘܐܠ‬. 59 See also Pitra, who translated with atque de cruce in aeternitatem but admitted that a literal translation should read et de ligne in abyssum. Pitra, Analecta Sacra 4, 310 n. 5. Smith refers to the Syriac Hippolytus extracts 56 57

ANNOTATION PRACTICES

239

Finally, Add. 12156 introduces the note on the searching women, as the third extract, merely with the words “by the same,” whereas the corresponding extract in Vat. Syr. 103 mentions Hippolytus by name. According to Simeon’s version of the note, in contrast to Add. 12156, an angel addressed the women, which may have been an addition to make clear who informed the women that Christ is not to be found in the grave.60 It seems possible that most of these differences regarding order, delineation and vocabulary resulted from changes introduced by Simeon as he accorded the extracts fitting positions in the margin of the commentary on the Song. On the other hand, the anonymous appearance of the extract on the leap is hard to explain, and it is not clear whether Simeon was aware of its Hippolytan background. Therefore, the possibility cannot be excluded that the extracts were available to Simeon in a different form than how they appear in Add. 12156. Either way, Simeon’s use of the Hippolytus extracts in his marginal notes has the double effect of not only further elucidating parts of the Song, but also adding themes with a special relevance for a Miaphysite audience. For example, both the extracts on the spices and on the leaping present a Miaphysite view of Christ and the incarnation. In the former, spices serve as a sign for Christ, and Mary is figured as the bearer of the Word. The latter presents Mary as giving birth to the heavenly Word.61 Thus, the inclusion of Hippolytus’s extracts may well have served to strengthen the Miaphysite character of the Collection of Simeon.

collected by Pitra in a footnote, although the text he cites contains several mistakes. Smith, The Mystery of Anointing, 518 n. 332. 60 Although the extract itself reflects the tomb account of Luke 24 (including the note “Why are you seeking the living among the dead?”), the reference to the angel perhaps rather reflects Matthew 28. 61 This representation of Mary as not only the one who gave birth to the earthly Jesus but as the mother of God is taken further in the eleventh-century copy of the Collection of Simeon, where the phrase ‫ܝܠܕܬ ܐܠܗܐ‬ is added to Hippolytus’s note, a qualification absent from Add. 12156 and Vat. Syr. 103. BL Add. 12144 f. 58r.

240

MARION PRAGT

4 CONCLUSION

This paper aimed to investigate the colophon and notes preserved in the Collection of Simeon, which reveal traces of the history of its composition and use. Questions remain concerning with what aims the Collection was created and what needs it fulfilled. The Collection of Simeon is one of the works in which late antique and medieval Syriac compilers brought together the biblical interpretations of their Greek and Syriac predecessors. It was argued by Romeny that exegetical compilations like the Collection of Simeon played an important role in the identity formation of West Syrians as compilers collected knowledge for transmission to future generations.62 This is also visible in Simeon’s marginal notes. This paper has illustrated the importance of taking into account the marginal notes of the Collection of Simeon; for although its main text largely relies on the trusted authorities of Ephrem and Jacob of Edessa, the margins draw on a wider array of authors and sources. In particular, Cyril of Alexandria, Severus of Antioch and Daniel of Salah are cited, which shows Simeon’s predilection for authors regarded as fitting in a West Syriac context. In this light, Hippolytus may seem an unexpected choice. However, by virtue of their transmission in Add. 12156, which was designed to aid doctrinal debate against Chalcedonians and Dyophysites, the extracts may have found a favourable reception among West Syriac audiences. Thus, Simeon’s marginal notes, including the extracts from Hippolytus, not only discuss themes not addressed in the main text but, more specifically, suggest an interest in adding to the Miaphysite character of the Collection. Nonetheless, further, more extensive study is required to draw firmer conclusions on the interests which Simeon’s notes reflect and the question of whether he excerpted full works or only presented already selected extracts. Finally, the colophons and marginal notes allow for some suggestions as to the possible audience and aims of the Collection. The Collection was composed and used in a monastic context. Simeon See Bas ter Haar Romeny, “The Formation of a Communal Identity among West Syrian Christians: Results and Conclusions of the Leiden Project,” Church History and Religious Culture 89, no. 1 (2009), 1–52. 62

ANNOTATION PRACTICES

241

loosely described the potential audience of his work as “all who come across it.” That the Collection contains material written near Edessa, was copied and expanded in Eastern Turkey, and ended up in Deir al-Surian in Egypt, where it was inscribed by a visitor from Tur ʿAbdin, already shows that use of the work was not restricted to a single area. Vat. Syr. 103 contains the requests for prayer, appeals to remember a scribe’s notes and warnings against not returning a borrowed work, which are common to Syriac manuscripts. While such notes are part of what Heal has called “the poetics of the colophon,” at the same time, they perhaps also testify to the expected use of the manuscript.63 For example, the invocation of “each reader and listener” in the donation note of Simeon of Takrit may reflect the expectation that the Collection of Simeon would be read and used by the monks of Deir al-Surian. Moreover, the nature of the commentaries and marginal notes preserved in the Collection, which reflect exegetical, theological and philological interests, suggests that the work was used for the purposes of study. In this way, Simeon’s marginal notes provide insight into how interpretations of scripture, both Greek and Syriac, continued to be used and interacted with in the Syriac world.

APPENDIX: THE SYRIAC EXTRACTS FROM HIPPOLYTUS ON THE SONG OF SONGS

Vat. Syr. 10364

BL Add. 1215665

‫ܕܐܝܦܘܠܛܘܣ ܡܢ ܦܘܫܩܐ ܕܫܐܪܬ ܕܐܝܦܘܠܝܛܝܣ ܡܢ ܦܘܫܩܐ ܕܫܐܪܬ‬ ‫ ܠܗܠܝܢ ܗ̈ܪܘܡܐ ܟܕ‬.‫ܫܐ̈ܪܝܢ‬ ‫ ܠܗܠܝܢ ܗ̈ܪܘܡܐ ܟܕ‬.‫ܫܐ̈ܪܝܢ‬ ‫ ܒܪ ܡ ܼܠܟܐ‬.‫ܐܬܪܓܪܓ ܝܘܣܦ‬ ‫ ܒܪ ܡ ̣ܠܟܐ‬.‫ܐܬܪܓܪܓ ܝܘܣܦ‬ ‫ ܒܗܠܝܢ‬.‫ ܒܗܠܝܢ ܒܬܘܠܬܐ ܕܐܠܗܐ ̣ܗܘܐ‬.‫ܕܐܠܗܐ ̣ܗܘܐ‬ ̇ ‫ܐܬܡܫܚܬ‬ ‫ ܡܪܝܡ ܟܕ‬66‫ܒܬܘܠܬܐ‬ ‫ܒܟܪܣܗ‬ .‫ܐܬܡܫܚܬ‬ ‫ܡܪܝܡ ܟܕ‬ ̣ ̣ ̄‫ ܐܘ‬.‫ܒܟܪܣܗ ܠܡܠܬܐ ܒܛܢܬ‬ ܿ ‫ ܐܘ ܠ̈ܪܐܙܐ‬.‫ܠܡܠܬܐ ܒܛܢܬ‬ ̣ Heal, “Catalogues and the Poetics,” 375–417. Simeon’s punctuation and signs are followed here. 65 Lagarde, Analecta Syriaca, 87; Pitra, Analecta Sacra, 40–41; Rucker, Florilegium Edessenum, 8–10. See note 54 above. 66 At this stage, Add. 12144 adds the phrase ‫ܝܠܕܬ ܐܠܗܐ‬. 63 64

‫‪MARION PRAGT‬‬

‫̈ܚܕܬܐ‪ .‬ܘܠܫܪܪܐ ܕܟܐܢܐܝܬ‬ ‫ܡܬܚܘܐ܀‬ ‫‪Song 1:3‬‬ ‫‪f. 70r‬‬

‫‪242‬‬

‫ܠ̈ܪܐܙܐ ̈ܚܕܬܐ܇ ܘܠܫܪܪܐ‬ ‫ܡܬܚܘܐ‪ .‬ܐܘ̄‬ ‫̇‬ ‫ܕܚܕܬܐܝܬ‬ ‫ܠܡܕܒܪܢܘܬܐ ܚܕܬܐ‪ .‬ܐܘ‬ ‫̈‬ ‫ܡܬܡܠܠܢܐ‬ ‫ܠ̈ܪܐܙܐ ̈ܪܘܪܒܐ ܘܐܠ‬

‫‪Song 4:14‬‬ ‫‪f. 179r‬‬

‫ܡܢܐ ܗܝ ̇ܫܘܪܬܐ‪ .‬ܡܠܬܐ ̣ܫܘܪ‬ ‫ܘܒܬܪ ܚ̈ܪܢܝܬܐ ܕܝܠܗ ܟܕ ܕܝܠܗ܀‬ ‫ܗܘܐ ܡܢ ܫܡܝܐ ܠܡܪܒܥܐ‬ ‫ܐܘ ܠܡܕܒܪܢܘܬܐ ܚܕܬܐ‪ .‬ܘܐ‬ ‫ܘܫܘܪ ܗܘܐ ܡܢ ܟܪܣܐ‬ ‫ܠ̈ܪܙܐ ̈ܪܘܪܒܐ‪ .‬ܗܐ ܒܪ ܕܕܝ ܟܕ ܡܫܘܪ ܕܒܬܘܠܬܐ‪̣ .‬‬ ‫ܐܬܐ‪ .‬ܡܢܐ ܗܝ ܫܘܪܬܐ‪ .‬ܡܠܬܐ ܠܩܝܣܐ‪ .‬ܘܡܢ ܩܝܣܐ ܠܩܒܪܐ‬ ‫ܫܘܪ ܗܘܐ ܡܢ ܫܡܝܐ ܠܡܪܒܥܐ‬ ‫ܕܒܬܘܠܬܐ‪ .‬ܘܫܘܪ ܗܘܐ ܡܢ‬ ‫ܡܪܒܥܐ ܠܩܝܣܐ‪ .‬ܘܫܘܪ ܗܘܐ ܡܢ‬ ‫ܩܝܣܐ ܠܬܗܘܡܐ܀‬ ‫‪Song 2:8‬‬ ‫‪f. 70r‬‬

‫‪Song 2:8‬‬ ‫‪f. 177v‬‬

‫ܐܬܝ ̈‬ ‫ܐܬܝ ̈‬ ‫̈‬ ‫̈‬ ‫ܢܫܐ ܒܠܠܝܐ‬ ‫ܢܫܐ ܒܠܠܝܐ ܕܐܝܦܘܠܝܛܝܣ܀‬ ‫ܕܝܠܗ ܟܕ ܕܝܠܗ‪.‬‬ ‫̈‬ ‫ܕܒܥܝܢ ܠܗ ܒܩܒܪܐ‪ .‬ܒܥܝܬܗ‬ ‫ܕܒܥܝܢ ܠܗ ܒܩܒܪܐ‪ .‬ܒܥܝܬܗ‬ ‫ܐܡܪܐ ܘܐܠ ܐܫܟܚܬܗ‪ .‬ܡܢܐ‬ ‫ܐܡܪܐ ܘܐܠ ܐܫܟܚܬܗ‪ .‬ܡܢܐ‬ ‫̈‬ ‫ܓܝܪ ܐܡܪ ܡܐܠܟܐ‪ .‬ܒܥܝܢ ܐܢܬܝܢ‬ ‫ܓܝܪ ܒܥܝܢ ܐܢܬܝܢ ܠܚܝܐ ܒܝܬ‬ ‫̈ܡܝܬܐ‪ .‬ܐܠ ܐܫܟܚܢܝܗܝ ܬܡܢ‪ .‬ܐܠ ܚܝܐ ܒܝܬ ̈ܡܝܬܐ‪ .‬ܐܠ ܐܫܟܚܢܝܗܝ‬ ‫ܬܡܢ‪ .‬ܕܐܠ ܗܘܐ ܓܝܪ ܩܒܪܐ ܗܘ‬ ‫ܗܘܐ ܓܝܪ ܩܒܪܐ ܐܝܬܘܗܝ‬ ‫̣‬ ‫ܐܬܪܗ ܕܗܢܐ܆ ܐܐܠ ܫܡܝܐ‪ .‬ܡܢܐ ܐܬܪܗ ܕܗܢܐ‪ .‬ܐܐܠ ܫܡܝܐ‪.‬‬ ‫ܒܥܝܢ ܐܢܬܝܢ ̇‬ ‫ܒܥܝܢ ܐܢܬܝܢ ܥܠ ܐܪܥܐ܆ ̇‬ ‫ܠܡܢܐ ̈‬ ‫ܠܗܘ ܕܥܠ‬ ‫ܠܗܘ‬ ‫̇‬ ‫ܟ̈ܪܘܒܐ ܝܬܒ܀‬ ‫ܕܗܐ ܥܠ ܟ̈ܪܘܒܐ ܝܬܝܒ܀‬

‫‪Song 3:1‬‬ ‫‪f. 70r‬‬

‫‪Song 5:2-6‬‬ ‫‪f. 179v‬‬

ANNOTATION PRACTICES

243

BIBLIOGRAPHY

Abramowski, Luise. “Zur geplanten Ausgabe von Brit. Mus. add. 12156.” In Texte und Textkritik. Eine Aufsatzsammlung, edited by Jürgen Dummer. Berlin: Akademie-Verlag, 1987.

Assemanus, J.S. and S.E. Assemanus. Bibliothecae Apostolicae Vaticanae Codicum Manuscriptorum Catalogus in tres partes distributus. Rome, 1758–1759. Baumstark, Anton. Geschichte der syrischen Literatur mit Ausschluß der christlich-palästinensischen Texte. Bonn: A. Marcus und E. Webers Verlag, 1922.

Briquel-Chatonnet, Françoise, Pier Giorgio Borbone, and Ewa Balicka-Witakowska. “Syriac Codicology.” In Comparative Oriental Manuscript Studies: An Introduction, edited by Alessandro Bausi, Pier Giorgio Borbone, Françoise Briquel-Chatonnet, Paola Buzi, Eugenia Sokolinski, Jos Gippert, Caroline Macé, Witold Witakowski, Laura Emilia Parodi. Hamburg: Tredition, 2015: 252–266. Brock, Sebastian. “Some New Syriac Texts Attributed to Hippolytus.” Le Muséon 94, no. 1–2 (1981): 177–200.

———. “Without Moses of Nisibis, Where Would We Be? Some Reflections on the Transmission of Syriac Literature.” Journal of Eastern Christian Studies 56 (2004): 15–24.

Brock, Sebastian and Lucas Van Rompay. Catalogue of the Syriac Manuscripts and Fragments in the Library of Deir al-Surian, Wadi alNatrun (Egypt). Leuven: Peeters, 2014.

Ernst, Allie M. Martha from the Margins: The Authority of Martha in Early Christian Tradition. Leiden: Brill, 2009.

Evelyn-White, Hugh G. The Monasteries of the Wadi ‘n Natrun. The History of the Monasteries of Nitria and of Scetis. New York: The Metropolitan Museum of Art Egyptian Expedition, 1932. Friedrich, Michael and Cosima Schwarke (eds.). One-Volume Libraries: Composite and Multiple-Text Manuscripts. Berlin: De Gruyter, 2016.

244

MARION PRAGT

Garitte, Gérard. Traités d’Hippolyte sur David et Goliath, sur le Cantique des cantiques et sur l’Antéchrist. CSCO 263-264. Leuven: Peeters, 1965.

Gumbert, Jan Peter. Codicologische eenheden. Opzet voor een terminologie. Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen, 2004.

Halleux, André de. “Une version syriaque révisée du commentaire d’Hippolyte sur Suzanne.” Le Muséon 101, no. 3–4 (1988): 297–341. ———. “Hippolyte en version syriaque.” Le Muséon 102, no. 1–2 (1989): 19–42. Heal, Kristian S. “Catalogues and the Poetics of Syriac Manuscript Cultures.” Hugoye: Journal of Syriac Studies 20.3 (2017): 375–417.

Johnston, Michael and Michael van Dussen (eds.). The Medieval Manuscript Book: Cultural Approaches. Cambridge: Cambridge University Press, 2015.

Kruisheer, Dirk. “Ephrem, Jacob of Edessa, and the Monk Severus. An Analysis of Ms. Vat. Syr. 103, ff. 1–72.” In Symposium Syriacum VII, edited by René Lavenant. Rome: Pontificio Instituto Orientale, 1998. ———. “Reconstructing Jacob of Edessa’s Scholia.” In The Book of Genesis in Jewish and Oriental Christian Interpretation: A Collection of Essays, edited by Judith Frishman and Lucas Van Rompay. Leuven: Peeters, 1997).

Kwakkel, Erik. “Towards a Terminology for the Analysis of Composite Manuscripts.” Gazette du livre médiéval 41 (2002): 12–19. Lagarde, Paul de. Analecta Syriaca. Leipzig: Teubner, 1858.

Langerbeck, Hermannus. Gregorii Nysseni In Canticum Canticorum. Leiden: Brill, 1960. Lied, Liv Ingeborg and Hugo Lundhaug (eds.) Snapshots of Evolving Traditions: Jewish and Christian Manuscript Culture, Textual Fluidity, and New Philology. Berlin: De Gruyter, 2017.

McNamee, Kathleen. “Sigla in Late Greek Literary Papyri.” In Signes dans les textes, textes sur les signes. Érudition, lecture et écriture

ANNOTATION PRACTICES

245

dans le monde gréco-romain, edited by Gabriel Nocchi Macedo and Maria Chiara Scappaticcio. Liège: Presses universitaires de Liège, 2017.

Norris, Richard A. Gregory of Nyssa: Homilies on the Song of Songs. Atlanta: Society of Biblical Literature, 2012.

Pitra, J.B. Analecta Sacra Spicilegio Solesmensi parata. Paris: A. Jouby et Roger, Bibliopolis, 1883.

Romeny, Bas ter Haar. “The Identity Formation of Syrian Orthodox Christians as Reflected in Two Exegetical Collections.” Parole de l’Orient 29 (2004): 103–121. ———. “Greek or Syriac? Chapters in the Establishment of a Syrian Orthodox Exegetical Tradition.” In Studia Patristica XLI: Papers Presented at the Fourteenth International Conference on Patristic Studies Held in Oxford 2003, edited by Frances Young, Mark Edwards and Paul Parvis. Leuven: Peeters, 2006.

———. “The Peshitta of Isaiah. Evidence from the Syriac Fathers.” In Text, Translation and Tradition: Studies on the Peshitta and Its Use in the Syriac Tradition Presented to Konrad D. Jenner on the Occasion of His Sixty-Fifth Birthday, edited by Wido Van Peursen and Bas ter Haar Romeny. Leiden: Brill, 2006.

———. “The Greek vs. the Peshitta in a West Syrian Exegetical Collection.” In The Peshitta: Its Use in Literature and Liturgy. Papers Read at the Third Peshitta Symposium, edited by Bas ter Haar Romeny. Leiden: Brill, 2006.

———. “Ephrem and Jacob of Edessa in the Commentary of the Monk Severus.” In Malphono w-Rabo d-Malphone. Studies in Honor of Sebastian P. Brock, edited by George A. Kiraz. Piscataway, NJ: Gorgias Press, 2008.

———. “The Formation of a Communal Identity among West Syrian Christians: Results and Conclusions of the Leiden Project.” Church History and Religious Culture 89, no. 1 (2009): 1–52. Rucker, Ignaz. Florilegium Edessenum anonymum. Munich: Verlag der Bayerischen Akademie der Wissenschaften, 1933.

246

MARION PRAGT

Smith, Yancy, The Mystery of Anointing: Hippolytus' Commentary on the Song of Songs in Social and Critical Contexts. Texts, Translations, and Comprehensive Study (Piscataway: Gorgias Press, 2015).

Steinová, Eva, Notam superponere studui. The Use of Annotation Symbols in the Early Middle Ages (Turnhout: Brepols, 2019).

Van den Eynde, Ceslas, La version syriaque du Commentaire de Grégoire de Nysse sur le Cantique des Cantiques: ses origines, ses témoins, son influence (Leuven: Bureaux du Muséon, 1939).

Wright, William. Catalogue of Syriac Manuscripts in the British Museum Acquired Since the Year 1838. London: British Museum, 1870– 1872.

SCRIBES AND THE BOOK OF REVELATION IN EASTERN NEW TESTAMENTS T. C. SCHMIDT This paper examines the reception of the book of Revelation in the hands of eastern Christian scribes. It first explores Revelation’s troubled standing within many eastern churches and then discusses the role scribes played in authorizing, preserving and even influencing Revelation’s status and textual makeup. Such included decisions over whether Revelation should be bound with a biblical or nonbiblical text, or if it should be bound alone; and whether, how, and where a scribe should add a colophon, a preface, or a title. Other choices concerned the use of accents, vocalizations, artwork, script style and a host of still more possibilities as to whether Revelation should be transcribed with accompanying interpretive aids like scholia, marginal notes or even fulllength commentaries. While some of these decisions may seem quite innocuous, this paper shows that even the most minor could promote or demote Revelation’s authority in the eyes of later eastern Christian readers.

The book of Revelation was often viewed suspiciously by eastern churches and was hence excluded from many of their New Testament manuscripts. In the Greek churches, for example, Oecumenius (c. 550 CE) explains that the “nonsense of the majority” reject Revelation; and Nicephoras, the Patriarch of Constantinople (c. 814 CE), admits as much when he classifies Revelation as a “disput-

247

248

T. C. SCHMIDT

ed” book.1 Doubts were even greater in the Syriac church, where Dionysius bar Ṣalibi (c. 1171 CE) tells us that “many of the teachers are in doubt concerning the Revelation of John, and they say that it is not his.”2 So Bar Hebraeus (c. 1286 CE), the Catholicos of the Syriac Orthodox Church, asserts that “Revelation, which is ascribed under the name of John the Apostle, is not his.”3 Revelation received similar treatment in the Caucuses. There it was not even translated into Georgian until the tenth century,4 and the Armenian writer Nerses of Lambron (c. 1179/80 CE) claims that Revelation was despised by many “who had ridiculous thoughts about it.”5 All these doubts meant that Revelation was not copied frequently in these churches, leaving us with far fewer extant manuscripts of Revelation than of any other New Testament text. But Oecumenius, Commentary on the Apocalypse 12.20.4, translation from John N. Suggit, trans., Oecumenius: Commentary on the Apocalypse, FC 112 (Washington, D.C.: Catholic University of America Press, 2006), 203. Nicephoras, Chronographia Brevis (PG 100.1055–1060), also known as “The Stichometry of Nicephorus.” See also Brooke Foss Westcott, A General Survey of the History of the Canon of the New Testament, 4th ed. (London: Macmillan, 1875), 552–554, Appendix D xix; Wilhelm Schneemelcher, New Testament Apocrypha: Gospels and Related Writings. I, trans. Robert McLachlan Wilson (Cambridge, UK: James Clarke Company, 1991), 41–42. 2 Dionysius bar Ṣalibi, Commentary on the Apocalypse 3.12–14; my translation from Jaroslav Sedláček, ed., Dionysius Bar Salibi: In Apocalypsim, Actus et Epistulas Catholicas, CSCO/Scriptores Syri, 53/18 (syriac), 60/20 (Latin) (Paris: Typographeo Reipublicae, 1909), vol 18 p.3. 3 Bar Hebraeus, Nomocanon 7.9; my translation from Giuseppe Simone Assemani, Bibliotheca orientalis Clementino-Vaticana: scriptoribus syris Nestorianis (Rome: Sacra Congregationis de Propaganda Fide, 1725), 3.1.7, p.15 note 5. 4 Jeff W. Childers, “The Georgian Version of the New Testament,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes, 2nd ed. (Leiden: Brill, 2013), 293–328 at 297. 5 This passage is from a colophon made by Nerses and is translated in Robert W. Thomson, trans., Nerses of Lambron: Commentary on the Revelation of Saint John (Leuven and Dudley, MA: Peeters, 2007), 4–5. 1

SCRIBES AND THE BOOK OF REVELATION

249

those Revelation manuscripts that do remain can tell us much about the practices employed by scribes when copying Revelation and, therefore, can also tell us about why scribes chose to copy Revelation, the peculiar challenges they faced when doing so, and, especially, how they promoted Revelation as an authentic New Testament text. Hence in this paper I examine the reception of Revelation in the hands of eastern scribes in order to identify how they interacted with Revelation’s text and how they promoted its status. These interactions include decisions about whether Revelation should be bound with a biblical or nonbiblical text, or if it should be bound alone; and whether, how, and where a scribe should add a colophon, a preface, or a title. Other choices include the use of accents, vocalizations, artwork, and the transcription of Greek names and cryptograms. While some of these decisions seem quite innocuous, even the most minor could promote or demote Revelation’s authority in the eyes of later readers. A sloppy or staid transcription of Revelation that poorly communicated certain subtleties of the text and that was bound between a random assortment of nonbiblical works would of course not appear as authoritative as would a copy of Revelation that was lavishly prepared in beautiful style and accompanied by a fitting preface, accents, colophon, and perhaps even a commentary promoting and defending Revelation’s inclusion in the New Testament.

THE DANGER OF REVELATION

But before examining eastern scribal practice, I should first say a word about the special danger and the special difficulty that Revelation held for any scribe, no matter how skilled. This is expressed in Revelation 22:18–19, which promises to add “the plagues described in this book” to the one who adds to the text, and further warns that one’s share “of the tree of life and of the holy city” will be taken away from any who take away from it.6 So when discussing a All quotations from the New Testament are my own from the Greek text found in UBS 5. 6

250

T. C. SCHMIDT

textual variant in Revelation, this warning prompted Irenaeus (c.185 CE) to declare that “one shall have no light punishment [inflicted] who adds or subtracts (προσθέντος ἢ ἀφελόντος) anything from Scripture.”7 Irenaeus was not wrong to call for care even in these minor matters, for the smallest slip could indeed change the meaning of a passage. Manuscript 1248 of Revelation provides a case in point; there, at Revelation 12:10, a scribe has not written “Christ” but “Antichrist,8 thus quite drastically altering the sense of the passage. Transcribing Revelation was therefore a fearful enterprise for the scribe. Accordingly, a Greek scribe named Theodosius appended a colophon shortly after Revelation’s warning. There he begs forgiveness “if I have tripped in any sentence or phrase or in the order of words or in an accent or in a single word or in any other thing, unwittingly or wittingly—pity Theodosius, O Lord, sinner above all men, and my spiritual children, friends and brethren. Amen.”9

THE DECISION TO COPY REVELATION

Now eastern scribes appear to have avoided the danger of copying Revelation by simply choosing not to copy it at all. This decision is the first aspect of eastern scribal tendencies, or lack of tendencies, that I want to highlight in regards to Revelation’s reception in the East. But the choice to neglect Revelation was likely not because scribes feared the warning of Revelation, but more probably because they did not consider it to be a member document of the New Testament. Irenaeus, Against Heresies 5.30.1; ANCF translation modified with reference to the Greek of Adelin Rousseau, Louis Doutreleau, and Charles Mercier, eds., Irénée de Lyon: Contre les hérésies livre V, Sources Chrétiennes 153 (Paris: Les Editions du Cerf, 1969). Here Irenaeus appears to have exchanged Revelation’s unusual, albeit synonymous, ἐπιτίθημι (to add) for the more usual idiom προστίθημι (to add). 8 H. C. Hoskier, ed., The Complete Commentary of Oecumenius on the Apocalypse (University of Michigan: Ann Arbor, 1928), vol 1.746 MS 250. I am grateful for Maurice Robinson, who first made me aware of this passage. 9 Translation from Hoskier, vol. 1 p. 637 MS 200. 7

SCRIBES AND THE BOOK OF REVELATION

251

Consequently, as I mentioned above, relatively few eastern manuscripts of Revelation even exist. This is in part due to the fact that it was not included in the first biblical translation projects of many eastern languages. In Syriac, it seems to have been passed over with the Old Syriac (3rd century) and Peshitta (4/5th century) versions and had to wait until Philoxenus of Mabbug (c. 507/8 CE) and Thomas of Harkel (c. 615/6 CE) translated it for their versions of the New Testament.10 Despite these efforts, the earliest relatively complete Syriac manuscript of Revelation dates from 1088 CE and is actually an anonymous commentary on the book. Aside from a few fragments from a sixth-century commentary manuscript, 11 only one other Syriac manuscript dates before the printing press, and fewer than two dozen manuscripts of Revelation in Syriac even exist.12 As expected, Peshitta Syriac manuscripts of the whole New It is not totally certain that Philoxenus actually did include Revelation in his Syriac translation of the New Testament; see Martin Heide, “Die syrische Apokalypse oder Offenbarung an Johannes: Kritische Edition der harklensischen Textzeugen,” in Studien zum Text der Apokalypse II, ed. Marcus Sigismund and Darius Müller, Arbeiten zur Neutestamenilichen Textforschung 50 (Germany: De Gruyter, 2017), 106; Peter J. Williams, “The Syriac Versions of the New Testament,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes, 2nd ed. (Leiden: Brill, 2012), 153. But Gwynn gives many good reasons for supposing that Philoxenus did include Revelation; see John Gwynn, The Apocalypse of St. John in a Syriac Version Hitherto Unknown, edited from a MS. in the Library of the Earl of Crawford and Balcarres (Dublin: Hodges, 1897), xciv–xcix; T. K. Abbott, review of The Apocalypse of St. John, in a Syriac Version Hitherto Unknown, by John Gwynn, Hermathena 10, no. 23 (1897): 27–35. 11 These are cataloged and discussed in Sebastian P. Brock, Catalogue of Syriac Fragments (New Finds) in the Library of the Monastery of Saint Catherine, Mount Sinai (Athens: Mount Sinai Foundation, 1995), Sp. 23 p.17–19, 92– 94 (Syriac), 94–96 (English), 197–199 (plates). 12 According to Heide, there are 15 Syriac manuscripts of Revelation, only two of which date before the printing press and some of which can no longer be found; see Martin Heide, “Die Syrische Johannes Apokalypse. Zum Gegenwaertigen Stand der Forschung,” in Die Johan10

252

T. C. SCHMIDT

Testament dating from the seventh, eighth, and ninth centuries omit Revelation.13 Of special note are two manuscripts from the twelfth to thirteenth centuries: both of these contain the Harklean New Testament, but in what was likely a conscious scribal choice, the scribes of both manuscripts omitted Revelation, even though Thomas of Harkel had originally included it.14 The same pattern of scribal neglect emerges in other eastern traditions. The earliest Revelation manuscript in Georgian comes from 978 CE when Euthymius first translated it; with Armenian, the earliest dates to the twelfth century; and with Slavonic, we must go to the thirteenth century to find a Revelation manuscript.15 The nesoffenbarung: ihr Text und ihre Auslegung, ed. Michael Labahn and Martin Karrer, Arbeiten zur Bibel und ihrer Geschichte 38 (Leipzig: Evangelische Verlagsanstalt, 2012), 71–81. Heide only lists one commentary manuscript (an anonymous Syriac commentary) and excludes five manuscripts of Dionysius bar Ṣalibi’s commentary (four of which date from the 15th century and the other from the 19th century). For the manuscripts of Dionysius bar Ṣalibi’s commentary, see Sedláček, Dionysius Bar Salibi: In Apocalypsim, Actus et Epistulas Catholicas, vol 18 p. 1. More manuscripts of Revelation can no doubt be found; see for example the following catalog: http://syri.ac/digimss/sortable?keys=Revelation&items_per_page=100. 13 Andreas Juckel, “Bible, New Testament Manuscripts,” in Gorgias Encyclopedic Dictionary of the Syriac Heritage: Electronic Edition, ed. Sebastian P. Brock et al. (Piscataway, NJ: Gorgias Press, 2011; online ed. Beth Mardutho, 2018), https://gedsh.bethmardutho.org/Bible-New-Testamentmanuscripts. 14 MS. Cambridge, Add. 1700 (1069/70 CE) and MS. Oxford, New Coll. 333 (12th/13th cent.); see Juckel, sec. 5c. But note, MS. New Coll. 333 technically breaks off after Hebrews 11:27, so further investigation should be made to determine whether it originally included Revelation. 15 Childers, “The Georgian Version of the New Testament,” 316 note 128; Frederick Cornwallis Conybeare, The Armenian Version of Revelation (London: Text and Translation Society, 1907), 67–77; Robert W. Thomson, trans., Nerses of Lambron, 21–22; Robert W. Thomson, trans., Nersēs of Lambron: Commentary on the Dormition of Saint John: Armenian Text and Annotated Translation (Leiden: Brill, 2017), 25; Thomas Hilary Oller, “The Nikol’skij Apocalypse Codex and Its Place in the Textual History of Medieval Slavic Apocalypse Manuscripts” (Ph.D. Diss., Brown University,

SCRIBES AND THE BOOK OF REVELATION

253

most recent survey of Arabic manuscripts of Revelation was done by Georg Graf in 1929 and 1947. He reports that the earliest manuscripts date from the thirteenth century, though there are abundant earlier manuscripts for other New Testament works.16 The two earliest Arabic Revelation manuscripts I have found date from the twelfth or thirteenth centuries and are of Egyptian origin: the first is an Arabic-Bohairic Coptic diglot mentioned by Horner, which dates “probably not later than 1200,” and the other is from St. Catherine’s monastery, dating from the twelfth to thirteenth centuries.17 Coptic is the only eastern language which has manuscripts of Revelation from before the tenth century, going back to perhaps as early as the fifth or sixth centuries, though dating them is difficult.18 Coptic Revelation manuscripts also evidence a diversity of kind. Several of them are miniature codices, suggesting personal, not ecclesiastical use, but others are liturgical in nature; and some are large-format codices, a few of which have evidence of lectionary notations.19 Revelation also existed in at least three different Coptic 1993), 518, http://search.proquest.com/pqdtglobal/docview/304048523 /abstract/E5B9BBDF886949E6PQ/1. 16 Georg Graf, “Arabische Übersetzungen Der Apokalypse,” Biblica 10, no. 2 (1929): 170–94; Georg Graf, Geschichte der Christlichen Arabischen Literaturur (Vatican City: Biblioteca apostolica vaticana, 1947), 1.182–184. 17 George William Horner, The Coptic Version of the New Testament in the Northern Dialect (London: Clarendon Press, 1898), vol. 3 p.lvii-lviii. The shelf number is given by Horner as Petersburg Bibl. Caesariensis Orient. 625. For the manuscript at Catherine’s monastery, see Arabic MS 85, Microfilm 5014, found at https://www.loc.gov/resource/amedmonastery.0 0279386693–ms/?sp=10. 18 There are, however, a few small Syriac fragments of one commentary on Revelation, whose manuscript dates to the sixth century; see Brock, Catalogue of Syriac Fragments. Of course, I exclude Greek Revelation manuscripts from among eastern manuscripts, some of which are extremely early. 19 Christian Askeland, “The Sahidic Apocalypse in Early Islamic Egypt,” in Studien zum Text der Apokalypse, ed. Marcus Sigismund, Martin

254

T. C. SCHMIDT

dialects: Sahidic, Bohairic and Fayumic.20 There are even two fragments of Revelation found in Nubian, attesting to the esteem scribes held for Revelation in the southern Egyptian orbit.21 A translation of Revelation was also made into Ethiopic, and at least two commentaries were written; but dates for their representative manuscripts are unknown and may be quite late.22

TEXTUAL CHANGES

However, when eastern scribes did choose to copy Revelation, they then encountered difficulties that their Greek counterparts did not face pertaining to how one should transcribe Greek numerals and Greek names found within the text of Revelation. A prime example may be found with the name “Alpha and the Omega,” which is mentioned three or four times in Revelation.23 Not only were these divine titles that translators would have approached with much caution, but their meaning was also inextricably linked with “Greekness,” in that they formed part of the Greek alphabet. A further difficulty is that Revelation seems to have used a full word for Alpha (Ἄλφα), but a simple alphabetic character for Omega (Ω.) How does one communicate all of this in a non-Greek language? The Syriac translator Thomas of Harkel decided to substitute the first and last letters of the Syriac alphabet. His translation thus reads, “I am the Alaf and the Taw” (‫)ܐܢܐ ܐܠܦ ܘܬܘ‬, with either Thomas or a later scribe fully spelling out the letters.24 Philoxenus Karrer, and Ulrich Schmid, Arbeiten zur Neutestamentlichen Textforschung 47 (Berlin: De Gruyter, 2015), 271–87 at 274–75. 20 Askeland, 274–75. 21 Askeland, 274–75. 22 Roger W. Cowley, The Traditional Interpretation of the Apocalypse of St John in the Ethiopian Orthodox Church (Cambridge: Cambridge University Press, 1983), 73–74, 164. 23 Revelation 1:8, 21:6, 22:13; certain Byzantine texts also have the title at 1:11. 24 My translation of Revelation 22:13 is from Heide, “Die syrische Apokalypse.” Owing to the paucity of manuscript witnesses for these two translations, it is possible that later scribes could have altered this reading and the one given below.

SCRIBES AND THE BOOK OF REVELATION

255

chose a similar method, and then either he or a later scribe spelled out the Syriac letters (‫)ܐܢܐ ܐܠܦ ܘܐܢܐ ܬܘ‬.25 But in Coptic this kind of translational (and transcriptional) decision would have been especially problematic because the language has the letters Alpha and Omega in its own alphabet, but Omega is usually not in the final position.26 Even so, the Sahidic translator kept the Greek letters as is, though certain scribes fully spelled out the second term instead of using a character: ⲁⲛⲟⲕ ⲡⲉ ⲁⲗⲫⲁ ⲁⲩⲱ ⲱⲙⲉⲅⲁ.27 The meaning of these divine names may therefore have been confusing to Coptic audiences, who may not have realized that Omega was the final letter of the Greek alphabet. In his Armenian translation, however, Nerses opted to transliterate the Greek term “Alpha” rather than to substitute the Armenian “Ayb.” But then, for the Greek Omega (Ω), he curiously substituted the Armenian letter Օ (pronounced like “Oh”), so as to say “I am the Alpha and the O (Ես եմ Ալֆա և Օ).” But Nerses’s choice to deploy an Armenian character caused him to interpret the Revelation 22:13 found in Manchester MS. Syr. 2 f. 134r lines 7–8. My translation from pictures of the manuscript, in consultation with Gwynn, The Apocalypse of St. John in a Syriac Version. For more information on this manuscript, see below, note 60. 26 On the Coptic Alphabet, see Rodolophe Kasser, “Alphabets, Coptic,” in The Coptic Encyclopedia, ed. Aziz S. Atiya (New York: Macmillan, 1991). For an example of a Coptic abecedarium, see Princeton MS 97127 at https://catalog.princeton.edu/catalog/5182612, which, according to the catalog, ends with Coptic letters. See also the abecedarium in Harry Reginald Hall, Coptic and Greek Texts of the Christian Period from Ostraka, Stelae, Etc. in the British Museum (London: Longmans, 1905), 35–36, plate 29 #21247 (which appears to end with the Coptic letter Shai). 27 I use the Coptic text found in Christian Askeland, “An Eclectic Edition of the Sahidic Apocalypse of John,” in Studien zum Text der Apokalypse II, ed. Marcus Sigismund and Darius Müller, Arbeiten zur Neutestamenilichen Textforschung 50 (Germany: De Gruyter, 2017), 33–79. For an example of how a Coptic scribe transcribed these titles, see SA 32 158r (NTVMR 601032), found at http://ntvmr.uni-muenster.de/community/ modules/papyri/?zoom=116&left=-4&top=-118&site=INTF&image=6 01032/0/3160/10/632. 25

256

T. C. SCHMIDT

meaning behind the Greek “Omega” in a way different from what a Greek audience may have done. He writes, “For just as Alpha [comes] before the other letters, so he is prior to created things. Also, like Օ, everything is summed up in Him.”28 It seems that the circular structure of the Armenian character led him to conclude that all things are summed up in Christ, rather than to emphasize Christ’s finality as one might do when interpreting in Greek. But Nerses’s choice becomes more interesting still—for happenstance has given to us a manuscript of Revelation in Nerses’s own hand. Yet this version was written before Nerses made his own translation and therefore represents the older Armenian version of the text.29 So here we have the opportunity to view Nerses’s choices as a scribe and to compare them to his later choices as a translator. Turning to Revelation 1:8, Nerses has instead recorded “I am the Ayb (Այբ) and I am the K'ē (Քէ),”30 this time using both the first and the last Armenian letters and spelling them out. This underscores a challenge eastern scribes faced when confronted with passages like Revelation 1:8. Some of them worked in languages that did not have standardized alphabets; consequently when considering how to translate and transcribe the Greek letter Omega, Nerses seems to have been at liberty to choose between the Armenian letter K'ē (Ք) or O (O), both of which might be in the last place of the Armenian alphabet. It seems plausible, therefore, that when he later made a translation of Revelation, he preferred the Armenian character O because he thought that it had a somewhat similar shape as the Greek Omega and that the shape of the letter, not its position in the alphabet, held its primary meaning. Nerses, Commentary on the Apocalypse 1.1 [31–32]; translation from Thomson, Nerses of Lambron: Commentary on the Revelation, 49–50. Armenian text from Meknut’iwn Yaytnut’ean S. Awetaranč’in Yovhannu arareal S. Andrēi ew Aritasay episkoposac’n Kesaru. Jerusalem, 1855. 29 This manuscript is MS. Paris arm. 27. For discussion, see Conybeare, The Armenian Version of Revelation, 68–76. 30 Ես եմ Այբ եւ ես եմ Քէ. MS. Paris arm. 27 (GA 256) f. 293v lines 12–13. Pictures from NTVMR 30256 found at http://ntvmr.unimuenster.de/community/modules/papyri/?zoom=58&left=7&top=108&site=INTF&image=30256/0/5960/10/456 28

SCRIBES AND THE BOOK OF REVELATION

257

Ibn Kātib Qayṣar further highlights many of these difficulties. He decided to translate the two terms into Arabic as “al-ʾAlifā” and “al-ʾAʾ,” which are known in English translation as the Arabic letters “Alif” and “Hamza.” He, or a later scribe, also added Coptic orthography by inserting the Coptic letters Alpha and Omega (Ⲁ and Ⲱ). So his translation reads, “I am al-ʾAlifā Ⲁ and al-ʾAʾ Ⲱ.” In a manuscript dating just a few decades after Ibn Kātib Qayṣar’s death, however, a scribe on one occasion dropped out the Coptic Alpha but kept the Omega as a superscript.31 Yet both the Arabic and the Coptic ortography may have been extraordinarily confusing for contemporary audiences; for while the Arabic word “al-ʾAʾ” can refer to the Arabic letter “hamza” (‫)ء‬, this letter does not always occur at the end of the Arabic alphabet and in fact is often omitted from the alphabet entirely. More confusingly, however, is that “al-ʾAʾ” can actually be a synonym for the first letter of the Arabic alphabet “ʾAlifā” and can even be a word meaning “madness.”32 And, to make a terribly confusing situation much worse, the letter “Omega” in Coptic is not typically the last letter in its alphabet, as I pointed out above. Thus, both the Arabic and the Coptic of Ibn Kātib Qayṣar’s translation of “Alpha and Omega” may not even have contained the last letter of either language’s alphabet. Likely realizing such translational and scribal constraints, Ibn Kātib Qayṣar explains: In this section, with respect to what precedes, what comes after, and his decree, the Lord God says, I am al-ʾAlifā and the al-ʾAʾ. He is the first and the last, by way of analogy, so that he might make this understood. Just as the al-ʾAlifā is the first Greek letter and the al-ʾAʾ is the last, so too God (may he be praised and exalted) is the first and last of all existing things.33

MS. Paris ar. 67 f.7r lines 13 and 16 read (Ⲱ ‫ )أﻧﺎ اﻷﻟﻔﺎⲀو أﻧﺎ اﻷء‬with both the Alpha and Omega written in small font above the line only on line 16. 32 Edward William Lane, Arabic–English Lexicon (London: Williams and Norgate, 1863), 1 col 1; 2901 cols 1–2. 33 Ibn Kātib Qayṣar, Commentary on the Apocalypse §5, modified from the following translation: Stephen J. Davis and T.C. Schmidt, trans., “Ibn Kātib Qayṣar, Commentary on the Apocalypse of John, Ch. 1–3,” in Reve31

258

T. C. SCHMIDT

CLOAKING THE TEXT

Though eastern scribes were hampered in some respects compared to their Greek counterparts, they did hold an advantage in that they could more easily disguise the text of Revelation and hide potentially damaging information from prying eyes. And they might do this by recording information in a second language, something that the frequently monolingual Greek exegetes and scribes would have had trouble doing. Cicero, for instance, was known to have done this by quoting Greek in his Latin letters to ensure that only the initiated could understand what he meant.34 With Revelation, we find a similar practice put to work regarding the number of the beast, 666, in Revelation 13:18. In this passage, the reader is confronted by a terrible beast “coming up out of the earth” who performs wondrous signs and whose name is signified by a number. The narrator declares “let the one who has wisdom calculate the number of the beast, for it is the number of a man, and his number is 666.” Textual critics believe that, originally, this number was not represented by Greek numerals (χξϛ), but instead was fully spelled out as “six hundred sixty-six” (ἑξακόσιοι ἑξήκοντα ἕξ).35 Now Greek numerals are simply alphabetic characters that represent a given number, so here Revelation presents a cipher that hides the name of the earth beast. To solve it and discover the beast’s name, one must first find a name, then convert the letters of the name to numerical values and then, final-

lation 1–3 in Christian Arabic Commentary: John’s First Vision and the Letters to the Seven Churches, by Stephen J. Davis, T.C. Schmidt, and Shawqi Talia, Christian Arabic Texts in Translation 1 (New York: Fordham University Press, 2019), 94. 34 Han Baltussen, “Cicero’s Translation of Greek Philosophy: Personal Mission or Public Service?,” in Complicating the History of Western Translation: The Ancient Mediterranean in Perspective, ed. Siobhán McElduff and Enrica Sciarrino (New York: Routledge, 2014), 37–48 at 42. 35 But Greek manuscripts do differ on whether or not the number is spelled out or presented in numeral form.

SCRIBES AND THE BOOK OF REVELATION

259

ly, add them together. Any name whose sum is 666 could qualify as the name of the earth beast.36 Notable for us is that this passage requires the reader to know how Greek numerals functioned, a great difficulty for the translator and the non-Greek scribe alike. Bar Ṣalibi thus cautions the reader that such a calculation must be made “in the alphabet of the Greeks” and not “in the alphabet of the Syrians.” He, like many other exegetes, runs through a list of possible names, one of which is “Titan.” Bar Ṣalibi pauses here and warns his audience once again not to calculate according to Syriac numerals, “for Tau in the Greek language is three hundred, while with us Tet is nine.”37 Clearly, his non-Greek audience would have been at a severe disadvantage when it came to interpreting this passage. This difficulty of determining how to properly treat Greek numerals is quite evident in the Arabic-speaking world. There, the number of the beast was of special interest because the Quran also speaks of a “beast of the earth” (‫ض‬ ِ ‫ن ٱ ْ� �� ْر‬ َ ِ ّ‫ )دَآب َّة ً م‬that performs signs.38 Certain Quranic commentators claim that the name of this beast is ḥasāsah (‫)حساسة‬, “the sensual,” which seems a mistaken attempt at vocalizing the Greek number 666 (χξϛ) as if it spelled a name.39 The Arabic commentator Būlus al-Būshī was not, however, misled by this approach and knew well that the number called for a mathematical solution. Like Cicero so many centuries before him, In a famous passage, the Greek writer Irenaeus goes through this process and suggests various possibilities; see Irenaeus, Against Heresies 5.29–30. 37 Dionysius bar Ṣalibi, Commentary on the Apocalypse 23.11–13; my translation from Sedláček. For similar discussions, see also Nerses of Lambron, Commentary on the Apocalypse 13.38 [192]; Ibn Kātib Qayṣar, Commentary on the Apocalypse §64 found in al-Qummuṣ Armāniyūs Ḥabashī Shattā al-Birmāwī, ed., Tafsīr Sifr Al-Ru’yā Li-l-Qiddīs Yūḥannā al-Lahūtī LiIbn Kātib Qayṣar (Cairo: Maktabat al-Maḥabbah, 1939), 274–75. 38 Qu’ran 27.82; Arabic text from https://al-quran.info. 39 David Brady, “The Book of Revelation and the Qur’an: Is There a Possible Literary Relationship?,” Journal of Semitic Studies 23, no. 2 (September 21, 1978): 216–25 at 222–23 and notes, https://doi.org/ 10.1093/jss/23.2.216. 36

260

T. C. SCHMIDT

he used his scribal tools to impart a secret to his readers. He writes in Arabic that John “means that when you combine the letters, count each letter separately and then add the total, the result is this understanding: his name is,” and then Būlus al-Būshī switches to Coptic to write “ⲙⲁⲙⲉⲧⲓⲟⲥ” or “Mohammed.”40 Of course to declare that the Islamic prophet Mohammed was the earth beast of Revelation would have been quite risky in Būlus al-Būshī’s day, so our commentator insulated himself from such danger by recording the name of Mohammed in a language other than Arabic. Any Muslim reader would likely not see through the disguise, though his Coptic audience would not have had so much trouble.41 Still, more than one Christian scribe simply left the term blank, either out of an inability to write in Coptic or perhaps a fear of being caught.42 Two other scribes did the opposite and went so far as to record underneath each Coptic letter its valuation so that the numerically challenged Coptic reader would be able to add up the name of Mohammed to 666.43

Būlus al-Būshī, Commentary on the Apocalypse §24, translation from Shawqi Talia, “Būlus Al-Būshi’s Arabic Commentary on the Apocalypse of St. John: An English Translation and a Commentary” (Ph.D. Diss., Catholic University of America, 1987), 198. At least one manuscript contains the whole phrase “his name is Mohammed” in Coptic; see Nagi Edelby, “Commentaire de l’apocalypse de Būlus Al-Būšī (évêque du Caire en 1240 AD): Étude, édition critique, traduction et index exhaustif” (Ph.D. Diss., Université Saint-Joseph, 2015), 97. 41 For further discussion, see Stephen J. Davis, “Ibn Kātib Qayṣar on Visions, Angels, Prophets, and Dreams,” in Revelation 1–3 in Christian Arabic Commentary: John’s First Vision and the Letters to the Seven Churches, by Stephen J. Davis, T.C. Schmidt, and Shawqi Talia, Christian Arabic Texts in Translation 1 (New York: Fordham University Press, 2019), 22–40 at 24–27. 42 Talia, “Būlus Al-Būshi’s Arabic Commentary on the Apocalypse of St. John: An English Translation and a Commentary,” 199. 43 Talia, 199. 40

SCRIBES AND THE BOOK OF REVELATION

261

BINDING

The above discussions describe the challenges and opportunities faced by eastern scribes when copying the text of Revelation. But eastern scribes would also have had the opportunity to influence the status of Revelation in ways beyond mere transcriptional practices; and these include deciding what biblical or nonbiblical texts to place before and after Revelation, or whether to let it stand on its own. If, for example, a scribe decided to place Revelation alongside miscellaneous Christian writings, and not with other New Testament texts, a reader might be persuaded to think that Revelation was not an indispensable part of the New Testament. As Bruce Metzger has pointed out, of the approximately 300 Greek Revelation manuscripts known to scholars, more than 40 have the “unusual” phenomenon of placing alongside Revelation nonbiblical materials from authors like Justin Martyr, Basil the Great, Gregory of Nyssa, Gregory Nazianzus, John Chrysostom, Theodoret, and John of Damascus.44 J.K. Elliott further shows that 139 other Greek manuscripts have Revelation bound alone,45 thus meaning that more than half of Greek Revelation manuscripts are not bound with any other New Testament text. This is quite different from other New Testament texts, which, according to Von Soden’s list, are normally bound with other parts of the New Testament.46 Bruce M. Metzger, “The Future of New Testament Textual Studies,” in The Bible as Book: The Transmission of the Greek Text, ed. Scot McKendrick and Orlaith O’Sullivan (London: Oak Knoll Press, 2003), 201–8 at 205–6. 45 J. K. Elliott, “The Distinctiveness of the Greek Manuscripts of the Book of Revelation,” The Journal of Theological Studies 48, no. 1 (1997): 116– 24 at 117. Metzger claims that these manuscripts “are the quires containing Revelation taken out of the middle of some general theological book”; Metzger, “The Future of New Testament Textual Studies,” 205. It is unclear, however, how Metzger knows that these were taken from general theological books. 46 See Hermann Freiherr von Soden, Die Schriften des Neuen Testaments: Untersuchungen, vol. 1.1 (Vandenhoeck und Ruprecht, 1911), 102– 248. 44

262

T. C. SCHMIDT

When turning to non-Greek manuscripts, we find that their scribes made similar choices as their Greek counterparts. Of the two dozen or so Syriac manuscripts of Revelation, three are bound with miscellaneous materials, and another two are bound alone.47 No Georgian version of Revelation can be found bound with another biblical text.48 In Armenian, there are numerous New Testament manuscripts that contain Nerses’s translation of Revelation; but of those that contain the older pre-Nersian translation of Revelation, four out of six are bound with miscellaneous texts rather than New Testament texts.49 In Slavonic, the majority of Revelation manuscripts are not bound with biblical texts.50 All this suggests that Revelation had a relatively uncertain relationship with the New Testament of many eastern churches because scribes so frequently did not place it amongst New Testament writings. Such examples also highlight the powerful role scribes played in promoting or demoting the status of Revelation by, at times, quite literally including or excluding it from New Testament manuscripts.51 Ian Beacham, “The Harklean Syriac Version of Revelation: Manuscripts, Text and Methodology of Translation from Greek” (Ph.D. Diss., University of Birmingham, 1990), 21–68. 48 I confirmed this via personal correspondence with Jeff Childers on January 9, 2019. 49 Conybeare, The Armenian Version of Revelation, 67–77. 50 For this, see the discussion of Slavonic manuscripts given in Oller, “The Nikol’skij Apocalypse Codex and Its Place in the Textual History of Medieval Slavic Apocalypse Manuscripts,” 518–49. 51 A similar phenomenon can be observed in Syriac, Georgian, and Slavonic manuscripts which frequently include Revelation alongside nonbiblical materials; see J. N. Birdsall, “The Georgian Version of the Book of Revelation,” Le Muséon 91 (1978): 356, 359; Beacham, “The Harklean Syriac Version of Revelation,” 21–68; Oller, “The Nikol’skij Apocalypse Codex and Its Place in the Textual History of Medieval Slavic Apocalypse Manuscripts,” 518–49. Thomson argues that the Armenian version of Revelation was not normally handed down alongside other biblical texts in its early transmission history; Thomson, Nerses of Lambron: Commentary on the Revelation, 3. Later manuscripts, however, do frequently place Revela47

SCRIBES AND THE BOOK OF REVELATION

263

INCLUDING COMMENTARY

Certain scribes might also choose to influence the status of Revelation by placing alongside its text a formal commentary that endeavored to make Revelation more understandable and palatable, or that even overtly defended it. Notably, Revelation scribes frequently chose to do just that. Hence, approximately 30% of Greek manuscripts of Revelation contain commentaries. Similarly, about one third of Syriac manuscripts, including the two earliest, contain commentaries.52 A large percentage of Armenian, Arabic, and Slavonic manuscripts also contain commentaries,53 and all Georgian manuscripts are commentary manuscripts.54 And these commentaries frequently defend the text of Revelation from criticisms involving its authorship, coherency, and doctrinal content, something that no doubt would have caused many readers to think more highly of Revelation. tion alongside biblical texts; see Thomson, Nersēs of Lambron: Commentary on the Dormition, 24–25. 52 For Greek manuscripts of Revelation and their relationship to commentaries, see this author’s dissertation: Thomas C. Schmidt, “The Last Book: Revelation, Commentaries, and the Making of the New Testament” (New Haven, CT, Yale University, 2020), 296. For Revelation manuscripts in Syriac, see note 12. 53 Oller writes regarding the Slavonic tradition, “until Gennadius and his subordinate clergy undertook the compilation of a complete Bible at the end of the 15th c., no one in the Orthodox Church seems to have felt the need to produce a text of Revelation without commentary.” Oller, “The Nikol’skij Apocalypse Codex and Its Place in the Textual History of Medieval Slavic Apocalypse Manuscripts,” 517. Though the Aleksian New Testament is an exception to this, see Oller, 519–20. Nearly 100 Revelation manuscripts in Armenian contain the commentary of Nerses of Lambron; see Thomson, Nerses of Lambron: Commentary on the Revelation, 20. No complete catalog of Arabic Revelation manuscripts has been completed, but the partial catalog published by Graf shows that many include commentaries; see Graf, “Arabische Übersetzungen Der Apokalypse.” 54 I confirmed this via personal correspondence with Jeff Childers on January 9, 2019.

264

T. C. SCHMIDT

WRITING MATERIALS AND PARATEXTUAL ELEMENTS

Eastern Revelation scribes also had the opportunity to influence the status of Revelation through the practical decision of what material should be used as a writing surface. In ancient times, the costconscious scribe would have used papyrus while those interested in a longer lasting, more luxurious product would have selected parchment, “by which” Pliny the Elder (c. 79 CE) states, “immortality is ensured to man.”55 After the Islamic conquests, paper was also available; this was even cheaper than papyrus, but more given to decay. Important for us here is that material choices do more than merely unveil how a particular scribe viewed a text; such choices may actually affect the perceived status of the text. Thus, a reused, low-quality material might cause a reader to despise the status of the text it contains while a luxurious parchment manuscript could do quite the opposite. Aside from choices over material format, a scribe would also have needed to make a host of paratextual decisions, which again could directly affect the text and its status. These include the use of accents (or vocalizations for Syriac and Arabic), punctuation, and many other possibilities like artwork, script style, paragraphing, line breaks, text critical notation, interlinear notation, marginal notes, and abbreviations. Though working amid Latin and Greek tradition in the sixth century, Cassiodorus is one of our only sources who gives instructions on many of these paratextual decisions and, remarkably, acknowledges that such could indeed change how a text was interpreted. When discussing punctuation, for example, he writes, “O eager reader these marks or points of punctuation are, as it were, paths for thoughts and beacon lights for words, and render the Pliny, Natural History 13.21, translation from John Bostock and Henry Thomas Riley, trans., The Natural History of Pliny (London, H.G. Bohn, 1855), 186. Pliny also tells us that parchment was invented because Ptolemy of Egypt forbid the export of papyrus; Pliny, Natural History 13.21; Jerome, Letter 7.2. For discussion on the use of papyrus as a writing material, see Strabo, Geography 17.1.15. 55

SCRIBES AND THE BOOK OF REVELATION

265

readers as easily taught as they would be if they were being instructed by the most intelligible interpreters.”56 Cassiodorus also talks of scribal decisions like paragraphing, line breaks, line spacing, and indenting—indeed, he employed such features himself57—and further demanded that a scribe’s activities should be carried out in a beautiful fashion;58 or, in his words, “a handsome external form [should] clothe the beauty of sacred letters.”59

TWO CASE STUDIES

Many of the above scribal practices can be seen at work in Revelation’s eastern manuscript tradition, where scribes used them to promote the status of Revelation in the eyes of readers. The following are two short case studies. ʾSṭepanús: A Syriac scribe A prominent example of Revelation in Syriac is exhibited by what is likely the one remaining Revelation manuscript of Philoxenus’ Syriac translation.60 This parchment manuscript is located at the John Ryland’s Library, shelfmark Syr.2. As Gwynn and Beacham ably describe, the manuscript dates to the twelfth century, measures Cassiodorus, Institutions, 1.15.12; translation by Leslie Weber Jones, trans., Cassiodorus Senator: An Introduction to Divine and Human Readings (New York: Columbia University Press, 1946). 57 Cassiodorus, Institutions, 1.12.4 (Jones trans.). 58 Cassiodorus, Institutions 1.15.12, 15 (Jones trans.). 59 Cassiodorus, Institutions 1.30.3 (Jones trans., modified). 60 For descriptions of this manuscript, see Gwynn, The Apocalypse of St. John in a Syriac Version, cvi–cxiv; Beacham, “The Harklean Syriac Version of Revelation,” 36–44; J.F. Coakley, “A Catalogue of the Syriac Manuscripts in the John Rylands Library,” Bulletin of the John Rylands University Library Manchester 75, no. 2 (1993): 105–207 at 118–19. For a transcription of the Syriac text of Revelation with accompanying notes, see Gwynn, The Apocalypse of St. John in a Syriac Version, 1–100. I am grateful for Dukhrana Biblical Research, through whom I was able to access pictures of this manuscript: http://www.dukhrana.com/crawford/files/syriac_ms_2–book _of_revelation.pdf. 56

266

T. C. SCHMIDT

27.7 x 18.8cm, and contains around 250 folios.61 Its text is written in two columns of 31 lines each and, according to a colophon at the end of the manuscript, was copied by a scribe named ʾSṭepanús (‫)ܐܣܛܦܢܘܣ‬.62 ʾSṭepanús rarely used vocalizations but occasionally employed them to clarify ambiguities in the text.63 The manuscript contains a 27-book New Testament in the following order: Gospels, Revelation, Acts, Seven Catholic Epistles, and Pauline Epistles (including Hebrews). Revelation is found on folios 120r–134v. At some point in the manuscript’s history, 11 folios containing a document harmonizing the Gospel versions of Jesus’s passion were inserted after the Gospels and before Revelation. Originally, however, Revelation was seamlessly situated between the Gospels and Acts.64 Such an ordering of biblical books is rare amongst Syriac New Testament manuscripts, which typically omit Revelation. The handful that do contain Revelation more often place it at the end of the New Testament, or even bind it alone. As mentioned above, Revelation may also be bound with miscellaneous theological materials.65 Therefore, ʾSṭepanús appears to have promoted Revelation by placing it between New Testament books. And not only that: he further situated it directly after the Gospel of John and then recorded a superscription attributing Revelation to “St. John the Evangelist.” He then likewise concluded Revelation with the attribution “St. John Apostle and Evangelist.”66 Clearly, ʾSṭepanús, or at least his patron, believed that John the Evangelist wrote Revelation and thought that it should be given full membership in the New Beacham, “The Harklean Syriac Version of Revelation,” 36–37; Gwynn, The Apocalypse of St. John in a Syriac Version, cvi–cvii, cxi–cxix. Beacham notes that Gwynn claimed the manuscript has 249 leaves while Gottstein claimed 250; see Beacham, “The Harklean Syriac Version of Revelation,” 36. 62 For a transcription and discussion of Stephen’s colophon, see Gwynn, The Apocalypse of St. John in a Syriac Version, 31–32, 98–100. 63 Gwynn, cvii. 64 Gwynn, cvii–cviii. 65 See note 51. 66 Gwynn, The Apocalypse of St. John in a Syriac Version, cviii. 61

SCRIBES AND THE BOOK OF REVELATION

267

Testament. ʾSṭepanús, therefore, concretely endorsed this view by both his placement of the text and by the two attributions he attaches to it. Despite these efforts, however, one gleans that ʾSṭepanús’s church may not always have agreed with his assessment of Revelation and perhaps disagreed still, for in the margins of his manuscript there are two different series of numbers. One of these divides the manuscript into lectionary lessons; but Revelation, along with James, 2 Peter, 2 John, and 3 John, lack these. The second ̈ 67 Yet series divides the manuscript into 165 sections (‫)ܨܚܚܐ‬. Revelation and the aforementioned Catholic epistles once again lack these. Such omissions hint at a troubled reception in Revelation’s ecclesiastical past; for the book was evidently passed over when ʾSṭepanús’s church divided the New Testament into sections and was likewise omitted when his church’s liturgy was implemented. A similar scenario occurs in the Eastern Orthodox Church today, where Revelation is typically omitted from liturgical readings even though it is considered part of the New Testament.68 Such a situation has arisen, it seems, because when such liturgical matters were determined, Revelation was not considered part of the Eastern Orthodox New Testament.69 A further indication of such a troubled reception in ʾSṭepanús’ church can be seen on the recto of the final folio, where ʾSṭepanús has written “the book of the New Testament; in which there are 165 sections; besides the Revelation and the four epistles.”70 While the phrasing is a bit ambiguous, it appears that ʾSṭepanús is explainGwynn, cviii–cix. Stephen Shoemaker, “The Afterlife of the Apocalypse of John in Byzantine Apocalyptic Literature and Commentary,” in The New Testament in Byzantium, ed. Derek Krueger and Robert Nelson (Washington D.C.: Dumbarton Oaks, 2016), 301–16 at 301 note 2. 69 It is, of course, possible that a New Testament book could be omitted from the liturgy for other reasons, perhaps even because of its high level of sacredness; but I am speaking here of tendencies, not all possibilities. 70 Translation from Gwynn, The Apocalypse of St. John in a Syriac Version, cix. 67 68

268

T. C. SCHMIDT

ing that Revelation is excluded from the 165 sections, not from the New Testament itself.71 But still, our scribe seems to have worried that a reader might think that Revelation was omitted from the New Testament because it was omitted from the sectional numbering; thus ʾSṭepanús felt the need to assert its full-fledged membership. Such apprehension implies that Revelation had only recently settled into place in the New Testament of ʾSṭepanús’ church, if it had settled into place at all. Step‛anos: An Armenian scribe As Cassiodorus suggested to us, beauty was a tool for affirming and elevating the status of a work, and such is well exemplified by the Armenian scribe Step‛anos Goyneric‛anc‛. Step‛anos worked in Cyprus between 1284–1312 CE; and in his colophon to University of Chicago Goodspeed Ms. 229, he writes that the manuscript included, ‘…the holy Gospels, also the Acts of the Apostles, and the Epistles, including the fourteen Epistles of Paul…as well as the frightful and holy Revelation of the Evangelist John….’72 Many folios are now missing, but those that remain are well preserved. The pages of the manuscript itself are skillfully written in dark ink and with beautiful selections of bright red, rose, brown, and teal colorings that mark emphasis and sectional divisions. Before each New Testament book, Step‛anos has included a preface and a table of contents listing chapters. Occasionally other materials are inserted, such as an itinerary of the apostle Paul and a description of his martyrdom. Revelation is the penultimate text in the manuscript, coming immediately after 3 Corinthians and before the Dormition of

Gwynn, cix–cx. My observations for this manuscript and the translation of the colophon are derived from Avedis Krikor Sanjian, A Catalogue of Medieval Armenian Manuscripts in the United States (Berkeley: University of California Press, 1976), 191–200. I also made use of pictures found at http://goodspeed.lib.uchicago.edu/ms/index.php?doc=0229. The manuscript was formally known as MS. Goodspeed Arm. 58. 71 72

SCRIBES AND THE BOOK OF REVELATION

269

John, both of which had a disputed standing in the Armenian New Testament.73 Before transcribing Revelation, Step‛anos inserts a table of contents beginning on folio 227v that enumerates seventy-two chapters of Revelation. In doing so, Step‛anos clearly follows the divisional rubric in the Armenian Revelation commentary of Nerses, who himself had followed Andrew of Caesarea’s popular Greek commentary.74 Step‛anos also transcribes Nerses’s own translation of Revelation and not the older Armenian translation. At folio 230r, Step‛anos marks the beginning of Revelation with beautiful red lettering accompanied by a skillfully drawn bird and floral arrangements, both of which are placed in the margin. Throughout the text of Revelation, Step‛anos utilizes abbreviations in fine ligatures to emphasize the sacredness of certain names like “Jesus Christ” and “God” and employs marginal notes to indicate chapter numbers. The manuscript is a great pleasure to behold, and clearly Step‛anos expended much time, energy, and skill. His effort fulfills Cassiodorus’s belief that the “beauty of the sacred letters” should be clothed in “a handsome external form,”75 and no doubt the beauty of the volume would have tended to raise the status of its contents in the eyes of readers. Step‛anos has thus promoted the Revelation by beautifying its appearance, by binding it amongst New Testament books, and by explicitly labeling it as part of the New Testament in a colophon. But even with all these efforts, the careful reader will still notice how Step‛anos has placed Revelation at the end of the New Testament, sandwiched between two works of disputed status in the Armenian church. This, once again, hints that Revelation’s standing in the Armenian church of Step‛anos’s For a brief discussion on the place of the Dormition of John in the Armenian New Testament, see Thomson, Nersēs of Lambron: Commentary on the Dormition, 24–25. 74 On this, see Robert W Thomson, trans., Nerses of Lambron: Commentary on the Revelation of Saint John (Leuven; Dudley, MA: Peeters, 2007), 14–27 75 Cassiodorus, Institutions 1.30.3 (Jones trans.). 73

270

T. C. SCHMIDT

day may not have been as secure as it seems in his beautiful manuscript.

CONCLUSION

In the words of Cassiodorus, scribal activities were “paths for thought” and “beacon lights” that would guide readers to appropriate understanding of text.76 Hence, eastern scribes could promote the status of a troubled text like Revelation by placing it next to other New Testament texts or by leaving notes indicating the authorship of Revelation, all of which would help the reader to embrace Revelation as a truly apostolic work. Eastern scribes further influenced Revelation’s status by choosing to copy the text on expensive material and through various aesthetic choices such as beautifying its text with colored inks and drawings. Scribes might also include a commentary that defended Revelation’s content or might even compose a colophon explicitly affirming its inclusion in the New Testament, both of which would raise the prestige of Revelation in the eyes of readers. These choices were likely necessary for a text like Revelation, which was difficult enough to understand in its original language but even more so in translation. Such scribal choices therefore also reveal the uncertain status that Revelation had in many eastern churches insofar as Revelation evidently needed such scribal assistance in order for it to be welcomed as a full member of the New Testament.

BIBLIOGRAPHY

Abbott, T. K. Review of The Apocalypse of St. John, in a Syriac Version Hitherto Unknown, by John Gwynn. Hermathena 10, no. 23 (1897): 27–35.

Askeland, Christian. “The Sahidic Apocalypse in Early Islamic Egypt.” Pp. 865–876 in Studien zum Text der Apokalypse, edited by Marcus Sigismund, Martin Karrer, and Ulrich Schmid. Arbeiten zur Neutestamentlichen Textforschung 47. Berlin: De Gruyter, 2015. 76

Cassiodorus, Institutions, 1.15.12 (Jones trans.).

SCRIBES AND THE BOOK OF REVELATION

271

———. “An Eclectic Edition of the Sahidic Apocalypse of John.” Pp. 33–79 in Studien zum Text der Apokalypse II, edited by Marcus Sigismund and Darius Müller. Arbeiten zur Neutestamenilichen Textforschung 50. Germany: De Gruyter, 2017.

Assemani, Giuseppe Simone. Bibliotheca orientalis Clementino-Vaticana: scriptoribus syris Nestorianis. Rome: Sacra Congregationis de Propaganda Fide, 1725. Baltussen, Han. “Cicero’s Translation of Greek Philosophy: Personal Mission or Public Service?” Pp. 37–48 in Complicating the History of Western Translation: The Ancient Mediterranean in Perspective, edited by Siobhán McElduff and Enrica Sciarrino. New York: Routledge, 2014. Beacham, Ian. “The Harklean Syriac Version of Revelation: Manuscripts, Text and Methodology of Translation from Greek.” Ph.D. Diss., University of Birmingham, 1990. Birdsall, J. N. “The Georgian Version of the Book of Revelation.” Le Muséon 91 (1978): 133–144.

Brady, David. “The Book of Revelation and the Qur’an: Is There a Possible Literary Relationship?” Journal of Semitic Studies 23, no. 2 (September 21, 1978): 216–225. Brock, Sebastian P. Catalogue of Syriac Fragments (New Finds) in the Library of the Monastery of Saint Catherine, Mount Sinai. Athens: Mount Sinai Foundation, 1995.

Cassiodorus. Cassiodorus Senator: An Introduction to Divine and Human Readings. Translated by Leslie Weber Jones. New York: Columbia university press, 1946.

Childers, Jeff W. “The Georgian Version of the New Testament.” In The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, edited by Bart D. Ehrman and Michael W. Holmes, second edition. Leiden: Brill, 2013. Coakley, J.F. “A Catalogue of the Syriac Manuscripts in the John Rylands Library.” Bulletin of the John Rylands University Library Manchester 75, no. 2 (1993): 105–207. Conybeare, Frederick Cornwallis. The Armenian Version of Revelation. London: Text and Translation Society, 1907.

272

T. C. SCHMIDT

Cowley, Roger W. The Traditional Interpretation of the Apocalypse of St John in the Ethiopian Orthodox Church. Cambridge: Cambridge University Press, 1983.

Davis, Stephen J. “Ibn Kātib Qayṣar on Visions, Angels, Prophets, and Dreams.” Pp. 22–40 in Revelation 1–3 in Christian Arabic Commentary: John’s First Vision and the Letters to the Seven Churches, by Stephen J. Davis, T.C. Schmidt, and Shawqi Talia. Christian Arabic Texts in Translation 1. New York: Fordham University Press, 2019. Davis, Stephen J. and T.C. Schmidt, trans. “Ibn Kātib Qayṣar, Commentary on the Apocalypse of John, Ch. 1–3.” Pp. 86– 148 in Revelation 1–3 in Christian Arabic Commentary: John’s First Vision and the Letters to the Seven Churches, by Stephen J. Davis, T.C. Schmidt, and Shawqi Talia. Christian Arabic Texts in Translation 1. New York: Fordham University Press, 2019.

Edelby, Nagi. “Commentaire de l’apocalypse de Būlus Al-Būšī (évêque du Caire en 1240 AD): Étude, édition critique, traduction et index exhaustif.” Ph.D. Diss., Université Saint-Joseph, 2015.

Elliott, J. K. “The Distinctiveness of the Greek Manuscripts of the Book of Revelation.” The Journal of Theological Studies 48, no. 1 (1997): 116–24. Foss Westcott, Brooke. A General Survey of the History of the Canon of the New Testament, fourth edition. London: Macmillan, 1875.

Graf, Georg. “Arabische Übersetzungen der Apokalypse.” Biblica 10, no. 2 (1929): 170–94. ———. Geschichte der Christlichen Arabischen Literaturur. Vatican City: Biblioteca apostolica vaticana, 1947.

Gwynn, John. The Apocalypse of St. John in a Syriac Version Hitherto Unknown, edited from a MS. in the Library of the Earl of Crawford and Balcarres. Dublin: Hodges, 1897.

Hall, Harry Reginald. Coptic and Greek Texts of the Christian Period from Ostraka, Stelae, etc. in the British Museum. London: Longmans, 1905.

SCRIBES AND THE BOOK OF REVELATION

273

Heide, Martin. “Die Syrische Johannes Apokalypse. Zum Gegenwaertigen Stand der Forschung.” Pp. 71–81 in Die Johannesoffenbarung: ihr Text und ihre Auslegung, edited by Michael Labahn and Martin Karrer. Arbeiten zur Bibel und ihrer Geschichte 38. Leipzig: Evangelische Verlagsanstalt, 2012. ———. “Die syrische Apokalypse oder Offenbarung an Johannes: Kritische Edition der harklensischen Textzeugen.” Pp. 81–187 in Studien zum Text der Apokalypse II, edited by Marcus Sigismund and Darius Müller. Arbeiten zur Neutestamentlichen Textforschung 50. Germany: De Gruyter, 2017.

Horner, George William. The Coptic Version of the New Testament in the Northern Dialect. London: Clarendon Press, 1898.

Hoskier, H. C., ed. The Complete Commentary of Oecumenius on the Apocalypse. University of Michigan: Ann Arbor, 1928. Juckel, Andreas. “Bible, New Testament Manuscripts.” in Gorgias Encyclopedic Dictionary of the Syriac Heritage: Electronic Edition, edited by Sebastian P. Brock, Aaron M. Butts, George A. Kiraz, and Lucas van Rompay. Piscataway, NJ: Gorgias Press, 2011; online ed. Beth Mardutho, 2018. https://gedsh.bethmardutho.org/Bible-New-Testament-manuscripts.

Lane, Edward William. Arabic–English Lexicon. London: Williams and Norgate, 1863.

Metzger, Bruce M. “The Future of New Testament Textual Studies.” Pp. 201–208 in The Bible as Book: The Transmission of the Greek Text, edited by Scot McKendrick and Orlaith O’Sullivan. London: Oak Knoll Press, 2003.

Nerses of Lambron. Nerses of Lambron: Commentary on the Revelation of Saint John. Translated by Robert W. Thomson. Leuven: Peeters, 2007. ———. Nersēs of Lambron: Commentary on the Dormition of Saint John: Armenian Text and Annotated Translation. Translated by Robert W. Thomson. Leiden: Brill, 2017.

Meknut’iwn Yaytnut’ean S. Awetaranč’in Yovhannu arareal S. Andrēi ew Aritasay episkoposac’n Kesaru. Jerusalem, 1855.

274

T. C. SCHMIDT

Oecumenius. Oecumenius: Commentary on the Apocalypse. Translated by John N. Suggit. Washington, D.C.: Catholic University of America Press, 2006.

Oller, Thomas Hilary. “The Nikol’skij Apocalypse Codex and Its Place in the Textual History of Medieval Slavic Apocalypse Manuscripts.” Ph.D. Diss., Brown University, 1993. Pliny the Elder. The Natural History of Pliny. Translated by John Bostock and Henry Thomas Riley. London: H.G. Bohn, 1855.

Rousseau, Adelin, Louis Doutreleau, and Charles Mercier, eds. Irénée de Lyon: Contre les hérésies livre V. Sources Chrétiennes 153. Paris: Les Editions du Cerf, 1969. Sanjian, Avedis Krikor. A Catalogue of Medieval Armenian Manuscripts in the United States. Berkeley: University of California Press, 1976.

Schmidt, Thomas C. “The Last Book: Revelation, Commentaries, and the Making of the New Testament.” Ph.D. Diss., Yale University, 2020. Schneemelcher, Wilhelm. New Testament Apocrypha: Gospels and Related Writings. I. Translated by Robert McLachlan Wilson. Cambridge, UK: James Clarke Company, 1991.

Sedláček, Jaroslav, ed. Dionysius Bar Salibi: In Apocalypsim, Actus et Epistulas Catholicas. Paris: Typographeo Reipublicae, 1909.

Shoemaker, Stephen. “The Afterlife of the Apocalypse of John in Byzantine Apocalyptic Literature and Commentary.” Pp. 301– 316 in The New Testament in Byzantium, edited by Derek Krueger and Robert Nelson. Washington D.C.: Dumbarton Oaks, 2016. Talia, Shawqi. “Būlus Al-Būshi’s Arabic Commentary on the Apocalypse of St. John: An English Translation and a Commentary.” Ph.D. Diss., Catholic University of America, 1987.

von Soden, Hermann Freiherr. Die Schriften des Neuen Testaments: Untersuchungen. Volume 1.1 Berlin: A. Glaue, 1911.

Williams, Peter J. “The Syriac Versions of the New Testament.” Pp. 143–166 in The Text of the New Testament in Contemporary Re-

SCRIBES AND THE BOOK OF REVELATION

275

search: Essays on the Status Quaestionis, edited by Bart D. Ehrman and Michael W. Holmes, second edition. Leiden: Brill, 2012.

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION SZILVIA SÖVEGJÁRTÓ Glosses in Sumerian literary manuscripts have long been the ugly stepchild of Assyriologists. Text editors were not particularly fascinated by these annotations; they are hard to decipher and usually not included in the composite text edition. Glosses were thus mostly relegated to footnotes or commentaries; however, in some cases, they even remained unmentioned or unnoticed. Despite their relative low number, however, these annotations shed new light on Old Babylonian scribal practices and on one of the earliest hermeneutical practices in the history of writing. This brief introduction to the Sumerian glossographic tradition will focus on the glossators as well as on scribal practices related to the glossing of literary manuscripts.

THE LINGUISTIC BACKGROUND

“The art of writing begins with a single wedge; it has six vocalizations. One of them is ‘60.’ Do you know its name? Whatever you know in Sumerian, do you know how to interpret its hidden meaning? [… Do you know] its synonym, antonym, polysemy? And do you know that all those three cannot be harmonized in the Akkadian language?” With these questions, a Sumerian literary composition reports that a scribe examined his son.1 The challenges preÅke Sjöberg, “Der Examenstext A,” Zeitschrift für Assyriologie und Vorderasiatische Archäologie 64 (1975), 140–150 for ll. 12–13 and 15. The 1

277

278

SZILVIA SÖVEGJÁRTÓ

sented in this passage were indeed the main concerns of scribes copying and interpreting Sumerian literary texts in the Old Babylonian period (ca. 2000–1600 BCE), based on the testimony of the glosses preserved in numerous manuscripts. The origins of the Sumerian glossographic tradition go back to the third millennium BCE: the first glosses are known from Sumerian lexical manuscripts of the Ebla archives (ca. 2500–2250 BCE). Though by that time Sumerian was still a spoken language in the Mesopotamian heartland, the archives of Ebla came into being in a different linguistic milieu. The city of Ebla, located on the Mesopotamian periphery, was inhabited by a population speaking a Semitic language, Eblaite, which is closely related to Akkadian. The Sumerian language enjoyed a special status, though, being the language the cuneiform writing system was based on. Quite similarly, during the Old Babylonian period, a functional diglossia was already perceptible between the Sumerian and Akkadian languages in the Mesopotamian heartland: while Sumerian was mostly restricted to academic, scholarly and cultic purposes, Akkadian was the native tongue of the scribes and the language of daily life. Akkadian slowly replaced Sumerian from the Old Babylonian period on, overtaking it in more and more domains, including literature and scholarship—though Sumerian preserved its prominent status in the cult up to the end of cuneiform culture (first century CE). Consequently, in the Old Babylonian period, educational background determined one’s access to written texts, and especially to texts written in Sumerian. Only those who took part in academic education were able to read and write—and to comment on— Sumerian manuscripts.

THE GLOSSATORS

The glosses of Sumerian literary manuscripts are the products of individual engagement with the texts; glossed manuscripts were English translation presented above is based on, but does not follow in every aspect, Sjöberg’s German rendering.

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION

279

primarily intended for personal use. Even if the body of the text and the gloss were not necessarily written by the same person, both should have been written within a narrow time frame, as clay tablets dry rapidly. In some lexical manuscripts, it is likely that the glosses were added by a teacher2—similar cases are not known from the literary corpus. The glossators of Sumerian literary manuscripts are just as many-faceted as the glossed manuscripts themselves. Glosses were included in several manuscripts of lexical lists3 that formed part of the Old Babylonian curriculum—for example the lists Izi or Ura. Similarly, glosses can be found sporadically throughout elementary school compositions. Apparently, pupils learned to use this scholarly and exegetical technique at an early stage of their education. On the other hand, the overall number of glossed lexical and literary manuscripts is rather low.4 Therefore, it is likely that this skill was taught only in a few schools or by a few instructors and thus practiced only by a couple of scribes and apprentice scribes throughout Mesopotamia.

Jay C. Crisostomo, “Bilingual Education and Innovations in Scholarshiop: The Old Babylonian Word List Izi,” (Ph.D. Diss., University of California, Berkeley, 2014), 95–96. 3 Lexical lists are lists of cuneiform signs and word lists developed to describe, transmit and research the writing system and the Sumerian language and its vocabulary. The history of these lists started with the beginning of writing in the 4th millennium BCE and ended in the first century CE. The Old Babylonian period has a rich lexical tradition; the estimated number of lexical tablets from this period is three to four thousand. Quite importantly, “Old Babylonian lexical texts form a corpus that is structured by the needs of a systematic introduction to writing and its history”; Niek Veldhuis, History of the Cuneiform Lexical Tradition (Münster: Ugarit-Verlag, 2014), 143. The Old Babylonian material, therefore, served primarily teaching purposes. 4 Ca. 2.5% of the presently known Old Babylonian Sumerian literary manuscripts contain glosses. 2

280

SZILVIA SÖVEGJÁRTÓ

In a few cases, it is apparent that several manuscripts with glosses have been produced by the same scribe.5 These manuscripts usually belong to small archives of five to twenty tablets with a related complex of themes, as well as similar format and script—though not all of them contain glosses. These internship archives6 point out that glosses were in use from the earliest to the potentially latest stage of scribal education. Furthermore, these manuscripts show that glossing was a personal choice or habit of the scribe: either glosses are completely missing or the proportion of glossed manuscripts is exceptionally high in such internship archives. The question of whether some of the glossed manuscripts were produced by professionals is rather problematic. Glosses seem to be present in highly specialized and advanced scholarly texts, like collections of incantations or long and complex pieces of literature. It is, however, dubious whether these texts were studied in the course of the apprenticeship and then stored in a professional’s library or whether they were copied in the course of the professional’s praxis and incorporated in his library at a later date.

The cohesion of these groups of manuscripts is partly known from colophons mentioning the scribe’s name; additionally, such groups can sometimes be reconstructed on the basis of the manuscripts’ physical features (format, ruling and indention techniques, graphic variants of signs) as well as noteworthy orthographic peculiarities. 6 Steve Tinney, “Tablets of Schools and Scholars: A Portrait of the Old Babylonian Corpus,” in The Oxford Handbook of Cuneiform Culture, ed. Karen Radner and Eleanor Robson (New York: Oxford University Press, 2011), 585. 5

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION

GLOSSES AND GLOSSING PRACTICES OF CUNEIFORM MANUSCRIPTS

The heavily glossed manuscript HS 1486 (P345620) from the Frau Professor Hilprecht Collection of Babylonian Antiquities, Jena. Courtesy of the curator, Prof. Dr. Manfred Krebernik

281

282

SZILVIA SÖVEGJÁRTÓ

Glosses in Old Babylonian cuneiform manuscripts are typically pressed in the wet clay simultaneously or nearly simultaneously to the writing of the main text. Scratches on tablets resembling glosses are exceptional and hard to identify with certainty as the signs usually cannot be deciphered. Except for the time factor, practices related to writing glosses on clay are similar to those related to writing glosses on parchment or paper. In the following, I would like to highlight some scribal practices known from cuneiform tablets of literary and scholarly content, along with the key characteristics of Old Babylonian textual hermeneutics. Glosses are usually written with a significantly smaller script than the main text. They can be placed on a cuneiform tablet based on the availability of free space: infralinear (within a line), interlinear (between two lines), or on any of the tablet edges. The main principle determining the position of a gloss is the proximity to its antecedent. Consequently, glosses on a manuscript might have various positions; sometimes there is even a functional distribution between the glosses according to their position. The number of glosses on a Sumerian manuscript is often low; a glossed manuscript usually contains up to five glosses. There are, however, a few examples of heavily glossed manuscripts containing up to eighty glosses filling the empty spaces on the tablet and mostly leading to a disorganized visual appearance of the text. Glosses were used for various purposes in Sumerian literary texts. Most of them intended to render a Sumerian word or expression in Akkadian. Phonetic glosses providing the vocalization of a sign with multiple values in a given context are also frequently attested. However, a small number of glosses had rather unique functions: they were used as figurative, iconic or graphic hints, or even as paraphrases or short commentaries to their antecedent. Beyond that, some glosses had no hermeneutical function but rather a redactional one: they contained text variants or corrections to the text or indicated an incomplete or broken Vorlage. Furthermore, omitted text lines could also be added to a manuscript with a smaller script, especially when they were inserted between two lines or positioned at the tablet’s edge; thus such additions might also appear formally as glosses. Glosses, as a rule, were not transmitted together with the main text during the Old Babylonian period. Consequently, no manuscripts are known so far from this period which share their glosses.

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION

283

However, a few manuscripts prove that glosses may have been misinterpreted as part of the main text and thus taken over while copying a manuscript. Such frozen glosses no longer stand out for their smaller script size and are included into the text line. Frozen glosses originated solely from phonetic glosses; thus these glosses were potentially mistaken as corrections or phonetic complements.

GLOSSES AS SOURCES ON ARTIFACT BIOGRAPHIES

In some cases, glosses also provide information on the praxeology of a given manuscript. However, only the consideration of a set of glosses preserved on the manuscript might be conclusive when the potential use of a written artifact is to be determined. In the following, I would like to present three manuscripts as case studies to demonstrate that merely considering the glosses preserved on a given tablet might lead to different conclusions than considering the whole of the text together with the glosses. The unprovenanced manuscript MS 3401 (P252342) is a small excerpt tablet containing, without doubt, a school exercise.7 The tablet has a landscape format, which is rather rare among Old Babylonian literary tablets, and contains nine lines from the composition Šulgi A [ETCSL 2.4.2.01], one of the elementary texts of the Old Babylonian school curriculum. The length of the excerpt, first of all, is unusual for an exercise tablet; normally, the daily pensum of a schoolboy consisted of 30–35 lines of a composition. The reverse of the tablet is blank. Thus it is questionable whether this tablet is a usual exercise tablet or whether this glossed manuscript indeed served a different purpose. On the tablet, four glosses and two further Akkadian annotations have been added to the main text. One of the annotations is located on the left edge, the other right after the excerpt on the lower edge and on the reverse of the tablet. The glosses on the tablet are heterogenous: an interpretative gloss in Sumerian, an AkkaLudek Vacín, “New Duplicates of Hymn Šulgi A in the Schøyen Collection,” in Wiener Zeitschrift für die Kunde des Morgenlandes 101 (2011), 461–462. 7

284

SZILVIA SÖVEGJÁRTÓ

dian translation, and two redactional glosses have been added to the text. The latter either are variants or were intended as corrections to the text. All things considered, it is unlikely that this excerpt tablet was once used to support the memorization of the text passage, in line with the praxeology attributed to most Old Babylonian school exercise tablets. This manuscript was rather intended for practicing the glossing of a text. Short glosses were positioned interlinearly, between text lines. Longer annotations have been placed on the tablet edge. The function of the four glosses is also diverse. The scribe was apparently experimenting with the possibilities by supplementing the text with glosses of different form and content. The text chosen by the scribe, the hymn Šulgi A, which was studied among the first literary compositions, was ideally suited for exercising this hermeneutical instrument on the basis of a well-known text passage. This suggestion, however, has a further implication for the function of exercise tablets: this tablet type was not only appropriate for memorizing a text passage, but it had some other uses as well. Though further glossed manuscripts with a similar use are likely, they are hard to identify. Especially short collections of proverbs were fitting for such an exercise. This small tablet, then, is the only evidence beyond glossed lexical lists that writing glosses was a skill which apprentice scribes had to master during their studies. The manuscript L 1489 (P343582), probably from Larsa, contains a liturgical composition rich in phonetic spellings. Moreover, the manuscript contains a number of phonetic glosses. Though both phonetic writings and phonetic glosses are quite frequent in liturgical manuscripts, it is rather exceptional that the two occur together in one tablet. The two methods are likely mutually exclusive because they fulfill the same function: they provide a simplified, phonetic spelling to a lexeme written logographically. The scribe, being aware of the standard orthography of a given lexeme, supplemented it by the correct pronunciation in the form of a

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION

285

gloss.8 It is not obvious, therefore, why the two methods were used side by side in the same manuscript. The conventional interpretation of this phenomenon, especially in case of a liturgical composition, would be that the glosses facilitated the correct pronunciation of the text in the course of its performance. In light of this interpretation, the glosses would point to the performative use of the manuscript, drafted by an expert in the course of his preparations.9 It remains, however, open why the two methods of providing phonetic hints to a text were mixed in this case. In case the correct pronunciation of the text was the central concern of the scribe, phonetic writings were more suitable to support the composition’s performance. Furthermore, phonetic glosses to the same lexeme in consecutive lines were both superfluous and unusual in glossed manuscripts. Another option is that the glosses of the manuscript were composed as corrections to a phonetically written Vorlage. This alleged prototype, as is common among liturgical manuscripts, included elements in both the standard orthography and the phonetic orthography mixed through the main text. The scribe thus followed a single strategy to render elements of the text phonetically wherever he needed such an aid. It is possible that the scribe who was copying the original manuscript attempted to reconstruct the standard orthography of the phonetically written elements in the text and noted the phonetic elements of the Vorlage in the form of glosses. Wherever the phonetic writings were preserved in the manuscript, he did not succeed with his task. This procedure was best suited for teaching purposes. It also confirms the hermeneutical interest of the glossator, who attempted to reconstruct an intelligible text on the basis of a phonetic Vorlage. Old Babylonian manuscripts with redactional glosses containing textual variants or corrections are quite similar in their structure to the present manuscript: the correction was included in the main Joachim Krecher, “Glossen,” in Reallexikon der Assyriologie und Vorderasiatischen Archäologie, vol. 3 (Berlin: De Gruyter, 1971), 431. 9 Paul Delnero, “Scholarship and Inquiry in Early Mesopotamia,” Journal of Ancient Near Eastern History 2 (2016), 115–116. 8

286

SZILVIA SÖVEGJÁRTÓ

text, and the corrected form was preserved in form of a gloss. It was, however, not obligatory to note the removed variant in the form of an annotation; thus corrections often remain unnoticed. As this manuscript indicates, it is not sufficient to consider the types of glosses in a given manuscript when determining its praxeology. Even plain phonetic glosses might serve other purposes than to note the pronunciation of a lexeme. Such features are only perceptible, however, when an ambiguity is present in a manuscript, as in the present case. A further unique manuscript that should be discussed here briefly is VAT 6705 (P342964), also of unknown provenance. The manuscript contains two compositions, the hymn Ninkasi A [ETCSL 4.23.1] and A drinking song [ETCSL 5.5.a]. The manuscript contains, moreover, several glosses; among them, the annotations on the left tablet edge10 are especially relevant for the praxeology of the tablet. Other glosses on the tablet, all placed interlinearly, contain textual variants. The three glosses on the left edge refer to the beginning of each line: two of the three glosses simply repeat the first signs of the line. The third gloss should, therefore, also refer to the beginning of the line containing the names of the rivers Euphrates and Tigris. The gloss contains the sign TAB, consisting of two parallel horizontal wedges, written twice. Although the sign TAB has the meaning “to double, to repeat,” it is likely that, in this case, rather the shape of the sign indicated the content meant by the glosses: namely, the reduplicated sign represented the two rivers iconically or graphically. The names of the rivers were replaced by this abbreviated form since their names consist of a set of signs which could not fit on the thin margin of the tablet. The annotations on the tablet edge likely had a mnemotechnical function: they were used as orientation aids within the manuscript. Thus this manuscript may have been used to memorize a text or to perform a text that was not memorized completely. A clearer conclusion on the manuscript’s use cannot be given, as such a mnemotechnical aid might also be used for different purposes. The cuneiform script goes from left to right; thus the left edge is the side of the tablet where the lines start. 10

ON THE SUMERIAN GLOSSOGRAPHIC TRADITION

287

CONCLUSION

In this brief overview of the Sumerian glossographic tradition, I argued that textual glosses can be found in manuscripts of any kind, in the products of students as well as in the reference libraries of scholars. The broad range of textual glosses that were used can be explained by artifact biographies. Some manuscripts testify to the fact that glossing was a hermeneutical method learned in the course of scribal education. In the later stages of one’s studies, even up to the phase where special knowledge related to a field of expertise has been acquired, this methodology was still in use. Glosses are thus to be interpreted as aids; these aids were not used due to the lack of knowledge of the glossator but to support the study of an unknown composition. The study of a new text was realized through copying, interpreting, reading or translating the text into Akkadian—and beyond that, also glossing it whenever necessary. As the case studies presented above clearly show, glosses do not provide sufficient information in most cases to determine the glossator or the setting of a manuscript. However, they provide important hints on the education and knowledge level of the glossator, leading to a better overall understanding of Old Babylonian education as well as of the artifact biographies of cuneiform manuscripts. Moreover, glosses also provide an insight into Old Babylonian textual hermeneutics and text criticism, as they are unique sources from the period when this scholarly method was developed.

BIBLIOGRAPHY

Crisostomo, Jay C. Bilingual Education and Innovations in Scholarship: The Old Babylonian Word List Izi. Ph.D. Diss., University of California, Berkeley, 2014.

Delnero, Paul. “Scholarship and Inquiry in Early Mesopotamia.” Journal of Ancient Near Eastern History 2 (2016).

Krecher, Joachim. “Glossen.” In Reallexikon der Assyriologie und Vorderasiatischen Archäologie Vol. 3. Berlin/New York: De Gruyter, 1971.

Sjöberg, Åke W. “Der Examenstext A.” Zeitschrift für Assyriologie und Vorderasiatische Archäologie 64 (1975).

288

SZILVIA SÖVEGJÁRTÓ

Tinney, Steve. “Tablets of Schools and Scholars: A Portrait of the Old Babylonian Corpus.” In Oxford Handbook of Cuneiform Culture, edited by Karen Radner and Eleanor Robson. New York: Oxford University Press, 2011.

Vacín, Ludek. “New Duplicates of Hymn Šulgi A in the Schøyen Collection.” Wiener Zeitschrift für die Kunde des Morgenlandes 101 (2011). Veldhuis, Niek. History of the Cuneiform Lexical Tradition. Guides to the Mesopotamian Textual Record, 6. Münster: Ugarit-Verlag, 2014.

CAN MANUSCRIPT HEADINGS PROVE THAT THERE WERE ARABIC GOSPELS BEFORE THE QURʾĀN? ROBERT TURNBULL Concrete evidence for pre-Islamic translations of the Bible into Arabic remains elusive. However, Anton Baumstark believed that the liturgical rubrics in some Arabic Gospel manuscripts proved that they were translated before the Arab conquests. He argued that these headings reflect a time in the development of the lectionary calendar at Jerusalem prior to the introduction of pre-Lent fasting in ‘Cheesefare week’ which Baumstark believed occurred immediately after the emperor Heraclius’ victory over Persia (622–628 CE). Baumstark’s study was based on three manuscripts in European libraries. At the present time, however, many more Arabic Gospel manuscripts with the old Jerusalemite lectionary calendar are known. A comprehensive study of all these manuscripts reveals a number of inter-related lectionary systems. When all the manuscripts are considered, it is clear that all the lectionary systems do indeed incorporate Cheesefare week before Lent. Therefore, the headings cannot be used to prove that the translations were made before Islam. Nevertheless, the lectionary headings are still highly significant historically as they provide a window into the beginnings of Christian use of Arabic in worship.

289

290

ROBERT TURNBULL

AN ARABIC BIBLE BEFORE ISLAM?

Were there Arabic translations of the Bible before Islam?1 Perhaps when Mavia, the fourth-century Arab queen, fought the Arian Roman emperor Valens to install her own Orthodox Arab bishop, parts of the liturgy were celebrated in Arabic.2 Or potentially the fifth-century Arab King Dāwūd facilitated translations of the Bible in the monastery he founded.3 Or possibly in the sixth century, at the monastery of Queen Hind near al-Ḥīra, the Bible was read in Arabic at the church where the Arabic inscription bearing her name once stood over its door.4 However, in the absence of concrete evidence in the surviving sources, many scholars have been hesitant to affirm that written pre-Islamic biblical translations existed.5 Nevertheless, there stands one giant figure in the 20th century who attempted to prove that there were multiple translations of sections of the Bible into Arabic before Islam: Anton Baumstark (1872– 1948).

I thank Alexander Treiger for his comments on an earlier version of this chapter. All remaining errors are my own. 2 Irfan Shahîd, Byzantium and the Arabs in the Fourth Century (Washington, D.C.: Dumbarton Oaks, 1984), 439–40. 3 Irfan Shahîd, Byzantium and the Arabs in the Fifth Century (Washington, D.C.: Dumbarton Oaks, 1989), 297–300, 426–27. 4 Arthur Vööbus, Early Versions of the New Testament: Manuscript Studies (Stockholm: impr. de L. Durbecq, 1954), 275–76. For the text of the inscription, see Gustav Rothstein, Die Dynastie der Lahmiden in al-Ḥīra: Ein Versuch zur arabisch-persichen Geschichte zur Zeit der Sasaniden (Berlin: Verlag von Reuther & Reichard, 1899), 23–24 n. 2. 5 Sidney H. Griffith, “When Did the Bible Become an Arabic Scripture?,” Intellectual History of the Islamicate World 1, no. 1–2 (2013), 10. Also: Samuel Noble and Alexander Treiger, The Orthodox Church in the Arab World, 700–1700: An Anthology of Sources (DeKalb: Northern Illinois University Press, 2014), 11. 1

ARABIC GOSPELS BEFORE THE QURʾĀN

291

A MATHEMATICAL PROOF

In the 1930s, Baumstark published a series of articles outlining arguments for the existence of pre-Islamic Arabic Bible translations.6 The most comprehensive discussion came in an address that Baumstark delivered at the 6th Deutscher Orientalistentag in Vienna in 1930.7 In this lecture, Baumstark gave various arguments for evidence which he believed pointed to pre-Islamic translations of liturgical books (particularly the Gospels and the Psalms) into Arabic. Most of his arguments are based on hints in Islamic sources or extrapolations from extant Arabic manuscripts and can be disputed.8 However, his final argument is one which Baumstark ventured to say “mathematically proved” his thesis.9 The thrust of this argument is that the lectionary headings in Arabic manuscripts of the Gospels were out of date after the Islamic conquests and so must have been translated beforehand. Let’s explore these manuscripts in more detail.

BAUMSTARK’S MANUSCRIPTS

Before Baumstark’s address in Vienna, he wrote an article in honor of August Heisenberg, professor of Byzantine studies (and father Several of these are now translated into English and collected in: Ibn Warraq, ed., Christmas in the Koran: Luxenberg, Syriac, and the Near Eastern and Judeo-Christian Background of Islam (New York: Prometheus Books, 2014), 265–352. 7 Later published in 1932 under the title: “Das Problem eines vorislamischen christlich-kirchlichen Schrifttums in arabischer Sprache.’ Translated into Spanish by Magdalena López Pérez as ‘El problema de la producción escrita cristiana-eclesiástica en árabe de época preislámica.” Translated into English by Elizabeth Puin, “The Problem of Pre-lslamic Ecclesiastical Christian Literature in Arabic,” in Ibn Warraq, Christmas in the Koran, 285–95. 8 For example, Georg Graf, Geschichte der christlichen arabischen Literatur: Bd. 1. Die Übersetzungen (Rome: Biblioteca Apostelica Vaticana, 1944), 39–40. Also: Gregor Schoeler, Écrire et transmettre dans les débuts de l’Islam (Paris: Presses universitaires de France, 2002), 26–29. 9 Baumstark, “Das Problem eines vor-islamischen christlichkirchlichen Schrifttums in arabischer Sprache,” 570. 6

292

ROBERT TURNBULL

of the pioneer of quantum mechanics Werner Heisenberg).10 In this article, Baumstark referred to three Arabic Gospel manuscripts that had found their way to Europe.

Fig. 1. Matt 14:31–35a from folio 2v of Leipzig Vollers 1059A.11

One was just a two-leaf parchment fragment, Leipzig Vollers 1059a (fig. 1), dated to the 9th century.12 It would have been the outer bifolio of a quire containing two sections of Matthew’s Gospel. Constantine Tischendorf had taken it from St. Catherine’s Monastery on his second visit to Sinai in 1853. This fragment remained at the Leipzig University Library for many years but was reported as missPublished under the title “Die sonntägliche Evangelienlesung im vorbyzantinischen Jerusalem.” 11 Image taken from the facsimile for ms. II in H.L. Fleischer, “Beschreibung der von Prof. Dr. Tischendorf im J. 1853 aus dem Morgenlande zurückgebrachten christlicharabischen Handschriften,” Zeitschrift der Deutschen Morgenländschen Gesellschaft 8.3 (1854), 584–87. 12 Fleischer, ‘Beschreibung,’ 585–86. Originally labelled Cod. Tisch. XXXI (A) in Constantin von Tischendorf, Anecdota sacra et profana (Leipzig: H. Fries, 1855), 70. Later given the catalogue number 1059a in Karl Vollers, Katalog der islamischen, christlich-orientalischen, jüdischen und samaritanischen Handschriften der Universitäts-Bibliothek zu Leipzig (Leipzig: Otto Harrassowitz, 1906), 373–74. 10

ARABIC GOSPELS BEFORE THE QURʾĀN

293

ing in 1938.13 Lamentably, if Tischendorf had left it in the hands of the monks of St. Catherine’s, then in all likelihood it would still be available today.14 The fragment has rubricated headings that specify when the following passage is to be read.

Fig. 2. Mt 9:37b–10:2a from folio 16v of Vatican Borg. ar. 95.15

Baumstark’s second manuscript, Vatican Borg. ar. 95 (fig. 2), was also dated to the 9th century. It has the same main text as the Tischendorf fragment, but it does not contain the lectionary headings, except for occasional liturgical indications, added by a later hand.16 In the place of the rubrics are empty spaces, indicating that the original scribe was aware of the location of the lectionary headings.

Bernhard Levin, Die griechisch-arabische Evangelien-Übersetzung Vat. Borg. ar. 95 und Ber. orient. oct. 1108 (Uppsala: Almqvist & Wiksells Boktryckeri-a.-b., 1938), 12. 14 Fortunately, in 1904 Georg Graf made a careful transcription of the fragment which is still available in his Nachlass: Graf, GCAL 1, 143 n. 3. 15 Image taken from Eugene Tisserant, Specimina Codicum Orientalium (Bonn: A. Marcus et E. Weber, 1914), 55. 16 Levin, Evangelien-Übersetzung, 18. Also: Graf, GCAL 1, 144. 13

294

ROBERT TURNBULL

Fig. 3. Luke 7:34–37a from folio 110v of Berlin MS. or. oct. 1108.17

A third manuscript, Berlin Ms. or. oct. 1108 (fig. 3), was brought back from the East by Paul Kahle, Baumstark’s colleague at Bonn University, and deposited at the Staatsbibliothek in Berlin. Though it was much younger (dated to 1047 CE), it contained the lectionary rubrics like the Tischendorf fragment.18 The first part of the manuscript (Matt 1:1–6:18a) is missing. It was the liturgical headings in this manuscript that were the focus of Baumstark’s argument. To understand their significance, we must take a detour via the city so significant for the development of Christian liturgy, Jerusalem.

JERUSALEM AND ITS LECTIONARY CYCLE

In Jerusalem, Christians can transcend the gap separating them from the Gospel events by reading Scripture in the same location where the events unfolded and at the same time of year in which they occurred. This feature is reflected in the development of the Image ©2019 STAATSBIBLIOTHEK ZU BERLIN - Preussischer Kulturbesitz, Orientabteilung. Used by permission. 18 The colophon on 206r gives the date 438 AH, and so Baumstark (and others after him) state the date of the manuscript as 1046–1047 CE: Baumstark, “Die sonntägliche Evangelienlesung im vorbyzantinischen Jerusalem,” Byzantinische Zeitschrift 30.1 (1930), 352. However, the colophon also mentions that it was finished in the middle of Lent, and therefore the date is unambiguously 1047 CE. 17

ARABIC GOSPELS BEFORE THE QURʾĀN

295

unique lectionary cycle of the Jerusalem church. From 381 to 384 CE, a pilgrim from the West named Egeria came to visit the Holy Land; and she wrote for her sisters back home a marvelous account of her travels.19 Many of the details that she gave about the unique liturgical practice in Jerusalem were corroborated by the discovery of the old Armenian lectionary deriving from the 5th century, also based on the Jerusalemite tradition.20 By the 8th century, the Jerusalem lectionary traditions had been translated into Georgian, representing a later stage of evolution beyond the Armenian lectionary.21 One significant development is that the Georgian lectionary includes Gospel readings for regular Sundays through the year. Baumstark noted that the lectionary headings in the Arabic manuscripts of the Gospels mentioned above included similar Sunday Gospel readings for the entire year. However, he believed that these Arabic manuscripts preserved a stage in the development of the Jerusalemite liturgical calendar prior to that recorded by the Georgian lectionary and could even be firmly dated to before Byzantium’s victorious campaign over Persia under Heraclius (622– 628 CE).22 This dating is based on Alfred Rahlfs’s understanding of the introduction of Cheesefare week.23 Cheesefare week is a preparatory week of fasting immediately before Lent during which Christians abstain from meat, yet animal products (such as cheese) are still permitted (unlike in Lent proper). While it is not certain that Rahlfs was correct in his view of the origin of Cheesefare week, nevertheless, the practice of abstaining from meat in the

English edition: John Wilkinson, Egeria’s Travels: Newly Translated with Supporting Documents and Notes, (London: S.P.C.K., 1971), 89–147. 20 Daniel Galadza, Liturgy and Byzantinization in Jerusalem, (Oxford: Oxford University Press, 2018), 46–49. 21 Galadza, Liturgy and Byzantinization, 49–52. 22 Baumstark, “Die sonntägliche Evangelienlesung,” 358; Baumstark, “Das Problem,” 571. 23 Alfred Rahlfs, Die alttestamentlichen Lektionen der griechischen Kirche, (Berlin: Weidmann, 1915), 201–5. See also: Anton Baumstark, Comparative Liturgy, trans. F.L. Cross (London: A.R. Mowbray & Co., 1958), 197. 19

296

ROBERT TURNBULL

week before Lent was probably present in Palestine at least by the time of Heraclius’s Persian campaign, if not earlier.24

A MISSING SUNDAY

To show that the Arabic lectionary headings knew nothing of Cheesefare week, Baumstark tabulated the passages for Luke which in the old Jerusalemite tradition were read after Epiphany and continued through Lent. Baumstark noted that the headings in the Arabic Gospel manuscript correspond to eight Sundays after Epiphany (January 6) and six Sundays in Lent. This adds to 14 Sundays, Rahlfs rejected any testimony which suggested the existence of pre-Lent fasting before Heraclius. For example, Rahlfs regarded the mention of eight weeks of fasting by Dorotheus of Gaza (6th century) as a later interpolation: Rahlfs, Die alttestamentlichen Lektionen, 196–97. However, Rahlfs’s conjecture has been regarded as unnecessary and is unsupported by the manuscript tradition: Vittorio Peri, “La durata e la struttura della Quaresima nell’antico uso ecclesiastico gerosolimitano,” Aevum 37 (1963), 45; Vassa Conticello, “La Quarantaine hiérosolymitaine dans le De sacris ieiuniis de Jean Damascène,” in Θυσία αἰνέσεως: mélanges liturgiques offerts à la mémoire de l’archevêque Georges Wagner (1930–1993, ed. Job Getcha and André Lossky (Paris: Presses Saint-Serge, Institut de théologie orthodoxe, 2005), 86–87. Furthermore, while he accepted Eutychius of Alexandria’s association of the pre-Lent fast with Heraclius, Rahlfs rejected the presupposition in Eutychius’s work that pre-Lent fasting had already been practiced prior to Heraclius’s entry into Jerusalem since Eutychius seems to inaccurately describe the background of Coptic practice. However, an earlier recension of Eutychius’s Annals has been discovered (perhaps the author’s own autograph itself) without the anti-Coptic polemic: M. Breydy, “Mamila ou Maqella? La prise de Jérusalem et ses conséquences (614 AD) selon la récension alexandrine des Annales d’Eutychès,” Oriens Christianus 65 (1981), 62–86. In this earlier recension, Eutychius’s description closely parallels the practice of pre-Lent fasting in Jerusalem described by John of Damascus: Alexander Treiger, “Макарий Синаит, «О посте на сырной седмице»,” Vestnik Pravoslavnogo Sviato-Tikhonovskogo Gumanitarnogo Universiteta III 53 (2017), 107–8. One can therefore justifiably regard both Dorotheus and Eutychius as witnesses to the practice of Cheesefare week in Palestine prior to the 7th century. 24

ARABIC GOSPELS BEFORE THE QURʾĀN

297

which is the maximum possible number of Sundays between Epiphany and Palm Sunday (since the latest possible date for Easter is April 25). The Georgian lectionary used by Baumstark did not have readings for weeks 6 to 8 after Epiphany; instead, Baumstark noted that Matt 6:34 (and following) is read in the Georgian lectionary at the beginning of Cheesefare week. This reading would have taken up one of the weeks between Epiphany and the beginning of Lent. Because Cheesefare week does not correspond to any of the entries in the tables, and since it is seemingly complete in that it gives the maximum number of weeks allowable, Baumstark concluded that the Arabic Gospel headings reflect a time before Heraclius initiated the practice, and thus before the Islamic conquest.

MORE MANUSCRIPTS

In 1972, Gérard Garitte made his own study of the rubrics in Jerusalemite Arabic Gospel manuscripts.25 Garitte described Baumstark’s theory as untenable since it was based only on the particularities of the three manuscripts to which Baumstark had access.26 In his study, Garitte edited the text of the rubrics in Mark’s Gospel from five other Arabic Gospel manuscripts at St. Catherine’s Monastery at Sinai. He concluded by recommending that a further study of these rubrics be published based on all known witnesses.27 Since Gérard Garitte, “Les rubriques liturgiques de quelques anciens tétraévangiles arabes,” in Scripta Disiecta, vol. 2 (Louvain: Institut Orientaliste de Louvain, 1980), 722–37. 26 Garitte, “Les rubriques liturgiques,” 723–24. Garitte did not elaborate further on these particularities since his focus was on the manuscripts in the Sinai collection. For other responses to Baumstark’s argument see: Graf, GCAL 1, 143–46; Griffith, “The Gospel in Arabic: An Inquiry into its Appearance in the First Abbasid Century,” Oriens Christianus 69 (1985), 154–57; Griffith, The Bible in Arabic: The Scriptures of the ‘People of the Book’ in the Language of Islam (Princeton: Princeton University Press, 2013), 47; Joshua Blau, “Sind uns Reste arabischer Bibelübersetzungen aus vorislamischer Zeit erhalten geblieben?,” Le Muséon 86 (1973), 67–72. 27 Garitte, “Les rubriques liturgiques,” 736. 25

298

ROBERT TURNBULL

Garitte’s study, a number of other Arabic Gospel manuscripts have been discovered from the Sinai New Finds that also contain the old Jerusalemite lectionary headings, including what is probably much of the rest of the codex of Tischendorf’s fragment.28

LECTIONARY SYSTEMS

In light of this these new discoveries, I thought it appropriate for my contribution to the “Dots, Marginalia and Peritexts” workshop to edit all the lectionary headings from all known Arabic Gospel manuscripts with the old Jerusalemite lectionary. This amounted to 15 manuscripts. These belong to four families, which Hikmat Kashouh labeled as families A, B, C, and D and which were translated into Arabic from Greek (families A, B, and C) and from Syriac (family D).29 They were digitally transcribed into a database and collated using a software program called D-Codex, which I currently have under development.30 As I collated the headings, I identified nine distinct lectionary systems. I have designated these lectionary systems by using Greek letters that correspond to the Latin letters given to the textual families by Kashouh, with subtypes being distinguished by numerals. For example, lectionary system “α1” is the first subtype associated with textual family “A,” and so forth. To my surprise, the transmission of the lectionary headings in the manuscripts did not always align with the textual family to which the manuscript belongs. For example, Sinai ar. 72 is part of family A, yet the liturgical rubrics correspond mainly to those of family D (i.e. Sinai ar. 70) and yet still retain some characteristics of family A. Consequently, I categorized it as Lectionary System αδ. Dmitry Morozov, “К датировке древнейшей арабской рукописи Евангелия,” in Каптеревские чтения: сборник статей (Moscow: Институт всеобщей истории РАН, 2008), 20. For the lectionary headings of other Arabic Gospel manuscripts from the New Finds, see Robert Turnbull, “Codex Sinaiticus Arabicus and its Family,” forthcoming. 29 Hikmat Kashouh, The Arabic Versions of the Gospels: The Manuscripts and Their Families (Berlin: De Gruyter, 2011), 86–125. 30 Robert Turnbull, “D-Codex: Software for Digital Manuscript Transcription, Analysis, and Publication,” forthcoming. 28

ARABIC GOSPELS BEFORE THE QURʾĀN

299

Siglum

Catalogue Number

Date (CE)

Textual Family

Lectionary System

CSA

Sinai ar. NF Parch. 8, 27, 28

9th century

B

β

873

A

α3

Sinai ar. NF Parch. 5, 6, 63

9th century

C+P

γ1

901–902

A

α2

B

L+ Ν6+ N7

N15+ S54 S70

S71+ S72 S74 S75 S97 S98 V

Berlin Ms. or. oct. 1108

Leipzig, Vollers 1059A; Sinai ar. NF Parch. 14, 16; БАН Q537

Sinai ar. NF Parch. 7

1047

Sinai ar. NF Parch. 15, 24, 36, 64

10th

Sinai ar. 70

9th

Sinai ar. 54

century

10th century century

Leipzig, Vollers 1059Β; Sinai ar. 71; Sinai ar. NF Parch. 44

10th

Sinai ar. 74

9th

Sinai ar. 97

1125–1126

Sinai ar. 72 Sinai ar. 75 Sinai ar. 98

Vatican Borg. ar. 95

9th

century

A

B+C A

D

α2

β + γ2 + αγ α1 δ

B+A

β+α

897

A

αδ

century

C

γ2

century

13th century 9th century

A A

A

A

Table 1. The Arabic Gospel manuscripts with old Jerusalemite lectionary headings

α1

α1

α2

α2

These lectionary systems appear to evolve from each other, as is diagrammed in fig. 4. On three occasions, the differences are explained by the addition of Saturdays to an existing lectionary sys-

300

ROBERT TURNBULL

tem. I will elaborate on these relationships in the forthcoming edition and analysis of the rubrics, but for now it is clear that the situation in Palestinian monasteries and churches was rich and varied, as is characteristic of the old Jerusalemite tradition.31

Fig. 4. The relationships between the nine lectionary systems of the old Jerusalemite Arabic Gospels

WAS BAUMSTARK CORRECT?

Now that the headings of all the appropriate manuscripts are collated, what can we say about Baumstark’s argument for a preIslamic Arabic Gospel? When all the manuscripts are considered, it appears that all of the lectionary systems include readings for the Sunday before Cheesefare week at Luke 7:36, as can be seen in Table 2 with the headings for this week.32

Galadza, Liturgy and Byzantinization, 348. The β is not attested in the manuscripts at Luke 7:36, but it here would have most probably aligned with system δ given its likely dependence on that system and since it agrees with δ for the first Sunday of Lent (Table 3). 31 32

ARABIC GOSPELS BEFORE THE QURʾĀN System α1–3

Arabic Rubric

‫تقرا �ي �د ا��روم ا��ول‬

txt B N7 S54 S74 S97 S98

Empty Space: S71+ V ¦ Lacuna: L+

δ

‫ىقرا �ي ا���د ا��روم ا��ول‬

txt S70

αδ

‫يقرا ���روم ا��ول‬

txt S72

γ1

‫قرايه ايضا ��د ��ازه ا���م‬

txt N6+

γ2

‫���د ��ازه ا���م �ي القداس‬

txt S75

αγ

��‫يقرا �ي �د ال���ازه ا��و‬

txt N15+

301

English Translation

Read on the first Sunday of the prohibitions. Read on the first Sunday [of] the prohibitions.

Read for the first prohibitions.

Reading also for the Sunday of the prohibition33 of meat. For the Sunday of the prohibition of meat ([read] in the Liturgy).

Read on the Sunday of the first prohibition.

Table 2. The rubrics at Luke 7:36 for the Sunday before Cheesefare Week (i.e. Meatfare Sunday).

Baumstark misunderstood Luke 7:36 in the Berlin manuscript as being read on the first Sunday of Lent instead of the Sunday before Cheesefare Week.34 He misinterpreted the word translated as “prohibitions” (al-ḥurūm) as “anathemas,” which he understood as referring to the Feast of the Triumph of Orthodoxy initiated in Constantinople in 843 CE, during which anathemas are declared against iconoclasm.35 This would be the incorrect week in Lent for the Feast of Orthodoxy, and The usage here of the Arabic ‫ ��ازه‬derives from the Syriac root ‫ܟܪܙ‬ meaning “to proscribe.” 34 The old Jerusalemite convention enumerated the weeks in Lent from the Sunday before the first week of proper fasting, whereas the Byzantine convention names this Sunday as “Cheesefare Sunday” and counts the following Sunday as the first in Lent: Galadza, Liturgy and Byzantinization, 218. 35 Baumstark, “Die sonntägliche Evangelienlesung,” 358–59. 33

302

ROBERT TURNBULL

it is strange to enumerate such a feast the “first” because it is a feast of only one day; but Baumstark saw this as a clumsy redaction from an original heading, which he conjectured would have read “the first Sunday of Lent,” with the word for “anathemas” later substituted for “Lent.” This interpretation was only possible because Baumstark’s manuscripts had no other reading for the first Sunday of Lent. However, when the other manuscripts are considered (Table 3), it is clear that Matt 6:16 was to be read on the first Sunday of Lent. System α1

Arabic Rubric

‫ىقرا �ى �د الصوم ا��ول‬

txt S97 ¦ S74 ‫ �ي‬omitted, ‫ الصوم‬illegible

α2–3

Lacuna: S54

‫تقرا �ي ا���د ا��ول من الصوم‬

txt N7 S98

Empty Space: V ¦ Lacuna: B L+

β

‫يقرا �ي �د الصوم ا��ول ا��روم‬

txt N15+

No heading: CSA ¦ Lacuna: S71+

δ

‫قراه يوم ا���د ا��روم ا��ول من الصوم‬

txt S70

αδ

‫ىقرا �ى ا���د ا��ول من الصوم‬

txt S72

γ2

‫قرايه ��ول ا�د من الصوم‬

txt S75

English Translation

Read on the first Sunday of Lent. Read on the first Sunday in Lent. Read on the first Sunday of Lent, the prohibitions.

Reading of the first Sunday [of] the prohibitions in Lent. Read on the first Sunday in Lent.

Read for [the] first Sunday in Lent.

Table 3. The rubrics at Matt 6:16 for the Sunday before the first week of Lent.36 36

The γ1 and αγ systems are not attested at Matt 6:16.

ARABIC GOSPELS BEFORE THE QURʾĀN

303

This heading at Matt 6:16 for the first Sunday in Lent was not known to Baumstark because the Berlin manuscript, which was his main source, is lacunose until Matt 6:18b, just two verses and three words later. Therefore, Luke 7:36 in all these lectionary systems should be understood as being read on the Sunday before Cheesefare week, as is stated explicitly in the γ1 and γ2 systems, as well as in a number of other old Jerusalemite sources.37 Though Baumstark was correct that the eight weeks of Sunday readings after Epiphany ought not to have left room for an extra Sunday in Lent, it is not unusual for extra lections to be specified, even when they will not be read. For example, Sinai geo. 38 (979 CE) also outlines eight Sunday readings after Epiphany as well as the Sunday before Cheesefare week.38 The superfluous eighth Sunday after Epiphany may be a remnant of a time in the Jerusalemite lectionary before pre-Lent fasting existed. Subsequently, when the Sunday before Cheesefare week was added, the 8th Sunday after Epiphany was not removed.

CONCLUSION

Baumstark’s argument for a pre-Islamic Arabic Gospel translation was bold. It brought together many threads in political and liturgical history. Ultimately, though, it was incorrect because it relied too heavily on a single defective manuscript. Even though he was mistaken, Baumstark’s argument serves as a stimulus to investigate the liturgical peritexts of these and other manuscripts so we can understand their transmission and use far more deeply. Though the lectionary headings cannot prove that the Gospels were translated into Arabic before Islam, they nevertheless provide an important window into the early history of Arabic in Christian liturgy. The lectionary systems of these Arabic Gospel manuscripts must be compared with other related traditions to reveal the evolution of For other Jerusalemite sources see: Gabriel Bertonière, The Sundays of Lent in the Tri[o]dion: The Sundays Without a Commemoration (Rome: Pontifical Oriental Institute, 1997), 46–47; Galadza, Liturgy and Byzantinization, 323–24. 38 Galadza, Liturgy and Byzantinization, 323–24. 37

304

ROBERT TURNBULL

the liturgy in Jerusalem in the years after the Arab conquest. Perhaps vestiges of pre-Islamic Arabic biblical translations, if indeed they existed, do lie somehow in the manuscripts that have come down to us. If so, then it would take an emulation of Baumstark’s ingenuity, attention to detail, and ability to synthesize various traditions to be able to conclusively demonstrate the case.

BIBLIOGRAPHY

Baumstark, Anton. Comparative Liturgy. Translated by F. L. Cross. London: A. R. Mowbray & Co, 1958. ———. “Das Problem eines vor-islamischen christlich-kirchlichen Schrifttums in arabischer Sprache.” Islamica 4 (1931): 562–75.

———. “Die sonntägliche Evangelienlesung im vorbyzantinischen Jerusalem.” Byzantinische Zeitschrift 30.1 (1930): 350–359.

———. “El problema de la producción escrita cristiana-eclesiástica en árabe de época preislámica.” Translated by Magdalena López Pérez. Collectanea Christiana Orientalia (2006): 351–62.

Bertonière, Gabriel. The Sundays of Lent in the Tri[o]dion: The Sundays Without a Commemoration. Rome: Pontifical Oriental Institute, 1997. Blau, Joshua. “Sind uns Reste arabischer Bibelübersetzungen aus vorislamischer Zeit erhalten geblieben?” Le Muséon: Revue d’Études Orientales 86 (1973): 67–72.

Breydy, M. “Mamila ou Maqella? La prise de Jérusalem et ses conséquences (614 AD) selon la récension alexandrine des Annales d’Eutychès.” Oriens Christianus 65 (1981): 62–86.

Conticello, Vassa. “La Quarantaine hiérosolymitaine dans le De sacris ieiuniis de Jean Damascène.” In Θυσία αἰνέσεως: mélanges liturgiques offerts à la mémoire de l’archevêque Georges Wagner (1930– 1993). Edited by Job Getcha and André Lossky. Paris: Presses Saint-Serge, Institut de théologie orthodoxe, 2005.

Fleischer, H. L. “Beschreibung der von Prof. Dr. Tischendorf im J. 1853 aus dem Morgenlande zurückgebrachten christlicharabischen Handschriften.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 8.3 (1854): 584–87.

ARABIC GOSPELS BEFORE THE QURʾĀN

305

Galadza, Daniel. Liturgy and Byzantinization in Jerusalem. Oxford: Oxford University Press, 2018. Garitte, Gérard. “Les rubriques liturgiques de quelques anciens tétraévangiles arabes.” In Scripta Disiecta. Vol. 2 of. Louvain: Institut Orientaliste de Louvain, 1980. Graf, Georg. Geschichte der christlichen arabischen Literatur: Bd. 1. Die Übersetzungen. Rome: Biblioteca Apostolica Vaticana, 1944.

Griffith, Sidney H. The Bible in Arabic: The Scriptures of the ‘People of the Book’ in the Language of Islam. Princeton: Princeton University Press, 2013. ———. “The Gospel in Arabic: An Inquiry into its Appearance in the First Abbasid Century.” Oriens Christianus 69 (1985): 126– 67. ———. “When Did the Bible Become an Arabic Scripture?” Intellectual History of the Islamicate World 1.1–2 (2013): 7–23.

Ibn Warraq, ed. Christmas in the Koran: Luxenberg, Syriac, and the Near Eastern and Judeo-Christian Background of Islam. New York: Prometheus Books, 2014. Kashouh, Hikmat. The Arabic Versions of the Gospels: The Manuscripts and Their Families. Berlin: De Gruyter, 2011.

Levin, Bernhard. Die griechisch-arabische Evangelien-Übersetzung Vat. Borg. ar. 95 und Ber. orient. oct. 1108. Uppsala: Almqvist & Wiksells Boktryckeri-a.-b., 1938.

Morozov, Dmitry. “К датировке древнейшей арабской рукописи Евангелия.” Pages 19–23 in Каптеревские чтения: сборник статей. Moscow: Институт всеобщей истории РАН, 2008.

Noble, Samuel, and Alexander Treiger, eds. The Orthodox Church in the Arab world, 700–1700: An Anthology of Sources. DeKalb: Northern Illinois University Press, 2014. Peri, Vittorio. “La durata e la struttura della Quaresima nell’antico uso ecclesiastico gerosolimitano.” Aevum 37 (1963): 31–62.

Rahlfs, Alfred. Die alttestamentlichen Lektionen der griechischen Kirche. Berlin: Weidmann, 1915.

306

ROBERT TURNBULL

Rothstein, Gustav. Die Dynastie der Lahmiden in al-Ḥīra: Ein Versuch zur arabisch-persichen Geschichte zur Zeit der Sasaniden. Berlin: Verlag von Reuther & Reichard, 1899. Schoeler, Gregor. Écrire et transmettre dans les débuts de l’Islam. Paris: Presses universitaires de France, 2002. Shahîd, Irfan. Byzantium and the Arabs in the Fifth Century. Washington DC: Dumbarton Oaks, 1989.

———. Byzantium and the Arabs in the Fourth Century. Washington DC: Dumbarton Oaks, 1984.

Tischendorf, Constantin von. Anecdota sacra et profana. Leipzig: H. Fries, 1855. Tisserant, Eugene. Specimina Codicum Orientalium. Bonn: A. Marcus et E. Weber, 1914.

Treiger, Alexander. “Макарий Синаит, «О посте на сырной седмице».” Vestnik Pravoslavnogo Sviato-Tikhonovskogo Gumanitarnogo Universiteta III 53 (2017): 103–34.

Vollers, Karl. Katalog der islamischen, christlich-orientalischen, jüdischen und samaritanischen Handschriften der Universitäts-Bibliothek zu Leipzig. Leipzig: Otto Harrassowitz, 1906. Vööbus, Arthur. Early Versions of the New Testament: Manuscript Studies. Stockholm: impr. de L. Durbecq, 1954.

Wilkinson, John. Egeria’s Travels: Newly translated with supporting documents and notes. London: S.P.C.K, 1971.