Studies on the Paratextual Features of Early New Testament Manuscripts: Texts and Editions of the New Testament 9004537945, 9789004537941

Most studies of ancient New Testament manuscripts focus on individual readings and textual variants. This book, however,

477 135 8MB

English Pages 379 [396] Year 2023

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
16 Studies On The Paratextual Features Of Early New Testament Manuscripts Texts and Editions of the New Testament
16 Studies On The Paratextual Features Of Early New Testament Manuscripts Texts and Editions of the New Testament
Contents
Preface
Figures and Tables
Abbreviations
Contributors
Introduction: Paratextual Features of Early Greek Manuscripts
Chapter 1 What Is Paratext? In Search of an Elusive Category
Chapter 2 Missing the Point: Modern Punctuation Practice as Authoritative but Possibly Problematic Decision-Making
Chapter 3 Pointers to Persons and Pericopes? A Study of the Intermarginal Signs in Sahidic Manuscripts of the Gospel of John
Chapter 4 But for Me, the Scriptures Are Jesus Christ (ΙC􁐊 ΧC􁐊 ; Ign. Phld. 8:2): Creedal Text-Coding and the Early Scribal System of Nomina Sacra
Chapter 5 Segmentation and Interpretation of Early Pauline Manuscripts
Chapter 6 Can Papyri Correspondence Help Us to Understand Paul’s “Large Letters” in Galatians?
Chapter 7 The Tradition and Development of the Subscriptions to 1 Timothy
Chapter 8 Second Timothy: When and Where? Text and Traditions in the Subscriptions
Chapter 9 Composite Citations in New Testament Greek Manuscripts
Chapter 10 Titus in P32 and Early Majuscules: Textual Reliability and Scribal Design
Chapter 11 The Scribal Use of Ekthesis as a Paragraph Marker? The Galatians Text in Codex Sinaiticus as a Test Case
Chapter 12 Miniature Codices in Early Christianity
Chapter 13 Marginalia in New Testament Greek Papyri: Implications for Scribal Practice and Textual Transmission
Conclusion: Paratextual Features: Summary and Prospects
Index of Ancient Sources
Index of Modern Authors
Recommend Papers

Studies on the Paratextual Features of Early New Testament Manuscripts: Texts and Editions of the New Testament
 9004537945, 9789004537941

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Studies on the Paratextual Features of Early New Testament Manuscripts

Texts and Editions for New Testament Study Edited by Stanley E. Porter and Wendy J. Porter

volume 16

The titles published in this series are listed at brill.com/tent

Studies on the Paratextual Features of Early New Testament Manuscripts Texts and Editions of the New Testament Edited by

Stanley E. Porter Chris S. Stevens David I. Yoon

LEIDEN | BOSTON

Library of Congress Cataloging-in-Publication Data Names: Porter, Stanley E., 1956– editor. | Yoon, David I., editor. | Stevens, Chris S., editor. Title: Studies on the paratextual features of early New Testament manuscripts : texts and editions of the New Testament / edited by Stanley E. Porter, Chris S. Stevens, David I. Yoon. Description: Leiden ; Boston : Brill, [2023] | Series: Texts and editions for New Testament study, 1574–7085 ; volume 16 | Identifiers: LCCN 2023002053 (print) | LCCN 2023002054 (ebook) | ISBN 9789004537941 (hardback ; acid-free paper) | ISBN 9789004537972 (ebook) Subjects: LCSH: Bible. New Testament—Criticism, Textual. | Bible. New Testament—Versions—History. | Bible. New Testament—Manuscripts—History. Classification: LCC BS2325 .S7795 2023 (print) | LCC BS2325 (ebook) | DDC 225.4/8—dc23/eng/20230317 LC record available at https://lccn.loc.gov/2023002053 LC ebook record available at https://lccn.loc.gov/2023002054

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface. isSn 1574-7085 isbn 978-90-04-53794-1 (hardback) isbn 978-90-04-53797-2 (e-book) Copyright 2023 by Stanley E. Porter, Chris S. Stevens, and David I. Yoon. Published by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Nijhoff, Brill Hotei, Brill Schöningh, Brill Fink, Brill mentis, Vandenhoeck & Ruprecht, Böhlau, V&R unipress and Wageningen Academic. Koninklijke Brill NV reserves the right to protect this publication against unauthorized use. Requests for re-use and/or translations must be addressed to Koninklijke Brill NV via brill.com or copyright.com. This book is printed on acid-free paper and produced in a sustainable manner.

Contents

Preface vii List of Figures and Tables ix Abbreviations xiii Contributors xvi



Introduction: Paratextual Features of Early Greek Manuscripts 1 Stanley E. Porter, Chris S. Stevens, and David I. Yoon

1

What Is Paratext? In Search of an Elusive Category 13 Stanley E. Porter

2

Missing the Point: Modern Punctuation Practice as Authoritative but Possibly Problematic Decision-Making 33 Hans Förster

3

Pointers to Persons and Pericopes? A Study of the Intermarginal Signs in Sahidic Manuscripts of the Gospel of John 59 Matthias H. O. Schulz

4

But for Me, the Scriptures Are Jesus Christ (Ιϲ� Χϲ� ; Ign. Phld. 8:2): Creedal Text-Coding and the Early Scribal System of Nomina Sacra 89 Tomas Bokedal

5

Segmentation and Interpretation of Early Pauline Manuscripts 123 S. Matthew Solomon

6

Can Papyri Correspondence Help Us to Understand Paul’s “Large Letters” in Galatians? 146 William Varner

7

The Tradition and Development of the Subscriptions to 1 Timothy 172 Linnea Thorp and Tommy Wasserman

8

Second Timothy: When and Where? Text and Traditions in the Subscriptions 202 Conrad Thorup Elmelund and Tommy Wasserman

vi

Contents

9

Composite Citations in New Testament Greek Manuscripts 227 Sean A. Adams and Seth M. Ehorn

10

Titus in P32 and Early Majuscules: Textual Reliability and Scribal Design 267 Chris S. Stevens

11

The Scribal Use of Ekthesis as a Paragraph Marker? The Galatians Text in Codex Sinaiticus as a Test Case 288 David I. Yoon

12

Miniature Codices in Early Christianity 310 Michael J. Kruger

13

Marginalia in New Testament Greek Papyri: Implications for Scribal Practice and Textual Transmission 330 Michael P. Theophilos



Conclusion: Paratextual Features: Summary and Prospects 358 Stanley E. Porter, Chris S. Stevens, and David I. Yoon Index of Ancient Sources 361 Index of Modern Authors 375

Preface This project was conceived as an effort to draw attention to a variety of textual phenomena often underappreciated by New Testament scholars. Common practices of New Testament textual criticism often reduce ancient manuscripts to mere repositories of textual variants. These textual variants are isolated, often as individual words or at best small groups of words, and then debated at individual points within the text of the New Testament. Single-minded approaches to manuscripts overlook the wealth of information the manuscripts contain about their textual history and church history. One important modern axiom is that an ancient manuscript is not just a repository of ancient readings, but it is a representation of the production and reception of the text as an entity that transcends the simple wording. How texts are represented within such manuscripts varies greatly but includes various elements often called paratextual features (for more on the terminology, see the Introduction and Chapter 1). This volume seeks to draw attention to various paratextual features by providing some introductory articles and making some advances in the field. A second modern axiom is that manuscripts are embedded within a context of social meaning. This context is complex and represents the collusion of a variety of social, cultural, historical, and even personal factors. However, the production of the manuscript does not restrict the meaning potential of a manuscript. Many manuscripts become living entities that continue to be received and interpreted, as evidenced by many editorial and scribal fingerprints upon them. This volume also attempts to capture how these social contexts play an important role in the study of ancient manuscripts. The editors agree that defining and exploring paratextual features is not easy, and the various essays will show that there are many different approaches to manuscript analysis. The paratextual features are perhaps distinguishable but inseparable from the complex matrix of textual production, reception, and interpretation, sometimes in harmony with and sometimes in opposition to each other. The essays within this volume provide a snapshot of some of these issues and how they are captured and then interpreted in various examples. The editors wish to thank all the contributors to this project. The final appearance of this volume has been unfortunately delayed by a variety of factors, not least the COVID-19 pandemic. Nevertheless, despite the turmoil and problems around us, the global scholarly community continues to explore ideas and make advances in research. The editors of this volume wish to thank

viii

Preface

our contributors for their diligence, patience, and willingness to be a part of this volume. We are optimistic that this volume will make a positive contribution to an emerging area within scholarship that merits much further recognition and devoted scholarship. Stanley E. Porter, Chris S. Stevens and David I. Yoon

Figures and Tables Figures 2.1

Basel University Library A. N. IV. 2 (GA1; fol. 277r). Source: http://ntvmr .uni-muenster.de; accessed 2020/7/27 36 2.2 Erasmus Roterodamus, Novum Instrumentum (Basel: Johannes Froben, 1516), 214 36 2.3 Constantinus Tischendorf, Novum Testamentum Graece ad antiquissimos testes denuo recensuit apparatum criticum omni studio perfectum (Leipzig: Giesecke & Devrient, 1869), 839 36 2.4 P75 Fol. 55r (Lines 9–11). Source: ntvmr.uni-muenster.de/community /modules/papyri/?zoom=39&left=2&top=49&site=INTF&image =10075/0/850/10/725 (accessed 2017/6/19) 38 2.5 P66: page 54 lines 12–14. Source: ntvmr.uni-muenster.de/community /modules/papyri/?zoom=30&left=2&top=69&site=INTF&image=10066 /0/500/10/66 (accessed 2017/6/16) 39 2.6 D/05 135v lines 25–27. Source: ntvmr.uni-muenster.de/community/modules /papyri/?zoom=30&left=2&top=69&site=INTF&image=20005/0/2720/10/965 (accessed 2017/6/19) 39 2.7 Novum Instrumentum omne p. 215 41 2.8 Tischendorf p. 844 42 2.9 Codex Bezae Cantabrigiensis John 8:45. Source: http://ntvmr.uni-muenster .de/community/modules/papyri/?zoom=71&left=3&top=-222&site=INTF &image=20005/0/2760/10/685; accessed 2017/06/27 43 2.10 Codex Basiliensis John 8:45. Source: http://ntvmr.uni-muenster.de/community /modules/papyri/?zoom=32&left=-5&top=-722&site=INTF&image=20007 /0/5590/10/3237; accessed 2017/06/27 44 2.11 Codex Sangallensis (037/Δ); John 8:43 (p. 352; l. 5–6). Source: http://ntvmr.uni -muenster.de; accessed 2017/07/04; for the Latin transcript cf. http://www .iohannes.com/XML/transcriptions/latin/04_VL27.xml#V-B04K8V43 46 2.12 Novum Instrumentum omne p. 215 47 2.13 Codex Colbertinus (Fol. 74r; col. 2; l. 32–35). Source: http://gallica.bnf.fr /ark:/12148/btv1b8426051s/f151.image.r=latin%20254; accessed 2017/07/05 48 2.14 Complutensian Polyglot John 8:43 48 2.15 Codex Bezae Cantabrigiensis John 8:43 (Fol. 137v 11–13). Source: http://ntvmr.uni-muenster.de; accessed 2017/7/3 49

x

Figures and Tables

2.16 Codex Basiliensis John 8:43. Source: http://ntvmr.uni-muenster.de; accessed 2017/7/3 49 2.17 Codex Claromontanus (D/06) 2 Cor 11:11 (Fol 238v). Source: http://ntvmr.uni -muenster.de; accessed 2017/7/3 50 2.18 Codex Augiensis (Fp/010; fol 63r). Source: http://ntvmr.uni-muenster.de; accessed 2017/7/3 50 2.19 Novum Instrumentum omne p. 227 52 2.20 Complutensian Polyglot John 12:40 53 3.1 sa 1 (p. 128, John 18:28) 86 3.2 sa 4 (p. 5, “Amen”-saying 1 at John 1:51 and coronis at 2:1) 86 3.3 sa 5 (p. 66, John 10:22) 86 3.4 sa 102 (p. 176, John 10:6–7) 86 3.5 sa 106 (p. 170, John 3:3) 86 3.6 sa 106 (p. 210, John 8:48–49) 86 3.7 sa 112 (p. 265, John 7:45–46) 87 3.8 sa 115 (p. 150, John 10:18–22) 87 3.9 sa 116 (p. 165, John 19:13) 87 3.10 sa 127 (p. 29, John 9:25–27) 87 3.11 sa 141 (p. 43, John 6:30–32) 87 3.12 sa 285 (fol. 139r, John 12:12) 87 3.13 sa 12 (John 9:1) 88 6.1 P.Oxy. 246 157 6.2 P. Col. VIII 216 158 6.3 P. Mich. V 351 159 6.4 P.Oxy. 3472 161 6.5 P.Oxy. 2770 162 6.6 P.Oxy. 3487 163 6.7 P. Fayum 110 165 6.8 P46 Gal 6:10–Phil 1:1 166 6.9 P46 Gal 6:10–12 167 6.10 Scroll Showing Galatians 171 9.1 Acts 16:26–32 in P74 229 9.2 Romans 11:36–12:1 in P46 (f. 16v) 229 9.3 Romans 3:10 in GA 012 (f. 4v) 230 9.4 Romans 10:6 in P46 (f. 13r) 230 9.5 Matthew 6:7 in Codex Sinaiticus (f. 74v) 232 9.6 Romans 3:10–12 in GA 012 (f. 4v) 233 9.7 1 Corinthians 14:21 in GA 012 (f. 36r) 234 9.8 Romans 11:26–27 in P46 (f. 15v) 235

Figures and Tables 9.9 Hebrews 10:37–38 in P13 (f. 64v) 235 9.10 John 12:13 in P66 (f. 87) 236 9.11 Romans 9:25 in P46 (f. 13v) 237 9.12 Hebrews 10:37–38 in P46 (32v) 237 9.13 Marginal Notation at Romans 3:10 in Codex Sinaiticus (f. 63r) 241 9.14 Matthew 21:5 in Codex Sinaiticus (f. 12v) 243 9.15 John 12:39 in Codex Vaticanus (p. 1369) 245 9.16 John 19:36–37 in Codex Vaticanus (p. 1379) 245 9.17 Acts 7:6–7 in Codex Vaticanus (p. 1390) 246 9.18 Matthew 21:5 in Codex Vaticanus (p. 1262) 247 9.19 2 Corinthians 6:16 in Codex Vaticanus (p. 1481) 248 9.20 Mark 1:2–3 in Codex Alexandrinus (f. 30r) 250 9.21 Acts 3:22–23 in Codex Alexandrinus (f. 83v) 251 9.22 Romans 9:25–26 in Codex Alexandrinus (f. 115r) 251 9.23 Dieresis use in John 12:41 in Codex Bezae (f. 154) 252 9.24 John 12:15 in Codex Bezae (f. 152) 252 9.25 Mark 1:2–3 in Codex Bezae (f. 286) 253 9.26 Romans 9:27–28 in Codex Claromontanus (p. 53) 255 9.27 Romans 11:26–27 in Codex Claromontanus (p. 65) 255 9.28 1 Corinthians 2:9 in Codex Claromontanus (p. 100) 256 9.29 Hebrews 10:37–38 in Codex Claromontanus (p. 514) 256 9.30 Romans 3:12a in GA 012 (f. 4v) 259 9.31 Romans 3:12b–18 in GA 012 (f. 5r) 259 9.32 Attribution Signal in GA 012 260 9.33 Romans 7:7 in GA 012 (f. 9v) 261 9.34 Romans 10:20 in GA 012 (f. 14r) 261 9.35 Romans 9:9 (f. 12v) 262 9.36 Romans 9:20 (f. 12v) 262 9.37 1 Timothy 5:18 (f. 89v) 262 9.38 1 Corinthians 15:54–55 (f. 39v) 262 9.39 1 Corinthians 14:21 (f. 36r) 262 9.40 Romans 13:9 (f. 17r) 262 9.41 Romans 10:15 (f. 15r) 262 9.42 1 Corinthians 2:9 (f. 22r) 262 9.43 2 Corinthians 6:16–18 in GA 012 (f. 46r) 263 9.44 Romans 11:34–35 in GA 012 (f. 15v) 263

xi

xii

Figures and Tables

Tables 4.1 Significant Jewish and biblical symbolic numbers 97 4.2a NS word-group frequencies in Ignatius’s letters 100 4.2b NS word-group frequencies in Ignatius’s letters 101 4.3 NS word-group frequencies in Barnabas and 1–2 Clement 104 4.4 Nomina Sacra in Eusebius’s Ecclesiastical History 106 4.5 NrN and Other Word frequencies in Early Christian and Jewish Literature 108 4.6 15/17/26-multiples from the NS word-group (THGNT) 119 4.7 22/24/27-multiples from the NS word-group (THGNT) 121 7.1 Subscription Collation and Apparatus 187 8.1 Subscription Collation and Apparatus 206 13.1 McNamee’s classifications indicating the subject matter of annotations 332 13.2 Marginalia in Papyri 352

Abbreviations 01, ℵ 02 03 05, D ABD

Codex Sinaiticus Codex Alexandrinus Codex Vaticanus Codex Bezae Cantabrigiensis Anchor Bible Dictionary, ed. David Noel Freedman, 6 vols. New York: Doubleday, 1992 AJP American Journal of Philology ANF Ante-Nicene Fathers ANTF Arbeiten zur neutestamentlichen Textforschung ANRW Aufstieg und Niedergang der römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung APF Archiv für Papyrusforschung ASP American Studies in Papyrology AThR Anglican Theological Review BASP Bulletin of the American Society of Papyrologists BDAG Danker, Frederick W., Walter Bauer, William F. Arndt, and F. Wilbur Gingrich. Greek-English Lexicon of the New Testament and Other Early Christian Literature. 3rd ed. Chicago: University of Chicago Press, 2000 BECNT Baker Exegetical Commentary on the New Testament BETL Bibliotheca Ephemeridum Theologicarum Lovaniensium BHGNT Baylor Handbook on the Greek New Testament BHT Beiträge zur historischen Theologie Bib Biblica BibTrans Bible Translator BICS Bulletin of the Institute of Classical Studies BIS Biblical Interpretation Series BJRL Bulletin of the John Rylands Library BSac Bibliotheca Sacra BN Biblische Notizen CBNTS Coniectanea Biblica New Testament Series CSNTM Center for Study of New Testament Manuscripts COMSt Comparative Oriental Manuscript Studies CQ Classical Quarterly DBW Dietrich Bonhoeffer Werken EC Early Christianity ECM Editio Critica Maior

xiv

Abbreviations

ExpTim The Expository Times ESV English Standard Version FN Filología Neotestamentaria FRLANT Forschungen zur Religion und Literatur des Alten und Neuen Testaments GA Gregory-Aland Numbers GNB Good News Bible GNT Greek New Testament HTR Harvard Theological Review HTS Harvard Theological Studies Hermeneutische Untersuchungen zur Theologie HUT Journal of Biblical Literature JBL Journal of the Evangelical Theological Society JETS Journal of Greco-Roman Christianity and Judaism JGRChJ Journal of Translation and Textlinguistics JOTT Journal of Reformed Theology JRT Journal for the Study of the New Testament JSNT JSNTSup JSNT Supplement Series JSOT Journal for the Study of the Old Testament Journal of Theological Interpretation JTI JTS Journal of Theological Studies King James Version KJV Linguistic Biblical Studies LBS LDAB Leuven Database of Ancient Books LNTS The Library of New Testament Studies LSJ Liddell, Henry George, Robert Scott, Henry Stuart Jones. A Greek-English Lexicon. 9th ed. with revised supplement. Oxford: Clarendon, 1996 Library of Second Temple Studies LSTS Novum Testamentum Graece, Nestle-Aland, 28th ed. NA28 NASB New American Standard Bible NewDocs New Documents Illustrating Early Christianity Nag Hammadi and Manichaean Studies NHMS NICNT New International Commentary on the New Testament New International Greek Testament Commentary NIGTC New International Version NIV NovT Novum Testamentum NovTSup Novum Testamentum Supplements NPNF Nicene and Post-Nicene Fathers NRSV New Revised Standard Version NovT Novum Testamentum

Abbreviations NTS NTTSD NT.VMR PAST P. Bodmer PNTC PTS RB RILP RP SBLGNT SNTSMS SNTU SSLL STAC STDJ TCSt TCNT TDNT

TENTS THGNT TrinJ TS TSAJ TUGAL TynBul UBS VC VT WBC WH WUNT ZAC ZNW ZPE

xv

New Testament Studies New Testament Tools, Studies, and Documents New Testament Virtual Manuscript Room Pauline Studies Bodmer Papyri Pillar New Testament Commentary Patristische Texte und Studien Revue biblique Roehampton Institute London Papers Robinson-Pierpont Greek New Testament The Greek New Testament: SBL Edition Society for New Testament Studies Monograph Series Studien zum Neuen Testament und seiner Umwelt Studies in Semitic Languages and Linguistics Studien und Texte zu Antike und Christentum Studies on the Texts of the Desert of Judah Text Critical Studies Tyndale Commentary on the New Testament Theological Dictionary of the New Testament. Edited by Gerhard Kittel and Gerhard Friedrich. Translated by Geoffrey W. Bromiley. 10 vols. Grand Rapids: Eerdmans, 1964–1976 Texts and Editions for New Testament Study Tyndale House Greek New Testament Trinity Journal Texts and Studies Texts and Studies in Ancient Judaism Texte und Untersuchungen zur Geschichte der altchristlichen Literatur Tyndale Bulletin The Greek New Testament, United Bible Societies, 5th ed. Vigiliae Christianae Vetus Testamentum Word Biblical Commentary Westcott-Hort Greek New Testament Wissenschaftliche Untersuchungen zum Neuen Testament Journal of Ancient Christianity/Zeitschrift für Antikes Christentum Zeitschrift für die neutestamentliche Wissenschaft Zeitschrift für Papyrologie und Epigraphik

Contributors Sean A. Adams, University of Glasgow, Scotland, UK Tomas Bokedal, University of Aberdeen, Scotland, UK Seth M. Ehorn, Wheaton College, IL, USA Conrad Thorup Elmelund, University of Copenhagen, Denmark Hans Förster, University of Vienna, Austria Michael J. Kruger, Reformed Theological Seminary, Charlotte, NC, USA Stanley E. Porter, McMaster Divinity College, Hamilton, ON, Canada Matthias H. O. Schulz, University of Vienna, Austria S. Matthew Solomon, Luther Rice College and Seminary, Lithonia, GA, USA Chris S. Stevens, McMaster Divinity College, Hamilton, ON, Canada Michael P. Theophilos, Australian Catholic University, Victoria, Australia Linnea Thorp, Ansgar University College and Theological Seminary, Kristian­ sand, Norway William Varner, The Master’s University, Santa Clarita, CA, USA Tommy Wasserman, Ansgar University College and Theological Seminary, Kristiansand, Norway David I. Yoon, McMaster Divinity College, Hamilton, ON, Canada | Emmanuel Bible College, Kitchener, ON, Canada

Introduction

Paratextual Features of Early Greek Manuscripts Stanley E. Porter, Chris S. Stevens, and David I. Yoon Holding a book in one’s hands is a multisensory experience. It involves touch, sound, smell, and sight. The experimental psychologist Charles Spence explains that there is a “multisensory stimulation that interacting with a print book necessarily delivers.”1 While the modern book is different from documents from the ancient world – to say nothing of how we experience the sensations of digital technology – encountering an ancient manuscript is also multisensory. Beyond the aesthetic experience, which itself is not to be minimized, the act of reading an ancient manuscript is a multifaceted and intricate experience. Readers usually anticipate encountering a text, but they are also confronted with much more than just words. A viewer also confronts many other visual elements, many of which are consciously crafted by the creators to contribute to the overall potential to convey meaning. One is confronted with a font or fonts of various sizes and shapes, ink color, page dimensions, varying margins, intratextual headings, sometimes headers or footers or marginalia, artwork such as illuminations of various types, and a myriad of other bibliographic features. Examining an ancient manuscript by means of high-resolution devices adds to the intensity of the multisensory and multifaceted experience. All these features are part of the meaning resource of an ancient manuscript. Ancient manuscripts were created with these features relatively early in the human writing experience and these features persist to the present. Such features are especially associated with handwritten manuscripts, but they also persist to the present, in the form of not just printed but manual additions to the modern printed book, such as underlining and notations. Even within the scope of ancient manuscripts, there is an abundance of data to consider. We cannot consider all such manuscripts, nor all the possible kinds of phenomena. The focus of this volume is the myriad of features in ancient biblical manuscripts. Even within that scope, there is an abundance of features to consider. In this introduction, we introduce some of the important concepts surrounding the notion of those features of a manuscript that are not the text itself. Even defining such features raises a host of interesting questions. 1 Charles Spence, “The Multisensory Experience of Handling and Reading Books,” Multisensory Research 33 (2020): 902–928, here 904.

© Stanley E. Porter et al., 2023 | doi:10.1163/9789004537972_002

2

Porter, Stevens and Yoon

1 Paratexts A traditional approach in the field of biblical textual criticism is to distinguish between the textual and extratextual features. These extratextual features are often referred to as the paratextual features. In the broadest sense, “all contents in biblical manuscripts, except the biblical text itself, are a priori paratexts.”2 However, this distinction is problematic. Following along the lines of the New Philology, Stanley E. Porter rightly calls into question attempts to distinguish between, or worse erect a dichotomy between, text and paratext. In the first article in this volume, Porter provocatively contends that the textual and paratextual cannot and should not be separated. The reason for erasing such distinctions is aptly stated by the Old Norse textual scholar Matthew Driscoll, namely that “literary works do not exist independently of their material embodiments.”3 Similarly, Mary A. Rouse contends that There is a relationship between the physical form of an artifact and the function that it is meant to serve; therefore, the physical form of a manuscript is evidence of the purpose for which the manuscript was made. This fact, simple and direct, is the inevitable starting point of any study of the manuscript as an object.4 While approaches vary, the contributors to this volume believe that it is vitally important to thoroughly analyze every aspect of ancient biblical manuscripts for their meaning potential. Our contributors believe that the features that have often been neglected have much to teach us about the meaning and function of a manuscript. Therefore, we contend that all features of a manuscript are necessary to analyze. These may be heuristically distinguished in relation to, but not separated into, three categories: physical object, literary text, and paratextual. The physical category concerns the material used as the writing surface: 2 Patrick Andrist and Martin Wallraff, “ParaTexBib: An ERC Project Dedicated to Paratexts in Greek Manuscripts of the Bible,” COMSt Bulletin 2 (2016): 63–68, here 63. Similarly, Garrick V. Allen and Anthony P. Royle (“Paratexts Seeking Understanding: Manuscripts and Aesthetic Cognitivism,” Religions 11.523 [2020]: 1–25, here 3) contend that paratexts are “every part of a manuscript beyond the main text itself, especially features that have aesthetic qualities insofar as they negotiate a text to readers.” 3 Matthew J. Driscoll, “The Words on the Page: Thoughts on Philology, Old and New,” in Creating the Medieval Saga: Versions, Variability and Editorial Interpretations of Old Norse Saga Literature, ed. J. Quinn and E. Lethbridge (Odense: University Press of Southern Denmark, 2010), 87–104, here 90. 4 Mary A. Rouse and Richard H. Rouse, Authentic Witnesses: Approaches to Medieval Texts and Manuscripts (Notre Dame: University of Notre Dame Press, 1991), 4.

Introduction

3

papyrus, type of animal skin, clay, stone, bone, or centuries later paper.5 The textual category is the most familiar to those in the humanities and biblical studies, namely the text that is represented through the characters written on the physical material and representing a unit of meaning. The third category is the paratextual. The focus of this book is on this often-underappreciated category, namely the extratextual or the paratextual elements that are found on the object other than the notional text. The manuscript features beyond the literary text represent choices made by scribes, choices and resulting elements in addition to the representation of the main textual content. The reader of an ancient manuscript is confronted by a potentially wide array of such paratextual features. These designed features are the product of choices before, during, and even after the writing of the text. In that sense, an ancient manuscript is an organic and changing, and even developing, artifact that expresses its text in new and developing ways. The most common paratextual features that are found in ancient biblical manuscripts are: – Headings – Introductions – Marginal comments (prima manus and post-publication) – Colophons, occasional inclusion of the scribe’s name, the date of production, location, and sometimes additional information about the manuscript production – Features of layout: margin, header, footer, number of columns, space between columns, and script size and type – Reading aids: segmentation, paragraphos, ekthesis, eisthesis or indent, obelis, coronis, ekphonetic notation, diplai, rubrication, and even new nomenclature that is proposed in this volume. All these paratextual phenomena are the results of decisions and they confront a reader when the reader examines a page of an ancient manuscript. As Harry Gamble notes, “it is the physical presentation of the text that is most immediately evident and effective for its readers.”6 Modern experience illustrates the 5 Andrist and Wallraff (“ParaTexBib,” 63) explain that, in the newly launched ParaTexBib project, “decorative elements and codicological features such as the page numbers or quire signatures are not considered paratexts, even though they may have a paratextual valence.” These widely recognized paratextual features are products of scribal work that have value in codicological research. The focus of this book is on scribal work that is not so obvious in its communicative value, although we would not want to minimize the significance of these features that Andrist and Wallraff too quickly dismiss. 6 Harry Y. Gamble, Books and Readers in the Early Church: A History of the Early Christian Texts (New Haven, CT: Yale University Press, 1995), 42.

4

Porter, Stevens and Yoon

point. A student opens a used book and is confronted with highlighting, underlining, or marginal notes. These visual signals draw attention to how a previous reader responded to and even enhanced or embellished the manuscript, to the point of affecting the text itself. These visual additions are post-publication means of emphasizing, shaping, and even interpreting discrete textual content. The visual additions have effectively changed the reading experience for subsequent readers. Similarly, the creators of texts curate the reading experience by making a myriad of decisions that encode layers of interpretation. The scribal decisions are means of their creator to communicate to the reader or reading community how they desire the manuscript to be received and how they desire for the text to be read as a meaningful text. History results in potential layers of interaction represented and recorded in a single manuscript. Some of these layers are created at the time of creation of the manuscript, while others are part of a continuing and extended process of reception and interpretation of the manuscript and its text. The focus of this volume is on the pre-publication crafting of a manuscript, although we recognize that there is much that could and probably should be said about subsequent paratextual elaboration.7 2

Why are Paratextual Features Important?

Ancient manuscripts are ancient artifacts that empower our studying the past because they provide windows into history. Unfortunately, the full potential of manuscripts is often neglected. As Peter Van Minnen notes, “most papyrologists are mainly interested in the texts written on papyri … most of them study papyri because they have a passionate interest in the past and because they are trained to construe that past from texts.”8 However, there is far more information conveyed through a manuscript than the text.

7 ParaTextBib is a project that explores both pre- and post-publication features. As Wallraff and Andrist (“Paratexts of the Bible: A New Research Project on Greek Textual Transmission,” EC 6 [2015]: 237–243, here 239) contend, the “paratexts need not be part of the original editorial project of a manuscript.” However, the textual features examined in this project are the ones created by the prima manus and the editorial work done prior to publication or circulation. Of course, it is difficult sometimes to distinguish the debated categories of pre-publication and post-publication markings on ancient manuscripts, but ink, palaeographical details, and location are essential to the investigation. 8 Peter Van Minnen, “The Future of Papyrology,” in The Oxford Handbook of Papyrology, ed. Roger S. Bagnall (Oxford: Oxford University Press, 2009), 644–660, here 644–645.

Introduction

5

One of the primary goals of this volume is to explore how textual criticism is more than the study of the biblical text; it is the study of history and of the history of a text by means of examining and interpreting the paratextual features of manuscripts.9 Textual critics have long noted that the history of the text is interwoven with what is usually called church history.10 However, this label can be deceiving, because church history is not a form of history apart from other types of history, but concerns matters of the church within the wider scope of history, in particular social history. Manuscripts are the products of communities, whether these are seen as ecclesial, scholarly, scribal, or otherwise. The paratextual features are fundamentally historical elements requiring interpretation to divulge information concerning the scribes, the text, the communities involved, that is, the intent and use of the manuscript, as witnesses to the acts of writing and reading. To neglect these paratextual features is to neglect a vital resource that opens a window into the early communities around these texts, in both their production and reception. Michael Kruger describes the concern well: These ancient Christian books are artifacts – physical objects that real Christians manufactured and used in the second and third centuries. Not only do these ancient documents give valuable insight into the texts of the biblical books they contain, but they are equally valuable in giving insight into the manner in which early Christians used these writings. This information can be gleaned by examining the ‘paratextual’ features.11

9

Wayne C. Kannaday (Apologetic Discourse and the Scribal Tradition: Evidence of the Influence of Apologetic Interests on the Text of the Canonical Gospels, TCSt 5 [Atlanta: SBL, 2004], 238) notes that “textual criticism is at its essence not only a literary enterprise, but a historical discipline as well.” 10 J. Keith Elliott (“The New Testament Text in the Second-Century: A Challenge for the Twenty-First Century,” New Testament Textual Research Update [2000]: 13–27, here 9) contends, “textual criticism of the New Testament should never be practiced outside church history.” See also Kenneth W. Clark, “The Theological Relevance of Textual Variation in Current Criticism of the Greek New Testament,” JBL 85 (1966): 1–16, here 16; Kirsopp Lake, The Influence of Textual Criticism on the Exegesis of the New Testament (Oxford: Parker and Son, 1904); Donald Wayne Riddle, “Textual Criticism as a Historical Discipline,” AThR 18 (1936): 220–233. 11 Michael J. Kruger, Canon Revisited: Establishing the Origins and Authority of the New Testament Books (Wheaton, IL: Crossway, 2012), 234. For an exploration of this quotation, see Timothy N. Mitchell, “Christian Papryi and the Ancient Church,” BSac 173 (2016): 182–200.

6

Porter, Stevens and Yoon

The crafting of a manuscript in the ancient world was not like the use of a modern copy machine. One did not take in one’s hand a manuscript and proceed to produce machine-created replicas. The manuscripts that we are concerned with were human products, and therefore were the literal result of manual labor. But they are more than that. There is far more work and many more decisions required that are not much different from the efforts of an artist working with their medium. For instance, consider the now-common use of headings. A scribe may choose to provide a heading at the beginning of a text, a section, or a subsection. The inclusion of a heading previews the following content by shaping expectations. One must first decide that there is to be a heading, and then the content and wording of that heading. Some headings are traditional while others are occasional. One must also decide where to place the heading. The choice of placement of a heading also interjects a palpable break or at least a transition from what precedes. Similarly, a line of text may be bracketed in some fashion to subordinate it, or possibly highlight it, from the main text for editorial comment, marginal note, or indication of the uncertainty of the textual reading. Conversely, a portion of text may be spatially segmented from the surrounding text to draw attention, like highlighting, bolding, or blocking quotations in modern books. Progress has been made in recent decades by researchers asking more of manuscripts than in times past. One popular – and perhaps even infamous – area of investigation concerns the social history of textual variations. In times past, textual readings were discriminated into the binary category of wrong or right, with the wrong ones being discarded as useless in the aim of genealogically recovering the original textform, or Ausgangstext. In the last four decades, textual variants have come to be appreciated as more than chaff to be blown away by the wind. Instead, textual variants may at times, in the words of New Testament textual critic Bart Ehrman, “provide data for the social history of early Christianity.”12 While Ehrman has a controversial aim, he is correct to conclude that the “NT MSS. can thus serve as a window into the social world of early Christianity.”13 Gamble stresses the same point but extends the argument. Gamble sees manuscripts as vital to historical research, concluding that “the physical object is also a social artifact.”14

12 Bart D. Ehrman, “The Text as Window: New Testament Manuscripts and the Social History of Early Christianity,” in Studies in the Textual Criticism of the New Testament (Leiden: Brill, 2006), 101. 13 Ehrman, “Text as Window,” 101. 14 Gamble, Books and Readers, 43.

Introduction

7

Biblical studies unfortunately has generally given little attention to detailed textual criticism and practically no attention to paratextual features. As Martin Wallraff and Patrick Andrist note, “research up till now has never developed a systematic interest in these elements (paratextual) since the manuscripts were taken mainly as witnesses for textual criticism – that is, as a means to reconstruct some distant Urtext.”15 For those doing biblical studies, the manuscripts are primarily only of interest in establishing the Ausgangstext. However, neglecting to use the manuscripts as artifacts is a detriment to biblical studies. Gamble warns, The failure to consider the extent to which the physical medium of the written word contributes to its meaning – how its outward aspect informs the way a text is approached and read – perpetuates a largely abstract, often unhistorical, and even anachronistic conception of early Christian literature and its transmission.16 Perhaps the best way to explain the importance of paratextual features, and the power they wield (whether rightly or not), is to use a recent example of paratextual examination. Over the last few years, a couple of dots, in nontechnical language, have received heated debate. The heart of the question is whether these couple of dots are evidence for the inclusive or exclusive role of women in the ancient church. In 1995, Philip Barton Payne argued that two horizontally-aligned dots at mid-character in 1 Corinthians in Codex Vaticanus were indicators of the non-originality of 1 Cor 14:34–35.17 In 2000 and 2009, Payne and Paul Canart further defended their contention that the distigmai-obelos, as they came to be called, were from the original hand (prima manus) of Vaticanus as indicated by the matching ink.18 More recently, in 2017,

15 Wallraff and Andrist, “Paratexts of the Bible,” 238. 16 Gamble, Books and Readers, 42. 17 Philip B. Payne, “Fuldensis, Sigla for Variants in Vaticanus, and 1 Cor 14.34–35,” NTS 41 (1995): 240–262. 18 Philip B. Payne and Paul Canart, “The Originality of Text-Critical Symbols in Codex Vaticanus,” NovT 42 (2000): 105–113; Philip B. Payne and Paul Canart, “Distigmai Matching the Original Ink of Codex Vaticanus: Do They Mark the Location of Textual Variants?” in Le manuscript B de la Bible (Vaticanus Graecus 1209): Introduction au facsimilé Actes du Colloque de Genève (11 juin 2001): Contributions supplémentaires, ed. Patrick Andrist (Lausanne: Zèbre, 2009), 199–206. Payne also extensively discusses 1 Cor 14:34–35 in his book, Man and Woman, One in Christ: An Exegetical and Theological Study of Paul’s Letters (Grand Rapids: Zondervan, 2009), 217–270.

8

Porter, Stevens and Yoon

Payne further refined his argument, concluding that “the two-dot-plus-bar ‘distigme-obelos’ symbols in Vaticanus signal added text.”19 While the merits of Payne’s argument were vigorously repudiated by respondents, the potential impact of his proposal is noteworthy, to say the least. Important for our purposes here is to show that interpretation of paratextual features has impact upon the field of biblical studies. If Payne were correct in his interpretation of the distigmai-obelos, then some of our understandings of Pauline theology, New Testament theology, and early ecclesiastical practice are directly affected. According to Payne, the paratextual feature indicates that Paul did not write in 1 Cor 14:34–35 that women are to keep silent in the church. More importantly, according to Payne’s argument, the verse should be removed from this place in the New Testament, with all the implications of such removal within wider ecclesiastical discussions. Unsurprisingly, Payne’s work has garnered much attention, with most scholars not accepting his interpretation of this paratextual feature or its implications. Richard Fellows and Jan Krans both published rebuttals in their reassessments of the evidence.20 Fellows concludes, The crux of Payne’s argument is that in eight cases, including Cor., the horizontal line contains features that distinguish it from a normal paragraphos.… However, he has made systematic measurement errors.… the bar at the start of Cor. is indistinguishable from other paragraphoi. Payne’s counter-arguments result from errors of measurement and method.21 Krans also acknowledges that “the communicative value of the distigmai is an important issue as well. They signal something, but for whom?”22 However, Krans concludes, “Payne’s views are almost entirely to be rejected.”23 Despite the strongly held opinions on either side, what is clear is that this paratextual feature has clearly aroused opinions and responses that possibly threaten established interpretations, so it is not always easy to decide whether

19 Philip B. Payne, “Vaticanus Distigme-Obelos Symbols Marking Added Text, Including 1 Corinthians 14:34–35,” NTS 63 (2017): 604–625, here 604. 20 See also Jennifer Shack, “A Text without 1 Corinthians 14.34–35? Not according to the Manuscript Evidence,” JGRChJ 10 (2014): 90–112. 21 Richard G. Fellows, “Are There Distigme-Obelos Symbols in Vaticanus?” NTS 65 (2019): 246–251, here 250–251. 22 Jan Krans, “Paragraphos, Not Obelos, in Codex Vaticanus,” NTS 65 (2019): 252–257, here 253. 23 Krans, “Paragraphos, Not Obelos,” 252.

Introduction

9

arguments or emotions are offered for consideration. While not all paratextual features are as potentially impactful, paratextual features in ancient biblical manuscripts should shape and continually reformulate our understanding of the history of Christianity, not least of which are church development and canon formation. 3

The Contribution of This Volume

Even though not all paratextual features have the potential implications of the instance noted above, there are still many that merit further consideration. This volume attempts to address some of the major issues in the discussion. In the first essay, Stanley E. Porter offers a provocative challenge to the distinction between textual and paratextual features. As we have already intimated in comments above, we may all believe that we can identify a manuscript and distinguish text from paratext. This essay recognizes the complexity but begins by discussing the notion of text before turning to paratext. Even such a notion as text is subject to reconsideration before one can explore paratext. However, the definition and identification of paratext is of much more recent vintage than many recognize and it is not altogether clear how one differentiates paratextual from textual and other features. As a necessary introduction to the complex of the issues involved, this essay explores the notion of paratext in relationship to various views of text as an attempt to problematize an issue that, in fact, not only merits problematization but benefits from it. In Chapter 2, “Missing the Point,” Hans Förster highlights examples of how differences in punctuation have an impact on the interpretation of a text. This chapter addresses an important feature of ancient manuscripts – they are written in continuous script or scriptio continua. Without spaces between letters, or except in rare places segmentation, there arises some ambiguity in particular constructions. For instance, in scriptio continua οτι could be ὅ τι or ὅτι. Similarly, where one sentence ends and another begins is often a matter of interpretation. Förster explores John 8:25, 43, 45, and 12:40 in the ancient manuscripts and the history of interpretation to illustrate. Matthias H. O. Schulz turns his attention to Sahidic and minor Coptic dialect manuscripts of the Gospel of John. Schulz notes that the chapter divisions of modern Bibles are relatively new in the history of editions of the New Testament. For nearly fifteen hundred years the Gospel of John did not have the modern divisions made by the sixteenth-century printer and scholar Robert Estienne (Robertus Stephanus). In a first of its kind study, Schulz examines intermarginal signs within Sahidic Coptic manuscripts of John to determine

10

Porter, Stevens and Yoon

their function. He examines all 173 Coptic manuscripts to document and explore all indicators of textual division. In an exploration of a fascinating feature of ancient manuscripts, Tomas Bokedal in “But for Me, the Scriptures are Jesus Christ” explores the so-called sacred names or nomina sacra. Bokedal gives detailed calculations on the use of these unique scribal features in biblical and sub-apostolic writings. Bokedal sees these short forms of particular New Testament figures as having a role within trinitarian theological discussions. Furthermore, Bokedal hopes that the numerological evidence he finds will renew nomina sacra debates in new ways. Bokedal concludes that perhaps modern printed Bibles should resume the inclusion of the nomina sacra for visual and theological significance. In an essay that continues discussion from above, S. Matthew Solomon also investigates the interpretive challenges of scriptio continua in “Segmentation and Interpretation.” The division of strings of letters into discrete words, or more importantly the lack thereof, is a factor that scribes had to contend with in the transmission of the text. Solomon discusses the use and absence of textual segmentation in the Pauline papyri. He focuses on exegetically significant examples in 1 Corinthians 7, 14, and Ephesians. William Varner addresses a detail of ancient letters that could explain an interesting feature in Paul’s correspondence. In Gal 6:11, the Apostle Paul writes, Ἴδετε πηλίκοις ὑμῖν γράμμασιν ἔγραψα τῇ ἐμῇ χειρί, “see with what large letters I am writing to you in/with my hand.” These words may be, and in fact have been, understood in multiple ways. Varner examines other ancient epistolary writings to conclude that Paul was following a common practice for authentication of his letter. In a further essay on Pauline manuscripts, Tommy Wasserman and Linnea Thorp look at the role and function of subscriptions in “Tradition and Development of the Subscriptions to 1 Timothy.” Their detailed work examines the subscription in over 300 manuscripts to trace the history and development of the paratextual feature closing the letter. The article will serve as a resource for both 1 Timothy studies and subscriptions generally. The authors conclude that the subscriptions provide important paratextual details about the inclusion and role of 1 Timothy within the Pauline corpus. In this next essay, Tommy Wasserman and Conrad Thorup Elmelund continue a study on subscriptions in their “Second Timothy: When and Where? Text and Traditions in the Subscriptions.” The authors build on the research method conducted on the manuscripts of 1 Timothy, by turning to the second letter in this mini-corpus. No previous work has examined all

Introduction

11

manuscripts of 2 Timothy, making this chapter a valuable addition to scholarship on subscriptions. Sean A. Adams and Seth M. Ehorn offer a detailed exploration of scribal work in multiple manuscripts in “Composite Citations in New Testament Manuscripts.” The authors explore the various ways scribes communicate readers aids through blank spaces, punctuation marks, ekthesis and indentation, paragraphoi, diplai, rubrication (ink color changes), and other small scribal additions. The authors offer a wealth of details that highlight the nature and use of ancient biblical manuscripts. In a third essay on one of the Pastoral Epistles, Chris S. Stevens brings together multiple lines of analysis in “Titus in P32 and Early Majuscules.” Stevens conducts manuscript criticism, a term from poetics, to examine the physical details of a manuscript from multiple angles without interpretation of the textual content. The article compares details concerning the layout, spatial segmentation, margins, and overall scribal design of P32 with the early majuscules. Stevens concludes that P32 is part of a multi-text codex and that Sinaiticus has distinctive segmentation features in 1 Timothy and Titus to aid in public reading. David I. Yoon explores in further detail the scribal practice of ekthesis in Sinaiticus. The protruding text into the left margins of columns has elicited various interpretations. A dominant line of thinking is that the scribal feature represents a paragraph marker. However, Yoon finds that the function of these features has to date not been properly defined, including the problematic definition of an ancient paragraph. Drawing on Systemic Functional Linguistics, and focusing on Galatians in Sinaiticus, Yoon offers robust, linguistically defined criteria for paragraphs and ekthetic markers. Turning to a less familiar area within biblical studies, Michael J. Kruger examines paratextual features in “Miniature Codices in Early Christianity.” Kruger gives plenty of evidence to demonstrate that miniature codices are mistakenly overlooked resources for understanding early Christianity and its developing textual tradition. Kruger explores what miniature codices tell us about ancient book culture within Christian communities and the fascinating possibilities of their function. Michael P. Theophilos gives detailed analysis to marginal notes in a variety of New Testament Greek papyri. The scribal and reader additions of notes are a testament to the use of ancient papyri, as they record additional information that is incorporated into the manuscript. Some marginalia give evidence of early interaction with scribal variants and the way early Christians arrived

12

Porter, Stevens and Yoon

at conclusions concerning selection of textual readings. While many New Testament papyri do not preserve marginalia in their extant form, Theophilos documents and comments on all those that do. The systematic analysis of paratextual features in biblical manuscripts is a developing field of research. It unites traditional textual analytic techniques from such fields as papyrology with new questions about history, social setting, and context. The following essays showcase various ways that paratextual features offer insights into the use and function of ancient manuscripts within their socio-historical setting. The individual authors provide a rich application of traditional manuscript study and pioneer new methods of analysis. These are offered as an encouragement to further study of not just the text but that which is not the text.

CHAPTER 1

What Is Paratext? In Search of an Elusive Category Stanley E. Porter 1 Introduction This paper is concerned with the phenomenon often called paratext or sometimes metatext. The term paratext is of recent provenance, as we will discuss in the next section, but the kind of phenomena that are often captured under the term have been noticed in textual studies for a long time. Even though the term paratext was originally derived as a means of describing phenomena in modern printed books, this paper is at least as, if not more, interested in ancient manuscripts. There are many similarities between the two as artifacts that present texts in text-like environments to be read and understood. However, there are also significant differences between the two. These differences revolve around a variety of factors, with some linked to the technology used in their production (modern printing verses scribal writing) but others attached to the purposes, use, and even interpretation of the artifacts involved. As a result, there are particular issues raised by each that, despite their similarities, distinguish the two in discussion. In this paper, I wish to address the question of what constitutes paratext. I will do so by first examining the origins and definition of the term. I will then examine the concept of text that is implied by the use of the term paratext. I will conclude by revisiting the question of paratext in an attempt to arrive at a definition of this problematic term in textual studies. 2

The Origin of Paratext

The definition of what constitutes a paratext is one that continues to bedevil discussion of ancient (and even modern) manuscripts.1 I can only imagine 1 I have written on related topics several times before, but never addressing the specific question of what constitutes paratext. In the past, I have tended to assume that there is such a thing as paratext and have even provided a definition of what constitutes paratext. I am thankful for the opportunity to write this paper, as it has encouraged me to look in more detail at what paratext is and, even more importantly, whether there is such a thing as paratext.

© Stanley E. Porter, 2023 | doi:10.1163/9789004537972_003

14

Porter

that some may even raise questions about the definition of the term on the basis of the some of the other essays that are found within this volume. The term “paratext” was apparently created by the French narratologist and literary scholar Gérard Genette in the course of writing a number of works during the late 1970s and 1980s.2 Although he appears to have developed in his understanding of the term, he appears to have used the term paratext in relation to a number of other concepts.3 The first distinction is that paratext is a form of “pastiche” or “parody.”4 Genette further refines the notion of paratext by seeing it as one of five types of transtextuality, that is, “the textual transcendence of the text” or its complex of relationships to other texts.5 At this stage, Genette defines paratext as “a title, a subtitle, intertitles; prefaces, postfaces, notices, forewords, etc.; marginal, infrapaginal, terminal notes; epigraphs; illustrations; blurbs, book covers, dust jackets, and many other kinds of secondary signals, whether allographic or autographic.”6 However, Genette apparently extends the notion of paratextuality by making a further distinction between what he 2 The major works are Gérard Genette, Introduction à l’architexte (Paris: Seuil, 1979; ET The Architext: An Introduction, trans. Jane E. Lewin [Berkeley: University of California Press, 1992]); Genette, Palimpsestes: La littérature au second degré (Paris: Seuil, 1982; ET Palimpsests: Literature in the Second Degree, trans. Channa Newman and Claude Doubinsky [Lincoln: University of Nebraska Press, 1997]); and Genette, Seuils (Paris: Seuil, 1987; ET Paratexts: Thresholds of Interpretation, trans. Jane E. Lewin [Cambridge: Cambridge University Press, 1997]). 3 See Patrick Andrist, “Toward a Definition of Paratexts and Paratextuality: The Case of Ancient Greek Manuscripts,” in Bible as Notepad: Tracing Annotations and Annotation Practices in Late Antique and Medieval Biblical Manuscripts, ed. Liv Ingeborg Lied and Marilena Maniaci, Manuscripta biblica 3 (Berlin: De Gruyter, 2018), 130–149, esp. 130–132. 4 Genette, Introduction, 82 (cited by Andrist, “Toward a Definition,” 131 n. 5). 5 Genette, Palimpsests, 1. The other four are intertextuality, the “copresence” among two or more texts; metatextuality, or commentary; hypertextuality, a hyponymous relationship among texts; and architextuality, or relationships of genre (1–5). Genette’s language raises questions about the text, ontology, and essentialism. 6 Genette, Palimpsests, 3. Cf. Stanley E. Porter, “The Domains of Textual Criticism and the Future of Textual Scholarship,” in The Future of New Testament Textual Scholarship: From H. C. Hoskier to the Editio Critica Maior and Beyond, ed. Garrick V. Allen, WUNT 417 (Tübingen: Mohr Siebeck, 2019), 131–153, esp. 151, where I list the following as paratext: “the ink, the lettering not just in its broad style but in the particular ways letters are formed, the material peculiarities of this artefact …, spacing, letter sizing, placement upon the sheet (margins), lines per page or column, number and types of columns, scroll or codex and what kinds and sizes and contents, the length of material on a sheet and other codicological features, corrections, any markings indicating paragraphing or completing of lines, etc., supra- and sublinear lines, the use of nu-lines and sometimes smaller letter, nomina sacra, other forms of figuration such as staurograms, abbreviations and other kinds of shortened or reduced or otherwise transcribed forms, phonetic notation of any type, accentuation of any type, punctuation of any type, or any related features, headings, notes, marginalia of any type, or prosodic and

What Is Paratext?

15

calls “peritext” and “epitext,” although Genette offers the formula: “paratext = peritext + epitext.”7 This distinction is made in relation to that which is outside and inside the published product, so that epitext is concerned with what is outside and extends the text (e.g. “interviews, conversations … letters diaries, and others”), and peritext is concerned with what is inside (e.g., “the title or the preface and sometimes elements inserted into the interstices of the text, such as chapter titles or certain notes”).8 Patrick Andrist rightly finds problems with Genette’s definition of paratext, besides the fact that it presents itself as a concept in a constantly developing process, as admitted by the author, to the point where it becomes difficult to specify exactly what is meant by paratext. Is it to be equated with both peritext and epitext,9 or is it the more restricted definition that seems to be associated with epitext?10 Genette seems to use the terms ambiguously. Andrist, however, believes that the terminology, at least as Genette defines it, is not usable for the study of ancient texts, because it was explicitly developed in relationship to the modern printed book, as Genette states. The difficulty that this provides, at least for Andrist, is seen in how Genette links paratext to authorial control. Paratext is defined, at least by Genette, in terms of authors or their representatives being responsible for it as something at least implicitly sanctioned by the author.11 However, to add a further wrinkle, Genette further differentiates between authorial and publisher peritext.12 Authorial peritext encompasses “his preface, titles, subtitles, etc.,” while publisher peritext encompasses “the cover, all the writing on the cover pages of the book, the title page, etc.,” as well as “the format, the typesetting and the paper” and “illustrations.”13 Genette recognizes that his theory was focused upon the modern printed book, but also recognizes that there were peritext elements in the pre-printed book

discourse features, whether they were part of the original scribal hand or added later and for whatever reason, and no doubt numerous other features that I simply have not identified.” 7 Genette, Paratexts, 5. 8 Genette, Paratexts, 5. 9 As it appears to be by Genette in Paratexts, 5. 10 As it is in the earlier definition offered by Genette in Palimpsests, 3. 11 Genette, Paratexts, 9–10, where he states, “By definition, something is not a paratext unless the author or one of his associates accepts responsibility for it, although the degree of responsibility may vary” (9). 12 Genette, Paratexts, 16–36, in a chapter entitled “The Publisher’s Peritext.” However, Andrist, “Toward a Definition,” 131, refers to the difference between authorial and publisher paratext when Genette uses the term peritext. 13 Andrist, “Toward a Definition,” 131, summarizing Genette, Paratexts, 17–36, 406.

16

Porter

period, although he does not elucidate on how his theory would apply to this earlier period. There are several abiding issues raised by Genette’s attempt to define paratext. The first is that Genette seems to assume the notion of a text, against which he defines the notions of paratext, epitext, and peritext. He seems to do so on the basis of the modern published book as his example of text. As we shall discuss below, the notion of text is anything but an easy assumption to make. A second is that Genette appears to be closely indebted to the notion of an author in defining and utilizing the notion of paratext, with authorial control central to his definition. Besides the notion of what is entailed in this definition of control and authority, this raises a number of questions in relation to ancient manuscripts. Authorship is a complex concept, even more so for ancient manuscripts. Are we referring to the originating author who dictated or wrote or supervised the “publication” of the text, or are we referring to a community activity that was responsible for generating a text, or, as we often think, are we referring to the scribe responsible for the transcribing of the manuscript that we are studying, or are we using the concept of author in another sense? These various definitions have implications for how we define paratext in relation to text and authorial involvement. If we use a definition of author as textual originator, then any subsequent rendition of the manuscript is paratext, which either makes the concept vacuous or forces us to define it more carefully in relationship to an author.14 A third issue is the relationship between formal and functional features of manuscripts. Most discussions of paratext seem to involve an identification of formal features of manuscripts, including the various lists that are discussed by Genette, encompassing such things as prefaces, titles, subtitles and the like, with publisher attributed items even more of the same (e.g. cover, typesetting, etc.), whereas there are others who contend that paratext should be a functional distinction. For example, Andrist, after surveying various views of text, approaches the subject on the basis of content, providing a second form of definition. He observes that the concept of “text” is used “to describe a piece of content in a manuscript.”15 After noting that even paratext, by definition, implies text, he reinterprets Genette’s analysis in terms of all of the features mentioned as providing content for the book and concludes that “a paratext represents a piece of content which 14 Andrist, “Toward a Definition,” 140, ends up positing that those who produce a manuscript are surrogates for the author. Whether this is an acceptable definition or not is subject to question, although many who use the term paratext in relation to ancient manuscripts seem to accept this definition, whether explicitly or consciously or otherwise. 15 Andrist, “Toward a Definition,” 137.

What Is Paratext?

17

distinguishes itself from other pieces of content on the basis of its subordinate position in the greater scheme of the overarching book project.”16 However, his own examples do not address the major questions that have been raised above, and his own analysis, despite his definition (which I will assess more fully below), is less content based than it is formally based. The question of what constitutes paratext remains largely unresolved, or at least largely unsatisfactorily analyzed. In what remains of this chapter, I will examine various definitions of text, because in both formal and functional analyses of paratext the definition of text, whether implicit or explicit, seems to stand behind what is meant by paratext. Once I have examined various definitions of text, I will then turn to a reconsideration of paratext. 3

Defining Text

As stated in the above section, the definition of text is one that is central to definition of paratext. It is very difficult to identify a paratext if one does not have an understanding of what is text, since paratext is dependent, by definition, on text. Much of the problem stems from the fact that prior to the late 1960s or the early 1970s, especially with the advent of post-structuralism (see below for further discussion), the concept of text had been a relatively consistent one, even if this concept was usually an assumed and pre-critical understanding. Andrist recognizes the problematic nature and offers several definitions of texts,17 but he uses them primarily to conclude that paratext and text have in common their being varying degrees of presentation of content. There is much more to be said about the concept of text – in fact the concept is one that is itself 16 Andrist, “Toward a Definition,” 137 (essentially repeated on 146). One cannot help but noticing that, in the end, Andrist’s identification of paratext still seems to resemble a formal set of features, such as titles, added statements, marginalia, which are credited to the producer of the book (141–143). He shies away from authorial paratext (e.g., Luke 1:1–4), leaving it to what he calls “text historians,” a term that leaves plenty of ambiguity. He also identifies paratext added after production of the manuscript, including additional marginalia, notes of ownership, and the like (144–145). See the figure on 147 for a summary of these features. Cf. Garrick V. Allen and Anthony P. Royle, “Paratexts Seeking Understanding: Manuscripts and Aesthetic Cognitivism,” Religions 11.523 (2020): 1–25, here 3: “every part of a manuscript beyond the main text itself, especially features that have aesthetic qualities insofar as they negotiate a text to readers.” They tend to follow Andrist. 17 Andrist, “Toward a Definition,” 135–136, where he identifies (1) “any meaningful sequence of words in a manuscript” (135), (2) any manuscript content including pictures, (3) philological text, and (4) what he calls “a book-historical usage of the word ‘text’” (136).

18

Porter

highly problematic. The attempt to clarify the notion is one of the reasons that I have returned to this topic on several occasions – but I do so also because I do not believe that we in biblical studies have come to terms with the notion of text, and so continued discussion is warranted until we do arrive at further understanding. The philosopher Jorge Gracia has explored the notion of text more than most. He arrives at the following definition as a useful starting point: “texts are semantically significant artifacts in relation to which authors, audiences, and contexts play interesting and idiosyncratic roles.”18 This is a useful definition, not least because it at least makes clear what at least one person thinks a text is. Each of the elements of this definition merits elucidation. Even if Gracia does not here specify his theory of meaning or the standard by which he establishes significance (the rest of his book is dedicated to such questions), he notes that texts are “semantically significant.” The implication of semantic significance, however, may involve content, but it may also involve much more than content, as semantics or meaning in language goes far beyond simply content to include a variety of functions that a text may perform.19 The text is equated with being an artifact, by which Gracia seems to indicate that a text is realized in or by a particular instance of semantic significance. The way that texts as artifacts relate to their environment is seen by Gracia as complex. He recognizes that this relationship is “interesting” and “idiosyncratic,” apparently because a text plays a variety of “roles” in this set of relationships. In other words, a text does not have a singular function and it does not relate to its environment in altogether predictable ways, but in ways that are individualized on the basis of the text being a text, and as a result these are interesting and worth discussing further. Gracia identifies three situational variables related to texts, authors, audiences, and contexts. We often equate texts with authors, as has been noted above, even if such identification is problematic because the notion of authorship – especially in ancient texts – cannot always be equated with a person. The inclusion of reference to audiences and contexts adds a, perhaps surprising, set of dimensions to consideration of what constitutes a text, as it encompasses not just those who are thought to be originators and propounders of texts but also those who respond to texts and the environments in which the text and they respond to it. Gracia elsewhere considers 18 Jorge J. E. Gracia, A Theory of Textuality: The Logic and Epistemology (Albany: State University of New York Pres, 1995), xv. Cf. also Ronald L. Troxel, “What is the ‘Text’ in Textual Criticism?” VT 66 (2016): 603–626. 19 Gracia, for example, examines texts in light of their linguistic functions, in which he draws upon speech-act theory to identify texts as informative, directive, expressive, evaluative, and performative (Theory of Textuality, 87–89).

What Is Paratext?

19

other factors influencing understanding of texts (e.g. society, language, etc.), but his definition here focuses upon these factors as having a role to play in what a text is. In the end, Gracia argues for a communal and functional view of texts, in which one requires the signs of the text but also those who interpret the signs in a meaningful context. Gracia’s definition – which begs for further elucidation that he attempts to provide in the rest of his volume – is a useful starting point for consideration of a text, especially as we approach the question of paratext in relationship to it.20 The above definition, as helpful as it is, only goes so far in helping us to understand text, even if we can appreciate a number of identifiable features that constitute a text. As a result, there have been various ways in which these artifacts, texts, have been defined. Gracia, for example, identifies at least twelve texts on the basis of their cultural functions.21 This is a useful list so far as contemporary culture is concerned. However, another way of identifying texts – and the fact that they can be identified in a multitude of ways is itself commensurate with Gracia’s definition above – is according to how they function within various academic or intellectual disciplines. There are at least seven conceptions of text that merit examination here. These are: ontological, phenomenological, philological, literary, linguistic, semiological, and text-critical texts.22 Each of these merits a brief explanation. I begin with what might at first appear to be the most abstract notion and proceed to the one closest to the concerns of this paper and the volume in which it appears. The ontological text is the most abstract way of describing a text and is associated with a range of views of text as a representation of being and thereby serving as

20

So far as I can determine, Gracia does not address the question of paratext, which may not be surprising in light of the date of publication of his work in relationship to the work of Genette, who is not mentioned in his bibliography. 21 Gracia, Theory of Textuality, 89–96, including legal, literary, philosophical, scientific, religious, historical, political, pedagogical, confessional, entertaining, inspirational, and pneumonic texts, a non-exhaustive list. 22 The basis of these categories is found in John Mowitt, Text: The Genealogy of an Antidisciplinary Object (Durham, NC: Duke University Press, 1992), esp. 223–224, but also 2–10 and passim, who provides some of the references to secondary literature cited below as examples. I offer more detail on these types of texts in Stanley E. Porter, “Text as Artifact: An Introduction,” in Scribes and Their Remains, ed. Craig A. Evans and Jeremiah J. Johnston, LSTS 94 (London: T&T Clark, 2020), 1–14 esp. 3–6, upon which I draw in the subsequent discussion, although I refine and revise my understanding of text, especially in light of the purpose of this paper. I encourage others to engage the discussion as well, rather than simply dismissing it.

20

Porter

a representation of logical, epistemological, or metaphysical properties or characteristics.23 The notion of being is defined differently depending upon one’s philosophical orientation. In that sense, an ontological text (to be distinguished from a philosophical text as Gracia defines it) represents a view of reality, how it is constituted, how one knows it, and how it is logically organized. The ontological text may at first glance appear to be too abstract and without a significant place in contemporary thought, especially textual studies, but the resilience of the ontological text is surprisingly high. It is possible that we are dealing with an ontological text, or at least a text that is being viewed as ontological, when we may respond by saying that a text reflects a particular view of reality or is illogical or is incoherent.24 Various philosophical and theological positions make use of the concept of the ontological text. The phenomenological text is a text that maintains characteristics of both an ontological text, or is seen as a sub-category of ontological text, and a literary text.25 From the ontological side, the phenomenological text is concerned with experience of the thing itself, emphasizing the relationship of the experiencer of the phenomenon to the text. From the literary side, the phenomenological text is reflected in various literary movements, such as Russian Formalism (itself related to phenomenology), Prague School structuralism (heavily influenced by Russian Formalism), the Anglo-focused New Criticism, and forms of reception in which text as phenomenon is to be experienced by textual participants. The focus is upon the phenomenon of text, with the reader responding diachronically in terms of the history of the text’s reception or synchronically in productively responding to the text to complete its meaning. The phenomenological text, especially in its relationship to the literary text, continues to influence biblical studies through various forms of literary study such as narrative criticism or even reader-oriented criticism. The philological text has many similarities to the text-critical text that I will be discussing last in this list, because traditional philology is primarily concerned with establishing ancient texts for study and especially for appreciation, and in that sense is an idealization of what is valued in classical studies.26 Classical studies has traditionally based itself upon this concept 23 Cf. Gracia, Theory of Textuality, xxii–xxiv, who defends the epistemological and metaphysical textual features. 24 Cf. Gracia, Theory of Textuality, xx–xxi. 25 See, e.g., Wolfgang Iser, The Implied Reader: Patterns of Communication in Prose Fiction from Bunyan to Beckett (Baltimore: Johns Hopkins University Press, 1974); and Paul Ricoeur, Hermeneutics and the Human Sciences, ed. and trans. John B. Thompson (Cambridge: Cambridge University Press, 1981). Cf. Mowitt, Text, 224, who offers these as examples. 26 Cf. Mowitt, Text, 2–6 and 223–224.

What Is Paratext?

21

of text, with the study of its texts in this way described as classical philology. The philological text, while often making use of textual criticism (although not always, especially when authors are treated in translation), is a more narrowly construed concept of text than in textual criticism, because it is concerned with the preservation and presentation of the best representatives of the literature of those who produced such texts. Such texts are studied for a variety of reasons, but the focus is often upon their artistic or literary merits, appreciation of their language, and even their exemplification of various generic qualities. Culture is increasingly being brought into consideration of philological texts, but the study of philological texts has not emphasized the importance of defining culture apart from how it is represented in the text. The field of biblical studies has been greatly influenced by the concept of the philological text, with the commentary tradition sharing with classical studies the notion of a philological text. The literary text is very closely related to the philological text27 and in many ways may be seen as the natural descendant of the philological text in its representation of various literary works that go beyond the scope of the philological text. If the philological text focuses upon ancient idealizations, the literary text encompasses those texts and others since, although in similar idealized form. The notion of a literary text has undergone a significant shift over the course of the last several centuries, to the point where the literary text is often conceived of as a semiological text (see the discussion below).28 The literary text was conceived of as an authorial text up to the beginning of the twentieth century, when the literary text was transformed into a text very similar to the phenomenological text with emphasis upon the experience of the text itself,29 and then in the latter part of the twentieth century the focus shifted to various reader- or audience-oriented perspectives on text (in the extreme forms resulting in Derridian-inspired deconstruction). As a result of shifts in the perceived nature of the literary text, there were changes in the valued features. Those characteristics associated with authors included such things as veracity, 27

Both of them may be seen in relation to other concepts of text, such as rhetorical texts or stylistic texts. For the sake of this discussion, these two may be subsumed under the notion of literary text, or possibly philological text. See Mowitt, Text, 4. 28 See the discussion in Mario J. Valdés and Owen Miller, eds., Identity of the Literary Text (Toronto: University of Toronto Press, 1985), with essays on intertextuality, deconstruction, and other issues. 29 In many ways the image from the literary critic Cleanth Brooks’s The Well Wrought Urn: Studies in the Structure of Poetry (San Diego: Harcourt Brace Jovanovich, 1974 [1947]), captures the image of what is represented by the literary text throughout a good portion of the last century.

22

Porter

realism, and certain conceptions of form, structure, and content. The shift to the text itself resulted in a move away from authorial values, with emphasis upon internal characteristics, including irony but also different views of form and structure. The conceptualization of the literary text as reader-oriented has resulted in further changes in the notion of text, with characteristics attributed to readers and their communities as giving value and significance to the text. In biblical studies, the literary text has in some ways supplanted the philological text in its recent influence upon the discipline, even if its views of the importance of the reader have not gone as far as have those of others who have utilized the notion of the literary text. The linguistic text is related to the philological and the literary text, by means of their common association with the use of language.30 However, the linguistic text, from its inception, has also had a more explicitly metaphorical view of text, with priority in linguistics being given to spoken over written language, whereas both philological and literary texts have predominantly emphasized the written text as the emphasized medium of transmission. The distinguishing characteristic of a linguistic text is that it is an instance of another entity, a linguistic code. If philological, literary, and other texts place their emphasis upon the concept of the written text, the linguistic text is seen as a representation of choices within the system of language that is represented. The system of choices represents the meaning potential of the language that is realized in text. The text is but an instance, an example, that represents the linguistic code with its potential fulfilled. Linguists have various ways of characterizing the way in which the linguistic code is realized in text and speak of this relationship with terms such as generatively, structurally, functionally, cognitively, and otherwise. The structural conception of text, in which the text is seen as a representation of fundamental structures in relationship, has had the greatest amount of influence upon the concept of the linguistic text so far. The linguistic text has not been nearly as important in biblical studies as have some other conceptions of text, such as the philological and literary text. The semiological text is closely related to the linguistic text, but arguably identifying a broader conception of text that encompasses language and other semiotic (or sign) systems within a culture.31 The semiological text has developed over the years. Its earliest conceptualization occurred with the advent of modern linguistics, in which language was seen to constitute a system of signs 30 31

See Mowitt, Text, 8, but also 2–10. See Mowitt, Text, 6–7, 10, but whose entire volume is dedicated to discussing the semiological text. The language and concepts of semiology are deeply embedded within the language of post-structuralism.

What Is Paratext?

23

that maintain arbitrary relationships between the sign and what they signified. Ferdinand de Saussure, often credited with being the originator of modern linguistics, was primarily concerned with semiology or semiotics.32 However, the notion of the semiological text has undergone a significant reconceptualization in light of post-structuralism. Post-structuralism, although variously valuated, is a movement that took root in the 1970s as a reaction to but also an expansion upon structuralism, most clearly reflected in the writings of Jacques Derrida, along with others.33 Rejecting the ontological textual characteristics that some semiological texts assumed, post-structuralism further dissolved the relationship between the sign and its signification to the point of advocating freedom and play within sign systems and hence textual deconstruction.34 The result was semiological texts as socially and culturally embedded and unstable semiotic systems, with shifting relationships to society, culture, author, reader, and even language itself. The concept of intertextuality – in which language becomes a pastiche of other language – has become an important concept in understanding the semiological text.35 The semiological text has only recently become important in biblical studies as more and more interpreters embrace the post-modern and post-structural agenda. The so-called New Philology, which is textual criticism in the light of post-structuralism, has in many instances embraced the semiological text.36 32 See Ferdinand de Saussure, Course in General Linguistics, ed. Charles Bally and Albert Sechehaye with Albert Riedlinger, trans. Wade Baskin (New York: McGraw-Hill, 1959 [1916]). Cf. Alexandros Ph. Lagopoulos, “A Meta-theoretical Approach to the History and Theory of Semiotics,” Semiotica 213 (2016): 1–42, esp. 6–9. 33 See Jacques Derrida, Writing and Difference, trans. Alan Bass (Chicago: University of Chicago Press, 1978), which includes Derrida’s well-known paper, “Structure, Sign, and Play in the Discourse of the Human Sciences” (278–294), delivered at a conference at Johns Hopkins University in 1966 that many see as beginning the post-structuralist movement. 34 For recognition of Derrida as in some ways the successor to Saussure, see Samuel Weber, “The Future of Saussure,” Semiotica 217 (2017): 9–12, esp. 11, with his analogy of the change from a French-English to a French-French dictionary illustrating the move from arbitrary sign to endless signification. 35 The concept of intertextuality (outside of most New Testament and other biblical studies) involves the ways in which “any text is constructed as a mosaic of quotations; any text is the absorption and transformation of another.” The origination of the term, in response to Mikhail Bakhtin, was first developed by Julia Kristeva. See Kristeva, “Word, Dialogue and Novel,” repr. in Toril Moi, ed., The Kristeva Reader (New York: Columbia University Press, 1986), 34–61, citing 37 in the quotation above. 36 The movement was launched by the 1990 issue of Speculum. See Stephen G. Nichols, “Introduction: Philology in a Manuscript Culture,” Speculum 65 (1990): 1–10. See, e.g., Matthew J. Driscoll, “The Words on the Page: Thoughts on Philology, Old and New,” in Creating the Medieval Saga: Versions, Variability and Editorial Interpretations of Old Norse

24

Porter

The final text is the text-critical text.37 The text-critical text has some characteristics of the philological text (or perhaps better, the philological text has elements of the text-critical text), but the text-critical text has an ambivalent relationship to the actual and the ideal. On the one hand, the text-critical text places high emphasis upon an actual, physical text, since textual criticism consists of comparing not just texts or passages or meanings but specific instances of texts for their similarities and differences. The texts of textual criticism are recognizably finite and admittedly flawed, and because of this feature they are suitable for comparison at the most minute levels for their similarities and differences. On the other hand, this comparison reveals that the text-critical text uses the imperfections of its physical text as a point of comparison with and as a means of establishing an admittedly elusive earlier text (whether that text is conceived of as the original text, Urtext, Ausgangstext, or something else). In that sense a text-critical text is a representation, in a flawed form as it is realized in the text, of the source text.38 This text-critical text has been, at least notionally, challenged by the semiological text of the New Philology. However, from what I have seen in much New Testament textual criticism, the New Philology and its conception of text still resides on the distinct periphery of New Testament text-critical discussion. The traditional text-critical text may have shifted in some ways from its persistent goal of attempting to reconstruct the original text, however that is defined, to a more modest goal Saga Literature, ed. J. Quinn and E. Lethbridge (Odense: University Press of Southern Denmark, 2010), 87–104, who identifies some of the characteristics of the New or Material Philology in response to post-structuralism: it “de-emphasised the importance of the author, focusing instead on the inevitably collaborative nature of literary production, dissemination and reception and the cultural, historical and ideological forces at work in these processes” (93). Contra Allen and Royle, “Paratexts Seeking Understanding,” who appear to embrace the New Philology but seem to endorse the traditional text-critical text. They are not alone, from what I can see, in not recognizing the full impact of the New Philology on textual criticism. A casual perusal of Bart D. Ehrman and Michael W. Holmes, eds., The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, 2nd ed., NTTSD 42 (Leiden: Brill, 2013), only shows possible hints but no references (in the index to Derrida, Nichols, Driscoll, the New or Material Philology, or post-structuralism). This is not necessarily true of non-biblical textual criticism. For example, the textual scholar D. C. Greetham, in Textual Scholarship: An Introduction (New York: Garland, 1994), 295–346, already in the 1990s saw the New Philology as “symptomatic of a general shift in critical theory from a reliance on an author’s imputed meaning to the free play of meaning associated with post-structuralism” (341). 37 Cf. Mowitt, Text, 223–224. 38 See Troxel, “What is the ‘Text,’” 619–620, who uses the language of type and tokens, from Ron Hendel, “What is a Biblical Book?” in From Author to Copyist: Composition, Redaction, Transmission of the Hebrew Bible, ed. C. Werman (Winona Lake, IN: Eisenbrauns, 2015), 260.

What Is Paratext?

25

of reconstructing the earliest recoverable text as advocated by narrative or sociohistorical textual criticism, but the goal remains relatively stable and the means of doing so remain much the same.39 There are no doubt other conceptions of text that might be suggested. However, these concepts of text, all of them pertinent to the discussion of text within New Testament studies and especially in relationship to textual criticism, offer sufficient insight to provide a basis for discussion of paratext. However, this reveals some of the problems. One of these problems is the lack of ability to define a text. If by text we are looking for a singular definition, and certainly an essentialist definition (of immutable essential characteristics), then we are to be seriously disappointed, for there is not a standard definition of text. The concept of text varies, both with reference to various disciplines and within given disciplines. As a result, even though there is some significant overlap among views of text (e.g., ontological and phenomenological text, phenomenological and literary text, philological and literary text, literary and semiological text, linguistic and semiological text, philological and text-critical text, etc.), the differentiation of notions of text in areas of intellectual and academic interest illustrates – as do their descriptions – that there are varied views of what constitutes a text. Further, within each of these disciplines, the conceptions of text are not static. I have not explored individual and even idiosyncratic views of text within these disciplines but have attempted to capture some of the broad sweeps that are represented within these conceptions. The second problem is that, if we are pressed to find what each of these definitions does have in common, it is that text has become only a tentative notion. By that, I mean that in each definition, the physical artifact that we may identify as a text is not tantamount to what is meant by text within any given discipline. For each of the disciplines discussed above, the notion of text is as a representation of something else. I do not deny that texts as physical objects exist. They no doubt do. However, when one refers to a text, if one persists in using such terminology (and the above discussion makes one wonder), one must mean more than simply the physical object that one has in one’s hand. Once one begins to describe and attempts to define that object, one realizes that the notion of text becomes much more difficult to capture, to the point that whatever it is that is represented on the physical artifact is always pointing 39 See Stanley E. Porter and Andrew W. Pitts, Fundamentals of New Testament Textual Criticism (Grand Rapids: Eerdmans, 2015), 1–6, for a discussion of the goals of textual criticism and their proponents. Cf. Eldon Jay Epp, “It’s All about Variants: A Variant-Conscious Approach to New Testament Textual Criticism,” HTR 100.3 (2007): 275–308, who, taking a narrative approach, reduces textual criticism to variants; Epp, “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism,” HTR 92.3 (1999): 245–281.

26

Porter

to something else, something that may not even be text. That something may be a logical or metaphysical or epistemological conception, an experience of something, an idealized view of what constitutes literary beauty, an authorial or structural or readerly process and product, a realization of a linguistic code, an unstable set of signs, or, even in the case of the text-critical text, a flawed representation that points to a representative or idealized earlier conceptual entity, even if graphically represented. 4

Defining Paratext

In light of these considerations regarding text, are we in a better position to define paratext? I am not sure. Paratext has been defined according to physical features in the definition of Genette but content in the definition of Andrist. In other words, paratext has been defined according to form or function. Each is problematic for a number of reasons. One of these reasons is the difficulty in defining either. Even Genette apparently had difficulty in defining the terms. If we were strictly to follow Genette, we in New Testament textual criticism would probably call whatever it is that we are discussing “epitext.” Nevertheless, whatever we call it, epitext or paratext, the concept is still problematic. Genette considers everything but the text (which he attributes to the author) as paratext, including epitext and peritext, even if he later differentiates authorial and publisher epitext. However, we must note that Genette is concerned with a published work, not a handwritten manuscript. In other words, paratext encompasses everything but the authorial text. This is highly problematic, even for a published book, as the published text, as a literary text, is itself merely a representation of the authorial text; that is, it is a printed version of the text written by the author. In that regard, every published book, and by implication every copied manuscript, is entirely paratext or epitext. Nevertheless, we may concede for the sake of argument that the printed text includes the author’s text, or in the case of a written manuscript that the scribe is to be equated with the author. Some may accept this equivalence of scribe and author, as it attributes agency to the scribe and potential responsibility for some dimensions of the manuscript. However, this is not entirely satisfactory, because few will probably wish to argue that the scribe is the author. The scribe may be responsible for some variations in a particular manuscript (the variations being the paratext), but the scribe cannot be the same as the author (however this is defined so far as originating author is concerned) unless the scribe is actually the author by copying one’s own work. The major issue with scribes and ancient manuscripts in text-critical texts is

What Is Paratext?

27

concern with how they unintentionally or intentionally depart from the text that they are copying, from their archetype. The assessment of their work in relationship to their Urtext indicates both their role as scribe and their role as not being the author but something else, whatever role we wish to attribute to such a process of copying. So equating the scribe with the author of the text does not seem to be valid, apart from the expedience of finding a way to introduce an authorial presence into the manuscript as manuscript. Genette himself differentiates between authorial and publisher paratext or epitext, as a means of differentiating those features with direct authorial control and those without. This seems to be a way of extending the notion of authorial text into the published work although it does so by equating the extension not with the authorial text but the response to the text. However, this too is a problematic distinction. What if the author self-published the work? Is all of the work text or is all of it authorial epitext? The differentiation of two categories of epitext, authorial and publisher, adds to the complications. Genette relies upon the notion of direct and indirect authorial responsibility. As a result, Genette leaves the author in charge of the title, subtitle, and preface, while the publisher is responsible for the cover and jacket design, etc. However, even this does not solve the problem. Authors are not necessarily responsible for their titles, and sometimes perhaps not even the subtitles, at least in modern publishing conventions. There are also a whole number of features that Genette does not address, such as ink, paper, font, margins, etc. These are all the responsibility of the publisher. Once the publisher epitext is removed, what exactly remains of the text? Even if the text notionally exists in the published work, or in a copied manuscript for that matter, once the ink and paper and font and page dimensions are removed, all that is left is an idealization of a text, something not to be found within the book or manuscript. One is forced to conclude that, however one defines paratext in relationship to physical features, the concept is problematic. If one recognizes that the text within such a book or manuscript is the representation of a text and only an idealization, then all else is paratext. The result may be pragmatically useful for those wishing to focus upon paratext, but it hardly answers the question adequately regarding text and paratext. The other approach to paratext is to equate it with content or function. There are hints of this viewpoint in Genette himself, although his own enumeration of features tends to be formal. However, the shift to function is itself problematic. In this conception of paratext, the differentiation is not around physical features of the manuscript as additions to a text (those things located around the text), but text is defined as a semantic notion and in that respect the notion of a linguistic or semiological text seems to be used. As Andrist states,

28

Porter

the term text is used “to describe a piece of content in a manuscript.”40 This seems to indicate that text is a quantum of semiological substance located and identifiable in a manuscript. There is some merit in this definition, as a means of overcoming the issue of the disappearance of the text when the formal definition is accepted. Whereas the formal features of the text may be gone, the semantic features may be present and represented within the published book or manuscript. This is no doubt helpful in enabling us to understand text as a semantic concept represented by means of language, or possibly even some other meaningful system of signs or indicators. However, the definition is not without its own problems. The first problem is one of delimitation and the other of quantification. It is entirely unclear what it means to say that the text is a “piece” of something else and what it means to have or be content. There is nothing in Andrist’s definition to indicate what is meant by content or what its relationship is to anything else. I am not sure how one can even use such a definition to talk about text and paratext. Would a single letter in a manuscript be a text? It is certainly a piece of content. Or is it an entire manuscript with all of its features as text, a semiological text in a full and complete sense? This too is certainly a piece of content, one that is equated with the entire manuscript (one piece that constitutes the whole). According to this semantic definition, however, paratext is also a semantic notion, but one that is based upon the definition of text – an admittedly difficult concept as we have seen. Andrist defines paratext as representing “a piece of content which distinguishes itself from other pieces of content on the basis of its subordinate position in the greater scheme of the overarching book project.”41 There are several observations to be made here, each of them problematic within the definition. The first is that the same definition is used of paratext as of text, that is, it is “a piece of content.” A second problem is that a necessary distinction must be made between these pieces of content, as they cannot be the same content (although they are defined that way). A third problem is that their relationship is apparently decided on the basis of hierarchical ordering. And a fourth problem is that there is a conception of the overarching nature of the individual project, in this case a book or manuscript. Such a definition must die the death of a thousand qualifications and redefinitions. Who decides the overall purpose of a book project? This probably depends upon how one is defining text. Is a given manuscript meant to be a text-critical text, a literary text, a linguistic text, a philological text, or what? More specifically, is this text a repository of textual variants, a theological 40 Andrist, “Toward a Definition,” 137. 41 Andrist, “Toward a Definition,” 137; cf. 14.

What Is Paratext?

29

document, an ecclesial record, a historical record of the development of the text, an aesthetic artifact, or what exactly? The developing Christian church used its manuscripts – including its non-canonical documents – in a variety of ways that indicated its cultural literacy, its literary creativity (e.g., the so-called apocryphal gospels, especially the early fragmentary ones found on papyrus), its theological understanding, its theological interpretation, and the nature of its worship – all of which are purposes and priorities reflected in various ways by the artifacts of early Christianity.42 In this scheme, text and paratext are constantly shifting depending on purpose. Further, who decides what is superior and what is subordinate? Is the use of a nomen sacrum subordinate or more significant in importance to the resolved form? It is arguable, especially in light of the fervent discussion of nomina sacra in recent scholarship, that the nomina sacra are more important and indicate the greatest amount of content and insight regarding a given artifact.43 Is an illuminated manuscript highlighting the illuminations or the text? For many ancient, illuminated manuscripts, the illuminations are arguably more revealing than the texts that they contain, as we have these texts in a variety of other forms. Can the ekphonetic notation of manuscripts be meaningful content apart from the individual words to which the neumes are joined, and do the individual words comprise a liturgical text in the same way without the ekphonetic notation?44 Furthermore, how does 42

43 44

See some similar questions raised in Stanley E. Porter, “What Do We Know and How Do We Know It? Reconstructing Early Christianity from its Manuscripts,” in Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament, ed. Stanley E. Porter and Andrew W. Pitts, ECHC 1, TENTS 9 (Leiden: Brill, 2013), 41–70. That theology is more and more recognized to be involved in textual criticism, see Daniel B. Wallace, “Challenges in New Testament Textual Criticism for the Twenty-First Century,” JETS 52.1 (2009): 79–100, esp. 92–96; David Parker, “Textual Criticism and Theology,” ExpTim 118.12 (2007): 583–589; and Nathan C. Johnson, “Living, Active, Elusive: Toward a Theology of Textual Criticism,” JRT 12 (2018): 83–102. See Porter, “What Do We Know,” 64–66, where pertinent recent literature is noted. An excellent instance of this problem is found in the ninth- or tenth-century manuscript Australian National Library Suppl. Gr. 121 (0105). This parchment of John 6–7 has portions of four pericopes. The first, Jesus going to the Feast of Tabernacles, and the second, Jesus teaching during that Feast, have ekphonetic notation. The third, discussion about the identity of Jesus, does not have ekphonetic notation, and the fourth, Jesus’s statement on thirst and living water, has ekphonetic notation. If one attempts to examine the four episodes as simply a repository of variants (the late manuscript has very few), then one misses perhaps the most significant feature of the manuscript, the notation of three but not all of the pericopes. See Stanley E. Porter and Wendy J. Porter, New Testament Greek Papyri and Parchments: New Editions: Texts, MPER 29 (Berlin: De Gruyter, 2008), no. 40 (162–186); and further description in W. Porter, “Sacred Music at the Turn of the Millennia,” in Faith in the Millennium, ed. Stanley E. Porter, Michael A. Hayes, and David Tombs, RILP 7 (Sheffield: Sheffield Academic, 2001), 423–444, esp. 432–438.

30

Porter

one distinguish one piece of content from another? What are the boundaries of these pieces and how does one define content in relationship to them? Are they discrete pieces of content or are they overlapping pieces of content? How can they be discrete pieces of content if they are part of the same artifact, since the definition of paratext places them in interdependent relationship? The numerous hermeneia Johannine manuscripts well illustrate how various parts of the content are dependent upon each other to give the manuscript its significance.45 And if these parts of content are overlapping, how then can one distinguish one from the other, as two separate categories? Finally, since the same definition of text and paratext is used, how does this actually distinguish them from each other? One might well conclude, on the basis of the functional definition, that all is text, as both text and paratext are a piece of content joined by virtue of their being constituents of the same manuscript. In other words, we cannot adequately define text and paratext in any meaningful way, because the two concepts are so related to each other that they are arguably inseparable. As a result, I wonder if we should use the term paratext (or epitext, if we wish to be true to Genette) at all, as it seems to carry unsupportable connotations that cannot be relied upon for clarity in our analysis. But if we must use the term, as we do in this volume, whether because of tradition or because of the need for some kind of a term, how then should we think of using the notion of paratext in relation to text, especially with regard to New Testament textual criticism? I think that if the notion of paratext is to be used in relation to text (assuming we use the term text), paratext must be clearly defined within each context in which it is used. In other words, the categories cannot simply be assumed or treated as self-evident but the opposition of text and paratext must be flexible depending upon the artifact in which it is being described. This may seem to be self-evident according to the practice of many if not most textual critics but bears repeating nevertheless. If by text we mean the part of content that we wish to focus upon and in relation to which 45 See Stanley E. Porter, “The Use of Hermeneia and Johannine Papyrus Manuscripts,” in Akten des 23. Internationalen Papyrologenkongresses, Wien, 22.–28. Juli 2001, ed. Bernhard Palme (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2007), 573–580, esp. 579–580; cf. Porter, “Reconstructing Early Christianity,” 60–63. Cf. Wally V. Cirafesi, “The Bilingual Character and Liturgical Function of ‘Hermeneia’ in Johannine Papyrus Manuscripts: A New Proposal,” NovT 56.1 (2014): 45–67, who posits a bilingual environment for their use. This may describe the use of some of the manuscripts, but it probably does not describe the function of all of them, especially those written only in Greek or only in Coptic. See Brice C. Jones, “A Coptic Fragment of the Gospel of John with Hermeneiai (P.CtYBR inv. 4641),” NTS 60.2 (2014): 202–214. The use of hermeneia is more likely an interpretive device that happened to be used in both mono- and multi-lingual textual environments.

What Is Paratext?

31

the other part, however this is defined, is subordinate (if such distinctions can be made and defended), then we must be open to various ways in which this opposition may be framed depending upon the artifact itself. A useful example is a manuscript like the so-called Purple Codex (Codex Purpureus, N or 022).46 This sixth-century manuscript of the Old and New Testaments (incomplete) contains a beautifully prepared purple parchment on which are written silver letters, with nomina sacra in gold. The major difficulty in differentiating text and paratext for this manuscript is that if we take either a formal or functional view, we end up speaking about the same thing. Since the lettering of Luke’s Gospel is written in the silver and gold ink, including the nomina sacra, it is impossible to separate the two and retain each as a physical entity. In such an instance, the text is only an idealized form, as the realized form is in silver and gold. If we focus on content, we probably have an even more difficult situation. We must begin with the purpose of our investigation. Is the object of investigation the biblical text represented in the Purple Codex or is it the beautiful purple parchment and its incredible use of silver and gold lettering? Even if it is the silver and gold lettering, then, at least according to traditional criteria, we have a manuscript that is tantamount to and without remainder paratext. 5 Conclusion This chapter has not simply attempted to define paratext but has identified some of the difficulties in defining it. Although there are surprisingly few definitions of paratext, perhaps because of the term being of recent vintage, there appears to be a tacit, assumed, or implied understanding of it. That, however, reveals the problem. As with so many topics within intellectual investigation, assumptions often belie a failure to come to terms with important technical details, such as those of definition. So, instead of defining the term paratext, I have attempted to problematize the term. In order to do so, however, I have had to begin with an earlier theoretical notion, that of text. The notion of text – which appears to provide the foundation for definition of paratext – is itself a problematic concept that fails to provide the kind of essentialist or firm foundation required for many of the notions popularly associated with paratext. We may even find the notion of text an unusable one. Once we investigate further, we not only call text into question, but we also call paratext into question, with the result that it is pragmatically exceptionally difficult to define 46

See Porter and Porter, New Testament Greek Papyri, no. 37 (149–157), where the portion of the manuscript located in Vienna that I have examined is described in detail.

32

Porter

or differentiate the two. In that regard, I realize that my conclusion reflects (although perhaps even more radically) one of the major tenets of the New Philology, that “literary works do not exist independently of their material embodiments,” thus rendering important (even if not altogether clear) “the relationships between the text and such features as form and layout, illumination, rubrics, and other paratextual features, and, not least, the surrounding texts”; textual artifacts are “consumed in ways which are also socially, economically and intellectually [and theologically] determined.”47 The most important question regarding paratext appears to be, not what is paratext, but paratext for whom and in what context? As I stated above, if the notion of paratext is to be used in relation to text, paratext must be clearly defined within each context in which it is used, since the categories cannot simply be assumed or treated as self-evident but the opposition of text and paratext must be flexible depending upon the artifact in which it is being described. If we accept a definition along these lines, without attempting to establish absolutist values for it, there is abundant scope for discussion of both text and paratext and probably much more besides. We may not agree on exactly what we are observing, but there will nevertheless be plenty of room for insightful and instructive comment. 47 Driscoll, “Words on the Page,” 90, 91. Parenthetical comment my addition.

CHAPTER 2

Missing the Point: Modern Punctuation Practice as Authoritative but Possibly Problematic Decision-Making Hans Förster 1 Introduction1 The current standard edition of the Greek New Testament – the 28th edition of the Nestle-Aland – uses punctuation. These are editorial decisions that have a long and venerable tradition. Already the editio princeps of the Greek New Testament – the novum instrumentum omne of Erasmus (published by Johann Froben in Basel in 1516)2 – used punctuation to guide its readers.3 The edition used punctuation differently than ancient manuscripts did. Many ancient Greek manuscripts of the New Testament do not have punctuation; others use it seemingly random. There are, however, manuscripts having some kind of logic to their punctuation, such as the punctuation in Latin manuscripts that becomes more pronounced in the eighth century. On first glance, it appears to be of minor importance whether some small signs are added to a text in order to guide the reader and make it easier to understand what is written. The practice of thorough punctuation is certainly not as important as other elements. By analogy, today it is usually the task of 1 This chapter arose from a research project funded by the Austrian Science Fund (FWF-project P28821). I would like to thank the anonymous reviewer for his helpful suggestions and comments. 2 J. K. Elliott, “‘Novum Testamentum editum est’: The Five-Hundredth Anniversary of Erasmus’s New Testament,” BibTrans 67 (2016): 9–28, there 10; for this edition cf. also Valentina Sebastiani, “The Impact of Erasmus’ New Testament on the European Market (1516–1527): Considerations Regarding the Production and Distribution of a Publishing Success,” in Basel 1516: Erasmus’ Edition of the New Testament, ed. Martin Wallraff, Silvana Seidel Menchi, and Kaspar von Greyerz (Tübingen: Mohr Siebeck, 2016), 225–237, esp. 230–233. 3 The inclusion of punctuation should be of interest for New Testament scholars. See Martin Wallraff, “Paratexte der Bibel: Was Erasmus edierte außer dem Neuen Testament,” in Basel 1516: Erasmus’ Edition of the New Testament, 145–173; 146: “Begreiflicherweise ist für ErasmusForschung relevant, was Erasmus geschrieben hat, während für Neutestamentler der von Erasmus konstituierte Text nur in gewissen historischen Zusammenhängen noch von Interesse ist.”

© Hans Förster, 2023 | doi:10.1163/9789004537972_004

34

Förster

student assistants to check the punctuation and the spelling of a paper written by an authority in his or her field.4 Thus, punctuation is comparable to the invisible man of Gilbert Keith Chesterton’s Father Brown.5 The postman is obviously a “mentally invisible man” not recognized as anybody important. Who would suspect a postman to be an imposter or punctuation to miss the point? Consider then, if the New Testament were written today in the same manner as P66 or P75, this would trouble many readers. The text has been written in the so-called scriptio continua, a writing style that does not have spatial segmentation between words. The continuous sequence of characters required the reader to divide the characters into words. The reading of scriptio continua, therefore, required a technique different from the written texts of today; namely, ancient readers read aloud in order to understand the text. Consequently, modern punctuation and separation of words makes reading and understanding easier but there is a catch. These modern reader aides are added to the text of the Greek New Testament much later and not by the original authors.6 The critical edition of the New Testament contains texts written by authors in antiquity while their authoritative interpretation, which is what punctuation does, is from a later time and persons different from the author(s). While making things easier for readers this could – at least in some cases – be an easy way to lead the reader astray. The punctuation of the printed editions of the New Testament has to date not been systematically researched or criticised. This paper begins to rectify that. Using examples from the Gospel of John, the paper argues there are interpretive problems arising from punctuation indicating that editorial punctuation choices have led readers in directions that cause problems for both

4 And indeed, I would like to thank Chris S. Stevens, co-editor of this volume, for editing this contribution, especially for improving my English. 5 Alison Milbank, Chesterton and Tolkien as Theologians. The Fantasy of the Real (London/New York: T&T Clark, 2009), 92: “In ‘The Invisible Man,’ for example, the tale’s title is its central paradox: four witnesses swear that no-one has entered Himalaya Mansions and Smythe’s flat, yet his body has disappeared from it, to be found eventually in the canal; there is blood on the apartment’s floor and footprints in the snow. … Father Brown can ‘see’ the invisible man … a postman whose function as daily mediator of letters means that he is totally overlooked.” For the short story “The Invisible Man,” cf. Gilbert Keith Chesterton, The Innocence of Father Brown (New York: Lane, 1911), 119–146. 6 The assumption is that every text has some sort of original author. It does not matter whether this author is known or not, whether this author is a single person or a collective writing the text or whether, the author was actually a succession of different persons writing and reshaping the text till it acquired its final form.

Missing the Point

35

grammar and interpretation.7 The problems indicated would not exist if punctuation had been used in a more restrictive manner. Furthermore, modern punctuation and the problems caused by it are demonstrated to be a hereditary disease, which was first introduced into the printed editions by Erasmus. He in turn seems to depend – at least in some of his decisions – upon the Latin punctuation from the time of Charlemagne. 2

Colon and Sentence Structure: John 8:25

The first example is taken from the Gospel of John 8. The example highlights the problems of the implicit but authoritative construction of a sentence structure by means of punctuation. According to Roger L. Omanson, John 8:25 is highly problematic and grammatically troubled. He comments that “the words of Jesus, Τὴν ἀρχὴν ὅ τι καὶ λαλῶ ὑμῖν, are an incomplete sentence. They have been called‚ the most obscure sentence in the Gospel and the most uncertain how to translate.”8 Τὴν ἀρχήν stands not within the sub-clause introduced with ὅ τι. Consequently, it is impossible to understand τὴν ἀρχήν as modifier of the verb λαλῶ. The first step to understanding the persistence of editorial punctuation is to start with the earliest printed edition of the Greek New Testament, the Novum instrumentum omne published by Erasmus of Rotterdam. It is, in fact, not just a Greek New Testament but a bilingual edition having in the left-hand column the Greek and in the right-hand column the Latin. While it is well known that the editio princeps of 1516 contains multiple errors, such errors do not concern the matter at hand.9 In fact, the examples discussed in this paper have not been indicated or corrected in any printed critical edition after Erasmus’s editio princeps. It is, therefore, legitimate to start with the edition of 1516 despite its 7 I have dealt with the grammatical problems of each example elsewhere and so focus in this paper on the punctuation matters. 8 Roger L. Omanson, A Textual Guide to the Greek New Testament. An Adaptation of Bruce M. Metzger’s Textual Commentary for the Needs of Translators, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 2012), 185; with reference to George Raymond Beasley-Murray, John, WBC 36 (Waco, TX: Word, 1987), 125; and D. A. Carson, The Gospel According to John, PNTC (Grand Rapids: Eerdmans, 1991), 345–346. 9 Cf. Kurt Aland and Barbara Aland, Der Text des Neuen Testaments: Einführung in die wissenschaftlichen Ausgaben sowie in Theorie und Praxis der modernen Textkritik (Stuttgart: Deutsche Bibelgesellschaft, 1982), 12, comments that those errors have been pointed out to Erasmus during his lifetime and that it was an English critic of the 19th century who described this book as the most carelessly prepared book ever printed.

36

Förster

known shortcomings as it provides insight into how punctuation was inserted and how it might change the perception of what is written. Erasmus apparently followed in his punctuation of John 8:25 the sentence division as he discovered it in at least one of the Greek manuscripts he used for his edition. Such a sentence structure is clearly visible in the minuscule Basel University Library A. N. IV. 2 (GA1) which was at his disposal in compiling his critical edition. GA1 dates from the 12th century. Greek manuscript used by Erasmus in preparing his edition of John 8:25 (See Figure 2.1).

Figure 2.1

Basel University Library A. N. IV. 2 (GA1; fol. 277r)

The text of John 8:25 has changed little in printed editions that appeared after Erasmus’s ground-breaking publication. There are not many textual variants, but the punctuation shows an obvious unease of the editors with parts of the text of John 8:25. Erasmus’s editio princeps John 8:25:

Figure 2.2

Novum Instrumentum omne

The edition of Tischendorf:

Figure 2.3

Tischendorf p. 839

The Novum Testamentum Graece28 reads ἔλεγον οὖν αὐτῷ· σὺ τίς εἶ; εἶπεν αὐτοῖς ὁ Ἰησοῦς· τὴν ἀρχὴν ὅ τι καὶ λαλῶ ὑμῖν; There are only minor differences between Erasmus, Tischendorf, and the Nestle-Aland edition; they differ only in punctuation. Erasmus has a full stop after Ἰησοῦς, while Tischendorf and Nestle-Aland offer a colon. It is a matter of taste whether one separates the direct speech from the narrative frame by means of a colon or a full stop.

Missing the Point

37

However, the punctuation in Erasmus’s edition at the end of John 8:25 is a comma, Tischendorf has a full stop, and the Nestle-Aland uses a semicolon. This sentence has apparently troubled editors leading them to different editorial decisions. The comma and full stop have similar functions in editions of classical texts and in modern texts. A semicolon, however, is a distinctive choice to use in the edition of a Greek text. It changes the discourse since a semicolon signifies a question mark in Greek. Thus, the different editions used as examples here have either a declarative (Erasmus/Tischendorf) or an interrogative sentence (Nestle-Aland). The shift from a declarative sentence to an interrogative sentence has the potential to fundamentally change the understanding of the text. The translations follow the pattern of the editions. They offer either a declarative sentence or an interrogative sentence. The King James Version (KJV) opts for the declarative sentence, “Then said they unto him, who art thou? And Jesus saith unto them, even the same that I said unto you from the beginning.” The Good News Bible (GNB) offers, “‘Who are you?’ They asked him. Jesus answered, ‘What I have told you from the very beginning.’”10 Conversely, the New Revised Standard Version (NRSV) opts to translate the direct speech of Jesus in John 8:25 as a question, it reads as “They said to him, ‘Who are you?’ Jesus said to them, ‘Why do I speak to you at all?’”11 It may be that the differences between declarative and interrogative sentences are comparatively small in this case, but they are noticeable. For instance, the NRSV introduces a notion of rejection. Jesus raises the rhetorical question as to why he is speaking at all to his Jewish audience. No answer is expected, and the intention is to show that talking to this audience is without merit. Therefore, given that the rhetorical question expresses the notion of exasperation in light of a stupid/obstinate audience, this rhetorical question expresses an important point. Even a notion of vengeance can be inferred in the translation phrase “speak to you at all.”12 To use such a phrase as translation 10 The English Standard Version (ESV) is close to the GNB: “So they said to him, ‘Who are you?’ Jesus said to them, ‘Just what I have been telling you from the beginning.’” This pertains also to the New International Version (NIV): “‘Who are you?’ they asked. ‘Just what I have been telling you from the beginning,’ Jesus replied.” 11 The revised Einheitsübersetzung (2016) is close to GNB and has: “Da fragten sie ihn: Wer bist du denn? Jesus antwortete: Warum rede ich überhaupt noch mit euch?” The revised Lutherbibel (2017) opts for: “Da fragten sie ihn: Wer bist du denn? Und Jesus sprach zu ihnen: Was soll ich euch zuerst sagen?” 12 In order to show that exasperation is expressed by this phrase, one can quote the way Mrs. Bernstein uses a similar phrase in her letter to the famous editor Max Perkins, who discovered authors like Ernest Hemmingway, F. Scott Fitzgerald, and Thomas Wolfe (the letter is quoted in Scott Berg’s work describing the editor Max Perkins): “‘It is not easy for me to be hard, and I have to push myself in that direction, in order to speak to you at all of

38

Förster

contributes, therefore, most certainly to what Wheaton has called an “undercurrent of judgment” upon the Jews.13 The obvious grammatical problem is that the direct speech of Jesus starts with an incomplete and obscure sentence, in fact, it is actually unhelpful to call it a sentence since it lacks a subject and predicate. What has been called a sentence consists of two words, an article and an accusative noun. There is nothing like an absolute accusative in Greek and this is, therefore, hardly a sentence. A sub-clause immediately follows, which is introduced by either a relative pronoun (ὅ τι) or a conjunction (ὅτι). Both words are in most of the ancient manuscripts isomorphic. It is – at least from the point of view of Greek philology – inappropriate (or even incorrect) to call the beginning of the direct speech of Jesus a sentence (τὴν ἀρχήν). These are just two words in the accusative, and they do not constitute a sentence. In consequence, while this sequel of words had been deemed the most obscure sentence in the Gospel, it should be called a phrase. It is not a new or a modern insight to say that the beginning of the direct speech in John 8:25 needs help to be understood properly. There is evidence in the manuscripts of the perceived difficultly of the phrase motivating adjustment. Manuscripts (or rather their scribes) deal with such and similar problems by correcting or changing the text either superlinearly or in the margins, thereby creating textual variants. Consider that P75 is scriptio continua and offers no markers as to how the text is to be read. It is, therefore, the task of the educated reader to make up his or her mind as to how to understand the text. This is, at least according to David Parker, actually still today the task of every reader.14

Figure 2.4

P75 Fol. 55r (Lines 9–11)

what was in my mind.’ She was not acting out of vengeance, she explained.” Andrew Scott Berg, Max Perkins: Editor of a Genius (repr., New York: New American Library, 2016), 272. 13 Gerry Wheaton, The Role of Jewish Feasts in John’s Gospel, SNTSMS 162 (Cambridge: Cambridge University Press, 2015), 45. 14 David C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997), 212: “There is a manifold tradition to be studied and from which we may learn. But once that is done the people of God have to make up their own minds. There is no authoritative text to provide a short-cut.”

Missing the Point

39

Furthermore, the scribe of P66 seems to have the impression that something is amiss and decides there is actually something missing. To rectify matters the scribe adds words at the left hand side of line 13 and adds a mark or stroke after the nomen sacrum ΙΣ indicating the words written in the margin should be inserted there when reading the text. The scribal reading of John 8:25 becomes ἔλεγον αὐτῷ σὺ τίς εἶ εἶπεν αὐτοῖς ὁ Ἰ(ησοῦ)ς εἶπον ὑμῖν τὴν ἀρχὴν ὅ τι κ(αὶ) λαλῶ ὑμῖν.

Figure 2.5

P66: page 54 lines 12–14

The insertion of εἶπον ὑμῖν at the beginning of the direct speech makes this a complete and functioning sentence. The scribal correction of P66 suggests the modern editorial punctuation is correct. However, other manuscripts deal differently with the problem. The presentation of the text in two (rather important) manuscripts actually suggests that modern punctuation might miss the point. Codex Bezae Cantabrigiensis (D/05) is written in a manner to make reading for an audience easier. Therefore, the lines of Codex Bezae Cantabrigiensis segment into units that belong together.

Figure 2.6

D/05 135v lines 25–27

One possible interpretation is that the scribe of D/05 sees the noun in the accusative as part of a larger structure. In standard New Testament editions τὴν ἀρχήν constitutes the beginning of the direct speech. In D/05 it may be seen as the very end of the narrative frame. The same holds true for Codex Basiliensis (07). Thus, on the basis of manuscript evidence the following punctuation of John 8:25 is appropriate, at least if one were to edit the text based on its presentation in D/05: ἔλεγον οὖν αὐτῷ· σὺ τίς εἶ; εἶπεν αὐτοῖς ὁ Ἰησοῦς τὴν ἀρχήν· ὅ τι καὶ λαλῶ ὑμῖν.

40

Förster

Such an edition of the text following the layout of a codex like D/05 makes the natural breaks in the text apparent. John 8:25 is comprised of three units. Unit one has two parts, the first is the narrative frame introducing the reader to the fact that direct speech of Jesus’s Jewish opponents will follow ἔλεγον οὖν αὐτῷ, the second part is their direct speech σὺ τίς εἶ;. Unit two consists of a longer text forming part of the narrative frame. It states the fact that Jesus is immediately answering εἶπεν αὐτοῖς ὁ Ἰησοῦς τὴν ἀρχήν. In unit three, the reported direct speech begins: ὅ τι καὶ λαλῶ ὑμῖν. This segmentation makes an understandable Greek sentence without any grammatical problem. It is well known that the accusative of a temporal noun can function like an adverb and here it occurs at the end of the sentence. The adverbial use of τὴν ἀρχήν at the end of a sentence can also be found in other Greek writings, e.g. in John Chrysostom.15 It is possible to translate the Greek text then as follows, “They said to him: ‘Who are you?’ Jesus spoke to them immediately: ‘And I am telling you this.’”16 As shown above, the NRSV – on the basis of the text as presented in the critical editions – creates a notion of rejection in the translation. However, an edition of the text that considers the different options of punctuation offered by the manuscripts can present syntactical configurations that are faithful to both the Gospel as handed down in early textual tradition and also in accordance with the basic rules of Greek grammar. Furthermore, it may be the case that the Greek of John 8:25 exhibits Semitic influence.17 By extension the author of the Gospel of John might be using language subtleties that were not always obvious to scholars trained mainly in classical Greek.18 Consequently, it appears modern punctuation has contributed to the perception that the text itself has problems. The input of manuscript evidence concerning John 8:25 is a rather good example of problems created by modern critical edition punctuation and also 15

For full documentation of similar structures, see Hans Förster, “Überlegungen zur Grammatik von Joh 8,25 im Lichte der handschriftlichen Überlieferung,” ZNW 107 (2016): 1–29. 16 Cf. Hans Förster, “Possible Similarities of the Linguistic Structure of John 8.25b and John 8.45a,” BibTrans 68 (2017): 164–178, there 176. 17 Cf. Förster, “Similarities,” 175: “It appears that both John 8.25b and John 8.45a exhibit sentence structures which show Semitic influence and hint at the linguistic background of the author of John’s Gospel.” 18 Cf. Hans Förster, “Selbstoffenbarung und Identität: Zur grammatikalischen Struktur der ‘absoluten’ Ich-Bin-Worte Jesu im Johannesevangelium,” ZNW 108 (2017): 57–89; Förster, “Überlegungen zur angeblichen Textverderbnis von Apg 17,27 (Codex Bezae Cantabrigiensis) und zu Joh 8,25,” in The Language and Literature of the New Testament: Essays in Honor of Stanley E. Porter’s 60th Birthday, ed. Lois K. Fuller Dow, Craig A. Evans, and Andrew W. Pitts, BIS 150 (Leiden: Brill, 2017), 138–157.

Missing the Point

41

the remedy. Punctuation in critical editions must be deemed a guideline – and fallible at that – by the editors for the intended reader. The punctuation reveals how the editors intend the readers of the Greek New Testament to understand the text. As noted above these interpretive guidelines were already used in printed editions in the 16th century and are based on punctuation in Latin manuscripts, which have never been systematically evaluated in light of the evidence of Greek New Testament manuscripts. In consequence, punctuation preserves decisions of the early editors of the New Testament. One such decision which might even be deemed an error emerges in the punctuation of John 8:25. It becomes further evident the division into verses is something that happened very late in the textual history of the New Testament. It is only in the 16th century that editions of the New Testament start to use chapter and verse (the verse-structure was introduced by the edition of the Frenchman Robert Estienne called in a Latinized form of his name Stephanus).19 On the basis of the presentation of John 8:25 in D/05 one might be tempted to raise the question whether the beginning of John 8:26 is at the right place, at least if the scribe of D/05 were to have his say. 3

A Missing Comma and a Point Missed? The Example of John 8:45

To date there is no discussion about the syntactical problems John 8:45 exhibits. The Novum Testamentum Graece offers the following text ἐγὼ δὲ ὅτι τὴν ἀλήθειαν λέγω, οὐ πιστεύετέ μοι. The punctuation in modern editions differs slightly from the punctuation given in Erasmus’s editio princeps, which has only a final full stop but avoids any other form of sentence division.

Figure 2.7

Novum Instrumentum omne p. 215

The difference between the modern critical edition and Erasmus’s text is apparent. In the critical edition of Novum Testamentum Graece there is a comma between two sentences while Erasmus avoids any division within the sentence. Tischendorf offers a division of the sentence similar to what modern critical editions have. 19 Cf. Aland and Aland, Text, 14–16. See also Elliott, “Novum Testamentum,” 23: “This 1551 edition was the first to introduce versification….”

42

Figure 2.8

Förster

Tischendorf p. 844

The translation of the sentence appears straightforward. The NRSV suggests, “But because I tell the truth, you do not believe me.” The KJV offers, “And because I tell you the truth, ye believe me not.”20 The difference between the edito princeps and modern editions is small but decisive. While Erasmus refrains from punctuation between sub-clause and main clause, modern editions add some punctuation thereby possibly disguising a problem of the sentence structure. The structure of the sub-clause is more problematic than hitherto acknowledged. It seems the author placed the subject of the causal sub-clause ἐγὼ δέ outside of the sub-clause, which opens with the conjunction ὅτι. The explicit subject of the sub-clause is, therefore, not part of the sub-clause. As a result, the beginning of John 8:45 may be described as a defective nominal sentence since it contains only a personal pronoun and a conjunction. The problem here is not the sub-clause but the defective nominal clause implicitly presupposed for the beginning of John 8:45. Translations often obscure the problematic structure by relocating the personal pronoun into the sub-clause. One would actually expect something like the following structure as the Greek source text for the translations in NRSV or KJV: ὅτι δὲ ἐγὼ τὴν ἀλήθειαν λέγω, οὐ πιστεύετέ μοι. The difference is subtle. Given that some argue John’s Gospel is rather simple Greek,21 the seemingly strange sentence structure of the subject of a sub-clause placed proleptically 20 NIV has: “Yet because I tell the truth, you do not believe me!” ESV offers: “But because I tell the truth, you do not believe me.” Lutherbibel 2017 has: “Weil ich aber die Wahrheit sage, glaubt ihr mir nicht.” Revised Einheitsübersetzung (2016) reverts the structure of the sentence and positions the causal sub-clause after the main clause translating: “Mir aber glaubt ihr nicht, weil ich die Wahrheit sage.” 21 Norman Hugh Young, “The Use and Purpose of Three Frequent Syntactical Forms in the Fourth Gospel,” BibTrans 67 (2016): 315–330, there 315: “The Fourth Gospel (FG) uses three quite sophisticated forms of syntax more frequently than one would expect given its otherwise simple Greek.” The concept that the style of the New Testament lacks refinement has a long tradition; cf. Erika Rummel, “Biblical Humanism,” in Basel 1516: Erasmus’ Edition of the New Testament, 27–42, there 28–29: “In the eyes of the humanists, the style of the biblical authors was primitive, to say the least.… The discussion over biblical style was by no means new. The humanists followed the example of Saint Jerome and often cited him to validate their own judgment.”

Missing the Point

43

before the conjunction, one might be tempted to see this as a mistake of the author. The grammar of Blass, Debrunner, and Rehkopf, sees narrators/authors of the Gospels as comparable to “folk-narrators,” implying that the language of the texts is an easy vernacular.22 Thus, it seems permissible to posit an error of the author to explain the seemingly strange and irregular construction. The problem continues in the Latin version of the text. The Vulgate and Erasmus’s Latin text have, ego autem quia ueritatem dico non creditis mihi. The Vulgate and Erasmus interpret the sentence structure in the way described above, with the personal pronoun standing outside the sub-clause but needing to be read within the sub-clause. There is no other referent for the personal pronoun but the verb in the sub-clause since the sub-clause has a verb in the first person singular while the main clause has a verb in the second person plural. While outside of the structure, the proleptic personal pronoun needs to be placed within the sub-clause for translation. Therefore, standard interpretation of the sentence structure exhibited in critical editions of the Greek New Testament of John’s Gospel presents the sentence as grammatically problematic. There are two manuscripts that can be mentioned as offering some insight. D/05 and 07 (Codex Basiliensis) appear to be similar to what modern critical editions have. However, appearances may – in case of manuscripts – be deceptive. Figure 2.9

Codex Bezae Cantabrigiensis John 8:45

The transcript of Codex Bezae Cantabrigiensis (John 8:45) in the virtual manuscript reading room of the INTF (Institut für Neutestamentliche Textforschung; University of Münster/Germany) transcribes the text: εγω οτι την αληθειαν λαλω· ου πιστευετε μοι. The transcription offers a colon instead of the comma but looks quite similar to the critical edition. The adversative conjunction (δέ) 22 Friedrich Blass and Albert Debrunner, Grammatik des neutestamentlichen Griechisch, bearbeitet von Friedrich Rehkopf, 18th ed. (Göttingen: Vandenhoeck & Ruprecht, 2001), §470.1: “Da die indirekte Redeform volkstümlichen Erzählern und so auch den ntl Erzählern nicht liegt, folgt auf ὅτι gewöhnlich nicht nur Ind. statt Opt., sondern auch genaue Nachbildung direkter Redeform, so daß dieses ὅτι die Rolle unseres Anführungszeichens vertritt (ὅτι recitativum).” Cf. Friedrich Blass and Albert Debrunner, A Greek Grammar of the New Testament and Other Early Christian Literature, trans. Robert W. Funk (Chicago: University of Chicago Press, 1967), §470.1: “Since indirect discourse, whether it be with and the optative or with the acc. (nom.) and infinitive, is not at all congenial to the NT narrators any more than it is to folk-narrators in general….”

44

Förster

is missing in this manuscript, and its omission is not relevant for our purposes. Codex Bezae Cantabrigiensis has two clauses divided by punctuation. Since D/05 uses punctuation rather restrictively, to find here a sparingly used mark is telling to the reader.

Figure 2.10

Codex Basiliensis John 8:45

Codex Basiliensis is a bit more generous in its use of punctuation. As can be seen in Figure 2.10, a sign resembling a comma or a full stop is placed at the end of John 8:44. In all probability it indicates the beginning of a new syntactic unit. A similar sign is used at the end of John 8:45a separating the sub-clause from the main clause. The end of the sentence also has a colon. Some breathing marks and accents are used in this manuscript too. What is of interest is the fact that the sentence of both D/05 and 07 could also be transcribed, εγω (δε) ο τι την αληθειαν λαλω/λεγω · ου πιστευετε μοι. This presentation of the Greek text is in accordance with the manuscripts shown above, which have no (D/05) or only marginal (07) word divisions in the sections shown above. As seen in the image, D/05 has no conjunction δέ here. Also, the proposed transcript transforms the conjunction ὅτι into a pronoun ὅ τι. Again, in scriptio continua both the conjunction and the pronoun are isomorph. It is the decision of the editor either to use the pronoun or the conjunction. In my proposed reconstruction the pronoun functions as the direct object making τὴν ἀλήθειαν – which was a direct object in a sub-clause in the critical edition – function adverbially. The use of the noun ἀλήθεια in the accusative in place of an adverb is very common in Septuagintal Greek. This rather small emendation of the text resolves a troubled sentence making it grammatically correct Greek, albeit with traits of Semitic influence. Furthermore, if interpreted like this John 8:45 does not contain anymore hypotactic sub-clause to main-clause relationships but only two independent and asyndetically coordinated main-clauses. The paratactic main clauses and the use of an accusative in adverbial function could support the notion of Semitic influence upon the Greek language structure.23 Such an influence is 23

The very fact that the sentence structure of John 8:45 has been deemed in critical editions implicitly a troubled sentence structure (placing the subject of the perceived sub-clause proleptically outside the sub-clause) points to intended readers who could understand

Missing the Point

45

to be expected in a text written by an author who is able to translate independently from Hebrew to Greek (as he did with the quotation from Isa 6:10 in John 12:40; cf. below). While the suggested change may be small it has the potential to change the interpretation of the passage. Traditional translations introduce causality by presenting Jesus as speaking “the truth” and for this very reason “the Jews” do not believe. The GNB is as dynamic equivalent translation able to convey this understanding of the text aptly in English, “But I tell the truth, and that is why you do not believe me.” Additionally, TLB (The Living Bible) is also very explicit in the interpretive translation of John 8:45, “And so when I tell the truth, you just naturally don’t believe it!” Bruner suggests that translation should be “as clear and contemporary as possible in order to be already a commentary on the passage and its meaning.”24 He arrives in John 8:45 at the following translation: “It is precisely because I, I myself, am telling you the truth that you cannot believe me.”25 He then comments on the dangers of this translation: “This dualism can, if we are not very careful, venomize [sic] our Christian relation with our co-religionists or our theological enemies ….”26 Bruner is not the only one to describe the language of John’s Gospel in its phrasings concerning “the Jews” as poison and venom.27 Such an “undercurrent of judgment,” or even rejection, finds support in traditional translations based on current critical editions. If one were to use the sentence as structured naturally one would arrive at the following translation (07), “But I am saying this (ὅ τι) honestly/truthfully (τὴν ἀλήθειαν): you do not believe me.” Or (D/05): “I am saying this (ὅ τι) honestly/truthfully (τὴν ἀλήθειαν): (And) you do not this type of Greek with Semitic influence; cf. Henry S. Gehman, “The Hebraic Character of Septuagint Greek”, in The Language of the New Testament: Classic Essays, ed. Stanley E. Porter, JSNTSup 60 (Sheffield: Sheffield Academic, 1991) 163–173, there 163: “The object of a translator obviously is to render a document clearly into the vernacular. Upon reading the LXX, however, it is often difficult to obtain the sense without comparing the Hebrew text.… If the Jews who read the LXX did not understand Hebrew, we may infer at least that the translation made sense to them and that it was intelligible when it was read in the synagogue.” 24 Frederick Dale Bruner, The Gospel of John: A Commentary (Grand Rapids: Eerdmans, 2012), xiii. 25 Bruner. Gospel of John, 543. 26 Bruner. Gospel of John, 543. 27 Gerd Theißen, “Aporien im Umgang mit den Antijudaismen des Neuen Testaments,” in Die Hebräische Bibel und ihre zweifache Nachgeschichte: FS Rolf Rendtorff zum 65. Geburtstag, ed. Erhard Blum, Christian Macholz, and Ekkehard W. Stegemann (Neukirchen: Neukirchener, 1990), 535–553, there 535.

46

Förster

believe me.” In both translations (either translating the adversative conjunction but of 07 or transforming the asyndetic coordination of 05 into a copulative coordination using the common conjunction and) the exasperation of an orator is apparent, namely of one trying to orally convince his audience while the audience does not heed his words. Thus, the rather small change of the edited text changes the meaning of the text. It is not unusual that human beings do not always believe honest people. Admittedly, John 8:44 is highly anti-Jewish and might support a rather polemical understanding of John 8:45 and John 8:43. As current discussion shows, John 8:44 has problems of syntax, translation, and interpretation which make its interpretation the subject of lively debate.28 Standard translations of John 8:44 have even been said to reflect the “suicide of Greek philology.”29 4

The Full Stop in John 8:43

The presentation of John 8:43 in a bilingual manuscript from the 9th century – 037/Δ Codex Sangallensis – will be used as starting point to discuss the problems of the coordination of the two sentences contained in this verse.

Figure 2.11

Codex Sangallensis (037/Δ); John 8:43 (p. 352; l. 5–6)

It is apparent that Codex Sangallensis does not use punctuation in the same way as it is done today. Further, Codex Sangallensis presents the relationship 28 Cf. e.g. April DeConick, “Why are the Heavens Closed? The Johannine Revelation of the Father in the Catholic-Gnostic Debate,” in John’s Gospel and Intimations of Apocalyptic, ed. Catrin H. Williams and Christopher Rowland (London: Bloomsbury, 2014), 147–179, and Stephen Robert Llewelyn, Alexandra Robinson, and Blake Edward Wassell, “Does John 8:44 Imply that the Devil Has a Father? Contesting the Pro-Gnostic Reading,” NovT 60 (2018): 14–23. 29 Emanuel Hirsch, Studien zum vierten Evangelium (Text/Literarkritik/Entstehungsgeschichte), BHT 11 (Tübingen: Mohr [Siebeck], 1936), 78–79: “Der Schluß des Satzes kann, sprachlich korrekt verstanden, nur sagen: ‘Denn auch sein Vater ist ein Lügner,’ schreibt also wiederum dem, von dem geredet wird, einen Vater zu. Da es einen Vater des Teufels nicht gibt, kann der Satzschluß nur durch Korrektur der ganzen Aussage seinen jetzigen falschen Sinn empfangen haben. Die durch Luthers Übersetzung vertretene Deutung des Satzschlusses ist Selbstmord der Philologie.”

Missing the Point

47

between Greek and Latin text in a rather interesting way. In direct contradistinction to most of the manuscripts – and also to the presentation in the Novum Instrumentum omne of Erasmus of Rotterdam or the Complutensian Polyglot – this manuscript has an interlinear structure. The Greek text is written as majuscule and the comparatively smaller Latin script is spaced and written in a way that looks very much like a modern interlinear edition. The most important difference is that modern interlinear translations tend to present the modern translation below the Greek text and not above. It is easy to see the punctuation is rather erratic. There are dots after a few words without any particular method. Thus, in this case punctuation of the Greek text does not seem to have any impact on its understanding but the presentation of the text is important. The presentation implies that the Latin text follows closely the Greek text. The direct equivalent of the respective word in Greek is written in Latin above it: διὰ τί is translated as quare, τὴν λαλιάν corresponds to loquela(m), τὴν ἐμήν to mea(m), etc. The presentation conveys the following message, the scribe intends the user to understand that the Latin version of the text is a faithful rendering of the Greek text, translating every detail literally. It will be argued that punctuation of John 8:43 has the potential to change the understanding of the text. The punctuation in John 8:43 is especially interesting for research since it imposes a decision on the text, which the early scribes of New Testament did not and even could not make. This concerns the use of a symbol indicating that a sentence is not to be read as a declarative sentence but as an interrogative sentence. The “question mark” was not used in Greek manuscripts in antiquity. The reader was required to make the decision whether a sentence was to be read either as declarative or as interrogative. It is, once again, the edition of Erasmus that sets the standard for all other editions to follow concerning the relationship between the two sentences in John 8:43 and its interpretation. The edition of 1516 uses a semicolon for the Greek text and has in the Latin translation a – compared with modern question marks – rather rudimentary question mark.

Figure 2.12

Novum Instrumentum omne p. 215

Erasmus is obviously following the practice as attested in older Latin manuscripts. One of the earliest manuscripts attesting a question mark at John 8:43 in Latin is Codex Colbertinus (Vetus Latina).

48

Figure 2.13

Förster

Codex Colbertinus (Fol. 74r; col. 2; l. 32–35)

Elliott remarks that it was not so much a Greek New Testament but rather “a purified enhanced Latin New Testament that Erasmus wished to promote in his pious attempt to encourage a greater understanding of the Christian faith.”30 Consequently, it does not come as a surprise that Latin interpretation of the biblical text, as attested in the Carolingian manuscript used here as an example Codex Colbertinus, had influence on Erasmus’s understanding and punctuation of the Greek text. Additionally, the Complutensian Polyglot of 1520 avoids the use of the semicolon in the Greek text but gives a question mark in the Latin translation of John 8:43.

Figure 2.14

Complutensian Polyglot John 8:43

The later editions like Tischendorf or Novum Testamentum Graece 28th edition follow Erasmus’s editorial decision. The text of John 8:43 appears straightforward: διὰ τί τὴν λαλιὰν τὴν ἐμὴν οὐ γινώσκετε; ὅτι οὐ δύνασθε ἀκούειν τὸν λόγον τὸν ἐμόν. The NRSV translates as, “Why do you not understand what I say? It is because you cannot accept my word.” The Good News Bible (GNB) strengthens the interpretation of the text by translation, “Why do you not understand what I say? It is because you cannot bear to listen to my message.”31 The rhetorical devise used by Jesus is apparent. He asks – just as in some translations of John 8:25 – a rhetorical question and then he answers it. The answer indicates that he does not see any possibility that his audience might 30 Elliott, “Novum Testamentum,” 13. 31 NIV translates: “Why is my language not clear to you? Because you are unable to hear what I say.” ESV offers: “Why do you not understand what I say? It is because you cannot bear to hear my word.” Lutherbibel 2017 has: “Warum versteht ihr meine Rede nicht? Weil ihr mein Wort nicht hören könnt!” The revised Einheitsübersetzung opts for: “Warum versteht ihr nicht, was ich sage? Weil ihr nicht imstande seid, mein Wort zu hören.”

Missing the Point

49

actually listen to him, specifically they “cannot accept” Jesus’s word, they “cannot bear to listen” to his message. The GNB handles John 6:60 slightly different, translating Πολλοὶ οὖν ἀκούσαντες ἐκ τῶν μαθητῶν αὐτοῦ εἶπαν· σκληρός ἐστιν ὁ λόγος οὗτος· τίς δύναται αὐτοῦ ἀκούειν; as “Many of his followers heard this and said, ‘This teaching is too hard. Who can listen to it?’” The authorial intent is apparent: the audience cannot listen. In order to clarify the nature of the two sentences in John 8:43 punctuation makes the first sentence an interrogative sentence and the second sentence a declarative sentence. As mentioned, question marks do not exist in Greek manuscripts and this raises intriguing questions. Neither Codex Bezae Cantabrigiensis nor Codex Basiliensis have punctuation comparable to modern punctuation.

Figure 2.15

Codex Bezae Cantabrigiensis John 8:43 (Fol. 137v 11–13)

Note the transcription, δια τι την αληθειαν (corr: λαλειαν) την εμην ου γεινωσκετε οτι ου δυνασθαι ακουειν τον λογον τον εμον. It is easy to see in the reproduction that the scribe wrote first truth αληθειαν and then corrected it to speech λαλειαν. There is no punctuation whatsoever in the passage of Codex Bezae Cantabrigiensis. Codex Basisliensis, however, does exhibit punctuation.

Figure 2.16

Codex Basiliensis John 8:43

Codex Basiliensis divides the two sentences by a mark that has similarities with a comma to mark a shorter break while reading the text aloud and ends with a colon. Only the prepositional phrase at the beginning of John 8:43a, διὰ τί, marks this sentence as a question. The following sentence, John 8:43b, is the answer. But is this really the correct solution to the grammatical structure presented in John 8:43? In search for similar structures one can point to the manner Paul phrases one of his letters as present in Codex Claromontanus (D/06):

50

Förster

Figure 2.17

Codex Claromontanus (D/06) 2 Cor 11:11 (Fol 238v)

According to modern editions 2 Cor 11:11 is punctuated as: διὰ τί; ὅτι οὐκ ἀγαπῶ ὑμᾶς; ὁ θεὸς οἶδεν, which is commonly translated into English (ex. NRSV): “And why? Because I do not love you? God knows I do!” If the translational method of combining a question and answer used in John 8:43 were used here then 2 Cor 11:11 would read: “And why? It is because I do not love you! God knows this.” However, it is apparent that this transforms the text and reverses the meaning. Since it is commonly – probably correctly – assumed Paul does not want to say in 2 Cor 11:11 that he does not love those to whom he has written his letter, the traditional translational choice is accepted as the correct one. Furthermore, Codex Augiensis, another a bilingual codex from the 9th century (University of Cambridge), uses question marks in the Latin text. In fact, it is one of the earliest manuscripts to use a question mark . This form of punctuation mark came into use in Carolingian minuscule manuscripts around 800.32 Its use here indicating the scribe of the Latin text sees the two sentences as two questions, meaning that the Latin text and its Greek source text were both understood as a question followed by a second question.

Figure 2.18

Codex Augiensis (Fp/010; fol 63r)

There is actually not much difference concerning the sentence structure of 2 Cor 11:11 and John 8:43. Both consist of two main clauses. In both, the first main clause is treated as a question introduced by διὰ τί. Also, in both examples the second clause begins with ὅτι. However, in 2 Cor 11:11 the ὅτι is interpreted as introducing a second question, while editors of John 8:43b interpret the 32

Alexandra Wiebelt, Symmetrie bei Schriftsystemen: Ein Lesbarkeitsproblem, Linguistische Arbeiten 488 (Tübingen: Max Niemeyer, 2004), 838 n. 221; see also Paul Edward Dutton, Charlemagne’s Mustache: And Other Cultural Clusters of a Dark Age (New York: Palgrave Macmillan, 2004), 86: “Charlemagne himself or those writing on his behalf had already begun to fuss over punctuation (the question mark being one of their new signs) and were especially attentive to details of legibility, standardization, and correctness.”

Missing the Point

51

ὅτι as introducing the answer to the proceeding question of John 8:43a. It is, therefore, the authoritative decision of the editors to decree that John 8:43b constitutes the answer to the rhetorical question in John 8:43a, even though the same syntactic structure in 2 Cor 11:11b is treated as a second question following the first. It seems probable that modern translations are in this instance still influenced by the Vulgate. The Vulgate appears to exhibit a deliberate translational choice here.33 There is an answer to the puzzling question why two grammatically closely related (if not identical) ways of coordinating two sentences should lead to mutually exclusive interpretations. The reader makes up his or her mind as to how to understand a given text. Umberto Eco describes this as the “uncontrollable drives of the reader.”34 Since some view the author of the Gospel to be “intentionally hostile”35 against the Jews, the use of two questions to translate the text for John 8:43 conflicts with common sense and theological knowledge, even though it seems to be more consistent with Greek grammar. However, John 8:43 can – and probably should – be translated: “Why do you not understand what I say? Is it because you cannot hear my word?” The consequences for interpretation are apparent: this sentence structure and its semantics would raise the rather intriguing question whether the problems of John 8:44 mentioned above could possibly point to a different meaning of John 8:44. Whereas traditional translations imply an outright rejection of the audience, the syntactically possible and possibly more appropriate translation suggests an intensive dialogue between Jesus and his Jewish audience. Jesus basically asks his audience whether there is anything (loud voices or whatever) that might prevent his audience from hearing Jesus’s words. It should be noted that ἀκούω concerns first and foremost the physical act of hearing. Thus, neither “accept” nor “listen” are the best lexical choices to represent the semantics of the Greek word appropriately. These later glosses are translational choices supporting traditional interpretation of the text. In consequence, any 33 See Hans Förster, “Martin Luther und die Veritas Graeca – Eine Positionsbestimmung,” KuD 66 (2020): 195–219, here 208–209. 34 Umberto Eco, “Overinterpreting Texts,” in Umberto Eco: Interpretation and Overinterpretation, ed. Stefan Collini (Cambridge: Cambridge University Press, 1992), 45–66, there 65: “How to prove a conjecture about the intentio operis? The only way is to check it upon the text as a coherent whole. This idea, too, is an old one and comes from Augustine (De doctrina christiana): any interpretation given of a certain portion of a text can be accepted if it is confirmed by, and must be rejected if it is challenged by, another portion of the same text. In this sense the internal textual coherence controls the otherwise uncontrollable drives of the reader.” 35 Sonya Shetty Cronin, Raymond Brown, ‘The Jews,’ and the Gospel of John: From Apologia to Apology, LNTS 504 (London: T&T Clark, 2016), 104.

52

Förster

user of the Greek New Testament does well to remember that question marks did not exist in antiquity. The insertion of a question mark is an editorial decision introducing authoritative editorial interpretation, which might replace and obliterate authorial intent.36 5

What Difference Does It Make? The Use of a Comma Instead of a Full Stop in John 12:40

The next example, John 12:40, discusses the possible impact of the use of a comma in today’s critical editions as opposed to a full stop which could and probably should be used here. It is again a rather innocuous mark differentiating two possible interpretations of a passage that have seen a lot of discussion and comment. John 12:40 is a quotation from Isa 6:10. It is considered one of the most problematic passages of John’s Gospel since the quotation from the Hebrew Scriptures in the author’s translation introduces a notion of negative predestination. The differences in translation between the version created by the author of the Gospel and the Septuagint are significant. The Gospel of John presents a unique text indicating that it has been translated for the occasion by the author of John’s Gospel. This proves the author of the Gospel of John to have had an outstanding command of both languages, Hebrew and Greek. The common understanding of this quotation is problematic, since the quotation seems to indicate that God himself has hardened the hearts of the Jews. The edited text has not changed much over the centuries.

Figure 2.19

Novum Instrumentum omne p. 227

A similar structure of the sentence is observed in the Complutensian Polyglot.

36 Cf. for a full philological analysis Hans Förster, “Die syntaktische Funktion von ὅτι in Joh 8.47,” NTS 62 (2016): 157–166.

Missing the Point

Figure 2.20

53

Complutensian Polyglot John 12:40

The structure in Novum Testamentum Graece 28th edition is also similar, τετύφλωκεν αὐτῶν τοὺς ὀφθαλμοὺς καὶ ἐπώρωσεν αὐτῶν τὴν καρδίαν, ἵνα μὴ ἴδωσιν τοῖς ὀφθαλμοῖς καὶ νοήσωσιν τῇ καρδίᾳ καὶ στραφῶσιν, καὶ ἰάσομαι αὐτούς.37 The text in the KJV translation is, “He hath blinded their eyes, and hardened their heart; that they should not see with their eyes, nor understand with their heart, and be converted, and I should heal them.” The Good News Bible offers, “God has blinded their eyes and closed their minds, so that their eyes would not see, and their minds would not understand, and they would not turn to me, says God, for me to heal them.”38 The GNB mirrors the interpretation offered by the articles in the Theological Dictionary of the New Testament, which identifies God as the cause of the hardening of the heart.39 Theological grammars like Blass/Debrunner/Rehkopf support this notion that the last sentence belongs to the structure beginning with the conjunction ἵνα. It has escaped observation that the sentence introduced with ἵνα is negated. In consequence, the introduction of the last sentence καὶ ἰάσομαι αὐτούς is grammatically problematic for the following reasons: (1) The conjunction ἵνα typically needs a subjunctive and not a future. (2) The sub-clause has three verbs in the third person plural (ἴδωσιν, νοήσωσιν and στραφῶσιν). The last 37 Isa 6:10 (LXX) has: ἐπαχύνθη γὰρ ἡ καρδία τοῦ λαοῦ τούτου, καὶ τοῖς ὠσὶν αὐτῶν βαρέως ἤκουσαν καὶ τοὺς ὀφθαλμοὺς αὐτῶν ἐκάμμυσαν, μήποτε ἴδωσιν τοῖς ὀφθαλμοῖς καὶ τοῖς ὠσὶν ἀκούσωσιν καὶ τῇ καρδίᾳ συνῶσιν καὶ ἐπιστρέψωσιν καὶ ἰάσομαι αὐτούς. 38 The ESV offers: “He has blinded their eyes and hardened their heart, lest they see with their eyes, and understand with their heart, and turn, and I would heal them.” NIV has: “He has blinded their eyes and hardened their hearts, so they can neither see with their eyes, nor understand with their hearts, nor turn – and I would heal them.” Lutherbibel 2017 translates: “Er hat ihre Augen verblendet und ihr Herz verstockt, dass sie mit den Augen nicht sehen noch mit dem Herzen verstehen und sich bekehren und ich ihnen helfe.” The revised Einheitsübersetzung opts for: “Er hat ihre Augen blind gemacht und ihr Herz hart, damit sie mit ihren Augen nicht sehen und mit ihrem Herzen nicht zur Einsicht kommen, damit sie sich nicht bekehren und ich sie nicht heile.” 39 The authors of two entries in TDNT state explicitly that it is God who causes blindness in John 12:40 (cf. Wolfgang Schrage, “τυφλός κτλ.,” TDNT 8: 270–295, there 292); Karl Ludwig and Martin Anton Schmidt (“πωρόω [πηρόω],” TDNT 5: 1025–1028, there 1026) posit: “In Jn. 12:40 God is expressly called the author of the hardening.”

54

Förster

sentence has a verb in the first person. The tense, mode, and person, actually speak against the inclusion of the last sentence in the sub-clause. For grammatical reasons, which have been discussed elsewhere extensively,40 the sentence should be printed thus, τετύφλωκεν αὐτῶν τοὺς ὀφθαλμοὺς καὶ ἐπώρωσεν αὐτῶν τὴν καρδίαν, ἵνα μὴ ἴδωσιν τοῖς ὀφθαλμοῖς καὶ νοήσωσιν τῇ καρδίᾳ καὶ στραφῶσιν. Καὶ ἰάσομαι αὐτούς, which can be translated as, “He has blinded their eyes and hardened their heart, so that they might not look with their eyes, and understand with their heart and turn. And then I will heal them.” Arguably, it is the Latin of the Vulgate which supports the introduction of the last sentence (καὶ ἰάσομαι αὐτούς) into the sub-clause.41 Admittedly, the addition of a full stop also introduces an authoritative interpretation of the sentence structure. The introduction of a full stop would make all translators and interpreters of this passage aware that the structure might – and possibly should – be understood differently. Consequently, one could see it as preferable not to use punctuation here. Furthermore, the revised Luther Bible (revision of 2017) has heeded the argument put forward in ZNW42 and has added in its newest print run (10/2021) a note to John 12:40 stating: “Andere Übersetzung: ‘aber ich (Gott) werde ihnen helfen’ (vgl. Apg 28,27).” This might be a model for dealing with this verse in other Bible translations. In context it appears probable that “He, who has blinded” might be Jesus. It is the prophet who is told in Isaiah’s vision that he (i.e. the prophet) will cause blindness, and it is Jesus whose preaching was not successful leading to his rejection and crucifixion. In case that a full stop is placed after the last verb of the negated sub-clause, nobody would think to see the new sentence as part of the negated sub-clause. Since editorial practice always made this sentence part of the sub-clause, even grammars of the New Testament argue for a “paradigm” that “functions usually like this.”43 It may be problematic to make the sentence in the future tense part of a ἵνα μή sub-clause. Such a move possibly inverts the meaning of the text at hand and causes problems of interpretation as seen in John 12:40 and the theological discussion caused by traditional interpretation of the sentence. The traditional 40

Hans Förster, “Ein Vorschlag für ein neues Verständnis von Joh 12,39–40,” ZNW 109.1 (2018): 51–75. 41 Cf. Förster, “Martin Luther und die Veritas Graeca,” 209–214. 42 Förster, “Ein Vorschlag für ein neues Verständnis.” 43 Standard practice has no regard for the following problem: it is only possible in the case the ἵνα-sentence is not negated, to make such a directly following sentence in the future tense part of the sub-clause. Such would not change the meaning of the text. This pertains only in case that neither of the two sentences, sub-clause and following main clause in the future tense, are negated.

Missing the Point

55

understanding of John 12:40 is the basis for the medieval typology of “Ecclesia and Synagoga” which may be seen as anti-Jewish.44 Differences of understanding become evident also if one looks at the use of this quotation from Isaiah in Acts (Acts 28:26–27). Acts 28:28 may be understood as a clarification of the meaning of the quotation. A common translational decision is as follows (Acts 28:28/KJV): “Be it known therefore unto you, that the salvation of God is sent unto the Gentiles, and that they will hear it.” Arguably this translation would need a slight but important revision of the Greek source text to be an accurate rendering.45 The Greek text as published in critical editions seems to indicate the following meaning: “Be it known therefore unto you, that the salvation of God is sent unto the Gentiles; they too will hear it.” While the standard rendering seems to indicate that non-Jews instead of Jews believe, a more faithful rendering of the Greek text would indicate that both non-Jews and Jews believe. 6 Summary Modern interpretation and translation has avoided the problem that there are apparently quite a few passages that might be ambiguous in Greek. Biblical scholarship follows, therefore, one of the important aims of Erasmus’s intention in publishing the Novum Instrumentum omne: “… it was one of Erasmus’s aims to emend textual corruptions in the Latin tradition from Greek manuscripts and to correct passages inaccurately translated in the Vulgate.… he sought to reflect the original text ‘more faithfully, more clearly, and more expressively.’ The clarity of the translation demanded that ambiguous expressions be replaced and avoided.”46 It appears that Erasmus had switched in some cases from the task of the editor to the role of author. Erasmus claimed for himself the right to decide which ambiguous phrase should be disambiguated to “reflect the original text ‘more faithfully, more clearly, and more expressively.’” 44 Cf. Nina Rowe, The Jew, the Cathedral, and the Medieval City: Synagoga and Ecclesia in the Thirteenth Century (Cambridge: Cambridge University Press, 2011). 45 Cf. Acts 28:28: γνωστὸν οὖν ἔστω ὑμῖν ὅτι τοῖς ἔθνεσιν ἀπεστάλη τοῦτο τὸ σωτήριον τοῦ θεοῦ· αὐτοὶ καὶ ἀκούσονται. Arguably, for the translation of KJV to be in line with the source text, two words need to be inverted (the translation presupposes καὶ αὐτοί instead of αὐτοὶ καί). 46 Henk Jan de Jonge, “Erasmus’s Translation of the New Testament: Aim and Method,” BibTrans 67 (2016): 29–41, there 34; cf. also Marie Barral-Baron, “Erasmus and the New Testament. Innovation and Subversion,” in Basel 1516: Erasmus’ Edition of the New Testament, 239–254, there 253: “Erasmus so earnestly hoped to enable each Christian to know the word of God in all its purity and exactitude that he forgot the reality of history and the historical consequences of philological inquiry.”

56

Förster

This practice disregards the possibility of a Greek text which might be on purpose ambiguous. This paper has clearly expressed the notion that a faithful edition should not domineer over the author and should interfere as little as possible with the text in order not to alter what the author wanted to say. It seems acceptable and justified to posit that editors of important texts like the New Testament should not impose their own understanding of the text on the edition of the text. As demonstrated, punctuation is a powerful tool to introduce the notion of how a text at hand should be understood and constitutes an activity similar to what has been described by Emmanuel Tov, “… many scribes actually took an active role in the shaping of the final form of the text, and therefore the general term ‘scribe’ is more appropriate for them than ‘copyist’, since it covers additional aspects of scribal activity and could easily include creative elements.”47 Alain Kirk states that a scribe’s actual task and way of interaction with the text “accounts for the blurring of the distinction between author and scribe in a work’s transmission. Transmission qua cultivation entails that scribes are participants in the on-going formulation of a work and its constituent traditions  ….”48 It becomes apparent that such a scribal intervention is also taking place in the critical edition of the New Testament. The interpretive function of punctuation inserted by the editors of the critical text has a far stronger impact upon understanding and translation of the original text than hitherto acknowledged. Punctuation and its insertion into a text written without punctuation – at least without punctuation as used today – has the potential to make important or even decisive interpretive decisions that may side with a certain interpretation of a sentence, thereby deconstructing multiple meanings of an originally ambiguous sentence. Punctuation may also create an understanding of the text that deviates from the way an uninformed reader would read such a text if the reader were just to follow Greek grammar. The inclusion of the last three words into the negated sub-clause in John 12:40 makes the quotation from Isaiah anti-Jewish. The entire discussion about the problematic use of the quotation from Isaiah by an “intentionally hostile” author is pointless since there is no problematic use if punctuation were used to clearly separate the last three words from the sub-clause. Furthermore, in the case of John 12:40, the use of one comma 47 48

Emanuel Tov, Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert (Leiden: Brill, 2004), 7–8. Alain Kirk, “The Scribe as Tradent,” in Scribal Practices and Social Structures Among Jesus Adherents: Essays in Honour of John S. Kloppenborg, ed. William E. Arnal et al., BETL 2285 (Leuven: Peeters, 2016), 97–115, there 108.

Missing the Point

57

instead of a full stop is an example that the editors might possibly become usurpers of an authorial function. This, in consequence, forces the interpreters of the text to fault the author for something he neither did nor intended. It is not the fault of the author that others assume that he/she is an “ignorant author.” If editors or interpreters use the concept of an “ignorant author” to rephrase the text this is not the fault of the author. The change in mode, tense, and person would seem to speak against the inclusion of the last three words of John 12:40 into the negated sub-clause. This appears to hold true also in cases like John 8:25 where the Greek of the Gospel of John appears to exhibit characteristics of a sociolect. A central characteristic of its language seems to be Semitic influence. If punctuated with disregard to the peculiarities of Greek under Semitic influence the sentence structure of John 8:25 as constructed in critical editions may and must be deemed “incomplete.” Arguably, editions could refrain from using punctuation at all but this is rather impractical. Consequently, this contribution would like to raise awareness that punctuation is an editorial decision designed to guide the reader. This function might be compared to electronic devices leading vehicles to their destination. In most cases these devices are very helpful. However, rare exceptions have been documented where such devices have misled drivers blindly trusting them. There are similar cases. James Hardy Ropes saw some “scribal mistakes” of Codex Bezae Cantabrigiensis as proving the scribe “ignorant.”49 Well, the “scribal errors” can be understood to be a textual variant in case one were willing to allow the scribe to actually know his own tongue better than modern scholars and to deal creatively with the text.50 This paper has indeed pointed to quite a few instances were the victim is blamed for what editors forced the text to do. The work on the edition of the Greek New Testament started in the 16th century, a time when Jews were not well treated in Europe. In consequence, one should probably exclude the examples given in this contribution from those passages proving the author of John’s Gospel to be “intentionally hostile.” While acknowledging it is the well-meant intention of the critical edition to create an easily readable text, this motivation unwittingly introduced hostility within the following passages: John 8:25, 43, 45, and 12:40. It may be a matter of fact that it is hard to live with ambiguity; however, in light of the reception 49 James Hardy Ropes, The Beginnings of Christianity, Vol. 1/3: The Acts of the Apostles: The Text of the Acts (London: Macmillan, 1926), lxx. 50 Cf. Hans Förster, “‘An ignorant, if not a careless, scribe.’ Überlegungen zum Text von Apg 3,26 in der von Codex Bezae Cantabrigiensis (D 05) gebotenen Version,” Glotta 94 (2018): 153–162.

58

Förster

history of John’s Gospel it is important to avoid transforming a possibly ambiguous or polyvalent passage into a hostile text. Adele Reinhartz comments aptly, “… the gospel of love has also been an instrument of hate, not once, not occasionally, but frequently and pervasively in the history of Jewish-Christian relations.”51 In light of the examples in this paper one is tempted to suggest that a slight editorial emendation in Adele Reinhartz’s comment is called for. It is far more correct that the “gospel of love has … been made an instrument of hate.” As argued, it is not always the long-deceased author of a text with a long reception history who is to fault as to how his or her text has been understood. All examples collected for this essay concern the Gospel of John and were recently identified during text-critical research in cooperation with the Editio Critica Maior of the Gospel of John. The decisive influence of the punctuation on the interpretation and understanding of the text of John’s Gospel raises the question as to how many other instances might be waiting in Novum Testamentum Graece to be identified as originally ambiguous sentences that could be understood differently if traditional punctuation is challenged.52 Thus, modern punctuation might be an obstacle for the task of a translator as described by Umberto Eco: “Translators are duty-bound not to say more than the original text….”53 It appears that systematic research into the punctuation of the critical edition(s) of the New Testament might open new perspectives for the understanding of this important corpus of literature. 51 Adele Reinhartz, “The Grammar of Hate in the Gospel of John: Reading John in the Twenty-First Century,” in Israel und seine Heilstraditionen im Johannesevangelium: Festgabe für Johannes Beutler SJ zum 70. Geburtstag, ed. Michael Labahn, Klaus Scholtissek, and Angelika Strotmann (Paderborn: Ferdinand Schönigh, 2004), 416–427, there 416. 52 For an example of an “antisemitic punctuation” in Paul’s letter to the Thessalonians, cf. Frank. D. Gilliard, “The Problem of the Antisemitic Comma between 1 Thessalonians 2.14 and 15,” NTS 35 (1989): 481–502; Frank. D. Gilliard, “Paul and the Killing of the Prophets in 1 Thess 2:15,” NovT 36 (1994): 259–270; and Stanley E. Porter, “Translation, Exegesis and 1 Thessalonians 2.14–15: Could a Comma Have Changed the Course of History?” BibTrans 64 (2013): 82–98. 53 Umberto Eco, Mouse or Rat? Translation as Negotiation (London: Phoenix, 2003), 170.

CHAPTER 3

Pointers to Persons and Pericopes? A Study of the Intermarginal Signs in Sahidic Manuscripts of the Gospel of John Matthias H. O. Schulz 1 Introduction1 We are used to finding the text of the Bible divided into chapters and verses. Modern translations, editions of the text in Hebrew, Greek, or other old languages, manuscript descriptions in catalogues, and scholarly articles on the Bible use this division. It is almost a secret that the French printer Robert Estienne introduced the modern text division into chapters and verses only in the middle of the sixteenth century.2 For nearly 1,500 years, Christianity did not know our modern division that is so natural for us today. Older text divisions for the New Testament such as the sections of Ammonius of Alexandria, the canon tables of Eusebius of Caesarea (Eusebian canons as described in the Epistula ad Carpianum), or the kephalaia in Greek codices are today a marginal note even in text-critical editions. Other possibilities of textual division through paragraph marks or initials, which are perhaps limited only to individual manuscripts and individual textual traditions, remain unmentioned for the most part even in descriptive manuscript catalogues. However, it is precisely the intermarginalia that can illuminate the Sitz im 1 This independent research was originally planned as part of my PhD on text and music of the Coptic-Orthodox liturgy (supervisor: Ralf M. Jäger, Institute for Musicology, University of Münster), but I have excluded the chapter, since the amount of material was in no justifiable relation to the goals of my PhD. While working on the Coptic Gospel of John in projects led by Hans Förster, he encouraged me to resume the work in my spare time but limited to the Gospel of John. The complete data will be included as an appendix to the critical edition of the Sahidic Gospel of John, a project aided by the Austrian Science Fund (FWF-project P25082). This appendix will contain tables that include the positions of all paragraph marks, initials, and glosses of the manuscripts. 2 To be correct, the division into chapters goes back to Stephen Langton († 9th of July 1228), the division into verses to Estienne, see Joop H. A. van Banning S. J., “Reflections upon the chapter Divisions of Stephan Langton,” in Method in Unit Delimitation, ed. Marjo C. A. Korpel, Josef M. Oesch, and Stanley E. Porter (Leiden: Brill, 2007), 141–61, who gives an account on the introduction of the chapter-division before and in the time of Langton.

© Matthias H. O. Schulz, 2023 | doi:10.1163/9789004537972_005

60

Schulz

Leben of a manuscript. Intermarginal symbols can provide decisively clues as to how the text was read and used. The following remarks are limited to the manuscript tradition of the Gospel of John in Sahidic Coptic and minor Coptic dialects. Since paragraph marks and initials in Coptic manuscripts have not yet been examined for their function, this study is meant as a starting point for further research.3 All 173 manuscripts have been analysed, containing verses from the Gospel of John in Sahidic Coptic and minor Coptic dialects, that are currently listed in the SMR-online-database of the Institute for New Testament Textual Research, Westfälische Wilhelms-Universität Münster.4 Forty-two of them are lectionaries or liturgical typika. Not yet included in the online database are manuscripts in Proto-Bohairic (1), Lycopolitan (2), and Fayyumic (10) as well as two Sahidic pericope lists in the Coptic Museum, Cairo (Call.-No. 911 and CM 3804 (624)), and five further fragments. They will be added in the near future but are included in this study as well.5 Lectionaries and liturgical typika are not listed in this paper because for them as genuinely liturgical manuscripts the Sitz im Leben is known and a separate study seems necessary. However, lectionaries and liturgical typika are used for comparation. Some manuscripts, attesting verses of the Gospel of John in Sahidic and minor Coptic dialects, are not genuine Bible manuscripts but amulets, magical texts, or private copies or they do not show intermarginalia and initials, which may, first, be due to their fragmentary state of preservation: cw 1 || fa 16 [= GA 0260] || fa 19 (private copy) || fa 20 || fa 22 (palimpsest, scriptio inferior illegible on scans and photographs) || ly 1 || ly 2 || mae 5 3 A very useful work on the development of paragraph marks from the point of view of art history is Theodore Petersen, “The Paragraph Mark in Coptic Illuminated Ornament,” in Studies in Art and Literature for Belle da Costa Greene, ed. Dorothy Miner (Princeton, NJ: Princeton University Press, 1954), 295‒330. Petersen counts only obelos, diple, and coronis as paragraph marks attested in Coptic manuscripts. However, it is possible to distinguish the paragraphos (simple horizontal stroke) from the obelos, normaly accompanied by dots in Coptic manuscrits. 4 For sigla and bibliographical information, see http://intf.uni-muenster.de/smr/ (SchmitzMink-Richter-database). For the sigla of Greek and Greek-Coptic bilingual manuscripts (Gregory-Aland-sigla, GA) see Kurt Aland et al., eds., Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments. Zweite, neubearbeitete und ergänzte Auflage, 2nd ed., ANTF 1 (Berlin: De Gruyter, 1994). 5 Transcriptions and images of the Fayyumic manuscripts are available via the Institute for Textual Scholarship and Electronic Editing (ITSEE), University of Birmingham. A complete set of images of the Proto-Bohairic manuscript is included in Daniel B. Sharp, Papyrus Bodmer III. An Early Coptic Version of the Gospel of John and Genesis 1–4:2, ANTF 48 (Berlin: De Gruyter, 2016), 58‒321.

Pointers to Persons and Pericopes?

61

|| sa 7 || sa 8 || sa 22 || sa 56 || sa 60 || sa 66 (private copy) || sa 80 || sa 81 || sa 82 || sa 84 || sa 86 (Greek-Sahidic bilingual, no GA-number) || sa 89 || sa 90 || sa 163 || sa 168 || sa 177 || sa 178 || sa 179 || sa 180 || sa 184 || sa 185 || sa 193 || sa 194 || sa 195 (however, paragraph mark at Matt 25:1) || sa 196 || sa 205 || sa 207 || sa 209 || sa 215 || sa 216 || sa 217 || sa 218 || sa 219 || sa 231 (part of sa 299L) || sa 243 || sa 244 || sa 246 || sa 251 || sa 259 || sa 266 || sa 268 || sa 275 || sa 276 || sa 393var (amulet) || sa 402 (fragment with ἑρμηνεία, inner and outer margin lost) || sa 411 (private copy for liturgical use) || sa 452var (magical text) || sa 453 (quote in a homily) || sa 454 || sa 455 || sa 649 (amulet) || Cambridge, University Library, Michaelides Collection, No. 1282/1f || El-Ashmunein, Italian Archaeological Mission at Sheikh Abadah, Storeroom of the Supreme Council of Antiquities, Inv. Nr. 1017 Some further manuscripts have been excluded and are mentioned only occasionally in support to data collected from other manuscripts since they contain less than four initials or intermarginal signs and, therefore, do not provide enough data for any conclusion on their own.6 Most of these manuscripts are extremely fragmentary, having only one or two initials or paragraph marks: ac 1 [= GA P6] (Greek-Achmimic ms., languages alternately, fragmentary; paragraph marks at least at John 11:20, 11:45, and 11:55) || fa 15 (paragraph mark at John 1:6, most likely also at John *1:1) || fa 17 (kephalaion 1 at John 2:1, paragraph marks at John 2:6 and 2:17) || fa 18 (paragraph mark at John 3:1) || fa 21 (initial at John 17:10, starting with ⲁⲩⲱ ⲁⲓϫⲓ ⲉⲁⲩ) || sa 13 (paragraph mark at John 8:40) || sa 47 (nearly illegible, script mutilated and faded, at least initials at John 1:23 and with paragraph mark at John 2:1, paragraph mark at John 2:12) || sa 57 (paragraph mark at John 13:35) || sa 65 (initial at John 21:20) || sa 74 (*initial at John 18:5 starting at ⲛⲉϥⲁϩⲉ ⲇⲉ) || sa 83 (initials with diples at John 18:36 and 18:37) || sa 85 (initial with section number 26 at John 11:53) || sa 87 (initial with diple at John 16:30) || sa 88 (initial at John 3:1) || sa 137 (initial at John 3:16; however, more initials attested for the other gospels) || sa 143 (paragraph mark at John 20:11; however, more paragraph marks attested for the other gospels) || sa 149 (no image; however, paragraph marks in Mark 12:2‒5) || sa 159 (initial at John 8:12) || 6 Of course, a lower number of initials or paragraph marks does not mean that the marked verses of these manuscripts are less interesting – note e.g. sa 57 (paragraph mark at John 13:35: “By this shall all men know that ye are my disciples, if ye have love one to another” [KJV]), which might point to a liturgical context.

62

Schulz

sa 187 (paragraph mark at John 4:1) || sa 210 (paragraph marks at John 8:12 and 8:17) || sa 237 (initial at John 21:20) || sa 255 (initial at John 21:15) || sa 256 (initial at John 21:26) || sa 270 (initials with obelos-like ornaments at John 7:28, 7:30, and 7:32) || sa 271 (paragraph mark at John 21:1) || sa 370 (paragraph mark at John 14:31, starting with ⲙⲁⲣⲟⲛ ⲉⲃⲟⲗ ϩⲙ ⲡⲉⲓⲙⲁ) || sa 647 (paragraph mark at John 10:25) || Cairo, Coptic Museum, Call-No. 931 [G 809 Fragment 1] (ornamented initials at John 16:29 and *17:5) || Cairo, Coptic Museum, Call-No. 948 [G 805, Fragment 2] (ornamented initial at John *20:1) || Oxford, Bodleian Library, Copt. c.67 [formerly Copt. g.3] (not part of sa 109, initial at John 11:28) 2

Manuscripts Analysed

fa 6

[= GA 086] || palimpsest || scriptio inferior (John): dating unknown || White Monastery (Amélineau) || parchment, 18 fol. || 2 col. || c *18‒21 ll. || John 1:19‒26; 3:5‒4:19, 23‒35, 45‒49 || initials with paragraph marks

fa 7

9th‒10th c. (Boud’hors) || White Monastery (Amélineau) || parchment, 15 fol. (fragm.) || 2 col. || 34‒40 ll. || John 4:28‒34, 36‒40; 5:43‒53; 6:13‒15, 35‒37; 8:59‒9:3, 18‒21; 13:38‒14:3, 7‒10, 13‒17, 21‒23; 17:26‒18:3, 6‒10 || initials

pbo 4th c. (Kasser) || Dishna (Robinson) || papyrus, *84 fol. (Kasser) || 1 col. || c *21‒25 ll. (mostly 22 ll.) || John 1:1‒25, 40‒45; 2:9‒16; 3:33; 4:5‒7:52; 8:12‒21:25 || paragraph marks (decorative at the head of the written area of nearly all pages, but also marking specific passages even within the written area) sa 1

1st half 5th c. (Quecke), 5th c. (Orlandi) || provenance unknown || parchment, 230 fol. || 2 col. || 23 ll. || John 1:1‒5:3, 5‒7:52; 8:12‒9:37, 39‒16:14, 16‒21:25 || paragraph marks, initials, citations marked with diples

sa 3

5th c. (Luft), 6th c. (Beltz, Crum), 6th‒8th c. (Schubart) || provenance unknown || parchment, 18 fol. || 1 col. || 17‒21 ll. || John 1:1‒3:2; 3:10‒36; 4:1‒29, 39‒47; 5:36‒47; 6:1‒9 || paragraph marks

sa 4

c 600 (Thompson, Buchthal / Kurz), 9th c.? (Kasser) || Monastery of St. Jeremiah, Saqqara || parchment, 201 fol. || 1 col. || 29 ll. || John 1:1‒5:3,

Pointers to Persons and Pericopes?

63

5‒7:52; 8:12‒21:25 || paragraph marks, initials, citations marked with diples, numbering of “Amen, amen”-sayings (1‒25) sa 5

6th c. (Hintze / Schenke), c 600 (Buchthal / Kurz), 7th c. (Thompson); 9th c.? (Kasser) || Monastery of St. Jeremiah, Saqqara || parchment, 168, fol. || 1 col. || 19‒28 ll. || John 1:1‒5:3, 5‒7:52; 8:12‒21:25 || paragraph marks, initials (few), citations marked with diples

sa 9

7th / 8th c.? (Petersen), 8th c. (Schenke), 8th‒9th c. (Cramer), 9th c. (Landschoot), 9th‒10th c. (Kasser) || Harabt Hamuli || parchment, 113 fol. (*127) || 2 col. || 36‒39 ll. || John 1:1‒5:3, 5‒7:52; 8:12‒21:25 || paragraph marks, initials, kephalaia (No. 8 and 12 not indicated at the text, perhaps John 6:15 and 19:28)

sa 10

578 Anno Martyrum (colophon; Julian calendar: 29th of August 861‒ 1st of September 862) || Harabt Hamuli || parchment, 49 fol. || 2 col. || 33 ll. || John 1:1, 3‒35, *36, 37‒40, *41, 42‒2:3, 5‒4:3, *4, 5‒42, *43, 44‒5:2, *3, 5‒21, *22, 23‒39, *40, 41‒6:13, *14, 15‒19, 21‒32, *33, 34‒7:5, *6, 7‒20, *21, 22‒24, *25, 26‒41, *42, 43‒52; 8:12‒44, 46‒49, 51‒57, *58, 59‒9:13, 15‒10:1, *2, 3‒6, *7, 8‒29, *30, *31, 32‒39, 41‒11:18, *19, 20‒25, *26, 27‒34, 36‒44, *45, 46‒12:27, *28, 29‒36, *37, 38‒13:27, *28, 29‒15:11, 13‒19, *20, 21‒24, *25, 26‒16:13, 15‒17, *18, 19‒17:16, 19‒21, 23‒18:6, 8‒22, *23, 25‒38, *39, 40‒19:14, 16‒22, *23, 24‒25, *27, 28‒29, 31‒34, *35, 36‒42; 20:1‒22, 24‒21:2, *3, 4‒9, *11, 12‒14, *15, 16‒19, 21‒23, 25 || initials (enlarged and ornamented), section-Nos. (perhaps by second hand), pericope mark (dotted paragraph mark and double cross, John 19:13‒22), citations marked with diples, kephalaia (recent, written with pencil, not visible in the facsimile edition)

sa 12

5th‒6th c. (Bethge) || provenance unknown || papyrus, 110 fragm. (60 identified) || 2 col. || c *30 ll. || John 1:8‒9, 12‒14, 22, 23, 26, 27, 33, *34, *35, 36, 39‒40, 42, *43, *44, 45, 46, *47, 48, 49, *50, 51; 2:1, 5, 6, 9, 10, 12, 15, 16, 18, 19, 22, 23; 3:1, 2, 4, 6‒8, 10, 11, 13, 14, 16‒17, 19, 20, 23, 24, 26, 27, 29, 32, *33, 36; 4:1, 5, 6, 9‒10, 10‒12, 14, 15, 19‒20, 22, 24‒25, 28, 29, 33, 34, 36, 37, 39, 40, 43, 44‒45, 46‒47, 50‒51, 54; 5:1, 5, 6, 9, 13, 16‒18, 19, 22, 24‒25, 27‒28, 30, 34‒35, 36‒37, 39; 8:58, 59; 9:1, 4, 5, 8, 11‒12, *16, *18, *19, 20, 22, 25, 28, 29, 33; 10:*12, *13, 16, 19‒21, 33‒36; 11:4‒5, *6, *7, 30‒31, *41‒*43, 45, 46, 48‒49, *52, 54‒55, 57; 12:1, 2‒4, 8‒9, 12, 13, 16, 17, 24, 28, 29, 32,

64

Schulz

36, 40, 42‒44, *47; 13:37, 38; 14:10, 11, 13‒15, 23‒24, 27; 15:5, 6, 9, 10, 19 || paragraph marks, initials, section-Nos. sa 19

493 Anno Hijra (colophon; Julian calendar: 17th of November 1099‒ 6th of November 1100) || Naqlun, Fayyum || paper, 72 fol. || 1 col. || 23‒26 ll. || J 1:1‒5:3, 5‒7:52; 8:12‒10:41, 43‒12:9, 11‒44, 46‒21:25 || initials

sa 40 9th / 10th c. (Schüssler) || White Monastery, Sohag (Hyvernat) || parchment, 7 fol. (fragm.) || 2 col. || *45 ll. || John 1:1‒7, 17‒25, 31‒41, *42, 45‒2:1 || initials sa 101 6th / 7th c. (Crum, Hyvernat), 8th c. (Till, Balestri) || White Monastery, Sohag (Amélineau) || parchment, 24 fol. || 2 col. || 32 ll. || John 1:11‒12, 15; 4:23‒25, 28‒29, 33‒35, 37‒40, 44‒45, 47‒49, 52‒54; 5:3, 5‒7, 18‒22, 32‒34; 6:11, 24; 7:6‒10, 12, 17, 20‒26, 28, 32‒33, 35‒37; 8:35‒37, 44‒46; 10:*11, *12, *13, 14‒16, *17, *18, 19‒29; 12:39‒43, 47‒48, 50‒13:1, 3‒5 || paragraph marks, initials, citation marked with diples sa 102 6th or 7th c. (Balestri, Cramer), 7th c. (Horner, Till), 8th or 9th c. (Crum, Hyvernat) || White Monastery, Sohag (Amélineau) || parchment, 49 fol. (partly fragm.) || 2 col. || 32‒35 ll. || John 4:17‒36; 9:15‒10:7; 16:31‒32; 17:2‒3, 7‒8, 11 || paragraph marks, initials, end-mark sa 103 [part U = GA 0299] || 8th c. (Beltz), 8th / 9th c. (Crum, Hyvernat), 10th c. (Balestri), 10th / 11th c.? (Aland for part U), 11th c. (Till), 11th / 12th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 39 fol. || 2 col. || 31‒41 ll. || John 1:*7, 13‒14, 18‒19, *23, *24, *25, 32‒34, 38‒40; 4:42‒48, 50‒5:3, 5‒11; 6:12‒47, 50‒7:18, 41‒52; 8:12‒23; 9:7‒11:52; 12:38‒13:31; 14:29‒16:10, 29‒17:15; 18:23‒20:30; 21:15‒25 || paragraph marks, initials, glosses (liturgical references) sa 105 [= GA 070 +[0110, 0124, 0178, 0179, 0180, 0190, 0191, 0193, 0202]] || 6th c. (Aland), 7th / 8th c. (Woide), 8th / 9th c. (Horner), 9th c. (Till) || White Monastery, Sohag (Amélineau) || parchment, 44 fol. (partly fragm.) + 1 fragm. || 2 col. || 35 ll. || John 3:18‒20; 5:13‒22, 24‒25; 6:63‒7:1, 42‒52; 8:12‒13, 22‒9:28; 11:*41, *42, *43, *44, 45‒48; 12:19‒21, 36‒46; 16:28‒29 || paragraph marks, initials

Pointers to Persons and Pericopes?

65

sa 106 8th c. (Horner, Petersen), 8th or 9th c. (Balestri), 9th c. (Till) White Monastery, Sohag (Amélineau) || parchment, 23 fol. (partly fragm.) + 1 fragm. || 2 col. || 30‒32 ll. || John 1:1‒6; 2:13‒3:18; 8:38‒9:28; 10:2‒16; 12:34‒13:8; 16:2‒11, 13, 20‒23; 20:26‒31; 21:2, 4‒7 || paragraph marks, initials sa 108 10th / 11th c. (Till), 11th c. (Horner), c 11th c. (Balestri) || White Monastery, Sohag (Amélineau) || parchment, 20 fol. (fragm.) || 2 col. || 27‒30 ll. || John 9:22‒40; 10:*40, *41; 11:3, 10‒12, 15‒17, 20‒22, 26‒28, 31‒32, 34‒37, 40‒42, 44‒45, 47‒48, 51‒53, 55‒56, 57‒12:1; 13:2‒15:20; 16:10‒19:38 || initials, section-Nos. sa 109 10th c. (Horner, Balestri, Till) || White Monastery, Sohag (Amélineau) || parchment, 26 fol. || 2 col. || 30‒33 ll. || John 8:54‒9:18; 15:11‒12, 16‒17; 16:2‒6, 9‒13 || fragment Oxford, Bodleian Library, Copt. c.67 (formerly Copt. g.3) does, against Horner, not belong to this ms. (John 11:27‒29, 32‒34, initial at John 11:28) || paragraph marks, initials, section-No., gloss sa 110 6th c. (Beltz), 10th c. (Balestri, Horner, Till), 10th / 11th c. (Crum, Hyvernat) || White Monastery, Sohag (Amélineau) || parchment, 12 fol. (partly fragm.) || 1 col. || 23‒27 ll. || John 1:25‒2:18; 3:27‒4:3; 8:37‒46; 9:12‒22; 11:22‒28, 30‒48; 18:38‒40; 19:4‒6, 18‒27; 21:2‒14 || paragraph marks (obeloi), initials, section-Nos. sa 112 11th c. (Schüssler according to Kraus), 11th / 12th c. (Till), 12th c. (Balestri), 13th and 14th c. (Horner for different parts of the ms.) || White Monastery, Sohag (Amélineau) || parchment, 57 fol. || 2 col. || 26‒35 ll. || John 3:22‒26, 29‒30, *31, 36; 4:1, 6‒9, 12‒36; 6:26‒46; 7:23‒52; 8:12‒23; 11:51‒52, 57; 12:1, 6‒7, 12‒13 || paragraph marks (obeloi, paragraphoi, and circles in a different silvery ink), initials sa 115 10th c. (Balestri, Crum, Hyvernat, Horner, Till) || White Monastery, Sohag (Amélineau) || parchment, 26 fol. || 2 col. || 34‒35 ll. || John 1:25‒4:42; 7:25‒45; 8:25‒44; 9:7‒27; 10:8‒11:10, 18‒19, 24‒25, 37‒57; 16:22‒17:8; 20:15‒21:2 || paragraph marks, initials (many, some only slightly ekthetic), section-Nos.

66

Schulz

sa 116 10th c. (Till), 11th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 57 fol. + 2 fragm. || 2 col. || 34‒40 ll. || John 1:23‒5:3, 5‒7:52; 8:12‒44; 9:6‒26; 11:42‒12:17, 19‒29, 29‒13:2; 14:10‒24, 26‒29; 18:36‒19:31, 40‒41, *42; 20:2‒4, 11‒21:17 || paragraph marks (coronis, obelos, and ornament), initials, section-Nos. (by second hand), citations marked with diples or dots, glosses sa 117 [= GA 029 +[0113, 0125, 0139]] || 4th c. (Georgi), 5th c. (Aland, Gregory, Hyvernat [for a part of the ms.]), 5th / 6th c. (Cavalieri, Irigoin, Tisserant, Treu), 6th c. (Balestri), 6th‒8th c. (Clark, Hyvernat [for a part of the ms.]), 10th c. (Amélineau) || White Monastery, Sohag (Amélineau) || parchment, 31 fol. + 1 fragm. || 2 col. || 27‒33 ll. || John 1:7‒11, 15‒23; 3:2‒10; 4:45‒52; 6:21‒58, 68‒7:52; 8:12‒23 || paragraph marks, initials sa 121 7th c. or older, 8th c., and 11th c. (Horner for different parts of the ms.), 7th / 8th c. (Crum, Hyvernat), 8th c. (Balestri, Beltz, Leipoldt), 9th c. (Till), 10th / 11th c. (Schüssler) || White Monastery, Sohag (Amélineau) || parchment, 48 fol. (partly fragm.) || 2 col. || 36‒38 ll. || John 4:27‒5:3, 5; 7:21‒22, 26‒34; 10:36‒37, 41‒11:1; 13:18‒20, 22‒26, 29‒31, 33‒36; 14:29‒30; 15:4; 16:*4, 5‒8, *9‒12, 15, *19, 21; 19:18‒35; 20:1, 5‒*6 || paragraph mark, initials sa 123 8th c. or older and 10th c. (Horner for different parts of the ms.), 10th c. (Crum, Hyvernat, Balestri, Till) || White Monastery, Sohag (Amélineau) || parchment, 28 fol. || 2 col. || 37‒49 ll. || John 6:71‒7:5, 10‒16, 21‒26, 30‒35; 8:33‒42; 9:2‒11; 15:3‒27; 16:1‒17 || paragraph mark, initials sa 124 11th c. (Horner, Balestri, Till) || White Monastery, Sohag (Amélineau) || parchment, 26 fol. || 2 col. || 27‒31 ll. || John 4:9‒28; 13:20‒14:1, *20, 21‒18:15, 18‒19, *31, 32‒19:24 || paragraph mark, initials, kephalaion-No., section-Nos. sa 125 8th / 9th c.? (Petersen), 10th c. (Till), 10th or 11th c. (Balestri), 11th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 17 fol. || 2 col. || 37‒40 ll. || John 7:29‒52; 8:12‒9:41 || paragraph marks (some by second hand), initials (with coronis, unless stated otherwise), section-Nos.

Pointers to Persons and Pericopes?

67

sa 126 10th / 11th c. (Crum, Hyvernat), 11th c. (Horner, Balestri, Till) || White Monastery, Sohag (Amélineau) || parchment, 8 fol. + 1 fragm. || 2 col. || 25‒29 ll. || John 1:24‒26, 29‒32, 33‒36, 38‒40, 42‒3:4, 27‒28; 4:9‒10; 8:16‒34; 9:31‒10:10; 12:48‒13:17; 14:*17, 18‒20, 29‒31; 17:13‒15, 18‒18:1 || paragraph marks (obeloi, coronis), initials sa 127 11th c. (Till), 12th c. (Balestri, Horner) || White Monastery, Sohag (Amélineau) || parchment, 28 fol. || 2 col. || 39‒44 ll. || J 8:48‒13:1; 17:11‒12; 18:3, *4, 5‒19:40; 20:5‒6, *7, 12‒14 || paragraph marks (obeloi, coronis), initials, glosses sa 131 6th c. (Amélineau), 6th‒7th c. (Kasser), 7th c. (Schüssler), 7th‒8th c. (Crum), 8th c. (Horner), 9th‒10th c. (Till) || White Monastery, Sohag (Amélineau) || parchment, 30 fol. (partly fragm.) || 2 col. || 35 ll. || John 11:46‒12:12, 16‒*18, 21‒24, 27‒29; 16:32‒17:1, *2, 3‒4, *5‒*6, 7‒19; 21:*13, 14‒25 || paragraph marks, initials (some ornamented), kephalaia-Nos., canon-Nos. sa 132 13th c. (Horner, Balestri) || White Monastery, Sohag (Amélineau) || parchment, 24 fol. || 1 col. || 19‒26 ll. || John 7:18‒28; 12:48‒13:8 || initials, kephalaion-No., section-No. sa 134 10th / 11th c. (Till), 11th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 22 fol. || 2 col. || 37‒40 ll. || John 1:1‒27; 7:29‒33, 39‒43, 48‒52; 8:12‒14; 9:19‒10:10; 11:40‒12:13 || initials sa 136 8th c. (Kasser), 9th c. (Till) || White Monastery, Sohag (Amélineau) || parchment, 18 fol. (partly fragm.) || 2 col. || 32 ll. || John 1:43‒51; 2:1‒4:9; 5:14‒30; 6:43, *44, 58, *61, *62, 63‒65, *66, 67, 68‒7:10, 28‒52; 8:12‒10:12; 11:50‒12:4, 6‒13, 16‒19, 21‒28, 30‒34, 36, 40‒41, 44, 46; 20:31; 21:16 || paragraph marks, initials sa 139 9th c. (Till), 9th‒10th c. (Kasser), 11th c.? (Horner) || White Monastery, Sohag (Amélineau) || parchment, 17 (partly fragm.) || 2 col. || 26‒32 ll. || John 1:2‒11, 13‒16, 18‒25, 28‒30, 33‒34, 38‒40; 3:23‒4:9; 5:*42, 43‒45, 47; 6:1‒13, 37‒38, 51‒52, 65‒7:10, 44‒52; 8:12‒20, 36‒49; 11:44‒12:2, 21‒29, 31‒38; 13:5‒15:5, 22‒16:16; 18:29‒19:3, 35‒36, 39‒42; 20:1‒7, *11, 12 || initials, section-Nos.

68

Schulz

sa 140 6th‒7th c. (Crum, Hyvernat), 8th c. and 10th c. (Horner for different parts of the ms.), 8th / 9th c. (von Lemm), 10th c. (Till) || White Monastery, Sohag (Amélineau) || parchment, 23 fol. || 2 col. || 36‒40 ll. || John 17:3‒19:23 || initials (only initial at John 19:7 not ornamented) sa 141 9th c. (Elanskaja, von Lemm), 10th c. (Till, Hyvernat), 10th / 11th c. (Schüssler), 11th c. (Horner), 11th‒12th c. (Kasser) || White Monastery, Sohag (Amélineau) || parchment, 15 fol. (partly fragm.) || 1 col. || 19‒22 ll. || John 3:17‒21, 22‒27, 28‒4:5, 27‒39; 6:29‒41; 7:4‒18; 8:37‒40, 42‒44, 58‒9:1, 6‒7, 40‒10:11; 12:25‒35, 48‒13:9; 14:17‒18, *19‒20, 22‒24; 16:1‒24; 18:12‒22, 40‒19:11 || paragraph marks, initials sa 142 palimpsest || scriptio superior (John): 10th (Kasser), 11th c. (Till), 12th / 13th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 15 fol. || 1 col. || 19‒24 ll. || John 4:6‒11, 12‒17, *18; 9:28‒30, *31, 34‒38; 10:16, 21‒11:6, 34‒46, 48, *49, *50, *53, 54‒55; 12:8‒18, 36‒37, 41‒42; 13:7‒38; 14:1‒2, 3‒20; 17:11‒13, *14, 15‒20; 18:36‒37, 40; 19:1‒4, *24, 25‒26, *27, *29, 30‒31 || initials, section-Nos. sa 154 10th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 7 fol. || 2 col. || 29 ll. || John 4:36‒38, *39, 40‒41, 44‒45, 47‒49, *50; 11:56‒57; 12:1, 2‒3, 5‒7, 9‒11, 19, 23‒*24, 40‒42, 43‒46, 47‒49, 50; 13:1‒2, *17, 18‒31; 16:*15, 16‒27, 30, *31, 33; 17:3‒4 || paragraph mark, initials, liturgical glosses (by second hand) sa 155 10th c. (Till), 11th c. (Balestri, Horner) || White Monastery, Sohag (Amélineau) || parchment, 6 fol. (partly fragm.) || 2 col. || 27‒32 ll. || John 6:35‒37, 39‒42, 44‒46, 48‒51; 8:23‒35, 37‒39; 11:32‒33, 36‒39; 18:31‒19:4, 17‒20:2 || initials, section-Nos. sa 162 10th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 5 fol. (fragm.) || 2 col. || 32‒36 ll. || John 6:61‒7:1, 3‒8, *9, 10‒23, *24, *25, 26‒28, *29‒*30, 31‒32, *33‒*34, 35‒40; 15:6‒10, 13‒16, 18‒20, 22‒26 || paragraph marks, section-Nos. sa 167 13th c. (Horner) || White Monastery, Sohag (Amélineau) || parchment, 4 fol. (partly fragm.) || 2 col. || 35‒42 ll. || John 1:47‒4:22; 5:24‒6:5 || initials, section-Nos.

Pointers to Persons and Pericopes?

69

sa 173 c 10th c. (Balestri), 10th c. (Horner) || White Monastery, Sohag (Orlandi) || parchment, 3 fol. || 1 col. || 27‒28 ll. || John 9:11‒27; 10:25‒11:2; 12:23‒39 || initials sa 174 11th c. (Balestri, Horner) || White Monastery, Sohag (Orlandi) || parchment, 3 fol. || 2 col. || 28‒30 ll. || John 2:2‒20, 23‒3:6; 9:3‒10:1 || initials, section-Nos. sa 181 10th c.? and 11th c. (Horner for different parts of the ms.) || provenance unknown || parchment, 7 fol. (fragm.) + 7 fragm. || 2 col. || 40‒42 || John 1:4‒12, 13‒17, 19‒25, 29‒33, 37‒42, 44‒49, *50, 51‒2:8, 10‒15, 19‒24; 3:2‒6, 8‒15, 18‒23, 27‒31, 34‒4:5, 9‒13, 18‒22, 25‒30, 34‒38 || paragraph mark, initials (data according to edition by von Lemm) sa 182 palimpsest || scriptio inferior: 10th c. (Balestri), 11th c. (Crum); scriptio superior (John): 13th c. (Horner), 13th / 14th c. (Balestri) || White Monastery, Sohag (Orlandi) || parchment, 6 fol. || 2 col. || 21‒28 ll. || John 12:36‒13:2; 19:38‒21:23 || initials, section-Nos. (by second hand) sa 190 12th c. (Horner) || White Monastery, Sohag (Schüssler) || parchment, 3 fol. (fragm.) || 2 col. || *35‒38 ll. || John 2:14‒15; 3:14‒16, 17‒18; 4:7‒9, 14‒21, 23‒31, 32‒38, 40‒46 || initials sa 197 10th c. (Balestri, Horner) || White Monastery, Sohag (Orlandi) || parchment, 2 fol. (fragm.) || 2 col. || 36‒39 ll. || John 7:35‒52; 8:12‒25; 10:21‒11:13 || initials, kephalaion-No. sa 199 11th c. (Horner) || White Monastery, Sohag (Schüssler) || parchment, 2 fol. (fragm.) || 2 col. || *35 ll. || John 11:27‒47; 19:18‒30, 31‒36 || paragraph mark, initials, obeloi at the end of phrases sa 234 11th c.? (Horner), 11th c. (Schüssler) || White Monastery, Sohag (Amélineau) || parchment, 1 fol. (fragm.) || 2 col. || 30‒31 ll. || John 7:18‒40 || paragraph marks, initials sa 235 11th c.? (Horner) || White Monastery, Sohag (Amélineau) || parchment, 1 fol. || 2 col. || 32‒33 ll. || John 18:3‒26 || initials

70

Schulz

sa 238 11th c. (Till), 11th‒12th c. (Kasser, Schüssler) || provenance unknown || parchment, 1 fol. || 2 col. || 30‒31 ll. || John 8:33‒54 || initials sa 250 7th c. (Schüssler) || provenance unknown || papyrus, 1 fol. (fragm.) || 2 col. || *32 ll. || John 7:*29, 30‒49 || paragraph marks, initials sa 260 11th c. (Schüssler) || White Monastery, Sohag (Schüssler) || parchment, 1 fol. (fragm.) || 2 col. || *max. 47 ll. || John 18:8‒23, 24‒38 || initials sa 285 14th c. (Horner) || provenance unknown || paper, 23 fol. || 1 col. || 17‒19 ll. || John 7:13‒19; 8:20‒33, 40‒11:32; 12:9‒25 || initials, kephalaia, sectionand canon-Nos. sa 286 13th‒14th c. (Burmester) || Deir Abu Maqar, Wadi Natrun (Schüssler) || parchment, 10 fol. || 1 col. || 14‒18 ll. || John 13:10‒15; 14:21‒26, 30‒16:13 || initials, section-Nos. sa 369 11th c. (Schüssler) || White Monastery, Sohag (Schüssler) || parchment, 1 fol. (fragm.) || 2 col. || c *47 ll. || John 20:4, 12‒26; 21:*1 || initials, section-No. CGC 8086 [= Cairo, Coptic Museum, CM 3770 (654); G 338b] || dating unknown || provenance unknown || parchment, 1 fol. (fragm.) || 2 col. || c 12‒15 ll. (*?) || John 2:12‒13, 15, 17‒18, *19, 20‒22 || Walter E. Crum, Coptic monuments. Catalogue général des antiquités égyptiennes du musée du Caire. Nos 8001–8741 (Cairo: Service des antiquités de l’Égypte, 1902), 27 No. 8086. || initials In Coptic manuscripts, paragraph marks occur mostly in the form of a coronis.7 Only in very few instances coronis- and paragraphos- or diple-shaped forms are used side by side, see e.g., sa 106 (fig. 3.5‒6). Often paragraph marks accompany initials as ornamental elements, e.g., sa 141, where small and enlarged initials are accompanied by diples (fig. 3.11). For other possible ways of ornamenting initials see sa 1 (fig. 3.1), sa 4 (fig. 3.2), sa 5 (fig. 3.3), sa 12 (fig. 3.13), sa 102 (fig. 3.4), sa 106 (fig. 3.5‒6), sa 112 (fig. 3.7), sa 115 (fig. 3.8), sa 116 (fig. 3.9), sa 127 (fig. 3.10). 7 See Petersen, “Paragraph Mark,” 297. According to Petersen the coronis “offered the Coptic caligrapher his chief inducement and pretext for experimenting with further marginal decoration” (297).

Pointers to Persons and Pericopes?

71

In sa 127, nearly every initial is accompanied by either a small coronis or an obelos. It is not certain that all these paragraph marks have a specific meaning (fig. 3.10). It is noteworthy that some pages show only initials with coronis (e.g. Cairo, Coptic Museum, CM 2710r), others only initials with obelos (e.g. Venice, Bibliotheca Marciana, Ms. Or. 192 (9) Fragmenum II (Nan.), f. 14r). However, at least the position of the obeloi might have a special relevance since some obeloi precede initials or are placed up to two lines beneath initials heading parts of verses in a meaningful manner.8 For the time being, a satisfying interpretation cannot be given for the positions of many paragraph marks in sa 127. The size of initials can vary to a high degree, compare sa 1 (fig. 3.1) with sa 5 (fig. 3.3).9 In some manuscripts, the difference in scale of initials as compared to normal size letters of the written area is so fluid that a systematisation seems to be arbitrary. In particular, some rather unskilfully written manuscripts such as sa 108, sa 115 (fig. 3.8), or sa 123 require a separate investigation, which can not be done in this article. Especially sa 115 is so nuanced in terms of ekthesis and size of initials that it is often impossible to say what is accidental and what is intentional and thus perhaps meaningful in interpretation (see fig. 3.8). 3 Analysis It is not intended to try to reconstruct one or more systems of the original text division for the Gospel of John in Sahidic Coptic or some other Coptic dialects.10 Since most of the manuscripts are preserved fragmentarily, such an attempt is prohibitive from the outset. We are dealing with sources written over a timespan of more than a millennium. It is very likely that different text divisions have developed over time. Furthermore, the number of preserved manuscripts varies considerably for each verse, as can be seen in the manuscript list. 8

In sa 127 obeloi are sometimes placed between the lines making it rather hard to say, which sentence, verse, or part of speech they indicate. 9 Note that verses follow each other directly if there is space left in a line at the end of a verse. In such cases, it is normal for the first letter of the following line to be in ekthesis or ornamented even if a word is separated. 10 Since the function of numbering systems (kephalaia, section-Nos., Eusebian canon-/ section-Nos.) is self-evident these features of the manuscripts will not be discussed. At least for the section-Nos. that are the normal dividing system of the text in Bohairic manuscripts, see Adolph Hebbelynck, “Les κεφαλαία et les τίτλοι des Évangiles dans les MSS: Bohaïriques Paris Bibl. Nation. Copte 16, Vat. Copte 8 et le fragm. Bohaïrique Brit. Mus. Add. 14740a fol. 9. La Lettre d’Eusèbe à Carpianus d’après le MS. Bohaïrique Vat. Copte 9. Textes inédits et traduction”, Les Muséon 41 (1928): 81–120. For the Gospel of John, see especially pp. 100–102, 109–110.

72

Schulz

Only sa 1, sa 4, sa 5, sa 9, and sa 19 are complete copies of the Gospel of John and only pbo, sa 10, sa 103, sa 106, sa 110, and sa 116 preserve the major part of the text. Interestingly, these codices – except sa 19 and sa 116 – are assumed to be among the oldest surviving manuscripts. Even if the palaeographical dating of Coptic manuscripts has hardly yielded satisfactory results so far (compare the divergent opinions in the manuscript list above), it can still be seen as a tendency that the number of initials and ornamentations increases the younger a manuscript is dated (partly proven by colophons). However, an assignment to a specific century should be avoided until a comprehensive, reliable palaeography has been developed in Coptology.11 Even if the fragmented condition of most manuscripts permits only the identification of general tendencies, they help to gain a better understanding for the reasons of specific ways of text divisions. For practical reasons, it is best to start with those manuscripts that highlight or emphasise only a small number of verses and thus forming a loose group. To serve as examples, the verses are analysed that are marked by paragraph marks or ekthesis/initials in sa 1, sa 4, sa 5, sa 9, and sa 10.12 Even if the dating of the manuscripts is not completely certain, none of them is assumed to be later than the 10th century: sa 1: John 1:6; 2:1; 2:12; 3:1; 3:22; 4:46; 5:1; 6:1; 7:31; 9:1; 10:22; 11:1; 13:1; 17:1; 18:1; 18:28; 21:113

11 See Bentley Layton, “Towards a New Coptic Paleography,” in Acts of the Second International Congress of Coptic Studies, Roma, 22–26 September 1980, ed. Tito Orlandi and Frederik Wisse (Rome: C.I.M., 1985), 149–158. 12 In a recent article, Stanley Porter analysed, inter alia, the verses marked in New Testament papyri, for the Gospel of John P66, P75, and P90; see Stanley E. Porter, “Pericope Markers in Some Early Greek New Testament Manuscripts,” in Layout Markers in Biblical Manuscripts and Ugaritic Tablets, ed. Marjo C. A. Korpel and Josef M. Oesch, Pericope 5 (Assen: Koniklijke van Gorcum, 2005): 161–176, especially 165 (P90), 167–168 (P66), 168–170 (P75). Nearly all verses marked in the three papyri reoccur in the manuscripts mentioned below; for missing verses see, e.g., pbo, sa 101, sa 105, sa 115, and especially sa 116. The verses in the papyri are: GA P66: *1:6; 1:24; 1:28; 2:11; 2:23; 3:22; 4:1; 7:1; 10:6; 10:22; 11:53; 12:14 GA P75: 1:6; 1:29; 1:35; 1:43; 2:1; 2:23; 3:1; 3:22; 4:1; 4:5; 4:31; 4:43; 6:1; 6:3; 7:37; 7:40; 8:1; 8:28; 8:30; 8:51; 9:1; 10:19; 11:1; 12:30 GA P90: 18:37 (2×, beginning and ἀπεκρίθη ὁ Ἰησοῦς); 18:40; 19:1. 13 Compare Ch. Hill, The First Chapters: Dividing the Text of Scripture in Codex Vaticanus and Its Predecessors (Oxford: Oxford University Press, 2022), table 7.4–the divisions marked by paragraphoi in sa 1 match nearly precisely the numbered chapter divisions in Codex Vaticanus (GA B 02).

Pointers to Persons and Pericopes?

73

sa 4: John 1:1; 1:14; 1:29; 1:35; 1:43; 2:1; 2:12; 3:1; 3:11; 3:22; 4:43; 4:46; 5:1; 5:24; 5:25; 6:3; 6:15; 6:59; 7:1; 7:14; 7:28; 7:37; 8:12; 8:21; 8:51; 8:56; 9:1; 10:1; 10:7; 10:11; 10:16; 10:22; 11:1; 11:18; 11:38; 12:1; 12:3; 12:12; 12:20; 12:24; 12:35; 12:44; 13:1; 13:20; 13:33; 14:12; 14:18; 14:21; 14:27; 14:31 (at ⲧⲟⲩⲛⲧⲏⲩⲧⲛ); 15:7; 15:11; 16:20; 16:23; 17:1; 17:14; 18:1; 18:28; 19:25; 19:28; 19:41; 20:1; 20:11; 20:19; 20:21; 20:24; 20:26; 21:1; 21:15; 21:18 sa 5: John 1:1; 1:29; 1:35; 2:12; 2:13; 3:1; 3:22; 3:29: 4:1; 4:46; 5:1; 5:24; 5:25; 6:1; 6:15; 6:41; 6:47; 7:1; 7:37; 8:12; 8:45; 9:1; 10:22; 11:1; 12:12; 13:1; 13:21; 13:33; 13:38; 17:1; 18:1; 18:28; 20:1; 20:26; 21:1; 21:18 sa 9: John 1:1; 1:6; 1:15; 1:18; 1:28; 1:29; 1:35; 1:43; 2:1; 2:12; 2:23; 3:1; 3:16; 3:22; 3:25; 4:1; 4:5; 4:31; 4:40; 4:43; 4:46; 5:1; 5:17; 5:19; 5:21; 5:25; 5:45; 6:1; 6:15; 6:22; 6:27; 6:32; 6:34; 6:47; 6:59; 6:70; 7:1; 7:2; 7:9; 7:14; 7:21; 7:25; 7:28; 7:32; 7:37; 7:50; 8:12; 8:20; 8:30; 8:31; 8:34; 8:44; 8:51; 9:1; 9:4; 10:1; 10:6; 10:7; 10:11; 10:19; 10:22; 11:1; 11:49; 11:54; 11:55; 12:1; 12:4; 12:12; 12:20; 12:44; 13:1; 13:16; 13:20; 13:21; 13:33; 14:1; 14:5; 14:8; 14:12; 14:18; 14:26; 14:27; 15:1; 15:11; 15:26; 16:1; 16:4; 16:20; 16:23; 16:25; 16:33; 16:23; 17:1; 18:1; 18:4; 18:10; 18:15; 18:25; 18:28; 19:23; 19:28; 19:38; 20:1; 20:18; 20:19; 20:24; 20:26; 21:1; 21:14; 21:18 sa 10: John 1:1; 1:6; 1:15; 1:19; 1:28; 1:35; 1:38; 1:43; 2:1; 2:12; 2:23; 3:1; 3:9; 3:16; 3:22; 4:1; 4:5; 4:21; 4:31; 4:43; 4:46; 5:1; 5:19; 5:24; 5:31; 6:1; 6:15; 6:24; 6:47; 6:59; 6:70; 7:14; 7:37; 7:46; 8:30; 9:1; 10:1; 10:11; 10:19; 10:22; 11:1; 11:54; 12:1; 12:12; 12:14; 12:20; 12:29; 12:36; 13:1; 13:21; 14:1; 14:26; 14:27; 15:7; *15:26; 17:1; 18:1; 18:15; *18:28; 19:1; 19:13; 19:17; 19:38; 20:19; 21:1; 21:15 If we compare the data, it is obvious that each manuscript marks a different selection of verses, even if many verses reoccur. Furthermore, there are quite a few verse markings at positions similar to ancient Greek chapter breaks, which might lead to the conclusion that the Sahidic textual tradition follows text divisions attested in Greek manuscripts. However, a comparison with starting points of pericopes in Sahidic liturgical manuscripts, i.e., lectionaries and liturgical typika, points in a different direction. Since the liturgical manuscripts were written over a period of more than a millennium as are the continuous-text manuscripts, there is, of course, a high degree of uncertainty. In order to determine whether the text-dividing function of initials or paragraph marks in continuous-text manuscripts could have a liturgical function, the following lectionaries and liturgical typika with lections from the Gospel of John were used for comparison:

74

Schulz

sa 14L || sa 15L || sa 16L || sa 58L || sa 59L || sa 264 (list of pericope initials by second hand) || sa 291L || sa 292L || sa 293L || sa 294L || sa 295L || sa 296L || sa 297L || sa 298L || sa 300L || sa 301L || sa 302L || sa 305L || sa 306L || sa 308L || sa 310L || sa 313L || sa 317L || sa 318L || sa 324L || sa 326L || sa 330L || sa 335L || sa 336L || sa 337L || sa 339L || sa 340L || sa 342L || sa 343L || sa 347L || sa 349L || sa 350 || sa 352 || sa 396L (liturgical typikon) || sa 407L || sa 408L || sa 411 || sa 440L || sa 663L (liturgical typikon) || Cairo, CM, Call.-No. 911, fragm. 2 [G 836] (list of pericope initials) || Cairo, CM, CM 3804 (624) [G 691] (list of pericope initials) Of the verses marked in sa 1, sa 4, sa 5, sa 9, and sa 10 the major part is attested in liturgical manuscripts as starting points of pericopes. In the following overview, testified verses are given with the manuscripts that have pericopes starting with them. Here are also included liturgical glosses of continuous-text manuscripts that indicate starting points of pericopes. However, the list is incomplete since only two fragmentary liturgical typika are included (sa 396L and sa 663L) and only three lists of incipits (sa 264; Cairo, CM, Call.-No. 911, fragm. 2 [G 836]; Cairo, CM, CM 3804 (624) [G 691]). In each case, only the beginning of a pericope is given, regardless of its end14 and the liturgical occasion, since only a much more detailed study can adequately take these aspects into account: John 1:1 (sa 15L, sa 293L, sa 295L, sa 411); 1:6 (sa 295L, sa 396L); 1:15 (sa 264); 1:19 (sa 264); 1:29 (sa 110); 1:35 (sa 14L, sa 264); 1:40 (sa 110); 1:43 (sa 15L, sa 396L, sa 110, sa 126, sa 264; see as well lac 1:45 in sa 297L); 2:1 (sa 14L, sa 15L, sa 110; see as well lac 2:5 in sa 326L and lac 2:6 in sa 294L); 2:12 (sa 295L, sa 264); 2:13 (sa 264); 2:18 (sa 294L); 2:23 (sa 264); 3:1 (sa 14L, sa 15L, sa 306L, Cairo, CM, CM 3804 (624) [G 691]); 3:22 (see lac 3,26 in sa 296L); 4:5 (sa 291L, sa 301L, sa 339L; see as well lac 4:6 in sa 310L); 5:1 (sa 291L, Cairo, CM, CM 3804 (624) [G 691]); 5:21 (sa 15L, Cairo, CM, Call.-No. 911, fragm. 2 [G 836]); 5:39 (sa 264); 6:1 (sa 347L); 6:15 (sa 294L, sa 301L); 6:22 (sa 116); 6:27 (sa 16L); 6:47 (see lac 6:47 in sa 318L); 7:1 (sa 103); 7:14 (sa 15L, sa 103, sa 291L, Cairo, CM, CM 3804 (624) [G 691]); 7:28 (sa 234); 7:37 (sa 302L, sa 305L; see as well lac 7:38 in sa 343L); 8:12 (sa 16L, sa 103); 8:17 (sa 264); 8:21 (sa 16L); 8:51 (sa 16L); 9:1 (Cairo, CM, Call.-No. 911, fragm. 2 [G 836]); 14

For a continuous-text manuscript marking the ends of pericopes compare the marks in sa 102 (fig. 3.4) at John 4:24; 9:38; 10:5. Most likely this is an abbreviation for τέλος. In the overview, pericopes are marked by lac in front of the verse number or after it, whose beginnings or endings are not preserved.

Pointers to Persons and Pericopes?

75

10:1 (sa 103); 10:7 (sa 15L; see as well 10:9 in sa 14L); 10:11 (sa 14L; see as well lac 10:12 in sa 339L and sa 407L); 10:22 (sa 115, sa 264; 10:23 in sa 16L); 11:1 (Cairo, CM, CM 3804 (624) [G 691]; see as well 11:5–9 in sa 324L); 11:54 (sa 264; see as well 11:55); 12:1 (sa 16L, sa 154, Cairo, CM, CM 3804 (624) [G 691]; see as well lac 12:6 in sa 296L); 12:12 (sa 14L, sa 15L, sa 301L, sa 345L, sa 440L); 12:20 (sa 15L); 12:24 (sa 663L); 12:29 (sa 396L); 12:35 (sa 16L, sa 305L; see as well 12:36); 12:44 (sa 103, sa 264); 13:1 (sa 16L); 13:20 (see 13:21 in sa 16L and sa 349L); 13:23 (sa 154); 13:33 (sa 16L); 14:26 (sa 16L, sa 297L, sa 336L; see as well 14:27); 14:31d (at ⲧⲟⲩⲛⲧⲏⲩⲧⲛ, sa 294L, Cairo, CM, CM 3804 (624) [G 691]; in sa 294L the pericope seems to occur twice [see lac 15:3]; in sa 14L at beginning of verse); 15:1 (see lac 15,3 in sa 294L); 15:7 (sa 15L; see as well lac 15:10 sa 299L); 15:26 (sa 14L, sa 15L, sa 103, sa 297L, sa 299L); 16:20 (sa 59L); 17:1 (sa 15L); 17:4 (sa 14L, sa 103, sa 294L, sa 302L; see as well sa 292L and sa 297L); 18:1 (sa 292L; see as well sa 16L and sa 292L); 18:10 (sa 16L, sa 127, sa 292L; gloss in sa 124); 18:15 (sa 16L, sa 292L); 18:28 (sa 16L, sa 127, sa 140, sa 292L; see as well lac 18:37 lac in sa 330L); 19:1 (sa 16L, sa 116, sa 127, sa 292L, sa 298L, sa 663L; see as well lac 19:7 in sa 58L); 19:13 (sa 16L; see as well sa 116 [fig. 3.9] and 19:14 in sa 140); 19:16 (sa 337L, sa 103, sa 124, sa 127, sa 352; see as well 19:15 in sa 140, lac 19:19 in sa 14L and lac 19:20 in sa 335L; see as well 19:17); 19:17 (sa 103; see 19:16); 19:23 (see lac 19,23 in sa 337L); 19:25 (sa 15L, sa 127; see as well lac 19:25 in sa 408L); 19:28 (sa 16L; see as well lac 19:30 in sa 294L); 19:38 (sa 16L, sa 127); 20:1 (sa 14L, sa 15L, sa 103, sa 313L, sa 335L, sa 340L, Cairo, CM, CM 3804 (624) [G 691]; see as well lac 20:2 in sa 308L); 20:19 (sa 15L, sa 663L; see perhaps as well sa 302L, but see also 20:18 and 20:21; see as well the Greek pericope lac 20,18–20 in sa 339L); 20:24 (sa 14L, sa 335L, sa 342L; see as well lac 20:25 in sa 340L and perhaps lac 20:28 in sa 336L); 21:1 (sa 293L; see as well lac 21:7 in sa 317L); 21:15 (sa 293L; see as well 21:14) For our selected manuscripts, only a small number of verses remains that are not attested otherwise: sa 1: John 4:46; 7:31 sa 4: John 1:14; 3:11; 4:43; 4:46; 5:24; 5:25; 6:3; 6:59; 8:56; 10:16; 11:18; 11:38; 12:3; 14:12; 14:18; 14:21; 15:11; 16:23; 17:14; 19:41; 20:11; 20:26; 21:18 sa 5: John 3:29: 4:1; 4:46; 5:24; 5:25; 6:41; 8:45; 13:38; 20:26; 21:18

76

Schulz

sa 9: John 1:18; 3:16; 3:25; 4:1; 4:31; 4:40; 4:43; 4:46; 5:17; 5:19; 5:25; 5:45; 6:32; 6:34; 6:59; 6:70; 7:2; 7:9; 7:21; 7:25; 7:32; 7:50; 8:20; 8:30; 8:31; 8:34; 8:44; 9:4; 10:6; 10:19; 11:49; 12:4; 13:16; 14:1; 14:5; 14:8; 14:12; 14:18; 15:11; 16:1; 16:4; 16:23; 16:25; 16:33; 16:23; 18:4; 18:25; 20:26; 21:18 sa 10: John 3:9; 3:16; 4:1; 4:21; 4:31; 4:43; 4:46; 5:1; 5:19; 5:24; 5:31; 6:24; 6:59; 6:70; 7:46; 8:30; 9:1; 10:19; 12:14; 14:1 It is possible that some points of contact can no longer be deduced today, since there are simply not enough liturgical manuscript sources preserved to give the starting points of all pericopes for all occasions or to determine with certainty the degree of variation for the beginning (and end) of a pericope. However, in some cases, continuous-text manuscripts might indicate further starting points of pericopes not preserved elsewhere. Other manuscripts, in which paragraph marks and initials are likely to have a liturgical meaning, are sa 3, sa 12, sa 40, sa 101, sa 102, the Greek-Sahidic bilingual sa 105, sa 110, sa 116, sa 117, the Greek-Sahidic bilingual sa 121 (all paragraph marks are at positions as in sa 1, sa 4, and sa 5, except for John 19:32, but see sa 16L; furthermore, eight initials are accompanied by paragraph-like ornaments, seven times at Jesus-sayings [John 5:19, 24, 25; 6:26, 32, 47, 53]), sa 181, and sa 190, perhaps as well sa 103, sa 155, and sa 162.15 For fa 6 and fa 7 a liturgical function of initials and paragraph marks is possible as well, but both manuscripts mark verses not highlighted in the Sahidic textual tradition. However, in Fayyumic Coptic only very few liturgical manuscripts are preserved, which prohibits any judgment. In the majority of the manuscripts, paragraph marks and initials occur more frequently and are used to structure the text much finer. It is only possible to determine their specific meaning if the text itself is included in a detailed analysis. Since the text is divided into sections too small to be explained as liturgical lections, the division follows obviously other criteria. As an example of a very clear structuring, sa 112 may serve (fig. 3.7). Obeloi and paragraphoi are marked by ÷, simple initials by #: 15 The following list is certainly incomplete, and an in-depth study is much needed. The data is based on clusters of paragraph marks in the manuscripts mentioned: John 1:14 or 1:15; 1:19; 1:27 (see as well 1:28 and 1:29 in sa 110); 2:23; 3:3; 3:5; 3:11; 3:16; 3:22 (see lac 3:26 in sa 296L); 4:1; 4:31; 4:43; 4:46 (see as well sa 103); 5:19 (see 5:21 in sa 15L); 5:24; 5:25; 6:3 (see 6:5 in sa 15L); 6:32; 6:34; 6:41; 6:43; 6:59; 7:1 (see sa 103 and lac 7:7 lac in sa 350); 7:28; 7:31; 8:56; 9:1; 10:1; 10:16 (see 10:14–21 in sa 16L); 11:18; 11:38; 12:3; 12:29; 12:44; 14:12; 14:18; 14:21; 14:27 (see as well 14:26); 15:11; 16:23; 17:14; 18:25 (see pericope end at 18:26 in sa 140); 19:31 (but see 19:28 and 19:32 in sa 16L); 19:41 (but see 19:38); 20:11; 20:21 (see perhaps lac 20:22 in sa 302L); 20:26; 21:18.

Pointers to Persons and Pericopes?

77

÷ John 3:25 (account), ÷ John 3:26–28 (account, afterwards the disciples of John the Baptist as well as John’s answer), ÷ John 3:29–30 (at ⲉⲧⲉ ⲡⲁⲓ ⲡⲉ, account), ÷ John 3:31 (John the Baptist), [John 3:32–4:6 lacuna], # John 4:7–8 (at ⲡⲉϫⲉ ⲓ︦︤ⲥ︦ ⲛⲁⲥ, Jesus) ÷ John 4:9 (Samaritan woman), [John 4:10–12 lacuna], ÷ John 4,13–14 (Jesus), ÷ John 4:15–16 (Samaritan woman, afterwards Jesus), ÷ John 4:16 (at ⲁⲧⲉⲥϩⲓⲙⲉ, Samaritan woman), [John 4:17–18 lacuna], ÷ John 4:19–20 (Samaritan woman), ÷ John 4:21–22 (Jesus), ÷ John 4:22–23 (at ⲁⲛⲟⲛ ⲇⲉ ⲉⲛⲟⲩⲱϣⲧ, Jesus), ÷ John 4:23 (at ⲉⲣⲉⲡⲉⲓⲱⲧ ⲅⲁⲣ ϣⲓⲛⲉ, Jesus), [John 4:24–28 lacuna], # John 4:28 (at ⲡⲉϫⲁⲥ ⲛⲛⲣⲱⲙⲉ, Samaritan woman), [John 4:29–30 lacuna], # John 4:31–33 (account, disciples, only briefly Jesus), # John 4:34–35 (Jesus), # John 4:35–36 (at ⲉⲓⲥ ϩⲏⲏⲧⲉ, Jesus), ÷ John 4:36 (at ⲁⲩⲱ ⲛϥⲥⲱⲟⲩϩ ⲉϩⲟⲩⲛ, Jesus), [John 4:37–6,25 lacuna], ÷ John 6:27 (at ⲁⲗⲗⲁ ⲧⲉϩⲣⲉ ⲛⲧⲟϥ, Jesus), ÷ John 6:28–29 (crowd in Capernaum, Jesus), ÷ John 6:30–31 (crowd in Capernaum), ÷ John 6:32 (Jesus), ÷ John 6:33 (Jesus), ÷ John 6:34 (crowd in Capernaum), ÷ John 6:35 (Jesus), ÷ John 6:36–37 (Jesus), ÷ John 6:37–39 (at ⲁⲩⲱ ⲡⲉⲧⲛⲏⲩ ϣⲁⲣⲟⲓ, Jesus), ÷ John 6:40 (Jesus), # John 6:41 (indirect speech), ÷ John 6:42 (Jews), ÷ John 6:43–44 (Jesus), ÷ John 6:45 (Jesus, OT-citation), ÷ John 6:45 (at ⲟⲩⲟⲛ ⲛⲓⲙ ⲛⲧⲁϥⲥⲱⲧⲙ, Jesus),

78

Schulz

[John 7:23–29, page skipped by scribe] # John 7:30–31 (account) # John 7:32 (account on Pharisees and chief priests) # John 7:33–34 (Jesus) ÷ John 7:35–36 (Jews) # John 7:37–39 (Jesus) ÷ John 7:40 (some) ÷ John 7:41 (others) ÷ John 7:42 (OT-citation) ÷ John 7:43 (account) # John 7:44 (some; scribe started ÷, but left it unfinished) ÷ John 7:45 (officers) ÷ John 7:45 (at ⲡⲉϫⲉ ⲛⲏ ⲛⲁⲩ, officers) ÷ John 7:46–47 (officers and Pharisees, perhaps till John 7:49) [John 7:48–49 ÷ lacuna] ÷ John 7:50–51 (Nicodemus) ÷ John 7:52 (they [some out of the crowd]) (John 7:53–8:11 not attested in Sahidic) ÷ John 8:12 (Jesus) ÷ John 8:13 (Pharisees) End of preserved text. It is obvious that the text division follows closely the biblical narrative, which coincides with the distribution of speech, emphasis on important sections, and delineation of explanatory parts. The most probable explanation is that the manuscript was meant for reading aloud, as such a fine division of the text is not necessary for silent reading. At least partially, the text division was already carried out by the copyist as part of the layout, as can be seen by the ekthetic initials. However, a second hand added marks in a different silvery ink. The same ink was used for ornaments in the upper margin and to decorate enlarged initials. However, it was used only for paragraph marks and punctuation in the Gospels of Luke and John, not for Mark and Matthew. Other manuscripts with paragraph marks or initials, which structure the text according to the distribution of speech, emphasis, or delineation of explanatory parts, are sa 106, sa 109 (also for the Gospel of Matthew), sa 115, sa 124, sa 125, sa 134, sa 139, sa 174, and CGC 8086. In sa 19, sa 103, sa 108, sa 123, sa 126, sa 127, sa 131, sa 132, sa 136, sa 140, sa 141, sa 154, sa 167, sa 173, and sa 199 initials and paragraph marks occur more irregular. It is possible, but not certain, that they have a similar function and serve to structure the text for reading as well. Even if the question of dating Coptic manuscripts has not yet been conclusively

Pointers to Persons and Pericopes?

79

clarified, it can be cautiously noted that the manuscripts are dated from the 8th to 14th century in the literature, with a focus on the 10th to the 12th century. Although it is only possible to speculate on how the lecture took place, the division is in favour of a lecture with role allocation, which might indicate a scenic performance of the biblical narratives. Little more can be said about the liturgical context than that the assigned different roles indicate longer excerpts than the short pericopes of the liturgy. For sa 142, sa 182, sa 197, sa 234, sa 235, sa 238, sa 250, sa 260, sa 285, sa 286, and sa 369 the data is inconclusive. However, a function in order to structure the text according to distribution of speech or emphasis is more likely than a function in order to indicate pericopes for the liturgy since to many paragraph marks occur. In a few cases paragraph marks, but numbers as well, are set to designate verses or perhaps text units, which are of importance for the interpretation of the text so that an exegetical function can be assumed. The most striking examples are sa 4 and pbo. In sa 4 (fig. 3.2), which was discussed above as part of the manuscripts with initials and paragraph marks of liturgical value, there is a special numbering system to designate all “Amen”-sayings by Jesus (KJV: Verily, verily, I say unto you …). The script of the numbers is extremely close to the script of the text and most likely by the same scribe. In total, twenty-five places are labelled.16 Even though it is not possible to say exactly why the sayings were numbered, a deeper meaning is unquestionable. A possible explanation might be that the codex was not only read liturgically, but also used for the creation of sermons. Two different systems of setting paragraphoi seem to exist in pbo, which, however, overlap and are not distinguished by their shape. Furthermore, the written area of nearly all pages is headed by a paragraphos, but this is certainly a decorative element or part of the page layout. Interestingly, paragraphoi occur even within the written area, not only on the margins. On the one hand, there are paragraphoi that mark section breaks – e.g., chapter breaks – but on the other hand, there are paragraphoi that denote the direct speech of Jesus or emphasise parts of it.17 The paragraphoi certainly served the user of the codex to find text passages more quickly, but why these specific verses have been marked is unknown, even if the choice suggests an underlying system. 16 17

This was noted, but not commented upon, by Franz-Jürgen Schmitz and Gerd Mink, Liste der koptischen Handschriften des Neuen Testaments 1: Die sahidischen Handschriften der Evangelien, 1. Teil, ANTF 8 (Berlin: De Gruyter, 1986), 9. See Sharp, “Papyrus Bodmer III,” 99, l. 5 (John 5:19; p. 31 of the manuscript), 100, l. 6 and 13 (John 5:24, 25; p. 32 of the manuscript), 107, l. 14 and 20 (John 5:45 at ⲁⲗⲗⲁ ϥϣⲟⲡ and 6:1; p. 35 of the manuscript), etc.

80

Schulz

Further examples are sa 106 and sa 108, in which verses are especially highlighted, which suggests a deeper meaning. In sa 106 (fig. 3.5‒6), there are three cases, in which paragraph marks are clearly distinguished from others by their shape and by small strokes covering them; they are found at John 3:3; 8:46; 9:27. Although the reason for marking these verses cannot be given, the combination is certainly not arbitrary.18 Also in sa 108, a single verse is particularly marked (John 17:9), but the reason is unknown. 4 Conclusion The text of the Bible in the Coptic tradition is structured by paragraph marks and initials. Based on the complete manuscript tradition of the Gospel of John in Sahidic and minor Coptic dialects, it becomes apparent that the function allows a grouping of manuscripts. The oldest manuscripts, according to dates given in the literature, have only relatively few marked text units and show a strong connection to starting points of pericopes in liturgical manuscripts. The uncertainties arising from the fragmentary condition of most manuscripts, as well as from dating codices, will have to be taken into account in future research. The question of whether the liturgical use caused the emergence of text divisions or whether existing text divisions were transferred into the liturgical sphere cannot be easily decided. Nonetheless, the markings of the oldest Sahidic manuscripts might tilt the scale pan in favour of an originally liturgical text division in Coptic, only later refined by numbering text units according to kephalaia or section numbers. In some cases, these numbers were added by a second scribe as in sa 12 (fig. 3.13) or sa 116 (fig. 3.9). Taking into account the manuscripts preserved in Sahidic, the text division according to the canonand section-numbers of Eusebius is of subordinate importance; only sa 131 and sa 285 (fig. 3.12) have them. In manuscripts dated rather late in the literature, the number of initials and paragraph marks increases. Even if it is not possible to give a definite conclusion for the position of every initial and every paragraph mark, there is evidence that they structure the text to highlight the distribution of speech, to put emphasis on important sections, and to delineate explanatory parts. The 18 John 3:3 (KJV): “Jesus answered and said unto him, Verily, verily, I say unto thee, except a man be born again, he cannot see the kingdom of God.” John 8:46 (KJV): “Which of you convinceth me of sin? And if I say the truth, why do ye not believe me?” John 9:27 (KJV): “He (man blind from birth) answered them, I have told you already, and ye did not hear: wherefore would ye hear it again? will ye also be his disciples?” See perhaps as well sa 57 (John 13:35) [excluded manuscript].

Pointers to Persons and Pericopes?

81

setting of initials and paragraph marks suggests that the manuscripts were set up for reading aloud, probably with allocated roles. With regard to the Greek textual tradition and its text divisions, the statements made here allow the question of how they are connected to the liturgical use of the text. Furthermore, especially for the papyri of the New Testament that are almost all from Egypt but as well for some uncial manuscripts, it is necessary to ask whether the reference to the Greek or so-called Byzantine tradition makes sense, or if rather an Egyptian context should form the framework of investigations as at least P66, P75, and P90 suggest.

Appendix: Pericopes of the Gospel of John in Sahidic Liturgical Manuscripts

The following list contains all Sahidic pericopes from John attested in Sahidic lectionaries and typikon-manuscripts currently listed in the SMR-online database of Coptic manuscripts; Greek and Arabic pericopes are not included. Since most of the manuscripts are fragmentary, a lacuna is indicated by lac before or after a verse number. For further information, see the SMR-online database. John

1:1–5 lac 1:1–13 1:1–14 1:6 1:6–9 lac lac 1:10–11, *12, 14 1:15 1:19 1:35 1:35–51 1:43 1:43–51 lac 1:45–46 lac 2:1–11 lac 2:5–11 lac 2:6–11 2:12

sa 411L sa 293L sa 15L | sa 295L sa 396L (incipit) sa 295L sa 14L sa 264 (incipit) sa 264 (incipit) sa 264 (incipit) sa 14L sa 264 (incipit) | sa 396L (incipit) sa 15L sa 297L sa 14L | sa 15L sa 326L sa 294L sa 264 (incipit)

82

Schulz

(cont.)

John

2:12–22 2:13 2:18–22 lac 2:23–25 lac 2:23 3:1 3:1–8 lac 3:1–11 3:1–15 lac 3:26–28 4:5–12 lac 4:5–30 4:5, 6–11, 13–15 lac 4:6–14 4:19–29 lac 4:19–20 lac 5:1 5:1–3, 5–14 5:21 5:21–29 5:33 5:39 6:1–3, *4, 5–14 6:5–14 6:8–9 lac 6:15–24 lac 6:15–26 6:27–58 lac 6:47–48 lac 6:62–69 lac 7:7 lac 7:14 7:14–17 lac

sa 295L sa 264 (incipit) sa 294L sa 294L sa 264 (incipit) Cairo, CM, CM 3804 (624) [G 691] (incipit) sa 306L sa 14L sa 15L sa 296L sa 339L sa 291L sa 301L sa 310L sa 293L sa 297L Cairo, CM, CM 3804 (624) [G 691] (incipit) sa 291L Cairo, CM, Call.-No. 911, fragm. 2 [G 836] (incipit) sa 15L sa 396L (incipit) sa 264 (incipit) sa 347L sa 15L sa 294L sa 301L sa 294L sa 16L sa 318L sa 293L sa 350 Cairo, CM, CM 3804 (624) [G 691] (incipit) sa 291L

Pointers to Persons and Pericopes? (cont.)

John

7:14–24 7:37–45 lac 7:37–46 lac lac 7:38–40, 42–44 lac 8:12–20 8:17 8:21–28 8:51–59 9:1 10:7, 11–18 10:9–18 10:11–18 lac 10:12–18 lac 10:12–15, 16–18 10:14–21 10:22 10:23–28 10:29–42 lac 10:29–30 lac 11:1 11:5–9 11:46–53 11:54 11:55–57 12:1 12:1–8 lac 12:6–8 12:12–19 12:12–28 12:12–36 12:12–17, 18–23 12:12–14, *15 lac 12:20–32

sa 15L sa 302L sa 305L sa 343L sa 16L sa 264 (incipit) sa 16L sa 16L Cairo, CM, Call.-No. 911, fragm. 2 [G 836] (incipit) sa 15L sa 14L sa 14L sa 407L sa 339L sa 16L sa 264 (incipit) sa 16L sa 16L sa 294L Cairo, CM, CM 3804 (624) [G 691] (incipit) sa 324L sa 16L sa 264 (incipit) sa 16L Cairo, CM, CM 3804 (624) [G 691] (incipit) sa 16L sa 296L sa 15L sa 14L sa 440L sa 301L sa 345L sa 15L

83

84

Schulz

(cont.)

John

12:24 sa 663L (incipit) 12:26–27 lac sa 300L 12:29 sa 396L (incipit) 12:35–42 sa 305L 12:35–43 sa 16L 12:44 sa 264 (incipit) 13:1–20 sa 16L 13:21–30 sa 16L | sa 349L 13:33–18:2 (parts: 13:33–14:25; sa 16L 14:26–15:25; 15:26–16:33; 17:1–26; 18:1–2) 14:26–15:1, 3–11 lac sa 297L 14:26–15:9 lac sa 336L 14:31–15:16 sa 14L 14:31d–1 Cairo, CM, CM 3804 (624) [G 691] (at ⲧⲟⲩⲛⲧⲏⲩⲧⲛ, incipit) 14:31d–15:5 sa 294L (at ⲧⲟⲩⲛⲧⲏⲩⲧⲛ) lac 15:3–4 sa 294L 15:7–17 sa 15L lac 15:10–25 sa 299L 15:15 sa 663L (incipit) 15:26–16:7 lac sa 299L 15:26–16:13 sa 15L 15:26–16:15 sa 14L 15:26–16:2, 4–9, 13–15, 19 sa 297L 16:20–23 lac sa 59L 17:1–11 sa 15L 17:4–10 lac sa 302L 17:4–17 sa 14L 17:4–5, 8–9, 11–12 sa 294L lac 17:6–26 sa 292L lac 17:8–9 lac sa 297L 17:17–26 sa 349L 18:1–2 sa 292L 18:3–9 sa 16L lac 18:6–9 sa 292L 18,10 lac sa 292L 18:10–14 sa 16L

Pointers to Persons and Pericopes? (cont.)

John

18:15–27 18:28–40 lac 18:37 lac 19:1–12 19:1‒16 lac 19:1, 16 lac 19:7–17 lac 19:13–27 19:16–17, 20–21 lac 19:16–24 lac lac 19:19–20, 22–23, *25, 27 lac 19:20–21, 23–24 lac 19:25–37 lac 19:25–27 lac 19:28–30 lac 19:30–31 lac 19:31–37 19:38–42 20:1–2 lac 20:1–18 20:1, 3–5, 7–9, 12–13, 15–18 lac 20:1 lac lac 20:2–3, 15–16 lac 20:8–18 lac 20:19 20:19–31 lac 20:22–23 20:24–31 lac 20:25–26, 29–30 lac lac 20:28–31 21:1–11 lac lac 21:7–12 lac 21:12 21:15–20

sa 16L | sa 292L sa 16L | sa 292L sa 330L sa 16L | sa 292L sa 298L sa 663L (incipit and desinit) sa 58L sa 16L sa 352 sa 337L sa 14L sa 335L sa 15L sa 408L sa 16L sa 294L sa 16L sa 16L sa 313L sa 14L | sa 15L | Cairo, CM, CM 3804 (624) [G 691] (kephalaion 44) sa 335L sa 340L sa 308L sa 339L sa 663L (incipit) sa 15L sa 302L sa 14L | sa 335L | sa 342L sa 340L sa 336L sa 293L sa 317Lr sa 663L (incipit) sa 293L

85

86

Schulz

Plates

Figure 3.1 sa 1 (p. 128, John 18:28) Figure 3.2 sa 4 (p. 5, “Amen”-saying 1 at John 1:51 and coronis at 2:1)

Figure 3.4 sa 102 (p. 176, John 10:6–7)

Figure 3.3 sa 5 (p. 66, John 10:22)

Figure 3.5 sa 106 (p. 170, John 3:3)

Figure 3.6 sa 106 (p. 210, John 8:48–49)

87

Pointers to Persons and Pericopes?

Figure 3.9 sa 116 (p. 165, John 19:13)

Figure 3.7 sa 112 (p. 265, John 7:45–46)

Figure 3.8 sa 115 (p. 150, John 10:18–22)

Figure 3.12 sa 285 (fol. 139r, John 12:12)

Figure 3.10 sa 127 (p. 29, John 9:25–27)

Figure 3.11 sa 141 (p. 43, John 6:30–32)

88

Figure 3.13

Schulz

sa 12 (John 9:1)

CHAPTER 4

But for Me, the Scriptures Are Jesus Christ (Ι ϲ� Χϲ; � Ign. Phld. 8:2): Creedal Text-Coding and the Early Scribal System of Nomina Sacra Tomas Bokedal 1

Nomina Sacra Demarcations: from Ancient Manuscripts to Modern Bibles1

One of the leading theologians of the twentieth century, Dietrich Bonhoeffer (1906–45), once made an interesting comment concerning the biblical manuscript tradition: It is good, now and then, to remind oneself of the times of the handwritten and illustrated bibles, in which the name of Jesus was rendered with special reverence and beauty. By printing the Bible, it has perhaps become precisely such a despised book.2 Bonhoeffer here seems to be referring to an aspect of the Christian scribal nomina sacra (sacred names; henceforth NS) practice, namely the graphic emphasis given to the name of Jesus through its contraction (ΙΗϲΟΥϲ, Ι c).3 �

1 Funding for the research was provided by NLA University College, Norway. 2 DBW 14 (Gütersloh: Gütersloher Verlagshaus, 1996), 510. I am grateful to Rune Imberg for pointing me to this passage in Bonhoeffer and to Jutta Leonhardt-Balzer for assistance with translation. 3 Ignatius of Antioch (or his scribe), ca. AD 110, arguably made use of nomina sacra contractions. A fifth-century papyrus containing Ignatius’s letters includes the following nomina sacra: Θ� ϲ� , Θ̅ Υ̅ , Θ̅ Ω̅ , Θ̅ Ν̅ , Ι̅c� , Ι̅Υ̅, Κ̅ Υ̅, Κ̅ Ω̅ , Κ̅ Ν̅ , Π̅ Ρ̅ Ι ̅, Π̅ Ν̅ Α̅ , Χ� c� and Χ̅ Υ̅ (Ign. Smyrn. III, 3–XII, 1). Kurt Aland, Repertorium der griechischen christlichen Papyri, II: Kirchenväter – Papyri, Teil 1: Beschreibungen (Berlin: de Gruyter, 1995). Regarding the rendering of ΕΜΟΙ ΔΕ ΑΡΧΕΙΑ ΕcΤΙΝ Ι ϲ� x�c� , see W. R. Schoedel, “Ignatius and the Archives,” HTR 71 (1978): 97–106. In what follows, New Testament manuscript datings are taken from NA28, and data from word searches are acquired using Oaktree Software, Inc., Accordance XII (June 2017). I want to thank David I. Yoon, Chris S. Stevens, Johanna Bokedal, and the anonymous peer reviewer for helpful comments on previous drafts of the chapter.

© Tomas Bokedal, 2023 | doi:10.1163/9789004537972_006

90

Bokedal

This is usually found together with the Greek4 names for Christ (ΧΡΙϲΤΟϲ, Χϲ� � ), Lord (ΚΥΡΙΟϲ, Χϲ� � ), God (ΘΕΟϲ, Θ� ϲ )� and Spirit (ΠΝΕΥΜΑ, Π̅ Ν̅ Α̅ ). Some additional words are also written in contracted or suspended form in Old and New Testament manuscripts (see note 17), and from the second and third century onwards the NS short-forms are observable in a broader range of texts.5 Particular attention was often given to early NS forms for Jesus or Christ, e.g., the suspension Ι Η ̅ ̅ for ΙΗϲΟΥϲ in the Epistle of Barnabas 9.7–8 (AD 70–135), which seems to presuppose the emergent scribal practice,6 and – with regard to ornamentation – in manuscripts such as the Latin Book of Kells (ca. AD 800), with its magnified Greek nomen sacrum for ΧΡΙϲΤΟϲ (the Chi Rho, ΧΡ).7 One striking thing – which Bonhoeffer seems to have noticed – about these specially written short-forms, usually supplied with a horizontal supra-script line, is their almost universal presence in the Greek and Latin manuscript tradition, in stark contrast to their next to total absence in modern editions of the New Testament.8 This notable discrepancy between old and present-day Bibles may perhaps be primarily explained by one particular event in the history of book manufacturing, namely the transition from manuscript production to book printing.9 In the first edition of Biblia Germanica, the printed Luther Bible of 1534, three of the NS demarcations from the manuscript tradition were kept; the German words for Lord, God and Jesus (HERR/HErr, GOtt and JEsus).10 However, in the second edition of 1545, published around the time of Luther’s death, strangely, 4

NS are used also in the Latin, Coptic, Gothic, Armenian, Slavonic and other versions of the Christian Scriptures. 5 E.g., P. Egerton 2 (unknown Gospel); P. Geneva 253 (Christian homily); P.Oxy. III 405 (Iren. Haer. III, 9.2–3; apologetical literature). 6 Cf. Bart D. Ehrman, “The Text as Window: New Testament Manuscripts and the Social History of Early Christianity,” 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., NTTSD 42 (Leiden: Brill, 2013), 823. 7 See Traube, Nomina, 286, concerning the Latin tradition. 8 About 97% of verifiably Christian Greek manuscripts dated before AD 300 contain NS (Scott Charlesworth, “Consensus Standardization in the Systematic Approach to Nomina Sacra in Second- and Third-Century Gospel Manuscripts,” Aegyptus 86 [2006]: 37). Cf. note 3 above. 9 Cf. Tomas Bokedal, The Formation and Significance of the Christian Biblical Canon: A Study in Text, Ritual and Interpretation (London: Bloomsbury T&T Clark, 2014), 84, 86 n. 11, 122–123, 363; and Bokedal, “Notes on the Nomina Sacra and Biblical Interpretation,” in Beyond Biblical Theologies, ed. H. Assel, S. Beyerle, and C. Böttrich, WUNT 295 (Tübingen: Mohr Siebeck, 2012), 294. Cf. Ludwig Traube, Nomina sacra: Versuch einer Geschichte der christlichen Kürzung (Munich: C. H. Beck, 1907), 285–286. 10 Cf. Traube, Nomina, 285–286.

the Early Scribal System of Nomina Sacra

91

all NS capitalizations were removed from the New Testament section, except the German rendering of the Greek Κϲ� (ΚΥΡΙΟϲ) as HErr/HERR, corresponding to the English LOrd/LORD.11 Sixty-six years later, the English King James version of 1611 followed suit with the second edition of Biblia Germanica, however, with the difference that the King James Bible did not demarcate any NS whatsoever in its New Testament portion, including the divine Name, which was rendered as LORD/LOrd only in the Old Testament. Here I make an appeal, noting these editorial and translational shortcomings in preserving a 1400-year old scribal – editorial NS convention,12 to reintroduce – in one way or another – these highlighted keywords in the manuscripts also into present-day Bible or New Testament editions and translations. For example, by demarcating the core group of four or five13 NS in capitals/small caps; or alternatively, to mark as NS the words that the scribes of our best Greek manuscripts have chosen to demarcate as such, or in accordance with their literary production standards. In the latter case, appropriate text-critical principles ought to be employed with the editorial decisions perhaps ending up somewhere in the range between the restrictive employment by one copyist of Codex Vaticanus (use of four demarcated NS) and the more generous application by the scribes of Sinaiticus and Alexandrinus (15 or 17 NS contractions). Following some introductory matters on the scribal NS practice (in the next section), the heart of this chapter explores a sophisticated, but previously forgotten, embedded numerical symbolism in the biblical text connected to the NS word-group (third and fourth sections of this chapter). To my knowledge, this has not previously been pursued in scholarship. I argue that 12 or 1414 of the altogether 15 or 17 standard terms in the NS word-group were specially treated in the Christian communities, possibly already around AD 100, or perhaps even earlier (see below). As a consequence, selections of 4 to 17 of the words from the wider group were, with varying consistency, to be specially treated as NS short-forms in Greek Christian biblical manuscripts thereafter 11 Bokedal, Formation, 86 n. 11. 12 Regarding the Latin tradition, Traube (Nomina, 286) estimates that in the fifth century “fängt man an, das sakrale Prinzip zu vergessen, das den ersten Bildungen zu Grunde lag.” 13 Cf. the sixth-century Codex Purpureus Petropolitanus (N 022), where the core group of four primary NS compounds, together with the short-forms for ΠΑΤΗΡ, ΥΙΟϹ and ΠΝΕΥΜΑ, are written in gold, and the remaining NS in silver. 14 Except for ΔΑΥΙΔ and ΟΥΡΑΝΟϹ, these 12 or14 words belong to the first three groups of nomina sacra treated below (section four). ΙΕΡΟΥϹΑΛΗΜ, however, the short-form of which is absent from some early manuscripts, is not included (C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt [Oxford: Oxford University Press, 1979], 28 n. 1).

92

Bokedal

(sections two and four). Behind the evolving scribal practice, which from an early date seems to have been occupied with Jewish-Christian and Christian Name theology,15 lay a divine-identity16 or title Christology that was somehow mapped onto early Jewish scribal practices, in particular the special treatment of the divine Name(s). 2

Introducing the NS – “Embryonic Creed of the First Church”

Present in basically all Greek Christian biblical manuscripts,17 the so-called NS are specially written short-forms for ΘΕΟϲ, ΚΥΡΙΟϲ, ΙΗϲΟΥϲ, ΧΡΙϲΤΟϲ and ΠΝΕΥΜΑ, often including up to 10 or 12 additional words.18 Probably originating in early, markedly influential, segments of Christianity with specially written ‘abbreviations’ for ΙΗϲΟΥϲ (Ι ̅Η̅ , Ι ̅ϲ� , Ι ̅Η̅ ϲ� ),19 the scribal phenomenon involved the highlighting of central figures of the Christian faith, especially as these featured within a devotional, Name-theological, and/or early creedal

15 Roberts, Manuscript, 41–42. 16 Terminology borrowed from Richard Bauckham, Jesus and the God of Israel: God Crucified and Other Studies on the New Testament’s Christology of Divine Identity (Grand Rapids: 2008); and Bauckham, God Crucified: Monotheism and Christology in the New Testament (Grand Rapids; Eerdmans, 1999), viii: “The earliest Christology was already the highest Christology. I call it a Christology of divine identity, proposing this as a way to move beyond the standard distinction between ‘functional’ and ‘ontic’ Christology, a distinction which does not correspond to early Jewish thinking about God and has therefore seriously distorted our understanding of New Testament Christology. When we think in terms of divine identity, rather than divine essence or nature, which are not the primary categories for Jewish theology, we can see that the so-called divine functions which Jesus exercises are intrinsic to who God is.” 17 Cf. note 7 above. 18 Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 95–134. For a discussion of 15/17 standard NS – ΙΗϹΟΥϹ, ΧΡΙϹΤΟϹ, ΚΥΡΙΟϹ, ΘΕΟϹ, ΠΝΕΥΜΑ/ΠΝΕΥΜΑΤΙΚΟϹ, ϹΤΑΥΡΟϹ/ϹΤΑΥΡΟΩ, ΠΑΤΗΡ, ΑΝΘΡΩΠΟϹ, ΥΙΟϹ, ΙϹΡΑΗΛ, ΙΕΡΟΥϹΑΛΗΜ, ΟΥΡΑΝΟϹ, ϹΩΤΗΡ, ΔΑΥΕΙΔ and ΜΗΤΗΡ – see Bokedal, “Notes,” 263–295. Beyond the Greek biblical manuscript tradition, where the short-forms are most consistently demarcated, NS are found also in other literary and documentary Christian texts and related material (e.g., Christian Gnostic texts), as well as on a broader selection of artefacts, such as mosaics, frescoes, graffiti, and icons. 19 See Roberts (Manuscript, 35–48) and Hurtado (“The Origin of the Nomina Sacra: A Proposal,” JBL 117 (1998): 664–671), who propose Ι Η̅ ̅ as the first nomen sacrum. Cf. A. H. R. E. Paap, Nomina Sacra in the Greek Papyri of the First Five Centuries A.D.: The Sources and Some Deductions (Leiden: Brill, 1959), 107–109.

the Early Scribal System of Nomina Sacra

93

linguistic framework.20 Thus, the practice, involving a delimited number of words, was not designed to save space or the scribe’s time.21 Along these lines of argumentation, the renowned palaeographer C. H. Roberts famously proposed that the NS “may be plausibly viewed as the creation of the primitive Christian community, representing what might be regarded as the embryonic creed of the first Church; the four primary terms … together with πατήρ, σταυρός, and πνεῦμα represent the beliefs common to all Christians.”22 First Corinthians 8:5–6 provides a good illustration in this regard, since the passage, throughout the manuscript tradition, upholds a sacral/profane distinction for Θ̅ ϲ� /ΘΕΟΙ, as well as for Κ̅ϲ� /ΚΥΡΙΟΙ, and features as recurrent framework for later more elaborate creedal formulations.23 In Greek palaeography and biblical studies, the inclusive 4 to 17 intrinsically or contextually religious terms are usually called NS. However, applied to the above-mentioned four or five core names (which may or may not include Spirit),24 scholars in the past have also referred to them as nomina dei or nomina divina (“divine names”).25 Usually written as contractions – first and final letter, sometimes including middle letters – the standard appearance of these in Greek Bible manuscripts is as follows: Θ̅ ϲ� , Κ̅ϲ� , Ι ̅ϲ� , Χ̅ϲ� , and Π̅ Ν̅ Α̅ ; and in their inflected genitive forms: Θ̅ Υ̅, Κ̅ Υ̅, Ι Υ̅ ̅ , Χ̅ Υ̅, and Π̅ Ν̅ ϲ� . Early renderings, especially 20

See below; for creedal and devotional dimensions of earliest Christianity, see, e.g., Vernon H. Neufeld, The Earliest Christian Confessions (Leiden: Brill, 1963); and Larry W. Hurtado, One God, One Lord: Early Christian Devotion and Ancient Jewish Monotheism, 2nd ed. (Edinburgh: T&T Clark, 1998), 99–114. 21 See Roberts, Manuscript, 27; Bokedal, Formation, 88 n. 20; and section three below. 22 Roberts, Manuscript, 46; cf. Charlesworth, “Consensus,” 67; and John Barton, The Spirit and the Letter: Studies in the Biblical Canon (London: SPCK, 1997), 122. Arguably all the earliest nomina sacra/divina had to do with the Father, the Son, or the Spirit, and with names that were symbolically attached to the Son or the Father (cf. the Primary to Tertiary Groups of nomina sacra, section four below). 23 Paap, Nomina Sacra, 100–101; and Bokedal, “The Rule of Faith: Tracing Its Origins,” JTI 7 (2013): 244–245. 24 In early Latin manuscripts, only the first four NS are “everywhere and always” contracted (C. H. Turner, “The Nomina Sacra in Early Latin Christian MSS,” Studi e Testi 40 [1924]: 64); regarding statistics on the frequency of ΠΝΕΥΜΑ as nomen sacrum, see Joel D. Estes, “Reading for the Spirit of the Text: nomina sacra and πνεῦμα Language in P46,” NTS 61 (2015): 566–594; and Bokedal, “Notes,” 272–276. 25 Cf. Traube, Nomina, 6, 17–18, 33; and Schuyler Brown, “Concerning the Origin of the Nomina Sacra,” Studia Papyrologica (1970): 7–19. For a different view, see C. M. Tuckett, “Nomina sacra – Yes and No?,” in The Biblical Canons, ed. J.-M. Auwers and H. J. de Jonge, BETL 163 (Leuven: Leuven University Press, Peeters, 2003), 431–458. For criticism of Tuckett’s interpretation, see Hurtado, Artifacts, 122–34; and Jane Heath, “Nomina Sacra and Sacra Memoria Before the Monastic Age,” JTS 61 (2010): 518–523.

94

Bokedal

of ΙΗϲΟΥϲ, could also occur as suspension (Ι Η ̅ ̅ ), or, for ΙΗϲΟΥϲ, ΧΡΙϲΤΟϲ, and ΠΝΕΥΜΑ, as a longer contraction (Ι ̅Η̅ ϲ� , Χ̅ Ρ̅ ϲ� , and Π̅ Ν̅ Α̅ ). In basic agreement with suggestions made by Roberts and Hurtado, in the first instance, I here argue that the scribal practice developed roughly in the following order (see further section four below): (1) Commencing with the suspension of ΙΗϲΟΥϲ,26 probably indicating a numerical value, as in Barn. 9.7–8 (Ι Η ̅ ̅ );27 (2) a core group of words was designed into a system of four28 nomina divina compounds; (3) these soon became an extended system – being introduced into the Jewish and new Christian s/Scriptures – including also additional words, written mainly as longer contractions. A fine illustration of the extended NS system from Codex Alexandrinus is the Triune formulations in Matt 28:19, ΟΝΟΜΑ ΤΟΥ Π̅ Ρ̅ ϲ� … Υ̅ Υ̅ … Π̅ Ν̅ ϲ� , and 1 Clem. 46.6 (arguably alluding to 1 Cor 8:6 and Eph 4:4), Η ΟΥΧΙ ΕΝΑ Θ̅ Ν̅ ΕΧΟΜΕΝ ΚΑΙ ΕΝΑ Χ̅ Ν̅ ΚΑΙ ΕΝ Π̅ Ν̅ Α̅ ΤΗϲ ΧΑΡΙΤΟϲ ΤΟ ΕΚΧΥΘΕΝ ΕΦ ΗΜΑϲ ΚΑΙ ΜΙΑ ΚΛΗϲΙϲ ΕΝ Χ̅ Ω̅ (“Do we not have one God and one Christ and one Spirit of grace that has been poured out upon us? And is there not one calling in Christ?”).29 This chapter supports the claim that an early grammar of divine-identity30 or title Christology – as well as a Triune Christological system of numerical text-coding (see below) – lay behind the key developmental stages of the NS system, especially as represented by points (2) and (3) above – beginning with the graphic scribal inclusion of the short-forms for ΙΗϲΟΥϲ (Ι Η ̅ ̅ , Ι ϲ̅ � , Ι Η ̅ ̅ ϲ� ) and ΧΡΙϲΤΟϲ (Χ̅ Ρ̅, Χ̅ϲ� , Χ̅ Ρ̅ ϲ� ) among two specially written Greek compounds of the 26 Ι Η̅ ̅ – the only nomen sacrum commented on in the early Christian literature – is present in at least six early manuscripts. Cf. Paap, Nomina Sacra, 93, 109; Roberts, Manuscript, 36–37; and Scott D. Charlesworth, Early Christian Gospels: Their Production and Transmission, Papyrologica Florentina 47 (Florence: Edizioni Gonnelli, 2016), 97 n. 6 – thanks to Felicia Caratù, Libreria già Nardecchia s.r.l., Rome, for generous help with scanning a chapter from the pre-ordered copy of the book. 27 Ι Η̅ ̅ was perhaps soon followed by suspensions also of ΧΡΙϲΤΟϲ (see P45 and P18). Regarding the development of the ‘abbreviation’ for ΙΗϲΟΥϲ, from Ι Η̅ ̅ to Ι ϲ̅ � (the standard form after AD 300) and the conflation Ι ̅Η̅ ϲ� , see Charlesworth, Gospels, 97; cf. Paap, Nomina Sacra, 109. 28 Among the NS in New Testament manuscripts before the fourth-century codices (when the possibility of using Δ̅ Δ̅ is added), the only ones that can be written with just two letters are those designating God and Christ, i.e., Θ̅ ϲ� , Κ̅ϲ� , Ι ̅Η̅ , Ι ̅ϲ� , Χ̅ ϲ� , Χ̅ Ρ̅, Π̅ Ρ̅ (nominative and vocative forms), and Υ̅ ϲ� – which all relate straightforwardly to the earliest four NS. 29 The Apostolic Fathers: Greek Texts and English Translations, ed. and rev. Michael W. Holmes, 3rd ed. (Grand Rapids: Baker, 2007), 106–107. 30 See note 15.

the Early Scribal System of Nomina Sacra

95

divine Name (Κ̅ϲ� and Θ̅ ϲ� ).31 A unique, yet common Christian understanding of the Name was thus graphically visualized.32 One important implication of such early visual representation of the Name, as Ι Η̅ ̅ /Ι ̅ϲ� /Ι ̅Η̅ ϲ� , Χ̅ Ρ̅ /Χ̅ ϲ� /Χ̅ Ρ̅ ϲ� , Κ̅ϲ� and Θ̅ ϲ� , is that attributes and devotional aspects that were associated with Κ̅ϲ� and Θ̅ ϲ� in the Jewish Scriptures could also be easily and adaptably predicated of Ι Η̅ ̅ /Ι ̅ϲ� /Ι ̅Η̅ ϲ� in the new apostolic literature.33 Whether the earliest origin of the Christian scribal system – which hereafter were to be employed throughout the major strands of the biblical manuscript tradition, commenced with the two contractions Κ̅ϲ� and Θ̅ ϲ� on Jewish ground, as has been argued by some (cf. section three below),34 or with Christian short-forms for ΙΗϲΟΥϲ – does not significantly affect the present argument. In either case, in light of the expected early Jewish and Christian special treatment of the Name, the Christian inclusion of ΙΗϲΟΥϲ and ΧΡΙϲΤΟϲ among the earliest NS “signifies a momentous religious development.”35 Along these lines, Hurtado has convincingly maintained that the core group of the four earliest NS “collectively manifest one noteworthy expression” of what he calls the “binitarian,” or “dyadic,” shape of earliest Christian piety and devotion.36 However, as I shall argue below, the closely associated numerical text-coding that was introduced by influential Christian circles probably no later than the late first century had already attained a Triune Christological form. Again, if, with Roberts, we stress the creedal devotional aspect, the development of the system of NS appears to have emerged not least out of a creedal interest, in creedal-language style. In what follows I shall point to the Christological logic (section four) and specially designed text-coding (section three) behind this development. On the way towards a clearer grasp of the 31 See, e.g., Hurtado, Artifacts, 95–134; and Bokedal, Formation, 83–123. 32 Cf. Larry W. Hurtado, God in New Testament Theology, Library of Biblical Theology (Nashville: Abingdon Press, 2011), 43–44: “‘God’ is so closely linked with Jesus and Jesus so closely linked with ‘God’ that one cannot adequately identify the one without reference to the other.” 33 Cf., e.g., Rom 10:9–13; David B. Capes, Old Testament Yahweh Texts in Paul’s Christology, WUNT 2/47 (Tübingen: Mohr Siebeck, 1992); and Charles Gieschen, “The Divine Name in Ante-Nicene Christology,” Vigiliae Christianae 57 (2003): 117: “The confession Κύριος Ἰησοῦς Χριστός … can be seen to reflect Jewish identification of Jesus with YHWH.” 34 So Kurt Treu, “Die Bedeutung des Griechischen für die Juden im römischen Reich,” Kairos 15 (1973): 123–144; George Howard, “The Tetragram and the New Testament,” JBL 96 (1977): 63–83; and Robert Kraft, “The ‘Textual Mechanics’ of Early Jewish LXX/OG Papyri and Fragments,” in The Bible as Book: The Transmission of the Greek Text, ed. Scot McKendrick and Orlaith O’Sullivan (London: British Library, 2003), 51–72. Cf. Hurtado, “Origin,” 665; and Hurtado, Artifacts, 107ff. 35 Hurtado, Artifacts, 106. Similarly, Charlesworth, “Consensus,” 39. 36 E.g., Hurtado, Artifacts, 106.

96

Bokedal

scribal practice, we shall particularly notice an authorial – editorial arithmetical occupation with the terms in the NS word-group, as they tend to feature with certain frequencies, e.g., as multiples of the significant numbers 3, 4, 7, 17 and 22. 3

Editorial and/or Authorial Integration of NS Numerical Coding into the Text

Greek-Hebrew gematria, that is, number symbolism or alphanumeric coding, in Barn. 9.7–8 (AD 70–135), and possibly also in John 20:31,37 may indicate a first-century origin for the NS practice. Roberts and Hurtado have explained the Johannine phrase “to have life in his name” (John 20:31) with reference to such a Jewish-Christian gematria, where the Hebrew word for life, chai, is thought to have the numerical value 18 [‫ = חי‬18],38 the same as that of the early nomen sacrum for ΙΗϲΟΥϲ – Ι Η ̅ ̅ [= 18] (a nomen sacrum form used in P45 and P18). If this interpretation of John 20:31 is correct – “these are written so that you may come to believe that Jesus (Ι Η ̅ ̅ ) is the Christ (Χ̅ Ρ̅ ), the son of God (Θ̅ Υ̅ ), and that through believing you may have life in his name”39 – it is possible, or even likely, that NS demarcations were employed by the author(s) and/or early editor(s) of John’s Gospel. In any case, Roberts, Hengel, Hurtado, and others do, for reasons other than this, suggest a first-century dating for the origin of the scribal practice.40 Furthermore, Barnabas refers to the early Christian manner of writing numbers with Greek letters supplied with a horizontal stroke, when in 9.7–8 reference is made to Ι Η ̅ ̅ = 18, signifying both the name of Jesus and the numerical value 18. The authorial comment, “the Ι is ten and the Η is eight; thus you have ΙΗ(ϲΟΥ�),41” was arguably intended to refer to an already existing or developing NS practice.42 Now, in addition to this possible, or probable, Christian gematria in Barnabas and John, counting how often certain words occur in the text, numerical considerations linked to the NS seem to have occupied early Christianity as they 37 For caution as to such interpretation of John 20:31, see Heath, “Memoria,” 539 n. 67. 38 Chet = 8; yod = 10. 39 Hurtado, “Origin,” 665–669; and Hurtado, Artifacts, 114–120. Vaticanus includes Ι ϲ̅ � , Χ̅ ϲ� , and Θ̅Υ ̅ in John 20:31. 40 Roberts and Martin Hengel have proposed a pre-70 date, and Hurtado (“Origin,” 660) a date no later than the end of the first century AD. 41 Sinaiticus has Ι Ν̅ ̅ . 42 See note 5 above; cf. Bokedal, Formation, 95 n. 45, 106–107.

the Early Scribal System of Nomina Sacra

97

composed their literature around the turn of the century (AD 100) and probably also earlier (AD 50–100). I shall now comment on some of these tendencies in the Apostolic Fathers, Eusebius of Caesarea, the Septuagint, and the New Testament manuscript tradition. For an overview of the symbolic numbers used in my experimental numerological approach,43 I refer the reader to the basic definitions provided in Table 4.1. Table 4.1 Significant Jewish and biblical symbolic numbers

Jewish/biblical Basic symbolic meaning symbolic numbers 3

4 7 [3 + 4]

8 [7 + 1] 12 [3 × 4]

One of the most significant of all sacred or symbolical numbers, God’s own number (Rev 1:4), representing divine fulness, and the vertical dimension of the world (ΤΡΕΙϹ features 88 times [4 × 22 = 8 × 11] in the Pentateuch (LXX Rahlfs); and 33 times [3 × 11] in the four Gospels). Representing the horizontal dimension of the world, including the four cardinal directions (Rev 7:1). The frequently occurring number of fulness, abundance, and completeness, representing cosmic ordera (ΕΠΤΑ occurs 154 times [7 × 22 = 14 × 11] in the Pentateuch (LXX Rahlfs); and 88 times [4 × 22 = 8 × 11] in the New Testament (NA28)). Symbolising resurrection and new creation in early Christianity (ΟΚΤΩ occurs 8 times in the New Testament). Signifying completeness, perfection (of the kingdom of God), and totality (the product of 3 and 4, representing God and the world, or the vertical and the horizontal dimensions of the world, respectively).b

a Labuschagne, Numerical, 26–31. b Labuschagne, Numerical, 24. 43

On early Jewish and Christian number symbolism/numerical theology, see, e.g., Crispin Fletcher-Louis, Jesus Monotheism, Vol. 1: Christological Origins: The Emerging Consensus and Beyond (Eugene, OR: Cascade, 2015), 39–49; and Casper J. Labuschagne, Numerical Secrets of the Bible: Introduction to Biblical Arithmology (Eugene, OR: Wipf & Stock, 2016). Cf. F. Acerbi, “A Reference to Perfect Numbers in Plato’s Theaetetus,” Archive for History of Exact Sciences 59 (2005): 319–348.

98

Bokedal

Table 4.1 Significant Jewish and biblical symbolic numbers (cont.)

Jewish/biblical Basic symbolic meaning symbolic numbers 11 [4 + 7]

17

26 15

23 32 13 39

22 [2 × 11] 24 [2 × 12] 27 [3 × 3 × 3]

The sum of 4 as the number of extensiveness and 7 as the number of fulness, 11 seems to have developed in the course of time a separate status as a number expressing fulfilment.c Representing the numerical value of the divine Name YHWH/AHWH: yod [1 + 0]/aleph [1] + he [5] + waw [6] + he [5] = 17 (which is the seventh prime number). Representing the numerical value of the divine Name YHWH, the Tetragrammaton: yod [10] + he [5] + waw [6] + he [5] = 26 = 15 [YH] + 11 [WH]. Both 17 and 26 (and 15) represent, each in its own way, the presence of God through his name YHWH.d Both also represent the Hebrew word kabod, “glory” (kbd = 17 or 26; or, when spelled kbwd = 23 or 32). Kabod symbolism in the Bible means that God and his glory are “regarded as belonging inextricably together” (cf. Exod 33:17–23;e and Barn. 12.7: “the glory of Jesus”). We may add here as well the numerical value for God as one (echad), which is 13; thus, YHWH [26)] + echad [13], “The LORD is one” = 39. The number of letters in the Hellenistic Greek (24) and Hebrew alphabets (22, or 27 when including the five end-consonant forms; cf. Jerome, Prologue to the Books of Samuel and Kings), representing completeness and totality.

c Labuschagne, Numerical, 70–73. d Labuschagne, Numerical, 89–90. Cf. also Carlos del Valle, “Abraham Ibn Ezra’s Mathematical Speculations on the Divine Name,” in Mystics of the Book: Themes, Topics, and Typologies, ed. Robert A. Herrera (New York: Peter Lang, 1993), 159–176; 161: “According to Ibn Ezra, there are three proper names (Shem-ha-ʿesem) of God: Yah, Ehyeh, and YHWH. ‘The prophets,’ he declares, ‘established the name of the two letters [Yah], that of the four letters [YHWH], and the name Ehyeh [as] the proper names of the Most High… The remainder of the divine names found in the Bible are adjectives [shemot ha-toʾar].’ In the commentary on Ex 3:15 he expressly states, ‘These three names are proper names.’ Ibn Ezra attributes the same derivation to the three proper divine names: Yah, Ehyeh, and YHWH. Obviously he is thinking of the root hayah (‘to be,’ ‘to exist’) as the ultimate root from which the three names are derived.” I am grateful to Jamie Grant for directing me to del Valle’s article. e Labuschagne, Numerical, 90.

the Early Scribal System of Nomina Sacra

99

3.1 Word-Frequency Coding in the Letters of Ignatius of Antioch Ignatius’s famous phrasing, which provides the title for this chapter – “But for me, the Scriptures are Jesus Christ (Ι ̅ϲ� Χ̅ ϲ� )” (Phld. 8.2) – is well suited for our present purposes.44 As it turns out, Ignatius, his scribe, and/or the editor(s) of the seven Ignatian letters (Middle Recension),45 appear to have utilised specific numbers linked to word-frequencies as an organizing compositional principle (Table 4.2b). The name ΙΗϲΟΥϲ, for example, occurs 22 times in Philadelphians and 22 times in Magnesians, thus equalling the number of letters in the Hebrew alphabet, presumably representing completeness (alluding also to the number 11 [22 = 2 × 11], which may signify fulfilment). The eight terms in the NS word-group used by Ignatius occur as multiples of the central symbolical numbers 3, 4, 7 and 11, the alphabetical 22/24/27 and the numerical values associated with the Tetragrammaton, 15/17/26 (multiples of the latter two groups of numerals are marked in boldface in Tables 4.2, 4.3 and 4.4). As Table 4.2b illustrates – especially the consistent results in the column to the farthest right – particular emphasis seems to be placed on what we might call a nascent creedal matrix made up of the included eight NS terms in individual letters, on the one hand, and in the letter corpus as a whole, on the other. Word-frequencies linked to the two groups of numbers this chapter ascribes particular significance, i.e., multiples of 15/17/26 or 22/24/27, occur with unexpectedly high consistency in the individual letters and in the Ignatian Corpus as a literary whole. In our search through the Corpus for these multiples, at 15 or higher (which is the lowest of these numbers), the outcome is surprisingly high – far above the ca. 26% occurrences we would normally expect if these multiples were not intended, featuring randomly. Instead, from Tables 4.2a and 4.2b we get the following results for multiples of our six basic numbers, pertaining to the divine Name (15/17/26 and 22/24/27) – in the following referred to as Name-related Numeral(s) (henceforth NrN; cf. Rev 1:8; 21:6; and 22:13 for alphabetical fullness associations linked to the Name) – in total an average between 77% and 97% NrN for the NS terms listed below:46

44 Cf. note 51, below. 45 The analysis below is based on the Accordance electronic version 12.3.1 (Oaktree Software, 2018) of Michael W. Holmes’s third edition of The Apostolic Fathers (2007) and the Tyndale House edition of The Greek New Testament (Cambridge: Cambridge University Press; Wheaton, IL: Crossway, 2017) (THGNT), if not indicated otherwise. 46 Only figures at 15 or above included; 77% (lowest percentage included for each of the 12 listed terms/items), 97% (highest percentage included for each item). If, instead, we count the full letter-corpus as one item beside the seven individual letters, i.e., as altogether eight items, it gives an average of 79%.

100

Bokedal

Table 4.2a NS word-group frequencies in Ignatius’s letters

NS(-related) Terms

NrN in the Full Ignatian Corpus NrN in individual letters

ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ ΘΕΟϹ Ο ΘΕΟϹ ΘΕΟ* ΚΥΡΙΟϹ ΠΝΕΥΜΑ ΠΝΕΥΜΑ + ΠΝΕΥΜΑΤΙΚΟϹ ΠΑΤΗΡ Ο ΠΑΤΗΡ ΑΝΘΡΩΠ*

0% (0 NrN out of 1, i.e. 0/1) 100% (130x [5 × 26], i.e. 1/1) 100% (108x [4 × 27]) 100% (176x [8 × 22]) 100% (30x [2 × 15]) 100% (195x [13 × 15]) 100% (34x [2 × 17]) 100% (22x) 100% (34x [2 × 18])

60% (3 NrN out of 5, i.e. 3/5) 20% (1 NrN out of 5, i.e. 1/5) 60% (3/5) 50% (3/6) 0% (no figures above 8) 33% (2/6) 0% (no figures above 9) 0% (no figures above 5) 0% (no figures above 9)

100% (45x [3 × 15]) 100% (22x) 100% (17x)

0% (no figures above 12) 0% (no figures above 6) 0% (no figures above 7)

(ΘΕΟ* = words beginning with the three Greek letters ΘΕΟ-; ΑΝΘΡΩΠ* = words beginning with the six Greek letters ΑΝΘΡΩΠ-).

As presented in Table 4.2b, the underscored triadic structure provided by the words Jesus, Christ, God, Lord and Spirit (the core group of five NS; cf. notes 1 and 63) is here worth noticing. ΠΝΕΥΜΑ occurs in the seven letters altogether 22 times, as does ΘΕΟϹ in Philadelphians and each of the words ΙΗϹΟΥϹ and ΧΡΙϹΤΟϹ in Magnesians. Significantly, the presumed dually coded expression ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ features 108 times [4 × 27] in the full letter corpus,47 and its occurrences in individual letters point to the numerical values associated with the Tetragrammaton (Ephesians 26x, Trallians 17x, Romans 15x). We can also observe (Table 4.2b) the seemingly significant appearance of multiples of 11 for ΑΓΑΠΗ/ΑΓΑΠ* (11x), ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ (11x), ΘΕΟϹ (22x = 2 × 11), etc., and,

47 Presumably representing ΙΗcΟΥc ΧΡΙcΤΟc, Ι* (words commencing with the letter Ι) occurs 225 times [15 × 15] and Χ* 225 times [15 × 15] in the Ignatian Corpus, i.e., altogether Ι* + Χ* feature 450 times [2 × 15 × 15]. NrN figures for Ι* + Χ* are found as well in Exodus, Judges, Sirach, Micah, Joel, Jonah, Nahum, Habakkuk, Haggai, Jeremiah, Lamentations, Daniel (including Susanna and Bel and the Dragon); Mark, the fourfold Gospel, 1 Thessalonians, Colossians, 1 Timothy, 1 John, 2 Peter, Jude, Revelation; Ign. Eph., Trall., Pol.; Barn.; and 1 Clem. Cf. notes 48, 52 and 53.

101

the Early Scribal System of Nomina Sacra Table 4.2b NS word-group frequencies in Ignatius’s letters NS WordGroup

Eph.

Magn.

Trall.

Rom. Phld.

Smyrn. Pol.

1. ΙΗϹΟΥϹ

 33 [3 × 11] 22 [2 × 11] 19

17

22 [2 × 11] 12

3

128 [4 × 32]

Ι*

 54 [2 × 27] 35

28

31

34 [2 × 17] 26

17

225 [15 × 15]

2. ΧΡΙϹΤΟϹ

 33 [3 × 11] 22 [2 × 11] 19

18

21

14

3

130 [5 × 26]

The Ignatian 7-Letter Corpus

Χ*

50

39

32

30

30

31

13

225 [15 × 15]

ΧΡ*

39

27

24

23

24

15

4

156 [6 × 26]

ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ 26

18

17

15

18

11

3

108 [4 × 27]

Ι* + Χ* 3. ΚΥΡΙΟϹ 4. ΘΕΟϹ Ο ΘΕΟϹ ΘΕΟ*

104 [4 × 26] 74

60 61 [4 × 15]

64 [2 × 32] 57

30 450 [2 × 15] [2 × 15 × 15]

9

2

2

1

7

5

8

 34 [2 × 17]

43

23

13

26

22a

25

24

176 [8 × 22]

8

1

3

8

3

4

3

 30 [2 × 15]

 45 [3 × 15] 25

14

27

25

31

28

3

2

4

3

2

22

2

2

12

195 [13 × 15]

5. ΠΝΕΥΜΑ

3

5

ΠΝΕΥΜΑΤΙΚΟϹ

6

2

ΠΝΕΥΜΑ + ΠΝΕΥΜΑΤΙΚΟϹ

9

7

3

2

4

5

4

 34 [2 × 17]

ΠΝΕΥΜΑ*

10

7

3

2

4

6

4

 36 [2 × 18]

6. ΠΑΤΗΡ

9

12

7

6

5

5

1

 45 [3 × 15]

Ο ΠΑΤΗΡ

6

4

4

3

2

3

22

7. ΑΝΘΡΩΠΟϹ

5

1

2

3

1

12

ΑΝΘΡΩΠ*

7

1

3

4

2

17

8. ΥΙΟϹ

3

2

1

8

(ΑΓΑΠΗ)

11

6

6

7

5

7

3

 45 [3 × 15]

(ΑΓΑΠ*)

16

9

11

8

9

11

8

 72 [3 × 24]

(ΠΙϹΤΙϹ)

14

4

1

1

2

4

1

(ΠΙϹΤ*)

15

9

4

5

8

8

2

2

a Following Latin and Armenian manuscripts, omitting ΘΕΟϹ in Phld. 5.1.

27  51 [3 × 17]

102

Bokedal

for the full letter corpus, ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ (108x) + ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ (13x) = 121x [11 × 11].48 Moreover, the number of instances of ΧΡΙϹΤΟϹ in the seven letters is a multiple of a numerical value associated with the divine Name (130x = 5 × 26), which is the case also for three of the frequencies for ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ (see above), ΘΕΟϹ (26x), ΑΓΑΠΗ (45x = 3 × 15), ΠΑΤΗΡ (45x), ΚΥΡΙΟϹ (34x = 2 × 17), ΠΝΕΥΜΑ + ΠΝΕΥΜΑΤΙΚΟϹ (34x), ΙΗϹΟΥϹ (17x), ΑΝΘΡΩΠ* (17x), and ΠΙϹΤ* (51x = 3 × 17)/ΠΙϹΤΙϹ (27x), perhaps thus making a numerical creedal notification. As we shall see, the Christological – triadic pattern is conspicuous also as we move on to 1–2 Clement and other writings among the Apostolic Fathers.49 3.2 1–2 Clement and Barnabas As shown in Table 4.3, possible correlation between the NS word-group and the frequencies with which these words occur appears to be embraced also by 1–2 Clement and Barnabas. Accordingly, in 1 Clement we find apparent alphabetical numerical allusions across the following word occurrences: 22 instances of ΠΝΕΥΜΑ and 66 [3 × 22] of ΚΥΡΙΟϹ; in 2 Clement and Barnabas, 22 appearances of Ο ΘΕΟϹ and 108 [4 × 27] of ΚΥΡΙΟϹ, respectively; and, in all three texts, a few Christological or triadic ‘alphabetical word-frequency-sums’, e.g., ΚΥΡΙΟϹ (108x) + ΙΗϹΟΥϹ (21x) + ΧΡΙϹΤΟϹ (3x) = 132x [6 × 22] in Barnabas.50 For comparison, potentially intended alphabetical-numerical features can be observed as well in the New Testament writings (THGNT, if not indicated otherwise), such as 22 instances of ΠΝΕΥΜΑ in Mark’s Gospel (Vaticanus) and 24 in Revelation; 81 occurrences [3 × 27] of ΘΕΟϹ in John, 96 [4 × 24] in Revelation and 22 in 1 Timothy; 108 appearances [4 × 27] of ΚΥΡΙΟϹ in Acts, 24 in 1 Thessalonians and 22 in 2 Thessalonians. Similar alphabetical NrN patterns characterise as well several of the NS found in the full 27-book Greek New Testament: ϹΤΑΥΡΟϹ 27x, ϹΩΤΗΡ 24x, ΤΟ ΠΝΕΥΜΑ 176x [8 × 22], ΠΝΕΥΜΑΤΙΚΟϹ 26x, Ο ΚΥΡΙΟϹ 351x [13 × 27], ΙΗϹΟΥϹ 912x [38 × 24], ΑΝΘΡΩΠΟϹ 550x [25 × 22], ΥΙΟϹ 378x [14 × 27] and ΧΡΙ* 540x [20 × 27]; and the following words in the singular: ΠΑΤΗΡ 360x [15 × 24; NA28], ΚΥΡΙΟϹ 702x [26 × 27; NA28] and ΜΗΤΗΡ 81x [3 × 27].

48 Corresponding NrN figures for ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ + ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ in Romans, 1–2 Timothy, 1 Clement, 1–2 Clement, and Eusebius’s Ecclesiastical History are 30x [2 × 15; NA28], 27x, 24x, 27x and 45x [3 × 15], respectively. 49 For triadic/Triune numerical patterns in Ignatius, see Table 4.2B. 50 Cf. 1 Clem. 46.6, quoted above.

the Early Scribal System of Nomina Sacra

103

In regard to triadic patterns, perhaps it is of significance that just as in 1–2 Clement the word-frequencies of ΠΑΤΗΡ (29x) + ΥΙΟϹ (11x) + ΠΝΕΥΜΑ (30x) add up to 70 [10 × 7], also in Barnabas the corresponding sum of ΘΕΟϹ (49x) + ΧΡΙϹΤΟϹ (3x) + ΠΝΕΥΜΑ (18x) amounts to 70; that of ΚΥΡΙΟϹ (108x) + ΙΗϹΟΥϹ (21x) + ΧΡΙϹΤΟϹ (3x) to 132 [6 × 22]; and that of the triadic ΠΑΤΗΡ (8x) + ΥΙΟϹ (23x) + ΠΝΕΥΜΑ (18x) to 49 [7 × 7], thus equalling the word-frequency in Barnabas of ΘΕΟϹ which is 49 [7 × 7], and that of ΠΑΤΗΡ (18x) + ΥΙΟϹ (9x) + ΠΝΕΥΜΑ (22x) in 1 Clement, which is 49 [7 × 7].51 Noteworthy parallels to these NS figures are found as well in the New Testament (see below). In our numerical observations so far, all or most instances of a particular word have been included – in the singular as well as in the plural, whether with sacred or profane meaning – which appears to be the case also with the soon standardized NS convention. That is, the scribes of the late second century (and probably also earlier) routinely demarcated certain words, such as ΠΝΕΥΜΑ, as NS. Scott Charlesworth has helpfully referred to this all-inclusive scribal routine phenomenon as “consensus standardization in the systematic approach to NS.”52 As for some more advanced, potentially intended, associations between words occurring with specified frequencies in Barnabas, embracing multiples of three and four, perhaps the letter writer and/or later scribes and editors wanted readers to be able to connect the Christ of the House of David (Christ (3x) × David (4x) = 12) with a perfect man/human being (12 = 3 × 4) of the new people of Israel (12 = 3 × 4), crucified on a cross (12 = 3 + 3 × 3) – the divine sign.53 Additionally, perhaps we are further invited to think along the following lines: that Barnabas might intend an early form of Christological correspondence between the incorporated multiples of the four NS ΚΥΡΙΟϹ (4 × 3 × 3 × 3), ΔΑΥΙΔ (4x), ΧΡΙϹΤΟϹ (3x) and ϹΤΑΥΡΟϹ (3 × 3), in which the former of the four (4 × 3 × 3 × 3) arithmetically connects, or overlaps, with the latter three (4 × 3 × 3 × 3).

51 Similar triadic/Triune arithmetical patterns surface in some of the Ignatian letters, e.g., Eph. ΠΑΤΗΡ (9x) + ΥΙΟϹ (3x) + ΠΝΕΥΜΑ (3x) = 15, alternatively ΘΕΟϹ (43x) + ΠΑΤΗΡ (9x) + ΘΕΟϹ (43x) + ΥΙΟϹ (3x) + ΘΕΟϹ (43x) + ΠΝΕΥΜΑ (3x) = 52 [2 × 26] + 46 [2 × 23] + 46 [2 × 23] = 144 [6 × 24] (cf. Tables 4.1 and 4.2b). 52 Charlesworth, “Consensus”; and Charlesworth, Gospels, 97–119. 53 Cf., e.g., Rev 7:2–4; see further Benjamin Overcash, “Σταυρος as a Nomen Sacrum: A Suggestion,” unpublished conference paper presented at the SBL Papyrology and Early Christian Backgrounds Group, San Antonio, TX, 22 Nov 2016.

104

Bokedal

Table 4.3 NS word-group frequencies in Barnabas and 1–2 Clement

NS Word-Group Barn.

NS Word-Group

1 Clem.

2 Clem.

1–2 Clem.

ΠΝΕΥΜΑ ΠΝΕΥΜΑ*

18 21

ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ ΚΥΡΙΟϹa

3 24

ΠΝΕΥΜΑ ΠΝΕΥΜΑ* ΤΟ ΠΝΕΥΜΑ ΧΡΙϹΤΟϹ Ο ΧΡΙϹΤΟϹ

22 23 9 46 [2 × 23] 13

8 11 6 12 5

 30 [2 × 15]  34 [2 × 17] 15 58 18

66 [3 × 22] 30 [2 × 15] 9 18 109 90 [6 × 15]

20 86 15  45 [3 × 15] 2 11 11 29 33 [3 × 11] 142 22 112 [8 × 14] 153 [9 × 17] 21  70 [10 × 7]

ΥΙΟϹ ΠΑΤΗΡ ΘΕΟϹ ΘΕΟϹ + ΥΙΟϹ ΠΑΤΗΡ + ΥΙΟϹ + ΠΝΕΥΜΑ Ο ΘΕΟϹ + Ο ΙΗϹΟΥϹ + ΤΟ ΠΝΕΥΜΑ ΔΑΥΙΔ ΙϹΡΑΗΛ ΙΗϹΟΥϹ ΑΝΘΡΩΠΟϹ ϹΤΑΥΡΟΩ ϹΤΑΥΡΟϹ ϹΤΑΥΡΟϹ + ϹΤΑΥΡΟΩ

Ι* + Χ*

108 [4 × 27] ΚΥΡΙΟϹ Ο ΚΥΡΙΟϹ 23 ΥΙΟϹ 8 ΠΑΤΗΡ  49 [7 × 7] ΘΕΟϹ Ο ΘΕΟϹ  72 [3 × 24] ΘΕΟϹ + ΥΙΟϹ  49 [7 × 7] ΠΑΤΗΡ + ΥΙΟϹ + ΠΝΕΥΜΑ  51 [3 × 17] ΟΝΟΜΑ + ΠΑΤΗΡ + ΥΙΟϹ + ΠΝΕΥΜΑ 4 12 21 12 3 9 12

ΔΑΥΙΔ ΙϹΡΑΗΛ ΙΗϹΟΥϹ ΑΝΘΡΩΠΟϹ ΑΝΘΡΩΠ* ΥΙΟϹ + ΔΑΥΙΔ ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ

ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ + ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ 195 [13 × 15] Ι* + Χ* ΧΡ*

49 66 [3 × 22]

4 7 28 24 26 13 20 4

25

 91 [7 × 13]

4 8 8 10 2 3

24 242 [11 × 22]

a Barn. 2.10 and 12.11 modified in Holmes’s Greek text provided by Accordance.

3

 36 [2 × 18] 32 36 15 23 4 27

 81 [3 × 27]

105

the Early Scribal System of Nomina Sacra Table 4.3 NS word-group frequencies in Barnabas and 1–2 Clement (cont.)

NS Word-Group Barn. ΚΥΡΙΟϹ + ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ + ΥΙΟϹ + ΘΕΟϹ (ΑΓΑΠ*-words)

NS Word-Group

132 [6 × 22] ΚΥΡΙΟϹ + ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ

1 Clem.

2 Clem.

140 [10 × 14] 40

1–2 Clem. 180 [12 × 15]

 96 [4 × 24] ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ + 192 [8 × 24] ΥΙΟϹ + ΘΕΟϹ 21

(ΑΓΑΠ*-words)  55 [5 × 11] 11 (ΠΙϹΤΙϹ + ΠΙϹΤΕΥΩ) 27 7 (ΠΙϹΤ*-words) 39 10 (ΑΛΗΘ*-words) 14

 66 [3 × 22]  34 [2 × 17]  49 [7 × 7]

3.3 Additional Frequency-Coded Text from the Apostolic Fathers Similar numerical patterns are found also in other writings among the Apostolic Fathers. Didache includes potentially significant figures of fulness for ΠΝΕΥΜΑ (7x), ΚΥΡΙΟϹ (24x) and ΠΑΤΗΡ (7x); ΥΙΟϹ and ΙΗϹΟΥϹ both occur four times (as do ΟΥΡΑΝΟϹ, ΕΥΑΓΓΕΛΙΟΝ, ΑΠΟϹΤΟΛΟϹ and ΕΚΚΛΗϹΙΑ);54 and ΘΕΟϹ appears 13 times (perhaps representing divine oneness). The Shepherd of Hermas features at least a handful potentially significant figures, boasting a perfect 343 instances [7 × 7 × 7] of ΚΥΡΙΟϹ, 49 [7 × 7] of ΥΙΟϹ and 103 (= the 27th prime number) of ΠΝΕΥΜΑ, the triadic sum of which equals 495 [3 × 11 × 15], with a potential allusion to fulfilment (11) and the short-form of the Tetragrammaton (YH = 15). As is the case with the term ΚΥΡΙΟϹ, the phrasing ΤΟ ΟΝΟΜΑ, too, is repeated 49 times [7 × 7] in The Shepherd, and Ο ΚΥΡΙΟϹ, ΠΝΕΥΜΑ* and Ο ΘΕΟϹ appear 192 [8 × 24], 104 [4 × 26] and 234 [9 × 26] times, respectively. The 246 occurrences of ΘΕΟϹ, on the other hand, may perhaps not be particularly significant if we do not assume that the ancients were familiar with so-called untouchable numbers, as 246 happens to be the 15th untouchable number.55 However, ΠΑΤΗΡ (4x) + ΥΙΟϹ (49x) + ΠΝΕΥΜΑ (103x) amounts to 156x [6 × 26 = 4 × 39, potentially signifying the universal reign (4x) of the one and only God (39x)], i.e., another NrN. 54 Cf. Irenaeus’s emphasis on the number ‘four’ in Haer. 3.11.8. 55 See, e.g., https://www.easycalculation.com/other/untouchable-numbers-chart.php (accessed 8 July 2018). Eight of the first 15 untouchable numbers are NrN. I extend my thanks to Naomi Bokedal for assistance with mathematical figures.

106

Bokedal

Similar arithmetical patterns occur in the works of later church writers, such as Eusebius, to whom we now turn. 3.4 Eusebius’s Ecclesiastical History The scriptorium at Caesarea enabled Eusebius, bishop of Caesarea, to supply Emperor Constantine with fifty copies of Scripture (Vita Constantini 4.34–37). Eusebius’s further involvement in the craftmanship of Scripture and other literary production can be seen in what appears to be employment of word-frequency coding also in his own writings. Below, I have listed the figures for the NS word-group found in Eusebius’s 10-volume Ecclesiastical History. The figures that especially interest us are, again, the NrN, i.e., multiples of two main groups of significant numbers, 15/17/26 and 22/24/27. The following results appear, for figures at 15 or above, based on Accordance (Oaktree Software, 2009): Table 4.4 Nomina Sacra in Eusebius’s Ecclesiastical History

NS(-related) Terms

NrN in Euseb. Hist. eccl., vol. 1–10

NrN in individual volumes

Ο ΧΡΙϹΤΟϹ

0% (0 NrN out of 1, i.e. 0/1)

ΧΡΙϹΤΟϹ ΧΡΙ* ΧΡΙϹ* Ο ΙΗϹΟΥϹ ΙΗϹΟΥϹ ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ Ι* + Χ* ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ + ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ Ο ΧΡΙϹΤΙΑΝΟϹ ΘΕΟϹ Ο ΚΥΡΙΟϹ ΚΥΡΙΟϹ ΠΝΕΥΜΑ* ΤΟ ΠΝΕΥΜΑ ΑΓΙΟΝ ΠΝΕΥΜΑ

0% (0/1) 100% (440x [20 × 22]) 100% (432x [16 × 27]) 100% (26x) 0% (0/1) 100% (405x [15 × 27]) 100% (2772 [9 × 14 × 22]) 100% (45x = 3 × 15)

60% (3 NrN out of 5, i.e. 3/5) 44% (4/9) 30% (3/10) 40% (4/10)

100% (44x [2 × 22]) 0% (0/1) 0% (0/1) 0% (0/1) 100% (90x [6 × 15]) 100% (22x) 100% (15x)

50% (1/2) 33% (3/9) 20% (2/10)

60% (6/10) 100% (3/3) 50% (4/8) 0% (0/1)

107

the Early Scribal System of Nomina Sacra (cont.)

NS(-related) Terms

NrN in Euseb. Hist. eccl., vol. 1–10

NrN in individual volumes

Ο ΑΝΘΡΩΠΟϹ ΑΝΘΡΩΠΟϹ Ο ΠΑΤΗΡ ΥΙΟϹ ΥΙ* ΥΙΟϹ + ΑΝΘΡΩΠΟϹ ΙΕΡΟΥ* ΟΥΡΑΝ* ΠΑΤΗΡ + ΥΙΟϹ ΘΕΟϹ + ΧΡΙϹΤΟϹ + ΠΝΕΥΜΑ Ο ΘΕΟϹ + Ο ΧΡΙϹΤΟϹ + ΤΟ ΠΝΕΥΜΑ ΟΝΟΜΑ + ΠΑΤΗΡ + ΥΙΟϹ + ΠΝΕΥΜΑ ΙΗϹΟΥϹ + ΧΡΙϹΤΟϹ + ΥΙΟϹ + ΘΕΟϹ

100% (34x [2 × 17]) 0% (0/1) 0% (0/1) 100% (52x [2 × 26]) 100% (54x [2 × 27]) 100% (216x [8 × 27]) 100% (24x) 100% (81x [3 × 27]) 100% (170x [10 × 17]) 100% (816x [2 × 24 × 17]) 100% (384x [16 × 24])

37.5% (3/8)

100% (374x [22 × 17])

25% (2/8)b

100% (900x [4 × 15 × 15])

33% (3/10)c

40% (2/5) 100% (1/1)

50% (1/2) 40% (2/4) 30% (3/10) 44.5% (4/9)a

a Cf. note 28. b Cf. Matt 28:19. Corresponding figures for ΤΟ ΟΝΟΜΑ + Ο ΠΑΤΗΡ + Ο ΥΙΟϹ + ΤΟ ΠΝΕΥΜΑ: Ignatian Corpus 34x [2 × 17]; John 198x [9 × 22], Gospels 507x [3 × 132 = 13 × 39], Romans 30x [2 × 15], 1 Corinthians 26x, Hebrews 22x, 1–3 John 52x [2 × 26], New Testament 867x [3 × 172]; 2 Samuel 104x [4 × 26], 1 Chronicles 135x [5 × 27], Sirach 34 [2 × 17], Isaiah 105x [7 × 15], Daniel 34 [2 × 17], Twelve Prophets 78x [3 × 26], the Pentateuch 1274x [72 × 26], Historical Books 1647x [61 × 27], LXX Rahlfs 3718x [22 × 132 = 11 × 13 × 26], perhaps indicating the fulfilment/ fulness (11/22) of the oneness (13) of the divine Name (26). c Cf. 3.5 (John 20:31) and Table 4.3.

The consistent appearance of NrN also in Eusebius assists in underscoring one of the main points in this chapter, namely that multiples of the two selected groups of numerals, 15/17/26, on the one hand, and 22/24/27, on the other, often are associated with highly significant terminology in early Christian (and Jewish) literature, such as the nomina sacra word-group, and their wordor phrase frequencies. Other significant terms, e.g., Η ΠΙϹΤΙϹ, ΕΛΠΙϹ and ΑΓΑΠΗ (100% NrN in Historia Ecclesiastica), are involved as well as indicated in Table 4.5.

108

Bokedal

Table 4.5 NrN and Other Word frequencies in Early Christian and Jewish Literature

Term

Book

Wordfrequency

Term

ΖΩΗ

Hist. eccl. Shepherd of Hermas Josephus Antiquitates New Testament Gospels Gospels + Acts Praxapostolos Corpus Paulinum Rev Prophets

 30 [2 × 15]  54 [2 × 27] 17 135 [5 × 27]  52 [2 × 26]  60 [4 × 15] 17 39 17  45 [3 × 15] (LXX Rahlfs) 45 22  48 [2 × 24] 15 15  30 [2 × 15]  45 [3 × 15]  75 [5 × 15]

ΟΔΟϹ

Η ΖΩΗ

ΖΑΩ

ΟΔΟϹ

Sir Eze New Testament Gospels Gen Pentateuch (LXX Rahlfs) Hist. eccl. Shepherd of Hermas Ignatian Letter Corpus Josephus War New Testament John Hist. eccl. Shepherd of Hermas

Book

Matt Gen Deut 1–2 Sam 1 Kgs 2 Chron Tob Jer + Bar Eze LXX Rahlfs (full text) Η ΟΔΟϹ Josephus Works New Testament Gospels + Acts Matt Luke Praxapostolos Acts Pentateuch (LXX Rahlfs) 22 Gen  81 [3 × 27] Deut 140 [10 × 14] Judg 17 1–2 Sam  30 [2 × 15] Sir 17 Twelve Prophets (LXX Rahlfs)

Wordfrequency 22 30 [2 × 15] 48 [2 × 24] 48 48 22 15 68 [4 × 17] 72 [3 × 24] 845 [5 × 13 × 13] 88 [4 × 22] 78 [3 × 26] 66 [3 × 22] 17 15 22 17 75 [5 × 15] 24 34 17 30 [2 × 15] 15 26

the Early Scribal System of Nomina Sacra

109

3.5 The Septuagint and the New Testament After a review of Ignatius’s and some of his contemporaries’ – as well as later church teachers’ – use of terms from the NS word-group, it makes sense to investigate closer also the bishop’s claim about finding/having Christ in the Scriptures. In his own phrasing: The ‘archives’, or the ‘Old Testament’ Scriptures, are Jesus Christ (ἀρχεῖα; Phld. 8.2). Was it perhaps the case that the visible introduction of the earliest Greek NS practice involved, not least, the Jewish Scriptures? Could one of the motivating factors behind the evolving scribal practice perhaps have been to find, and graphically visualize, Christ and creedal patterns in the Scriptures? When posing these kinds of questions, we may be initially encouraged to see that the LXX Odes of Solomon and the Psalms, often viewed closely together,56 together contain 14 instances of ΧΡΙϹΤΟϹ, which may remind readers of the early Christian interest in the numerical value of the Hebrew name for David (dalet [4] + waw [6] + dalet [4] = 14), most likely involved also in the composition of Matthew’s genealogy (Matt 1:17; David appears as the 14th name in the genealogy) as well as in the 14 instances in the New Testament/Gospels of the phrases ΥΙΟϹ ΔΑΥΙΔ (14x) and Ο ΒΑϹΙΛΕΥϹ ΤΩΝ ΙΟΥΔΑΙΩΝ (14x). Our curiosity may be further fuelled by the following: according to Rahlfs’s 2006 Septuagint edition, ΧΡΙϹΤΟϹ features 51 times [3 × 17] and ΙΗϹΟΥϹ 272 times [16 × 17] in the full eclectic text of LXX Rahlfs, and ΥΙΟϹ 34057 times [20 × 17] in Genesis, which all, numerically, may allude to the divine Name. On a basic level, our two primary groups of numerals, 15/17/26 and 22/24/27, seem to be linked to the divine Name, which may be indicated in the alphabetical reference involving the Name in Revelation 1:8, 21:6 and 22:13: ΕΓΩ (ΕΙΜΙ) ΤΟ ΑΛΦΑ ΚΑΙ ΤΟ Ω (cf. Exod 3:14, Isa 41:4). Moreover, the phrasing “ΕΓΩ ΕΙΜΙ” occurs 24 times, respectively, in Leviticus and in John, and 48 times [2 × 24] in the Pentateuch, in the Historical Books, as well as in the 27-book New Testament (NA28).58 In addition, the expression ΕΓΩ ΕΙΜΙ (all forms) features in: LXX Rahlfs (full text) 378x [14 × 27], Poetical Books59 66x [3 × 22], Exodus 15x, Jeremiah 22x and 2 Samuel 22x, and the New Testament 96x [4 × 24] (THGNT). 56 See J. A. Emerton, “Introduction” to “The Odes of Solomon,” in The Apocryphal Old Testament, ed. H. F. D. Sparks (Oxford: Clarendon, 1984), 685: “The two Greek lists of Biblical books, the pseudo-Athanasian Synopsis and the Stichometry of Nicephorus, both refer to the ‘Psalms and Ode(s) of Solomon.’” 57 341x in LXX Göttingen. 58 47x in THGNT, which has ΚΑΓΩ ΕΙΜΙ in Acts 26:29 instead of ΕΓΩ ΕΙΜΙ. In LXX Rahlfs, “ΕΓΩ ΕΙΜΙ” appears 180 times [12 × 15]; moreover, 27x in 1–2 Samuel and 22x in Isaiah. 59 LXX Rahlfs (Odes of Solomon and Psalms of Solomon not included).

110

Bokedal

The pronoun ΕΓΩ (all forms), too, features as NrN according to our two main groups of word-frequencies (LXX Rahlfs and THGNT, if not otherwise indicated): Genesis 850x [50 × 17], Numbers 220x [10 × 22], Sirach 150x [10 × 15], Isaiah 680x [40 × 17], Twelve Prophets 504x [21 × 24], Matthew 270x [10 × 27], Luke 288x [12 × 24], Gospels + Acts 1540x [70 × 22], Praxapostolos 435x [29 × 15], Romans 150x [10 × 15], Colossians 24x, 1 Thessalonians 51x [3 × 17], 2 Thessalonians 26x, Philemon 22x, Hebrews 66x [3 × 22], etc. In the Catholic Epistles of Codex Sinaiticus (‫)א‬, NrN further appear as follows for ΕΓΩ (all forms): Jam 26x, 1–2 Pet 27x, 1 John 60x [4 × 15], 1–2 John 66x [3 × 22], 1–3 John 72x [3 × 24], and the full Catholic Epistles corpus 132x [6 × 22]. The type of arithmetical sensibility that seems to lie behind these word-frequency patterns may go back to Hebrew scribal traditions, as we find similar NrN patterns for fundamental Hebrew terminology: The Tetragrammaton, ‫יהוה‬, e.g., appears as NrN 1820x [70 × 26] in the Pentateuch, 165x [11 × 15] in Genesis, 396x [18 × 22] in Numbers, 550x [25 × 22] in Deuteronomy, 153x [9 × 17] in 2 Samuel, 450x [3 × 15 × 15] in Isaiah, 26x in Jonah, 17x in Nehemiah, etc. Whereas the shortform ‫ יה‬appears 49x [7 × 7] in the Masoretic text, ‫ *יה‬frequently features according to the NrN pattern: In the Pentateuch 1914x [87 × 22], 1–2 Samuel 598x [23 × 26], Isaiah 486x [18 × 27], Jeremiah 960x [40 × 24], Ezekiel 450x [2 × 15 × 15], Amos 85x [5 × 17], Jonah 26x, Micah 44x [2 × 22], Esther 60x [4 × 15], Nehemiah 60x and 2 Chronicles 646x [38 × 17].60 Elohim, ‫ֹלהים‬ ִ ‫ ֱא‬, moreover, appears 2600 times [10 × 10 × 26] in the Masoretic Text (BHS). New Testament frequency-coding in accordance with our two groups of NrN includes the following figures: The 27-book New Testament corpus: Ο ΕΡΧΟΜΕΝΟϹ 17x, Η ΕΙΡΗΝΗ 17x (same as in LXX Rahlfs (full text), 17x), Η ΕΛΠΙϹ 17x, ΠΙϹΤΙϹ 243x [9 × 27] (Gospels 24x, Acts 15x, Luke – Acts 26x, Gal 22x, 1–2 Tim 27x), Ο ΔΙΔΑϹΚΑΛΟϹ 17x, ΡΑΒΒΙ 15x, ΡΑΒΒ* 17x, ΒΑϹΙΛΕΥϹ ΤΩΝ ΙΟΥΔΑΙΩΝ 17x, ΟΜΟΛΟΓΕΩ 26x, ΚΥΡΙΟϹ ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ 24x (NA28), ΚΥΡΙΟϹ Ο ΘΕΟϹ 24x, ΒΑϹΙΛΕΙΑ 162x [6 × 27], Η ΒΑϹΙΛΕΙΑ 136x [8 × 17],61 ϹΠΕΡΜΑ 44x [2 × 22] (Historical Books 45x [3 × 15], Isa 34x [2 × 17], Prophets 52x [2 × 26]), ΤΟ ϹΠΕΡΜΑ 17x (Isa 15x, Prophets 27x), ϹΠΕΡΜ* 45x [3 × 15], ΗϹΑΙΑϹ 22x, ΠΡΟΦΗΤΗϹ 144x [6 × 24] (the latter two demarcated as NS in P. Egerton 2), ΕΝ ΟΥΡΑΝΟϹ 26x (Matt 15x), ΕΝ ΠΝΕΥΜΑ 34x [2 × 17], ΕΝ ΑΥΤΟϹ 144x [6 × 24], ΕΝ ΕΓΩ (all forms) 68x [4 × 17], ΕΝ ϹΥ (all forms) 60 Similarly ‫מֹושׁ ַיע‬ ִ : 27x in the Hebrew Bible; ‫רוּח‬ ַ : Latter Prophets 154x [7 × 22], Isa 51x [3 × 17], Eze 52x [2 × 26], Writings 135x [5 × 27], Eccl 24x, Hebrew Bible 378x [14 × 27]. 61 Η ΒΑϹΙΛΕΙΑ appears in Matt 51x [3 × 17], and in Gospels + Acts 121x [11 × 11].

the Early Scribal System of Nomina Sacra

111

108x [4 × 27], ΕΝ ΚΥΡΙΟϹ 48x [2 × 24], Ο ΔΙΔΑϹΚΑΛΟϹ 17x, ΔΙΔΑϹΚΩ 96x [4 × 24], ΔΙΔΑΧΗ 30x [2 × 15], ΔΙΔΑ* 210x [14 × 15], ΙΔΕ 30x [2 × 15], ΚΑΙ ΕΥΘΥϹ 30x, ΕΥΘΕΩϹ 34x [2 × 17], ΟΥ ΓΑΡ 81x [3 × 27] (Pentateuch 34x [2 × 17], Sir 17x, Gospels 26x, Rom 17x), ΑΚΟΗ 24x, Η ΟΙΚΙΑ 66x [3 × 22] (Matt 22x, Luke 17x, Luke – Acts 24x), Ο ΟΙΚΟϹ 68 [4 × 17] (Acts 17x), ΕΙϹΕΡΧΟΜΑΙ 192x [8 × 24] (Lev 17x, Josh 15x, Judg 27x, 1–2 Sam 72x [3 × 24], 2 Kings 48x [2 × 24], 1–2 Kings 88x [4 × 22], Twelve Prophets 22x, Mark 30x [2 × 15], John 15x, Heb 17x), ΑΝΑΒΑΙΝΩ 81x [3 × 27] (Num 22x, Deut 24x), ΔΟΞΑΖΩ 60x [4 × 15] (John 22x), ΔΙΚΑΙΟϹΥΝΗ (92x) + ΘΕΟϹ (1312x) = 1404x [2 × 26 × 27] (Rom 187x [11 × 17]), ΠΕΤΡΑ 15x, ΕΠΑΓΓΕΛΙΑ 52x [2 × 26], Η ΕΠΑΓΓΕΛΙΑ 34x [2 × 17], Ο ΘΕΟϹ 986x [58 × 17], O ��Ρ��Ϲ 351x [13 × 27], “Ο ΚΥΡΙΟϹ” 110x [5 × 22], Ο ΧΡΙϹΤΟϹ 150x [10 × 15], ΤΟ ΠΝΕΥΜΑ 176x [8 × 22], ΑΡΝΙΟΝ 30x [2 × 15],62 ΑΦΕϹΙϹ 17x, ΧΑΡΙϹ 154x [7 × 22], ΤΟ ϹΗΜΕΙΟΝ 17x, ΜΕΤΑ ϹΥ 27x, ΛΟΓΟϹ (see below), Η ΟΔΟϹ (see Table 4.5), Ο ΥΙΟϹ ΤΟΥ ΘΕΟΥ 27x, ΕϹΧΑΤΟϹ 52x [2 × 26], ΠΑΡΟΥϹΙΑ 24x, ΥΙΟϹ ΤΟΥ ΑΝΘΡΩΠΟΥ (forms in the singular) 81x [3 × 27], Ο ΥΙΟϹ Ο ΑΝΘΡΩΠΟϹ (all forms) 78x [3 × 26]. Four Gospels: ΥΙΟϹ Ο ΑΝΘΡΩΠΟϹ (all forms) 81x [3 × 27],63 ΤΟ ΟΝΟΜΑ 60x [4 × 15] (Rev 24x), “ΓΕΓΡΑΠΤΑΙ” 27x, ΚΑΤΑΒΑΙΝΩ 45x [3 × 15] (John 17x), ϹΠΕΡΜΑ 17x, Η ΔΟΞΑ 26x (John 15x), ΔΟΞ* 78x [3 × 26] (Luke 22x), ΠΑϹΧΑ 26x, ΤΟ ΕΡΓΟΝ 27x, ΜΑΚΑΡΙΟϹ 30x [2 × 15] (Luke 15x, Luke – Acts 17x), ΔΙΔΑϹΚΑΛΟϹ 48x [2 × 24] (Luke 17x), Ο ΙΗϹΟΥϹ ΛΕΓΩ 60x [4 × 15] (Mark 15x, Luke 15x), ΕΓΩ ΔΕ 24x, ΕΝ ΤΗ ϹΥΝΑΓΩΓΗ 17x (all forms). Gospels + Acts: ΥΙΟϹ ΤΟΥ ΘΕΟΥ 17x, ΕΝ ΤΗ ϹΥΝΑΓΩΓΗ 24x (all forms), ΠΝΕΥΜΑ ΑΓΙΟΝ 30x [2 × 15] (Acts 17x), ΕΡΧΟΜΑΙ ΠΡΟϹ ΑΥΤΟΝ (all forms) 24x. John (cf. 20:31): ΙΗϹΟΥϹ (240x) + ΧΡΙϹΤΟϹ (19x) + ΘΕΟϹ (81x) + ΥΙΟϹ (56x) = 396x [18 × 22],64 probably indicating, arithmetically, the significance of the number 18, discussed above (see notes 37 and 38). The corresponding sum for 1 John is 104x [4 × 26], i.e., the sum of four NS and NrN averaging at 26: ΙΗϹΟΥϹ (12x) + ΧΡΙϹΤΟϹ (8x) + ΘΕΟϹ (62x) + ΥΙΟϹ (22x) = 104x [4 × 26]; “ΕΓΩ ΕΙΜΙ” 24x.

62 ΤΟ ΑΡΝΙΟΝ appears 27x in Rev. 63 Phrase distribution: Matt 30x [2 × 15], Mark 15x, Luke 24x, John 12x. 64 416x [16 × 26] in TR.

112

Bokedal

Corpus Paulinum: ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ 88x [4 × 22], ΧΡΙ* 396x [18 × 22], ΕΝ ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ 45x [3 × 15], ΟΝΟΜΑ 26x, Ο ΑΝΗΡ 27x, ΧΑΡΙϹ Ο ΘΕΟϹ 15x, ΕΥΑΓΓΕΛΙ* 85x [5 × 17], ΔΙΚΑΙΟΩ 27x, ΘΥϹΙΑ* 26x (Heb 17x, New Testament 51x [3 × 17], Gospels + Acts 15x; Gen 17x, Lev 170x [10 × 17]), ΕΚ ΠΙϹΤΕΩϹ 22x, ΚΗΡΥ* 27x, Η ΕΚΚΛΗϹΙΑ 45x [3 × 15], ΕΙΡΗΝ* 52 [2 × 26], ΚΑΙ ΕΙΡΗΝΗ 17x, ΓΡΑΦ* 78x [3 × 26], ΕΠΙϹΤΟΛΗ 17x, ΔΙΑΘΗΚΗ 26x (Heb 17x), ΒΑϹΙΛΕΙΑ 17x, ϹΩΤΗΡΙΑ 26x, ΔΙΔΑϹΚΑΛ* 27x, ΔΙΔΑϹΚ* 45 [3 × 15]. Moreover, the occurrences in New Testament writings of the two words ΙΗϹΟΥϹ and ΧΡΙϹΤΟϹ add up to the following NrN: Matthew 168x [7 × 24], Mark 88x [4 × 22], 1 Corinthians 90x [6 × 15], 2 Corinthians 66x [3 × 22], Ephesians 66x [3 × 22], Hebrews 26x, 2 Peter 17x and 1–2 Peter 48x [2 × 24]. Corresponding figures for ΠΑΤΗΡ + ΥΙΟϹ + ΠΝΕΥΜΑ are: Luke 169x [13 × 13], John 216x [8 × 27 = 9 × 24], Romans 60x [4 × 15], 1 Corinthians 48x [2 × 24], 2 Corinthians 26x, Hebrews 45x [3 × 15], 1 John 48x [2 × 24], 1–3 John 54x [2 × 27]; and for ΘΕΟϹ + ΧΡΙϹΤΟϹ + ΠΝΕΥΜΑ (cf. note 28) a next to perfect sequence of NrN: Luke 170x [10 × 17], Acts 260x [10 × 26], Gospels + Acts 720x [2 × 15 × 24] (Rom 252x [14 × 18]), 1–2 Corinthians 352x [16 × 22], Colossians 48x [2 × 24], 1 Thessalonians 51x [3 × 17], 1–2 Timothy 68x [4 × 17], 1–2 Peter 85x [5 × 17], 1–3 John 90x [6 × 15], New Testament 2220x [148 × 15]. It seems as well that the classic acronym ΙΧΘΥϲ, made up of the initial letter of five NS, has gained an integrated place in some (late layers of) New Testament writings. The sum of Ι* + Χ* + Θ* + Υ* + ϲ* emerges as NrN in the following New Testament texts: Ephesians 308x [14 × 22], 2 Thessalonians 130x [5 × 26], 1–2 Timothy 312x [12 × 26], Titus 90x [6 × 15], Philemon 66x [3 × 22], James 192x [8 × 24], Gospels 6615x [5 × 7 × 7 × 27], Praxapostolos 2834x [109 × 26], Corpus Paulinum 4824x [201 × 24] and the New Testament 15198x [894 × 17]. The NrN pertaining to the Old Testament (two examples) and New Testament distribution of the word ΛΟΓΟϹ, here deserve particular attention: ΛΟΓΟϹ: the Pentateuch 54x [2 × 27], Isaiah 44x [2 × 22], Mark 24x, 1 Corinthians 17x, 1–2 Corinthians 26x, Corpus Paulinum 96x [4 × 24], New Testament 330x [15 × 22]; Ο ΛΟΓΟϹ: the Pentateuch 45x [3 × 15], Isaiah 26x, Luke 22x, Luke – Acts 72x [3 × 24], Gospels + Acts 150x [10 × 15] (John 39x; cf. Table 4.1), Corpus Paulinum 48x [2 × 24]; ΛΟΓ*: Gospels 132x [6 × 22], Acts 68x [4 × 17], 1 Corinthians 22x, New Testament 384x [16 × 24]; and ΛΟΓΟϹ ΤΟΥ ΘΕΟΥ: Luke – Acts 15x, and the 27-book New Testament 34x [2 × 17]. Before proceeding with an additional historical overview of the NS system, we shall look briefly also at the biblical NrN associated with one of the NS, namely “Israel,” ΙϹΡΑΗΛ:

the Early Scribal System of Nomina Sacra

113

ΙϹΡΑΗΛ: New Testament 68x [4 × 17], Gospels 30x [2 × 15], Acts 15x, Luke – Acts 27x, Gospels + Acts 45x [3 × 15], Praxapostolos 15x; Leviticus 66x [3 × 22], Judges 187x [11 × 17], 2 Samuel 120x [5 × 24], 1 Kings 192x [8 × 24], 1 Chronicles 108x [4 × 27], Nehemiah 22x, Poetical Books 85x [see note 67; 5 × 17], Prophets 510x [2 × 15 × 17], Twelve Prophets 105x [7 × 15], Hosea 44x [2 × 22], Amos 30x [2 × 15], Jeremiah 88x [4 × 22], Ezekiel 187x [11 × 17], LXX Rahlfs + New Testament 2816x [4 × 32 × 22]. Ο ΙϹΡΑΗΛ: Gospels + Acts 22x; LXX Rahlfs (full text) 480x [20 × 24], 1 Samuel 15x, 2 Samuel 17x, 1 Kings 26x, 2 Kings 30x [2 × 15], the Prophets 165x [11 × 15], Jeremiah 15x, Ezekiel 78x [3 × 26], ΙϹΡΑΗΛ*: Gospels + Acts 51x [3 × 17]; LXX Rahlfs (full text) 2760x [5 × 23 × 24 = 8 × 23 × 15], Pentateuch 600x [40 × 15], Prophets 510x [2 × 15 × 17]; Deuteronomy 75x [5 × 15], 1–2 Samuel 270x [10 × 27]. Ο ΟΙΚΟϹ ΙϹΡΑΗΛ: LXX Rahlfs (full text) 48x [2 × 24], Prophets 34x [2 × 17], Ezekiel 24x. ΘΕΟϹ ΙϹΡΑΗΛ: Joshua 15x, 2 Chronicles 22x; 1–2 Chronicles 34x [2 × 17], Jeremiah + Baruch 22x, LXX Rahlfs + New Testament 189x [7 × 27]. Together these NrN seem to indicate a numerically largely hidden, but nevertheless significant unity of the Scriptures. 4

Creedal Christological Logic: the NS System in Retrospect

4.1 From the Fifth to the First Century AD When studying the NS origin and development of the NS as a scribal and communal textual phenomenon, we can observe a close connection between early divine-identity or title Christology and the emerging NS practice. The Greek NS word-group, which in its developed fourth- and fifth-century format came to include 15 or 17 terms – arguably thus representing numerical values associated with the Tetragrammaton (AHWH = 17; YH = 15; cf. Table 4.1 above) – appears to belong to the most important group of significant scriptural terms characterised by their frequencies of occurrence in individual New Testament writings, sub-collections of New Testament writings, the complete New Testament, and elsewhere.

114

Bokedal

To get an after-the-fact, or reversed historical, perspective of how the Greek editorial – scribal practice developed between the first and fifth century AD, I shall take my point of departure in the fully developed Byzantine system present in Sinaiticus and Alexandrinus and then work my way backwards towards a probable origin in the first century.65 I will begin by discussing what I choose to designate the Quarternary Group,66 which contains the four words demarcated as NS in 0–5% of the earliest New Testament manuscript occurrences of these words. I will then treat, in order, the Tertiary Group of seven (23–62%), and, very briefly, also the Secondary Group of two (89–95%) and Primary Group of four NS (97–100%; 99–100% for NS forms in the singular).67 4.2 The Quarternary Group: Messianic-Soteriological Finish To get a third- to fifth-century glimpse of how the system may have been perceived, I shall start by discussing the four NS contractions that seem to have been added latest to the list of 15/17 standard terms. The words in this group – ΔΑΥΕΙΔ (Δ̅ Α̅ Δ̅ , Δ̅ Δ̅ ), ϲΩΤΗΡ (ϲ Η � ̅ Ρ̅ ), ΜΗΤΗΡ (Μ̅ Η̅ Ρ̅ ) and ΟΥΡΑΝΟϲ (Ο̅ Υ̅ Ν̅ Ο̅ ϲ� ) – are visibly demarcated as NS in only 0–5% of the earliest New Testament manuscript occurrences of the terms, and the figures are in agreement also with extant Old Testament manuscript evidence.68 Three of these terms relate intuitively to Jesus’s family, or ‘incarnational,’ identity (ΜΗΤΗΡ and ΔΑΥΕΙΔ) or to basic messianic-soteriological themes (ΔΑΥΕΙΔ and ϹΩΤΗΡ). For our purposes here – where the words included in the NS word-group typically signify, or point to, the person of God or Christ – we note that the fourth word, ΟΥΡΑΝΟϹ, in early Christian and rabbinic literature functions as a metonym for the divine Name.69 Along these lines we can also notice what Justin Martyr (ca. AD 100–165) underscores, namely that “the name Jesus in the Hebrew language means Savior in the Greek tongue. Therefore 65 See, e.g., Hurtado, “Origin”, 660, 672. 66 The Quarternary Group (ΔΑΥΕΙΔ, ϹΩΤΗΡ, ΜΗΤΗΡ and ΟΥΡΑΝΟϹ) is found, e.g., in Sinaiticus and Alexandrinus (Bokedal, Formation, 90). The other groups treated are the Primary Group (ΙΗϹΟΥϹ, ΧΡΙϹΤΟϹ, ΚΥΡΙΟϹ and ΘΕΟϹ), the Secondary Group (ϹΤΑΥΡΟϹ and ΠΝΕΥΜΑ), and the Tertiary Group (ϹΤΑΥΡΟΩ, ΠΑΤΗΡ, ΑΝΘΡΩΠΟϹ, ΙΕΡΟΥϹΑΛΗΜ, ΥΙΟϹ, ΙϹΡΑΗΛ and ΠΝΕΥΜΑΤΙΚΟϹ). 67 For statistics, see Bokedal, “Notes,” 273–276; and Bokedal, Formation, 88–91. Similar figures in Overcash, “Σταυρος.” 68 See Bokedal, Formation, 88–91, which is based on an analysis of 74 early New Testament manuscripts, second to fourth century (excluding Vaticanus and Sinaiticus), up to and including P123. Old Testament figures in note below. 69 Cf. Ludwig Traube, “οὐρανός,” TDNT, 5: 497–543, esp. 512 and 521; and Traube, Nomina, 99.

the Early Scribal System of Nomina Sacra

115

also the angel said to the virgin, ‘And you shall call his name Jesus, for he will save his people from their sins’” (1 Apol. 33). In terms of manuscript evidence, the first short-forms for ΜΗΤΗΡ and ΟΥΡΑΝΟϲ to appear are present in the third-century P121 (one occurrence of Μ̅ Η̅ Ρ̅ in John 19:26) and the third-/fourth-century P115 (Ο̅Υ̅ Ν̅ Ο̅ ϲ� in Revelation).70 A somewhat different Christological emphasis can be inferred from our next term, the late addition of ϲΩΤΗΡ (ϲ Η � ̅ Ρ̅ ), appearing in extant fourth-century New Testament manuscripts onwards – 33% of the occurrences of the word contracted in Washingtonianus (1 of 3 instances), 38% in Sinaiticus (10 of 26 instances) and 72% in Alexandrinus (18 of 25 instances).71 The obvious soteriological function of the word ϲΩΤΗΡ – beyond its use as pointer to the Hebrew meaning of Jesus’s name (Justin Martyr)72 and in early creedal formulations (Irenaeus)73 – surfaces as well in principal contemporary creeds (e.g., The Creed of the Synod of Antioch from AD 325).74 Of potential interest from a canonical vantage point is the fact that in THGNT and NA28 ϲΩΤΗΡ (ϲ Η � ̅ Ρ̅ ) features 24 times in the New Testament – and also 24 times in the Old Testament (LXX Rahlfs).75 Messianic-incarnational leanings can further be seen from the emerging fourth- and fifth-century New Testament contraction of ΔΑΥΕΙΔ (Δ̅ Α̅ Δ̅ ): Vaticanus (0%), Bezae (0%), Washingtonianus (22%; 8 of 36 instances), Sinaiticus (93%; 55 of 59 instances) and Alexandrinus (98%; 40 of 41 instances). In Sinaiticus and Alexandrinus, we encounter Jesus, Ι ̅Ν̅ Χ̅ Ν̅ … ΕΚ ϲΠΕΡΜΑΤΟϲ Δ̅ Α̅ Δ̅ (2 Tim 2:8), and Ι ̅ϲ� … Η ΡΙΖΑ ΚΑΙ ΤΟ ΓΕΝΟϲ Δ̅ Α̅ Δ̅ (Rev 22:16, cf. 5:5). It is here worth keeping in mind, however, that emphasis on Jesus’s Davidic descent 70

Extant Old Testament papyri, too, confirm the lack of the nomen sacrum for ΜΗΤΗΡ. The earliest occurrence in Kurt Aland, “Repertorium der griechischen christlichen Papyri,” in Biblische Papyri, Altes Testament, Neues Testament, Varia, Apokryphen, PTS 18 (Berlin: De Gruyter, 1976) is from the third and fourth centuries, AT53 (III/IV) and AT59 (IV). The situation changes in W (4th/5th century) with ca. 40% of the appearances of ΜΗΤΗΡ demarcated, e.g., Μ̅ Η̅ Ρ̅ ΤΟΥ Ι̅Υ̅ in John 2:1, 3. As for ΟΥΡΑΝΟϲ as nomen sacrum, AT100 (Aland) from around AD 220 (Stegmüller) may be the earliest Old Testament occurrence. 71 Cf. Aland, “Repertorium,” where the earliest Old Testament occurrence of ϹΩΤΗΡ as a nomen sacrum is found in AT123 (5th–7th century). 72 See above. 73 E.g., Iren. Haer. I, 10.1: “Christ Jesus, our Lord, God, Savior, and King.” 74 Jaroslav Pelikan and Valerie Hotchkiss, Creeds and Confessions of Faith in the Christian Tradition, vol. 1: Early, Eastern, and Medieval (New Haven: Yale University Press, 2003), 84–86. 75 Old Testament Apocrypha excluded (except Baruch as part of Jeremiah in line with early church practice).

116

Bokedal

became important already two or three centuries earlier as part of traditional creedal vocabulary (Rom 1:3; Ign. Eph. 18.2, 20.2; Trall. 9:1–2; Just. Dial. 100.3).76 4.3 The Tertiary Group: Securing a Broader Creedal Visibility Two of the seven words in the Tertiary Group – ΠΝΕΥΜΑΤΙΚΟϲ (Π̅ Ν̅ Ι ̅ΚΟϲ) and ϲΤΑΥΡΟΩ (Ε̅ ϲ� Τ̅ Ρ̅ Θ̅ Η̅ and other verbal forms) – are derived from the two NS contractions in the Secondary Group (Π̅ Ν̅ Α̅ and c� Τ̅ Ρ̅ Ο̅ c� ), and thus perhaps originally introduced also to offer further numerical options for the NS frequency-matrices discussed above (section three). As in the case with the noun ΠΝΕΥΜΑ, the derived adjectival (or adverbal) forms are demarcated as NS with perfect consistency in manuscripts like Sinaiticus:77 24 NS contractions in the New Testament as a whole. However, the standard number of Νew Testament occurrences of ΠΝΕΥΜΑΤΙΚΟϲ (Pauline letters and 1 Pet 2:5) in the wider manuscript tradition is actually 26 (the numerical value of the Tetragrammaton). Although one of the key designations for the first person of the Godhead, ΠΑΤΗΡ (Π̅ Η̅ Ρ̅ ), emerges as visible nomen sacrum earlier than the important Christological title ΥΙΟϲ (Υ̅ ϲ� ), in creedal structures developing throughout the second century, the two appear together as central NS, as can be gathered from early manuscripts, such as P66 (e.g., John 5:20–23), which contains nine regularly employed NS. Already in several of the New Testament writings, “God” as “Father” to Jesus is emphasized, indicating that “Jesus’ divine sonship is unique.”78 This can be seen, for example, in greetings and salutations in the letters of Paul, echoing liturgical formulae (e.g., 1 Thess 1:1–3 and 1 Cor 1:3).79 Jesus’s unique relationship to the Father in this way serves as the basis also for believers to “enter into filial relationship with ‘God’ and so to appeal to ‘God’ as their ‘Father’ as well.”80 Hurtado here stresses an important insight from the New Testament horizon, namely that “[t]he NT does not present ‘God’ as ‘Father’ to believers through creation or in some universalizing sentimental sense. Instead … the Christian practice of addressing ‘God’ as ‘Father’ originates as a profoundly christological statement”81 (cf. Col 1:3).

76 In Ignatius, Tertullian, et al., Jesus’s Davidic descent is thought to be from the line of Mary (Ign. Eph. 18.2). In this connection we may also mention the later appearing nomen sacrum for ΘΕΟΤΟΚΟϲ, Θ̅ Κ̅ Ο̅ ϲ� , in the Greek textual tradition (Traube, Nomina, 8, 125–26). 77 Cf. Dirk Jongkind, Scribal Habits of Codex Sinaiticus (Piscataway, NJ: Gorgias, 2007), 67, 77. 78 Hurtado, God in New Testament Theology, 40. 79 Hurtado, God in New Testament Theology, 41. 80 Hurtado, God in New Testament Theology, 40. 81 Hurtado, God in New Testament Theology, 41.

the Early Scribal System of Nomina Sacra

117

As for the remaining NS in the Tertiary Group we can note Justin Martyr’s comment that “Christ is called both ‘Jacob’ and ‘Israel’” (Just. Dial. 100). Again, this association to ΙϹΡΑΗΛ by Justin – e.g., Christ as the representation or recapitulation of Israel – works well with the observation that most NS signify, or are related to significations of, the person of God or Christ. Illustrative examples include the NS demarcations of ΠΑΤΗΡ,82 ΥΙΟϹ,83 ΟΥΡΑΝΟϹ,84 ϹΩΤΗΡ, ΙϹΡΑΗΛ, and ΑΝΘΡΩΠΟϹ,85 with its specific incarnational quality (Rom 5:15; Ign. Eph. 7.2). Further creedal breadth is secured by the eschatological ΙΕΡΟΥϹΑΛΗΜ (Just. Dial. 113.3–5), which, in a way, is the most flexible NS term, due to its interchangeability with ΙΕΡΟϹΟΛΥΜΑ, a word not included among the 15/17 NS. 4.4 The Secondary Group: Triune Credo and the Cross of Christ When assessing the next group of two words, the Secondary Group, it is important to stress the fact that ΠΝΕΥΜΑ becomes a primary nomen sacrum in the great fourth- and fifth-century Bible codices, in addition to the core group of four.86 Prior to that, however, in the earliest New Testament manuscripts, ΠΝΕΥΜΑ is confidently highlighted as a nomen sacrum in circa 90% of the occurrences of the word and is, in various ways, promoting a triadic or Triune textual profile. The second word, ϹΤΑΥΡΟϹ, significantly appears 27 times [3 × 3 × 3] in the New Testament of Sinaiticus, as well as in key eclectic New Testament texts (THGNT, NA28). A contribution of the Secondary Group is to highlight the Triune (see above) and ‘cruciform’ quality of the NS system. 4.5 The Primary Group: Earliest Christian Nomina Divina The Primary Group was our point of departure in sections one and two above. I understand this core group of the first four specially treated short-forms as Christologically framed nomina divina.87 The mature system of 15/17 NS terms, arguably with a reference to alphanumerical values associated with the Tetragrammaton (15 and 17; cf. Table 4.1) – the nomen divinum – thus seems to connect back to this beginning. 82 ΠΑΤΗΡ in LXX Rahlfs: Genesis 216x [9 × 24], Leviticus 24x, Psalms 22x, etc.; Ο ΠΑΤΗΡ Numbers 24x, Isaiah 15x, etc. 83 ΥΙΟϹ in LXX Rahlfs: Genesis 340x [20 × 17], Exodus 216x [8 × 27], Joshua 220x [10 × 22], Twelve Prophets 54x [2 × 27], etc.; Ο ΥΙΟϹ Twelve Prophets 24x, etc. 84 ΟΥΡΑΝΟϹ in LXX Rahlfs: Genesis 44x [2 × 22], Exodus 15x, Prophets 162x [6 × 27], Twelve Prophets 27x; Ο ΟΥΡΑΝΟϹ: Pentateuch 96x [4 × 24], Twelve Prophets 24x, etc. 85 ΑΝΘΡΩΠΟϹ in LXX Rahlfs: Deuteronomy 51x [3 × 17], Historical Books 234x [9 × 26], Twelve Prophets 44x [2 × 22], etc.; Ο ΑΝΘΡΩΠΟϹ Exodus 17x, Psalms 26x, etc. 86 Jongkind, Scribal Habits, 68. 87 Cf. Hurtado, Artifacts, 95–134.

118

Bokedal

4.6 The NS Word-Group and the Wider Class of Frequency-Coded New Testament Vocabulary The NS word-group, which in its developed fourth- and fifth-century format came to include 15 or 17 terms, appears from the outset as part of a larger class of frequency-coded New Testament vocabulary. Presumably the NS are (among) the most high-profiled terms within this wider class of significant New Testament vocabulary characterized by their frequencies of occurrence. As we have seen, two of the frequencies that stand out are the ones that relate either to the numerical values associated with the Tetragrammaton (15/17/26) or to alphabetical completeness (22/24/27). So, whether or not a particular nomen sacrum is graphically highlighted in the text, it carries this additional faith-vocabulary-related, arithmetical quality. For the NS word-group we can thus note the interesting figures in Tables 4.5 and 4.6, pertaining to multiples of the numbers 15/17/26 and 22/24/27, respectively – with data acquired from THGNT. If we make the reasonable assumption that New Testament terms found in the two NS word-groups listed in Tables 4.5 and 4.6 are of the ‘highest arithmetical order’ (cf. Rev 1:8), the first thing to notice is that all 15/17 NS occurring in the New Testament are included in these two groups. Secondly, comparable NrN New Testament vocabulary listed in section three above include additional, textually related, high-profiled words of Christian faith, such as ΔΙΔΑϹΚΑΛΟϹ/ΔΙΔΑΧΗ, ΛΟΓΟϹ, ΠΑϹΧΑ, ΑΦΕϹΙϹ, ΠΑΡΟΥϹΙΑ, and Η ΕΙΡΗΝΗ.88 Now, if we presume that Table 4.6 with its alphanumerical Tetragrammatoncoding contains the highest-profiled NS terms, we see that 12 (words in full without the article) or 16 (ΙΕΡΟΥϹΑΛΗΜ excluded) of the 17 NS are included. Within Table 4.6 some additional textual – creedal features may be detected as well, such as (a) emphasis on crucifixion-words (ϹΤΑΥΡΟ*), on the one hand, and “mother”- or “human being”-words, arguably alluding to the Incarnation, on the other; (b) inclusion of terms constituting the early Triune pattern God, Christ, and Spirit (cf. 1 Clem. 46.6), rather than the supposedly somewhat later appearing Father, Son, and Spirit; (c) inclusion of a quite broad selection of creedal vocabulary representing messianic – Christological Triune faith, with the earliest, Pauline, layer of seven 15/17/26-NS-words (without the article) represented. These include ΙΗϹΟΥϹ (1 Corinthians, 1–2 Corinthians, Galatians), ΘΕΟϹ (Romans, 1 Corinthians, 1–2 Timothy, Hebrews), ΚΥΡΙΟϹ (Ephesians, Philippians, 2 Timothy), ΠΝΕΥΜΑ (Romans, 2 Corinthians), ΠΝΕΥΜΑΤΙΚΟϹ 88

The extensive number of words involved in such New Testament frequency-coding may also have (had) the function of helping to secure an authentic, editorially approved, New Testament text.

119

the Early Scribal System of Nomina Sacra

(1 Corinthians), ΧΡΙϹΤΟϹ (1 Timothy), and ΑΝΘΡΩΠΟϹ (1–2 Timothy, Pauline Corpus).89 From Tables 4.5 and 4.6 we can also connect some of the NS terms with particular New Testament books (beyond the Pauline link for seven or eight early NS) such as ΠΑΤΗΡ (John), ΜΗΤΗΡ (Synoptic Gospels), ΔΑΥΙΔ (Matthew), ΙϹΡΑΗΛ (Gospels and Acts), ΟΥΡΑΝΟϹ (Mark, Gospels, Acts, and Revelation), ΙΕΡΟΥϹΑΛΗΜ (Luke), and ϹΤΑΥΡΟ* (Matthew, John, Gospels). As we have used various models throughout this chapter to capture the development of the NS word-group in New Testament manuscripts and elsewhere, the five words that stand out as a solid core group, from the canonical edition of the Corpus Paulinum onwards, are: ΙΗϲΟΥϲ (Ι ϲ̅ � ), ΧΡΙϲΤΟϲ (Χ̅ ϲ� ), ΚΥΡΙΟϲ (Κ̅ϲ� ), ΘΕΟϲ (Θ̅ ϲ� ) and ΠΝΕΥΜΑ (Π̅ Ν̅ Α̅ ). In view of the wider canonical horizon, however, the full system of 15/17 (or 12/14) NS, and the wider textual word-frequency matrix, provide an equally important key to scriptural understanding. Table 4.6 15/17/26-multiples from the NS word-group (THGNT)

Term

Text

Frequency

Term

ΘΕΟϹ

Matt Acts Rom 1 Cor 1–2 Tim Heb Rom Gal 1 Thess 1–2 Tim New Testament Gospels Praxapostolos

 52 [2 × 26] 165 [11 × 15] 153 [9 × 17] 105 [7 × 15]  34 [2 × 17]  68 [4 × 17] 105 [7 × 15] 17 26 15 986 [58 × 17] 306 [18 × 17] 300 [20 × 15]

Gospels + Acts Gospels Rom 2 Cor ΤΟ ΠΝΕΥΜΑ Mark John Gospels ΠΝΕΥΜΑΤΙΚΟϹ 1 Cor New Testament ΠΝΕΥΜΑ* Eph ΠΝΕΥΜΑ ΑΓΙΟΝ Acts Gospels + Acts ϹΤΑΥΡΟ* Matt

Ο ΘΕΟϹ

ΘΕΟ*

89

ΠΝΕΥΜΑ

Text

Frequency 17 102 [6 × 17]  34 [2 × 17] 17 15 15  51 [3 × 17] 15 26 17 17  30 [2 × 15] 15

Corresponding data from Table 4.7 include ΙΗϹΟΥϹ (Philippians, 1–2 Timothy), ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ (Pauline Corpus), ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ (1–2 Tim), ΘΕΟϹ (1–2 Thessalonians, 1 Timothy), ΚΥΡΙΟϹ (1–2 Corinthians, 1–2 Thessalonians), ΑΝΘΡΩΠΟϹ (Romans), ΥΙΟϹ (Hebrews), and ΠΝΕΥΜΑΤΙΚΟϹ (Pauline Corpus). For possible Pauline usage of NS, see also Heath, “Memoria,” 543–47.

120

Bokedal

Table 4.6 15/17/26-multiples from the NS word-group (THGNT) (cont.)

Term

Text

Frequency

ΚΥΡΙΟϹ

Luke John Eph Phil 2 Tim Praxapostolos Matt Luke John Acts 2 Cor 2 Thess 1–2 Tim Luke

104 [4 × 26]  52 [2 × 26] 26 15 17 150 [10 × 15] 26  45 [3 × 15] 17  68 [4 × 17] 15 15 17 105 [7 × 15]

Ο ΚΥΡΙΟϹ

ΚΥΡΙ*

Gospels 1 Cor 2 Cor Col 2 Pet ΙΗϹΟΥϹ 1 Cor 1–2 Cor Gal ΧΡΙϹΤΟϹ 1 Tim 1–2 Pet Ο ΧΡΙϹΤΟϹ 2 Cor New Testament ΙΗϹΟΥϹ 1–2 Pet ΧΡΙϹΤΟϹ ΚΥΡΙΟϹ Acts ΙΗϹΟΥϹ

255 [15 × 17]  68 [4 × 17]  30 [2 × 15] 17 15 26  45 [3 × 15] 17 15  30 [2 × 15] 15 150 [10 × 15] 17 15

Term

ΠΑΤΗΡ Ο ΠΑΤΗΡ

ΑΝΘΡΩΠΟϹ Ο ΑΝΘΡΩΠΟϹ Ο ΥΙΟϹ ΥΙ* ΙϹΡΑΗΛ

ΙϹΡΑΗΛ* ΟΥΡΑΝΟϹ

Ο ΟΥΡΑΝΟϹ ΜΗΤΗΡ

ΔΑΥΙΔ ϹΩΤΗΡ*

Text

Frequency

John Gospels John Mark Gospels 1–3 John 1–2 Tim Pauline Corpus John Rev Luke Rom Gospels Acts/ Praxapostolos Gospels + Acts New Testament Gospels + Acts Gospels Acts Rev Pauline Corpus Matt Mark Luke Gospels + Acts Matt Luke – Acts

15  51 [3 × 17] 136 [8 × 17] 15 221 [13 × 17] 15 15 136 [8 × 17]  34 [2 × 17] 15  45 [3 × 15] 15  30 [2 × 15] 15  45 [3 × 15]  68 [4 × 17]  51 [3 × 17] 153 [9 × 17] 26  52 [2 × 26] 17 26 17 17  75 [5 × 15] 17 17

New Testament  75 [5 × 15]

121

the Early Scribal System of Nomina Sacra Table 4.7 22/24/27-multiples from the NS word-group (THGNT)

Term ΘΕΟϹ

Text

John 1–2 Thess 1 Tim Rev Ο ΘΕΟϹ Mark Eph ΘΕΟ* Rom Col ΚΥΡΙΟϹ Acts 1–2 Cor 1 Thess 2 Thess 1–2 Pet Ο ΚΥΡΙΟϹ New Testament

Frequency

 81 [3 × 27]  54 [2 × 27] 22  96 [4 × 24]  44 [2 × 22] 22 154 [7 × 22] 22 108 [4 × 27]  96 [4 × 24] 24 22 22 351 [13× 27 = 32 × 39] ΚΥΡΙ* Eph 27 1–2 Tim 24 Praxapostolos 154 [7 × 22] Rev 24 ΙΗϹΟΥϹ Mark  81 [3 × 27] John 240 [10 × 24] Phil 22 1–2 Tim 27 Praxapostolos 110 [5 × 22] New Testament 912 [38 × 24] Ο ΙΗϹΟΥϹ Matt 110 [5 × 22] Gospels 384 [16 × 24] ΧΡΙϹΤΟϹ Gospels  54 [2 × 27] 1 Pet 22 Ο ΧΡΙϹΤΟϹ Gospels + Acts  48 [2 × 24] ΧΡΙ* 2 Cor  48 [2 × 24] Corpus Paulinum 396 [18 × 22] New Testament 540 [20 × 27]

Term

Text

Frequency

ΙΗϹΟΥϹ ΧΡΙϹΤΟϹ Pauline Corpus Praxapostolos ΧΡΙϹΤΟϹ ΙΗϹΟΥϹ 1–2 Tim ΠΝΕΥΜΑ John Rev ΤΟ ΠΝΕΥΜΑ Gospels + Acts New Testament ΠΝΕΥΜΑΤΙΚΟϹ Pauline Corpus ΠΝΕΥΜΑ ΑΓΙΟΝ Luke – Acts ϹΤΑΥΡΟϹ NT ΠΑΤΗΡ Gospels + Acts ΑΝΘΡΩΠΟϹ Rom New Testament Ο ΑΝΘΡΩΠΟϹ Matt

 88 [4 × 22]  44 [2 × 22] 22 24 24  88 [4 × 22] 176 [8 × 22] 24  24 27 308 [14 × 22] 27 550 [25 × 22]  72 [3 × 24]

Acts Luke – Acts Heb 1 John 1–3 John Praxapostolos New Testament 1 John 1–3 John Luke – Acts Gospels + Acts Luke Acts Rev Praxapostolos Luke – Acts Νew Testament Praxapostolos

22  72 [3 × 24] 24 22 24  48 [2 × 24] 378 [14 × 27] 22 24 27 22 27 24  48 [2 × 24] 27 24 24 22

ΥΙΟϹ

Ο ΥΙΟϹ ΙϹΡΑΗΛ Ο ΙϹΡΑΗΛ ΙΕΡΟΥϹΑΛΗΜ Ο ΟΥΡΑΝΟϹ

ΔΑΥΙΔ ϹΩΤΗΡ ϹΩΤΗΡ*

122

Bokedal

5 Conclusion In my discussion I hope to have successfully demonstrated a few crucial things pertaining to the NS in Christian biblical and early literature. First, it is important to make the NS visible again on the written page, not only in archives, in museums, or in electronic format online, but in published Bibles of all formats and for church usage, no less than for academic purposes. Second, such visibility will probably have the potential of affecting the important ongoing dialogue between historians and theologians interested in early Christology and creedal-Triune structuring of the faith, on the one hand, and biblical scholars, on the other. Along these lines, the NS system seems to have functioned as a tacit bridge-builder in the past, with meta-textual significance, while still being very much part of the manifest text itself. Third, in light of the extensive numerological investment that arguably have been linked to the NS word-group, the potential for forceful, creedal miniature memos, presumably perceived by scribes and other groups of early Christians, might hopefully reignite the NS discussion. Fourth, by reviewing the development of the scribal practice, from a core group of four names to a wider selection of 15 or 17 words, I hope to have provided some tools for future research, not least with regard to neglected creedal textual dimensions engrafted into the biblical material, as well as the potential significance of the related wider selection of frequency-coded New Testament terms and phrases (e.g., ΠΙϹΤΙϹ occurring 24 times in the Gospels, 26 times in Luke – Acts, 39 times in the fourfold Gospel + Acts, 22 times in Galatians, 27 times in 1–2 Timothy, 243 times [9 × 27] in the New Testament, Η ΠΙϹΤΙϹ 17 times in Romans, and ΕΚ ΠΙϹΤΕΩϹ 22 times in Corpus Paulinum). Fifth, basic Old Testament and New Testament text-critical principles probably need some adaptation to reflect significant, arithmetically motivated, textual variants – involving the NS word-group as well as the wider range of frequency-coded vocabulary – in the early Christian manuscript tradition. If there is an advanced Christian ‘Bible code’ involving the Greek (and Hebrew) text, this is probably it – which may have implications also for other areas of biblical and related scholarship. Overall, it can be claimed with some confidence that the Greek Christian Scriptures were designed with the NS as key linguistic building blocks, helping to set the textual agenda for early Scripture production, reading and interpretation.

CHAPTER 5

Segmentation and Interpretation of Early Pauline Manuscripts S. Matthew Solomon 1 Introduction Segmentation features in early manuscripts (MSS; sing. MS) differ in some ways from modern critical editions of the Greek New Testament (GNT) as well as English translations and commentaries. Punctuation in critical editions is generally accepted as well established. Later MSS, though not earlier ones, seem to drive modern segmentation. What would Pauline corpus segmentation features look like if we were to examine only MSS from before the year 400 ce? I suggest that our modern critical editions and translations would punctuate and segment the text differently in certain key places. Modern and ancient writers vary in the use of punctuation and paragraphing. In the same way, modern and ancient readers vary in the amount of prosodic information provided to them. While ancient scribes tend to prefer scriptio continua, early Christian writings tend to exhibit punctuation marking sense units (e.g., space intervals or mid-level dot in the text) and sometimes even paragraphing (e.g., paragraphos or ekthesis). Larry Hurtado suggests that early Christian codices were constructed in a manner that turned away from the skills of elite pagan readers to more socially diverse early Christian readers.1 As such, we inherit a text that has a small level of interpretation already built into the text with segmentation features, whereas ancient elite reading culture places the onus on readers. Modern critical editions and translations sometimes differ with ancient segmentation features offering different interpretive contours. This essay will examine paratextual segmentation features to gain interpretive insight into early Christian reading practices. Modern critical editions differ from ancient manuscripts in terms of the purpose of segmenting their texts. As a result, modern punctuation and paragraphing does not always align with ancient punctuation and paragraphing. 1 Larry W. Hurtado, “Manuscripts and the Sociology of Early Christian Reading,” in The Early Text of the New Testament, ed. Charles E. Hill and Michael J. Kruger (Oxford: Oxford University Press, 2012), 58–59.

© S. Matthew Solomon, 2023 | doi:10.1163/9789004537972_007

124

Solomon

In this chapter, all early manuscripts of the Pauline letter collection (including Hebrews) will be examined for punctuation and paragraphing then compared to modern critical editions with an eye toward how to properly divide the text. Modern interpreters can sometimes take for granted the punctuation of modern critical editions without considering the earliest manuscript evidence. The purpose of this chapter is to highlight the differences between concepts of ancient and modern paragraphing and suggest divisions for certain texts including 1 Cor 7:39, 14:33, Eph 5:21, 1 Tim 3:1, and Phlm 7. 2 Dividing Texts – Theories Before considering examples from MSS of the Pauline letter collection, the nature of composition and subsequent dividing of sense units needs to be briefly examined. In this section, the methods of composing a unit of meaning as well has how to divide that unit will be considered, including both ancient and modern conceptions. Some exegetical implications of dividing texts will be examined as well. 2.1 Ancient Versus Modern Textual Delimitation Paul’s letters are written compositions, but they show indications that they were written for oral use.2 One might consider these compositions as hybrids, a type of writing meant for oral communication. To make matters more complicated, these hybrid compositions were most likely written originally in scriptio continua format with very little division of sense units or larger pericopes. As these texts were copied over the centuries, more reading aids were added by copyists and/or readers, and by the time of the printing press some of these paratextual features were cemented in place. The crux of the segmentation discussion centers on the consideration of ancient and modern views of delimiting the text, to which we now turn our attention. One of the more distinctive features of ancient versus modern composition comes in the form of visually representing larger blocks of sense units or pericopes.3 Whereas most ancient writings initially were composed in

2 A great example can be seen in Betz’s rhetorical breakdown of Paul’s letter to the Galatians in Hans Dieter Betz, Galatians: A Commentary on Paul’s Letter to the Churches in Galatia, Hermeneia (Minneapolis: Fortress, 1989). See also Eph 3:4; Col 4:16; 1 Thess 5:27; and 1 Tim 4:13. 3 From this point forward the discussion will center on ancient writings not necessarily ancient oral compositions.

Early Pauline Manuscripts

125

pure scriptio continua format with no word or sense unit divisions,4 modern writings are mostly composed with paragraphs and larger sections in mind.5 Ancient writings tended to feature more rhetorical elements as most reading in the ancient world tended to be aloud. As such, compositions must make use of rhetoric and discourse markers to communicate. 2.2 What Is a Paragraph? Defining a paragraph is tricky business.6 As a matter of fact, linguists for the most part have not devoted much time to analyzing paragraphs.7 There are, of course, some features of paragraphs that may be discernible in modern composition. First, there is a visual component, meaning the paragraph is set off by some sort of visual cue like indentation or a full blank line before and after. Second, there is the notion that the paragraph should communicate a single idea, perhaps even indicated by a topic sentence to begin the paragraph.8 Beyond these two features, there is plenty of room for ambiguity. The fact is that the paragraph exists structurally somewhere between the sentence and the entire discourse.9 Stanley Porter adds that the paragraph’s function is to demonstrate that a group of sentences are more like each other than other sentences in the immediate and global context of a discourse.10 In this way, paragraphs are most effective when they can be defined by the original author or scribe and not necessarily by later readers. The whole idea of paragraphs may be even more tricky if one considers that modern writers compose writings with paragraphs in mind whereas ancient 4

Harry Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven, CT: Yale University Press, 1995), 48. 5 For instance, in the composing of this essay, headings, subheadings, and paragraphs dictate to the reader the beginning, ending, and transition of individual units of thought that have been carefully constructed to guide the reader to understanding. These paratextual features are in addition to the textual content, but as the author I choose to provide them as helps for the reader to understand the thoughts being communicated. 6 A popular anecdote may be applicable here: perhaps defining paragraphs is as tricky as defining pornography. In both cases, definitions can become blurred and there will be exceptions, but in both cases you know it when you see it. 7 Stanley E. Porter, “Pericope Markers and the Paragraph: Textual and Linguistic Implications,” in The Impact of Unit Delimitation on Exegesis, ed. Raymond de Hoop, Marjo C. A. Korpel, and Stanley E. Porter, Pericope 7 (Leiden: Brill, 2009), 176. Much of this section will summarize and interact with Porter’s chapter on paragraphs. 8 Porter, “Pericope Markers” 180–182, includes a list of seven generalizations regarding the composition and nature of a paragraph after his section discussing two schools of thought. 9 Porter, “Pericope Markers” 177. 10 Porter, “Pericope Markers,” 177.

126

Solomon

writers may not have composed writings with any consideration to the same visual and/or structural cues. In other words, if ancient writers did not compose their discourses into smaller sections of cohesive micro units intended to be visually separated from the linear flow of the whole discourse, but rather composed their discourses with oral/aural cues in mind, then how can modern editors, interpreters, translators, and readers feel confident in dividing ancient texts into paragraphs? In some mss. readers and copyists have added visual elements to divide texts, so we have something with which to work. The idea, generally, though is that MSS of NT texts were initially written in scriptio continua format with very little to no punctuation or paragraphs. This possible disconnect may be a challenge going forward. 2.3 Exegetical Implications Paratextual features of writings brought to the fore by Gérard Genette paradoxically provide subtle yet overt interpretive frameworks for reading texts.11 What we take for granted as headings, subheadings, paragraphs, sentences, layout on a page, medium material, etc. all provide some sort of interpretive framework for how a text is to be read. The question becomes how should modern editors and readers divide ancient texts which initially had no division? The act of reading ancient texts placed a greater burden of interpretation and prosodic division on readers than the act of modern reading. Modern writings can be delimited as part of an author’s intentional interpretive framework offered to potential readers. The challenge for the Pauline epistles comes from the fact that modern editors are in charge of delimiting the text of ancient MSS, which had little to no segmentation and in many cases differ in how they delimit texts. Modern critical editions (and subsequent translations) are based on these MSS, which are copies of what could lack authorial segmentation features in their initial forms. Therefore, paragraphing and punctuation decisions of modern editors of critical editions of Greek texts as well as translators for modern English NTs impact how a text is interpreted. Examples of these interpretive features are further explored below, but one example at this point will help illustrate the point. In 1 Tim 3:1, the verse begins “this is a trustworthy saying.” The “this” of the clause is potentially ambiguous as the clause is situated between two possible referents requiring the reader to determine the referent. Most critical editions and English translations begin a new paragraph at 3:1, thereby presented a decision to which paragraph “this is a trustworthy saying” belongs. Some posit 11 Gérard Genette, Paratexts: Thresholds of Interpretation (Cambridge: Cambridge University Press, 1997).

Early Pauline Manuscripts

127

the referent as 2:15, which is anaphoric, while others have the referent functioning cataphorically to the rest of 3:1; the difference is two completely different statements with exegetical implications. In other words, the placement of a new paragraph break directly affects the interpretation of the passages. The importance, therefore, of considering all possible data, including especially the oldest MS tradition, when making editorial decisions cannot be stressed enough. 3

Segmentation Features of Early Pauline Manuscripts

Consideration must be given to segmentation features of Pauline MSS prior to the Great Codices. In this section, all pre-400 ce MSS are examined for segmentation features. It should be noted that most MSS containing Paul’s letters before 400 are fragmentary in nature outside of 𝔓46. These remaining fragmentary MSS can be categorized by the presence or absence of segmentation features.12 3.1 Segmentation Features Present Thirteen early, fragmentary Pauline MSS contain segmentation features. Overall, minimal segmentation is observed. This lack of segmentation might exist for two reasons. First, the total available data represents such a small portion of what were once whole MSS. In other words, the fragmentary nature of these MSS could cloud the assessment of how scribes might have been delimiting texts. Second, the early date of these MSS could signal a lack of development in terms of reading aids. The lack of segmentation features possibly signals an early preference for scriptio continua in contrast to later preference for fully divided texts. The following early Pauline MSS are to be examined for their segmentation features: 𝔓13, 𝔓15, 𝔓16, 𝔓17, 𝔓49, 𝔓65, 𝔓113, 𝔓118, 𝔓123, 𝔓132, 𝔓133, 𝔓139, and 0220. Dating to ca. 300 ce, 𝔓13 is made up of multiple fragments of the book of Hebrews as part of Paul’s letter collection.13 This manuscript was originally a 12

Of course, with the evidence being fragmentary this is not to say that the ones deemed unhelpful never contained any segmentation features but that what we have extant is not helpful. Therefore, this survey does not prove one way or the other how scribes might have segmented biblical texts during this period merely that we only have very little with which to work. 13 Hebrews, in the early period, was included in Paul’s letter collection. Authorship views aside, Hebrews must be examined for segmentation features as the Pauline letter collection was most likely copied as a whole, segmentation features and all.

128

Solomon

scroll (one of only four NT papyri from a scroll), with a Latin epitome of Livy appearing on the recto and Hebrews appearing on the verso.14 The text has been copied in mostly scriptio continua format with single-column justification on both sides giving the appearance of a rectangular block of text. In terms of segmentation, 𝔓13 contains one notable feature. The text has been broken up regularly by two in line dots resembling a modern colon with letter-sized gaps frequently occurring concurrently. These dots often align with modern versification in Hebrews but also occur in the middle of verses. No discernible pattern emerges in terms of clause, sentence, or paragraph delimitation.15 In some instances, a multiple-letter sized gap occurs before Old Testament quotations, such as before Heb 10:37. Neither ekthesis nor paragraphoi are present where left-hand margins are extant. Also, no lines are left blank in order to begin a new segment on the next line. In terms of determining paragraphs for modern editions and translation, 𝔓13 is not helpful. 𝔓15 and 𝔓16 look similar and could be part of one codex.16 Whether or not this is the case, the segmentation features of both of these MSS are remarkably similar. 𝔓15 is a single-leaf, fragmentary MS dating to the early fourth century containing the text of 1 Corinthians 7.17 𝔓15 is written in scritptio continua format with single-letter-sized spaces included where one might expect a pause. These spaces align with what is present in NA28/UBS5 critical editions. Interestingly, a larger gap is in between 7:24 and 7:25, which is where NA28/UBS5 have decided that a new paragraph should start. This same larger gap is observed between 7:35 an 7:36 on the recto, where NA28 has a minor break and UBS5 has a hard return to start a new paragraph.18 Since the left-hand margin is extant for this portion of the MS, a diple and ekthesis can be observed on the line below 14 James R. Royse, “The Early Text of Paul (and Hebrews),” in The Early Text of the New Testament, ed. Charles Hill and Michael Kruger (Oxford: Oxford University Press, 2014), 184. 15 This could be due to the small sample size with which we are presented because of the fragmentary nature of the MS. 16 Royse, “The Early Text of Paul,” 189, who adds in a footnote that Philip Comfort and David Barrett (The Text of the Earliest New Testament Greek Manuscripts [Wheaton, IL: Tyndale House, 2001], 95), argue for 𝔓15 and 𝔓16 originating from the same codex. Conversely, Junack et al. (Das Neue Testament auf Papyrus. II. Die Paulinischen Briefe, 1 [Berlin: De Gruyter, 1989], xli) doubt that interpretation due to the size of the letters and the color of the ink. Alan Mugridge (Copying Early Christian Texts: A Study of Scribal Habits [Tübingen: Mohr Siebeck, 2016], 262) includes both 𝔓15 and 𝔓16 in the same entry arguing they come from the same scribe. 17 Royse, “The Early Text of Paul,” 189. 18 The Tyndale House Greek New Testament includes a hard return and new paragraph here.

Early Pauline Manuscripts

129

where 7:39 begins, suggesting a major break compared to the more minor gaps previously observed. The same is the case between 7:40 and 8:1 where a larger gap exists followed by a marginal diple and ekthesis on the line below; a new paragraph is started in NA28/UBS5 at this point too. 𝔓16 similarly is a single-leaf, fragmentary MS dating to ca. 300 ce19 of Philippians written in scriptio continua format with what appear to be singleletter sized gaps. The MS is rarely extant where one might normally expect segmentation. No medial dots, ekthesis, or paragraphoi can be seen in the extant portion of this MS. Taking the papyri together, only 𝔓15 has provided two places of which modern critical editions should be aware for current formatting between 1 Cor 7:35–36 and 7:38–39. 𝔓17 is a small fragment of Heb 9 dating to the fourth century.20 The text is written in scriptio continua format. The text has been divided in two extant places with two medial dots resembling a modern colon. These two markings appear between Heb 9:12 and 9:13 as well as Heb 9:15 and 9:16.21 In terms of paragraphing, 𝔓17 is not helpful. 𝔓49 is a single-leaf fragment containing portions of Eph 4 and 5. Dating from the third century, this leaf suffers from a large section of damage in the middle of the leaf.22 The text is written in scriptio continua format. There is no sign of ekthesis or paragraphoi in this MS. The text has been divided in some places by the presence of two raised dots resembling modern colons. These two raised dots align often with modern versification but sometimes simply break up the text with no concern for grammatical function. For instance, in the middle of 4:19, two raised dots can be observed between ασελγειᾳ and εις εργασιαν, which merely breaks up the clause.23 On the other hand, two raised dots near the beginning of 4:28 separate two independent clauses.24 In any case, these raised dots occur frequently throughout the text preserved on this leaf, sometimes totaling two occasions per verse of extant text. 19 Royse, “The Early Text of Paul,” 189. 20 Royse, “The Early Text of Paul,” 190. 21 The marking between 9:12 and 9:13 appears to be three dots, but upon further inspection is more than likely two dots with what might be a third middle dot that is the end of the final sigma stroke. 22 Royse, “The Early Text of Paul,” 194. Mugridge (Copying Early, 265) allows for a late third to early fourth century date. 23 This is merely speculation, but the raised dots here could also function to break up three words in a row that could be a bit of a tongue twister with sigmas, nus, gammas, and the vowels. 24 Unfortunately, the MS is damaged between 4:27 and 4:28 so we cannot know if there were raised dots separating these two verses.

130

Solomon

𝔓65 is a single-leaf fragment containing very sparse portions of 1 Thessalonians 1 and 2 dating to the third century.25 The text seems to have been copied in scriptio continua format with lines justified to the left margin but aligning slightly out of line in the right margin. The papyrus fragment is mostly unremarkable in terms of punctuation and paragraphing in that such a small amount of text survives, but the MS is notable in that the margins are preserved. There are no signs of ekthesis or paragraphoi, even in places where one might expect like 1 Thess 2:13. Interestingly, the copyist has squeezed εν τῃ αχαιᾳ (1:7) at the end of a line so it sticks out further than the other lines in what might be an attempt to begin 1:8 on a new line. Other than these two features, 𝔓65 is largely unhelpful in determining paragraphing. 𝔓113 is a very small fragment of one leaf from a MS containing a portion of Romans 2 dating from the third century.26 The text is sparse, containing only four whole words and a total of thirty-six letters are extant.27 The extant text seems to occupy a narrow column, suggesting it was part of a codex that arranged the text into multiple columns.28 For such a small fragment, however, three individual raised dots appear: one between 2:12 and 2:13 on one side and two surrounding and thus setting apart the phrase ου γραμματι in 2:29. These raised dots are aligned near the tops of letters and are very similar to the reading marks found in 𝔓46. With such a scant amount of evidence, no patterns can emerge. However, the presence of three dots with only thirty-six extant letters suggests that this text might have been frequently broken up. 𝔓118 is a single-leaf fragmentary MS containing portions of Romans 15 and 16 dating to the third century written in what appears to be two columns.29 The fragmentary nature of the MS limits the scope of paratextual features extant, but a couple spots should be considered. First, the nomen sacrum for Jerusalem at the end of 15:26 appears to stick out further than the other letters in the right-hand margin perhaps to allow 15:27 to begin on a new line. Second, two raised dots appear in between αυτων and ει γαρ in 15:27 separating two clauses in a place where NA28 and UBS5 provide a raised dot. Lastly, there is a one- or two-sized letter gap between 16:5 and 16:6 but with no medial dots. This gap with no mark occurs in the last part of the letter and falls naturally between two sets of greetings. 25 Royse, “The Early Text of Paul,” 194–195. Mugridge (Copying Early, 266) allows for a late third to early fourth century date. 26 Royse, “The Early Text of Paul,” 197. 27 Even these are only two letter words. 28 Royse, “The Early Text of Paul,” 197. 29 Royse, “The Early Text of Paul,” 198.

Early Pauline Manuscripts

131

𝔓123 is a single-leaf fragment containing portions of 1 Corinthians 14 and 15 dating to the earlier part of the fourth century.30 Very little of the text remains. One possible reading mark may be present between 14:33 and 14:34, which will be examined in more detail in a later section. The dot in question would be a raised dot found above the letters, but the difficulty in confirming its presence as a segmentation feature comes from the fact that this line is lacunose except for the very tops of the letters. Further, if this were a raised dot used for segmentation, the definite article would most likely be missing, which would be a singular reading at this place. 𝔓132 is a single-leaf fragment containing portions of Ephesians 3 and 4 dating to the third or fourth century.31 Very little of the text survives. One notable feature is a raised dot between 4:1 and 4:2 on the verso. The recto contains no extant paratextual features. 𝔓133 is a single-leaf fragment containing portions of 1 Timothy 3 and 4 dating to the third century.32 The surviving text shows signs of segmentation including single-letter-sized gaps, which separate what appears to be a creedal statement in 3:16, and one instance of a raised dot, which separates two clauses in 4:8. Other than these features, the MS lacks any extant portion that would reveal ektheses, paragraphoi, and lengthy gaps at the end of lines. 𝔓139 dates to the fourth century and contains portions of Philemon. The single-leaf fragment includes vv. 6–8 and 18–20 preserving the bottom of the folio.33 The surviving text has been segmented with raised dots in the middle of v. 7 and v. 19 where one might expect. Also, a forked paragraphos appears in the margin in the middle of v. 20. The text is damaged after what appears to be a nomen sacrum for χριστος. Two possibilities exist here: either the paragraphos marks the beginning of v. 21 and v. 20b has been omitted due to a slip of the eye, or the nomen sacrum for χριστος is a singular reading and v. 20b was included after, which would make the placement of the paragraphos suspect. It is more likely that the copyist made an error and instead of writing κς then v. 20b, their eye skipped to the end of v. 20 in their exemplar copying χς and then v. 21. The forked paragraphos appears near v. 21, which is where one might expect the text of Paul’s letter to be divided.

30 Royse, “The Early Text of Paul,” 199. 31 Geoff S. Smith, “5258. Ephesians 3:21–4:2, 14–16,” in The Oxyrhynchus Papyri LXXXI, ed. J. H. Brusuelas and C. Meccariello (London: Egypt Exploration Society, 2016), 1–3. 32 Jessica Shao, “5259. 1 Timothy 3:13–4:8,” in The Oxyrhynchus Papyri LXXXI, 4–8. 33 David Lincicum, “5347. Philemon 6–8, 18–20,” in The Oxyrhynchus Papyri LXXXIII, ed. P. J. Parsons and N. Gonis (London: Egypt Exploration Society, 2018), 11–14.

132

Solomon

GA 0220 is a single-leaf fragmentary parchment MS containing the oldest extant portions of Romans 4 and 5 dating to the third century.34 The text is written in scriptio continua format and contains raised dots that align with modern versification in all extant verses. Although the left-hand margin is present on the back, the leaf is far too damaged to observe any paratextual features that would help with paragraphing or punctuation. As observed thus far, the fragmentary nature of the MSS dating to this period should provide caution in discerning patterns for the Pauline collection. We can note, however, that the MSS from this period (outside of 𝔓15’s diples and 𝔓139’s paragraphos) generally do not include extant portions with ekthesis, paragraphoi, or significant gaps to end lines. The text of these MSS has been copied mostly in scriptio continua format ranging from professional calligraphic handwriting to unskilled handwriting and has been aligned to be justified to both margins.35 The dominant segmentation feature of these MSS surveyed has been single- and multiple-letter gaps as well as medial dots to break up the text. Absent Segmentation Features 3.2 Seven early, fragmentary Pauline MSS do not contain segmentation features: 𝔓12, 𝔓27, 𝔓30, 𝔓32, 𝔓40, 𝔓87, 𝔓92, and 𝔓114. This could be because of the fragmentary nature of the MSS in question, i.e., the MSS in question in their fully extant form could have contained segmentation features now lost to us. This lack of segmentation could also be due to these MSS belonging to the earliest stages of the copying process for New Testament MSS. The New Testament text of 𝔓12, Heb 1:1, can be found written at the top of the recto of a single-leaf portion of a scroll containing a Christian letter written from Rome. The LXX version of Gen 1:1–5 and some extra-biblical text is found on the verso. The text of Heb 1:1 is written in a different hand than the text of the Christian letter and may date to the late third or early fourth century.36 No segmentation features are present in the extant New Testament text. 𝔓30 is a multiple-leaf fragment of 1 Thessalonians 4–5 dating to the third century.37 The text is written in scriptio continua format with no noticeable places of punctuation or paragraphing. 34 Royse, “The Early Text of Paul,” 199. This MS is of course famous as a witness to the textual variant in Rom 5:1. 35 These designations have been assigned by Mugridge (Copying Early, 156), who provides his rating scale in his catalogue of papyri. 36 Royse, “The Early Text of Paul,” 188. 37 Royse, “The Early Text of Paul,” 191.

Early Pauline Manuscripts

133

𝔓32 is a single-leaf third-century fragment containing portions of Titus 1 and 2.38 Even though the surviving fragment preserves the margins on its recto and verso side, no evidence is extant for any segmentation features. 𝔓40, likewise, is a third-century MS made up of multiple fragments of leaves with no apparent segmentation, even in extant portions of text where one might expect some sort of segmentation. 𝔓87 is the earliest MS containing text from Paul’s letter to Philemon dating to the early third century.39 This fragment preserves very little text and certainly no punctuation or paragraphing. The most notable feature is that this tiny fragment preserves the ending of Philemon, which is indicated by blank space to the end of the line with no markings. Neither 𝔓92 (dating to the late third to early fourth century) nor 𝔓114 (dating to the third century) display any sort of segmentation features.40 As a matter of fact, both MSS preserve very little text, and even though both have a small portion that contains a left-hand margin, neither has any indication of paragraphing or punctuation. 3.3 Questionable One of the earliest Pauline fragments can be labeled as questionable in terms of segmentation features. The papyrus fragment 𝔓10 contains Rom 1:1–7 but only occupies the front half of a sheet of papyrus with the back left blank minus some stray marks and the word αποστολος. While these odd characteristics are enough to warrant caution in evaluating segmentation features, true hesitation arises from the notion that 𝔓10 has been observed to be a school writing exercise.41 Any hesitations can be dismissed considering that the exercise is not one of composition but one of copying. Segmentation features, therefore, can be examined with the hope that the student copied the writing faithfully. With this understanding in mind, 𝔓10 contains only one medial dot in the middle of 1:1 with no other signs of punctuation or paragraphing.42 Another questionable example can be found in 𝔓27. The text of 𝔓27 has been preserved on two fragments from a single leaf dating to the third century.43 38 39 40 41

Royse, “The Early Text of Paul,” 192. Royse, “The Early Text of Paul,” 195. Royse, “The Early Text of Paul,” 196–197. Annemarie Luijendijk, “A New Testament Papyrus and Its Documentary Context: An Early Christian Writing Exercise from the Archive of Leonides (P.Oxy. II 209/ 𝔓10),” JBL 129.3 (2010): 589. 42 It is always possible that the student performed the exercise poorly and did not incorporate segmentation from his or her exemplar. 43 Royse, “The Early Text of Paul,” 191.

134

Solomon

The remaining fragments do not preserve any obvious segmentation features in its extant text of portions of Romans 8 and 9. The fragments do preserve a left-hand margin, which may show a paragraphos mark below the line where 9:1 begins. This mark may simply be some sort of imperfection or damage to the MS, which is very difficult to discern given the poor state of the photographs.44 It is unfortunate that the MS is lacunose before the beginning of 9:1 where some sort of gap or medial dot might exist to add to the validity of this possible paragraphos. In either case, this mark is the only visible segmentation feature in the MS. 3.4 𝔓46 The earliest, most complete MS of Paul’s letter collection is 𝔓46. Much has been written about this MS and its importance, so the goal of this section is not to rehearse the work of previous scholars.45 Rather, this section will focus on paratextual segmentation features. Although dates can range from the first century46 to the third,47 the most sensible date for the MS seems to be sometime during the third century.48 The text is written in a single column with anywhere from twenty-six to thirty-two lines per page with minimal adornment. Although 𝔓46 is the earliest MS of Paul’s letter collection, the codex is missing 2 Thessalonians, the Pastoral Epistles, Philemon, and more than likely would not have had space at least for the Pastorals.49 In terms of segmentation features, the text of 𝔓46 contains mostly space intervals50 original to the MS and reading marks in certain books that were

44 45

This could also be a place where one sees what one wants to see even if it is not obvious. For recently extensive work on 𝔓46 see James Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Atlanta: SBL, 2010); Edgar Ebojo, “A Scribe and His Manuscript: An Investigation into the Scribal Habits of Papyrus 46” (PhD thesis, University of Birmingham, 2014). For more, see J. K. Elliott, A Bibliography of Greek New Testament Mansucripts, 3rd ed. (Leiden: Brill, 2015), 27–29. 46 Young Kyu Kim, “Palaeographical Dating of P46 to the Later First Century,” Biblica 69.2 (1988): 248–257. Kim’s arguments have been refuted for the most part. For a summary of the arguments for dating, see Min Seok Jang, “A Reconsideration of the Date of Papyrus 46” (PhD dissertation, New Orleans Baptist Theological Seminary, 2010), 118–143. 47 For a brief discussion of the historical debate surrounding the date of 𝔓46, see Brent Nongbri, God’s Library: The Archaeology of the Earliest Christian Manuscripts (New Haven, CT: Yale University Press, 2018), 141–145. 48 Royse, Scribal Habits, 199–201. 49 Ebojo, “A Scribe,” 204–235. 50 The term “space intervals” here is borrowed from Ebojo, “A Scribe,” 182.

Early Pauline Manuscripts

135

more than likely added later.51 No ekthesis, eisthesis, or paragraphoi are present.52 The space intervals, although very slight in nature, are noticeable and functional at breaking up the text on the phrase, clause, and paragraph level with no distinction between the three. The space intervals have been supplemented at a seemingly later date by what have been termed “reading marks,” which mostly look like bolded acute accents. The space intervals and reading marks do not always align. These paratextual features more than likely signal the fact that 𝔓46 was read publicly. Unfortunately, since there is no distinction between the phrase, clause, and paragraph level even our earliest and most complete manuscript of the Pauline epistles does not provide us with a comprehensive guide to paragraphing the text. 3.5 Summary Each MS outside of the Great Codices dating to before 400 ce has been examined. Thirteen early fragments plus 𝔓46 contain segmentation features, seven early fragments do not contain segmentation features, and two fragments are questionable in terms of their features. Several observations can be deduced. First, during this early period, the text of Pauline MSS was mainly observed to be divided by single-letter sized gaps, with larger gaps sometimes being employed. Second, sometimes raised single dots or dicolons were used to divide the text. Third, we do have an example (𝔓15) where a diple and ekthesis indicate a new paragraph. Fourth, we also have one example (𝔓139) where a paragraphos has been employed to indicate what we can assume is a new paragraph where the text is lacunose. Lastly, and perhaps most importantly, we must remember that due to the fragmentary nature of these earliest MSS, broad-sweeping conclusions cannot be sustained, we must proceed with caution. 4

Reconsidering Paragraph Divisions in Pauline Letters

This section examines places where paragraph breaks in modern critical editions need to be reconsidered. This is not to say that modern critical editions are monolithic in how their texts are divided, but a few key prominent examples should lead to a more widespread reconsideration of how texts should 51 52

These reading marks appear to be a different color ink and can be found in Romans, 1 Corinthians, Philippians, and Hebrews. Their presence in only four of the letters adds to the likelihood that they were added later. A few instances of paragraphoi occur to mark the transition from book to book but never appear in the text portion, as per Ebojo, “A Scribe,” 177.

136

Solomon

be divided at the paragraph level, especially considering some of the earliest MSS in the tradition. Unfortunately, due to the fragmentary nature of the early MS tradition of Paul’s letters, the oldest MSS to witness to the text of some of the examples below date to the mid-third century and later. Even still, the need to reconsider how Pauline texts have been divided over the years should be clear.53 4.1 1 Corinthians 7 The text of 1 Cor 7:39 is attested in the Greek MS tradition before the year 400 ce by four MSS: 𝔓15, 𝔓46, 01, and 03. In ch. 7, Paul is addressing issues concerning marriage. Throughout the chapter, he is giving advice to differing groups. Editors of critical editions over the years have differed on how to divide the text of the last half of the chapter. For instance, UBS5, Tyndale House GNT (THGNT), and Society of Biblical Literature GNT (SBLGNT) include a paragraph break at 7:39. Robinson-Pierpont GNT (RP), Tischendorf, and Westcott and Hort (WH) do not include a paragraph break at 7:39. The NA28, however, includes a large gap in the line before 7:39 but no hard return to the next line for a paragraph break.54 These differences in paragraphing in modern critical editions are curious, especially the difference between the UBS5 and the NA28 who share an identical Greek text. Examination of the earliest Greek MS tradition for 1 Corinthians 7 sheds light on how the text should be divided at 7:39. In 𝔓15, the text is damaged at the beginning of 7:39, so it is impossible to determine whether or not there was a gap in the text, which is usually what signals some sort of segmentation in the earliest Pauline MS tradition.55 We are fortunate, however, to observe in the left-hand margin a paragraphos diple and slight ekthesis indicating a major break in the text. The same marking occurs below where the beginning of 8:1 occurs, which is a major shift in content (from marriage to food sacrificed to idols). A letter-sized gap can be observed in 𝔓46 between 7:38 and 7:39, which indicates a break, even though paratextual segmentation features are limited in 𝔓46, this gap is the same sort of gap found in many other locations. The gap does signal a break, which is important for paragraphing consideration. Codex Sinaiticus’s text has been frequently segmented using ekthesis to begin a new 53 I must acknowledge early on that punctuation and paragraphing issues ought not be solely based on MS features. However, it seems as though we have places where the testimony of the earliest MSS has been ignored or overruled, which will hopefully be demonstrated in this section. 54 Interestingly, major English translations (NIV, ESV, NASB, etc.) all include a paragraph break at 7:39. This may be due to the use of UBS GNT for translation activities. 55 See the survey of early Pauline MSS above.

Early Pauline Manuscripts

137

sense unit. This is indeed the case at 7:39 where the line above contains a large gap and 7:39 begins with ekthesis on the next line. Codex Vaticanus’s text is also highly segmented using paragraphoi.56 The end of 7:38 contains a raised dot and 7:39 begins on the next line with a paragraphos mark. So, the early MS tradition for this passage is clear: the text of 7:39–40 should be included in a new paragraph. The exegetical implications of whether a paragraph is included here are minor when compared to some other passages. Gordon Fee notes that this final paragraph is a bit of a puzzle in terms of why it is the last word in this section to begin with due to the fact that it assumes that the woman is married, which is different than 1 Cor 7:25–38. In his estimation, the paragraph is functioning as a final word for the whole chapter instead of simply the last half.57 If 7:39–40 was not a separate paragraph functioning as a conclusion to the entire chapter, then it could be read as concluding only the section contained in 7:25–40. In this way, the women under consideration would not be married, and this paragraph is further commentary on 7:36–38, which would not make much sense in terms of Paul’s logic throughout ch. 7. 4.2 1 Corinthians 14 The text of 1 Cor 14:33 is attested in the Greek MS tradition before the year 400 ce by four MSS: 𝔓46, 𝔓123, 01, and 03.58 In chapter 14, Paul is addressing issues of orderliness in worship gatherings, including speaking in tongues as well as other aspects of worship such as singing and prophesying. When arriving at 14:33, it is important to note early on that some have theorized over the years that 14:34–35 makes up an interpolation that entered the MS tradition

56

While this is not the case throughout the entire MS, this is the case for the Pauline epistles under consideration. See Charles E. Hill, “Rightly Dividing the Word: Uncovering an Early Template for Textual Division in John’s Gospel,” in Studies on the Text of the New Testament and Early Christianity: Essays in Honour of Michael W. Holmes on the Occasion of his 65th Birthday, ed. Daniel M. Gurtner, Juan Hernández Jr., and Paul Foster, NTTSD 50 (Leiden: Brill, 2015), 222. 57 Gordon D. Fee, The First Epistle to the Corinthians, NICNT (Grand Rapids: Eerdmans, 1987), 355. 58 Although the issue of external evidence dealing with where to place the paragraph in 14:33 has been handled recently by Alesja Lavrinovica, “1 Cor 14.34–35 without ‘in All the Churches of the Saints’: External Evidence,” NTS 63 (2017): 370–389, this issue first came to the forefront for me in 2007 with Delio DelRio and Bill Warren, “Orderly Worship or Silent Women: A Study of 1 Corinthians 14:33” (paper presented at the annual meeting of the Evangelical Theological Society, San Diego, CA, 16 November 2007).

138

Solomon

after the original letter was sent out.59 Editors of critical editions over the years have differed on how to divide the text at 14:33. For instance, NA28, UBS5, and Tischendorf begin a new paragraph in the middle of 14:33, connecting the phrase ὡς ἐν πάσαις ταῖς ἐκκλησίαις τῶν ἁγίων with 14:34–36. The editors of the THGNT, SBLGNT, RP, and WH begin a new paragraph at 14:34, making sure that 14:33 stays together as one sentence. Again, we see a split in how modern critical editions indicate paragraphs in the text of Paul’s epistles.60 Examination of the earliest Greek MS tradition for 1 Corinthians 14, once again, sheds light on how the text should be divided in 14:33. In 𝔓46, reading marks appear on the folio containing 14:33–34. Unfortunately, the bottom of the leaf is damaged where these verses are present. Fortunately, we are left with some data with which to work. First, reading marks are present between 14:33a and 14:33b, but no gap is present at this juncture. Second, a reading mark can be observed at the end of 14:33, but the mark is at the end of the line with 14:34 beginning on the next line in the lacunose portion of the leaf. The evidence is not conclusive. First, there is speculation on whether the reading marks were original to the MS or not. Second, we would need to see the beginning of 14:34 to be sure about any possible paragraphing issues. Third, although a reading mark appears between vv. 33a and 33b, there is no space interval, which we know to be original to the MS. Space intervals are one of the more dominant segmentation features found throughout 𝔓46. This lack of a space interval between vv. 33a and 33b suggests that the original copyist did not see a new break or paragraph beginning in the middle of v. 33, but we must be cautious in this conclusion because of the fragmentary nature of the leaf. Although 𝔓123 is a very small fragment, the middle of 1 Cor 14:33 is extant. Of note, there are no paratextual segmentation features present to suggest a new paragraph break at v. 33b. Unfortunately, 14:34 begins on the next line where only the very tops of letters can be observed with no visible paratextual segmentation features present. A small dot is present but the dot is more than likely the top of the iota of αἱ before γυναικες. Otherwise the dot would be either between the article and noun or the article would be omitted giving us a singular reading. Both Sinaiticus and Vaticanus divide the text at 14:34 keeping 14:33 together as one clause. Sinaiticus ends 14:33 with a large gap on the line before beginning 14:34 with ekthesis. Vaticanus ends 14:33 with a raised dot on the line before beginning 14:34 with a paragraphos mark. So, the early MS tradition 59 See Fee, First Corinthians, 699–705, for a very thorough discussion of the text-critical issue. 60 Major English translations seem just as split as critical editions on where to include a paragraph break.

Early Pauline Manuscripts

139

for this passage offers no evidence that 14:33b should begin a new paragraph. Sinaiticus and Vaticanus both clearly begin a new paragraph at 14:34, while 𝔓46 and 𝔓123 give us no reason to think that v. 33b belongs with v. 34.61 The exegetical implications of whether a paragraph is included at 1 Cor 14:33b or at 14:34 has been debated.62 The difference in meaning could be major, so this is a place where modern critical editions and major English translations need to reconsider how the text is divided. Beyond the discussion of whether 14:34–35 is a later interpolation, the fact remains that when these verses are found elsewhere in Pauline MSS, 14:33 is always intact and v. 33b is never attached to vv. 34–35. If 14:33b begins a new paragraph, the implication is that Paul’s command for women to be silent in churches is a universal command, which seems to contradict other places where Paul assumes women are not silent in churches (e.g., 1 Cor 11:5).63 If, on the other hand, 14:33b is included with 14:33a wrapping up a paragraph and 14:34 begins a new paragraph, then the universal nature of the phrase modifies the character of God being a god of peace and not one of disorder. Given the examination of the earliest Pauline epistles as well as how much is at stake exegetically, modern critical editions should consider the early scribal interpretation or at least give indication of the possible ambiguity of the punctuation here to have a new paragraph begin at 14:34. 4.3 Ephesians 5 The text of Eph 5:21 is attested in the Greek MS tradition before the year 400 ce by three MSS: 𝔓46, 01, and 03.64 In the section of ch. 5 under consideration, the author is interacting with household codes in order to elaborate what it means to live ὡς σοφοί.65 Editors of critical editions over the years have differed 61 Lavrinovica, as well as DelRio and Warren, also examined other MSS, which point to the same conclusion that 1 Cor 14:33 should remain as one sentence and 14:34 should begin a new paragraph. 62 See Marshall Janzen, “Orderly Participation or Silenced Women? Clashing Views on Decent Worship in 1 Corinthians 14,” Direction 42 (2013): 55–70, and Philip B. Payne, “Is 1 Corinthians 14: 34–35 a Marginal Comment or a Quotation? A Response to Kirk MacGregor,” Priscilla Papers 33 (2019): 24–30. 63 It may be argued that since Paul includes ἐν ταῖς ἐκκλησίαις in v. 34, the emphasis is still the same whether or not v. 33b is attached to v. 34 or not. A counterpoint obviously would be that Paul did not include the adjective πάσαις in v. 34 in the same way he included the adjective in v. 33b. 64 I am indebted to Katherine Morgan for first bringing this punctuation issue to my attention in a course on Ephesians at New Orleans Baptist Theological Seminary. 65 Andrew T. Lincoln (Ephesians, WBC 42 [Grand Rapids: Zondervan, 1990], 386) adds further that this is the paraenesis of Ephesians, which begins at 5:15. The fact that this

140

Solomon

on how to divide the text, whether to insert a new paragraph at 5:21 or 5:22. For example, the NA28 and the UBS5 include a paragraph break at 5:21, signifying that the phrase “submitting to one another in fear of Christ” is an introduction to the household code of 5:22–6:9. On the other hand, the THGNT and the SBLGNT have divided the text at 5:22, signifying that the participle in 5:21 belongs with the other four participles that are fleshing out what it means to be filled with the spirit in 5:18. Interestingly, RP and Tischendorf do not have a paragraph break at all (only including ones at 5:15 and 6:1) while WH does not have a paragraph break from 5:15 all the way until 6:21. Major English translations are also split in terms of where to include a paragraph break, with the CEV, NIV, NLT, RSV, and NRSV including one at 5:21 while the ASV, CSB, HCSB ESV, NASB, NET, and NKJV include a paragraph break at 5:22. It is clear there is no consensus on where to include a paragraph division here. Examination of the earliest Greek MS tradition for Ephesians 5 may be suggestive how the text should be divided. In 𝔓46, there are no space intervals between either 5:20 and 5:21 or 5:21 and 5:22. Although a small space exists, these spaces follow nomina sacra, which regularly have a little space on either side to set them apart visually. The lack of space intervals may seem not helpful but also may signal that this section should not contain paragraphs at all, like RP, Tischendorf, and WH. Sinaiticus is only slightly more helpful, with ekthesis at the beginning of Eph 5:18, 5:20, 5:21, and 5:22. The lack of cohesion of segmentation features for the five participles that flesh out what it means to be filled with the spirt does not help the case that 5:21 automatically belongs with 5:19–20. On the other hand, Sinaiticus includes these segmentation features at a much higher rate than other MSS during this early period. In the end, the fact that Sinaiticus breaks the text at 5:22 is not in and of itself significant as the text is segmented frequently in this passage. Turning to Vaticanus, however, we are provided with a clearer instance. The copyist has started a new paragraph at Eph 5:22, including a gap between 5:21 and 5:22 as well as a paragraphos mark in the left-hand margin of the line below. Just as with the critical editions and English translations, there is no consensus from this early group of MSS as to where to divide the text. The exegetical implications of the paragraphing of this text can be far reaching. The inclusion of Eph 5:21 as an introduction to the household code provides the guidelines for all parties involved. Everyone is to be submitting to one another – wives and husbands, children and parents, slaves and masters. The paragraphing issue appears at the paraenesis will have exegetical implications as the author is calling his audience to action.

Early Pauline Manuscripts

141

mutual submission would remove some of the exaggerated force of the imperative given to women to submit to their husbands in 5:22, especially since an imperative verb has more than likely been added over the years and is certainly included in English translations to make the meaning clearer.66 On the other hand, the participle ὑποτασσόμενοι in 5:21 can be seen as the fifth of five participles that indicate what it means to be filled with the spirit in 5:18. In this way, keeping 5:21 with 5:18–20 makes sense. Beginning in 5:22, then, the text is not completely disconnected from 5:21 but would be a further description of what it means to be submitting to one another. In either case, it seems as though 5:21 both looks backward to 5:18–20 and forward into 5:22–6:9. The connective nature of 5:21 explains why some have not provided a paragraph break at all but have kept everything together in one unit. The dearth of early MSS as well as the lack of consensus among the ones available makes this the hardest unit to segment when considering the earliest MSS.67 4.4 1 Timothy 3 The text of 1 Tim 3:1 is attested in the Greek MS tradition before the year 400 ce by only one MS: 01. In light of this unfortunate dearth of early MS attestation, 02 (fifth century) and 06 (sixth century) will also be examined for this passage. In ch. 3, the author is advising his mentee Timothy in matters of church leadership. First Timothy 3:1 marks a transition from the roles of men and women in worship to overseers and deacons. The beginning of 3:1 includes the statement πιστὸς ὁ λόγος, which could either point back to 2:15 or forward to the rest of 3:1. Editors of critical editions over the years have differed on how to divide the text, whether to include πιστὸς ὁ λόγος with the end of 2:15 or with the rest of 3:1. For example, the NA28, the UBS5, and WH keep the phrase with 2:15 and begin a new paragraph. On the other hand, Tischendorf, RP, the THGNT, and the SBLGNT all attach the phrase to the rest of 3:1 with a new paragraph beginning at 3:1.68 Examination of the earliest Greek MS tradition for 1 Tim 3:1 may shed light on how the text should be divided. Unfortunately, there is a lack of early MSS 66 67 68

Not a few witnesses include an imperative verb that was not likely to be part of the initial text in Eph 5:22: 01, 02, 33, 1175, 1505, 1739, 1881, et al. add υποτασσεσθωσαν, 06, 010, 012, MT, et al. add υποτασσεσθε, while 𝔓46 and 03 do not include a verb in 5:22. It is, of course, acknowledged that there are other linguistic considerations that go into dividing a text. In this instance, a survey of major English translations has revealed that the editors have all included πιστὸς ὁ λόγος with the rest of 3:1 beginning a new paragraph. This may be due to the fact that the phrase is used four other times in the New Testament (all in the Pastoral Epistles) and each other instance seems to look forward instead of behind.

142

Solomon

for this part of 1 Timothy. Sinaiticus and Alexandrinus both begin the phrase πιστος ὁ λογος on a new line with the pi extending into the margin. On the other hand, Claromontanus begins a new paragraph after the phrase on both the Greek and Latin side, visually signaling that πιστος ὁ λογος should be read with 2:15. This lack of consensus does not shed much light on the situation as more early MSS would be desired in this instance. The two earliest witnesses to the text of 1 Tim 3:1, however, add weight to the idea that πιστος ὁ λογος should be included as an introduction to the rest of 3:1 beginning a new paragraph. If this is the case, in future editions of NA/UBS texts a change could be made. The exegetical implications of this paragraphing issue are significant. The end of 1 Timothy 2 includes some ambiguous remarks about why a woman should be of a quiet demeanor and not wield authority over a man, including the act of teaching. The author includes reasons why a woman should act in this way by appealing to the story of Adam and Eve adding that women will be saved through childbearing if she also meets certain other criteria.69 Following up this notoriously difficult to interpret logic of the author with the phrase “this is a trustworthy statement” adds to the force of the claims made in 2:9–15. Improper paragraphing at this point has the ability to lend support for women to be excluded from positions of leadership in churches and even in society. The correct placement of the phrase and the beginning of a new paragraph should help readers of NA/UBS, which are two of the most widely used modern critical editions, better interpret the text of 1 Timothy 2 and 3. 4.5 Philemon 7 The text of Phlm 7 is attested in the Greek MS tradition before the year 400 ce by only two MSS: 𝔓139 and 01. In light of this unfortunate dearth of early MS attestation, 02 (fifth century), 04 (fifth century), and 06 (sixth century) will also be examined for this passage. The issue at hand here is whether v. 7 belongs with the typical thanksgiving section or belongs to the body of the letter. Unlike the above examples, editors of critical editions over the years have only slightly differed on how to divide the text.70 For instance, the NA28, the UBS5, the SBLGNT, the THGNT, and RP all include a paragraph break at v. 8. Both Tischendorf and WH, however, do not include a paragraph break at all at v. 7 or v. 8. These two editions include a paragraph at v. 4 and then at

69 Much has been written about these verses, and it is outside the scope of this essay on paratextual segmentation features to solve the issue here. 70 Major English translations are monolithic in including a new paragraph at Phlm 8 and keeping v. 7 with the thanksgiving section of the letter.

Early Pauline Manuscripts

143

v. 21. Tischendorf, however, includes a large gap between v. 7 and v. 8 indicating some sort of break not on the same level as a paragraph. Examination of the earliest Greek MS tradition for Phlm 7 may indicate how the text should be divided. Interestingly, none of the earliest Greek witnesses include a paragraph break only at v. 8. 𝔓139 is lacunose for the punctuation considerations in question, but it can be noted that it appears v. 7 begins a new line but v. 8 does not.71 Both Sinaiticus and Claromontanus include breaks in the form of ekthesis extending the beginning of both vv. 7 and 8 into the left-hand margin. While these two are witnesses to a break at v. 8, the break at v. 7 as well is curious. Both Alexandrinus and Ephraemi Rescriptus contain a paragraph break at v. 7, indicated by ekthesis into the left-hand margin. Interestingly, Alexandrinus also includes an inline gap the size of a letter before v. 8 with what could be a medial dot.72 There is no segmentation feature between vv. 7 and 8 in Ephraemi Rescriptus. In any case, none of these early MSS resemble exactly the modern critical editions in terms of paragraphing. The exegetical implications of this paragraphing issue could be significant. David Russell has examined Philemon’s structure, and although he maintains traditional paragraphing as discussed above, he demonstrates that the hortatory schema of the letter moves from authority in vv. 1–3 to the problem in vv. 7–13 to commands in vv. 14–17 and finally to motivation in vv. 18–22.73 So, while not suggesting that the letter should be divided into these paragraphs, he demonstrates that v. 7 can get along just fine with the body of the letter in terms of scheme. In this way, v. 7 is introducing the problem. The problem for Paul is that Philemon normally “refreshes the hearts of the saints,” but now he is not, which becomes the reason Paul is writing. Additionally, if v. 7 is included with the body of the letter, then the combination of the verb ἀναπαύω and the noun τὰ σπλάγχνα could act as bookends to the main portion of the body, occurring in both v. 7 and v. 20. Paul, then, would be surrounding his imperative verbs with the idea that Philemon must refresh his heart in the same way he has refreshed the hearts of the holy ones. Paul’s purpose in sending this letter to Philemon is to move Philemon to action, 71 The right-hand margin is visible in the surviving text portion. The end of Phlm 6 is visible at the end of the right-hand margin, so one can assume v. 7 begins on the next line. However, it appears that the end of v. 7 must be included on the same line as v. 8 when counting letters. There is always the possibility that a paragraphos was included in one of these places, or not at all. 72 There is a very faint mark of some sort, but it is difficult to discern whether this is an imperfection or a faded dot. 73 David Russell, “The Strategy of a First-Century Appeals Letter: A Discourse Reading of Paul’s Epistle to Philemon,” JOTT 11 (1998): 8–11.

144

Solomon

to change his behavior. Paul, then, would be layering the rhetorical force of his letter by surrounding his commands with motivation, the motivation to refresh both his heart and the hearts of the saints by receiving Onesimus as he would Paul himself. Conversely, one could also make the argument that Phlm 6 and 7 flesh out his prayer in vv. 4 and 5, which suggests that v. 7 definitively belongs with the thanksgiving section.74 In v. 5, Paul gives a reason why he is thankful for Philemon’s church when he remembers them in his prayers: Paul has heard of their faith in the Lord and love for all the saints. In v. 6 he mentions faith again and in v. 7 he mentions love. This repetition and chiastic structure could signal that v. 7 belongs with the thanksgiving paragraph. In either case, this example reveals yet another instance in which the paragraphing of modern critical editions could and should be reexamined. 4.6 Summary In this section, specific examples from the Pauline letter collection were examined for how the text has been divided in the earliest Greek MSS. Keeping in mind that paragraphing for critical editions and English translations should not solely be done in accordance with the MS tradition, the following conclusions can be drawn. First, critical editions are not uniform in terms of their paragraphing. We even have an example where the UBS5 and the NA28 disagree (1 Cor 7:39). While the goal should not be editorial homogeneity in critical editions or translations, this diversity of textual division encourages questioning how we divide texts and upon what criteria we use to make such determinations. From the above examples, it seems as though there are instances where the ancient MS tradition is telling a different story than what is found in modern critical editions. Second, based on the examination of the earliest MSS for the above five passages, more consideration should be given to the earliest MS tradition, which upon examination suggests the following segmentation features. Editors should include a new paragraph break at 1 Cor 7:39. Editors should also join the text of 1 Cor 14:33 together in one sentence with a new paragraph break beginning at 14:34. Additionally, editors should either include a paragraph break at Eph 5:22 or simply keep the text of 5:15 through 6:9 together as one long paragraph. Editors can also consider keeping Philemon 7 with the body instead of with the thanksgiving section as has traditionally been done. Along with these suggestions, scholars and editors should also revisit many other places in the NT in terms of the placement of paragraphs as exegetical implications could be significant. 74 Russell, “The Strategy,” 16–17.

Early Pauline Manuscripts

145

5 Conclusion In this essay, I have demonstrated the need to reconsider the paragraphing of the text of the New Testament based upon an examination of early Greek New Testament MSS. The segmentation features of these MSS sometimes differ from modern critical editions. While these paratextual segmentation features are not the only criteria by which to divide the text, consideration must be given to some of these ancient features of our earliest MSS. For the Pauline letter collection, most MSS predating 400 ce are extraordinarily fragmentary, with 𝔓46 being our most complete MS. Μost of our earliest MSS of the Pauline letter collection contain segmentation features of some sort acting as reading aids for the early church. In the five specific passages examined in this essay, I have suggested changes to some of our critical editions, including the NA28 and the UBS5, where the earliest MSS of Paul’s letter collection indicate new paragraphing features should be included in future editions. While this essay is not comprehensive, the hope is that it will ignite more scholarly study on segmentation features with an eye toward including more data from early MSS in future critical editions of the GNT.

CHAPTER 6

Can Papyri Correspondence Help Us to Understand Paul’s “Large Letters” in Galatians? William Varner 1 Introduction In Gal 6:11 the Apostle Paul begins to conclude his impassioned letter by calling attention to his own handwriting, Ἴδετε πηλίκοις ὑμῖν γράμμασιν ἔγραψα τῇ ἐμῇ χειρί (“see with what large letters I am writing to you with my own hand”). While the reference to his writing the subscription “with my own hand” is not unique among his epistles (see 1 Cor 16:21; Col 4:18; 2 Thess 3:17; and Phlm 1:19), his calling attention to the size of his letters does not appear elsewhere in his writings. The issue of the epistolary conventions of the epistles of Paul and other New Testament writers has been given some serious attention in the last generation.1 Some of these works have also focused on the subscriptions in these epistles.2 The issues to be addressed in this chapter are the purpose(s) of these epistolary subscriptions and the roles of both the “secretary” and the “author” in the composition of the letters. While we will survey these subjects, the main 1 Limiting our bibliography to just the last thirty years, monographs in English include the following works on the form and function of the Pauline epistles: J. L. White, Light from Ancient Letters (Philadelphia: Fortress, 1986); E. Randolph Richards, The Secretary in the Letters of Paul, WUNT 2/42 (Tubingen: Mohr, 1991); Jerome Murphy-O’Connor, Paul the Letter Writer (Collegeville: Liturgical, 1995); E. Randolph Richards, Paul and First Century Letter Writing: Secretaries, Composition and Collection (Downers Grove, IL: InterVarsity, 2004); Hans-Josef Klauck, Ancient Letters and the New Testament (Waco, TX: Baylor University Press, 2006); Stanley E. Porter and Sean A. Adams, eds., Paul and the Ancient Letter Form, PAST 6 (Leiden: Brill, 2010); Jeffrey A. D. Weima, Paul the Ancient Letter Writer: An Introduction to Epistolary Analysis (Grand Rapids: Baker, 2016). Klauck’s volume also provides adequate attention to monographs in German. Many of these books also list the dozens of related articles on this subject in scholarly journals. 2 Jeffrey A. D. Weima, Neglected Endings: The Significance of the Pauline Letter Closings, JSNTSup 101 (Sheffield: JSOT Press, 1994); Steve Reece, Paul’s Large Letters: Paul’s Autographic Subscriptions in the Light of Ancient Epistolary Conventions, LNTS 561 (London: Bloomsbury T&T Clark, 2017). The appearance of Reece’s volume during the research for this chapter was a special boon, and it has influenced my own approach to this subject.

© William Varner, 2023 | doi:10.1163/9789004537972_008

Paul ’ s “ Large Letters ” in Galatians

147

purpose of this chapter is to suggest why Paul called attention to his “large letters” in his letter to the Galatians. In concert with the purpose of this volume, the task of this chapter is to explore what help can be found in the abundance of papyri letters that survive from antiquity. Are there patterns in these documentary papyri that can shed light on what Paul intended to convey by calling attention to his “large letters” in the Galatians subscription?3 2

Interpretations of Paul’s Large Letters in His Galatian Subscription

Before considering what light the papyri letters might shed on these issues, it would be helpful to examine the various ideas that have been offered by Christian commentators through the centuries about the meaning of Paul’s “large letters.” This survey will list a representative number of commentators from various stages in church history. Apart from a few rather idiosyncratic ideas, there have been three main categories of interpretations about the unique expression used by Paul in Gal 6:11. A sampling of scholars advocating these positions will be noted. 1. The apostle is commenting on the size of his handwriting that begins at this point after he has taken over the writing from the secretary/ scribe/amanuensis to whom he has dictated the previous portions of the epistle.4

3 Papyrology concerns itself with a range of texts often grouped into two main categories: documentary and literary. Documentary papyri include official writings such as public edicts, announcements, official petitions, as well as private affairs such as letters, legal contracts, and accounts. Literary papyri are composed of literature such as poetry (Homer, lyric poems, tragedy, etc.) and prose (historiography, philosophy, rhetoric, etc.) as well as a class of texts sometimes referred to as “subliterary” (grammatical treatises, school exercises, magic and astronomical texts). In this chapter, we will be concerned with documentary papyri, more specifically the letters and contracts in letter form. We also will not concern ourselves with a discussion of the various types of ancient letters but settle for Cicero’s brief explanation of two types of letters: public and private letters (Pro Flacco 37). Later we will consider the nature of personal and official letters and heir bearing on the letters of the New Testament, especially those of Paul. 4 J. B. Lightfoot, St. Paul’s Epistle to the Galatians (Cambridge: Macmillan, 1865), 211–212; C. I. Scofield, The Scofield Reference Bible (New York: Oxford University Press, 1909), 1248; Donald Guthrie, Galatians (London: Nelson, 1969), 148.; F. F. Bruce, The Epistle to the Galatians: A Commentary on the Greek Text, NIGTC (Grand Rapids: Eerdmans, 1982), 268.

148

Varner

2.

The apostle is commenting on his handwriting throughout the entirety of his epistle, which he has written in his own hand.5 3. The apostle is commenting, not on his handwriting, but rather about the “large” length of the epistle to the Galatians.6 As mentioned, there are a few other novel views, but these three approaches represent the vast number of commentators through the years. I have cited the commentators by name and approximate date, and their comment about the “large letters” can be found at the point in their respective commentary where Gal 6:11 is addressed. As samples of these explanations, there follows a few more detailed explanations from some of the writers mentioned above. Interpretation One: That Paul at This Point Is Taking Over from His Secretary “The boldness of the handwriting answers to the force of the apostle’s convictions. The size of the characters will arrest the attention of his readers in spite of themselves.”7 “Paul added the concluding remarks in his own hand to bring a warm touch to his concluding appeal. The largeness of his letters would have contrasted with the smaller and neater writing of his scribe. He explicitly mentions the change of hands because most of his audience would have heard rather than seen the text.”8 “Paul took up the pen in order to write his concluding comments in large letters for emphasis.”9 An interpretation was set forth by C. I. Scofield and has greatly influenced popular interpretations ever since, “the apostle was … afflicted with ophthalmia, a common disease in the East, to the point of almost total blindness (e.g., Gal 4:13–15). We cannot know with what 2.1

5 John Chrysostom, Nicene and Post-Nicene Fathers, vol. 13, ed. Philip Schaff (New York: Christian Literature Company, 1889), 45–46; Henry Alford, The Greek New Testament, 3 vols. (London: F&J Rivington, 1849), 3:64; Theodor Zahn, Introduction to the New Testament, 3 vols. (Edinburgh: T&T Clark, 1909) 1:172–173; Kenneth S. Wuest, Wuest’s Word Studies from the Greek New Testament, 4 vols. (Grand Rapids: Eerdmans, 1973) 1:175; David A. deSilva, Galatians: A Handbook on the Greek Text, BHGNT (Waco, TX: Baylor University Press, 2014), 139. 6 Marius Victorinus, In Epistulam Pauli ad Galatas Commentariorum Libri Duo, on Gal 6:11; Desiderius Erasmus, Novum Testamentum (Basel: Johann Froben, 1516), 411; Matthew Henry, An Exposition of the Old and New Testament (London: J. Stratford, 1706), 2305; Albert Barnes, Notes on the New Testament. II Corinthians and Galatians (London: Blackie & Son, 1884), 396. I have not been able to find a published commentator from the last century and a half who has espoused this view. 7 Lightfoot, Galatians, 211. 8 Guthrie, Galatians, 148. Guthrie’s last comment on the “oral” purpose of Paul’s statement about the “large letters” is significant and we will return to it later in this chapter. Guthrie does not, however, mention this specific point in his later discussion of Galatians in his New Testament Introduction, 3rd ed. (Downers Grove, IL: InterVarsity, 1970), 450–471. 9 Bruce, Galatians, 268.

Paul ’ s “ Large Letters ” in Galatians

149

pain and difficulty, with his own hand, in the ‘large letters’ his darkened vision compelled him to use.”10 Interpretation Two: That Paul Wrote the Entire Epistle in His Own Hand Paul wrote the entire letter himself “because of the urgency of the situation in the Galatian church. By the expression ‘with what large letters’ he seems to me to signify not the size but the formlessness of his letters, as if he had said, ‘Although skilled in writing, nevertheless I have been compelled to write with my own hand to stop the mouth of these sycophants.”11 “Paul meant that he wrote the entire epistle in his large (i.e., unsightly) handwriting, perhaps from the weakness of his eyes, perhaps by choice to show that he has no desire to make a fair show outwardly (which is his theme throughout the conclusion of his letter).”12 “Paul wrote the entire epistle himself. There are two possible reasons for Paul’s large handwriting: Paul was not accustomed to writing, and therefore was not as refined as that of a scribe; Paul’s large letters were the result of his hand being injured by the flogging he had received in Philippi (Acts 16:22–23).”13 There are contemporary advocates of this view as well. “The adjective πηλίκοις refers to the relatively large size of Paul’s handwriting, perhaps a reflection of his anxiety and emotional agitation as he wrote this particular letter.”14 Another interpretation that again has influenced popular views of Paul’s comment was set forth by Kenneth Wuest: “Paul wrote the entire epistle in large uncial letters. His large letters are attributable to an oriental eye disease that not only gave him a repulsive appearance but rendered him almost totally blind.”15 2.2

Interpretation Three: That Paul Is Commenting on the Greater Length of Galatians Victorinus (ca. 400) translated Gal 6:11 as follows: “Behold with how many letters [quantis litteris] I wrote to you with my own hand.”16 Erasmus, on the 2.3

10 11 12 13 14 15 16

Scofield, Scofield Reference Bible, 1248. Chrysostom, Galatians, NPNF 13:45–46. Alford, Greek New Testament, 3:64. Zahn, Introduction, 1:172–173. deSilva, Galatians, 139. Wuest, Word Studies, 1:175. Victorinus, on Gal 6:11. His translation of πηλίκοις … γράμμασιν differs from the Vulgate of his contemporary Jerome, which reads: “See with what kind of letters (qualibus litteris) I have written to you.” Victorinus’s translation emphasized the number of letters while Jerome’s emphasized the kind of letters.

150

Varner

other hand, translated the expression as singular: “You Galatians see how dear you are to me, who wrote so long a letter [proxilam epistolam] to you with my own hand”17 Perhaps the influence of Erasmus was seen in the influential (mis)translation of Gal 6:11 witnessed in the venerable Authorized Version of 1611: “Ye see how large a letter I have written unto you with mine own hand.” This confusion also appeared in a number of English and American popular authors in the subsequent centuries. “As a particular mark of his respect for the Galatians, Paul had written this large letter with his own hand, and had not made use of another as his amanuensis, and only subscribed his name to it, as he was wont to do in his other epistles.”18 “According to this, it was proof of special interest in them, and regard for them, that he had written to them a whole letter with his own hand. Usually he employed an amanuensis, and added his name, with a brief benediction or remark at the close.”19 There is a reason why the works that suggest this view are all from the nineteenth century and earlier. As we will see, the evidence of the papyri letters, which were not published until the early twentieth century, renders this third view highly unlikely.20 3

The Role of the Secretary in Ancient Letter-Writing and the “Secretaries” of Paul

As was mentioned in the opening paragraph, the last generation has witnessed serious attention to how ancient letters (and other documents for that matter) were composed, written, and sent. Limiting our consideration to Greek papyrus letters, the following is a simple summary of the conclusions reached by most of these articles and monographs.21 The everyday realia of letter composition involved the writing materials, and here in Gal 6:11 we have reference even in the New Testament to the material used (papyrus), the medium used (the ink), and the writing instrument employed (the pen). Second John 12 and 3 John 13 refer to the κάλαμος (the 17 Erasmus, Novum Testamentum, 411. 18 Henry, Exposition, 2305. 19 Barnes, Notes, 396. This is a later edition of his commentary (1884) that was originally published in 1847. 20 Although not addressing the specific issues addressed in this chapter, an engaging account of early Christian manuscripts and their modern discoverers is given by Brent Nongbri, God’s Library: The Archaeology of the earliest Christian Manuscripts (New Haven, CT: Yale University Press, 2018). 21 For convenience, we will basically follow the conclusions of Richards, Porter and Adams, Reece, and Weima in the works cited in footnotes two and three.

Paul ’ s “ Large Letters ” in Galatians

151

reed pen), the μέλαν (the black ink), and the χάρτης (the papyrus sheet).22 The evidence is also overwhelming from antiquity about the active involvement of secretaries (paid letter writers), who were an almost universal part of the ancient letter-writing process. The secretary wielded a wide sphere of influence from being little more than a transcriber to sometimes becoming the complete composer in the sender’s name. “The role played by the secretary depended on how much control the author exercised at that particular moment in that particular letter, even shifting roles with the same letter.”23 The Pauline letters explicitly mention or imply the use of secretaries six times (Rom 16:22; 1 Cor 16:21; Gal 6:11; Col 4:18; 2 Thess 3:17; Phlm 19). In the Romans reference, the secretary, Tertius, identifies himself and contributes his own greeting. While this illustrates how Paul’s secretaries may have injected their own material, it is safer to conclude that it becomes obvious to the reader when this takes place, as in the Romans reference. Paul’s letters, even a short one like the one to Philemon, still convey a sense of authority and in this way do not exactly reflect the multitude of personal letters exchanged in antiquity often dealing with insignificant personal matters. Paul’s letters were intended to be read not just privately but publicly and even to be exchanged between the churches. As an example, see Col 4:16: “When this letter has been read among you, have it read also in the church of the Laodiceans; and see that you also read the letter from Laodicea.”24 Paul’s extended epistolary style must recognize the length of time involved in letter composition, the use of secretaries, and the probability of various drafts before the final letter was sent. Most of these studies argue that at least two final copies of each letter were made – one sent to the recipients and the other kept for Paul’s personal records. Secretaries were paid, and as a conservative estimate Richards thinks Paul’s longest letter (Romans) would have cost as much as the equivalent of $2,200 and his shortest (Philemon) around $100. “Weeks, if not months, of work likely went into a letter” in addition to “considerable 22

Cicero referred to the Latin terms for these three types of material as calamus, atramentum, and charta, respectively (Cicero, ad Quintum Frtarem 2.15.1). Almost all of what is referred to in Greek letter writing is confirmed by the letters of Cicero which survive in copies, but we will limit our investigation to the actual surviving copies of Greek letter papyri. 23 Richards, Secretary, 80. See also Jeffrey A. D. Weima, “Sincerely, Paul: The Significance of the Pauline Letter Closings,” in Paul and the Ancient Letter Form, 307–345, here 337–338; Reece, Paul’s Large Letters, 13–14, 204–207; and Weima, Paul the Ancient Letter Writer, 193–194. 24 Even if “Colossians” was not actually written by Paul, it likely reflects the practice of exchanging and circulating his letters.

152

Varner

expense.”25 Richard’s comments should be nuanced if Paul utilized co-workers for these tasks. In other words, if he used professional secretaries, these would have been the costs. Having been written, the ancient letters had to be sent. Despite the postal system developed by Caesar Augustus in the Roman Empire, many letters were sent through “occasional” travelers who were already going to the intended destination or who were sent privately at the expense of the sender. An example of this is the role of Phoebe who is mentioned in Rom 16:1–2 as one such carrier of the letter from Cenchreae to Rome. Travel in the ancient world shared no convenience with modern standards, and letters were usually conveyed by friends. Depending on the time of year and difficulties of the journey, a letter could take a few days or multiple weeks to be delivered.26 Paul probably used these “occasional” carriers to deliver his early letters like Galatians and 1–2 Thessalonians but then began using members of his “apostolic team” as private letter carriers to ensure safe delivery and to serve as interpretive readers to the congregations. The question of how Paul’s letters eventually came together into a collection is not entirely clear. Richards suggests, based on the common practice of authors keeping copies of their letters, that Paul himself was responsible for collecting all his letters in a possible notebook format.27 These notebooks were then posthumously circulated as a collection (by Luke?). Could Paul be referring to this collection in his request to Timothy in 2 Tim 4:13?28 He writes: φέρε

25 Richards, Secretary, 169. We must admit, however, that we cannot be dogmatic about these costs. See also Reece, Paul’s Large Letters, 14–15. 26 Eldon Epp’s study of ancient letter-carrying suggests that people actually expected letters to be delivered quickly. See Eldon Jay Epp, “New Testament Papyrus Manuscripts and Letter Carrying in Greco-Roman Times,” in The Future of Early Christianity: Essays in Honor of Helmut Koester, ed. Birger A. Pearson (Minneapolis: Fortress, 1991), 35–56. 27 Richards, Secretary, 218–223; Reece, Paul’s Large Letters, 209–212. Stanley Porter has offered a sustained argument that Paul and his “literary team” kept copies of his letters and these formed the group from which emerged the “Pauline canon.” See Stanley Porter, The Apostle Paul: His Life, Thought, and Letters (Grand Rapids: Eerdmans, 2016), 169–179, which draws on a number of his previous articles that are cited there. Later in this chapter, I will offer physical evidence from the papyri letters about ancient writers keeping copies of their correspondence. 28 I recognize the controversy over the Pauline authorship of some of the traditionally Pauline writings such as Colossians and 2 Timothy. While I personally affirm that Paul was responsible also for these letters, it must be acknowledged by all that these letters reflect ancient epistolary practice and the existence of an early collection of his letters.

Paul ’ s “ Large Letters ” in Galatians

153

καὶ τὰ βιβλία, μάλιστα τὰς μεμβράνας (“bring also the rolls [his personal copies of the letters?], especially the parchments” [various Old Testament texts?]).29 4

The Papyri Letters and Where to Find Them

The discovery of papyrus documents in the drier climates of Judea and Egypt began in the nineteenth century and continues until today.30 The pioneering discoveries of Bernard P. Grenfell and Arthur S. Hunt in Egypt, especially at Oxyrhynchus, led to the monumental series, The Oxyrhynchus Papyri, which has now reached over eighty volumes.31 It should not be overlooked, however, that a number of other volumes and series have published the papyri finds at other sites in Egypt and Judea.32 Furthermore, the present physical locations of these papyri are spread over the world in dozens of academic institutions, museums, and private collections.33 For the purposes of this chapter, among the over one hundred thousand papyri are thousands of letters, in various levels of preservation.34 How can the researcher even begin to investigate this large number of papyri to see what 29 Skeat (and others) have proposed that the μεμβράνας here were codices: T. C. Skeat, “‘Especially the Parchments:’ A Note on 2 Timothy 4:13,” in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup 113 (Leiden: Brill, 2004), 262–266, originally in JTS 39 (1979): 173–177. My thanks to Larry Hurtado for calling my attention to this article. 30 An indispensable source for the overall study of this subject is Roger S. Bagnall, ed., The Oxford Handbook of Papyrology (Oxford: Oxford University Press, 2009). 31 Bernard P. Grenfell et al., eds., The Oxyrhynchus Papyri (London: Egypt Exploration Fund, 1898–2018). 32 Rather than burdening this chapter with bibliography found elsewhere in this volume, I will simply mention that the story of the discovery of papyri in various locations and their subsequent publication, is summarized well by Hélène Cuvigny, “The Finds of Papyri: The Archaeology of Papyrology,” in The Oxford Handbook of Papyrology, 30–58. 33 Ancient letters survive also on such material as wax in wooden tablets, such as the Vindolanda Tablets found in a Roman encampment at Hadrian’s Wall. While not written on the papyri which concern us, certain features such as a secretary’s hand and a cruder subscription hand illustrate some of the features that we will examine in this chapter. See the volumes by A. K. Bowman and J. D. Thomas, The Vindolanda Writing-Tablets, 3 vols. (London: British Museum Press, 1983–2003). 34 “About 550 letters from the Ptolemaic period have been edited, but half of these are from the Zenon archive. By contrast, about 1900 have been published from the first three centuries of Roman rule” (Bernhard Palme, “The Range of Documentary Texts: Types and Categories,” in The Oxford Handbook of Papyrology, 358–394, here 361). While it is difficult to be precise due to the inclusion at times of letters from other literary contexts, the following summary is probably as close as we can get to the current situation. “There are extant several dozen letters from the caves … of eastern Judea, several hundred from

154

Varner

light they may possibly shed on Pauline epistolography? Furthermore, even the dozens of published papyri collections, many with excellent transcriptions and translations, only occasionally provide actual images of the papyri being published. The physical and visual features of the scribal habits in these letters cannot be “seen” even in good transcriptions. Finally, while there are many published (printed) images of these letters, there are still thousands more that have not appeared in printed books. Thus the researcher must depend on the databases published online that often include high resolution images of the documents. Therefore, I have not only examined the photographs of papyri in printed collections but have heavily depended on such valuable online resources as the Duke Database of Documentary Papyri, as well as other sources mentioned below.35 In recent years a site called “The Papyrological Navigator” (www.papyri.info) has emerged that seeks to provide access to all these databases under one portal. “Papyri.info aggregates material from the Advanced Papyrological Information System (APIS), Duke Databank of Documentary Papyri (DDbDP), Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens (HGV), Bibliographie Papyrologique (BP), and depends on close collaboration with Trismegistos, for rigorous maintenance of relationship mapping and unique identifiers.”36 In even a cursory overview of this material, it becomes obvious that many types of papyri survive other than letters, such as legal documents (sometimes in the form of letters), contracts, economic texts of all sorts, government announcements, petitions, and declarations of all sorts.37 We will be concerned only with the letters in this chapter, but the number that have survived is not exactly clear. A number of other ancient letters have come down to us, but not their originals. While the correspondence cited by authors like Josephus as well as the letter collections of such ancient authors as Cicero and Cato, provide some fascinating insights into ancient epistolography, we will limit our examination to the physical features in surviving papyri letters which can be

Vindolanda (in Latin), and several thousand from the sands of Middle and Upper Egypt (in Greek, Latin, and Demotic)” (Reece, Paul’s Large Letters, 11). 35 The Logos Bible Study System, published by Faithlife, offers the Duke Database of Documentary Papyri as a downloadable resource. The transcriptions are clear but the visual features of the various hands in the documents cannot be discerned without accompanying images of the papyri. 36 www.papyri.info, accessed 26 September 2017. 37 For a convenient summary of these types of papyrus texts, see Palme, “The Range of Documentary Texts,” 358–394.

Paul ’ s “ Large Letters ” in Galatians

155

“seen” through the images.38 While early Christians were also letter writers, and official letters have come down to us in the New Testament and through such early church writers as Ignatius and Polycarp, the earliest surviving originals date from the mid- to late-third century from Oxyrhynchus.39 However, the figure of “several thousand” of surviving letters (see n. 35) may not be that helpful because many do not survive with their concluding subscriptions clearly visible. This is important because the focus of this study is what the handwriting in the subscription reveals to us about the author’s relationship to the one who penned the body of the letter. The handwriting of the subscription and of the body of the letter are important also in addressing the issue of Paul’s “large letters” in Gal 6:11. Therefore, of the less than one thousand letters that have legible subscriptions and are photographed, a representative sample will be provided from the four hundred letters that I have personally examined through published books and through their images available in “The Papyrological Navigator.” This examination is based on a survey of the 80 volumes of the Oxyrhynchus Papyri that have been published as of January 2018.40 It is acknowledged that some letter writers did not make use of a secretary, especially the brief personal letters between family members. These letters indicate no separate hand that writes the subscription. It is quite surprising, however, how many letters do indicate a different hand, thus signaling a secretary or amanuensis who was utilized to write the body of the letter. Furthermore, and also anticipating my conclusions, the second hand is uniformly cruder and less skilled in its orthography. That second hand, however, is not always larger and sometimes is smaller or about the same size. But in every example that I have examined, the second hand is not as skilled as the hand that wrote the body of the letter. I have focused in this study on letters with larger second hands because they help to illustrate Paul’s “large letters” in Gal 6:11.

38

The publication of Antonia Sarri, Material Aspects of Letter Writing in the Graeco-Roman World: c.500 BC–c. AD 300, Materiale Textkulturen 12 (Berlin: Walter de Gruyter, 2018), took place at the very end of my research. It is an invaluable resource on the realia of letter writing from 500 BC to AD 300. I will mention her observations on the hand of the secretary in a later section of the chapter. 39 See the excellent compendium of Christian texts from Oxyrhynchus by Lincoln Blumell and Thomas Wayment, Christian Oxyrhynchus: Texts, Documents, and Sources (Waco, TX: Baylor University Press, 2015). The authors cite the earliest distinctively Christian letters found there as dating from the mid- to late-third century (459–609). 40 See n. 32.

156

Varner

In the following section, some representative examples of Greek letters from the same general period as those written by Paul will be cited, although some of the examples date from the second and early third century. Unfortunately, space does not allow for a survey of the “Jewish” papyrus letters written in Hebrew and Greek that were discovered in such Judean locations as the Wadi Murabaat, Nahal Hever, and Masada. Suffice it to observe at this point that the physical features of these letters are also consistent with what is observed in the Greek letters recovered in Egypt. 5

The Visual Evidence of the Papyri Letters

Because this is only a chapter length study, I will first offer some deductive conclusions and then seek to illustrate them from the evidence which we have. These conclusions are as follows: (1) ancient letter authors utilized secretaries, with Paul giving evidence he did so in Gal 6:11ff.; (2) the hand of the author was cruder and often larger than the secretary’s, and Paul’s larger hand in Gal 6:11ff. is evidence of his own personal handwriting at this point in the letter; (3) consistent with the evidence of these ancient letters, there is nothing unexpected or unique in the larger letters that Paul uses in Gal 6:11ff.; (4) the purpose of his using larger letters is explained in the following context of Gal 6:12–18 and should not be sought in “modern” analogies or reasons; and (5) the unique statement of calling attention to his larger letters is best understood as a subtle rhetorical irony by Paul that is also explained in the following context. Following are some examples of ancient letters and subscriptions from the first to the third centuries, with brief comments on what can be “seen” in the handwritings. This letter (Figure 6.1)41 dates from an internal reference to AD 66 and is from Harmiysis to Papiscus. In the last four lines of the 26-line letter, one can discern in the last three lines of text that the “author” swears by “Nero Claudius Caesar Augustus [σεβαστον] Germanikos, Emperor [αυτοκρατορ].” Then follows a four-line subscription in a larger and cruder hand. “I, Apollonius, one of the men of Papiscus the strategos, who notes the seven lambs.”

41 Adolf Deissmann, Light from the Ancient East (repr., Grand Rapids: Baker, 1978), 173. The text of the papyrus was published as P.Oxy. II 246. The manuscript is #4053 in the Cambridge University Library. Image is courtesy of Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens.

157

Paul ’ s “ Large Letters ” in Galatians

Figure 6.1 P.Oxy. 246

158

Varner

What is most surprising, however, is the comment on the letter by Deissmann: The handwriting of this document is interesting on account of the clear, almost literary uncials of the main text, sharply distinguished from the cursive signatures of the attesting officials. We must imagine this state of things reversed in the case of the Epistle to the Galatians: the handwriting of the amanuensis of Gal. 1:1–6:10 was probably cursive, and the autograph signature of St. Paul the stiff heavy uncials of a manual laborer; the contrast was just as great.42 What is surprising, with due respect to the great Deissmann, is that the format of this papyrus is more likely the case for Paul’s letter! We should rather envision that Gal 1:1–6:10 would have looked like the uncial letters of the secretary writing for Harmiysis to Papiscus, and Paul’s hand like the cursive subscription of Apollonius! Whatever be the final judgment on Deissmann’s comment, the change from the finer hand of the secretary to the slightly larger and cruder hand of the subscription is evident.

Figure 6.2 P. Col. VIII 216

This letter (Figure 6.2), with an Alexandrian provenance, dates from the later first century AD.43 The most significant feature is the one-line subscription: 42 Deissman, Light from the Ancient East, 173–174. 43 P. Col. VIII 216. Courtesy of Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens.

159

Paul ’ s “ Large Letters ” in Galatians

ἐρρῶσθαί σε εὐ̣ ̣χόμεθα, “we pray you are well.” Varying forms of this greeting/prayer were quite common in the first three centuries and almost every surviving letter has some form of this verb, here as a complementary infinitive but often simply as the imperative ἐρρῶσσο. Reece comments: “A private letter

Figure 6.3 P. Mich. V 351

160

Varner

from Severianus to his brother Ammonius, with body of the letter in a small, regular, bilinear hand, followed by a three-word farewell formula in a much larger, sprawling yet graceful hand.”44 Dating from AD 44, this papyrus (Figure 6.3) is a formal contract regarding some properties with twenty-two lines in the small and linear hand of a professional secretary with the subscription in a much larger and cruder hand.45 Not only does this date closer to the end of Paul’s letter to the Galatians, it may very well approximate the appearance of the two styles of handwriting at the end of the letter. This letter (Figure 6.4) dating from the twelfth year of Antoniunus Pius (AD 149) is from Ailios Aphrodisias to Ammonius. I include this letter, even though the size of the hand in the subscription is not much different than that of the body; it is simply slanting and a bit cruder. There are, however, other more important features. While the subscription is in a hand different from the more professional hand in the body, the secretary identifies himself. Σαραπάμ̣[μων] γρ[αμματεὺς] ἐπ̣ή[νεγκα] (“The secretary Sarpammon, brought this”). This act of self-identification of the secretary parallels the role of Tertius, who mentions himself in Rom 16:22 (Peter also mentions Silvanus in 1 Pet 5:12). It also exemplifies that the secretary was the bearer of the letter to Ammonius, a task that some of Paul’s secretaries may have performed (Epaphroditus to the Philippians?). Finally, since the letter was sent from Oxyrhynchus, it exemplifies how a copy of the letter must have been kept, because this letter was found in Oxyrhynchus. This supports the suggestion that Paul kept copies of his letter (see earlier discussion and footnote 27). The secretary takes the pen again in the last two lines to date the letter in the twelfth year of Emperor Antoninus Pius (AD 149). This letter (Figure 6.5), although dating later than the New Testament (ca. AD 300), is an excellent example of the precise, although cursive, hand of a secretary compared to the scrawling larger hand of the “author” Heracles, as can be discerned at beginning of the six-line subscription.46 We have no certainty that Heracles was of poorer education; only that his handwriting was simply not as skilled as that of his secretary. Many believe that the wealthier 44 Reece, Paul’s Large Letters, 168. It is surprising, even perplexing, that Sarri questions that this is a separate hand (Sarri, Material Aspects, 162). She argues that the subscription was not by the “author” as different from the secretary but was the same person. Closer examination of both sections of the brief letter simply does not support her contention. 45 P. Mich. V 351. Papyrology Collection, University of Michigan. Image courtesy of Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens. 46 P.Oxy. XXXVI 2770. Image taken from Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens.

Paul ’ s “ Large Letters ” in Galatians

Figure 6.4

P.Oxy. XLIX 3472 Image taken from Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens

161

162

Varner

Figure 6.5 P.Oxy. 2770

an author was, the less likely he was to write his own letters. He could hire a secretary like Heracles did here. Two more examples of letters from the first century follow. The first has only a slightly larger but cruder hand in the subscription, while the second has a smaller but cruder hand.

163

Paul ’ s “ Large Letters ” in Galatians

Figure 6.6 P.Oxy. 3487

164

Varner

This letter (Figure 6.6) is from AD 65.47 There are two second hands, one of 9 lines in cursive and one in uncials with larger and very crude block letters stating: “I, Kefalas, her husband, have added my signature.” Lucius Bellenus Gemellus writes to his own Epagathos, AD 94 (Figure 6.7).48 Sarri writes of this letter: “It displays a different professional-looking hand and layout, suggesting that this letter was penned by a professional writer.”49 Although letters like these could be multiplied, these seven examples have been included to make the point that ancient letters, especially the more official the letter, were written by secretaries hired by the author and that the author then added their own subscription in a cruder and often in a larger hand.50 The last image that I wish to include is not the original letter but is a copy that dates probably a century and a half after Paul’s letter to the Galatians. The papyrus is the earliest copy of Galatians, as well as of at least nine other letters of Paul, and it is a part of the Chester Beatty Papyri, P46. I include this even though it is different from the type of epistolary evidence that I have submitted this far. Upon closely examining the last recto leaf of Galatians, I noticed something that I submit for the reader’s consideration. Here follows the page of the end of Galatians and the very beginning of the next letter in the collection, Philippians. I was closely examining the image of this leaf (Figure 6.8)51 and noticed the variant reading that it contains: ηλικοις instead of the better supported πηλίκοις in line 3.52 I looked at the text in P46 and wondered if the scribe would be bold enough to enlarge his own hand in imitation of Paul. As I looked closer, I became convinced that he did. Further examination of the words before and after the verse confirmed my initial impression. While I did see some words in other lines that seemed close to the size of the letters in 6:11 overall, the first and last words of 6:11 seem to be slightly larger than the surrounding words! Perhaps taking a closer look at the three lines that end 6:10 and then start 6:12 will also convince the reader (Figure 6.9). 47 P.Oxy. XLIX 3487. Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens. 48 P. Fayum 110. Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens. 49 Sarri, Material Aspects, 132. 50 For additional examples, see the dozens of letters discussed in Reece as well as numerous other examples in Sarri. 51 The manuscript is in the Chester Beatty Library, Dublin. The image of P46 is courtesy of The Center for the Study of New Testament Manuscripts. 52 Although the P46 spelling also appears in Vaticanus and family 33, the rest of the manuscript tradition supports πηλίκοις.

165

Paul ’ s “ Large Letters ” in Galatians

Figure 6.7 P. Fayum 110

166

Figure 6.8 P46 Gal 6:10–Phil 1:1

Varner

Paul ’ s “ Large Letters ” in Galatians

167

Figure 6.9 P46 Gal 6:10–12

The last word on the first line of Figure 6.9 is the first word in 6:11, the imperative Ἴδετε. The first word in the third line here is the last word in 6:11, the noun χειρί. One should easily discern the size difference from the preceding and subsequent words. If the viewer is not convinced, perhaps there is another approach that is more convincing. The scribe may have intentionally decreased the size of the letters immediately before Ἴδετε and after the final χειρί. This would have made the beginning and end of the verse appear to be larger than the preceding and following words. In any case, my suggestion is that the scribe of P46 is nodding his head to Paul by attempting to portray what Paul called his “large letters.”53 To keep this suggestion from becoming a sidebar to a more serious question, I believe that this scribal action may have at least one interpretive takeaway relative to the main subject of this chapter. While it does not answer the question of why Paul may have called attention to his “large letters,” it may keep us from following one wrong path. The scribe did not think that Paul wrote his entire epistle in large letters like those interpreters in “Interpretation Two” above, nor that his reference referred to the large size of the entire epistle to the Galatians as those advocating “Interpretation Three.” The scribe believed that Paul began his large letters in 6:11, although Paul did not limit them to 6:11 as the P46 scribe does but continued to use that hand through to the end of the epistle. 6

Why Paul Called Attention to His Large Letters

So what has our survey of ancient letters taught us about the meaning of Gal 6:11? I suggest that the value of this study has been largely negative. In 53

None of the works analyzing the habits of the scribe(s) who prepared P46 call attention to this characteristic in Gal 6:11. See, e.g., the thorough discussion of the codex in James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill, 2008), 199–358; and Gunther Zuntz, The Text of the Epistles (London: Oxford University Press, 1953), 58–159.

168

Varner

other words, our examination of ancient epistolary subscription negates most of the reasons that interpreters have suggested for the large letters. In his subscription, Paul was doing what the authors of ancient letters normally did when they picked up the pen from the recording secretary. He wrote his final subscription in an untrained hand, and for Paul at least, that was a larger hand. We need to reject the many suggestions that often are anachronistic reasons borrowed from modern practices. The predominant view among commentators on Galatians is that Paul wrote in large letters to emphasize the importance of what he is writing. This could be compared to the modern practice of typing in all capital letters or bolding the font.54 The problem with this suggestion is two-fold. (1) It imports back into the ancient world a modern practice, assuming that ancients also emphasized their points in this way. (2) It ignores that fact that no example has ever been found from the ancient world of writing in large letters to stress the importance of the words. Another suggestion that is often made by those who even recognize the above evidence is that Paul wrote in large letters to authenticate this letter. There is evidence that the apostle had to deal with fraudulent letters written in his name (cf. 2 Thess 2:2 and 3:17). This is the most logical reason that has been offered, but I warn about one often overlooked fact. The authentication in this letter is not based on Paul’s large letters, a statement utilized only here in his epistles. The authentication was based on his statement that he was writing from this point onward “in my hand” (τῇ ἐμῇ χειρί). This was also the point that he makes in the four other times he uses this expression (1 Cor 16:21; Col 4:18; 2 Thess 3:17; and Phlm 19). The authenticating act of the epistle was not the large letters; it was his writing in his own hand! Paul does not mention any large letters in those other occasions when he mentions his hand.55 We must settle for the unexciting but necessary explanation that there was nothing special at all in his writing in larger letters, because that was the expected result of an author untrained in the writing skills of a secretary. The most pressing issue in this regard is not his writing in larger letters but why he called attention to his own large letters, a statement that as far as we can tell has not been uncovered elsewhere in the thousands of papyrus letters 54 Reece lists at least thirty-three proponents of this view, including 20th–21st century authors, of all theological stripes (Reece, Paul’s Large Letters, 103–104 n. 47). This list could easily be expanded. 55 “Perhaps Paul simply draws attention to the script different from that of the scribe’s so that the Galatians will recognize his special effort here.” Craig S. Keener, Galatians: A Commentary (Baker Academic, 2019), 563.

Paul ’ s “ Large Letters ” in Galatians

169

that have been uncovered.56 Some have suggested that he called attention to the size because the epistle would be read out in congregations and the auditors would not see his larger hand. This suggestion has some merit, but it may assume a large congregation in a large building rather than the more probable small gatherings of communities. Furthermore, if that was the reason, why did he not do so in the other letters? And again, why would it be important that all see his larger letters? Chris Keith draws us closer to a reasonable answer by suggesting that this is a rhetorical move of irony on the part of the apostle.57 Keith argues that Paul wrote this to show that he was capable of writing, but that he had dictated the earlier part of the letter to a secretary, even though he was quite capable of writing it himself. In this way he demonstrates to his readers that he could choose not to write whenever he so decided. In other words, by deciding to write he was asserting the status of his education, and by deciding not to write he was asserting his social status. Keith’s suggestion is creative, but I believe there is a better approach that also discerns the role of irony on the part of the apostle. In my opinion, the reason it is better is because it pays closer attention to the context of the rest of the words that Paul wrote in Gal 6:12–18. The time-honored hermeneutical principle of context is the key to understanding, not Paul’s large letters, but why he called attention to them. Strangely the context of this statement has often been ignored, even though it lies in the passage before us. Let us peruse the English text that follows and then try to develop a satisfactory conclusion to this discussion. 12 Those who want to impress people by means of the flesh are trying to compel you to be circumcised. The only reason they do this is to avoid being persecuted for the cross of Christ. 13 Not even those who are circumcised keep the law, yet they want you to be circumcised that they may boast about your circumcision in the flesh. 14 May I never boast except in the cross of our Lord Jesus Christ, through which the world has been crucified to me, and I to the world. 15 Neither circumcision nor uncircumcision means anything; what counts is the new creation. 16 Peace and mercy to all who follow this rule – to the Israel of God. 17 From now on, 56 57

Plutarch states that Cato wrote in large letters for the sake of his young son who was just learning to read (Cato Maior 20.5–7). My thanks to Steve Reece for calling this reference to my attention (cf. Reece, Paul’s Large Letters, 89). Chris Keith, “In My Own Hand: Grapho-Literacy and the Apostle Paul,” Biblica 89 (2008): 39–58; see esp. 46, 54–56.

170

Varner

let no one cause me trouble, for I bear on my body the marks of Jesus. 18 The grace of our Lord Jesus Christ be with your spirit, brothers and sisters. Amen. (NIV; italics mine) A number of authors have called attention to the references to the body expressed and implied in this passage. These references of course add to that body member that Paul mentions in Gal 6:11: his hand. The references to impressing by a visual show (6:12) and boasting in bodily circumcision and in the cross should also be noted (6:13–14), as well as the bodily marks of Jesus and Paul (6:17). The important point is that all these things can be visibly “seen.” So how do these later references illuminate Paul’s earlier calling attention to his larger letters? The very next statement after Gal 6:11 is about those who “want to impress” others in a visible and fleshly way (6:12). The verb used (εὐπροσωπέω) appears only here in the New Testament. Appropriately for this discussion, it has only been found elsewhere in one papyrus letter dating from the first century BC.58 The letter is exhorting a relative to prepare their estate to show to others for a sale. The translation is as follows: “You will be right in not diminishing the report compared with the first one, in order that we may make a good show (ὅπως εὐπροσωπῶμεν), and for the rest please hurry on the collection of taxes.” It is obvious that the sender wants to impress others with their possessions (even if they are not as valuable as they are priced!). In the context of Gal 6:12–18, Paul condemns those Judaizers who want to impress others with the number of circumcised converts. By making a good outward show they will impress others. Paul refuses to do this and as he writes in his larger untrained hand, he sees this as an analogy with his own life. He calls attention to something he cannot boast about – skilled handwriting – and as far as he is concerned, there is nothing else in him to impress or “make a good show.” (Parallels closer to another autobiographical confession in Phil 3:5–12 come to mind.) So just as he cannot boast about his handwriting, he will not boast in anything but the cross of Christ (Gal 6:14). Paul is not out to impress others; his only boast is in Jesus, whose bodily wounds he shares, and not with shame but with justifiable pride.59

58 P. Teb. 1.19 (line 12). 59 The only commentator that I have found who recognizes Paul’s irony in these terms is R. A. Cole, Galatians: An Introduction and Commentary, 2nd ed., TCNT (Downers Grove, IL: InterVarsity, 1989), 233.

Paul ’ s “ Large Letters ” in Galatians

Figure 6.10

171

Scroll Showing Galatians This photo was published in black and white in Reece, Paul’s Large Letters, 215. This color image was personally provided to the author by Reece. Although we have reached different conclusions about the reason why Paul called attention to his large letters, I acknowledge with gratitude the influence that his book has had on my thinking and approach to this subject.

7 Conclusion Our survey of ancient epistolary subscriptions has led to some not so dramatic conclusions. Paul in using larger letters in his handwriting simply was following the epistolary practices of the period. There is no need to import back into his day anachronistic comparisons to modern writing and typing and printing practices. By appealing to “my own hand” Paul was certainly appealing for authentication. His inviting us to look at his large letters, however, is an ironic appeal to his own humility that brags not in his accomplishments (including handwriting), but in the redemptive work of his Savior, Jesus. As a final gift to my readers, I here add another image (Figure 6.10). This photograph is of a reconstruction of a papyrus roll that creatively portrays the end of Galatians in both a finer secretarial hand and concluding in the imagined cruder hand of the apostle. Acknowledgments My gratitude is expressed to Professor Larry Hurtado for reviewing this chapter and offering some valuable suggestions. It is dedicated to his memory.

CHAPTER 7

The Tradition and Development of the Subscriptions to 1 Timothy Linnea Thorp and Tommy Wasserman 1 Introduction The subscriptions to New Testament texts are not part of the original writings. Hence, our purpose is not to reconstruct the subscription to 1 Timothy as a part of the initial letter, nor to examine how it is connected to the original letter situation. In other words, we are not primarily concerned with the questions of historicity or authenticity.1 Instead, we will survey all the extant subscriptions (in over 300 manuscripts [MSS; sing. MS]), attempt to trace the origin, character and development of the various traditions reflected in these subscriptions and observe how they illuminate the history of interpretation of the letter and of the people and different places associated with it. As we hope to demonstrate, the subscriptions, though often neglected, are interesting paratexts and source material for a variety of ancient church traditions connected to the New Testament text, in our case specifically to the text of 1 Timothy. William H. P. Hatch provides a good definition and starting point for our survey: A subscription is appended to each book of the New Testament. In the oldest codices the subscriptions, like the titles, are short and simple. They give no information about the book in question, but merely indicate that it is ended. The superscription at the beginning and the subscription at the end are often identical.2 1 For an historical approach to the subscriptions, see G. Andrew Payne, “A Textual Analysis, Critical Reconstruction, and Evaluation of the Superscriptions and Subscriptions to the Corpus Paulinum” (PhD dissertation, Southwestern Baptist Theological Seminary, 2002). 2 William H. P. Hatch, Facsimiles and Descriptions of Minuscule Manuscripts of the New Testament (Cambridge, MA: Harvard University Press, 1951), 33. The term “subscription” here is not used (as it sometimes is) to connote the end of a letter, which an author writes by his own hand, as opposed to the rest of the letter, copied by an amanuensis. The term is so used by Gordon J. Bahr, “The Subscriptions in the Pauline Letters,” JBL 87 (1968): 27–41; and Steve Reece, Paul’s Large Letters: Paul’s Autographic Subscription in the Light of Ancient

© Tommy Wasserman and Linnea Thorp, 2023 | doi:10.1163/9789004537972_009

the Subscriptions to 1 Timothy

173

Bruce M. Metzger further highlights the development of the subscriptions: As time passed, these became more elaborate and often included traditional information regarding the place at which the book was thought to be written and sometimes the name of the amanuensis. The King James Version includes the subscriptions to the Pauline Epistles.3 As is apparent from Metzger’s widely used Textual Commentary, these long subscriptions exhibit a large textual variety and sometimes give conflicting information.4 2

Survey of Research

Scholars and editors of the Greek New Testament have largely neglected or paid less attention to the subscriptions, since they are regarded as secondary to the text proper.5 In 1790, William Paley devoted a chapter to the subscriptions in his Horae Paulinae as part of a larger effort to examine the authenticity and historicity of the Pauline Letters.6 Paley concluded that six of the subscriptions were “false or improbable,” i.e., they were either contradicted by the content of the respective letter (1 Corinthians, Galatians, 1 Timothy) or difficult to reconcile with it (1–2 Thessalonians, Titus).7 Of course, Paley evaluated the fully developed subscriptions in the Textus Receptus, based on late MSS, and argued against the indications of places of origin. We will discuss these and other geographical markers as part of the textual transmission and history of interpretation.

3 4 5 6 7

Epistolary Conventions, LNTS 561 (London: Bloomsbury T&T Clark, 2017). Further, stichometric notations are not considered as part of subscriptions proper and have been left out of this study. On stichometry, see Charles Graux, “Nouvelles recherches sur la stichométrie,” Revue de Philologie de Littératur et d’Histoire Anciennes 2 (1878): 97–143; Rendel J. Harris, Stichometry (London: C. J. Clay & Sons, 1893). Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption and Restoration, 4th ed. (Oxford: Oxford University Press, 2005), 41. Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 1994). This attitude is reflected in the introduction to the most recent Nestle-Aland edition (NA28): “For the sake of focusing on essentials, the subscriptions will no longer appear in the NA apparatus.” William Paley, Horae Paulinae: Or, The Truth of the Scripture History of St. Paul Evinced by a Comparison of the Epistles (London: J. Davis, 1790), 378–385. Paley, Horae Paulinae, 378.

174

Thorp and Wasserman

2.1 Subscriptions in Critical Editions Editors like Constantine von Tischendorf and Herrmann von Soden included a selection of superscriptions and subscriptions in their editions and von Soden offered a few comments and examples.8 In more recent critical editions, Reuben Swanson has collated some subscriptions (he uses the term “postscripts”) in the Gospels and four Pauline Letters.9 The subscriptions are excluded from UBS GNT editions (although a selection is listed in Metzger’s accompanying Textual Commentary) and from the most recent 28th Nestle-Aland edition, whereas the Novum Testamentum Graece: Editio Critica Maior of Acts and the Catholic Letters include a large selection of subscriptions for each of book.10 Tommy Wasserman included all extant subscriptions in his edition of Jude and commented briefly on them, noting that they were added to the initial text and expanded over time.11 2.2 A Pauline Letter Collection A number of scholars have studied various paratexts in relationship to the development of a Pauline letter collection.12 Among them, David Trobisch has 8

Constantin von Tischendorf, Novum Testamentum Graece, 3 vols., 8th ed. (Leipzig: Hinrichs, 1869–1894); Hermann von Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte, 2 parts in 4 vols., 2nd unchanged ed. (Göttingen: Vandenhoeck & Ruprecht, 1911–1913), 1:294–300 (subscriptions to Pauline Letters on p. 300). 9 Reuben Swanson, New Testament Greek Manuscripts: Variant Readings Arranged in Horizontal Lines against Codex Vaticanus (Sheffield: Sheffield Academic; Pasadena, CA: William Carey International University Press, 1995–2005) (Galatians, 1999; Romans, 2001; 1 Corinthians, 2003; 2 Corinthians, 2005). 10 Barbara Aland et al., eds., Novum Testamentum Graece: Editio Critica Maior IV: Catholic Letters, 2nd ed., 2 vols. (Stuttgart: Deutsche Bibelgesellschaft, 2013); Holger Strutwolf et al., eds., Novum Testamentum Graece: Editio Critica Maior III: The Acts of the Apostles, 3 parts in 4 vols. (Stuttgart: Deutsche Bibelgesellschaft, 2017). 11 Tommy Wasserman, The Epistle of Jude: Its Text and Transmission, CBNTS 43 (Stockholm: Almqvist & Wiksell International, 2006), 208–209 (apparatus), 339 (commentary). Wasserman notes a higher error rate of the transcriptions of subscriptions in the ECM of the Catholic Letters (Wasserman, Epistle of Jude, 223). 12 For an assessment of the various theories of how the Pauline collection originated and developed, see Stanley E. Porter, “When and How Was the Pauline Canon Compiled? An Assessment of Theories,” in The Pauline Canon, ed. Stanley E. Porter, PAST 1 (Leiden: Brill, 2004), 95–128. See also Stanley E. Porter, “Paul and the Pauline Letter Collection,” in Paul and the Second Century, ed. Michael F. Bird and Joseph R. Dodson, LNTS 412 (London: Bloomsbury T&T Clark, 2011), 19–36. For a recent study of Paul’s letter collection including attention to texts and paratexts in the manuscript tradition, see Eric W. Scherbenske, Canonizing Paul: Ancient Editorial Practice and the Corpus Paulinum (Oxford: Oxford University Press, 2013). Another recent study by C.-B. Amphoux, “Les lieux de rédaction

the Subscriptions to 1 Timothy

175

presented the most elaborate thesis that there was an early “canonical edition” of the fourteen letters of Paul and, in extension, a canonical edition of the New Testament by the second century.13 In support of his thesis, he points to several common redactional traits in early Christian MSS that point to an early archetype: “the notation of the nomina sacra, the codex form, the uniform arrangement and number of writings in the manuscript tradition, the formulation of the titles, and the evidence indicating that the collection was called ‘New Testament’ from the very beginning.”14 The uniformity of the titles, or superscriptions, throughout the Corpus Paulinum is relevant for our study, since we think the earliest subscriptions were identical to the superscriptions. Trobisch points out that the full titles of the works in the collection are longer than the short titles, e.g. πρòς Ῥωμαίους, which presupposes a first part of the title including the author and genre designation, Παύλου ἐπιστολή.15 Thus, Trobisch thinks the ancient title of the letter collection was “Letters of Paul.”16 Trobisch’s thesis of a single early “canonical edition” of the Pauline Letters, and of the New Testament, as the source of the manuscript tradition has been variously received, and we will not attempt to evaluate it here.17 On the other hand, the superscriptions and subscriptions reflect undeniable traces of redactional activity. For example, the numbering of the letters with the same address, e.g. πρὸς Τιμόθεον ᾱ, must be a result of their being brought together and published in a collection, regardless of the question whether this derives from a single archetype collection.18

13

14 15 16 17

18

des lettres de Paul d’après la tradition manuscrite,” Babelao 2 (2013): 87–104, refers to the places of origin in the subscriptions (in a selection of textual witnesses) in order to trace different redactional stages. David Trobisch, Die Entstehung der Paulusbriefsammlung: Studien zu den Anfängen christlicher Publizistik, NTOA 10 (Göttingen: Vandenhoeck & Ruprecht, 1989); Trobisch, The First Edition of the New Testament (New York: Oxford University Press, 2000 [2nd ed., 2011]); Trobisch, Paul’s Letter Collection: Tracing the Origins (Minneapolis: Fortress, 1994; repr., Bolivar, MI: Quiet Waters Publications, 2001). Trobisch, First Edition, 44. Trobisch, First Edition, 38–41. Trobisch, Paul’s Letter Collection, 24. For critical evaluations of Trobisch’s thesis, see Jason T. Larson’s review in TC: A Journal of Biblical Textual Criticism 6 (2001) (http://rosetta.reltech.org/TC/v06/Trobisch2001rev -x.html); David Parker, “Review of David Trobisch, The First Edition of the New Testament,” JTS 53.1 (2002): 298–305. For further objections, see Michael W. Holmes, “Text and Transmission in the Second Century,” in The Reliability of the New Testament, ed. Robert W. Stewart (Minneapolis: Fortress, 2011), 62–65. Cf. Trobisch, First Edition, 40.

176

Thorp and Wasserman

2.3 The Euthalian Apparatus Some scholars have discussed the relationship of superscriptions and subscriptions to the Euthalian apparatus, a collection of ancillary paratextual material, such as divisions of texts, lists of chapter headings and summaries of books (Acts and the Letters), and, significantly, a table that lists places from which the letters were sent. In the case of 1 Timothy, the locations’ table indicates “Laodicea of Phrygia” as the place of origin.19 The apparatus, circulating under the name of “Euthalius,” is generally dated to the fourth century when it was likely added to an ancient edition of the New Testament.20 As Nils A. Dahl explains, the Euthalian apparatus “is attested in a very impressive number of Middle Age minuscule manuscripts from the ninth century onwards” although “the Syrian and Armenian translations of parts of the apparatus indicate that the apparatus for Paul’s letters was known much earlier.”21 In this connection, it should be noted that the sixth-century Codex Coislinianus (H 015) also includes Euthalian material.22 In his extensive study of the Euthalian apparatus, Louis Charles Willard thinks Caesarea is the most likely place of its origin and proposes a date between 380–396 CE.23 Metzger has suggested that the longer subscriptions to the Pauline Letters probably are the work of Euthalius.24 Willard, however, has argued that there is no direct link between the apparatus and the subscriptions, i.e., there is no evidence that the subscriptions derive from Euthalius’s original edition.25 At the same time, he highlights the fact that the locations’ table “agrees with the subscriptions in the points of origin with minor variation in the elaboration of a place name.”26 The question of dependence is further related to another paratext in the manuscript, the so-called hypotheses, or argumenta.

19 Louis Charles Willard, A Critical Study of the Euthalian Apparatus, ANTF 41 (Berlin: De Gruyter, 2009), 80. 20 Bruce M. Metzger, Manuscripts of the Greek Bible: An Introduction to Greek Palaeography (Oxford: Oxford University Press, 1981), 42–43. 21 Nils A. Dahl, Studies in Ephesians: Interpretation of Texts and Themes, ed. David Hellholm, Vemund Blomkvist, and Tord Fornberg, WUNT 131 (Tübingen: Mohr Siebeck), 231. 22 Dahl, Studies, 244 n. 47, refers briefly to H (015) in a discussion of the earliest format of the Euthalian Apparatus and whether it was written colometrically, as in H (015). For an extensive discussion of Coislinianus, see Scherbenske, Canonizing Paul, 116–174. 23 Willard, Euthalian Apparatus, 131. 24 Metzger, Manuscripts, 40. 25 Willard, Euthalian Apparatus, 79. 26 Willard, Euthalian Apparatus, 80.

the Subscriptions to 1 Timothy

177

2.4 Hypotheses/Argumenta In the Greek MSS, the books are often prefaced by a brief summary, ὑπόθεσις (Greek) or argumentum (Latin). As Dahl points out, most argumenta start “by naming the place from which Paul writes.”27 The argumenta were once regarded as part of the Euthalian tradition, but Willard explains that from the time of Lorenzo Alessandro Zacagni, who printed the first edition of the Euthalian apparatus in 1698, a connection between the argumenta and “the primary body of the Euthalian apparatus” has been denied.28 Thus, Willard presents a number of important arguments for the independence of the argumenta; most importantly, his survey of Greek MSS shows that many include argumenta but lack any part of the Euthalian apparatus and, conversely, a few with Euthalian material omit the argumenta (015 81 330 1243 1720).29 Dahl suggests that the argumenta was fused together with chapter-lists and the Euthalian apparatus in an edition at some point, but minuscule 81 and the Euthalian edition in Syriac translation, both lacking the argumenta, clearly presuppose an older edition.30 The authorship of the argumenta in the Pauline Letters is “complicated,” in the words of Vemund Blomkvist, since “[t]hey appear in Euthalius, in the commentary of Oecumenius (PG 118–119) and in the pseudo-Athanasian Synopsis Scripturae Sacrae (PG 28:282–438).”31 Blomkvist and David Hellholm think that the Synopsis Scripturae Sacrae is the original source of the argumenta for linguistic reasons.32 In a recent study of the superscriptions and subscriptions to the Pauline Letters, G. Andrew Payne discusses the literary relationship between the Euthalian apparatus and the subscriptions in Greek MSS and concludes that it is indeed complicated: in case the subscriptions were derived from the 27 Dahl, Studies, 255. 28 Willard, Euthalian Apparatus, 70. See Lorenzo Alessandro Zacagni, Collectanea monumentorum veterum ecclesiae Graece, vol. 1 (Rome: Typis Sacrae Gongreg, de propag, fide, 1698). 29 Willard, Euthalian Apparatus, 70–72 (manuscript survey on 158–169); cf. Dahl, Studies, 253. 30 Dahl, Studies, 254, 257. For the Euthalian tradition in Syriac, see Ernst von Dobschütz, “Euthaliusstudien,” Zeitschrift für Kirchengeschichte 19 (1899): 107–134; and Sebastian Brock, “The Syriac Euthalian Material and the Philoxenian Version of the NT,” ZNW 70 (1979): 120–130. 31 Vemund Blomkvist, Euthalian Traditions: Text, Translation and Commentary, TUGAL 170 (Berlin: Walter de Gruyter, 2012), 147. 32 David Hellholm and Vemund Blomkvist, “‘Paraenesis’ as an Ancient Genre-Designation: The Case of the ‘Euthalian Apparatus’ and the ‘Affiliated Argumenta,’” in Early Christian Paraenesis in Context, ed. James Starr and Troels Engberg-Pedersen (Berlin: Walter de Gruyter, 2004), 510 n. 221.

178

Thorp and Wasserman

Euthalian traditions, “[i]t would call for someone to draw from three different locations – the argumenta, the locations’ table, and the co-writers’ table – to successfully reconstruct each subscription. Even so, one could successfully reconstruct only one subscription: the one to Galatians.”33 On the other hand, Payne does not think that the literary dependence goes in the other direction, from subscription to Euthalian apparatus, but instead suggests that the subscriptions and the hypotheses/argumenta draw from a common tradition.34 Similarly, Willard refers to the discrepancy between the information in some argumenta and the corresponding subscriptions. For example, he explains, “manuscripts frequently show Macedonia as the place of origin of 1 Timothy in the argumentum; the subscription and the appended note [the locations’ table], together with a regular variant in the text of the argumentum, show Laodicea.”35 2.5 The Latin Prologues The information about letter carriers and places of origin can also be found in the so-called Latin prologues to the Pauline Letters found in various Latin MSS including Codex Fuldensis (6th cent.). In 1907, Donatien de Bruyne proposed that these prologues were Marcionite in origin (hence “Marcionite Prologues”), an issue that has been debated since then.36 The origin of these prologues is of marginal interest to us. As Dahl points out, the information in the prologues about letter carriers and places from which Paul sent his letters are “secondary accretions,” and “likely to have been inspired by, in some cases taken over from, similar pieces of information in subscriptions and argumenta (ὑποθέσεις) to the Pauline letters in Greek manuscripts” – additions which likely took place in the later part of the of the fourth or early fifth century “when the Greek influence upon the Latin Bible was strong.”37 In this connection, he notes that there is a considerable textual variety concerning “the places from which Paul wrote Timothy and Titus.”38 Whereas a few important manuscript witnesses lack the addition altogether (G, Z*, P), there are textual variants that ascribe the letter 33 34 35 36

Payne, “Textual Analysis,” 138. Payne, “Textual Analysis,” 138–139. Willard, Euthalian Apparatus, 71. Donatien de Bruyne, “Prologues bibliques d’origin marcionite,” RB 24 (1907): 1–24; Peter Corssen, “Zur Überlieferungsgeschichte des Römerbriefes,” ZNW 10 (1909): 1–45, 97–102. For recent discussions, see Nils A. Dahl, “The Origin of the Earliest Prologues to the Pauline Letters,” Semeia 12 (1978): 233–277 (republished in Dahl, Studies, 179–210); Scherbenske, Canonizing Paul, 237–242 (containing an appendix evaluating Dahl’s argument against a Marcionite origin of the argumenta). 37 Dahl, Studies, 186. 38 Dahl, Studies, 187.

the Subscriptions to 1 Timothy

179

to Macedonia, Laodicea, Nicopolis, or Rome (the latter is unattested in Greek manuscripts).39 As we will see, the three first locations occur in the subscriptions of the Greek manuscript tradition, whereas Rome is unattested. 2.6 Special Studies of the Subscriptions In the two recent decades, two unpublished dissertations have appeared which are devoted specifically to the superscriptions and subscriptions to the Pauline Letters. In 2002, Payne presented an analysis of superscriptions and subscriptions in the Corpus Paulinum including a collation of superscriptions and subscriptions in fifty-six MSS including all papyri and majuscules and a small selection of minuscule MSS, evaluating the data from literary-critical, text-critical, and historical-critical perspectives.40 He presented the evidence in a critical apparatus, generally preferring the formats πρὸς Ῥωμαίους or πρὸς Τιμόθεον ᾱ in both superscription and subscriptions.41 Payne further examined the literary relationship between these paratexts and the Marcionite (or Latin) Prologues and the Euthalian apparatus, noting similarities and differences, and concluded that there is no direct relationship.42 Finally, he evaluated the historical information conveyed in the subscriptions finding, on the one hand, “some apparent erroneous information,” but, on the other hand, some useful data “to both confirm and possibly fill in the gaps in our historical knowledge of Paul.”43 In 2012, David G. Champagne presented his thesis on the scribal habits reflected in the superscription and subscription traditions.44 Champagne cast the net wider in his analysis of these paratexts by including evidence from over one hundred MSS in each part of the New Testament (Gospels, Acts, Pauline 39 Dahl, Studies, 186–187 (esp. n. 46). 40 Payne, “Textual Analysis.” 41 Payne, “Textual Analysis,” 81–119. In 2 Thessalonians, Payne preferred the unique subscription of Codex Vaticanus, πρὸς Θεσσαλονίκεις β̅ ἐγράφη ἀπὸ Ἀθηνῶν (Payne, “Textual Analysis,” 104), without realizing that the second half, ἐγράφη ἀπὸ Ἀθηνῶν, was added by a later hand, probably in the sixth century. Constantin von Tischendorf pointed out that the uncial script used for these subscriptions was different (than the first hand, the corrector, and a third hand) and ascribed it to the sixth century (Novum Testamentum Vaticanum [Leipzig: Giesecke und Devrient, 1867], xxviiii). This one deviation from the standard format is particularly odd, since Payne assumed that the original subscriptions date back either to the time of Paul himself, or to the time of the compilation of the Corpus Paulinum (Payne, “Textual Analysis,” 2, 163–167, 170–171). 42 Payne, “Textual Analysis,” 138–139. 43 Payne, “Textual Analysis,” 161. 44 David G. Champagne, “Scribal Habits Within the Superscription and Subscription Traditions of Greek New Testament Manuscripts” (PhD dissertation, New Orleans Baptist Theological Seminary, 2012).

180

Thorp and Wasserman

Letters, Catholic Letters, and Revelation).45 In contrast to Payne, Champagne did not attempt to investigate the origins of the superscriptions and subscriptions, nor the historical reliability of the information they conveyed, but focused on the content and chronological development of these paratexts and the question of what they reveal about the scribal habits and the genealogical relationships between MSS. Champagne’s study confirmed the earlier observation that the superscriptions and subscriptions expanded over time. Further, it demonstrated that the average frequency of variation is higher in these paratexts as compared to the New Testament text, i.e., there is a “greater editorial freedom within the superscription and subscription traditions.”46 In this connection, Champagne found significant differences between the distinct parts of the New Testament. In general, the Gospels were the “most stable in both the superscription and subscription,” whereas “Revelation was the least stable in the superscription traditions, and the Paulines [sic] were least stable in the subscription traditions.”47 Undoubtedly, this instability in the Pauline letter subscriptions is due to the addition of geographical locations and letter carriers. In spite of editorial freedom, Champagne observed that the scribes worked within some parameters including a limited set of elements. He identified the following common elements in the subscription traditions: Each of the collections includes references to authorship, genre/content [e.g. ἐπιστολή], the reverential modifier “holy,” authorship modifiers [e.g. τοῦ ἁγίου ἀποστόλου], divine-agency modifiers [e.g. σὺν θω� ], connections with doxologies or colophons, terminal modifiers [e.g. τέλος], and rare elements. The reference to authorship in each of the collections except the Pauline follows the same pattern found in the superscription traditions.48 On the other hand, Champagne pointed to a specific difference between subscriptions to the Pauline Letters and to Catholic Letters in that “[r]eferences to recipients and letter carriers are core elements in the Pauline tradition and rare elements in the Catholic tradition [sic].”49 Champagne’s study of superscriptions and subscriptions in the New Testament textual tradition remains the most comprehensive to this date, and we will refer to it as we 45 For Champagne’s method of selection, see his “Scribal Habits,” 28–29. 46 Champagne, “Scribal Habits,” 434. The rich textual variation in superscriptions and subscriptions is apparent by looking at editions that include them. For example, Wasserman’s apparatus of Jude displays seventy-five variants of the superscription and sixty-seven variants of the subscription. 47 Champagne, “Scribal Habits,” 433. 48 Champagne, “Scribal Habits,” 400–401. 49 Champagne, “Scribal Habits,” 404.

the Subscriptions to 1 Timothy

181

compare the development of subscriptions to 1 Timothy to those of the rest of the New Testament in general, and to the subscriptions to the Pauline Letters in particular.50 We will also use some of Champagne’s terms to describe the elements in the subscriptions. 3

The Subscriptions to 1 Timothy: a Background

3.1 First Timothy: Traditional Letter Situation The traditions concerning where the Pastorals were written are connected to various traditions about Paul’s travels and activities. The issues of authenticity have been much debated in exegetical scholarship, since it is difficult to situate these letters in Paul’s life, not to mention other problems relating to their language and theology. Furthermore, the Pastoral Letters are missing from P46 and Marcion’s canon, and Tatian apparently rejected 1 Timothy according to Jerome (Comm. Tit. praef.).51 For these reasons most critical scholars have viewed the Pastoral Letters as pseudepigraphical.52 The question of authenticity is not our primary concern but rather the tracing of ancient traditions connected to Paul’s life and travels and the attempts to place the Pastoral Letters into this context, as reflected in the various subscriptions. First, Clement of Rome says that Paul had “reached the farthest limits of the west” (1 Clem. 5.7), which suggests that he went beyond Italy to Spain.53 Further, Eusebius of Caesarea refers to a passage in Origen’s Commentary on Genesis (3:1): There is evidence that, having then been brought to trial, the apostle again set out on the ministry of preaching, and having appeared a second time in the same city found fulfilment in his martyrdom. In the course of

50

It is to be noted that Champagne sometimes omits reference to the crucial evidence in Codex Alexandrinus in the subscriptions to the Pauline Letters. 51 Edmon L. Gallagher and John D. Meade, The Biblical Canon Lists from Early Christianity: Texts and Analysis (Oxford: Oxford University Press, 2017), 43. 52 For a new perspective (in particular on 2 Timothy and Titus), see Jens Herzer, “Zwischen Mythos und Wahrheit: Neue Perspektiven auf die sogenannten Pastoralbriefe,” NTS 63 (2017): 428–450. 53 Translation in Michael W. Holmes, The Apostolic Fathers: Greek Texts and English Translations, 3rd ed. (Grand Rapids: Baker, 2007), 53. In the note to the translation, Holmes takes “the farthest limits of the west” to refer to the Straits of Gibraltar (Holmes, Apostolic Fathers, 53).

182

Thorp and Wasserman

this imprisonment he composed the second Epistle to Timothy, referring both to his earlier trial and to his impending fulfilment (Hist. eccl. 2.22).54 The origin of this tradition is apparently 2 Tim 4:16–17, where Paul first comments, “at my first defence no one came to my support,” and then that he was eventually “rescued from the lion’s mouth,” which Eusebius evidently takes as a reference to Paul’s first trial in Rome from which he was released. As Philip Towner explains, the second-imprisonment theory “creates the space needed for the travels and ministry indicated by 1 Timothy and Titus, but unmentioned in Acts, to take place.”55 Apparently, Eusebius was followed by Euthalius in his prologue to the Pauline Letters. Neither of them, however, reports where Paul was during the time between the first and second imprisonment.56 This left room for the growth of tradition concerning 1 Timothy. The Growth of Tradition: Curiosity, Authorization, and Legitimization In a study of the growth of Christian tradition concerned with names, Metzger points out that there are many people and places mentioned in the New Testament unaccompanied by their names, and because of curiosity about these characters and places early Christians sought to identify them.57 “Tradition,” he says, “provided names for all of these – sometimes several different names.”58 For example, Eusebius (4th cent.) took an interest in the names of the seventy disciples, which Jesus sent out (Luke 10:1), but stated that there was no written record of them. Nevertheless, he proceeded to identify and write down five of them on the basis of oral tradition. In subsequent centuries this tradition grew, so that “more than one list was circulated that included these and sixty-five other names!”59 Metzger of course regarded most of these traditions as historically unreliable, and rather “a testimony to the fertility of pious imagination down 3.2

54 English translation by G. A. Williamson, Eusebius: The History of the Church from Christ to Constantine, rev. and ed. Andrew Louth (London: Penguin, 1989), 57. Cf. Origen’s Commentary on Genesis 3.1. 55 Philip H. Towner, The Letters to Timothy and Titus, NICNT (Grand Rapids: Eerdmans, 2006), 11. For Towner’s attempts to connect 1 Timothy to the story line of Acts, see 13– 15, 107. 56 Willard, Euthalian Apparatus, 18. 57 Bruce M. Metzger, “Names for the Nameless in the New Testament: A Study in the Growth of Christian Tradition,” in Kyriakon: Festschrift Johannes Quasten, ed. Patrick Granfield and Josef Andreas Jungmann, vol. 1 (Münster: Aschendorff, 1970), 79–99. 58 Metzger, “Names,” 79. 59 Metzger, “Names,” 79.

the Subscriptions to 1 Timothy

183

through the centuries and the reluctance to respect the silence of the New Testament narratives.”60 Though not discussed by Metzger, the subscriptions to the New Testament letters are analogous in several respects to the growing traditions that name characters and places in the narratives – human curiosity is one of the main factors behind the origin and growth of traditions in the subscriptions. In relation to the Euthalian edition of the Corpus Paulinum, Eric W. Scherbenske points to another important factor in that the extracanonical traditions shaped the way Paul was read: The sources deemed authentic were used to reconstruct Paul’s life, which in turn authenticated them as sources for readers of this edition. In a related manner, these extracanonical traditions supplied a proto-orthodox metanarrative legitimating their own claims of apostolicity and orthodoxy. The inclusion of such traditions in paratexts (prologues, subscriptions, etc.) even ensured their transmission as part of the very scripture they sought to authenticate; for example, the tradition that Titus was the first bishop of the Cretian church, which happens to be found in the subscription to this letter in Codex Coislinianus, was transmitted through the Majority text to the King James Version.61 In a recent article, Tobias Nicklas demonstrates how canonical texts played a role in constructing what he terms as ancient Christian “landscapes of memory,” which could initiate “a starting point of later speculation, interpretation, and the creation of new stories connected to various places.”62 Nicklas points to the Acts of Titus as such a new story building on the book of Acts (and the Acts of Paul), and one of the goals of this text is apparently “to legitimize the Cretan Church: the apostles ordain Titus himself (Acts of Titus 4), thus placing him into an apostolic succession” (cf. Titus 1:5).63 Nicklas does not discuss the possibility of competing traditions – different “maps of the memory landscape,” so to speak. It is an open question whether the origin and further transmission of

60 Metzger, “Names,” 98. 61 Scherbenske, Canonizing Paul, 127. 62 Tobias Nicklas, “New Testament Canon and Ancient Christian ‘Landscapes of Memory,’” EC 7 (2016): 23. 63 Nicklas, “New Testament,” 20. As we will see, the tradition about Titus being the first bishop of Crete is attested in the subscriptions, first in Codex Coislinianus (H 015), which reads παυλου αποστολου επιστολη προς τιτον της κρητων εκκλησιας πρωτον επισκοπον χειροτονηθεντα εγραφη απο νικοπολεως της μακεδονιας.

184

Thorp and Wasserman

some of the competing variants in the subscriptions to 1 Timothy concerning its place of origin could be related to the influence of local traditions. 4

Collation and Apparatus of Subscriptions to 1 Timothy

The present survey of subscriptions is based on Linnea Thorp’s (née Arvedal) collation of a large number of extant Greek MSS that contain the ending of 1 Timothy as far as digital images are available.64 Dating follows Kurt Aland’s Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments. A total of 310 out of 415 examined MSS with the ending of 1 Timothy contain a subscription to the letter, whereas 103 omit the subscription.65 In addition to these, another 29 MSS were lacunose, and two were illegible.66 We have added a limited range of versional evidence to the apparatus including significant bilingual (Greek-Latin) MSS. There is a very limited number of early MSS that preserve a subscription to 1 Timothy: Codex Sinaiticus (ℵ 01) from the fourth century; Codex Alexandrinus (A 02) from the fifth century; and the Greek-Latin diglot Codex Claromontanus (D 06) from the sixth century. In addition, 048, a palimpsest from the fifth century, contains the ending of 1 Timothy but is unfortunately illegible at this 64

65 66

Linnea Arvedal, “The Subscriptions to 1 Timothy: An Investigation of Their Traditions and Development” (Bachelor’s Thesis, Örebro School of Theology, 2016). The MSS are referred to with their Gregory-Aland number. See Kurt Aland et al., eds., Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments, 2nd ed., ANTF 1 (Berlin: Walter de Gruyter, 1994). We have used digital images in the following repositories: New Testament Virtual Manuscript Room (NT.VMR maintained by the Institut für Neutestamentliche Textforschung, Münster; http://ntvmr.uni-muenster.de/); Center for the Study of New Testament Manuscripts (CSNTM; http://csntm.org); British Library (“Digitised Manuscripts”; http://www.bl.uk/manuscripts/). We would like to thank Conrad Thorup Elmelund for consulting Gregory-Aland 2009 at the Royal Library of Copenhagen for us, for correcting some errors in our apparatus, and for bringing to our attention the following manuscripts with subscription to 1 Timothy which have subsequently been made available in these repositories but were not included in the study: 172 309 312 356 378 384 421 440 479 483 489 616 639 823 914 997 1022 1069 1099 1106 1115 1149 1161 1398 1424 1425 1598 1636 1721 1722 1753 1759 1766 1839 1843 1847 1850 1851 1868 1870 1886 1891 1907 1910 1951 1962 1970 1971 1972 1981 1986 2191 2194 2257 2289 2310 2374 2482 2587 2629 2690 2691 2705 2817 2909. From Payne’s consultation of critical editions, he suggests that 195 MSS of the total number of MSS of the Corpus Paulinum contain superscriptions and/or subscriptions (“Textual Analysis,” 10). The following manuscripts are lacunose in all or part of 1 Timothy (including the ending): P133 04 016 061 0151 0241 0259 0262 0272 0285 339 400 480 498 602 720 1506 1729 1738 1864 1918 1942 1965 2423 2425 2596 2716 2736 2892. MS 048 and 1727 were illegible.

the Subscriptions to 1 Timothy

185

location.67 There are no more extant MSS with this part of the letter until the ninth century.68 We have divided the subscriptions in the apparatus below into different types according to content and wording. These types and subtypes roughly follow Hatch’s description of the development of subscriptions, according to how elaborate they are. Each manuscript is represented under one type (unless it has been corrected to another type). Hatch describes the earliest subscriptions in the Pauline Letters as very short and simple, e.g., Πρὸς Ῥωμαίους (identical to the superscription).69 These simple subscriptions are called “Type 1” in the collation. Over time, the simple subscriptions became more elaborate, one type of which “also only marked the ending of the book,” e.g., τέλος τῆς πρὸς Τιμόθεον ᾱ ἐπιστολῆς.70 This semi-elaborated type is labelled “Type 2” in the collation, and commonly contain a genre reference, i.e. ἐπιστολή, and/or the terminal modifiers τέλος or ἐπληρώθη (ἐτελέσθη/explicit).71 The type with a verb form is standard in Latin books in general.72 Thus, it appears in the fourth-century bilingual Codex Bezae in 3 John and in the Gospels, e.g., εὐαγγέλιον κατὰ Μαθθαίον ἐτελέσθη, a Greek translation of the more familiar Latin Euangelium

67

68 69 70 71

72

It is difficult to read the underlying text from the images of 048. Dale Eldon Heath, “The Text of Manuscript Gregory 048 (Vatican Greek 2061)” (PhD Dissertation, Michigan State University, 1965), 15, comments that “the writing which follows [the end of the letter] is so nearly obliterated as to render assertions impossible; but apparently the next division begins in line three after a symbol [image on p. 153]…” Aland, Κurzgefasste, passim. Also NT.VMR, “Liste” (http://ntvmr.uni-muenster.de/liste). Hatch, Facsimiles, 33. Hatch, Fascimiles, 33–34. Vito Lorusso says concerning Byzantine colophons, “Some frequently occurring expressions in Greek colophons and subscriptions to indicate that the manuscript was completed are, for example, ἐγράφη (egráphē, ‘it was written’), ἐπληρώθη (eplērṓthē ‘it was completed’), ἐτελειώθη (eteleiṓthē ‘it was brought to a close’), τετελείωται (teteleíōtai ‘it has been brought to a close’), ἐτελέσθη (etelésthē ‘it was accomplished’), εἴληφε τέρμα (éilēphe térma ‘has reached the end’) and τέλος εἴληφε (télos éilēphe ‘has come to the end’). In all these cases, the subject of the verb is obviously the manuscript itself.” See Vito Lorusso, “Locating Greek Manuscripts through Paratexts: Examples from the Library of Cardinal Bessarion and other Manuscript Collections,” in Tracing Manuscripts in Time and Space through Paratexts, ed. Giovanni Ciotti and Hang Lin, Studies in Manuscript Culture (Berlin: De Gruyter, 2016), 246. See David C. Parker, Codex Bezae: An Early Christian Manuscript and Its Text (Cambridge: Cambridge University Press, 1992), 11. We cannot say at this point whether the terminal modifier τέλος originates from Latin usage.

186

Thorp and Wasserman

sec(undum) Mattheum explicit (which can be translated, “here ends the Gospel of Matthew”).73 According to Champagne’s study, the terminal modifier τέλος can be found in subscriptions to Mark and Luke from the ninth century onwards and in most other books from the tenth century.74 However, we have found it in a subscription to 1 Timothy in the ninth-century codex 1900. Champagne believes that “[t]he actual introduction of τελος into the tradition may have stemmed from the use of lectionary markings, which were standard in many MSS.”75 This is uncertain, but possibly the use of τέλος to mark the end of lections influenced its more frequent use in the later manuscript tradition. The subscriptions of “Type 3” in our typology are those that convey to “the reader in succinct form certain information concerning the book in question” – in the case of Paul’s Letters, the place of origin and sometimes the letter carrier.76 In 1 Timothy, these subscriptions contain the place of origin and various other additions. The earliest example of this type is the form attested in fifth-century Codex Alexandrinus (A 02), προς Τιμοθεον ᾱ εγραφη απο Λαοδικειας. It is to be noted that the subscription in this form may well predate most or all Type 2-subscriptions. Some erroneous readings are recorded in the apparatus. However, misspellings have not been recorded if they are of the most frequent kind of vowel interchange (αι-ε, ε-η[ει], ε-ι-η-υ-ι-οι, ο-ω). Neither is the vowel interchange of ει-ι in Λαοδικείας recorded. Variants for πρώτη and πρώτης are only recorded if they are different from what can be expected. Both πρώτη and πρώτης are represented by ᾱ. Prima is similarly represented by I̅ in the Latin readings. Key to Signs in the Apparatus substitution ⸀ ⸂ ⸃ substituted words (between the signs) ⸆ addition o omission ⸉ ⸊ alternative word order (between the signs) txt variants which agree with the text. * identifies first-hand reading when a correction has been made. 73

The subscription in Bezae is combined with the superscription to the next book in a colophon (which is also common in Latin MSS). 74 Champagne, “Scribal Habits,” 32–388. 75 Champagne, “Scribal Habits,” 114–115. We have observed that some lections terminate with τέλος τῆς έ (for the last reading in 1 Timothy) which seem to combine the function of the τέλος. 76 Hatch, Fascimiles, 34.

187

the Subscriptions to 1 Timothy

c identifies a correction made by a later hand. vid placed after a siglum to mark that there are uncertain letters or lacunae in the MS which means that the reading is uncertain. (⸀ ⸁ ⸀1 or o, o1 etc., for signs that occur successively in a single unit.)77 Table 7.1

Subscription Collation and Apparatus

Type

Reading

0

Omit 056 1 5 38 57 61 69 110 149 216 226 254 263 319 322 323 327 336 337 365 383 454 455 456 463 491 582 606 608 612 620 627 628 629 630 676 680 796 886 891 941 996 1245 1315 1319 1354 1359 1390 1490 1505 1609 1611 1622 1646 1661 1718 1731 1798 1827 1852 1867 1889 1890 1925 1929 1943 1945 1948 1950 1961 1969 1973 1984 1985 1987 1991 1994 1995 1998 2000 2002 2005 2086 2102 2105 2127 2143 2200 2289 2344 2400 2492 2516 2523 2544 2576 2659 2674 2718 2739 2774 2816 2889 2899 l2024 προς τιμοθεον oᾱ o1836 | txt 01 044 33 460 1739* προς τιμοθεον επιστολη ᾱ 1854 τελος ⸆ ⸆ε και της επιστολης 1367 | txt 90. τελος της προς τιμοθεον ᾱ επιστολης⸆ ⸆παυλου του αποστολου 1501 | txt 94 1509 1930 1947 1978 1992 2011 2197 2248 2495 2936 επληρωθη επιστολη προς τιμοθεον ᾱ 010 012 f g i explicit epistola ad timotheum I̅ προς τιμοθεον ᾱ επληρωθη ⸆scribens a ladicia c dc g1c ad timotheum I̅ ⸆ explicit Vggc | txt 06 0319 d* g1* Vgg* ii ⸀εγραφη απο λαοδικειας γραφη 2243 | 6vid iii 363 876 935 1127 1251 1832 2138 2494

1.1 1.2 2.1 2.2

2.3.1 2.3.2

3.1

Apparatus

i Codex Augiensis (F 010) = Old Latin f; Codex Boernerianus G 012 = Old Latin g. ii Codex Claromontanus (D 06) = Old Latin d. Codex Sangermanensis (Dabs1 0319) = Old Latin g1. iii 6vid could attest to any place.

77

Signs are based on the critical signs in the Nestle-Aland apparatus; see NA28, 11*–13*.

188 Table 7.1

Thorp and Wasserman Subscription Collation and Apparatus (cont.)

Type

Reading

Apparatus

3.1.1

⸆προς τιμοθεον ⸇ oᾱ ⸆1 εγραφη απο λαοδικειας

3.1.2

εγραφη απο λαοδικειας ητις εστι μρ̄οπολις φρυγιας ⸀η προς τιμοθεον ᾱ o ⸁επιστολη εγραφη απο λαοδικειας ⸂ητις εστι⸃ μρ̄οπολις φρυγιας

⸆η 999 2147 2652 ¦ τελος της 2085 ¦ τελος της ε και της 636 ¦ επιστολη ενδεκατη αυτη η 1728 | ⸇επιστολη 2147 | o 999| ⸆1 επιστολη 496 636 1728 1897 2085 ¦ επιστολης 496 | txt 02 105 241 1841 2865 1873 1404

3.1.3

3.1.4

3.1.5

⸀τελος της 142 | o 547 | ⸁επιστολης 142 | ⸂της 228vid | txt 1642c 1831 1939 1963 1996 1999 2012 1524

τελος της επιτομης της εις προς τιμοθεον πρωτης επιστολης εγραφη δε η επιστολη απο λαοδικειας ητις εστιν μητροπολις φρυγιας ⸆ oεγραφη ⸀απο λαοδικειας ⸂ητις εστι ⸁μητροπολις⸃ ⸇ o1 φρυγιας της ⸀1 καπατιανης ⸆ η τοιαυτη επιστολη 1241 ¦ αυτη η επιστολη 808 2501 ¦ τελος της επιστολης 1977 | o 1858 | ⸀πο 1858 | ⸂ μρ̄οπολεως 35 218 444 664 801 1040 1075 1101 1248 1352 1503 1617 1628 1637 1652 1740 1745 1746 1771 1865 1892 1899 2218 2431 2723 ¦ μητροπολεως 18 201 204 386 432 522 757 824 928 959 986 1072 1100 1105 1250 1482 1573 1704 1725 1732 1855 1856 1876 1960 2009 2080 2221 2261 2352 2501 2554 2626 | ⸁ μρ̄οπολις 76 465 808 1003 1626 1977 2201 | ⸇της 1003vid 1977 | o12183 1892 | ⸀1 καταπατιανης 1250 1876 ¦ πακτιανης 1242 1750 ¦ παγκρατιανης 1977vid ¦ πακατιανης 0142 296 1003 1626 1869 1919 2004 2183 | txt 436 607 808 918 2125

189

the Subscriptions to 1 Timothy Table 7.1

Subscription Collation and Apparatus (cont.)

Type

Reading

3.1.6.a

⸆ ⸀προς τιμοθεον ⸇ o ⸁ ᾱ ⸆1 ⸀1 εγραφη ⸆2 απο ⸀2 λαοδικειας⸃ ⸄ ητις εστι ⸀3 μητροπολις⸅ ⸆3 ⸀4 φρυγιας της ⸀5 πακατιανης ⸆ η 205 209 592 614 641 927 1384 1409 1769 1903 1916 1933 2412 2886 ¦ αυτη η 2558 ¦ τελος 457 ¦ τελος της 42 51 103 223 234 367 390 425 644 909 1456 1594 1595 1678 1872 1900 ¦ τελος συν θ̅ω της 91 1840 2008vid ¦ επιστολη 075 | ⸀ρος 665 | ⸇ επιστολη 1094 | o 62 175 1243 1311 1720 1830 1857 1881 | ⸁πρωτης 1724 1772 | ⸆1 επιστολη 141 205 209 326 592 605 1409 1769 1837 1860 1877 1933 2558 2886 ¦ επιστολης 42 51 91 103 223 234 367 390 425 644 1456 1594 1595 1678 1735 1840 1900 ¦ επιστολης του αγιου αποστολου παυλου ητις 1872 | ⸀1 εγραφει 42 189 223 425 614 1297 1735 1874 | ⸆2 δε | ⸀2 λαοδικει 2003 ¦ λαοδικαιας 614 | ⸄ μητροπολεως 1247 1768 ¦ της μρ̄οπολεως 1830 | ⸀3 μρ̄οπολις 51 122 141 189 205 209 223 234 367 390 425 468 592 614 632 641 644 794 901 909 911 1094 1270 1297 1311 1360 1384 1456 1597 1678 1739c 1757 1769 1770 1772 1780 1795 1857 1860 1871 1877 1881 1893 1894 1896 1916 1976 2412 2886 2893 | ⸆3 της 141 1384 2404 ¦ τη 1881 | ⸀4 φυγιας 1719 | ⸀5 παρακατιανης 425 441 442 1881 ¦ πακατανης 452 ¦ παγκατιανης 103 ¦ παγκρατιανης 794 ¦ πακτιανης 122 ¦ πευκατιανης 2412 ¦ παυατιανης 1830 ¦ καπιανης 605 ¦ καταπατιανης 1735 ¦ καππατιανης 623 ¦ καπατιανης 018 0150 97 141 205 209 223 314 1247 1311 1384 1595 1719 1724 1760 1768 1770 1772 1780 1795 1845 1849 1857 1860 1955 2404 2558 2886 ¦ καπατιανης or πακατιανης 1903vid 1976vid | txt 43 81 82 93 131 133 172 177 181 221 250 256 321 325 424 458 517 617 619 622 642 699 910 919 920 922 1162 1175 1244 1448 1734 1862 1879 1880 1888 1905 1906 1908 1912 1917 1920 1921 1922 1923 1927 1934 1954 1956 1997 2007 2110 2298 2374 2401 2541 2625 εγραφη προς τιμοθεον ᾱ επιστολη απο λαοδι431 κειας ητις εστι μρ̄οπολις φρυγιας της κατανης 1751 τελος της ᾱ επιστολης τιμοθεον ητις εγραφη απο λαοδικειας ητις εστιν μρ̄οπολις φρυγια της καπατινης ⸆η 1828 | o 1838 | ⸇ επιστολη ⸆προς τιμοθεον oᾱ ⸇ εγραφη απο λαοδικειας ⸀ητις εστι ⸁μητροπολις φρυγιας της ⸀1 πακατια- 1828 |⸀ητι 1838 | ⸁μρ̄οπολις 459 | νης εκ προσωπου παυλου μονου ⸀1 καπατιανης 1828

3.1.6.b 3.1.6.c

3.1.6.1

Apparatus

190

Thorp and Wasserman

Table 7.1

Subscription Collation and Apparatus (cont.)

Type

Reading

Apparatus

3.1.6.2

η προς τιμοθεον ⊤ εγραφη απο λαοδικειας ητις εστι ⸀μρ̄οπολις φρυγιας της πακατιανης ιδου δη και η εκ λαοδικειας ⸆ παυλου αποστολου ⸉επιστολη προς τιμοθεον ⸁α εγραφη⸊ ⸁απο λαοδικειας ⸂ητις εστι ⸀1 μητροπολις⸃ φρυγιας της ⸀2 πακατιανης ⸇

⊤ πρωτη επιστολη 1758 | ⸀μητροπολις 429 ¦ txt 206

3.1.6.3.a

3.1.6.3.b

3.2 3.3.1 3.3.2

5

προς τιμοθεον πρωτης επιστολης παυλου εγραφει απο λαοδικειας ητις εστι μρ̄οπολις φρυγιας της πακατιανης προς τιμοθεον ᾱ εγραφη απο νικοπολεως ⸆ η προς τιμοθεον ᾱ oεπιστολη εγραφη απο μακεδονιας τελος της ⸆ επιστολης εγραφη απο μακεδονιας ⸆

⸆του αγιου 020 104 ¦ τελος 330 398 451 | ⸉προς τιμοθεον επιστολη ᾱ εγραφη L (020) 104 ¦ επιστολη ᾱ̄ εγραφη προς τιμοθεον 404 | ⸀πρωτης ητις 398 451 ¦ ητις 330 | ⸁επι 467 1959 | ⸂της μρ̄οπολεως 330 398 451 | ⸀1 μρ̄οπολις 104 404 462 467 1959 | ⸀2 καπατιανης 020 462 | ⸇ προτης εις ρωμην παραστασεως 2805 | txt 88 915 1875

025 102 ⸆εγραφη 945 | o1240 | txt 3 469 1642* 2772 ⸆ προς τιμοθεον ᾱ 203 506 | ⸆ ητις εστι μητροπολις φρυγιας της καπατιανης 203 (cf. 3.1.6. a) | txt 1848

The Development of Subscriptions in 1 Timothy and the Pauline Letters

In order to understand the origin, characteristics and development of the subscriptions to 1 Timothy, we will examine them in relationship to (1) subscriptions to other New Testament letters, in particular in the Corpus Paulinum;

the Subscriptions to 1 Timothy

191

(2) versional subscriptions to 1 Timothy;78 (3) other paratexts in the MSS, such as titles, hypotheses and Euthalian material which contain similar information; (4) internal evidence in 1 Timothy and other Pauline letters and Acts; and (5) the wider church history. 5.1 Omission Of the 415 1 Timothy MSS examined, 105 had positive evidence of omission of the subscription, the earliest of which date from the tenth century. This result is not unique to 1 Timothy – in Champagne’s wider analysis of subscriptions, he concludes that “[u]nlike in the superscription tradition, a large number of scribes omitted the subscription.”79 The earliest example of positive evidence of omission of several subscriptions is in P46. Simple Subscriptions (Type 1) 5.2 The subscription of Type 1 in 1 Timothy, πρὸς τιμόθεον ᾱ, is attested in seven MSS (01 044 33 460 1739* 1836 1854). The earliest attestations of Type 1 subscriptions to the Pauline Letters are found in two papyri from the third century: P30 and P46. P30, a fragment that contains 1 Thess 4:12 through 2 Thess 1:2, includes the subscription to 1 Thessalonians: [προς] [θεσσα]λ̣ον̣ε[ικε]ις̣ ᾱ.80 P46 contains the ends of the letters of Romans through Colossians, including Hebrews. Although every extant letter has a superscription, there is only one subscription, namely that of 2 Corinthians which reads [π]ρ̣ος κορινθιους β̅.81 It is likely that 1 Corinthians also contained a subscription but the MS is lacunose where the subscription might have been.82 The MS has an unusual order (with Hebrews following Romans, and Ephesians prior to Galatians), which has “left no traces in the later manuscript tradition.”83 In contrast to the rather sporadic use of subscriptions in P46, there are subscriptions to all the Pauline Letters in the fourth-century Codex Sinaiticus and in the fifth-century Codex Alexandrinus, whereas the fourth-century 78 Metzger, Textual Commentary, 578. 79 Champagne, “Scribal Habits,” 56. 80 Transcription of P30 and P46 from NT.VMR, “Manuscript Workspace” (http://ntvmr.uni -muenster.de/manuscript-workspace). 81 It should be noted that the end of Colossians is fragmentary, but there does not seem to be enough room for a subscription like the one in 2 Corinthians in the lacunose parts. 82 Cf. Edgar Ebojo, “A Scribe and His Manuscript: An Investigation into the Scribal Habits of Papyrus 46 (P. Chester Beatty II – P. Mich. Inv. 6238)” (PhD thesis, University of Birmingham, 2014), 214: “It might be possible also that in its original state, f60v contained the subscription to 1Cor, since there is still sufficient space left on the page.” 83 Trobisch, First Edition, 34.

192

Thorp and Wasserman

Codex Vaticanus is lacunose in several parts (including 1 Timothy).84 The two fourth-century codices, where extant, reflect Type 1 subscriptions consistently, but in Vaticanus a different hand, dated by Tischendorf to the sixth century, has added the place of origin to the extant Pauline subscriptions.85 Codex Alexandrinus contains Type 1 in Romans to Philippians, but then there is a shift to Type 3 from Colossians. The subscription in the important minuscule 1739 has evidently developed in two steps from the simple προς τιμοθεον ᾱ written in semi-uncial to a more elaborate subscription where the scribe did not have enough space and divided the addition (in minuscule hand): εγραφη απο λαοδι προς τιμοθεον ᾱ κιας ητις εστι μρ̄οπολις φρυγιας της πακατιανης (our emphasis).86 The subscription to Romans in 1739 suggests the use of several exemplars: the first hand wrote προς ρωμαιους in semi-uncial, and a later hand added a distinct subscription in minuscule, τελος της προς ρωμαιους επιστολης ητις εγραφη απο κορινθου δια φοιβης διακονου.87 One Type 1 subscription (1854) includes a genre modifier ἐπιστολή. Semi-Elaborated Subscriptions (Type 2) 5.3 Type 2 subscriptions are somewhat more elaborate and include a terminal modifier. Unlike our Type 3, however, this type does not give any additional background information concerning the book. The earliest witness to this type in 1 Timothy is Codex Claromontanus (D 06) in the sixth century: προς τιμοθεον ᾱ επληρωθη, ad timotheum I explicit, where the terminal modifier possibly derives from Latin.88 The subscription of the ninth-century Codex Sangermanensis 84 The subscription to Philemon in Codex Alexandrinus has been cut away. For an overview, see W. Andrew Smith, A Study of the Gospels in Codex Alexandrinus: Codicology, Palaeography, and Scribal Hands, NTTSD 48 (Leiden: Brill, 2014), 130–139. In Alexandrinus, some subscriptions are copied by a different hand than the respective book, but all are ancient. See B. H. Cowper, Codex Alexandrinus. Η ΚΑΙΝΗ ΔΙΑΘΗΚΗ: Novum Testamentum graece ex antiquissimo codice alexandrino a C. G. Woide olim descriptum: ad fidem ipsius codicis (London: David Nutt and Williams & Norgate, 1860), vi. Codex Vaticanus is lacunose in 1 Timothy – Philemon (Heb 9:14–13:25 is Gregory-Aland 1957). 85 See note 41 above. In virtually all recent studies the elaborate subscriptions in Codex Vaticanus are assigned to the fourth century. See, e.g., Payne, “Textual Analysis,” 104; Champagne, “Scribal Habits,” 189; Benjamin Laird, “Early Titles of the Pauline Letters and the Formation of the Pauline Corpus,” Biblische Notizen 175 (2017): 62, 66. Amphoux, “Les lieux,” 89, is an exception: “mais les souscriptions sont postérieures à la copie (7e s.?).” 86 Cf. Kirsopp Lake and Silva New, Six Collations of New Testament Manuscripts, HTS 17 (Cambridge, MA: Harvard University Press, 1932), 191. 87 Cf. Lake and New, Six Collations, 171. 88 The terminal modifier explicit is attested in Codex Bezae in 3 John and in the Gospels where it is translated ἐτελέσθη (“was accomplished”).

the Subscriptions to 1 Timothy

193

(Dabs1 0319) is identical and derives from Claromontanus. The ninth-century diglots 010 and 012 are akin to D and 0319 but add the genre modifier ἐπιστολή (epistola). There are also a number of other Latin MSS that attest to this kind of subscription to 1 Timothy. The Vulgate MSS Codex Fuldensis (between 541–546 CE) and Codex Amiatinus (7th–8th cent.) read explicit epistola ad timotheum I (Fuldensis omits epistola).89 Elsewhere in the Pauline Letters, the genre modifier is attested in the elaborate Type 3 subscription to Titus in the sixth-century Codex Coislinianus (H 015). The earliest attestation of the terminal modifier τέλος in the subscriptions to 1 Timothy is in the ninth-century codex 1900. It is attested in both Type 2 and Type 3 subscriptions. Manuscripts 496 and 636 combine its function as marking the end of a lection as well as the letter, e.g., τελος της ε και της προς τιμοθεον πρωτης επιστολης εγραφη απο λαοδικιας (MS 496).90 5.4 Elaborated Subscriptions (Type 3) Type 3 subscriptions typically contain the place of origin, a reference to the letter carrier in the Pauline Letters, and sometimes additional information as well. According to Champagne, reference to recipients and letter carriers are “core elements” in the subscription tradition, although our investigation confirms that this information is missing in Greek witnesses to 1 Timothy.91 The earliest attestations of Type 3 subscriptions are found in two fifth-century codices. First, Codex Alexandrinus includes the place of origin in Colossians through Titus (including Hebrews which follows 2 Thessalonians here), while Romans through Philippians contain subscriptions of Type 1.92 Second, the palimpsest 048 preserves a subscription to Philemon with the information that it was written from Rome.93 The sixth-century Codex Coislinianus (H 015), also known as Codex Euthalianus, preserves two subscriptions; the first to Titus reads: παυλου αποστολου επιστολη προς τιτον της κρητων εκκλησιας πρωτον επισκοπον χειροτονηθεντα εγραφη απο νικοπολεως της μακεδονιας, “The Letter of Paul, the apostle, 89 Ernst Ranke, ed., Codex Fuldensis: Novum Testamentum latine interprete Hieronymo (Marburg: N. G. Elwert, 1868), 800; Constantin von Tischendorf, ed., Codex Amiatinus: Novum Testamentum Latine: Interprete Hieronymo (Leipzig: Avenarius et Mendelssohn, 1854), 354. Dates: Metzger, The Text, 106–108. 90 Champagne observed this habit in subscriptions to other letters (Champagne, “Scribal Habits,” 404). 91 Champagne, “Scribal Habits,” 404. 92 The British Library, “Royal” (http://www.bl.uk/manuscripts/FullDisplay.aspx?index=0 &ref=Royal_MS_1_D_VIII). 93 Heath, “Gregory 048,” 16, 168 (]αφη απο ρωμ[ ]ς).

194

Thorp and Wasserman

to Titus, ordained the first bishop of Crete, written from Nicopolis in Macedonia” (our translation); the second to Hebrews reads, παυλου αποστολου επιστολη προς εβραιους εγραφη απο ιταλιας δια τιμοθεου, “The Letter of Paul to the Hebrews, written from Italy, delivered by Timothy” (our translation).94 In Codex Vaticanus, a different hand, dated by Tischendorf to the sixth century, has added the place of origin in the subscriptions to the extant Pauline Letters.95 Similarly, a later scribe of the sixth-century diglot Codex Claromontanus has added the place of origin in all the Pauline Letters except Philippians in the Latin text (d), and a different scribe has added the place of origin to the subscription of Romans and 1 Corinthians in the Greek text (D 06).96 Most likely, these additions were made from separate traditions by two different scribes who did not seek to harmonise the Greek and Latin texts.97 Interestingly, the textual tradition to 1 Timothy reflects the largest number of alternative places of origin for a Pauline letter: Laodicea, Nicopolis, Macedonia, and Athens. In the following, we will consider the textual evidence for these places of origin and then briefly discuss them in the context of tradition and history. 5.4.1 Place of Origin: Laodicea Laodicea is by far the most common place of origin in subscriptions to 1 Timothy, the earliest witness being Alexandrinus (5th cent.). It is attested by several versional witnesses, for example, Codex Colbertinus (c), the corrected Claromontanus (dc), some Vulgate MSS and Bohairic Coptic MSS.98 Subsequently, this city became specified first as the capital of Phrygia 94 Images of H (015) are available in the NTVMR (http://ntvmr.uni-muenster.de/manu script-workspace). For an extensive discussion of Codex Coislinianus and the Euthalian edition of the Corpus Paulinum, see Scherbenske, Canonizing Paul, 116–174. Scherbenske does not discuss the subscriptions but points out that several hands worked on the codex. It was retraced with a corrosive ink, probably in the 8th–9th century, since the later hand shows tendencies towards minuscule script (Canonizing Paul, 157–158). In our opinion, the two extant subscriptions to Titus and Hebrews, in particular, preserve the beauty of the original hand (they are only partially retraced), and here it is evident that the later hand added breathing marks, accents, and punctuation. 95 Tischendorf, Novum Testamentum Vaticanum, xxviiii. 96 The additions of the place of origin are in a different hand and color than the rest of the text in the subscriptions, and the corrector has written over the word incipit in some cases and rewritten it after the addition. 97 Tischendorf uses the siglum D** for the corrector who added the places of origin to the Greek text, and d**c for the corrector who added them to the Latin. See Constantin von Tischendorf, ed., Codex Claromontanus, sive, Epistulae Pauli omnes (Leipzig: F. A. Brockhaus, 1852), xxv, 550. 98 Metzger, Textual Commentary, 578; Amphoux, “Les lieux,” 89.

the Subscriptions to 1 Timothy

195

(Type 3.1.2–3.1.4) – this is the form in the Euthalian apparatus – then as Phrygia Pacatiana (3.1.5–3.1.6), probably because Phrygia was divided.99 Part of the textual tradition of Theodoret’s commentary on Paul’s Letters preserves subscriptions to each letter, where the subscription to 1 Timothy reads: ῾Η πρός Τιμόθεον ᾱ ᾽Επιστολὴ ἐγράφη ἀπὸ Λαοδικείας, ἥτις ἐστὶ μητρόπολις Φρυγίας (PG 82:829), lacking the addition of Pacatiana.100 It is likely, however, that these subscriptions in the commentary were added at a later stage in conformity with the Euthalian material.101 The transmission history of the name Pacatiana (Πακατιανή) is complex, since scribes, unaware of the geographic location, probably misspelled and corrected the name several times independently throughout the centuries. For example, it is spelled Καπατιανῆς in the subscriptions to K (018), L (020) and 0150.102 One MS indicates that Laodicea was “formerly a Roman province” (2805).

99

At some point, the emperor Diocletian (244–311 CE) divided Phrygia into Phrygia Prima and Secunda, which are the older terms for Phrygia Pacatiana and Phrygia Salutaris. The Council of Nicaea (325 CE), mentions a Phrygia which contains both Synnada and Laodicea. This seems to suggest that the province, at that point, was yet to be divided, but W. M. Ramsay argues that there is conclusive evidence that this division had already taken place in 325 CE. W. M. Ramsay, The Cities and Bishoprics of Phrygia: Being an Essay of the Local History of Phrygia from the Earliest Times to the Turkish Conquest, vol. 1: The Lycos Valley and South-Western Phrygia (Oxford: Clarendon, 1895), 81–82. 100 Blomqvist, Euthalian Traditions, 171 n. 188. Migne’s edition goes back to a compilation made from Codex Augustanus and Codex Bavaricus. Agnès Lorrain has prepared a commentary on Theodoret’s work on Romans, Le Commentaire de Théodoret de Cyr sur l’Épître aux Romain: Études philologiques et historiques, TUGAL 179 (Berlin: Walter de Gruyter, 2018) and her edition of the commentary is forthcoming in the GCS-series (De Gruyter). In private correspondence, Lorrain reports that the Euthalian subscriptions are preserved in J (GA 012), M (GA 1999) and in its direct copies (GA 1939, GA 1963, GA 1996, München BSB gr 18, Ottob. Gr. 74), S (GA 2242, where it was added by the scribe because it was missing in the Vorlage V = GA 1945). We are grateful for this information. 101 According to Robert C. Hill, who translated the commentary into English, the Greek text in general in Migne’s edition is reliable (Theodoret of Cyrus: Commentary on the Letters of St. Paul, 2 vols. [Brookline, MA: Holy Cross Orthodox Press, 2001], 1:4–5). However, Agnès Lorrain informs us in private correspondence (January 15, 2018), “Indeed looking at the manuscripts of Theodoret there is no doubt that these subscriptions are added by some of the scribes in conformity with the Euthalian edition.” In her thesis, Lorrain further draws attention to a discrepancy in the provenance of 1 Corinthians in Theodoret’s prologue (placing it in Ephesus) and the subscription (Philippi) (“Théodoret de Cyr, Interpretatio in Epistulam ad Romanos Édition, traduction et commentaire” [PhD dissertation, Université Paris-Sorbonne, 2015], 57 n. 126). 102 Similar misspellings of this location have occurred elsewhere. Sergey A. Ivanov, “An Anonymous Byzantine Geographical Treatise,” Revue des Études Byzantines 60 (2002): 174

196

Thorp and Wasserman

In our opinion, Laodicea as the place of origin reflects the earliest attempt to locate the letter, and we agree with J. B. Lightfoot’s proposal that the tradition likely originated from 1 Timothy being identified as the letter from Laodicea mentioned in Colossians 4:16.103 The passage implies that the church had received a letter from Paul and the Colossians were to exchange the two writings.104 The Council of Laodicea (c.363–365) is evidence of the city’s continued importance to the church in post-biblical times. The tradition to place the letter in Laodicea entered the subscriptions to 1 Timothy, spread to other paratexts, and was later advocated by patristic writers like John of Damascus (676–749) and Theophylact (1055–ca. 1108).105 Interestingly, a few MSS (206 429 1758) add to the subscription, ἰδοὺ δὴ καὶ ἡ ἐκ λαοδικείας, “see also the (letter) from Laodicea,” an obvious reference to Col 4:16, implying that the letter mentioned there was part of a correspondence.106 Further, Champagne draws attention to the subscription to Colossians in 424, which reads, “and he said about the one from Laodicea, concerning the first letter to Timothy, see that you read from that (letter); for to Laodicea the second letter to Timothy was written” (our translation).107 On the other hand, this location seems to contradict the internal evidence of the letter in 1 Tim 1:3, “I urge you, as I did when I was on my way to Macedonia, to remain in Ephesus” (NRSV). If Paul had left Ephesus for Macedonia, he would not pass by Laodicea, at least not Laodicea in Phrygia.108 And if Paul had returned from Macedonia going to Laodicea, he would very likely have n. 57, regards Καπατιανή, as found in several sources of an anonymous Byzantine geographical treatise, as a “corrupt reading by metathesis.” 103 J. B. Lightfoot, Saint Paul’s Epistles to the Colossians and Philemon, rev. ed. (London: Macmillan, 1879), 276–277: “It is quite possible that this subscription was prior to the theory respecting the interpretation of Col. iv. 16, and gave rise to it; but the converse is more probably, and in some MSS (ascr 74) the bearing of this subscription on Col. iv. 16 is emphasized, ἰδοὺ δὴ καὶ ἠ ἐκ Λαοδικείας.” 104 F. F. Bruce suggests that the city was “probably evangelized, during Paul’s Ephesian ministry (Acts 19:10) – not by Paul in person but, it appears, by his colleague Epaphras (Col 4:[12–]13).” See F. F. Bruce, “Laodicea,” ABD 4:230. 105 We know that Tertullian was aware of the Marcionite tradition to connect Ephesians with the letter from Laodicea (Col 4:16), which he dismissed (Marc. 5.17), but he shows no awareness of the connection of 1 Timothy to Laodicea; cf. Can. Mur. 64. 106 Cf. Lightfoot, Saint Paul, 277. 107 Champagne, “Scribal Habits,” 226, translates the Greek text as “and he said the one from Laodicea concerning the first epistle to Timothy that you read there, for in reference to the Laodiceans the second epistle Timothy was written” (λεγει δε την εκ λαοδικε περι της προς τιμοθεον α επιστολης οτι αναγνωτε εκεισε εις λαοδικειαν γαρ εγραφη η προς τιμοθεον β επιστολη). 108 Laodicea in Syria is even more out of the way.

the Subscriptions to 1 Timothy

197

passed by Ephesus, as John Calvin notes, “especially since there were many reasons that urged him to visit it.”109 Perhaps it is this difficulty with Laodicea that partly explains the origin of the rival locations in the textual tradition (including other paratexts), especially Macedonia. 5.4.2 Place of Origin: Macedonia, Nicopolis, or Athens Nine MSS (3 203 469 506 945 1240 1642* 1848 2772), the earliest of which is 945 (11th cent.), indicate Macedonia as the place of origin.110 Macedonia is also attested by several MSS with the Euthalian apparatus, two Bohairic Coptic MSS, and more widely in the ὑπόθεσις/argumenta.111 The internal evidence suggests that Paul is once again urging Timothy to remain in Ephesus, now that he has reached his destination in Macedonia (1:3). Only two MSS, the ninth-century Codex Porphyrianus (P 025) and a later minuscule 102, indicate that the letter was written from Nicopolis. On the other hand, this location is consistently attested in the subscriptions to Titus.112 In Titus, it is first attested in Codex Alexandrinus from the fifth century, and from the sixth century in Codex Coislinianus (015) with the further specification, “Nicopolis in Macedonia,” since there were several Nicopolises in the Roman Empire (in Epirus, in Bithynia, ad Nestum, or ad Istrum).113 The tradition is likely based on Titus 3:12, “… do your best to come to me at Nicopolis, for I have decided to spend the winter there” (NRSV). Robert W. Smith points out that “it is widely held” that, after a first imprisonment in Rome, “Paul made his way to Nicopolis, where he was arrested and returned to Rome in a second Roman imprisonment.”114 Smith assumes this must be 109 John Calvin, Commentaries on the Epistles to Timothy, Titus, and Philemon, trans. William Pringle (Edinburgh: The Calvin Translation Society, 1856), 16; cf. Payne, “Textual Analysis,” 158. Calvin writes, “If Paul left for Macedonia then traveled to Laodicea before writing 1 Timothy, he would have traveled almost 700 miles, and the return trip from Macedonia to Laodicea would most likely have gone through Ephesus. This does not appear to fit with the biblical text” (Calvin, Commentaries, 16). 110 Codex 203 attests to a conflation, placing Macedonia in Phrygia Pacatiana. 111 Metzger, Textual Commentary, 578; Willard, Euthalian Apparatus, 71. 112 Metzger, Textual Commentary, 587. 113 “Nicopolis in Macedonia” is further attested in the Euthalian locations’ table as well as in many later subscriptions, but the simple “Nicopolis” is the earliest and most widespread reading (contra Champagne, “Scribal Habits,” 254, who neglects Codex Alexandrinus altogether and assumes that “scribes simplified the tradition by including only an abbreviated form of the reference to the provenance, which omitted the modifier της μακεδονιας”). Further, Smith’s statement that the subscription to Titus in Codex Alexandrinus, “indicating that the letter was sent from Nicopolis … is not attested in other ancient manuscripts,” is misleading (Robert W. Smith, “Nicopolis,” ABD 4:1108). 114 Smith, “Nicopolis,” 4:1108.

198

Thorp and Wasserman

Nicopolis in Epirus (in Achaia), but adds that the note in Titus 3:12 “suggests that Paul was near Nicopolis, not in Nicopolis” (note the use of ἐκεί, “there”) and that “Paul is usually thought to have written the letter from Corinth [in Achaia].”115 From the viewpoint of transcriptional evidence, the tradition that Paul wrote Titus from Nicopolis, widely attested in the subscriptions and other paratexts, likely influenced the subscriptions to 1 Timothy in P (025) and 102. This probably happened at a stage of the text where Macedonia had not yet been added, as reflected in the subscription to Titus in Porphyrianus (P 025), and before that in Alexandrinus (A 01). According to the Bohairic Coptic MS K (14th cent.), the letter was sent “from Athens, [carried] by his disciple Titus” (our translation).116 Titus as letter carrier is also attested in Bohairic Coptic F.117 This location is unique and difficult to explain. 5.4.3 Other Additions and Omissions The subscriptions to the Pauline Letters contain various other additions, attested from the sixth century by MS Codex Coislinianus (H 015), which preserves subscriptions to Titus and Hebrews (see above). In the subscription to 1 Timothy many MSS add παύλου or παύλου ἀποστόλου. The curious reading ἐκ προσώπου παύλου μόνου, “from the person (or presence) of Paul alone,” can be found in a few subscriptions to 1 Timothy in the eleventh century (459 1828 1838).118 This type of addition is also attested in a subscription to Romans from the ninth century (0151).119 Subscriptions that commence with the definite article ἡ are attested from tenth century onwards (075). Some MSS supply a demonstrative pronoun, αυτη η ἐπιστολη (808 1728 2501 2558), whereas 1241 reads η τοιαυτη επιστολη. The subscription to 1728 indicates that 1 Timothy is Paul’s “eleventh letter,” although it is only the tenth in the order. The location of a similar phrase concerning the order is fluctuating in other letters. This information was 115 Smith, “Nicopolis,” 4:1108. 116 George Horner, The Coptic Version of the New Testament in the Northern Dialect: Volume III (Oxford: Clarendon, 1905), 587. 117 Horner, Coptic Version, 586–587. The Bohairic Coptic K also departs from the common Bohairic order by placing Hebrews after Philemon. See F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament: For the Use of Biblical Students, 2 vols. (London: George Bell & Sons, 1894), 2:121 (n. 16). 118 As opposed to εκ προσοπου παυλου και, “from Paul and” which is present in some subscriptions to other letters (Metzger, Textual Commentary, 589; cf. Champagne, “Scribal Habits,” 186–275). 119 Champagne, “Scribal Habits,” 192.

the Subscriptions to 1 Timothy

199

apparently added irregularly by the scribe: at first it occurs in the beginning of some letters, then in the margin, until, finally, it is placed in the subscription.120 Three subscriptions contain the divine-agency modifier, σὺν θω� (91 1840 2008). To our knowledge, this element is attested in the Gospels, Acts, Catholic Letters, and Pauline Letters.121 Finally, several Type 3-subscriptions in 1 Timothy omit the core element πρὸς τιμόθεον ᾱ and often commence with the words ἐγράφη ἀπό instead.122 Probably, a repetition of the title πρὸς τιμόθεον ᾱ was perceived as superfluous by some scribes. 6

The Relationship of Subscriptions to Other Paratexts

Metzger has suggested that the longer subscriptions to the Pauline Letters are probably the work of Euthalius, who created an edition in the fourth century.123 The sixth-century Codex Coislinianus is considered “the oldest extant representative of Euthalius’s editorial work.”124 As we have seen, this codex has elaborated subscriptions to Titus and Hebrews, which mention the place of origin and provides additional information (Titus as bishop in Crete; Timothy as letter carrier of Hebrews). On the other hand, there are a number of subscriptions in the fifth-century Codex Alexandrinus (starting from Colossians) which include place of origin without additional information, and this type is also attested in the fifth-century palimpsest 048. This evidence suggests that the elaborated subscriptions we find in Coislinianus may have developed in stages. This is further confirmed by the rival traditions we have observed in the subscriptions to 1 Timothy. We think that the indication of a place of origin predates the Euthalian edition. Further, as Willard’s more detailed analysis shows, the Euthalian material in general has grown over time; the ὑποθέσεις/argumenta have a distinct origin, and there is sometimes a discrepancy between the indicated place of origin in some argumenta and the corresponding subscriptions in the same manuscript. In 1 Timothy, Macedonia dominates in the argumenta although Laodicea is a 120 The phrase is written in uncial script except in the subscription to 1 Timothy. 121 Champagne, “Scribal Habits,” 398, is unaware of the attestation in the Pauline Letters. To our knowledge, the earliest attestation of the phrase occurs in subscriptions to the Catholic Letters in the ninth-century K (018). See Wasserman, Epistle of Jude, 209; Robert W. Yarbrough, 1–3 John, BECNT (Grand Rapids: Baker, 2008), 13. 122 Champagne’s data confirms that similar omissions are common in all Pauline Letters (“Scribal Habits,” 186–275). 123 Metzger, Greek Bible, 40. 124 Scherbenske, Canonizing Paul, 174.

200

Thorp and Wasserman

regular variant, whereas Laodicea dominates in the subscriptions and the note in the locations’ table. As we compare the earliest Type 3 subscriptions in Alexandrinus with the reconstructed locations’ table in Euthalius’s edition, there are a number of differences between the two.125 With regard to 1 Timothy, the particular form in Euthalius’s locations’ table, “Laodicea of Phrygia,” is clearly secondary to the simple “Laodicea” in Alexandrinus. Thus, we think that the earliest forms of the Type 3 subscription, as reflected in Codex Alexandrinus, predate both the ὑπόθεσις/argumenta, the Euthalian apparatus, and the Latin prologues, although further research is necessary. In our opinion, Metzger, Willard, Payne, and other authors have failed to draw out another important implication of the discrepancies and textual variation in these sources: it is apparent that the various paratextual traditions – the subscriptions, the ὑποθέσεις/argumenta, the Euthalian locations’ table, and the Latin prologues – have affected each other at various points during their history of transmission.126 It would not be surprising if scribes, noting discrepancies between these parts in an exemplar, would attempt to harmonize one tradition to another with cross contamination as a result. 7 Conclusions The earliest attestations of the Type 1 subscriptions to the Pauline Letters are preserved in two third-century papyri (P30 and P46). In 1 Timothy this kind of subscription, προς τιμοθεον ᾱ, is attested in seven MSS (01 044 33 460 1739* 1836 1854) including the fourth-century Codex Sinaiticus (ℵ 01). These early subscriptions correspond to the earliest superscriptions to the Pauline Letters, which ultimately derive from the time when the letters were collected together, regardless of the potential existence of a single canonical edition of Paul’s letters (or the New Testament). The subscriptions became more elaborate through the centuries. At some point, terminal modifiers were added to the subscriptions (Type 2), as reflected in several Greek-Latin diglots, where ἐπληρώθη is the common Greek 125 Willard, Euthalian Apparatus, 80, claims, “This table agrees with the subscriptions in the points of origin with only a minor variation in the elaboration of a place name.” The question remains with which set of subscriptions Willard has compared his Euthalian locations’ table. 126 Cf. Payne (“Textual Analysis,” 138), who suggests that it would be necessary to draw from several sources, the ὑποθέσεις/argumenta, the locations’ table, and the co-writers’ table, to reconstruct each subscription, as if these sources were fixed.

the Subscriptions to 1 Timothy

201

equivalent of the explicit which was standard in Latin books. Many later Greek MSS contain the terminal modifier τέλος. The genre modifier ἐπιστολή, only implied in the earliest form of the titles (and possibly part of the title of the full collection) and subscriptions, was later added to some subscriptions. At some point in the fourth century, a certain Euthalius edited a collection which contain more elaborate subscriptions including the place of origin and additional information in some letters. However, this brief survey suggests that the indication of a place of origin probably predates Euthalius edition. The Type 3 subscriptions to 1 Timothy in the Greek MSS attest to either Laodicea, Macedonia, or Nicopolis as the place of origin with or without further specification. Laodicea is the earliest, and by far the most widespread location. This tradition is likely based on a connection to the letter from Laodicea mentioned in Colossians 4:16 (and several MSS make this connection explicit), an identification which further secures the place of the letter in the collection. Macedonia as the location of the letter is attested in eight MSS and agrees with the note in 1 Tim 1:3, where Paul is said to have left for Macedonia. This reading, however, is rare in the subscriptions and may represent influence from other paratexts; the location is widely attested in the ὑπόθεσις/argumenta. Finally, Nicopolis as the location of the letter (025 102) probably reflects influence from the letter to Titus where Nicopolis is widely attested in subscriptions and other paratexts. The continuous elaboration of subscriptions to 1 Timothy and other Pauline Letters, shaping the way they are read, not only satisfies human curiosity but serves to further authenticate the writings by connecting them to Paul and his circle of co-workers. The location of the letters in time and space reflect the ongoing construction of a “landscape of memory” in the early church.

CHAPTER 8

Second Timothy: When and Where? Text and Traditions in the Subscriptions Conrad Thorup Elmelund and Tommy Wasserman 1 Introduction: Subscriptions – A Neglected Paratext of the New Testament The paratexts found alongside the text in the New Testament manuscripts do not attract nearly as much interest as the text itself, which is reflected in the fact that the subscriptions are excluded from the dominant text critical hand editions (NA28, UBS5). On the other hand, paratextual features may offer a window into the social history of late antique and medieval Christianity and help trace the reception history of the New Testament writings. This article will focus on the subscriptions to 2 Timothy, which Conrad Thorup Elmelund has collated in 485 extant manuscripts under Tommy Wasserman’s supervision. There is a great benefit just in the presentation of the extensive textual data in as precise a manner as possible. In fact, Bruce M. Metzger’s very brief collation of subscriptions to 2 Timothy in the widely used Textual Commentary contains many inaccuracies, perhaps symptomatic of the relative negligence of paratexts in general.1 In addition to a critical apparatus which presents the subscriptions, we will attempt to trace the origin, character, and development of the various traditions reflected in this paratext and see how they illuminate the history of interpretation of 2 Timothy and of the people and different places associated with it. 1 In the case of the six important majuscules (01, 02, 04, 06, 010, and 020), Metzger’s transcriptions have been verified. However, he has MS 33 read πρὸς Τιμόθεον, whereas the present work records προς τιμοθεον β̅; he indicates πρὸς Τιμόθεον δευτέρα for minuscule 90, where the present work records positive evidence of omission, and ἐτελέσθη πρὸς Τιμόθεον β̅ for 012 where the present work records ετελες προς τιμοθεον β̅. In the latter case, Metzger (or his source) likely considered ετελες to represent ετελεσθη. The ending of the participle χειροτονηθέντα, “installed”/“ordained,” is a volatile part in the subscription tradition, and although 018 gets it right, Metzger’s apparatus gives χειροτονηθένα, which seems to be a typographical error. Finally, Metzger’s apparatus, which differentiates between β̅ and δευτέρα, wrongly indicates δευτέρα for 018 (it reads β̅) (Bruce M. Metzger, A Textual Commentary on the Greek New Testament: A Companion Volume to the United Bible Societies’ Greek New Testament, 2nd ed. [Stuttgart: Deutsche Bibelgesellschaft, 1994], 583).

© Conrad Thorup Elmelund and Tommy Wasserman, 2023 | doi:10.1163/97890045

Second Timothy: When and Where?

203

To avoid repetition, we refer to the general survey of research on subscriptions as presented by Linnea Thorp and Tommy Wasserman in the previous article in this volume.2 As Thorp and Wasserman explain, Metzger suggests that the longer subscriptions to the Pauline Letters probably derive from Euthalius’s fourth-century edition. The oldest extant representative is the sixth-century Codex Coislinianus (H 015) which preserves subscriptions that indicate the place of origin and additional information (Titus as bishop of Crete; Timothy as letter carrier of Hebrews). However, Louis Charles Willard, who examined the Euthalian apparatus, has argued persuasively that there is no direct link between the Euthalian apparatus and the subscriptions.3 The fact that locations of Pauline letters were assigned before Euthalius produced his edition is confirmed by Thorp and Wasserman who have demonstrated that the Type 3 subscription to 1 Timothy in Codex Alexandrinus, indicating that the letter was “written in Laodicea,” is older than the indication in the Euthalian apparatus that the letter was written in “Laodicea of Phrygia.”4 Thus, in light of recent research, we conclude that the history of these paratexts, the Euthalian apparatus and the subscriptions, overlap. 2

Materials and Method

We have consulted a total of 521 manuscripts that could contain a subscription to 2 Timothy. Of these, thirty-one were lacunose, one was illegible (from the images), and another four could not be consulted because images were missing.5 Thus, 485 manuscripts are included in the apparatus. In bilingual manuscripts, only Greek subscriptions are included. Most transcriptions have been made from images available through the New Testament Virtual Manuscript Room (NT.VMR).6 In addition, images of eight manuscripts were acquired through the Center for the Study of New Testament Manuscripts 2 See Chapter 7 in this volume by Linnea Thorp and Tommy Wasserman, “The Tradition and Development of the Subscriptions to 1 Timothy.” 3 Louis Charles Willard, A Critical Study of the Euthalian Apparatus (Berlin: Walter de Gruyter, 2009), 79. See further Thorp and Wasserman, “Tradition and Development.” 4 Thorp and Wasserman, “Tradition and Development.” 5 We would like to thank Justin Soderquist for consulting the relevant image of 016 for us confirming that the manuscript is lacunose (the folio ends in 2 Tim 4:20). See further his new edition, Justin J. Soderquist and Thomas A. Wayment, A New Edition of Codex I (016): The Washington Pauline Manuscript, TS 20 (Piscataway, NJ: Gorgias, 2019), 126. 6 Institut für neutestamentliche Textforschung, Münster. “New Testament Virtual Manuscript Room – INTF,” NT.VMR, http://ntvmr.uni-muenster.de/.

204

Elmelund and Wasserman

(CSNTM),7 one each from the University of Chicago Library,8 the Bibliothèque Nationale de France,9 the British Library,10 the Leimonos Monastery,11 the Library of Congress,12 and a private person.13 Finally, minuscule 2009 has been consulted in person by Elmelund at the Royal Library in Copenhagen. Itacisms, iota adscripts, and misspellings have been noted in the transcriptions but are not included in the collation and resulting apparatus below.14 In relation to ordinal numbers, no distinction is made between β̅ and δευτερα(ς) when used in the title of the letter. However, the distinction is retained between δευτερoυ and β̅ when it refers to Paul’s second appearance before Nero (where δευτερoυ is the common form). The text-critical symbols follow NA28, and the dating of manuscripts follow the Kurzgefasste Liste, which is available online in the NT.VMR.15 In four cases, concerning majuscule 0150, minuscules 614, 1739 and 2110, we have distinguished between two distinct subscriptions. In two instances the subscription of the first hand (614T and 1739T) has been supplemented by a second subscription added either in the top margin (614Z), or below by a later hand (1739Z).16 In the case of 0150 and 2110 we find the two different subscriptions (Type 3.2.3 and 3.5.2) present in both manuscripts. Both 7 8 9 10 11 12 13 14 15

16

The Center for the Study of New Testament Manuscripts. “Manuscripts – CSNTM,” CSNTM’s digital library, http://www.csntm.org/Manuscript. University of Chicago Library, Ms 142 (GA 2401), http://goodspeed.lib.uchicago.edu/view /index.php?doc=0142&obj=215. Bibliothèque Nationale de France, Gr. 102 (GA 2298), http://gallica.bnf.fr/ark:/12148/btv1b 8478950g. British Library, Add. 38538 (GA 2484), http://www.bl.uk/manuscripts/FullDisplay.aspx ?ref=add_ms_38538. Leimonos Monastery, Ms. Lesbiacus Leimonos 195 (GA 1758), http://84.205.233.134/lei monos/pdfs/200873095317.pdf (accessed 17 May 2018, now defunct). Athos, Vatopediu 857 (GA 2289), photographed by Library of Congress, https://www .loc.gov/item/00271052054-ma/. We want to thank Hugh A. H. Houghton for sharing the image of the relevant page in GA 81 (Alexandria, Greek Patriarchate 59). The transcriptions are available (open access) through the Data Archiving and Networked Services (DANS), https://doi.org/10.17026/dans-xk3-dabp. Barbara Aland et al., eds., Novum Testamentum Graece, 28th ed. (Stuttgart: Deutsche Bibelgesellschaft, 2012), 56*–57*; Kurt Aland et al., Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments, 2nd ed., ANTF 1 (Berlin: Walter de Gruyter, 1994). All quotations from the Greek New Testament are taken from NA28. Z = “Zusatz” refers to an additional subscription which is not clearly a correction or an alternative reading. On fol. 100r in 1739, a later scribe has added a second more elaborate subscription, Type 3.4.1, under the decoration below the first subscription of Type 3.2.3. On fol. 212r in 614, which attests to Type 3.4.1, a later scribe has added a subscription of Type 2.3 with the terminal modifier in the top margin.

Second Timothy: When and Where?

205

subscriptions belong to the original hand and appear successively divided by a decorative line.17 Other related paratextual materials, such as prayers, blessings, commentary material, stichometric notes, and lectionary markings, are included in our collation only when integrated in the actual subscription, e.g., τελος της δ̄ και της προς τιμοθεον επιστολης, “end of the fourth (reading) and the letter to Timothy” (part of the subscription of 496). 3 Typology The subscription tradition is particularly rich and flexible in the case of the Pauline Letters, as scribes and interpreters tried to provide a setting in time and place for each letter.18 In their survey of the subscriptions to 1 Timothy, Thorp and Wasserman have proposed a typology that largely reflect the development of subscriptions, one that has been applied in this study as well. They distinguish between three main types: (1) Simple subscriptions: The earliest subscriptions are short and often identical to the superscriptions (πρὸς Τιμόθεον β̅). (2) Semi-elaborated subscriptions: Subsequently, a terminal modifier was added to mark the end of the book (e.g., ἐπληρώθη, ἐτελέσθη, τέλος);19 sometimes the author, otherwise implied, was made explicit (e.g., Παύλου τοῦ ἀποστόλου). (3) Elaborated subscriptions: Subscriptions of this type can add (to the above elements) a geographical modifier, i.e., the place of origin, a chronological modifier (e.g., how a letter fits in Paul’s biography), a reference to the

17 The two manuscripts, 0150 and 2110, form a family together with 1506 (lacunose in 2 Timothy), which is currently being studied by David Flood at the University of Edinburgh (see http://cal-itsee.bham.ac.uk/itseeweb/igntp/theses.html). 18 As evident from David G. Champagne, “Scribal Habits Within the Superscription and Subscription Traditions of Greek New Testament Manuscripts” (PhD dissertation, New Orleans Baptist Seminary 2012), 270–274. A striking example in Thorp and Wasserman’s survey of subscriptions to 1 Timothy is the number of alternative places of origin for that letter (“Tradition and Development”). 19 Cf. William H. P. Hatch, Facsimiles and Descriptions of Minuscule Manuscripts of the New Testament (Cambridge, MA: Harvard University Press, 1951), 33. Various elements connected to the terminal modifier could then be elaborated (e.g., with a divine-agency modifier σὺν θεῷ) or an indication of what was ended apart from the letter as such (e.g., the fourth and final reading; the exegetical comments; or, in the case of Jude, the whole corpus of the Catholic Letters).

206

Elmelund and Wasserman

letter carrier (in the Pauline Letters), and sometimes additional information as well.20 As Thorp and Wasserman have shown, the earliest attestation of a Type 3 subscription with a short geographical modifier (in the fifth-century codices 02 and 048) precedes the Type 2 subscriptions by a century.21 In other words, this typology, which takes into account various elements of the subscriptions and their degree of elaboration, may not correspond precisely to the historic development. In this study, particular to the textual tradition of 2 Timothy, the subscriptions of Type 2 have been divided into five sub-groups and Type 3 into six sub-groups according to which elements they include. Type 2.1–2 contain terminal modifiers as they appear in bilingual manuscripts, whereas Type 2.3–5 contain the terminal modifier τέλος with or without further elements. The first two groups (Type 3.1–2) contain distinct (and rival) geographical modifiers; the third group (Type 3.3) a chronological modifier; the fourth group (Type 3.4) a tradition about Timothy’s episcopate; the fifth group (Type 3.5) quotations from Eusebius; and the sixth group (Type 3.6) an indication of the letter-carrier. As far as we can determine, the sequence of all the subtypes (2.1–5 and 3.1–6) largely follows the chronological order of their appearance in the manuscript tradition, as will be shown in the analysis.22 Manuscripts that give positive evidence of omission of subscription (not as the result of lacuna) have been assigned to Type 0. 4

Subscription Collation and Apparatus

Table 8.1 Subscription collation and apparatus

Type Reading



20 21 22

Manuscripts

Subscription unknown Lacunose 015 016 048 061 0151 0241 0259 0262 0272 0285 218 339 400 480 602 612 642 1506 1729 1738 1836 1858 1870 1918 1942 2003 2257 2596 2716 2772 2903 Thorp and Wasserman, “Tradition and Development”; cf. Hatch, Facsimiles, 34. Thorp and Wasserman, “Tradition and Development.” It is to be noted that all the subtypes 3.2–3.5 begin to appear in the ninth century among the extant manuscripts.

207

Second Timothy: When and Where? Table 8.1 Subscription collation and apparatus (cont.)

Type Reading – –

Manuscripts

No images Images illegible

025 241 464 1731 1759

Positive evidence of omission 0

Omit

056 5 38 57 61 69 90 94 110 149 216 226 254 263 319 322 323 327 365 383 454 455 456 491 582 606 608 620 627 628 629 630 676 680 720 796 823 886 891 941 996 1022 1099 1106 1115 1149 1243 1245 1315 1319 1354 1359 1425 1505 1609 1622 1646 1661 1718 1721 1757 1798 1827 1850 1851 1852 1867 1889 1890 1919 1925 1929 1943 1945 1948 1950 1961 1965 1969 1973 1977 1984 1985 1987 1991 1994 1995 2000 2002 2005 2008 2086 2102 2105 2127 2143 2191 2200 2289 2344 2400 2482 2492 2516 2523 2544 2576 2629 2659 2674 2690 2718 2739vid 2774 2816 2817 2889 2893 2899 l680 l884 l2024

Type 1 – Typical inscription is repeated as subscription 1

προς τιμοθεον ⸆ oβ̅

⸆ επιστολη 1854 | o 01 04 | txt 044 33vid

Type 2 – Semi-elaborate subscriptions containing a terminal modifier (e.g. τελος) and potentially an attribution to Paul as author 2.1 2.2 2.3

2.4 2.5

προς τιμοθεον β̅ επληρωθη ⸀ετελεσθη προς τιμοθεον β̅ τελος της ⸆ προς τιμοθεον β̅ ⸀επιστολης ⸇

06 0319 ⸀ ετελες 012 | txt 010 ⸆ προς 2495 | ⸀ επιστολη 1509 ¦ – 2011 | ⸇ παυλου του αποστολου 1501 | txt 103 614Ζ 1930 1947 1978 1992 2197 2248 τελος των εις την προς τιμοθεον β̅ επιστολην 1910 εξηγητικων εκλογων τελος ⸆ ⸆ δ̄ και της επιστολης 1367 | txt 1727

208

Elmelund and Wasserman

Table 8.1 Subscription collation and apparatus (cont.)

Type Reading

Manuscripts

Type 3.1 and 3.2 – Including a geographical modifier (λαοδικειας/ρωμης) 3.1 3.2.1

προς τιμοθεον β̅ εγραφη απο λαοδικειας εγραφη ⸆ απο ρωμης

3.2.2

τελος εγραφη απο ρωμης

3.2.3

⸆ oπρος τιμοθεον o1 β̅ ⸇ εγραφη απο ρωμης

3.2.4

τελος της επιτομης των εξηγητικων των εις τον προς τιμοθεον β̅ εγραφη δε η επιστολη απο ρωμης

02 ⸆ αυτη 1127 | txt 6 35 76 201 204 432 639 664 928 935 959 1003 1075 1100 1101 1248 1250 1251 1503 1628 1636 1652 1746 1848 1860 1865 1876 1892 1899 1960 2004 2218 2494 2501 2626 2691 2723 2865 18 386 444 522 757 801 824 986 1040 1072 1105 1482 1573 1617 1637 1704 1725 1732 1740 1745 1766 1771 1855 1856 1864 2009 2080 2221 2261 2352 2431 2554 2587 ⸆ η 141 469 1069 1722 1893 ¦ τελος συν θ�ω� της 91 | o 1881 | o1 1868 | ⸇ επιστολη 141 469 479 1069 1384 1722 1897 2404 ¦ επιστολης 91 | txt 0150T 312 1739T 1843 1873 2110T 1524

Type 3.3 – Including a chronological modifier (οτε εκ δευτερου παρεστη …) 3.3.1

⸆ προς ⸆1 τιμοθεον o β̅ ⸆2 ⸆3 εγραφη ⸆4 απο ρωμης ⸀οτε ⸂εκ δευτερου⸃ παρεστη ⸀1παυλος ⸂1τω καισαρι ρωμης νερωνι⸃1 ⸆ η 337 378 999 1490 1642 1751 1758 1831 1877 1939 1963 1996 1999 2012 2147 2558 2652 ¦ αυτη η 1728 ¦ τελος της 142 384 483 644 909 2085 ¦ τελος παυλου αποστολου επιστολης 330 451 1398 ¦ η δε δευτερα η 489 927 | ⸆1 αυτον τον 489 927 | o 489 927 1751 | ⸆2 επιστολη 378 1490 1642 1728 1758 1831 1877 2147 ¦ επιστολης 142 384 483 644 909 2085 | ⸆3 ητις 330 451 1398 ¦ η 1728vid | ⸆4 δε 384 | ⸀ οτι 489 927 1751 | ⸂ το δευτερον 1758 ¦ – 1642 | ⸀1 ο παυλος 378 1751 2147 ¦ παυλος ο αποστολος 1877 ¦ – 330 451 1398 | ⸂1 4 547 2147 ¦ 1 4 378 999 1751 1758 1841 1981 1986 2652 ¦ 1–2 4 337 384 ¦ 1–2 ρωμαιων 4 489 927 ¦ 2 330 451 1398 ¦ 1–2 922 2374 ¦ 4 1–2 εκ δευτερου 1642 ¦ 1 καισσαρι 3–4 1963 1999vid ¦ 1–3 νερον 2558 | txt 172 221 314 321 1360 1862 1874 1888 1900 1956

209

Second Timothy: When and Where? Table 8.1 Subscription collation and apparatus (cont.)

Type Reading

Manuscripts

3.3.2

⸆ εγραφη απο ρωμης ⸀οτε ⸉εκ δευτερου παρεστη⸊ παυλος ⸂τω καισαρι ρωμης νερωνι⸃

3.3.3

⸆ ⸀εγραφη απο ρωμης η προς τιμοθεον oβ̅ ⸆1 οτε παρεστη παυλος τω καισαρι νερωνι εκ δευτερου o1παλιν

⸆ τελος 1242 | ⸀ οτι 2183 | ⸉ 3 1–2 436 945 ¦ 1–2 1242 1750 | ⸂ 1 4 1404 ¦ 1–2 4 363 436 ¦ 1–2 εν 3 945 ¦ 1–4 παρεστη 1242 1750 ¦ 1–3 1795 ¦ 1–2 ρωμαιων 4 2183 | txt 607 794 ⸆ εγραφη απο ρωμης η προς τιμοθεον β̅ οτε παρεστη παυλος τω καισαρι νερωνι εκ δευτερου παλιν 3 ¦ τελος της δ̄ και της προς τιμοθεον επιστολης 496 636 | ⸀ εγραφ 496 | o 1886 | ⸆1 επιστολη 496 636 914 1240 1886 | o1 1886 | txt 205 209 296 1869 2736 2886 2909

Type 3.4 – Including ecclesiastical tradition about Timothy (της εφεσιων εκκλησιας επισκοπον χειροτονηθεντα) 3.4.1

⸆ προς ⸀τιμοθεον ⸆1 ⸀1β̅ ⸆2 ⸂της ⸀2εφεσιων ⸀3εκκλησιας ⸆3 ⸀4επισκοπον ⸆4 ⸀5χειροτονηθεντα⸃ ⸆5 ⸋ ⸀6εγραφη ⸆6 ⸀7απο ρωμης⸌ ⸋1 ⸀8οτε ⸆7 ⸂1εκ δευτερου⸃1 παρεστη ⸀9παυλος ⸂2τω καισαρι ρωμης νερωνι⸃2 ⸌1 ⸆ η 206 429 641 1409 1998 1626 1769 1828 1857 1933 2425 ¦ τελος 441 442 457 1916 ¦ τελος της 51 203vid 223 234 367 390 1456 1594 1678 1739Z 1753 1768 1830 1891 2194 ¦ τελος συν θ�ω� της 1840 ¦ τελος της δ̄ και της επιστολης η 614Τ 2412 ¦ τελος της δ̄ και της προς τιμοθεον β̅ επιστολης 2484 ¦ β̅ 1879 ¦ β̅ επιστολη 1880 ¦ επιστολη β̅ 326 1352 1837 ¦ παυλου αποστολου επιστολη 2805 ¦ παυλου αποστολου επιστολη β̅ 404 462 467 1959 ¦ του αγιου αποστολου παυλου επιστολη 020 104 ¦ η τοιαυτη επιστολη εγραφη 1241 | ⸀ τιμοθεν 1959 | ⸆1 επιστολη 336 | ⸀1 ᾱ 131* vid ¦ – 175 326 404 462 467 614T 997 1094 1241 1352 1448 1837 1839 1879 1880 1891 1959 1971 2412 ¦ txt 131c vid | ⸆2 επιστολη 425 460 468 592vid 915 1409 2194 ¦ επιστολης 51 203vid 223 234 367 390 1456 1594 1678 1739Z 1753 1768 1830 1840 1891 | ⸂ 1–2 4 3 5 592vid 614T ¦ 4 1–3 5 97 203vid ¦ 1 3 2 4–5 459 ¦ 4–5 1–3 1891 2194 ¦ 5 1–4 1678vid ¦ 4–5 πρωτον 1–3 632 ¦ 4–5 πρωτον 1–4 2625 ¦ 1 3 2 4 πρωτον 5 1838 ¦ 1–3 5 4 309 | ⸀2 εν εφεσω 431 614T 918 1448 1879 1907 1920 1922 1971 2125 2201 2412 ¦ – 1839* ¦ txt 1839c | ⸀3 μητροπολεως 43 1780 ¦ – 997 1768 | ⸆3 πρωτον 020 104 398 441 442 468 1903 | ⸀4 επισκοπος 460 1247

210

Elmelund and Wasserman

Table 8.1 Subscription collation and apparatus (cont.)

Type Reading

3.4.2

3.4.3

3.4.4 3.4.5 3.4.6

Manuscripts

¦ επισκοπω 1772 1921 ¦ επισκοπω 1951 ¦ επισκον 1971 | ⸆4 πρωτον 88 122 206 250 326 336 404 424 429 431 459 462 467 614T 616 623 915 1161 1175 1352 1448 1611 1626 1720 1735 1760 1770 1837 1845 1847 1879 1907 1912 1920 1922 1959 1971 2125 2201 2412 2425 2805 | ⸀5 χειροτονιθεντα 463 ¦ χειροτονηθεντο 131vid 1311 1998 ¦ χειροτονηθεντι 1921 ¦ χειροτονηθεντοι 1927 ¦ χειροτονηθεντος 2007 ¦ χειροτονoθεντα 1857 ¦ – 97 1839 | ⸆5 β̅ επιστολη 614T 2412 | ⸋ 1241 | ⸀6 εγραφε 463 | ⸆6 δε 223 425 | ⸀7 οπο 1896 | ⸋1 918 1896 | ⸀8 οτι 256 465 623 641 1270 1297 1595 1719 1840 1923 1933 2705 | ⸆7 δε 614T | ⸂1 2 460 ¦ το δευτερον 203vid | ⸀9 παυλω 915 1424 ¦ ο παυλος 2805 ¦ – 460 | ⸂2 1–3 1880 ¦ 4 431 1907 2625 ¦ 2–4 1954 ¦ 1–2 4 102 177 398 465 1839 1903 1976 ¦ 1–2 4 3 1678 2194 ¦ 1 4 1 2 1409 ¦ εν ρωμη 1–4 1241 ¦ 1–2 ρωμαιων 4 88 250 326 424 459 614T 623 915 1611 1352 1448 1735 1760 1769 1837 1879 1912 1920 1922 1971 2125 2201 2412 ¦ 4 2 ρωμαιων 404 467 1959 ¦ 4 1–2 ρωμαιων 441 442 462 632 ¦ 1–2 ρωμαιων 1 4 1175 1838 ¦ 1–2 4 των ρωμαιων 2805 ¦ 1 καισαρ 3–4 234 390 ¦ 1 καισσαρι 3–4 421 ¦ 1 κεσσαρι 3–4 1770 ¦ 1 κεσαρι 3–4 1780 ¦ 1 κεσαρι 4 3 1891 ¦ 1 καισαρος 3–4 1871 2705 ¦ 1–2 νερωνη 3 1875 ¦ 1–3 νερων 42 43 1311 1424 1719 ¦ 1–3 νερονι 463 ¦ 1–2 ρωμαιων νερω 616 | txt 018 075 1 62vid 82 93 133 181 189 325 356vid 440 452 458 498 517 605 617 619 622 665 699 901 910 911 919 920 1162 1244 1270 1297 1390 1595 1597 1598vid 1724 1734 1849 1872 1894 1905 1906 1908 1917 1934 1955 1970 1972 1997 2298 2310 2401 2541 προς τιμοθεον β̅ εγραφη απο ρωμης οτε εκ δευτερου παρεστη 105 παυλος τω καισαρι νερωνι εν ρωμη του τιμοθεου επισκοπου εφεσιων χειροτονηθεντος 0142 χειροτονηθεντα πρωτον της εφεσιων εκκλησιας εγραφη απο ρωμης οτε εκ δευτερου παρεστη παυλος νερωνι τω καισαρι ρωμαιων 808 αυτη εγραφη απο ρωμης προς τιμοθεον της εφεσιων εκκλησιας επισκοπον οτε εκ δευτερου παρεστη παυλος νερωνι τελος της προς τιμοθεον β̅ επιστολης επισκοπον 506vid εγραφη απο ρωμης προς τιμοθεον ᾱ επισκοπον εφεσον εκκλη- 1962 σιας οτε δευτερον παρεστη νερων καισαρι ρωμων

211

Second Timothy: When and Where? Table 8.1 Subscription collation and apparatus (cont.)

Type Reading 3.4.7

Manuscripts

‹[13]› εγραφη απο ρωμης οτε παρεστη εκ δευτερου τω κα‹σ̣α̣ρ̣ι›̣ ‹[18]› χειροτονηθεις υπο του απο ‹[15]› εφεσιων εκκλησιας

228vid

Type 3.5 – Including Eusebian material from Hist. eccl. 2.22.2 (Type 3.5.2 incl. 2 Tim 4:6) 3.5.1

3.5.2

⸀εγραφη απο ρωμης εν δεσμοις ⸁εχομενος και ⸀ γραφη 2243 | ⸁ εχομενοις 876 1832 σημαινων την τε πρωτην απολογιαν και την | txt 2138 παρα ποδας τελειωσιν 0150Z 2110Z απο ρωμης γραφη την επιστολην εν δεσμοις εχομενος ομου σημαινων την τε προτεραν αυτω γενομενην απολογιαν και την παρα ποδας τελειωσιν εγω γαρ φησιν ηδη σπενδομαι και ο καιρος της αναλυσεως μου εφεστηκεν

Type 3.6 – Including a tradition about Mark as the letter-carrier 3.6

5

προς τιμοθεον β̅ εγραφη απο ρωμης εις εφεσον υπο παυλου μαλλον δια μαρκου απεσταλη

81

The Setting of 2 Timothy

5.1 The Letter Situation of 2 Timothy The ancient traditions concerning the place of writing of the Pastoral Letters are connected to various traditions about Paul’s travels and activities. The issues of authenticity have been much debated in exegetical scholarship, since it is difficult to situate these letters in Paul’s life, not to mention other problems relating to their language, theology, and attestation in the earliest sources (for

212

Elmelund and Wasserman

example, they are missing in P46 and Marcion’s canon).23 For these reasons, a vast majority of critical scholars have viewed the letters as pseudepigraphical.24 The question of authenticity is not our concern here but rather the tracing of ancient, sometimes conflicting, traditions connected to Paul’s life and travels and the attempts to place the Pastoral Letters into this context, as reflected in the various subscriptions.25 Since the setting of 2 Timothy, as it was understood in antiquity (as reflected in the subscriptions), is our primary consideration, we will not distinguish further between the historical Paul and the implied author of 2 Timothy. Some information regarding the historical situation can be drawn from the letter itself: It is sent from the apostle Paul to his co-worker Timothy (1:1–2); Paul is imprisoned (1:16; 2:9) and expects the imminent end of his life (4:6–8, 18); The imprisonment seems to be in Rome, as Paul blesses Onesiphoros for finding him there and not being ashamed of Paul’s imprisonment (1:16);26 Paul refers to his first defense when the Lord saved him from the mouth of a lion (4:16–17); Paul is alone except for Luke (4:11); and he urges Timothy to come quickly before the winter and to bring Mark with him (4:9, 11, 21). The question is whether Paul’s first defense (4:16) took place during the same or a previous imprisonment than the one of 2 Timothy. The passage in Acts 28:14–31 records a two-year period in Rome where Paul is under house arrest (vv. 16, 30) guarded by a soldier and a chain (vv. 16, 20). For this reason, Michael Prior sees nothing in Acts to justify that “Paul’s death followed immediately on the course of events described at the end of the Acts,” had Luke known Paul’s death to be the outcome of the trial he recorded.27 Paul’s anticipation 23

See Edmond L. Gallagher and John D. Meade, The Biblical Canon Lists from Early Christianity: Texts and Analysis (Oxford: Oxford University Press, 2017), 43. 24 For a new perspective (in particular on 2 Timothy and Titus) challenging the consensus, see Jens Herzer, “Zwischen Mythos und Wahrheit: Neue Perspektiven auf die sogenannten Pastoralbriefe,” NTS 63 (2017): 428–450. 25 As Thorp and Wasserman explain, Tobias Nicklas has demonstrated how canonical texts played a role in constructing what he terms as an ancient Christian “landscapes of memory,” which could initiate “a starting point of later speculation, interpretation, and the creation of new stories connected to various places.” See Thorp and Wasserman, “Tradition and Development”; Tobias Nicklas, “New Testament Canon and Ancient Christian ‘Landscapes of Memory,’” EC 7 (2016): 23; cf. Nicklas, “Neutestamentlicher Kanon, christliche Apokryphen und antik-christliche ‘Erinnerungskulturen,’” NTS 62 (2016): 588–605. 26 Michael Prior, Paul the Letter-Writer and the Second Letter to Timothy (Sheffield: Sheffield Academic, 1989), 67–68; Prior describes this understanding of 2 Tim 1:16 as referring to Rome as “the almost universally agreed reading of the text” (68). 27 Prior, Paul the Letter-Writer, 79.

Second Timothy: When and Where?

213

of a positive end to his trial is likewise reflected in two of the Captivity Letters (Philippians and Philemon) although their place of origin is uncertain.28 In any case, we agree with Prior’s conclusion that Paul was released from the first imprisonment in Rome. However, his further assumption that 2 Timothy was written during this same imprisonment, “after the first difficult hearing of his case (after which he was, nevertheless, confident of his release), and before his acquittal, at a time when his eyes were fixed on another missionary journey,” is unconvincing in light of the words of Paul’s imminent death in 2 Tim 4:6–8.29 5.2 A Survey of the Early Reception History of 2 Timothy All ancient sources that comment on the manner of Paul’s death indicate that it was execution by decapitation, and the location, when mentioned, is Rome.30 The problem with 2 Timothy, however, is how to fit it with the situation of Paul’s first imprisonment in Rome, as presented in Acts. In the fourth century, Eusebius of Caesarea connected Paul’s death to a second imprisonment in Rome after a fourth missionary journey westward. This solution would allow for 2 Timothy to have been written from a second Roman imprisonment. Certainly, Eusebius based his two-imprisonment theory on information from other earlier sources. For example, Clement of Rome mentions that Paul “reached the farthest limits of the west” (ἐπὶ τὸ τέρμα τῆς δύσεως ἐλθών; 1 Clem. 5.7).31 The reference of this expression is disputed among modern scholars, some of which take it to mean Spain since Paul reveals his plans to travel there in Rom 15:24, and from a Roman perspective, Spain is the farthest limits of the west (as 1 Clement is written in Rome).32 From the perspective of ancient reception, it is significant 28 29 30 31

Prior, Paul the Letter-Writer, 80–81. Prior, Paul the Letter-Writer, 84. Prior, Paul the Letter-Writer, 87. Greek text and translation in Michael W. Holmes, The Apostolic Fathers: Greek Texts and English Translations, 3rd ed. (Grand Rapids: Baker, 2007), 52–53. 32 For an extensive argument for Paul’s missionary trip to Spain and the second imprisonment in Rome, see Rainer Riesner, “The Pastoral Epistles and Paul in Spain (2 Timothy 4:16–18),” in Rastreando los orígenes: lengua y exegesis en el Nuevo Testamento, ed. J. M. G. Pérez (Madrid: Encuentro, 2011), 316–335; Riesner, “Paul’s Trial and End According to Second Timothy, 1 Clement, the Canon Muratori, and the Apocryphal Acts,” in The Last Years of Paul, ed. A. Puig I Tàrrech, J. M. G. Barclay, and J. Frey, WUNT 352 (Tübingen: Mohr Siebeck, 2015), 391–410. In the note to the translation of 1 Clem. 5.7, Holmes takes “the farthest limits of the west” to refer to the Straits of Gibraltar (Holmes, Apostolic Fathers, 53). Several scholars take the reference to mean Rome, e.g., Andreas Lindemann, Paulus in ältesten Christentum: Das Bild des Apostels und die Rezeption der

214

Elmelund and Wasserman

that the apocryphal Acts of Peter and the Muratorian fragment attest to the tradition that Paul traveled to Spain.33 5.2.1 Eusebius Eusebius quotes 2 Tim 4:6, as evidence for the letter being written shortly before Paul’s death (Hist. eccl. 2.22.5). He connects Paul’s first defense (2 Tim 4:16) with the first Roman imprisonment (Hist. eccl. 2.22.2), and he understands Paul’s words about being alone during that defense as the reason for Luke not reporting the end of Paul’s trial in canonical Acts (Hist. eccl. 2.22.6–8). Eusebius was the first to narrate that Paul was freed from his first imprisonment in Rome, later to die during a second one. He locates the writing of 2 Timothy during this second imprisonment, where Nero was less gentle than before (Hist. eccl. 2.22.8). He also understands Paul’s reference to a lion as a reference to the bruality of emperor Nero. Eusebius sees proof for an additional (fourth) missionary journey in between the two imprisonments in Paul’s words about the reason for the Lord saving him initially: ἵνα δι’ ἐμοῦ τὸ κήρυγμα πληροφορηθῇ καὶ ἀκούσωσιν πάντα τὰ ἔθνη (“so that by me the proclamation should be completed, and all the nations hear”; 2 Tim 4:17, our translation). 5.2.2 Theodore of Mopsuestia Theodore, Bishop of Mopsuestia in Cilicia (c.350–428 CE), wrote a commentary on the Pauline Letters expanding on the two-imprisonment theory of Eusebius. This commentary should probably be dated to 404–428 CE, since it may reflect knowledge of the controversy between Jerome and John Chrysostom.34 The work is known today mainly through an obscure Latin version. However, excerpts from the commentary are preserved in Greek catenae, by which the quality of the translation has been determined.35 paulinischen Theologie in der frühchristlichen Literatur bis Marcion, BHT 58 (Tübingen: Mohr Siebeck, 1979), 74–80; Udo Schnelle, Apostle Paul: His Life and Theology (Grand Rapids: Baker, 2005), 385; Claire K. Rotschild, New Essays on the Apostolic Fathers, WUNT 375 (Tübingen: Mohr Siebeck, 2017), 63. 33 Acts Pet. ll. 1–3, 40; Canon Muratori l. 39. For a brief review of the discussions and debate over the Muratorian fragment and its text, see Gallagher and Meade, Biblical Canon Lists, 175–83 (including bibliography); cf. Clare K. Rothschild, “The Muratorian Fragment as Roman Fake,” NovT 60 (2018): 55–82. 34 Alternatively, even to 412–429 CE, because it possibly reveals an awareness of Pelagius’s writings (Rowan A. Greer, Theodore of Mopsuestia: The Commentaries on the Minor Epistles of Paul [Atlanta: SBL Press, 2010], xi). 35 The Latin text along with the corresponding Greek excerpts has been published in Henry Barclay Swete, Théodore de Mopsueste: Theodori Episcopi Mopsuesteni in Epistolas B. Pauli

Second Timothy: When and Where?

215

Theodore assumes the framework of Acts and argues in his introduction to 2 Timothy that the letter was written during Paul’s second imprisonment in Rome.36 Since Paul in 1 Timothy expects to join Timothy soon (1 Tim 3:14), while in 2 Timothy is ready to die (2 Tim 4:6), the second letter, Theodore explains, does not follow immediately upon the first. Further, Timothy was apparently situated in Ephesus when he received the first letter (1 Tim 1:3). In contrast, Paul writes in 2 Timothy that he sent Tychicus to Ephesus (2 Tim 4:12), presupposing that Timothy is elsewhere. This requires some time in between the letters. Thus, Theodore makes it clear that Paul was released after he had first appeared before Nero so that he could make another missionary journey. Subsequently, he was brought back to Rome, where he wrote 2 Timothy in captivity before Nero eventually beheaded him. Theodoret of Cyrrhus 5.2.3 The tradition of Rome as the place of writing is repeated by Theodoret of Cyrrhus (c.393–c.458/466 CE) in his Commentary on the Letters of St. Paul. In the introduction to 2 Timothy, he wrote Ἀπὸ τῆς Ῥώμης τὴν δευτέραν τῷ τρισμακαρίῳ Τιμοθέῳ γέγραφεν Ἐπιστολήν ὁ θεῖος Ἀπόστολος (“The divine apostle has written the second letter to the three-fold blessed Timothy from Rome”; our translation).37 Theodoret then goes on to quote 2 Tim 4:6 in line with Eusebius and Theodore, but contrary to them, he does not seem to assume two imprisonments. Further, in his introduction to Ephesians, Theodoret concludes from the passage in 2 Tim 4:12 that Tychicus was the letter-carrier of Ephesians and that both letters were written relatively close, during the same imprisonment, and sent from Rome.38 Whereas the location of the letter to Rome was common ground, there were apparently different views on the question of chronology.

Commentarii: The Latin Version with the Greek Fragments Vol. II (Cambridge: Cambridge University Press, 1882) and reprinted with translation into English by Greer, Theodore of Mopsuestia. 36 Greer, Theodore of Mopsuestia, xviii, 682–689. 37 Migne PG 82:675. Agnès Lorrain is currently preparing a new text-critical edition of Theodoret’s commentary. 38 Migne PG 82:401. For further discussion, see Agnès Lorrain, Le Commentaire de Théodoret de Cyr sur l’Épître aux Romains: Études philologiques et historiques, TUGAL 179 (Berlin: Walter de Gruyter, 2018), 137.

216 6

Elmelund and Wasserman

Analysis of the Subscriptions to 2 Timothy

6.1 Positive Evidence of Omission, Lacunose, and Inaccessible (Type 0) Out of 485 manuscripts extant at the ending of 2 Timothy, 122 attest positive evidence of omission of the subscription (25%).39 The earliest positive evidence of omission is from the tenth century (056 454 456 627 1851). Another 36 manuscripts were either lacunose or inaccessible.40 6.2 Simple Subscription (Type 1) Type 1 subscriptions are found in a handful of manuscripts. The earliest attestation is in Codex Sinaiticus (01), which dates to the fourth century.41 Subscriptions of this type are virtually identical to the superscriptions which may go back to the initial collection of the Corpus Paulinum. Note, however, that the numeral β̅ is omitted from the earliest uncials (01, 04). The subscription of 1854, προς τιμοθεον επιστολη β̅, with the genre-modifier ἐπιστολή is assigned to Type 1.42 Semi-Elaborate Subscription (Type 2) 6.3 Nineteen manuscripts include a Type 2 subscription. The earliest attestation of this type is found in Codex Claromontanus (06), dating to the sixth century. In our typology, Type 2 subscriptions are set apart from Type 1 because they include a terminal modifier, and from Type 3 since they provide no background information about the book/letter. The reason for the development from Type 1 to Type 2 is unclear. David C. Parker mentions that terminal modifiers with a verb form are standard elements in ancient Latin books.43 Thus, 39 40 41

42

43

This figure is entirely in line with Thorp and Wasserman who documented positive evidence of omission of subscriptions in 105 out of 415 examined manuscripts of 1 Timothy (25%). Hopefully, images of inaccessible manuscripts will be made available through digitization in the future. Outside of 2 Timothy, two papyri from the third century attest to this type: P30, a fragment that contains 1 Thess 4:12 through to 2 Thess 1:2, includes the subscription to 1 Thessalonians: [προς] [θεσσα]λ̣ον̣ε[ικε]ις̣ ᾱ; and P46 preserves a subscription to 2 Corinthians that reads [π]ρ̣ος κορινθιους β̅. See Thorp and Wasserman, “Tradition and Development.” David Trobisch suggests that the titles in the initial Corpus Paulinum (a “Canonical Edition”) consisted of two elements: (1) the genre designation (and title of the collection), Παύλου ἐπιστολή; and (2) the addresse(s) of each letter (e.g., πρὸς Τιμοθεον β̅). See David Trobisch, The First Edition of the New Testament (Oxford: Oxford University Press: 2012), 39–40. David C. Parker, Codex Bezae: An Early Christian Manuscript and Its Text (Cambridge: Cambridge University Press, 1992), 11.

Second Timothy: When and Where?

217

it is possible that such modifiers as used in Type 2.1–2 (ἐπληρώθη, ἐτέλεσθη), uniquely attested in bilingual manuscripts, were initially introduced into Latin and then bilingual New Testament manuscripts. The verb ἐπληρώθη is the common Greek equivalent of the explicit which was standard in Latin books. Many later Greek MSS, however, contain the alternative terminal modifier τέλος (Type 2.3–5). David G. Champagne has suggested that the introduction of τέλος in the subscription may have stemmed from the use of the noun in lectionary markings (the same abbreviation of the word occur most frequently).44 This is uncertain, but the use of τέλος to mark the end of lections possibly influenced its more frequent use in the later manuscript tradition. Interestingly, six manuscripts in the present collation include terminal modifiers that explicitly terminate both the final lesson and the letter.45 Written from Laodicea or Rome (Type 3.1 and 3.2) 6.4 The fifth-century Codex Alexandrinus (A 02) uniquely places the writing of the letter in Laodicea, ἐγράφη ἀπὸ Λαοδικειάς (Type 3.1). In light of the information in 2 Tim 1:17, however, it is unlikely that an ancient scribe or redactor would locate the letter in any other place than Rome.46 Thus, the singular reading in Alexandrinus is most likely an error on the part of the scribe who perhaps accidentally repeated the location of 1 Timothy from an exemplar, an error that could have happened more easily if the subscriptions were copied from a list.47 The Type 3.2 subscription that contains the expected geographical modifier, ἐγράφη ἀπὸ ᾽Ρώμης, indicating Rome as the place of writing, is attested from the ninth century in majuscules 025 (according to Tischendorf) and 0150T.48 44 Champagne, “Scribal Habits,” 114–115. 45 Manuscripts 1367 (Type 2.5) 496 636 (Type 3.3.3) and 614T 2412 2484 (Type 3.4.1). 46 G. S. Duncan’s attempt to place Paul and 2 Timothy in Laodicea based on the unique subscription in Alexandrinus is unconvincing (G. S. Duncan, “Paul’s Ministry in Asia – the Last Phase,” NTS 3 [1957]: 216–218). 47 Interestingly, W. Andrew Smith has identified a “cascading error” in the Eusebian Apparatus of Codex Alexandrinus, a paratext which he believes “was copied from a second exemplar” and an error which he has convincingly shown “is a result of reading and copying the canon number values in the way a list is read and copied.” See W. Andrew Smith, A Study of the Gospels in Codex Alexandrinus: Codicology, Palaeography, and Scribal Hands (Leiden: Brill, 2014), 153. Thus, it seems to us that the scribe of Alexandrinus may have copied one or several exemplars which were lacking some of these paratexts which were then added from a different source. The use of different exemplars may also explain the discrepancy between the superscriptions to the letters (which are uniform across each corpus) and the subscriptions which change form from Colossians onwards. 48 Codex Porphyrianus (P 025), ed. pr. in Constantin von Tischendorf, Monumenta sacra inedita, Vol. 5: Epistulae Pauli et Catholicae palimpsestae (Leipzig: Giesecke & Devrient, 1865), 355.

218

Elmelund and Wasserman

Subscriptions of Types 3.3–6 also contain this same geographical modifier but have added several other elements. The earliest attestation of the geographical modifier in these types is also in the ninth century.49 Interestingly, Thorp and Wasserman have shown that in the case of 1 Timothy, the Type 3 subscription including a geographical modifier is attested one century earlier than Type 2, in Alexandrinus and 048.50 Therefore, we think the subscriptions attesting a Roman origination are much earlier than the other forms of attestion from the ninth-century manuscripts. Notably, the important tenth-century minuscule 1739, the text of which likely reflects a much older uncial archetype, also attests to Type 3.2 in the original subscription. 6.5 Written When Paul Appeared before Nero a Second Time (Type 3.3) At some point in the transmission of the subscriptions, a chronological modifier was added, serving to place the letter into a chronology of the life of Paul.51 This addition is first attested in the two ninth-century minuscules 1862 and 1900 as well as the uncials 018 and 020 (Type 3.4).52 The modifier reads ὅτε ἐκ δευτέρου παρέστη Παῦλος τῷ Καίσαρι Ῥώμης Νέρωνι, “When, for a second time, Paul stood before Nero, the Emperor of Rome” (our translation). This subscription presupposes a two-imprisonment chronology as held by Eusebius and Theodore of Mopsuestia. Although Eusebius and Theodore do not mention a second appearance or trial, Paul’s martyrdom under Nero presupposes it. A second trial, however, is explicitly mentioned in the Euthalian apparatus. 6.6 The Euthalian Apparatus The Euthalian material is generally thought to be a fourth-century work of a deacon Euthalius, about whom very little is known. The work consists of various ancillary material to Acts and the New Testament letters, such as three distinct prologues (for Acts, Paul and the Catholic Letters), chapter lists, book summaries (ὑπόθεσις/argumenta), quotation lists, lection tables, and other material including a location table that lists places from which each letter

49 50 51

Manuscripts 1862 and 1900 (Types 3.3.1), and 018 and 020 (Type 3.4.1). Thorp and Wasserman, “Tradition and Development.” Fourteen manuscripts read οτι instead of οτε. They have been assigned to types 3.3.1, 3.3.2 and 3.4.1, respectively. The change from οτε to οτι might be due to itacism but the latter reading significantly changes the meaning of the phrase to indicate the cause of the letter. 52 This is one century earlier than the manuscripts included in Champagne, “Scribal Habits,” 245.

Second Timothy: When and Where?

219

originated.53 This material is thought to have been composed in an original two-volume work – one volume of Acts and the Catholic Letters, and one volume of the Pauline Letters – which has not survived. The Euthalian material was henceforth transmitted alongside the text of the New Testament in the manuscripts.54 Since the Euthalian material had a tremendous influence on interpretation, it is of great significance when evaluating the reception history of the New Testament in general and its affect on other paratexts, including subscriptions in particular.55 The Euthalian ὑπόθεσις to 2 Timothy opens, ταύτην ἐπιστέλλει πάλιν ἀπὸ Ῥώ­ μης, “This (letter) in turn, he sends from Rome” (our translation).56 As Vemund Blomkvist notes, the ὑπόθεσις reflects no knowledge of two imprisonments.57 In regard to the Prologue to the Life of Paul, Blomkvist points to contradictory information regarding the end of Paul’s life, which is a sign of later editorial activity.58 He labels the earliest section the Life of Paul and the later addition the Chronicle of the Preaching of Paul.59 In the former section Paul’s journey and witness in Rome are followed by the statement: “For when the emperor Nero shortly afterward (μικρὸν ὕστερον) wanted to lead him out of this life, he in fact bestowed true and genuine life upon him” (Blomkvist’s translation).60 This gives the impression that the martyrdom of Paul followed shortly after Paul’s journey from Jerusalem to Rome, seemingly without room for two imprisonments. The Chronicle section, on the other hand, refers to Eusebius’s history as its source: “But Eusebius, who has accurately described the following period [following Acts], has told us also the story of his [Paul’s] martyrdom in the second book of his History of the Church” (Blomkvist’s translation).61 The Chronicle section further indicates a period of ten years of preaching after Paul’s first defense and then explicitly brings up the second appearance before Nero: μεταστειλάμενoς γὰρ τὸν Παῦλον, αὖθις τῷ βήματι παριστᾷ. συνῆλθε δὲ πάλιν ὁ 53 The most comprehensive examination of the Euthalian apparatus remains Willard, A Critical Study of the Euthalian Apparatus. 54 The earliest witness to the Euthalian apparatus is the sixth-century Codex Coislinianus (H 015). 55 Vemund Blomkvist, Euthalian Traditions: Text, Translation and Commentary, TUGAL 170 (Berlin: Walter de Gruyter, 2012), 3–6. 56 Blomkvist, Euthalian Traditions, 87. 57 Blomkvist Euthalian Traditions, 172; Blomkvist refers to our subscription Type 3.4 as “[t]he Byzantine subscription” (172 n. 190). 58 Blomkvist, Euthalian Traditions, 196–197. 59 Blomkvist, Euthalian Traditions, 100–111, 196. 60 Blomkvist, Euthalian Traditions, 103. 61 Blomkvist, Euthalian Traditions, 109; cf. Hist. eccl. 2.22.3–6.

220

Elmelund and Wasserman

Λουκᾶς αὐτῷ (“For when he [Nero] had summoned Paul, he again placed him before the judgement seat, and Luke was with him again”; our translation and italics).62 Thus, this passage in the Chronicle section of the Prologue to the Life of Paul is the first source to explicity mention Paul’s second imprisonment in Rome as a second appearance or trial before Nero as in the subscription of Type 3.3. Further, the use of the verb παριστᾷ (“bring before”) in a juridicial sense with the dative referring to the judge or judgement seat also aligns with the subscription.63 Hence, it is quite possible that this passage is the immediate source behind the chronological modifier in the subscription, if the two do not derive independently from Eusebius’s account or other sources.64 Timothy Installed as the Bishop of the Church of the Ephesians (Type 3.4) The widely attested addition of the ecclesiastical tradition that Timothy, the addressee of the letter, was “installed (the first) bishop of the church of the Ephesians” (τῆς Ἐφεσίων ἐκκλησίας [πρῶτον] ἐπίσκοπον χειροτονηθέντα), is an interesting development in the transmission history of the subscriptions. The earliest attestation of this expansion is found in two majuscule manuscripts from the ninth century, 018 and 020, and is widely attested in the tenth century by 0142 and many minuscules.65 The question is where and when this tradition about the episcopate of Timothy originated.

6.7

6.7.1 Eusebius and the Acts of Timothy Eusebius mentions that Timothy was the first to receive the episcopate in Ephesus: Τιμόθεός γε μὴν τῆς ἐν Ἐφέσῳ παροικίας ἱστορεῖται πρῶτος τὴν ἐπισκοπὴν εἰληχέναι (“It is recorded that Timothy was indeed the first to receive the episcopate of the congregation in Ephesus”; Hist. eccl. 3.4.5, our translation).66 A similar tradition is reflected in the Acts of Timothy, which accounts for Timothy’s installment in Ephesus as patriarch/bishop and his succeeding 62 Blomkvist, Euthalian Traditions, 110. The πάλιν (“again”) does not likely refer to Luke being present with Paul under his first defense, since Eusebius also cites 2 Tim 4:16–17, where Paul recalls being alone at that time. Eusebius also cites 4:11 where Luke is said to be with Paul as he writes the letter. 63 The verb (παρίστημι/παριστάνω) is used in a different sense in 2 Tim 2:15 and 4:17. 64 The complex relationship between the Euthalian material and the subscriptions is debated. See Thorp and Wasserman, “Tradition and Development.” 65 Champagne, “Scribal Habits,” 246, who identified the first attestation of the phrase in the tenth century, apparently overlooked the well-known ninth-century majuscules 018 and 020 cited also in Metzger’s Textual Commentary, 583. 66 Eusebius of Caesarea, Hist. eccl. 3.4.5 (Lake et al., LCL).

Second Timothy: When and Where?

221

martyrdom as archbishop in that city.67 This work cannot have been composed earlier than the second half of the fourth century and may well be later.68 In the opening title, Timothy is either called “bishop” or “patriarch.” Claudio Zamagni who recently edited the text prefers the latter, “The Martyrdom of the Holy Timothy, who was a disciple of the holy apostle Paul and who was the first established as partriarch (κατασταθὲντος πατριάρχου) of the Ephesians, the metropolis of Asia.”69 Several manuscripts, however, read πρώτου ἐπισκόπου, “first bishop” (O, P, B sine πρώτου). Later in v. 2, the oldest Greek witness (P) reads του πρωτου επισκοπησαντος instead of του πρωτου πατριαρχησαντος.70 Further, in v. 6 where Zamagni prints ὁ αὐτὸς ἁγιώτατος ἀπόστολος καὶ πατριάρκης τιμόθεος, “this same Timothy, the most holy apostle and patriarch,” the oldest witness (P) omits και πατριαρκης, “and patriarch.”71 From the viewpoint of internal evidence, the reading “bishop” as attested in the earliest Greek witness is preferable – it seems easier to explain a change from bishop to patriarch (in the title and elsewhere) than vice versa. Titles like this in Christian literature typically expanded in the course of textual transmission. On the other hand, Zamagni who studied the tradition in detail and proposed a stemma codicum preferred the readings with “patriarch.”72 67 Cavan W. Concannon, “The Acts of Timothy: A New Translation and Introduction,” in New Testament Apocrypha: More Noncanonical Scriptures, vol. 1, ed. Brent Landau and Tony Burke (Grand Rapids: Eerdmans, 2016), 395–405. 68 The debates about the date of the Acts of Timothy are reviewed by Concannon in the introduction to his translation (“Acts of Timothy,” 396–397). See also Jennifer Knust and Tommy Wasserman, “The Wondrous Gospel of John: Jesus’s Miraculous Deeds in Late Ancient Editorial and Scholarly Practice,” in Healing and Exorcism in Second Temple Judaism and Early Christianity, ed. Mikael Tellbe and Tommy Wasserman, WUNT 2/511 (Tübingen: Mohr Siebeck, 2019), 186–187. Claudio Zamagni has recently provided a new edition with French translation in: “Passion (ou Actes) de Timothée: Étude des traditions anciennes et édition de la forme BHG 1487,” in Poussières de christianisme et de judaïsme Antiques: Études réunies en l’honneur de Jean-Daniel Kaestli et Éric Junod, ed. Albert Frey and Rémi Gounelle (Prahins, Switzerland: Publications de l’Institut Romand des Sciences Bibliques, 2007), 341–375 (Greek text and translation, 366–375). In the introduction to his edition, he concludes that, though the Acts cannot have been written until after 350 CE, it may well contain earlier traditions (352–357). 69 English translation by Concannon, “Acts of Timothy,” 402. 70 Paris, Bibliothèque nationale de France, gr. 1219 (= P), fol. 69v, https://gallica.bnf.fr /ark:/12148/btv1b10722538m/f69.image. 71 When Photius refers to the Martyrdom of Timothy, he also refers to Timothy as the first bishop of Ephesus rather than patriarch (see next section below). 72 Zamagni, “Passion (ou Actes) de Timothée,” 364. It is the be noted that the reading ἐπισκόπου in the title is attested in independent branches in Zamagni’s stemma, as is πατριάρχου, so there must be contamination or independent textual changes at this point.

222

Elmelund and Wasserman

Nevertheless, all manuscripts agree that Timothy is said to be “the first to sit as bishop” (ἐπισκοπεῖν) in v. 4. Then, at the time of his martyrdom, he is said to be “archbishop” (ἀρχιεπίσκοπος) of Ephesus (v. 13). Whereas the subscriptions exclusively use the participle χειροτονηθέντα (appointed, ordained), the Acts of Timothy uses another participle (κατασταθέντος from καθίστημι) for Timothy’s ordination in the title.73 The difference in wording between the Acts of Timothy and the subscription speaks against any direct literary dependence between the two. It is more likely that both depend on Eusebius or an earlier source (cf. Hist. eccl. 3.4.5, “it is recorded”). It is difficult to ascertain whether the variant readings της εν εφεσω (see Type 3.4.1 ⸀2), and the more widely attested πρωτον (Type 3.4.1, both ⸂, ⸆3 and ⸆4), reflect harmonization to the passage in Eusebius. The designation of Ephesus as a “metropolis” in a few manuscripts, της εφεσιων μητροπολεως (Type 3.4.1 ⸀3), occurs frequently in the Acts of Timothy as well, and reflects the fact that the city became the see of a metropolis after the First Council of Nicaea (325). 6.7.2 Photius and the Martyrdom of Timothy Photius, Patriarch of Constantinople in the latter half of the ninth century, wrote Bibliotheca, a collection of surveys of works of ancient Christian and Pagan authors. Photius probably finished this collection after his retirement from public life in 867 CE. In his 254th book survey, Photius reviews the Martyrdom of Timothy (τῆς μαρτυρικῆς Τιμοθέου). From his survey, it is evident that the work is closely related to the Acts of Timothy, as discussed above. Notwithstanding the close similarities, Ilaria Ramelli proposes that the work surveyed by Photius is an earlier form of the text called Martyrdom of Timothy.74 73 Cf. BDAG s.v. καθίστημι 2; s.v. χειροτονέω 2 (interestingly, BDAG here refers to subscriptions to 2 Timothy and Titus). 74 Ilaria Ramelli, “John the Evangelist’s Work: An Overlooked Redaktionsgeschichtliche Theory from the Patristic Age,” in The Origins of John’s Gospel, ed. Stanley E. Porter and Hughson T. Ong, JOST 2 (Leiden: Brill, 2016), 30–52. Meira Z. Kensky discusses the Acts of Timothy in the context of religious travel in late antiquity. She dates the the work to the fifth century and suggests that the claim that Timothy had been the first bishop of Ephesus is one but not the only reason for its composition. She points out that the entire second section of the text “attempts to fix the date of Timothy’s death and the spot of his martyrion on the Pion [in Ephesus]”–hence, another reason was to reimagine a certain place in Ephesus as loca sancta. See Meira Z. Kensky, “Ephesus, Loca Sancta: The Acts of Timothy and Religious Travel in Late Antiquity,” in The Narrative Self in Early Christianity: Essays in Honor of Judith Perkins, ed. Janet E. Spittler (Atlanta: SBL Press, 2019), 94. We agree that the final redaction of the work probably served several functions. However, Kensky does not take into account Photius’s review of the Martyrdom of Timothy, and the possibility that the Acts of Timothy builds on at least one earlier work (the first editor Hermann Usener, Acta S. Timothei [Bonn: Caroli Georgi Universitas, 1877], 36, assumed

Second Timothy: When and Where?

223

One sentence in Photius’s survey is of particular interest in relation to the subscription Type 3.4. About halfway through, Photius repeats what he already mentioned in the beginning, that Timothy was the first bishop in Ephesus, although with a different wording: ὅτι ὁ ἀπόστολος Τιμόθεος ὑπὸ τοῦ μεγάλου Παύλου καὶ χειροτονεῖται τῆς Ἐφεσίων μητροπόλεως ἐπίσκοπος καὶ ἐνθρονίζεται (“That the apostle Timothy was under the great Paul and was installed and enthroned as bishop of the metropolis of the Ephesians”; our translation).75 The fact that Photius uses χειροτονεῖται may suggest that either his work, or more likely the Martyrdom of Timothy, is the source behind the subscription Type 3.4, which is first attested in manuscripts contemporary to Photius. 6.8 The Eusebian Subscription (Type 3.5) Six subscriptions include excerpts from the latter part of Hist. eccl. 2.22.2. The earliest of these is majuscule 0150 dating to the ninth century.76 The subscription is found in two versions. The short version (Type 3.5.1) combines the usual geographical modifier, known from type 3.2–4 (εγραφη απο ρωμης) with two parts from Hist. eccl. 2.22.2: εν δεσμοις εχομενος και σημαινων την τε πρωτην απολογιαν και την παρα ποδας τελειωσιν (“While being in bonds he also reports about the first defense and the near fulfillment”; our translation). The longer version (Type 3.5.2) is present as a second subscription in 0150 and 2110 which both also contain a Type 3.2.3 subscription (προς τιμοθεον β̅ εγραφη απο ρωμης). The second subscription reads απο ρωμης [ε]γραφη την επιστολην, before including the above-mentioned quotation from Hist. eccl. 2.22.2 in slightly different form. The subscription ends with 2 Tim 4:6 which is also cited in Eusebius’s account (Hist. eccl. 2.22.5). These six subscriptions, in particular, highlight the influence that Eusebius’s writings have had on the subscriptions.

that the work Photius reviewed was earlier). Kensky traces the tradition that Timothy was the first bishop of Ephesus to Eusebius (Hist. eccl. 3.4) and suggests that “he clearly obtained [this information] from the Pastoral Epistles since in the same section he states that Titus was appointed bishop of Crete, information unattested anywhere other than the Letter to Titus” (94 n. 10). However, the Pastoral Letters do not mention that either Timothy or Titus was appointed bishop (although the subscriptions do). We think these traditions developed after the Pastoral Letters but before Eusebius. 75 Photius, Bibliotheca (254). Immanuel Bekker, Photii Bibliotheca ex recensione Immanuelis Bekkeri (Berlin: Reimer: 1824), 468b. 76 Champagne, “Scribal Habits,” 248. Champagne was only able to date this phrase to the twelfth century and seems unaware of the material being a quotation from Eusebius.

224

Elmelund and Wasserman

6.9 Mark as Letter-Carrier (Type 3.6) According to Champagne, traditions about letter-carriers are “core elements in the Pauline tradition,”77 but this is hardly the case in 2 Timothy where only a single subscription includes such information.78 The unique addition in minuscule 81 (Type 3.6) reads, εις εφεσον υπο παυλου μαλλον δια μαρκου απεσταλη (“to Ephesus by Paul, but rather sent through Mark”; our translation). The indication that the letter was sent to Ephesus conflicts with the information from Theodore of Mopsuestia, as discussed above. Furthermore, Paul urges Timothy to come to him immediately (2 Tim 4:9), and since only Luke is with him, Paul tells Timothy to bring Mark along (4:11), which at least implies that Mark was not with Paul at the time when Timothy received the letter.79 Did an ancient interpreter deduct that Mark was the letter-carrier and that Timothy should bring him back to Paul? It is true that 2 Tim 4:11 presupposes that Mark was in Paul’s service, but it may seem awkward that Paul should tell Timothy to bring Mark back, an instruction he could have given himself if Mark was the letter-carrier. At this point, we are unable to trace the tradition about Mark as the letter-carrier beyond minuscule 81, copied by the scribe Monk Ioannes in the year 1044 CE.80 It should be noted, however, that this manuscript is known to preserve a very ancient text in Acts and the New Testament letters.81 6.10 Subscriptions in Catena Manuscripts Two catena (commentary) manuscripts, 1524 and 1910, include subscriptions stating that the commentary to the letter has ended. The two subscriptions are not identical and have thus been assigned to different types in the apparatus. The short version in 1910, τελος των εις την προς τιμοθεον β̅ επιστολην εξηγητικων εκλογων (“End of the commentary extracts to the second letter to Timothy”; our translation), was printed in the catena text published by John Antony Cramer in 1844.82 The subscription dates to the eleventh or twelfth 77 Champagne, “Scribal Habits,” 404–405. 78 Cf. Thorp and Wasserman, “Tradition and Development,” who conclude that no lettercarrier is mentioned in the subscriptions to 1 Timothy. 79 It is worth noting that minuscule 81 agrees with NA28 in 2 Tim 4:11. 80 The British Library. “Acts of the Apostles (Gregory-Aland 81),” Digitised Manuscripts, http://www.bl.uk/manuscripts/FullDisplay.aspx?ref=add_ms_20003. 81 Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 4th ed. (New York: Oxford University Press: 2005), 88; Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, trans. Erroll F. Rhodes (Grands Rapids: Eerdmans, 1987), 129. 82 John Antony Cramer, ed., Catenæ Græcorum Patrum in Novum Testamentum: Tomus VII In epistolas S. Pauli ad Timotheum, Titum, Philemona, et af Hebræos (Oxford: E Typographeo

Second Timothy: When and Where?

225

century and has been assigned to Type 2.4, because it does not give any background information. Minuscule 1524 has a longer version, τελος της επιτομης των εξηγητικων των εις τον προς τιμοθεον β̅ εγραφη δε η επιστολη απο ρωμης (“End of the epitome of the comments to 2 Timothy. The letter was written from Rome”; our translation). The subscription, dating to the fourteenth century, has been assigned to Type 3.2.4 since it also includes the geographical modifier. 7 Conclusions All 485 extant and accessible subscriptions to 2 Timothy in Greek New Testament manuscripts have been transcribed and collated into the apparatus presented. Whereas 122 (25%) of the extant manuscripts contain positive evidence of omission, a handful contained Type 1 subscriptions, nineteen contained Type 2 subscriptions but the large bulk of manuscripts attest to various subscriptions of Type 3, which in turn is divided into six subtypes. Except for the miscellaneous reading of Codex Alexandrinus, Type 3 subscriptions are attested from the ninth century onwards. The early reception history of 2 Timothy locates the letter to Rome, a location which is also widely attested in the subscriptions from the ninth century. This tradition is only challenged by the fifth-century pandect Codex Alexandrinus, which uniquely indicates Laodicea – a reading probably due to a scribal error. In contrast, the various sources are divided on the occasion of the letter and how it fits into Paul’s biography. On the one hand, the Euthalian ὑπόθεσις, the Life section of the Euthalian Prologue to the Life of Paul, and Theodoret of Cyrrhus all seem to presuppose a one-imprisonment chronology of the life of Paul. On the other hand, Eusebius, Theodore of Mopsuestia, and the Chronicle section of the Euthalian Prologue to the Life of Paul (which depends on Eusebius) all agree on a two-imprisonment chronology. The latter view is in turn reflected in the chronological modifier of Type 3.3 and 3.4 from the ninth century, which dates the writing of 2 Timothy to Paul’s second appearance before Nero. The chronological modifier seems to be influenced by the Chronicle section of the

Academico, 1844), 82; cf. Hugh A. G. Houghton and David C. Parker, “An Introduction to Greek New Testament Commentaries with a Preliminary Checklist of New Testament Catena Manuscripts,” in Commentaries, Catenae and Biblical Tradition, ed. Hugh A. G. Houghton, Texts and Studies 3.13 (Piscataway, NJ: Gorgias, 2016), 21, “For the rest of Paul [including 2 Tim], he prints the text of Paris, BnF, Coislin grec 204 (GA 1910).”

226

Elmelund and Wasserman

Euthalian Prologue to the Life of Paul but the information likely originated from Eusebius’s Hist. eccl. 2.22.2–8. In conclusion, it seems that the tradition of the episcopate of Timothy, first mentioned by Eusebius, was included in the Acts of Timothy, and although no direct literary relationship between the Acts of Timothy and the subscription of Type 3.4 can be determined, the very wording is found in Photius’s survey of an earlier version of the Acts of Timothy namely the Martyrdom of Timothy. Thus, it is possible that this earlier writing or, less likely, Photius’s own survey is the immediate source of the tradition reflected in Type 3.4. The fact that Eusebius’s writings have been highly influential on the subscriptions is further reflected by the fact that six subscriptions quote directly from Hist. eccl. 2.22.2 (Type 3.5). Minuscule 81 is the only manuscript to include an indication of a letter-carrier (Type 3.6) but in light of internal evidence, the identification of Mark as the letter-carrier is unexpected. Letter-carriers are otherwise completely absent in the subscriptions to 1 and 2 Timothy in the Greek manuscripts. Presumably, more Greek New Testament manuscripts can be collated in the future so that additional subscriptions can be added to this collation. Now that nearly full collations and analysis of the subscriptions to 1–2 Timothy have been completed, the road is paved for someone to work on Titus, the final Pastoral Letter.

CHAPTER 9

Composite Citations in New Testament Greek Manuscripts Sean A. Adams and Seth M. Ehorn 1 Introduction Ancient scribes employed a range of paratextual features to signal quotations in Greek manuscripts, which in turn provide a fascinating window into early reading practices. Although we explore the diversity of scribal practices for indicating material taken from another text, we are primarily interested in looking at a particular sub-set of quotations, namely composite citations, and how they may or may not be differentiated from “normal” citations by ancient scribes/readers. Composite citations may be defined as “when literary borrowing occurs in a manner that includes two or more passages (from the same or different authors) fused together and conveyed as though they are only one.”1 Some composite citations are easy to identify (e.g., Mark 1:2–3), and others less so (e.g., Mark 11:17 / Matt 21:13 / Luke 19:46). In the conclusion of our recent volume, Composite Citations in Antiquity: Volume 2: New Testament Uses, we observed that the determination of a composite citation depended upon various factors, including reception history: “[T]he way that the text has been read by early readers.”2 For this project we intend to explore this aspect further. The data we examine here builds upon the results from our project, utilizing the texts we identified as composite as the chapter’s primary focus. We checked all the Greek papyri currently listed on the Institut für Neutestamentliche Textforschung (INTF) database3 and several Greek codices (ℵ, A, B, D05, D06, 012) where they include composite citations. Most of the texts we accessed through the Center for Study of New Testament Manuscripts (CSNTM), but if they were missing from CSNTM, we consulted the editio princeps of the relevant 1 Sean A. Adams and Seth M. Ehorn, “What is a Composite Citation?” in Composite Citations in Antiquity: Vol. 1: Jewish, Graeco-Roman, and Early Christian Uses, ed. Sean A. Adams and Seth M. Ehorn, LNTS 525 (London: Bloomsbury T&T Clark, 2015), 4. 2 Sean A. Adams and Seth M. Ehorn, “Composite Citations in Antiquity: A Conclusion,” in Composite Citations in Antiquity: Vol. 2: New Testament Uses, ed. Sean A. Adams and Seth M. Ehorn, LNTS 593 (London: Bloomsbury T&T Clark, 2018), 235. 3 http://ntvmr.uni-muenster.de/home.

© Sean A. Adams and Seth M. Ehorn, 2023 | doi:10.1163/9789004537972_011

228

Adams and Ehorn

text.4 Our purpose was to determine what readers and copyists of these Greek manuscripts understood regarding composite citations within the text of the New Testament. To determine this, we looked at the different ways that scribes (and readers) signaled text citations using paratextual features. What we found was that the occurrence of these features increased over time and that subsequent readers were more likely to include paratextual elements to signal the citation of a text. Furthermore, the sophistication of their readings, as well as their recognition of composite citations, also increased over the centuries. 2

Paratextual Features

Before analyzing the data from Greek manuscripts, we will introduce briefly the paratextual features that will be discussed below. René Falkenberg notes that “[a]ll paratextual elements supplement the main text either for communicative purposes or aesthetical purposes, in often overlapping ways.”5 Most of the following features may be classified as “readers’ aids,” which reflects a concern “to guide and facilitate reading of the texts.”6 Several of these readers’ aids can be classified as a form of punctuation, which, as Jocelyn Penny Small reminds us, functions as a kind of commentary within an ancient manuscript,7 a perspective also found in ancient authors (cf. Galen, Ind. 14). But, especially in the later manuscripts, the features we discuss also clearly have an aesthetic element to their presentation. The types of paratextual features we introduce here are those that appear in the citations we examine below. This includes the use of spacing, punctuation marks, ekthesis and indentation, paragraphoi, diplai, rubrication, and source attributions. 4 E.g., Victor Martin and J. W. B. Barns, Papyrus Bodmer II. Supplement. Évangile de Jean chap. 14–21 (Cologny-Geneva: Bibliothèque Bodmer, 1962), for full plates of P66. 5 René Falkenberg, “The Making of a Secret Book of John: Nag Hammadi Codex III in Light of New Philology,” in Snapshots of Evolving Traditions: Jewish and Christian Manuscript Culture, ed. Liv Ingeborg Lied and Hugo Lundhaug (Berlin: De Gruyter, 2017), 106. 6 Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 177. Cf. William A. Johnson, “Toward a Sociology of Reading in Classical Antiquity,” AJP 121 (2000): 593–627. 7 Jocelyn Penny Small, Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity (New York: Routledge, 1997), 22. Johannes de Vries and Martin Karrer, “Early Christian Quotations and the Textual History of the Septuagint: A Summary of the Wuppertal Research Project and Introduction to the Volume,” in Textual History and the Reception of Scripture in Early Christianity / Textgeschichte und Schriftrezeption im frühen Christentum, ed. Johannes de Vries and Martin Karrer, SCS 60 (Atlanta: Society of Biblical Literature, 2013), 10, make a similar point about the function of diplai as marking “the transition from quotation to interpretation” in a manuscript.

Composite Citations in New Testament Greek Manuscripts

229

2.1 Blank Space(s) Because most ancient Greek texts were written in scriptio continua, the use of a blank space within a line of text is the most basic way of delimiting a unit of text. Turner indicates that “[a] space deliberately left blank is also to be considered as a mode of punctuation.”8 Many of the other features described below were able to be added by scribes or readers well after the text was copied, but the inclusion of spacing must be added (or simply reproduced) by the scribe at the time of text production. Sometimes spacing is used at the end of a line and the text continues on the following line, as in Figure 9.1.

Figure 9.1 Acts 16:26–32 in P74 (f. 74a)

Here we see that a space large enough for three or four letters was left, signaling a clear break. The high point (στιγμὴ τελεία) is also a form of punctuation that we will discuss below. There are also examples where spacing is used intralinear, as seen in Figure 9.2.

Figure 9.2 Romans 11:36–12:1 in P46 (f. 16v)

Here a large, intralinear space serves the function of marking a sense unit. In this example, the text of Rom 11:36 is followed by a large space before the beginning of Rom 12:1. It is significant that this space appears at the transition to paraenetic material within the letter body (Rom 12:1–15:32), suggesting that intralinear spacing can be used to mark a significant sense unit in the text.9 Below we will see that sometimes quotations are made using blank spaces. 8 E. G. Turner, Greek Manuscripts of the Ancient World (Princeton, NJ: Princeton University Press, 1971), 10. 9 See Sean A. Adams, “Mark, Manuscripts, and Paragraphs: Sense-Unit Divisions in Mark 14–16,” in Mark, Manuscripts, and Monotheism: Essays in Honor of Larry Hurtado, ed. Chris Keith and Dieter T. Roth, LNTS 528 (London: Bloomsbury T&T Clark, 2014), 61–78; Jeffrey A. D.

230

Adams and Ehorn

Related to the use of spacing, sometimes scribes used “filler marks” (e.g., resembling multiple diplai, “>>”) to mark sense units.10 These marks would appear near the right margin of the manuscript, as may be seen in Figure 9.3.

Figure 9.3 Romans 3:10 in GA 012 (f. 4v)

2.2 Punctuation Marks Points, dots, or other signs within the text were used by scribes to mark sense units. A uniform description of their function is not possible across time, but the marks may be described briefly as follows. A high point (στιγμὴ τελεία) functions as the equivalent of a full stop (see Figure 9.1 above).11 The middle point (στιτμὴ μέση) was used with varying values by scribes but generally signaled a pause less severe than στιγμὴ τελεία (perhaps equivalent to a modern comma). Another paratextual feature added to Greek manuscripts is the use of an upward stroke similar to an acute accent. In P46, these upward strokes appear mainly in Romans and Hebrews, perhaps because these texts were marked for liturgical reading or personal study by a later reader.12 In the Figure 9.4, upward strokes (ll. 1 and 4) mark a sense unit that introduces and includes a scriptural quotation.

Figure 9.4 Romans 10:6 in P46 (f. 13r)

Weima, Paul the Ancient Letter Writer: An Introduction to Epistolary Analysis (Grand Rapids: Eerdmans, 2016), esp. 91–164. 10 Discussed in Hurtado, Earliest Christian Artifacts, 179. 11 Bruce M. Metzger, Manuscripts of the Greek Bible: An Introduction to Palaeography (Oxford: Oxford University Press, 1981), 31–32; Edward M. Thompson, An Introduction to Greek and Latin Palaeography (Oxford: Clarendon, 1912), 60. 12 A brief, but helpful, summary of diacritical marks in P46 is available on the University of Michigan’s website: https://www.lib.umich.edu/reading/Paul/diacritics.html.

Composite Citations in New Testament Greek Manuscripts

231

A final mark of punctuation is the dicolon, which resembles a modern colon mark (“:”). Like the other punctuation marks above, it functions to divide the text into small units. While the dicolon could be used for a range of features (e.g., marking change of speaker), its most common use was as a strong pause.13 See Figure 9.9 below. 2.3 Ekthesis and Indentation Ekthesis refers to when the first letter of a line protrudes into the left margin. Turner refers to it as “reverse indentation” and further notes that it may be used to mark the presence of metrical units, the start of new sections, or new entries in a list or commentary.14 Roberts notes that ekthesis is a scribal practice taken over in Christian manuscripts in imitation of the formatting found in documentary texts.15 In Figure 9.5, ekthesis is used to mark a new section and is often paired with blank space in the previous line. Indentation is a similar scribal practice to ekthesis but refers to the indentation of a line or lines of text away from the left margins. Unlike ekthesis, indentation is rarely used to signal new sense units,16 but is used in some New Testament manuscripts to visually mark scriptural citations (e.g., Bezae, Claromontanus).17 2.4 Paragraphos A paragraphos typically appears as a horizontal mark that “served to designate the end of a preceding section.”18 This paratextual feature, when written by the original scribe, rarely occurs on its own, but often appears in conjunction 13 See Turner, Greek Manuscripts, 10, for a summary of the function(s) of the dicolon. 14 E. G. Turner, Greek Papyri: An Introduction (Oxford: Oxford University Press, 1968), 14; cf. Turner, Greek Manuscripts, 9. 15 Colin H. Roberts, Manuscript, Society and Belief in Early Christian Egypt: The Schweich Lectures 1977 (Oxford: Oxford University Press, 1979), 17. 16 For some examples of indentation in Septuagint manuscripts and Qumran, see Emanuel Tov, “The Background of the Sense Divisions in the Biblical Texts,” in Delimitation Criticism: A New Tool in Biblical Scholarship, ed. M. Christina, A. Korpel, and J. M. Oesch, Pericope 1 (Assen: Van Gorcum, 2000), 312–350. According to Tov, most paratextual signs used in Greek scholarship (e.g., asteriskos, obelos, anchora) are not evidenced in Judean Desert texts. This includes the diple, although the “fish hook” shaped paragraphos might function in this way. Emanuel Tov, Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert, STDJ 54 (Leiden: Brill, 2004), 173, 259. 17 For a discussion of this feature in modern texts/editions, but with minimal discussion of ancient manuscripts, see M. Peppard, “‘Poetry,’ ‘Hymns,’ and ‘Traditional Material’ in New Testament Epistles, or How to Do Things with Indentations,” JSNT 30 (2008): 319–342. 18 Hurtado, Earliest Christian Artifacts, 184.

232

Adams and Ehorn

with line spaces and/or ekthesis.19 The paragraphos provides an additional visual cue to the reader and can be used strategically to reinforce a text break. Again, this feature was not unique to New Testament scribes, but was adopted by them, having been used previously (among other signals) to mark the change in speaker in dramatic texts (e.g., P.Mil.Vogl. VI 263).20 As may be seen in Figure 9.5, a paragraphos typically appears above wording in the left part of the text column and here is accompanied by other features (e.g., ekthesis).

Figure 9.5 Matthew 6:7 in Codex Sinaiticus (f. 74v)

Because paragraphoi are written between lines, without interrupting the text itself, this paratextual feature can be introduced either at the time of copying or as a later addition. 2.5 Diplai in Ancient Manuscripts The diple is a wedge-shaped symbol (>) that appears in the left margin of manuscripts.21 The development of this sign traces back to Aristarchus (b. 215 BC) and forms part of a transition to a genre of scholarly commentary on Greek literature.22 For Aristarchus, the diple was a multi-functional sign, often employed to mark a place where a text required further commentary 19

20 21 22

Jongkind asserts that a break is required to have at least two features in order to be considered a paragraph. Although this is part of his definition for Sinaiticus, we would argue that any break in the text, including, but not limited to, space within a line, functions as a delimitating marker for the reader and so defines sense-units. Cf. Dirk Jongkind, Scribal Habits of Codex Sinaiticus, TS 5 (Piscataway, NJ: Gorgias, 2007), 95–96. For Greek and Jewish use of paragraphoi, see Turner, Greek Manuscripts, 8–9; Tov, Scribal Practices, 178–87. Raymond Clemens and Timothy Graham, Introduction to Manuscript Studies (Ithaca, NY: Cornell University Press, 2007), 86–87; Thompson, Greek and Latin Palaeography, 66; Turner, Greek Manuscripts, 17; Turner, Greek Papyri, 117. Rudolf Pfeiffer, History of Classical Scholarship: From the Beginnings to the End of the Hellenistic Age (Oxford: Clarendon, 1968), esp. 210–33; L. D. Reynolds and N. G. Wilson, Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature, 4th ed.

Composite Citations in New Testament Greek Manuscripts

233

beyond what the margin of a manuscript could contain. In later commentaries the use of a diple could signal some confusion in the text. For example, in P.Oxy. VIII 1086 1.11 the commentator employs the diple to signal “reverse order” in the Homeric text (cf. 1086 1.27; 2.21–23). Similarly, in P48 the scribe used a diple to note a correction in Acts 23:27. By the second century CE, we have a clear example of a source text being marked through the use of diplai from a commentary on Plato’s Theaeteus (see P.Berol. 9782). The most common use of diplai in biblical manuscripts was for marking quotations (cf. Isidore of Seville, Orig. 1.21.13).23 In Figure 9.6, diplai appear in the left margin to mark a quotation.

Figure 9.6 Romans 3:10–12 in GA 012 (f. 4v)

2.6 Rubrication Rubrication of the text by the scribe is an important method of signaling something to the reader. For almost all ancient manuscripts the text is written in black ink. Accordingly, the change of text color (e.g., from black to red) stands out and communicates something to the reader. In some manuscripts (Codex Bezae) red ink is a decorative feature that signals the beginning of a new biblical book,24 while in other codices, it marks a scriptural quotation (e.g., Codex Claromontanus).25 This feature of marking a citation is less common than other means, especially diplai, as it requires access to red ink and the willingness (Oxford: Oxford University Press, 2013), 11. The letter chi (Χ) had the same function as the diple but was used for tragedy scholia. 23 Several important studies on diplai in New Testament manuscripts (‫א‬, A, B) are published in Martin Karrer, Siegfried Kreuzer, and Marcus Sigismund, eds., Von der Septuaginta zum Neuen Testament: Textgeschichtliche Erörterungen, ANTF 43 (Berlin: De Gruyter, 2010). See also Thompson, Greek and Latin Palaeography, 63; Turner, Greek Manuscripts, 12 (see pp. 117–118 for a list of manuscripts with diplai); Clemens and Graham, Manuscript Studies, 86–87. 24 Metzger, Greek Manuscripts, 15. 25 On rubrication in ancient manuscripts, see some of the studies in Richard Hunter and S. P. Oakley, Latin Literature and its Transmission, CCS (Cambridge: Cambridge University Press, 2016).

234

Adams and Ehorn

to switch color.26 However, its potency for drawing the reader’s attention is high, as may be seen in Figure 9.25 (below). It also allows for greater accuracy for signaling which words are or are not part of a quotation, as will be discussed below. 2.7 Source Attributions Because we are interested in studying how ancient readers understood composite citations in Greek manuscripts, the most significant paratextual feature available is an attribution to one (or more) author(s) in the margin of the text. This practice has an ancient pedigree, perhaps attested as early as the first century BC in Sophocles’s Inachus (P.Oxy. XXIII 2369).27 Yet, marginal attributions are not witnessed in any of the surviving New Testament papyri and are found in only a few of the early Greek codices (esp. Sinaiticus). However, by the 9th century, a well-developed source attribution practice had emerged in some manuscript traditions. Figure 9.7 includes an attribution to “Jeremiah.”

Figure 9.7 1 Corinthians 14:21 in GA 012 (f. 36r)

Marginal attributions provide some of the best evidence for understanding how ancient scribes, copyists, or readers understood the origin of a quotation as well as if multiple scriptural passages were brought together in a composite citation. 3

Greek Papyri

Having briefly outlined several paratextual features, we turn our attention to the Greek papyri of the New Testament. Because of the rareness of paratextual features in Greek papyri, the following discussion is organized by feature (e.g., spacing, punctuation, etc.). This allows us to consider similar features across the papyri in a systematic manner.

26 See discussion of inks in Metzger, Greek Manuscripts, 17–18. 27 Turner, Greek Manuscripts, 15.

Composite Citations in New Testament Greek Manuscripts

235

3.1 Spaces in the Papyri Several quotations we examined include spaces to mark sense units, both to open and close quotations and also within the quoted text. For example, a quotation in Rom 11:26–27 (cf. Isa 59:20–21; 27:9) of P46 begins about half-way through the first line. Following the introductory phrase καθὼς γέγραπται, there is a space of approximately one character.

Figure 9.8 Romans 11:26–27 in P46 (f. 15v)

The use of a space just prior to quoted text is not unusual in P46, but neither is it common. In every instance where an intralinear space is used prior to a quotation, it follows a specific attribution or introductory formula that marks the quotation (cf. Rom 11:8; 1 Cor 2:9; 15:54–55; Gal 3:13; Heb 10:37–38; 13:5; cf. John 12:40 in P75). As Kenyon noted, the use of spacing to mark a sense unit “suggest[s] … some perception by the scribe of the sense of what he was writing.”28 A particularly fascinating example of intralinear spacing appears within the composite quotation of Heb 10:37–38 in P13 (ca. 3rd or 4th cent.), a passage that quotes Hab 2:3–4 with a conflation from Isa 26:20. The opening of the quotation at Heb 10:37 is preceded by a space, but the wording that begins Heb 10:38 (ὁ δὲ δίκαιός μου) is also demarcated in this way, as may be seen in Figure 9.9.

Figure 9.9 Hebrews 10:37–38 in P13 (f. 64v)

28 F. G. Kenyon, ed., Supplement: Pauline Epistles, Text. Vol. 3 of The Chester Beatty Biblical Papyri (London: Emery Walker, 1936), xiv. Discussed also by James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill, 2007), 207.

236

Adams and Ehorn

Before we comment on the spacing, it is important to note the use of dicola in P13 (cf. Matt 5:21 in P64). According to Grenfell and Hunt, these correspond to the stichoi of its exemplar.29 The use of dicola corresponds to the spacing in Heb 10:37–38, but a comparison with the rest of the manuscript does not demonstrate a normal pattern of pairing both features together. This suggests that we should treat the spacing here as an independent feature. Having observed this, it is noteworthy that Heb 10:37b–38 includes continuous wording from Hab 2:3–4. So, whatever break was intended by the scribe does not corresponded to the textual conflation of wording from Isa 26:20 that appears at the opening of Heb 10:37.30 Given the paucity of data, it is difficult to know why the spacing appears at Heb 10:38, which separates Hab 2:3 from 2:4. One possibility is that the copyist understood the source as deriving from (or belong to) Pauline tradition (cf. Rom 1:17; Gal 3:11) or perhaps because it belonged to a “traditional testimonia from the earliest period.”31 This theory implies that Hab 2:3 was less well known and so was not recognized as being part of the quotation with 2:4. Such an explanation is plausible, especially given the lack of explicit markers indicating a citation within the text. 3.2 Punctuation Marks Punctuation marks are used infrequently in early Greek papyri, although the high point (στιγμὴ τελεία) to mark sense units is not uncommon.32 In the combined citation at John 12:13 (cf. Ps 118[117]:25–26; Zeph 3:15; Isa 44:6) in P66 (Figure 9.10), the word ὡσαννά is followed by a middle point (στιγμὴ μέση) and the remainder of the citation is kept as a singular unit. At the end of the citation a dicolon appears in the text and the rest of the line is left empty. This is followed by ekthesis of the next word, suggesting a major sense unit division.

Figure 9.10

John 12:13 in P66 (f. 87)

29 Noted by B. P. Grenfell and A. S. Hunt, The Oxyrhynchus Papyri, Part IV (London: Egypt Exploration Fund, 1904), 37. 30 On the citation itself, see Gert J. Steyn, A Quest for the Assumed LXX Vorlage of the Explicit Quotations in Hebrews, FRLANT 235 (Göttingen: Vandenhoeck & Ruprecht, 2011), 310–325. 31 C. H. Dodd, According to the Scriptures: The Substructure of New Testament Theology (London: Nisbet, 1952), 49–51. 32 See brief discussion in Hurtado, Earliest Christian Artifacts, 182.

Composite Citations in New Testament Greek Manuscripts

237

In the quotation at John 12:40 (cf. Isa 6:10; 42:18–20; 44:18) in P66, a high point precedes each use of καί within the quotation. This feature is not used consistently elsewhere in the manuscript, suggesting that the scribe subdivided the quotation into a perceived sense unit. But this scribal practice is not associated with the recognition of multiple source texts, as the pairing of καί and a high point to mark a sense unit is evidenced in other non-composite citations (e.g., John 12:38 which quotes Isa 53:1). Similar punctuation occurs for sense units in P74, including high points that appear before and after a quotation (Acts 13:22), high points following a quotation (Acts 3:13; 7:5), and high points marking sense units within a quotation (Acts 7:6–7; 15:16–18). Another paratextual feature added to Greek manuscripts, particularly P46, is the use of an upward stroke similar to an acute accent. Their irregular placement makes them difficult to describe in a uniform manner, but they sometimes appear over the source attribution or introductory words of a quotation (e.g., “Hosea” in Rom 9:25; “Isaiah” in 9:27; citation formula in Rom 10:6). In other examples, the stroke appears as an annotation just prior to (e.g., Rom 13:9; 14:21; Heb 7:1; 11:21) or just after a quotation (e.g., Rom 11:8; Heb 7:3; 12:21). As an example of this, Figure 9.11 shows where the stroke is placed over the omega in the name Ὡσηέ.

Figure 9.11

Romans 9:25 in P46 (f. 13v)

Given that sometimes the stroke is used to mark the source attribution within the text, it seems reasonable to conclude that the reader(s) who added these sometimes intended to mark a quotation for his or her convenience. The important word here is sometimes because the practice was not carried out consistently. Similar features appear at Heb 10:37–38 in P46, where two upward strokes were added to mark a sense unit within the quotation.

Figure 9.12

Hebrews 10:37–38 in P46 (32v)

238

Adams and Ehorn

As with the quotation in Heb 10:37–38 from P13 (see Figure 9.9), a sense unit division appears in the middle of the quotation from Hab 2:3–4, marking out the wording ὁ δὲ δίκαιός μου ἐκ πίστεως ζήσεται from Hab 2:4. Unlike P13, this division appears in the form of punctuation that was added later by a reader rather than being incorporated by the scribe at the time of copying. Finally, an interesting punctuation mark appears in P72 just prior to the conflated quotation of Isa 53:4 and 53:12 in 1 Pet 2:24.33 Two high points appear in an intralinear space. Perhaps these points were added by the scribe at the time of copying, but it is also possible that these were added to a pre-existing space left by the scribe. In any case, the spacing and/or punctuation demarcate a sense unit at the beginning of the quotation. 3.3 Summary The earliest New Testament papyri provide limited examples of paratextual features to mark quotations. The most common feature was the division of the text into sense units using spaces, although different types of punctuation were also used, often in conjunction. Other paratextual features mentioned above, such as indentation, rubrication, and marginal attributions, are absent in New Testament papyri, potentially indicating a limited set of signaling tools available to the scribe. Different paratextual features potentially represent different stages of text production and reading. Spacing must have happened at the time of copying, either reflecting the scribe’s exemplar or, alternatively, demonstrating the scribe’s inclusion of a textual division in or around a quotation. Conversely, other forms of punctuation (e.g., high points) could be added to the text later. Both paratextual features – those added at the time of copying and those added by later readers – provide insights into how the text was understood. The most basic conclusion we can make about composite citations in the papyri is that they were treated very similarly to other quotations. Perhaps the most interesting example from the papyri comes from P13, where Heb 10:37–38 is divided with a sense unit in the middle of the quotation. This may provide evidence of knowledge of a source within the quotation. There is other evidence that readers perceived a sense unit at the same place, notably the added stroke marks in P46 at this exact location. But this division does not align with

33

Whether this is a quotation or an allusion is a complex question. For present purposes, we can note that many early Christian writers viewed Isaiah 53 as important, increasing the likelihood of recognition of the text here. See discussion in Adams and Ehorn, “Introduction,” in Composite Citations in Antiquity: Vol. 2, 13–14.

Composite Citations in New Testament Greek Manuscripts

239

the multiple sources that actually comprise the composite quotation (i.e., Hab 2:3–4; Isa 26:20). Rather, it is better taken to represent limited knowledge of Habakkuk by the scribes with possible influence from the Pauline tradition. That is, here we may have some evidence that the scribe (thought he) understood the boundaries of a quotation. However, even here we have no basis for claiming that a scribe or reader perceived that this text was a composite citation. 4

Early Greek Codices

Having looked at the composite citations in New Testament papyri, we turn our attention to several of the earliest Greek codices: Sinaiticus, Vaticanus, Alexandrinus, Bezae, and Claromontanus.34 Their relative similarity in date provides us with an important perspective of how certain scribal practices were enacted during the fourth–sixth centuries and how scribal habits might have developed. The previous section was structured by paratextual feature rather than by individual manuscript. In this section, we move away from identifying specific paratextual features to looking at how these features are deployed in individual manuscripts and for what purpose. One of the benefits of going through whole codices is that we have the opportunity to evaluate a fuller range of quotations, which in turn allows us to determine with greater accuracy the scribe(s)’s awareness and handling of composite citations. 4.1 Codex Sinaiticus Codex Sinaiticus is a fourth-century Greek uncial manuscript that originally contained both the Septuagint and the New Testament.35 The New Testament portion, most of which is extant, is thought to have two to three scribes and

34 Codex Ephraemi Rescriptus was not included in this study because reading the text is problematic. Unfortunately, Tischendorf, although mentioning them, did not faithfully reproduce the diplai in his reproduction. Of relevance for this discussion is the use of two types of diplai, the traditional chevron (>) as well as the inverse N (И). For a brief discussion of this codex, along with a list of diple in the manuscript, see Ulrich Schmid, “Diplés im Codex Ephraemi rescriptus: Eine Problemanzeige,” in Von der Septuaginta zum Neuen Testament, 145–147. Cf. Constantin von Tischendorf, Codex Ephraemi Syri rescriptus, sive Fragmenta Novi Testamenti e codice Graeco Parisiensi celeberrimo quinti ut videtur post Christum seculi (Leipzig, 1843), §8 (pp. 10–11). 35 For a critical introduction to Sinaiticus, see Jongkind, Scribal Habits.

240

Adams and Ehorn

correctors.36 Regarding paratextual signs and manuscript layout for the citations of Scripture, the scribes did not have a consistent practice.37 Many Old Testament quotations are distinguished through the use of sense-unit divisions (e.g., Mark 14:62, f. 28r; Rom 9:25–26, f. 65v), although a number are also incorporated into larger paragraphs (e.g., Mark 12:10–11, f. 26r; Luke 4:18–19, f. 31v–32r). More visually helpful is the signaling of a citation through the use of diplai (e.g., Matt 27:9–10, f. 17r).38 This feature was not consistently applied and diplai are more likely found for quotations that are explicitly signaled in the text as well as those that are longer in size.39 The pairing of paragraphoi with diple is also not consistent, even within the books that regularly include both features (e.g., Matthew). An important paratextual feature in Sinaiticus, which is not found in many contemporary manuscripts, is the occasional use of source attribution to indicate where the quotation could be found in Scripture. This practice is limited to specific New Testament books (i.e., Matthew, Luke, Romans, Acts, and 1 Peter), is concentrated in the opening chapters, and is not consistently or thoroughly applied.40 In a majority of these instances, the scribe writes the source attribution in the left margin, although occasionally the scribe writes in the right

36 For a discussion of the scribes and their correctors, see H. J. M. Milne and T. C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938), 40–50; Jongkind, Scribal Habits, 9–18. 37 For example, source attribution and diple are well attested in the first few chapters of Matthew but diminish as the book progresses. 38 One of the major issues with the discussion of diple is determining whether or not they are from the hand of the scribe or later readers. For a detailed discussion of this, see K. McNamee, “Marginalia and Commentaries in Greek Literary Papyri” (PhD thesis, University of Durham, 1977), 404–443. It is widely thought that most diple and other paratextual features (such as book references) are part of the production of the text and so can be attributed to the scribe and/or his exemplar. 39 Scribes are not always accurate with their use of diplai. For example, Acts 3:22–23, which is a citation of Deut 18:15–16, 19 and Lev 23:29, is marked with diplai. However, Acts 3:24 is also marked with diple and it is not a quotation (f. 101v). In addition, Acts 26:23 is marked by diplai, but is not a quotation. Cf. Ulrich Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” in Von der Septuaginta zum Neuen Testament, 89. 40 Schmid concludes, we think rightly, that the inclusion of source attributions and most diplai were not added at the time of transcription, but soon after in the scriptorium. Cf. Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” 91. The concentration of the source references to the beginning of Matthew, Romans, and Acts corresponds to the beginning of different groupings of texts (i.e., Gospels, Paul, and Apostolic books) (so Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” 94) and corresponds well to the idea of division of labour in the creation of the manuscript (cf. Jongkind, Scribal Habits, 39–51).

Composite Citations in New Testament Greek Manuscripts

241

margin or incorporates the attributions into the text (e.g., Matt 13:35, which reads διὰ ’Ησαΐου τοῦ προφήτου against the majority reading, διὰ τοῦ προφήτου, f. 7v). Generally, these book attributions are correct, although there are some errors.41 Specific chapters are not provided (nor could they be); however, with the Psalms the scribe was able to specify which Psalm is being cited. More important for our discussion is how the scribe might have indicated multiple sources for specific quotations. One of the issues for this evaluation is that a scribe could indicate (correctly) the book from which a quotation was taken but might not be able to indicate that it came from multiple parts (i.e., two or more different passages) of that book. One example that allows for multiple attributions is the quotation in Matt 2:15. Although signaled in the text of Matthew as coming διὰ τοῦ προφήτου and so most likely from Hos 11:1, the scribe indicates that the quotation can be found εν αριθμοις (f. 1v), signaling its association with Numbers, most likely 23:22 and/or 24:8. The challenge here is determining whether or not the scribe knew and/or could have signaled that multiple texts, either in Numbers or another book, could be alluded to in Matt 2:15. Both options seem possible, although there is no example of either in Sinaiticus. As noted above, the exception to the constraint of multiple references from one book is Psalms. A good example of this distinction by the scribe of Sinaiticus is found at Rom 3:10–18, where the author signals that the quotation comes from Psalms 13 and 52: ψαλμω ΙΓ κ(αι) NB (f. 63r).42

Figure 9.13 Marginal Notation at Romans 3:10 in Codex Sinaiticus (f. 63r)

41 For example, the scribe attributes the quotation of Gen 22:18 and 26:4 in Acts 3:25 to Deuteronomy (f. 101v, δευτερ) and the quotation of Hab 1:5 at Acts 13:41 to Joel (f. 108v, ιωηλ). 42 The marginal reference at Rom 3:5 has a different pattern of source attribution and diple pairing, which might suggest a different scribe, although one using a similar, though not identical, method. Cf. Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” 87.

242

Adams and Ehorn

This is an important example for our study as Rom 3:10–18 has been identified as a composite quotation.43 In this instance, the composite citation of 3:15–18 is part of a larger paragraph (3:10–18), the entirety of which is marked with diplai. In this catena, a number of passages have been recently discussed by Mark Reasoner (i.e., in order, Eccl 7:20; Ps 13:1–3; Ps 5:10; Ps 139:4; Ps 10:7; Isa 59:7–8; Prov 1:16; and Ps 35:2);44 however, only one of these passages was noted by the scribe (i.e., Psalm 13). The other scribal attribution is to Psalm 52, which also has textual similarities and can understandably be seen as a possible source text.45 This multiple attribution indicates that the scribe of Sinaiticus and possibly others were close readers of Scripture and were sophisticated enough to make accurate source attributions.46 Another example of scribal awareness of composite citations is found in Matt 21:5. Here, the first part of the quotation, which is a phrase from Isa 62:11 (εἴπατε τῇ θυγατρὶ Σιών), is divided from the remainder of the quotation, which is taken from Zech 9:9, through the use of a large space at the fusion of the passages and diplai only for the last section (i.e., for Zechariah). This suggests a clear recognition of the limitations of the Zechariah material as well as the possibility that the introductory line might not be scriptural (or at least not part of the Zechariah passage). Another example of a subdivided quotation is found in Mark 1:2–3 (f. 18v). In this manuscript, the scribe created a paragraph break at end of 1:2 and united vv. 3 and 4 into one paragraph. This is an odd division, not only because a break between vv. 2 and 3 is grammatically undesirable, but also because 1:4 begins with ἐγένετο and this verb is regularly associated with the opening of sense-unit divisions, and so would not normally be found in the middle of a paragraph. Most importantly for this study is the fact that the seam between the quotation from Isaiah and Malachi occurs at the end of verse 2 and the beginning of v. 3. Could it be that the scribe recognized that the content of 1:3 43 See, for example, Mark Reasoner, ‘“Promised Beforehand Through His Prophets in the Holy Scriptures”: Composite Citations in Romans,’ in Composite Citations in Antiquity: Vol. 2, 129–136. 44 For a recent discussion, see Reasoner, “Promised Beforehand,” 129; Christopher D. Stanley, “Composite Citations: Retrospect and Prospect,” in Composite Citations in Antiquity: Vol. 1, 204. 45 E.g., Leander E. Keck, “The Function of Rom 3:10–18: Observations and Suggestions,” in God’s Christ and His People: Studies in Honour of Nils Alstrup Dahl, ed. Jacob Jervell and Wayne A. Meeks (Oslo: Universitetsforlaget, 1977), 144. 46 Such attributions, however, were not always accurate. For example, at Matt 2:6 the scribe correctly uses diplai to note a quotation from Scripture. However, the scribe ascribes the text to Isaiah (f. 1v), although the passage is actually a combination of Mic 5:1–3 and 2 Sam 5:2. See Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” 84–85.

Composite Citations in New Testament Greek Manuscripts

Figure 9.14

243

Matthew 21:5 in Codex Sinaiticus (f. 12v)

was not part of Isaiah? As there are no diplai or other paratextual features in Sinaiticus it is difficult to come to a firm conclusion.47 Nevertheless, in light of the examples from other manuscripts (e.g., GA 021, which not only has diplai for 1:2 and not 1:3, but also has a paragraph break at 1:3) this is an important piece of evidence for understanding the use of textual features by scribes to shape reading practices.48

47

Schmid (“Diplés und Quellenangaben im Codex Sinaiticus,” 96–97) raised the possibility that the standard reading in Mark 1:2 (ἐν τῷ Ἠσαΐᾳ τῷ προφήτη) might be secondary based on the growing tradition of adding source references. Although possible, we would suggest that the reading of prophets in the plural (προφηταις) is more likely secondary based on the tradition of identifying and disambiguating sources. 48 A similar example is Rom 11:26–27, in which a later scribe/reader inserted high dots that break the combined citation into its parts (f. 66v). This implies that the original scribe and his exemplar did not recognize the different components of the quotation or that they did not think it warranted signalling. There are two other examples of divided quotations in Sinaiticus. In Rom 9:27–28 (f. 65v) the scribe is aware that the Pauline quotation comes from Isa 10:22 and adds ἐν δικαιοσύνῃ ὅτι λόγον συντετμημένον, a phrase that brings the quotation into greater alignment with the source text. This phrase is not integrated into the quotation but added through a marginal note. This scribal activity does not impact the composite citation and does not indicate any awareness of combined quotations in 9:27–28. What is worth noting here is that a similar insertion is found in Claromontanus, but in that manuscript the addition is fully integrated into the text (see below). A similar example is found in Luke 20:28 (f. 44r), in which the scribe adds κ(αι) ουτοϲ ατεκνοϲ η ϊνα λαβη ο αδελφοϲ αυτου

244

Adams and Ehorn

4.2 Codex Vaticanus Codex Vaticanus is a fourth-century Greek majuscule. Originally containing the whole of the New Testament, Vaticanus now lacks the Pastorals, Philemon, and Revelation. Also, after Heb 9:13 the document is written in a much later, minuscule hand. The text is of good quality and the scribe who copied it showed strong awareness of text-critical issues.49 The primary method of signaling quoted text in Vaticanus is the use of diplai, and although not all quotations are signaled, there is a consistency in Vaticanus that is not found in other manuscripts.50 For example, passages that quote Scripture within speech (e.g., Matt 5:21; 11:5; Mark 14:62; Luke 20:28) regularly lack diplai. Passages in which the narrator or author quotes scripture (e.g., Matt 2:6; Rom 3:10–18) are more likely to have diplai, although, again, this is not consistently applied. When diplai are added to the text they appear in the left-hand margin and mark the lines in which the quoted text can be found. On a couple of occasions, the scribe marks a line with a diple that is not a quotation from Scripture. For example, the scribe rightly places diplai for John 12:38 and 12:40, as both of them include citations from Scripture. However, the scribe also marks John 12:39 with diplai, which does not contain cited text. This gives the false impression that 12:38–40 is one large quotation by visually joining these two citations. Similarly, in John 19:36–37 the intermediate introductory line of 19:37 (καὶ πάλιν ἑτέρα γραφὴ λέγει) is also given a diple. The inclusion of introductory την γυναικα, showing clear understanding of the meaning behind the passage (most likely Deut 25:5) but not signaling a specific source text through verbatim citation. 49 For the scribe’s awareness of textual issues (signalled by an umlaut/“marginal colon”/distigmai) and how this corresponds to their reading practices, see Philip Barton Payne and Paul Canart, “The Originality of Text-Critical Symbols in Codex Vaticanus,” NovT 42 (2000): 105–113. Cf. Edward D. Gravely, “The Text Critical Sigla in Codex Vaticanus” (PhD dissertation, Southeastern Baptist Theological Seminary, 2009). 50 One of the major issues when dealing with Vaticanus is that the text has been retraced and so the original ink is obscure. However, the diplai were not retraced. This is helpful for this study as it places their insertion prior to the refreshing of the text. The issue here is that the diplai are sometimes very faded and so are difficult to identity and assign to specific scribes. For the most recent studies of this phenomenon, see Ulrich Schmid, “Diplés im Codex Vaticanus,” in Von der Septuaginta zum Neuen Testament, 99–113; A. Stokowski, “Diplé-Auszeichnungen im Codex Vaticanus Graecus 1209 (B): Liste nebst einigen Beobachtungen,” in Textual History and the Reception of Scripture in Early Christianity / Textgeschichte und Schriftrezeption im frühen Christentum, 93–113 (who posits multiple scribes who inserted diplai, p. 96); C. E. Hill, The First Chapters: Dividing the Text of Scripture in Codex Vaticanus and its Predecessors (Oxford: Oxford University Press, 2022), 9–10; Jesse R. Grenz, “The Scribes and Correctors of Codex Vaticanus: A Study on the Codicology, Paleography, and Text of B(03),” TynBul 73 (2022): i–iv.

Composite Citations in New Testament Greek Manuscripts

Figure 9.15

245

John 12:39 in Codex Vaticanus (p. 1369)

phrases under the diplai in Vaticanus is not unique to this verse, but it does make John 19:36–37 look like one quotation.

Figure 9.16

John 19:36–37 in Codex Vaticanus (p. 1379)

These two examples do not address the issue of how scribes marked composite citations. However, in their own way these scribes are uniting disparate citations and presenting them as though they were one. This has the potential to influence the reader and shape how they understand these now-united citations. There are a few instances in Vaticanus where the scribe appears to show some awareness of a composite citation by attempting to distinguish the fused texts.51 One example occurs in the citation at Acts 7:6–7, which is a composite quotation of Exod 2:22, 3:12, and Gen 15:13–14. Here the last half of Acts 7:7 (καὶ μετὰ ταῦτα ἐξελεύσονται καὶ λατρεύσουσίν μοι ἐν τῷ τόπῳ τούτῳ) is not marked with diplai, which stop following ὁ θεὸς εἶπεν, a phrase that might have signaled to the scribe that the quotation was finished. However, this action excludes 51

One marginal example is found at 1 Cor 15:54–55, where the scribe and/or his exemplar place a high dot between 1 Cor 15:54 and 15:55. This use of punctuation suggests a break within the text, although it might not necessarily be read as distinguishing the composite citation.

246

Adams and Ehorn

Figure 9.17

Acts 7:6–7 in Codex Vaticanus (p. 1390)

the subsequent material from Exod 3:12, thus visually dividing the composite quotation. This action might suggest an awareness of the source texts but could also be an expression of scribal expectations of how texts from Jewish Scripture are cited. Another example of scribal use of diplai to signal a specific interpretation is found in Matt 21:5. Here the author makes an interesting use of this paratextual feature by placing under the diple the opening line (εἴπατε τῇ θυγατρὶ Σιών), which is taken from Isa 62:11, but does not mark with a diple the final line of the quotation (καὶ ἐπὶ πῶλον υἱὸν ὑποζυγίου), which is a continuation of Zech 9:9. This use of diplai in Matt 21:5 is opposite to Sinaiticus above, which showed potential awareness of the differences between the two cited passages. In contrast, the scribe of Vaticanus does not display knowledge that the quotation is composite.52 52

This does not explain why the last line of Zech 9:9 was not given a diple. It could indicate that the scribe did not have a good knowledge of the source text. Another explanation is that the scribe accidentally omitted the final diple. The latter position is supported by the

Composite Citations in New Testament Greek Manuscripts

Figure 9.18

247

Matthew 21:5 in Codex Vaticanus (p. 1262)

Although the above examples do not indicate scribal recognition of composite citations, it is clear from other passages in Vaticanus that the scribe did have knowledge of the original context of the citation. For instance, in Rom 14:11, a text that is marked by diplai, the scribe switched the Pauline word order to better conform to its source text, Isa 45:23: καὶ πᾶσα γλῶσσα ἐξομολογήσεται to ἐξομολογήσεται πᾶσα γλῶσσα. Another example of scribal awareness of Scripture has been proposed by Stokowski, who notes that 2 Cor 6:16–18 has two diplai (one at the opening of 6:16 and the other and the end of 6:18).53 One way to understand the evidence is that one diple marks the beginning of the quotation and another diple marks the end (so Stokowski). Alternatively, each diple may mark (in the mind of the scribe) a new quotation. A survey of the scribal practice must help us discern between these options. The first diple does not correspond to any quoted text but marks the introductory sentence in 2 Cor 6:16 (καθὼς εἶπεν ὁ θεὸς ὅτι). The second diple marks the final line of 6:18 (λέγει κύριος παντοκράτωρ), which is the end of the quoted text. The use of a singular diple to mark the opening of a quotation is found in other parts of Vaticanus (e.g., Hebrews 1–4), thus providing some support for the first part of Stokowski’s theory.54 The issue, however, is that Stokowski does not provide evidence, either from Vaticanus or elsewhere, that scribes fact that the sense unit concludes at the end of Matt 21:5, which is also the conclusion of the quotation. 53 Stokowski, “Diplé-Auszeichnungen im Codex Vaticanus Graecus 1209 (B),” 109. Interestingly, Schmid (“Diplés im Codex Vaticanus,” 107) has an independent entry for 2 Cor 6:17, although he views it as part of the larger citation of 6:16–18. 54 In P.Oxy. III 405, a second-century fragment of Irenaeus’s Adversus Haereses, the scribe also uses a single diple to mark the opening line of quotations from Matthew.

248

Figure 9.19

Adams and Ehorn

2 Corinthians 6:16 in Codex Vaticanus (p. 1481)

marked the end of a quotation by a single diple. There is no strong evidence as to why the scribe changed his practice. Schmidt has tentatively suggested that the length of quotation might have been a determining factor for the change in behavior.55 However, this did not appear to be an issue for the longer quotation at Rom 3:10–18, all of which has diplai. In contrast, it is also plausible that the scribe viewed the phrase λέγει κύριος παντοκράτωρ as the indication of a new quotation, even though this would be in error as Paul resumes his argument through the use of the vocative (ἀγαπητοί; 2 Cor 7:1). This interpretation would imply that the scribe marked the quotation but did so without engaging with the text’s contents. What is problematic to both theories is the omission of a diple for the formula in 6:17 (λέγει κύριος), which would also likely signal a new quotation. Overall, Stokowski’s theory is 55 Schmid, “Diplés im Codex Vaticanus,” 110.

Composite Citations in New Testament Greek Manuscripts

249

plausible, although it does not fully explain all of the evidence.56 The implication of his argument is that the scribe had a high awareness of the text’s contents and that scribal practices were not consistent. Overall, Codex Vaticanus provides some examples of scribal awareness of composite citations and the willingness to use sense-unit divisions and diplai to signal how a reader should understand and identify scriptural quotations. Once again, the marking of composite citations by the scribe who inserted the diplai in Vaticanus does not differ from the regular practice of marking citations from singular sources, but he does show a strong knowledge of source texts. 4.3 Codex Alexandrinus Codex Alexandrinus is a fifth-century Greek uncial manuscript of both the Septuagint and the New Testament. Similar to the codices above, when the scribe of Alexandrinus isolated cited texts, he used sense-unit divisions and, less frequently, diplai to signal visually to the reader the occurrence of a scriptural quotation. These diplai are unlikely to have been copied by the original scribe but were inserted by at least three different hands, thus showing how the text was read in late antiquity.57 In Alexandrinus we find that there is little differentiation between regular and composite citations; both are often (but not always) distinguished by sense-unit divisions (i.e., interlinear spacing and ekthesis). Similarly, some composite citations (e.g., Luke 4:18–19, f. 47r; John 12:40, f. 76r) are also marked with diplai. The inconsistent employment of diplai for composite citations is parallel to the variable use of this paratextual feature for “normal” citations, even when they are in close proximity (e.g., Romans 9, ff. 113v–114r; 11:8–10, f. 115v; Gal 3:8, 10, f. 128v). However, there are a few passages in Alexandrinus that indicate an awareness of the composite nature of quotations along with the need to structure the text and/or signal to the reader that multiple passages are being quoted. The first example of this practice is Mark 1:2, in which the scribe (or his exemplar) changed the singular προφήτῃ to the plural προφηταιϲ, indicating that the quotation derived from multiple prophetic texts. This recognition and change

56 57

One question for future consideration would be how scribes marked quotations when the source referent is at the end of the quotation and/or subsequent to the cited text. For a discussion of the different diplai markings and their use in Alexandrinus, see Marcus Sigismund, “Formen und Verwendung der Diplé im Codex Alexandrinus,” in Von der Septuaginta zum Neuen Testament, 117–22, 142. Cf. W. Andrew Smith, A Study of the Gospels in Codex Alexandrinus: Codicology, Palaeography, and Scribal Habits, NTTSD 48 (Leiden: Brill, 2014), 207–210.

250

Adams and Ehorn

is found in a number of manuscripts and so is not unique to Alexandrinus.58 However, in addition to the pluralizing of prophets, the scribe also divided the citation into two paragraphs 1:2 and 1:3 through the use of ekthesis and interlinear space in previous line. The division implies not only a close reading of the Markan text and an awareness of the Old Testament source passages by the scribe, it also displays a desire to mark visually the passages for the benefit of subsequent readers.

Figure 9.20

Mark 1:2–3 in Codex Alexandrinus (f. 30r)

Similar divisions of composite citations are also found in Acts 3:22–23 and Rom 9:25–26. In the former, the composite citation of Deut 18:15–16, 19, and Lev 23:29 is broken into two parts; 3:22 is a separate paragraph, marked by ekthesis, interlinear space, and diplai, whereas 3:23 starts a new paragraph (3:23–25), which is also marked by ekthesis and an interlinear space. Here, however, the scribe only provides diplai for 3:22 and 3:25 and not 3:23. This arrangement creates a visual break in the middle of the quotation, creating a cognitive seam at 3:23 where the material from Lev 23:29 appears. Furthermore, this action associates the material in 3:23 more closely with the citation of Gen 22:18 in Acts 3:25. Similarly, the composite citation of Hos 2:25 and 2:1 in Rom 9:25–26 is part of larger paragraph (9:18–26). However, 9:25 is marked by diplai, but 9:26 is not. This selection by the author is interesting as the material in 9:26 is essentially verbatim to Hos 2:1, whereas the quotation in 9:25 is less closely aligned with Hos 2:25.59 The introductory formula clearly signaled to the scribe a quotation; 58 59

This change in found in A K P W Π f13 28 1009 1010 1079 1195 1216 1230 1242 1253 1344 1365 1546 1646 2148 Byz Lectm syrh copboms mg arm eth. There is variation in the Greek text of Hos 2:25, with the version found in codex Vaticanus close to that cited by Paul. Cf. Joseph Ziegler, ed., Septuaginta: Vetus Testamentum

Composite Citations in New Testament Greek Manuscripts

Figure 9.21

251

Acts 3:22–23 in Codex Alexandrinus (f. 83v)

however, it is somewhat surprising that the scribe stopped marking the citation at 9:26. This suggests that the scribe viewed the material in 9:26 as a commentary on the quotation and not part of the quotation itself. Such an action likely implies that the author did not recognize the composite quotation (including the citation of Hosea) and divided it.

Figure 9.22

Romans 9:25–26 in Codex Alexandrinus (f. 115r)

These divisions of citations either suggest that the scribe (or his exemplar) did not recognize the quotation as a tightly knit unit and/or only saw the first half as being the citation and not the latter part. This not only provides insight into early reading practices but also suggests that composite citations were not always recognizable to ancient readers.

Graecum. Vol. XIII, Duodecim Prophetae, 3rd ed. (Göttingen: Vandenhoeck & Ruprecht, 1984), 152.

252

Adams and Ehorn

4.4 Codex Bezae Codex Bezae is a fifth-century uncial, Greek-Latin diglot manuscript containing the Gospels and Acts. The codex is the work of one scribe, but shows evidence of a number of different hands.60 Both the Greek and Latin New Testament texts are divided colometrically, and so interlinear spaces are infrequent, especially in the earlier folios.61 However, as the work continues and colons are no longer given their own line, the text is more often divided by an interlinear space, which occasionally is paired with a colon “:” and dieresis above the following letter (even if it is a consonant and not a potential diphthong; see below) to signal verse divisions.62

Figure 9.23

Dieresis use in John 12:41 in Codex Bezae (f. 154)

The use of colometric divisions and the less frequent use of interlinear spacing might suggest that scribes would use more paratextual features to signal elements within the text. This, however, is not the case. There is essentially no use of diplai or other paratextual indicators to signal citations of Scripture. One possible exception is the inclusion of a dot in left margin to mark the citation of Zech 9:9 and Zeph 3:16 in John 12:15.

Figure 9.24

John 12:15 in Codex Bezae (f. 152)

The more common way to signal cited texts is through the use of indentation (e.g., Matt 21:5; 27:9–10), although this only occurs in the last third of Matthew, the beginning of Mark, and the first section of Acts.63 An interesting example of this is Mark 1:2–3 (see Fig. 9.25).

60 E.g. D. C. Parker, Codex Bezae: An Early Christian Manuscript and its Text (Cambridge: Cambridge University Press, 1992), 7–30, 35–49. 61 In different sections of the manuscript the colons are placed on the same line, but distinguished by a raised dot, which often looks like an apostrophe (e.g., f. 174). For more on sense-lines, see Parker, Codex Bezae, 73–75, 78–81. 62 F. H. Scrivener, Bezae Codex Cantabrigiensis (Cambridge: Deighton, Bell, 1864), xix–xx. 63 See E. A. Lowe, “More Facts about Our Oldest Latin Manuscripts,” CQ 22 (1928): 61; Parker, Codex Bezae, 31–34.

Composite Citations in New Testament Greek Manuscripts

Figure 9.25

253

Mark 1:2–3 in Codex Bezae (f. 286)

Here the author follows the practice of writing the first three lines of the biblical book in red ink and switches to black ink for the remainder. More important for our focus is the division between vv. 2 and 3. There is evidence of a raised dot from the original scribe (this is faint and located between the two dots of the colon). However, a subsequent reader/scribe has added a colon and diaresis to mark the new verse, but this is not original (there is also a lack of interlinear space). As this is a consistent feature of verse division in the codex it is unlikely that the inclusion of the colon and dieresis is meaningful for our discussion of composite citations, even though it does fall right at the seam of two quoted verses. Overall, the scribe of Bezae makes minimal use of paratextual features to signal quoted material. In addition, there is essentially no indication that the scribe or subsequent readers of Bezae noted composite citations, or if they did they refrained from signaling them in any way. 4.5 Codex Claromontanus (D06) Our final codex to be evaluated is Codex Claromontanus, which is a sixthcentury diglot of the Pauline Epistles written in majuscule script on parchment. Sense units in both the Greek and Latin texts of Claromontanus are marked by ekthesis and an oversized initial letter. Similar to other manuscripts, the scribe of Claromontanus regularly created a sense-unit division immediately following a scriptural quotation (e.g., Romans 10–11), though this practice fades towards the end of Romans.64 Unlike the other codices discussed above, the scribe of Claromontanus did not employ diplai for marking cited text. Rather, quotations from Scripture are written in red ink and are regularly indented 64 Cf. Sean A. Adams, “Sense Units and Manuscript Families: A Test Proposal,” in The Language and Literature of the New Testament: Essays in Honor of Stanley E. Porter’s 60th Birthday, ed. Lois K. Fuller Dow, Craig A. Evans, and Andrew W. Pitts, BIS 150 (Leiden: Brill, 2017), 179–180.

254

Adams and Ehorn

from the left margin (e.g., Rom 3:10–18, pp. 16–17; 9:9, p. 49; 9:20, p. 51; 9:25–26, pp. 52–53; although quotations in Hebrews are not in red ink, e.g., p. 480).65 In Claromontanus (perhaps reproducing its Vorlage) we see a scribal tendency to bring the quoted text into conformity with its source.66 For example, in Rom 9:33 (p. 54) the scribe changed the ending of the quotation from οὐ καταισχυνθήσεται to οὐ μὴ καταισχυνθῇ to match the wording of Isa 28:16. Similarly, in Rom 14:11 (p. 77) the word order was changed from καὶ πᾶσα γλῶσσα ἐξομολογήσεται τῷ θεῷ to καὶ ἐξομολογήσεται πᾶσα γλῶσσα τῷ θεῷ to align with the source text (Isa 45:23). Such examples indicate that sometimes the scribe made an effort to find the cited text and was willing to adjust the Pauline text to bring it into conformity with his version of the Septuagint.67 A particularly interesting example of this is found in Rom 9:27–28. Here the scribe divides the quotation in the middle of 9:28 through the insertion of ἐν δικαιοσύνῃ in black ink, which is taken from the source text of Isa 10:22. However, following this addition he adds ὅτι λόγον συντετμημένον, also from Isa 10:22, but this time in red ink before finishing the Romans quotation, ποιήσει κύριος ἐπὶ τῆς γῆς. The same addition and alternation between colors is also found in the Latin, which supports the view that this was an intentional adjustment of the Pauline text by the scribe to correspond better with Isaiah 10. This addition is also found in Sinaiticus (see above), although there the addition appeared in the margin rather than being integrated into the text. The scribe of Claromontanus does not omit any aspect of Romans, and there is no indication that the scribe recognized the Hosean elements of the composite citation, even though it is clear that at some point a scribe or reader explicitly checked the source of the main quotation and made additions to the manuscript. Despite this lack of awareness, the use of black ink in the middle of the quotation signals to the reader that text contains two separate quotations and not one large one. This disrupts the unity of the quotation and has the potential to shape how the text is read by later readers. 65

66 67

For a discussion of indentation of quotations in ancient manuscripts, see Patrick McGurk, “Citation Marks in Early Latin Manuscripts (with a list of citation marks in manuscripts earlier than AD 800 in English and Irish libraries),” Scriptorium 15 (1961): 6. For a discussion of Hebrews, see E. A. Lowe, “Some Facts about our Oldest Latin Manuscripts,” CQ 19 (1925): 204. Some changes, such as those in 2 Cor 6:16 (p. 215), do not match a specific known version of a Greek biblical text. This practice was not consistently enacted, nor is there always explicit awareness of the scriptural text. For instance, in Gal 3:10 (p. 268) the scribe inserted ἐν to mirror Deut 27:26 but did not change the rest of the verse in order to bring it in line with Deut 27:26 (e.g., did not add λόγοις or τούτου). In addition, elements foreign to Deut 30:10 that were added by Paul were allowed to remain.

Composite Citations in New Testament Greek Manuscripts

Figure 9.26

255

Romans 9:27–28 in Codex Claromontanus (p. 53)

The awareness of the source text occasionally results in the deconstruction of composite citations. One instance of this is found in Rom 11:26–27 (p. 65) in which the scribe adds καί to align the quotation with its source text (Isa 59:20–21). However, the final section of the quotation (ὅταν ἀφέλωμαι τὰς ἁμαρτίας αὐτῶν) that was included from Isa 27:9a is written in black ink and not indented and so is not presented as part of the quotation.

Figure 9.27

Romans 11:26–27 in Codex Claromontanus (p. 65)

Similarly, in 1 Cor 2:9 (p. 100) the scribe only wrote the first part of the composite citation in red (Isa 65:12), but copied the material from Isa 64:3, including aspects that do not have a clear scriptural referent (e.g., ἃ ἡτοίμασεν ὁ θεὸς τοῖς ἀγαπῶσιν αὐτόν) in black ink and without indentation. Hebrews 10:37–38 (p. 514) provides a third example of scribal interpretation of citations. Not only did the scribe omit part of the quotation (καὶ ἐὰν ὑποστείληται) that is part of the source passage (Hab 2:3–4), but he also put the subsequent portion of Habakkuk flush against the margin. These three passages provide insight into the reading practices of the scribe and/or his Vorlage. The first two indicate a clear awareness of the contents of

256

Adams and Ehorn

Figure 9.28

1 Corinthians 2:9 in Codex Claromontanus (p. 100)

Figure 9.29

Hebrews 10:37–38 in Codex Claromontanus (p. 514)

specific passages of Isaiah by the scribe but surprisingly minimize the potential contribution of secondary Isaianic passages. In contrast, the changes in the third example do not indicate a strong awareness of Habakkuk. All of these examples show effort by scribes to disentangle New Testament quotations and provide a window into their interpretation(s) of the text. 4.6 Summary The first preliminary conclusion to draw from this investigation is that most New Testament scribes in the fourth–sixth centuries display some method of signaling cited text. Many scribes, such as those of Sinaiticus, Vaticanus, Alexandrinus, and Ephraemi used diplai, but others, such as Bezae and Claromontanus, used indentation.68 A few manuscripts were distinctive in their practices, namely, Codex Sinaiticus’s marginal source attributions and Claromontanus’s red ink. For the practices of indentation and diplai, the scribe

68

The multiple types of diple in Alexandrinus and Ephraemi indicate that multiple scribes and possibly later readers recognised this convention and that this practice continued over time.

Composite Citations in New Testament Greek Manuscripts

257

regularly marked the whole of the citation and did so accurately (although see some exceptions above). Second, none of the scribes evaluated are consistent in their practice. This is not to say that these scribes changed the means by which they mark citations (e.g., diple, indentation, ink color, etc.). Rather, these scribal features are not applied to all occurrences of cited texts. For some instances, such as paraphrases or quotations (or near quotations) in speech, it might be that the scribe did not recognize or view the passage as a quotation.69 However, there are a number of other instances in which a cited text could have been marked but was not. The inclusion of an explicit citation formula within the text (e.g., γέγραπται, καθὼς γέγραπται, etc.) increased the likeliness that a scribe would employ paratextual or other visual features to mark a cited text.70 In addition, there is no gradation of citation signaling within the text, such as to indicate an allusion rather than a citation or a more secure citation. Similarly, there is no difference in scribal habits for signaling normal or composite citations. Third, a number of the examples above indicate that the scribe was aware of the scriptural source of at least some citations, most often those from Isaiah and the Psalms. This is not to imply that knowledge of the source text/book was held throughout, as a majority of citations include differences from the original. For Sinaiticus, the origin was explicitly indicated by paratextual source attributions in 20 instances.71 For Sinaiticus and other manuscripts, this knowledge was evidenced through changes in the New Testament texts to conform to known versions of Old Testament books. This recognition of the origin of quoted material leads to a fourth finding: some scribes recognized foreign material in certain composite citations and took action to separate it from the “true” quotation. In a number of cases this partition was accomplished through the use of sense-unit divisions, which physically separated the two texts (e.g., Mark 1:2–3). In other instances, the scribe used the absence of markers (e.g., diple or red ink), which, in effect, indicates to the reader that this line is not cited Scripture. This directly impacts how the text would be read. Both of these scribal actions provide evidence of reading practices and interpretations found within early Christian communities. 69

Although outside the purview of this chapter, it would be interesting to see if there is a minimum set of criteria for a text to be marked paratextually as a quotation by a scribe. Similarly, are there elements (e.g., inserted words, paraphrase, citation in speech) that, if included, deter a scribe from marking the passage as a citation? 70 This collocation was also identified by Marcus Sigismund, “Die Diple als Zitatmarkierung in den ‘großen’ Unzialcodices: Versuch eines Fazits,” in Von der Septuaginta zum Neuen Testament, 149. 71 Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” 85–87.

258 5

Adams and Ehorn

Composite Citations in Later New Testament Codices

Although not specifically within the central purview of this volume, later New Testament codices show how certain paratextual features developed and stabilized over time. Because this volume’s primary focus is on earlier Greek manuscripts, we will consider only one later codex: Codex Boernerianus. The reason for selecting this manuscript is that it includes a range of paratextual features, especially the blossoming of marking and annotating quotations. 5.1 Codex Boernerianus (GA 012) Codex Boernerianus is a ninth-century Greek-Latin diglot of the Pauline letters written on parchment.72 The text was written with spacing between words, but without breathing marks or accents. There are also regular sense-unit divisions, indicated by the use of punctuation (e.g., high points). Metzger notes that the exemplar of Boernerianus was likely arranged by stichoi: “If the capital letters be assumed to commence the lines of the exemplar, the text divides itself into regular στίχοι.”73 Although the first book, Romans, begins with ornate letters and contains color (yellow, red, pink) embellishments, these features disappear after Rom 5:15 (folio 8) in the editio princeps.74 The manuscript itself contains color embellishments throughout.75 In addition to stichoi divisions, the scribe occasionally utilized filler marks, resembling double (>>) or triple (>>>) diplai to mark a sense unit near the right margin of the manuscript (e.g., after Rom 1:7, 24; 3:7, 21; Col 3:20; 2 Tim 2:19) and also intralinearly (e.g., at Rom 4:12, 18; 5:5, 9; Phlm 3, 17).76 Of importance for this discussion is the occasional use of these marks just prior to a quotation (e.g., at Rom 4:6; 2 Tim 2:19). Although the primary method of marking citations in Boernerianus is the use of diplai in the left margin (to be discussed below), the addition of these filler marks implies an additional layer of interpretation. A particularly fascinating example of this is found in the extended quotation in Rom 3:10–18, which is split between two pages of the codex. What is noteworthy about this quotation is that it is marked by diplai in the left margin of the text as well as by filler marks in the right margin. The double 72 This manuscript was damaged during WWII, making Reichardt’s facsimile edition an important witness to the text. Alexander Reichardt, Der Codex Bœrnerianus: Der Briefe des Apostels Paulus (Msc. Dresd. A 145h) (Leipzig: Hiersemann, 1909). 73 Metzger, Manuscripts of the Greek Bible, 104. 74 There is some discussion about the origin and the extent of the coloring in the manuscript itself. See the information available on the CSNTM website. 75 A digital copy is available at http://digital.slub-dresden.de/werkansicht/dlf/2966/17/0/. 76 See P.Oxy. VIII 1086 (1st c. BC) for an example of intralinear diple.

Composite Citations in New Testament Greek Manuscripts

Figure 9.30

Romans 3:12a in GA 012 (f. 4v)

Figure 9.31

Romans 3:12b–18 in GA 012 (f. 5r)

259

diplai appear at the beginning of the quotation in Rom 3:10 (l. 1 in Figure 9.30) and the triple diplai appear at the end of 3:14 (l. 4 in Figure 9.31), precisely where wording from Ps 13:1–3 and Ps 9:28 stops and where wording from Isa 59:7–8 begins.77 This is a fascinating example because it seems to distinguish sources (e.g., Psalms and Isaiah) within the larger composite quotation.78 Moreover, because Rom 3:13–18 is found in some LXX Psalm manuscripts after Ps 13:3 (e.g., A), this citation in Boernerianus may represent a different scholarly reading tradition of the quotation that understood Rom 3:10–14 as a unit (rather than 3:13–18).79 77 On the citation in Rom 3:10–18, see Reasoner, “Promised Beforehand,” 128–158. 78 This observation is further underscored by the fact that the triple diplai at the conclusion of Rom 3:14 are colored, following the same pattern of coloring the single diplai in the left margin. Only two other triple diplai are colored: see Rom 1:7 and Rom 3:21 in the manuscript. 79 On Rom 3:13–18 in LXX A, see Robert A. Kraft, “Christian Transmission of Greek Jewish Scriptures: A Methodological Probe,” in Paganisme, Judaïsme, Christianisme: Influences

260

Adams and Ehorn

Turning to the use of diplai in Boernerianus, we find a well-defined practice throughout the manuscript.80 For example, if we use Richard Longenecker’s statistic that there are 83 quotations in the Pauline letters,81 it is striking that this codex uses diplai at 77 places in Paul’s letters, resulting in a percentage of just under 93 percent. Most of Longenecker’s list overlaps with the diplai in Boernerianus.82 This figure, of course, depends upon Longenecker’s definition of “quotation” (which is certainly disputable), but the point here is simply to illustrate the extensive use of diplai to mark quotations in this manuscript. The most important paratextual feature in Boernerianus for our purposes is the regular use of source attribution in the left column of the manuscript, similar to that found in Codex Sinaiticus above. These attributions are not uniform in their presentation. For example, sometimes an attribution appears in an abbreviated format (e.g., δευτ at Rom 15:9; δευτο at Rom 10:20) and sometimes with fuller spelling (e.g., δευτερονομιω at Rom 10:6). However, the attributions are typically placed next to the first diple that marks the quotation. Moreover, nearly all of the attributions are preceded by the same scribal feature: an abbreviated form of the Latin word “in.”83 The abbreviation appears as the letter “I” with a superlinear stroke. Figure 9.32 Attribution Signal in GA 012

This understanding of the abbreviation is confirmed by other examples in the manuscript where the scribe failed to abbreviate the word prior to the attribution (cf. Rom 7:7; 2 Cor 6:16–18, see Figure 9.33 below).84

80

81 82

83 84

et affrontements dans le monde antique, ed. A. Benoit, M. Philonenko, and C. Vogel (Paris: E. de Boccard, 1978), 207–226. Diplai are found at: Rom 1:17; 3:4; 3:10–18; 4:3, 7–8, 17; 7:7; 8:36; 9:7, 9, 12, 13, 15, 17, 25–26, 27–28, 29, 33; 10:6, 8, 11, 13, 15, 16, 18, 19, 20–21; 11:3, 4, 8, 9–10, 26–27, 34–35; 12:19–20; 13:9; 14:11; 15:3; 15:9, 10, 11, 12, 21; 1 Cor 1:19, 31; 2:9, 16; 3:19, 20; 9:9; 10:7, 26; 14:21; 15:32, 45, 54–55; 2 Cor 4:13; 6:2; 6:16–18; 8:15; 9:7, 9; 10:17; 11:19; Gal 3:6, 8, 10, 11, 13; 4:22–23, 27, 30; 5:14; Eph 4:8; 5:14, 31; 1 Tim 5:18; 2 Tim 2:19. Richard N. Longenecker, Biblical Exegesis in the Apostolic Period, 2nd ed. (Grand Rapids: Eerdmans, 1999), 92–95. Longenecker, but not Boernerianus, includes Rom 2:24; 4:18; 10:5; 1 Cor 6:16; 15:27; 2 Cor 13:1; Gal 3:12, 16; Eph 6:2–3. These differences highlight the fact that Boernerianus marks several other texts with diplai that are not included in Longenecker’s list. Nevertheless, the overlap is striking. See Adriano Cappelli, The Elements of Abbreviation in Medieval Latin Paleography (Lawrence: University of Kansas Libraries, 1982), 14. We wish to thank Jeffrey Kloha for assistance with the Latin abbreviations in this manuscript.

Composite Citations in New Testament Greek Manuscripts

261

Figure 9.33 Romans 7:7 in GA 012 (f. 9v)

There are several mistakes in Boernerianus’s attributions (e.g., Rom 10:20; 15:9; 1 Cor 3:19, 20; 2 Cor 10:17; Gal 3:11; 4:27), which suggests that the scribe who first added these did not crosscheck the quotations in the relevant sources. A particularly egregious example appears at Rom 10:20, where the attribution to δευτο (i.e., Deuteronomy) appears in the left margin but the text itself provides an attribution to ησαιας (i.e., Isaiah).

Figure 9.34

Romans 10:20 in GA 012 (f. 14r)

Nevertheless, the practice of adding attributions to the manuscript does indicate how some readers understood the quotations in the text. The following attributions appear for the composite citations in Paul’s letters: Rom 9:9 Rom 9:20 Rom 10:6 Rom 10:15 Rom 11:8 Rom 11:26–27 Rom 11:34–35 Rom 13:9 Rom 14:11 Rom 14:21 1 Cor 2:9 1 Cor 15:54–55 2 Cor 6:16–18 Gal 3:10 Gal 3:13 1 Tim 5:18

Attribution γενεσεως εϲαια δευτερονομιω εϲαια, naum εϲαια εϲαια εϲαια εχοδος, λευιτικον εϲαια ιερεμια εϲαια, αποκαλιψι ενοχ, και ηλιας osee εϲαια, εϲαια, λευιτικον ηῑῑῑς δευτερονομιω δευτερονομιω δευτερονομιω

Diplai in Left Margin? yes no yes yes yes yes yes yes yes yes yes yes yes yes yes yes

262

Adams and Ehorn

Figure 9.35 Romans 9:9 (f. 12v)

Figure 9.36 Romans 9:20 (f. 12v)

Figure 9.37 1 Timothy 5:18 (f. 89v)

Figure 9.38 1 Corinthians 15:54–55 (f. 39v)

Figure 9.39 1 Corinthians 14:21 (f. 36r)

Figure 9.40 Romans 13:9 (f. 17r)

Figure 9.41 Romans 10:15 (f. 15r)

Figure 9.42 1 Corinthians 2:9 (f. 22r)

Figures 35–42

Source Attributions in GA 012

Several observations follow. First, like the main text of the manuscript itself, the attributions appear in both Greek and Latin.85 Second, the use of a source attribution almost always corresponds to the use of diplai in the left margin of the manuscript. The exception to this is Rom 9:20, which may simply be an irregularity.86 Third, three of the composite citations receive more than one source attribution in the margin. For example, 2 Cor 6:16–18 includes multiple source attributions that appear, uncharacteristically, in both the left and right margins of the manuscript.87 These attributions correspond to some, but not all, of the known sources for the quotation (i.e., Lev 26:11–12; Isa 52:11; etc.).88 Significantly for our purposes, these multiple source attributions demonstrate an awareness of the composite nature of the quotation. We can also observe the unusual placement of the attribution to Isaiah at the fourth diple in the left margin. Given that 85 Contra Metzger, Manuscripts of the Greek Bible, 104, who states that the source attributions in GA 012 were written in Latin. 86 Another possibility, although less likely, is that Rom 9:20 was not considered a quotation but, perhaps, was still considered to have language reminiscent of Isaiah. 87 According to the Reichardt, the symbols next to the first and last diplai are cantillation marks and not textual notations relating to content. See Reichardt, Der Codex Bœrnerianus, 4–23. 88 See Roy Ciampa, “Composite Citations in 1–2 Corinthians and Galatians,” in Composite Citations in Antiquity: Vol. 1, 160–174.

Composite Citations in New Testament Greek Manuscripts

Figure 9.43

263

2 Corinthians 6:16–18 in GA 012 (f. 46r)

attributions in Boernerianus are almost always placed beside the first diple, this may suggest that the scribe recognized material from Isaiah starting at this point and did not see the first half of the quotation as coming from the same source.89 The lack of marginal attribution, however, was remedied by a later scribe, who identified material from Isaiah and Leviticus in the line where the quotation begins, although this time in the right margin (see Figure 9.43). Another quotation with multiple attributions comes from 1 Cor 2:9. Here Boernerianus has attributions to “Isaiah,” “Apocalypse of Enoch,” and “[Apocalypse of] Elijah” in the margin. These attributions align with both modern and ancient discussion about the source(s) that inform 1 Cor 2:9.90 For our purposes, these multiple attributions clearly demonstrate that attribution to “Isaiah” alone was not sufficient to explain the derivation of the quotation. Another citation of particular interest is the quotation at Rom 11:34–35 (cf. Isa 40:13; Job 41:3).91 In this composite text the scribe marked sense units in several distinct ways, as may be seen in Figure 9.44.

Figure 9.44

89 90 91

Romans 11:34–35 in GA 012 (f. 15v)

Another example when the attribution is not placed by the first diple appears at 1 Cor 1:31. However, it is likely that this attribution was displaced by another paratextual feature in context. See Ciampa, “Composite Citations in 1–2 Corinthians and Galatians,” 186–187, for discussion on this passage and its sources. See Reasoner, “Promised Beforehand,” 138–140.

264

Adams and Ehorn

The quotation is preceded by two middle points, signaling a stronger break than the single point that is more commonly used throughout the manuscript. As expected, diplai mark the quotation in the left column; however, only two (out of three) lines of the quotation are marked in this way. Unexpectedly, following the opening line (cf. Isa 40:13a), the scribe indented the rest of the quotation, which consists of wording from Isa 40:13b and Job 41:3 (although the final indented line did not receive a diple). It is not clear why the scribe did this, but indenting lines is very rare in this manuscript (cf. 1 Cor 10:1; Gal 4:6; Col 3:21; Titus 1:2; Phlm 11).92 Finally, immediately following the quotation, the text of Rom 11:36 continues and is re-justified to the left margin. 5.2 Summary The paratextual features in Codex Boernerianus convey a significant level of interest in reading Paul’s letters. The extensive use of diplai and source attribution represent a strong interest in quotation practice, including understanding the derivation of quotations. Source attributions present the most significant data for this study on composite quotations. In particular, three quotations (i.e., Rom 13:9; 1 Cor 2:9; 2 Cor 6:16–18) include multiple source attributions, demonstrating an awareness that multiple sources inform these quotations. 6 Conclusion A few observations can be made from this study of citation paratextual features. First, beginning at the fourth century, scribes regularly used some type of paratextual feature to signal the citation of a text. This practice was not commonly found in the papyri, although there is limited evidence, but became more prevalent over the centuries. Second, although there was wide adoption of paratextual features in general, there was much less consistency among scribes as to which feature(s) they would adopt. By far the most common signal was the diple, for which we saw evidence in most New Testament codices in this study (e.g., Sinaiticus, Vaticanus, Alexandrinus, Ephraemi, and Boernerianus). However, other features were also employed in different manuscripts (e.g., rubrication, indentation, attribution).

92

These passages do not correspond to Old Testament citations. Perhaps the poetic nature of Job led to this layout in Rom 11:34–35. However, this explanation is not very strong given that the Greek versions of Job use prose (rather than poetic) features.

Composite Citations in New Testament Greek Manuscripts

265

Third, paratextual features were often used in clusters and not just individually. For example, citations for which diplai were used often had corresponding sense-unit divisions, either at the beginning or the end of the citation, marked by intralinear spaces, ekthesis, and/or paragraphoi. The concentration of paratextual features acts to isolate the citation from its textual surroundings and draws the readers’ attention. Fourth, the occurrence of paratextual features for marking citations is strongly correlated to the use of introductory phrases within the text (e.g., γέγραπται, καθὼς γέγραπται, etc.). This collocation suggests that overt claims of citation within the text by the author influenced the practices of subsequent scribes. It also implies that scribes did not copy the text without consideration of the text’s content but noted places where citations occurred and used paratextual features to mark them accordingly. Fifth, none of the scribes evaluated in this study was consistent in their practice. This is not to say that these scribes changed the means by which they marked citations (e.g., diple, indentation, ink color, etc.). Rather, these scribal features are not applied to all occurrences of cited texts. For some instances, such as paraphrases or quotations (or near quotations) in speech, it might be that the scribe did not recognize or view the passage as a quotation. More commonly, we see that the scribe lost motivation for some reason (e.g., boredom, fatigue, impatience) and so the inclusion of some paratextual features fades as the text progresses (e.g., source attributions in Sinaiticus). Sixth, the scribe regularly marked the whole of the citation and did so accurately. When a scribe noted a citation within the manuscript and decided to include paratextual features (e.g., diplai) to mark its occurrence he did so with high levels of accuracy. This suggests that the scribe reviewed the text to ensure that the diplai or indentation covered the entire quotation. Accordingly, when certain citations are not fully marked/indented/colored/etc., there is scope for further investigation into possible reasons that go beyond the standard claim of scribal error. Seventh, we found that there was a gradual increase in sense-unit divisions and paratextual elements over the centuries. This accretion over the centuries indicates that subsequent scribes and readers built on the findings of previous generations. In particular, source attributions provide important insight into ancient reading practices. Their gradual development and increased use over time indicates a growing sophistication in reading and exegetical strategies.93

93

For a similar conclusion with other examples, see Sigismund, “Die Diple als Zitatmarkierung in den ‘großen’ Unzialcodices,” 152.

266

Adams and Ehorn

Finally, and related to the above point, we found that awareness of composite citations by scribes increased alongside the development of more refined reading/signaling practices. Initially, there was minimal evidence that scribes recognized composite citations. However, in many of the codices, there are instances in which the scribe made adjustments, either to the text, the layout, or through paratextual features, to differentiate foreign material in composite citations and separate it from the “true” quotation. In a number of cases this partition was accomplished through the use of sense-unit divisions, which physically separated the two texts (e.g., Mark 1:2–3). In other instances, the scribe used the absence of markers (e.g., diple or red ink), which, in effect, indicates to the reader that this line is not cited Scripture. These actions directly impact how the text would be read and provide evidence of reading practices and interpretations found within early Christian communities. These changes in text presentation are valuable for evaluating proposed composite citations by scholars. For example, Mark Reasoner suggested that Rom 11:34–35 was a possible composite citation of Isa 40:13 LXX and a non-LXX version of Job 41:3.94 In our evaluation of manuscripts, we have found that scribes did not have a consistent view of the unity of Rom 11:34–35 as a citation. One example is that of Codex Claromontanus, in which only 11:34 is written in red ink. On the other hand, Claromontanus presents 1 Tim 5:18 as a unified citation rather than dividing it into two quotations divided by καί. Such evidence is not conclusive but does provide empirical support for or against a proposition. In the Latin (L) version of Barnabas (Codex Petropolitanus Bibl. Publ. Q. v. I 39), a composite citation is sometimes re-divided by adding a new, distinguishing introductory formula (e.g., Barn. 5.12; 6.6; 6.16; 9.2). Although this practice was not carried out with consistency, we see the willingness of one scribe/translator to divide certain composite citations into their discrete sources and provide each with its own source attribution. Occasionally, the scribe of L omits divisions that result in new composite citations that are not found in the Greek versions (e.g., Barn. 6:4; 9:3; 11:4). 94 Cf. Reasoner, “Promised Beforehand,” 138–140.

CHAPTER 10

Titus in P32 and Early Majuscules: Textual Reliability and Scribal Design Chris S. Stevens 1 Introduction Critical examination of the physical manuscripts is often treated as peripheral in biblical studies. Attention is given to the texts contained in manuscripts, or more narrowly to the modern critical text, but surprisingly little effort is given to teaching students to interact directly with the historical documents. The state of affairs is unfortunate, as Harry Gamble contends that “the physical object is also a social artifact…. All aspects of the production, distribution, and use of texts presuppose social functions and forces … the design of the text as a concrete medium, physical object.”1 The manuscripts are significant historical artifacts and are essentially physical pieces of church history that should influence the interpretation of the text. In this chapter I will conduct a process of manuscript criticism. While the term is not used in textual criticism or biblical studies, the aim of manuscript criticism is to examine the physical manuscript from multiple angles without interpretation of the textual content.2 The primary goal is deriving as much information as possible from the physical features, namely layout, spatial segmentation, margins, and overall scribal design. While many of these elements appear stereotyped in different centuries and scriptoriums, there is enough variation in the design choices by individual scribes that require further attention. The impetus for manuscript criticism is the contention by Mary A. Rouse, There is a relationship between the physical form of an artifact and the function that it is meant to serve; therefore, the physical form of a manuscript is evidence of the purpose for which the manuscript was made.

1 Harry Y. Gamble, Books and Readers in the Early Church: A History of the Early Christian Texts (New Haven, CT: Yale University Press, 1995), 43. 2 Michael S. Batts, “Poetic Form as a Criterion in Manuscript Criticism,” The Modern Language Review 55 (1960): 543–552.

© Chris S. Stevens, 2023 | doi:10.1163/9789004537972_012

268

Stevens

This fact, simple and direct, is the inevitable starting point of any study of the manuscript as an object.3 Scribal design choices have intentionality behind them, and the goal of manuscript criticism is to attempt to examine those implications from the physical medium. The chapter will consist of three sections. The first will examine various aspects of P32. The second section examines P32 using a comparative method. The section will highlight scribal choices of the paratextual features. Third, the scribal design of Titus in Sinaiticus is explored for possible liturgical purposes, or at least communal reading.4 I conclude that P32 and the majuscules indicate the text of Titus was transmitted in a highly uniform manner and the scribe of Sinaiticus used layout features to facilitate public reading. 2

P32 Investigation

Manuscript (henceforth MS; pl. MSS) criticism aims to investigate all profitable angles, traditionally beginning with the history, provenance, physical measurements, and content being examined first. The oldest MS of Titus is P32, dated from ca. 200 to the third century.5 Its provenance is mostly unknown, but it was recovered at Oxyrhynchus. It is currently housed in Manchester, John Rylands University Library, Shelf Number Gr. P. 5. Physically it is not an exceptional document like its contemporary P46. The extant fragment is only 4.9 × 10.5 cm. Given the number of lines and the space necessary for the missing text, Comfort, Turner, K. Aland, and Gathergood, estimate the MS would originally

3 Mary A. Rouse and Richard H. Rouse, Authentic Witnesses: Approaches to Medieval Texts and Manuscripts (Notre Dame, IN: University of Notre Dame Press, 1991), 4. 4 For a lengthy explanation of communal reading and its widespread occurrence in the early church, see Brian J. Wright, Communal Reading in the Time of Jesus: A Window into Early Christian Reading Practices (Minneapolis: Fortress, 2017). 5 For a ca. 200 dating see Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, 2nd ed., trans. Errol F. Rhodes (Grand Rapids: Eerdmans, 1995), 57. The date is adopted by the NA28 apparatus. For dating to the second half of the second century, see Philip Wesley Comfort and David P. Barrett, The Text of the Earliest New Testament Greek Manuscripts (Wheaton, IL: Tyndale House, 2001), 134. For dating in the third century see Emily J. Gathergood, “Papyrus 32 (Titus) as a Multi-Text Codex: A New Reconstruction,” NTS 59.4 (2013): 588–606, esp. 592.

Titus in P32 and Early Majuscules

269

have been 12–15 × 16–20 cm.6 The double-sided papyrus was written as a single column of typical text size. Folio 1 begins with Titus 1:11b–15b7 ζειν οιτ πουσι διδ κερδους χα ϊδιος αυτω 5 ψευσται κ η μαρτυρι ην αιτιαν ϊν̣ ̣ ̣ ̣ ̣ α̣ ιν προ̣ εχον 10 και εντολ μενων τ ̣ θαρα το̣ ̣ις̣ ̣ μεν ̣ ̣ ̣ κ̣ Folio 2 begins with Titus 2:3c–8b λους ϊνα ας φιλαν ους σωφρο ς αγαθας 5 ιδ̣ ̣ ιοις ̣ αν ς ̣ του Θ̅ Υ̅ βλα τερ̣ ους ει̣ ̣ σ̣ωφρο το̣ν παρε

6 Comfort and Barrett, Earliest New Testament, 134; E. G. Turner, The Typology of the Early Codex, Haney Foundation Series 18 (Philadelphia: University of Philadelphia Press, 1977), 147; Kurt Aland, Repertorium der griechischen christlichen Papyri, I: Biblische Papyri, PTS 18 (Munich: De Gruyter, 1976), 253; Gathergood, “Papyrus 32,” 595. 7 The text of P32 was determined using images in the NT.VMR provided by the Institut für Neutestamentliche Textforschung. The only adjustments made in this reconstruction are graphical. I have used a modern sigma rather than a lunate, full letters rather than superscripted, and kept the corrector’s insertions in Sinaiticus for reasons that will be explored below. My transcription of P32 agrees with the one by Gathergood (“Papyrus 32,” 593).

270 10

Stevens

ω̣ν εργων θονιαν σε α̣καταγνω α̣ς εντρα

There are nine different nomen sacrum with a macron in the majuscules of Titus: contracted ΣΡΣ – σωτῆρος, ΠΝΣ – πνεύματος, ΠΡΣ – πατρός, and suspended ΘΥ – θεοῦ, ΙΥ – ἰησοῦ, ΧΥ – χριστοῦ, ΘΝ – θεόν, ΘΩ – θεῶ, ΘΣ – θεός.8 In P32 there is only one location of a nomina sacra Θ̅ Υ̅ , and that is in line six of the verso.9 While there is only a single occurrence of this peculiar feature, it is consistent with scribal practices for Christian MSS. The hand is also typical of other documents with New Testament texts, namely a common rounded script with minimal decorative strokes.10 Interestingly, the top stroke of the letter π in lines two and nine on the recto extend into the left margin. It is unlikely to be used for ekthetic purposes since the first occurrence is mid-word and the latter is not a natural break – topical, addressee, etc. – of any sort. P32 contains roughly twenty-six lines of Titus, which corresponds to eight verses in the modern NA28 text. P32, therefore, explicitly attests to less than ten percent of Titus. However, such a small papyrus does not leave the investigator with a negligible amount of data. Rather, when effort is spent investigating the document in full detail, further information can be gained.

8

For instance, see Sinaiticus ΣΡΣ in line 16; ΠΡΣ in line 21; ΘΣ in line 9. For a description of the nomina sacra generally see Ludwig Traube, Nomina Sacra: Versuch einer Geschite der Christlichen Kürzung, Quellen und Untersuchungen zur lateinischen Philogie des Mittelalters (Munich: Oskar Beck, 1907); A. H. R. E. Paap, Nomina Sacra in the Greek Papyri of the First Five Centuries A.D.: The Sources and Some Deductions, Papyrologica LugdunoBatava 8 (Leiden: Brill, 1959). For specifics of the nomina sacra in Codex Sinaiticus, see Gregory Scott Paulson, “Scribal Habits in Codex Sinaiticus, Vaticanus, Ephraemi, Bezae, and Washingtonianus in the Gospel of Matthew” (PhD thesis, University of Edinburgh, 2013), 28–30, 125–126; Scott D. Charlesworth, “Consensus Standardization in the Systematic Approach to Nomina Sacra in Second- and Third-Century Gospel Manuscripts,” Aeg 86 (2006): 37–68, here 40. See Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 100–101. 9 I am using the terms front/back and recto/verso interchangeably rather than horizontal and vertical. See Gamble, Books and Readers, 265 n. 9. 10 See especially the deltas in recto line 2 and the ligature of αι in recto line 8. P32 also has typical textual features observed elsewhere. See Michael J. Kruger, Canon Revisited: Establishing the Origins and Authority of the New Testament Books (Wheaton, IL: Crossway, 2012), 234; Timothy N. Mitchell, “Christian Papryi and the Ancient Church,” BSac 173 (2016): 182–200, esp. 82–185.

Titus in P32 and Early Majuscules

3

271

Reconstructing P32 for Investigation

One of the more practical steps in examining fragmentary papyri is to create a conjectural reconstruction. Given the nature of having a front and back to the papyrus, the precise amount of missing text is calculable. The calculation is determinative for the number of letters per line and lines per page, thereby providing the ability to reasonably consider the non-extant portions. Gathergood makes excellent use of the reconstruction of P32 leading her to conclude that Titus 1:1–6 would occupy nineteen lines on its first page.11 Quite interestingly, having only nineteen lines would render about one-fifth of the page blank. Since leaving a large portion of a page blank is uncommon for a starting page, Gathergood concludes that the evidence suggests there was a different text preceding Titus 1:1 on the same column. Combined with other features, Emily Gathergood posits that P32 was originally part of a multi-text codex, with Titus not being the first text.12 Her conclusion posits P32 being a partially extant codex containing Titus as early as the second to third century. While it cannot be codicologically determined what the other texts were, she notes that the “canonical reception of Titus at the turn of the third century” makes it likely it was “in a collection of thirteen or fourteen Pauline scriptures.”13 While the object of this study is not the canonical history or reception status of Titus, the information and means used by Gathergood are instructive methodologically and suggestive for further analysis. I have beneficially used its methodology for the earliest documents of James, P20, P23, and P100.14 There I concluded that P20 and P23 both circulated as multi-text codices and P100 likely did too. Therefore, as with P32, the benefit to such an investigative tool is knowing that the earliest extant documents for both Titus and James circulated in a manner consistent with other known biblical codices. Furthermore, the work by Gathergood is suggestive for further means of analysis, namely exploration of the textual supplementation. While Gathergood did excellent work in her reconstruction, I think more can be done. For greater usability, the surviving portion of P32 needed supplementation and Gathergood chose NA28 to do so.15 However, there is no explanation why a modern critical edition is the best choice, or even a sound choice, for 11 12 13 14

Gathergood, “Papyrus 32,” 595–596. Gathergood, “Papyrus 32,” 597. Gathergood, “Papyrus 32,” 606. Chris S. Stevens, “Does Neglect Mean Rejection? Canonical Reception History of James,” JETS 60 (2017): 767–780, esp. 777–778. 15 Gathergood, “Papyrus 32,” 592.

272

Stevens

supplementation, especially given adjustments needing to be made for the nomina sacra – as Gathergood notes – and I would add that spelling variability would also make a difference for larger textual reconstructions.16 The modern academically created eclectic text is often used on account of it being convenient and likely sufficient. However, it is not the best choice for further MS criticism. The supplementation of the NA text into P32 is at best an evaluation of the text in P32 and at worst creating a circular argument.17 The supplementation and comparison are also limited to the textual content apart from form. I propose using a contemporary MS as a means of supplementing P32 for both textual and paratextual features. The advantages are manifold. Most relevant for this study is the similar scribal practices such as spelling, nomina sacra, and scribal formatting.18 The MS most closely dated to P32 in complete form is the famous Sinaiticus, which helpfully contains the entirety of Titus. Reconstructing P32 with the text of Sinaiticus, including textual supplementation, reveals an incredible textual agreement rate of 99.85 percent of the letters or 99.1 percent of the words. If supplemented with Alexandrinus or Claromontanus, not graphically depicted here, then P32 has a 100 percent agreement rate with them. Comparison with P32 also indicates the two scribal additions in Ephraemi Rescriptus were not part of the prima manus of the MS.19 Therefore, P32 is in nearly perfect textual uniformity with the earliest available and contemporary MSS down to around the sixth century. 16 No critique of Gathergood is intended. Her conclusions would not be greatly affected by using a slightly different textform. Gathergood is likely adopting the suggestion by Barbara Aland that the NA text be used as a base given that it is nearly identical with the “Ausgangstext der Überlieferung” (source of the tradition). Barbara ⁠Aland even calls the eclectic text a hypothetischer Ausgangstext in Barbara Aland, “Kriterien zur Beurteilung kleinerer Papyrus Fragmente des Neuen Testaments,” in New Testament Texual Criticism and Exegesis: Festschrift J. Delobel, ed. A. Denaux (Leuven: Leuven University Press and Peeters, 2002), 2; Barbara Aland, “Der textkritische und textgeschichtliche Nutzen früher Papyri, demonstriert am Johannesevangelium,” in Recent Developments in Textual Criticism: New Testament, Other Early Christian and Jewish Literature, ed. Wim Weren and Dietrich-Alex Koch (Assen: Royal van Gorcum, 2003), 19–38, here 19–20. Furthermore, Comfort and Barrett (Earliest New Testament, 135–137) do their reconstructions by supplementation but it is not always precisely clear what text they are following. 17 Bart D. Ehrman, “A Problem of Textual Circulatiry: The Alands on the Classification of New Testament Manuscripts,” Bib 70 (1989): 377–388, esp. 382–383. Additionally, no attempt is made here concerning the standard goal of textual criticism, which is to determine the earliest textform or Ausgangstext. 18 The use of contemporary MSS also avoids diachronic leaps and matters of circularity. 19 In the reconstructed section, Ephraemi Rescriptus has the additions of καὶ ἠ διδασκαλία at 2:5 and ἀγνειάν at 2:7. Furthermore, while Claromontanus has moved ἐστίν (Titus 1:13)

273

Titus in P32 and Early Majuscules

3.1

Layout of Titus 1:11b–15b in P32 Supplemented with Sinaiticus Lines 65–90 ζειν οιτ[ινες ολους οικους ανατρε πουσι διδ[ασκοντες α μη δι αισχρου κερδους χα[ριν ειπεν τις εξ αυτων ϊδιος αυτω[ν προφητης·κρητες·αει ψευσται κ[ακα θηρια γαστερες αργαι η μαρτυρι[α αυτη εστιν αληθης δι ην αιτιαν [ελεγχε αυτους αποτομως ϊνα ϋγιαιν[ωσιν εν τη πιστι μη προσεχον[τες ϊουδαϊκοις μυθοις και εντολ[αις ανθρωπων αποστρεφο μενων τ[ην αληθιαν παντα κα θαρα τοις[καθαροις τοις δε μεμιαμ μενοις κα[ι απιστοις ουδεν

3.2

10

Layout of Titus 2:3c–8b in P32 Supplemented with Sinaiticus Lines 114–137 λωμενας καλοδιδασκα]λους ϊνα σωφρονιζωσι τας νε]ας φιλαν δρους ειναι φιλοτεκν]ους σωφρο νας·αγνας οικουργου]ς αγαθας ϋποτασσομενας τοις] ϊδιοις αν δρασιν ϊνα μη ο λογο]ς του Θ̅ Υ̅ βλα σφημηται τους νεω]τερους ωσαυτως παρακαλ]ει σωφρο νειν περι παντα σεαυτ]ον παρε χομενος τυπον καλ]ων εργων εν τη δι δα σκ αλ ια αδιαφ]θονιαν σε20 μνοτητα λογον ϋγιη]ακαταγνω στον ϊνα ο εξ εναντι]ας εντρα

20

5

5

10

back one slot in comparison to the other majuscules, its location cannot be established in P32. The letter N in P32 is confirmed by the width of space available and the few remaining elements of ink. Visually it can be compared to the N in lines 2, 5, and 10. More will be discussed about the first three letters αδι.

274

Stevens

The above reconstruction using Sinaiticus is useful for further critical inquiries. First, contemporary spelling confirms the average number of letters per line is 25.27, and the textual supplementation precisely fits with the textual layout of P32.21 Second, the deletion of δέ in Titus 1:12 in Sinaiticus matches the line length in the reconstruction of P32 line 3. Third, in Titus 1:13 at line 32 of Sinaiticus, the corrector has superscripted the preposition ἐν, which the NA28 text adopts. The reconstruction of P32 lends support to the inclusion of ἐν. Third, in Titus 2:7 at line 35–36 of Sinaiticus, diacritical marks are signifying an error occurred. A scribe noticed the error and corrected it to indicate that τύπον should be read after παρεχόμενος. Again, the move is confirmed by the width of the resulting reconstructed column of P32 (see above line 10). A fourth matter, also in Titus 2:7, is more involved. At lines 37–38 of Sinaiticus a number of issues concern the word φθορίαν, or ἀφθορίαν, or ἀδιαφθοίαν. P32 is lacunose at the beginning of the word so the preformatives α- or αδια- are not verifiable. However, if ἀφθορίαν (eight letters) were the reading in P32 then the reconstruction would be twenty-five letters for the line, perfectly matching the average. Conversely, if P32 had the preformative αδια-, as the later correction in Sinaiticus does, then it is the longest line on the verso with twenty-eight letters. While line length is a variable matter, resulting in the longest line suggests against being the correct reading. Furthermore, the next MS to contain αδια- is not until the ninth century Codex Athous Lavrensis. The longer preformative does not become widespread until the eleventh century.22 Consequently, it is quite likely that the preformative was added to Sinaiticus late and was not part of the exemplar or the published form of Sinaiticus. There is a further variable in determining precisely what letters were added. Did the later edit add the three letters δια- or four letters αδια-? It is likely that a scribal error is what prompted the editor to add something in the first place. It is entirely possible the scribe of the prima manus wrote the previous word ending with -αλια then failed to write another alpha before moving straight onto the much more common φθορά.23 However, without the alpha negation, the reading prescribes corrupt teaching as a positive qualification for overseers, which is contrary to the clear intention of the co-text and themes of Titus. 21 Gathergood (“Papyrus 32,” 595) agrees in her rounded average of 25 letters. However, Gathergood does use an average of twenty-three letters per line for the intervening portion of 1:15–2:3, “Papyrus 32,” 592. 22 A few examples from the eleventh century are 1243, 1854, 1244, and 131. 23 φθοια is used nine times in the New Testament: Rom 8:21; 1 Cor 15:42, 50; Gal 6:8; Col 2:22; 2 Pet 1:4, 2:12, 2:19.

Titus in P32 and Early Majuscules

275

Therefore, if haplography is the root cause behind the scribal error of the prima manus, then someone subsequently recognized the problem and added not merely the missing alpha-negation but four letters αδια-. The scenario is more explanatory of the visual evidence since the four letters are more faded and further extending into the right column. It also explains why the reading of αδια- does not show up again in the extant evidence until half a millennium later. However, the complication in a definitive conclusion is, of course, both ἀφθορίαν and ἀδιαφθορίαν would be hapax legomena in the New Testament, with only the former being used in Esth 2:2 LXX. The last textual challenge to mention also concerns Titus 2:7. P32 reads -θονιαν, which differs from Sinaiticus’s -θοριαν by a single letter. Both words, ἀφθορίαν and ἀφθονίαν, have a similar meaning in the given co-text.24 Bruce Metzger claims the Sinaiticus reading should be accepted because it is widely read in the Alexandrian and Western texts, ‫ *א‬A C D* 33, and explains the alternative reading of αδια-.25 However, while the reading ἀδιαφθορίαν is in more diverse MSS, they are much later: Codex Athous Lavrensis, 1243, 1244, 1319, 1854, etc. Conversely, ἀφθορίαν is also quite diverse but is in the oldest reading of P32 and Codex Augiensis, Codex Boernerianus, 1735, 1881.26 In fact, the only feature supporting Metzger’s contention is the switch to a less common word in Sinaiticus, but both words are hapax legomena in the New Testament, so it is somewhat a moot point. A decision concerning the Ausgangstext is not an objective of this paper. Instead, using contemporary MSS for textual and paratextual comparison aides in an entirely different kind of investigation. The textual reconstruction

24 BDAG has ἀφθορία as soundness, incorruption, pure doctrine; LSJ also has incorruption. For ἀφθονία, BDAG LSJ have ‘freedom from envy.’ ἀδιαφθορία BDAG has sincerity, integrity. LSJM lists it under the heading ἀδιάφθαρτος as uncorrupted or incorruptible. 25 The letter N is confirmed by the width of space provided and the few strokes that remain. The N from line 2, 5, and 10, fit in the space as confirmed by the remaining ink in P32. This conclusion is contrary to the transcription in Comfort and Barrett, Earliest New Testament, 137. It is confirmed by Bruce M. Metzger, A Textual Commentary on the Greek New Testament: A Companion Volume to the United Bible Societies’ Greek New Testament (London: United Bible Societies, 1994), 585; James R. Royse, “The Early Text of Paul (and Hebrews),” in The Early Text of the New Testament, ed. Charles E. Hill and Michael J. Kruger (New York: Oxford University Press, 2012), 175–203, esp. 192; Gathergood, “Papyrus 32,” 59; “The Center for New Testament Textual Studies NT Critical Apparatus,” ed. New Orleans Baptist Theological Seminary: Center for the New Testament Textual Studies (CNTTS, 2004), Titus 2:7. 26 Royse (“Early Text of Paul,” 192) finds the reading in P32 of ἀφθορίαν to be interesting as it pushes the reading back six centuries before the next.

276

Stevens

of papyri, therefore, is useful even for their non-extant locations by textual supplementation. 4

Paratextual Formatting and Design

Using Sinaiticus for textual comparison and supplementation with P32 offers insight to the textual dimensions. The significant level of agreement – including Alexandrinus, Claromontanus, and Ephraemi Rescriptus – reinforces the decision of comparing P32 with contemporary texts. Moving beyond the textual, there is an additional benefit to a contemporary comparison, namely, the physical features of the MSS. Sinaiticus, and other ancient MSS for that matter, have design features that greatly differ from the design of the NA text. The most apparent difference between ancient and modern editions of the biblical texts is the results of the printing press. The modern eclectic text has become a standardized form concerning versification, spelling, punctuation, textual apparatus, materials, and so forth. Conversely, the ancient MSS do not have the same degree of stereotyping and consequently display greater variability from scribe to scribe and location to location. Such different features are not without consequence. Therefore, having looked at the textual features, critical attention is now given to the design and paratextual comparison of P32 with other MSS. The place to begin paratextual examination can be the same as the textual, namely assessing if the contemporary MSS confirm interpreting P32 as a multi-text codex in the manner Gathergood contends. She used the NA text to create a reconstruction of P32 that has twenty-seven words per line with twenty-six or twenty-seven lines.27 Using that figure and “assuming the scribe employed a similar amount of text,” Gathergood concludes seven or eight lines of blank space on the starting page.28 While the assumption is reasonable according to the textual examination, is it supported by a paratextual examination? 4.1 Design of Titus in Sinaiticus Dirk Jongkind notes that the “first impression one gets when looking at Sinaiticus is one of regularity,” but there is in no way complete visual

27

Gathergood (“Papyrus 32,” 592) also arrives at twelve lines though she uses an average of twenty-three letters with a range of twenty-one to twenty-six. 28 Gathergood, “Papyrus 32,” 595.

Titus in P32 and Early Majuscules

277

uniformity.29 Concerning Titus, there are six five-centimeter-wide columns. Quite interesting for our purposes is that not every line utilizes the common width. The most striking section is Titus 1:7b–8, where the scribe writes twelve lines significantly shorter than the rest; eleven of which contain a single word or a single word with a particle of negation. A few prima facie reasons indicating the scribal design of these short lines require further examination and have potentially significant ramifications for P32. First, the short lines are not sense-lines (colometrical) such as in Claromontanus, Boernerianus, and 0319. In those codices, each line contains only “the sense of the phrase in focus,” resulting in many blank spaces.30 By contrast, the short lines in Sinaiticus are markedly different, with each containing only a single word. Second, nothing in the physical medium – such as damage or imperfections to the materials – necessitated the design.31 Third, nothing about the lexemes or syntax is graphically difficult leading the scribe to reduce the potential for errors in writing or reading.32 Before exploring the scribal motivation and effects of the short lines in Sinaiticus, its ramifications for P32 are explored. 4.2 P32 Paratextually Consistent with Sinaiticus? If the short lines in Sinaiticus of Titus 1:7b–8 were also in P32, it would significantly change its reconstruction. The verses in question consist of 112 words, which at an average of twenty-five letters per line is about 4.4 lines. However, the design of Sinaiticus uses twelve lines for the verses. The additional seven or eight lines is interestingly the exact number of blank lines proposed by Gathergood’s reconstruction and confirmed by my calculations.33 Given that the blank space is the key evidence for the multi-text codex theory, the possibility of the paratextual segmentation at Titus 1:7b–8 adding seven or eight lines would be detrimental to the theory. Unfortunately, Gathergood does not comment on paratextual features or segmentation. 29 Dirk Jongkind, Scribal Habits of Codex Sinaiticus, TS 5 (Piscataway, NJ: Gorgias, 2007), 59. 30 Sean A. Adams, “Sense Units and Manuscript Families: A Test Proposal,” in The Language and Literature of the New Testament: Essays in Honour of Stanley E. Porter’s 60th Birthday, ed. Lois K. Fuller Dow, Craig A. Evans, and Andrew W. Pitts, BIS 150 (Leiden: Brill, 2017), 174–204, here 179. 31 Jones notes that “ancient scribes negotiated faults in their writing medium, whether they were holes, relaxed fibers, creases, or the like.” Brice C. Jones, “Scribes Avoiding Imperfections in their Writing Materials,” APF 62 (2016): 371–383, here 372. 32 Compare Sinaiticus to Alexandrinus and Vaticanus at Rom 13:7. The latter two use segmentation and other paratextual features to remove potential for ambiguities and errors. 33 Gathergood, “Papyrus 32,” 596.

278

Stevens

While the possibility of P32 having similarly segmented text must, at least on theoretical grounds, be left open, there are strong reasons to doubt P32 had short lines at Titus 1:7b–8. While it is the paratextual comparison that leads to such a possibility, further paratextual features also serve to doubt the probability. First and foremost, there are no indications that P32 had any spatial segmentation. Second, there are no visible short lines as the verso is right justified. Also, the reconstruction – supplemented with Sinaiticus – does not create lines varying much from the average letters per line. Third, there is only one explicit indicator of any such paratextual features, namely the macron bar over the nomen sacrum Θ̅ Υ̅ in line six of the verso. Furthermore, there is no subsequent support for the transmissional history of the short lines in Sinaiticus. The support for this reasoning is that the short lines in Sinaiticus are not replicated in any other extant MS. All other contemporary majuscules do not have any lines as short in Titus. Additionally, beyond the majuscules, no document in the first millennium has the same layout as Sinaiticus.34 There is only one MS that has segmentation similarities to Sinaiticus. In 088, a fifth-/sixth-century MS, eight lines match Sinaiticus from μὴ αὐθάδη to σώφρονα (Titus 1:7–8).35 The lines μὴ αἰσχροκερδῆ and ἀλλὰ φιλόξενον are slightly above average letters per line in 088 and match perfectly with Sinaiticus. However, while the two MSS have similarities, there is a fundamental reason for doubting 088 is attempting to copy the formatting of Sinaiticus. In 088 the lines of Titus 1:7–8 are largely consistent with the average letters per line of the rest of the MS. They are not substantially shorter as in the case of Sinaiticus. Therefore, the only MS that reasonably matches Sinaiticus is not a good connection. I conclude that while P32 could have contained short lines like Sinaiticus and, therefore, not had any blank space at the beginning, it is highly improbable and lacks supporting evidence. The findings by Gathergood that Titus of P32 is not the first text in a multi-text codex is, therefore, corroborated by paratextual features and scribal design of other MSS. Consequently, Sinaiticus uses a spatial segmentation that is unlikely to be in any available exemplar and is the originator of the design choice. 4.3 Sinaiticus and Scribal Design The examination of P32 using comparative paratextual features confirmed previous findings but has highlighted a peculiar feature of Sinaiticus. If the 34 35

Exploration was not conducted of documents dated after AD 1000. 088 contains 1 Cor 15:53–16:9; Titus 1:1–13. It is housed in St. Petersburg and shelf number Gr. 6 II, fol. 5–6.

Titus in P32 and Early Majuscules

279

short lines in Titus 1:7–8 are not part of transmissional history, preceding or subsequent to Sinaiticus, then there are two important and interrelated questions to ask. Why did the scribe make the lines short, and what purpose does the design serve? The questions become more perplexing when recognizing how anomalous the design is. No other part of Titus has similarly shortened lines. Also, such a concentration of short lines is not typical of Sinaiticus. In commenting on distinguishing features of scribes, Milne and Skeat point to tendencies such as compressed letters (Scribe A), filling marks (Scribe B), and the καί compendium (Scribe D), but they make no mention of leaving large blank segmented space.36 Clearly, the choice of concentrated segmentation blanks is no accident or typical of the scribes of Sinaiticus. Turning to internal considerations, the content does not immediately offer a resolution. The qualifications for an overseer τὸν ἐπίσκοπον are listed in Titus 1:7–8. However, the parallel passage at 1 Tim 3:2–4 does not have short lines in Sinaiticus or the other majuscules. One might suspect that if the content, syntax, or lexemes were the motivation for the short lines in Titus, the same would be true in 1 Timothy. By all accounts, the two texts have the same scribe.37 However, there are other locations in the Pauline corpus that does have a similar design of spatial segmentation.38 In 2 Tim 3:2–3, there are ten short lines, each with only a single word. The segmentation design there, as with Titus, has no parallel in other MSS. It seems that whatever the motivation for the short lines was it compelled the scribe to do the same in both locations. Interestingly, the two passages do not have the same topic. Also, similarities in syntax do not readily resolve the design anomaly. While both locations have a concentration of adjectives, there are also five adjectives in a row at Rom 1:30–31, Col 1:15, and 1 Tim 3:2, but no short lines at either of those

36 H. J. M. Milne and T. C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938), 22–24. 37 Milne and Skeat (Scribes and Correctors, 29) contend Scribe A wrote the Pauline corpus except for 1 Thess 2:14–3:18. See also Kirsopp Lake and Helen Lake, Codex Sinaiticus Petropolitanus: The New Testament, the Epistle of Barnabas and the Shepherd of Hermas Preserved in the Imperial Library of St. Petersburg (Oxford: Clarendon, 1911), xix; Jongkind, Scribal Habits, 39–44. 38 Examining the Pauline corpus makes no claim as to authorship. Reference and examination concern the thirteen letters known as the canonical Pauline corpus. Similar scribal design is used in listing the fruit of the Spirit in Gal 5:22–23 and the list of sins in 1 Cor 6:9–10.

280

Stevens

locations.39 Therefore, an explanation for the shared scribal design of Titus, 2 Timothy, Gal 5:22–23, and 1 Cor 6:9–10 need to look at other factors, namely, what does the segmentation accomplish. 4.4 Consequences of Spatial Segmentation In modern texts, such as this book, the punctuation, word, paragraph, and essay sections are based on silent reading practices that became prevalent in thirteenth-century Europe.40 Ancient practices are different from modern ones. In his dissertation, Jeongseop Ahn examines how the majuscules, P66, and P75, used spatial segmentation and punctuation differently than today. He concludes that the breaks in the majuscules – mainly focusing on ekthesis – are intended to assist with the public reading in worship services.41 Specifically, Ahn notes that visual spacing is used in Sinaiticus to mark both major and minor shifts.42 The conclusion appears reasonable. For instance, the uses of ekthesis in Titus 1:10 line 11, Titus 1:12 line 22, and Titus 1:15 line 39, are all at locations of topical shifts in addressing different social groups. The scribe shortens the preceding line and adds an ekthetic letter to accentuate the feature. Larry Hurtado believes Christian writings contain such features in order to make “the texts accessible to a wider range of reader-competence, with fewer demands made on readers to engage and deliver them.”43 Further attention will, therefore, turn to ancient reading as the possible explanation for the unique feature of Sinaiticus.

39 Titus 1:7–8 has eleven adjectives in twenty-six words and 2 Tim 3:2–4 has twelve adjectives in nineteen words. 40 Paul Saenger, Space between Words: The Origins of Silent Reading (Stanford, CA: Stanford University Press, 1997), 256–257. 41 Jeongseop Ahn, “Segmentation Features in New Testament Manuscripts: An Overlooked Resource for Editors and Translators” (PhD dissertation, New Orleans Baptist Theological Seminary, 2012), 31. Cf. Dan Nässelqvist, Public Reading in Early Christianity: Lectors, Manuscripts, and Sound in the Oral Delivery of John 1–4, NovTSup 163 (Leiden: Brill, 2016), 48. Similarly, Sean Adams (“Sense Units,” 177) found a high consistency of using ekthesis to distinguish sense units in Arethas Codex and Codex Mutinensis. See also David I. Yoon, “Scribal Use of Ekthesis,” in this volume. 42 Ahn, “Segmentation Features,” 33. While the work of Ahn is thorough he gives far too much attention to ekthesis at the expense of other paratextual features. For instance, Ahn (“Segmentation Features,” 33) marks Titus 1:10 with a M (major paragraph shift) since it has an ekthesis but 1:7–8 as a m (minor). Ahn believes there is a paragraph break – possibly indicated by the use of γάρ – and ignores the more extensive spatial segmentation. 43 Larry W. Hurtado, “Manuscripts and the Sociology of Early Christian Reading,” in The Early Text of the New Testament, 49–62, here 59.

Titus in P32 and Early Majuscules

5

281

Ancient Reading

The ancient use of scriptio continua and absence of punctuation made reading difficult. Quintilian implies it could take years to obtain adequate skills to read the form.44 The challenge was not merely the act of performing aloud, which a trained lector or ἀναγιγνώστης could fulfill.45 The challenge was interpreting the text and then reading aloud. Raffaella Cribiore notes that “ancient manuscripts did not make many concessions to readers. A passage made of words written without separation in continuous blocks and containing almost no punctuation was only an ensemble of letters in need of interpretation. Reading at first sight was practically impossible.”46 Whoever was going to read had to make prosodic decisions but only after deciding what the text is. For instance, the simple string of letters ΟΤΙ could be either ὅ τι or ὅτι. While communal reading was difficult in the ancient world, in more than one place within the Pauline corpus there is an exhortation to consistently maintain public reading. In 1 Tim 4:13 the church is commended to be consistently engaged in the public reading of the Scriptures πρόσεχε τῇ ἀναγνώσει, which is similarly stated in 1 Thess 5:27 and Col 4:16.47

44 Quintilian, Institutio oratoria 1.1.31. See also William A. Johnson, Readers and Reading Culture in the High Roman Empire: A Study of Elite Communities, Classical Culture and Society (New York: Oxford University Press, 2010), 28–31. 45 For an examination of the use of ἀναγιγνώσκω, see Donald J. Allan, “ΑΝΑΓΙΓΝΩΣΚΩ and Some Cognate Words,” CQ 30 (1980): 244–251. By the end of the second century the office of reader was clearly established; Gamble, Books and Readers, 218. See Tertullian, Praescr. 41. Also, Eusebius, Hist. Ecc. 1.11, lists the reader along with other offices, “that in it there were forty-six presbyters, seven deacons, seven sub-deacons, forty-two acolyths, fifty-two exorcists, readers, and janitors.” 46 Raffaella Cribiore, Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt (Princeton, NJ: Princeton University Press, 2001), 189–190. See also Bruce M. Metzger, Manuscripts of the Greek Bible (Oxford: Oxford University Press, 1981), 20. 47 The reading aloud of texts in early Christian gathers is well documented. For instance, Justin Martyr Apol. 1.67, on the “day of the sun there is an assembly … memoirs of the apostles or the writings of the prophets are read for as long as time permits.” Also, the Muratorian Canon, mentions the public reading of the Pauline letters in the church, explicitly mentioning Titus (see Muratorian fragment trans. Bruce M. Metzger, The Canon of the New Testament: Its Origin, Development, and Significance (Oxford: Oxford University Press, 1987), 307). Other letters were also read in church. See Eusebius, Haer. 4.23.11; especially consider Dionysius of Corinth writing to Roman bishop Soter, about his previous letter. Wright (Communal Reading in the Time of Jesus: A Window into Early Christian Reading Practices [Minneapolis: Fortress, 2017], 181, 203) says concerning Titus, “Paul assumes this letter will be read communally. … Paul is fully aware of the social context in which his letters will be used: communal reading events.” See also Philip H. Towner, “The

282

Stevens

Amidst the tension of needing to have texts read communally and the difficulty of reading, the early church used the not altogether unique office of reader. Tertullian is the earliest to distinguish between the role of the ἀναγιγνώστης, lector, and reader.48 Cyprian mentions one entering into training for the “office of a reader.”49 Eusebius also lists various roles in the church including presbyters, deacons, sub-deacons, acolytes, exorcists, readers, and janitors.50 Likewise, Theodoret mentions a certain Julian being “enrolled in the order of Readers, and used to read aloud the sacred books to the people in the assemblies of the church.”51 Having a role within the church community for readers highlights the need to address the inherent difficulty in the act of communal reading. The scribes had a plethora of reasons motivating developing concessions for readers. In Christian MSS particularly, the codex form itself was an asset. The codices containing Christian texts typically used wider margins, fewer lines per page, large uniform letters, and spatial segmentation.52 A familiar example can illustrate the nexus between scribal design and reading aids. Hurtado has long argued that the “nomina sacra can be thought of as ‘hybrid’ phenomena that combine textual and iconographic features and functions, with particular sacred words presented in a special written form that was intended to mark them off from the surrounding text and express special reverence for them as visual signs.”53 The scribe set “apart these words visually in the way they are written,” which affects both the appearance and its reading.54 While not intending specifically to engage issues surrounding the nomina sacra, I believe that Hurtado is undoubtedly correct about the visual and spatial

48 49 50 51

52 53 54

Function of the Public reading of Scripture in 1 Timothy 4:13 and in the Biblical Tradition,” Southern Baptist Journal of Theology 7 (2003): 44–54. Pres. Praescr. 41 (ANF 3:263). Epist. of Cyprian 32 (ANF 5:312). Eusebius: Hist. eccl. 6.63 (NPNF2 1:288). Theodoret, Hist. eccl. 3.1 (NPNF2 3:94). A contrary position by Paul F. Bradshaw, Maxwell E. Johnson, and L. Edward Phillips (The Apostolic Tradition: A Commentary, Hermeneia [Minneapolis: Fortress, 2002], 74) contend, “we have no clear evidence for the existence of the office of reader before the third century.” They believe that the so-called reader in Justin in 1 Apol. 67.4 and 2 Clem. 19.1 is not an official position. In contrast, Wright (Communal Reading, 201) concludes his monograph on ancient reading by stating that “communal reading events were widespread in the first century.” Nässelqvist, Public Reading, 29–30; Larry W. Hurtado, “Oral Fixation and New Testament Studies? ‘Orality,’ ‘Performance’ and Reading Texts in Early Christianity,” NTS 60.3 (2014): 321–340, esp. 328–329; Gamble, Books, 229; Kruger, Canon Revisited, 234. Larry W. Hurtado, “Origin of the Nomina Sacra: A Proposal,” JBL 117 (1998): 655–673, here 672–673. Hurtado, “Origin,” 659.

Titus in P32 and Early Majuscules

283

consequences of the scribal design. The nomina sacra stand out by being shortened and having a macron over the letters. The highlighting of names and religiously significant matters by scribal design serves as a support for interpreting other scribal design choices for meaning and intentionality. Taken together, Scott Charlesworth contends that the large size, bilinearity of the margins, uniform columns, professional uncial hand, and overall clarity of the writing suggest that Sinaiticus was a public document. He believes it would be easier to read than many ancient MSS. Such features indicate that the early Christian MSS “were used in two general settings – publically in corporate worship, and privately by individuals.”55 Furthermore, Towner notes the obvious, “it is easier to read aloud to an assembly – a congregation – from a book in which the letters are written at a fair size than from one in which they are small.”56 While Charlesworth and Towner are mostly correct, it is unnecessary for our purposes to distinguish between public assembly and private reading, especially since the evidence does not clearly indicate such a dichotomy.57 It is better to follow Brian Wright and use the category of communal reading, which highlights both the social aspect of reading and defines the reading event as one in which two or more persons are involved. In other words, “communal reading can be public or private, but not individualistic.”58

55 Scott D. Charlesworth, “Public and Private – Second- and Third-Century Gospel Manuscripts,” in Jewish and Christian Scripture as Artifact and Canon, ed. Craig A. Evans and H. Daniel Zacharias (London: T&T Clark, 2009), 148–175, here 148. The background to public reading of Scripture is argued to be from the practices of the Jewish synagogue and stated in the New Testament at 1 Tim 4:13. See Towner, “Public Reading,” 44; Charlesworth, “Public and Private,” 149–150; Philo, Som. 2.127; Josephus, Ag. Ap. 2.175; Ahn, “Segmentation Features,” 31. While Dan Batovici (“The Reception of Early Christian Apocrypha and of the Apostolic Fathers: Reassessing the Late-Antique Manuscript Tradition and the Patristic Witnesses” [PhD dissertation, Katholieke Universiteit Leuven, 2015], 175) cautions against points of over interpretation by Charlesworth, the main point made here is these visual features do aid in the act of reading. 56 Turner, Typology, 84. 57 In a study of ancient Egyptian liturgical papyri, Ágnes T. Mihálykó (“Writing the Christian Liturgy in Egypt [Third to Ninth Century]” [PhD dissertation, University of Oslo, 2016], 16) finds “early Christians did not draw a clear distinction between ‘private prayer’ and ‘communal worship.’” 58 Wright, Communal Reading, 12. Similarly, Nässelqvist (Public Reading, 15) finds public and private reading should not be distinguished but considered a sliding scale: (a) “reading aloud for a single listener,” (b) “reading aloud in a small, controlled context, namely in a semi-private setting,” (c) “reading aloud in a limited, but less controlled context,” and (d) “reading aloud in a large-scale, open context.”

284

Stevens

The ancient scribes producing MSS like Sinaiticus developed features to aid in these communal settings. As stated at the beginning, Rouse is entirely correct that “there is a relationship between the physical form of an artifact and the function that it is meant to serve; therefore, the physical form of a manuscript is evidence of the purpose for which the manuscript was made.59” In a similar vein, Nässelqvist explains, “an important feature of reading culture is how the text appears to the lector. This includes its layout as well as the characteristic of the format in which it is found. … Since manuscript culture and reading culture are interconnected, the practical aspects and the physical appearance of manuscripts – and the texts found within them – affect how they are used for reading.”60 Moreover, Ahn concludes that “the scribes of the uncial and papyri MSS placed breaks for dramatic reading.”61 The conclusions are consistent with a growing consensus that scribal features like spatial segmentation were intended to aid in reading. 6

Interpretations and Proposals

Before proposing the motivation behind the short lines in Sinaiticus, it is essential to draw together their aesthetic consequences. First, similar to the design of the nomina sacra and ekthesis, the short lines of Titus and 2 Timothy visually draw attention to the text. The short lines function as the ancient equivalent of highlighted or underlined text. Second, the short portion is easier to find. When looking at the whole page, or two pages in codex form, the short lines are easily found. In a period without verse numbers, subheadings, or large paragraph breaks, turning to a specific location in a codex would be difficult, so making the short lines easier to locate further highlights them. A third consequence of the design is the short, uncluttered lines make reading easier. Having a single word without wrapping on one line removes the burden and responsibility from the reader to determine how to separate the letters into distinct words. Fourth, the scribe has endorsed if not outright prescribed a specific manner of reading. William Johnson notes that the medium of writing is inadequate at “conveying prosodic and paralinguistic features like tone of voice, facial expression, eye contact, body language, and other elements.”62 Consequently, standard 59 60 61 62

Rouse and Rouse, Authentic Witnesses, 4. Nässelqvist, Public Reading, 18. See also Mihálykò, “Writing the Christian Liturgy,” 112. Ahn, “Segmentation Features,” 91. William A. Johnson, “Toward a Sociology of Reading in Classical Antiquity,” AJP 121 (2000): 593–627, here 607.

Titus in P32 and Early Majuscules

285

written MSS do not contain a clear indication of the author’s locutionary act and illocutionary force.63 Therefore, it becomes the responsibility of the reader to function as a reader-performer in interpreting and projecting the illocutionary force by enacting prosodic and paralinguistic features.64 Even when one is cautious to not argue for a theatrical performance, reading aloud in the ancient world undoubtedly put, in the words of Johnson, a “surprising amount of the burden to interpret the text” on the reader.65 Consequently, by shortening the lines, the scribe signals the intended prosodic manner of reading, which is an interpretation by the scribe. The lines are to be read differently than the preceding and subsequent lines. Each line, or in this case each word, is intended to receive a full stop, just like the short lines before an ekthetic topic shift marker. The aural effect of individualized prosody presents each line in a more pronounced manner than reading in a continuous manner. Taking these features together, I propose the scribal layouts of Titus 1:7–8, 2 Tim 3:2–4, Gal 5:22–23, and 1 Cor 6:9–10 suggest the scribe designed the text for an ecclesiastical setting if not outright liturgical use. For Titus, the list of overseer qualifications, both negative and positive, is easy to find and easy to read, and the reading is in a didactic layout to give prosodic emphasis. The portion of text is appropriate for ordination or examination of an overseer (τὸν ἐπίσκοπον). Likewise, it could be easily read often to the overseers as an exhortation of their duties. Likewise, the 2 Timothy section, while not about qualifications for overseers, does discuss similar content, but in the negative. It serves as a warning against devolving morality in the last days (ἐν ἐσχάταις ἡμέραις) as a warning to the church. The scribal design there too makes it easy to find and read often. Its form serves to aid in an exhortational context. There is a distinctive challenge to this proposal, namely, if the design of the sections in Titus and 2 Timothy aid in ecclesiastical use, then why these passages and not others? The most obvious challenge to my proposal is that there are no short lines in the parallel list of elder qualifications in 1 Tim 3:2–3. The scribe does not have this same layout. Nothing at the location sets apart the qualifications list.

63 Johnson, “Toward a Sociology,” 607. 64 Utilizing categories from speech-act theory, Olson finds that written texts, especially in the ancient world, leave significant elements of interpretation to the reader. David R. Olson, The World on Paper: The Conceptual and Cognitive Implications of Reading and Writing (Cambridge: Cambridge University Press, 1994), 91–114. See also Johnson, “Toward a Sociology,” 607, 620; Hurtado, “Manuscripts and the Sociology,” 52. 65 Johnson, “Toward a Sociology,” 620; Hurtado, “Oral Fixation,” 321, 337.

286

Stevens

I propose a simple explanation. The scribes of Sinaiticus were making an experimentally large MS with new design features and did so in an inconsistent manner. The proposal has three primary strengths. First, the experimental nature of the oversized codex does explain the inconsistent use of spatial segmentation to highlight text. Second, the proposal is consistent with the history and possible provenance of Sinaiticus. Whether the oldest Greek majuscule was produced for the edict of Constantine is not needed to explore here, but what is clear is the form and design are new, inconsistent, and possibly, as proposed by some, a failure. Christfried Böttrich, following T. C. Skeat, sees the excessively large format of Sinaiticus as suggestive of being a test case.66 Skeat states in no uncertain terms that “I do not, therefore, think that there can be any doubt that it was the huge format of Sinaiticus which caused its abandonment.”67 Jongkind expands on the theory of Skeat and contends that three scribes were working simultaneously to speed up the production process; the scriptorium was, in his words, “experimenting with the correct approach.”68 A third support of the proposal is the lack of subsequent adoption. Other scribes do not adopt the experimental nature of the segmentation since they are working within more standard parameters of codex design and size. However, this idea does not serve to disprove the contentions of Sinaiticus serving as a master copy. Peter Cresswell has strongly contended for Sinaiticus serving as a master copy, especially evidenced by its size and extensive editing.69 In fact, the theory of being a master copy serves to support my proposal, as master copies are distinct. 66

Christfried Böttrich, “Codex Sinaiticus and the Use of Manuscripts in the Early Church,” ExpTim 128 (2017): 469–78, here 475. 67 T. C. Skeat, “The Codex Sinaiticus, The Codex Vaticanus and Constantine,” JTS 50 (1999): 583–625, here 613. Elliott agrees and believes that is why Sinaiticus was abandonded. See J. K. Elliott, “T. C. Skeat on the Dating and Origin of Codex Vaticanus,” in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup 113 (Leiden: Brill, 2004), 282–294, esp. 289, 293. 68 Dirk Jongkind, “One Codex, Three Scribes, and Many Books: Struggle with Space in Codex Sinaiticus,” in New Testament Manuscripts: Their Texts and their World, ed. Thomas J. Kraus and Tobias Nicklas, TENTS 2 (Leiden: Brill, 2006), 121–135, here 134–135. 69 Peter Cresswell, The Invention of Jesus: How the Church Rewrote the New Testament (London: Watkins, 2013), 84–88; Peter Cresswell, “Another Scribe or Another Exemplar? Examining Textual Patterns in Codex Sinaiticus Matthew and Mark,” JGRChJ 13 (2017): 111–128, here 125. See also Klaus Wachtel, “The Corrected New Testament Text of Codex Sinaiticus,” in Codex Sinaiticus: New Perspectives on the Ancient Biblical Manuscript, ed. Scot McKendrick, D. C. Parker, Amy Myshrall, and Cillian O’Hogan (Peabody, MA: Hendrickson, 2015), 97–106; Böttrich, “Codex Sinaiticus,” 475.

Titus in P32 and Early Majuscules

287

Admittedly, the precise nature of the motivation behind the scribal design cannot be verified or proven. Likewise, the proposal cannot be disproven, but it does account for much of the available evidence and is consistent with other theories about Sinaiticus as a whole. 7 Conclusion The reconstruction of P32 with supplementation from Sinaiticus is insightful for a number of factors as explored in this chapter. The reconstruction indicates P32 agrees with Sinaiticus in 99.85 percent of the letters and 99.1 percent of the words.70 Metzger, followed by James Royse, contends that P32 shows affinities with Sinaiticus.71 However, at 99.85 percent agreement, the contention of textual similarities appears to be an understatement. The textual relationship between the two is great enough to suggest P32, or at least its text, had a more direct relationship to Sinaiticus. It is likely that the text in P32 served as a potential parent in the genealogy of the corrected text in Titus as confirmed by the level of agreement, the addition of ἐν, the deletion of δέ, and the change in word order. For comparison, the degree of uniformity between P32 and Sinaiticus is eight times greater than the agreement between Alexandrinus and Sinaiticus for the same amount of text.72 Additionally, the reconstruction of the textual and examination of the paratextual features provide means for further study. First, the paratextual comparison with other MSS confirms that P32 was likely not the first text in a multi-text codex. The findings confirm the examination by Gathergood. Second, the paratextual comparison points to the scribal work in Sinaiticus as a creative act and one that was not subsequently adopted. The shortening of lines is concluded to be for the purpose of reading aids at both Titus 1:7–8 and 2 Tim 3:2–4. In accord with the rest of this volume, this chapter also confirms the importance of paratextual features and scribal design. Specifically, this chapter demonstrates that scribal features can be informatively compared with other MSS for greater insight into the unique decisions a particular scribe makes. The work here hopefully contributes to the understanding of P32 and Sinaiticus. To a greater extent, I hope to have encouraged paratextual examination through a comparative lens with contemporary MSS. 70

The reconstruction of P32 has 657 letters when the addition of δια is removed, and there are 114 whole words. 71 Metzger, Text of the New Testament, 250; Royse, “Early Text of Paul,” 192–193. 72 There are eight letters different in Alexandrinus (1:15; 2:3, 4, 7).

CHAPTER 11

The Scribal Use of Ekthesis as a Paragraph Marker? The Galatians Text in Codex Sinaiticus as a Test Case David I. Yoon 1 Introduction The earliest extant New Testament manuscripts do not contain chapter, verse, and section divisions as our modern Bibles do. Many of them, such as P46, are written in scriptio continua, without punctuation or other discourse markers,1 written simply from the left side of the folio to the right side continuously, sometimes marking only the end of a book and the beginning of a new one. Other manuscripts, however, contain markers of unit delimitation, such as paragraphos, ekthesis, and spacing. It appears that these markers start appearing in the third and fourth centuries; Codex Sinaiticus is one such manuscript that contains such features.2 Sinaiticus is one of the most significant early Greek manuscripts containing all of the books of the New Testament. Discovered in the nineteenth century by Constantine Tischendorf, Sinaiticus contains a significant portion of the Greek Old Testament, as well as apocryphal books, both Old Testament and New Testament apocrypha, as well as most of the New Testament.3 There is some confusion and debate regarding the circumstances surrounding its discovery, but it is accepted that Tischendorf is the figure responsible for 1 A few, however, such as P4, P66, P75, and P90, appear to have section divisions. Cf. Scott Charlesworth, “Indicators of ‘Catholicity’ in Early Gospel Manuscripts,” in The Early Text of the New Testament, ed. Charles E. Hill and Michael J. Kruger (Oxford: Oxford University Press, 2012), 43 (37–48). 2 Some of the other major, early manuscripts, such as Vaticanus and the Chester Beatty Papyri, ostensibly do not contain ekthesis. Codex Bezae contains ekthesis and Alexandrinus has a variation of it, where the first letter is significantly larger than the rest of the line of what seems to be the paragraph. 3 Constantine Tischendorf, Codex Sinaiticus: The Ancient Biblical Manuscript Now in the British Museum, 8th ed. (London: Lutterworth, 1934); see also Stanley E. Porter, Constantine Tischendorf: The Life and Work of a 19th Century Bible Hunter (New York: Bloomsbury, 2015); D. C. Parker, Codex Sinaiticus: The Story of the World’s Oldest Bible (Peabody, MA: Hendrickson, 2010).

© David I. Yoon, 2023 | doi:10.1163/9789004537972_013

The Scribal Use of Ekthesis as a Paragraph Marker?

289

publishing it.4 One notable distinctive feature of Sinaiticus is that it contains ekthesis throughout the codex, which on first impression appears to indicate paragraph divisions. In light of the number of ekthetic occurrences in Sinaiticus and their implication as potential paragraph markers, this chapter seeks to analyze the scribal use of ekthesis in Sinaiticus as compared with the linguistic description of paragraph and other levels of discourse and rank scale to see if these instances of ekthesis do indeed function as paragraph markers or markers of other levels or ranks. While the convention in biblical studies is to assume that ekthesis represents some sort of a paragraph marker, and this according to modern notions of the paragraph, there are very few studies that identify what a paragraph is and how ekthesis relates to it. This study seeks to confirm (or deny) the common assumption that ekthesis is a paragraph marker through analyzing the linguistic notion of a paragraph. Since analyzing every instance of ekthesis in Sinaiticus would require a monograph-length (or multi-monograph-length) study, I will apply my analysis to the text of Galatians as a test case. 2

Ekthesis, Rank Scale, and Levels of Discourse

Ekthesis describes the occurrence of a protruding line in the left margin of a manuscript, a hanging outdent.5 In other words, it is a reverse indentation (reverse of eisthesis), which is probably a more familiar textual feature for modern English readers. Occurrences of ekthesis may evoke the question of function, whether or not there is a consistent, determinable function of ekthesis by the scribe(s). Although ekthesis is discussed often as a well-observed scribal phenomenon, there has been no extensive study on its scribal function to date, especially in terms of its linguistic function.6 The presence of ekthesis may also evoke the idea of paragraphing, but it is important not to simply assume modern conventions and functions for ancient manuscripts. 4 The debate surrounds the fact of whether Tischendorf discovered these manuscripts at St. Catherine’s monastery in Sinai or whether he took them from the monks who knew about them. For the latter position, see Parker, Codex Sinaiticus, 129–32; for the former, see Porter, Constantine Tischendorf, 26–35. 5 William A. Johnson, Bookrolls and Scribes in Oxyrhynchus, Studies in Book and Print Culture (Toronto: University of Toronto Press, 2004), 342; see also E. G. Turner, Greek Manuscripts of the Ancient World (Princeton: Princeton University Press, 1971), 9, 14. 6 However, see Dirk Jongkind, Scribal Habits of Codex Sinaiticus, TS 5 (Piscataway, NJ; Gorgias, 2007), 95–109, who assumes paragraphing by use of ekthesis; James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill, 2008).

290

Yoon

The notions of the rank scale and levels of discourse are helpful for identifying a possible function of ekthesis. The rank scale was a part of Michael Halliday’s Scale and Category Grammar, which later developed into a fuller linguistic theory called Systemic Functional Linguistics (SFL). In SFL, the rank scale represents the units of grammar organized hierarchically, the lowest rank being morpheme, then word, then group, and the highest rank being clause (some may include clause complex or sentence as a higher rank; see below).7 From highest rank to lowest, the relationship between units is that of constituency. In other words, clause consists of items in the unit below it, such as word groups; word group consists of items below it, such as words; and word consisting of items below it, such as morphemes. In addition, rank shifting is when a unit is shifted to a lower rank, such as a word group functioning as a word. According to Christopher Butler, upward rank shifting, when a unit is shifted to a higher rank, does not occur.8 However, there are instances where upward rank shifting does in fact occur. In English, for example, “Pretty good” is itself a word group which is shifted to a higher rank to function as a clause or sentence, when in answer to a question such as, “How are you?” The rank scale is a description within the lexicogrammar, and thus it stops at its highest unit, the clause complex (regarding clause complex, see below). Levels of discourse, on the other hand, represent the stratum above the lexicogrammar, semantics, which are represented by logical relations rather than lexicogrammatical relations (explained further below in the section on clause complex and sentence).9 The base level of discourse is clause – clause is the basic unit of a discourse. Clause overlaps as the highest rank in the rank scale and the lowest level of discourse. Levels of discourse contain the clause, followed by clause complex and/or sentence, followed by paragraph, which is finally followed by (an entire) discourse or text. Although level is sometimes synonymous with rank or stratum, here I use it to refer to semantic units above the rank scale, which is found in the semantic stratum. For purposes of clarity, 7 M. A. K. Halliday, System and Function in Language: Selected Papers edited by G. R. Kress (London: Oxford University Press, 1976), 70–71; Christopher S. Butler, Systemic Linguistics: Theory and Application (London: Batsford, 1985), 16–17. Margaret Berry, Introduction to Systemic Linguistics: 1 Structure and Systems (London: Batsford, 1975), 104–106, offers a few different options to determine the rank scale, including complexes (such as clause complex or word complex). See below on clause complex. An SFL approach to rank scale is applied to textual criticism by Chris S. Stevens, History of the Pauline Corpus in Texts, Transmissions, and Trajectories: A Textual Analysis of Manuscripts from the Second to the Fifth Century, TENT 14 (Leiden: Brill, 2020), esp. 50–75. 8 Butler, Systemic Linguistics, 16–17. 9 On a concise description of stratification, see David I. Yoon, A Discourse Analysis of Galatians and the New Perspective on Paul, LBS 17 (Leiden: Brill, 2019), 71–72.

The Scribal Use of Ekthesis as a Paragraph Marker?

291

the definitions of clause, clause complex, sentence, and paragraph for Greek are provided below. 2.1 Clause The Greek clause is difficult to define, especially in comparison with the English clause, which is usually defined as a unit containing a subject and a predicate. But because Greek does not require a subject, having person and number encoded in the predicate, this definition does not fit. In addition, the Greek clause may arguably also not require a predicate. Few Greek grammars explicitly define a Greek clause. Daniel Wallace, however, defines a Greek clause as “units of thought forming part of a compound or complex sentence. Each clause normally contains a subject and predicate or a nonfinite verbal form (i.e., either an infinitive or participle).”10 This definition, however, is insufficient in that (1) it describes a clause in terms of being a part of something else, even if it identifies what it “normally contains” (subject and predicate), and (2) “unit of thought” is vague and can apply to almost any linguistic phenomenon, including paragraph. The transitivity network is useful for defining a Greek clause. Halliday, in defining the transitivity network, identifies various types of processes in the language and the structures that help express them.11 These elements include Process (usually realized by the predicate), Participant (usually realized by the subject and/or objects), and Circumstance (usually realized by adjuncts and complements). Among these three transitivity components, Circumstance is not a required element in the clause, but Process and Participant are necessary, at least in English. Considering the Greek language system, Participant can be encoded in the verbal system through person and number, and Process can be implied through ellipsis or adjectival predicate structure. Considering these factors, as well as the rank scale, a Greek clause can be defined as a unit of grammar above the word group level that contains one or more Participants, whether explicit or implicit from the co-text, and a Process, whether explicit or implicit from the co-text. Perhaps a simpler description is to state that a Greek clause is one that contains an explicit or implicit subject and an explicit or implicit predicate. But the transitivity network identifies semantic categories rather than grammatical ones (i.e., subject and predicate). 10 Daniel B. Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996), 656. 11 M. A. K. Halliday, Introduction to Functional Grammar, 1st ed. (London: Edward Arnold, 1985), 101–157. I use the first edition since it reflects Halliday’s early thoughts on functional grammar.

292

Yoon

There are three types of clauses identified in Greek: primary (or main), secondary (or relative or subordinate), and embedded. Primary clauses stand alone, secondary clauses are dependent on another clause (usually identified by a subordinate conjunction, participle, or infinitive), and embedded clauses are subordinate clauses that are embedded within another clause. 2.2 Clause Complex and Sentence Defining clause complex and sentence also proves to be difficult for Greek. Modern editions contain interpretive decisions due to the lack of punctuation in the earliest New Testament manuscripts (see Hans Förster’s chapter in this book). In English, a sentence is easily discernible: it begins with a capital letter and ends with a period (full stop) and has often been described as a “complete thought.” This, however, is a description mostly based on orthographic features and is not helpful for defining a sentence in terms of its grammatical and syntactic components, especially when considering an ancient language which evinces no such orthographic distinctions in the earliest manuscripts. Kenneth McKay describes a (Greek) sentence as containing at least one finite verb with either an expressed or implied subject in the nominative form but also indicates that verbs may be implied by usually a “to be” verb.12 Sentence, here for McKay, probably means clause, and his description is similar to mine above. But since “sentence” is largely orthographically defined, it is difficult to define a Greek sentence; orthographic demarcations are reflected in later editions of the Greek New Testament. On the other hand, the term clause complex was introduced by Halliday in his Functional Grammar (IFG) and is defined as “a Head clause together with other clauses that modify it.”13 In other words, a clause complex consists of a primary clause with secondary or embedded clauses attached to it, hypotactically. Regarding the difference between clause complex and sentence, although he functionally equates them, Halliday notes that a sentence is a constituent of writing, as it is defined by orthographic features (capital letter and full stop), while clause complex is a constituent of grammar, as it is defined in terms of grammatical categories. Thus, he prefers using the term clause complex over sentence. A problem with this definition of clause complex (and sentence), however, is that sentences and clause complexes, at least in English, may contain two or more primary clauses. For example, in the sentence “Jonny went to the park and Bill went to the store,” there are two paratactic clauses joined 12 K. L. McKay, A New Syntax of the Verb in New Testament Greek, SBG 5 (New York: Peter Lang, 1994), 5. 13 Halliday, Functional Grammar, 192.

The Scribal Use of Ekthesis as a Paragraph Marker?

293

together rather than one head clause with modifying (hypotactic) clauses. Can this be considered a clause complex according to Halliday’s definition above? Nguyen Thi Minh Tâm, on the other hand, sees a clause complex simply as a “combination of two or more clauses into a larger unit,”14 reflecting the logico-semantic combination of clauses. Nguyen identifies three basic types of clause complexes: (1) a primary clause with embedded clauses (i.e., simple sentence), (2) a primary clause with secondary clauses (i.e., complex sentence), and (3) two or more primary clauses (i.e., compound sentence). As clauses reflect different types – primary secondary, and embedded – it seems reasonable for clause complexes to contain different types, based on the relationships between the clauses which compose it – paratactic clause complexes and hypotactic clause complexes. But the problem that arises with this notion of clause complex for Greek is that clauses often begin with a conjunction (e.g., δέ or γάρ, which are paratactic), so paratactic clause complexes, by this definition, can be composed of a dozen clauses. This does not seem to be helpful in identifying clause complexes if a complex contains more than, say, three or four clauses. In light of this evaluation, it seems that Halliday’s original definition of a clause complex, as consisting of a head (primary) clause with other modifying (secondary or embedded) clauses attached to it, is suitable for Greek. And since sentence is usually defined orthographically, it seems best to avoid using this term for Greek, unless we are referring to a modern edition with punctuations (again, see Förster’s chapter in this book on punctuations as interpretative decisions by editors). Thus, a clause complex in Greek should be viewed as a complex of clauses consisting of a primary clause and any secondary or embedded clause attached to it. 2.3 Paragraph The paragraph, on the other hand, is even more challenging to define semantically. Clause and clause complex were identified above, since they are constituents of paragraph, and paragraph should be defined in light of its constituents. Like clause and possibly clause complex, language users tend to have a given or assumed understanding of what a paragraph is, and some have probably put forth a definition of a paragraph as something like a string of sentences that expound a single thought or key idea.15 But this definition is linguistically 14 Nguyεn Thι Minh Tâm, “The Notion of Clause Complex in Systemic Functional Linguistics,” Journal of Foreign Studies 29 (2013): 29 (25–36). 15 Louise J. Ravelli, “Signalling the Organization of Written Texts: Hyper-Themes in Management and History Essays,” in Analysing Academic Writing: Contextualized Frameworks,

294

Yoon

vague, like the definition of sentence being “a single thought or idea.” On the other hand, a paragraph often is – just like a sentence – identified simply orthographically, its primary feature being that it begins with an indentation and ends with empty space for the rest of the line. This, however, is not helpful either, since it simply identifies what a paragraph may look like, without defining its function(s). But what about its constituents define a paragraph as a paragraph? And what exactly is “a single thought or idea” (which is the same definition given for a sentence)? This definition is at best nebulous, at worst, meaningless, and specificity is needed in order to define a paragraph apart from orthographic features. It seems that attempts at more rigorous, linguistically based definitions of a paragraph are either too difficult or perhaps seemingly impossible; and so these studies are lacking.16 This may be due to the fact that most linguistic analyses have been traditionally limited to the sentence or clause level, but there has been a resurgence of discourse analysis in the past several decades, which focuses on a larger discourse (or paragraph) level. However, much work is still needed in developing discourse analysis in biblical studies, as much of it is still focused on singular items like pointing devices or connectives.17 A recent attempt at identifying the function(s) of a paragraph is by Stanley Porter, who draws from Halliday and Robert Longacre’s linguistic descriptions of a paragraph. Porter identifies seven general characteristics of a paragraph.18 (1) The presence of conjunctions, particles, and temporal and spatial references mark the beginning (and sometimes end) of the paragraph. The paragraph marker, in this case, can be a conjunctive device, or a reference to a new temporal or spatial transition in a narrative. (2) Paragraphs are characterized by cohesion and segmentation. They are cohesive units and paragraph breaks are reflected by segmenting these cohesive units. (3) Paragraphs contain participants, full reference, pronouns, and anaphora. The use of pronominal references to anaphoric participants and other referents indicate a paragraph unit. (4) Paragraphs usually begin with the fronting of the noun group and a high referential distance (length of continued reference to the noun group) ed. Louise J. Ravelli and Robert A. Ellis, Open Linguistics Series (New York: Continuum, 2004), 105–106 (104–130). 16 Cf. Stanley E. Porter, “Pericope Markers and the Paragraph: Textual and Linguistic Implications,” in The Impact of Unit Delimitation in Exegesis, ed. Raymond de Hoop, Marjo Korpel, and Stanley E. Porter, Pericope 7 (Leiden: Brill, 2009), 176–182. 17 Discourse analysis in biblical studies is still somewhat atomistic, although SFL has provided a way forward in developing tools to extract various categories of meaning (see Yoon, Discourse Analysis of Galatians, 59–60, 66–135). 18 The following is a summary of Porter, “Pericope Markers and the Paragraph,” 180–182.

The Scribal Use of Ekthesis as a Paragraph Marker?

295

throughout the paragraph. (5) Paragraphs usually reflect a single main topic and are distinguished from each other by topics shifts. The main topic is usually introduced in the beginning of the paragraph and developed throughout the rest of the paragraph. (6) Paragraphs also begin with an introduction of a thematic element (topic), and this is developed throughout the rest of the paragraph (comment).19 (7) Finally, literary text types play a role in the function of a paragraph, as different types may reflect different criteria for paragraphing. This list of main features of a paragraph may be considered to be on a cline from configurational to functional. After evaluating Mark and Romans in the text of Codex Sinaiticus, he concludes that ancient and modern notions of the paragraph have the following in common: initial conjunctive devices, cohesion and segmentation, topicality and thematization, grammaticality of reference, and word order and referential distance. Paragraphs also differ, depending on the literary text-type in which they are found (e.g., conjunctions having a greater role in paragraph transitions in narrative texts than expositional texts). Porter’s analysis of the paragraph is one of the most comprehensive yet concise ones, including both formal and semantic features of the paragraph, attending to the fact that text type (or genre) plays a role in paragraph formation. His use of cohesion, thematization, and topicality are especially helpful in defining a paragraph. Elizabeth Robar is another recent work that attempts to define a paragraph, albeit in biblical Hebrew.20 Although her study is not for Greek, her application of paragraph to Hebrew is helpful to consider for this study. Defining the paragraph as “the minimal complete discourse unit, namely, the linguistic reflection of the semi-active consciousness of a mental space,”21 using language that reflects a cognitive linguistic framework, she states that a paragraph contains coherence in structure and theme in the sense that every element in the paragraph relates to the theme in some way. As with previous definitions of paragraph, outside of orthographic features, she includes the idea that a paragraph contains a single theme or idea (or topic or thought).

19 Although Porter uses the terms theme and rheme to describe thematic elements at the paragraph level in “Pericope Markers and the Paragraph,” he later labels these as topic and comment for the paragraph/discourse level, and theme and rheme for the clause complex or sentence level. Cf. Stanley E. Porter and Matthew Brook O’Donnell, Discourse Analysis and the Greek New Testament (forthcoming). 20 Elizabeth Robar, The Verb and the Paragraph in Biblical Hebrew: A Cognitive-Linguistic Approach, SSLL 78 (Leiden: Brill, 2015). 21 Robar, Verb and Paragraph, 72.

296

Yoon

Finally, the most recent work on the paragraph and paragraphing is from Iain McGee.22 Approaching the issue from a more descriptive rather than prescriptive method, McGee provides one of the most comprehensive surveys of the paragraph, which includes both the paragraph unit and the paragraph break. This is of course from a modern perspective. Concluding his survey, he identifies some of the major functions of the paragraph: as a big period, a discourse managing technique, a highlighting technique, an aesthetic device, a formal grammatical unit between a sentence and an essay, an aid to text structuring, a cohesive unit, and a cognitive-easing device. This is a comprehensive description. McGee then offers three definitions of a paragraph, each reflecting the well-known text-reader-writer triad: (1) in the text-oriented definition, the paragraph is defined as an orthographic unit; (2) in the reader-oriented definition, it is defined as a unit that helps facilitate reading, expectation, and comprehension; (3) in the writer-oriented definition, it is defined in terms of the structuring of the text, including aesthetic reasons.23 McGee’s study is comprehensive and offers a descriptive summary of the various views on the paragraph, but it does not solve the question of a functional description of a paragraph – what the function(s) is/are of a paragraph as distinguished from other levels of discourse. And while drawing out the definition of a paragraph based on the three viewpoints of text, reader, and writer provides a multi-perspective, it seems to simply capture all aspects of how a paragraph has been conceived without any concluding synthesis. This brief survey of the paragraph illustrates the fact that narrowing the definition of the paragraph to a single or unified criterion is a difficult if not an impossible task, especially when trying to define it according to functional features. Very little consensus is reached on what a paragraph is, apart from orthographic features, and attempts to define a paragraph or identify features of a paragraph result in multiple criteria. Nevertheless, the paragraph is not completely nebulous, so the question remains: what makes a paragraph a paragraph? Drawing from this survey, I conclude that there are three minimal criteria for the Greek New Testament paragraph: (1) orthographic features delimit paragraphs, (2) paragraphs function in the semantic stratum (not in the lexicogrammar), so attempts to define a paragraph according to lexicogrammatical categories will not work, and (3) the function of the paragraph is more or less arbitrary (i.e., semi-arbitrary), based on the subjective decision of the writer (or editor or scribe) to divide the text into meaningful, cohesive units.

22 Iain McGee, Understanding the Paragraph and Paragraphing (Sheffield: Equinox, 2018). 23 McGee, Understanding the Paragraph, 349–53.

The Scribal Use of Ekthesis as a Paragraph Marker?

297

To elaborate the first point, orthographic features of a paragraph – for modern English, the indentation and the end-line space, and for Greek, ekthesis and paragraphos – are formal methods of delimiting the text, whatever the function might be (point 3), and so modern interpreters are probably justified in identifying ekthetic units in New Testament manuscripts as paragraphs. The problem is that the ancient paragraph is assumed to be parallel to the modern paragraph. This is where my second and third criteria provide further description. Regarding the second suggestion, since writers have almost unconstrained liberty to create paragraph divisions where they may – there are no “rules” or conventions as with clauses and other elements in the rank scale – it is a semantic feature rather than a lexicogrammatical one. This means that a paragraph consists not of formal features but of cohesive, prominent, and thematic features, all of which represent the textual metafunction and which function in the semantic stratum. Paragraphs are formed by the writer’s decision to group clauses and clause complexes into more or less cohesive units, sometimes using the paragraph division to reflect prominence of a part of a discourse and sometimes grouping clauses and clause complexes according to self-perceived thematic units. Paragraph divisions can have various functions, but the unifying feature of these functions is to facilitate the structure of a text according to cohesive, prominent, and thematic elements. Rather than being lexicogrammatical, the paragraph is semantic. Finally, the fact that there is no scholarly consensus on the semantic function of the paragraph points to the semi-arbitrary nature of the paragraph as a subjective choice by writers. In fact, this subjective choice of paragraph division is a meaning-making choice of the writer – or in this case, the scribe. Here is where McGee’s writer-oriented definition is helpful. Paragraph divisions are not completely arbitrary if a text is meaningful, but the assumption is that there is some motivation of writers (whether conscious or subconscious) to group the text into paragraphs in the way they have done so; hence, semi-arbitrary. This semi-arbitrariness of paragraph divisions can be illustrated from Halliday and Hasan’s comment that readers can expect to find a higher degree of cohesion within a paragraph than between paragraphs, and that in some instances, the reverse can also be true, where there is a high cluster of cohesive ties at the end of one paragraph and the beginning of the next, with relatively loose texture within the paragraph itself.24 To support this semi-arbitrary nature of paragraphing, Dirk Jongkind observes that the scribes of Sinaiticus seem to 24 M. A. K. Halliday and Ruqaiya Hasan, Cohesion in English, English Language Series (London: Longman, 1976), 296–97.

298

Yoon

vary in their use of ekthesis or other paragraph markers, especially considering variations in length and other criteria within a single book.25 He also observes that in Romans, which is densely paragraphed in Sinaiticus, certain signal words seem to be used to decide paragraph breaks, while in other books, such as Revelation, the frequency of paragraphs is relatively low. 2.4 Conclusion Considering the above criteria and definitions for the levels of discourse, including clause, clause complex, and paragraph, I will examine the placements of ekthesis in the text of Galatians in Sinaiticus (Galatians, for short) to analyze what the scribe was doing. A related issue involves the identity of the scribe(s) responsible for Sinaiticus. Tischendorf concluded, in spite of the relative uniformity in script throughout the entire manuscript, that there were four scribes that worked on Sinaiticus, identified as Scribe A, Scribe B, Scribe C, and Scribe D.26 It is asserted that Scribe A was responsible for a vast majority of the New Testament (excluding a few folios written by Scribe D), so the text of Galatians is considered to be entirely by one scribe, A. Thus, I will simply refer to “the scribe” as the responsible party for the Galatians text of Sinaiticus. 3

Ekthesis in Galatians27

There are 72 instances of ekthesis in the Galatians.28 This is compared to 19 paragraph markers in NA28 (more or less, depending on whether or 25 Jongkind, Scribal Habits, 108–109. 26 Constantin Tischendorf, ed., Novum Testamentum Sinaiticum: sive Novum Testamentum cum epistula Barnabae et fragmentis Pastoris (Leipzig: F. A. Brockhaus, 1863), xxi. See also Jongkind, Scribal Habits, 9–18; Peter M. Head, “The Presence of a Fourth Scribe?,” in Codex Sinaiticus: New Perspectives on the Ancient Manuscript, eds. Scot McKendrick, David C. Parker, Amy Myshrall, and Cillian O’Hogan (Peabody, MA: Hendrickson, 2015), 127 (127–38). See, however, H. J. M. Milne, and T. C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938), 22–50, who asserted Tischendorf was wrong about scribe C. 27 My analysis is based on my own observation of ekthesis in the electronic edition of Sinaiticus at www.codexsinaiticus.org. 28 Some of these instances might be debatable as to whether or not ekthesis exists, due to the shortness of length of some of them. The debate is usually resolved, however, by looking at the previous line and noting whether the line ends before the right margin. Both a protrusion of the left margin of the first line and the ending of the previous line before the right margin would signify the paragraph marker here.

The Scribal Use of Ekthesis as a Paragraph Marker?

299

not certain chapter divisions are new paragraphs) and 28 in UBSGNT5. Furthermore, when compared to the modern chapter-and-verse divisions in our modern Bibles, there is a total of 149 verses in Galatians to the 72 instances of ekthesis, which makes the ratio of ekthesis to verse roughly 1:2. More could be said about the chapter-and-verse editorial decisions that were made, but that is beyond the scope of this chapter. These statistics call into question whether or not ekthesis is indeed a paragraph marker, or whether they are markers of a lower level or rank, such as clause complex. It is readily apparent that ekthesis does not demarcate any element below the clause complex rank, such as clause, word group, or word, although rank shifting can occur. To determine whether ekthetic units are clause complexes or paragraphs is easily attainable by simply analyzing each unit. The first ekthetic unit in Galatians (1:1–5), the opening of the letter, consists of two clause complexes, the first beginning in v. 1 and the second beginning in v. 3. The second ekthetic unit (1:6–8) contains one long clause complex (vv. 6–8) and one short clause (v. 9). The third ekthetic unit (1:9) is a single verse containing two clause complexes. The fourth ekthetic unit (1:10), again a single verse, contains a clause, followed by a short clause complex, followed by a short clause. The fifth ekthetic unit (1:11–12) begins with a clause complex, followed by three clauses. The sixth ekthetic unit (1:13–14) begins a clause complex, which is interrupted by an ekthetic marker at v. 15 (οτε δε). In this case, the ekthesis breaks up the clause complex midway through. I can continue to analyze the rest of the ekthetic units, but it seems clear already that ekthetic units in Galatians do not demarcate clause complexes, so paragraph is the appropriate label for them, even if these paragraphs may not reflect modern paragraph conventions (hence, the semi-arbitrariness of paragraph formation). About 45 of the 72 (or 62.5%) of these ekthetic markers begin with conjunctions (including postpositives). Of these 45, 19 occurrences are with δε, 11 with γαρ, 4 with αλλα, and the rest of other conjunctions once or twice each, including και, ουτως και, αλλα τουναντιον, ουν, οτι δε, ωστε, and διο. These various conjunctions, as it can be seen, serve to connect in some way with the previous co-text, whether it be inferential, logical, causal, resultant, continuous, disjunction, or emphatic disjunction. This is related to Porter’s first criterion regarding conjunctions and particles (initial and final) and temporal and spatial references. They serve to indicate some sort of continuity or discontinuity with the previous paragraph. They also serve as cohesive devices in its connective function, whether continuous or discontinuous. The percentage here is slightly higher than Porter’s findings in Romans, where less than a third of the paragraphs used a major conjunction (και, δε, or γαρ) as a transition, although

300

Yoon

he did not count non-major conjunctions in his analysis.29 This means that roughly 27 of the 72 (or 37.5%) of the occurrences of ekthesis do not begin with any conjunctions. Of these 27, three begin with the temporal marker επειτα, (1:18; 1:21; 2:1) which is not a conjunction per se but an adverb and signifies continuity with the previous co-text as a temporal marker. There are also 11 indicative statements, seven imperative statements, and two conditional statements. Two other instances of ekthesis occur mid-verse (4:12; 5:13), both beginning with the nominative of address (or so-called vocative), αδελφοι, and another nominative of address occurs in 3:1 (ω ανοητοι γαλαται). The last is an interjection (ιδε; 5:2). It has been established that, at least in Galatians, a majority of ekthetic instances use conjunctive means to connect in some way the previous co-text, usually by way of a topic shift or a further development of the previous co-text, indicating some sort of transition. For example, ekthesis occurs in 4:1, 3, 6, and 8, all using a connective of some sort (δε, ουτως και, δε, and αλλα respectively). In 4:1, δε is used to transition the discussion from equality in Christ to heirship. In 4:3, ουτως καὶ is used to transition the discussion from heirship to sonship. In 4:6, δε is used to transition the discussion from sonship to the Spirit of sonship and back to how heirship relates. Then in 4:8, αλλὰ is used to transition the discussion to slavery and knowing God. While some of these subtopics can be subsumed under one paragraph (as might be expected in a modern paragraph), the scribal use of ekthesis can be an interpretive tool to understand the text. If it is accepted that ancient practices of paragraph divisions may be different than modern ones, this should not be a problem, and to equate the two would be an imposition of a modern conception onto an ancient one. Of the 27 instances of ekthesis without conjunctions or particles, most of them seem to signify some sort of topic or sub-topic shift that is consistent with a new paragraph. There are a few seeming exceptions, however, that warrant consideration. First is the ekthetic line in Gal 2:15: ημεις φυσει ιουδαιοι και ουκ εξ εθνων αμαρτωλοι (“we ourselves are Jews by birth and not of Gentiles”). This is an interesting occurrence because ekthesis here may be interpreted as signifying the end of Paul’s speech to Peter (as many interpreters have understood) and a return of direct discourse to the Galatians, but given the large number of ekthesis in this letter, it may signify simply a shift in topic.30 It is

29 Porter, “Pericope Markers and the Paragraph,” 187–88. 30 Cf. David I. Yoon, “Identifying the End of Paul’s Speech to Peter in Galatians 2: Register Analysis as a Heuristic Tool,” FN 28–29 (2015–2016): 57–79.

The Scribal Use of Ekthesis as a Paragraph Marker?

301

also interesting to note that even many English translations do not mark off a new paragraph here (in spite of a section heading). The next apparently dissonant occurrence of ekthesis is in Gal 3:6: καθως αβρααμ επιστευσεν τω θω (just as Abraham believed God). This is a dependent clause upon the rhetorical question that Paul asks: “Does he who supplies the Spirit to you and works miracles among you do so by works of the law, or by hearing with faith?” It seems that the use of ekthesis here is dissonant to what is expected of a (modern) paragraph marker in this verse, but it can be explained by the scribe’s desire to (1) highlight a Scripture reference, and (2) link 3:6 with the following co-text, where 3:7 describes Abraham again. Another ekthetic line occurs in 3:8, with another Scripture reference, and so possibly the scribe wanted to highlight this by using ekthesis. Another example is in Gal 4:12, where there is an instance of ekthesis “mid-verse” (as a total of two ekthetic lines in the verse). It reads: “–Become as I for I also as you – brothers, I urge you; no one has done me wrong” (–γινεσθε ως εγω οτι καγω ως υμις – αδελφοι δεομαι υμων ουδεν με), with the en-dash representing ekthesis.31 It seems fairly clear that the clause “brothers I urge you” refers to the previous clause “become as I,” so why did the scribe use ekthesis there? A possible answer is that the second ekthetic occurrence begins on a new folio, but this is not a consistent practice by the scribe to begin a new folio with ekthesis. Another possible answer is that the scribe wanted to highlight 4:12a, the imperative for the Galatians to become like Paul as he has become like them, so he placed it within its own paragraph. A final instance is actually an omission of ekthesis where it may be expected, especially if one has been exposed to modern editions of the Bible: Gal 5:1. There is no ekthesis at the beginning of this verse, but it does occur in the middle of the verse: τη ελευθερια ημας χριστος ημελυθερωσεν – στηκετε ουν και … (“for freedom Christ set you free – stand therefore and …”). The previous verse, however, consists of an indicative statement (we are not children of the slave woman but of the free), and thus 5:1a may cohere with 4:31 than with 5:1b, at least according to the scribe. Gal 5:1b contains an imperative, which may reflect why the scribe used ekthesis there. These are a few examples in which ekthesis may not at first glance seem to be used as a paragraph marker. However, it is best to take ekthesis as the scribe’s interpretation of the text and to view them as paragraph markers in light of their reflections of cohesion, thematization, and prominence.

31 Interestingly, B(03) also begins a new paragraph with αδελφοι. Many thanks to the anonymous reviewer of this manuscript in pointing this out.

302 4

Yoon

Conclusion: Implications for Understanding Galatians in Sinaiticus

Although not too much has been written on the function of ekthesis in ancient Greek manuscripts, there is a consistent observation that scribes did not have a standard use of them.32 This coheres with my notion of the “semi-arbitrary” paragraph. One suggestion for its function is that they were used for aesthetic or reading purposes. Larry Hurtado, for example, posits that visual aids such as ekthesis, eisthesis, and other markers reflect “efforts to facilitate the public/liturgical usage of texts.”33 In other words, their function is primarily stated as reading markers, and most have simply accepted this theory. And while this may be one potential function of ekthesis, is there more to it and other such features of the text (or paratext) than simply visual or oratory aids? If my suggestions above regarding a linguistic notion of a paragraph – (1) that orthographic features such as ekthesis, paragraphos, or eisthesis should be considered to be some sort of paragraph marker, (2) that paragraphs be considered in the semantic stratum rather than the lexicogrammar, and (3) that paragraphs are semi-arbitrary in that they are based on the writer’s subjective decision to reflect various features of cohesion, prominence, and thematization – as well as taking caution against misplacing modern notions of the paragraph onto ancient ones, then there really is no problem with viewing ekthesis as ancient paragraph markers. Many have simply assumed this, but this chapter has attempted to provide some methodological and analytical criteria for this thought. It is necessary, however, to go even further than this. If ekthesis is seen not simply as editorial decisions for aesthetic or oratory purposes but as an ancient paragraph marker, modern interpreters can understand a significantly early interpretation of the structure and texture of Galatians. While analyzing each ekthetic occurrence is beyond the scope of this chapter, one final example is provided here from 3:10–13. Each of these verses is an ekthetic unit of its own, while NA28 and UBSGNT5 combine them within the same paragraph, continuing to 3:14. The reason why the scribe of Sinaiticus divides each of these verses into its own ekthetic unit, apparently, is that each of these verses contains a Scripture reference (NA28 marks Scriptures references in italics and UBSGNT5 marks them in bold). So while these verses may, in their content, reflect a 32 Lance Jenott, The Gospel of Judas: Coptic Text, Translation, and Historical Interpretation of ‘the Betrayer’s Gospel’, STAC 64 (Tübingen: Mohr Siebeck, 2011), 118; Turner, Greek Manuscripts, 9, 14. 33 Larry Hurtado, Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 178.

The Scribal Use of Ekthesis as a Paragraph Marker?

303

continuous line of thought according to modern paragraphing conventions, the scribe seems to want to highlight (indicating prominence) these Scripture references through the use of ekthesis. So while the ancient scribe may not have practiced paragraphing the way modern editors might, there seems to be an interpretive schema in their use of ekthesis as paragraph markers. Potential for further analysis of scribal interpretation is promising given these criteria. Finally, it should be evident, but it is still worth mentioning, that modern Greek New Testament editions are interpretations of the Greek text in using not only punctuation (again, see Förster’s chapter) but paragraph and section divisions, not to mention section headings. These are all aids for readability and interpretation, to be sure. But as much as modern editions of the Greek New Testament attempt to capture the Ausgangstext, paratextual features such as paragraphing reflect modern interpretations of the text. Thus, for the interpreter of the Greek New Testament, it is necessary to identify editors’ interpretations as editors’ interpretations, whether ancient or modern, and use them to better understand the history of interpretation of the New Testament. Ekthesis is not only an aesthetic feature or an oratory aid but an interpretive tool to understand how the scribe interpreted the structure of the text. 1:1

6

9 10

Appendix: Ekthetic Units in the Galatians Text (NA28) of Sinaiticus Παῦλος ἀπόστολος οὐκ ἀπ’ ἀνθρώπων οὐδὲ δι’ ἀνθρώπου ἀλλὰ διὰ Ἰησοῦ Χριστοῦ καὶ θεοῦ πατρὸς τοῦ ἐγείραντος αὐτὸν ἐκ νεκρῶν, 2 καὶ οἱ σὺν ἐμοὶ πάντες ἀδελφοὶ ταῖς ἐκκλησίαις τῆς Γαλατίας, 3 χάρις ὑμῖν καὶ εἰρήνη ἀπὸ θεοῦ πατρὸς ἡμῶν καὶ κυρίου Ἰησοῦ Χριστοῦ 4 τοῦ δόντος ἑαυτὸν ὑπὲρ τῶν ἁμαρτιῶν ἡμῶν, ὅπως ἐξέληται ἡμᾶς ἐκ τοῦ αἰῶνος τοῦ ἐνεστῶτος πονηροῦ κατὰ τὸ θέλημα τοῦ θεοῦ καὶ πατρὸς ἡμῶν, 5 ᾧ ἡ δόξα εἰς τοὺς αἰῶνας τῶν αἰώνων, ἀμήν. Θαυμάζω ὅτι οὕτως ταχέως μετατίθεσθε ἀπὸ τοῦ καλέσαντος ὑμᾶς ἐν χάριτι [Χριστοῦ] εἰς ἕτερον εὐαγγέλιον, 7 ὃ οὐκ ἔστιν ἄλλο, εἰ μή τινές εἰσιν οἱ ταράσσοντες ὑμᾶς καὶ θέλοντες μεταστρέψαι τὸ εὐαγγέλιον τοῦ Χριστοῦ. 8ἀλλὰ καὶ ἐὰν ἡμεῖς ἢ ἄγγελος ἐξ οὐρανοῦ εὐαγγελίζηται [ὑμῖν] παρ’ ὃ εὐηγγελισάμεθα ὑμῖν, ἀνάθεμα ἔστω. ὡς προειρήκαμεν καὶ ἄρτι πάλιν λέγω· εἴ τις ὑμᾶς εὐαγγελίζεται παρ’ ὃ παρελάβετε, ἀνάθεμα ἔστω. Ἄρτι γὰρ ἀνθρώπους πείθω ἢ τὸν θεόν; ἢ ζητῶ ἀνθρώποις ἀρέσκειν; εἰ ἔτι ἀνθρώποις ἤρεσκον, Χριστοῦ δοῦλος οὐκ ἂν ἤμην.

304 11

Yoon

Γνωρίζω γὰρ ὑμῖν, ἀδελφοί, τὸ εὐαγγέλιον τὸ εὐαγγελισθὲν ὑπ’ ἐμοῦ ὅτι οὐκ ἔστιν κατὰ ἄνθρωπον· 12 οὐδὲ γὰρ ἐγὼ παρὰ ἀνθρώπου παρέλαβον αὐτὸ οὔτε ἐδιδάχθην, ἀλλὰ δι’ ἀποκαλύψεως Ἰησοῦ Χριστοῦ.  13 Ἠκούσατε γὰρ τὴν ἐμὴν ἀναστροφήν ποτε ἐν τῷ Ἰουδαϊσμῷ, ὅτι καθ’ ὑπερβολὴν ἐδίωκον τὴν ἐκκλησίαν τοῦ θεοῦ καὶ ἐπόρθουν αὐτήν, 14 καὶ προέκοπτον ἐν τῷ Ἰουδαϊσμῷ ὑπὲρ πολλοὺς συνηλικιώτας ἐν τῷ γένει μου, περισσοτέρως ζηλωτὴς ὑπάρχων τῶν πατρικῶν μου παραδόσεων.  15 Ὅτε δὲ εὐδόκησεν [ὁ θεὸς] ὁ ἀφορίσας με ἐκ κοιλίας μητρός μου καὶ καλέσας διὰ τῆς χάριτος αὐτοῦ 16 ἀποκαλύψαι τὸν υἱὸν αὐτοῦ ἐν ἐμοί, ἵνα εὐαγγελίζωμαι αὐτὸν ἐν τοῖς ἔθνεσιν, εὐθέως οὐ προσανεθέμην σαρκὶ καὶ αἵματι 17 οὐδὲ ἀνῆλθον εἰς Ἱεροσόλυμα πρὸς τοὺς πρὸ ἐμοῦ ἀποστόλους, ἀλλ’ ἀπῆλθον εἰς Ἀραβίαν καὶ πάλιν ὑπέστρεψα εἰς Δαμασκόν. 18 Ἔπειτα μετὰ ἔτη τρία ἀνῆλθον εἰς Ἱεροσόλυμα ἱστορῆσαι Κηφᾶν καὶ ἐπέμεινα πρὸς αὐτὸν ἡμέρας δεκαπέντε, 19 ἕτερον δὲ τῶν ἀποστόλων οὐκ εἶδον εἰ μὴ Ἰάκωβον τὸν ἀδελφὸν τοῦ κυρίου. 20 ἃ δὲ γράφω ὑμῖν, ἰδοὺ ἐνώπιον τοῦ θεοῦ ὅτι οὐ ψεύδομαι. 21 Ἔπειτα ἦλθον εἰς τὰ κλίματα τῆς Συρίας καὶ τῆς Κιλικίας· 22 ἤμην δὲ ἀγνοούμενος τῷ προσώπῳ ταῖς ἐκκλησίαις τῆς Ἰουδαίας ταῖς ἐν Χριστῷ. 23 μόνον δὲ ἀκούοντες ἦσαν ὅτι ὁ διώκων ἡμᾶς ποτε νῦν εὐαγγελίζεται τὴν πίστιν ἥν ποτε ἐπόρθει, 24 καὶ ἐδόξαζον ἐν ἐμοὶ τὸν θεόν. 2:1 Ἔπειτα διὰ δεκατεσσάρων ἐτῶν πάλιν ἀνέβην εἰς Ἱεροσόλυμα μετὰ Βαρναβᾶ συμπαραλαβὼν καὶ Τίτον· 2 ἀνέβην δὲ κατὰ ἀποκάλυψιν· καὶ ἀνεθέμην αὐτοῖς τὸ εὐαγγέλιον ὃ κηρύσσω ἐν τοῖς ἔθνεσιν, κατ’ ἰδίαν δὲ τοῖς δοκοῦσιν, μή πως εἰς κενὸν τρέχω ἢ ἔδραμον. 3 ἀλλ’ οὐδὲ Τίτος ὁ σὺν ἐμοί, Ἕλλην ὤν, ἠναγκάσθη περιτμηθῆναι· 4 διὰ δὲ τοὺς παρεισάκτους ψευδαδέλφους, οἵτινες παρεισῆλθον κατασκοπῆσαι τὴν ἐλευθερίαν ἡμῶν ἣν ἔχομεν ἐν Χριστῷ Ἰησοῦ, ἵνα ἡμᾶς καταδουλώσουσιν, 5 οἷς οὐδὲ πρὸς ὥραν εἴξαμεν τῇ ὑποταγῇ, ἵνα ἡ ἀλήθεια τοῦ εὐαγγελίου διαμείνῃ πρὸς ὑμᾶς. 6 Ἀπὸ δὲ τῶν δοκούντων εἶναί τι, – ὁποῖοί ποτε ἦσαν οὐδέν μοι διαφέρει· πρόσωπον [ὁ] θεὸς ἀνθρώπου οὐ λαμβάνει – ἐμοὶ γὰρ οἱ δοκοῦντες οὐδὲν προσανέθεντο, 7 ἀλλὰ τοὐναντίον ἰδόντες ὅτι πεπίστευμαι τὸ εὐαγγέλιον τῆς ἀκροβυστίας καθὼς Πέτρος τῆς περιτομῆς, 8 ὁ γὰρ ἐνεργήσας Πέτρῳ εἰς ἀποστολὴν τῆς περιτομῆς ἐνήργησεν καὶ ἐμοὶ εἰς τὰ ἔθνη, 9 καὶ γνόντες τὴν χάριν τὴν δοθεῖσάν μοι, Ἰάκωβος καὶ Κηφᾶς καὶ Ἰωάννης, οἱ δοκοῦντες στῦλοι εἶναι, δεξιὰς ἔδωκαν ἐμοὶ καὶ Βαρναβᾷ κοινωνίας, ἵνα ἡμεῖς εἰς τὰ ἔθνη, αὐτοὶ δὲ εἰς τὴν περιτομήν· 10 μόνον τῶν πτωχῶν ἵνα μνημονεύωμεν, ὃ καὶ ἐσπούδασα αὐτὸ τοῦτο ποιῆσαι.

The Scribal Use of Ekthesis as a Paragraph Marker?

11

305

Ὅτε δὲ ἦλθεν Κηφᾶς εἰς Ἀντιόχειαν, κατὰ πρόσωπον αὐτῷ ἀντέστην, ὅτι κατεγνωσμένος ἦν. 12 πρὸ τοῦ γὰρ ἐλθεῖν τινας ἀπὸ Ἰακώβου μετὰ τῶν ἐθνῶν συνήσθιεν· ὅτε δὲ ἦλθον, ὑπέστελλεν καὶ ἀφώριζεν ἑαυτὸν φοβούμενος τοὺς ἐκ περιτομῆς. 13 καὶ συνυπεκρίθησαν αὐτῷ [καὶ] οἱ λοιποὶ Ἰουδαῖοι, ὥστε καὶ Βαρναβᾶς συναπήχθη αὐτῶν τῇ ὑποκρίσει. 14 ἀλλ’ ὅτε εἶδον ὅτι οὐκ ὀρθοποδοῦσιν πρὸς τὴν ἀλήθειαν τοῦ εὐαγγελίου, εἶπον τῷ Κηφᾷ ἔμπροσθεν πάντων· εἰ σὺ Ἰουδαῖος ὑπάρχων ἐθνικῶς καὶ οὐχὶ Ἰουδαϊκῶς ζῇς, πῶς τὰ ἔθνη ἀναγκάζεις ἰουδαΐζειν;  15 Ἡμεῖς φύσει Ἰουδαῖοι καὶ οὐκ ἐξ ἐθνῶν ἁμαρτωλοί· 16 εἰδότες [δὲ] ὅτι οὐ δικαιοῦται ἄνθρωπος ἐξ ἔργων νόμου ἐὰν μὴ διὰ πίστεως Ἰησοῦ Χριστοῦ, καὶ ἡμεῖς εἰς Χριστὸν Ἰησοῦν ἐπιστεύσαμεν, ἵνα δικαιωθῶμεν ἐκ πίστεως Χριστοῦ καὶ οὐκ ἐξ ἔργων νόμου, ὅτι ἐξ ἔργων νόμου οὐ δικαιωθήσεται πᾶσα σάρξ. 17 εἰ δὲ ζητοῦντες δικαιωθῆναι ἐν Χριστῷ εὑρέθημεν καὶ αὐτοὶ ἁμαρτωλοί, ἆρα Χριστὸς ἁμαρτίας διάκονος; μὴ γένοιτο. 18 εἰ γὰρ ἃ κατέλυσα ταῦτα πάλιν οἰκοδομῶ, παραβάτην ἐμαυτὸν συνιστάνω. 19 ἐγὼ γὰρ διὰ νόμου νόμῳ ἀπέθανον, ἵνα θεῷ ζήσω. Χριστῷ συνεσταύρωμαι· 20 ζῶ δὲ οὐκέτι ἐγώ, ζῇ δὲ ἐν ἐμοὶ Χριστός· ὃ δὲ νῦν ζῶ ἐν σαρκί, ἐν πίστει ζῶ τῇ τοῦ υἱοῦ τοῦ θεοῦ τοῦ ἀγαπήσαντός με καὶ παραδόντος ἑαυτὸν ὑπὲρ ἐμοῦ. 21 Οὐκ ἀθετῶ τὴν χάριν τοῦ θεοῦ· εἰ γὰρ διὰ νόμου δικαιοσύνη, ἄρα Χριστὸς δωρεὰν ἀπέθανεν. 3:1 Ὦ ἀνόητοι Γαλάται, τίς ὑμᾶς ἐβάσκανεν, οἷς κατ’ ὀφθαλμοὺς Ἰησοῦς Χριστὸς προεγράφη ἐσταυρωμένος; 2 τοῦτο μόνον θέλω μαθεῖν ἀφ’ ὑμῶν· ἐξ ἔργων νόμου τὸ πνεῦμα ἐλάβετε ἢ ἐξ ἀκοῆς πίστεως; 3 οὕτως ἀνόητοί ἐστε, ἐναρξάμενοι πνεύματι νῦν σαρκὶ ἐπιτελεῖσθε; 4 τοσαῦτα ἐπάθετε εἰκῇ; εἴ γε καὶ εἰκῇ. 5 ὁ οὖν ἐπιχορηγῶν ὑμῖν τὸ πνεῦμα καὶ ἐνεργῶν δυνάμεις ἐν ὑμῖν, ἐξ ἔργων νόμου ἢ ἐξ ἀκοῆς πίστεως; 6 Καθὼς Ἀβραὰμ ἐπίστευσεν τῷ θεῷ, καὶ ἐλογίσθη αὐτῷ εἰς δικαιοσύνην· 7 γινώσκετε ἄρα ὅτι οἱ ἐκ πίστεως, οὗτοι υἱοί εἰσιν Ἀβραάμ. 8 προϊδοῦσα δὲ ἡ γραφὴ ὅτι ἐκ πίστεως δικαιοῖ τὰ ἔθνη ὁ θεός, προευηγγελίσατο τῷ Ἀβραὰμ ὅτι ἐνευλογηθήσονται ἐν σοὶ πάντα τὰ ἔθνη· 9 ὥστε οἱ ἐκ πίστεως εὐλογοῦνται σὺν τῷ πιστῷ Ἀβραάμ. 10 Ὅσοι γὰρ ἐξ ἔργων νόμου εἰσίν, ὑπὸ κατάραν εἰσίν· γέγραπται γὰρ ὅτι ἐπικατάρατος πᾶς ὃς οὐκ ἐμμένει πᾶσιν τοῖς γεγραμμένοις ἐν τῷ βιβλίῳ τοῦ νόμου τοῦ ποιῆσαι αὐτά. 11 ὅτι δὲ ἐν νόμῳ οὐδεὶς δικαιοῦται παρὰ τῷ θεῷ δῆλον, ὅτι ὁ δίκαιος ἐκ πίστεως ζήσεται· 12 ὁ δὲ νόμος οὐκ ἔστιν ἐκ πίστεως, ἀλλ’ ὁ ποιήσας αὐτὰ ζήσεται ἐν αὐτοῖς. 13 Χριστὸς ἡμᾶς ἐξηγόρασεν ἐκ τῆς κατάρας τοῦ νόμου γενόμενος ὑπὲρ ἡμῶν κατάρα, ὅτι γέγραπται· ἐπικατάρατος πᾶς ὁ κρεμάμενος ἐπὶ ξύλου, 14 ἵνα εἰς

306

Yoon

τὰ ἔθνη ἡ εὐλογία τοῦ Ἀβραὰμ γένηται ἐν Χριστῷ Ἰησοῦ, ἵνα τὴν ἐπαγγελίαν τοῦ πνεύματος λάβωμεν διὰ τῆς πίστεως. 15 Ἀδελφοί, κατὰ ἄνθρωπον λέγω· ὅμως ἀνθρώπου κεκυρωμένην διαθήκην οὐδεὶς ἀθετεῖ ἢ ἐπιδιατάσσεται. 16 τῷ δὲ Ἀβραὰμ ἐρρέθησαν αἱ ἐπαγγελίαι καὶ τῷ σπέρματι αὐτοῦ. οὐ λέγει· καὶ τοῖς σπέρμασιν, ὡς ἐπὶ πολλῶν ἀλλ’ ὡς ἐφ’ ἑνός· καὶ τῷ σπέρματί σου, ὅς ἐστιν Χριστός. 17 τοῦτο δὲ λέγω· διαθήκην προκεκυρωμένην ὑπὸ τοῦ θεοῦ ὁ μετὰ τετρακόσια καὶ τριάκοντα ἔτη γεγονὼς νόμος οὐκ ἀκυροῖ εἰς τὸ καταργῆσαι τὴν ἐπαγγελίαν. 18 εἰ γὰρ ἐκ νόμου ἡ κληρονομία, οὐκέτι ἐξ ἐπαγγελίας· τῷ δὲ Ἀβραὰμ δι’ ἐπαγγελίας κεχάρισται ὁ θεός. 19 Τί οὖν ὁ νόμος; τῶν παραβάσεων χάριν προσετέθη, ἄχρις οὗ ἔλθῃ τὸ σπέρμα ᾧ ἐπήγγελται, διαταγεὶς δι’ ἀγγέλων ἐν χειρὶ μεσίτου. 20 ὁ δὲ μεσίτης ἑνὸς οὐκ ἔστιν, ὁ δὲ θεὸς εἷς ἐστιν. 21 ὁ οὖν νόμος κατὰ τῶν ἐπαγγελιῶν [τοῦ θεοῦ]; μὴ γένοιτο. εἰ γὰρ ἐδόθη νόμος ὁ δυνάμενος ζῳοποιῆσαι, ὄντως ἐκ νόμου ἂν ἦν ἡ δικαιοσύνη· 22 ἀλλὰ συνέκλεισεν ἡ γραφὴ τὰ πάντα ὑπὸ ἁμαρτίαν, ἵνα ἡ ἐπαγγελία ἐκ πίστεως Ἰησοῦ Χριστοῦ δοθῇ τοῖς πιστεύουσιν. 23 Πρὸ τοῦ δὲ ἐλθεῖν τὴν πίστιν ὑπὸ νόμον ἐφρουρούμεθα συγκλειόμενοι εἰς τὴν μέλλουσαν πίστιν ἀποκαλυφθῆναι, 24 ὥστε ὁ νόμος παιδαγωγὸς ἡμῶν γέγονεν εἰς Χριστόν, ἵνα ἐκ πίστεως δικαιωθῶμεν· 25 ἐλθούσης δὲ τῆς πίστεως οὐκέτι ὑπὸ παιδαγωγόν ἐσμεν. 26 Πάντες γὰρ υἱοὶ θεοῦ ἐστε διὰ τῆς πίστεως ἐν Χριστῷ Ἰησοῦ· 27 ὅσοι γὰρ εἰς Χριστὸν ἐβαπτίσθητε, Χριστὸν ἐνεδύσασθε. 28 οὐκ ἔνι Ἰουδαῖος οὐδὲ Ἕλλην, οὐκ ἔνι δοῦλος οὐδὲ ἐλεύθερος, οὐκ ἔνι ἄρσεν καὶ θῆλυ· πάντες γὰρ ὑμεῖς εἷς ἐστε ἐν Χριστῷ Ἰησοῦ. 29 εἰ δὲ ὑμεῖς Χριστοῦ, ἄρα τοῦ Ἀβραὰμ σπέρμα ἐστέ, κατ’ ἐπαγγελίαν κληρονόμοι. 4:1 Λέγω δέ, ἐφ’ ὅσον χρόνον ὁ κληρονόμος νήπιός ἐστιν, οὐδὲν διαφέρει δούλου κύριος πάντων ὤν, 2 ἀλλ’ ὑπὸ ἐπιτρόπους ἐστὶν καὶ οἰκονόμους ἄχρι τῆς προθεσμίας τοῦ πατρός. 3 οὕτως καὶ ἡμεῖς, ὅτε ἦμεν νήπιοι, ὑπὸ τὰ στοιχεῖα τοῦ κόσμου ἤμεθα δεδουλωμένοι· 4 ὅτε δὲ ἦλθεν τὸ πλήρωμα τοῦ χρόνου, ἐξαπέστειλεν ὁ θεὸς τὸν υἱὸν αὐτοῦ, γενόμενον ἐκ γυναικός, γενόμενον ὑπὸ νόμον, 5 ἵνα τοὺς ὑπὸ νόμον ἐξαγοράσῃ, ἵνα τὴν υἱοθεσίαν ἀπολάβωμεν. 6 Ὅτι δέ ἐστε υἱοί, ἐξαπέστειλεν ὁ θεὸς τὸ πνεῦμα τοῦ υἱοῦ αὐτοῦ εἰς τὰς καρδίας ἡμῶν κρᾶζον· αββα ὁ πατήρ. 7 ὥστε οὐκέτι εἶ δοῦλος ἀλλ’ υἱός· εἰ δὲ υἱός, καὶ κληρονόμος διὰ θεοῦ. 8 Ἀλλὰ τότε μὲν οὐκ εἰδότες θεὸν ἐδουλεύσατε τοῖς φύσει μὴ οὖσιν θεοῖς· 9 νῦν δὲ γνόντες θεόν, μᾶλλον δὲ γνωσθέντες ὑπὸ θεοῦ, πῶς ἐπιστρέφετε πάλιν ἐπὶ τὰ ἀσθενῆ καὶ πτωχὰ στοιχεῖα οἷς πάλιν ἄνωθεν δουλεύειν θέλετε; 10 ἡμέρας

The Scribal Use of Ekthesis as a Paragraph Marker?

12

15 16 19 20 21

24 25 26 27 28 29 31 5:2

5 6

307

παρατηρεῖσθε καὶ μῆνας καὶ καιροὺς καὶ ἐνιαυτούς, 11 φοβοῦμαι ὑμᾶς μή πως εἰκῇ κεκοπίακα εἰς ὑμᾶς. Γίνεσθε ὡς ἐγώ, ὅτι κἀγὼ ὡς ὑμεῖς, ἀδελφοί, δέομαι ὑμῶν. οὐδέν με ἠδικήσατε· 13 οἴδατε δὲ ὅτι δι’ ἀσθένειαν τῆς σαρκὸς εὐηγγελισάμην ὑμῖν τὸ πρότερον, 14 καὶ τὸν πειρασμὸν ὑμῶν ἐν τῇ σαρκί μου οὐκ ἐξουθενήσατε οὐδὲ ἐξεπτύσατε, ἀλλ’ ὡς ἄγγελον θεοῦ ἐδέξασθέ με, ὡς Χριστὸν Ἰησοῦν. ποῦ οὖν ὁ μακαρισμὸς ὑμῶν; μαρτυρῶ γὰρ ὑμῖν ὅτι εἰ δυνατὸν τοὺς ὀφθαλμοὺς ὑμῶν ἐξορύξαντες ἐδώκατέ μοι. ὥστε ἐχθρὸς ὑμῶν γέγονα ἀληθεύων ὑμῖν; 17 ζηλοῦσιν ὑμᾶς οὐ καλῶς, ἀλλ’ ἐκκλεῖσαι ὑμᾶς θέλουσιν, ἵνα αὐτοὺς ζηλοῦτε· 18 καλὸν δὲ ζηλοῦσθαι ἐν καλῷ πάντοτε καὶ μὴ μόνον ἐν τῷ παρεῖναί με πρὸς ὑμᾶς. τέκνα μου, οὓς πάλιν ὠδίνω μέχρις οὗ μορφωθῇ Χριστὸς ἐν ὑμῖν· ἤθελον δὲ παρεῖναι πρὸς ὑμᾶς ἄρτι καὶ ἀλλάξαι τὴν φωνήν μου, ὅτι ἀποροῦμαι ἐν ὑμῖν. Λέγετέ μοι, οἱ ὑπὸ νόμον θέλοντες εἶναι, τὸν νόμον οὐκ ἀκούετε; 22 γέγραπται γὰρ ὅτι Ἀβραὰμ δύο υἱοὺς ἔσχεν, ἕνα ἐκ τῆς παιδίσκης καὶ ἕνα ἐκ τῆς ἐλευθέρας. 23 ἀλλ’ ὁ μὲν ἐκ τῆς παιδίσκης κατὰ σάρκα γεγέννηται, ὁ δὲ ἐκ τῆς ἐλευθέρας δι’ ἐπαγγελίας. ἅτινά ἐστιν ἀλληγορούμενα· αὗται γάρ εἰσιν δύο διαθῆκαι, μία μὲν ἀπὸ ὄρους Σινᾶ εἰς δουλείαν γεννῶσα, ἥτις ἐστὶν Ἁγάρ. τὸ δὲ Ἁγὰρ Σινᾶ ὄρος ἐστὶν ἐν τῇ Ἀραβίᾳ· συστοιχεῖ δὲ τῇ νῦν Ἰερουσαλήμ, δουλεύει γὰρ μετὰ τῶν τέκνων αὐτῆς. ἡ δὲ ἄνω Ἰερουσαλὴμ ἐλευθέρα ἐστίν, ἥτις ἐστὶν μήτηρ ἡμῶν· γέγραπται γάρ· εὐφράνθητι, στεῖρα ἡ οὐ τίκτουσα, ῥῆξον καὶ βόησον, ἡ οὐκ ὠδίνουσα· ὅτι πολλὰ τὰ τέκνα τῆς ἐρήμου μᾶλλον ἢ τῆς ἐχούσης τὸν ἄνδρα. Ὑμεῖς δέ, ἀδελφοί, κατὰ Ἰσαὰκ ἐπαγγελίας τέκνα ἐστέ. ἀλλ’ ὥσπερ τότε ὁ κατὰ σάρκα γεννηθεὶς ἐδίωκεν τὸν κατὰ πνεῦμα, οὕτως καὶ νῦν. 30 ἀλλὰ τί λέγει ἡ γραφή; ἔκβαλε τὴν παιδίσκην καὶ τὸν υἱὸν αὐτῆς· οὐ γὰρ μὴ κληρονομήσει ὁ υἱὸς τῆς παιδίσκης μετὰ τοῦ υἱοῦ τῆς ἐλευθέρας. διό, ἀδελφοί, οὐκ ἐσμὲν παιδίσκης τέκνα ἀλλὰ τῆς ἐλευθέρας. 5:1 Τῇ ἐλευθερίᾳ ἡμᾶς Χριστὸς ἠλευθέρωσεν· στήκετε οὖν καὶ μὴ πάλιν ζυγῷ δουλείας ἐνέχεσθε. Ἴδε ἐγὼ Παῦλος λέγω ὑμῖν ὅτι ἐὰν περιτέμνησθε, Χριστὸς ὑμᾶς οὐδὲν ὠφελήσει. 3 μαρτύρομαι δὲ πάλιν παντὶ ἀνθρώπῳ περιτεμνομένῳ ὅτι ὀφειλέτης ἐστὶν ὅλον τὸν νόμον ποιῆσαι. 4 κατηργήθητε ἀπὸ Χριστοῦ, οἵτινες ἐν νόμῳ δικαιοῦσθε, τῆς χάριτος ἐξεπέσατε. ἡμεῖς γὰρ πνεύματι ἐκ πίστεως ἐλπίδα δικαιοσύνης ἀπεκδεχόμεθα. ἐν γὰρ Χριστῷ Ἰησοῦ οὔτε περιτομή τι ἰσχύει οὔτε ἀκροβυστία ἀλλὰ πίστις δι’ ἀγάπης ἐνεργουμένη.

308 7

Yoon

Ἐτρέχετε καλῶς· τίς ὑμᾶς ἐνέκοψεν [τῇ] ἀληθείᾳ μὴ πείθεσθαι; 8 ἡ πεισμονὴ οὐκ ἐκ τοῦ καλοῦντος ὑμᾶς. 9 μικρὰ ζύμη ὅλον τὸ φύραμα ζυμοῖ. 10 ἐγὼ πέποιθα εἰς ὑμᾶς ἐν κυρίῳ ὅτι οὐδὲν ἄλλο φρονήσετε· ὁ δὲ ταράσσων ὑμᾶς βαστάσει τὸ κρίμα, ὅστις ἐὰν ᾖ. 11 Ἐγὼ δέ, ἀδελφοί, εἰ περιτομὴν ἔτι κηρύσσω, τί ἔτι διώκομαι; ἄρα κατήργηται τὸ σκάνδαλον τοῦ σταυροῦ. 12 Ὄφελον καὶ ἀποκόψονται οἱ ἀναστατοῦντες ὑμᾶς. 13 Ὑμεῖς γὰρ ἐπ’ ἐλευθερίᾳ ἐκλήθητε, ἀδελφοί· μόνον μὴ τὴν ἐλευθερίαν εἰς ἀφορμὴν τῇ σαρκί, ἀλλὰ διὰ τῆς ἀγάπης δουλεύετε ἀλλήλοις. 14 ὁ γὰρ πᾶς νόμος ἐν ἑνὶ λόγῳ πεπλήρωται, ἐν τῷ· ἀγαπήσεις τὸν πλησίον σου ὡς σεαυτόν. 15 εἰ δὲ ἀλλήλους δάκνετε καὶ κατεσθίετε, βλέπετε μὴ ὑπ’ ἀλλήλων ἀναλωθῆτε.  16 Λέγω δέ, πνεύματι περιπατεῖτε καὶ ἐπιθυμίαν σαρκὸς οὐ μὴ τελέσητε. 17 ἡ γὰρ σὰρξ ἐπιθυμεῖ κατὰ τοῦ πνεύματος, τὸ δὲ πνεῦμα κατὰ τῆς σαρκός, ταῦτα γὰρ ἀλλήλοις ἀντίκειται, ἵνα μὴ ἃ ἐὰν θέλητε ταῦτα ποιῆτε. 18 εἰ δὲ πνεύματι ἄγεσθε, οὐκ ἐστὲ ὑπὸ νόμον. 19 φανερὰ δέ ἐστιν τὰ ἔργα τῆς σαρκός, ἅτινά ἐστιν πορνεία, ἀκαθαρσία, ἀσέλγεια, 20 εἰδωλολατρία, φαρμακεία, ἔχθραι, ἔρις, ζῆλος, θυμοί, ἐριθεῖαι, διχοστασίαι, αἱρέσεις, 21 φθόνοι, μέθαι, κῶμοι καὶ τὰ ὅμοια τούτοις, ἃ προλέγω ὑμῖν, καθὼς προεῖπον ὅτι οἱ τὰ τοιαῦτα πράσσοντες βασιλείαν θεοῦ οὐ κληρονομήσουσιν. 22 ὁ δὲ καρπὸς τοῦ πνεύματός ἐστιν ἀγάπη χαρὰ εἰρήνη, μακροθυμία χρηστότης ἀγαθωσύνη, πίστις 23 πραΰτης ἐγκράτεια· κατὰ τῶν τοιούτων οὐκ ἔστιν νόμος. 24 οἱ δὲ τοῦ Χριστοῦ [Ἰησοῦ] τὴν σάρκα ἐσταύρωσαν σὺν τοῖς παθήμασιν καὶ ταῖς ἐπιθυμίαις. 25 Εἰ ζῶμεν πνεύματι, πνεύματι καὶ στοιχῶμεν. 26 μὴ γινώμεθα κενόδοξοι, ἀλλήλους προκαλούμενοι, ἀλλήλοις φθονοῦντες. 6:1 Ἀδελφοί, ἐὰν καὶ προλημφθῇ ἄνθρωπος ἔν τινι παραπτώματι, ὑμεῖς οἱ πνευματικοὶ καταρτίζετε τὸν τοιοῦτον ἐν πνεύματι πραΰτητος, σκοπῶν σεαυτὸν μὴ καὶ σὺ πειρασθῇς. 2 Ἀλλήλων τὰ βάρη βαστάζετε καὶ οὕτως ἀναπληρώσετε τὸν νόμον τοῦ Χριστοῦ. 3 εἰ γὰρ δοκεῖ τις εἶναί τι μηδὲν ὤν, φρεναπατᾷ ἑαυτόν. 4 τὸ δὲ ἔργον ἑαυτοῦ δοκιμαζέτω ἕκαστος, καὶ τότε εἰς ἑαυτὸν μόνον τὸ καύχημα ἕξει καὶ οὐκ εἰς τὸν ἕτερον· 5 ἕκαστος γὰρ τὸ ἴδιον φορτίον βαστάσει. 6 Κοινωνείτω δὲ ὁ κατηχούμενος τὸν λόγον τῷ κατηχοῦντι ἐν πᾶσιν ἀγαθοῖς. 7 Μὴ πλανᾶσθε, θεὸς οὐ μυκτηρίζεται. ὃ γὰρ ἐὰν σπείρῃ ἄνθρωπος, τοῦτο καὶ θερίσει· 8 ὅτι ὁ σπείρων εἰς τὴν σάρκα ἑαυτοῦ ἐκ τῆς σαρκὸς θερίσει φθοράν, ὁ δὲ σπείρων εἰς τὸ πνεῦμα ἐκ τοῦ πνεύματος θερίσει ζωὴν αἰώνιον.

The Scribal Use of Ekthesis as a Paragraph Marker?

9 11

14 15 17 18

309

τὸ δὲ καλὸν ποιοῦντες μὴ ἐγκακῶμεν, καιρῷ γὰρ ἰδίῳ θερίσομεν μὴ ἐκλυόμενοι. 10 Ἄρα οὖν ὡς καιρὸν ἔχομεν, ἐργαζώμεθα τὸ ἀγαθὸν πρὸς πάντας, μάλιστα δὲ πρὸς τοὺς οἰκείους τῆς πίστεως. Ἴδετε πηλίκοις ὑμῖν γράμμασιν ἔγραψα τῇ ἐμῇ χειρί. 12 Ὅσοι θέλουσιν εὐπροσωπῆσαι ἐν σαρκί, οὗτοι ἀναγκάζουσιν ὑμᾶς περιτέμνεσθαι, μόνον ἵνα τῷ σταυρῷ τοῦ Χριστοῦ μὴ διώκωνται. 13 οὐδὲ γὰρ οἱ περιτεμνόμενοι αὐτοὶ νόμον φυλάσσουσιν ἀλλὰ θέλουσιν ὑμᾶς περιτέμνεσθαι, ἵνα ἐν τῇ ὑμετέρᾳ σαρκὶ καυχήσωνται. Ἐμοὶ δὲ μὴ γένοιτο καυχᾶσθαι εἰ μὴ ἐν τῷ σταυρῷ τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ, δι’ οὗ ἐμοὶ κόσμος ἐσταύρωται κἀγὼ κόσμῳ. οὔτε γὰρ περιτομή τί ἐστιν οὔτε ἀκροβυστία ἀλλὰ καινὴ κτίσις. 16 καὶ ὅσοι τῷ κανόνι τούτῳ στοιχήσουσιν, εἰρήνη ἐπ’ αὐτοὺς καὶ ἔλεος καὶ ἐπὶ τὸν Ἰσραὴλ τοῦ θεοῦ. Τοῦ λοιποῦ κόπους μοι μηδεὶς παρεχέτω· ἐγὼ γὰρ τὰ στίγματα τοῦ Ἰησοῦ ἐν τῷ σώματί μου βαστάζω. Ἡ χάρις τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ μετὰ τοῦ πνεύματος ὑμῶν, ἀδελφοί· ἀμήν.

CHAPTER 12

Miniature Codices in Early Christianity Michael J. Kruger 1 Introduction In recent years, scholars have increasingly begun to recognize the importance of paratextual features of early Christian manuscripts.1 While no doubt the content of these manuscripts continues to be paramount, the various features of the physical medium by which these texts are delivered – e.g., scribal conventions, page layout, punctuation, abbreviations, quality of the hand – have helped illuminate the function of books within the early Christian movement. Rather than seeing manuscripts as merely disposable “husks” that carry an author’s story, it is becoming more evident that the visual features of these manuscripts can (and do) tell their own story.2 This chapter will focus on the small physical size of some early Christian Greek manuscripts, particularly the tiny format known as the miniature codex.3 This distinctive book format, though present in non-Christian circles 1 Harry Y. Gamble, Books and Readers in the Early Church (New Haven, CT: Yale University Press, 1995); Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006); Thomas J. Kraus, Ad Fontes: Original Manuscripts and Their Significance for Studying Early Christianity, TENTS 3 (Leiden: Brill, 2007); Thomas J. Kraus and Tobias Nicklas, eds., New Testament Manuscripts: Their Texts and Their World, TENTS 2 (Leiden: Brill, 2006); Michal J. Kruger, “Manuscripts, Scribes, and Book Production within Early Christianity,” in Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament, ed. Stanley E. Porter and Andrew W. Pitts, TENTS 9 (Leiden: Brill, 2012), 15–40. 2 Thomas Kraus and Tobias Nicklas, “The World of New Testament Manuscripts: ‘Every Manuscript Tells a Story.’” in New Testament Manuscripts, 1–11. 3 Some prior studies on the miniature codex include, Michael J. Kruger, “P.Oxy. 840: Amulet or Miniature Codex?,” JTS 53 (2002): 81–94; Thomas J. Kraus, “P.Oxy. V 840–Amulett oder Miniaturkodex? Grundsätzliche und ergänzende Anmerkungen zu zwei Termini,” ZAC 8 (2004): 485–497 (ET Kraus, “P.Oxy. V 840–Amulet or Miniature Codex? Principal and Additional Remarks on Two Terms,” in Ad Fontes, 47–67); Kraus, “Die Welt der Miniaturbücher in der Antike und Spätantike: Prolegomena und erste methodische Annäherungen für eine Datensammlung,” SNTU 35 (2010): 79–110; Kraus, “Miniature Codices in Late Antiquity: Preliminary Remarks and Tendencies about a Specific Book Format,” EC 7 (2016): 134–152. For more general comments, see L. Amundsen, “Christian Papyri from the Oslo Collection,” Symbolae Osloenses 24 (1945): 121–147, at 126–128; E. G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977), 22, 25, 29–30; C. H. Roberts,

© Michael J. Kruger, 2023 | doi:10.1163/9789004537972_014

Miniature Codices in Early Christianity

311

as early as the first century,4 began to be used by Christians by the third century and reached the height of its popularity in the fourth and fifth centuries (though it continued to be used thereafter).5 So established was this format among Christians during this time period that C. H. Roberts once surmised (incorrectly, it turns out) that, “the miniature codex would seem to be a Christian invention.”6 Many of these tiny books were elegant, well-crafted, and could contain a surprising number of pages.7 According to the catalog of Thomas Kraus, we currently have 30 such codices, which contain a wide range of Christian literature including New Testament books, patristic, and non-canonical writings (e.g., the Didache, Acts of Paul, and even apocryphal gospels).8 The popularity of the miniature codex raises a number of intriguing questions for our understanding of early Christian literature. Why were books put in this format at all? Was this format preferred for some types of literature and not others? What does this format say about the owner’s socio-cultural context? And, of course, before we can even address these sorts of questions, we must determine what counts as a miniature codex in the first place. What are the characteristics that define this book format? Needless to say, such questions cannot be answered at length in a single chapter such as this one. But the intent here is to provide an introductory overview, which hopefully can lay the groundwork for further research and discussion.

4

5 6 7 8

Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University Press, 1979), 10–12; Gamble, Books and Readers, 235–237; Hurtado, Earliest Christian Artifacts, 160–163. C. H. Roberts and T. C. Skeat, The Birth of the Codex (London: Oxford University Press, 1987), 27. The Roman poet Martial, writing ca. 84–86, mentions that his own poems could be obtained in codices with small parchment pages (brevibus membrana tabellis) that could be held in one hand (Ep. 1.2). Kruger, “P.Oxy. 840,” 90. Roberts, Manuscript, 12. Gamble (Books and Readers, 236) offers a more balanced assessment when he states, “The miniature format was, if not a uniquely Christian phenomenon, one heavily favored by the Christians.” E.g., P.Ant. 12, a fifth-century codex of 2 John, has numbers 164 and 165 in the upper margins. Kraus, “Die Welt der Miniaturbücher,” 98–99; Kraus, “Miniature Codices,” 144–145. The number 30 does not include Old Testament texts, nor codices in Coptic (though it does include some in Latin). There are a number of manuscripts that could be added to Kraus’s list (depending on whether one applies the 10 cm. limit to height as well as breadth), including P.Oxy. 849 (Acts of Peter), P.Grenf. 1.8 (Protevangelium of James), P.Vindob. G. 39212/GA 0217 (John), P.Vindob. G. 39780/GA 0101 (John); P. Rainer 4.45/MPER N. S. 29 51 (2 Corinthians), P.Oxy. 1602 (Acts of Paul), and P.Oxy. 401 (Matthew). For an example of an interesting Coptic miniature codex, see Annemarie Luijendijk, Forbidden Oracles? The Gospel of the Lots of Mary, STAC 89 (Tübingen: Mohr Siebeck, 2014).

312 2

Kruger

Characteristics of Miniature Codices

Given the fluidity of book production within the early Christian world, it is important that we acknowledge the complexity of defining a particular format like “miniature codex.” While some documents would have fit neatly into this category, others would have not – and still others would have hovered on the border. As we shall see below, there is a degree of fuzziness to any modern categorizations of ancient books. That said, we can still observe the broader characteristics of miniature codices, particularly as it pertains to their form and content. 2.1 Form I deal first with the physical form of the miniature codex which involves size (a width of 10 cm. or less) and the traditional codex format with writing on both sides of the page. 2.1.1 Size As for the physical form of miniature codices, we begin with their most obvious feature: tiny size. Since early Christian codices often exceeded 20 cm. in either height or width – e.g., P459 (20.4 × 25.4 cm.10), P5211 (18 × 21.3 cm.), P75 (13 × 26 cm.) – a smaller format would have stood out as distinctive. Historically, papyrologists have used 10 cm. in width (reconstructed size) as the outside range for what counts as “miniature.”12 Of course, this number is obviously arbitrary and scholars have rightly pointed out the artificial nature of its limitations.13 But the definition of miniature requires some number, and thus 10 cm. remains a convenient starting place for any discussion (as long as it is not enforced too rigidly).14

9

For more on P45, see T. C. Skeat, “A Codicological Analysis of the Chester Beatty Papyrus Codex of the Gospels and Acts (P45),” Hermathena 155 (1993): 27–43. 10 Following Turner’s practice (Typology of the Codex, 14–22), we list width before height. 11 For a helpful introduction to P52, see C. H. Roberts, “An Unpublished Fragment of the Fourth Gospel in the John Rylands Library,” BJRL 20 (1936): 45–55. 12 Turner, Typology of the Codex, 22, 29. Kraus (“Miniature Codices,” 136 n. 7) argues that this standard must have an older history because it also appears in J. van Haelst, Catalogue des papyrus littéraires juifs et chrétiens (Paris: Publications de la Sorbonne, 1976). 13 Kraus, “P.Oxy. V 840” (English), 57–58. 14 For instance, P.Oxy. 1602 (fourth/fifth century copy of Acts of Paul) is 10.5 cm. wide! This codex is not on Kraus’s list, though he includes codices much wider (e.g., P.Lond.Lit. 216 [1 Cor] is 15 cm. × 15 cm.).

Miniature Codices in Early Christianity

313

2.1.2 Codex The second physical feature that distinguishes miniature codices is that they are in the typical codex book format.15 The codex was created by taking a stack of papyrus or parchment leaves, folding them in half, and binding them at the spine.16 This format formed a traditional leaf book with writing on the recto and the verso.17 We even know of a number of miniature codices that are multi-quire, including P.Oxy. 1782, a fourth century copy of the Didache whose sheets were folded into a quire that would have originally held eight pages (the total number of quires is unknown).18 These two physical features – size and codex format – allow us to distinguish miniature codices from other tiny documents in the ancient world (though, as we shall see, this distinction is not absolute). For one, these features distinguish miniature codices from miniature scrolls which are rolled up with writing on only one side.19 More importantly, they distinguish miniature codices from most (though not all) amulets.20 Despite scruples from some church 15 In addition to classics like Turner and Roberts and Skeat, other overviews of the codex include A. Blanchard, ed., Les début du codes (Turnhout: Brepols, 1989); S. R. Llewelyn, “The Development of the Codex,” NewDocs 7 (1994): 249–256; Gamble, Books and Readers, 49–81; Hurtado, Earliest Christian Artifacts, 43–93; G. N. Stanton, “Why Were Early Christians Addicted to the Codex?” in Jesus and Gospel (Cambridge: Cambridge University Press, 2004), 165–191; Eldon J. Epp, “The Codex and Literacy in Early Christianity at Oxyrhynchus: Issues Raised by Harry Y. Gamble’s Books and Readers in the Early Church,” in Critical Reivew of Books in Religion 1997 (Atlanta: American Academy of Religion and Society of Biblical Literature, 1997), 15–37. 16 It is worth noting that the vast majority of miniature codices (approximately eighty percent) are on parchment, whereas the reverse is true for amulets (approximately seventy percent are on papyrus). See Kruger, “P.Oxy. 840,” 85–91 (these numbers are confirmed by Kraus’s list where 23 out of 30 miniature codices are parchment; “Miniature Codices,” 144–145). 17 I am using the terms “recto” and “verso” not as a reference to flesh/horizontal or the hair/vertical side of the manuscript, but simply as reference to the front and back of the folio (see Gamble, Books and Readers, 265 n. 9). 18 Lincoln H. Blumell and Thomas A. Wayment, eds., Christian Oxyrhynchus: Texts, Documents, and Sources (Waco, TX: Baylor University Press, 2015), 282–285. P.Grenf. 1.8 (Protevangelium of James) also appears to have an eight-page quire. Of course, there is also the famous Mani codex which had eight quires of twenty-four pages each, for a total of 192 pages: A. Henrichs and L. Koenen, “Ein Griechischer Mani-Codex (P. Colon. Inv. Nr. 4780),” ZPE 5 (1970): 97–216. 19 For more on miniature rolls, see W. Schubart, Das Buch bei den Griechen und Römern (Heidelberg: Schneider, 1962), 49–51; Frederic G. Kenyon, Books and Readers in Ancient Greece and Rome (Oxford: Clarendon, 1932), 49; and Kraus, “Die Welt der Miniaturbücher,” 96–97. 20 For more on Christian amulets, see Joseph E. Sanzo, Scriptural Incipits on Amulets from Late Antique Egypt, STAC 84 (Tübingen: Mohr Siebeck, 2014); Brice C. Jones, New Testament

314

Kruger

authorities,21 amulets were popular within early Christianity and were often hung or carried on the body in hopes of offering protection from things such as illness, bad dreams, demonic activity, or even wild animals.22 While many papyrus or parchment amulets were quite small, the vast majority of them are not in a codex format.23 This is evidenced by the fact that the verso is typically blank24 or contains unrelated content from a different hand (indicating either an opisthograph or some other reused manuscript).25 For instance, PSI 6.719 has the incipits of the four gospels, Ps 90:1, and some other texts on one side, and has an unrelated Byzantine protocol on the other (suggesting it is a portion of a roll).26 As another example, P.Amh. 1.3 (P12) has the text of Gen 1:1–5 on one side, and an unrelated Christian letter on the other (with Heb 1:1 in the upper margin).27 And Van Haelst 121 (LDAB 3142) has an inscription of Ps 19:7–8 on the one side and unrelated (and largely illegible) scribbling on the other.28

21 22

23 24 25

26 27

28

Texts on Greek Amulets from Late Antiquity, LNTS 554 (London: Bloomsbury T&T Clark, 2016); and T. de Bruyn, Making Amulets Christian: Artefacts, Scribes, and Contexts (Oxford: Oxford University Press, 2017). E.g., Synod of Laodicea, 36; Augustine, Doct. chr. 2.29. E.g., Christian amulets against fever include P.Princ. II 107; P.Oxy. 924; and P.Oxy. 1151. See also P.Princeton 159: Bruce M. Metzger, “A Magical Amulet for Curing Fevers,” in Studies in the History and Text of the New Testament (Salt Lake City: University of Utah Press, 1967), 89–94; and P.L.Bat. 20: G. H. R. Horsley, “Creedal Formula in a Christian Amulet Against Fever,” NewDocs 3 (1983): 114–119. Examples of other ailments include healing eye pain (P.Berl. 21911), protection from pain/distress (P.Rainer 5[13b]), and protection of a house from vermin (P.Oxy. 1060). For more such texts (in Greek and Coptic), see Marv W. Meyer and Richard Smith, Ancient Christian Magic: Coptic Texts of Ritual Power (San Francisco: Harper, 1994). Kruger, “P.Oxy. 840,” 86–87. A standard example is P.Vindob. G. 348, a papyrus amulet containing the four incipits of the canonical Gospels and Psalm 90 on one side, but nothing on the other. For more on this amulet, see Robert W. Daniel, “A Christian Amulet on Papyrus,” VC 37 (1983): 400–404. In the list by de Bruyn of documents that were either “certainly” or “probably” amulets, approximately seventy percent had either a blank verso or unrelated writing on the verso. See T. de Bruyn, “Papyri, Parchments, Ostraca, and Tablets Written with Biblical Texts in Greek and Used as Amulets: A Preliminary List,” in Early Christian Manuscripts: Examples of Applied Method and Approach, ed. Thomas J. Kraus and Tobias Nicklas, TENTS 5 (Leiden: Brill, 2010), 145–189, at 166–174. See discussion in Jones, Greek Amulets, 77–80. The status of P.Amh. 1.3 as an amulet has been challenged: Claire Clivaz, “The New Testament at the Time of the Egyptian Papyri: Reflection Based on P12, P75 and P126 (P.Amh. 3b, P.Bod. XIV–XV and PSI 1497),” in Reading New Testament Papyri in Context, ed. Claire Clivaz and J. Zumstein (Leuven: Peeters, 2011), 15–55. Robert A. Kraft and Antonia Tripolitis, “Some Uncatalogued Papyri of Theological and Other Interest in the John Rylands Library,” BJRL 51 (1968): 137–149, at 138–142.

Miniature Codices in Early Christianity

315

2.2 Content If the first distinguishing feature of miniature codices is their physical form (codex, 10 cm. wide or less), we now turn to the second feature, namely the kind of content they typically contain. Again, these features are not absolute. We only offer a general pattern here – as we will see, not all miniature codices share the same type of content. 2.2.1 Continuous Text The content of miniature codices is often characterized by continuous texts that begin on a prior page and continue onto future ones.29 Thus, it is not unusual for miniature codices to be fairly lengthy (sometimes even with page numbers in the upper margins), whereas amulets are comparably very brief. For example, P.Ant. 12, a miniature codex of 2 John dated between the third and fifth century, has the pages 164 and 165 in the upper margins.30 This would have been a sizeable codex, perhaps containing the other Catholic letters and the book of Hebrews.31 Other examples include: (a) P.Oxy. 1010 (fourth century), a continuous text of 4 Ezra (6 Ezra) with page 40 in the upper margin of the recto;32 (b) P.Oxy. 1080 (fourth/fifth century) a continuous text of Revelation with pages 33 and 34 in the upper margins;33 (c) P.Oxy. 840 (fourth century) an apocryphal gospel with a continuous text that began on prior pages and would have continued onto future ones;34 (d) P.Oxy. 849 (fourth century), a continuous text of the Acts of Peter with pages 167 and 168 in the upper margins;35 (e) PSI 1.5 (fourth/fifth century) is a continuous text of James with pages 17 and 18 in upper margins;36 (f) P.Vindob. G. 39756 + Bodl. MS Gr. th. f. 4 [P] 29 Of course, given the fragmentary nature of many of the miniature codices discussed below we cannot always be sure a text is continuous even if it spans multiple pages. For more on the definition of “continuous” and its limitations, see Jones, Greek Amulets, 25–41. 30 C. H. Roberts, ed., The Antinoopolis Papyri, Part I (London: Egypt Exploration Society, 1950), 24–26. 31 For a more detailed reconstruction of this miniature codex, see Michael J. Kruger, “The Date and Content of P. Antinoopolis 12 (0232),” NTS 58 (2012): 254–271. 32 A. S. Hunt, ed., The Oxyrhynchus Papyri, Vol. VII (London: Egypt Exploration Society, 1910), 11–15. There is no page number visible on the verso, but Hunt suggests this may be due to damaged condition of that side of the leaf. 33 Blumell and Wayment, Christian Oxyrhynchus, 181–184. Other miniature codices of Revelation include P.Oxy. 848 (fifth century) and P.Oxy. 4500 (third/fourth century). 34 Michael J. Kruger, The Gospel of the Savior: An Analysis of P.Oxy. 840 and Its Place in the Gospel Traditions of Early Christianity, TENTS 1 (Leiden: Brill, 2005). 35 B. P. Grenfell and A. S. Hunt, eds., The Oxyrhynchus Papyri, Vol. VI (London: Egypt Exploration Society, 1908), 6–12. 36 Blumell and Wayment, Christian Oxyrhynchus, 161, 179–181.

316

Kruger

(fifth century), a continuous text of the Apocalypse of Peter;37 (g) P.Ant. 13 (fourth century), a continuous text of the Acts of Paul;38 (h) P.Grenf. 1.8 (fifth/sixth century), a continuous text of the Protoevangelium of James.39 This feature of miniature codices stands in contrast to the content of most amulets which is typically non-continuous.40 Indeed, amulets are known for having a mish-mash of passages – often the Psalms, the Lord’s Prayer, incipits of the four Gospels – and also a variety of prayers and Christian symbols.41 And much of the content was chosen because it had an obvious medical, protective, or apotropaic purpose – again, something typically lacking in the content of miniature codices.42 This explains the popularity of Psalm and prayers in Christian amulets.43 These texts were typically used to protect, bless or heal the wearer of the amulet. Scribal Quality 2.2.2 In addition to (normally) having continuous texts, miniature codices often exhibit a competent scribal hand (and sometimes even standard scribal conventions; e.g., corrections, punctuation, color, coronis, ekthesis). Kraus has argued that such palaeographical features can help distinguish miniature 37 Thomas J. Kraus, “P.Vindob.G 39756 + Bodl. MS Gr. Th. f. 4 [P]: Fragmente eines Codex der griechischen Petrus-Apokalypse,” BASP 40 (2003): 45–61. 38 Roberts, ed., Antinoopolis Papyri, 26–27; Alan Mugridge, Copying Early Christian Texts: A Study of Scribal Practice, WUNT 362 (Tübingen: Mohr Siebeck, 2016), 300. See also P.Oxy. 6 (fifth century), another miniature codex of the Acts of Paul. 39 B. P. Grenfell, An Alexandrian Erotic Fragment and Greek Papyri Chiefly Ptolemaic (Oxford: Clarendon, 1896), 13–19. 40 de Bruyn, “A Preliminary List,” 149. There are quasi exceptions to this pattern, such as P.Oxy. 2684 (P78) which appeared to carry half of Jude (1–13), but Wasserman argues it was an amulet (see discussion below). This raises the question of whether half of a book is a substantive enough portion to count as “continuous” or whether it has to be the entire book. For more on that point, see Stanley E. Porter, “Why So Many Holes in the Papyrological Evidence for the Greek New Testament?” in The Bible as Book: The Transmission of the Greek Text, ed. Scot McKendrick and Orlaith A. Sullivan (New Castle: Oak Knoll, 2003), 167–186. It should also be acknowledged that non-continuous text does not automatically make a document an amulet; other options exists (e.g., lectionary). For example, see Amundsen, “Christian Papyri,” 121–147. 41 de Bruyn, “A Preliminary List,” 149. 42 de Bruyn, “A Preliminary List,” 147. 43 Psalm 90 was arguably the most popular Psalm in the amulets, and the Lord’s Prayer (in some version) was one of the most popular prayers. On the former, see Thomas Kraus, “Septuaginta: Psalm 90 in apotropäischer Verwendung: Vorüberlegungen für eine kritische Edition und (bisheriges) Datenmaterial,” BN 125 (2005): 39–73. On the latter, see Kraus, “Manuscripts with the Lord’s Prayer – They Are More than Simply Witnesses to That Text Itself,” in New Testament Manuscripts, 227–266.

Miniature Codices in Early Christianity

317

codices from amulets since the latter are typically “written in a coarse and inexperienced hand.”44 Thus, he argues that, “the classification ‘miniature codex’ … should actually be defined with the help of the palaeographical details of the manuscripts to be considered.”45 Roberts makes the same point, arguing that miniature codices “were often elegant and calligraphic …. They are best regarded not as amulets but as devotional handbooks for the well-to-do.”46 Of course, this observation is not without its exceptions – not all miniature codices share a high level of scribal quality,47 and even some amulets have a competent bookhand.48 Although there is not space here for a detailed analysis of the palaeography of all miniature codices, a brief sampling will establish Kraus’s point: P.Ant. 12 (2 John), has a hand which is “delicate and rounded of considerable individuality and elegance”;49 P.Oxy. 1080 (Revelation) has “good-sized upright uncial, fairly regular and having a certain ornamental finish”;50 P.Oxy. 1010 (4 Ezra) was “written by a competent scribe who knew what to do”;51 P.Oxy. 848 (Revelation) has “bold upright uncials … similar in style to those of the codex Alexandrinus”;52 P.Ant. 13 (Acts of Paul), is described as a “de luxe codex” where “the material … and the script, in its regularity and delicacy reminiscent of the great Biblical codices, are of exceptional quality”;53 P.Vindob. G. 39756 (Apocalypse of Peter) has a hand that “belongs to the Alexandrian type of the upright Biblical majuscule”;54 P.Oxy. 4500 (Revelation) is “written by a very skilled scribe in a rounded uncial that is largely bilinear”;55 P.Oxy. 840 (apocryphal gospel) is a deluxe codex (enlarged first letter, use of color, spaces

44 45 46 47

48 49 50 51 52 53 54 55

Kraus, “P.Oxy. V 840,” 53. Kraus, “P.Oxy. V 840,” 58. Roberts, Manuscript, 10–11. E.g., P.Berol. 1710 (unknown gospel); see Thomas Kraus, Michael J. Kruger, and Tobias Nicklas, eds., Gospel Fragments: The ‘Unknown Gospel’ on Papyrus Egerton 2, Papyrus Oxyrhynchus 840, Other Gospel Fragments, Oxford Early Christian Gospel Texts (Oxford: Oxford University Press, 2009), 228–230 (of course, it is not altogether clear this fragment is a miniature codex and not an amulet); see also P.Ryl. 463 (Gospel of Mary); see Christopher Tuckett, The Gospel of Mary (Oxford: Oxford University Press, 2007), 83–85. E.g., P.Yale. inv. 989; P.Vindob. G. 29831. Roberts, ed., Antinoopolis Papyri, 24. A. S. Hunt, ed., The Oxyrhynchus Papyri, vol. VIII (London: Egypt Exploration Society, 1911), 15. Kraus, “Miniature Codices,” 136. Grenfell and Hunt, eds., Oxyrhynchus Papyri, vol. VI, 6. Roberts, ed., Antinoopolis Papyri, 26. Kraus, “P.Oxy. V 840,” 56. Blumell and Wayment, Christian Oxyrhynchus, 161.

318

Kruger

between sentences) with a scribal hand in “upright pointed majuscule”;56 PSI 1.5 (James) has a palaeography that “lends the overall impression that it was a professional production.”57 3

Examples of Miniature Codices: Probing Borderline Cases

Thus far, we have seen the two kinds of characteristics that typically mark miniature codices and distinguish them from amulets: (1) their form (codex format, small size); and (2) their content (continuous text, high scribal quality). However, simply because these characteristics are typical does not mean they always hold. As noted above, book production in the ancient world was a fluid affair – and did not always conform to modern categories. As might be expected, we have some documents in the ancient world that are “hybrids” of sorts; they look like miniature codices in some ways and look like amulets in other ways. These sorts of documents typically have the physical form of a miniature codex but are filled with the content of an amulet. One might say that they are miniature codices on the outside and amulets on the inside. Such documents remind us that miniature codices and amulets, though usually distinguishable, sometimes overlap.58 Let us briefly consider a few examples. 3.1 P.Ant. 54 This tiny piece of papyrus (5.2 × 4 cm.) is folded in half to form a bifolium and is dated to the third/fourth century.59 It contains a small portion of the Lord’s 56 Kruger, Gospel of the Savior, 40–45. 57 Blumell and Wayment, Christian Oxyrhynchus, 179. 58 I continue to be confused by the charge of Kraus (followed by Jones) that my 2002 article created too sharp a distinction (what Kraus calls a “polarization”) between amulets and miniature codices (“P.Oxy. V 840,” 57; Jones, Greek Amulets, 120–122). I was quite plain in that article that I do not see these as absolute categories: “The boundaries between the categories of miniature codex and amulet are not absolute. The function of literature in the ancient world was fluid and ever-changing and we should not be surprised, therefore, if we occasionally find miniature codices that are amulets, or amulets that are in the form of miniature codices … [they] occasionally overlap” (“P.Oxy. 840,” 82; emphasis mine). I even provided additional clarity later: “these two categories are not mutually exclusive – it is possible (though rare) for a document to be both a codex and an amulet at the same time” (“P.Oxy. 840,” 91; emphasis mine). In light of these comments, I am equally confused about how Jones can say, “I contend that it is reductionistic to argue that an amulet must never be a miniature codex or vice versa” (Greek Amulets, 121; emphasis his). My comments above show that I have said no such thing. 59 J. W. B. Barns and H. Zilliacus, eds., The Antinoopolis Papyri, Part II (London: Egypt Exploration Society, 1960), 6–7.

Miniature Codices in Early Christianity

319

Prayer (Matt 6:10–12) on the first three pages – though the text stops abruptly in the middle of a word – and the last page is blank.60 Although the external form fits with the category of miniature codex (less than 10 cm. in breadth, codex form), the content fits with that of an amulet (Lord’s Prayer,61 unprofessional scribal hand, text breaks mid-word).62 This overlap might explain the wide range of scholarly opinion about how to categorize this manuscript, including child’s toy book,63 miniature notebook,64 a failed attempt at a miniature codex,65 and, of course, amulet.66 3.2 P.Oxy. 2684 (P78) This papyrus manuscript is another bifolium sheet folded in half to form an unusually shaped codex where the width is almost twice the height (5.3 × 2.9 cm.).67 Dated to the third/fourth century, the manuscript contains Jude 4–5, 7–8, but would have likely contained several more leaves depending on the details of reconstruction (single or multi-quire? Did it hold the entire book of Jude?). Wasserman argues that it was single-quire and originally contained just Jude 1–13, a passage, he argues, that has particularly apotropaic value consistent with its usage as an amulet.68 This possibility finds support in the poor scribal hand of the manuscript which is “informal documentary”69 or “hastily written semi-cursive,”70 along with an overall appearance which is “mundane.”71 Thus, 60 61 62

63 64 65 66 67

68 69 70 71

de Bruyn (Making Amulets, 163) explores the possibility that P.Ant. 54 may have originally contained additional pages. Kraus, “Manuscripts with the Lord’s Prayer,” 227–266. According to Jones, other amulets have texts that break off unexpectedly: P.Oxy. 5073; P. Vindob. G. 29831; P.Berlin 11710; and P.Vindob. G. 2312 (Greek Amulets, 119). However, de Bruyn (Making Amulets, 163) observes that this is less common in amulets that contain the Lord’s Prayer. Barns and Zilliachus, eds., Antinoopolis Papyri, 6–7. Kraus, “Manuscripts with the Lord’s Prayer,” 234–235. Raffaella Cribiore (Writing, Teachers, and Students in Graeco-Roman Egypt [Atlanta: Scholars, 1996], 72) argues P.Ant. 54 is a “tiny school book” for a young student. G. H. R. Horsley, “Reconstructing a Biblical Codex: The Prehistory of MPER XVII. 10 (P.Vindob. G 29831),” in Akten des 21. Internationalen Papyrologenkongresses, Berlin 13–19.8.1995, ed. Bärbel Kramer, 2 vols. (Leipzig: B. G. Teubner, 1997), 2:473–481, at 480. Jones, Greek Amulets, 117–20; Roberts, Manuscript, 82. L. Ingrams, P. Kingston, and P. Parsons, eds., The Oxyrhynchus Papyri, Vol. XXXIV (London: Egypt Exploration Society, 1968), 4–6. For an updated analysis of this fragment, see Tommy Wasserman, “P78 (P.Oxy. XXXIV 2684): The Epistle of Jude on an Amulet?” in New Testament Manuscripts, 137–160. Wasserman, “P78,” 151–158. Jones, Greek Amulets, 177. Blumell and Wayment, Christian Oxyrhynchus, 139. Wasserman, “P78,” 137.

320

Kruger

it seems we likely have another example of a manuscript that bears the form of a miniature codex, along with the content of an amulet.72 3.3 P.Vindob. G. 29831 (MPER N.S. 1710) Measuring just 6.5 × 4.2 cm., this is another bifolium sheet (this time parchment) containing four pages of text dated to the sixth/seventh century.73 The front and back of the first leaf contains a prayer of protection for the one carrying the document, and the front and back of the second leaf contains a citation from John 1:5–6 which abruptly cuts off. While the content clearly points in the direction of an amulet – prayer of protection along with scriptural citation – the hand of the scribe is unexpectedly elegant with an impressive biblical majuscule script (even including a number of developed scribal features such as an enlarged first letter). This manuscript, therefore, is a curious hybrid, looking very much like a miniature codex in external form, but possessing the kind of content that is unequivocally linked to amulets.74 3.4 P. Yale 1.3 (P50) This bifolium papyrus codex measures 8.8 × 13.8 cm. and is dated to the fourth/fifth century.75 It contains the text of Acts 8:26–32 followed immediately by Acts 10:28–31 (with the transition occurring in the middle of the second page), and there is no indication the codex originally held more pages.76 The fact that it contains an obvious discontinuous text, along with evidence that it had been folded multiple times, has led a number of scholars to label

72

The role of P.Oxy. 2684 as an amulet is less certain than P.Ant. 54 above given the ambiguity surrounding the original content of the codex (it may have been multi-quire and held the entire book of Jude). Thus, de Bruyn (“Preliminary List,” 172) rightly puts it into the “probably” rather than “certain” category for amulets. 73 Kurt Treu and Johannes M. Diethart, eds., Griechische Literarische Papyri Christlichen Inhaltes II, vol. 1 (Vienna: Hollinek, 1993), 23. For additional discussion, see Horsley, “Reconstructing a Biblical Codex,” 473–481. 74 The tension between these two categories led Horsley to suggest that the manuscript may have been originally intended as a miniature codex of John but after the scribe made a mistake, he turned it into an amulet (“Reconstructing a Biblical Codex,” 473–481). 75 Carl H. Kraeling, “P 50. Two Selections from Acts,” in Quantulacumque: Studies Presented to Kirsopp Lake by Pupils, Colleagues, and Friends, ed. Robert P. Casey, Silva Lake, and Agnes K. Lake (London: Christophers, 1937), 163–172. For a more updated discussion, see John Granger Cook, “P50 (P.Yale I 3) and the Question of Its Function,” in New Testament Manuscripts, 115–128. 76 Jones, Greek Amulets, 26.

Miniature Codices in Early Christianity

321

it an amulet.77 However, John Granger Cook has argued that the folds are not necessarily an indication of an amulet because there are numerous examples of non-amulet documents that were folded simply so they could be carried around more easily.78 As for the odd combination of texts, Cook makes the case, following Carl Kraeling,79 that it does not have a magical purpose but rather a “missionary” one.80 This document may have been used as “preacher’s notes for use in a worship service or as a Christian’s travelers note for use in teaching.”81 On the other hand, Stanley Porter argues that P50 might still have amulet-like qualities because the texts of Acts 8 and 10 may have been chosen “for the talismanic value related to Philip and Peter.”82 P50, then, is an example of the ambiguous nature of some documents. If Cook’s assessment is correct, it is also a reminder that some manuscripts could have contained discontinuous texts (and, in some cases, even folded) and yet still should be considered miniature codices.83 4

The Function of Miniature Codices

Having explored the characteristics of miniature codices, as well as some examples of borderline cases, we now turn our attention to their function within the early Christian movement. Why was this tiny format utilized? What practical functions did these books serve? Given that we have very little direct evidence that speaks to the function of these books, we can only reconstruct probable scenarios for how they were used. Here we survey a number of the 77 78 79 80 81 82 83

Van Haelst, Catalogue, 482; Kurt Aland, Repertorium der griechischen christlichen Papyri, I, Biblische Papyri (Berlin: De Gruyter, 1976), 280. Following Van Haelst, I originally placed P50 in the amulet category (“P.Oxy. 840,” 88), but now see it more as a miniature codex. Cook, “P50,” 120. Kraeling, “P50,” 171. Cook, “P50,” 122. Cook, “P50,” 125. Stanley E. Porter, “Textual Criticism in the Light of Diverse Textual Evidence for the Greek New Testament: A Proposal,” in New Testament Manuscripts, 305–337, at 320. Cook’s proposal is analogous to the (possible) use of Christian testimonia books which would have contained excerpts from the Old Testament useful for missionary activities and journeys. For more, see Stanton, “Why Were Early Christians Addicted to the Codex,” 182–185; Martin C. Albl, And Scripture Cannot Be Broken: The Form and Function of the Early Christian Testimonia Collections, NovTSup 96 (Leiden: Brill, 1999); Timothy H. Lim, Holy Scripture in the Qumran Commentaries and Pauline Letters (Oxford: Clarendon, 1997), 55.

322

Kruger

leading possibilities, none of which are mutually exclusive. Indeed, Christians may have utilized this tiny format for a multiplicity of reasons. 4.1 Private Reading There is near-unanimity among scholars that one function of miniature codices was to facilitate the private reading of books.84 Such usage would stand in contrast to those larger documents in early Christianity that were prepared for “public” reading.85 Of course, in order to take advantage of such private books, a person would not only have to be able to read, but would need to have access to such books (either by purchase or by borrowing). In the ancient world neither of these requirements would have been easy to meet. It is likely that literacy rates were quite low (probably no more than ten to fifteen percent), suggesting very few Christians could read these books even if they could get a hold of one.86 In terms of access, we can reasonably assume that some of the deluxe miniature codices could only have been purchased by the more well-to-do.87 However, Gamble argues that the repeated exhortations of patristic writers towards private reading suggests books may have been more accessible than we think.88 And we do have evidence that Christians would have borrowed books on occasion, as opposed to purchasing them.89 Regardless, it seems 84 E.g., Gamble, Books and Readers, 235–236; Hurtado, Earliest Christian Artifacts, 160–163; Roberts, Manuscript, 11. 85 Scott Charlesworth, “Public and Private – Second- and Third-Century Gospel Manuscripts,” in Jewish and Christian Scripture as Artifact and Canon, ed. Craig A. Evans and H. Daniel Zacharias, LNTS 70 (London: T&T Clark, 2009), 148–175. 86 W. V. Harris, Ancient Literacy (Cambridge, MA: Harvard University Press, 1989); Catherine Hezser, Jewish Literacy in Roman Palestine, TSAJ 81 (Tübingen: Mohr Siebeck, 2001). For arguments for a higher literacy rate, see Alan Millard, Reading and Writing in the Time of Jesus (New York: New York University Press, 2000); and Brian J. Wright, “Ancient Literacy in New Testament Research: Incorporating a Few More Lines of Enquiry,” TrinJ 36 (2015): 161–189. 87 Roberts, Manuscript, 11. Jerome laments the fact that some manuscripts were decorated beautifully while the poor are overlooked, implying that at least those manuscripts would have been costly (Ep. 22.32; cf. 107.11). In addition, the Apostolic Constitutions assumes that the private readers are those who “are in good circumstances” (2.4), that is affluent. For more on the cost of book production, see Roger S. Bagnall, Early Christian Books in Egypt (Princeton, NJ: Princeton University Press, 2009), 50–69. 88 Gamble, Books and Readers, 233–234. Cf. Chrysostom, Hom. Jo. 11, 32; Hom. princ. Act. 3. 89 Apparently Pamphilus of Caesarea (died ca. 309) had quite an extensive library and was quite willing to lend books (Jerome, Ruf. 1.9), with Jerome himself being one of the later benefactors (Vir. ill. 75). See also P.Oxy. 4365, a Christian letter showing the exchange/ borrowing of books, presumably for private reading; Annemarie Luijendijk, Greetings in the Lord: Early Christians in the Oxyrhynchus Papyri (Cambridge, MA: Harvard University Press, 2008), 70–74.

Miniature Codices in Early Christianity

323

that miniature codices served a relatively narrow strand of early Christianity, the literate (and probably the upper) class who had the leisure time for private reading. As much as patristic writers encouraged private reading, they also were aware of its dangers. The content of public reading in the church was controlled by ecclesiastical authorities, whereas there were no such controls over private reading. This may explain the language of the Apostolic Constitutions where affluent private readers are encouraged to avoid books that are “out of the church” such as books by “false prophets” and “strange and diabolical books.”90 Later, Cyril of Jerusalem clamps down on the looseness of private reading when he declares, “what is not read in the church should not be read privately.”91 Such warnings may prove justified by the significant number of apocryphal writings we find in the miniature codex format: e.g., unknown gospels, Gospel of Mary, Acts of Peter, Acts of Paul, Protoevangelium of James, Apocalypse of Peter, 4 Ezra, and (at least for some) Revelation.92 4.2 Portability A second function of miniature codices is that they would have allowed books to be taken on journeys with convenience and ease. Members of the early Christian movement would certainly not be characterized as home-bound, but were quite willing to travel, especially for missionary or ministry related purposes.93 Indeed, the new Christian movement could be characterized as one with a “transient life-style” as believers sought to evangelize the world around them.94 Thus, portable books would have been particularly attractive 90 Apostolic Constitutions 1.4–6. 91 Catech. 4.36. 92 Gamble, Books and Readers, 236. Kraus objects to my characterization of miniature codices as containing an “abundance” of apocryphal literature (“Manuscripts,” 27) on the basis that distinctions between canonical and apocryphal are “superficial and they result from modern reasoning” (“Miniature Codices,” 141). However, by the time of the fourth through sixth centuries – when miniature codices were most popular – the boundaries of the New Testament canon were largely solidified. Thus, the fact that a large portion of our miniature codices are apocryphal in nature seems noteworthy. 93 Larry W. Hurtado, “Interactive Diversity: A Proposed Model for Christian Origins,” JTS 64 (2013): 445–462; Michael B. Thompson, “The Holy Internet: Communication Between Churches in the First Christian Generation,” in The Gospel for All Christians: Rethinking the Gospel Audiences, ed. Richard Bauckham (Grand Rapids: Eerdmans, 1998), 113–145; Eldon J. Epp, “New Testament Papyrus Manuscripts and Letter Carrying in Greco-Roman Times,” in The Future of Early Christianity: Essays in Honor of Helmut Koester, ed. A. T. Kraabel et al. (Minneapolis: Fortress, 1991), 35–56. 94 Michael McCormick, “The Birth of the Codex and the Apostolic Life-Style,” Scriptorium 39 (1985): 150–158, at 158.

324

Kruger

for those engaged in such activities.95 We already see a hint of this purpose for miniature books in the first century when the Roman poet Martial explains that his books are in this tiny format for those “who want my little books to keep you company where you may be and desire their companionship on a long journey.”96 Similarly, the rhetorician Libanius (fourth century) mentions a copy of Thucydides in small letters that he could easily take on his journeys, carrying it himself instead of having his servant do so.97 And the sixth-century monastic document, Regula Magistri, tells of monks who bring a “modestsized little codex” on their long journeys.98 Thus, we have good reasons to think that portability would have been a significant reason for the production of miniature codices. On this score, it is also worth noting that a number of our miniature codices contain portions from books about travelling Christians: either canonical Acts (P.Yale 1.3) or apocryphal acts such the Acts of Paul (P.Ant. 13; P.Oxy. 6; P.Oxy. 1602) or the Acts of Peter (P.Oxy. 849). Given that these various acts catalogue the missionary journeys of the apostles, is it possible that they were put into a tiny format so that Christians could take them on their own missionary journeys? Put another way, might travelling Christians have particularly enjoyed reading pocket-sized books about the travels of Peter and Paul as a means of inspiration and encouragement while far from home? Of course, there is no way to know the answer to such questions, but they are intriguing, nonetheless. 4.3 Expression of Devotion In addition to reading miniature codices, it seems that these tiny books may have also been used to express one’s devotion to Christ or as a means to draw nearer to Christ. Thus, these books were often carried or displayed on the body as a visual expression of one’s Christian identity, or as visual reminder of Christ’s presence and power.99 As Caroline Humfress observes, “the physical presence of particular Christian codices actually invoked the spiritual presence of God.”100 95 Epp, “The Codex and Literacy in Early Christianity”, 21. Epp is referring to the adoption of the codex in general, but his comments certainly apply to miniature codices. 96 Ep. 1.2. 97 Libanius, Oratio 1.148. 98 Regula Magistri 57.4. English translation from Claudia Rapp, “Holy Texts, Holy Men, and Holy Scribes: Aspects of Scriptural Holiness in Late Antiquity,” in The Early Christian Book, ed. William E. Klingshirn and Linda Safran (Washington, DC: Catholic University Press, 2007), 194–222, at 204. 99 Caroline Humfress, “Judging by the Book: Christian Codices and Late Antique Legal Culture,” in The Early Christian Book, 141–158, at 152. 100 Humfress, “Judging by the Book,” 148. Cf. Sanzo, Scriptural Incipits, 165.

Miniature Codices in Early Christianity

325

We know Christians often treated scriptural books in just such a fashion.101 A biblical codex was placed on a throne at the Council of Ephesus (c. 431) as a powerful visual representation of the presence of Christ.102 Similarly, Epiphanius describes the power of just seeing scriptural books, “the mere sight of these books renders us less inclined to sin, and incites us to believe more firmly in righteousness.”103 The monk Philoxenus of Mabbug (fifth century) shows a similar devotion to the Gospels as a physical object: “Take the Gospel in your hands. Place it on your eyes and your heart … place the Gospel on the cushion and prostrate yourself before it up to ten times … [and] you will conceive in your heart the internal adoration and the effect of divine grace.”104 Given this background, it seems that miniature codices might have served this same sort of devotional function. Of course, we cannot always tell why a person carried or wore a miniature book, but we have a few possible examples of this devotional use. For instance, Maximus, disciple of the fourth-century Martin of Tours, hung a “book of the Gospels” around his neck during his travels, along with the instruments of the sacrament (a small paten and chalice).105 There is no indication these items functioned as an amulet, but apparently as a way for Maximus to put the tools of his divine trade on full visual display; i.e., to identify himself as minister. Similarly, an early deacon named Euplus was martyred in Catania in 304 under the Governor Calvisianus with the “book of the Gospels” hung around his neck.106 Indeed, he presented this very book to the authorities as proof that he was a Christian, even reading aloud portions from the Gospels.107 The fact that the possession of such a Gospel book was regarded as a violation of “the emperor’s edict” suggests we are dealing with a miniature codex and not an amulet.108 We also have the story of the 101 Rapp, “Holy Texts,” 194–222. 102 Rapp, “Holy Texts,” 197. 103 Benedicta Ward, The Sayings of the Desert Fathers: The Alphabetical Collection (London: Liturgical, 1984), 58 (Epiphanius 8). 104 Pierre Graffin, “La lettre de Philoxène de Mabboug à un supérieur de monastère sur la vie monastique,” Orient Syr. 6 (1961): 463–464, cited in Rapp, “Holy Texts,” 198. 105 Raymond Van Dam, Gregory of Tours: Glory of the Confessors (Liverpool: Liverpool University Press, 1988), 37. 106 Herbert Musurillo, Acts of the Christian Martyrs (Oxford: Clarendon, 1972), 319 (Latin recension). 107 Euplus is said (in the Greek recension) to have read from all four Gospels – something unlikely in a book that could be hung around the neck. But it would make for a more idealized account. The Latin recension has him reading texts from just two Gospels (Matt 5:10; Mark 8:34), neither of which had apotropaic characteristics. 108 Musurillo, Acts of the Christian Martyrs, 315. While the historical reliability of some martyrdom accounts is questionable (and often exaggerated), they likely reflect some actual practices of early Christians.

326

Kruger

fourth-century hermit Amoun of Nitria who carried a codex around his neck from which he read “the words of the Apostle and the Savior” to his bride on their wedding night to convince her to live a chaste lifestyle.109 Again, a likely reference to a miniature codex of some sort. Of course, not every document carried around the neck was necessarily a miniature codex. Chrysostom notes that “many of our women now wear Gospels hung from their necks,”110 but it is difficult to know the format of these “Gospels” – some consider them to be folded strips of papyrus/parchment (perhaps in containers),111 and others are open to the possibility these Gospels are tiny codices (whether complete Gospels or just portions thereof).112 However, it is noteworthy that Chrysostom appears to approve of these women wearing these Gospels. de Bruyn observes, “John Chrysostom refers to Christian women who wear the Gospel around their neck without any criticism, indeed with approbation.”113 The fact that elsewhere Chrysostom plainly condemns the use of amulets (περιάπτα) – calling them “a mere cheat and mockery” and “idolatry”114 – suggests that he did not view the Gospels around these women’s neck as amulets. 4.4 Protection/Healing If, as noted above, one regards miniature codices as symbolizing (and sometimes even embodying) the very presence of Christ, it is not difficult to see how some Christians might have taken a further step and begun to believe that the mere presence of such books could provide miraculous protection or 109 Palladius, Hist. Laus. 8.1–3. 110 Hom. Matt. 72.2; cf. Stat. 19.14. 111 Kraus, “Miniature Codices,” 38. Cf. Alcuin, Ep. 290. It should be noted that “strips” of parchment could still form a bifolium and thus be regarded as a miniature codex. 112 Roberts, Manuscript, 11. Sanzo argues these “Gospels” contain just portions of the text (Scriptural Incipits, 161–165). 113 de Bruyn, Making Amulets, 30. It is noteworthy, however, that other scholars argue Chrysostom is doing the opposite and condemning these women, e.g., Luijendijk, Forbidden Oracles, 55 n. 100; and Sanzo, Scriptural Incipits, 161–162. For more, see Robert Matthew Calhoun, “The Gospel (-Amulet) as God’s Power for Salvation,” EC 10 (2019): 21–55. 114 Hom. Col. 8. Although it might be suggested that Chrysostom is only condemning “pagan” amulets and not “Christian” ones, the text seems to indicate that Christianized amulets are in view. Indeed, Chrysostom anticipates how Christians rationalize their usage: “[They] are forever rationalizing about them [amulets], and saying, ‘we call upon God, and do nothing extraordinary,’ and the like; and ‘the old woman is a Christian,’ says he, ‘and one of the faithful’” (Col. 8). Moreover, other writers condemn amulets even that contain explicit Christian content or that are produced by priests or monks (e.g., Augustine, Tract. Ev. Jo. 7.12; Caesarius of Arles, Serm. 50.1).

Miniature Codices in Early Christianity

327

healing. Whether one regards such a usage as “magical” in some fashion is not something we need to resolve here.115 But we might say that these books, for some Christians, functioned as “relics” of sorts – physical objects that could communicate divine power.116 Chrysostom hints at such a usage when he tells us that “the devil will not dare to approach a house where a Gospel is lying”117 – a possible allusion to a miniature codex (but we cannot be sure). He is quick to explain, however, that real protection is achieved when “from those books you convey the letters and sense into your understanding.”118 Chrysostom also notes that it is not unusual for Christians to have a “Gospel hanging by our bed” – another possible allusion to a miniature codex that is thought to offer protection.119 But, he is again quick to emphasize that the mere presence of a Gospel book by one’s bed does nothing.120 It is a person’s commitment to help the poor that “give(s) wings to your prayer.”121 The language of Augustine is very similar. Augustine condemns those who use amulets to relieve “fever and pain” and instead suggests it is better to have a Gospel book by their head.122 The fact that Augustine contrasts this Gospel book with an amulet suggests that he may have a miniature codex in mind. After all, according to Augustine, these two sorts of documents can (and ought to be) differentiated from one another.123 Indeed, it seems likely that Augustine (and possibly Chrysostom) is encouraging the usage of these tiny books as an 115 For overviews of Christianity and magic, see David E. Aune, “Magic in Early Christianity,” in ANRW 2.23.2 (1980), 1507–1557; Edwin M. Yamauchi, “Magic in the Biblical World,” TynBul 34 (1983): 169–200; Christopher A. Faraone and Dirk Obbink, eds., Magika Hiera: Ancient Greek Magic and Religion (Oxford: Oxford University Press, 1991); Peter Schäfer and Hans G. Kippenberg, eds., Envisioning Magic: A Princeton Seminar and Symposium (Leiden: Brill, 1997). For more on distinction between magic and religion, see Alan F. Segal, “Hellenistic Magic: Some Questions of Definition,” in Studies in Gnosticim and Hellenistic Religion Presented to Gilles Quispel on the Occasion of His 65th Birthday, ed. Roel van den Broek and M. J. Vermaseren, NHMS 39 (Leiden: Brill, 1981), 349–375; H. S. Versnel, “Some Reflections on the Relationship Magic-Religion,” Numen 38 (1991): 177–197; Jens Braarvig, “Magic: Reconsidering the Grand Dichotomy,” in The World of Ancient Magic: Papers from the First International Samson Eitrem Seminar at the Norwegian Institute at Athens, ed. David R. Jordan, Hugo Montgomery, and Einar Thomassen (Bergen: Åströms, 1999), 21–54. 116 Rapp, “Holy Texts,” 200–201. 117 Hom. Jo. 32.3. 118 Hom. Jo. 32.3 (emphasis mine). 119 Hom. 1 Cor. 43.7. 120 Hom. 1 Cor. 43.7. 121 Hom. 1 Cor. 43.7. 122 Tract. Ev. Jo. 7.12; see also Doct. chr. 2.20; Serm. 286.7. 123 Of course, it is impossible to know whether the difference pertains to form or content. As noted above, some books were in the form of a codex but had the content of an amulet.

328

Kruger

orthodox alternative to the usage of amulets.124 While Augustine’s language suggests such a Gospel book might provide a cure for illness, he is quick to clarify that physical healing is not the purpose of such books; instead the purpose is to heal the heart.125 We also see the protective use of miniature books in the Acts of Andrew where a Christian woman named Trophima was falsely accused and sentenced to work as a prostitute in a brothel. The text tells us that Trophima “had the Gospel on her bosom [i.e., around her neck] and no one could approach her.”126 Whether this was a complete Gospel or just portions thereof is of not material to our discussion here – as note above, there was some fluidity in the content of such books. Regardless, the text presents this Gospel as something that protected Trophima from being attacked. Of course, some Christians used miniature codices for healing/protective purposes in even more explicit ways. As already noted above, we have examples of miniature codices that were produced intentionally as amulets (e.g., P.Vindob. G. 29831). Such “hybrid” documents would have been in the physical form of a codex, but the content would have included the classic adjurations, prayers, and texts designed for healing or protection. No doubt, such amulets would have gone beyond what church leaders like Chrysostom and Augustine would have regarded as the legitimate use of books.127 This complicated array of options reminds us that the different functions of the miniature codex are not always easily distinguished from one another – they often blend together and overlap. They are not so much separate options but rather points on a continuum. After all, what is the material difference between a miniature codex protecting a house and a “hybrid” miniature codex (with the content of an amulet) healing an illness?128 Why does Chrysostom seem to be comfortable with the former but not the latter? Is the difference perhaps that Chrysostom labored to emphasize that these books must still be applied to the heart, whereas amulets tended toward “the reduction of Scripture to a 124 de Bruyn, Making Amulets, 29. Not all in the early church would have agreed with this move (e.g., Jerome, Matt. 4.23). 125 Tract. Ev. Jo. 7.12. See discussion in de Bruyn, Christian Amulets, 30. 126 Gregory of Tour’s Epitome 23. 127 For a general overview of how the early church fathers viewed incantations, amulets, and other “magical” practices, see F. C. R. Thee, Julius Africanus and the Early Christian View of Magic, HUT 19 (Tübingen: Mohr Siebeck, 1984); and de Bruyn, Making Amulets, 17–42. 128 We are reminded that miniature codices lacking the explicit content of an amulet could still have functioned like an amulet depending on the intent of the user. In other words, a document can be an “amulet” if it used like one, regardless of whether it has the typical content associated with amulets.

Miniature Codices in Early Christianity

329

formal device”?129 Or does he object to explicit amulet-like content such as adjurations and prayers for healing – something lacking in “ordinary” miniature codices? Moreover, we should remember that these different functions are not mutually exclusive. For instance, a person may have worn a Gospel around their neck to express their Christian identity and to get protection or healing. Thus, we are reminded again that there are fuzzy boundaries between these various functions of miniature codices. 5 Conclusion The purpose of this chapter has been to explore the fascinating world of miniature codices in order to illumine our understanding of the way these tiny books functioned within the early Christian movement. First, we examined the characteristics that (generally speaking) distinguish miniature codices from other tiny documents, namely their physical form (codex, 10 cm. or less in breath), and their distinctive content (continuous text, high scribal quality). Of course, these characteristics are not absolute, with some documents being a miniature codex and an amulet at the same time. Second, and in light of this overlap between these two categories, we explored a number of “hybrid” documents that, one might say, are miniature codices on the outside and amulets on the inside. This reminded us of the fluidity of books in the ancient world and the complexity of categorizing certain manuscripts. Third, and finally, we explored the function of miniature codices. Why did Christians utilize this format? What purpose did it serve? Four purposes for miniature codices emerged: private reading, portability, expression of devotion, and for protection/healing. These categories are not mutually exclusive – a document can be created for multiple purposes. Moreover, the last two categories overlap considerably, demonstrating that it is difficult to draw a sharp distinction between using a book devotionally and using it as an amulet within the early Christian movement. 129 E. A. Judge, “The Magical Use of Scripture in the Papyri,” in Perspectives on Language and Text: Essays and Poems in Honor of Francis I. Anderson’s Sixtieth Birthday July 28, 1985, ed. Edgar W. Conrad and Edward G. Newing (Winona Lake, IN: Eisenbrauns, 1987), 339–349, at 341.

CHAPTER 13

Marginalia in New Testament Greek Papyri: Implications for Scribal Practice and Textual Transmission Michael P. Theophilos 1 Introduction By way of introduction, albeit not papyrological, it is of some note that Codex Corsendoncensis (GA 3 [f. 376v], twelfth century) includes, either by way of immediate exemplar or more distant predecessor, a marginal note within the actual text of 2 Cor 8:4, ἐν πολλοῖς τῶν ἀντιγράφων οὕτως εὕρηται (“in many of the copies it was thus found”), presumably referring to the addition of δέξασθαι ἡμᾶς in some manuscripts.1 Although there are no known extant manuscripts which include this as a marginal gloss, per se, its incorporation into the text of GA 3 is indicative of the importance of paying attention to marginalia and the manner in which they can exert an effect upon the text of the New Testament.2 The analysis of marginalia in New Testament Greek papyri provides a fascinating and distinct window into the process of textual transmission and scribal practice throughout the early centuries of the first millennium CE. The phenomenon of marginalia is not unique, of course, to New Testament papyri. 1 The weight of both external and internal evidence supports the shorter reading for 2 Cor 8:4, τῆς διακονίας τῆς εἰς τοὺς ἁγίους (“the ministry to the saints”) supported by P46 ‫ א‬B C D L, as opposed to τῆς διακονίας τῆς εἰς τοὺς ἁγίους δεξάσθαι ἡμᾶς (“for us to receive the ministry to the saints”) supported by 6 (13th century), 209 margin (14th century), 489 (14th century), 927 (12th century), 945 (11th century), 1244 margin (14th century), 1249 (15th century), 1628 (16th century), 1768 (16th century), 1876 (15th century), 2374 (14th century). The motivation for the later manuscripts to add δεξάσθαι ἡμᾶς (“for us to receive”) could be to provide further clarification, but what is most intriguing about this variation unit is what was in the margin. 2 Erasmus refers to this example from GA 3 in his 1519 edition. See further Bruce M. Metzger, The Text of the New Testament, 2nd ed. (New York: Oxford University Press, 1968) 194, who refers to Bengel who in turn refers to Erasmus in stating “apud Er” (trans. “with Erasmus”) cited in Bengel, Greek New Testament, 1734, 682. Also see discussion in Jan Krans, Beyond What is Written: Erasmus and Beza as Conjectural Critics of the New Testament, NTTS 35 (Leiden: Brill, 2006), 18 n. 29. For other examples of the same phenomenon in the Latin tradition see F. A. Knittel, New Criticisms on the Celebrated Text, 1 John V.7, trans. W. A. Evenson (London: J. Hatchard and Son, 1829), 98.

© Michael P. Theophilos, 2023 | doi:10.1163/9789004537972_015

Marginalia in New Testament Greek Papyri

331

Marginalia in Greek papyri are most regularly attested in the fragments of classical authors such as Homer, Euripides, and Menander;3 they also frequently appear in medical treatises in the form of practical explanations of technical terms, works on Roman law as sources of bilingual aids for Greek readers, and other cultural texts on music and astronomy.4 J. Krüger calculates that over 10% of literary finds at Oxyrhynchus include marginalia.5 These and other papyri provide a rich resource of comparanda for New Testament manuscripts. Both interlinear and marginal glosses range in scope and extent as much as they do in genre and scribal motivation for inclusion. Portions of this material have been assessed in the first editions of the New Testament papyri and other publications; however, thus far no consensus has emerged in light of the evidence of marginalia as a whole within the New Testament Greek papyri. This paper will provide (1) a catalogue of marginalia in New Testament Greek papyri, (2) categories for classification of these marginalia, (3) discussion of implications for scribal practice and textual transmission, and (4) a discussion that highlights some of the exegetical contributions of the marginalia. It is hoped that the present discussion will provide rationale and further motivation to carefully consider marginalia in future discussion of New Testament papyri. 2

Marginalia in Greco-Roman Documentary and Literary Texts

Kathleen McNamee’s voluminous corpus of marginal and interlinear notes in 293 Greek and Latin texts from Egypt, ranging in date from the third century BCE to the seventh century CE, has made significant advances in the way terminology, selection criteria, and types of evidence might be applied to the issue of marginalia in papyri.6 McNamee’s volume is the first to assemble and assess marginalia on such a large group of papyri as a whole. She notes that “without a reliable corpus of ancient annotations, editors of new texts lack useful comparanda, and scholars of classical antiquity lack convenient access to information in the ancient notes.”7 Her discussion is divided into four parts: (1) codicology (pp. 5–30), (2) context (pp. 31–92), (3) analysis of selected 3 S. F. Bonner, Education in Ancient Rome: From the Elder Cato to the Younger Pliny (London: Methuen, 1977), 160–161. 4 Kathleen McNamee, Annotations in Greek and Latin Texts from Egypt, ASP 45 (Oxford: American Society of Papyrologists, 2007), 12. 5 Julien Krüger, Oxyrhynchos der Kaiserzeit: Studien zur Topographie und Literaturrezeption (London: Peter Lang, 1990), 150. 6 McNamee, Annotations. 7 McNamee, Annotations, 1.

332

Theophilos

annotated texts (pp. 93–128), and (4) corpus of marginal and interlinear notes (pp. 129–530). Surprisingly, McNamee excludes all biblical and Christian texts from her discussion: “Biblical and Christian texts have been excluded [among other ancient texts and phenomena] … partly in order to limit the mass of material.”8 McNamee also excludes secondary additions such as corrections, unattributed variants, sigla of various kinds, titles, speakers, or stichometric numerals. The exclusion of these exiguous phenomena is a significant lacuna in her otherwise outstanding study – hence the need for the present work.9 Especially important for our task in the present paper is the classification system McNamee proposes for the different types of annotations in the 293 papyri she examines. Four types of annotations are identified: textual notes, elementary exegesis, notes on specific subjects, and quasi-personal notes, each with its various sub-divisions. This classification is extremely useful to analyze the content of marginalia. McNamee’s proposed categories are not all applicable to New Testament papyri, nor, as we shall see, is the classification system exhaustive. Nonetheless, McNamee’s proposed categories (see Table 13.1) offer direction on the systematic classification of these features in ancient scholarship across the broad filed of papyrology.10 This paper will seek to integrate (and expand on) McNamee’s classifications by incorporating marginalia in New Testament Greek papyri. Table 13.1 McNamee’s classifications indicating the subject matter of annotations

A Textual notes A1 Confirmation of a reading A2 Query A3 Attributed variant A4 Other textual comments B Elementary exegesis B1 Metaphrase B2 Interpretations and explanations

8

McNamee, Annotations, 2. The exclusion of this material is particularly striking given that she has included both literary and sub-literary texts. However, McNamee does seemingly have a preference for papyrus rolls (201 examples) rather than papyrus codices (69 examples), parchment or vellum codices (19 examples), or wooden tablets (4 examples). 9 Another fruitful area of research that would benefit from further analysis and contribute to the current discussion is a full assessm nt of similar phenomena in the ostraca. 10 McNamee, Annotations, 130.

Marginalia in New Testament Greek Papyri

333

Table 13.1 McNamee’s classifications indicating the subject matter of annotations (cont.)

C Notes on specific topics C1 Notes pertaining to organization or production of the work a Identification of speaker, contents, addressee b Stage directions and the like c The circumstances of production d The organization of the work e Didascalia, hypotheses f References to commentaries C2 Language a Morphology; discussion of dialect b Etymology c Expression, tone, phrasing d Syntax e Poetic figures, poetic expression f Meter C3 Matters of Fact a Myth b Historical or topical information c Geography d Custom e Proverbs f Ethnography g Ritual, religion, festivals h Astronomy i Botany, zoology, and similar scientific matters j Identification of sources or parallels used by the author k Medicine l Music m Law D Quasi-personal notes D1 Indications of useful passages D2 Other

McNamee’s findings in chapter 6 are particularly relevant for further work in the area of the utility of marginalia for interpretation, namely, that verbal correspondences between the marginalia and medieval scholia are striking. Although she draws back from suggesting that the marginalia are the direct

334

Theophilos

predecessors of medieval scholia, this may well prove to be a fertile area for further investigation, that is, identifying the relationship between the marginalia and later New Testament textual traditions. Also of interest is McNamee’s consideration of the types of authors in antiquity whose texts were annotated (ch. 7). It is perhaps not surprising that Homer is preeminent and that Pindar and Alcaeus are the most common lyric poets. She suggests that the readers were students at a secondary level of education (grammatical stage) and tabulates the standard expressions used.11 Also telling are the great variety of grammatical data. McNamee finds five of the six grammatical definitions in Dionysius’s Τέχνη Γραμματική attested in the 293 texts under discussion, including: (1) trained reading with due regard to prosody, (2) explanation according to poetical figures, (3) statement of dialectical peculiarities and allusions, (4) etymology, and (5) an accurate account of analogies. Conspicuously absent is the κρίσις ποιημάων (criticism of poetical productions), which Dionysius labels as the “noblest part of grammatic art.”12 One solution to this phenomenon is that the papyri reflect the need to understand the text rather than to comment on it, although the two may not be mutually exclusive.13 Furthermore, McNamee highlights an important distinction between registers of Greek expression. For example, archaic lyric and iambic poets such as Pindar and Bacchylides,14 as well as Hellenistic poets such as Callimachus and Theocritus,15 operate differently from literary and sub-literary prose.16 The main distinction being that in prose texts marginalia are relatively rare and elementary, a trajectory we also detect in New Testament papyri. Our preliminary analysis of the marginalia of New Testament papyri also generates a range of other interrelated questions: Does the width of the margin affect the quality or length of the marginal note? Are notes more regularly made in a particular margin (left, right lower, upper)? What role does interlinear space contribute to marginalia? Are marginalia accompanied by any regular lemmata, paragraphoi, or other sigla? Is there any way, methodologically, to determine whether the marginalia have derived from a textual or oral source, and what does paleography contribute to the investigation? We will not assume, as McNamee does, that “annotations are roughly contemporary with the copying of the book,”17 and would caution her tacit approval that “given 11 12 13 14 15 16 17

McNamee, Annotations, 61. Dionysius Thrax, Τέχνη Γραμματική §1.6. See further discussion below in relation to select New Testament papyri. McNamee, Annotations, 95–104. McNamee, Annotations, 105–116. McNamee, Annotations, 117–128. McNamee, Annotations, 5.

Marginalia in New Testament Greek Papyri

335

the difficulty of dating formal literary hands, editors tend in fact to depend on the more easily datable cursive marginalia to determine when a book was written.”18 Additionally, we will explore the question of whether annotations were more regularly penned by professional scribes or a later reader/owner of the text, and how multiple annotators are recognizable. We will return to address several of these questions below. We thus proceed with caution, sensitive to the reality that it is “impossible to say with certainty just how representative of ancient literary production this group of annotated texts actually is,”19 but optimistic that the marginalia in New Testament papyri shed valuable light on the process of textual transmission and scribal practice. 3

New Testament Papyri

3.1 P1 P1 (P.Oxy. 2) is a fragment of a bifolium containing Matt 1:1–9, 12, 14–20 dated to approximately the mid-third-century CE. The right-most portion of the recto preserves the beginnings of three lines: εγεν̣[ | παρ[ | μητ[̣ |. Bernard Grenfell and Arthur Hunt, the editors of the first edition, suggested that these traces “cannot be determined”;20 however, “the difference in the handwriting and the greater margin above the three broken lines distinguish them from the text of St. Matthew … [and] they may have formed a title of some kind.”21 Inspired by this possibility, José O’Callaghan suggested in 1971 that these lines summarize Matt 2:14, ἐγερθεὶς παρέλαβεν τὸ παιδίον καὶ τὴν μητέρα αὐτοῦ (“he arose, took the child and his mother by night and went to Egypt”).22 However, Matt 2:14 is a rather selective and obscure summary of Matthew’s material, and the identification is highly dubious. In 2001 Philip Comfort and David Barrett suggested that “it could be conjectured that it was not so much a title as it was a kind of subhead descriptor”23 and suggest the following: εγεν̣[νεθη (was born [Jesus Christ, the son of David,] | παρ[α (from [the Holy Spirit coming upon]) | μητ[ρος (his mother [Mary, the ̣ 18 McNamee, Annotations, 5. 19 McNamee, Annotations, 2. 20 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri: Part I (London: Egyptian Exploration Fund, 1898), 4. 21 Grenfell and Hunt, eds., The Oxyrhynchus Papyri: Part I, 4. 22 José O’Callaghan, “¿Mt 2,14 en el fragmento adéspota de P1?” Studia Papyrologica 10 (1971): 87–92. 23 Philip W. Comfort and David P. Barrett, The Text of the Earliest New Testament Greek Manuscripts (Wheaton, IL: Tyndale House, 2001), 39.

336

Theophilos

wife of Joseph]). Although this proposal is more probable than O’Callaghan’s, it still fails to find any further corroborating support regarding a subhead descriptor. More recently, Brice C. Jones has suggested ἐγ[εννήθη Ἰησοῦς] (Jesus was born) | παρ[ὰ πνεύματος ἁγίου καὶ τῆς] (by the Holy Spirit and) | μητ[ρὸς αὐτοῦ ̣ Μαρίας] (his mother, Mary).24 This proposal is the most attractive and has additional support from the Nicene Creed from the First Council of Constantinople (381 CE): σαρκωθέντα (he was made flesh) | ἐκ πνεύματος ἁγίου καὶ (by the Holy Spirit and) | Μαρίας τῆς παρθένου (Mary, the Virgin), corresponding to the tripartite formula proposed by Jones. Although it is the subject of much debate, the letters ΧΜΓ are regularly found in Christian papyri and are generally accepted as the abbreviation for (ὁ) (ἐκ) Μ(αρίας) γ(εννηθείς), and by implication emphasise the human nature of Christ.25 Contemporaneous examples abound and show the preference for this in Christian literature in the fourth and fifth centuries.26 Returning to Grenfell and Hunt’s assertion, “what relation, if any, the beginnings of the three lines … have to the St. Matthew fragment cannot be determined,”27 we can confidently note that both the paleographic and formatting features of the three lines distinguishes the text as a later development of the Matthean fragment. One in which a summary title was added to most probably offer a summary of the idea that Jesus was born of Mary. 3.2 P2 Currently housed in the Museo Archeologico, Florence, P2 (Inv. Nr. 7134) is a sixth-century papyrus Gospel manuscript with thirteen lines of Coptic text on the verso and two on the recto (Luke 7:22–26, 50), and a further ten lines of

24 Brice C. Jones, “The Mysterious Flyleaf of P.Oxy. 2 (P1): An Odd Gospel Title” (https:// www.bricecjones.com/blog/the-mysterious-flyleaf-of-poxy-2-p1-an-odd-gospel-title; accessed 5 April 2014). 25 See Philippe Le Bas and William Henry Waddington, Voyage archéologique en Grèce et en Asie Mineure … pendant 1834 et 1844, III, Part 6, ed. W. H. Waddington, Inscriptiones grecques et latines de la Syrie (Paris, 1870), no. 2145. De Rossi proposed the alternative Χ(ριστός), Μ(ιχαήλ), Γ(αβριήλ) and suggests that the form highlights the association of Christ with the archangels Michael and Gabriel, cited in William K. Prentice, “ΧΜΓ, a Symbol of Christ,” Classical Philology 9 (1914): 410. See Prentice’s erudite critique of the angel hypothesis on pp. 412–416 of the same article. 26 BGU 3.948.1; 12.2144.1; 17.2690a.1; 17.2727.1; 19.2780.1; CPR 1.30.54; 9.41.1; 9.43.1; 10.56.1; 19:62.1; P.Harr. 1.158.1; P.Haun. 2.25.1; P.Horak 21.1; P.Kellis 4.96.1; P.Köln 13.541.1; 13.542.2; P. L.Bat. 62.1; P.Merton 2.92.1. 27 Grenfell and Hunt, eds., The Oxyrhynchus Papyri: Part I, 4.

Marginalia in New Testament Greek Papyri

337

Greek text on the recto28 containing John 12:12–15.29 The fragment measures 25 × 15 cm and on the recto preserves the left margin. Of particular interest is the marginal depiction of a cross in the left column which indicates a transition to a different text, and in this case, a transition from Coptic to Greek. 3.3 P3 P3 (P.Vindob. G 2323) is a fragment measuring 24.5 × 11.5 cm. and consists of 15 verses from the Gospel of Luke (recto Luke 7:36–45; verso Luke 10:38–42). Karl Wessely, author of the editio princeps, concludes that “Nach den Schriftzügen scbliesse ich mit ziemlicher Sicherheit, dass” (“according to the handwriting it is almost certainly around the year 600”).30 Wessely argues that it formed a lectionary of sorts, bound by its common theme of New Testament Marys, despite the fact that reading the “sinful woman” in Luke 7 as Mary Magdalene is an unsubstantiated assumption that it is the same figure as the Mary of Luke 8, as well as the fact that the passages are not both found in the Synaxarion.31 Nonetheless, of interest for our discussion are the apparent titles in line 1 of both sides, εὐ]αγγ[έλ]ιον (recto), and εὐαγγέλιο]ν τοῦ ἀγίου Λούκα (verso). The second line on both sides seems to have “defeated his [Wessely’s] attempt at decipherment since he transcribed only part of what is visible, giving on the recto τὸν κύριον Ἰησοῦ and on the verso the single word κωμη, neither of which prove upon closer perusal to be correct.”32 J. Neville Birdsall, however, reconstructed the second lines on the recto as περι της αλειψ]α[σ]η[ς] τον κν μυρ[ω (“concerning the anointing of the Lord with oil”), and on the verso π]ερι μαρθας και μαρια[ς (“Concerning Martha and Mary”).33 The origin of these κεφάλαια

28 ευαγγελ[| ☩ τη επαυριον οχλος .π.[ολυς ο ελθων εις την] | εορτην ακουσαν.τ.[ες οτι ερχεται ο ι̅ς]̅ | εις ιερου[σαλ].ημ..ε.[λαβον τα βαια των] | φοινικων κ[αι εξηλθον εις υπαντησιν αυτω | κα]ι εκραυ[γαζον ωσαννα ευλογημενος | ο] ερχομε[νος εν ονοματι κ̅υ̅ και ο | βασιλευς τ]ου ι̅η̅λ̅ ε[υρων δε ο ιησους οναριον | εκαθισεν ε]π αυτ[ο καθως εστι γεγραμμενον] |[μη] .φ.[οβου θυγατηρ σιων ιδου … επι πωλον ονου]. 29 Astorre Pellegrini, “Piccoli testi coptp-saidici del Museo archeologico di Firenze,” Sphinx: Revue critique embrassant le domaine entier de l’egyptologie 10 (1906): 153–154 n. 19. 30 Karl Wessely, “Evangelien-Fragmente auf Papyrus,” Wiener Studien 4 (1882): 198. 31 Synaxarion (Συναξάριον) equates the Eastern Orthodox, Oriental Orthodox, and Eastern Catholic Churches to a compilation of hagiographies corresponding roughly to the martyrology (catalogue or list of martyrs and other saints and beati arranged in the calendar order of their anniversaries or feasts) of the Roman Church. 32 J. Neville Birdsall, Collected Papers in Greek and Georgian Textual Criticism (Piscataway, NJ: Gorgias, 2006), 79. 33 Birdsall, Collected Papers, 79.

338

Theophilos

and associated τίτλοι is still unknown but it almost certainly means we are dealing with a selection of Gospel material and not necessarily a lectionary.34 3.4 P4 P4 (Paris, Bibl. Nat., Suppl. Gr. 1120) contains ninety-five verses from Luke 1–6 (1:58–59; 1:62–2:1, 6–7; 3:8–4:2, 29–32, 34–35; 5:3–8; 5:30–6:16) and is typically, following C. H. Roberts, dated late second century CE,35 or following Kurt Aland, dated to the third century CE.36 Of interest here is not only the frequent high-point, mid-point, base-point, and colon for new sections (Luke 3:14; 6:8), or paragraphs marked with an outdent (Luke 1:76, 80; 2:1; 3:19, 23; 5:36; 6:12) but rather the small “fly leaf” fragment which reads ευαγγελιον κατα μαθ᾿θαιον (“Gospel according to Matthew”). The first edition of P4 in 1892 by Vincent Scheil neglected to even mention the existence of this piece.37 Attention to this small fragment of P4 has some significant implications. Simon Gathercole’s 2012 article is a good example of exploiting the “fly-leaf” for papyrological concerns including that “letter forms have close parallels in second and third centuries”38 and that “the apostrophe may suggest a date post-200 CE.”39 As Gathercole has explored, issues pertaining to the convention of the apostrophe as a consonant divider between double mutes or liquids is not necessarily a feature unique to the third century.40 Neither of these implications, however, is decisive, but attention to the extant marginalia is the starting point of the discussion. 3.5 P5 P5 is composed of two manuscripts from Oxyrhynchus: P.Oxy. 208 consists of two codex pages containing John 1:23–31, 33–40 on one leaf (recto and verso) and John 20:11–17, 19–20, 22–25 on another (recto and verso) published in 1899. 34 For full and updated editions of the relevant Vienna material see Stanley E. Porter and Wendy J. Porter, New Testament Greek Papyri and Parchments: New Editions: Texts. Mitteilungen Aus der Papyrussammlung der Österreichischen Nationalbibliothek (Papyrus Erzherzog Rainer) Neue Serie XXIX XXIX (Berlin: De Gruyter, 2008). 35 C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University Press, 1979), 12–13. 36 Kurt Aland, ed., Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments, 2nd ed. (Berlin: De Gruyter, 1994), 29. 37 Vincent Scheil, “Fragments de l’Évangile selon saint Luc, recueillis en Égypte,” Revue Biblique 1 (1892): 113–115. 38 Simon J. Gathercole, “The Earliest Manuscript Title of Matthew’s Gospel (BnF Suppl. gr. 1120 ii 3 / P4),” NovT 54 (2012): 209–235, here 234. 39 Gathercole, “The Earliest Manuscript,” 234. 40 Gathercole, “The Earliest Manuscript,” 209–235.

Marginalia in New Testament Greek Papyri

339

Over two decades later (1922) a further fragment of the codex was published as P.Oxy. 1781 and contained John 16:14–30 (recto and verso). In addition to the standard marginalia indicating superlinear additions (οι δε added to John 1:38; ο added in John 16:19; αυτω in John 16:29; και in John 20:19), deletion by superlinear dots (αυ in John 1:38), or a combination of elements (λοπηθη changed to λυπηθη by diagonal line through ο and υ added superlinearly in John 16:20; λοι changed to λυ with superlinear υ in John 16:21), John 16:24 is a particularly striking example where an apparent scribal error in the text has been rectified and amended in the lower margin by a second hand, εν τω ονοματι [μου εως αρτι ουκ η]τ[ησατε ουδεν εν. Interestingly, we never would have known that there was a variation here in the manuscript because line 8 does not preserve enough text to identify any variation (line 8 reads εν τω ον̣[ο]μ̣α̣[τι μου αιτειτε και]). 3.6 P6 The codex which P6 is from contains both Greek and Coptic (Akmimic dialect) text and is preserved in over 200 fragments (P.Copt. 379, 381, 382, 384 [Strasbourg, Bibl. Nat. et Univ.]). The Greek portion consists of John 10:1–2, 4–7, 9–10; 11:1–8, 45–52, while the Coptic preserves 1 Clem. 1.1–26.2; John 10:1–12.20; 13:1–2, 11–12; James 1:13–5:20. The 15 fragments from four leaves containing portions of the Greek New Testament (John) are commonly dated to the fourth century CE. Images from the published editio princeps (which appeared in 1910)41 and the high-resolution modern images (accessible through the INTF portal) have seriously misaligned the fragments, even mounting one upside down. It does not take much effort, however, to realign these fragments into a coherent reconstruction. The most interesting feature in this manuscript, with respect to marginalia, is the transitional marker either as a paragraphos divider or a more elaborate divider (similar to the coronis in other earlier classical Greek texts). The paragraphos indicates a transition from John 10:5 to John 10:6. Later the mark indicates a transition from the Coptic of John 11:44, to the Greek of John 11:45, and again later, the mark indicates a transition from the Coptic of John 10:42, to the Greek of John 11:1. Additionally, Rösch finds traces of the same sign indicating a transition from the Greek of John 11:52, to the Coptic of John 11:45.42 In sum, the horizontal line with S-shaped mark below from mid-point of the line indicates a significant division in the text and within P6 indicates a transition from Greek to Coptic and vice-versa. 41 See further Friedrich Rösch, Bruchstücke des ersten Clemensbriefes: Nach dem achmimischen Papyrus der Strassburger Universitäts- und Landesbibliothek (Strasbourg: Schlesier & Schweikhardt, 1910). 42 Rösch, Bruchstücke, 119–126.

340

Theophilos

3.7 P7 P7 is a fifth century fragment of Luke 4 and was formerly located in the Library of the Ukrainian Academy of Sciences (Shelf mark F. 301 [KDA], p. 553); however, it has since been lost. Caspar René Gregory records the manuscript with minimal details in 1908,43 and Aland discusses the fragment with a full transcription and brief notes in his 1957 survey.44 Gregory notes that the fragment preserves “den Schluss von einer Rede, einer Homile, oder eines Abschnittes eines Kommentars” (“the conclusion of a speech, a homily, or a section of a commentary”),45 and is then followed by Luke 4:1–2. Gregory also reports that there was another manuscript missing from the Archaeological Museum of the Spiritual Academy of Kiev with the same inventory number, of which he recorded a transcription in 1903 and which was reminiscent of the Sermon on the Mount (Matt 6:33–34; 7:2). Sadly, none of the fragments, as far as we are aware, were photographed before being lost during the wartime German occupation. Of interest in the present discussion is that the transcription records of the Lukan text preserve a paragraphos at the mid-point between the interlinear space separating the homily and biblical text. Of note are the angular right pointing brackets at the beginning of each line of the Lukan text, of which ten are preserved. 3.8 P9 P9 (P.Oxy. 402) is a fragment of 1 John 4:11–12, 14–17 dated to the fourth century. The handwriting is crude and irregular and there are some unintelligible spellings which the editio princeps describes as “extremely careless copying.”46 One point of interest is the interlinear addition of “ις” on line 5 of the verso which illustrates that a scribe could make use of the interlinear area when space was restricted. 3.9 P10 Since the erudite discussion of the social context of P.Oxy. 209 (P10) by Annemarie Luijendijk in 2010,47 it has been generally accepted that the 43

Caspar René Gregory, Die griechischen Handschriften des Neuen Testament (Leipzig: J. C. Hinrichs, 1908), 46. 44 Kurt Aland, “Neue neutestamentliche Papyri,” NTS 3 (1956–1957), 261–286, esp. 262–265. 45 Gregory, Die griechischen Handschriften, 46. 46 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri. Part 3: Nos. 401–653 (London: Egypt Exploration Fund, 1903), 2. 47 Annemarie Luijendijk, “A New Testament Papyrus and Its Documentary Context: An Early Christian Writing Exercise from the Archive of Leonides (P.Oxy. II 209/P10),” JBL 129 (2010): 575–596.

Marginalia in New Testament Greek Papyri

341

manuscript was a school exercise of the early fourth century, building of course on the previous work of Grenfell and Hunt,48 Aland,49 Guglielmo Cavallo and Herwig Maehler,50 Klaus Junack,51 and Raffaela Cribiore.52 Regarding the para-New Testament-textual material, Luijendijk notes that, “underneath the biblical passage a different hand has penned a couple of random phrases in cursive writing. On the back it reads ‘apostle.’”53 On the recto: Αὐρήλιος Παῦλο[ς ..]νυνισιου τῶν παρὰ γενήματος | περὶ τῶν γενημάτων […]ου ἐπὶ τοῦ λογείας ..[.] των | χιτ; on the verso: .πι[…]ση ἀπόστολος | (1st hand) are traces of ink. Discernable here is the name “Aurelius Paulus,” but beyond that the remainder seem to contain “ungrammatical expressions.”54 The full name of the main figure within the Leonides archive is, “Aurelius Leonides, son of Theon,” and although there is nothing explicit in the attested archive (P.Oxy. 3254, 2585, 3255, 103, 3256, 3257, 3258, 3259, 3260, 3261, 3262, PSI 469) that suggests explicit Christian affiliation (nomina sacra, Christian names), with the linking of P10 to that archive, we have strong evidence that Aurelius was a Christian. Despite the suspicion that the marginal gloss simply “may have served to test the pen,”55 the gloss preserves crucial information for identifying the social context of the manuscript. The marginal comment helps to identify both the owner (through mention of the name Αὐρήλιος with Παῦλο[ς?/υ?] possibly in the genitive referring to the Pauline letter quoted above?) and the “mercantile environment”56 (through mention of terms “produce” [γενήματος / γενημάτων] and “account” [λογείας]).57 The recto also contain a page number in the center top margin: alpha, page 1. This is certainly consistent with Eric G. Turner’s observation, “the favorite place for [pagination] is undoubtedly the center of the upper margin.”58 This 48 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri. Part 2: Nos. 208–400 (London: Egypt Exploration Fund, 1899), 8. 49 Kurt Aland, Biblische Papyri: Altes Testament, Neues Testament, Varia, Apokryphen, vol. 1 of Repertorium der griechischen christlichen Papyri, PTS 18 (Berlin: De Gruyter, 1976), “Var 33 [NT10],” 357–358. 50 Guglielmo Cavallo and Herwig Maehler, Greek Bookhands of the Early Byzantine Period A.D. 300–800, BICS Bulletin Supplement 47 (London: Institute of Classical Studies, 1987), 8 (no. 1a). 51 Klaus Junack, Das Neue Testament auf Papyrus (Berlin: De Gruyter, 1989), 21. 52 Raffaela Cribiore, Writing, Teachers, and Students in Graeco-Roman Egypt, ASP 36 (Atlanta: Scholars, 1996), 246–247. 53 Luijendijk, “P.Oxy. II 209/P10,” 576. 54 Luijendijk, “P.Oxy. II 209/P10,” 576. 55 Cited by Luijendijk, “P.Oxy. II 209/P10,” 576. 56 Luijendijk, “P.Oxy. II 209/P10,” 576. 57 Luijendijk, “P.Oxy. II 209/P10,” 576. 58 Eric G. Turner, The Typology of the Early Codex, Haney Foundation Series 18 (Philadelphia: University of Pennsylvania Press, 1977), 76.

342

Theophilos

may suggest that the pagination was extant in the Vorlage and, indeed, was a collection of Pauline material.59 3.10 P12 Unique among the New Testament papyri is P12, for the entirety of the New Testament portion of the text of is found within the margin.60 The text reads, Πολυμερῶς κ(αὶ) πολυ[τρό]πως | πάλε ὁ (θεὸ)ς λαλήσ[α]ς το[ῖς π]ατρά- | [σιν] ἡμ[ῶ]ν ἐν τοῖς προ[φ]ήτα[ις]. Line 2 has πάλε for πάλαι, and line 3 adds ἡμῶν which is not found in the broader manuscript tradition (although a corrector of P46 does include ἡμῶν superlinearly). The recto of P.Amh. 1.3 is a late third century Christian letter, possibly written during the era of Diocletian or even earlier (250–285 CE), to a Christian community in the Arsinoite nome, in the Fayum, concerning the payment of a sum of money by the addressees. The quotation from Hebrews occurs at the top margin of the second column of the letter and is assumed to have been added some time after the composition of the letter. The purpose of the quotation from Hebrews, however, remains a mystery, in part due to the fragmentary nature and mutilated condition of the manuscript. The verso consists of Gen 1:1–5 in a more cursive hand, dated to approximately the time of Constantine. One tentative suggestion derives from the tracing the linkage of Heb 1:1 and Gen 1:1–5 through Ambrose (340–397 CE), Bishop of Milan, who in De Fide 4.102, while arguing against the Arians who claimed that the Son had a beginning of existence, are considered and refuted on the grounds that, since He is the Word, He is not a work. Now the Father has spoken in various manners, whence it follows that He has begotten many Sons, if He has spoken His Word, not created it as a work of His hands. O fools, talking as though they knew not the difference between the word uttered and the Divine Word, abiding eternally, born of the Father – born, I say, not uttered only – in Whom is no combination of syllables, but the fullness of the eternal Godhead and life without end! That the writer to the Hebrews was interested in such matters is evident through such sayings as Heb 11:3, “By faith we understand that the worlds were prepared by the word of God.”

59 Cf. P46 which also starts with Romans. 60 Bernard P. Grenfell and Arthur S. Hunt, eds., The Amherst Papyri I 3B and 1 plate (London: H. Frowde, 1900), 28–31 and plate XXV in II (London: H. Frowde, 1901).

Marginalia in New Testament Greek Papyri

343

3.11 P13 P13 (P.Oxy. 657 + PSI 1292) is a third- to fourth-century scroll of 12 columns of 23–27 lines each including portions from the book of Hebrews (2:14–5:5; 10:8–22; 10:29–11:13; 11:28–12:17). Marginalia include pagination 47–50, 61–65, and 67–69, which may imply it was preceded by the book of Romans or a text(s) of similar length (approximately 36,000 characters). Editorial activity in the same hand of the scribe, often in the margin, is also evident in corrections, additions, and changes: χαυχη corrected to καυχη (Heb 3:6), ε was added superlinearly in λειτουργῶν (Heb 10:11), α has been erased (Heb 10:17), φενομενων changed to φαινομενων by diagonal through ε and addition of αι superlinearly (Heb 11:3), αυτου changed to αυτω (Heb 11:4), ιρηνικον changed to ειρηνικον (Heb 12:11), and αυτοις was changed to αυτης (Heb 12:11). 3.12 P15 P15 (P.Oxy. 1008) consists of a fourth-century single codex leaf of 1 Cor 7:18–8:4. Of interest for our discussion are the two occasions (7:39; 8:1) where a new paragraph division in the text is indicated with a marginal bird-beak coronis and ekthesis (recto, ll. 25, 32).61 Although many commentators identify a division at 8:1 as Paul transitions to his discourse on fleeing idolatry and glorifying God in worship (8:1–14:40), a division at 7:39 is harder to assimilate.62 Given that Paul is presumably associating ideas discussed earlier in the chapter pertaining to changes in a marital status, a division here seems unnecessary. As Garland notes, Paul simply “continue[s] the pattern of alternating instructions to males and females.”63 But again, attention to the marginalia has revealed some interesting dimensions of the textual tradition and scribal activity. 3.13 P23 P23 is a late-second- to early-third-century fragment of James from Oxyrhynchus (P.Oxy. 1229) which displays several compelling textual affinities with Alexandrinus and Vaticanus. The pagination, β (2) on the verso, and γ (3) on the recto, indicate that the text commenced on the second page of the preceding leaf, and it is assumed that the first page was either blank or contained a title.64 A correction occurs in Jas 1:10 on l. 1 of the verso where ταπεινουσι was 61 Arthur S. Hunt, ed., The Oxyrhynchus Papyri. Part 7: Nos 1007–1072 (London: Egypt Exploration Fund, 1910), 4. 62 However, see Solomon’s essay in this volume who notes a paragraphos break at 7:39 in 03, and ekthesis in 01 and 02. 63 David E. Garland, 1 Corinthians, BECNT (Grand Rapids: Baker Academic, 2003), 343. 64 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri. Part 10: Nos 1224–1350 (London: Egypt Exploration Fund, 1914), 16–18.

344

Theophilos

corrected to ταπεινωσι by a diagonal bar through ου and then addition of ω superlinearly in the top margin. James 1:11 on l. 7 also adds και superlinearly. 3.14 P27 P27 (P.Oxy. 1355) is a mid-third-century fragmentary manuscript of Romans (8:12–22, 24–27, 8:33–9:3, 5–9) and includes two superlinear additions in 8:21. Both the addition of σ]εται to ελευθερωθη to form ελευθερωθη[σ]εται, and the preposition απο, immediately following, were added superlinearly. 3.15 P30 P30 (P.Oxy. 1598) includes portions of 1 Thess 4:12–13, 16–17, 5:3, 8–10, 12–18; 2 Thess 1:1–2 in two consecutive leaves in five fragments, and is dated to the mid-third-century. Pagination is preserved for pages 207–208, which the author of the editio princeps suggests is evidence for a collection of Pauline epistles.65 Right angular signs (>) are used to fill short lines. 3.16 P34 P34 (P.Vindob. G 39784) is a mid-seventh-century text containing portions of 1 Cor 16:4–7, 10; 2 Cor 5:18–21; 10:13–14; 11:2, 4, 6–7 in a double sheet with two columns per side. At the bottom of column 1 of fragment A, after ἐν τῷ εὐαγγελίῳ τοῦ Χριστοῦ of 2 Cor 10:14, there are four lines of decoration. The first three consist of a sequence of three angular brackets followed by what appears as an elongated tilde. This sequence is repeated three times for each line for three lines. The fourth line begins within the text block of the column and extends into the lower margin with six enlarged decorative shapes resembling a shallow Y shaped with a descending vertical which curves to the left. The purpose of this decoration is not known, but it is intriguing that, in precisely the same page location on the reverse side, there is a string of unidentifiable letter traces, if indeed they are letters. 3.17 P35 P35 (PSI 1) is a third- to fourth-century fragment of the middle-upper portion of a single column codex leaf containing portions of Matt 25:12–15, 20–23. The top margin is partially preserved and there are small traces of ink on the right of the recto top margin with a partially abraded vertical of 3.5 mm. There are corresponding traces on the verso top margin (far left) consisting of a small 65 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri. Part 13: Nos 1594–1625 (London: Egypt Exploration Fund, 1919), 12–14.

Marginalia in New Testament Greek Papyri

345

descending vertical of approximately 1 mm. While the state of preservation of the papyrus does not allow us to be definitive, the traces are consistent with the location of page numbers on other papyri of the same era. 3.18 P36 P36 (PSI 3) consists of two fragments from the sixth century of a single column copy of John’s Gospel (John 3:14–17, 17–18, 31–32, 34–35). Pagination notation �� (= 25) is extant in the mid-point of the top margin of fragment A. 3.19 P37 P37 (P.Mich. 137) is a third- to fourth-century leaf of papyrus codex of Matthew’s Gospel (Matt 26:19–37, 26:37–52) with mostly cursive letter forms. On occasion, there are reading marks inserted by a later hand at the end of phrases. Henry Sanders, author of the editio princeps states, “at the ends of certain phrases a dot or stroke appears above, apparently not by the first hand. The leaf is too fragmentary for one to be sure that these marks occurred regularly. They seem to be marks put in to assist reading.”66 Superlinear corrections include ν added to εγενηθη to form εγεννηθη (Matt 26:24), εκαλεσεν changed to εκλασεν (Matt 26:26), ν added to εκχυνομενον to form εκχυννομενον (Matt 26:28), and του added superlinearly before γενηματος (Matt 26:28). Both the letters forms and appearance of ink suggest these corrections were by a second hand. 3.20 P39 P39 (P.Oxy. 1780) is a third-century single column leaf with portions of John’s Gospel (John 8:14–22).67 The upper margin of the recto has pagination, ΟΔ (=74) which may imply that the codex only contained John’s Gospel. The number of letters missing from verso to recto would imply this codicological reconstruction. 3.21 P40 On one occasion P40 (P.Heid. Inv. G. 645), a third century copy of portions of Romans (1:24–27; 1:31–2:3; 3:21–4:8; 6:4–5, 15–16; 9:16–17, 27), adds an ν by superlinear addition for εσμεν in Rom 6:15.

66 Henry A. Sanders, “An Early Papyrus Fragment of the Gospel of Matthew in the Michigan Collection,” HTR 19 (1926): 217. 67 Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri. Part 15: Nos 1778–1828 (London: Egypt Exploration Fund, 1922), 7–8.

346

Theophilos

3.22 P41 P41 consists of 15 fragments68 from the eighth century and has portions of texts from the book of Acts 17–22. It regularly employs a decorative paragraphos in the left margin of the extant fragments in what resembles a Y with a cursive tail: K7541 (Acts 17:28–31, 17:34–18:2), three examples; K7542 (Acts 18:17–18, 22–23), two examples; K7544 (Acts 19:13–16, 18–19), two examples; K7546 (Acts 20:22–24, 26–28), one example; and K7548 (Acts 22:11–14, 16–17), one example. 3.23 P45 P45 is a third-century codex consists of 30 extant fragments including extensive portions from the Gospels and Acts (Chester Beatty I + P.Vindob. G 31974).69 On nine occasions, the superlinear space on the manuscript is used for corrections or additions: εορακεν was changed to εορακαν by the addition of α above (Luke 9:36); ειπεν τις προς αυτον ακολουθησω σοι was deleted by dots above (Luke 9:58); γεινωσκουσι becomes γεινωσκει by deletion of ουσι with diagonal strokes and addition of ει above (John 10:14); ποιησομεν becomes ποιουμεν by adding ου above ησο (John 11:47); σ is added to καταλυθηεται to form καταλυθησεται (Acts 5:38); ς was added to indicate τους before πατερας (Acts 7:12); συνγενειαν was changed to συγγενειαν by the addition of γ above (Acts 7:14); μετ αυτου is written above (Acts 10:38), and γεγονως was changed to γεγονος by adding an ο above (Acts 13:12). 3.24 P46 According to James R. Royse, there are 56 corrections by the second hand of P46 (Chester Beatty II + P.Mich. Inv. 6238), 42 of which occur in Hebrews.70 On three occasions, omissions by parablepsis are corrected with an interlinear or marginal placement of an ancora. At Heb 8:8, “Kenyon misread the sign marking the addition as a sigma, and thought that the second hand changed 68

These fragments are identified and distinguished by their inventory number in the Austrian National Library as P.Vindob. K7541, K7384, K7396, K7914, K7542, K7543, K7544, K7545, K7546, K7426, K7731, K7377, K7547, K7912, and K7548. 69 Matt 20:24–32, 21:13–19, 25:41–26:39; Mark 4:36–40, 5:15–26, 5:38–6:3, 16–25, 36–50, 7:3–15, 7:25–8:1, 10–26, 8:34–9:9, 18–31, 11:27–12:1, 5–8, 13–19, 24–28; Luke 6:31–41, 6:45–7:7, 9:26–41, 9:45–10:1, 6–22, 10:26–11:1, 6–25, 28–46, 11:50–12:12, 18–37, 12:42–13:1, 6–24; 13:29–14:10, 17–33; John 4:51, 54, 5:21, 24, 10:7–25, 10:30–11:10, 18–36, 42–57; Acts 4:27–36, 5:10–21, 30–39, 6:7–7:2, 10–21, 32–41, 7:52–8:1, 14–25, 8:34–9:6, 16–27, 9:35–10:2, 10–23, 31–41, 11:2–14, 11:24–12:5, 13–22, 13:6–16, 25–36, 13:46–14:3, 15–23, 15:2–7, 19–27, 15:38–16. 70 James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill, 2007), 235.

Marginalia in New Testament Greek Papyri

347

λεγει to λεγεις.”71 The lower margin, now lost, probably contained ἰδοὺ ἡμέραι ἔρχονται, λέγει (“behold, days are coming, it says”). At Heb 9:14b, an interlinear ancora is evident, which marks the place where the now lost lower margin contained καθαριεῖ τὴν συνείδησιν ἡμῶν ἀπὸ νεκρῶν ἔργων εἰς τὸ λατρεύειν θεῷ ζῶντι (“cleanse your conscience from dead works to serve the living God”). At Heb 12:6–7, the ancora by the second hand also marks the original scribe’s longer omission, which was also presumably in the, now lost, lower margin, μαστιγοῖ δὲ πάντα υἱὸν ὃν παραδέχεται. εἰς παιδείαν ὑπομένετε, ὡς υἱοῖς ὑμῖν προσφέρεται ὁ θεός. τίς γὰρ υἱὸς ὃν οὐ παιδεύει (“and chastises every child whom he accepts. Endure trials for the sake of discipline. God is treating you as children; for what child is there whom [a parent] does not discipline?”). 3.25 P47 P47 (Chester Beatty III) is a ten-fragment third-century codex of portions of Revelation (9:10–11:3; 11:5–16:15; 16:17–17:2) and includes at least eighteen superlinear additions: χαλκα becomes χαλκεα by addition of ε above (Rev 9:20); an additional υ was added in the right margin for ουρανου (Rev 10:1); the numeral ζ with superlinear bar was added superlinearly (Rev 10:4); there are three occasions where the scribe either adds a letter-size mark (Rev 11:7), or writes ς above (Rev 11:11, two examples, above υ and above α) in an attempt to maintain a regular text-block justification. Other examples include ο above ο in μικρους (Rev 11:18), τ above τ in γαστρι (Rev 12:2), παι changed to και by addition of κ above (Rev 12:2), ο above ο in τριτον (Rev 12:4), ω above ω in των (Rev 12:4), γ above γ in καταφαγη (Rev 12:4), μ above μ in ερημον (Rev 12:6), την υναικι to τη γυναικι by erasure of ν and addition of γ above (Rev 12:16), μου was changed to του by erasure of μ and addition of τ above (Rev 14:10), μωσεως to μωυσεως by addition of υ above (Rev 15:3), εβλασφημησαν to εβλασφημουν by erasure of ησα and adding ου superlinearly (Rev 16:11), and ιδου to ειδου by addition of ε above (Rev 16:15). 3.26 P50 P50 (P.Yale I 3)72 is a bifolium containing two stories from Acts regarding an early church leader approaching a non-Jewish prospective convert (Philip and Ethiopian in Acts 8; Peter and Cornelius in Acts 10). Some have suggested that 71 Royse, Scribal Habits, 235 n. 175. 72 Carl H. Kraeling, “P50: Two Selections from Acts,” in Quantulacumque: Studies Presented to Kirsopp Lake, ed. Robert P. Casey, Silva Lake, and Agnes K. Lake (London: Christophers, 1937), 163–172; John Oates, Alan Samuel, and C. Bradford Welles, Yale Papyri in the Beinecke Rare Book and Manuscript Library (Atlanta: Scholars, 1967), 15–21.

348

Theophilos

this text “may have been designed to promote evangelism to Gentiles.”73 The scribe seems to have prematurely drawn a horizontal line to separate the passages, but then realized he omitted a word (ἄφωνος, “silent”), which was subsequently added below the line and then re-drew the horizontal line. 3.27 P55 P55 (P.Vindob. G 26214) consists of Jn 1:31–33 and Jn 1:35–38 (sixth/seventh century CE), and is a good example of ερμηνει[α] (ἑρμηνεῖαι) manuscripts, which consist of (1) a biblical passage, (2) the word ἑρμηνεία (“interpretation,” “explanation”) centered underneath, and (3) explanatory comments on the biblical passage. The implications that P55 and other manuscripts of John like it74 are biblical quotations from hermeneiai, rather than continuous texts of the New Testament, have not yet been appreciated by New Testament textual critics. The hermeneia manuscripts of John in Greek include five papyri: P55, P59, P63, P76, and P80.75 3.28 P66 P66 (P.Bodmer II + Chester Beatty XIX + P.Köln Inv. Nr. 4274/4298) consists of 39 folios (= 78 leaves, 156 pages) containing John 1:1–6:11; 6:35–14:26, 29–30; 15:2–26; 16:2–4, 6–7; 16:10–20:20, 22–23; 20:25–21:9, 12, 17. Royse enumerates 465 corrections,76 but even he admits the difficulty of certainty here, echoing Gordon Fee who states, “At other places one cannot tell whether a blank space is the result of the crossing out of a letter, or simply a rough spot on the papyrus.”77 It is possible that there were two correctors at work within the manuscript. Fee, for example, notes 13:19a where απ αρτι λεγω υμιν προ is 73 74

Comfort and Barrett, eds., Text of the Earliest New Testament Greek Manuscripts, 363. The hermeneia manuscripts of John in Greek include five papyrus manuscripts: P.Vindob. G 26214 [P55], P.Ness. 2 3 [P59], P.Berlin 11914 [P63], P.Vindob. G 36102 [P76], P.Monts. Roca 83 [P80], and three parchment fragments, the lost parchment from Damascus [0145], P.Berlin 3607 + 3623 [0210], and P.Berlin 21315 [0302]. 75 See further discussion and research findings in Stanley E. Porter, “Textual Criticism in the Light of Diverse Textual Evidence for the Greek New Testament: An Expanded Proposal,” in New Testament Manuscripts, ed. Thomas J. Kraus and Tobias Nicklas, TENTS 2 (Leiden: Brill, 2006), 305–337; Wally V. Cirafesi, “The Bilingual Character and Liturgical Function of ‘Hermeneia’ in Johannine Papyrus Manuscripts: A New Proposal,” NovT 56.1 (2014): 45–67; and Stanley E. Porter, “The Use of Hermeneia and Johannine Papyrus Manuscripts,” in Akten des 23. Internationalen Papyrologenkongresses. Wien, 22–28. Juli 2001, ed. Bernhard Palme (Vienna: Österreichischen Akademie der Wissenschaften, 2007), 573–580. 76 Royse, Scribal Habits, 409. 77 Fee, Papyrus Bodmer II, 57 n. 2.

Marginalia in New Testament Greek Papyri

349

an interlinear insertion. In this case the υ and μ are so dissimilar to the text that they “simply demand a second hand.”78 Either way, marginalia are used frequently in conjunction with insertion marks and arrows. For example, in P66, L1r l. 18 at John 1:19 has an insert marker, which presumably intended to include προς αυτον, the lower margin fragment is unfortunately missing. L7r l. 11 at John 3:17 has an insert mark with accompanying marginalia in top margin consisting of a 17-letter omission, ινα κρινα τον κοσμον. L8v l. 16 at John 3:31 has an insert mark with [και] εκ της γης λαλει in the partly preserved margin. L12r l. 5 at John 4:40 has an insert mark with εποιησα in the right margin. L15r l. 5 at John 5:28 has the marginal addition of σοι in the left margin, which then underwent a further change to οι by erasure of the first letter. 3.29 P72 P72 (P.Bodmer VII–VIII) contains the entire text of 1–2 Peter and Jude.79 Dated to the third/fourth century, it is the earliest extant manuscript of this portion of the New Testament.80 With respect to marginalia and paratextual elements, there are four features worthy of note. First, the numbering of pages is included as paratextual information, namely pagination for 1–2 Peter as 1–36, and for Jude as 62–68. Second, there are 34 instances of the superlinear addition of one or more letters, including the addition of α (1 Pet 1:21; Jude 12); δ (Jude 3); ε (2 Pet 3:3; Jude 12); εις (2 Pet 2:6); η (2 Pet 2:9); ι (1 Pet 2:7); λ (1 Pet 2:21; 2 Pet 3:9; Jude 10); μ (1 Pet 4:12); ν (1 Pet 1:17, 22; 3:18; Jude 4, 18); να (2 Pet 3:18); ον (1 Pet 1:8); ς (1 Pet 1:7, 8; 4:11, 15; Jude 11, 17); τ (1 Pet 2:22); τι (2 Pet 3:9); το (Jude 4); των (Jude 15); υ (1 Pet 1:7; 2 Pet 2:4; Jude 20); υμας (2 Pet 3:11) ω (2 Pet 3:14). Third, there is evidence of deletion by erasure (1 Pet 1:22; 3:18; 5:5; Jude 4), superlinear dot (1 Pet 3:9; 4:11; 5:5; 2 Pet 1:5, 13; 2:18; 3:6; Jude 7, 12, 13, 17, 25), and strikethrough (Jude 3). Fourth, on 16 occasions, P72 includes a marginal note. Of these, 12 consist of a phrase beginning with περι (“concerning”), which function as topic descriptions and highlighting particular themes: περι αγειοσυνη (1 Pet 1:15); περι αγνια (1 Pet 1:22); περι ιερατευμα αγιον (1 Pet 2:5); περι γενος εγλεκτον βασιλιον ιερατευμα εθνος αγ[ι]ον λαον περι ποησιν (1 Pet 2:9); περι χ̅ρυ̅ ̅ παθος εν σαρκι (1 Pet 4:1); περ σαρκος (1 Pet 4:6); περ 78 Fee, Papyrus Bodmer II, 59. 79 Michel Testuz, Papyrus Bodmer VII–IX: L’Epître de Jude, les deux Epîtres de Pierre, les Psaumes 33 et 34 (Cologny-Geneva: Bibliotheca Bodmeriana, 1959); Carlo M. Martini, Beati Petri Apostoloi Epistulae, Ex Papyro Bodmeriano VIII (Milan: Pizzi, 1968); Sakae Kubo, P72 and the Codex Vaticanus, Studies and Documents 27 (Salt Lake City: University of Utah Press, 1965). 80 Cf. P78 also dated to third to fourth centuries.

350

Theophilos

αγαπη (1 Pet 4:8); περι θ̅υ̅ κτειστη (1 Pet 4:19); περ ψεδοδιδασκαλοι (2 Pet 2:1); περι τεκνα καταρας (2 Pet 2:14); περι εμπεκται (2 Pet 3:3); περι ειρηνη (2 Pet 3:15). Notably, these marginal topical markers only appear in 1–2 Peter and are completely absent from Jude. Even so, they are applied inconsistently throughout the text and do not seem to be related to the later kephalaia. Four other marks are attested: in 2 Pet 2:22 there is an S-shaped letter above the α and also one in the right margin, which contains πμει (= true?) in Coptic; in 2 Pet 2:6, the left margin contains κατεπ[ρ]ησεν; in 2 Pet 2:8, the left margin contains ὁρασις; and in Jude 16 there is an insert mark with the lower margin containing κατα τας επιθυμιας εαυτω πορεομανοι. 3.30 P75 Amid dozens of possible examples in the 50 fragments of the second-tothird-century manuscript P75 (P.Bodmer XIV–XV), I will endeavor to highlight the most significant marginalia attested. P75 has ten small letters above δε πονηρος in Luke 11:34, possibly Coptic. At Luke 11:37, in the left margin, there are two arrows (>>) next to σαι, and a line above presumably indicating a new paragraph. Interestingly, however, in other manuscripts such arrows typically indicate insertion marks (for example P66). At Luke 17:14, an insert mark appears in text and lower margin contains θελω καθαρισθητε κ ευθεως εκαθαρισθησαν written by a second hand. Also in a second hand is γρα̥[π]τα εν τοι ουν[…] written in large letters in the upper margin at Luke 10:20. In the lower margin at John 8:22, there is a curious note in a different hand which Martin and Kasser restore as τον υον ως ̣ [κ]υ̣ρ̣ιο̣ ̣[ν] απο της τ[ρα]πεζ [ης].81 Although, ὑίος is often spelt as υον in ̣ ̣ the earlier inscriptions (most frequently Attic it is not found, as far as I can tell as an abbreviation or nomen sacrum. Like the note at 11:34, could this upside down scribal addition under 8:22 also be something in Coptic? Other interesting examples of marginalia include υπερεκχυνομενον to υπερεκχυννομενον by addition of ν above (Luke 6:38), εκβαλειν to εκβαλλειν by addition of λ above (Luke 11:18), σα to νοτου by erasure and superlinear addition (Luke 11:31), ηρξατο to ηρξαντο by addition of ν above (Luke 11:53), φοβητε to φοβηθητε by addition of θη above (Luke 12:4), κλιβανον to κλειβανον by addition of ε above (Luke 12:28),

81 Victor Martin and Rodolphe Kasser, eds., Papyrus Bodmer XIV–XV: Evangiles de Luc et Jean, Vol. II. Evangile de Jean, chap. 1–15 (PapyBod XV) (Cologny–Geneva: Bibliotheca Bodmeriana, 1961).

Marginalia in New Testament Greek Papyri

351

δε above τουτο γινωσκετε (Luke 12:39), and γαλαλιας to γαλειλαιας by addition of ε above (Luke 17:11). On eight occasions a paragraphos in the margin indicates a new textual division: above παρεγενετο (Luke 8:19), above γους (Luke 9:28), above ρουσθαι (Luke 9:51), above κ̅ς̅ (Luke 10:1), above εξει (Luke 11:5), above σαι (Luke 11:35), above ο (John 4:1), and above το (John 10:19). 3.31 P126 P126 (PSI 1497) is a mid-fourth-century single column fragment containing parts of Hebrews 13 (13:12–13, 19–20).82 The extant papyrus preserves a small portion (9.1 × 3.7 cm.) of a single codex leaf which preserves a 2 cm. external margin, and 2.4 cm. upper margin. Pagination is present on both recto (ΡΞΑ = 161) and verso (ΡΞΒ = 162). Guido Bastianini calculates that the entire Epistle to the Hebrews would have taken up 45 pages, starting on page 118 and ending on 162.83 Claire Clivaz suggests that “the numbering of the pages indicates a place for Hebrews among the Pauline letters”84 as the preceding 117 pages are too numerous for the inclusion of Romans, which Clivaz calculates would occupy only half the available pages (60/117 according to the spacing of the writing).85 4

Summary of Findings

Table 13.2 summarizes the findings of our analysis. Details provided include (1) reference to the P number as allocated by the Institut für neutestamentliche Textforschung (INTF) (Institute for New Testament Textual Research) at the University of Münster, Germany, (2) the published papyrological reference, (3) the generally accepted date (approximate only), (4) the form of the manuscript as codex or roll, (5) the provenance within Egypt, if it is known, (6) the type of marginalia attested in the manuscript under discussion, and (7) the location of the marginalia on the physical surface of the papyrus.

82

Guido Bastianini, ed., “1497. NT HEBR. 13:12–13; 19–20”, in PSI: Papiri Greci e Latini.Volume 15 (Florence: Istituto papirologico, 2008), 171–172; Claire Clivaz, “A New NT Papyrus: P126 (PSI 1497),” EC 1 (2010): 158–162. 83 Bastianini, “1497,” 171. 84 Clivaz, “NT Papyrus,” 159. 85 Clivaz, “NT Papyrus,” 159.

352

Theophilos

Table 13.2 Marginalia in Papyri

Reference New Testament Papyrus

Date (circa)

Form

Egyptian Provenance (if known)

Marginalia

Location

P1

P.Oxy. 2

250

Codex

Oxyrhynchus

fly-leaf

P2

Inv. 7134

550

Codex

P3

Pap. G. 2323 600

Codex

P4

Codex

Coptus

Codex

Oxyrhynchus

P7

Suppl. Gr. 200 1120 P.Oxy. 208, 250 1781 P. Copt. 379, 350 381, 382, 384 Petrov 553 300

Codex

P9 P10

P.Oxy. 402 P.Oxy. 209

250 350

Codex Codex

Oxyrhynchus Oxyrhynchus

P12 P13

P. Amherst 3b250 P.Oxy. 657, 300 PSI 1292

Codex Roll

Oxyrhynchus

P15

P.Oxy. 1008 250

Codex

Oxyrhynchus

P23

P.Oxy. 1229 225

Codex

Oxyrhynchus

P27

P.Oxy. 1355 250

Codex

Oxyrhynchus

title, summary of content cross indicating transition to a different language title, and or pericope heading punctuation, ekthesis, title textual varient, punctuation transitional marker paragraphos, right pointing angular brackets correction named individual, mercantile reference, pagination quotation pagination, corrections, additions bird-beak coronis, exthesis, pagination, corrections corrections

P5 P6

Codex

left

top

fly-leaf interlinear interlinear, lower left left

interlinear below text but above lower margin, top top top, interlinear, left left

top, left, interlinear interlinear

353

Marginalia in New Testament Greek Papyri Table 13.2 Marginalia in Papyri (cont.)

Reference New Testament Papyrus

Date (circa)

Form

Egyptian Provenance (if known)

Marginalia

Location

Oxyrhynchus

pagination, angular line fillers decoration pagination pagination reading marks pagination correction

top, right

P30

P.Oxy. 1598 250

Codex

P34 P35 P36 P37 P39 P40

Pap. G. 39784 650 PSI 1 450 PSI 3 550 P. Mich. 137 300 P.Oxy. 1780 P. Bad. 57 250

Codex Codex Codex Codex Codex Codex

P45

P. Chest. B. I 250

Codex

P46

P. Chest. B. II 200

Codex

P47

P. Chest. B. III250

Codex

P50

P. Yale I 3

400

Codex

P55

Pap. G. 26214 600

Codex

P66

P. Bodmer II 200

Codex

Dishnah

P72

P. Bodmer VII; P. Bodmer VIII

300

Codex

Dishnah

Oxyrhynchus Fayum Oxyrhynchus

correction, textual varient textual addition, textual correction, ancora textual correction textual division, textual addition textual division, heading textual correction, insertion marks, textual addition pagination, textual corrections, topic descriptions

bottom top top interlinear top interlinear interlinear interlinear, left

interlinear left, interlinear left, interlinear interlinear, left, top

top, interlinear, left, right

354

Theophilos

Table 13.2 Marginalia in Papyri (cont.)

Reference New Testament Papyrus P75

P126

Date (circa)

Form

Egyptian Provenance (if known)

Marginalia

Location

textual addition, left, double angular interlinear bracket, insert marker, textual correction, textual division PSI inv. 1479 350 Codex pagination top Manuscripts with no discernible extant marginalia: P8, P11, P14, P16, P17, P18, P19, P20, P21, P22, P24, P25, P26, P28, P29, P30, P31, P32, P33, P38, P40, P42, P43, P44, P48, P49, P51, P52, P53, P54, P56, P57, P58, P59, P60, P61, P62, P63, P64, P65, P67, P68, P69, P70, P71, P73, P74, P76, P77, P78, P79, P80, P81, P82, P83, P84, P85, P86, P87, P88, P89, P90, P91, P92, P93, P94, P95, P96, P97, P98, P99, P100, P101, P102, P103, P104, P105, P106, P107, P108, P109, P110, P111, P112, P113, P114, P115, P116, P117, P118, P119, P120, P121, P122, P123, P124, P125, P127, P128, P129, P130, P131, P132, P133, P134, P135, P136. P. Bodmer XIV; P. Bodmer XV

200

Codex

Dishnah

5 Conclusion In conclusion, we return to several questions raised in the second section of our discussion. Although the fragmentary nature of the papyri can make measurements imprecise, it was found that the width of the margin did not affect the frequency, quality, or length of the marginal note. This observation is consistent with McNamee’s sample of Greco-Roman texts, that not a single manuscript “has marginalia that come anywhere near filling the available space.”86 It seems then that the purpose of a broad margin is not found in the desire or capacity of marginal annotations.87 86 McNamee, Annotations, 13. 87 William Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004), 156–157, suggests that a broad margin is a mark of luxury production.

Marginalia in New Testament Greek Papyri

355

In Greco-Roman codices, the annotation most regularly occurs in the left margin;88 for example, titles of poems (Bacchylides [MP3 175], Pindar [MP3 1361], and prose paraphrase of Homer’s Iliad [MP3 1169]89). This tendency was also observed in the New Testament papyri. However, marginalia can also occur at the top (P3, P10, P12, P13, P23, P30, P35, P36, P39, P66, P72, P126), bottom (P34), or right (P72) margin. In classical book rolls, marginalia are also attested in the top margin [MP3 63, 166, 473, 485, 1506].90 Greco-Roman texts are often marked with special sigla such as the antisigma (Ↄ) for variants and the ancora (⸔) for corrections,91 often accompanied by paragraphoi (MP3 63, 78, 117, 119, 145, 157, 179, 186, 201, 332, 474, 500.6, 536, 543.3, 616, 998, 1237, 1356, 1360, 1361, 1369, 1414, 1420, 1473, 1485.1, 1485.5, 1487, 1506, 1534, 1551, 1957.1, 2054, 2866, 2925, 2953, 2966.1, 2974).92 William Johnson also notes that a single dot can function in the same manner.93 Related features of these aspects were also found in the New Testament papyri. Classical marginalia also regularly included textual variants, especially in the popular literary works such as Homer’s Odyssey [MP3 1127].94 Similar attestation of textual variation and scribal correction was found in the New Testament papyri (P5, P45, P46, P47, P50, P55, P66, P72, P75). One conspicuous absence in the marginalia of the New Testament papyri that was a common feature of the classical texts was the lack of a form of οὕτως (“thus,” “in this manner” in the marginalia; cf. MP3 145, 360, 1360) or ζήτει (“look it up”; cf. MP3 62, 103, 362, 1360, 1361, 1473, 1910).95 Interlinear space was also noted as being commonly used for annotation, including glosses, metaphrases, textual observations, or comments, although the latter is infrequent because of the lack of physical space to do so (P4, P5,

88 In book rolls, however, the annotation more regularly occurs in the right margin. 89 MP3 refers to Mertens and M.-H. Marganne, Mertens-Pack3 online (http://promethee .philo.ulg.ac.be/cedopal!indexsimple.asp [accessed on July 2017]). Cited in McNamee, Annotations, 15. 90 Cited in McNamee, Annotations, 17 n. 39. 91 Kathleen McNamee, Sigla and Select Marginalia in Greek Literary Papyri (Brussels: Fondation Égyptologique Reine Élisabeth, 1992), 11–15 and Table 2. 92 Cited in McNamee, Annotations, 19 n. 49. 93 Johnson, Bookrolls, 266. 94 Cited in McNamee, Annotations, 15. 95 Cited in McNamee, Annotations, 15, 23, 37, 40, 50, 52, 61, 66, 67, 68, 72, 75, 97, 124. An annotation that appears exclusively in Greco-Roman book rolls is the chi-rho monogram, which most commentators take as an abbreviation for χρ(ηστόν) or perhaps another word derived from the same root, and indicates passages which the reader found helpful; see McNamee, Annotations, 23.

356

Theophilos

P9, P13, P23, P27, P37, P39, P40, P45, P46, P47, P50, P55, P66, P72, P75; cf. MP3 55, 1840, 1895).96 The source marginalia in Greco-Roman texts is often assumed to be traditional commentaries, glossaries, or student annotation (see Table 13.1 above). One intriguing example, MP3 2958, includes marginalia such as ὡραῖον (“good point”) and σημείωσαι (“take note”).97 The situation is considerably more opaque with the New Testament papyri. It may be the case that, on occasion, the marginalia were copied directly from an exemplar, but on other occasions it is clear that the marginalia are secondary and postdate the manuscript by up to two centuries.98 To address these questions one must appeal primarily to paleography. However, more recent advances in imaging technology and ink analysis methods (e.g., multi-spectral imaging) often have the capacity to identify multiple hands at work based on the distinctive spectral signature obtained from the analysis.99 More work remains to be done in this area. We began our discussion with an extraordinary example of a marginal note finding its way into the actual text of the New Testament (2 Cor 8:4) in a parchment codex, namely Codex Corsendoncensis (GA 3, 12th century). We conclude in a similar fashion by noting one of the most distinctive (and unsympathetic) marginal comments in the extant New Testament (parchment) codices, namely the marginal reading at Heb 1:3 in Codex Vaticanus (GA 03, fourth century). The marginal note concerns, on the one hand, the variation between B* and B2, which read φανερῶν τε τὰ πάντα τῷ ῥήματι τῆς δυνάμεως αὐτοῦ (“manifesting all things by the word of his power”), and, on the other, B1, which reads φέρων τε τὰ πάντα τῷ ῥήματι τῆς δυνάμεως αὐτοῦ (“upholding all things by the word of his power”). B2 then adds the marginalia of interest, ἀμαθέστατε καὶ κακέ, 96 97

Classical sources cited in McNamee, Annotations, 16. Other later examples from the fifth century include ἀπὸ φωνῆς, indicating they were transcribed from oral teaching. M. Richard, “ἀπὸ φωνῆς,” Byzantion 20 (1950): 191–222, notes that by the ninth century however the phrase meant “according to” without the implication of orality. I am grateful for an anonymous reviewer who noted some examples of marginalia on later Christian codices such as 03, whose later hands have added ωρ(αιον), ση(μειωσαι), and χρ(ηστον). 98 Classical texts abound with scribes making annotations in their text during the process of composition (MP3 78, 81, 157, 202, 373.2, 412.21, 917.3, 1368). For more substantial comments of three or more lines, the upper and lower margins are often used (MP3 47, 55, 59, 71.1, 78, 79, 157, 361, 371, 473, 517.1, 551, 998, 1039,1237, 1297.6, 1360, 1361, 1391). Cited in McNamee, Annotations, 16 n. 35. 99 Michael P. Theophilos, “Multispectral Imaging of Greek Papyrus Fragments from Oxyrhynchus,” in Reading New Testament Papyri in Context/Lire les papyrus du Nouveau Testament dans leur context, ed. Clare Clivaz and Jean Zumstein, BETL 242 (Leuven: Peeters, 2011), 269–280.

Marginalia in New Testament Greek Papyri

357

ἄφες τὸν παλαιόν, μὴ μεταποίει (“you most ignorant and wicked person; leave the old [reading] alone; do not change it”). However, as the above examples in the papyri have attested, later readers can be grateful for the illumination that the paratextual material provides in discerning the textual history of the New Testament text. This chapter has attempted to demonstrate that features appearing in the margins and interlinear portions of a New Testament papyrus, first, have been neglected in discussions of the value, nature, and weight of textual support, and second, can have significant implications for the nature, use, and genre of the papyrus concerned. An exploratory analysis of the marginalia in New Testament papyri revealed their varied forms and functions: document title, pericope title, sigla indicating transition from one language to another, textual amendment, textual amendment through interlinear addition, insert marker, subject heading, or scribal glosses. Attention to the marginalia provides a fascinating, if fragmentary, glimpse into the complicated history of the transmission of not just the textual component of the papyrus but also the function, reception, use, re-use, and editorial activities that a manuscript has undergone.

Conclusion

Paratextual Features: Summary and Prospects Stanley E. Porter, Chris S. Stevens, and David I. Yoon The editors and contributors of this volume have sought to address various areas within the fields of textual criticism and papyrology. Each chapter seeks to give detailed attention to the physicality of the manuscripts by drawing attention to their paratextual features rather than treating them as texts divorced from their medium. Lamentably, the study of the New Testament is often myopically fixated on the original reading, or source text, while paying little to no attention to the manuscripts themselves and the value of paratextual features. However, as expressed in Chapter 1, this raises a more fundamental question of the appropriateness of making a significant divide between text and paratext. The first chapter delves into this discussion of definitions and the difficulties faced in such attempts. Despite this difficulty, each contributor has been at liberty to address features of New Testament manuscripts that have been more or less neglected in place of individual readings. Manuscripts are treated not simply as witnesses to the “original,” “source,” or Ausgangstext text, but as texts in and of themselves that merit interpretations. There are at least three concluding observations to be made from these chapters regarding paratextual features of New Testament manuscripts. First, paratextual features are codified representations of the interpretations of the biblical text by the scribes, copyists, or any editor with a marking device. For example, at the end of Chapter 13, an example is given of scribal interactions from Codex Vaticanus at Heb 1:3. First, B* reads φανερῶν τε τὰ πάντα τῷ ῥήματι τῆς δυνάμεως αὐτοῦ (“manifesting all things by the word of his power”). Then B1 reads φέρων τε τὰ πάντα τῷ ῥήματι τῆς δυνάμεως αὐτοῦ (“upholding all things by the word of his power”). Finally, B2 adds the marginalia, ἀμαθέστατε καὶ κακέ, ἄφες τὸν παλαιόν, μὴ μεταποίει (“you most ignorant and wicked person; leave the old [reading] alone; do not change it”). Apparently, the third corrector was angry with the second corrector and contended the original manuscript reading (prima manus) should be retained. This brief intrascribal banter is part of the history of interpretation for Heb 1:3. Another fine example concerns the subscriptions that exist in over 300 manuscripts, which serve to help readers understand the history of interpretation of the biblical text, and the various

© Stanley E. Porter et al., 2023 | doi:10.1163/9789004537972_016

Conclusion

359

scribal markings such as diple that signal a scriptural citation or a composite citation in the manuscripts. Second, some paratextual features can serve as reading aids, whether public, private, or communal, and may not reflect the earliest extant versions of the biblical text. For instance, the addition of punctuation in modern editions of the Greek New Testament aids in reading in contrast to earlier manuscripts written in scriptio continua but is founded on sometimes disputable interpretive decisions. Similarly, Coptic manuscripts and other ancient translations of the Greek New Testament may evince features such as paragraph divisions and initials for public reading and interpretations. Furthermore, the functions of segmentation, short lines, ekthesis, and paragraphing can be considered to facilitate public or private reading and performance, although they may also reflect interpretations of the text. While these features may serve to aid in public reading, there also exist miniature codices that contain shorter passages of Scripture, such as the Lord’s Prayer, and these would have made mobility and accessibility much easier for private reading. Finally, paratextual features, as they are typically called, should be considered as a part of the text of the manuscript and reading itself. For example, the use of nomina sacra that characterizes many of the majuscules reflects a potential system that demarcated certain words or names as “sacred,” although the parameters of this system have been debated in biblical scholarship. As another example, Paul’s well-known ending of certain letters, such as Galatians with “large letters,” and subsequent scribes reflecting the size of these large letters, are parts of the manuscripts themselves. The fact that certain scribes reflected in their copies the large letters of Paul may indicate their fidelity to maintaining their exemplar’s sense, meaning, and paratextual features. This volume presents a collection of essays on textual and paratextual features of New Testament manuscripts that hopefully fosters further scholarly discussion and interest in these important historical artifacts. The editors believe this modest goal has been achieved through the important contributions of our colleagues.

Index of Ancient Sources Old Testament Genesis 1:1–4:2 1:1–5 15:13–14 22:18 26:4

60 132, 314, 342 245 241, 250 241

Exodus 2:22 3:12 3:14 33:17–23

245 245–246 109 98

Leviticus 23:29 26:11–12

240, 250 262

Numbers 23:22 24:8

241 241

Deuteronomy 18:15–16 18:19 25:5 27:26 30:10

240, 250 240, 250 244 254 254

2 Samuel 5:2

242

Esther 2:2

275

Job

41:3

Psalms 5:10 9:28 10:7 13 13:1–3 13:3 19:7–8

263–264, 266 242 259 242 241–242 242, 259 259 314

35:2 52 90 90:1 118(117):25–26 139:4 Proverbs 1:16

242 241–242 316 314 236 242 242

Ecclesiastes 7:20 242 Isaiah 6:10 45, 52–53, 237 10 254 10:22 243, 254 26:20 235, 239 235, 255 27:9 28:16 254 40:13 263–264, 266 41:4 109 42:18–20 237 44:6 236 44:18 237 45:23 247, 254 52:11 262 53 238 53:1 237 53:4 238 53:12 238 59:7–8 242, 259 59:20–21 235, 255 62:11 246 64:3 255 65:12 255 Hosea 2:1 2:25

250 250

Micah 5:1–3

242

Habakkuk 1:5

241

362

Index of Ancient Sources

Habakkuk (cont.) 2:3–4 235–236, 238–239, 255 2:4 238 Zaphaniah 3:15 3:16

236 252

Zechariah 9:9

246, 252

New Testament Matthew 1:1–9 335 1:12 335 1:14–20 335 1:17 109 2:6 242, 244 2:14 335 2:15 241 5:21 236, 244 6:7 232 6:10–12 318 6:33–34 340 340 7:2 13:35 241 20:24–32 346 21:5 242–243, 246–247, 252 21:13–19 346 21:13 227 61 25:1 25:12–15 344 25:20–23 344 25:41–26:39 346 26:19–37 345 26:24 345 26:26 345 26:37–52 345 26:38 345 27:9–10 240, 252 28:19 94, 107 Mark 1:2–3 1:2 1:3

252–253, 257, 266 242–243, 249–250 242–243, 250

242 1:4 4:36–40 346 346 5:15–26 5:38–6:3 346 346 6:16–25 6:36–50 346 7:3–15 346 346 7:25–8:1 8:10–26 346 8:34–9:9 346 346 9:18–31 11:17 227 11:27–12:1 346 12:2–5 61 12:5–8 346 12:10–11 240 12:13–19 346 12:24–28 346 14–16 229 14:62 240, 244 Luke 338 1–6 1:58–59 338 1:62–2:1 338 1:76 338 1:80 338 2:1 338 2:6–7 338 3:8–4:2 338 338 3:14 3:19 338 3:23 338 4:1–2 340 4:18–19 240, 249 4:29–32 338 4:34–35 338 5:3–8 338 5:30–6:16 338 5:36 338 6:8 338 6:12 338 6:31–41 346 6:45–7:7 346 7 337 7:22–26 336 7:36–45 337 7:50 336

Index of Ancient Sources Luke (cont.) 8 8:19 9:26–41 9:28 9:36 9:45–10:1 9:51 9:58 10:1 10:6–22 10:26–11:1 10:38–42 11:5 11:6–25 11:28–46 11:31 11:34 11:35 11:37 11:50–12:12 11:53 12:4 12:18–37 12:18 12:42–13:1 13:6–24 13:29–14:10 14:17–33 17:11 17:14 19:46 20:28

337 351 346 351 346 346 351 346 182, 351 346 346 337 351 346 346 350 350 351 350 346 350 350 346 350 346 346 346 346 351 350 227 243, 244

John 1–4 1:1–6:11 1:1 1:5–6 1:23–31 1:31–33 1:33–40 1:35–38 1:38 2:18 3:14–17 3:17–18 3:17

280 348 61, 63, 73–74 320 338 348 338 348 339 7, 4, 63 345 345 349

363 345 3:31–32 3:31 349 345 3:34–35 4:1 351 349 4:40 4:51 346 4:54 346 116 5:20–23 5:21 346 5:24 346 349 5:28 6:35–14:26 348 8:14–22 345 8:22 350 8:25 9, 35–38, 40–41, 48, 57 8:43 9, 46–51, 57 8:44 44, 46, 51 8:45 9, 41–46, 57 10:1–12:20 339 10:1–2 339 10:4–7 339 10:7–25 346 10:9–10 339 10:14 346 10:19 351 10:30–11:10 346 10:42 339 11:1–8 339 11:1 339 346 11:18–36 11:42–57 346 11:44 339 11:45–52 339 11:45 339 11:47 346 11:52 339 12:12–15 337 12:13 236 12:38 237, 244 12:39 244­–245 12:40 9, 45, 52–57, 235, 237, 244, 249 12:41 252 13:1–2 339 13:11–12 339 14:29–30 348 15:2–26 348

364 John (cont.) 16:2–4 348 348 16:6–7 16:10–20:20 348 339 16:14–30 16:19 339 16:20 339 339 16:21 19:26 115 19:36–37 244­–245 20:11–17 338 20:19–20 338 20:22–25 338 20:22–23 348 20:25–21:9 348 20:31 96 21:12 348 21:17 348 various 59–85 Acts 3:13 237 3:25 241, 250 3:24 240 4:27–36 346 346 5:10–21 5:30–39 346 5:38 346 6:7–7:2 346 6:10–21 346 6:32–41 346 7:6–7 245–246 7:7 245 7:12 346 7:14–25 346 7:14 346 7:52–8:1 346 8 321 8:16–27 346 8:26–32 320 8:34–9:6 346 9:35–10:2 346 10 321 346 10:10–23 10:28–31 320 10:31–41 346 10:38 346 11:2–14 346

Index of Ancient Sources 346 11:24–12:5 12:13–22 346 346 13:6–16 13:12 346 237 13:22 13:25–36 346 13:41 241 346 13:46–14:3 14:15–23 346 15:2–7 346 237 15:16–18 15:19–27 346 15:38–16 346 16:22–23 149 16:26–32 229 17–22 346 17:28–31 346 17:34–18:2 346 18:17–18 346 18:22–23 346 19:10 196 19:13–16 346 19:18–19 346 20:22–24 346 20:26–28 346 22:11–14 346 22:16–17 346 23:27 233 26:23 240 26:29 109 212 28:14–31 28:16 212 28:20 212 28:26–27 55 28:28 55 28:30 212 Romans 1:1–7 1:7 1:17 1:24–27 1:24 1:30–31 1:31–2:3 2 2:12 2:13

133 258–259 236, 260 345 258 279 345 130 130 130

Index of Ancient Sources Romans (cont.) 2:24 260 130 2:29 3:4 260 3:7 258 3:10–18 241–242, 244, 248, 254, 260 3:10–14 259 3:10–12 233 3:10 230, 259 3:12–18 259 3:12 259 3:13–18 259 3:14 259 3:15–18 242 3:21–4:8 345 3:21 258–259 4 132 4:3 260 4:6 258 4:7–8 260 4:12 258 4:17 260 4:18 258, 260 5 132 5:5 258 5:9 258 5:15 117, 258 6:4–5 345 6:15–16 345 7:7 260–261 8 134 8:12–22 344 8:21 274 8:24–27 344 8:33–9:3 344 8:36 260 9 134, 249 9:1 134 9:5–9 344 9:7 260 9:9 254–262 9:12 260 9:13 260 9:15 260 9:16–17 345 9:17 260 9:18–26 250

365 254, 261–262 9:20 9:25–26 240, 250–251, 254, 260 9:25 237, 250 9:26 251 243, 254–255, 260 9:27–28 9:27 237, 345 260 9:29 9:33 254, 260 10–11 253 10:5 260 10:6 230, 237, 260–261 10:8 260 10:11 260 10:13 260 10:15 260, 261–262 10:16 260 10:18 260 10:19 260 10:20–21 260 10:20 260–261 11:3 260 11:4 260 11:8–10 249 11:8 235, 237, 260–261 11:9–10 260 11:26–27 235, 243, 255, 260–261 11:34–35 260–261, 263–264, 266 11:36–12:1 229 229 12:1–15:32 12:1 229 12:19–20 260 13:7 277 13:9 237, 260–262, 264 14:11 254, 260–261 14:21 237, 261 15 130 15:3 260 15:9 260–261 15:10 260 15:11 260 15:12 260 15:21 260 15:24 213 15:26 130 15:27 130

366 Romans (cont.) 16 16:1–2 16:5 16:6 16:22

Index of Ancient Sources 130 152 130 130 151

1 Corinthians 1:3 116 1:19 260 1:31 260, 263 2:9 235, 255–256, 260–264 2:16 260 3:19 260–261 3:20 260–261 6:9–10 279–280, 285 6:16 260 7 10, 128, 136 7:18–8:4 343 7:24 128 7:25–40 137 7:25–38 137 7:25 128 7:35–36 129 7:35 128 7:36–38 137 7:36 128 7:38–39 129, 137 7:38 136–137 7:39–40 137 7:39 124, 129, 136–137, 144, 343 7:40 129 8:1–14:40 343 8:1 129, 343 8:5–6 93 8:6 94 9:9 260 10:1 264 10:7 260 10:26 260 11:5 139 14 10, 131, 137–139 14:21 234, 260–262 14:33–34 138 14:33 124, 131, 137–139, 144 14:34–36 138 14:34–35 7, 8, 137, 139

14:34 15 15:27 15:32 15:42 15:45 15:50 15:53–16:9 15:54–55 16:4–7 16:10 16:21

131, 138–139, 144 131 260 260 274 260 274 278 235, 245, 260–262 344 344 146, 151, 168

2 Corinthians 4:13 5:18–21 6:2 6:16–18 6:16 6:17 7:1 8:4 8:15 9:7 9:9 10:13–14 10:17 11:2 11:4 11:6–7 11:11 11:19 13:1

260 344 260 247, 260–264 248, 254 248 248 330, 356 260 260 260 344 260–261 344 344 344 50–51 260 260

Galatians 1:1–6:10 1:1–5 1:1 1:3 1:6–8 1:9 1:10 1:11–12 1:13–14 1:15 1:18 1:21 2 2:1

158 299 299 299 299 299 299 299 299 299 300 300 300 300

367

Index of Ancient Sources Galatians (cont.) 3:6 260, 301 301 3:7 3:8 249, 260, 301 302 3:10–13 3:10 249, 254, 260–261 3:11 236, 260–261 260 3:12 3:13 235, 260–261 3:14 302 3:16 260 4:1 300 4:3 300 4:6 264, 300 4:8 300 4:12 300–301 4:13–15 148 4:22–23 260 4:27 260–261 4:30 260 5:1 301 5:2 300 5:13 300 5:14 260 5:22–23 279–280, 285 6:8 274 6:10–12 167 6:10 166 6:11 10, 146, 149–151, 155–156, 167 6:12–18 156, 170 Ephesians 3 3:4 3:21–4:2 4 4:1 4:2 4:4 4:8 4:14–16 4:19 4:27 4:28 5 5:14 5:15 5:18–20

131 124 131 129, 131 131 131 94 260 131 129 129 129 129, 139–140 260 140, 144 141

5:18 140–141 5:19–20 140 140 5:20 5:21 124, 139 5:22–6:9 140–141, 144 140–141 5:22 5:31 260 140 6:1 6:2–3 260 6:21 140 Philippians 1:1

166

Colossians 116 1:3 1:15 279 2:22 274 3:20 258 3:21 264 4:12–13 196 4:16 124, 196, 201, 281 4:18 146, 151, 168 1 Thessalonians 1 1:1–3 1:7 1:8 2 2:13 2:14–3:18 2:14–15 2:15 4–5 4:2 4:12–13 4:12 4:16–17 5:3 5:8–10 5:12–18 5:27

130 116 130 130 130 130 279 58 58 132 191 344 216 344 344 344 344 124, 281

2 Thessalonians 1:1–2 1:2 2:2 3:17

344 191, 216 168 146, 151, 168

368 1 Timothy 1:3 2 2:9–15 2:15 3 3:1 3:2–4 3:2–3 3:2 3:13–4:8 3:14 3:16 4 4:8 4:13 5:18

Index of Ancient Sources 196–197, 201, 215 142 142 127, 141–142 131, 141–142 124, 126–127 279 285 279 131 215 131 131 131 124, 281, 283 260–262, 266

2 Timothy 1:1–2 212 1:16 212 1:17 217 2:9 212 2:15 220 2:19 258, 260 3:2–4 280, 285, 287 3:2–3 279 4:6–8 212–213 4:6 214–216 4:9 212, 224 4:11 212, 224 4:12 215–216 4:13 152–153 4:16–17 182, 212, 220 4:16 212, 214 4:17 214, 220 4:18 212 4:20 203 4:21 212 Titus 1 1:1–13 1:1–6 1:1 1:2 1:5 1:7–8 1:10

133 278 271 271 264 183 277–280, 285, 287 280

1:11–15 1:12 1:13 1:15 2 2:3–8 2:3 2:4 2:7 3:12

269, 273 274, 280 274 280, 287 133 269, 273 287 287 274–275, 287 197–198

Philemon 1–3 3 4 5 6–8 6 7–13 7 8 11 14–17 17 18–22 18–20 19 20 21

143 258 142, 144 144 131 143–144 143 124, 131, 142–144 142–143 264 143 258 143 131 131, 146, 151, 168 131, 143 131, 143

Hebrews 1–4 1:1 1:3 2:14–5:5 3:6 7:1 7:3 8:8 9:12 9:13 9:14 9:15 9:16 10:8–22 10:11 10:17 10:29–11:13 10:37–38

247 132, 314, 342 356, 358 343 343 237 237 346 129 129, 244 347 129 129 343 343 343 343 235–238, 255–256

369

Index of Ancient Sources Hebrews (cont.) 10:37 10:38 11:3 11:4 11:21 11:28–12:17 12:6–7 12:11 12:21 13:5 13:12–13 13:19–20

128, 235–236 235–236 342–343 343 237 343 347 343 237 235 351 351

James 1:10 1:11 1:13–5:20

343 344 339

1 Peter 1:7 349 1:8 349 1:15 349 1:17 349 1:21 349 1:22 349 2:5 116 2:7 349 2:9 349 2:21 349 2:22 349 2:24 238 3:9 349 3:18 349 4:1 349 4:6 349 4:8 350 4:11 349 4:12 349 4:15 349 5:5 349 5:12 160 2 Peter 1:4 1:5 1:13 2:1 2:4

274 349 349 350 349

2:6 2:8 2:9 2:12 2:14 2:18 2:19 2:22 3:3 3:6 3:9 3:11 3:14 3:15 3:18

349–350 350 349 274 350 349 274 350 349–350 349 349 349 349 350 349

1 John 4:11–12 4:14–17

340 340

2 John 12

150

3 John 13

150

Jude 1–13 1–3 3 4–5 4 7–8 7 10 11 12 13 15 17 18 20 25

319 316 349 319 349 319 349 349 349 349 349 349 349 349 349 349

Revelations 1:4 1:8 7:1 7:2–4

97 99, 118 97 103

370 Revelations (cont.) 9:10–11:3 9:20 10:1 10:4 11:5–16:15 11:7 11:18 12:2 12:4 12:6 12:16 14:10 15:3 16:11 16:15 16:17–17:2 21:6 22:13

Index of Ancient Sources 347 347 347 347 347 347 347 347 347 347 347 347 347 347 347 347 99 99

New Testament Apocrypha Acts of Peter 2:1–3 214 2:40 214 Early Church Writings 1 Clement 1.1–26.2 5.7 46.6

339 181, 213 94, 102

2 Clement 19.1

282

Barnabas 2.10 5.12 6.4 6.6 6.16 9.2 9.3 9.7–8 11.4 12.7 12.11

104 266 266 266 266 266 266 90, 94, 96 266 98 104

Ambrose De Fide 4.102

342

Augustine Doct. chr. 2.20 Serm. 286.7 Tract. Ev. Jo. 7.12

327 327 326–328

Caesarius of Arles Serm. 50.1

326

Chrysostom Hom. 1 Cor. 43.7 Hom. Jo. 11 Hom. Jo. 32 Hom. princ. Act. 3

327 322 322, 327 322

Cyprian Epist. of Cyprian 32 282 Cyril of Jerusalem Catech. 4.36

323

Eusebius Hist. eccl. 1.11 281 Hist. eccl. 2.22 182, 214, 223, 226 Hist. eccl. 3.4 220, 222–223 Hist. eccl. 6.63 282 VC 4.34–37 106 Ignatius Eph. 7.2 Eph. 18.2 Eph. 20.2 Haer. 1.10.1 Haer. 3.9.2–3 Haer. 3.11.8 Haer. 4.23.11 Phld. 8.2 Trall. 9.1–2

117 116 116 115 90 105 281 99 116

Jerome Comm. Tit. praef Ep. 22.32 Ep. 107.11 Ruf. 1.9 Vir. ill. 75

181 322 322 322 322

Josephus Ag. Ap. 2.175

283

Index of Ancient Sources Justin Martyr 1 Apol. 33 1 Apol. 67 1 Apol. 67.4 Dial. 100 Dial. 100.3 Dial. 113.3–5

115 281 282 117 116 117

Philo Som. 2.127

283

Tertullian Can. Mur. 64 Marc. 5.17 Pres. Praescr. 41

196 196 282

Theodoret Hist. eccl. 3.1

282

New Testament Papyri P1 335–336 P2 336 P3 337, 355 P4 338, 355 P5 338, 355 P6 61, 339 P7 340 P9 340–341, 356 P10 133, 340–341, 355 P12 132, 314, 342, 355 P13 127–128, 235–236, 238, 343, 355–356 P15 127–128, 132, 135–136, 343 P16 127–129 P17 127, 129 P18 96 P20 271 P23 271, 343, 355–356 P27 132–133, 356 P30 132, 191, 200, 216, 355 P32 11, 132, 133, 268–278, 286 P34 344, 355 P35 344–345, 355 P36 345, 355 P37 345, 356

371 345, 355–356 P39 132–133, 345, 356 P40 P41 346 P45 96, 312, 346, 355–356 P46 93, 127, 130, 134–140, 164, 166–167, 181, 191, 200, 212, 216, 229–230, 235, 237–238, 268, 288, 330, 342, 346, 355–356 P47 347, 355–356 P48 233 P49 127, 129 P50 320–321, 347, 355–356 P52 312 P55 348, 355–356 P59 348 P63 348 P64 236 P65 127, 130 P66 34, 39, 72, 116, 236–237, 280, 288, 348, 350, 355–356 P72 238, 349, 355–356 P74 229, 237 P75 34, 38, 72, 235, 280, 288, 312, 314, 319, 350, 356 P76 348 P78 316, 319, 349 P80 348 P87 132–133 P90 72, 288 P92 132–133 P100 271 P113 127, 130 P114 132–133 P115 115 P118 127, 130 P121 115 P123 127, 131, 137–139 P126 314, 351, 355 P132 127, 131 P133 127, 131, 184 P139 127, 131–132, 135, 143 various 352–354

372 Other Papyri, Codices, and Amulets 314, 342 P.Amh. 1.3 P.Ant. 12 311, 315, 317 P.Ant. 13 316–317, 324 318–320 P.Ant. 54 P.Berlin 3607 348 348 P.Berlin 3623 P.Berlin 11710 319 P.Berlin 11914 348 P.Berlin 21315 348 P.Berol. 1710 317 P.Berol. 9782 233 P.Col. VIII 216 158 P.Egerton 2 90, 317 P.Fayum 110 164–165 P.Grenf. 1.8 311, 313, 316 P.Heid. Inv. G. 645 345 P.L.Bat. 20 314 P.Lond.Lit. 216 312 P.Mich. III 137 345 P.Mich. V 351 159–160 P.Mich. Inv. 6238 191, 346 P.Mil.Vogl. VI 263 232 P.Monts.Roca 83 348 P.Ness. 2.3 348 P.Oxy. I 2 335–336 P.Oxy. I 6 316, 324 P.Oxy. I 103 341 P.Oxy. II 209 133, 340–341 P.Oxy. II 246 156–157 P.Oxy. III 401 311 P.Oxy. III 402 340 P.Oxy. III 405 90, 247 P.Oxy. IV 657 343 P.Oxy. V 840 310–315, 317–318, 321 P.Oxy. VI 848 315, 317 P.Oxy. VI 849 311, 315, 324 P.Oxy. VI 924 314 P.Oxy. VII 1008 343 P.Oxy. VII 1010 315, 317 P.Oxy. VII 1060 314 P.Oxy. VIII 1080 315, 317 P.Oxy. VIII 1086 233, 258 P.Oxy. VIII 1151 314 P.Oxy. X 1229 343 P.Oxy. XII 1602 311, 324 P.Oxy. XV 1780 345

Index of Ancient Sources P.Oxy. XV 1781 P.Oxy. XV 1782 P.Oxy. XXIII 2369 P.Oxy. XXIII 2585 P.Oxy. XXXIV 2684 P.Oxy. XXXVI 2770 P.Oxy. XLV 3254 P.Oxy. XLV 3255 P.Oxy. XLV 3256 P.Oxy. XLV 3257 P.Oxy. XLV 3258 P.Oxy. XLV 3259 P.Oxy. XLV 3260 P.Oxy. XLV 3261 P.Oxy. XLV 3262 P.Oxy. XLIX 3472 P.Oxy. XLIX 3487 P.Oxy. LXIII 4365 P.Oxy. LXVI 4500 P.Oxy. LXXVI 5073 P.Princ. II 107 P.Princ. III 159 P.Rainer 4.45 P.Rainer 5[13b] P.Ryl. 463 P.Teb. 1.19 P.Vindob.G. 348 P.Vindob.G. 2312 P.Vindob.G. 2323 P.Vindob.G. 26214 P.Vindob.G. 29831 P.Vindob.G. 31974 P.Vindob.G 36102 P.Vindob.G. 39212 P.Vindob.G. 39756 P.Vindob.G. 39780 P.Yale 1.3 P.Yale. inv. 989

339 313 234 341 316, 319–320 160, 162 341 341 341 341 341 341 341 341 341 161 163–164 322 315, 317 319 314 314 311 314 317 170 314 319 337 348 317, 319–320, 328 346 348 311 315–317 311 320, 324, 347 317

New Testament Majuscules ℵ (01) 91, 115–116, 136–143, 184, 191, 200, 202, 207, 216, 227, 232–233, 239, 241, 243, 254, 264, 268, 270, 272–273, 275–276, 278–280, 286–289, 295, 298, 330

373

Index of Ancient Sources New Testament Majuscules (cont.) A (02) 91, 94, 115, 141–143, 181, 184, 186, 188, 191–194, 198–199, 202–203, 206, 208, 217, 225, 227, 233, 249–251, 264, 272, 275­–276, 288 B (03) 7–8, 72, 91, 102, 115, 136–141, 164, 179, 192, 194, 227, 233, 244, 245–248, 264, 270, 286, 288, 301, 330, 356, 358 C (04) 142–143, 184, 202, 207, 216, 239, 270, 272, 275–276, 330 D (05) 39–41, 43–46, 49, 57, 186, 193, 227, 231, 233, 239, 252–253, 264, 270, 288, 330 Dp (06) 24, 49–50, 141–143, 184, 187, 192–194, 202, 207, 216–227, 231, 233, 239, 253–256, 272, 275–277 E (07) 39, 43–44, 46 Fpp (010) 50, 187, 193, 202, 207, 275 Gp (012) 187, 193, 195, 202, 227, 230, 233–234, 258–264, 275, 277 H3 (015) 176–177, 183, 193–194, 197–199, 203, 206 I (016) 184, 203, 206 K (018) 189, 195, 202, 218, 250 L (020) 190, 195, 202, 209, 218, 330 M (021) 243 N (022) 91 P (025) 190, 197–198, 207, 217, 221, 250 W (032) 115, 250, 270 Δ (037) 46, 49 Π (041) 250

Ψ (044) 187, 191, 200, 207, 275 048 184, 193, 199, 206, 217 187, 207 056 061 184, 206 075 189, 198 088 278 0101 311 0142 210 0150 189, 195, 204–205, 208, 211, 217, 223 0151 184, 198, 206 0210 348 0217 311 0220 127, 132 0241 184, 206 0259 184, 206 0260 60 0262 184, 206 0272 184, 206 0285 184, 206 0302 348 0319 192–193, 277 New Testament Minuscules 1 36 3 197, 330, 356 6 330 28 250 141, 191, 200, 275 33 81 204, 224, 226 91 199 102 197 203 197 206 196 209 330 330 177 429 196 459 198 460 191, 200 469 197 489 330 496 193, 205, 217 506 197 614 204, 217 636 193, 217 808 198

374 New Testament Minuscules (cont.) 927 330 197, 330 945 1009 250 250 1010 1079 250 1086 233 141 1175 1216 250 1230 250 1240 197 1242 250 1243 177, 275 1244 275, 330 1249 330 1253 250 1319 275 1344 250 1365 250 1367 217 1506 205 1524 224–225 1546 250 1628 330 1642* 197 250 1646 1720 177 1728 198 1735 275 1739 141, 192, 204, 218 1739* 191, 200 1758 196, 198, 204 330 1768 1828 198 1836 191, 200 1838 198 1840 199 1848 197 1854 191, 200, 216, 275 1876 330 1881 275 1900 193 1910 224–225 195 1939 1945 195 1963 195

Index of Ancient Sources 195 1996 1999 195 199 2008 2009 204 204–205, 223 2110 250 2148 2242 195 204 2289 2374 330 2401 204 217 2412 2484 204, 217 2501 198 2558 198 2772 197 2805 195 f 13 250 f33 164 various lists 184, 187–190, 206–211 Lectionaries l680 207 l884 207 l2024 187, 207 Ancient Versions of the New Testament arm 250 copach 61 copbo 198, 250 61–62 copfay coppbo 62, 72 copsa 61–62, 70–72, 74–76, 78–85 250 eth itc 47–48 itf 178, 187 187 itd itg 178, 187 187 itgi 178 itp itz* 178 syrh 250 vul 193–194 A 193 F 193

Index of Modern Authors Acerbi, F. 97 Adams, Sean A. 146, 150, 227, 229, 238, 253, 277 Ahn, Jeongseop 280, 283–284 Aland, Barbara 35, 41, 174, 204, 224, 268, 272 Aland, Kurt 35, 41, 60, 89, 115, 184–185, 204, 224, 268–269, 321, 338, 340–341 Albl, Martin C. 321 Alford, Henry 148–150 Allan, Donald J. 281 Allen, Garrick V. 2, 14, 17, 24 Amphoux, C.-B. 174 Amundsen, L. 310 Andrist, Patrick 2–4, 7, 14–17, 28 Arvedal, Linnea 184 Aune, David E. 327 Bagnall, Roger S. 322 Bahr, Gordon J. 172 Bakhtin, Mikhail 23 Banning S. J., Joop H. A. van 59 Barnes, Albert 148, 150 Barns, J. W. B. 228, 318–319 Barral-Baron, Marie 55 Barrett, David 128, 268–269, 272, 275, 335, 348 Bas, Philippe Le 336 Baskin, Wade 23 Bastianini, Guido 351 Batovici, Dan 283 Batts, Michael S. 267 Bauckham, Richard 92 Beasley-Murray, George Raymond 35 Bekker, Immanuel 223 Betz, Hans Dieter 124 Birdsall, J. Neville 337 Blanchard, A. 313 Blass, Friedrich 43 Blomkvist, Vemund 177, 195, 219–220 Blumell, Lincoln 155, 313, 315, 317–318 Bokedal, Tomas 90–91, 96, 114 Bonner, S. F. 331 Böttrich, Christfried 286 Bowman, A. K. 153 Braarvig, Jens 327

Bradshaw, Paul F. 282 Brooks, Cleanth 21 Brown, Schuyler 93 Bruce, F. F. 147–148, 196 Bruner, Dale 45 Bruyn, T. de 314, 316, 319, 326, 328 Bruyne, Donatien de 178 Butler, Christopher S. 290 Canart, Paul 7, 244 Capes, David B. 95 Callaghan, José 335 Calvin, John 197 Cappelli, Adriano 260 Carson, D. A. 35 Cavallo, Guglielmo 341 Ciampa, Roy 262–263 Cirafesi, Wally V. 30, 348 Champagne, David G. 179–180, 186, 191, 193, 196–199, 205, 217–218, 220, 223–224 Charlesworth, Scott D. 94, 103, 270, 283, 288, 322 Chesterton, Gilbert Keith 34 Clark, Kenneth W. 5 Clemens, Raymond 232–233 Clivaz, Claire 314, 351 Cole, R. A. 170 Comfort, Philip 128, 268–269, 272, 275, 335, 348 Concannon, Cavan W. 221 Cook, John Granger 320–321 Cowper, B. H. 192 Cramer, John Antony 224 Cresswell, Peter 286 Cribiore, Raffaella 281, 341 Cronin, Sonya Shetty 51 Cuvigny, Hélèn 153 Dahl, Nils A. 176–179 Dam, Raymond Van 325 Daniel, Robert W. 314 de Jonge, Henk Jan 55 de Vries, Johannes 228 Debrunner, Albert 43 DeConick, April 46

376 Deissmann, Adolf 156, 158 DelRio, Delio 137 deSilva, David A. 148–149 Dodd, C. H. 236 Driscoll, Matthew J. 2, 23, 32 Duncan, G. S. 217 Dutton, Paul Edward 50 Ebojo, Edgar 134, 135, 191 Eco, Umberto 51, 58 Ehorn, Seth M. 227, 238 Ehrman, Bart D. 6, 24, 90, 173, 224, 272 Emerton, J. A. 109 Epp, Eldon Jay 25, 152, 313, 323–324 Evans, Craig A. 19 Elliott, J. Keith 5, 33, 48, 134, 153, 286 Falkenberg, René 228 Faraone, Christopher A. 327 Fee, Gordon D. 137–138, 348–349 Fellows, Richard G. 8 Fletcher-Louis, Crispin 97 Förster, Hans 40, 51–52, 54–55, 60 Gallagher, Edmon L. 181, 212 Gamble, Harry Y. 3, 6–7, 125, 267, 270, 281–282, 310–311, 313, 322–323 Garcia, Jorge J. E. 18–20 Garland, David E. 343 Gathercole, Simon J. 338 Gathergood, Emily J. 268–272, 274–276 Gehman, Henry S. 45 Genette, Gérard 14­–15, 126 Gieschen, Charles 95 Gilliard, Frank D. 58 Graffin, Pierre 325 Graham, Timothy 232–233 Graux, Charles 173 Gravely, Edward D. 244 Greer, Rowan A. 214–215 Greetham, D. C. 24 Gregory, Caspar René 340 Grenfell, Bernard P. 153, 236, 315–317, 335–336, 341–345 Grenz, Jesse R. 244 Guthrie, Donald 147–148 Halliday, M. A. K. 290–292, 297 Harris, Rendel J. 173

Index of Modern Authors Harris, W. V. 322 Hatch, William H. P. 172, 185–186, 205–6 Head, Peter M. 298 Heath, Eldon 185, 193 Heath, Jane 93, 95, 119 Hebbelynck, Adolph 71 Hellholm, David 177 Hendel, Ron 24 Henrichs, A. 313 Henry, Matthew 148 Herrera, Robert A. 98 Herzer, Jens 181, 212 Hill, Charles 72, 137, 244, 275 Hill, Robert C. 195 Hirsch, Emanuel 46 Holmes, Michael W. 24, 94, 99, 181, 213 Hotchkiss, Valerie 115 Horner, George 198 Horsley, G. H. R. 314, 319–320 Houghton, Hugh A. G. 225 Howard, George 95 Humfress, Caroline 324 Hunt, A. S. 236, 315, 317, 335–336, 341–345 Hunter, Richard 233 Hurtado, Larry W. 92–93, 95–96, 114, 116–117, 123, 228, 230–231, 236, 270, 280, 282, 285, 302, 310, 313, 322–323 Iser, Wolfgang 20 Ivanov, Sergey A. 195 Jang, Min Seok 134 Janzen, Marshall 139 Jenott, Lance 302 Johnson, Nathan C. 29 Johnson, William A. 228, 281, 284–285, 289, 354–355 Johnston, Jeremiah J. 19 Jones, Brice C. 30, 277, 313–315, 319–320, 336 Jongkind, Dirk 116–117, 232, 239–240, 277, 279, 286, 289, 298 Judge, E. A. 329 Junack, Klaus 128, 142, 341 Kannaday, Wayne C. 5 Karrer, Martin 228 Kasser, Rodolphe 350 Keck, Leander E. 242 Keener, Craig S. 168

377

Index of Modern Authors Keith, Chris 169 Kensky, Meira Z. 222 Kenyon, F. G. 235, 313 Kim, Young Kyu 134 Kirk Alain 56 Klauck, Hans-Josef 146 Knittel, F. A. 330 Knust, Jennifer 221 Koenen, L. 313 Kraeling, Carl H. 320–321, 347 Kraft, Robert 95, 259, 314 Krans, Jan 8, 330 Krause, Thomas J. 310–313, 316–317, 319, 323, 326 Kristeva, Julia 23 Krüger, Julien 331 Kruger, Michael J. 5, 270, 310, 313–315, 318 Kubo, Sakae 349 Labuschagne, Casper J. 97–98 Laird, Benjamin 192 Lagopoulos, Alexandros Ph. 24 Lake, Helen 279 Lake, Kirsopp 5, 192, 279 Langton, Stephen 59 Larson, Jason T. 175 Lavrinovica, Alesja 137 Layton, Bentley 72 Lewin, Jane E. 14 Lightfoot, J. B. 147–148, 196 Lim, Timothy H. 321 Lincicum, David 131 Lincoln, Andrew T. 139 Lindemann, Andreas 213 Llewelyn, Robert 46, 314 Longenecker, Richard N. 260 Lorrain, Agnès 195, 215 Lorusso, Vito 185 Lowe, E. A. 252, 254 Luijendijk, Annemarie 133, 311, 322, 326, 340–341 Maehler, Herwig 341 Martin, Victor 228, 350 Martini, Carlo M. 349 McCormick, Michael 323 McGee, Iain 296 McGurk, Patrick 254 McKay, K. L. 292

McNamee, K. 240, 331–332, 334, 354–356 Meade, John D. 181, 212 Metzger, Bruce M. 173, 176, 182–183, 191, 194, 197, 202, 220, 224, 230, 233, 258, 262, 275, 281, 287, 314, 330 Meyer, Marv W. 314 Mihálykó, Ágnes T. 283 Milbank, Alison 34 Millard, Alan 322 Miller, Owen 21 Milne, H. J. M. 240, 279, 298 Mink, Gerd 79 Mitchell, Timothy N. 5, 270 Mowitt, John 19–22, 24 Mugridge, Alan 128–129, 316 Murphy-O’Connor, Jerome 146 Musurillo, Herbert 325 Nässelqvist, Dan 280, 282–284 Neufeld, Vernon H. 93 New, Silva 192 Nichols, Stephen G. 23 Nicklas, Tobias 183, 212, 310 Nongbri, Brent 134, 150 Oakley, S. P. 233 Obbink, Dirk 327 O’Donnell, Matthew Brook 295 Olson, David R. 285 Omanson, Roger L. 35 Overcash, Benjamin 103 Paap, A. H. R. E. 92, 94, 270 Paley, William 173 Palme, Bernhard 153–154 Parker, David 29, 38, 185, 216, 225, 252, 289 Paulson, Gregory Scott 270 Payne, G. Andrew 172, 178–179, 184, 192, 200 Payne, Philip B. 7–8, 139 Pelikan, Jaroslav 115 Pellegrini, Astorre 337 Peppard, M. 231 Petersen, Theodore 60, 69 Pfeiffer, Rudolf 232 Pitts, Andrew W. 25, 29 Porter, Stanley E. 14, 19, 25, 29–31, 58, 72, 125, 146, 150, 152, 174, 288–289, 294–295, 300, 316, 321, 338, 348 Porter, Wendy J. 29, 31, 338

378 Prentice, William K. 336 Prior, Michael 212–213 Ramelli, Ilaria 222 Ramsay, W. M. 195 Ranke, Ernst 193 Rapp, Claudia 324–325, 327 Ravelli, Louise J. 293 Reasoner, Mark 242, 259, 263, 265 Reece, Steve 146, 150, 152, 154, 160, 168 Reichardt, Alexander 258, 262 Reinhartz, Adele 58 Reynolds, L. D. 232 Richards, E. Randolph 146, 150–152 Ricoeur, Paul 20 Riddle, Donald Wayne 5 Riesner, Rainer 213 Robar, Elizabeth 295 Roberts, C. H. 91–92, 94, 231, 310–311, 315–317, 322, 326, 338 Robinson, Alexandra 46 Ropes, James Hardy 57 Rösch, Friedrich 339 Rothschild, Claire K. 214 Rouse, Mary A. 2, 268, 284 Rouse, Richard H. 2, 268, 284 Rowe, Nina 55 Royle, Anthony P. 2, 14, 17, 24 Royse, James R. 128–134, 167, 275, 287, 289, 346–348 Rummel, Erika 42 Russell, David 143–144 Saenger, Paul 280 Sarri, Antonia 155, 160, 164 Sanders, Henry A. 345 Sanzo, Joseph E. 313, 324 Saussure, Ferdinand de 23 Scheil, Vincent 338 Scherbenske, Eric W. 174, 183, 194, 199 Schmid, Ulrich 239, 240–244, 248, 257 Schmitz, Franz-Jürgen 79 Schnelle, Udo 214 Schoedel, W. R. 89 Schubart, W. 313 Scofield, C. I. 147, 149 Scrivener, F. H. A. 198, 252 Sebastiani, Valentina 33

Index of Modern Authors Segal, Alan F. 327 Shack, Jennifer 8 Shao, Jessica 131 Sharp, Daniel B. 60, 79 Sigismund, Marcus 249, 257, 265 Skeat, T. C. 153, 240, 279, 286, 298, 310–312 Small, Jocelyn Penny 228 Smith, Geoff S. 131, 192, 198–199, 217 Smith, Richard 314 Smith, W. Andrew 249 Soderquist, Justin J. 203 Stanton, G. N. 313, 321 Stevens, Chris S. 34, 271 Steyn, Gert J. 236 Stokowski, A. 244, 247 Swanson, Reuben 174 Swete, Henry Barclay 214 Tâm, Nguyen Thi Minh 293 Testuz, Michel 349 Tischendorf, Constantin 174, 193–194, 217, 239, 288, 298 Thee, F. C. R. 328 Theißen, Gerd 45 Theophilos, Michael P. 356 Thomas, J. D. 153 Thompson, Edward M. 230, 232 Thompson, John B. 20 Thompson, Michael B. 323 Towner, Philip H. 182, 281, 283 Tov, Emanuel 56, 231–232 Traube, Ludwig 90, 93, 114, 270 Treu, Kurt 95 Tripolitis, Antonia 314 Trobisch, David 175, 191 Troxel, Ronald L. 18, 24 Turner, E. G. 229, 231–234, 283, 289, 310, 312, 341 Valdés, Mario J. 21 Van Haelst, J., 321 Van Minnen, Peter 4 Versnel, H. S. 327 Victorinus, Marius 148–149 Wachtel, Klaus 286 Waddington, William Henry 336 Wallace, Daniel B. 29, 291

379

Index of Modern Authors Wallraff, Martin 2–4, 7, 33 Ward, Benedicta 325 Warren, Bill 137 Wassell, Blake Edward 46 Wasserman, Tommy 174, 199, 216, 316, 319 Wayment, Thomas 155, 203, 313, 315, 317–318 Weber, Samuel 23 Weima, Jeffrey A. D. 146, 150–151 Wessely, Karl 337 Wheaton, Gerry 38 White, J. L. 146 Wiebelt, Alexandra 50 Willard, Louis Charles 176–178, 182, 200, 203 Williamson, G. A. 182

Wilson, N. G. 232 Wright, Brian J. 268, 281–283, 322 Wuest, Kenneth S. 148–149 Yamauchi, Edwin M. 327 Yarbrough, Robert W. 199 Yoon, David I. 280, 290, 294, 300 Young, Norman Hugh 42 Zahn, Theodor 148–149 Zamagni, Claudio 221 Ziegler, Joseph 250 Zilliacus, H. 318–319 Zuntz, Gunther 167